ROLE OF HETF AND PATU3 IN THE REGULATION OF …DOCTOR OF PHILOSOPHY IN MICROBIOLOGY AUGUST 2014 By Sasa K. Tom Dissertation Committee: Sean M. Callahan, Chairperson Maqsudul Alam Hongwei

ROLE OF HETF AND PATU3 IN THE REGULATION OF HETEROCYST

DEVELOPMENT IN ANABAENA SP. STRAIN PCC 7120

A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAI‘I AT MĀNOA IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN

MICROBIOLOGY

AUGUST 2014

By

Sasa K. Tom

Dissertation Committee:

Sean M. Callahan, Chairperson Maqsudul Alam

Hongwei Li Paul Q. Patek

Heinz Gert de Couet

ii

DEDICATION

This dissertation is dedicated to my father.

He silently held the weight of the world in his hands for the benefit of others. His passion for life,

compassion and empathy towards all, humility, kind and generous demeanor remain an

inspiration to me. It is through his sacrifice and strength, his wisdom, encouragement and

unconditional love that I can be me. For this, I am fortunate and forever grateful. Thank you,

Daddy.

iii

ACKNOWLEDGEMENTS

Many people were instrumental in the completion of this body of work. I thank my mentor Dr.

Sean Callahan for the opportunity to pursue my research interests (including the study of my

eponymous gene, sasA (all1758)), and for his guidance and support. I thank my committee

members Dr. Maqsudal Alam, Dr. Gert de Couet, Dr. Hongwei Li, and Dr. Paul Patek for their

time, encouragement and support. I am inspired by their steadfast dedication to scientific

research, teaching and aspiring students.

I am indebted to the kindness, fortitude, and noble sacrifice of my parents and my ancestors.

They remind me to not take anything for granted and to honor our collective past. I thank my

family and friends. My early interest in science is attributed to my Big Brother. I still remember

being in awe of the diverse components of a typical cell. How magical! I still remember

magnificent images of a fluorescing plant. To think that I could engineer cells with fluorescent

proteins one day!

I acknowledge past and current members of the Callahan Lab, including Michael Andonian,

Sylvia Beurmann, Dr. Pritty Borthakur, Andy Burger, Roy Nicholas Chang, Dr. Loralyn Cozy,

Rissa Fedora, Kelly Higa, Kathryn (Katie) Hurd, Tyler Law, Reid Oshiro, Dr. Ramya Rajagopalan,

Dr. Doug Risser, Orion Rivers, Chris Runyon, Amanda Shore-Maggio, Blake Ushijima, Patrick

Videau, Jasmine Young, and Shirley Young-Robbins. I also acknowledge members of the Alam

(Sun Ae Kim, Aaron Young), Donachie (Dr. Jimmy Saw), Douglas (Ami Ishihara-Mulligan), Huang

(Dr. Yun Kang), Li (Kimberley Hammond, Charles Hua, Deborah Lee, Dr. Baojin Yang), and

Patek (Dr. Van Luu, Jayson Masaki) Labs. I am grateful of the friendship and continued support

of fellow UC Berkeley alumnus Dr. Eric Umemoto.

Finally, I acknowledge support from the Microbiology staff. I acknowledge Roland Co, Dr.

Shaobin Hao, Debbie Morito and Dr. Xuehua (Ivy) Wan. I thank Dr. Susan Ayin for her guidance

with the Microbiology labs and our conversations over the years. I acknowledge the Microbiology

and Biology Departments for providing me with teaching assistantships every semester. I have

benefited from the opportunity to teach a diverse range of (six different) lower and upper division

courses in both Departments.

Two difficult life experiences unexpectedly transpired during my graduate career. The journey

back from great loss and pain was made much lighter through the combined support, affection

and patience of a caring group of family and friends. Aloha and mahalo nui loa!

iv

ABSTRACT

A central paradigm in developmental biology concerns the differentiation of cells despite the

fundamental sameness of the genetic complement shared by all cells in the organism. The

filamentous cyanobacterium Anabaena sp. strain PCC 7120 is an ideal model system for the

study of development. In response to nitrogen deprivation, Anabaena differentiates nitrogen-

fixing heterocyst cells in a periodic pattern. Anabaena research has practical applications in

human health, sustainable agriculture, and biofuel production. This study aimed to characterize

novel interactions to refine the current understanding of Anabaena development.

Investigations were performed to elucidate the hetF-dependent activation of differentiation. A

novel genetic regulatory network involving hetF, a CHF class protease, and the negative regulator

patU3 and other developmental genes was identified. A component of the hetZ-patU5-patU3

gene cluster, PatU3 was shown to directly interact with HetZ, another activator of differentiation.

Genetic epistasis analysis determined that PatU3 suppressed positive regulation by HetZ and

HetR. These negative feedback loops explain the elevated HetR-GFP concentrations in hetF-

dependent strains despite the paradoxical absence of heterocysts. The HetF-dependent pathway

may act as a control point prior to commitment to the heterocyst cell fate.

Lateral inhibition by PatS and HetN, which both contain the same RGSGR pentapeptide

sequence, involves regulation by HetR. The unique domains present in the structure of HetR

may relate to its activity as a transcriptional activator. In this study, genetic and cytological

approaches were used to identify residues in HetR necessary for interaction with PatS, HetN and

RGSGR. A related investigation demonstrated that the RGSGR-pentapeptide derived from HetN

directs pattern formation by direct cytoplasmic exchange.

Lastly, protein phosphorylation plays a prominent role in varied biological processes. The PP2C-

type protein phosphatase All1758 was characterized in this study. The corresponding all1758

gene is controlled by the developmental genes ntcA and hetR. All1758 affects later stages of

differentiation and may represent a critical link between cell growth, cell division and

morphogenesis by controlling putative sigma factor regulators (antisigma factors and antisigma

factor agonists) and cell division genes (including FtsZ and MinCE). Taken together, these

findings support additional regulatory mechanisms necessary for proper Anabaena development.

v

TABLE OF CONTENTS

Acknowledgements………………………………………………………………………………………..iii

Abstract……………………………………………………………………………………………………..iv

List of Tables……………………………………………………………………………………………….xii

List of Figures……………………………………………………………………………………………...xv

Chapter 1. Preamble on model systems of developmental biology including heterocyst development in the cyanobacterium Anabaena sp. strain PCC 7120……………………..…….1

Relevance of the regulation of gene expression in developmental biology………………...1

Genetic regulation of pattern formation in the eukaryotic model systems Arabidopsis thaliana and Drosophilia melanogaster...............................................................................1

Arabidopsis thaliana flower organ formation…………………………………………2

Arabidopsis thaliana trichome formation……………………………………………...3

Morphogens affect Drosophilia melanogaster cell fate……………………………..4

Genetic regulation of prokaryotic development in the Caulobacter crescentus model system couples a phosphorelay with the cell cycle…………………………………………...5

General background of heterocyst formation in the cyanobacterium Anabaena sp. strain PCC 7120…………………………………………..………………………………….7

A genetic network controls induction and patterning in Anabaena………………………….8

vi

PatA- and HetF- mediated regulation of heterocyst differentiation………………………...10

Sigma factors and Anabaena development…………………………………………………..11

Specific research aims and summary of study outcomes…………………………………..13

References cited………………………………………………………………………………...14

Chapter 2. Role of hetF and patU3 in the regulation of heterocyst development in Anabaena sp. strain PCC 7120………………………………………………………………………..18

Introduction………………………………………………………………………………………18

Materials and Methods………………………………………………………………………….20

Results……………………………………………………………………………………………39

Mutations in patU3 restore heterocyst formation in hetF and hetF patA backgrounds……………………………………………………………………………39

patU3 negatively regulates heterocyst development and contributes to normal cell size……………………..………………………………………………….40

Inactivation of internal transcription start sites present in hetZ do not affect the hetZ phenotype…………………………………………………………….42

Extra copies of patU5 and patU3 inhibit heterocyst differentiation……………….44

Cis-acting transcriptional and translational elements are important for function of the hetZ-patU5-patU3 gene cluster………………………….………….46

Overexpression of hetZ cannot functionally bypass deletion of either hetR or hetF…………………………………….……………………………………...48

vii

Regulation of heterocyst differentiation by patU3 does not require patS or hetN……………………………………………………………………………49

hetZ and patU3 are required for the timing of pattern formation………………….50

hetF patU3 and hetF patA patU3 genotypes have diminished HetR-GFP fluorescence in comparison to hetF and hetF patA genotypes…………………...52

hetF patU3 and hetF patA patU3 genotypes exhibit reduced hetR expression compared to the wild-type……………………………………………….53

Genetic epistasis analysis suggests hetR acts upstream of patU3 in the regulation of heterocyst differentiation………………………………………………54

Genetic epistasis analysis suggests patU3 acts upstream of hetZ in the regulation of heterocyst differentiation………………………………………………56

Direct interaction detected between HetZ and PatU3 by the bacterial two-hybrid system……………………………………………………………………..56

Discussion………………………………………………………………………………………..61

References cited………………………………………………………………………………...68

Chapter 3. Heterocyst development and the ser/thr protein phosphatase All1758 in the cyanobacterium Anabaena sp. strain PCC 7120…………………………………………………..71

Introduction………………………………………………………………………………………71

Materials and Methods………………………………………………………………………….73

Results……………………………………………………………………………………………98

viii

Mutation of all1758 prevents diazotrophic growth………………………………….98

Pleiotropic phenotype of all1758 null mutants……………………………………...99

Strain UHM184 (∆all1758) was Fix- and lacked the minor heterocyst-

specific glycolipid…………………………..…………………………………………102

all1758 is not required for the timing of pattern formation………………………104

all1758 is constitutively expressed and autoregulates its own expression…….105

NtcA and HetR repress all1758 expression……………………………………….106

Bypass of mutation of all1758 by extra copies of hetR…………………………..108

Genetic epistasis analysis supports role of all1758 in normal cell size and cell growth control……………………………….………………………………108

All1758 is required for proper localization of the cell division protein FtsZ and the cell division regulators MinC and MinE…………………………….110

Conserved aspartate residues of Motifs 2, 8 and 11 are required for PP2C phosphatase activity of All1758…………………………………………………….112

All1758 regulates the antisigma factor Alr3423…………………………………...113

NtcA activates expression of the all1758-dependent anti-antisigma factor all1087…………………………………………………………………………………115

Direct interaction of All1758 with itself and HetR detected by the bacterial two-hybrid system……………………………………………………………………116

Discussion………………………………………………………………………………………119

ix

References cited……………………………………………………………………………….124

Chapter 4. Overactive alleles of hetR in the cyanobacterium Anabaena sp. strain PCC 7120………………………………………………………………………………………………...130

Introduction……………………………………………………………………………………..130

Materials and Methods………………………………………………………………………..131

Results…………………………………………………………………………………………..144

Conservative substitutions at HetR residues 250−256 affect heterocyst formation and sensitivity to PatS-5………………………………………………....144

Conservative substitutions at HetR residues 250−256 affect heterocyst formation and sensitivity to PatS and HetN overexpression…………………….147

Identification of overactive alleles of hetR that bypass HetN function, but require a wild type copy of HetR for activity…………………………………..147

Identification of overactive alleles of hetR that bypass HetN function………….149

Identification of overactive alleles of hetR that bypass PatS and HetN function………………………………………………………………………………..150

Overactive alleles of hetR affect sensitivity to PatS and HetN overexpression……………………………………………………………………….151

HetR(mutant)-GFP localization predicts role in HetR turnover………………….152

Overactive alleles of hetR fail to complement hetF- and patA-deficient strains………………………………………………………………………………....157

x

alr9018, a gene that complemented a mutant of a PpetE-hetN ∆patS strain, is not essential for heterocyst differentiation………….…………………………..157

Discussion………………………………………………………………………………………158


Chapter 5. The intercellular signaling protein, hetN, in Anabaena sp. strain PCC 7120……………………………………………………………………………………………………….162

Introduction……………………………………………………………………………………..162

Materials and Methods………………………………………………………………………..164

Results…………………………………………………………………………………………..168

HetN-YFP localizes to heterocysts in stains expressing overactive

alleles of hetR………………………………….……………………………………..168

Localization of HetN-YFP to the cell periphery……………………………………169

hetN mutagenesis studies to identify necessary sites for function…………….. 172

Discussion………………………………………………………………………………………172


xi

Appendices: Peer-reviewed publications

A. The putative phosphatase All1758 is necessary for normal growth, cell size and synthesis of the minor heterocyst-specific glycolipid in the cyanobacterium Anabaena sp. strain PCC 7120………………………...……..………….180

B. Evidence for direct binding between HetR from Anabaena sp. PCC 7120 and PatS-5……………………………...………………………………………………………203

C. The RGSGR amino acid motif of the intercellular signaling protein, HetN, is required for patterning of heterocysts in Anabaena sp. strain PCC 7120…………….....234

xii

LIST OF TABLES

Table Page

1.1 The twelve sigma factors annotated in the Anabaena genome……………………………12

1.2 Description of the role of characterized sigma factors in the Anabaena genome………..12

2.1 Strains used in Chapter 2………………………………………………………………………31

2.2 Plasmids used in Chapter 2……………………………………………………………………33

2.3 Oligonucleotides used in Chapter 2…………………………………………………………...36

2.4 Location of spontaneous mutations isolated within the 258-amino acid protein PatU3 found to restore the ability of hetF and hetF patA background strains to form heterocysts……………………………………………………………………………………....40

2.5 Anabaena patU3 strains and strain description……………………………………………...42

2.6 Anabaena hetZ strains and strain description………………………………………………..43

2.7 Summary of plasmids used to test truncations of genes within the hetZ-patU5-patU3 gene cluster and effect on heterocyst development in wild-type and in ∆patA backgrounds……………………………………………………………………………………..45

2.8 Summary of plasmids used to overexpress individual genes or combinations of genes within the hetZ-patU5-patU3 gene cluster and effect on heterocyst development and cell size in wild-type and ∆patA backgrounds………………….…….....47

2.9 Summary of plasmids used to determine the role of tsp I and tsp II in patU5-patU3 expression and effect on heterocyst development in wild-type and ∆patA backgrounds……………………………………………………………………………………..48

xiii

2.10 Plasmids used for quantification of β-galactosidase activity (Figure 2.16) of positive

bacterial two-hybrid interactions, and average Miller units (with standard deviation) of three independent replicates……………………………………...………………………...61

3.1 Strains used in Chapter 3………………………………………………………………………88

3.2 Plasmids used in Chapter 3……………………………………………………………………90

3.3 Oligonucleotides used in Chapter 3…………………………………………………………...94

3.4 Plasmids used for quantification of β-galactosidase activity (Figure 3.14) of positive bacterial two-hybrid interactions, and average Miller units (with standard deviation) of three independent replicates………………………………...…………………………….118

4.1 Bacterial strains used in Chapter 4…………………………………………………………..137

4.2 Plasmids used in Chapter 4…………………………………………………………………..139

4.3 Oligonucleotides used in Chapter 4………………………………………………………….142

4.4 Summary of sensitivity of seven alleles of hetR (driven by PhetR) encoded on plasmids to PatS and HetN…………………………………………………………………...151

4.5 Summary of sensitivity to different allelic replacements of hetR in the chromosome to overexpression of PatS and HetN……………………………………...…………………152

4.6 Location of the overactive alleles described in this study within the domains of the 299-residue protein HetR………………………………………….……………………..158

5.1 Bacterial strains used in Chapter 5…………………………………………………………..166

5.2 Plasmids used in Chapter 5…………………………………………………………………..167

xiv

5.3 Oligonucleotides used in Chapter 5………………………………………………………….167

xv

LIST OF FIGURES

Figure Page

1.1 Schematic of the activator-inhibitor model of pattern formation proposed by Gierer and Meinhardt in 1972………………………………………………………………..………….4

1.2 Anabaena grown with nitrate in the medium or without a source of combined nitrogen…………………………………………………………………………………………….7

1.3 Working model for the genetic network involved in the regulation and patterning of heterocyst differentiation in Anabaena…………………………………………………………8

1.4 HetR, PatA and HetF are positive regulators of heterocyst development.........................10

2.1 The hetZ-patU5-patU3 gene cluster in Anabaena and locations of transcripts I and II…………………………………………………………………………………………….18

2.2 ∆hetF ∆patA strain with an additional mutation in patU3 restores the ability of ∆hetF ∆patA to form heterocysts………………………………………………………………………39

2.3 A model for the HetF-dependent regulation of PatU3……………………………………….40

2.4 Locations of putative tsps related to patU3 and sites of mutations isolated in the genetic screen described in this study…………………………………………....................40

2.5 Phenotype of patU3∆42-729 compared to the wild-type after 24 h in nitrate-deficient media……………………………………………………………………………………………..41

2.6 Locations of putative tsps (arrows) related to hetZ and site of the transposon insertion in mutant 1801…………………………………………..........................................43

2.7 Phenotype of ∆hetZ mutant after 24 h in nitrate-deficient media.………………………….44

xvi 2.8 Regions present on plasmids used to determine the role of tsp I and tsp II in

patU5-patU3 expression………………………………………………………………………..48

2.9 hetZ and patU3 are required for pattern formation…………………………………………..51

2.10 HetR-GFP is not detected in hetF patA patU3 genotypes…..……………………………...52

2.11 Expression of hetR is not detected in the hetF patA patU3 genotype……………………..53

2.12 Two proposed models of genetic interaction between hetR and patU3 in the control of heterocyst development……………………………………………………………………..54

2.13 Heterocyst formation is abrogated in ∆patU3 ∆hetR and ∆patU3 ∆hetZ double mutant strains……………………………………...………………………………………….…55

2.14 Two proposed models of genetic interaction between hetZ and patU3 in the control of heterocyst development……………………………………………………………………..56

2.15 Bacterial two-hybrid results indicate HetZ interacts with PatU3……………………………59

2.16 Average β-galactosidase activity of positive protein-protein interactions between HetR, HetF, HetF(C246A), HetZ, PatU3, PatA………………………………………………60

2.17 A preliminary model for genetic regulation of heterocyst differentiation in Anabaena involving the conversion of HetR to HetR* by developmental proteins…………….……...63

3.1 Phenotype of the all1758 deletion strain, UHM184 under varying conditions…………..101

3.2 Heterocyst-specific exopolysaccharide of UHM184………………………………………..102

3.3 Heterocyst-specific glycolipids of UHM184…………………………………………………103

xvii 3.4 Acetylene reduction assay for nitrogenase activity indicates strain UHM18

(∆all1758) is Fix-…………………………………….………………………………………….104

3.5 A functional all1758 gene is required for expression of all1758 but not for patterned expression of patS…………………..…………………………………………………………105

3.6 The global transcription factor NtcA negatively regulates all1758 expression………….106

3.7 The heterocyst differentiation gene HetR negatively regulates all1758 expression……107

3.8 All1758 is required for normal cell size and growth in Anabaena………………………...109

3.9 All1758 is involved in the proper localization of the cell division machinery components FtsZ, MinC and MinE…………………………………………………………..111

3.10 Schematic of the domains present in the 463-residue All1758 protein…………………..112

3.11 Overexpression of the antisigma factor Alr3423 in the wild type strain mimics the ∆all1758 phenotype at 24 hours……………………………………………………………...114

3.12 NtcA is required for the localized expression of the anti-antisigma factor All1087 in developing heterocysts……………..…………………………………………………………116

3.13 Bacterial two-hybrid results indicate All758 interacts with HetR and dimerizes with itself………………………………………………………….…………………………………..117

3.14 Average β-galactosidase activity of positive protein-protein interactions between All1758 and HetR and itself…………………………………………………………………..118

3.15 Partner-switch model of sigma factor activation……………………………………………122

3.16 A preliminary model for the regulation of heterocyst development by a modified partner-switch mechanism……………………………………………………………………123

xviii 4.1 Sensitivity to PatS-5 of strains with mutant alleles of hetR………………………………..146

4.2 Heterocyst frequencies of strains UHM170 to UHM182 and the wild-type……………...148

4.3 Localization of representative HetR(mutant)-GFP fusions………………………………..155

4.4 Localization of HetR(mutant)-GFP corresponding to the R250-D256 region……………156

5.1 Schematic of the domains present in the HetN protein……………………………………168

5.2 Localization of HetN-YFP……………………………………………………………………..169

5.3 Localization of HetN-YFP using confocal microscopy……………………………………..171

1

CHAPTER 1. PREAMBLE ON MODEL SYSTEMS OF DEVELOPMENTAL BIOLOGY

INCLUDING HETEROCYST DEVELOPMENT IN THE CYANOBACTERIUM ANABAENA SP.

STRAIN PCC 7120

Relevance of the regulation of gene expression in developmental biology

Multicellularity in nature arises through the elegant coordination of disparate cell types and a

dynamic interplay between the macromolecules of life. The developmental process encompasses

growth, differentiation into specialized cells, morphological modifications, and proper organization

of the differentiated cells. Over the course of evolutionary time, prokaryotic and eukaryotic

organisms have adapted common mechanisms to regulate development.

All development requires differential gene expression. Living organisms continually execute the

process of DNA replication. Yet while each cell of an organism contains the same genomic

content upon faithful partitioning of the chromosome during cellular division, cells nonetheless

express individual genes at different levels. This variance in gene expression is induced in

response to cellular requirements and environmental conditions, and can differ both temporally

and spatially. As a result of the products of these genes, organisms across all three domains of

life can adapt, grow and survive. Understanding the complexities of the regulation of gene

expression, processes underlying the control of the activation and inhibition of genes, is important

in the study of biological systems and evolution. The study of gene regulation not only extends

the knowledge of basic principles of biology. The understanding of when genes are switched on

or off also has implications for developmental impairment and disease pathologies in humans.

Further applications in cancer, cellular aging and senescence, birth defects and growth

abnormalities, pluripotent stem cell research, and pharmaceutical targets may ensue from these

studies.

Several eukaryotic and prokaryotic model systems are used for the study of developmental

biology including Arabidopsis thaliana, Drosophilia melanogaster, Caulobacter crescentus and

Anabaena sp. strain PCC 7120. Common themes that arise in developmental regulation involve

asymmetry (protein concentrations, cell division), cis- and trans-regulatory elements, and post-

translational modifications (proteolysis, phosphorylation).

Genetic regulation of pattern formation in the eukaryotic model systems Arabidopsis

thaliana and Drosophilia melanogaster

While differentiation into multiple cell types is important in the development of an organism,

subsequent organization of cell types, or pattern formation, is also a necessary component of

development. Examples of patterns abound in the biological world [1] and can manifest as

organizing regions or gradients (polyp hydra, Spemann organizer), segments (vertebrae, somite

2

formation), symmetry (left-right polarity, starfishes, anemones, flower petals), stripes (zebras,

zebrafishes), branching (bird feathers, leaf venations, insect trachea, vascularization), oscillations

(seashell pigmentation), periodicity (fly bristles, epidermal extensions, cilia, avian feathers) and

spots (leopards, cheetahs, giraffes, butterfly eyespots). Repeating patterns also frequently arise

in the natural environment: waves, sand dunes, cumulus clouds, honeycombs.

How cell fates are initially determined and the ongoing regulation of the process is a fundamental

question in developmental biology. Two eukaryotic model systems offer insight into this process.

With repeating flower and leaf units and tractable genetics, Arabidopsis thaliana is used as a

model organism for the study of plant development. Additionally, early segmentation studies

based on the invertebrate model organism Drosophilia melanogaster were later found to be

applicable to vertebrate systems.

Arabidopsis thaliana flower organ formation

In the model organism Arabidopsis thaliana, four cell types emerge from an undifferentiated

progenitor cell (the floral meristem) in response to seasonal changes. Together, these

differentiated cells form flower organs (sepals, petals, stamens and carpals) that exhibit

concentric growth in organized patterns called whorls. A genetic network of floral meristem-

identity genes establishes the cell fate of undifferentiated cells. Many of these regulatory genes

show similarity to the MADS domain of transcription factors, which share a conserved function in

yeast and mammalian muscle cells[2]. These transcription factors can dimerize and/or complex

with more than one protein. Among them, the master regulator AP1 (APETALA1) initiates flower

development. It primarily acts as a transcriptional repressor that downregulates genes normally

expressed in the floral meristem. Three pathways appear to upregulate AP1 expression. One

mechanism involves the binding of LFY, another main regulator of floral meristem identity, to the

AP1 promoter. Another involves a FD-dependent protein complex that also binds the AP1

promoter. Lastly, binding of the SPL family of transcription factors to the regulatory region of AP1

also mediates AP1 activation.

Multiple feedback loops positively and negatively regulate the master regulator of flower

development, AP1. Floral repressors inhibit the (repressive) activity of AP1. These repressors

include TEM1 and TEM2 (both are also transcription factors), AP2, TOE1-3, SMZ, SNZ, TLF1

and FLC (the central regulator of the vernalization pathway of flower initiation in response to cold

temperatures)[2]. AP1 repression may involve direct inhibition (SMZ), binding to the regulatory

region of AP1 (AP2), or competitively binding the activators of AP1 (TLF1). To counteract the

activity of floral repressors and begin floral initiation, AP1 represses floral meristem genes

(including SVP and the aforementioned TFL1, TEM1, TEM2, TOE1, TOE3, SNZ) and floral

transition genes (some of which are also activators of AP1; SOC1, AGL24, SPL9, FD, FDP).

3

Coupling of cellular events is another common biological theme. In addition to a key role in flower

development, other functions are associated with AP1. Upon interaction with SEP3 (a

component of different transcription factor complexes), the repressive function of AP1 is reversed.

In this context, AP1 acts as a genetic switch: the reversal of AP1 activity may allow for the

transition from the floral induction stage to the flower formation stage of development.

Furthermore, in addition to specifying organ identity during early flower development, AP1

appears to regulate additional cellular processes including hormone responses and meristem

pattern formation[2].

Arabidopsis thaliana trichome formation

Leaf trichomes represent another developmental model system to understand the regulation of

cell-fate determination, pattern formation and control of cellular morphology and polarity (during

cellular branching)[3, 4]. Trichomes, shoot-derived hair extensions of epidermal origin, are found

in most plants. They function to defend against insects, protect against UV irradiation, and

decrease the effects of transpiration and freezing temperatures. In Arabidopsis thaliana, trichome

spacing is nonrandom and occurs at an average interval of three cells. Four positive regulators of

patterning, GL1, TTGI, GL3 and EGL3, form a transcriptional-activation complex. The

transcriptional-activation domains present in GL1 and GL3 appear to mediate this activation.

Furthermore, these regulators, with the exception of TTG1, share sequence similarity to

transcription factors. TRY, CPC and ETC1 are negative regulators of trichome formation. The

redundant function exhibited by these three negative regulators led to proposal of a mutual-

inhibition mechanism of pattern formation. CPC appears to diffuse across cells. TRY competition

with the transcriptional activation complex, particularly by interacting with GL1 and GL3, is

sufficient to inactivate the complex. Finally, the trichome differentiation gene GL2 functions

downstream of these patterning genes; trichome-specific expression of GL2 is dependent on the

transcriptional-activation complex.

The trichome regulators exhibit additional functions, and may link different stages of development

and/or different cell types. For example, GL3 and TRY also positively and negatively control the

endoduplication stage of the trichome cell cycle, respectively. In this stage of development, DNA

replication continues without cell division to increase the ploidy of trichome cells. Additionally,

patterning of root-hairs involves a subset of the trichome patterning genes, as well as homologs

of these genes. A genetic regulatory network was demonstrated between the positive root-hair

patterning gene WER (WEREWOLF, a homolog of GL1) and the (trichome and root-hair)

negative regulator CPC[5].

Trichome formation conforms to the Gierer and Meinhardt activator-inhibitor model. Gierer and

Meinhardt coupled lateral inhibition with autocatalytic positive feedback loops into a mathematical

4

model to explain primary

pattern formation (Fig. 1.1)

[1, 6, 7]. Prior to Gierer and

Meinhardt, Turing’s

molecular reaction-diffusion

system demonstrated that

the interaction of two

components with differing

rates of diffusion resulted in

pattern formation[8].

However according to the

activator-inhibitor model, local and sustained self-activation (by an activator) is also necessary to

(i) drive differentiation in the presence of inhibitory factors (mutual-inhibition mechanism), and (ii)

introduce instability into a homogeneous system to incite pattern formation[1]. Another criterion

relates to the difference in diffusion rates of the autocatalytic activator and its cognate antagonist

(the inhibitor). The activator acts within in a short-range, while the lateral inhibitor has a longer-

range activity and more rapid turnover. This allows for localization of activator activity, which is

necessary to perturb the system towards an active or “on” state, followed by autocatalysis to

sustain signals for activation.

Trichome development involves activators and inhibitors. The transcriptional-activation complex

(consisting of GL1, TTGI, GL3 and EGL3) fulfil the requirements of an activator. Although

expressed initially in all epidermal cells, activator expression is increased in cells that commit to

the trichome cell fate. In addition, the negative regulators TRY, CPC and ETC1 fit the role of

inhibitors. CPC-GFP diffusion across trichoblasts has been observed. Because CPC is not

expressed in activator-deficient mutants, activator-dependent induction of the inhibitor CPC is

inferred, and consistent with the activator-inhibitor model.

Morphogens affect Drosophilia melanogaster cell fate

Cellular communication is a fundamental concept in biological systems. For example, hormones,

molecules that control varied physiological and behavioral processes over long distances in

plants and animals, represent one form of cell signaling. Examples in bacteria include quorum

sensing and biofilm production. In the context of morphogenesis, morphogens function as

diffusible signaling molecules that govern pattern formation during development. Morphogens

exhibit medium-range activity. Diffusion of morphogens from a source cell establishes a

concentration gradient among neighboring cells. Because levels of the morphogen are inversely

related to distance from the source cell, cells further away from the source cell are subjected to

Figure 1.1. Schematic of the activator-inhibitor model of

pattern formation proposed by Gierer and Meinhardt in 1972.

Inhibitors of the pathway represented with bars, activators

represented with arrows.

5

lower morphogen concentrations. Graded signals derived from morphogens therefore transmit

spatial information among neighboring cells. Furthermore due to the graded nature of the signal,

one signal can potentially induce the formation of multiple cell types due to different threshold

responses. Superposition of a diffusible morphogen under the regulation of an activator is one

criterion of the activator-inhibitor model.

Diffusion of signaling factors in Arabidopsis thaliana, including FT and CPC, influence flower

development and trichome formation respectively. In Drosophilia melanogaster, the Hedgehog

(Hh) morphogen gradient controls patterning of embryonic cuticles and adult appendages.

Vertebrate homologs to Hh have similar roles in development and cell fate specification. In

humans, Hh signaling is implicated in limb, neural tube and internal organ formation; height

determination; cell growth control; neural and hematopoetic stem cell maintenance; neuropathies,

cancer and neonatal abnormalities[9]. Hh signaling converges on the Gli family of zinc-finger

transcription factors [9]. The seven-span transmembrane protein Patched (Ptc) functions as the

Hh receptor. In the absence of Hh, Ptc inhibits the signaling pathway by targeting the G-coupled

protein receptor Smoothened (Smo) for endocytosis and subsequent degradation. Hh binding to

Ptc targets Ptc for lysosome degradation. Subsequent activation of Smo transduces the Hh

signal to the cytoplasm to generate the activator form of Gli.

The tight regulation of the process of Hh secretion relates to the intercellular transduction of the

Hh signal. Post-translational processing of Hh contributes to the activity of Hh as a morphogen.

After cleavage of the Hh signal sequence, the Hh C-terminal domain catalyzes the intramolecular

transfer of cholesterol to the N-terminal Hh signaling domain (HhN). The covalent attachment of

cholesterol increases the affinity of Hh for the plasma membrane and facilitates the addition of a

palmitic acid moiety at the N-terminus of HhN (to generate dually modified HhNp). These

modifications appear to allow for interaction with lipid rafts to target the Hh to the plasma

membrane. Stabilization of HhNp by heparin sulfate proteoglycans (HSPGs), some of which are

localized to lipid rafts, appears to facilitate HhNp diffusion[10]. HhNp secretion is further

facilitated by the 12-pass transmembrane protein Dispatched (Disp). Mechanisms also exist to

regulate gradient size[11]. Levels of Hh above a certain threshold can induce Hh expression.

Expression of Hh also leads to tandem expression of the Hh receptor, Ptc, thus forming a

negative feedback loop that restricts further transduction of the Hh signal.

Genetic regulation of prokaryotic development in the Caulobacter crescentus model

system couples a phosphorelay with the cell cycle

Development in the aquatic bacterium Caulobacter crescentus is intimately coordinated with cell

cycle events[12]. Asymmetric cell division of a progenitor stalked cell yields two morphologically

distinct daughter cells: another sessile stalked cell that can initiate DNA replication, and a motile

6

(flagellated) swarmer cell incapable of DNA replication. Swarmer cell differentiation involves loss

of the flagellum and subsequent stalk construction. Moreover, the developmental process

addresses common themes in biology. First, asymmetry in cell growth is a ubiquitous means to

alter the physiology of cells, and is observed in another model system of development, Bacillus

subtilis during spore formation. Outcomes of asymmetry also appear to influence cellular

senescence and population genetics[13]. Second, the stalked-to-swarmer cell transition

integrates phosphorylation and proteolysis, two mechanisms commonly used to regulate

biological systems, with progression of cell cycle events.

The DNA binding protein CtrA functions as the master regulator of Caulobacter crescentus cell

development. After initiation of DNA replication in late stalked and predivisional cells, ctrA is

expressed from the native P1 promoter. A phosphorelay involving the bifuncational kinase (and

phosphatase) CckA leads to CtrA phosphorylation at a conserved aspartate residue. CtrA

homodimerization coupled with phosphorylation leads to an active form of CtrA. Active CtrA

exhibits increased affinity for regions that contain or are adjacent to DnaA binding sites at the

origin of replication, thereby physically inhibiting DNA replication in swarmer cells. Accumulation

of phosphorylated CtrA activates the stronger P2 promoter, and represses expression from the

weaker P1 promoter. These positive autoregulatory loops rapidly enhance CtrA activity to

activate transcriptional events for transition to a predivisonal cell. Because CtrA concentrations

are in flux at this stage, varying promoter affinities for CtrA may be a means to regulate the timing

of CtrA-regulated gene expression. Approximately a hundred genes are regulated by CtrA,

mainly required for polar morphogenesis and cell division (initiation factors, topoisomerases, etc,).

For DNA replication to ensue, the physical inhibition of CtrA must be alleviated. Proteolysis of

CtrA occurs at the swarmer-to-stalked cell transitional stage upon cleavage by the ClpXP

protease, initiating swarmer cell differentiation. ClpXP localization to the stalked pole is facilitated

by the output domain deficient response regulator, CpdR. CpdR is active when

dephosphorylated. In another regulatory circuit, the CckA phosphorelay that activates CtrA by

phosphorylation also inhibits CpdR upon phosphorylation. CckA and CtrA are both active during

the mid- to late-predivisional stage. During replication initiation, CckA deactivation leads to the

non-phosphorylated states of CtrA (resulting in the non-active form of CtrA) and CpdR (active

form). Furthermore, CpdR activity involves the second messenger c-di-GMP. Induction of the

swarmer cell differentiation is induced by unphosphorylated CpdR-dependent degradation of the

c-di-GMP phosphodiesterase PdeA.

Regulation of the bifunctional (histidine kinase and phosphatase) protein CckA by phosphatases

(PleC) and kinases (DivK, DivJ, DivL) is necessary for the generation of a concentration gradient

of phosphorylated CtrA. During the swarmer-to-stalked cell transition, the bifunctional (histidine

7

kinase and phosphatase) protein CckA exhibits phosphatase activity at the stalked pole of the

predivisional cell. The phosphatase function contrasts with activation of kinase activity of CcKA

at the swarmer pole. Because concentrations of phosphorylated CtrA are thus elevated at the

swarmer pole compared to the stalker pole, a gradient of active CtrA is generated in the

predivisional cell[14]. Thus the CtrA gradient has consequences for development, and resembles

the diffusion properties of the morphogen Hedgehog in Drosophilia development localized to one

cell type.

General background of heterocyst formation in the cyanobacterium Anabaena sp. strain

PCC 7120

The proper orchestration of differentiation is fundamental to the development of multicellular

organisms. The transition to a specialized cell type, or differentiation, allows multicellular

organisms to carry out processes that are not accessible to a single cell. The filamentous

photoautotrophic cyanobacterium Anabaena sp. strain PCC 7120 (herein Anabaena) represents

another ideal model system for the study of differentiation. Anabaena represents one of the

simplest and most ancient examples of a multicellular organism. Anabaena filaments are

composed of vegetative cells and heterocyst cells. Heterocyst formation follows the stages of

development commonly used to describe more complex developmental systems: (1) induction of

differentiation, (2) specification of the cells in a pattern, (3) terminal commitment to the

differentiation process, and finally, (4) morphological and physiological alterations.

During conditions of combined

nitrogen availability, Anabaena

exists as linear filaments

composed of only vegetative

cells, often hundreds of cells in

length (Fig. 1.2A). After 24

hours of combined nitrogen

limitation (nitrogen step-down),

Anabaena filaments develop

heterocysts, which are

specialized cells capable of

molecular nitrogen fixation (Fig.

1.2B). Heterocyst development

occurs at regularly spaced

intervals; heterocysts total approximately 10% of the population. Within the heterocyst,

intracellular levels of oxygen are decreased through the physical elaboration of external glycolipid

Figure 1.2. Anabaena grown with nitrate in the medium (A)

or without a source of combined nitrogen (B). Carets

indicate heterocysts. Bar=10 µM.

8

and polysaccharide layers, deactivation of the oxygen-producing complex of photosystem II, and

enhancement of the respiratory rate to scavenge residual molecular oxygen[15, 16]. The ensuing

microoxic heterocyst provides the conditions necessary for the oxygen-labile nitrogenase

complex to reduce dinitrogen gas to ammonium, despite the continued oxygen evolution in

adjacent photosynthetic vegetative cells. Metabolite exchange occurs between these two cell

types. A bioactive, fixed form of nitrogen is supplied from heterocysts to vegetative cells, which in

turn provide a source of carbon and reductant to heterocysts.

A genetic regulatory network controls induction and patterning in Anabaena

A diverse array of signals governs the developmental program of heterocyst formation and

coordinates the subsequent periodic pattern of one heterocyst forming every tenth cell along a

filament (Fig. 1.3 and 1.2B). Proper regulation of heterocyst differentiation is critical to the

survival of the population because the heterocyst represents an endpoint in development due to

its terminally differentiated and non-dividing state.

Figure 1.3. Working model for the genetic network involved in the regulation and patterning

of heterocyst differentiation in Anabaena. The RGSGR pentapeptide motif is present in the

protein sequence of both PatS and HetN. Inhibitors of the pathway represented with bars,

activators represented with arrows. Figure adapted with permission from Dr. Sean M.

Callahan.

9

Induction occurs in response to nitrogen deprivation. Following limitation of combined nitrogen

(the “inducing factor”), levels of 2-oxoglutarate accumulate[17], ultimately triggering the induction

of heterocyst development upon binding to NtcA[18]. NtcA is a global transcriptional regulator of

the CAP family that regulates nitrogen metabolism in cyanobacteria. NtcA-mediated regulation

involves the consensus sequence GTAN8TAC[19, 20]. One of the genes activated by NtcA, the

nitrogen-dependent response regulator nrrA, produces a gene product that activates expression

of hetR[21]. Considered the master regulator of heterocyst differentiation, the DNA-binding

protein HetR is necessary for both heterocyst differentiation and patterning, largely through

positive regulation of other developmental genes. HetR has been shown to directly up-regulate

the developmental genes ntcA, patA, hepA, patS, hetP, hetZ as well as hetR itself[22-28]. Among

the developmental genes up-regulated by HetR, the protein also has been shown to directly bind

to the promoter regions of patS, hepA, hetP, hetZ, pknE and hetR[23, 25, 26, 29, 30], suggesting

that HetR may primarily function as the transcriptional regulator of developmental genes.

Inhibitory mechanisms are a central component of the regulatory network responsible for

patterning of heterocyst development. The unregulated differentiation of heterocysts would be

detrimental to the organism, as these cells represent a specialized cell type that no longer divides

and cannot fix carbon. Two proteins containing the pentapeptide RGSGR motif negatively

regulate the differentiation process. Exogenous addition of the pentapeptide alone is sufficient to

prevent differentiation of heterocysts [31] and moreover functionally inhibits binding of HetR to its

target sequence[23, 32, 33]. Thus HetR and the two RGSGR-containing inhibitory proteins fulfill

feedback loops in a manner consistent with the activator-inhibitor model (Fig.1.3). Furthermore,

the lateral diffusion patterns and consequence on heterocyst formation appears to follow the

properties of vertebrate morphogens.

Containing RGSGR at its C-terminus, the small (17 amino acid[34]) peptide PatS controls the

initial pattern of differentiation. Although dependent on HetR for up-regulation, PatS inhibits the

DNA-binding activity of HetR. A patS-null mutant exhibits a shorter interval between heterocysts

as well as a multiple contiguous heterocyst (“Mch”) phenotype due to increased differentiation

among adjacent cells. However, the Mch phenotype is resolved after additional rounds of cell

division, resulting in the restoration of the wild-type pattern to the mutant over time[35].

Pattern resolution in the patS-null mutant results from a second inhibitory protein containing an

internal RGSGR motif, HetN. This negative-acting regulator functions as a patterning protein

responsible for maintenance of the pattern initiated by PatS. The hetN-null mutant does not

exhibit perturbation of the wild-type pattern until 48 h after nitrogen step-down, at which point it

exhibits the Mch phenotype[36]. In a patS-null mutant conditionally lacking expression of hetN,

10

the Mch phenotype observed in the patS-null mutant fails to resolve over time, suggesting that

HetN is responsible for preserving the wild-type pattern[37].

PatA and HetF positively regulate heterocyst differentiation

Developmental checkpoints have evolved to control the genetic network of heterocyst formation.

PatS and HetN are regulated in part by patA. PatA contains a PATAN domain (“PatA N-terminus”)

present in many ligand signaling domains, a disrupted helix-turn-helix domain, and C-terminal

sequence similarity to the response regulator CheY[38]. PatA may attenuate the negative

inhibition of PatS and HetN[39]. PatA may also promote HetR activity. Inactivation of patA

results in the accumulation of HetR protein in filaments[40], but paradoxically, differentiation is

significantly reduced in the mutant as heterocyst formation is confined to filament termini (Fig.

1.4B) [41].

PatA-mediated promotion of

HetR activity appears to be

dependent upon HetF,

another positive regulator of

differentiation. As in the patA-

null mutant, inactivation of

hetF results in an accrual of

HetR,[40], but heterocysts do

not form in this strain (Fig.

1.3C) [42]. Excess levels of

HetR thus do not suffice to

induce the normal pattern of

heterocyst formation in the

absence of either PatA or

HetF. Although the effect of

gene inactivation and

presence in multicopy of hetF

closely mimics the effect of

hetR, irregularities in cell

morphology are associated

with the genetic manipulation of hetF. Both mutants no longer differentiate heterocysts, but the

hetF-null mutant grows poorly and exhibits aberrantly enlarged cells irrespective of nitrogen

status (Fig. 1.4A and 1.4C). This altered cell morphology is dependent on both PatA and HetR,

as evidenced by the restoration of normal cell size in the two double-deletion mutants, hetF patA

Figure 1.4. HetR, PatA and HetF are positive regulators of

heterocyst development. Light micrographs of background

strains deficient in (A) hetR, (B) patA, and (C) hetF 48 hours

after nitrogen removal. Carets indicate heterocysts.

11

and hetF hetR[40]. Additionally, overexpression of patA in the wild-type strain leads to enlarged

cells, similar to the hetF deletion phenotype[27]. The presence of the hetF gene in multicopy

results in increased heterocyst differentiation in the wild-type strain, similar to the multicopy effect

of hetR, but filaments with extra copies of hetF have a significant reduction in cell size[40].

A putative transmembrane domain and a conserved caspase-hemoglobinase fold (CHF) domain

are present in the protein sequence of HetF. CHF domains are responsible for the proteolytic

activity of cysteine-dependent endopeptidases that cleave after aspartate residues, including the

caspase family of apoptotic effector proteins. Cysteine proteases also share homology to

eukaryotic separases, universal mitotic regulatory proteins responsible for cleaving the cohesion

between sister chromatids to allow partitioning of chromosomes to daughter cells[43].

Metacaspases, the ancient family of caspase-like prokaryotic proteins, are particularly abundant

in Cyanobacteria and other bacteria that undergo complex developmental pathways[44]. This

observation may relate to the unique requirement for programmed cell death imposed upon

developmental systems that achieve a multicellular state.

Sigma factors and Anabaena development

Because the progression from gene to gene product to gene function involves many stages,

genes (and the output of genes) are controlled at multiple stages, allowing for diverse levels of

control. A common theme in biology concerns regulation of the first step in any multi-step

process. In bacteria, sigma subunits target and position RNA polymerase at specific promoter

sequences during transcription initiation. Because levels of RNA polymerase and sigma factors

are limited in the cell, promoters must compete for transcription initiation. Regulatory

mechanisms of this process include promoter DNA sequences, sigma factor availability (related

to sigma factor regulators), small ligands (i.e. ppGpp), transcription factors (activators and

repressors), and chromosomal conformation[45]. These diverse mechanisms modulate the level

of individual gene expression commensurate with changing cellular demands.

Complex responses and cellular processes are often globally organized by sigma factors in

mechanisms involving phosphorylation. Sigma factors regulate development (Bacillus subtilis

spore formation), control stress responses (Bacillus subtilis, Staphylococcus aureus), and

contribute to biofilm production (Pseudomonas aeruginosa). Sigma factors also regulate

heterocyst development and nitrogen fixation in Anabaena. Twelve sigma factors have been

annotated (Table 1.1), but not all sigma factors have been characterized. Mutant characterization

of individual sigma factors is hampered by the apparent redundancy in function shared among the

sigma factors. Nonetheless, sigma factors function at different stages of heterocyst development

(early, middle, late; Table 1.2), and moreover control stress responses (SigJ; Table 1.2) in

Anabaena. Orchestration of the progression of differentiation by distinct and/or overlapping

12

sigma factors would present an efficient means of developmental regulation in Anabaena, but

mechanisms of this process remain obscure.

Table 1.1. The twelve sigma factors annotated in the Anabaena genome. Characterized sigma

factors are described in Table 1.2.

σ70

family Sigma factor (gene annotation)

Genomic location

Group 1 sigA (all5263) chromosome

Group 2 sig B (all7615) β plasmid

sig B3 (all7608) β plasmid

sigB4 (all7179) α plasmid

sigB2 (alr3800) chromosome

sigC (all1692) chromosome

sigE (alr4219) chromosome

Group 3 sigF (all3853) chromosome

sigJ (alr0277) chromosome

Group 4 sigG (alr3280) chromosome

sigI (all2193) chromosome

Table 1.2. Description of the role of characterized sigma factors in the Anabaena genome.

‘Stages’ in selected cells of the far left column refers to the timing of heterocyst development

(separated into early, middle, and late stages).

Sigma factor Role in Anabaena development Reference

sigA Constitutively expressed, but transcription induced upon nitrogen limitation.

[46]

sigB Upregulated expression at 12 h, not essential for heterocyst differentiation or diazotrophic growth.

[47]

sigB2, sigD, sigF Not essential for heterocyst formation nor diazotrophic growth and not developmentally regulated. A sigB2 sigD double mutant only forms proheterocysts, cannot grow diazotrophically, and fragments extensively.

[48]

sigC (early-stage)

Expression induced upon nitrogen and sulfur limitation; upregulated by 4 h in differentiating cells. Not essential for heterocyst differentiation nor diazotrophic growth.

[47, 49]

sigG (middle-stage)

Expressed in vegetative cells grown in nitrate-replete conditions. Expression decreases after nitrogen removal, but by 10 h, individual differentiating cells increase expression

[49]

sigE (late-stage)

NtcA-dependent expression in differentiating cells by 16 h, required for expression of heterocyst-specific genes (nifH, fdxH, hglE2); mutant showed delayed heterocyst formation (at 32 h).

[48-50]

sigJ Key global regulator of desiccation tolerance; upregulated by drought stress. Regulates synthesis of extracellular heterocyst polysaccharides.

[51]

13

Specific research aims and summary of study outcomes

The synergistic relationship between the two cell types in Anabaena represents multiple levels of

genetic control involving elegant integration of positive- and negative-acting factors. The goal of

this research is to understand the genetic regulatory network governing different stages of

heterocyst development in Anabaena. The implications for these fundamental studies extend

beyond understanding the dynamic interplay of nitrogen fixation and cellular differentiation.

Anabaena research has practical applications in human health, sustainable agriculture, and

biofuel production. The objectives of this study and summary of research outcomes are outlined

below.

1. (Chapter 2) Investigate the interaction of hetF, a positive regulatory factor of heterocyst

development, with other developmental genes. A regulatory network involving the

negative regulator patU3 and hetF was identified through these studies, and extended to

include additional developmental genes (the positive regulators hetR and hetZ). These

studies characterize a regulatory mechanism involving a non-RGSGR containing inhibitor

during heterocyst patterning in Anabaena.

2. (Chapter 3) Characterize the role of the PP2C phosphatase All1758 in heterocyst

differentiation[52]. Phosphorylation is a common mechanism used to control a wide

range of biological processes. In addition to a function in nitrogen fixation, this novel

phosphatase was found to have a role in normal cellular processes including cell growth,

cell division, morphology and development. This aim targets understanding interactions

at advanced stages of heterocyst development including commitment and

morphogenesis (Fig. 1.3).

3. (Chapters 4 and 5) Characterize regions in HetR necessary for interaction with inhibitors

of heterocyst differentiation containing the RGSGR pentapeptide sequence (PatS,

HetN)[32], and investigate the intercellular signaling of HetN[53]. PatS and HetN regulate

the gradient size of HetR in the control of heterocyst development, in a manner

resembling vertebrate morphogens. This aim targets understanding interactions at the

pattern formation stage of heterocyst development (Fig. 1.3).

14

REFERENCES CITED

1. Gierer, A. and H. Meinhardt, A theory of biological pattern formation. Kybernetik, 1972.

12: p. 30-39.

2. Wellmer, F. and J.L. Riechmann, Gene networks controlling the initiation of flower

development. Trends Genet, 2010. 26(12): p. 519-27.

3. Schellmann, S. and M. Hulskamp, Epidermal differentiation: trichomes in Arabidopsis as

a model system. Int J Dev Biol, 2005. 49(5-6): p. 579-84.

4. Hulskamp, M., Plant trichomes: a model for cell differentiation. Nat Rev Mol Cell Biol,

2004. 5(6): p. 471-80.

5. Lee, M.M. and J. Schiefelbein, Cell pattern in the Arabidopsis root epidermis determined

by lateral inhibition with feedback. Plant Cell, 2002. 14(3): p. 611-8.

6. Meinhardt, H., Models of biological pattern formation: from elementary steps to the

organization of embryonic axes. Curr. Top. Dev. Biol., 2008. 81: p. 1-63.

7. Meinhardt, H. and A. Gierer, Pattern formation by local self-activation and lateral

inhibition. Bioessays, 2000. 22(8): p. 753-60.

8. Turing, A., The chemical basis of morphogenesis. Phil. Trans. R. Soc. Ser. B, 1952. 237:

p. 37-72.

9. Jiang, J. and C.C. Hui, Hedgehog signaling in development and cancer. Dev Cell, 2008.

15(6): p. 801-12.

10. Gallet, A., Hedgehog morphogen: from secretion to reception. Trends Cell Biol, 2011.

21(4): p. 238-46.

11. Varjosalo, M. and J. Taipale, Hedgehog: functions and mechanisms. Genes Dev, 2008.

22(18): p. 2454-72.

12. Curtis, P.D. and Y.V. Brun, Getting in the loop: regulation of development in Caulobacter

crescentus. Microbiol Mol Biol Rev, 2010. 74(1): p. 13-41.

13. Kysela, D.T., et al., Biological consequences and advantages of asymmetric bacterial

growth. Annu Rev Microbiol, 2013. 67: p. 417-35.

14. Tsokos, C.G. and M.T. Laub, Polarity and cell fate asymmetry in Caulobacter crescentus.

Curr Opin Microbiol, 2012. 15(6): p. 744-50.

15. Wolk, C.P., A. Ernst, and J. Elhai, Heterocyst metabolism and development, in The

Molecular Biology of Cyanobacteria, D.A. Bryant, Editor. 1994, Kluwer Academic

Publishers: Dordrecht, The Netherlands. p. 769-823.

16. Flores, E. and A. Herrero, Compartmentalized function through cell differentiation in

filamentous cyanobacteria. Nat Rev Microbiol, 2010. 8(1): p. 39-50.

17. Laurent, S., et al., Nonmetabolizable analogue of 2-oxoglutarate elicits heterocyst

differentiation under repressive conditions in Anabaena sp. PCC 7120. Proc. Natl. Acad.

Sci. U.S.A., 2005. 102: p. 9907-9912.

15

18. Vazquez-Bermudez, M.F., A. Herrero, and E. Flores, 2-Oxoglutarate increases the

binding affinity of the NtcA (nitrogen control) transcription factor for the Synechococcus

glnA promoter. FEBS letters, 2002. 512(1-3): p. 71-4.

19. Herrero, A., A.M. Muro-Pastor, and E. Flores, Nitrogen control in cyanobacteria. J.

Bacteriol., 2001. 183: p. 411-425.

20. Luque, I., E. Flores, and A. Herrero, Molecular mechanism for the operation of nitrogen

control in cyanobacteria. Embo J, 1994. 13(23): p. 2862.

21. Ehira, S. and M. Ohmori, NrrA, a nitrogen-responsive response regulator facilitates

heterocyst development in the cyanobacterium Anabaena sp. strain PCC 7120. Mol.

Microbiol., 2006. 59: p. 1692-1703.

22. Muro-Pastor, A.M., et al., Mutual dependence of the expression of the cell differentiation

regulatory protein HetR and the global nitrogen regulator NtcA during heterocyst

development. Mol. Microbiol., 2002. 44: p. 1377-1385.

23. Huang, X., Y. Dong, and J. Zhao, HetR homodimer is a DNA-binding protein required for

heterocyst differentiation, and the DNA-binding activity is inhibited by PatS. Proc. Natl.

Acad. Sci. U.S.A., 2004. 101(14): p. 4848-4853.

24. Black, T.A., Y. Cai, and C.P. Wolk, Spatial expression and autoregulation of hetR, a gene

involved in the control of heterocyst development in Anabaena. Mol. Microbiol., 1993. 9: p.

77-84.

25. Du, Y., et al., Identification of the HetR recognition sequence upstream of hetZ in

Anabaena sp. strain PCC 7120. J. Bacteriol., 2012. 194: p. 2297-2306.

26. Higa, K.C. and S.M. Callahan, Ectopic expression of hetP can partially bypass the need

for hetR in heterocyst differentiation by Anabaena sp strain PCC 7120. Mol. Microbiol.,

2010. 77: p. 562-574.

27. Young-Robbins, S.S., et al., Transcriptional regulation of the heterocyst patterning gene

patA from Anabaena sp. strain PCC 7120. J. Bacteriol., 2010. 192: p. 4732-4740.

28. Ramasubramanian, T.S., T.F. Wei, and J.W. Golden, Two Anabaena sp. strain PCC

7120 DNA-binding factors interact with vegetative cell- and heterocyst-specific genes. J.

Bacteriol., 1994. 176(5): p. 1214-1223.

29. Saha, S.K. and J.W. Golden, Overexpression of pknE blocks heterocyst development in

Anabaena sp. strain PCC 7120. Journal of bacteriology, 2011. 193(10): p. 2619-29.

30. Risser, D.D. and S.M. Callahan, Mutagenesis of hetR reveals amino acids necessary for

HetR function in the heterocystous cyanobacterium Anabaena sp. strain PCC 7120. J.

Bacteriol., 2007. 189: p. 2460-2467.

31. Yoon, H.-S. and J.W. Golden, Heterocyst pattern formation controlled by a diffusible

peptide. Science, 1998. 282: p. 935-938.

16

32. Feldman, E.A., et al., Evidence for direct binding between HetR from Anabaena sp. PCC

7120 and PatS-5. Biochmistry, 2011. 50: p. 9212-9224.

33. Feldman, E.A., et al., Differential binding between PatS C-terminal peptide fragments and

HetR from Anabaena sp. PCC 7120. Biochemistry, 2012. 51: p. 2436-2442.

34. Corrales-Guerrero, L., et al., Functional dissection and evidence for intercellular transfer

of the heterocyst-differentiation PatS morphogen. Mol Microbiol, 2013. 88(6): p. 1093-105.

35. Yoon, H.-S. and J.W. Golden, PatS and products of nitrogen fixation control heterocyst

pattern. J. Bacteriol., 2001. 183: p. 2605-2613.

36. Callahan, S.M. and W.J. Buikema, The role of HetN in maintenance of the heterocyst

pattern in Anabaena sp. PCC 7120. Mol. Microbiol., 2001. 40: p. 941-950.

37. Borthakur, P.B., et al., Inactivation of patS and hetN causes lethal levels of heterocyst

differentiation in the filamentous cyanobacterium Anabaena sp. PCC 7120. Mol.

Microbiol., 2005. 57: p. 111-123.

38. Makarova, K.S., et al., Cyanobacterial response regulator PatA contains a conserved N-

terminal domain (PATAN) with an alpha-helical insertion. Bioinformatics, 2006. 22(11): p.

1297-1301.

39. Orozco, C.C., D.D. Risser, and S.M. Callahan, Epistasis analysis of four genes from

Anabaena sp. strain PCC 7120 suggests a connection between PatA and PatS in

heterocyst pattern formation. J. Bacteriol., 2006. 188(5): p. 1808-1816.

40. Risser, D.D. and S.M. Callahan, HetF and PatA control levels of HetR in Anabaena sp.

strain PCC 7120. J. Bacteriol., 2008. 190: p. 7645-7654.

41. Liang, J., L. Scappino, and R. Haselkorn, The patA gene product, which contains a

region similar to CheY of Escherichia coli, controls heterocyst pattern formation in the

cyanobacterium Anabaena 7120. Proc. Natl. Acad. Sci, USA, 1992. 89: p. 5655-5659.

42. Wong, F.C.Y. and J.C. Meeks, The hetF gene product is essential to heterocyst

differentiation and affects HetR function in the cyanobacterium Nostoc punctiforme. J.

Bacteriol., 2001. 183: p. 2654-2661.

43. Aravind, L. and E.V. Koonin, Classification of the caspase-hemoglobinase fold: detection

of new families and implications for the origin of the eukaryotic separins. Proteins, 2002.

46: p. 355-367.

44. Uren, A.G., et al., Identification of paracaspases and metacaspases: two ancient families

of caspase-like proteins, one of which plays a key role in MALT lymphoma. Mol Cell,

2000. 6(4): p. 961-7.

45. Browning, D.F. and S.J. Busby, The regulation of bacterial transcription initiation. Nat Rev

Microbiol, 2004. 2(1): p. 57-65.

17

46. Brahamsha, B. and R. Haselkorn, Isolation and characterization of the gene encoding the

principal sigma factor of the vegetative cell RNA polymerase from the cyanobacterium

Anabaena sp. strain PCC 7120. J. Bacteriol., 1991. 173(8): p. 2442-2450.

47. Brahamsha, B. and R. Haselkorn, Identification of multiple RNA polymerase sigma factor

homologs in the cyanobacterium Anabaena sp. strain PCC 7120: Cloning, expression,

and inactivation of the sigB and sigC genes. J. Bacteriol., 1992. 174: p. 7273-7282.

48. Khudyakov, I. and J.W. Golden, Identification and inactivation of three group 2 sigma

factor genes in Anabaena sp. strain PCC 7120. J. Bacteriol., 2001. 183: p. 6667-6675.

49. Aldea, M.R., R.A. Mella-Herrera, and J.W. Golden, Sigma factor genes sigC, sigE and

sigG are upregulated in heterocysts of the cyanobacterium Anabaena sp. strain PCC

7120. J. Bacteriol., 2007. 189: p. 8392-8396.

50. Mella-Herrera, R.A., et al., The sigE gene is required for normal expression of heterocyst-

specific genes in Anabaena sp. strain PCC 7120. J Bacteriol, 2011. 193(8): p. 1823-32.

51. Yoshimura, H., et al., Group 3 sigma factor gene, sigJ, a key regulator of desiccation

tolerance, regulates the synthesis of extracellular polysaccharide in cyanobacterium

Anabaena sp. strain PCC 7120. DNA Res, 2007. 14(1): p. 13-24.

52. Tom, S.K. and S.M. Callahan, The putative phosphatase all1758 is necessary for normal

growth, cell size, and synthesis of the minor heterocyst-specific glycolipid in the

cyanobacterium Anabaena sp. strain PCC 7120. Microbiology, 2012. 158: p. 380-389.

53. Higa, K.C., et al., The RGSGR amino acid motif of the intercellular signaling protein,

HetN, is required for patterning of heterocysts in Anabaena sp. strain PCC 7120. Mol.

Microbiol., 2012. 83: p. 682-693.

18

CHAPTER 2. ROLE OF HETF AND PATU3 IN THE REGULATION OF HETEROCYST

DEVELOPMENT IN ANABAENA SP. STRAIN PCC 7120

INTRODUCTION

During development in multicellular organisms, the stages of growth, cell division and

morphogenesis occur in a tightly orchestrated manner. Programmed responses properly execute

the intricate progression of cellular differentiation. To this end, regulation of genes and the

products of genes are temporally and spatially coordinated. This regulation allows for the

generation of diverse cell types despite the fundamental sameness of the genetic complement

shared by all cells in the organism. Furthermore, regulation at the single cell level can both

promote and respond to extracellular signals or environmental cues. How this complex interplay

results in a pattern of diverse cell types remains a central question in developmental biology.

Cyanobacteria are key contributors of biologically fixed nitrogen in the environment, an essential

component of the macromolecules of life. Cellular differentiation of a periodic pattern of

specialized nitrogen-fixing cells in the cyanobacterium Anabaena sp. strain PCC 7120 (herein,

Anabaena) is a simple model system for the study of developmental biology. Proper control of

the genetic regulatory network governing heterocyst formation is critical to the development of

Anabaena. Maintenance of the periodicity of cellular differentiation, which results in

approximately one heterocyst separating every ten undifferentiated vegetative cells, is crucial for

the supply of fixed nitrogen to vegetative cells and reductant to heterocysts that supports robust

growth under fixing conditions. Populations of filaments genetically manipulated to give an

increased distribution of terminally differentiated heterocyst cells that no longer divide have a

slower growth rate than the wild-type. Likewise, filaments that elaborate a reduced number of

heterocysts have reduced growth under fixing conditions.

Developmental genes have been identified that regulate heterocyst differentiation and the pattern

of heterocyst formation (Chapter 1). In addition to the heterocyst differentiation genes hetR, hetF

and patA, the gene hetZ [1] also appears to positively regulate heterocyst differentiation.

Located in a gene cluster consisting of three genes, hetZ, patU5 and patU3, together these genes

Figure 2.1. The hetZ-patU5-patU3 gene cluster in Anabaena and locations of transcripts I

and II (tsp I and tsp II). The gene patU5 overlaps hetZ by eight base pairs.

19

coordinate heterocyst differentiation and pattern formation (Fig. 2.1). Deletion of the hetZ gene

abrogates heterocyst development. HetZ appears to regulate heterocyst differentiation by

promoting the expression of other heterocyst development genes such as patS, a negative

regulator of heterocyst formation containing the RGSGR pentapeptide sequence (“PatS-5”;

Chapter 1), and hetC, a predicted ABC transporter required for transition to the non-dividing,

terminally differentiated state of the heterocyst [1, 2]. Regulation of hetZ is connected to the

regulatory network that includes HetR and the RGSGR pentapeptide. HetR binds to a recognition

sequence upstream of the only transcriptional start site of hetZ to regulate hetZ expression[3].

The up-regulation of hetZ is induced within six hours of nitrogen step-down. Although expressed

in both cell types, transcription appears more pronounced in cells that will differentiate. In

addition to the dependence on HetR, expression of hetZ is enhanced by patU3 and repressed by

both hetZ itself as well as through exogenous addition of the RGSGR pentapeptide[3].

Arrangement of the hetZ-patU5-patU3 operon (annotated as alr0099, asr0100 and alr001

respectively) presumably facilitates co-transcription of these genes to coordinate some aspect of

heterocyst development and pattern formation. Two major transcripts, transcript I and transcript II

were originally described to be associated with the gene cluster[1]. Additional transcriptional start

points (tsps) have since been identified in this gene cluster[4] and are described as appropriate in

the Results section. The transcriptional start point for transcript I (“tsp I”) lies 425 bp upstream of

hetZ. A -10 element and a non-canonical NtcA binding site are located upstream of the -425

transcriptional start point. The tsp for transcript II (“tsp II”) lies upstream of patU5, at a site within

hetZ. This clustering of hetZ, patU5 and patU3 may facilitate both the concerted as well as

independent expression of the two major transcripts in order to coordinate heterocyst

development and patterning in Anabaena.

In Nostoc punctiforme, where this genetic locus was discovered, patU5 and patU3 form two parts

of the single patU gene, suggesting that these two gene products may work together in Anabaena.

The patU5 gene, the middle gene in the hetZ-patU5-patU3 gene cluster, is not well characterized.

Because it overlaps the upstream hetZ gene by eight base pairs, this genomic arrangement may

indicate the possibility for coordinated expression of hetZ-patU5. The phenotype of a strain with

patU5 inactivated has not been described. PatU5, and the gene product of the third gene in this

gene cluster, PatU3, do not exhibit homology to protein motifs of known biological function.

However the Mch (multiple contiguous heterocyst) phenotype observed upon inactivation of

patU3 in both the wild-type and patA mutant backgrounds suggests a role for PatU3 in the control

of heterocyst patterning. Overexpression of patU3 from an inducible promoter has been shown to

block the formation of terminal heterocysts in a patA-deletion mutant. Since the patU3 mutant

Mch phenotype parallels that of a patS mutant, PatU3 and PatS may interact through mutual

regulatory pathways. Both patU3 and patS are expressed during early heterocyst development.

20

In gfp reporter strains, patU3 is strongly expressed in developing heterocysts (“proheterocysts”)

and heterocysts, in a spatio-temporal manner that resembles that of patS expression. Unlike

hetZ, patU3 is only very weakly transcribed in vegetative cells, but up-regulation is also evident at

six hours and HetR-dependent[1].

Homologs of hetR, patS and the hetZ-patU5-patU3 gene cluster are found in both heterocyst-

forming and non-heterocyst forming cyanobacteria. These genes represent components of the

minimal gene set shared among filamentous cyanobacteria. As such, homologs of this minimal

gene set may have a more general role other than the control of differentiation and patterning.

In contrast, hetF-like genes are found only in heterocyst-forming strains. A relationship between

hetF and patU3 was identified in this study and integrated into a novel genetic regulatory network

along with other known developmental genes. The goal of this study is to understand how hetF

and patU3 function together as cellular differentiation regulatory factors in the filamentous

cyanobacterium Anabaena.

MATERIALS AND METHODS

Bacterial strains and growth conditions. Descriptions of strains constructed in this study are

summarized in Table 2.1. Growth of Escherichia coli and Anabaena sp. strain PCC 7120 and its

derivatives, concentrations of antibiotics; induction of heterocyst formation in BG-110 medium,

which lacks a combined-nitrogen source; regulation of PpetE and Pnir expression; and conditions

for photomicroscopy were as previously described[5]. Images were processed in Adobe

Photoshop CS2.

Construction of plasmids. Descriptions of plasmids constructed in this study are summarized

in Table 2.2 and oligonucleotides relevant to this study are summarized in Table 2.3. Constructs

derived by PCR were sequenced to verify the integrity of the sequence.

Suicide plasmids

Plasmid pST214 was used to replace most of the coding region of patU3 with an Ω interposon.

An 842 bp region containing the first 42-bp of the patU3 coding region and upstream DNA was

amplified from the chromosome (with primers patU3 up F and patU3 up R) and fused to an 855-

bp region containing the last 49-bp of the patU3 coding region and downstream DNA (using

primers patU3 down F and patU3 down R) via overlap extension PCR. The 1697-bp fragment

(containing patU3∆42-729) was moved into the EcoRV site of pBluescript SK+ (Stratagene) to

generate pST215. The 1697-bp fragment was moved from pST215 into the suicide vector

pRL278 as a BglII-SacI fragment using restriction sites introduced on the primers to create

21

pST213. To generate pST214, the 2082-bp Ω interposon, which confers resistance to

spectinomycin and streptinomycin [6], was introduced into pST213 at a SmaI site introduced on

the primers used for overlap extension PCR.

Plasmid pST373 was used to delete most of the coding region of patU3. The 1697-bp fragment

(containing patU3∆42-729) was moved from pST215 into pRL277 as a BglII-SacI fragment to

create pST373.

Plasmid pST502 was used to delete the coding region of patU3. An 836-bp region containing the

first 36-bp of the patU3 coding region and upstream DNA was amplified from the chromosome

(with primers patU3 up F and patU3 up R DHRB1156) and fused to an 943-bp region containing

the last 136-bp of the patU3 coding region and downstream DNA (using primers patU3 down F

DHRB1156 and patU3 down R) via overlap extension PCR. The 1779-bp fragment (containing

patU3∆36-641) was moved into the EcoRV site of pBluescript SK+ to generate pST501. The

1779-bp fragment was moved from pST501 into pRL277 as a BglII-SacI fragment using

restriction sites introduced on the primers to create pST502.

Plasmid pST504 was used to replace the coding region of patU3 with a Ω interposon, recreating

DHRB1156[1]. The 1779-bp fragment (containing patU3∆36-641) was moved from pST501 into

pRL278 as a BglII-SacI fragment using restriction sites introduced on the primers to create

pST503. To generate pST504, the 2082-bp Ω interposon was introduced into pST503 at a SmaI

site introduced on the primers used for overlap extension PCR.

Plasmid pST518 was used to construct patU3(E69stop). An 1010-bp region containing the first

210 bp of the patU3 coding region and upstream DNA was amplified from the chromosome (with

primers patU3 up F and patU3 glu69stop up R) and fused to an 1377-bp region containing the

last 570-bp of the patU3 coding region and downstream DNA (using primers patU3 glu69stop

down F and patU3 down R) via overlap extension PCR. The 2387-bp fragment was moved into

the EcoRV site of pBluescript SK+ to generate pST519. The 2387-bp fragment was moved from

pST519 into pRL277 as a BglII-SacI fragment using restriction sites introduced on the primers to

create pST518.

Plasmid pST370 was used to delete most of the coding region of hetZ. An 853-bp region

containing the first 33-bp of the hetZ coding region and upstream DNA was amplified from the

chromosome (with primers hetZ up F and hetZ up R) and fused to an 863-bp region containing

the last 48-bp of the hetZ coding region and downstream DNA (using primers hetZ down F and

hetZ down R) via overlap extension PCR. The 1716-bp fragment (containing hetZ∆33-1155) was

22

moved into the EcoRV site of pBluescript SK+ to generate pST369. The 1716-bp fragment was

moved from pST369 into pRL277 as a BglII-SacI fragment to create pST370.

Plasmid pST422 was used to delete the coding region of hetZ. An 1174-bp region containing the

first 354-bp of the hetZ coding region and upstream DNA was amplified from the chromosome

(with primers hetZ up F and hetZ +tsp up R) and fused to an 1259-bp region containing the last

444-bp of the hetZ coding region and downstream DNA (using primers hetZ +tsp II U5U3 down F

1 and hetZ down R) via overlap extension PCR. The 2243-bp fragment (containing hetZ∆354-

762) was moved into the EcoRV site of pBluescript SK+ to generate pST434. The 2243-bp

fragment was moved from pST421 into pRL277 as a BglII-SacI fragment to create pST422.

Plasmid pST424 was used to replace the coding region of hetZ with a Ω interposon. The 2243-

bp fragment was moved from pST434 into pRL278 as a BglII-SacI fragment using restriction sites

introduced on the primers to create pST423. To generate pST424, the 2082-bp Ω interposon

was introduced into pST423 at a SmaI site introduced on the primers used for overlap extension

PCR.

Plasmid pST594 was used to construct patU5(L12stop). An 938-bp region containing the first 36-

bp of the patU5 coding region and upstream DNA was amplified from the chromosome (with

primers patU5 up F and patU5 L12stop up R) and fused to an 975-bp region containing the last

144-bp of the patU5 coding region and downstream DNA (using primers patU5 L12stop down F

and patU5 down R) via overlap extension PCR. The 1917 bp fragment was moved into pRL277

as a BglII-SacI fragment using restriction sites introduced on the primers to create pST594.

Multicopy plasmids

The mobilizable shuttle vector, plasmid pST352 was created using primers alr0101 BamHI F and

asr0098 SacI R to amplify PhetZ-asr0098 from the chromosome. The 614-bp fragment was moved

as a BamHI-SacI fragment into the EcoRV site of pBluescript SK+ to generate pST348. To

create pST352, the fragment was subsequently moved with the same restriction sites into the

replicating plasmid pAM504.

Plasmid pST353 was created using primers alr0101 BamHI F and hetZ SacI R to amplify PhetZ-

hetZ from the chromosome. The 1683-bp fragment was moved as a BamHI-SacI fragment into

the EcoRV site of pBluescript SK+ to generate pST349. To create pST353, the fragment was

subsequently moved with the same restriction sites into pAM504.

23

Plasmid pST354 was created using primers alr0101 BamHI F and patU5 SacI R to amplify PhetZ-

patU5 from the chromosome. The 1851-bp fragment was moved as a BamHI-SacI fragment into

the EcoRV site of pBluescript SK+ to generate pST350. To create pST354 (containing PhetZ-hetZ-

patU5), the fragment was subsequently moved with the same restriction sites into pAM504.

Plasmid pST590 is a derivative of pST354, and bears a truncation in hetZ. The internal primers

hetZ up R SmaI and hetZ down F SmaI were used to amplify plasmid pST354 without the

hetZ∆354-762 region (to “knock out” hetZ). To give plasmid pST590 (carrying PhetZ-hetZ∆354-

762-patU5), the PCR product was moved into pAM504 using the SmaI restriction sites introduced

on the primers.

Plasmid pST355 was created using primers alr0101 BamHI F and patU3 SacI R to amplify PhetZ-

patU3 from the chromosome. The 2647 bp fragment was moved as a BamHI-SacI fragment into

the EcoRV site of pBluescript SK+ to generate pST351. To create pST355 (containing PhetZ-hetZ-

patU5-patU3), the fragment was subsequently moved with the same restriction sites into pAM504.

Plasmids pST429, pST430, pST431, pST587, pST589, pST591 and pST592 are derivatives of

plasmid pST355, carrying truncations in either one or two of the genes (hetZ, patU5 and/or patU3)

present on pST355. Plasmid pST429 carries a truncation in the hetZ coding region. The internal

primers hetZ up R SmaI and hetZ down F SmaI were used to amplify plasmid pST355 without the

hetZ∆354-762 region (to “knock out” hetZ). To give plasmid pST429 (carrying PhetZ-hetZ∆354-

762-patU5-patU3), the PCR product was moved into pAM504 using the SmaI restriction sites

introduced on the primers.

Plasmid pST430 carries a truncation in the patU5 coding region. The internal primers patU5 up R

SmaI and patU5 down F SmaI were used to amplify plasmid pST355 without the patU5∆45-126

region (to “knock out” patU5). To give plasmid pST430 (carrying PhetZ-hetZ-patU5∆45-126-patU3),

the PCR product was moved into pAM504 using the SmaI sites introduced on the primers.

A stop codon engineered at patU5(L12) is carried on plasmid pST596. A fragment containing the

first 36-bp of the patU5 coding region and upstream DNA was amplified from the chromosome

(with primers alr0101 BamHI F and patU5 L12stop up R) and fused to a fragment containing the

last 144-bp of the patU5 coding region and downstream DNA (using primers patU5 L12stop down

F and patU3 SacI R) via overlap extension PCR. The fragment was moved into pAM504 as a

BamHI-SacI fragment using restriction sites introduced on the primers to create pST596 (carrying

PhetZ-hetZ-patU5 L12stop-patU3).

24

Plasmid pST431 carries a truncation in the patU3 coding region. The internal primers patU3 up R

SmaI and patU3 down F SmaI were used to amplify plasmid pST355 without the patU3∆42-729

region (to “knock out” patU3). To give plasmid pST431 (carrying PhetZ-hetZ-patU5-patU3∆42-729),

the PCR product was moved into pAM504 using the SmaI sites introduced on the primers.

Plasmid pST591 carries truncations in the coding regions of hetZ and patU3. The internal

primers patU3 up R SmaI and patU3 down F SmaI were used to amplify plasmid pST429 without

the patU3∆42-729 region (to “knock out” patU3 from plasmid pST429). To give plasmid pST591

(carrying PhetZ-hetZ∆354-762-patU5-patU3∆42-729), the PCR product was moved into pAM504

using the SmaI sites introduced on the primers.

Plasmid pST592 carries truncations in the coding regions of hetZ and patU5. The internal

primers hetZ up R SmaI and hetZ down F SmaI were used to amplify plasmid pST430 without the

patU3∆42-729 region (to “knock out” hetZ from plasmid pST430). To give plasmid pST592

(carrying PhetZ-hetZ∆354-762-patU5∆45-126-patU3), the PCR product was moved into pAM504

using the SmaI sites introduced on the primers.

Plasmid pST586 carries patU5-patU3 under the control of tsp I. The internal primers PI hetZ up

R SmaI and PI hetZ down F SmaI were used to amplify plasmid pST355 without the hetZ∆33-

1155 region (to remove tsp II from pST355). To give plasmid pST596 (carrying hetZ∆33-1155-

patU5-patU3), the PCR product was moved into pAM504 using the SmaI sites introduced on the

primers.

Plasmid pST587 carries patU5-patU3 under the control of tsp I and tsp II. The internal primers PI

hetZ up R SmaI and hetZ down F SmaI were used to amplify plasmid pST355 without the

hetZ∆33-762 region. To give plasmid pST596 (carrying hetZ∆33-762-patU5-patU3), the PCR

product was moved into pAM504 using the SmaI sites introduced on the primers.

Plasmid pST589, carrying patU5-patU3 under the control of tsp II, was created using the primers

PII hetZ down F BamHI and patU3 SacI R and chromosomal DNA as template. To give plasmid

pST596 (carrying hetZ∆1-762-patU5-patU3), the PCR product was moved into pAM504 as a

BamHI-SacI fragment using the same sites introduced on the primers.

Overexpression vectors

Plasmids pST216 and pST255 carry a transcriptional fusion between the petE promoter and

patU3. The coding region of patU3 was amplified from the chromosome using primers patU3 OE-

25

EcoRI-F and patU3 OE-BamHI-R and ligated into the EcoRV site of pBluescript SK+ to generate

pST227. The fragment was moved from pST227 as an EcoRI-BamHI fragment into pKH256,

designed to create transcriptional fusions to PpetE [7], to generate pST216.

Plasmid pST255 differs from pST216 in the engineered ribosome-binding site that was introduced

by using primer patU3-EcoRI-rbs-F in place of patU3 OE-EcoRI-F to amplify patU3 from the

chromosome. The PCR fragment was moved into the EcoRV site of pBluescript SK+ to generate

pST256, and subsequently moved from pST256 as an EcoRI-BamHI fragment into pKH256 to

generate pST255.

Plasmid pST315 bears a transcriptional fusion between the petE promoter and patU5 and patU3.

The region containing patU5 and patU3 was amplified from the chromosome using primers patU5

EcoRI F (containing a ribosome-binding site) and patU3 OE-BamHI-R and ligated into the EcoRV

site of pBluescript SK+ to generate pST322. The fragment was moved from pST322 as an

EcoRI-BamHI fragment into pKH256 to generate pST315.

Plasmids pST414, pST415, pST416, pST417, pST418, pST419 and pST420 bear transcriptional

fusions to the nirA promoter. The coding region of hetZ was amplified from the chromosome

using primers hetZ SmaI F and hetZ KpnI R and ligated into the EcoRV site of pBluescript SK+ to

generate pST407. To make pST414, the fragment was moved from pST407 as a SmaI-KpnI

fragment into pDR311 (bearing a Pnir-lacZ transcriptional fusion[8]) digested with the same

enzymes to exchange lacZ for hetZ.

Primers hetZ SmaI F and patU5 KpnI R were used to amplify hetZ and patU5 from the

chromosome, and ligated into the EcoRV site of pBluescript SK+ to make pST408. To yield

pST415, the fragment was moved from pST408 as a SmaI-KpnI fragment into pDR311 digested

with the same enzymes.









26

Primers patU5 SmaI F and patU3 KpnI R were used to amplify patU5 and patU3 and the

intergenic region between these genes from the chromosome, and ligated into the EcoRV site of

pBluescript SK+ to make pST410. To yield pST417, the fragment was moved from pST410 as a

SmaI-KpnI fragment into pDR311 digested with the same enzymes.

Primers patU5 SmaI F and patU5 KpnI R were used to amplify patU5 from the chromosome, and

ligated into the EcoRV site of pBluescript SK+ to make pST411. To yield pST418, the fragment

was moved from pST411 as a SmaI-KpnI fragment into pDR311 digested with the same enzymes.







A fragment containing hetZ (amplified using primers hetZ SmaI F and hetZ patU3 up R) was

fused to a fragment containing patU3 (amplified using primers hetZ patU3 down F and patU3

KpnI R) via overlap extension PCR, and cloned into the EcoRV site of pBluescript SK+ to make

pST413. To yield pST420 (bearing Pnir-hetZ-patU3), the fragment was moved from pST413 as a

SmaI-KpnI fragment into pDR311 digested with the same enzymes.

Bacterial two-hybrid constructs

For each construct, a 32- or 33-bp linker domain was introduced between the adenylate cyclase

fragment and the protein of interest on either the forward or reverse primer.

The hetR coding region was amplified from the chromosome using the primers hetR BamHI F A

and hetR EcoRI R A, and cloned into the EcoRV site of pBluescript SK+ to make pST544. To

make pST558, the 900 bp fragment was moved as a BamHI-EcoRI fragment from pST544 into

pKT25. To make pST565, the fragment was moved as a BamHI-EcoRI fragment directly into

pUT18C.

The hetR coding region was amplified from the chromosome using the primers hetR BamHI F B

and hetR EcoRI R B, and cloned into the EcoRV site of pBluescript SK+ to make pST551. To

27

make pST572 and pST579, the 900 bp fragment was moved as a BamHI-EcoRI fragment from

pST551 into pKNT25 and pUT18 respectively.

The hetF coding region was amplified from the chromosome using the primers hetF BamHI F

GTG1 A and hetF EcoRI R A, and cloned into the EcoRV site of pBluescript SK+ to make

pST545. To make pST559, the 2487 bp fragment was moved as a BamHI-EcoRI fragment from

pST544 into pKT25. To make pST566, the fragment was moved as a BamHI-EcoRI fragment

directly into pUT18C.

Primers hetF BamHI F GTG1 A and hetF EcoRI R A were used to amplify hetF(C246A) from

pDR284[8]. To yield pST597 and pST598, the 2487 bp fragment was moved as a BamHI-EcoRI

fragment directly into pKT25 and pUT18C respectively.

The hetF coding region was amplified from the chromosome using the primers hetF BamHI F

GTG1 B and hetF EcoRI R B, and cloned into the EcoRV site of pBluescript SK+ to make

pST552. To make pST573 and pST580, the 2487 bp fragment was moved as a BamHI-EcoRI

fragment from pST552 into pKNT25 and pUT18 respectively.

Primers hetF BamHI F GTG1 B and hetF EcoRI R B were used to amplify hetF(C246A) from

pDR284. To yield pST599 and pST600, the 2487 bp fragment was moved as a BamHI-EcoRI

fragment directly into pKNT25 and pUT18 respectively.

The hetZ coding region was amplified from the chromosome using the primers hetZ BamHI F A

and hetZ EcoRI R A. To make pST560 and pST567, the 1206 bp fragment was moved as a

BamHI-EcoRI fragment directly into pKT25 and pUT18C respectively.

The hetZ coding region was amplified from the chromosome using the primers hetZ BamHI F B

and hetZ EcoRI R B. To make pST574 and pST581, the 1206 bp fragment was moved as a

BamHI-EcoRI fragment directly into pKNT25 and pUT18 respectively.

The patU5 coding region was amplified from the chromosome using the primers patU5 BamHI F

A and patU5 EcoRI R A. To make pST561 and pST568, the 177 bp fragment was moved as a

BamHI-EcoRI fragment into directly into pKT25 and pUT18C respectively.


B and patU5 EcoRI R B. To make pST575 and pST582, the 177 bp fragment was moved as a

BamHI-EcoRI fragment into directly into pKNT25 and pUT18 respectively.

28


A and patU3 EcoRI R A, and cloned into the EcoRV site of pBluescript SK+ to make pST548. To



pUT18C.


B and patU3 EcoRI R B, and cloned into the EcoRV site of pBluescript SK+ to make pST555. To


pST555 into pKNT25 and pUT18 respectively.

The patA coding region was amplified from the chromosome using the primers patA BamHI F A

and patA EcoRI R A, and cloned into the EcoRV site of pBluescript SK+ to make pST550. To



pUT18C.

The patA coding region was amplified from the chromosome using the primers patA BamHI F B

and patA EcoRI R B, and cloned into the EcoRV site of pBluescript SK+ to make pST557. To


pST557 into pKTN25 and pUT18 respectively.

Translational fusion constructs

The translational patU3-gfp reporter fusion under the control of the hetZ promoter was created

using primers alr0101 BamHI F and alr0101 SmaI R to amplify the 2624-bp fragment from the

chromosome. The fragment was ligated into the EcoRV site of pBluescript SK+ to make pST290.

The fragment was subsequently moved as a BamHI-EcoRI fragment into pSMC232, a replicating

plasmid used to generate translational fusions to gfp[8], to give pST291.

The translational patU3-gfp reporter fusion under the control of the hetZ promoter was created

using primers PpetE-BamHI-F and alr0101 SmaI R to amplify a 1119-bp fragment containing

PpetE-patU3 from pST216. The fragment was ligated into the EcoRV site of pBluescript SK+ to

make pST292. The fragment was subsequently moved as a BamHI-EcoRI fragment into

pSMC232 to give pST293.

29

Strain construction. Descriptions of strains constructed in this study are summarized in Table

2.1. Deletions and replacements of chromosomal DNA was performed as previously described [9].

All strains were screened via colony PCR with primers annealing outside of the chromosomal

region introduced on the suicide plasmids used for strain construction and tested for sensitivity

and resistance to the appropriate antibiotics. Additionally, strains with engineered stop codons in

patU5 and patU3 were sequenced to verify the presence of the mutation. For strains with

mutations in more than one gene, mutations were introduced in the order indicated in the strain

description.

Plasmid pST370 was used to cleanly delete nucleotides +33 to +1155 relative to the ATG

translational initiation codon of hetZ (denoted as hetZ∆33-1155) from strains PCC 7120, ∆hetF,

∆patA ∆hetF and ∆hetR ∆patA ∆hetF to create strains hetZ∆33-1155, ∆hetF hetZ∆33-1155,

∆patA ∆hetF hetZ∆33-1155 and ∆hetR ∆patA ∆hetF hetZ∆33-1155.

Strain patU3∆42-729 hetZ∆33-1155 was constructed by introduction of pST370 into strain

patU3∆42-729. Colonies representing single recombinants did not arise after repeated attempts

to introduce pST373 into hetZ∆33-1155 to generate hetZ∆33-1155 patU3∆42-729.


translational initiation codon of hetZ (denoted as hetZ∆354-762) from strains PCC 7120 and

∆hetF to create strains hetZ∆354-762 and ∆hetF hetZ∆354-762.

Plasmid pST424 was used to replace nucleotides +354 to +762 relative to the ATG translational

initiation codon of hetZ (denoted as hetZ∆354-762) with an Ω cassette. Plasmid pST424 was

introduced into strains ∆hetF, ∆patA ∆hetF and ∆hetR ∆patA ∆hetF to create strains ∆hetF

hetZ∆354-762 with Ω cassette, ∆patA ∆hetF hetZ∆354-762 with Ω cassette and ∆hetR ∆patA

∆hetF hetZ∆354-762 with Ω cassette.


translational initiation codon of patU3 (denoted as patU3∆42-729) from strains PCC 7120, ∆hetF,

∆patA ∆hetF and ∆hetR ∆patA ∆hetF to generate strains patU3∆42-72, ∆hetF patU3∆42-729,

∆patA ∆hetF patU3∆42-729 and ∆hetR ∆patA ∆hetF patU3∆42-729.

Strain patU3∆42-729 ∆hetR was constructed by introduction of pSMC104 into strain patU3∆42-

729. Colonies representing single recombinants did not arise after repeated attempts to

introduce pST373 into ∆hetR to generate ∆hetR patU3∆42-729.

30


initiation codon of patU3 (denoted as patU3∆42-729) with an Ω cassette. Plasmid pST214 was

introduced into strains PCC 7120, ∆hetF and ∆patA ∆hetF to generate strains patU3∆42-729 with

Ω cassette, ∆hetF patU3∆42-729 with Ω cassette and ∆patA ∆hetF patU3∆42-729 with Ω

cassette.


translational initiation codon of patU3 (denoted as patU3∆36-641) from strain PCC 7120 to

generate strain patU3∆36-641.


initiation codon of patU3 (denoted as patU3∆36-641) with an Ω cassette. Plasmid pST502 was

introduced into strain PCC 7120 to generate strain patU3∆36-641 with Ω cassette.

Plasmid pST518 was used to engineer a stop codon at residue E69 of the translated product of

patU3 (denoted as patU3(E69stop)) from strains PCC 7120, ∆hetF and ∆patA ∆hetF to generate

strains patU3(E69stop), ∆hetF patU3(E69stop), and ∆patA ∆hetF patU3(E69stop). All double

recombinants tested in the latter two parental strains did not have the E69stop mutation.

Plasmid pST594 was used to engineer a stop codon at residue L12 of the translated product of

patU5 (denoted as patU5(L12stop)) from strains PCC 7120, ∆hetF and ∆patA ∆hetF to generate

strains patU5(L12stop), ∆hetF patU5(L12stop), and ∆patA ∆hetF patU5(L12stop). All double

recombinants tested did not have the mutation irrespective of the parental strain that was used.

Genetic screens. To isolate spontaneous suppressor bypass mutants of parent strains ∆hetF

and ∆patA ∆hetF, 1 µL of plasmid pDR138 containing PhetR-hetR was introduced into the

hypermutator E. coli strain XL1-RED (Stratagene). The resulting library of mutations in PhetR-hetR

was subsequently introduced into both hetF and hetF patA background strains and plated directly

onto media lacking a source of combined nitrogen (BG-11˳) and antibiotics. Mutants that arose

were sequenced at specific sites in the chromosome (PhetR-hetR, asr0098, and PhetZ-hetZ-patU5-

patU3) to search for mutations in these regions. In addition, the absence of either or both hetF

and patA in these mutants as appropriate was verified by colony PCR.

Bacterial two-hybrid and β-galactosidase assays. Bacterial two-hybrid assays were

performed as previously described [10-12]. The average β-galactosidase-specific activity (three

replicates) of each positive protein-protein interaction was determined as previously described

[12].

31

Table 2.1. Strains used in Chapter 2

Anabaena sp. strain

Relevant characteristic(s) Source or reference (UHM designation)

PCC 7120 Wild-type Pasteur culture collection

∆hetF hetF-deletion strain [8] (UHM130)

∆patA ∆hetF patA-, hetF-deletion strain [8] (UHM135)

∆hetR ∆patA ∆hetF

hetR-, patA-, hetF-deletion strain [8] (UHM 134)

∆patS patS-deletion strain [9] (UHM114)

∆hetN with Ω cassette

hetN-deletion strain with Ω cassette [9] (UHM115)

∆hetN hetN-deletion strain [13] (UHM150)

hetZ∆33-1155 hetZ∆33-1155-deletion strain This study (UHM309)

∆hetF hetZ∆33-1155

hetF-, hetZ∆33-1155-deletion strain This study (UHM310)

∆patA ∆hetF hetZ∆33-1155

patA-, hetF-, hetZ∆33-1155-deletion strain This study (UHM311)

∆hetR ∆patA ∆hetF hetZ∆33-1155

hetR-, patA-, hetF-, hetZ∆33-1155-deletion strain This study (UHM312)

hetZ∆354-762 hetZ∆354-762-deletion strain This study (UHM313)

∆hetF hetZ∆354-762

hetF-, hetZ∆354-762-deletion strain This study (UHM314)

∆hetF hetZ∆354-762 with Ω cassette

hetF-, hetZ-strain with hetZ bp 354-762 replaced with Ω cassette

This study (UHM315)

∆patA ∆hetF hetZ∆354-762

with Ω cassette

patA-, hetF-, hetZ-strain with hetZ bp 354-762 replaced with Ω cassette

This study (UHM316)

∆hetR ∆patA ∆hetF hetZ∆354-762 with Ω cassette

hetR-, patA-, hetF-, hetZ-strain with hetZ bp 354-762 replaced with Ω cassette

This study (UHM317)

patU3∆42-729 patU3∆42-729-deletion strain This study (UHM319)

∆hetF patU3∆42-729

hetF-, patU3∆42-729-deletion strain This study (UHM320)

∆patA ∆hetF patU3∆42-729

patA-, hetF-, patU3∆42-729-deletion strain This study (UHM321)

∆hetR ∆patA ∆hetF patU3∆42-729

hetR-, patA-, hetF-, patU3∆42-729-deletion strain This study (UHM322)

patU3∆42-729 with Ω cassette

patU3-deletion strain with patU3 bp 42-729 replaced with Ω cassette

This study (UHM221)

∆hetF patU3∆42-729 with Ω cassette

hetF-, patU3-deletion strain with patU3 bp 42-729 replaced with Ω cassette

This study (UHM226)

32

Table 2.1. (Continued) Strains used in Chapter 2

Anabaena sp. strain


∆patA ∆hetF patU3∆42-729 with Ω cassette

patA-, hetF-, patU3-deletion strain with patU3 bp 42-729 replaced with Ω cassette

This study (UHM222)

patU3∆36-641 patU3∆36-641-deletion strain This study (UHM323)

patU3∆36-641 with Ω cassette

patU3-deletion strain with patU3 bp 36-641 replaced with Ω cassette

This study (UHM324)

patU3(E69stop) patU3 strain with an engineered stop codon at E69 This study (UHM332)

patU3∆42-729 ∆hetR

patU3∆42-729-, hetR-deletion strain This study (UHM338)

patU3∆42-729 hetZ∆33-1155

patU3∆42-729-, hetZ∆33-1155-deletion strain This study (UHM339)

33

Table 2.2. Plasmids used in Chapter 2

Plasmids Relevant characteristic(s) Source or reference

pAM504 Mobilizable shuttle vector for replication in E. coli and Anabaena; Km

r Neo

r [14]

pAM1951 pAM505 with PpatS-gfp [15]

pRL277 Suicide vector; Smr Sp

r [16]

pRL278 Suicide vector; Neor [16]

pROEX-1 Expression vector for generating polyhistidine epitope-tagged proteins; Ap

r

Life Technologies

pDR138 pAM504 carrying PhetR-hetR [17]

pDR211 pAM504 carrying PpetE-patS [18]

pDR320 pAM504 carrying PpetE-hetN [18]

pDR292 pAM504 carrying PhetR-hetR-gfp [18]

pDR293 pAM504 carrying PpetE-hetR-gfp [19]

pDR311 pAM504 carrying Pnir-lacZ [8]

pDR350 pAM504 carrying PpetE-lacZ [8]

pDR336 pET21a carrying hetFGTG1H6 [19]

pKH256 pAM504 bearing PpetE for transcriptional fusions [7]

pSMC104 Suicide vector used to delete hetR [9]

pSMC127 pAM504 carrying PhetR-gfp [20]

pSMC232 pAM504 bearing promotorless gfp for translational fusions [8]

pSMC257 pPROEX-1 carrying patU3 [19]

pSMC258 pPROEX-1 carrying hetZ [19]

pKT25 Plasmid carrying the T25 fragment of CyaA for C-terminal protein fusions

[10]

pKNT25 Plasmid carrying the T25 fragment of CyaA for N-terminal protein fusions

[10]

pUT18C Plasmid carrying the T18 fragment of CyaA for C-terminal protein fusions

[10]

pUT18 Plasmid carrying the T18 fragment of CyaA for N-terminal protein fusions

[10]

pST291 pAM504 carrying PhetZ-hetZ-patU5-patU3-gfp This study

pST293 pAM504 carrying PpetE-patU3-gfp This study

pST352 pAM504 carrying PhetZ-asr0098 This study

pST353 pAM504 carrying PhetZ-hetZ This study

pST354 pAM504 carrying PhetZ-hetZ-patU5 This study

pST355 pAM504 carrying PhetZ-hetZ-patU5-patU3 This study

pST429 pAM504 carrying PhetZ-hetZ∆354-762-patU5-patU3 This study

pST430 pAM504 carrying PhetZ-hetZ-patU5∆45-126-patU3 This study

pST431 pAM504 carrying PhetZ-hetZ-patU5-patU3∆42-729 This study

pST216 pAM504 carrying PpetE-patU3 This study

pST255 pAM504 carrying strong ribosomal binding site and PpetE-patU3

This study

pST315 pAM504 carrying PpetE-patU5-patU3 This study

34

Table 2.2. (Continued) Plasmids used in Chapter 2


pST414 pAM504 carrying Pnir-hetZ This study

pST415 pAM504 carrying Pnir-hetZ-patU5 This study

pST416 pAM504 carrying Pnir-hetZ-patU5-patU3 This study

pST417 pAM504 carrying Pnir-patU5-patU3 This study

pST418 pAM504 carrying Pnir-patU5 This study

pST419 pAM504 carrying Pnir-patU3 This study

pST420 pAM504 carrying Pnir-hetZ-patU3 This study

pST596 pAM504 carrying PhetZ-hetZ-patU5(L12stop)-patU3 This study

pST590 pAM504 carrying PhetZ-hetZ∆354-762-patU5 This study

pST591 pAM504 carrying PhetZ-hetZ∆354-762-patU5-patU3∆42-729 This study

pST592 pAM504 carrying PhetZ-hetZ∆354-762-patU5∆45-126-patU3 This study

pST586 pAM504 carrying tsp I-patU5-patU3 (hetZ∆33-1155) This study

pST587 pAM504 carrying tsp I-tsp II-patU5-patU3 (hetZ∆33-762) This study

pST589 pAM504 carrying tsp II-patU5-patU3 (hetZ∆1-762) This study

pST291 pSMC232 carrying PhetZ-patU3 This study

pST293 pSMC232 carrying PpetE-patU3 This study

pST370 Suicide plasmid used to delete hetZ∆33-1155 This study

pST422 Suicide plasmid used to delete hetZ∆354-762 This study

pST424 Suicide plasmid used to replace hetZ∆354-762 with an Ω interposon

This study

pST214 Suicide plasmid used to replace patU3∆42-729 with an Ω interposon

This study

pST502 Suicide plasmid used to delete patU3∆36-641 This study

pST504 Suicide plasmid used to replace patU3∆36-641 with an Ω interposon

This study

pST518 Suicide plasmid used to construct patU3(E69stop) This study

pST594 Suicide plasmid used to construct patU5(L12stop) This study

pST558 pKT25 carrying hetR This study

pST559 pKT25 carrying hetF This study

pST560 pKT25 carrying hetZ This study

pST561 pKT25 carrying patU5 This study

pST562 pKT25 carrying patU3 This study

pST564 pKT25 carrying patA This study

pST597 pKT25 carrying hetF(C246A) This study

pST565 pUT18C carrying hetR This study

pST566 pUT18C carrying hetF This study

pST567 pUT18C carrying hetZ This study

pST568 pUT18C carrying patU5 This study

pST569 pUT18C carrying patU3 This study

pST571 pUT18C carrying patA This study

pST598 pUT18C carrying hetF(C246A) This study

pJP42 pKNT25 carrying divIVA (Bacillus subtilis) [12]

35



pST572 pKNT25 carrying hetR This study

pST573 pKNT25 carrying hetF This study

pST574 pKNT25 carrying hetZ This study

pST575 pKNT25 carrying patU5 This study

pST576 pKNT25 carrying patU3 This study

pST578 pKNT25 carrying patA This study

pST599 pKNT25 carrying hetF(C246A) This study

pJP41 pUT18 carrying divIVA (Bacillus subtilis) [12]

pST579 pUT18 carrying hetR This study

pST580 pUT18 carrying hetF This study

pST581 pUT18 carrying hetZ This study

pST582 pUT18 carrying patU5 This study

pST583 pUT18 carrying patU3 This study

pST585 pUT18 carrying patA This study

pST600 pUT18 carrying hetF(C246A) This study

36

Table 2.3. Oligonucleotides used in Chapter 2

Primer no.

Primer name Sequence (5’ to 3’)

1 patU3 up F ATAAGATCTACCACCAGAAACCAACGTCGATATCG

2 patU3 up R ATCGAGTTCGCCCGGGCTGTAAACGACGCTTGATTACGGC

3 patU3 down F TCGTTTACAGCCCGGGCGAACTCGATCAGCAACCATTGC

4 patU3 down R ATAGAGCTCAGGTGCGAATGCCAGAGTTTGGC

5 patU3 up R DHRB1156 TGACTGGGCGCCCGGGACGACGCTTGATTACGGCTTG

6 patU3 down F DHRB1156 CAAGCGTCGTCCCGGGCGCCCAGTCATCAAGCCCAGGATG

7 patU3 glu69stop up R CTTTTCCAATATTTATTGGAAAACTCTTTCTGGTAG

8 patU3 glu69stop down F GAGTTTTCCAATAAATATTGGAAAAGTGTCAGCAG

9 hetZ up F ATAAGATCTGCACTGATCAAGATAGCTGAATATGC

10 hetZ up R GAATTACGAACCCGGGGGTTGGAATAGTTGCTGTTG

11 hetZ down F TATTCCAACCCCCGGGTTCGTAATTCCCTAGTGTCC

12 hetZ down R ATAGAGCTCCATTACCTTGCAGTGTCAAGATTCC

13 hetZ +tsp up R GAGTCAAACCCCCGGGGCGCTGAGGGGGATTAATGTA

14 hetZ +tspII U5U3 down F 1 TATTCCAACCCCCGGGATGGACTATCTGGAGCAGAAAC

15 patU5 up F ATAAGATCTAGTCGAACTACACAGTACACTCAGC

16 patU5 L12stop up R CAACAAACCGAGTCAGTACTGTTGCAAAGATTCTG AGTCAC

17 patU5 L12stop down F GACTCAGAATCTTTGCAACAGTACTGACTCGGTTTGTTGGCATCCACTGTCAACAAACG

18 hetZ flank up F CACCAAGTTAGCAATTGAGC

19 hetZ flank down R CTCTACACTCAGACATCCTG

20 patU3 flank up F CCAGGAAGACAACAACAGTTG

21 patU3 flank down R GATAACTAAAGTGGGAATGG

22 patU5 U3-1F GATTTTAGAACAAGCTTGG

23 patU5 U3-1R GCTTCTTTCTACCTTCATC

24 PI hetZ up R SmaI TATATCCCGGGGGTTGGAATAGTTGCTGTTG

25 PI hetZ down F SmaI ATATACCCGGGTTCGTAATTCCCTAGTGTCC

26 PII hetZ down F BamHI ATATAGGATCCATGGACTATCTGGAGCAGAAA

27 patU5 down R ATAGAGCTCCTACCTTCATCTCTATTGTGCTAGCAG

28 alr0101 BamHI F TATATGGATCCGGGATTAGAGAAACATCCTG

29 asr0098 SacI R TATATGAGCTCTAGTGATACGATTTGCTACATC

30 hetZ SacI R TATATGAGCTCGTTGCAAAGATTCTGAGTCA

37

Table 2.3. (Continued) Oligonucleotides used in Chapter 2

Primer no.


31 patU5 SacI R TATATGAGCTCCAGTAGGAATTTCTCCTAATC

32 patU3 SacI R TATATGAGCTCGCTAGCAGTGGTAATTTCAG

33 hetZ up R SmaI TATATCCCGGGGCGCTGAGGGGGATTAATGTA

34 hetZ down F SmaI ATATACCCGGGATGGACTATCTGGAGCAGAAA

35 patU5 up R SmaI TATATCCCGGGCAACAAACCGAGTAAGTACTG

36 patU5 down F SmaI ATATACCCGGGGCTACTTCACAAAGAGGTCAG

37 patU3 up R SmaI TATATCCCGGGCTGTAAACGACGCTTGATTACGGC

38 patU3 down F SmaI ATATACCCGGGGAACTCGATCAGCAACCATTGC

39 patU3 OE-EcoRI-F TATATGAATTCATGTGCAAGAACGTTTTCAAGC

40 patU3 OE-BamHI-R TATATGGATCCGCTAGCAGTGGTAATTTCAG

41 patU3-EcoRI-rbs-F TATATGAATTCAGGAGGTGATTGTGCAAGAACGTTTTCAAGC

42 patU5 EcoRI F TATATAAGGAGGAATTCATATGAATAGTGACTCAGAATC

43 hetZ SmaI F ATATACCCGGGAGGAAACAGCTATGAATTCAGCCGCAACAGC

44 hetZ KpnI R TATATGGTACCGTTGCAAAGATTCTGAGTCA

45 patU5 SmaI F ATATACCCGGGAGGAAACAGCTATGAATAGTGACTCAGAATC

46 patU5 KpnI R TATATGGTACCCAGTAGGAATTTCTCCTAATC

47 patU3 SmaI F ATATACCCGGGAGGAAACAGCTGTGCAAGAACGTTTTCAAGC

48 patU3 KpnI R TATATGGTACCGCTAGCAGTGGTAATTTCAG

49 hetZ patU3 up R GTTCTTGCACCTATTCATGAGTGGATGCACTT

50 hetZ patU3 down F TCATGAATAGGTGCAAGAAC GTTTTCAAGCC

51 hetR BamHI F A AGGAGGGATCCGGGTTCCGCTGGCTCCGCTGCTGGTTCTGGCAGTAACGACATCGATCTGATCAAG

52 hetR EcoRI R A CTCCTGAATTCTTAATCTTCTTTTCTACCAAACACCATTTGTAAAATC

53 hetR BamHI F B AGGAGGGATCCGAGTAACGACATCGATCTGATC

54 hetR EcoRI R B CTCCTGAATTCCCGCCAGAACCAGCAGCGGAGCCAGCGGAACCATCTTCTTTTCTACCAAACACCATTTGTAAAATC

55 hetF BamHI F GTG1 A AGGAGGGATCCGGGTTCCGCTGGCTCCGCTGCTGGTTCTGGC TCCCAGGAAT TTCACATTTC TGTAAC

56 hetF EcoRI R A CTCCTGAATTCCTACTTGGGGCTTTTTTGTTGC

57 hetF BamHI F GTG1 B AGGAGGGATCCGTCCCAGGAA TTTCACATTT CTG

58 hetF EcoRI R B CTCCTGAATTCCCGCCAGAACCAGCAGCGGAGCCAGCGGAACCCTTGGGGCTTTTTTGTTGCAG

59 hetZ BamHI F A AGGAGGGATCCGGGTTCCGCTGGCTCCGCTGCTGGTTCTGGCAATTCAGCCGCAACAGCAACTATTC

38


Primer no.


60 hetZ MunI R A CTCCTCAATTGCTATTCATGAGTGGATGCACTTG

61 hetZ BamHI F B AGGAGGGATCCGAATTCAGCCGCAACAGCAAC

62 hetZ MunI R B CTCCTCAATTGCCGCCAGAACCAGCAGCGGAGCCAGCGGAACCTTCATGAGTGGATGCACTTGATC

63 patU5 BamHI F A AGGAGGGATCCGGGTTCCGCTGGCTCCGCTGCTGGTTCTGGCAATAGGACTCAGAATCTTTGCAACAG

64 patU5 EcoRI R A CTCCTGAATTCTCAAAAGGTTTGGGGCTGCCTTTTC

65 patU5 BamHI F B AGGAGGGATCCGAATAGTGACTCAGAATCTTTGCAAC

66 patU5 EcoRI R B CTCCTGAATTCCCGCCAGAACCAGCAGCGGAGCCAGCGGAACCAAAGGTTTGGGGCTGCCTTTTC

67 patU3 BamHI F A AGGAGGGATCCGGGTTCCGCTGGCTCCGCTGCTGGTTCTGGCCAAGAACGTTTTCAAGCCGTAATC

68 patU3 EcoRI R A CTCCTGAATTC CTATGTTGCGGGATTAATGAC

69 patU3 BamHI F B AGGAGGGATCCGCAAGAACGTTTTCAAGCCGTAATC

70 patU3 EcoRI R B CTCCTGAATTCCCGCCAGAACCAGCAGCGGAGCCAGCGGAACCTGTTGCGGGATTAATGACAAATAGC

71 patA BamHI F A AGGAGGGATCCGGGTTCCGCTGGCTCCGCTGCTGGTTCTGGCAAAACACTTCCGATTACTAGATACAG

72 patA EcoRI R A CTCCTGAATTCTTACGTAATGTGTTTAAAAATTACTTTTCAAATCACC

73 patA BamHI F B AGGAGGGATCCGAAAACACTTCCGATTACTAGATACAG

74 patA EcoRI R B CTCCTGAATTCCCGCCAGAACCAGCAGCGGAGCCAGCGGAACCCGTAATGTGTTTAAAAATTACTTTTAGCAAATCACC

75 alr0101 SmaI R ATATACCCGGGATGTTGCGGGATTAATGACAAATAGC

76 PpetE-BamHI-F ATATAGGATCCCTGAGGTACTGAGTACACAG

77 hetZ up R CTTTCTACAGCACCAGAAGC

78 hetZ-F TCATGATGTAGCAAATCG

79 hetZ-R GCGTTTGTTGACAGTGGATG

39

RESULTS

Mutations in patU3 restore heterocyst formation in hetF and hetF patA backgrounds

As one of the key regulators of heterocyst differentiation, hetF promotes heterocyst development.

In an effort to elucidate the mechanism of HetF in the formation of heterocysts, a genetic screen

was performed to isolate spontaneous mutants that bypass the hetF-dependent regulation of

heterocyst development. Mutants arising from this screen may represent components of the

regulatory network controlling differentiation in heterocystous cyanobacteria.

hetF and hetF patA deficient strains both are incapable of forming heterocysts (Fig. 2.2A), but the

phenotypes of these strains differ dramatically. Because the hetF strain exhibits aberrant cell

morphology and growth impairment, it was unclear whether this phenotype would complicate the

screen. HetF and PatA work in the same pathway responsible for regulation of HetR turnover[8].

The hetF patA strain does not exhibit the characteristic morphological changes observed in the

hetF deficient strain. For this reason, the hetF patA deficient strain was also used in this screen.

Colonies that arose on N- medium (lacking a source of combined nitrogen) represented

spontaneous bypass mutants of hetF and hetF patA that can fix nitrogen. Colonies from bypass

of each background strain were subsequently screened for the presence of heterocysts in N-

medium. Approximately half of these colonies could form heterocysts in N- media, in a pattern

similar to that of the wild-type, although the timing of differentiation appeared to be delayed for

some mutants (Fig. 2.2B).

Mutants that fixed

nitrogen in both the hetF

as well as the hetF patA

mutant backgrounds

were isolated in this

screen. Ten mutants

were sequenced at

candidate regions in the

chromosome in an effort

to determine the identity

of mutations. As

summarized in Table 2.4,

mutations were not

found in the hetR promoter nor hetR coding frame, asr0098 (the gene annotated immediately

upstream of hetZ), the hetZ promoter nor hetZ coding frame, nor within patU5. Five of the ten

mutants isolated were found to have deletions or nonsense mutations in the coding frame of

Figure 2.2. ∆hetF ∆patA strain (A) with an additional mutation in

patU3 (B) restores the ability of ∆hetF ∆patA to form heterocysts in

nitrate-deficient media. Carets indicate heterocysts.

40

patU3. Thus in addition to patU3, other unidentified genes are involved in the observed bypass

phenotype.

Because mutations

in patU3 were

discovered to

restore the inability

of these

background strains to form functional heterocysts (Fig. 2.2), a genetic association may exist

between hetF and patA and patU3. The relationship between HetF and PatU3 suggested by the

suppression results is depicted in Figure 2.3.

Table 2.4. Location of spontaneous mutations isolated within the 258-amino acid protein PatU3

found to restore the ability of hetF and hetF patA background strains to form heterocysts.

Background strain Name of isolate Consequence of mutation in PatU3

∆hetF F 2 ∆61-66 (VNLPER)

∆hetF F 5 W170stop

∆hetF ∆patA F 10 R113stop

∆hetF ∆patA FA 2 E69stop

∆hetF ∆patA FA 3 S144stop

patU3 negatively regulates heterocyst development and contributes to normal cell size

The gene responsible for restoring heterocyst formation in hetF-deficient strains, patU3, is located

in a gene cluster along with hetZ and patU5 (Fig. 2.1). DRHB1156, the original patU3 mutant,

was described as having a Mch phenotype[1]. In this mutant, the first 36 and last 136 nucleotides

Figure 2.3. A model for the HetF-dependent regulation of PatU3.

Inhibitors of the interaction represented with bars.

Figure 2.4. Locations of putative tsps (arrows) related to patU3 and sites of mutations

(delta, stars) isolated in the genetic screen described in this study. Lines refer to regions

deleted (dash) or mutated (star) from the Anabaena chromosome in this study.

41

of patU3 flank a kanamycin-resistance cassette (C.K4), deleting all but one putative

transcriptional start point (tsp) located in patU3. The original DHRB1156 mutant was recreated in

our hands using the Ω interposon conferring spectinomycin and streptomycin resistance (strain

patU3∆36-641; Fig. 2.4). In addition, the same region was also cleanly deleted from the

chromosome. Both mutant strains exhibited the same Mch phenotype as originally described

(Table 2.5).

To assess the phenotype of

a patU3 mutant strain that

deleted all four putative tsps,

nucleotides 42-729 of

patU3 were cleanly deleted

(and also replaced with the

Ω interposon) from the

chromosome of different

genotypes (Table 2.5). The

corresponding strains,

patU3∆42-729, encode the

first 14 and last 16 residues

of the 259-residue patU3

gene product. Unlike

patU3∆36-641, this version

deleted all of the internal

tsps defined within patU3 (Fig. 2.4) [4]. In hetF and patA hetF backgrounds, the formation of

heterocysts was restored upon deletion patU3∆36-641 and patU3∆42-729 (Table 2.5, Figure 2.5),

consistent with the screen results.

However, deletions in targeted genes containing internal tsps or antisense tsps may potentially

also disrupt the expression of adjacent genes, leading to misinterpretation of the phenotype of

mutant knock-outs. The mutations described previously did not preserve the four internal tsps

found in patU3 (Fig. 2.4) [4]. To evaluate the inactivation of patU3 function without inactivation of

the internal tsps, one of the five mutant forms of patU3 originally isolated in the screen for bypass

mutants (Table 2.4, Figure 2.2), patU3(E69stop), was recreated in the wild-type strain. The

patU3(E69stop) mutation was chosen because it was not expected to influence any of the tsps in

patU3 while also inactivating patU3. Strain patU3(E69stop) exhibited the same phenotype as

patU3∆42-729 and patU3∆42-729 (Table 2.5) suggesting that the Mch phenotype is due to

inactivation of patU3 alone.

Figure 2.5. Phenotype of patU3∆42-729 (UHM318) (B)

compared to the wild-type (A) after 24 h in nitrate-deficient

media. The diminutive cell size and Mch phenotype is

observed for all other ∆patU3 mutations constructed (Table

2.5). Carets indicate heterocysts.

42

Interestingly, deletion of patU3 results in a diminutive cell phenotype that was not previously

described, and suggests a role in cell division (Fig. 2.5). Smaller filaments have been reported to

occur upon overexpression of hetF from an inducible promoter[8], again suggesting a relationship

between hetF and patU3.

Table 2.5. Anabaena patU3 strains and strain description. For strains with deletions in multiple

genes, mutations are listed in the order the deletions were made. Mch, multiple contiguous

heterocysts.

Name of strain

Strain description Strain characteristics

UHM221 patU3∆42-729 with Ω cassette Mch, small cell phenotype

UHM223 ∆hetF patU3∆42-729 with Ω cassette Mch, small cell phenotype

UHM222 ∆patA ∆hetF patU3∆42-729 with Ω cassette Mch, small cell phenotype

UHM318 patU3∆42-729 Mch, small cell phenotype

UHM319 ∆hetF patU3∆42-729 Mch, small cell phenotype

UHM320 ∆patA ∆hetF patU3∆42-729 Mch, small cell phenotype

UHM321 ∆hetR ∆patA ∆hetF patU3∆42-729 No heterocysts, small cell phenotype

UHM323 patU3∆36-641 Mch, small cell phenotype

UHM324 patU3∆36-641 with Ω cassette Mch, small cell phenotype

UHM332 patU3(E69stop) Mch, small cell phenotype

UHM338 patU3∆42-729 ∆hetR No heterocysts, small cell phenotype

UHM339 patU3∆42-729 hetZ∆33-1155 No heterocysts, small cell phenotype

Inactivation of internal transcription start sites present in hetZ do not affect the hetZ

phenotype

More is known about hetZ than patU5 and patU3; however available descriptions of the hetZ

phenotype are inconsistent and may not reflect inactivation of hetZ alone. Four mutant strains

with hetZ inactivated by a transposon insertion or interrupted by an antibiotic cassette have been

described, but only the transposon-derived mutant named 1801 was incapable of forming

heterocysts[1]. The other three hetZ mutants formed between 3% and 9% heterocysts within 24

hours of nitrogen step-down[1]. This difference in phenotypes may be explained by the presence

of internal transcriptional start sites. Using a differential RNA-seq approach, a genome-wide map

of >10,000 putative internal tsps has been defined[4]. Five internal tsps were found within hetZ

(Fig. 2.6, black arrows). None of these internal tsps correspond to the originally described tsp II

(Fig. 2.6, green arrow)[1].

43

In this study, different mutants were constructed to reflect the inactivation of HetZ function alone,

rather than deletion of regions containing internal tsps or antisense tsps. Using RNA-seq data

and the position of tsp II, different regions (Table 2.6) of hetZ were deleted based on the revised

annotated hetZ sequence [1]. Construction of hetZ strains both with and without the region

corresponding to the six internal tsps allowed for analysis of the function of this region.

Table 2.6. Anabaena hetZ strains and strain description. WT, wild-type.

Name of strain

Strain description Strain characteristics

UHM309 hetZ∆33-1155 No heterocysts; WT size and morphology

UHM310 ∆hetF hetZ∆33-1155 No heterocysts, small cell phenotype

UHM311 ∆patA ∆hetF hetZ∆33-1155 No heterocysts; WT size and morphology

UHM312 ∆hetR ∆patA ∆hetF hetZ∆33-1155 No heterocysts; WT size and morphology

UHM313 hetZ∆354-762 No heterocysts; WT size and morphology

UHM314 ∆hetF hetZ∆354-762 No heterocysts; WT size and morphology

UHM315 ∆hetF hetZ∆354-762 with Ω cassette No heterocysts; WT size and morphology

UHM316 ∆patA ∆hetF hetZ∆354-762 with Ω cassette No heterocysts; WT size and morphology

UHM317 ∆hetR ∆patA ∆hetF hetZ∆354-762 with Ω cassette

No heterocysts; WT size and morphology

Figure 2.6. Locations of putative tsps (arrows) related to hetZ and site of the transposon

insertion (Tn, green inverted triangle) in mutant 1801. Tsp I and tsp II are represented

with green arrows. Additional internal tsps represented with black arrows. Horizontal

dashed lines refer to regions deleted from the Anabaena chromosome in this study.

44

Nucleotides +33 to +1155 relative to the ATG translational initiation codon of hetZ were cleanly

deleted from the wild-type strain. The resulting in-frame deletion mutant, strain hetZ∆33-1155,

contained the first 11 and last 17 residues of the hetZ gene product, eliminating 375 residues of

the 402 residue gene product. Mutant hetZ∆33-1155 was incapable of forming heterocysts (UHM

309, Fig. 2.7) like the previously described mutant 1801. Deletion of the sequence removed all

five putative tsps in addition to tsp II (Fig. 2.6). For this reason, it was unclear whether the

phenotype of the resulting mutant was due solely to the absence of hetZ or polar effects that lead

to the loss of function of either or both downstream genes in the gene cluster, patU5 and patU3.

Strain hetZ∆354-762 deleted

nucleotides +354 to +762 from

the chromosome. Only 408-bp

of hetZ was removed in this

mutant to preserve all putative

tsps in the coding region of hetZ.

In addition, 321-bp upstream of

transcript II was left intact to

allow for interaction with any

unknown regulatory factors.

The resulting mutant contained

the first 118 and last 148 residues of the gene product of hetZ, and eliminated 136 residues of

HetZ from the chromosome. As with mutant 1801, strain hetZ∆354-762 was also incapable of

forming heterocysts. Since both strains hetZ∆354-762 and hetZ∆33-1155 exhibited the same

heterocyst-deficient phenotype, deletion of the tsps did not appear to disrupt the downstream

genes patU5 and patU3. However removal of the tsps from a ∆hetF (but not ∆patA ∆hetF) parent

strain resulted in the small cell phenotype observed in patU3 mutant strains (Table 2.6). The

relationship between these tsps and HetF remains unclear. Taken together, the hetZ ORF alone

appears to be required for coordinating the formation of heterocysts.

Extra copies of patU5 and patU3 inhibit heterocyst differentiation

HetZ and PatU3 regulate heterocyst development in a positive and negative manner, respectively.

In an effort to elucidate the role of the hetZ-patU5-patU3 operon in the coordination of heterocyst

formation, contributions of genes within the gene cluster were tested for their ability to inhibit or

stimulate heterocyst development in the wild-type and ∆patA strains. Genes within the hetZ-

patU5-patU3 gene cluster were introduced in multicopy in different combinations on plasmids

driven by the native promoter PhetZ (Table 2.7; tsp I in Fig. 2.1). As a result of these plasmid-

based studies of the hetZ-patU5-patU3 gene cluster, it was discovered that plasmid pST355

Figure 2.7. Phenotype of ∆hetZ mutant after 24 h in nitrate-

deficient media. hetZ∆33-1155 (UHM309; refer to Table

2.6) no longer differentiates heterocysts. The same

heterocyst-deficient phenotype is observed in the

hetZ∆354-762 mutants (data not shown).

45

carrying the entire gene cluster (PhetZ-hetZ-patU5-patU3) abrogated heterocyst development in all

background strains tested (Table 2.7; additional strains tested not listed). Because HetZ and

PatU3 have contrasting roles in heterocyst differentiation, it was unclear whether the internal start

sites present in hetZ (and thus PatU5 and PatU3) rather than HetZ function influenced the

inhibition results. A truncated version of hetZ was constructed on plasmid pST429 (carrying PhetZ-

hetZ∆354-762-patU5-patU3; Table 2.7) to preserve the internal start sites present in hetZ. The

region truncated corresponded to the same region of hetZ deleted from the chromosome (Table

2.6). Like plasmid pST355, plasmid pST429 also inhibited heterocyst formation in the wild-type

and patA strains. This suggests that the tsps present in hetZ rather than HetZ function are

necessary for inhibition of differentiation by extra copies of patU5 and patU3 on a plasmid.

Table 2.7. Summary of plasmids used to test truncations of genes within the hetZ-patU5-patU3

gene cluster and effect on heterocyst development in wild-type and in ∆patA backgrounds.

Regions deleted (“knocked-out”) in hetZ and patU3 correspond to the same regions deleted in the

chromosome (Fig. 2.4 and 2.6). Truncations in hetZ (hetZ∆354-762) preserved the internal tsps

within hetZ. The native promoter PhetZ was used.

Plasmid description wild-type ∆patA

pST352 PhetZ-asr0098 Unchanged Unchanged

pST353 PhetZ-hetZ Unchanged Unchanged

pST354 PhetZ-hetZ-patU5 Unchanged Unchanged

pST355 PhetZ-hetZ-patU5-patU3 No heterocysts No heterocysts

pST429 PhetZ-hetZ∆354-762-patU5-patU3 (knock-out hetZ)

No heterocysts No heterocysts

pST430 PhetZ-hetZ-patU5∆45-126-patU3 (knock-out patU5)

No heterocysts (rare heterocyst at 48 h)

Occasional terminal heterocyst

pST431 PhetZ-hetZ-patU5- patU3∆42-729 (knock-out patU3)

Wild-type at 72 h Unchanged

pST590 PhetZ-hetZ∆354-762-patU5 (knock out hetZ; PhetZ-patU5)

Unchanged Unchanged

pST591 PhetZ-hetZ∆354-762-patU5-patU3∆42-729 (knock-out hetZ and patU3; PhetZ-patU5)

Unchanged Unchanged

pST592 PhetZ-hetZ∆354-762-patU5∆45-126-patU3 (knock-out hetZ and patU5; PhetZ-patU3)

Terminal heterocysts at 48 h, wild-type at 72 h

Unchanged

pST596 PhetZ-hetZ-patU5(L12stop)-patU3 No heterocysts No heterocysts

The filamentous nitrogen-fixing cyanobacterium Nostoc punctiforme PCC 73102 (herein, “Nostoc”)

is genotypically related to Anabaena. Genes related to heterocyst development are shared

between the two species. In Nostoc however, a single gene patU corresponds to the two genes in

Anabaena, patU5 and patU3 (alluding to the 5’ and 3’ fragments of patU) [21]. Additional studies

were done to further determine if the plasmid-based inhibition results was due to the function of

46

PatU5, PatU3, or both PatU5 and PatU3. To ascertain if both patU5 and patU3 are required to

negatively regulate heterocyst development, plasmids containing either patU5 (pST590, pST591)

or patU3 (pST592) were introduced into wild-type and patA strains (Table 2.7). Inhibition of

heterocyst differentiation by either group of plasmid(s) would suggest that only the corresponding

portion of PatU is required. Either gene alone was not sufficient to inhibit heterocyst formation in

either strain. Thus both patU5 and patU3 in multicopy (driven by the native promoter) are

required for full inhibition of heterocyst development. However, PatU3 by itself appears to lower

the number of heterocysts when present in multicopy (pST592). These results are consistent

with the Mch phenotype of patU3 mutant strains (Table 2.5).

Cis-acting transcriptional and/or translational elements are important for function of the

hetZ-patU5-patU3 gene cluster

Overexpression studies attempted to discern the role of genes within the hetZ-patU5-patU3 gene

cluster, individually and in different combinations. The native promoter (PhetZ) was replaced by

two different inducible promoters (PpetE and Pnir). Despite contrasting positive and negative

regulatory roles for hetZ and patU3 respectfully, overexpression of either gene resulted in the

same phenotype. Excess hetZ or patU3 from the stronger Pnir promoter (plasmids pST414 and

pST419 respectively; Table 2.8) increased heterocyst frequency but also reduced cell size in a

wild-type background. An excess of both hetZ and patU3 on the same plasmid (plasmid pST420)

had no effect on heterocyst development or cell size. Taken together, the location, timing, and/or

level of expression appear to be important as indicated by the varying phenotypes observed.

47

Table 2.8. Summary of plasmids used to overexpress individual genes or combinations of genes

within the hetZ-patU5-patU3 gene cluster and effect on heterocyst development and cell size in

wild-type and ∆patA backgrounds. The inducible promoters PpetE and Pnir were used.

“Unchanged” indicates that the parent phenotype was observed; “↑ heterocysts” indicates that

supernumerary heterocysts were observed; “↓ cell size” indicates that cells had a smaller cell size

compared to the wild-type.


pST216 PpetE-patU3 Unchanged Unchanged

pST255 Strong ribosomal binding site and PpetE-patU3

Unchanged Unchanged

pST315 PpetE-patU5-patU3 Unchanged Unchanged

pST414 Pnir-hetZ ↑ heterocysts, ↓ cell size

Unchanged

pST415 Pnir-hetZ-patU5 ↑ heterocysts, ↓ cell size

Unchanged

pST416 Pnir-hetZ-patU5-patU3 No heterocysts No heterocysts

pST417 Pnir-patU5-patU3 Unchanged Unchanged

pST418 Pnir-patU5 Unchanged Unchanged

pST419 Pnir-patU3 ↑ heterocysts, ↓ cell size

Unchanged

pST420 Pnir-hetZ-patU3 Unchanged Unchanged

In Caulobacter crescentus two promoters, P1 and P2 are developmentally regulated by the DNA

binding protein CtrA (Chapter 1; [22]). Similarly, the tsps associated with the hetZ-patU5-patU3

gene cluster may control transcription of this region. Two tsps, located upstream and within hetZ

(tsp I and tsp II, respectively; Fig. 2.8, green arrows), were originally identified [1] prior to

identification of internal tsps within the hetZ-patU5-patU3 gene cluster[4, 23]. Plasmids carrying

regions corresponding to all tsps was necessary to drive expression of patU5 and patU3 (pST355

and pST429; Table 2.8), but the role of tsp I and tsp II in heterocyst formation has not been

explored. To evaluate the contribution of tsp I and tsp II towards the function of PatU5 and PatU3,

tsp I and tsp II were introduced separately and coordinately in front of patU5-patU3 on plasmids

(pST586, pST589 and pST587 respectively; Fig. 2.9, Table 2.9). The effect on heterocyst

differentiation was assessed in wild-type and patA backgrounds (Table 2.9). Heterocyst

formation in the wild-type and patA backgrounds was fully inhibited with plasmids that left all tsps

in the gene cluster intact (plasmids pST355, pST429). Plasmids bearing tsp I, tsp II or both tsp I

and tsp II (pST586, pST589, pST587) suppressed heterocyst formation at 24 hours after nitrogen

deprivation in the wild-type. However this inhibitory effect was lost after 48 hours of nitrogen

starvation, as heterocysts were observed in a regular pattern after this timepoint.

48

Table 2.9. Summary of plasmids used to determine the role of tsp I and tsp II in patU5-patU3

expression and effect on heterocyst development in wild-type and ∆patA backgrounds.

Parentheses indicate the truncated region of hetZ that corresponds to tsp I and/or tsp II and

depicted in Figure 2.9 (in the same order from top to bottom of Figure).


pST586 tsp I-patU5-patU3 (hetZ∆33-1155) Wild-type at 48h

No heterocysts

pST587 tsp I-tsp II-patU5-patU3 (hetZ∆33-762) Wild-type at 48h

No heterocysts

pST589 tsp II-patU5-patU3 (hetZ∆1-762) Wild-type at 48h

No heterocysts

pST429 PhetZ-hetZ∆354-762-patU5-patU3 No heterocysts No heterocysts

pST355 PhetZ-hetZ-patU5-patU3 No heterocysts No heterocysts

Overexpression of hetZ cannot functionally bypass deletion of either hetR or hetF

The developmental genes hetZ, hetR, and hetF positively regulate heterocyst formation. In

strains deficient in these genes, heterocyst formation is abrogated. Accordingly, overexpression

Figure 2.8. Regions present on plasmids used to determine the role of tsp I and tsp II in

patU5-patU3 expression. Putative tsps (arrows; tsp I and tsp II in green) carried on

plasmids described in Table 2.9. The hetZ ORF (1206 nucleotides) is denoted by green

text. Horizontal dashed lines refer to truncated regions upstream or within hetZ not

present on plasmids; these regions also correspond to regions deleted from the

Anabaena chromosome in this study (Table 2.6). Horizontal filled lines refer to the region

of PhetZ -hetZ-patU5-patU3 carried on plasmids (Table 2.9).

49

of these genes from inducible promoters induced increased heterocyst frequencies in the wild-

type. However, overexpression of hetR does not functionally bypass deletion of hetF [8] (nor

hetZ, data not shown) since HetF is required for transcription from the -271 (autoregulatory) start

point of the hetR promoter and proper modulation of HetR protein levels[8]. Additionally,

overexpression of hetF does not functionally bypass deletion of hetR [8] (nor hetZ, data not

shown).

To determine if overexpression of hetZ can functionally bypass deletion of either hetR or hetF, a

plasmid carrying Pnir-hetZ (plasmid pST414; Table 2.8) was introduced into both a hetR-deficient

strain and a hetF-deficient strain. Heterocyst formation was not restored to either background

strain carrying plasmid pST414 (data not shown). Thus in the absence of the normal

transcriptional regulation of hetZ (through use of the Pnir promoter in lieu of the native promoter of

hetZ), hetZ cannot functionally bypass the deletion of hetR or hetF. In addition, heterocyst

formation was not restored upon introduction of a plasmid carrying PhetZ-hetZ (plasmid pST353)

into the same strains (data not shown).

The enlarged cell morphology that results upon the inactivation of hetF is correlated with elevated

levels of HetR[8]. The aberrant phenotype is resolved upon exogenous addition of the

pentapeptide RGSGR (a motif present in PatS and HetN, negative regulators that work at the

level of HetR) and in the double mutants ∆hetR ∆hetF and ∆patA ∆hetF[8]. The hetZ gene

appears to also contribute to this phenotype because the ∆hetF ∆hetZ (as well as ∆patA ∆hetF

∆hetZ and ∆hetR ∆patA ∆hetF ∆hetZ; Table 2.6) strains also do not exhibit the aberrant

morphology associated with the inactivation of hetF alone.

Regulation of heterocyst differentiation by patU3 does not require patS or hetN

The multiple contiguous heterocyst (Mch) phenotype observed upon inactivation of patU3 has

been observed in strains deficient in two genes containing the RGSGR motif, patS and hetN.

Strains with deletions in either patS or hetN also result in a similar Mch phenotype, although the

phenotype is delayed in the hetN-deficient strain[20]. Due to the similar inhibitory effect on

heterocyst development by patU3, patS and hetN, it is unclear whether these three regulatory

genes are associated with the same regulatory circuit. To determine if a patU3 background strain

is sensitive to either inhibitor of differentiation, plasmids overexpressing patS or hetN (carried on

plasmids pDR211 and pDR320, respectively), were introduced into a patU3-deficient strain, and

the effect on heterocyst formation was assessed. Heterocyst formation was abrogated in the

patU3-deficient strain overexpressing either plasmid (data not shown). In the converse

experiment, a plasmid carrying PhetZ-hetZ-patU5-patU3 (plasmid pST355) was introduced into

strains deficient in either patS or hetN. Heterocyst formation was abrogated in both strains with

extra copies of the hetZ-patU5-patU3 gene cluster (data not shown). Thus, patS- and hetN-

50

dependent inhibition of heterocyst formation does not require patU3 for activity. In addition,

patU3-dependent inhibition of heterocyst formation does not require patS or hetN for activity.

Therefore PatU3 and PatS/HetN appear to regulate heterocyst differentiation independent of one

another.

hetZ and patU3 are required for the timing of pattern formation

Because deletions in hetZ and patU3 affect heterocyst frequencies, both genes would appear to

have a role in the patterned formation of heterocysts. Transcription of the developmental genes

hetR and patS are up-regulated in cells in a periodic pattern prior to morphological changes to

differentiating cells. Thus fluorescence emanating from fusions of the native promoters of hetR or

patS to gfp allows visualization of patterning approximately 3-8 hours after removal of combined

nitrogen even in strains that do not form heterocyts[17]. To investigate the role of hetZ and patU3

in pattern formation, plasmids carrying PhetR-gfp and PpatS-gfp (pSMC127 and pAM1951,

respectively) were individually introduced into the wild type (Fig. 2.9A, 2.9D), and into strains

deficient in either gene.

Transcription from the hetR promoter was below the level of detection in both the hetZ-deletion

strain and the patU3-deletion strain, suggesting that hetR expression is dependent on these

genes (Fig 2.9B, 2.9C, respectively). Transcription from the patS promoter was delayed in the

hetZ strain. Fluorescence was only weakly detected after 24 hours after nitrogen removal, but

the spacing of GFP appeared patterned (Fig. 2.9E). Similarly, transcription from the patS

promoter was also delayed in the patU3 strain, but the spacing of fluorescence appeared irregular.

Unlike transcription of patS in the wild-type (Fig. 2.9D), fluorescence corresponding to PpatS-gfp

did not eventually resolve only to differentiated cells. At 48 hours after nitrogen removal, PpatS-gfp

was observed in both vegetative cells and heterocysts (Fig. 2.9F). These results suggest that

both genes are involved in proper temporal and spatial control of pattern formation.

51

Figure 2.9. hetZ and patU3 are required for pattern formation. Visualization of a

transcriptional PhetR-gfp fusion carried on plasmid pSMC127 (A-C) and a transcriptional

PpatS-gfp fusion (D-F) carried on plasmid pAM1951 in the wild-type (A, D), hetZ (B, E),

and patU3 (C, F) genotypes. Micrographs show bright-field (left panels) and GFP

fluorescence images (right panels) images captured using identical microscope and

camera settings 24 h (A-C) and 48 h (D-F) after removal of combined nitrogen. Carets

indicate heterocysts.

52

hetF patU3 and hetF patA patU3 genotypes have diminished HetR-GFP fluorescence in

comparison to hetF and hetF patA genotypes

Inactivation of HetF

and PatA results in a

decrease in

heterocyst frequency,

despite the

concomitant

accumulation of

HetR protein [8].

Mutations in patU3

were found to restore

heterocyst

development in

∆hetF and ∆hetF

∆patA strains. In

order to observe the

fate of HetR protein

within these strains,

HetR-GFP

translational fusions

present on plasmids

pDR292 (bearing PhetR-hetR-gfp) and pDR293 (bearing PpetE-hetR-gfp) were used. Fluorescence

corresponding to HetR-GFP from both plasmids was markedly elevated in strains lacking HetF

and/or PatA (Fig. 2.10B), consistent with elevated levels of HetR in these parent strains.

In the wild-type background, extra heterocysts are observed due to multiple copies of hetR, but

fluorescence is below the level of detection (Fig. 2.10A). Fluorescence was also below the level

of detection for hetF patA patU3 (Fig. 2.10C) backgrounds in marked contrast to the robust

fluorescence seen with the hetF patA parent strain (Fig. 2.10B). Thus, inactivation of patU3

returned the HetR-GFP fluorescence intensity in hetF patA to levels comparable wild-type levels.

This suggests that PatU3 is required for increased levels of HetR in the absence of either or both

hetF and patA.

Figure 2.10. HetR-GFP is not detected in hetF patA patU3 genotypes.

Visualization of a translational PpetE-hetR-gfp fusion on plasmid

pDR293 in the wild-type (A), hetF patA (B), and hetF patA patU3 (C)

genotypes. Micrographs show bright-field (left panels) and

fluorescence images (right panels) images captured using identical

microscope and camera settings 24 h after removal of combined

nitrogen. Carets indicate heterocysts.

53

hetF patU3 and hetF patA patU3 genotypes exhibit reduced hetR expression compared to

the wild-type

The reduction of HetR-GFP in patU3-mutant strains may be explained by either a decrease in

transcription or increased protein turnover. The simplest explanation for decreased HetR-GFP is

decreased expression of hetR. Alternatively, the HetR protein may be rapidly degraded (for

example, perhaps in a hetF-dependent manner), thus resulting in the lack of observable HetR-

GFP fluorescence.

In the wild-type, a

pattern of hetR

expression can be

detected prior to

morphological

modifications to the cell.

Transcription of hetR,

initially up-regulated in

clusters of cells,

resolves to single cells

in a pattern reflecting

the eventual periodic

interval of heterocysts

(Fig 2.11A). To further

understand the

relationship between

hetR and patU3,

transcription of hetR

was observed through

use of plasmid pSMC127 bearing PhetR-gfp.

In the hetF and hetF patA deletion strains, transcription from the hetR promoter is weak (Fig.

2.11B), even though HetR protein is abundant in these strains (Fig. 2.10B). The level of

fluorescence is similar to that detected in vegetative cells of the wild-type (Fig. 2.11A). In mutant

strains lacking both hetF and patU3, there is a loss of HetR-GFP fluorescence altogether (Fig.

2.10C) and transcription is below the level of detection afforded by the GFP transcriptional fusion

(Fig. 2.11C). The question of how strains with hetF patU3 and hetF patA patU3 genotypes form

heterocysts seemingly in the absence of hetR transcription is enigmatic.

Figure 2.11. Expression of hetR is not detected in the hetF patA

patU3 genotype. Visualization of a transcriptional PhetR-gfp fusion on

plasmid pSMC127 in the wild-type (A), hetF patA (B), and hetF patA

patU3 (C) genotypes. Micrographs show bright-field (left panels) and

fluorescence images (right panels) images captured using identical

microscope and camera settings 24 h after removal of combined

nitrogen. Carets indicate heterocysts.

54

Genetic epistasis analysis suggests hetR acts upstream of patU3 in the regulation of

heterocyst differentiation

Despite the absence of PhetR-gfp and PpetE-hetR-gfp in hetF patA patU3 parent strains,

heterocysts form at a frequency far greater than observed in the wild-type (Fig. 2.10, 2.11). This

may suggest that the patU3 mutation in a hetF patA background bypasses the need for hetR in

heterocyst development. To determine if hetR was no longer required for the formation of

heterocysts in a strain without the genes hetF, patA or patU3, the quadruple mutant UHM321 was

constructed by creating an in-frame deletion of the patU3 gene (patU3∆42-729) from the ∆hetR

∆hetF ∆patA parent strain. The resultant ∆hetR ∆hetF ∆patA ∆patU3 (UHM321) strain did not

exhibit heterocyst formation (data not shown), indicating that mutations in patU3 cannot bypass

the requirement for hetR in a hetF patA background. The small cell phenotype of patU3 strains

was also observed (Table 2.5). Because some level of HetR appears to be required for

heterocyst to form in this quadruple mutant strain, deletion of hetR is epistatic to that of patU3.

To understand the relationship of hetR in the patU3-dependent regulation of heterocyst

development and allow placement of patU3 into the working model of genetic interactions

involved in heterocyst differentiation (Chapter 1; Fig. 1.3), the double mutant patU3∆42-729

∆hetR was constructed. Epistasis of mutations in PatU3 and HetR was used to discriminate

between the two models proposed in Figure 2.12. The absence of heterocysts in this strain

would support the model depicted in Figure 2.12A, whereas the presence of heterocysts would

support the model shown in Figure 2.12B.

Inactivation of patU3 did not restore heterocysts to a hetR mutant (Fig. 2.13E). The Mch

phenotype observed in patU3 background strains (Fig. 2.13D) was abrogated upon inactivation of

hetR in the double mutant. But the diminutive cell size characteristic of patU3 strains remained

(Fig. 2.13D; Table 2.5). Consistent with the heterocyst-deficient phenotype of the quadruple

mutant strain ∆hetR ∆hetF ∆patA ∆patU3, the model presented in Figure 2.12A is supported. The

patU3 gene acts upstream of hetR to inhibit heterocyst formation (Fig. 2.17).

Figure 2.12. Two proposed models of genetic interaction between hetR and patU3 in the

control of heterocyst development. Inhibitors of the interaction represented with bars,

activators represented with arrows.

55

Figure 2.13. Heterocyst formation is abrogated in ∆patU3 ∆hetR and ∆patU3 ∆hetZ

double mutant strains. The double mutant strains (E) patU3∆42-729∆hetR and (F)

patU3∆42-729hetZ∆33-1155 no longer exhibit the Mch phenotype of (D) strain

patU3∆42-729. Strains with patU3 inactivated (D-F) exhibit a small cell phenotype

compared to the (A) wild-type, (B) ∆hetR and (C) hetZ∆33-1155. Micrographs captured

after 48 h in nitrate-deficient media using identical microscope and camera settings.

Carets indicate heterocysts.

56

Genetic epistasis analysis suggests patU3 acts upstream of hetZ in the control of

heterocyst differentiation

The developmental genes hetZ and patU3 coordinate heterocyst development. However, the

interaction between them is not understood. To determine which mutation, ∆hetZ or ∆patU3 is

epistatic, double mutants of hetZ and patU3 were constructed. The resultant strain, patU3∆42-

729 hetZ∆33-1155, allowed for discrimination between the two models of genetic interactions

between hetZ and patU3 depicted in Figure 2.14. The presence of heterocysts in this strain would

support the model in Figure 2.14A, whereas the absence of heterocysts would support the model

in Figure 2.14B.

The patU3∆42-729 hetZ∆33-1155 double mutant was incapable of forming heterocysts (Fig.

2.13F), similar to the phenotype upon inactivation of hetZ (Fig. 2.13C), and exhibited the small

cell phenotype observed upon deletions in patU3 (Fig. 2.13D). This result supports the model

depicted in Figure 2.14B. Thus patU3 acts upstream of hetZ in the working model of genetic

interactions involved in heterocyst differentiation (Chapter 1, Fig. 1.3; Fig. 2.17).

Direct interaction detected between HetZ and PatU3 by the bacterial two-hybrid system

In this study a genetic relationship between hetF, patA, hetR and the hetZ-patU5-patU3 gene

cluster was discovered, but direct interactions between their protein products are unknown. To

investigate direct protein-protein interactions between HetZ, PatU5, PatU3, HetR, HetF,

HetF(C246A) and PatA, a bacterial two-hybrid system was employed. Previous bacterial two-

hybrid assays have reportedly failed to demonstrate the interaction of PatU3 with either PatA or

HetR, however a different two-hybrid reporter system than that proposed here was used[1]. In

the genetic assay performed in this study[10], fusion of proteins to two complementary fragments

of the Bordetella pertussis adenylate cyclase catalytic domain, T25 and T18, allowed for putative

protein-protein interactions to be tested (Table 2.2). Physical association between the two

proteins fused to T25 and T18 results in functional complementation of adenylate cyclase in an

Figure 2.14. Two proposed models of genetic interaction between hetZ and

patU3 in the control of heterocyst development. Inhibitors of the interaction

represented with bars, activators represented with arrows.

57

Escherichia coli cya strain. Subsequent synthesis of cAMP by the reconstituted adenylate

cyclase yields a distinctive blue-colored colony phenotype that signals the presence of a direct

interaction between the two proteins under investigation. The Bacillus subtilus cell division

protein DivIVA, shown in different reports to interact with itself[11, 12], was utilized as a positive

control for this study. Additionally, HetR was expected to interact with itself due to the reported

homodimer activity of HetR[24]. Consistent with these reports, the dimerization of HetR was

demonstrated with all four pairings of HetR to the N- and C-termini of T25 and T18 in this study

(Fig. 2.15, 2.16). The strength of the association between monomers of HetR was equivalent to

the robust interaction between DivIVA-DivIVA (Fig. 2.16).

A direct interaction between HetF and PatU3 was not identified in this assay. HetF, however, is

closely related to the CHF class of cysteine proteases[8], a family that includes caspases.

Proteolysis by initiator caspases leads to activation of effector caspases, ultimately leading to

apoptotic cell death. A C246A substitution in HetF fails to complement the heterocyst deficiency

of a hetF-deficient strain[8]. This inactive form of HetF, like the native (non-mutated) form of HetF,

dimerized with itself as well as native HetF. HetF(C246A) also interacted weakly with PatA.

These interactions are consistent with the dimer state of caspases, and the HetF-PatA regulatory

pathway, respectively. Although HetF was not shown to interact with PatU3, HetF(C246A)

interacted weakly with PatU3, as this interaction was barely above detection (Fig. 2.15, 2.16). No

interactions were observed with PatU5 in combination with all other proteins tested. It remains

unclear whether these results represent a limitation of the assay.

However, a robust interaction was observed between HetZ and PatU3 that was more pronounced

than the homodimer interactions between DivIVA-DivIVA and HetR-HetR (Fig 2.15, 2.16).

Positive interactions between HetZ and PatU3 are shown (Fig. 2.16), representing six of the eight

possible pairings between N- and C-terminal fusions to T25 and T18. While the molecular

mechanism remains unknown, a physical interaction between HetZ and PatU3 is supported, and

may represent a critical control point in the coordination of heterocyst development.

Of the 196 interactions tested between the cell division proteins HetZ, PatU5, PatU3, HetR, HetF,

HetF(C246A) and PatA in this study, only 22 (11.2%) demonstrated a significant level of

interaction (Fig. 2.16, Table 2.10). These 22 interactions generally exhibited three different

levels of interaction: weak, intermediate, and strong (Fig. 2.16; also assigned W, I, S in the left-

most column of Table 2.10), although some overlap is observed between a few interactions. The

weak interaction category represents results that appear statistically significant, but require

additional analysis to support the interactions listed. The PatU3-HetF(C246A) and PatA-

HetF(C246A) interactions in this grouping are predicted in the model shown in Figure 2.17.

Protein interactions with an intermediate level of interaction include interactions between all

58

combinations of HetF and HetF(C246A). Finally, strong interactions comparable or exceeding the

DivIVA-DivIVA interaction include the physical interaction between HetR and itself and HetZ-

PatU3. Evidence from biochemical or biophysical studies, including immunoprecipitation assays,

would further support the HetZ-PatU3 interaction.

59

Figure 2.15. Bacterial two-hybrid results indicate HetZ interacts with PatU3. Additional

interactions (HetR/HetR, HetF/HetF, HetF(C246A)/HetF, PatA/PatA, HetF(C246A)/PatA,

HetF/HetF(C246A), PatU3/HetF(C246A), PatA/HetF(C246A), and

HetF(C246A)/HetF(C246A)) also depicted. No interactions were observed with PatU5.

Colonies of E. coli cells expressing the indicated protein fusions to fragments of adenylate

cyclase carried on derivatives of plasmids pKT25 and pUT18C. Plus signs (lower right

corner) indicate a positive interaction indicated by the blue-colony phenotype.

60

Figure 2.16. Average β-galactosidase activity of positive protein-protein interactions between

HetR, HetF, HetF(C246A), HetZ, PatU3, PatA. Negative controls for the assay (empty vector

interactions) shown on the far left (first four bars in blue). Dashed vertical lines separate the

bar graph into three different levels of interaction as described in the Results section. A

“weak” level of interaction is represented in green (PatU3-HetF(C246A) and PatA-

HetF(C246A). An “intermediate” level of interaction is represented in purple (combinations of

HetF and HetF(C246A) interactions). Some HetR/HetR and HetZ/PatU3 interactions (far right;

red bars) have greater activity than the positive control for the assay (Bacillus subtilis

DivIVA/DivIVA interaction; blue bar towards right of graph) and are interpreted as “strong”.

Average activity of three replicates plotted in ascending order from left to right. Plasmid

identity for tested interactions detailed in Table 2.10 in the same order (interactions from left to

right of Figure corresponds to interactions listed from top to bottom of Table).

61

Table 2.10. Plasmids used for quantification of β-galactosidase activity (Figure 2.16) of positive

bacterial two-hybrid interactions, and average Miller units (with standard deviation) of three

independent replicates. Rows (top to bottom) correspond to ascending order (left to right) of

activity depicted in Figure 2.16. The first four bars (left side of bar graph) relating to empty vector

negative controls in Figure 2.16 are listed at the top of the Table. The positive control for the

assay (Bacillus subtilis DivIVA/DivIVA interaction) is identified in bold font. “W”, “I”, “S” in the left-

most column refers to an assigned level of interaction (“weak”, “intermediate”, and “strong”,

respectively) as described in the Results section. “NA” indicates “not applicable”.

Level Protein-protein interaction (T25 fragment/T18 fragment)

Plasmid (T25 fragment)


Average Miller Units (MU)

Standard deviation

NA None (negative control) pKNT25 pUT18C 56 12.6

NA None (negative control) pKNT25 pUT18 57 4.1

NA None (negative control) pKT25 pUT18 61 9.2

NA None (negative control) pKT25 pUT18C 61 3.8

W PatA/PatA pST564 pST571 81 7.4

W PatU3/HetF(C246A) pST562 pST598 91 7.0

W HetR/PatA pST558 pST571 94 22.4

W HetR/PatA pST578 pST579 125 14.5

W HetR/HetF pST572 pST580 128 21.2

W HetZ/PatU3 pST574 pST569 133 34.9

I PatA/HetF(C246A) pST564 pST598 139 28.4

I HetF/HetF pST573 pST580 145 12.9

I HetF(C246A)/PatA pST597 pST571 282 49.4

I HetF(C246A)/HetF pST597 pST566 405 145.5

I HetF/HetF pST559 pST566 408 47.6

I HetF(C246A)/HetF(C246A) pST597 pST598 461 123.6

I HetF/HetF(C246A) pST559 pST598 499 93.0

S HetR/HetR pST572 pST579 589 19.5

S PatU3/HetZ pST562 pST581 636 232.5

S HetR/HetR pST572 pSST565 1145 342.1



S DivIVA/DivIVA (positive control)

pJP42 pJP41 1579 109.6



S HetZ/PatU3 pST560 pST583 2301 69.4

S HetZ/PatU3 pST560 pST569 2347 55.6

DISCUSSION

Organisms orchestrate the differentiation of a subset of cells in often intricate patterns despite the

fundamental sameness of the genetic complement of all cells. How one cell type diverges into

different cell types during development remains a central question in biology. There is extensive

evidence for the roles of temporal and spatial variations in gene expression and environmental

62

cues that result in the morphological heterogeneity observed within a single multicellular

organism. The molecular mechanisms behind development involve trans-acting factors

(transcription factors, signal transduction systems, morphogens, etc.), cis-acting factors (non-

coding sequences; promoters, RNA elements), and stochasticity[25, 26]. Additionally, molecular

memory is imposed on the cellular infrastructure allowing for another layer of information

transmission across the generations. Together, these genetic programs facilitate organism

survival and have directed evolution over time.

Differentiation of heterocyst cells in Anabaena offers a unique opportunity to explore the

mechanisms of multicellularity and the organization of two cell types into a nonrandom, one-

dimensional arrangement. Several key players of heterocyst development have been identified.

Among them the positive regulator hetF is essential for differentiation, but aside from a

relationship with two positive regulators (hetR, patA)[8], a regulatory network encompassing hetF

has not been clearly defined. To investigate a HetF-dependent regulatory network, a genetic

screen utilizing hetF-deficient backgrounds was performed in this study. Mutations in patU3

repeatedly arose from these screens, suggesting that PatU3 and the gene cluster containing

patU3 represent components of a regulatory network involving hetF.

Mutations in HetF and PatU3 affect cell morphology. The enlarged cell morphology of hetF-

deficient strains is related to abundant HetR levels[8]. The diminutive cell phenotype of patU3-

deficient strains suggests another relationship to hetF. Due to the perturbations in morphology

associated with hetF and patU3, these genes may have additional roles in cell division. Coupling

of cell growth and cell division to the later stages of heterocyst development efficiently

coordinates biological processes in response to changes in the concentration of fixed nitrogen. In

Anabaena, cell division is required for heterocyst differentiation[27] and linked to cell cycle events,

including DNA synthesis[28]. However, the heterocyst represents an end point in development.

Cell division, while required for differentiation, ceases before the morphological changes are

complete. The key cell division gene FtsZ and its inhibitor MinC regulate cell size and

morphology in Anabaena[29]. Analysis of GFP translational fusions to these and other cell

division genes in patU3 and hetF backgrounds or bacterial two-hybrid assays between PatU3,

HetF and cell division proteins may elucidate a role for the hetF-patU3 regulatory circuit in cell

division.

63

The preliminary genetic network depicted in Figure 2.17 describes our current understanding of

the regulation of heterocyst development as described in this study. Although the PatA-HetF

interaction is inferred through previous studies[8] and only weakly shown in this study (Fig. 2.15,

2.16), a direct interaction between HetR and PatS has been demonstrated [30]. The additional

interactions shown are consistent with the (i) hetF and patU3, (ii) patU3 and hetZ, and (iii) hetR

and patU3 double mutant phenotypes. They represent the simplest explanation for the double

mutant phenotypes observed. Bacterial two-hybrid assays were used to further corroborate these

interactions as well as the other interactions shown in Fig. 2.17. A strong interaction was not

observed between (i) PatA and HetF, (ii) HetF and PatU3, nor (iii) HetZ and HetR, perhaps due to

additional requirements for the interaction not addressed owing to the limitations of the assay. It

is unclear whether the weak level of interaction between these proteins is biologically significant

in the control of heterocyst differentiation. However, a robust interaction was demonstrated

between PatU3 and HetZ, consistent with the model depicted. Currently the mechanism for the

PatU3-HetZ interaction remains unknown. Further characterization is necessary to identify the

Figure 2.17. A preliminary model for genetic regulation of heterocyst differentiation in

Anabaena involving the conversion of HetR to HetR* by regulatory factors. Inhibition of

the pathway represented with bars, activation represented with arrows. “T” indicates

transcriptional regulation; “P” represents post-transcriptional regulation; “P?” represents

putative post-transcriptional regulation.

64

specific residues or domains involved in the protein-protein interaction and to understand the

consequences of the interaction.

The relationship between hetF, patA and patU3 was extended to include the transcriptional

regulator hetR. The dramatic absence of HetR-GFP fluorescence and lack of hetR expression in

hetF patU3 and hetF patA patU3 genotypes is enigmatic due to the restored ability to form

heterocysts in these mutant strains. Despite the inability to detect both hetR expression and

translated product in these background strains, hetR appears to be required for heterocyst

formation: the quadruple mutant ∆hetR ∆hetF ∆patA patU3∆42-729 and the double mutant

patU3∆42-729 ∆hetR were both incapable of forming heterocysts. Based on these and other

observations, placement of HetR into the HetF-regulatory circuit is postulated (Fig. 2.17). In the

model as shown, HetR exists in two forms: HetR represents a stable inactive form and HetR*

represents the unstable but active form that is necessary for heterocyst formation. Conversion to

the active (HetR*) form involves the central components of the HetF-dependent pathway, HetZ

and PatU3. HetF (positively regulated by PatA) controls the level of PatU3. PatU3 negatively

regulates levels of (the activator) HetZ by direct physical interaction. HetZ is proposed to act late

in the pattern formation stage of development. Positive regulation of the activity of HetR* by HetZ

is necessary to commit the cell to the heterocyst cell fate during the later stages of pattern

formation. Multiple, integrated feedback loops exist to fine-tune the system and allow for many

potential points of regulation by additional unknown factors. Some putative factors may include

phosphorus, carbon, and cellular energy levels. Ultimately HetZ levels increase enough to

overcome the negative effects of PatU3 to allow conversion of HetR to HetR* and drive

heterocyst formation.

Localization of PatU3 currently remains elusive. Florescence from gfp translational fusions to

patU3 driven by the native promoter or an inducible promoter (plasmids pST291 and pST293,

respectively) was below the level of detection in the wild-type (data not shown). However, the

fusion was functional because a plasmid containing PhetZ-hetZ-patU5-patU3-gfp (like the plasmid

bearing PhetZ-hetZ-patU5-patU3) inhibited heterocysts in the wild-type. In addition, PatU3 (and

HetZ) appears to affect the patterned expression of hetR, consistent with the heterocyst

frequencies observed upon inactivation of patU3 (and hetZ). Additionally, plasmids pST291 and

pST293 (along with other combinations of the hetZ-patU5-patU3 gene cluster in multicopy or in

excess; plasmids pST255, pST315, pST355, pST429, pST430 and pST431) did not alter levels of

HetR-GFP when introduced into PhetR-hetR-gfp and PpetE-hetR-gfp strains (data not shown).

Levels of PatU3-GFP fluorescence are presumably below the levels of detection of our assay.

One universal mechanism used to control development involves asymmetry in concentrations of

positively acting factors (Chapter 1). The lateral inhibition of two RGSGR-based inhibitory

65

proteins (PatS and HetN) operates to guide a gradient of HetR formation in a manner consistent

with the activator-inhibitor model [18, 31]. The resultant asymmetry in the protein concentration

of HetR drives commitment to a particular cell fate. The model explaining the HetF-dependent

role in heterocyst development includes the HetR-PatS-dependent pathway (Fig. 2.17). It is

unclear whether the HetZ-PatU3-dependent pathway also conforms to the activator-inhibitor

model. The activator HetZ appears to autoregulate its own expression [1] but it is unknown if

HetZ leads to the production of the inhibitor PatU3. Rather, PatU3 has been shown to upregulate

the expression of hetZ[1]. Also, a gradient of PhetZ-gfp expression correlates to HetR

concentrations, and can be abolished by exogenous addition of the RGSGR pentapeptide [1], but

patU3 has not been examined under similar conditions. Furthermore, diffusion (nor localization)

of the inhibitor PatU3 has not been demonstrated. While both hetZ and patU3 are expressed in

heterocysts, expression of patU3 also occurs at a lower level in vegetative cells[3]. It remains

unclear whether this differential expression pattern affects the regulatory circuitry.

In this study, an additional regulatory pathway controlling the heterocyst cell fate decision was

characterized. The HetR-RGSGR and HetZ-PatU3 pathways can function in the absence of the

other, suggesting that these pathways act independently. Two independent pathways may

indicate that they act at different times during differentiation or they may allow for greater control

of the developmental process by overcoming background noise in the other system. The HetZ-

PatU3 pathway functions as another control point in the Anabaena cell fate decision. In this

context, the HetZ-PatU3 pathway may relate to the concept of a biological checkpoint[32]. In

eukaryotes, the cell cycle refers to an ordered four-stage process (G1, S, G2, M) that ultimately

generates two daughter cells upon cell division. A series of regulatory mechanisms known as

checkpoints ensures cell cycle progression in a unidirectional manner. One example is the intra-

S checkpoint responsible for inhibition of entry into mitosis (M-phase) until DNA replication is

completed. In Xenopus laevis, the kinase activity of ATR is activated upon association with

replication forks. Initiation of a signaling cascade upon ATR binding to the cell replication

machinery results in the inactivation of the Cdc25 phosphatase necessary for activating M-phase

events. Because events from earlier stages (DNA replication; S-phase) negatively regulate

events of a later stage (mitosis), mutations involved in checkpoints are identified by a “relief of

inhibition” [32]. Inactivation of PatU3 in the chromosome presumably relieved the PatU3-

dependent negative inhibition of the downstream activator HetZ, resulting in inappropriate levels

of differentiation (assessed as the Mch phenotype in patU3-deficient strains). These results fulfill

the minimal requirement (“relief of inhibition”) for a checkpoint regulatory mechanism. Additional

studies are necessary to characterize the HetZ-PatU3 regulatory mechanism as a checkpoint of

Anabaena development.

66

In the model as shown, commitment to the heterocyst cell fate is not executed until PatU3 and

HetZ detect and report on HetR-RGSGR concentrations. Integrated feedback mechanisms

between the two pathways presumably exist to ensure that inappropriate differentiation does not

occur. The HetZ-PatU3 pathway may act at the pattern formation stage of development,

controlling the transition to the commitment stage. In addition to controlling development, the

dual mechanisms may impose a biological timer on the heterocyst formation process. Heterocyst

commitment takes approximately 10 hours. The pattern of cells marked for differentiation is

established early in development (as early as 2 hours[33]). Presumably heterocyst commitment

can be completed in less than 10 hours. One explanation for the temporal restriction on the

process may relate to the cell fate decision: control points in Anabaena restrict development until

the last possible moment to allow for termination of the process if a sourced of fixed nitrogen

suddenly becomes available. Because heterocyst cells represent an energetically-expensive

developmental endpoint, feedback loops increase the robustness of the system by decreasing

sensitivity to noise. This delay or timing mechanism may be attributable to concentrations of

HetR. HetR levels must overcome inhibition by PatS, and furthermore inhibit PatU3 before HetZ

(and consequently HetR conversion to HetR*) can be activated. Because progression of

heterocyst development continues only after these threshold levels of HetR are met, the time

necessary for commitment of differentiation may be established by the regulatory circuitry

proposed in Figure 2.17. This allows for a delay mechanism governing heterocyst formation.

The genetic arrangement of the hetZ-patU5-patU3 gene cluster may also correspond to a role in

regulating heterocyst development. In prokaryotes, genes related in function are often efficiently

organized into genetic regions known as operons. Control of the lactose operon in Escherichia

coli by trans-acting repressors, for example, remains a seminal model for gene regulation[34].

Genetic dissection of the hetZ-patU5-patU3 operon, however, is nontrivial. Genetic inactivation

required preservation of the series of transcriptional start points (tsps) present within each gene

of the gene cluster in order to examine the representative function of hetZ, patU5 and patU3. In

addition, the overlap of the ORFs of hetZ and patU5 in the chromosome may be significant in

establishing cellular levels of hetZ, patU5 and/or patU3. Individual mutations in the downstream

genes asr0102, alr0103 and asr0104 did not affect the wild-type phenotype[1], suggesting that

these genes are not directly under the control of tsps present in the hetZ-patU5-patU3 operon.

However in plasmid-based studies, the tsps present at the 3’ end of hetZ (rather than the gene

product of hetZ), appeared to be required for the full inhibitory function of patU5-patU3 (plasmids

pST355, pST429). When the normal transcriptional regulation of the gene cluster was replaced

by an inducible promoter, patU5-patU3 no longer functionally inhibited heterocysts (plasmid

pST417). Incorporation of hetZ to the same plasmid backbone containing patU5-patU3 (driven by

the same inducible promoter; plasmid pST416) was necessary to restore the inhibition of

67

heterocyst formation. This effect was presumably due to incorporation of internal tsps within hetZ

in the latter plasmid. This requirement of the internal hetZ tsps for function of patU3 on plasmids

are consistent with studies aimed at deciphering the contribution of tsp I and tsp II in expression

of the gene cluster. In the tsp I-tsp II (plasmid-based) studies, the tsps present at the 3’ end of

hetZ appear to regulate the gene cluster. Plasmids without the 3’ tsps (plasmids pST586,

pST587 and pST589) upstream of patU5-patU3 no longer inhibited heterocyst formation in the

wild-type. Because of the inverse orientation of the 3’ tsps, antisense transcription or activation of

the gene all0097 directly upstream but in the opposite direction of hetZ may explain the

mechanism of control by these tsps. Taken together, combined chromosomal and plasmid-based

studies of the hetZ-patU5-patU3 gene cluster reveal the significance of location, timing, and/or

level of transcription of the operon in heterocyst differentiation. These and other findings support

the presence of additional levels of regulation superimposed on current models of Anabaena

development, and merits further investigation.

68

REFERENCES CITED

1. Zhang, W., et al., A gene cluster that regulates both heterocyst differentiation and pattern

formation in Anabaena sp. strain PCC 7120. Mol. Microbiol., 2007. 66: p. 1429-1443.

2. Xu, X. and C.P. Wolk, Role for hetC in the transition to a nondividing state during

heterocyst differentiation in Anabaena sp. J. Bacteriol., 2001. 183: p. 393-396.

3. Du, Y., et al., Identification of the HetR recognition sequence upstream of hetZ in


4. Mitschke, J., et al., Dynamics of transcriptional start site selection during nitrogen stress-

induced cell differentiation in Anabaena sp. PCC7120. Proc Natl Acad Sci U S A, 2011.

108(50): p. 20130-5.



Bacteriol., 2007. 189: p. 2460-2467.

6. Fellay, R., J. Frey, and H. Krisch, Interposon mutagenesis of soil and water bacteria: a

family of DNA fragments designed for in vitro insertional mutagenesis of Gram-negative

bacteria. Gene, 1987. 52: p. 147-154.



2010. 77: p. 562-574.





Microbiol., 2005. 57: p. 111-123.

10. Karimova, G., et al., A bacterial two-hybrid system based on a reconstituted signal

transduction pathway. Proc Natl Acad Sci U S A, 1998. 95(10): p. 5752-6.

11. van den Ent, F., et al., Dimeric structure of the cell shape protein MreC and its functional

implications. Mol Microbiol, 2006. 62(6): p. 1631-42.

12. Patrick, J.E. and D.B. Kearns, MinJ (YvjD) is a topological determinant of cell division in

Bacillus subtilis. Mol Microbiol, 2008. 70(5): p. 1166-79.



Microbiol., 2012. 83: p. 682-693.

14. Wei, T.-F., R. Ramasubramanian, and J.W. Golden, Anabaena sp. strain PCC 7120 ntcA

gene required for growth on nitrate and heterocyst development. J. Bacteriol., 1994. 176:

p. 4473-4482.

69





77-84.

17. Nayar, A.S., et al., FraG is necessary for filament integrity and heterocyst maturation in

the cyanobacterium Anabaena sp. strain PCC 7120. Microbiology, 2007. 153: p. 601-607.

18. Risser, D.D. and S.M. Callahan, Genetic and cytological evidence that heterocyst

patterning is regulated by inhibitor gradients that promote activator decay. Proc Natl Acad

Sci U S A, 2009. 106(47): p. 19884-8.

19. Callahan, S.M., unpublished.



21. Meeks, J.C., et al., Cellular differentiation in the cyanobacterium Nostoc punctiforme.

Arch Microbiol, 2002. 178(6): p. 395-403.

22. Spencer, W., et al., CtrA, a global response regulator, uses a distinct second category of

weak DNA binding sites for cell cycle transcription control in Caulobacter crescentus. J

Bacteriol, 2009. 191(17): p. 5458-70.

23. Flaherty, B.L., et al., Directional RNA deep sequencing sheds new light on the

transcriptional response of Anabaena sp. strain PCC 7120 to combined-nitrogen

deprivation. BMC Genomics, 2011. 12: p. 332.



Acad. Sci. U.S.A., 2004. 101(14): p. 4848-4853.

25. Mattick, J.S., A new paradigm for developmental biology. J Exp Biol, 2007. 210(Pt 9): p.

1526-47.

26. Kaern, M., et al., Stochasticity in gene expression: from theories to phenotypes. Nat Rev

Genet, 2005. 6(6): p. 451-64.

27. Sakr, S., et al., Inhibition of cell division suppresses heterocyst development in Anabaena

sp. strain PCC 7120. J Bacteriol, 2006. 188(4): p. 1396-404.


sp. strain PCC 7120. J. Bacteriol., 2006. 188: p. 1396-1404.

29. Sakr, S., et al., Relationship among several key cell cycle events in the developmental

cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol., 2006. 188: p. 5958-5965.



70


12: p. 30-39.

32. Hartwell, L.H. and T.A. Weinert, Checkpoints: controls that ensure the order of cell cycle

events. Science, 1989. 246(4930): p. 629-34.

33. Ionescu, D., et al., Heterocyst-specific transcription of NsiR1, a non-coding RNA encoded

in a tandem array of direct repeats in cyanobacteria. J Mol Biol, 2010. 398(2): p. 177-88.

34. Jacob, F. and J. Monod, Genetic regulatory mechanisms in the synthesis of proteins. J

Mol Biol, 1961. 3: p. 318-56.

71

CHAPTER 3. HETEROCYST DEVELOPMENT AND THE SER/THR PROTEIN

PHOSPHATASE ALL1758 IN THE CYANOBACTERIUM ANABAENA SP. STRAIN PCC 7120

INTRODUCTION

Heterocysts are terminally differentiated, non-dividing cells that allow the simultaneous execution

of two incompatible processes, fixation of nitrogen and oxygen-evolving photosynthesis, in some

filamentous cyanobacteria. They maintain a micro-oxic environment where oxygen-labile

nitrogenase can function. The envelope of heterocysts includes a layer of distinct glycolipid that

reduces the diffusion of molecular oxygen into cells, an outer layer of heterocyst-specific

exopolysaccharides that protect the integrity of the glycolipid layer, and increased respiration

reduces the level of molecular oxygen that does enter heterocysts [1]. Fixed nitrogen from

heterocysts supports the growth of vegetative cells in the filament, which in turn supply

heterocysts with fixed carbon [2, 3]. In the cyanobacterium Anabaena sp. PCC 7120 (herein,

Anabaena) heterocysts are arranged in a periodic pattern at intervals of approximately 10 cells

along unbranched filaments under conditions that require fixation of nitrogen for growth of the

organism [4].

Activation of the heterocyst differentiation signaling pathway in response to nitrogen limitation

leads to an accumulation of the metabolite 2-oxoglutarate (2-OG) in cells [5]. Binding of 2-OG

stimulates the DNA-binding activity of the global transcriptional regulator of nitrogen and carbon

metabolism in cyanobacteria, NtcA [6]. NtcA indirectly leads to upregulation of the master

regulator of heterocyst differentiation, hetR. Whereas heterocysts cannot develop in hetR-

deficient strains, in our hands, heterocysts occasionally arise in ntcA backgrounds.

Overexpression of hetR in the latter strain can also restore heterocyst formation to a wild-type

pattern. HetR feeds back on the regulation of transcription of ntcA [7], and together with NtcA,

directly or indirectly activates genes involved in the patterning and morphogenesis of heterocysts

[8]. Induction of differentiation, pattern formation, and morphogenesis involves the transcription of

hundreds of genes specific for the formation of heterocysts.

The reversibility of protein phosphorylation mediated by kinases and phosphatases is an

essential part of many signaling cascades and regulatory networks. Protein phosphatases (PPs)

are divided into three superfamilies. These include (i) protein aspartate phosphatases, (ii) protein

tyrosine phosphatases (divided into the conventional or class I family and the low molecular

weight or class II family), and (iii) protein serine/threonine phosphatases. Protein

serine/threonine phosphatases are subdivided into PPP or PPM (also referred to as PP2Cs)

families based on the dependence on divalent metal cations in the latter group [9, 10]. Although

structurally homologous to each other, PPPs and PP2Cs differ at the primary sequence level.

The approximately 290-residue catalytic domain of PPMs consists of eleven conserved motifs

72

that contain eight “absolutely” conserved amino acids [11, 12]. At the catalytic site of PP2Cs, four

invariant aspartate residues (in Motifs 1, 2, 8, 11) coordinate two metal ions. The metal ions have

been proposed to activate the nucleophilic activity of water molecules to catalyze

dephosphorylation of phosphoserine and phosphothreonine residues [13].

In Anabaena an estimated 4.2% of the annotated genome encodes two-component systems,

kinases, and phosphatases, suggesting increased signaling capability as compared to other

prokaryotes[14]. Genes with homology to signaling components have been reported to influence

nitrogen metabolism and the early regulatory stages of heterocyst development in Anabaena.

NrrA which encodes a response regulator of the OmpR family, acts early in the differentiation

process by directly upregulating hetR expression [15]. The PP2C-type protein phosphatases

PrpJ1 and PrpJ2 are also involved in the initiation of heterocyst development through mutual

regulation of both ntcA and hetR [16]. The protein kinases Pkn41 and Pkn42, which contain

Ser/Thr-kinase and His-kinase domains respectively, are cotranscribed specifically under iron

deficient conditions and regulated by NtcA [17]. The nitrogen regulatory protein PII encoded by

glnB, is differentially modified in the two cell types. It goes from a phosphorylated to non-

phosphorylated state during transition from vegetative cell to heterocyst [18]. The pknE gene lies

301 bp downstream from the protein phosphatase 1/2A/2B homolog encoded by prpA. Both

pknE and prpA inactivation mutants produce aberrant heterocysts [19]. Overexpression of the

putative Ser/Thr kinase encoded by pknE from its native promoter inhibited heterocyst

development, possibly through inhibition of HetR [20]. Another protein that appears to affect HetR,

PatA, contains a phosphoacceptor domain with homology to the response regulator CheY,

although a corresponding histidine kinase has not been isolated. It appears to attenuate the

negative regulation of heterocyst differentiation by PatS and HetN [21], while promoting the

activity, and limiting the accumulation, of HetR [22].

In addition, protein phosphorylation has also been implicated in the later stages of development,

particularly during the process of heterocyst maturation. Formation of the minor heterocyst

glycolipid involves two protein kinases of the hstK family, Pkn30 and Pkn44, which contain both

an N-terminal Ser/Thr kinase domain and a C-terminal His-kinase domain [23]. PrpJ1 has been

shown to regulate the synthesis of the major heterocyst glycolipid [24]. Alr0117 and HepK

(All4496) both encode putative two-component system sensory histidine kinases that appear to

be involved in the induction of hepA, one of the genes responsible for synthesis of the heterocyst

polysaccharide layer [25-27]. The manganese-dependent Ser/Thr protein phosphatase of the

PPP family DevT (Alr4674) accumulates in mature heterocysts, is not regulated by NtcA, and

appears to have a role in the later steps of heterocyst differentiation [28].

73

The coordination of development and other complex cellular processes is often regulated at the

level of transcription. In prokaryotes, transcription initiation is controlled by sigma factors. Sigma

factor association with RNA polymerase at specific promoters activates transcription from a

subset of genes. This allows for the rapid reprogramming of gene expression in response to

intracellular and extracellular signals.

One common mechanism for control of sigma factor activity involves reversible phosphorylation

of sigma factor regulators [29]. A sigma factor can oscillate between RNA polymerase (activating

transcription) and its cognate antisigma factor (inactivating transcription). Similarly, the antisigma

factor can also oscillate between its cognate sigma factor and anti-antisigma factor (or “antisigma

factor agonist”). Dephosphorylation of anti-antisigma factors by phosphatases is a central

component to the ‘partner-switch’ mechanism. This signaling network has been described in

other prokaryotes. In Bacillus subtilis the PP2Cs SpoIIE, RsbU and RsbP dephosphorylate their

cognate anti-antisigma factors (SpoIIAA and RsbV) in the control of sigma factors regulating cell

division and stress response, respectively [30-32]. In this study, the isolation and characterization

of a gene encoding a putative PP2C-type protein phosphatase all1758 is described.


Bacterial strains and growth conditions. Strains used in this study are described in Table 3.1.

The growth of Escherichia coli and Anabaena sp. strain PCC 7120 and its derivatives;

concentrations of antibiotics; and induction of heterocyst formation in BG-110, which lacks a

combined-nitrogen source; regulation of PpetE and Pnir expression; and conditions for

photomicroscopy were as previously described [33]. Images were processed in Adobe

Photoshop CS2. To avoid complications from the vacuolization phenotype observed with strains

UHM183 and UHM184 upon growth in BG-11 liquid medium, which contains a source of

combined nitrogen, these strains were transferred from solid BG-11 medium to liquid BG-110

medium for induction of heterocyst differentiation. Plasmids were conjugated from E. coli into

Anabaena strains as previously described [34].

Construction of plasmids used in this study. Plasmids used in this study are described in

Table 3.2. Oligonucleotides used in this study are described in Table 3.3. Constructs derived by

PCR were sequenced to verify the integrity of the sequence.

Suicide plasmids

Plasmid pST112 was used to delete most of the coding region of all1758. An 852-bp region

containing the first 30-bp of the all1758 coding region and upstream DNA was amplified from the

chromosome (with primers 1758 up F and 1758 up R) and fused to an 848-bp region containing

the last 30-bp of the all1758 coding region and downstream DNA (using primers 1758 down F

74

and 1758 down R) via overlap extension PCR. The 1700-bp fragment was cloned into pRL277

as a BglII-SacI fragment using restriction sites introduced on the primers to generate pST112.

Plasmid pST114 was used to replace most of the coding region of all1758 with an Ω interposon.

The 1700-bp fragment used for construction of pST112 was moved into the EcoRV site of

pBluescript SK+ (Stratagene) to generate pST110. The 1700-bp fragment was moved from

pST110 into pRL278 as a BlgII-SacI fragment to create pST113. To generate pST114, the 2082-

bp Ω interposon, which confers resistance to Sp and Sm [35], was introduced into pST113 at a

SmaI site generated during overlap extension PCR.

Plasmid pST426 was used to delete most of the coding region of all1087. An 815 bp region


chromosome (with primers alr1087-up-BamHI-F and alr1087-upR) and fused to a 799-bp region

containing the last 30-bp of the all1087 coding region and downstream DNA (using primers

alr1087-down-F and alr1087-down-SacI-R) via overlap extension PCR. The 1614-bp fragment

was ligated into the EcoRV site of pBluescript SK+ (Stratagene) to generate pST425 and

subsequently moved from pST426 into pRL277 as a BglII-SacI fragment using restriction sites

introduced on the primers to generate pST426.


A 1614-bp fragment was moved from pST425 into pRL278 as a BglII-SacI fragment to create



Plasmid pST448 was used to delete most of the coding region of alr3758. An 880-bp region

containing the first 34-bp of the alr3758 coding region and upstream DNA was amplified from the

chromosome (with primers alr3758 up F and alr3758 up R) and fused to an 886-bp region

containing the last 35-bp of the alr3758 coding region and downstream DNA (using primers

alr3758 down F and alr3758 down R) via overlap extension PCR. The 1766-bp fragment was

ligated into the EcoRV site of pBluescript SK+ (Stratagene) to generate pST447 and



Plasmid pST450 was used to replace most of the coding region of alr3758 with an Ω interposon.




Plasmid pST452 was used to delete most of the coding region of all0648. An 845-bp region


75

chromosome (with primers all0648 up F and all0648 up R) and fused to an 837-bp region

containing the last 42-bp of the all0648 coding region and downstream DNA (using primers

all0648 down F and all0648 down R) via overlap extension PCR. The 1682-bp fragment was








Plasmid pST520 was used to delete most of the coding region of alr3243. An 847-bp region

containing the first 27-bp of the alr3423 coding region and upstream DNA was amplified from the

chromosome (with primers alr3423 up F and alr3423 up R) and fused to an 826-bp region

containing the last 30-bp of the alr3423 coding region and downstream DNA (using primers

alr3423 down F and alr3423 down R) via overlap extension PCR. The 1763-bp fragment was




Plasmid pST522 was used to replace most of the coding region of alr3423 with an Ω interposon.




Multicopy plasmids

Plasmid pST141 is a mobilizable shuttle vector containing the putative promoter and coding

regions of all1758. An 1866-bp fragment was amplified from the chromosome (using primers

Pall1758 BamHI F and all1758 SacI R) and cloned into pAM504 as a BamHI-SacI fragment using

the same restriction sites engineered on the primers to generate pST141.


regions of all1087. A 1209-bp fragment was amplified from the chromosome (using primers

Pall1087 BamHI F and all1087 SacI R) and ligated into the EcoRV site of pBluescript SK+

(Stratagene) to generate pST491. To make pST494, the fragment was moved into pAM504 as a

BamHI-SacI fragment using restriction sites engineered on the primers.

76

pST530 is a mobilizable shuttle vector containing the putative promoter and coding regions of

alr3758. A 673-bp fragment was amplified from the chromosome (using primers Palr3758 BamHI

F and alr3758 SacI R) and ligated into the EcoRV site of pBluescript SK+ (Stratagene) to

generate pST529. To make pST530, the fragment was moved into pAM504 as a BamHI-SacI

fragment using the same restriction sites engineered on the primers.


regions of alr0648. A 783-bp fragment was amplified from the chromosome (using primers

Pall0648 BamHI F and all0648 SacI R) and cloned into pAM504 as a BamHI-SacI fragment using

the same restriction sites engineered on the primers to generate pST496.


regions of alr3423. A 414-bp fragment was amplified from the chromosome (using primers

Palr3423 BamHI F and alr3423 SacI R) and ligated into the EcoRV site of pBluescript SK+

(Stratagene) to generate pST523. To make pST524, the fragment was moved into pAM504 as a

BamHI-SacI fragment using the same restriction sites engineered on the primers.

Plasmids with nucleotide substitutions in all1758

Plasmid pST399 is a mobilizable shuttle vector containing the all1758 promoter and coding region

with the substitution D267A. An upstream fragment (amplified using primers Pall1758-Bam HI-F

and all1758 D267A-R) was fused to downstream fragment (amplified using primers all1758

D267A-F and all1758 Sac I-R) via overlap extension PCR. To yield pST399, the 1868-bp

fragment was cloned as a BamHI-SacI fragment into pAM504 using same restriction sites

engineered on the primers.








with the substitution A348G. An upstream fragment (amplified using primers Pall1758-Bam HI-F

and all1758 A348G-R) was fused to downstream fragment (amplified using primers all1758

A348G-F and all1758 Sac I-R) via overlap extension PCR, and ligated into the EcoRV site of

pBluescript SK+ to make pST396. To yield pST401, the 1868-bp fragment was moved from

77

pST395 as a BamHI-SacI fragment into pAM504 using same restriction sites engineered on the

primers.








with the substitution D453E. An upstream fragment (amplified using primers Pall1758-Bam HI-F

and all1758 D453E-R) was fused to downstream fragment (amplified using primers all1758

D453E-F and all1758 Sac I-R) via overlap extension PCR, and ligated into the EcoRV site of



primers.




D267E-F and all1758 Sac I-R) via overlap extension PCR and ligated into the EcoRV site of



primers.







primers.




78




primers.




D453E-F and all1758 Sac I-R) via overlap extension PCR, and ligated into the EcoRV site of



primers.


with the substitution T329S. An upstream fragment (amplified using primers Pall1758-Bam HI-F

and all1758 T329-R) was fused to downstream fragment (amplified using primers all1758 T329-F

and all1758 Sac I-R) via overlap extension PCR and ligated into the EcoRV site of pBluescript

SK+ to make pST466. To yield pST469, the 1868-bp fragment was moved from pST466 as a

BamHI-SacI fragment into pAM504 using same restriction sites engineered on the primers.


with the substitution G373A. An upstream fragment (amplified using primers Pall1758-Bam HI-F

and all1758 G373A-R) was fused to downstream fragment (amplified using primers all1758

G373A-F and all1758 Sac I-R) via overlap extension PCR and ligated into the EcoRV site of



primers.


with the substitution G399A. An upstream fragment (amplified using primers Pall1758-Bam HI-F

and all1758 G399A-R) was fused to downstream fragment (amplified using primers all1758

G399A-F and all1758 Sac I-R) via overlap extension PCR and ligated into the EcoRV site of



primers.

79

Overexpression plasmids

Two plasmids carrying a PpetE-all1758 transcriptional fusion were constructed. Plasmid pST156 is

a mobilizable shuttle vector carrying a transcriptional fusion between the petE promoter and

all1758. The 1392-bp coding region of all1758 was amplified from the chromosome using

primers all1758 OE EcoRI F and all1758 OE BamHI R and ligated into the EcoRV site of

pBluescript SK+ to create pST184. The fragment was moved from pST184 as an EcoRI-BamHI

fragment into pKH256 [36], which is designed to create transcriptional fusions to the petE

promoter, to generate pST156. Plasmid pST367 differs from pST156 in the engineered

ribosomal binding site that was introduced by using primer all1758 EcoRI rbs F in place of all1758

OE EcoRI F to amplify all1758 from the chromosome. The PCR fragment was moved into the

EcoRV site of pBluescript SK+ to generate pST365, and subsequently moved from pST365 as a


The transcriptional fusion Pnir-all1758 present on plasmid pST368 was created using the primers

all1758 rbs EcoRI F and all1758 OE BamHI R to amplify all1758 from the chromosome. The

fragment was ligated into the EcoRV site of pBluescript SK+ to generate pST366. To create

pST368, the fragment was moved as a EcoRI-BamHI fragment into pSMC188, a replicating

plasmid used to generate transcriptional fusions to the nirA promoter [33]

Plasmids pST460 and pST463 carry transcriptional fusions between all1087 and the petE

promoter and the nirA promoter respectively. The 390-bp coding region of all1087 was amplified

from the chromosome using primers all1087 SmaI F and all1087 KpnI R and ligated into the

EcoRV site of pBluescript SK+ to generate pST457. To make pST460, the fragment was moved

from pST457 as a SmaI-KpnI fragment into pDR311 (bearing a PpetE-lacZ transcriptional

fusion[22]) digested with the same enzymes to exchange lacZ for all1087. To make pST463, the

fragment was moved from pST457 as a SmaI-KpnI fragment into pDR311 (bearing a Pnir-lacZ

transcriptional fusion[22]) digested with the same enzymes to exchange lacZ for all1087.

Plasmids pST461 and pST464 carry transcriptional fusions between alr3758 and the petE

promoter and the nirA promoter respectively. The coding region of alr3758 was amplified from

the chromosome using primers alr3758 SmaI F and alr3758 KpnI R and ligated into the EcoRV

site of pBluescript SK+ to generate pST458. The 339-bp fragment was moved from pST458 as a

SmaI-KpnI fragment into pDR350 to yield pST461, and into pDR311 to yield pST464.


promoter and the nirA promoter respectively. The coding region of all0648 was amplified from

the chromosome using primers all0648 SmaI F and all0648 KpnI R and ligated into the EcoRV

80



Plasmids pST475 and pST478 carry transcriptional fusions between alr3423 and the petE

promoter and the nirA promoter respectively. The coding region of alr3423 was amplified from

the chromosome using primers alr3423 SmaI F and alr3423 KpnI R and ligated into the EcoRV




promoter and the nirA promoter respectively. The coding region of all1702 was amplified from

the chromosome using primers all1702 SmaI F and all1702 KpnI R and ligated into the EcoRV



Plasmids pST477 and pST480 carry transcriptional fusions between sigB (all7615) and the petE

promoter and the nirA promoter respectively. The coding region of sigB was amplified from the

chromosome using primers sigB SmaI F and sigB KpnI R and ligated into the EcoRV site of

pBluescript SK+ to generate pST474. The 999-bp fragment was moved from pST474 as a SmaI-

KpnI fragment into pDR350 to yield pST477, and into pDR311 to yield pST480.

Plasmids pST536 and pST537 carry transcriptional fusions between sigE (alr4259) and the petE

promoter and the nirA promoter respectively. The coding region of sigE was amplified from the

chromosome using primers sigE SmaI F and sigE KpnI R and ligated into the EcoRV site of

pBluescript SK+ to generate pST535. The 1194-bp fragment was moved from pST535 as a


Plasmids pST437 and pST456 are a mobilizable shuttle vectors carrying a transcriptional fusion

between the petE promoter and sigE (alr4249). The 1194-bp coding region of sigE was

amplified from the chromosome using primers sigE OE EcoRI F (PpetE) and sigE BamHI R and

ligated into the EcoRV site of pBluescript SK+ to create pST435. To yield pST437, the fragment

was moved from pST435 as an EcoRI-BamHI fragment into pKH256. To yield pST456, the

fragment was moved from pST435 as an EcoRI-BamHI fragment into pPAV213, another plasmid

designed to create transcriptional fusions to the petE promoter.

Plasmid pST438 is a mobilizable shuttle vector carrying a transcriptional fusion between the nir

promoter and sigE (alr4249). The 1194-bp coding region of sigE was amplified from the

chromosome using primers sigE OE EcoRI F (Pnir) and sigE BamHI R and ligated into the

EcoRV site of pBluescript SK+ to create pST436. To yield pST438, the fragment was moved

from pST435 as an EcoRI-BamHI fragment into pSMC188.

81

Transcriptional and translational fusion constructs

To create the gfp transcriptional reporter for all1758, a 474-bp region upstream of all1758 was

amplified from the chromosome with primers Pall1758 SacI F and all1758 15 bp up SmaI R and

cloned into pAM1956 as a SacI-SmaI fragment to create pST151.

The region cloned into plasmid pST151 was amplified using the same primers Pall1758 SacI F

and all1758 15 bp up SmaI R (used to clone pST151) and ligated into the EcoRV site of

pBluescript SK+ to generate pST374. To make pST375, the 474-bp fragment was subsequently

cloned into pAM504 as a SacI-SmaI fragment to generate pST375.


amplified from the chromosome using primers Pall1759-SacI-F and Pall1759-SmaI-R and moved

into the EcoRV site of pBluescript SK+ to generate pST252. The fragment was moved from

pST252 as a SacI-SmaI fragment into pAM1956 to create pST249.

To create the gfp transcriptional reporter for all1758 with a mutation at the putative -203 NtcA-

binding site, an upstream fragment of Pall1758 (amplified from pST151 with primers Pall1758 SacI F

and Pall1758 mut NtcA up R) was fused to a downstream fragment of Pall1758 (amplified from

pST151 with primers Pall1758 mut NtcA down F and all1758 15 bp up SmaI R) and ligated into

the EcoRV site of pBluescript SK+ to generate pST489. To make pST490, the 474-bp fragment

was subsequently cloned into pAM1956 as a SacI-SmaI fragment.



and Pall1758 mut NtcA up R -128) was fused to a downstream fragment of Pall1758 (amplified from

pST151 with primers Pall1758 mut NtcA down F -128 and all1758 15 bp up SmaI R) and ligated

into the EcoRV site of pBluescript SK+ to generate pST538. To make pST540, the 474-bp

fragment was subsequently cloned into pAM1956 as a SacI-SmaI fragment.



and Pall1758 mut NtcA up R -323) was fused to a downstream fragment of Pall1758 (amplified from

pST151 with primers Pall1758 mut NtcA down F -323 and all1758 15 bp up SmaI R) and ligated

into the EcoRV site of pBluescript SK+ to generate pST539. To make pST541, the 474-bp

fragment was subsequently cloned into pAM1956 as a SacI-SmaI fragment.

To create the gfp transcriptional reporter for all1087, an 820-bp region upstream of all1087 was

amplified from the chromosome with primers Pall1087 SacI F and Pall1087 SmaI R and ligated

82

into the EcoRV site of pBluescript SK+ to generate pST497. To make pST499, the fragment was

subsequently cloned into pAM1956 as a SacI-SmaI fragment.

To create the gfp transcriptional reporter for all1087 with a mutation at the putative NtcA-binding

site, an upstream fragment of Pall1087 (amplified from the chromosome with primers Pall1087 SacI

F and Palr1087-dNtcA-OEX-R) was fused to a downstream fragment of Pall1087 (amplified from the

chromosome with primers Palr1087-dNtcA-OEX-F and Pall1087 SmaI R) and ligated into the

EcoRV site of pBluescript SK+ to generate pST498. To make pST500, the 820-bp fragment was


To create the gfp transcriptional reporter for alr3758, a 334-bp region upstream of all1087 was

amplified from the chromosome with primers Palr3758 SacI F and Palr3758 SmaI R 1 and ligated



To create the gfp transcriptional reporter for alr3758 with a mutation at the putative NtcA-binding

site, an upstream fragment of Palr3758 (amplified from the chromosome with primers Palr3758 SacI

F and Palr358 mut ntcA R) was fused to a downstream fragment of Pall1087 (amplified from the

chromosome with primers Palr3758 mut ntcA F and Palr3758 SmaI R 1) and ligated into the

EcoRV site of pBluescript SK+ to generate pST482. To make pST486, the 820-bp fragment was










The translational all1758-gfp reporter fusion under the control of the all1758 promoter was

created using primers Pall1758 BamHI F and all1758 SmaI R to amplify the 1866-bp fragment

from the chromosome. The fragment was moved as a BamHI-SmaI fragment into pSMC232, a

replicating plasmid used generate translational fusions to gfp [22], to give pST150.

To make plasmid pST533, an upstream fragment of Pall1758 (amplified from the chromosome with

primers Pall1758 SacI F and Pall1758 mut ntcA R) was fused to a downstream fragment of Pall1087

(amplified from the chromosome with primers Pall1758 mut ntcA F and Pall1758 SmaI R) and

83

ligated into the EcoRV site of pBluescript SK+ to generate pST533. To make pST534, the 1866-

bp fragment was subsequently cloned into pSMC232 as a BamHI-SmaI fragment.

The translational all1087-gfp reporter fusion under the control of the all1087 promoter (amplified

from the chromosome using primers Pall1087 BamHI F and all1087 SmaI R) was ligated into the

EcoRV site of pBluescript SK+ to generate pST505. The 1209-bp fragment was subsequently

moved as a BamHI-SmaI fragment into pSMC232 to give pST511.

To make plasmid pST512, an upstream fragment of Pall1087 (amplified from the chromosome with

primers Pall1087 SacI F and Palr1087-dNtcA-OEX-R) was fused to a downstream fragment of

Pall1087 (amplified from the chromosome with primers Palr1087-dNtcA-OEX-F and Pall1087 SmaI

R) and ligated into the EcoRV site of pBluescript SK+ to generate pST506. To make pST512, the

1209-bp fragment was subsequently cloned into pSMC232 as a BamHI-SmaI fragment.

The translational alr3758-gfp reporter fusion under the control of the alr3758 promoter (amplified

from the chromosome using primers Palr3758 BamHI F and alr3758 SmaI R) was ligated into the



To make plasmid pST514, an upstream fragment of Palr3758 (amplified from the chromosome with

primers Palr3758 SacI F and Palr3758 mut ntcA R) was fused to a downstream fragment of

Palr3758 (amplified from the chromosome with primers Palr3758 mut ntcA F and Palr3758 SmaI R)

and ligated into the EcoRV site of pBluescript SK+ to generate pST532. To make pST514, the

673-bp fragment was subsequently cloned into pSMC232 as a BamHI-SmaI fragment.

The translational all0648-gfp reporter fusion under the control of the all0648 promoter (amplified

from the chromosome using primers Pall0648 BamHI F and all0648 SmaI R) was ligated into the



The translational alr3423-gfp reporter fusion under the control of the alr3423 promoter (amplified

from the chromosome using primers Palr3423 BamHI F and alr3423 SmaI R) was ligated into the



Protein expression vector

The pPROEX-1 (Life Technologies) derivative, pST188 was used to overexpress all1758 in E.

coli and facilitate protein purification. A 1392-bp region containing the coding region for all1758

was amplified from the chromosome using primers all1758-NdeI-F and all1758-NH6-NotIR, and

84

moved into the EcoRV site of pBluescript SK+ to create pST192. To generate pST188, the

fragment was moved as a NdeI-NotI fragment into the pPROEX-1 backbone derived from

pSMC148 [22] digested with NdeI-NotI. The resulting open reading frame, designated

all1758(H6), encodes a glycine, then six histidine codons preceding sixteen codons

(DYDIPTTENLYFQGAHAH, which contains a sequence recognized by the Tobacco Etch Virus

protease) fused to all1758 followed by a stop codon after the last codon of all1758.

Bacterial two-hybrid constructs

For each construct, a 32- or 33-bp linker domain was introduced between the adenylate cyclase

fragment and the protein of interest on either the forward or reverse primer. Details concerning

the construction of plasmids pST558-562, pST564, pST597, pST565-pST571, pST598, pST572-

pST576, pST578, pST599, pST579-pST583, pST585, pST600 for fusions to hetR, hetF, hetZ,

patU5, patU3, patA, hetF(C246A) are described in the Materials and Methods section of Chapter

2.

The all1758 coding region was amplified from the chromosome using the primers all1758 BamHI

F A and all1758 EcoRI R A, and cloned into the EcoRV site of pBluescript SK+ to make pST549.

To make pST563 and pST570, the 1392 bp fragment was moved as a BamHI-EcoRI fragment

from pST556 into pKT25 and pUT18C respectively.

The all1758 coding region was amplified from the chromosome using the primers all1758 BamHI

F B and all1758 EcoRI R B, and cloned into the EcoRV site of pBluescript SK+ to make pST556.

To make pST577 and pST584, the 1392 bp fragment was moved as a BamHI-EcoRI fragment

from pST556 into pKNT25 and pUT18 respectively.

Strain construction. Descriptions of strains constructed in this study are summarized in Table

3.1. Deletions and replacements of chromosomal DNA was performed as previously described

[37]. All strains were screened via colony PCR with primers annealing outside of the

chromosomal region introduced on the suicide plasmids used for strain construction and tested

for sensitivity and resistance to the appropriate antibiotics. For strains with mutations in more

than one gene, mutations were introduced in the order indicated in the strain description.

Deletion of the all1758 coding region was performed using plasmids pST112 and pST114 and

Anabaena sp. strain PCC 7120 to yield strains Δall1758 (UHM183) and Δall1758 with Ω cassette

(UHM184), respectively. The resultant strains were screened via colony PCR with primers

all1758 flank up F and all1758 flank down R, which anneal outside the chromosomal region

85

introduced onto pST112 and pST114 for deletion of all1758, and tested for sensitivity and

resistance to the appropriate antibiotics. PCR with primers all1758 3’ F and all1757 5’ R were

used to test if recombination between plasmid pST141, with harbors Pall1758-all1758, and the

chromosomal locus of all1758 of strain UHM183 had occurred.

Plasmid pST112 was introduced into the following parent strains to create unmarked deletions in

the all1758 coding region: (1) ΔhetR for construction of ΔhetR Δall1758, (2) ∆patA for

construction of ∆patA Δall1758, (3) ΔhetF for construction of ΔhetF Δall1758, (4) PhetR-hetR-gfp

for construction of PhetR-hetR-gfp Δall1758, (5) PpetE-hetR-gfp for construction of PpetE-hetR-gfp

Δall1758, (6) ∆patA PpetE-hetR-gfp for construction of ∆patA PpetE-hetR-gfp Δall1758, (7) PpetE-

hetR-cfp for construction of PpetE-hetR-cfp Δall1758 and (8) ∆patA PpetE-hetR-cfp for construction

of ∆patA PpetE-hetR-cfp Δall1758.

Strain ∆ntcA with Ω cassette was constructed by introduction of pSMC205 into Anabaena sp.

strain PCC 7120.

Deletion of the all1087 coding region was performed using plasmids pST426 and pST428 and

Anabaena sp. strain PCC 7120 to yield strains Δall1087 (UHM325) and Δall1087 with Ω cassette


all1087 flank up F and all1087 flank down R, which anneal outside the chromosomal region

introduced onto pST426 and pST428 for deletion of all1087, and tested for sensitivity and

resistance to the appropriate antibiotics.

Deletion of the alr3758 coding region was performed using plasmid pST448 and Anabaena sp.

strain PCC 7120 to yield strain Δalr3758 (UHM302). The resultant strain was screened via

colony PCR with primers alr3758 flank up F and alr3758 flank down R, which anneal outside the

chromosomal region introduced onto pST448 for deletion of alr3758, and tested for sensitivity and


Deletion of the all0648 coding region was performed using plasmid pST452 and Anabaena sp.

strain PCC 7120 to yield strain Δall0648 (UHM303). The resultant strain was screened via colony

PCR with primers all0648 flank up F and all0648 flank down R, which anneal outside the

chromosomal region introduced onto pST452 for deletion of all0648, and tested for sensitivity and


Deletion of the alr3423 coding region was performed using plasmids pST520 and pST522 and

Anabaena sp. strain PCC 7120 to yield strains Δalr3423 (UHM326) and Δalr3423 with Ω cassette


alr3423 flank up F and alr3423 flank down R, which anneal outside the chromosomal region

86

introduced onto pST520 and pST522 for deletion of alr3423, and tested for sensitivity and


Overexpression and purification of recombinant all1758 and phosphatase assays.

Recombinant All1758 was produced from E. coli BL21 (DE3) transformed with pST188 using Ni-

nitrilotriacetic acid affinity chromatography (Qiagen) as described previously [33] with the

following exception: 4 mL of wash buffer (50 mM NNaH2PO4, 300 mM NaCl, 20 mM imidazole,

20% glycerol) was added to the column twice, followed by four 500 µL volumes of elution buffer

(50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazole, 20% glycerol), each diluted four-fold directly

into a total volume of 1.5 mL of capture buffer (50 mM NaH2PO4, 300 mM NaCl) upon elution

from the column.

Assays for phosphatase activity were conducted as described by previously [38]. The substrate,

4-nitrophenol phosphate (pNPP), was dissolved in pNPP buffer containing 10 mM Tris-HCl, pH

8.5, 1 mM DDT, and 50 mM NaCl with addition of either 2 mM MnCl2 or 5 mM MgCl2 and

supplemented with either 10 or 100 µM cAMP or cGMP. As a test of hydrolysis of pNPP by the

putative phosphatase activity of All1758, 2 µg of the recombinant protein was added to the

cocktail at a final reaction volume of 1 mL. Absorbance at 400 nm was monitored as a

measurement of hydrolysis of pNPP at 25 °C every thirty seconds for the first thirty minutes after

addition of protein to the cocktail, then at 45 minutes, 1 hour, 2 hours, 5 hours, and 24 hours.

Measurement of cell size. Reported values for cell sizes were obtained by measuring the

“height” of one hundred randomly distributed cells of Anabaena sp. strains PCC 7120 and

UHM184. Height refers to the dimension perpendicular to the longitudinal axis of the filament

halfway between cell poles. Significance of height distributions was established using an

unpaired t-test with a two-tailed P value of less than 0.0001. Approximate cell volumes were

calculated by treating cells as spheres using the average cell height as the radius in the formula,

Volumesphere=4/3πr3.

Confocal microscopy. Cells were routinely viewed and imaged as described previously [37].

Confocal microscopy was performed using an Olympus Fluoview 1000 laser scanning confocal

mounted on an IX81 motorized inverted microscope. Fluorescence from Turbo-YFP was

detected with an excitation of 525 nm and an emission of 538 nm. All images were processed in

Adobe Photoshop CS2.

RNA isolation and RT-PCR. Total RNA was extracted as previously described [22]. For

reverse-transcription PCR (RT-PCR), 0.5 µg of total RNA was used for the synthesis of cDNA

with reverse transcriptase and the corresponding reverse primers for subsequent use in PCR

carried out for 27 cycles using primers 22 to 31 in a manner as previously described [22].

87

Acetylene reduction, glycolipid and exopolysaccharide assays. Acetylene reduction assays

measured from three independent cultures were performed as previously described [37].

Glycolipid extraction for thin layer chromatography and staining of exopolysaccharides also were

conducted as previously described [39].

Bacterial two-hybrid and β-galactosidase assays. Bacterial two-hybrid assays were

performed as previously described [40-42]. The average β-galactosidase-specific activity (three

replicates) of each positive protein-protein interaction was determined as previously described

[42].

88

Table 3.1. Strains used in Chapter 3

Anabaena sp. strain



ΔhetR hetR-deletion strain [43] (UHM103)

Δpbp6 pbp6-deletion strain [44] (UHM128)

PhetR-hetR-gfp PCC 7120 with chromosomal PhetR-hetR replaced by PhetR-hetR-gfp

[45] (UHM148)

PpetE-hetR-gfp PCC 7120 with chromosomal PhetR-hetR replaced by PpetE-hetR-gfp

[46] (UHM139)

∆patA PpetE-hetR-gfp

patA-deletion strain with chromosomal PhetR-hetR replaced by PpetE-hetR-gfp

[46] (UHM136)

PpetE-hetR-cfp PCC 7120 with chromosomal PhetR-hetR replaced by PpetE-hetR-cfp

[45](UHM217)

∆patA ∆hetF PpetE-hetR-cfp

patA-, hetF-deletion strain with chromosomal PhetR-hetR replaced by PpetE-hetR-cfp

[45] (UHM191)

Δall1758 all1758-deletion strain This study (UHM183)

Δall1758 with Ω cassette

all1758 replaced by Ω cassette This study (UHM184)

ΔhetR Δall1758 hetR-, all1758-deletion strain This study (UHM220)

∆patA Δall1758 patA-, all1758-deletion strain This study (UHM219)

∆hetF Δall1758 hetF-, all1758-deletion strain This study (UHM298)

PhetR-hetR-gfp Δall1758

PhetR-hetR-gfp-, all1758-deletion strain This study (UHM230)

PpetE-hetR-gfp Δall1758

PpetE-hetR-gfp-, all1758-deletion strain This study (UHM299)

∆patA PpetE-hetR-gfp Δall1758

patA-, PpetE-hetR-gfp-, all1758-deletion strain This study (UHM300)

PpetE-hetR-cfp Δall1758

PpetE-hetR-cfp-, all1758-deletion strain This study (UHM228)

∆patA ∆hetF PpetE-hetR-cfp Δall1758

patA-, hetF-, PpetE-hetR-cfp-, all1758-deletion strain

This study (UHM229)

ΔntcA with Ω cassette

ntcA replaced by Ω cassette

This study (UHM230)




Δalr3758 alr3758-deletion strain This study (UHM302)



89

Table 3.1. (Continued) Strains used in Chapter 3

Anabaena sp. strain




90

Table 3. 2. Plasmids used in Chapter 3



r Neo

r

[47]

pAM505 Shuttle vector pAM504 with inverted multiple cloning site [47]


pAM1956 pAM505 bearing promoterless gfp for transcriptional fusions [48]


pDR120 pAM504 carrying PpetE-hetR [21]



pDR267 pAM505 with PhetF-hetF [22]

pDR329 pAM505 with PpetE-hetFGTG1 [45]

pDR370 pSMC232 with PpetE-patA [49]

pDR350 pAM504 carrying PpetE-lacZ [22]

pDR311 pAM504 carrying Pnir-lacZ [22]

pDR292 pAM504 carrying PhetR-hetR-gfp [46]

pDR293 pAM504 carrying PpetE-hetR-gfp [45]

pKH256 pAM504 bearing PpetE for transcriptional fusions [50]

pKT25 Plasmid carrying the T25 fragment of CyaA for C-terminal protein fusions

[40]

pKNT25 Plasmid carrying the T25 fragment of CyaA for N-terminal protein fusions

[40]

pUT18C Plasmid carrying the T18 fragment of CyaA for C-terminal protein fusions

[40]

pUT18 Plasmid carrying the T18 fragment of CyaA for N-terminal protein fusions

[40]

pLMC115 pAM504 bearing ftsZ-yfp [45]

pLMC117 pAM504 bearing minC-yfp [45]

pLMC119 pAM504 bearing minE-yfp [45]


r

Life Technologies


r [51]



pSMC148 pPROEX-1 carrying hetR [22]

pSMC188 pAM504 bearing Pnir for transcriptional fusions [33]


pSSY125 pAM505 with PpetE-patA [49]

pST112 Suicide plasmid used to delete all1758 This study

pST114 Suicide plasmid used to replace all1758 with an interposon

This study

pSMC202 Suicide plasmid used to delete ntcA [45]

pSMC205 Suicide plasmid used to replace ntcA with an interposon [45]

91



pST141 pAM504 carrying Pall1758-all1758 This study

pST150 pSMC232 with Pall1758-all1758 This study

pST534 pSMC232 with with Pall1758(mutated ntcA binding site)-all1758 This study

pST151 pAM1956 with Pall1758 fused to gfp This study

pST375 pAM505 with Pall1758 This study

pST490 pAM1956 with Pall1758(mutated -189 Ntc-binding site) fused to gfp This study

pST540 pAM1956 with Pall1758(mutated -116 NtcA-binding site) fused to gfp This study

pST541 pAM1956 with Pall1758(mutated -321 NtcA-binding site) fused to gfp This study


pST188 pPROEX-1 carrying all1758 This study

pST156 pKH256 bearing PpetE-all1758 This study

pST367 pKH256 bearing strong ribosomal binding site and PpetE-all1758

This study

pST368 pSMC188 bearing Pnir-all1758 This study

pST399 pAM504 carrying all1758(D267A) This study

pST443 pAM504 carrying all1758(D267E) This study



pST469 pAM504 carrying all1758(T329S) This study

pST401 pAM504 carrying all1758(A348G) This study

pST470 pAM504 carrying all1758(G373A) This study



pST471 pAM504 carrying all1758(G399A) This study





This study

pST448 Suicide plasmid used to delete alr3758 This study

pST450 Suicide plasmid used to replace alr3758 with an interposon

This study



This study

pST520 Suicide plasmid used to delete alr3423 This study

pST521 Suicide plasmid used to replace alr3243 with an interposon

This study


pST530 pAM504 carrying Palr3758-alr3758 This study


pST524 pAM504 carrying Palr3423-alr3423 This study

92



pST460 pDR350 carrying all1087 This study

pST461 pDR350 carrying alr3758 This study




pST537 pDR350 carrying sigE (alr4249) This study

pST437 pKH256 bearing sigE This study

pST456 pPJAV213 bearing sigE This study

pST477 pDR350 carrying sigB (all7615) This study






pST536 pDR311 carrying sigE (alr4249) This study

pST438 pSMC188 bearing sigE This study

pST480 pDR311 carrying sigB (all7615) This study


pST500 pAM1956 with Pall1087(mutated NtcA-binding site) fused to gfp This study

pST528 pAM1956 Palr3758 fused to gfp This study

pST527 pAM1956 with Palr3758(mutated NtcA-binding site) fused to gfp This study

pST488 pAM1956 Palr3423 fused to gfp This study



pST512 pSMC232 with with Pall1087(mutated NtcA-binding site)-all1087 This study

pST513 pSMC232 with Palr3758-alr3758 This study

pST514 pSMC232 with Palr3758(mutated NtcA-binding site)-alr3758 This study

pST515 pSMC232 with Palr3243-alr3243 This study

pST558 pKT25 carrying hetR Chapter 2

pST559 pKT25 carrying hetF Chapter 2

pST560 pKT25 carrying hetZ Chapter 2

pST561 pKT25 carrying patU5 Chapter 2

pST562 pKT25 carrying patU3 Chapter 2

pST563 pKT25 carrying all1758 This study

pST564 pKT25 carrying patA Chapter 2

pST597 pKT25 carrying hetF(C246A) Chapter 2

pST565 pUT18C carrying hetR Chapter 2

pST566 pUT18C carrying hetF Chapter 2

pST567 pUT18C carrying hetZ Chapter 2

93



pST568 pUT18C carrying patU5 Chapter 2

pST569 pUT18C carrying patU3 Chapter 2

pST570 pUT18C carrying all1758 This study

pST571 pUT18C carrying patA Chapter 2

pST598 pUT18C carrying hetF(C246A) Chapter 2

pJP42 pKNT25 carrying divIVA (Bacillus subtilis) [42]

pST572 pKNT25 carrying hetR Chapter 2

pST573 pKNT25 carrying hetF Chapter 2

pST574 pKNT25 carrying hetZ Chapter 2

pST575 pKNT25 carrying patU5 Chapter 2

pST576 pKNT25 carrying patU3 Chapter 2

pST577 pKNT25 carrying all1758 This study

pST578 pKNT25 carrying patA Chapter 2

pST599 pKNT25 carrying hetF(C246A) Chapter 2

pJP41 pUT18 carrying divIVA (Bacillus subtilis) [42]

pST579 pUT18 carrying hetR Chapter 2

pST580 pUT18 carrying hetF Chapter 2

pST581 pUT18 carrying hetZ Chapter 2

pST582 pUT18 carrying patU5 Chapter 2

pST583 pUT18 carrying patU3 Chapter 2

pST584 pUT18 carrying all1758 This study

pST585 pUT18 carrying patA Chapter 2

pST600 pUT18 carrying hetF(C246A) Chapter 2

94


Primer no.


1 1758 up F AGATCTGGTGGTGCGATCGCTTGGAG

2 1758 up R GCCACGACTGTCCCCGGGTGCCTGTTGTTATTATCACG

3 1758 down F ACGGACAACAATAATCCCGGGCGGTGCTGACAG

4 1758 down R GAGGCTCGCAGCCGCCCAGCTGTATCTAC

5 all1758 flank up F GCCCAAGCACGAAATACAG

6 all1758 flank down R CTTCCCGGTAAACTGTTGG

7 all1758 3’ F GCTAGAACCTGGTGATACAG

8 all1757 5’ R ACAGACTGAGAGTTACGTGC

9 Pall 1758 BamHI F TATATGGATCCTAGAAGCTCTCTTGAGTGGC

10 all1758 SacI R ATATAGAGCTCCCCAGTCCCTTTTATACTCG

11 all1758 SmaI R ATATACCCGGGATTCGATCTGTAAGACCAC

12 Pall1758 SacI F TATATGAGCTCTAGAAGCTCTCTTGAGTGGC

13 all1758 15bp up SmaI R ATATACCCGGGAAAGGGGTTAATTTTAGATTGAC

14 Pall1759-SacI-F TATATGAGCTCCCAAAAAAAGGCGCTGAGTG

15 Pall1759-SmaI-R ATATACCCGGGTAGTTGTTACGAGTTATGAG

16 all1758 OE EcoRI F TATATGAATCCATGTGCCTGTGC CTCCATTTTC

17 all1758 OE BamHI R TATATGGATCCCCCAGTCCCTTTTATACTCG

18 all1758-NdeI-F ATATACATATGCCTGTGCCTCCATTTTCCTCTCAAC

19 all1758-NH6-NotIR TTACTGCGGCCGCTTTTCGATCTGTAAGACCACTAAAGTC

20 all1758 EcoRI rbs F TATATGAATTCAGGAGGTGATTGTGCCTGTGCCTCCATTTTC

21 all1758 rbs EcoRI F TATATAAGGAGGAATTCTGTGCCTGTGCCTCCATTTTC

22 asr5349 RT F AACTAAATCTAGTGAGCATGG

23 asr5349 RT R AAGCATATAAAAGGTGATGGT

24 asr5350 RT F ATGGCATTCATCAAGATACAG

25 asr5350 RT R AACGCTTGACTCTTGATATTG

26 all1757 RT F CTGCCGTCAGTACTGTTAC

27 all1757 RT R AGTCCAGGCTTGCTCTTTAG

28 all1759 RT F GGTTCCGTCGTCAAACTAACG

29 all1759 RT R CACGCCAGTTGTTTGTACTG

30 rnpB RT F ATAGTGCCACAGAAAAATACCG

31 rnpB RT R AAGCCGGGTTCTGTTCTCTG

32 alr1087-up-BamHI-F ATATAGGATCCTGTTGTGTACCCTTGTGTTTAAGCAAGTG

33 alr1087-up-R TGATTAAATTCCCCGGGACTAAATTTGCCAATATTATATAGCAACACACACAATTTTAG

95


Primer no.


35 alr1087-down-SacI-R ATATAGAGCTCGGATTGCGACTGCATCAATTAAGTCGAG

36 all1087 flank up F 1 TCATCGACTAAGTAGAACAGG

37 all1087 flank down R 1 GAGATGCAAGCAAGGGTTAG

38 alr3758 up F TATATAGATCTGTTATTTTTCCGGTTTTGCTACGC

39 alr3758 up R AGTACTTGCCCCCGGGCCGATACTTCTATGACTTG

40 alr3758 down F AAGTATCGGCCCGGGGGCAAGTACTGACAGCCAC

41 alr3758 down R ATATAGAGCTCGACTCAGGTTACATCAGCTCAATC

42 alr3758 flank up F CACGTTTTGGCAACGTCTGC

43 alr3758 flank down R TGGATGCTGGAGCTGATGAG

44 all0648 up F TATATAGATCTGCGAAGCAGCCAATTATAGTAATC

45 all0648 up R GGTACACCTTCCCGGGACCGTCTTGGGTTGTATATG

46 all0648 down F CCAAGACGGTCCCGGGAAGGTGTACCTCGTAGCATG

47 all0648 down R ATATAGAGCTCGCATATTTGCTATTGTGCCTCACAC

48 all0648 flank up F CGATTATCAAGATGGACAAGGTG

49 all0648 flank down R TATCCAGCCAAGGTGCAATG

50 alr3423 up F ATAAGATCTCGCATTCAAAAATGGCTACACAC

51 alr3423 up R CATACTTAACCCCGGGGATGGTCGTGCTGCATTATG

52 alr3423 down F CACGACCATCCCCGGGGTTAAGTATGCCAAGGAAGGAC

53 alr3423 down R ATAGAGCTCGTGTAACTTTGCTGCTGCGTAG

54 alr3423 flank up F CTAGTATCTGAAACTCAACGCTGTC

55 alr3423 flank down R CCACTACAGGCAAATAGACC

56 Pall1087 BamHI F TATATGGATCCAGGAAATAAGATTCATCGACTAAGTAGAAC

57 all1087 SacI R ATATAGAGCTCCTAACTATTCGCTAGTATAATTTG

58 Palr3758 BamHI F 1 TATATGGATCCATATTATTTATTGGTCGTTATTTTACAAAC AAAATTGATG TCC

59 alr3758 SacI R ATATAGAGCTCTTAGGAAATAACCCTTGTGGCTG

60 Pall0648 BamHI F TATATGGATCCTTCTGTAAATGTGGGGTTATTGATGTC

61 all0648 SacI R ATATAGAGCTC CTAGCTAGCCATGCTACGAG

62 Palr3423 BamHI F TATATGGA TCCGCTAGTCACCTAGAAGCTTC

63 alr3423 SacI R TATATGAGCTCCTAATATTGTCCTTCCTTGG

64 all1758 D267A-F AACTTGGTCAAGCTTCGGGTCGTTGGGGTTTGGTGATTG

96


Primer no.


65 all1758 D267A-R CAACGACCCGAAGCTTGACCAAGTTTGGTGTATG

66 all1758 D267E-F AACTTGGTCAAGAGTCGGGTCGTTGGGGTTTGGTGATTG

67 all1758 D267E-R CAACGACCCGACTCTTGACCAAGTTTGGTGTATG

68 all1758 D277A-F TGGTGATTGGTGCTGTCATGGGTAAGGGTGTTCCTGC

69 all1758 D277A-R TTACCCATGACAGCACCAATCACCAAACCCCAACG

70 all1758 D277E-F TGGTGATTGGTGAGGTCATGGGTAAGGGTGTTCCTGC

71 all1758 D277E-R TTACCCATGACCTCACCAATCACCAAACCCCAACG

72 all1758 T329S-F CCGCTTCATCTCGCTATTTTATTCTGAATATAATCC

73 all1758 T329S-R GAATAAAATAGCGAGATGAAGCGGTGAGAATTTTCC

74 all1758 A348G-F ATAGTAATGCAGGACATAATCCCCCTTTGTGGTGGC

75 all1758 A348G-R GGGGGATTATGTCCTGCATTACTATAGGACAAAATTC

76 all1758 G373A-F GATGTTAATTGCGCTGGATGCTAATAGCCAATACG

77 all1758 G373A-R CTATTAGCATCCAGCGCAATTAACATCCCCAGAGTATC

78 all1758 D398E-F TTATTATACGGAGGGTTTGACCGATGCGGCTGCG

79 all1758 D398E-R TCGGTCAAACCCTCCGTATAATAAATAACTGTATC

80 all1758 G399A-F TTATACGGATGCTTTGACCGATGCGGCTGCGG

81 all1758 G399A-R CATCGGTCAA AGCATCCGTATAATAAATAACTG

82 all1758 D453A-F GACAAAATAATGCTGATATGACTTTAGTGGTCTTAC

83 all1758 D453A-R AAAGTCATATCAGCATTATTTTGTCTGTCAGCACC

84 all1758 D453E-F GACAAAATAATGAGGATATGACTTTAGTGGTCTTAC

85 all1758 D453E-R AAAGTCATATCCTCATTATTTTGTCTGTCAGCACC

86 all1087 SmaI F ATATACCCGGGAGGAAACAGCTGTGTTGCTATATAATATTGGC

87 all1087 KpnI R TATATGGTACCCTAACTATTC GCTAGTATAA TTTG

88 alr3758 SmaI F ATATACCCGGGAGGAAACAGCTATGAATACAAACTTTCAAGTCATAG

89 alr3758 KpnI R TATATGGTACCTTAGGAAATAACCCTTGTGGCTG

90 all0648 SmaI F ATATACCCGGGAGGAAACAGCTGTGATTCAAA TAGAACAAAATTC

97


Primer no.


91 all0648 KpnI R TATATGGTACCCTAGCTAGCCATGCTACGAG

92 alr3423 SmaI F ATATACCCGGGAGGAAACAGCTATGCTTGGCATAATGCAGCAC

93 alr3423 KpnI R TATATGGTACCCTAATATTGT CCTTCCTTGG

94 all1702 SmaI F ATATACCCGGGAGGAAACAGCTATGAAAAGTGAGCTTCACGTACC

95 all1702 KpnI R TATATGGTACCGCTTTAATTATTA GCTAATTCC

96 sigB SmaI F ATATACCCGGGAGGAAACAGCTATGCCTAATTCAACATCCCAAG

97 sigB KpnI R TATATGGTACCTCAGTCAGCTAGATAGCTGC

98 sigE SmaI F ATATACCCGGGAGGAAACAGCTATGTACCAAACAAAGCATGAATCC

99 sigE KpnI R TATATGGTACCCTCTAATCCCTAATTCCTAGC

100 sigE EcoRI F (PpetE) TATATGAATTCAGGAGGTGATTATGTACCAAACAAAGCATGAATCC

101 sigE EcoRI F (Pnir) TATATAAGGAGGAATTCTATGTACCAAACAAAGCATGAATCC

102 sigE BamHI R ATATAGGATCCCTCTAATCCCTAATTCCTAGC

103 Pall1758 mut NtcA up R GAATTTCCATTAATTGCCTTCAATTTCATGATAATACATGAAGTTTTC

104 Pall1758 mut NtcA down F CATGTATTATCATGAAATTGAAGGCAATTAATGGAAATTCCTCTCC

105 Pall1758 mut NtcA up R -128

CAAGAGTATAAAGACTGCAGACTTAATCAAACATAAGTATTACATTTGC

106 Pall1758 mut NtcA down F -128

GTAATACTTATGTTTGATTAAGTCTGCAGTCTTTATACTCTTGCAAAA GC

107 Pall1758 mut NtcA up R -323 CAACTTCCTTTGTGCGCTATTAATCAGGCATTAGATTGAAATTTTA GC

108 Pall1758 mut NtcA 2 down F -323

CAATCTAATGCCTGATTAATAGCGCACAAAGGAAGTTGTTTAGATG C

109 Pall1087 SacI F TATATGAG CTC AGGAAATAAGATTCATCGAC TAAGTAGAAC

110 Palr1087-dNtcA-OEX-R TTTGCTATCTTGTCAACAAAAGACTAATTTTTTCGTAAATTTATCAAAC

111 Palr1087-dNtcA-OEX-F TGTTGACAAGATAGCAAAAATAATATTGAGGAATTAGGCTATC

112 Pall1087 SmaI R ATATACCCGGGAAGAGGCTTACTTATTGCTTATAGACTC

113 Palr3758 SacI F TATATGAGCTCATATTATTTATTGGTCGTTATTTTACAAACAAAATTGATGTCC

114 Palr3758 mut ntcA R TCTCGAATCATGCTTTTTTCATCAAAATAAAATTTAATTATTGG

115 Palr3758 mut ntcA F AATTTTATTTTGATGAAAAAAGCATGATTCGAGAAATTCTATAG

116 Palr3758 SmaI R 1 ATATACCCGGGTAAATAAATTGTTAGATTACTATCGGTTTG CTATTTAATT

117 Pall0648 SacI F TATATGAGCTCTTCTGTAAATGTGGGGTTAT TGATGTC

98


Primer no.


118 Pall0648 SmaI R ATATACCCGGGCACTGATCTCAAGATTGACTG

119 Pall3423 SacI F TATATGAGCTCGCTAGTCACCTAGAAGCTTC

120 Pall3423 SmaI R ATATACCCGGGTGTAATTATGAGTAAGAATTTTATGTAATCTATCG

121 all1087 SmaI R ATATACCCGGGAACTATTCGCTAGTATAATTTG

122 alr3758 SmaI R ATATACCCGGGAGGAAATA ACCCTTGTGG CTG

123 all0648 SmaI R ATATACCCGGGAGCTAGCCATGCTACGAG

124 alr3423 SmaI R ATATACCCGGGAATATTGTCCTTCCTTGG

125 all1758 BamHI F A AGGAGGGATCCGGGTTCCGCTGGCTCCGCTGCTGGTTCTGGCCCTGTGCCTCCATTTTCCTCTC

126 all1758 EcoRI R A CTCCTGAATTCTTATTCGATCTGTAAGACCACTAAAGTC

127 all1758 BamHI F B AGGAGGGATCCGCCTGTGCCTCCATTTTCCTC

128 all1758 EcoRI R B CTCCTGAATTCCCGCCAGAACCAGCAGCGGAGCCAGCGGAACCTTCGATCTGTAAGACCACTAAAGTC

RESULTS

Mutation of all1758 prevents diazotrophic growth

In an effort to identify genes necessary for diazotrophic growth of Anabaena sp. strain PCC 7120,

a genetic screen using transposon mutagenesis to both increase and mark the sites of mutations

was conducted as previously described [44]. In one mutant the transposon had disrupted open

reading frame all1758 in the annotated genome sequence, suggesting that all1758 was

necessary for diazotrophic growth. To verify a cause-and-effect relationship between inactivation

of all1758 and the phenotype of the mutant, an Ω interposon conferring spectinomycin and

streptomycin resistance was used to replace nucleotides +31 to +1358 relative to the GTG

translational initiation codon of all1758 in the wild-type strain. The resulting mutant, UHM184,

contained the first and last thirty nucleotides of all1758 flanking the Ω interposon and exhibited

impaired diazotrophic growth similar to that of the isolated transposon mutant. Complementation

of UHM184 with a plasmid bearing all1758 preceded by 497 bp upstream of all1758, including the

474 bp intergenic region between the end of upstream gene all1759 and the start of all1758,

restored diazotrophic growth to the mutant and suggested that inactivation of all1758 and not

polar effects of the transposon insertion was the cause of the mutant phenotype. The all1758

gene is located between two genes transcribed in the same orientation and predicted to have a

role in cell division, ftsX (all1757) and ftsY (all1759). To verify that inactivation of all1758 was

solely responsible for the phenotypes of the mutants, nucleotides +31 to +1358 were cleanly

deleted from the chromosome of the wild-type. The resulting strain, UHM183, had an in-frame

99

deletion of all1758 and differed from UHM184 only by the absence of the Ω interposon. The

phenotype of this strain was similar to that of UHM184. Strain UHM183 was complemented in

the same manner as for UHM184, and there was no evidence for recombination between the

plasmid and the chromosomal locus of all1758 as indicated by the lack of a PCR product using

primers corresponding to the region of all1758 that is deleted from the mutant chromosome and

the coding region of all1757. Lastly, transcripts of all1757 and all1759 were detected by RT-PCR

in RNA isolated from strains UHM184 and the wild-type (data not shown). Taken together, these

results indicated that a functional copy of all1758 was required for diazotrophic growth of

Anabaena sp. strain PCC 7120.

Pleiotropic phenotype of all1758 null mutants

Although all1758 was discovered in a genetic screen designed to isolate mutants incapable of

growth on nitrogen-deficient media, disruption of all1758 also affected viability on nitrogen-replete

media. Colonies of the mutant were pale green rather than the vibrant green of the wild-type,

and after about three weeks of growth on BG-11 solid media with either 17 mM nitrate or 2 mM

ammonia as a source of fixed nitrogen, growth appeared to cease, and the strain could not be

revived by subculturing onto fresh medium. In comparison, the wild-type strain remained viable

under the same conditions in excess of three months. However, the strain could apparently be

maintained indefinitely if repeatedly subcultured onto BG-11 solid medium every two weeks.

Despite the severity of the growth defect observed in strain UHM184, suppressor mutations could

not be isolated. To test if loss of viability was the result of a change in the medium over time,

non-viable filaments of UHM184 were scraped from a plate culture 3 weeks after its inoculation,

and the resulting “conditioned” solid medium was streaked with PCC 7120. Growth of the wild-

type strain was indistinguishable from that on unconditioned medium. Cells of UHM184 filaments

cultured on BG-11 solid medium were markedly smaller than those of the wild-type grown under

the same conditions. The average vegetative cell height (measured perpendicular to the

longitudinal axis of the filament) of mutant cells was 2.88 ± 0.25 m, compared to 4.34 ± 0.23 m

for cells of the wild-type, which corresponded to a more than 3-fold reduction in cell volume for

the mutant (Fig. 3.1A, 3.1B). This diminutive phenotype persisted until loss of viability and when

cells were transferred to liquid medium lacking fixed nitrogen.

Growth of the mutant in liquid BG-11 containing nitrate or ammonium as a source of fixed

nitrogen was different from that on solid medium. Within 24 hours of inoculation into liquid

medium, the cytoplasm was found on one side of the cell, and the majority of the cell volume was

occupied by what appears to be a single, large vacuole (Fig. 3.1C). Alternatively, the plasma

membrane may have pulled away from the cell wall in a manner resembling plasmolysis in plant

cells. The process was unlikely to be attributable to the formation of a large gas vesicle because

100

mutant cells settled more readily than those of the wild-type to the bottom of growth vessels.

Vacuole formation was accompanied by enlargement of cells and resulted in cell lysis within 48

hours. To test if cell lysis and/or vacuolization of cells was in response to a substance excreted

into the medium by the mutant, conditioned medium separated from cellular debris by

centrifugation was inoculated with PCC 7120. Growth of PCC 7120 in medium conditioned by

UHM184 was similar to that in unconditioned medium, indicating that lysis of UHM184 was not

caused by amendment of the medium by the mutant so as to make it unfit for growth of PCC

7120. The pH of medium conditioned by growth of UHM184 was similar to that conditioned by

the wild-type and unconditioned medium. In addition, the inverse experiment, growing UHM184 in

spent medium of the strain PCC 7120 did not rescue any of the mutant phenotypes. Finally strain

UHM184 was incapable of growth in the absence of fixed nitrogen, and morphologically distinct

heterocysts were not observed until 48 h after removal of fixed nitrogen, about twice the time for

the wild-type (Fig. 3.1D, 3.1E, 3.1F). Although delayed, the periodic pattern of heterocyts along

filaments was similar to that of the wild-type. Heterocysts of UHM184 had characteristic reduced

levels of auto-fluorescence, indicative of degradation of phycobiliproteins, and polar cyanophycin

granules associated with mature heterocysts. Note that all mutant phenotypes as described

above were complemented by introduction of all1758 on a plasmid to strain UHM184.

101

Figure 3.1. Phenotype of the all1758 deletion strain, UHM184 under varying conditions. The

diminutive cell size of UHM184 (B) as compared to the wild type strain Anabaena sp. strain

PCC 7120 (A) cultured on solid BG-11 medium. At 24 h after culturing in liquid BG-11 medium,

UHM184 develops intracellular vacuoles among all cells of a filament (C). At 24 h after transfer

from solid BG-11 medium to liquid BG-110 medium, which lacks a source of combined nitrogen,

heterocysts form in PCC 7120 (D) but not in UHM184 (E). After 48 h of nitrogen starvation,

UHM 184 exhibits heterocyst patterning among cells of diminutive size (F). Carets indicate

heterocysts. Bars = 10 µM.

102

Strain UHM184 (∆all1758) was Fix

- and lacked the minor heterocyst-specific glycolipid

During the late stages of heterocyst

development, deposition of

heterocyst-specific polysaccharides

contributes to the creation of the

microoxic environment necessary

for the function of nitrogenase, an

oxygen-labile enzyme. Heterocysts

of strain UHM184 were readily

stained by Alcian blue, indicative of

the presence of heterocyst

envelope exopolysaccharides (Fig.

3.2). Thus, polysaccharide

synthesis and deposition appears

unimpaired in the mutant.

Just interior to the layer of

exopolysaccharide, a layer of heterocyst-specific glycolipid also contributes to the microoxic

interior of a heterocyst. To determine if heterocyst-specific glycolipids (HGLs) were present in

mutant strain UHM184, lipids from strains PCC 7120, UHM184, and UHM103, a hetR mutant that

served as a negative control, were extracted after 72 hours of combined nitrogen deprivation, and

separated by thin-layer chromatography for visualization (Fig. 3.3). Two HGLs have been

characterized in the wild-type [52-54]. Although the more abundant, slower-migrating species

corresponding to 1-(0-a-D-glucopyranosyl)-3,25-hexacosanediol was present in strains UHM184

and PCC 7120 (the major HGL), the less abundant, faster migrating glycolipid, 1-(0-a-D-

glycopyranosyl)-3-keto-25-hexacosanol (the minor HGL) was absent from UHM184. As expected,

both HGLs were absent from strain UHM103, which does not make heterocysts. The absence of

the minor HGL, which is necessary for a functional heterocyst, could account for the inability of

UHM184 to grow under diazotrophic conditions. Typically strains affected in HGL synthesis lack

both the minor and major HGL, with one exception. A double kinase mutant was shown by RT-

PCR to lack expression of two glycolipid synthesis genes, asr5349 and asr5350 [23]. Conversely,

transcripts of asr5349 and asr5350 were detected by RT-PCR in RNA isolated from strain

UHM184 (data not shown).

Strains that lack one or both of the heterocyst specific glycolipids cannot fix nitrogen in the

presence of molecular oxygen (Fox- phenotype) but can in its absence. Other strains that are

Figure 3.2. Heterocyst-specific exopolysaccharide of

UHM184. Bright-field images of heterocyst-specific

exopolysaccharides stained by Alcian Blue in (A) PCC

7120 and (B) UHM184 at 48 h after removal of combined

nitrogen. Carets indicate heterocysts. Bars = 10 µM.

103

deficient in more than just the creation of microoxic heterocysts cannot fix under either condition

(Fix- phenotype). The nitrogenase activity of strain UHM184 at 48 h post-induction was assessed

Figure 3.3. Heterocyst-specific glycolipids of UHM184. The numbers 1, 2, 3 (top

horizontal header) denote lane numbers. Lane 1, lipid extract from the all1758 mutant,

UHM184, which lacks the minor heterocyst glycolipid (HGL); lane 2, PCC 7120; and lane

3, the ∆hetR mutant, UHM103, which lacks both the minor and major HGLs separated by

thin-layer chromatography 72 h after removal of combined nitrogen. The minor (1-(0-a-D-

glycopyranosyl)-3-keto-25-hexacosanol), and the major (1-(0-a-D-glucopyranosyl)-3,25-

hexacosanediol) heterocyst-specific glycolipids are indicated by the top and bottom

arrows, respectively. Lipid extracts were deposited at the origin, denoted by ‘*’.

104

by acetylene reduction assays under both oxic and anoxic conditions (Fig. 3.4) to determine if it

was a Fox- or Fix

- strain. Under oxic conditions, rates of acetylene reduction were 0.350 ± 0.064

nmol ethylene h-1

ml -1

(OD750 unit)-1

for PCC 7120 and undetectable for strains UHM184, UHM103

(∆hetR), a Fix- strain and UHM128 (∆pbp6), a previously characterized Fox

- strain [44, 55].

Under anoxic conditions rates of acetylene reduction were 0.073 ± 0.038 nmol ethylene h-1

ml-

1(OD750 unit)

-1 for strain UHM128 and undetectable for strains UHM184 and UHM103. Thus, the

phenotype of UHM184 was Fix-, suggesting that lack of a functional copy of all1758 disrupted

more than just the creation of microoxic conditions in heterocysts. For this reason, All1758 may

have a role in heterocyst development that extends beyond nitrogen fixation.

all1758 is not required for the timing of pattern formation

A null mutation in all1758 delays morphological differentiation of heterocysts. Prior to changes in

the morphology of differentiating cells, the cells of filaments that will differentiate are specified by

the generation of a pattern of gene expression. This pattern can be visualized using fusions of

the promoters of hetR or patS to gfp [48, 56]. Strain UHM184 carrying a plasmid with the

promoter of either hetR or patS fused to gfp exhibited a pattern of GFP-dependent fluorescence

Figure 3.4. Acetylene reduction assay for nitrogenase activity indicates strain UHM184

(∆all1758) is Fix-. The heterocysts that form in UHM184 are non-functional under oxic (left

panel) and anoxic (right panel) conditions. (+) and (-) indicates the positive and negative

controls for the assay, respectively.

105

that was similar to that observed in the wild-type carrying the same plasmid (Fig. 3.5C, 3.5D).

Both the number of cells between fluorescing cells and the timing of the appearance of

fluorescence were similar. The only noticeable difference was the smaller size of the cells in

filaments of UHM184. Thus, all1758 was not required for formation of the pattern that specifies

which cells will differentiate into heterocysts.

all1758 is constitutively expressed and autoregulates its own expression

Expression of all1758 was

assessed both temporally and

spatially using the gene for

green fluorescence protein

(GFP) as a reporter. A

uniform level of florescence

was observed in vegetative

cells and heterocysts when a

shuttle vector carrying the

putative promoter region of

all1758 fused to gfp (plasmid

pST151) was introduced into

the wild-type strain. Similar

levels of expression were seen

with and without nitrate in the

medium (Fig. 3.5A). However,

no GFP-dependent

fluorescence was observed

when the same plasmid was

introduced into strain UHM184,

which has all1758 deleted

from the chromosome (Fig.

3.5B).

No GFP-dependent

fluorescence was observed

from the wild-type strain

harboring a plasmid bearing

all1758 translationally fused to gfp under the control of its native promoter in the presence or

absence of nitrate (data not shown). UHM184 bearing the same plasmid also remained dim,

Figure 3.5. A functional all1758 gene is required for

expression of all1758 but not for patterned expression of

patS. Expression of all1758 as visualized with a

transcriptional Pall1758-gfp fusion on plasmid pST151 in strains

(A) PCC 7120 or (B) UHM184 at 24 h after removal of

combined nitrogen. Expression of patS in PCC 7120 (C) and

UHM184 (D) as visualized using the transcriptional fusion

patS-gfp on plasmid pAM1951 at 14 h after induction of

heterocyst formation. Micrographs show bright-field (left

panels) and fluorescence (right panels) images recorded

using identical microscope and camera settings. Carets

indicate heterocysts. Bars = 10 µM.

106

similar to the wild-type, although the cell size had returned to the wild-type dimensions,

apparently due to complementation by the fusion protein encoded on the plasmid. In addition,

overexpression of all1758 from the copper-inducible petE promoter in the wild-type strain had no

apparent effect on heterocyst differentiation or cell size, whereas extra copies of patS or hetN

(and patU3) expressed from the petE promoter on a plasmid in strain UHM184 prevented

heterocyst differentiation, but the diminutive cell phenotype remained (data not shown). Also,

patU3-deficient strains exhibit the diminutive cell morphology common to all1758-deficient strains

(Chapter 2). Overexpression of patU3 alone from the petE promoter on a plasmid was sufficient

to inhibit heterocysts in strain UHM184 (data not shown), but not the wild-type nor patA-deficient

strain. PatU3-GFP was occasionally detected in a few vegetative cells of strain UHM184 (data

not shown) at 30 hours post-induction, and not in the wild-type, perhaps due to spontaneous

mutations that arose under the conditions of this experiment. Overexpression of all1758 from the

nirA promoter, which is induced under both nitrate-replete conditions or conditions of nitrogen-

starvation [57], yielded the same results in the wild-type as those observed with use of the petE

promoter (data not shown).

NtcA and HetR negatively regulates all1758 expression

To understand the role of the heterocyst differentiation genes ntcA and hetR in the regulation of

the expression of all1758, a reporter plasmid carrying Pall1758-gfp was introduced into both mutant

backgrounds. Fluorescence in a ∆ntcA strain was increased in comparison to the wild-type (data

not shown) suggesting that the global transcription factor NtcA exhibits an inhibitory role on

all1758.

Figure 3.6. The global transcription factor NtcA negatively regulates all1758 expression. (A)

Expression of all1758 as visualized with a transcriptional Pall1758-gfp fusion on plasmid pST151

in strain UHM184. (B) Expression of all1758 as visualized with a transcriptional Pall1758-gfp

fusion with a mutated putative -189 binding site (on plasmid pST490) in strain UHM184.

Micrographs show bright-field (left panels) and fluorescence (right panels) images recorded

using identical microscope and camera settings at 24 h after removal of combined nitrogen.

107

To further support the regulatory role of NtcA in all1758 expression, three putative NtcA-binding

sites present in the promoter of all1758 were examined. Plasmids with mutations in the

consensus NtcA-binding site GTA-N8-TAC [58]starting at positions -189, -116 and -321 relative to

the putative translational start site of all1758 were individually introduced into all1758-deficient

strains. Whereas fluorescence was below the limit of detection when a plasmid bearing Pall1758-

gfp was examined in strain UHM184, fluorescence was restored to wild-type levels upon mutation

of the putative NtcA-binding site at position -189 in UHM184 (Fig. 3.6). This result supports a

negative role for NtcA in all1758 expression. However, only the putative -189 binding site

appears to be important for NtcA-directed regulation of all1758, because mutations in the

potential binding sites at positions -116 and -321 did not restore fluorescence from the Pall1758-gfp

construct in strain UHM184 (data not shown). Fluorescence was not detected upon introduction

of plasmids carrying mutations at -189, -116 and -321 potential binding sites into ntcA-deficient

strains (data not shown).

Like NtcA, HetR also has a negative role in all1758 expression. The Pall1758-gfp reporter plasmid

(without mutations in NtcA-binding sites) was lethal in a hetR background. No colonies arose

upon conjugation. However, the same region corresponding to Pall1758 carried on a plasmid could

be introduced into a hetR-deficient strain. Since the two plasmids differed only by fusion to the

gfp reporter gene, the level of gfp expression, rather than the promoter region, may account for

the lethality observed. To test this hypothesis, the Pall1758-gfp reporter plasmid was introduced

into a very small subset of cells along a filament of ∆hetR to create a genetic mosaic filament [22].

The mosaic filaments were examined at 24 hours and 48 hours after the conjugation event, to

limit the number of cell divisions. At 0 hours after nitrogen step-down, the mosaic filaments were

Figure 3.7. The heterocyst differentiation gene HetR negatively regulates all1758

expression. (A) Expression of all1758 as visualized with a transcriptional Pall1758-gfp fusion

carried on plasmid pST151 in a hetR-deficient strain. (B) The empty vector (plasmid

pAM1956) negative control in the same genetic background. Micrographs of genetic

mosaic filaments show bright-field (left panels) and fluorescence (right panels) images

recorded using identical microscope and camera settings at 48 h after removal of

combined nitrogen (48 hours after the conjugation event).

108

bright, and fluorescence intensified significantly over time (Fig. 3.7). These findings suggest that

in the absence of hetR, expression of all1758 increased dramatically. These findings imply that

NtcA as well as HetR negatively regulate all1758.

The regulatory role of the three NtcA binding sites was also explored in a ∆hetR background.

Plasmids carrying the mutated -116 and -321 binding sites did not result in detectable

fluorescence under the same conditions (data not shown). These sites appear to contribute to

the elevated levels of fluorescence observed in a hetR-deficient strain. However, expression of

all1758 with the mutated -189 binding site was observed without the need for mosaic filaments in

∆hetR (data not shown), unlike the case for plasmids bearing the non-mutated (normal) promoter.

It remains unclear why the -189 binding site appeared to lower expression in a ∆hetR background

but increase expression in a ∆all1758 background. But taken together, All1758 appears to have

a role in heterocyst development since two key developmental genes, NtcA and HetR, are

involved in the regulation of All1758.

Bypass of mutation of all1758 by extra copies of hetR

In an attempt to ascertain where in the regulatory network controlling heterocyst differentiation

all1758 may have been acting, plasmid-borne copies of hetR were put into UHM184 and the

resulting strain was examined for phenotypic differences. The strain with extra copies of hetR

had multiple contiguous heterocysts and reduced spacing between groups of heterocysts in

media with or without nitrate as a fixed nitrogen source (data not shown). The same Mch

phenotype was observed when the plasmid was introduced into the wild-type strain. Extra copies

of hetR bypassed the delay in morphological differentiation of heterocysts, but the diminutive cell

phenotype and other associated phenotypes remained.

Genetic epistasis analysis supports the role of all1758 in normal cell size and cell growth

control

Genes controlling heterocyst differentiation also have been shown to affect cell morphology.

Overexpression of hetF results in supernumerary heterocyst formation in the wild-type, but cells

are reduced in size. Heterocysts no longer form in a hetF-deficient strain, which also exhibits

aberrantly enlarged cells[22]. Similarly, overexpression in the wild-type of another positive

regulator of differentiation, patA, results in enlarged cell morphology[49].

Strain UHM184 overexpressing either hetF or patA was examined individually to determine if the

diminutive cell morphology of all1758 could be controlled by hetF and/or patA. UHM184 filaments

overexpressing hetF appeared smaller than the same strain without hetF overexpressed,

consistent with overexpression results in the wild-type. However, unlike the effect of patA in the

wild-type, no change in morphology was observed when patA was overexpressed in UHM184.

109

The all1758 gene appears to be necessary for the effect of overexpression of patA on cell

morphology.

In strain UHM184, GFP fluorescence from translational fusions to either hetF or patA was below

the level of detection, similar to levels of fluorescence in the wild-type without the use of confocal

microscopy[22, 49]. HetF-GFP was occasionally detected in 1-8 cells of strain UHM184 (data not

shown) at 24 hours post-induction, most likely due to spontaneous mutations that arose under the

conditions of this experiment.

The double-deletion mutants hetF all1758 and patA all1758 were created by deleting all1758 from

hetF or patA parent strains, respectively, to determine which gene is epistatic. In these strains

(and the six other all1758 double-deletion or replacement strains listed in Table 3.1), the

diminutive cell morphology and cell growth defect was observed, but the parental strain timing

and patterning of heterocyst development was unaffected (Fig. 3.8). Because the small cell and

growth defect phenotype is characteristic of ∆all1758 rather than the parent strain, all1758 is

epistatic with respect to cell size and growth. The results from epistasis analysis suggest that

all1758 is required for normal cell size and cell growth.

Figure 3.8. All1758 is required for normal cell size and growth in Anabaena. Parent

genetic background strains (left panels) deficient in (A) hetF, (C) patA, and (E) hetR.

The double-deletion mutants (B) ∆hetF ∆all1758, (D) ∆patA ∆all1758, and (F) ∆hetR

∆all1758, created by an additional in-frame deletion in all1758 (right panels) of parental

strains depicted in A, C, and E, respectively, exhibits the diminutive cell size and growth

defect common to UHM184. Micrographs were recorded using identical microscope and

camera settings 48 hours after nitrogen removal. Carets indicate heterocysts.

110

All1758 is required for proper localization of the cell division protein FtsZ and the cell

division regulators MinC and MinE

The diminutive cell morphology of strain UHM184 may relate to a cell division defect, particularly

because the all1758 gene is located in the same orientation between ftsX (all1757) and ftsY

(all1759), two predicted cell division genes. A similar genomic arrangement is observed in the

genome of Escherichia coli. In this organism, the cell division gene ftsE (b3463) lies between

ftsX (b3462) and ftsY (b3464). Another relationship involves SpoIIE phosphatase of Bacillus

subtilis. Similar in sequence to All1758, SpoIIE contributes to asymmetric cell division in Bacillus

subtilis by interacting with FtsZ[59].

The GTPase FtsZ initiates cell division in bacteria. It provides a scaffold, termed the ‘Z-ring’, at

the midpoint of a cell. Other division proteins, including FtsX and FtsY, are recruited to the

cytoskeletal scaffold, forming a division septum where cytokinesis eventually occurs[60]. In E.

coli, the localization of FtsZ at the cell midpoint is regulated by the min genes. MinC and MinD

oscillate between cell poles to inhibit FtsZ localization at the poles[61-64]. MinE directs the MinC-

MinD inhibitor complex away from the cell midpoint [65]. Translational fusions to YFP (yellow

fluorescent protein) were used to examine the relationship between the all1758-dependent

control of cell morphology and the cell division machinery. Plasmids containing ftsZ-yfp, minC-yfp

and minE-yfp driven by the PpetE promoter were introduced individually into strain UHM184 and

the wild-type and subsequently examined by confocal microscopy. Fusion of the cell division

genes to the inducible PpetE promoter rather than the native promoters removed any effects on

transcription from this study.

Deletion of the PP2C phosphatase spoIIE has been correlated with a significant reduction of FtsZ

rings in Bacillus subtilis cells [66]. Similarly, fluorescence corresponding to FtsZ-YFP was very

dim in strain UHM184 compared to the wild-type (Fig. 3.9A, 3.9D). Also, fewer cells in strain

UHM184 exhibited FtsZ rings. In a population of 500 cells, concentrated localization of FtsZ-YFP

to a mid-cell ring was observed for 49.0% of cells in the wild-type compared to 13.6% cells in

strain UHM184.

One simple explanation for decreased ftsZ-yfp in strain UHM184 would be increased levels of

MinC and/or MinE, the negative regulators of FtsZ. MinC-YFP is typically concentrated at the

poles of cells in the wild-type (Fig. 3.9B, 3.9E). However, fluorescence corresponding to minC

was barely detected in strain UHM184 (and when present, at a concentration that did not “cap”

the poles as fully as in the wild-type; Fig. 3.9C, 3.9F). Fluorescence corresponding to minE was

very dim in the wild-type, but not detected in strain UHM184 (data not shown). These results

suggest that All1758 is involved in the proper localization of the cell division machinery

components, FtsZ, MinC and MinE.

111

Figure 3.9. All1758 is involved in the proper localization of the cell division machinery

components FtsZ, MinC and MinE. (A-C) PCC 7120 and (D-E) UHM 184 with

plasmids carrying (A, D) PpetE-ftsZ-yfp, (B, E) PpetE-minC-yfp and (C, F) PpetE-minE-yfp.

Confocal micrographs show bright-field (left panels) and YFP fluorescence (right

panels) images recorded using identical microscope and camera settings at 48 hours


112

Conserved aspartate residues in the active sites of Motifs 2, 8 and 11 are required for

PP2C phosphatase activity of All1758

The predicted 463-

amino acid sequence

encoded by all1758

has similarities to

sequences of other

proteins in the Protein

Database maintained

by NCBI (Fig. 3.10).

Residues 57 to 191 appear to comprise a GAF domain, which is named after the types of proteins

containing the domain (cGMP-specific phosphodiesterases, adenylyl cyclases and FhlA). Many

of these proteins bind cyclic nucleotides. Residues 84 to 463 were similar in sequence to

serine/threonine protein phosphatases of the PP2C superfamily, which is comprised of metallo-

phosphatases that employ two metal ions in the catalytic site. PP2C proteins generally have a

catalytic domain of approximately 290 amino acids that includes eight “absolutely conserved”

residues within eleven conserved motifs [23, 67]. However, the predicted protein sequence of

All1758 lacked the invariant glycine present in Motif 5 and, instead, had an alanine in its place.

In an attempt to demonstrate phosphatase activity of All1758 using a biochemical approach, the

52 kD protein with an N-terminal polyhistidine epitope tag was purified from E. coli strain BL21 by

affinity chromatography. Attempts to characterize phosphatase activity using p-nitrophenol

(pNPP) as a substrate were unsuccessful even after addition of either 10 or 100 µM cAMP or

cGMP separately with the pNPP substrate, in an effort to support the potential role of the GAF

domain in phosphatase function.

A genetic approach was subsequently utilized to demonstrate the PP2C phosphatase function of

All1758. Conservative and non-conservative substitutions of the eight absolutely conserved

residues of PP2C phosphatases were cloned into plasmids pST399, pST443, pST400, pST444,

pST469, pST401, pST470, pST402, pST445, pST471, pST403 and pST446 corresponding to

all1758(D267A), all1758(D267E), all1758(D277A), all1758(D277E), all1758(T329S),

all1758(A348G), all1758(G373A) all1758(D398A), all1758(D398E), all1758(G399A),

all1758(D453A) and all1758(D453E). These alleles of all1758, representing mutations in

conserved motifs essential for PP2C phosphatase activity, were introduced into strain UHM184

and the wild-type. Lack of complementation of strain UHM184 to the wild-type phenotype was

observed for the D277 (pST400, pST444), D398 (pST402, pST445), and D453 (pST403, pST446)

substitutions corresponding to Motifs 2, 8, and 11. Because the aspartates in the active sites of

Figure 3.10. Schematic of the domains present in the 463-residue

All1758 protein.

113

Motifs 2, 8, and 11 are required for restoration of All1758 function, All1758 appears to act as a

PP2C phosphatase.

All1758 regulates the antisigma factor Alr3423

The sigma factor subunit of prokaryotic RNA polymerases is required for promoter sequence

recognition and transcription initiation. Partner-switch mechanisms that modulate the activity of

sigma factors are dependent upon phosphorylation activity. In particular, the phosphatase activity

of the PP2C phosphatases SpoIIE, RsbU, RsbP of Bacills subtilis [31, 32, 68] and RsbU of

Staphylococcus aureus [69] directly control sigma factors involved in sporulation or stress

response. The dual-function protein SpoIIE controls gene expression and morphogenesis.

SpoIIE promotes polar division by interacting with the cytokinetic machinery and directly activates

σE

(SigE) during sporulation[70]. Similarly, induction of the stress response by σB

(SigB) is

positively regulated by RsbU in Bacillus subtilis and Staphyloccus aureus [31, 69]. All1758

displays sequence homology to SpoIIE and RsbU (Fig. 3.10). For this reason, the possible role

of the PP2C phosphatase All1758 in a similar partner-switch mechanism was investigated.

Interactions of all1758 were strongly predicted in silico (STRING 9.1 database at http://string-

db.org) to antisigma factors (all1087, alr3758, all0648), and anti-antisigma factors (alr3423,

all1702).

In addition to sigma factor regulators, sigma factors (SigB and SigE) were investigated in relation

to All1758 The alternative sigma factor SigB is the master regulator of a large stress response

regulon in Bacillus subtilis and Staphylococcus aureus [71, 72]. Alr3423 and All1702 were

predicted to regulate SigB. Of the twelve sigma factor genes annotated in the Anabaena genome,

sigE appears to control the expression of genes involved in late-stage heterocyst

development[73]. The absence of the minor heterocyst glycolipid in ∆all1758, controlled by

heterocyst-development genes, may relate to sigE. The overlapping promoter specificities of

sigma factors in Anabaena have made it difficult to separate the role of sigma factors in gene

expression[74], but in this study, an All1758-dependent sigma factor controlling a developmental

regulon is proposed.

Individual deletions of sigma factor regulator genes (all1087, alr3758, all0648, alr3423) from the

chromosome were constructed in an effort to address the role of each gene. The resultant strains

(UHM325, UHM301, UHM302, UHM303, UHM326, UHM327) did not differ from in phenotype

from the wild-type with respect to cell morphology, cell growth, or heterocyst development.

Overexpression studies were done to address the role of sigma factors, antisigma factors and

antiantisigma factors in relationship to the phosphatase function of All1758. The phenotype of

114

each regulatory gene, fused to two different inducible promoters (PpetE and Pnir, Table 3.2), was

investigated upon introduction into strain UHM184 and the wild-type.

Overexpression of the sigma factor genes sigE and sigB did not result in any change to the

parent phenotypes tested. However this result may not be unexpected given reports that sigma

factor concentrations in E. coli generally exceed the number of RNA polymerase molecules [75].

In fact, gene expression is influenced by this competition between sigma factors for RNA

polymerase.

When sigma factor regulators (anti-antisigma factors and antisigma factors) were present in

multicopy under the control of native promoters, no phenotypic effect was observed in either

background. Overexpression of the anti-antisigma factors all1087, alr3758 and all0648 also did

not affect the phenotype of strain UHM184. However, in the wild-type, overexpression of anti-

antisigma factors disrupted cell growth to a greater extent than normally observed in strain

UHM184, but heterocyst formation and patterning was unimpaired (data not shown). This may

very weakly correlate the growth

defect of UHM184 to increased

anti-antisigma factor concentrations

in the absence of the phosphatase

activity of all1758.

The effect of overexpression of

antisigma factors (all1702, alr3423)

was investigated in the wild-type

and strain UHM184. The antisigma

factor all1702 when overexpressed

(carried on plasmid pST479) did

not affect the phenotype of the

wild-type or strain UHM184.

However overexpression of the

antisigma factor alr3423 (carried on

plasmid pST478) in the wild-type

mimics the phenotype of strain

UHM184. While no change was

observed in strain UHM184 (Fig.

3.11A), overexpression of alr3423

in the wild-type induced phenotypic

changes that resemble strain

Figure 3.11. Overexpression of the antisigma factor

Alr3423 in the wild type strain mimics the ∆all1758

phenotype at 24 hours. Plasmid bearing Pnir-alr3423 in

(A) UHM184, (B) wild type. The empty vector control in

the wild-type is shown in (C). Micrographs were recorded

using identical microscope and camera settings 24 hours


Bars = 10 µM.

115

UHM184. Like strain UHM184 at 24 hours post-induction, the cell growth defect, reduced cell

size, and deficiency in heterocyst development was observed (Fig. 3.11B). Heterocysts did not

form in this strain even after one week in nitrate-depleted conditions; however this may be due to

the severe growth defect observed before nitrogen step-down. Regulation of Alr3423 by All1758

could potentially account for the observed results.

NtcA activates expression of the all1758-dependent anti-antisigma factor all1087

Overexpression of anti-antisigma and antisigma factors in all1758-deficient strains did not alter

the growth defect phenotype normally associated with these strains. However, sigma factor

regulators impaired growth upon overexpression in the wild-type. The transcriptional control of

the sigma factor regulators was investigated to determine the relationship to All1758. Plasmids

containing transcriptional gfp fusions to antisigma factors (alr3423, all1702) and anti-antisigma

factors (all1087, alr3758, all0648) were introduced into strain UHM184 and the wild-type.

Additionally, the regulatory role of the global transcriptional regulator NtcA was also examined by

mutation of putative NtcA-binding sites found in the promoters of all1087 and alr3758.

In both strain UHM184 and the wild-type, GFP fluorescence was below the level of detection for

all antisigma factors and anti-antisigma factors tested, with the exception of the anti-antisigma

factor all1087. Expression of All1087 (carried on plasmid pST499) was heterocyst-specific in

both the wild-type and in strain UHM184 (Fig. 3.12A, 3.12B). Fluorescence was concentrated in

developing heterocysts and heterocysts by 24 hours of nitrogen step-down. In addition, NtcA was

required for this localized expression of All1087. Mutation of the putative NtcA-binding site in

Pall1087 on a plasmid (pST500) resulted in non-detectable levels of fluorescence in both strains

(Fig. 3.12C, 3.12D). This result implies that the heterocyst-specific expression of all1087 is NtcA-

dependent. To examine the spatial and temporal localization of sigma factor regulators within

filaments of strain UHM184 and the wild-type, translational fusions driven by the native promoters

for each regulator were examined. GFP fluorescence was below the level of detection for all

antisigma factors and anti-antisigma factors tested (data not shown), with the exception of the

anti-antisigma factors all1087 and all0648.Fluorescence from the translational fusion for all1087

to gfp was consistent with results from the transcriptional gfp reporter. Cells carrying All1087-

GFP were dimly fluorescent at 0 hour post-induction, but fluorescence eventually concentrated in

proheterocysts and heterocysts by 24 hours.

Localization of All0648 differed from that of All1087. Vegetative cells and proheterocysts in strain

UHM184 and the wild-type carrying All0648-GFP were fluorescent at 0 hour post-induction, but

the fluorescence dimmed in developed heterocysts at 24 hours. The dissimilar temporal and

spatial localization of All0648 and All1087 may relate to their respective roles in Anabaena.

116

Direct interaction of All1758 with itself and HetR detected by the bacterial two-hybrid

system

Dimerization of SpoIIE, a PP2C protein phosphatase with homology to All1758, has been shown

through a yeast two-hybrid system [76]. To determine if All1758 also interacts with itself and

other developmental proteins (HetR, HetF, HetF(C246A), HetZ, PatU5, PatU3, PatA), a bacterial

two-hybrid assay was performed (refer to Chapter 2 for more details). Putative protein-protein

interactions were tested by fusion of these proteins at the N- or C- termini of two complementary

fragments derived from the Bordetella pertussis adenylate cyclase catalytic domain, T25 and T18.

In this genetic assay, physical association of the two tested proteins result in the functional

complementation of adenylate cyclase in an Escherichia coli cya strain[40]. Subsequent

synthesis of cAMP by the reconstituted adenylate cyclase yields a distinctive blue cell phenotype

that serves to signal the presence of a direct interaction between the two proteins under

Figure 3.12. NtcA is required for the localized expression of the anti-antisigma factor All1087

in developing heterocysts. A plasmid bearing Pall1087-gfp in the (A) wild type, (B) UHM184. A

plasmid bearing Pall1087-gfp with a mutated putative NtcA binding site in the (C) wild type, (D)

UHM184. Bright-field (left panels) and fluorescent (right panels) micrographs were recorded

using identical microscope and camera settings 48 hours after nitrogen removal. Carets

indicate heterocysts. Bars = 10 µM.

117

investigation. The Bacillus subtilus cell division protein DivIVA, shown in different reports to

interact with itself[41, 42], was utilized as a positive control for this study.

Of the 60 interactions tested, four interactions (6%) were discovered (Fig. 3.14, Table 3.4). The

dimerization between All1758 and itself and a less robust interaction with HetR was detected.

Both interactions were not as strong as interactions between HetR-HetR or Bacillus subtilis

DivIVA-DivIVA but significantly greater than the negative (empty vector) controls. The E. coli

colonies that represent positive interactions from the assay pairing derivatives of plasmids

pKNT25 and pUT18 (Table 3.2) are depicted in Figure 3.13. Figure 3.13 summarizes

quantification of the β-galactosidase activity. The three positive interactions between All1758 and

All1758 are shown, representing three of the four possible pairings between N- and C-terminal

fusions to T25 and T18. Interaction of All1758 with the developmental protein HetR further

supports the role of the PP2C phosphatase All1758 in heterocyst differentiation.

Figure 3.13. Bacterial two-hybrid results indicate All758 interacts with (A) HetR and (B)

dimerizes with itself. Colonies of E. coli cells expressing the indicated protein fusions to

fragments of adenylate cyclase carried on derivatives of plasmids pKNT25 and pUT18. Plus

signs (lower right corner) indicate a positive interaction. Positive interactions for the assay

include dimerization of (D) Anabaena HetR and HetR and (E) Bacillus subtilis DivIVA and

DivIVA. The negative control for the assay (empty vector interactions) is depicted in (C).

118

Table 3.4. Plasmids used for quantification of β-galactosidase activity (Figure 3.14) of positive

bacterial two-hybrid interactions, and average Miller units (with standard deviation) of three

independent replicates. Rows (top to bottom) correspond to ascending order (left to right) of

activity depicted in Figure 3.14. The first four bars (left side of bar graph) relating to empty vector

negative controls in Figure 3.14 are listed at the top of the Table. The positive control for the

assay (Bacillus subtilis DivIVA/DivIVA interaction) is identified in bold font.

Protein-protein interaction (T25 fragment/T18 fragment)



Average Miller Units (MU)

Standard deviation

None (negative control) pKNT25 pUT18C 56 12.6

None (negative control) pKNT25 pUT18 57 4.2

None (negative control) pKT25 pUT18 61 9.2

None (negative control) pKT25 pUT18C 56 12.6

All1758/HetR pST577 pST579 133 24.0

All1758/All1758 pST577 pST570 262 38.5

All1758/All1758 pST577 pST584 291 36.2

All1758/All1758 pST563 pST570 466 14.9

DivIVA/DivIVA (positive control)

pJP42 pJP41 1579 109.6

HetR/HetR pST572 pST579 1565 94.9

Figure 3.14. Average β-galactosidase activity of positive protein-protein interactions between

All1758 and HetR and itself. Negative controls for the assay (empty vector interactions)

shown on the far left (first four bars) and positive controls shown on the far right (last two

bars; dimerization of Bacillus subtilis DivIVA and DivIVA and Anabaena HetR and HetR).

Average activity of three replicates plotted in ascending order from left to right. Plasmid

identity for tested interactions detailed in Table 3.4 in the same order (top to bottom of Table).

119

DISCUSSION

Phosphorylation of proteins is widely used in the post-translational regulation of various cellular

processes. All1758 is predicted to be a PP2C phosphatase, which is a member of the PPM

family of serine/threonine phosphatases that are dependent on the presence of the divalent

cation Mg2+

for catalytic function. However, repeated attempts to support the role of All1758 as a

phosphatase were unsuccessful. The simplest explanation for the lack of discernable

phosphatase activity in a biochemical assay relates to the absence of one of the eight absolutely

conserved residues in the protein sequence of all1758. PP2C proteins generally have a catalytic

domain of approximately 290 amino acids that includes eight absolutely conserved residues

within eleven conserved motifs [23, 67]. All1758 harbors an alanine residue, a conservative

substitution, in lieu of the glycine residue reported as invariant in Motif 5. Alternatively,

phosphatase activity of the protein encoded by all1758 may not be measurable under the

conditions of our assay due, for instance, to substrate specificity, lack of a cofactor, or improper

folding of the recombinant protein. Site-directed mutagenesis was used instead to support the

function of All1758 as a phosphatase. Conservative and non-conservative mutations were made

at eight residues important for PP2C function, and tested for complementation of the pleiotropic

phenotype of ∆all1758. Mutations at three of the four ‘invariant’ aspartate mutations failed to

complement the all1758 mutant strain, demonstrating that these residues are essential for

All1758 function as a PP2C phosphatase. Furthermore, like the PP2C phosphatase SpoIIE [76],

All1758 was shown to dimerize with itself in a bacterial two-hybrid assay.

The minor heterocyst glycolipid, which is essential for provision of microaerobic conditions

necessary for the activity of nitrogenase within the heterocyst, was absent in a strain lacking

all1758. However, the absence of nitrogen fixation in the mutant strain even under anoxic

conditions suggests that all1758 is required for some cellular process(es) related to fixation of

nitrogen not related to diminution of molecular oxygen in heterocysts. Indeed, a diverse and

seemingly unrelated range of phenotypes resulted from mutation of all1758. For instance, the

delay in heterocyst formation in UHM184 may suggest the involvement of all1758 in the timing of

heterocyst morphology development. The diminutive cell size and associated decrease in cell

volume apparent upon mutation of all1758 may be indicative of a change in the timing of cell

division in the mutant, which would prevent the cell from attaining a normal and/or enlarged cell

size. Inhibition of cell division has been shown to prevent differentiation of heterocysts, and

heterocysts have twice the DNA content of vegetative cells [77, 78]. The increased rate of cell

division in strains lacking all1758 may account for the defect in heterocyst differentiation in the

mutant. The location of all1758 between the cell division genes ftsX and ftsY may implicate

all1758 as a cell division gene as well. In Escherichia coli, for example, the cell division gene ftsE

(b3463) lies between ftsX (b3462) and ftsY (b3464). Or All1758 may have a role in activating the

120

cell division machinery. Levels of YFP corresponding to FtsZ, and the inhibitors of FtsZ (the

proteins MinC and MinD) were very low in all1758-deficient strains, despite the contrasting roles

of these proteins. Additionally, in Bacillus subtilis, localization of SpoIIE at the midcell (referred to

as the ‘E ring’; the E ring colocalizes with the Z-ring) is controlled by FtsZ[59, 79]. While this

phenomenon was not tested in Anabaena, it may be fruitful to determine if FtsZ regulates All1758,

and explore the direct interaction between these proteins by bacterial two-hybrid assays.

Alternatively, a defect in a process specific to heterocyst differentiation may influence cell division.

Also, the decrease in cell volume may have perturbed the concentration of specific intracellular

components, resulting in the observed phenotypes. This explanation has been used to describe

the process of sporulation in Bacillus. Despite localization of SpoIIE to both faces of the

sporulation septum, the smaller volume of the forespore leads to a higher concentration of

phosphatase activity in this structure[80]. The relationship between cell division and heterocyst

differentiation remains enigmatic.

Unpatterned expression of all1758 in all cells was evident under both nitrate-replete and depleted

conditions using a GFP transcriptional reporter fusion, suggesting that the gene is constitutively

expressed. In contrast, GFP-dependent fluorescence was not observed from an all1758

translational fusion in all background strains tested. However, the fusion protein did appear to

complement strain UHM184, suggesting the fusion protein was functional. The lack of detectable

fluorescence may indicate that levels of All1758 are very low in cells, and the associated

fluorescence from the fusion protein was below the level of detection, or the GFP portion of the

protein may have been non-functional. This phenomenon is not new. In our hands, plasmid

constructs bearing hetR-gfp and hetF-gfp translational fusions in hetR and hetF backgrounds,

respectively, also complement the inactivated gene, yet visible fluorescence is lacking[22]. Also,

plasmids bearing PhetZ-hetZ-patU5-patU3 and PhetZ-hetZ-patU5-patU3-gfp functionally inhibit

heterocyst formation in the wild-type, but GFP fluorescence is below the level of detection with

the latter plasmid (Chapter 2).

To understand the relationship between All1758 and HetR, post-translational control of All1758 by

HetR was investigated. HetR-GFP (and HetR-CFP; cerulean fluorescent protein) translational

fusions carried on a plasmid in the wild-type results in fluorescence below the level of detection.

Plasmid-borne copies of hetR-gfp in ∆all1758 strains result in punctate fluorescence, but the cells

aggregate, making it difficult to interpret the results (data not shown). To separate the roles of

these two proteins, strains with hetR-gfp (or hetR-cfp) encoded in the chromosome and all1758

deleted from the chromosome were constructed. In parent hetR-gfp (and hetR-cfp) strains (Table

3.1), GFP fluorescence was below the level of detection. No change in fluorescence intensity was

observed in hetR-gfp ∆all1758 and (hetR-cfp ∆all1758) strains (data not shown). The simplest

interpretation is that All1758 does not directly regulate HetR levels.

121

All1758 is required for normal cell size and cell growth. In double deletion or replacement strains

with deletions in all1758 (Table 3.1), the cell size and growth defect corresponding to ∆all1758 is

observed. Alterations in the levels of expression of other genes involved in heterocyst

differentiation have been shown to affect cell size. Inactivation of all2874, which encodes a

diguanylate cyclase results in a light intensity-dependent variability in heterocyst frequency and

reduction in vegetative cell size [81]. Conversely, a strain lacking hetF has enlarged cells, and

extra copies of hetF on a plasmid result in a diminutive cell phenotype [22]. In contrast with the

correlation between smaller cell size and a delay in heterocyst differentiation in the all1758

mutant, a strain with extra copies of hetF forms more heterocysts than the wild-type.

Overexpression of patA also results in enlarged cells [82]. There is no evidence for the

phosphorylation of HetF or PatA, but PatA is similar in sequence to response regulators of two-

component transduction systems [83].

Sequence homology to the PP2C phosphatases SpoIIE and RsbU may suggest a role for sigma

factor regulation by the PP2C phosphatase All1758. In this study, the hypothesized role of

All1758 in the regulation of a heterocyst-specific sigma factor was explored. The partner-switch

mechanism (Fig. 3.15) described for Bacillus and Staphylococcus involves different

phosphorylating states of sigma factor regulators to control regulons involved in sporulation and

stress response, respectively. A preliminary model to explain the genetic interactions between

putative sigma factor regulators, the heterocyst-development genes ntcA and hetR, and All1758

is proposed (Fig. 3.16). However these results represent a cursory study of sigma factor

regulation of heterocyst-specific gene expression and the model is speculative.

HetR appears to exhibit greater genetic control of all1758 than NtcA (Figs. 3.6, 3.7). Whereas

NtcA acts on the promoter of all1758, a direct interaction was shown between HetR and All1758.

Dual regulation by HetR and NtcA may fine-tune the temporal control of All1758 activity. In the

model as drawn, NtcA functions both as a negative and positive effector. NtcA (and HetR)

represses transcription of all1758 (the PP2C phosphatase) (Figures. 3.6, 3.7), but NtcA activates

transcription of all1087 (the anti-antisigma factor) (Fig. 3.12). The All1758-dependent

dephosphorylation of All1087 down-regulates All1087. This negative interaction between All1758

and All1087 deviates from the classical partner-switch model described for Bacillus and

Staphylococcus (Fig. 3.15). In the classical model, phosphatase-mediated dephosphorylation of

target anti-antisigma factors activates, rather than negatively regulates, the anti-antisigma factor.

Consistent with the classical model, All1087 negatively regulates its cognate partner, Alr3423 (an

antisigma factor with All1087-specific kinase activity) in the next stage of the partner-switch

mechanism. An additional mechanism is hypothesized to regulate Alr3423. In addition to

sequestration by All1087, an unidentified All1758-dependent protease is predicted to degrade

122

Alr3423. Degradation of the antisigma factor spoIIAB by the intracellular ClpCP protease

complex has been cited as a mechanism to compartmentalize σF (SigF) activity to the forespore

in Bacillus subtilis [84]. Here, the transient genetic asymmetry of alr3423 due to the All1758-

dependent protease may explain how the ∆all1758 phenotype is invoked in the wild-type upon

overexpression of alr3423.

The growing number of phosphatases and kinases implicated in heterocyst differentiation

indicates that protein phosphorylation plays a prominent role in the regulation of differentiation.

However, understanding the extent and nature of that role is dependent on uncovering the targets

of kinase and phosphatase activity. All1758 is not involved in pattern formation, and appears to

be involved in the later stages of heterocyst differentiation. The requirement of All1758 for normal

cell growth, cell division, morphology and development highlights the significance of

phosphatases in basic cellular processes. The complexities of phosphorylation in the regulation

of Anabaena development warrant further investigation.

Figure 3.15. Partner-switch model of sigma factor activation. (B) Regulation of σF

and σB in Bacillus subtilis and Staphylococcus aureus as typically depicted (bottom

panel; “•” indicates sigma factor sequestration, curved arrows represent direction of

cascading events) and shown with inhibitors (bars) and an activator (straight arrow)

of the pathway (top panel). (A) The preliminary model of interaction (detailed in

Figure 3.16) deviates from the classical pathway shown in (B).

123

Figure 3.16. A preliminary model for the regulation of heterocyst development by a

modified partner-switch mechanism. A speculative interpretation of the genetic

interactions between putative sigma factor regulators, a heterocyst-specific sigma factor, a

protease and the heterocyst-development genes ntcA and hetR with the PP2C

phosphatase All1758 is depicted. Dashed lines represent interaction with a hypothesized

protease. Inhibitors of the pathway represented with bars, activators represented with

arrows.

124

REFERENCES CITED

1. Murry, M.A. and C.P. Wolk, Evidence that the barrier to the penetration of oxygen into

heterocysts depends upon two layers of the cell envelope. Arch. Microbiol, 1989. 151: p.

469-474.

2. Meeks, J.C., et al., Pathways of assimilation of [13

N]N2 and 13

NH4+ by cyanobacteria with

and without heterocysts. J. Bacteriol., 1978. 134(1): p. 125-130.

3. Wolk, C.P., Movement of carbon from vegetative cells to heterocysts in Anabaena

cylindrica. J. Bacteriol., 1968. 96: p. 2138-2143.






Sci. U.S.A., 2005. 102: p. 9907-9912.







8. Olmedo-Verd, E., et al., HetR-dependent and -independent expression of heterocyst-

related genes in an Anabaena strain overproducing the NtcA transcription factor. Journal

of bacteriology, 2005. 187(6): p. 1985-91.

9. Kennelly, P.J., Protein phosphatases--a phylogenetic perspective. Chem Rev, 2001.

101(8): p. 2291-312.

10. Cohen, P., The structure and regulation of protein phosphatases. Annu Rev Biochem,

1989. 58: p. 453-508.

11. Bork, P., et al., The protein phosphatase 2C (PP2C) superfamily: detection of bacterial

homologues. Protein Sci, 1996. 5(7): p. 1421-5.

12. Das, A.K., et al., Crystal structure of the protein serine/threonine phosphatase 2C at 2.0

A resolution. EMBO J, 1996. 15(24): p. 6798-809.

13. Barford, D., Molecular mechanisms of the protein serine/threonine phosphatases. Trends

Biochem Sci, 1996. 21(11): p. 407-12.

14. Wang, L., et al., Genomic analysis of protein kinases, protein phosphatases and two-

component regulatory systems of the cyanobacterium Anabaena sp. strain PCC 7120.

FEMS Microbiol Lett, 2002. 217(2): p. 155-65.

125

15. Ehira, S. and M. Ohmori, NrrA directly regulates expression of hetR during heterocyst

differentiation in the cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol., 2006.

188: p. 8520-8525.

16. Jang, J., et al., Mutual regulation of ntcA and hetR during heterocyst differentiation

requires two similar PP2C-type protein phosphatases, PrpJ1 and PrpJ2, in Anabaena sp.

strain PCC 7120. Journal of bacteriology, 2009. 191(19): p. 6059-66.

17. Cheng, Y., et al., A pair of iron-responsive genes encoding protein kinases with a Ser/Thr

kinase domain and a His kinase domain are regulated by NtcA in the Cyanobacterium


18. Laurent, S., et al., Cell-type specific modification of PII is involved in the regulation of

nitrogen metabolism in the cyanobacterium Anabaena PCC 7120. FEBS letters, 2004.

576(1-2): p. 261-5.

19. Zhang, C.C., A. Friry, and L. Peng, Molecular and genetic analysis of two closely linked

genes that encode, respectively, a protein phosphatase 1/2A/2B homolog and a protein

kinase homolog in the cyanobacterium Anabaena sp. strain PCC 7120. Journal of

bacteriology, 1998. 180(10): p. 2616-22.








23. Shi, L., et al., Two genes encoding protein kinases of the HstK family are involved in

synthesis of the minor heterocyst-specific glycolipid in the cyanobacterium Anabaena sp.


24. Jang, J., et al., PrpJ, a PP2C-type protein phosphatase located on the plasma membrane,

is involved in heterocyst maturation in the cyanobacterium Anabaena sp. PCC 7120.

Molecular microbiology, 2007. 64(2): p. 347-58.

25. Ning, D. and X. Xu, alr0117, a two-component histidine kinase gene, is involved in

heterocyst development in Anabaena sp. PCC 7120. Microbiology, 2004. 150(Pt 2): p.

447-53.

26. Ramirez, M.E., et al., Anabaena sp. strain PCC 7120 gene devH is required for synthesis

of the heterocyst glycolipid layer. Journal of bacteriology, 2005. 187(7): p. 2326-31.

27. Xu, W.L., et al., pkn22 (alr2502) encoding a putative Ser/Thr kinase in the

cyanobacterium Anabaena sp. PCC 7120 is induced by both iron starvation and oxidative

stress and regulates the expression of isiA. FEBS letters, 2003. 553(1-2): p. 179-82.

126

28. Espinosa, J., et al., DevT (Alr4674), resembling a Ser/Thr protein phosphatase, is

essential for heterocyst function in the cyanobacterium Anabaena sp. PCC 7120.

Microbiology, 2010. 156(Pt 12): p. 3544-55.

29. Hughes, K.T. and K. Mathee, The anti-sigma factors. Annu Rev Microbiol, 1998. 52: p.

231-86.

30. Duncan, L., et al., Activation of cell-specific transcription by a serine phosphatase at the

site of asymmetric division. Science, 1995. 270(5236): p. 641-4.

31. Delumeau, O., et al., Functional and structural characterization of RsbU, a stress

signaling protein phosphatase 2C. J Biol Chem, 2004. 279(39): p. 40927-37.

32. Brody, M.S., V. Stewart, and C.W. Price, Bypass suppression analysis maps the

signalling pathway within a multidomain protein: the RsbP energy stress phosphatase 2C

from Bacillus subtilis. Mol Microbiol, 2009. 72(5): p. 1221-34.



Bacteriol., 2007. 189: p. 2460-2467.

34. Elhai, J. and C.P. Wolk, Conjugal transfer of DNA to cyanobacteria. Methods Enzymol.,

1988. 167: p. 747-754.



bacteria. Gene, 1987. 52: p. 147-154.


for hetR in heterocyst differentiation by Anabaena sp. strain PCC 7120. Mol. Microbiol.,

2010. 77: p. 562-574.



Microbiol., 2005. 57: p. 111-123.

38. Ruppert, U., et al., The novel protein phosphatase PphA from Synechocystis PCC 6803

controls dephosphorylation of the signalling protein PII. Molecular microbiology, 2002.

44(3): p. 855-64.



40. Karimova, G., et al., A bacterial two-hybrid system based on a reconstituted signal

transduction pathway. Proc Natl Acad Sci U S A, 1998. 95(10): p. 5752-6.

41. van den Ent, F., et al., Dimeric structure of the cell shape protein MreC and its functional

implications. Mol Microbiol, 2006. 62(6): p. 1631-42.

42. Patrick, J.E. and D.B. Kearns, MinJ (YvjD) is a topological determinant of cell division in

Bacillus subtilis. Mol Microbiol, 2008. 70(5): p. 1166-79.

127

43. Buikema, W.J. and R. Haselkorn, Expression of the Anabaena hetR gene from a copper-

regulated promoter leads to heterocyst differentiation under repressing conditions. Proc.

Natl. Acad. Sci., U.S.A., 2001. 98: p. 2729-2734.






Sci U S A, 2009. 106(47): p. 19884-8.



p. 4473-4482.







2010. 77: p. 562-574.



77-84.

52. Lambein, F. and C.P. Wolk, Structural studies on the glycolipids from the envelope of the

heterocyst of Anabaena cylindrica. Biochemistry, 1973. 12: p. 791-798.

53. Soriente, A., et al., Reinvestigation of heterocyst glycolipids from the cyanobacterium,

Anabaena cylindrica. Phytochemistry (Oxford), 1994. 38(3): p. 641-645.

54. Gambacorta, A., et al., Heterocyst glycolipids from five nitrogen-fixing cyanobacteria.

Gazz. Chim. Ital., 1996. 126: p. 653-656.

55. Leganés, F., et al., Wide variation in the cyanobacterial complement of presumptive

penicillin-binding proteins. Arch. Microbiol., 2005. 184: p. 234-248.


pattern in Anabaena sp. PCC 7120. Molecular microbiology, 2001. 40(4): p. 941-50.

57. Frias, J.E., E. Flores, and A. Herrero, Nitrate assimilation gene cluster from the

Heterocyst-forming cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol., 1997.

179: p. 477-486.

128

58. Vazquez-Bermudez, M.F., E. Flores, and A. Herrero, Analysis of binding sites for the

nitrogen-control transcription factor NtcA in the promoters of Synechococcus nitrogen-

regulated genes. Biochim Biophys Acta, 2002. 1578(1-3): p. 95-8.

59. Levin, P.A., et al., Localization of the sporulation protein SpoIIE in Bacillus subtilis is

dependent upon the cell division protein FtsZ. Mol Microbiol, 1997. 25(5): p. 839-46.

60. Errington, J., R.A. Daniel, and D.J. Scheffers, Cytokinesis in bacteria. Microbiol Mol Biol

Rev, 2003. 67(1): p. 52-65, table of contents.

61. Hu, Z., et al., The MinC component of the division site selection system in Escherichia

coli interacts with FtsZ to prevent polymerization. Proc Natl Acad Sci U S A, 1999. 96(26):

p. 14819-24.

62. Dajkovic, A., et al., MinC spatially controls bacterial cytokinesis by antagonizing the

scaffolding function of FtsZ. Curr Biol, 2008. 18(4): p. 235-44.

63. Hu, Z., E.P. Gogol, and J. Lutkenhaus, Dynamic assembly of MinD on phospholipid

vesicles regulated by ATP and MinE. Proc Natl Acad Sci U S A, 2002. 99(10): p. 6761-6.

64. Johnson, J.E., L.L. Lackner, and P.A. de Boer, Targeting of (D)MinC/MinD and

(D)MinC/DicB complexes to septal rings in Escherichia coli suggests a multistep

mechanism for MinC-mediated destruction of nascent FtsZ rings. J Bacteriol, 2002.

184(11): p. 2951-62.

65. Adams, D.W. and J. Errington, Bacterial cell division: assembly, maintenance and

disassembly of the Z ring. Nat Rev Microbiol, 2009. 7(9): p. 642-53.

66. Khvorova, A., et al., The spoIIE locus is involved in the Spo0A-dependent switch in the

location of FtsZ rings in Bacillus subtilis. J Bacteriol, 1998. 180(5): p. 1256-60.

67. Kennelly, P.J., Protein phosphatases--a phylogenetic perspective. Chemical reviews,

2001. 101(8): p. 2291-312.

68. Carniol, K., et al., Insulation of the sigmaF regulatory system in Bacillus subtilis. J

Bacteriol, 2004. 186(13): p. 4390-4.

69. Pane-Farre, J., et al., Role of RsbU in controlling SigB activity in Staphylococcus aureus

following alkaline stress. J Bacteriol, 2009. 191(8): p. 2561-73.

70. Feucht, A., et al., Bifunctional protein required for asymmetric cell division and cell-

specific transcription in Bacillus subtilis. Genes Dev, 1996. 10(7): p. 794-803.

71. Hecker, M., J. Pane-Farre, and U. Volker, SigB-dependent general stress response in

Bacillus subtilis and related gram-positive bacteria. Annu Rev Microbiol, 2007. 61: p. 215-

36.

72. Giachino, P., S. Engelmann, and M. Bischoff, Sigma(B) activity depends on RsbU in

Staphylococcus aureus. J Bacteriol, 2001. 183(6): p. 1843-52.

73. Mella-Herrera, R.A., et al., The sigE gene is required for normal expression of heterocyst-

specific genes in Anabaena sp. strain PCC 7120. J Bacteriol, 2011. 193(8): p. 1823-32.

129

74. Aldea, M.R., R.A. Mella-Herrera, and J.W. Golden, Sigma factor genes sigC, sigE and

sigG are upregulated in heterocysts of the cyanobacterium Anabaena sp. strain PCC

7120. J. Bacteriol., 2007. 189: p. 8392-8396.

75. Lee, D.J., S.D. Minchin, and S.J. Busby, Activating transcription in bacteria. Annu Rev

Microbiol, 2012. 66: p. 125-52.

76. Lucet, I., et al., Direct interaction between the cell division protein FtsZ and the cell

differentiation protein SpoIIE. EMBO J, 2000. 19(7): p. 1467-75.





79. Barak, I. and P. Youngman, SpoIIE mutants of Bacillus subtilis comprise two distinct

phenotypic classes consistent with a dual functional role for the SpoIIE protein. J

Bacteriol, 1996. 178(16): p. 4984-9.

80. Barak, I. and A.J. Wilkinson, Where asymmetry in gene expression originates. Mol

Microbiol, 2005. 57(3): p. 611-20.

81. Neunuebel, M.R. and J.W. Golden, The Anabaena sp. strain PCC 7120 gene all2874

encodes a diguanylate cyclase and is required for normal heterocyst development under

high-light growth conditions. J Bacteriol, 2008. 190(20): p. 6829-36.






84. Pan, Q., D.A. Garsin, and R. Losick, Self-reinforcing activation of a cell-specific

transcription factor by proteolysis of an anti-sigma factor in B. subtilis. Mol Cell, 2001.

8(4): p. 873-83.

130

CHAPTER 4. OVERACTIVE ALLELES OF HETR IN THE CYANOBACTERIUM ANABAENA

SP. STRAIN PCC 7120

INTRODUCTION

Composed of only two cell types under specific environmental conditions, the vegetative cell and

the differentiated heterocyst cell, Anabaena sp. strain PCC 7120 (herein Anabaena) represents

one of the most basic multicellular states present in nature. Anabaena executes the complexities

of developmental patterning common to higher organisms in a linear arrangement. The

assignment of cell type involves interaction between key developmental genes in response to a

starvation signal. Deprivation of nitrogen induces the regulatory cascade of heterocyst

differentiation in Anabaena. While each cell has the potential to become a differentiated cell, only

a subset of cells in Anabaena develops into a heterocyst. Moreover these heterocyst cells only

form at regular intervals (at an approximate frequency of one heterocyst per ten vegetative cells).

How genetically uniform cells diverge into different cell types and form a pattern is a central

question in developmental biology.

The activator-inhibitor model [1] has been applied to the molecular regulation of periodic

patterning in developmental systems including Anabaena. Central to this model is the dual role of

the activator, HetR, in this process. HetR auto-catalysis leads to the production of more HetR,

leading to self-enhancement of this activating signal, but simultaneously also produces the early-

stage developmental inhibitor, PatS, and the late-stage inhibitor, HetN. These two inhibitory

proteins, although temporally-restricted to different stages of development, contain the same

RGSGR-pentapeptide (“PatS-5”) motif in their primary sequence. The PatS peptide is

responsible for de novo pattern formation. Subsequent catalytic feedback between the activator

HetR and the inhibitor PatS leads to a spatial difference in the concentrations of both. Thus HetR

and PatS cooperatively reinforce random differences in HetR levels observed in each cell. The

distance between clusters of induced cells is a product of both the extent of intercellular transfer

of PatS and the corresponding effect on HetR. Across the filament, groups of cells are primed for

differentiation, but not in any predictable pattern. Over time, the groups of cells are arranged into

a periodic pattern foreshadowing the ultimate spacing of the differentiated cells. Eventually the

intervals of cell clusters resolve to a single cell which becomes the heterocyst. Whereas the

pattern is initiated by PatS, maintenance of the pattern of ten relies on lateral hetN-dependent

inhibition originating from differentiated cells. HetN concentration, and consequently HetN-

dependent inhibition of heterocyst formation, is inversely related to distance from heterocysts. As

cell division occurs in undifferentiated cells, the intervals between heterocysts are increased, and

HetN concentrations reach levels that no longer inhibit differentiation. Heterocysts arise at HetN-

131

concentration minima, following catalytic feedback between HetR and PatS, to restore the

periodic pattern.

Protein-mediated interplay between the developmental regulators HetR, PatS and HetN regulate

Anabaena development. HetR is essential in coordinating heterocyst development[2]. However,

the primary sequence of HetR does not show similarity to protein domains of known biological

function. Structural comparisons of HetR suggest that the N-terminal DNA-binding unit is related

to other DNA-binding proteins[3], consistent with its role in the transcriptional activation of

thousands of genes related to heterocyst development. The unique folds present in the structure

of HetR may to function as a physical scaffold for the organization of factors (RNA, RNA

polymerase and other proteins and peptides) necessary for the transcription of heterocyst-

development genes[3]. A genetic approach was used to identify interactions between HetR and

PatS-5, PatS, and HetN in an effort to further understand the molecular regulation of heterocyst

differentiation in Anabaena. Residues that contributed to the functional overactivity of HetR,

manifested as the enhanced activation of heterocyst formation irrespective of inhibitor

concentrations, were isolated. The region of HetR spanning amino acids 250-256 was found in

this study to be necessary for sensitivity of HetR to PatS-5. In addition, seven residues in HetR

were identified as necessary for function as part of a mutagenesis study.






for photomicroscopy were as previously described [4]. Images were processed in Adobe

Photoshop CS2.




Plasmids for making chromosomal alleles

Plasmids pST121, pST122, pST117, pST119, pST131, pST132, pST133, pST185, pST135,

pST143, pST144, pST145 and pST134 are suicide vectors used to replace the chromosomal

hetR-locus with hetR(Y51C), hetR(P206S), hetR(Y51F), hetR(Y51H), hetR(Q59P), hetR(D230G),

132

hetR(E56G) and hetR(C48R), hetR(L162I), hetR(Q59E), hetR(D230E), hetR(E56D) and

hetR(C48L), respectively. To generate each of the mutant hetR alleles, overlap extension PCR

was performed using the inner primers listed in Table 4.3 (primer numbers 4 to 27) with names

corresponding to that of the resulting substitution with the outer primers PhetR-BamHI-F and

hetR-SacI-SpeI-R. The resulting PCR products were cloned into plasmid pDR327 as NcoI−SpeI

fragments to create plasmids pST121, pST122, pST117, pST119, pST131, pST132, pST133,

pST185, pST135, pST143, pST144, pST145 and pST134.

Plasmids pJM100, pJM101, pJM102, pJM103, pST211, pJM104, pJM105 and pDR219 are

suicide vectors used to replace the chromosomal hetR-locus with hetR(R250K), hetR(A251G),

hetR(L252V), hetR(E253D), hetR(E254D), hetR(L255V), hetR(D256E) and hetR(E254G),

respectively. Overlap extension PCR was used to generate each of the mutant hetR alleles

except hetR(E254D) and hetR(E254G) that used the inner primers listed in Table 4.3 (primer

numbers 28 to 41) with names corresponding to that of the resulting substitution with the outer

primers PhetR-BamHI-F and hetR 3′ Seq. The resulting PCR products were cloned into plasmid

pDR327 as NcoI−SpeI fragments to create plasmids pJM100, pJM101, pJM102, pJM103,

pJM104, and pJM105. Plasmids pST211 and pDR219 were generated in a similar fashion, but

outer primer hetR-SacI-SpeI-R was used in place of hetR 3′ seq for pST211 and previously

published inner primers [4] were used to create hetR(E254G) in pDR219.

Plasmids pST376, pST377, pST378, pST379, pST380, pST381 and pST382 are suicide vectors

used to replace the chromosomal hetR-locus with PpetE-hetR(Y51C)-cfp, PpetE-hetR(P206S)-cfp,

PpetE-hetR(Q59P)-cfp, PpetE-hetR(D230G)-cfp, PpetE-hetR(C48R)-cfp, PpetE-hetR(L162I)-cfp,

respectively. To generate each of the mutant hetR alleles, primers PpetE-NcoI-F and cfp-SpeI-R

were used along with plasmids pST235, pST236, pST237, pST238, pST239, pST240 and

pST241 as template DNA for the reaction. The resulting PCR products were cloned directly into

plasmid pDR325 as NcoI−SpeI fragments to create plasmids pST376, pST377, pST378, pST379,

pST380, pST381 and pST382.

Plasmids for mutagenesis studies

A blunt end procedure was used to inactivate the BamHI site on plasmid pDR197 (bearing PpetE-

patS) to generate plasmid pST100 as an intermediate construct to create plasmid pST102.

Plasmid pST102 bears a transcriptional fusion between patS and the petE promoter. Using

primers PpetE-S-F and PpetE-S(ApR)-R, a 416-bp region containing PpetE-patS was amplified

from pST100 and cloned as an EcoRI fragment into pAM505 to yield pST102.

133

Plasmid pST103 bears PpetE-patS oriented divergently with PhetR-hetR. An 1848-bp fragment

containing PhetR-hetR was amplified from the chromosome using primers hetR-F-BamHI and

hetR-R-SacI and cloned into pST102 to generate pST103.

Multicopy plasmids

The mobilizable shuttle vectors plasmid pST125, pST126 and pST127 carrying hetR(Q59E),

hetR(D230E) and hetR(E56D) respectively were created by moving fragments containing each

hetR allele under the control of PhetR as BamHI−SacI fragments from pST143, pST144 and

pST145 into pAM504 to create pST125, pST126 and pST127.

Plasmids pST128, pST129, pST130 carrying hetR(Q59E), hetR(D230E) and hetR(E56D)

respectively were created by moving fragments containing each hetR allele under the control of

PhetR as BamHI−SacI fragments pST143, pST144 and pST145 into pST102 (bearing PpetE-patS)

to create pST128, pST129 and pST130.

Translational fusion constructs


pST218, pST219, pST220, pST221 and pST201 carry translational fusions to gfp with

hetR(Y51C), hetR(P206S), hetR(Y51F), hetR(Y51H), hetR(Q59P), hetR(D230G), hetR(E56G),

hetR(C48R), hetR(L162I), hetR(R250K), hetR(A251G), hetR(L252V), hetR(E253D),

hetR(E254D), hetR(L255V) and hetR(D256E), respectively, under the control of the petE

promoter. To generate each of the mutant hetR alleles, overlap extension PCR was performed

using pDR293 as template and the inner primers listed in Table 4.3 (primer numbers 4-7, 12-19,

28-41) with names corresponding to that of the resulting substitution with the outer primers

PhetR-BamHI-F and hetR-SmaI-R. The resulting PCR products were ligated into the EcoRV site

of pBluescript SK+ to make pST202, pST203, pST204, pST205, pST206, pST207, pST208,

pST209, pST222, pST223, pST224, pST225, pST226 and pST210 and subsequently moved as

BamHI−SmaI fragments into pSMC232 to create plasmids pST193, pST194, pST195, pST196,

pST197, pST198, pST199, pST200, pST217, pST218, pST219, pST220, pST221 and pST201.


pST230, pST231, pST232, pST233 and pST234 carry translational fusions to cfp with

hetR(Y51C), hetR(P206S), hetR(Y51F), hetR(Y51H), hetR(Q59P), hetR(D230G), hetR(E56G)

and hetR(C48R), hetR(L162I), hetR(R250K), hetR(A251G), hetR(L252V), hetR(E253D),


134

promoter. Fragments containing PpetE-hetR(mutant) were moved from plasmids pST193,

pST194, pST195, pST196, pST197, pST198, pST199, pST200, pST217, pST218, pST219,

pST220, pST221 and pST201 as BamHI−SmaI fragments into pSMC242 to create plasmids


pST231, pST232, pST233 and pST234.


pST271, pST272, pST273, pST274 and pST275 carry translational fusions to cfp with

hetR(Y51C), hetR(P206S), hetR(Y51F), hetR(Y51H), hetR(Q59P), hetR(D230G), hetR(E56G)

and hetR(C48R), hetR(L162I), hetR(R250K), hetR(A251G), hetR(L252V), hetR(E253D),


promoter. Fragments containing PpetE-hetR(mutant)-cfp were moved from plasmids pST235,


pST232, pST233 and pST234 as BamHI−SacI fragments into pCO100 to create plasmids


pST272, pST273, pST274 and pST275.

Protein expression vectors

The pET26b (Novagen) derivatives, pST307 to pST313 were constructed to overexpress

hetR(R250K) to hetR(D256E) in E. coli and facilitate protein purification. A 900-bp region

containing the coding region for hetR was amplified from the chromosome using primers hetR F

NdeI express and hetR R XhoI express and plasmids pST200, pST217, pST218, pST219,

pST220, pST221 and pST201 as template DNA for the reaction. The fragments were ligated into

the EcoRV site of pBluescript SK+ to make pST300, pST301, pST302, pST303, pST304,

pST305, pST306. The region corresponding to the hetR substitution was moved individually as

NdeI-XhoI fragments from pST300 to pST306 into pET26b to generate pST307 to pST313.

Construction of strains containing mutant alleles

Descriptions of strains constructed in this study are summarized in Table 4.1. Replacements of

chromosomal DNA was performed as previously described [5]. All strains were screened via

colony PCR with primers annealing outside of the chromosomal region introduced on the suicide

plasmids used for strain construction and tested for sensitivity and resistance to the appropriate

antibiotics.

135

Strains of Anabaena with mutant alleles of hetR in place of wild-type hetR(R250) to hetR(D256)

region were created using the hetR-deletion strain UHM103 and plasmids pJM100, pJM101,

pJM102, pJM103, pST211, pJM104, pJM105, and pDR219 to generate strains UHM163,

UHM164, UHM165, UHM166 and UHM318, UHM167 and UHM331, UHM168, UHM169,

UHM122, respectively. Strains UHM166 and UHM167 (hetR(E253D) and hetR(E254D)

respectively) are single recombinants in which the entire plasmid is in the hetR chromosomal

locus, whereas UHM318 and UHM331 (also hetR(E253D) and hetR(E254D)) and the other

strains are the same as PCC 7120 except for the change in hetR sequence.

Strains of Anabaena with mutant alleles of hetR in place of wild-type hetR(Y51C), hetR(P206S),

hetR(Y51F), hetR(Y51H), hetR(Q59P), hetR(D230G), hetR(E56G) and hetR(C48R), hetR(L162I),

hetR(Q59E), hetR(D230E), hetR(E56D) and hetR(C48L) were created using the hetR-deletion

strain UHM103 and plasmids pST121, pST122, pST117, pST119, pST131, pST132, pST133,

pST185, pST135, pST143, pST144, pST145 and pST134 to generate strains UHM179, UHM182,

UHM180, UHM181 and UHM170, UHM171, UHM172, UHM173, UHM174, UHM175, UHM176,

UHM177 and UHM178 respectively.

Strains of Anabaena with mutant alleles of hetR(Y51C), hetR(P206S), hetR(Q59P),

hetR(D230G), hetR(E56G) and hetR(C48R), hetR(L162I) driven by PpetE and translationally fused

to cfp were to be created using the hetR-deletion strain and the hetR- patA-deletion strain with

plasmids pST376, pST377, pST378, pST379, pST380, pST381 and pST382, respectively.

Strains UHM304, UHM305, UHM306 and UHM307 were generated using the hetR-deletion strain

and plasmids pST376, pST378, pST380 and pST381. Strain UHM308 was made using the hetR-

patA-deletion strain with plasmid pST381.

Deletion of the alr9018 coding region was performed using plasmids pSMC237 and Anabaena

sp. strain PCC 7120 to yield strains Δalr9018 with Ω cassette (UHM188), respectively. The

resultant strains were screened via colony PCR with primers alr9018 flank up F and alr9018 flank

down R, which anneal outside the chromosomal region introduced onto pSC237 for deletion of

alr9018, and tested for sensitivity and resistance to the appropriate antibiotics.

Mutagenesis studies

To create a library of mutations in the coding region of hetR, error-prone PCR was performed on

plasmid pDR204 (with primers 505-BamHI and 505-SacI) using a previously described method to

generate 0.5 mutations/kb[4], but with the exception of 2 mM dNTPs. Error-prone PCR of the

region carrying Pnir-hetR on plasmid pDR204 was subsequently moved into plasmids pST102 and

136

pAM505 and stored for conjugation. Introduction of the mutated pST102 plasmids into ∆hetR is

used for isolation of patS-bypass mutants of HetR, and similarly mutations present on plasmid

pAM505 introduced into pN∆hetR is used for isolation of hetN-bypass mutants of HetR.

The hypermutator Escherichia coli strain XL1-RED was employed to create a library of mutations

within plasmids pST103 and pDR138 in a method described previously[4]. The library of

mutations in plasmid pST103 was introduced into ∆hetR via conjugation. Similarly, the library of

mutations in plasmid pDR138 was introduced into strains pN and pN∆hetR via conjugation.

In vivo PatS-5 sensitivity assays

Duplicate cultures of Anabaena sp. strain PCC 7120 and the hetR mutant strains (UHM163,

UHM162, UHM165, UHM166, UHM167, UHM168 and UHM169) were grown to an approximate

optical density of 0.4 at 750 nm in BG-11 medium, which contains nitrate, a fixed form of nitrogen.

For one set of cultures, the culture medium was replaced with fresh BG-11, and replaced

thereafter every 48 h with BG-110, which lacks fixed nitrogen. For the other set of cultures, PatS-5

was included in the medium at a concentration of 1 μM. The percentage of 500 cells that were

heterocysts was determined microscopically and recorded after each change of medium. Reported

values are the average of three replicates with one standard deviation. Conditions for

photomicroscopy were as described previously[4]

137

Table 4.1. Bacterial strains used in Chapter 4

Anabaena sp. strain

Relevant characteristic(s) Source or reference


7120PN PCC 7120 with PhetN-hetN replaced by PpetE-hetN [6]

UHM110 Strain 7120PN with deletion in hetR [7]

UHM103 hetR-deletion strain [5]

UHM109 hetR-, patA-deletion strain [7]

UHM132 hetR-, hetF-deletion strain [8]

UHM135 patA-, hetF-deletion strain [8]

UHM101 patA-deletion strain [7]

UHM114 patS-deletion strain [5]

UHM179 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(Y51C)

This study

UHM182 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(P206S)

This study

UHM180 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(Y51F)

This study

UHM181 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(Y51H)

This study

UHM170 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(Q59P)

This study

UHM171 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(D230G)

This study

UHM172 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(E56G)

This study

UHM173 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(C48R)

This study

UHM174 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(L162I)

This study

UHM175 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(Q59E)

This study

UHM176 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(D230E)

This study

UHM177 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(E56D)

This study

UHM178 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(C48L)

This study

UHM163 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(R250K)

[9]; This study

UHM164 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(A251G)

[9]; This study

UHM165 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(L252V)

[9]; This study

UHM166 Plasmid pJM103 recombined into the hetR locus of strain UHM103

[9]; This study


This study

138

Table 4.1. (Continued) Bacterial strains used in Chapter 4

Anabaena sp. strain


UHM167 Plasmid pST211 recombined into the hetR locus of strain UHM103

[9]; This study


This study

UHM168 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(L255V)

[9]; This study

UHM169 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(D256G)

[9]; This study

UHM122 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(E254G)

[4]

UHM304 hetR-deletion strain with chromosomal PhetR-hetR replaced by PpetE-hetR(Y51C)-cfp

This study

UHM305 hetR-deletion strain with chromosomal PhetR-hetR replaced by PpetE-hetR(Q59P)-cfp

This study

UHM306 hetR-deletion strain with chromosomal PhetR-hetR replaced by PpetE-hetR(E56G)-cfp

This study

UHM307 hetR-deletion strain with chromosomal PhetR-hetR replaced by PpetE-hetR(C48R)-cfp

This study

UHM308 hetR- patA-deletion strain with chromosomal PhetR-hetR replaced by PpetE-hetR(C48R)-cfp

This study

UHM188 alr9018 replaced by Ω cassette This study

139




r Neo

r [10]

pCO100 Mobilizable shuttle vector based on pAM504 for replication in E. coli and Anabaena; Sm

r Sp

r

[11]

pDR197 pBluescript SK+ bearing PpetE-patS [11]


pDR204 pAM504 carrying Pnir-hetR [11]



pDR219 Suicide vector carrying PhetR-hetR(E254G) [4]

pDR293 pSMC232 carrying PpetE-hetR [8]

pDR325 Suicide plasmid based on pRL277, carrying hetR [11]

pDR327 Suicide plasmid based on pRL277, carrying PpetE-hetR-gfp [13]

pET28b Expression vector for generating polyhistidine epitope-tagged proteins; Km

r

Novagen


r [14]


pSMC242 pAM504 bearing promotorless cfp for translational fusions [8]

pSMC237 Suicide plasmid based on pRL278, carrying alr9018 [11]

pST100 pBluescript SK+ bearing PpetE-patS and inactivated BamHI restriction site

This study

pST102 pAM505 bearing PpetE-patS This study

pST103 pST102 bearing PhetR-hetR This study

pST121 Suicide vector carrying PhetR-hetR(Y51C) This study

pST122 Suicide vector carrying PhetR-hetR(P206S) This study

pST117 Suicide vector carrying PhetR-hetR(Y51F) This study

pST119 Suicide vector carrying PhetR-hetR(Y51H) This study

pST131 Suicide vector carrying PhetR-hetR(Q59P) This study

pST132 Suicide vector carrying PhetR-hetR(D230G) This study

pST133 Suicide vector carrying PhetR-hetR(E56G) This study

pST185 Suicide vector carrying PhetR-hetR(C48R) This study

pST135 Suicide vector carrying PhetR-hetR(L162I) This study

pST143 Suicide vector carrying PhetR-hetR(Q59E) This study

pST144 Suicide vector carrying PhetR-hetRD230E) This study

pST145 Suicide vector carrying PhetR-hetR (E56D) This study

pST134 Suicide vector carrying PhetR-hetR(C48L) This study

pJM100 Suicide vector carrying PhetR-hetR(R250K) [9]

pJM101 Suicide vector carrying PhetR-hetR(A251G) [9]

pJM102 Suicide vector carrying PhetR-hetR(L252V) [9]

pJM103 Suicide vector carrying PhetR-hetR(E253D) [9]

pST211 Suicide vector carrying PhetR-hetR(E254D) [9]

pJM104 Suicide vector carrying PhetR-hetR(L255V) [9]

140



pJM105 Suicide vector carrying PhetR-hetR(D256E) [9]

pST376 Suicide vector carrying PhetR-hetR(Y51C)-cfp This study

pST377 Suicide vector carrying PhetR-hetR(P206S)-cfp This study

pST378 Suicide vector carrying PhetR-hetR(Q59P)-cfp This study

pST379 Suicide vector carrying PhetR-hetR(D230G)-cfp This study

pST380 Suicide vector carrying PhetR-hetR(E56G)-cfp This study

pST381 Suicide vector carrying PhetR-hetR(C48R)-cfp This study

pST382 Suicide vector carrying PhetR-hetR(L162I)-cfp This study

pST108 pAM504 carrying PhetR-hetR (1-402) This study

pST109 pAM504 carrying PhetR-hetR (1-528) This study

pST121 pAM504 carrying PhetR-hetR(Y51C) This study

pST122 pAM504 carrying PhetR-hetR(P206S) This study

pST104 pAM504 carrying PhetR-hetR(Y51F) This study

pST105 pAM504 carrying PhetR-hetR(Y51H) This study

pST125 pAM504 carrying PhetR-hetR(Q59E) This study

pST126 pAM504 carrying PhetR-hetR(D230E) This study

pST127 pAM504 carrying PhetR-hetR(E56D) This study

pST128 pST102 carrying PhetR-hetR(Q59E) This study

pST129 pST102 carrying PhetR-hetR(D230E) This study

pST130 pST102 carrying PhetR-hetR(E56D) This study

pST193 pSMC232 carrying PhetR-hetR(Y51C) This study

pST194 pSMC232 carrying PhetR-hetR(P206S) This study

pST195 pSMC232 carrying PhetR-hetR(Q59P) This study

pST196 pSMC232 carrying PhetR-hetR(D230G) This study

pST197 pSMC232 carrying PhetR-hetR(E56G) This study

pST198 pSMC232 carrying PhetR-hetR(C48R) This study

pST199 pSMC232 carrying PhetR-hetR(L162I) This study

pST200 pSMC232 carrying PhetR-hetR(R250K) This study

pST217 pSMC232 carrying PhetR-hetR(A251G) This study

pST218 pSMC232 carrying PhetR-hetR(L252V) This study

pST219 pSMC232 carrying PhetR-hetR(E253D) This study




pST235 pSMC242 carrying PhetR-hetR(Y51C) This study

pST236 pSMC242 carrying PhetR-hetR(P206S) This study

pST237 pSMC242 carrying PhetR-hetR(Q59P) This study


pST239 pSMC242 carrying PhetR-hetR(E56G) This study

pST240 pSMC242 carrying PhetR-hetR(C48R) This study

141



pST241 pSMC242 carrying PhetR-hetR(L162I) This study

pST228 pSMC242 carrying PhetR-hetR(R250K) This study

pST229 pSMC242 carrying PhetR-hetR(A251G) This study






pST276 pCO100 carrying PhetR-hetR(Y51C)-cfp This study

pST277 pCO100 carrying PhetR-hetR(P206S)-cfp This study

pST278 pCO100 carrying PhetR-hetR(Q59P)-cfp This study

pST279 pCO100 carrying PhetR-hetR(D230G-cfp This study

pST280 pCO100 carrying PhetR-hetR(E56G)-cfp This study

pST281 pCO100 carrying PhetR-hetR(C48R)-cfp This study

pST282 pCO100 carrying PhetR-hetR(L162I)-cfp This study

pST269 pCO100 carrying PhetR-hetR(R250K)-cfp This study

pST270 pCO100 carrying PhetR-hetR(A251G)-cfp This study

pST271 pCO100 carrying PhetR-hetR(L252V)-cfp This study

pST272 pCO100 carrying PhetR-hetR(E253D)-cfp This study

pST273 pCO100 carrying PhetR-hetR(E254D)-cfp This study

pST274 pCO100 carrying PhetR-hetR(L255V)-cfp This study

pST275 pCO100 carrying PhetR-hetR(D256G)-cfp This study

pST284 pCO100 carrying PhetR-hetR-cfp This study

pST307 pET26b carrying hetR(R250K) This study

pST308 pET26b carrying hetR(A251G) This study

pST309 pET26b carrying hetR(L252V) This study

pST310 pET26b carrying hetR(E253D) This study

pST311 pET26b carrying hetR(E254D) This study

pST312 pET26b carrying hetR(L255V) This study

pST313 pET26b carrying hetR(D256E) This study

142


Primer no.


1 PhetR-BamHI-F ATATAGGATCCA ACCCTTATGACAAAGGAC

2 HetR 3’ Seq TGCTCTACACCACATTGGTTGG

3 hetR-SacI-SpeI-R TATATAGAGCTCACTAGTACTTTTATTCACTCTGGGTGC

4 hetR Y51C-F AGTGTGCCATTTGCATGACTTATCTAGAGCAAGGAC

5 hetR Y51C-R AGATAAGTCATGCAAATGGCACACTTAGCCGCCGT

6 hetR P206S-F CTTGGGGAATGTCCTTCTATGCCCTGACTCGTCCC

7 hetR P206S-R AGGGCATAGAAGGACATTCCCCAAGGAGAATCAATTCTTG

8 hetR Y51F-F AGTGTGCCATTTTCATGACTTATCTAGAGCAAGGAC

9 hetR Y51F-R AGATAAGTCATGAAAATGGCACACTTAGCCGCCGT

10 hetR Y51H-F AGTGTGCCATTCACATGACTTATCTAGAGCAAGGAC

11 hetR Y51H-R AGATAAGTCATGTGAATGGCACACTTAGCCGCCGT

12 hetR Q59P-F TAGAGCAAGGACCAAACCTCCGGATGACCGGACATTTGC

13 hetR Q59P-R ATCCGGAGGTTTGGTCCTTGCTCTAGATAAGTCATG

14 hetR D230G-F TTATGGTGGAAGGTACCGCTCGGTATTTCCGCATGATGAAA

15 hetR D230G-R TACCGAGCGGTACCTTCCACCATAATATAAGTCCGCTCTTG

16 hetR C48R-F1 CGGCGGCTAAGCGTGCCATTTACATGACTTATCTAGAGC

17 hetR C48R-R1 ATGTAAATGGCACGCTTAGCCGCCGTTGCTGCTGCATC

18 hetR L162I-F TGATCGAATTTATCCATAAGCGATCGCAAGAGGATCTG

19 hetR L162I-R GATCGCTTATGGATAAATTCGATCAACTCCAAAAACTC

20 hetR Q59E-F TAGAGCAAGGAAACAACCTCCGGATGACCGGACATTTGC

21 hetR Q59E-R ATCCGGAGGTTGTTTCCTTGCTCTAGATAAGTCATG

22 hetR D230E-F TTATGGTGGAAGAGACCGCTCGGTATTTCCGCATGATGAAA

23 hetR D230E-R TACCGAGCGGTCTCTTCCACCATAATATAAGTCCGCTCTTG

24 hetR E56D-F TGACTTATCTAGATCAAGGACAAAACCTCCGGATGACC

25 hetR E56D-R TTTTGTCCTTGATCTAGATAAGTCATGTAAATGGCACAC

26 hetR C48R-F CGGCGGCTAAGCTGGCCATTTACATGACTTATCTAGAGC

143


Primer no.


27 hetR C48R-R ATGTAAATGGCCAGCTTAGCCGCCGTTGCTGCTGCATC

28 hetRR250K-F CAAACGCTATGAAGGCCTTAGAAGAA

29 hetRR250K-R CTTCTAAGGCCTTCATAGCGTTTGGC

30 hetRA251G-F CGCTATGCGAGGATTAGAAGAACTCGATGTGCCAC

31 hetRA251G-R GTTCTTCTAATCCTCGCATAGCGTTTGGCCG

32 hetRL252V-F CTATGCGAGCCGTGGAAGAACTCGATGTGCCACC

33 hetRL252V-R CGAGTTCCTTCCACGCTCGCATAGCGTTTGG

34 hetRE253D-F CGCAGCCTTAGATGAACTCGATGTGCCACCAG

35 hetRE253D-R CATCGAGTTCATCTAAGGCTCGCATAGCGTTTG

36 hetR E254D-F CGAGCCTTAGAAGATCTCGATGTGCCACCAGAG

37 hetR E254D-R GCACATCGAGATCTTCATAGGCTCGCATAGCGTTTG

38 hetRL255V-F CCTTAGAAGAAGTGGATGTGCCACCA

39 hetRL255V-R GGTGGCACATCCACTTCTTCTAAGGC

40 hetRD256E-F CTTAGAAGAACTCGAAGTGCCACCAGAGCGCTG

41 hetRD256E-R CTGGTGGCACTTCGAGTTCTTCTAAGGCTCG

42 hetR-NdeI-R134-F CTCTT CATATGAAGACAAATTGAGCATAAG TTACCCAG

43 hetR-NdeI-M176-F CTCTTCATATGGAGTTAAGCGAAGCCCTGGCAGAGC

44 hetR-PstI-R ATATCTGCAGTTAATCTTCTTTTCTACCAAACACCATTTG

45 PpetE-S-F TATGAATTC GCTGAGGTACTGAGTACACAGC

46 PpetE-S(ApR)-R AAAGAATTCGGCGGCCGCTCTAGAACTAG

47 hetR-F-BamHI ATCCCGGATCCCCTGCCAATGCAGAAGGTTAAAC

48 hetR-R-SacI CATTAGAGCTCCTTTTATTCACTCTGGGTGC

49 505-BamHI CTACGGGGTCTGACGCTCAGTGG

50 505-SacI GTCGAACTGCGCGCTAACTAT TC

51 PpetE-NcoI-F ATATACCATGGCTGAGGTACTGAGTACACAG

52 CFP SpeI R AATAACTAGTTTACTTGTACAGCTCGTCCATGC

53 hetR F NdeI express ATCGATCGCATATGAGTAACGACATCGATCTGATC

54 hetR R XhoI express TGACTCTCGAGCTAATCTTCTTTTCTACCAAACACC

55 alr9018 flank up F TCAGGGATCAGAAAATCAGG

56 alr9018 flank down R GTATTTTGACGCAGAGCTTC

144

RESULTS

Conservative substitutions at HetR residues 250−256 affect heterocyst formation and

sensitivity to PatS-5

As part of a mutagenesis study designed to identify residues of HetR required for

function, an allele of hetR coding for an E254G substitution was found to cause

differentiation into heterocysts of essentially all cells containing a multicopy plasmid

carrying the mutant gene [15]. Introduction of a plasmid bearing the hetR(E254G) allele

under the control of the native promoter of hetR to PCC 7120 by conjugation resulted in

no viable transconjugants. To limit expression of the hetR(E254G) allele, the wild-type

promoter region was replaced with that of the copper-inducible petE promoter and the

plasmid was introduced into PCC 7120. Limited growth of a small number of

transconjugants on solid BG-11 medium containing ammonia and lacking copper was

observed. The resulting colonies lacked the green color of colonies of the wild-type, were

composed primarily of heterocysts, and showed little to no growth in liquid culture (data

not shown). By comparison, a strain with PpetE driving transcription of the wild-type allele

of hetR in place of the hetR(E254G) allele under the same conditions differentiated less

than 1% heterocysts. Replacement of the wild-type promoter region with a second

promoter, that of nirA, transcription from which is repressed in ammonia and induced in

nitrate or in the absence of fixed nitrogen [16], permitted the growth of filaments on solid

and liquid BG-11 medium with ammonia replacing nitrate as the nitrogen source.

Apparently, there is tighter on-off control of transcription with the nir promoter than with

petE in our hands. Transfer of filaments to BG-11 with nitrate or lacking a fixed source of

nitrogen resulted in the differentiation of greater than 90% of cells into heterocysts. By

comparison, a strain with Pnir driving transcription of the wild-type allele of hetR in place of

the hetR(E254G) allele under the same conditions differentiated about 30% heterocysts

(data not shown). When the native copy of hetR in PCC 7120 was replaced with an allele

encoding the E254G substitution, about 25% of cells in the resulting strain were

heterocysts 48 h after induction. The phenotype of this strain was indistinguishable from

that of a strain with a copy of hetR encoding the more conservative E254D substitution,

which is discussed in more detail below.

Differentiation of nearly all cells of a filament has been observed when both patS and

hetN are inactivated simultaneously or when an allele of hetR encoding protein less

sensitive to both inhibitors is overexpressed ectopically and the mutant strains are grown

in the absence of combined nitrogen [17, 18]. To determine if the more conservative

145

E254D substitution also resulted in an overactive allele of HetR and if residues in the

region of E254 were involved in the response of HetR to PatS-5, alleles of HetR encoding

individual conservative substitutions at residues R250 to D256 were used to substitute

hetR by allelic replacement in PCC 7120 (see Table 4.1). Filaments of strains with

alleles encoding R250K, E253D, E254D, L255V, and D256E substitutions consisted of

about 14 to 48% heterocysts, and the presence or absence of fixed nitrogen in the

medium had little effect on differentiation by an individual strain. By comparison, about

9% of cells in filaments of PCC 7120 were heterocysts in a medium that lacked fixed

nitrogen, and about 1% when fixed nitrogen was present. Conversely, A251G and L252V

substitutions prevented or reduced differentiation, respectively (Fig. 4.1). Strains that

differentiated an increased number of heterocysts were also less sensitive to PatS-5.

Addition of PatS-5 to the growth medium prevented differentiation of heterocysts by the

wild-type strain, PCC 7120 [19]. In contrast, 8 to 25% of cells in filaments of strains with

alleles encoding R250K, E253D, E254D, L255V, and D256E substitutions were

heterocysts in a medium that contained PatS-5 at a concentration of 1 M (Fig. 4.1). As

expected, PCC 7120 lacked heteroycsts under the same conditions. Taken together,

these results suggest that residues R250, E253, E254, L255 and D256 of HetR are

involved in sensitivity to PatS-dependent signals in vivo.

146

Figure 4.1. Sensitivity to PatS-5 of strains with mutant alleles of hetR. (A) Bar graph of the

percentage of cells that are heterocysts in the wild-type strain (PCC 7120) and strains with an

allele of hetR encoding the indicated amino-acid substitution (labeled on horizontal axis) 96 h

with a source of fixed nitrogen (green diagonal bars), and 96 h after removal of combined

nitrogen with (red horizontal bars) or without (solid blue bars) the addition of 1 M PatS-5 to

the medium. Values represent the average of 500 cells from three independent cultures.

Micrographs of strain PCC 7120 (B and C) and UHM167, which contained an allele of hetR

encoding an E254D substitution (D and E), 96 h after removal of combined nitrogen without (B

and D) or with (C and E) the addition of PatS-5 to the medium. Carets indicate heterocysts.

147

Conservative substitutions at HetR residues 250−256 affect heterocyst formation

and sensitivity to PatS and HetN overexpression

The genetic network controlling heterocyst development involves an intricate interplay between

positive and negative effectors. Coordination of heterocyst differentiation by the positive

regulator, HetR, is negatively regulated by two proteins encoding a RGSGR peptide sequence

(PatS-5), PatS and HetN. Heterocyst formation is abrogated upon introduction of plasmids

encoding patS or hetN independently into the wild-type, even during conditions of nitrogen

starvation[13]. Since residues 250-256 of HetR differed in sensitivity to PatS-5, response to full

length PatS and HetN was also evaluated (Table 4.5). Plasmids used for overexpression of patS

or hetN (pDR211 and pDR320, respectively) were introduced into strains encoding conservative

substitutions at HetR residues 250−256, and heterocyst formation was evaluated. Strains

HetR(E254D) and HetR(D256E) continued to form heterocysts when PpetE-patS was introduced.

Similarly, HetR(R250K), HetR(E253D), HetR(E254D) and HetR(D256E) continued to form

heterocysts when PpetE-hetN was introduced.

Identification of overactive alleles of hetR that bypass HetN function, but require a wild-

type copy of HetR for activity

To isolate additional alleles of hetR no longer responsive to regulation by the inhibitors PatS and

HetN, the hypermutator E. coli strain XL1-RED was used to introduce mutations into plasmids

carring PhetR-hetR (pDR138) alone and PhetR-hetR with PpetE-patS (pST103). The library of

mutations in plasmid pDR138 was introduced into a strain that has the native promoter of hetN

replaced by the copper-inducible promoter of petE (strain 7120PN) and a second hetR-deficient

strain that also has hetN driven by the PpeE promoter (strain UHM110; PpetE-hetN ∆hetR) to isolate

alleles of hetR that can overcome HetN inhibitory signals. The library of mutations in plasmid

pST103 was introduced into a hetR-deficient strain to isolate mutations that bypass inhibition by

PatS. A total of seven alleles of hetR were isolated from the XL1-RED mutagenesis studies.

Two substitutions were isolated independently in both the patS and hetN selection experiments.

Two truncated versions of hetR were also isolated with and without another substitution in hetR.

Finally, four plasmids containing an inversion of the multiple cloning sites were isolated,

presumably by a plasmid recombination event. These last four isolates were not pursued further

as they did not contain mutations in hetR.

The substitutions Y51C and P206S (arising from XL1-RED mutagenesis of plasmids carrying

PhetR-hetR), allowed heterocysts to form in an otherwise heterocyst-repressive PpetE-hetN

background. Additional mutations in PhetR-hetR were isolated. One mutation was identified in

148

the promoter region

of hetR. The

mutation at -772

relative to the

translational start site

is located near a

region that appears

to downregulate

expression from the

-728 and -696

transcriptional start

points of hetR when

a source of fixed

nitrogen is

available[20]. The

Y51C mutation was

also isolated with

additional mutations spanning R134-M176 at the C-terminus. Each of the mutations in the coding

region of hetR was isolated multiple times, but the Y51C mutation was the most frequently

encountered. Plasmids engineered with the Y51C and P206S mutations (plasmids pST121 and

pST122, respectively) did not form heterocysts upon reintroduction into the PpetE-hetN ∆hetR or

∆hetR strains. However, when introduced into the wild-type, heterocysts formed at a frequency

greater than that observed with wild-type hetR (on plasmid pDR138). On average, 62.8% for

Y51C, 56.0% for P206S as compared to 26.3% for wild-type hetR at 48 hours post-induction. For

this reason, the Y51C and P206S substitutions appear to require a copy of wild-type hetR in order

to overcome inhibition by HetN. Construction of Y51C and P206S in the chromosome resulted in

0% and 11.9% heterocysts, respectively (Fig. 4.2).

To understand the Y51C mutation of hetR, the conservative mutations Y51F and Y51H were

evaluated. The conservative substitutions Y51F and Y51H on plasmids (pST104 and pST105)

could not be introduced into the wild-type, ∆hetR or PpetE-hetN strains (but a few colonies taken to

be suppressor mutations eventually arose, but only in a PpetE-hetN background). To test if the

conservative mutations produced lethal levels of heterocysts, filaments were evaluated after 24

and 48 hours after conjugation with plasmids pST104 and pST105 (prior to selection with

appropriate antibiotics) for heterocyst formation. Most cells in strain ∆hetR carrying either

mutation on plasmids were heterocyst cells, suggesting that the conservative substitutions (Y51F,

Y51H) produced heterocysts at a frequency that impeded normal cell growth. To further

Figure 4.2. Heterocyst frequencies of strains UHM170 to UHM182

and the wild-type. Percentages represent the average of 500 cells

from three independent cultures of each strain (listed on horizontal

axis) 48 h after removal of combined nitrogen.

149

determine if the lethality of Y51F and Y51H alleles contributed to this phenomenon, these

mutations were recreated in the chromosome. The heterocyst frequency for Y51F and Y51H

(5.5% and 0%, respectively) in the chromosome (Fig. 4.2) was similar to Y51C (0%). It remains

unclear whether secondary site mutations that allow for survival of these strains with Y51C, Y51F

and Y51H substitutions (perhaps by dampening heterocyst frequencies) are responsible for this

phenotype.

A subset of plasmids isolated from the study resulted in the Y51C mutation accompanied with

additional mutations in the 299-residue HetR protein. These mutations were confined to residues

spanning R134 to M176, and did not arise without Y51C on the same XL1-RED-derived plasmid.

This correlation may underlie a function of the C-terminus on the activity of HetR(Y51C) or wild-

type HetR. To explore the role of the C-terminus of HetR in heterocyst development, residues

R134-D296 and M176-D296 were cloned into plasmids under the control of the PpetE promoter

and the effect of this truncated form of hetR on heterocyst frequency was accessed in different

genetic backgrounds. A plasmid with HetR R134-D296 (plasmid pST108) formed heterocysts at

a frequency well below wild-type hetR (carried on plasmid pDR138) in the wild-type and in a

patS-deficient strain. A plasmid with HetR M176-D296 (plasmid pST109) formed fewer

heterocysts in the wild-type compared to pST108 and a low frequency of heterocysts in a patS-

deficient strain. Also, plasmids pST108 and pST109 could not restore heterocyst formation in

∆hetR, pN nor pN∆hetR strains, unlike wild-type hetR carried on plasmid pDR138. The simplest

explanation for these results may relate to the inactivity of these truncations of HetR, a

phenomenon that is not directly dependent on the hetR(Y51C) allele.

Identification of overactive alleles of hetR that bypass HetN function

When a plasmid carrying wild-type hetR (plasmid pDR138) was mutagenized within XL1-RED E.

coli cells and introduced into strain PpetE-hetN ∆hetR, mutations in HetR were isolated that

represented bypass HetN-inhibitory signals. The mutations C48R, E56G, Q59P and D230G on

plasmid pDR138 individually conferred resistance to HetN overexpression, but the Q59P and

D230G substitutions were isolated the most frequently in this mutagenesis study. Unlike the

Y51C and P206 mutations previously described, the activities of these mutations were isolated in

the absence of a chromosomal copy of hetR.

The E56G, Q59P and D230G substitutions are non-conservative and may dramatically alter the

structure of HetR. Conservative substitutions for the E56G, Q59P and D230G mutations in HetR

were created on plasmids to more accurately address the role of these sites on HetR activity.

Compared to plasmids containing the E56G, Q59P and D230G non-conservative mutations, very

150

few colonies arose when plasmids containing HetR(E56D), HetR(Q59E) and HetR(D230E) on

plasmids (pST127, pST128 and pST129, respectively) were introduced into strain ∆hetR. In

addition, even fewer colonies arose when the same conservative substitutions were introduced

into strain PpetE-hetN ∆hetR. Colony formation in these strains was also delayed, suggestive of

secondary site mutations generated in response to the increased activity of the conservative

substitutions in comparison to the non-conservative substitutions. To determine if the

conservative mutations produced lethal levels of heterocysts in either strain, filaments were

evaluated at 24 and 48 hours after conjugation (prior to selection with appropriate antibiotics) for

heterocyst development. After conjugation, heterocyst cells were observed in strain ∆hetR.

However, because only vegetative cells were observed in strain PpetE-hetN ∆hetR, the

conservative substitutions were sensitive to HetN. In contrast, the non-conservative substitutions

produced heterocysts in both background strains under the same conditions, and also generated

a greater number of exconjugants as well. The discrepancy in heterocyst frequencies may

suggest that the non-conservative substitution is more active than the conservative substitution.

To further understand the E56, Q59 and D230 mutations, the corresponding non-conservative

and conservative substitutions were recreated in the chromosome. In strains with E56G and

Q59P substitutions, heterocyst frequency was 0%, but the D230G substitution resulted in 20.5%

heterocysts (Fig. 4.2). The corresponding E56D, Q59E and D230E conservative substitutions

resulted in heterocyst frequencies of 7.1%, 23.7% and 6.9% respectively (Fig. 4.2), suggesting

that only the non-conservative mutation at D230 is more active than the conservative mutation.

However, like the substitutions at the Y51 residue of HetR (Fig. 4.2), it remains unclear whether

heterocyst frequencies arising upon mutation of E56, Q59 and/or D230 are due to secondary site

mutations. Mutations that dampen lethal levels of heterocyst differentiation would allow these

strains to survive, and may explain the mutant phenotypes.

Identification of overactive alleles of hetR that bypass PatS and HetN function

When a plasmid carrying PhetR-hetR and PpetE-patS (plasmid pST103) was mutagenized by the E.

coli strain XL1-RED and introduced into strain ∆hetR, mutations were found in HetR that could

bypass PatS-inhibitory signals. Two of the substitutions found to bypass the activity of HetN were

found to also bypass the activity of PatS (Table 4.4). The mutations at residues Q59P, D230G

and L162I of HetR on plasmid pST103 individually conferred resistance to PatS overexpression.

Thus HetR with changes at residues at Q59 and D230 are resistant to both PatS- and HetN-

mediated repression of HetR, suggesting that the same RGSGR pentapeptide (also referred to as

‘PatS-5’) motif found in both inhibitory proteins may link these two pathways of heterocyst

151

repression. Like the E254G substitution, the Q59P and D230G alleles of hetR present on a

multicopy plasmid also caused nearly complete differentiation of all cells into heterocyst cells.

To determine the resistance of the remainder of the isolated hetR mutants (C48R, Y51C, E56G,

L162I and P206S) to both PatS and HetN, plasmids encoding each allele were individually

introduced into strains 1) PpetE-hetN ∆hetR or 2) ∆hetR, PpetE-hetN ∆hetR and the wild-type. A

plasmid containing the L162I mutation and PpetE-patS was introduced into strain PpetE-hetN ∆hetR.

Heterocysts were able to form, indicating bypass of HetN-overexpression in addition to PatS-

overexpression by the L162I substitution. Thus Q59P, L162I and D230G are less sensitive to

both negative regulators. However, the HetN-bypass mutants of HetR (at residues C48R, Y51C,

E56G, P206S) were sensitive to patS overexpression. Heterocyst development was abrogated

upon introduction of these corresponding plasmids into the different strains and grown in liquid

BG-110 (lacking a source of combined nitrogen) with 1 µM PatS-5 (the minimal inhibitory

concentration, or MIC), 2 µM PatS-5 and 6 µM PatS-5.

Table 4.4. Summary of sensitivity of seven alleles of hetR (driven by PhetR) encoded on plasmids

to PatS and HetN. Conditions include PpetE-patS (carried on derivatives of plasmids pST103), 1

to 6 µM PatS-5, and/or PpetE-hetN (from strains pN or pN∆hetR). ‘Yes’ indicates decreased

sensitivity (“bypass”) as evidenced by heterocyst formation. ‘No’ indicates sensitivity to the

inhibitors (PatS, middle two columns; or HetN, right column) tested as evidenced by the lack of

heterocyst formation. ‘ND’ indicates not determined.

HetR substitution on plasmid

PatS bypass (PpetE-patS on plasmid)

PatS bypass (1-6 µM PatS-5)

HetN bypass (PpetE-hetN in chromosome)

None (wild-type) No No No

C48R ND No Yes

Y51C ND No Yes

E56G ND No Yes

Q59P Yes ND Yes

L162I Yes ND Yes

P206S ND No Yes

D230G Yes ND Yes

Overactive alleles of hetR affect sensitivity to PatS and HetN overexpression

The seven mutations isolated in the bypass selection experiments were recreated in the

chromosome, along with six corresponding conservative substitutions (Fig. 4.2). Heterocyst

frequency of mutations in the chromosome was generally below the frequency observed on a

multicopy plasmid. One explanation for this discrepancy may relate to the presence of

suppressor mutations that arose in these strains to offset the increased lethal activity of these

152

alleles. Alternatively, the normal levels of PatS and HetN in the chromosome differ from the

selection conditions originally used to isolate these strains, since background strains as well as

plasmids that overexpress these inhibitors were used. This may explain sensitivity of all mutant

strains to overexpression of HetN (from plasmid pDR320), even though some of the isolated

alleles of hetR were insensitive to PpetE-hetN overexpression from the chromosome (Table 4.5).

Strains encoding Q59E, D230G, D230E and L162I were resistant to patS-overexpression (from

plasmid pDR211 and pST102), as expected (Table 4.5).

Table 4.5. Summary of sensitivity to different allelic replacements of hetR in the chromosome to

overexpression of PatS and HetN. Conditions include PpetE-patS (carried on plasmids pST102

and pDR211), 1 µM PatS-5, and PpetE-hetN (carried on plasmid pDR320). ‘Yes’ indicates

decreased sensitivity (“bypass”) as evidenced by heterocyst formation. ‘No’ indicates sensitivity

to the inhibitor (PatS or HetN) tested as evidenced by lack of heterocyst formation. ‘ND’ indicates

not determined.

HetR substitution in chromosome

PatS bypass (PpetE-patS on plasmid)

PatS bypass (1 µM PatS-5)

HetN bypass (PpetE-hetN on plasmid)

None (wild-type) No No No

Y51F No ND No

E56D No ND No

Q59E Yes ND No

L162I Yes ND No

P206S No ND No

D230G Yes ND No

D230E Yes ND No

R250K No Yes Yes

L252V No No No

E253D No Yes Yes

E254D Yes Yes Yes

L255V No Yes No

D256E Yes Yes Yes

HetR(mutant)-GFP localization predicts role in HetR turnover

Heterocysts are the source of the inhibitors of differentiation, PatS and HetN. Lateral diffusion of

the inhibitors from source cells results in inhibitor concentration gradients along a filament. In

turn, the inhibitor concentration gradients mediate patterning by promoting the posttranslational

decay of the activator, HetR. Thus concentrations of the activator and inhibitor are inversely

related. Activator decay can be indirectly observed by use of translational fusions of GFP to the

carboxy-terminal end of HetR. Levels of HetR-GFP decrease with increasing distance from

heterocysts due to PatS- and HetN-dependent degradation of HetR-GFP[13].

153

The mutations in hetR discussed previously are less sensitive to either or both PatS- and HetN-

inhibitory signals. The decreased sensitivity is hypothesized to be mediated by the absence of

HetR protein degradation. To determine if PatS- and HetN-dependent degradation is impaired by

the various isolated mutations in HetR, translational fusions of gfp to the different alleles of hetR

were constructed on plasmids, and subsequently introduced into a ∆hetR ∆patA strain.

Visualization of a concentration gradient of HetR-GFP is not expected with the mutant alleles of

hetR because this mechanism presumably allowed for bypass of inhibition by PatS and/or HetN.

Unexpectedly, three different outcomes were observed independent of the promoter used (PpetE

or PhetR). One substitution was chosen to represent each of the three outcomes in Figure 4.3.

HetR(mutant)-GFP corresponding to the R250-D256 region is depicted in Figure 4.3. The

different translational fusions resulted in the observation of: 1) a HetR-GFP concentration

gradient (Fig.4. 3C, 4.3D), 2) the absence of HetR-GFP concentration gradient (Fig. 4.3E, 4.3F),

or 3) HetR-GFP punctate fluorescence in all cells, independent of cell type (Fig. 4.3G, 4.3H).

A concentration gradient similar to wild-type HetR-GFP (Fig. 4.3A, 4.3B and Fig. 4.4A, 4.4B) was

observed upon the independent introduction of plasmids carrying HetR(L162I)-GFP (Fig. 4.3C,

4.3D), HetR(P206S)-GFP, HetR(A251G)-GFP (Fig. 4.4E, 4.4F), HetR(L252V)-GFP (Fig. 4.4G,

4.4H) and HetR(L255V)-GFP (Fig. 4.4M, 4.4N) into strain ∆hetR ∆patA. The A251G and L255V

substitutions appear to have a shorter gradient than the wild-type. Together, these results may

suggest that bypass of inhibitor activity by these alleles may not directly relate to HetR-dependent

degradation.

In contrast, plasmids containing HetR(Y51C)-GFP, HetR(Q59P)-GFP (Fig. 4.3E, 4.3F),

HetR(R250K)-GFP (Fig. 4.4C, 4.4D), and HetR(D256E)-GFP (Fig. 4.4O, 4.4P) resulted in uniform

and increased levels of GFP fluorescence. Heterocysts were rarely observed with plasmids

carrying the Y51C mutation (and C48R mutation, below), and fluorescence was uniform in

vegetative cells. It is unclear if the mutation in patA is responsible for limiting levels of

differentiation promoted by the Y51C (and C48R) overactive allele of HetR. This phenomenon of

elevated HetR-GFP levels accompanied by the absence of heterocyst formation has been

observed in hetF-deficient strains[8]. Overall, these results implicate specific amino acids of HetR

in affecting turnover of HetR. Adjusting HetR levels with respect to inhibitor levels is consistent

with a mechanism to turn on heterocyst formation in Anabaena.

Lastly, plasmids bearing HetR(C48R)-GFP, HetR(E56G)-GFP, HetR(D230G)-GFP (Fig. 4.3G,

4.3H), HetR(E253D)-GFP (Fig. 4.4I, 4.4J) and HetR(E254D)-GFP (Fig. 4.4K, 4.4L) resulted in

punctate fluorescence in heterocysts and adjacent vegetative cells. In addition to this

154

phenomenon, the E253D substitution appeared to result in a shorter gradient than the wild-type.

Finally, over time similar “foci” were occasionally observed with plasmids containing HetR(Y51C)-

GFP, HetR(Q59P)-GFP, HetR(A251G)-GFP, and HetR(L162I)-GFP in the same genetic

backgrounds. Thus HetR(mutant)-GFP accumulated in cells irrespective of inhibitor levels, but

the mutant HetR protein was not stable. The observed foci may relate to a HetR-degradation

pathway. PatS- and HetN-dependent degradation of HetR is presumably mediated by targeting

HetR for degradation. During the differentiation process, conversion of HetR to these

substitutions may signal an unidentified cellular protease to degrade HetR. This may allow for

post-translational temporal regulation of HetR levels during heterocyst development.

This preliminary work was later applied to studies aimed at uncovering cellular proteases

responsible for HetR turnover. However, strains with PpetE-hetR(mutant)-cfp in the chromosome

(generated through plasmids pST376 to pST382; Table 4.2) produced HetR-CFP fluorescence

below the level of detection and were not pursued further (K. L. Hurd, personal communication).

155

Figure 4.3. Localization of representative HetR(mutant)-GFP fusions. Bright-field (A, C, E,

G) and fluorescence (B, D, F, H) micrographs of filaments of strain ∆hetR ∆patA carrying

plasmids bearing translational fusions to hetR-gfp expressed from PpetE at 48 h after

removal of combined nitrogen. Plasmids bearing wild-type hetR (pDR293; A, B), and the

hetR substitutions L162I (pST199; C, D), Q59P (pST195; E, F) and D230G (pST196; G, H)

are shown. Micrographs were acquired using identical microscope and camera settings.

Carets indicate heterocysts.

156

Figure 4.4. Localization of HetR(mutant)-GFP corresponding to the R250-D256 region.

Bright-field (A, C, E, G, I, K, M, O) and fluorescence (B, D, F, H, J, L, N, P) micrographs of

filaments of strain ∆hetR ∆patA carrying plasmids bearing translational fusions to hetR-gfp

expressed from PpetE at 48 h after removal of combined nitrogen. Plasmids bearing wild-

type hetR (pDR293; A, B), and the hetR substitutions R250K (pST200; C, D), A251G

(pST217; E, F), L252V (pST218; G, H), E253D (pST219; I, J), E254D (pST220; K, L),

L255KV (pST221; M, N), and D256E (pST201; O, P) are shown. Micrographs were

acquired using identical microscope and camera settings. Carets indicate heterocysts.

157

Overactive alleles of hetR fail to complement hetF- and patA-deficient strains

The relationship between mutations in HetR that allow for bypass of HetN- and PatS-mediated

lateral inhibition and additional positive regulators of heterocyst differentiation was investigated.

To determine if overactive alleles of hetR could complement the heterocyst deficiency observed

in hetF and patA backgrounds, plasmids carrying the seven mutations responsible for bypass as

well as plasmids carrying conservative substitutions were introduced into strains ∆hetR∆hetF,

∆patA∆hetF and ∆patA. No change to the parent phenotype was observed (data not shown). To

further test if the different alleles of hetR produced lethal levels of heterocysts (resulting in

secondary site mutations that allowed for survival of these strains), filaments were evaluated after

24 and 48 hours after conjugation (without selection with appropriate antibiotics) for heterocyst

formation. Heterocyst cells were not observed under these conditions in the same background

strains (strains ∆hetR∆hetF, ∆patA∆hetF and ∆patA). Taken together, these results suggest that

function of the overactive hetR alleles in patS- and hetN-dependent pathways is independent of

hetF- and patA-mediated pathways.

alr9018, a gene that complemented a mutant of a PpetE-hetN ∆patS strain, is not essential

for heterocyst differentiation

As negative regulators of heterocyst differentiation, inactivation of either patS or hetN results in a

Mch phenotype. Heterocyst frequencies of either strain are increased in comparison to the wild-

type (approximately 20% in the mutant strains and 9% in the wild-type). Simultaneous

inactivation of both inhibitory genes in strain PpetE-hetN ∆patS (UHM100) results in nearly

complete differentiation of heterocysts (approximately 98%)[5]. The lethal level of heterocyst

differentiation in this strain was used to isolate spontaneous suppressor mutations that could

represent genes important in the control of the patS and hetN regulatory circuit. One of the

isolated bypass mutants of PpetE-hetN ∆patS, strain NSM6, was heterocyst-deficient (P.B.

Borthaker, unpublished). The mutation corresponded to the gene alr9018 located on the epsilon

plasmid of Anabaena. Reintroduction of this gene on a plasmid was found to restore the Mch

phenotype to strain NSM6. In addition, presence of alr9018 in multicopy on a plasmid increased

heterocyst frequencies in the wild-type, indicating that alr9018 may represent a novel activator of

heterocyst differentiation (P.B. Borthaker, unpublished). However, a null mutation in alr9018 to

reconstruct the original strain NSM6 carrying a mutation in alr9018 could not be made in earlier

studies (P.B. Borthaker, unpublished).

To understand the role of alr9018 in heterocyst development, alr9018 was inactivated from the

epsilon plasmid. Because heterocysts formed in a wild-type pattern without any discernable

158

temporal delay or alteration in phenotype, all9018 is not essential for heterocyst formation (data

not shown). However this result does not preclude alr9018 from an adjunct role in development

nor a role in stabilization of the lethal PpetE-hetN ∆patS mutant phenotype.

DISCUSSION

The developmental genes hetR, patS and hetN are essential in the coordination of heterocyst

patterning. Mutagenesis experiments were performed in this study to understand the relationship

between HetR and the inhibitors of HetR, the heterocyst-development regulators PatS and HetN.

Seven overactive alleles of hetR were isolated in the mutagenesis experiments described in this

study: C48R, Y51C, E56G, Q59P, L162I, P206S and D230G. Translational fusions of different

alleles of hetR to gfp were used to cytologically demonstrate the resistance of these versions of

HetR to inhibitor-mediated decay (Fig. 4.3). However, only some of the substitutions resulted in

an absence of HetR-GFP concentration gradient. Some of the substitutions corresponding to the

most resistance to PatS and/or HetN overexpression resulted in cells with punctate HetR-GFP

fluorescence. One hypothesis for this observation may relate to impeded protein decay.

Table 4.6. Location of the overactive alleles (far right column) described in this study within the

domains (far left column; corresponding residues in middle column) of the 299-residue protein

HetR.

HetR domains Residues of HetR Overactive HetR substitutions

N-terminal domain 1-98 C48R, Y51C, E56G, Q59P

Flap domain (middle) 99-213 L162I, P206S

Hood domain (C-terminal) 214-296 D230G, R250-D256 region

Mapping of the overactive hetR alleles identified in this study to one of the three distinct domains

in the structure of HetR [3] may elucidate the role of HetR in the coordination of heterocyst

differentiation (summarized in Table 4.6). The N-terminal DNA-binding domain contains highly

conserved residues in the dimerization core (residues 16-51), a region that includes C48R and

Y51C. Because dimerization of HetR appears to be required for heterocyst differentiation[21],

mutations at the dimer interface would presumably affect HetR function. Additional alleles

mapped to other highly conserved regions of HetR. In the flap domain, L162I and P206S

mapped to conserved residues spanning residues 158-168 and 178-212 respectively. In the

hood domain, D230G mapped to a conserved region containing residues 222-236. The flap and

hood domains represent unique protein folds and have been suggested to impart flexibility to the

159

HetR structure, which may be essential for HetR interaction with peptides, proteins, ions and

solvents[3]. The hood domain, positioned on top of the DNA-binding unit present in the N-

terminal domain, forms a cavity. This negatively-charged recess may act as a scaffold necessary

for the assembly of transcriptional machinery, including RNA polymerase[3]. In addition, the two

flap domains may function to enhance the interaction of HetR with DNA to regulate heterocyst

formation. Mutations in these conserved regions may alter the conformation of the flap and hood

domains, ultimately affecting the genetic regulation of genes involved in heterocyst formation.

Although previous reports suggest no interaction between HetR and Anabaena RNA polymerase

using yeast two-hybrid analysis[21], this result may represent a limitation of the assay.

The phenotypes of strains with alleles of hetR encoding substitutions at residues R250,

E253, E254, L255 and D256 are consistent with decreased sensitivity of the substituted

HetR proteins to PatS- and HetN-dependent inhibitory signals. First, both patS and hetN

null mutants as well as the strains with overactive alleles described here differentiate an

increased number of heterocysts relative to that of the wild-type, PCC 7120 [19, 22].

Second, a patS null mutant and the strains described here with overactive alleles

differentiate heterocysts in the presence of a fixed source of nitrogen. And third,

inactivation of both patS and hetN simultaneously results in the formation of a number of

heterocysts in excess of that made by either of the individual mutants [18], similar to the

strains with alleles of hetR encoding E253D, E254D, L255V, and D256E substitutions.

The in vivo studies (Fig. 4.1) with mutant alleles of hetR(250) to hetR(256) led to

biophysical data showing direct binding between HetR and PatS-5[9]. Two different and

independent techniques corroborated that PatS-5 binds directly to HetR, namely EPR

spectroscopy and isothermal titration calorimetry. The discovery of the 250-256 region

led to 1) evidence that PatS binds directly to HetR from cysteine scanning mutagenesis

combined with continuous wave (CW) electron paramagnetic resonance (EPR)

spectroscopy; 2) determination of the stoichiometry of PatS-5 binding to HetR and that

PatS-5 binds HetR as a dimer from double electron electron resonance EPR

experiments; and, finally, 3) corroboration that PatS-5 binds directly to HetR and

measurement of the dissociation constant for HetR binding to PatS-5 using isothermal

titration calorimetry.

160

REFERENCES CITED


12: p. 30-39.

2. Buikema, W.J. and R. Haselkorn, Characterization of a gene controlling heterocyst

development in the cyanobacterium Anabaena 7120. Genes Dev., 1991. 5: p. 321-330.

3. Kim, Y., et al., Structure of the transcription factor HetR required for heterocyst

differentiation in cyanobacteria. Proc. Natl. Acad. Sci., 2011. 108: p. 10109-10114.



Bacteriol., 2007. 189: p. 2460-2467.



Microbiol., 2005. 57: p. 111-123.












p. 4473-4482.






Sci U S A, 2009. 106(47): p. 19884-8.


involved in the control of heterocyst development in Anabaena. Mol. Microbiol., 1993. 9:

p. 77-84.

161


HetR function in the heterocystous cyanobacterium Anabaena sp. strain PCC 7120. J

Bacteriol, 2007. 189(6): p. 2460-7.

16. Frías, J.E., E. Flores, and A. Herrero, Nitrate assimilation gene cluster from the

heterocyst-forming cyanobacterium Anabaena sp. strain PCC 7120. J Bacteriol, 1997.

179(2): p. 477-86.

17. Khudyakov, I.Y. and J.W. Golden, Different functions of HetR, a master regulator of

heterocyst differentiation in Anabaena sp. PCC 7120, can be separated by mutation.

Proc Natl Acad Sci U S A, 2004. 101(45): p. 16040-5.


differentiation in the filamentous cyanobacterium Anabaena sp. PCC 7120. Mol Microbiol,

2005. 57(1): p. 111-23.

19. Yoon, H.S. and J.W. Golden, Heterocyst pattern formation controlled by a diffusible

peptide. Science, 1998. 282(5390): p. 935-8.

20. Rajagopalan, R. and S.M. Callahan, Temporal and spatial regulation of the four

transcription start sites of hetR from Anabaena sp. strain PCC 7120. J Bacteriol, 2010.

192(4): p. 1088-96.



Acad. Sci. U.S.A., 2004. 101(14): p. 4848-4853.


pattern in Anabaena sp. PCC 7120. Mol Microbiol, 2001. 40(4): p. 941-50.

162

CHAPTER 5. LOCALIZATION OF THE INTERCELLULAR SIGNALING PROTEIN, HETN, IN

ANABAENA SP. STRAIN PCC 7120

INTRODUCTION

Intercellular signaling regulates a variety of processes in bacteria, most notably as part of

quorum-sensing systems. In quorum sensing, individual cells communicate with short peptides or

other small molecules that are secreted into the surrounding medium to coordinate the behavior

of multiple cells. As a counterpoint to quorum sensing, signaling between cells in direct physical

contact has the potential to employ intercellular signals that pass directly between cells, without

an intermediate extracellular stage. Examples include signaling during the formation of transient

multicellular structures, such as the development of spores in Bacillus subtilis and in the fruiting

bodies elaborated by Myxococcus xanthus, or in bacteria that grow as multicellular organisms,

such as filamentous streptomycetes and cyanobacteria. In the latter of these examples, inhibitors

of cellular differentiation in filamentous cyanobacteria are thought to move from cell to cell and

govern the periodic patterning of heterocysts that form in response to deprivation of fixed nitrogen.

Anabaena sp. strain PCC 7120 is a filamentous cyanobacterium that can be induced to form a

periodic pattern of nitrogen-fixing heterocysts from an unbranched chain of undifferentiated

vegetative cells. Heterocysts occur, on average, at 10-cell intervals, are terminally differentiated,

and differ from vegetative cells morphologically, metabolically, and genetically [1, 2]. They have a

single known function, to supply the remaining vegetative cells of the filament with a bioactive

form of nitrogen, which allows growth of the organism in nitrogen-poor environments. The

elaboration of heterocysts facilitates the spatial separation of an oxygen-labile metabolic process,

nitrogen fixation, from one that evolves molecular oxygen, photosynthesis with photosystem II.

Fixed nitrogen is supplied to vegetative cells from heterocysts, and in return, heterocysts receive

a source of carbon and reductant to compensate for their lack of PS II and the Calvin cycle [3, 4].

Experimental evidence for the phenomenon of “lateral inhibition” of differentiation to explain how

Anabaena “counts to 10” was first described by Wolk in 1967 [5, 6]. More recently, patterning of

differentiation has been shown to be dependent on an activator of differentiation, HetR, and two

inhibitors of differentiation, PatS and HetN, that have properties of intercellular signaling

molecules. HetR is at the center of a regulatory circuit that shares the properties of biological

switches, which turn graded input signals into a binary output: when the switch is “off”, the cell

remains undifferentiated, but when the switch is “turned on”, the differentiation process begins

and eventually becomes irreversible and self-sustaining [7, 8]. HetN or PatS produced in one cell

lowers levels of HetR in adjacent cells [9], and it has been proposed, but not demonstrated, that

the relative positions of cells is conveyed by concentration gradients of PatS and/or HetN

extending from source cells [10, 11].

163

HetR has prominent roles in both the patterning and differentiation of heterocysts. In a wild-type

genetic background, expression of hetR is both necessary and sufficient for differentiation. Its

inactivation prevents development of heterocysts, and over-expression of hetR leads to the

formation of multiple contiguous heterocysts (Mch phenotype), even in the presence of

concentrations of nitrate or ammonia that suppress differentiation in the wild-type [12, 13]. HetR

has DNA-binding activity, which is necessary for heterocyst formation [14], suggesting that its

primary role in promoting differentiation is the regulation of transcription of other developmental

genes. The protein has also been shown to bind directly to the promoter regions of 5 genes, patS,

hepA, hetP, pknE and hetR [14-16]. Induction of hetR by HetR is an example of direct positive

auto-regulation, predicted by patterning models, and the mutual dependence of ntcA, which is

also necessary for differentiation, and hetR creates a second, indirect example of positive auto-

regulation [17]. The recent crystal structure of HetR reveals a dimer with a central DNA-binding

region, two outward facing flaps, and a hood-like domain over the core of the protein [18]. It was

suggested that HetR might serve as a scaffold for the assembly of transcription initiation

complexes.

The patS gene is predicted to encode 17 amino acid protein that is presumably processed to a

smaller, active form, perhaps during export from the cytoplasm to the periplasm of the cell or

directly to a neighboring cell [11]. The active form of PatS has yet to be identified, and a

presumed gradient of PatS has not been demonstrated. However, when confined to the

cytoplasm of the cell that produced it via a translational fusion to GFP, PatS is incapable of

restoring a normal pattern of heterocysts to a patS-mutant strain, suggesting that it must diffuse

from cell to cell to function properly [19]. A synthetic peptide corresponding to the predicted C-

terminal 5 amino acids of PatS (PatS-5; RGSGR) prevents DNA-binding activity of HetR in vitro

[14], prevents differentiation of heterocysts when added to the medium [11], and was recently

shown to bind directly to HetR with a 1:1 stoichiometry [20]. In a genetic selection to identify

residues of the 17 amino acid peptide that are necessary for activity of PatS, mutations were

found only in the RGSGR sequence, with mutations resulting in substitution of the underlined

residues reducing activity. A patS-null mutant has reduced spacing between heterocysts, and

adjacent cells often differentiate, the hallmark of the Mch phenotype (multiple contiguous

heterocysts). Transcription of patS is initially up-regulated in contiguous groups of cells, as shown

by a patS-gfp transcriptional reporter fusion, but between 10 and 12 hours after induction, at

about the same time that cells are irreversibly committed to differentiation, fluorescence has been

resolved to single cells in a pattern that predicts the eventual pattern of heterocysts along

filaments [21].

Interactions between PatS and HetR appear to control formation of the initial pattern of

differentiation, but HetN is necessary for stabilization of the initial pattern and prevents

164

differentiation of cells adjacent to existing heterocysts. It seems likely that it is also involved in

maintenance of the pattern as filaments lengthen. Several cell-generation times after the

formation of an initial Mch pattern by a patS-null mutant, the pattern normalizes to a wild-type-like

pattern [21], suggesting the presence of other patterning factors. In contrast, a patS mutant

conditionally lacking expression of hetN forms an Mch pattern initially like that of the patS mutant,

but the pattern fails to normalize over time, and eventually most cells of the filament differentiate

to form heterocysts [22]. A hetN single mutant has a delayed Mch phenotype; the pattern is

initially wild-type at 24 h after induction but is Mch after 48 h, approximately twice the time from

induction to the formation of an initial pattern of mature heterocysts [10]. The inhibitory activity of

HetN, as for that of PatS, includes blockage of the positive auto-regulation of hetR [10, 23].

The 287 amino-acid HetN protein is predicted to be a reductase in the short chain alcohol

dehydrogenase (ADH) family [24]. Within the sequence of HetN is the PatS-5 sequence, RGSGR,

at positions 132-136, raising the possibility that this same sequence is responsible for the

inhibitory activity of HetN. Experiments showing that the PatS- and HetN-dependent pathways

converge at or just before HetR and the identification of a point mutation in hetR that confers

resistance to suppression by PatS and HetN suggests that the RGSGR sequence of HetN may

be critical for its activity as it is in PatS [22, 25]. The RGSGR motif of HetN and not the predicted

ADH reductase activity is necessary for patterning of heterocysts[26]. When grown with

combined nitrogen, all cells of filaments have a low level of HetN in thylakoid and cytoplasmic

membranes, but upon nitrogen step-down, HetN is degraded [23]. After formation of

proheterocysts, HetN protein and expression of hetN is found exclusively in proheterocysts and

heterocysts. Localization of HetN-YFP fluorescence to the cell periphery is demonstrated in this

study.






for photomicroscopy were as previously described [27]. Images were processed in Adobe

Photoshop CS2 and ImageJ.




165

Plasmid pDR386 is a mobilizable shuttle vector based on pAM504 containing PpetE-hetR

translationally fused to YFP. This plasmid was used to make various HetN-YFP translational

fusion plasmids by replacing hetR with derivatives of hetN. Plasmid pDR386 was derived from

pDR293 [9] by replacing GFP with YFP. The coding region of YFP was amplified via PCR using

pUC57-PS12-yfp [28] as a template with primers YFP-TNL-F and YFP-SacI-R. The PCR product

was cloned as a SmaI-SacI fragment into pDR293 to replace gfp.

Plasmid pRR159 is a mobilizable shuttle vector containing PhetN-hetN translationally fused to YFP.

The insert was amplified via PCR using PCC 7120 chromosomal DNA as template with primers

PhetN-BamHI-Fwd and hetN-SmaI-R and cloned into pDR386 as a BamHI-SmaI fragment.

The plasmids pRR161, pRR162, pRR163 and pRR167 are mobilizable shuttle vectors containing

PpetE-hetN(1-46), PpetE-hetN(1-172), PpetE-hetN and PpetE-hetN(1-192), respectively, translationally

fused to YFP. The inserts were amplified via PCR using pDR320 as template with primers

PpetE-BamHI-F and hetN(1-46)-SmaI-R for plasmid pRR161, PpetE-BamHI-F and hetN(1-172)-

SmaI-R for plasmid pRR162, PpetE-BamHI-F and hetN-SmaI-R for plasmid pRR163, and PpetE-

BamHI-F and HetN(1-192)-SmaI-R for plasmid pRR167. The PCR products were cloned into

pDR386 as BamHI-SmaI fragments.

Plasmids pCO105, pCO107, pCO108, and pCO113 are shuttle vectors based on pAM505

carrying hetN encoding codon deletions Δ2-46, Δ47-176, Δ177-195, and Δ47-128, respectively,

transcriptionally fused to the petE promoter. The fragments used to make the hetN variants were

amplified via overlap extension PCR. The outside primers used to create the constructs were

hetN-EcoRI and hetN-BamHI-R. The following are the internal primers used: hetN-(Δ2-46)-F and

hetN-(Δ2-46)-R for pCO105; hetN-(Δ47-176)-F and hetN-(Δ47-176)-R for pCO107; hetN-(Δ177-

195)-F and hetN-(Δ177-195)-R for pCO108; and hetN-(Δ47-128)-F and hetN-(Δ47-128)-R for

pCO113. The overlap extension PCR products were cloned as EcoRI-BamHI fragments into

pKH256.

Plasmids pCO115, pCO117, pCO118, and pCO121 are mobilizable shuttle vectors containing

PpetE-hetN(Δ2-46), PpetE-hetN(Δ47-176), PpetE-hetN(Δ177-195), and PpetE-hetN(Δ47-128),

respectively, translationally fused to YFP. The fragments used to make the mutant alleles of hetN

were amplified via PCR using pCO105, pCO107, pCO108, and pCO113 as templates,

respectively. The primers used to create the constructs were PpetE-BamH1-F and hetN-Sma1-R.

The PCR products were cloned into pDR386 as BamHI-SmaI fragments to generate the fusions

with YFP.

166

Confocal microscopy

Cells were routinely viewed and imaged as described previously [22]. Confocal microscopy was

performed using an Olympus Fluoview 1000 laser scanning confocal mounted on an IX81

motorized inverted microscope. Fluorescence from Turbo-YFP was detected with an excitation of

525 nm and an emission of 538 nm. All images were processed in Adobe Photoshop CS2. To

estimate levels of fluorescence in parts of images, a two-dimensional plot profile of pixel

intensities along a line drawn on the image was constructed in ImageJ.

Mutagenesis studies

The hypermutator Escherichia coli strain XL1-RED was used to create a library of mutations

within plasmids pSMC115 in a method described previously[27] and subsequently introduced into

the wild-type via conjugation.

Table 5.1. Bacterial strains used in Chapter 5

Anabaena sp. strain



UHM150 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(R223W)

[29]

UHM163 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(R250K)

[30]

UHM169 hetR-deletion strain with chromosomal PhetR-hetR replaced by PhetR-hetR(D256E)

[30]

167




r Neo

r

[31]

pSMC115 pAM504 with PpetE-hetN [10]

pDR320 pAM504 with PpetE-hetN [32]

pDR386 pAM504 with PpetE-hetR-yfp This study

pRR159 pAM504 with PhetN-hetN-yfp This study

pRR163 pAM504 with PpetE-hetN-yfp This study

pRR161 pAM504 with PpetE-hetN(1-46)-yfp This study



pCO115 pAM504 with PpetE-hetN(∆2-46)-yfp This study





Primer no.


1 YFP-TNL-F ATATACCCGGGGATCGGCGTCAGCTAGCAGCGGCGCCTTGCTGTTC

2 YFP-SacI-R ATATAGAGCTCTCAGCTGGTGTCTCCGGAAC

3 PhetN-BamHI-Fwd ATATAGGATCCAGGAGAAGACGCGATGAATC

4 hetN-SmaI-R ATATACCCGGGATGAGCGATGAGACTCAACAG

5 PpetE-BamHI-F ATATAGGATCCCTGAGGTACTGAGTACACAG

6 hetN(1-46)-SmaI-R ATATACCCGGGACGTTTGGGCTAATCCTGATTG

7 hetN(1-172)-SmaI-R GGATCCTTATAATTTTTCTCTTTCTGTCAAAGCGGCGCGTGCG

8 HetN(1-192)-SmaI-R ATTATCCCGGGACCCAGTTTGCGAGACATAGCC

9 hetN-(Δ2-46)-F GGTTACAATGTGTAATGCGGTTAAGGCTGC

10 hetN-(Δ2-46)-R CCGCATTACACATTGTAACCTGCTAGTCTCTGTTGACAGCCGTTTGGGCTAATCCTGATTG

11 hetN-(Δ47-176)-F AGCCCAAACGGGTGTCAACATTTCGGTGGTTTG

12 hetN-(Δ47-176)-R TGTTGACAGCCGTTTGGGCTAATCCTGATTG

13 hetN-(Δ177-195)-F AGTTGGTACTGATACTCGTGTCTCTGCGCCCACGAGTATCAGTACCAACTAATTCCTGAC

14 hetN-(Δ177-195)-R CACGAGTATCAGTACCAACTAATTCCTGAC

15 hetN-(Δ47-128)-F AGCCCAAACGATGATGGAACGCGGTAGTG

16 hetN-(Δ47-128)-R GTTCCATCATCGTTTGGGCTAATCCTGATTG

168

RESULTS

HetN-YFP localizes to heterocysts in stains expressing overactive alleles of hetR

The 287-residue HetN

protein can be divided into

4 domains based on

hydrophobicity (Fig. 5.1).

An N-terminal

hydrophobic domain

(shown in gray), which

may serve as a leader

peptide, was found to consist of amino acids 1 – 46 (N-terminal hydrophobic domain). An N-

terminal hydrophilic domain containing the RGSGR motif and putative catalytic triad was found at

amino acids 47 – 176 (N-terminal hydrophilic domain). A second hydrophobic domain (shown in

gray), which has a length consistent with that of a central transmembrane domain, was found at

amino acids 177 – 195 (central hydrophobic domain). Lastly, a C-terminal hydrophilic domain

was found at amino acids 196 – 287 (C-terminal domain). The presence of two hydrophobic

domains is consistent with association of HetN to the cell membrane. In addition, HetN has been

detected in immunoblot experiments using purified thylakoid and plasma membranes[23].

Cytological evidence of the localization of HetN is hampered by the role of this protein in the

regulation of heterocyst development. Translational fusions of HetN to YFP (yellow fluorescent

protein) inhibit the formation of heterocysts (Fig.5.2A, 5.2B). For this reason, localization of HetN-

YFP to vegetative cells and heterocyst cells cannot be determined in the wild-type (Fig. 5.2A,

5.2B). To determine the localization of HetN-YFP in both cell types, variants of HetR less

sensitive to inhibition by PatS- and HetN-mediated inhibition of heterocyst differentiation were

used. Because the R223W, R250K, D256E forms of HetR differentiate heterocysts even when

HetN is overexpressed or present in multicopy (shown in Figure 5.2) [30, 33], fusion to YFP was

hypothesized to allow for visualization of HetN-YFP in both cell types. In strains with the

conservative substitutions R250K and D256E (UHM163 and UHM169, respectively), expression

of hetN-yfp fluorescence from the native hetN promoter was concentrated uniformly in

heterocysts (Fig.5. 2, C-F). In vegetative cells, fluorescence was below the level of detection

(Fig. 5.2, C-F). Thus the localization of HetN-YFP differs in vegetative cells and heterocysts.

Heterocysts did not form in a strain with the R223W substitution (data not shown). Introduction of

the empty vector plasmid into these strains did not result in any observable fluorescence, as

expected (data not shown). Finally, the formation of heterocysts in strains UHM163 and

Figure 5.1. Schematic of the domains present in the HetN protein.

169

UHM169, despite elevated levels of hetN on a plasmid, further confirms the overactive nature of

the R250K and D256E alleles of hetR.

Localization of HetN-YFP to the cell periphery

Further studies were done to examine the localization of HetN-YFP. Fluorescence from full-

length HetN fused at its C-terminus to YFP and driven by the petE promoter was observed

primarily at the periphery of cells in filaments using confocal microscopy (rather than fluorescence

microscopy), consistent with localization to the plasma membrane (Fig. 5.3A). In addition, a more

diffuse fluorescence signal was observed from the interior region of cells, consistent with previous

detection of HetN in thylakoid membranes. This result differs from previous studies

demonstrating the uniform concentration of HetN-YFP fluorescence in heterocysts (Fig. 5.2).

However, this discrepancy can be attributed to (i) the use of confocal microscopy (Fig. 5.3), which

achieves greater contrast and resolution than traditional fluorescence microscopy (Fig. 5.2), and

(ii) expression of hetN-yfp from the native hetN promoter in previous studies (Fig. 5.2).

Figure 5.2. Localization of HetN-YFP. Bright-field (A, C, E) and fluorescence

microscopy (B, D, F; false-colored green) of PCC 7120 (A, B), strain

UHM163 (C, D) and strain UHM169 (E, F) with a plasmid carrying PhetN-hetN-

yfp 48 hours after removal of combined nitrogen. Micrographs were acquired

using identical microscope and camera settings. Carets indicate heterocysts.

170

Again, since the fusion protein prevented the formation of heterocysts, it was not possible to view

the location of HetN in heterocysts in a wild-type genetic background. To facilitate visualization of

HetN-YFP in heterocysts, plasmid-borne hetN-yfp fusions were put into a derivative of PCC 7120

that continues to form heterocysts even when extra copies of hetN are introduced [20]. This

strain, UHM163, has the wild-type copy of hetR replaced by an allele that encodes a R250K

substitution. With transcription of hetN-yfp from either the native promoter or the petE promoter,

fluorescence from YFP was seen primarily at the periphery of heterocysts (Fig. 5.3B, 5.C). With

expression from the petE promoter, fluorescence was seen in both cell types, whereas with the

native hetN promoter fluorescence was seen only in heterocysts, consistent with heterocyst-

specific expression of hetN [10] and previous studies (Fig 5.2, C-F).

C-terminal translational fusions to YFP were also made with the following variants of HetN:

HetN(1-47), HetN(1-176), HetN(1-192), HetN(∆2-46), HetN(∆47-128), HetN(∆47-176), and

HetN(∆177-195), where the numbers in parentheses represent either the portion of HetN in the

fusion or, if preceded by a ∆ symbol, a protein lacking the amino acids indicated. Heterocyst

differentiation in a wild-type background was suppressed with all fusions in which the RGSGR

motif was present. However, with one exception, no fluorescence from YFP was observed. In

filaments with the construct encoding HetN(1-192)-YFP, which lacks the C-terminal domain of

HetN, YFP-dependent fluorescence was observed primarily at cell junctions (Fig. 5.3D). In

particular, rings of fluorescence circumscribed cell septa. Fusions were made with expression

from the native promoter of hetN or the petE promoter, but fluorescence was observed only when

the petE promoter was used. Rings of fluorescence were detected in vegetative cells but absent

from heterocysts. In addition, HetN-YFP-dependent fluorescence with both full-length HetN and

the C-terminal deletion was reduced from the cytoplasm of heterocysts relative to that from

vegetative cells (Fig. 5.3C, D). In a wild-type genetic background, the fusion protein inhibited

differentiation, so fluorescence could only be observed in vegetative cells (data not shown).

When the background strain was UHM163, which forms heterocysts even when hetN is over-

expressed ectopically, rings of fluorescence similar to those in the wild-type were observed in

vegetative cells but lacking in heterocysts (Fig. 5.3D).

171

Figure 5.3. Localization of HetN-YFP using confocal microscopy. Confocal micrographs of

PCC 7120 with a plasmid carrying PpetE-hetN-yfp (A). From left to right: bright-field, chlorophyll

autofluorescence, HetN-YFP fluorescence, and composite image. Arrows indicate areas

showing the lack of overlap between autofluorescence and HetN-YFP at the periphery of cells.

UHM163, a strain that forms heterocysts even when hetN is in extra copy or overexpressed, with

a plasmid carrying PpetE-hetN-yfp (B). UHM163 with a plasmid carrying PhetN-hetN-yfp (C). Top

panels are bright-field images and bottom panels are corresponding fluorescence images

corresponding to a 0.2 m section of the filament for (B) and (C). UHM163 with plasmid carrying

PpetE-hetN(1-192)-yfp (D). Bright-field (top-left) and fluorescence (top-right) images are shown.

Rings of fluorescence indicated by arrows are apparent when the three-dimensional image at

top-left is rotated 50 degrees horizontally. A 7x and 3x digital zoom on the confocal was used

for (A) and (D), respectively. Carets indicate heterocysts.

172

hetN mutagenesis studies to identify necessary sites for function

A positive selection was performed to identify necessary sites for HetN function. The

hypermutator E. coli strain XL1-RED was used to introduce mutations onto a plasmid carrying

PpetE-hetN. Overexpression of hetN in the wild-type using this plasmid inhibits heterocyst

development even in the absence of a source of fixed nitrogen. Mutations in hetN that allow the

wild-type to grow under these conditions were hypothesized to represent sites that functionally

suppress HetN activity. However, plasmids isolated from this study did not have a mutation in

hetN nor the PpetE promoter. Instead, all plasmids isolated from the study contained a sequence

with homology to a transposon. The sequence may have transposed onto the plasmid after

transformation into the hypermutator strain. The event appeared to have subsequently biased

the selection against the desired mutations in hetN. This finding may explain why a greater than

expected number of colonies arose under the selection conditions.

DISCUSSION

HetN and PatS are required for different stages of heterocyst patterning. PatS is necessary for

formation of the initial pattern of cells in a filament composed exclusively of vegetative cells that

will differentiate after a transition to conditions that require diazotrophy for growth. A hetN-mutant

strain is capable of forming this de novo pattern whereas a patS mutant is not [10, 11]. HetN is

necessary for stabilization of this initial pattern. The wild-type pattern of heterocysts initially

formed in a hetN-mutant turns Mch shortly thereafter [10]. Genetic and mathematical modeling

evidence relating to placement of additional heterocysts between existing ones as vegetative

cells divide to maintain the pattern and ratio of heterocysts to vegetative cells suggests that a

HetN-dependent signal produced in mature heterocysts is responsible for specification of a region

of cells at the midpoint between two heterocysts, and that PatS resolves this region to a single

cell that will differentiate [11, 34]. Despite the different roles of PatS and HetN in patterning, the

RGSGR motif is essential to the function of both. In over-expression studies, conservative

substitutions in any of the five amino acids reduced the ability of HetN to suppress differentiation.

Alteration of either of the two arginines appeared to eliminate suppression by HetN, even though

the gene for HetN was expressed at a level higher than that experienced in the wild-type

organism. These results are comparable to those found with substitutions in PatS. In this case, a

genetic selection identified four of the five amino acids of the RGSGR motif as necessary for

suppression of differentiation [11]. Whether or not substitution of the fifth has the same effect is

unknown. To take the analysis of HetN further, a single nucleotide or two corresponding to a

codon in the RGSGR motif was changed in the chromosome of the wild-type. Each of the strains

with an allele of hetN encoding a conservative substitution in one of the amino acids of the

173

RGSGR motif resembled the hetN-deletion strain more than the wild-type; the number of cells

that differentiated was higher than in wild-type and instead of a periodic pattern of single

heterocysts, the Mch phenotype was observed. As for the over-expression studies, substitution

of the two arginines had the most dramatic effect, producing strains with phenotypes

indistinguishable from that of the hetN-deletion strain. In contrast, conservative substitutions in

the putative catalytic triad of ADH activity had no effect on the function of HetN in suppression of

differentiation or patterning of heterocysts. The RGSGR motif of HetN and not predicted ADH

reductase activity was necessary for patterning of heterocysts.

Three previous studies have addressed the quality of HetN necessary for its ability to suppress

heterocyst differentiation, with varied interpretations. In the first, mutation of residues in the

RGSGR sequence of HetN was reported to have no effect on the ability of HetN to suppress

differentiation when the corresponding alleles of hetN were over-expressed [23]. In addition to

R132K and R136L substitutions, the report claims that G134S and S135D substitutions were

made and had no effect. However, this is confusing because the HetN sequence is S134 and

G135, not G134 and S135 as indicated, so what substitutions were actually made is unclear. In

the second study, substitution of one of the three amino acids in the predicted catalytic triad of

ADH activity in HetN prevented suppression of heterocyst differentiation by over-expression of

the corresponding mutant allele [35]. However, substitution of the other two had no effect. In the

another study, over-expression of hetN was shown to cause post-translational decay of HetR,

which presumably contributes to suppression of heterocyst formation. When the RGSGR motif

was replaced by RGDAR, over-expression of the corresponding allele of hetN did not lower levels

of HetR in filaments nor did it suppress differentiation [9], suggesting that the RGSGR motif is

necessary for suppression. The most recent study [26] is the most comprehensive with respect

to the regions of HetN that are necessary for inhibition and patterning. It clearly shows that the

RGSGR motif of HetN and not the predicted ADH reductase activity is necessary for patterning of

heterocysts.

The RGSGR motif as the primary functional group of HetN and PatS in patterning is consistent

with known activities of HetN and PatS. In electrophoretic mobility shift assays the synthetic

RGSGR peptide prevents the binding of HetR to a region of hetR-promoter DNA that includes the

autoregulated transcriptional start point at position -271, and over-expression of hetN or patS

prevents transcription from this same tsp [8, 14]. Addition of RGSGR peptide to medium causes

posttranslational decay of HetR protein in whole filaments, as does over-expression of hetN or

patS in all cells of filaments [9]. Over-expression of hetN or patS in individual cells of filaments

also causes decay of HetR levels in adjacent groups of cells, suggesting that the HetN- and PatS-

dependent signals that diffuse from cell to cell contain the RGSGR sequence.

174

Individual deletions of the predicted signal sequence, membrane-spanning hydrophobic domain,

and C-terminal domain had no effect on suppression of differentiation or patterning[26].

However, two pieces of evidence suggest that it is not solely the RGSGR motif of HetN that is

necessary for normal suppression and patterning of heterocysts. First, it has been reported that a

K159E substitution prevents the ability of HetN to suppress differentiation when the gene

encoding the protein is over-expressed [35]. This non-conservative substitution of a positively

charged amino acid with a negatively charged one, unlike the conservative K159R substitution

used in a later study, apparently alters HetN sufficiently to prevent it from suppressing formation

of heterocysts. The second piece of evidence is the formation of fewer heterocysts by a strain

with the wild-type hetN replaced with an allele encoding a deletion of amino acids 47 – 128, which

would lack the N-terminal hydrophilic domain up to three amino acids before the RGSGR

sequence. The resulting protein appears to lead to suppression of differentiation of cells in a

larger region of a filament than does the wild-type protein. Possible explanations of this

enhanced range of lateral inhibition include a more active form of the protein, a longer half-life, an

increased rate of diffusion or increased production of the mature form of HetN that presumably is

transferred between cells.

Despite sharing a functional protein motif, the forms of HetN and PatS that diffuse from cell to cell

and presumably interact with HetR may be different and remain unknown. Increased expression

of hetN in cells that will become heterocysts around the time of commitment to differentiation is

very different from that of patS, which occurs in groups of cells very soon after induction of

filaments to differentiate by growth-limiting levels of fixed nitrogen in the medium [21, 36]. The

difference in expression patterns reflects the respective roles of HetN and PatS in patterning.

The involvement of two separate proteins with a similar functional motif, rather than one protein

with two modes of expression, suggests that intrinsic differences in the mature HetN- and PatS-

dependent signals play roles in patterning. Recent mathematical modeling of heterocyst pattering

with two inhibitors having different rates of diffusion is consistent with this notion [34].

HetN has been reported to be a membrane protein that resides in both cytoplasmic and thylakoid

membranes. In this study, fluorescence from HetN-YFP was observed primarily in the cell

envelope, presumably in the cytoplasmic membrane, with a more diffuse, lower level of

fluorescence from the interior of the cell, suggesting that the concentration of HetN is higher than

that in thylakoid membranes. HetN-YFP-dependent fluorescence intensity in the cell membrane

was about twice that from thylakoid membranes, but given the much larger surface area of

thylakoid membranes, the majority of HetN in a cell is likely to be in thylakoid membranes,

consistent with earlier Western blot analysis [23]. The only fusions to YFP in this study for which

175

fluorescence was observed were those where YFP either replaced or was at the end of the C-

terminal domain, suggesting that this domain of the protein is located in the cytoplasm. Folding of

GFP is efficient only in the cytoplasm of bacterial cells, which has been used to map the topology

of inner-membrane proteins [37]. The YFP used in the fusions is a derivative of GFP and would

be expected to behave in a similar fashion. If the C-terminal domain is in the cytoplasm, the N-

terminal hydrophilic domain, which contains the RGSGR motif, may be located in the periplasm if

the central hydrophobic domain at amino acids 177 – 195 spans the cytoplasmic membrane.

However, it is difficult to interpret the lack of fluorescence with fusions to other parts of HetN,

which could be explained by the creation of an unstable protein with a short half-life.

Lateral inhibition implies the intercellular transfer of a signal that suppresses differentiation.

Based on the deletion studies, the essential part of HetN that comprises the suppression signal

appears to be little more than the RGSGR motif. This raises the possibility that the signal that

diffuses from cell to cell may be a peptide processed from the full protein. Location of HetN in the

cytoplasmic membrane suggests three potential routes of transfer between cells. First, the

periplasm of cells in filaments is contiguous, and there is evidence both for and against the

diffusion of proteins between cells via the periplasmic space [38, 39]. Processing of the N-

terminal domain after insertion in the membrane could release a soluble fragment of HetN that

contains the RGSGR motif and diffuses through the contiguous periplasmic space. Uptake into

the cytoplasm would be necessary to allow interaction with HetR. Inhibition of heterocyst

differentiation by addition of synthetic RGSGR peptide to the medium [11] suggests that transport

into the cytoplasm is possible for such a molecule or that it can act from the periplasm. The

second potential route of transfer is direct exchange between the membranes of adjacent cells.

Although the cytoplasmic membranes at cell septa are not shared by adjacent cells and so are

not continuous between cells, they are in close proximity [40], and could be bridged by an as yet

uncharacterized intermembrane transport system. The HetN-dependent patterning signal does

not appear to travel from cell to cell via one of these two potential routes because localization of

HetN to the membrane is not required for proper patterning of heterocysts. Deletion of the

putative transmembrane or signal sequence prevented HetN-YFP from localizing to the

membrane and had no effect on heterocyst patterning when deleted from the chromosomal copy

of hetN. In its simplest form, transfer of the RGSGR-containing portion of HetN via the

cytoplasmic membrane or the periplasmic space would likely require localization to the

membrane. The most likely route of transfer is via inter-cytoplasmic exchange mediated by SepJ,

FraC, and/or FraD, which are located at intercellular septa. Intercellular transfer of the

fluorescent molecular tracer calcein occurs in the wild-type, but is impaired in strains lacking one

of the three proteins [41, 42]. Calcein has a molecular weight of 623 Da, slightly more than that

of RGSGR peptide. An RGSGR-containing peptide could be processed from a subpopulation of

176

HetN prior to insertion into the membrane and be transferred via inter-cytoplasmic pores.

Deletion of the putative transmembrane and signal sequence domains of HetN-YFP prevented

not only localization to the membrane, but also detectable fluorescence. The lack of fluorescence

is likely attributable to proteolytic digestion of a protein that cannot fold properly. Processing of

an RGSGR-containing peptide from HetN in the cytoplasm prior to proteolytic digestion would be

consistent with retention of function of these truncated forms of HetN even in the absence of an

accumulation of HetN-YFP that can be detected by fluorescence microscopy. In addition,

preliminary results with a strain lacking SepJ, which has been shown to be necessary for inter-

cytoplasmic exchange of calcien, indicate that SepJ is necessary for decay of HetR in cells

adjacent to those overexpressing hetN in genetic mosaics [9]. Taken together, these results

suggest that an RGSGR-containing peptide derived from HetN diffuses between cells via direct

cytoplasmic exchange to direct pattering of heterocysts.

177

REFERENCES CITED

1. Golden, J.W. and H.-S. Yoon, Heterocyst development in Anabaena. Curr. Opin.

Microbiol., 2003. 6: p. 557-563.

2. Zhang, C.-C., et al., Heterocyst differentiation and pattern formation in cyanobacteria: a

chorus of signals. Mol. Microbiol., 2006. 59(2): p. 367-375.


N]N2 and 13





5. Haselkorn, R., How cyanobacteria count to 10. Science, 1998. 282: p. 891-892.

6. Wolk, C.P., Physiological basis of the pattern of vegetative growth of a blue-green alga.

Proc. Natl. Acad. Sci. U.S.A., 1967. 57: p. 1246-1251.

7. Ferrell, J.E.J., Self-perpetuating states in signal transduction: positive feedback, double-

negative feedback and bistability. Curr. Opin. Cell Bio., 2002. 14: p. 140-148.


transcription start sites of hetR from Anabaena sp. strain PCC 7120. J. Bacteriol., 2010.

192: p. 1088-1096.


patterning is regulated by inhibitor gradients that promote activator decay. Proc. Natl.

Acad. Sci., 2009. 106: p. 19884-19888.












Acad. Sci. U.S.A., 2004. 101(14): p. 4848-4853.



2010. 77: p. 562-574.

178








19. Wu, X., et al., patS minigenes inhibit heterocyst development of Anabaena sp. strain

PCC 7120. J. Bacteriol., 2004. 186: p. 6422-6429.


7120 and PatS-5. Biochemistry, 2011. 50: p. 9212-9224.





Microbiol., 2005. 57: p. 111-123.

23. Li, B., X. Huang, and J. Zhao, Expression of hetN during heterocyst differentiation and its

inhibition of hetR up-regulation in the cyanobacterium Anabaena sp. PCC 7120. FEBS

Lett., 2002. 517: p. 87-91.

24. Black, T., A. and C.P. Wolk, Analysis of a Het- mutation in Anabaena sp. strain PCC 7120

implicates a secondary metabolite in the regulation of heterocyst spacing. J. Bacteriol.,

1994. 176: p. 2282-2292.






Microbiol., 2012. 83: p. 682-693.



Bacteriol., 2007. 189: p. 2460-2467.

28. Norris, M.H., et al., Stable, site-specific fluorescent tagging constructs optimized for

Burkholderia species. Appl. Environ. Microbiol., 2010. 76: p. 7635-7640.




179



p. 4473-4482.



Sci U S A, 2009. 106(47): p. 19884-8.



Proc Natl Acad Sci U S A, 2004. 101(45): p. 16040-5.

34. Zhu, M., S.M. Callahan, and J.A. Allen, Maintenance of heterocyst patterning in a

filamentous cyanobacterium. J. Biol. Dynamics, 2010.

35. Liu, J. and W.-L. Chen, Characterization of HetN, a protein involved in heterocyst

differentiation in the cyanobacterium Anabaena sp. strain PCC 7120. FEMS Microbiol.

Lett., 2009. 297: p. 17-23.

36. Bauer, C.C., et al., Suppression of heterocyst differentiation in Anabaena sp. strain PCC

7120 by a cosmid carrying wild-type genes encoding enzymes for fatty acid synthesis.

FEMS Microbiol. Lett., 1995. 151: p. 23-30.

37. Drew, D., et al., Rapid topology mapping of Escherichia coli inner-membrane proteins by

prediction and PhoA/GFP fusion analysis. Proc. Natl. Acad. Sci., 2002. 99: p. 2690-2695.

38. Mariscal, V., A. Herrero, and E. Flores, Continuous periplasm in a filamentous,

heterocyst-forming cyanobacterium. Mol. Microbiol., 2007. 65: p. 1139-1145.

39. Zhang, L.-C., et al., Existence of periplasmic barriers preventing green fluorescent protein

diffusion from cell to cell in the cyanobacterium Anabaena sp. strain PCC 7120. Mol.

Microbiol., 2008. 70: p. 814-823.

40. Flores, E., et al., Is the periplasm continuous in filamentous multicellular cyanobacteria?

Trends Microbiol, 2005. 14: p. 439-443.

41. Merino-Puerto, V., et al., Fra proteins influencing filament integrity, diazotrophy and

localization of septal protein SepJ in the heterocyst forming cyanobacterium Anabaena

sp. Mol. Microbiol., 2010. 75: p. 1159-1170.

42. Mullineaux, C.W., et al., Mechanism of intercellular molecular exchange in heterocyst-

forming cyanobacteria. EMBO J, 2008. 27: p. 1299-1308.

180

APPENDIX A. PEER REVIEWED PUBLICATION: The putative phosphatase all1758 is

necessary for normal growth, cell size, and synthesis of the minor heterocyst-specific

glycolipid in the cyanobacterium Anabaena sp. strain PCC 7120

MICROBIOLOGY Vol. 158, p. 380-389

Copyright © 2012, Society for General Microbiology Journals. All Rights Reserved.

Sasa K. Tom and Sean M. Callahan

*

Department of Microbiology, University of Hawaii, Honolulu, HI 96822

Received 16 September 2011/Revised 30 October 2011/Accepted 1 November 2011

ABSTRACT

The filamentous cyanobacterium Anabaena sp. strain PCC 7120 differentiates nitrogen-fixing

heterocysts arranged in a periodic pattern when deprived of a fixed source of nitrogen. In a

genetic screen for mutations that prevent diazotrophic growth, open reading frame all1758, which

encodes a putative serine/threonine phosphatase, was identified. Mutation of all1758 resulted in

a number of seemingly disparate phenotypes that included a delay in the morphological

differentiation of heterocysts, reduced cell size, and lethality under certain conditions. The mutant

was incapable of fixing nitrogen under both oxic and anoxic conditions, and lacked the minor

heterocyst-specific glycolipid. Pattern formation, as indicated by the timing and pattern of

expression from the promoters of hetR and patS fused transcriptionally to the gene for GFP, was

unaffected by mutation of all1758, suggesting that its role in the formation of heterocysts is limited

to morphological differentiation. Transcription of all1758 was constitutive with respect to both cell

type and conditions of growth, but required a functional copy of all1758. The reduced cell size of

the all1758 mutant and the location of all1758 between the cell division genes ftsX and ftsY may

be indicative of a role for all1758 in cell division. Taken together, these results suggest that the

protein encoded by all1758 may represent a link between cell growth, division and regulation of

the morphological differentiation of heterocysts.

181

INTRODUCTION

Heterocysts are terminally differentiated, non-dividing cells that allow the simultaneous execution

of two incompatible processes, fixation of nitrogen and oxygen-evolving photosynthesis, in some

filamentous cyanobacteria. They maintain a micro-oxic environment where oxygen-labile

nitrogenase can function. The envelope of heterocysts includes a layer of distinct glycolipid that

reduces the diffusion of molecular oxygen into cells, an outer layer of heterocyst-specific

exopolysaccharides that protect the integrity of the glycolipid layer, and increased respiration

reduces the level of molecular oxygen that does enter heterocysts [1]. Fixed nitrogen from

heterocysts supports the growth of vegetative cells in the filament, which in turn supply

heterocysts with fixed carbon [2, 3]. In the cyanobacterium Anabaena sp. PCC 7120 heterocysts

are arranged in a periodic pattern at intervals of approximately 10 cells along unbranched

filaments under conditions that require fixation of nitrogen for growth of the organism [4].

Activation of the heterocyst differentiation signaling pathway in response to nitrogen limitation

leads to an accumulation of the metabolite 2-oxoglutarate (2-OG) in cells [5]. Binding of 2-OG

stimulates the DNA-binding activity of the global transcriptional regulator of nitrogen and carbon

metabolism in cyanobacteria, NtcA [6]. NtcA indirectly leads to upregulation of the master

regulator of heterocyst differentiation, hetR, which feeds back on the regulation of transcription of

ntcA [7], and together with NtcA, directly or indirectly activates genes involved in the patterning

and morphogenesis of heterocysts [8]. Induction of differentiation, pattern formation, and

morphogenesis involves the transcription of hundreds of genes specific for the formation of

heterocysts.

The reversibility of protein phosphorylation mediated by kinases and phosphatases is an

essential part of many signaling cascades and regulatory networks. In Anabaena sp. genes with

homology to signaling components have been reported to influence nitrogen metabolism and the

early regulatory stages of heterocyst development. NrrA which encodes a response regulator of

the OmpR family, acts early in the differentiation process by directly upregulating hetR expression

[9]. The PP2C-type protein phophatases PrpJ1 and PrpJ2 are also involved in the initiation of

heterocyst development through mutual regulation of both ntcA and hetR [10]. The protein

kinases Pkn41 and Pkn42, which contain Ser/Thr-kinase and His-kinase domains respectively,

are cotranscribed specifically under iron deficient conditions and regulated by NtcA [11]. The

nitrogen regulatory protein PII encoded by glnB, is differentially modified in the two cell types. It

goes from a phosphorylated to non-phosphorylated state during transition from vegetative cell to

heterocyst [12]. The pknE gene lies 301 bp downstream from the protein phosphatase 1/2A/2B

homolog encoded by prpA. Both pknE and prpA inactivation mutants produce aberrant

182

heterocysts [13]. Overexpression of the putative Ser/Thr kinase encoded by pknE from its native

promoter inhibited heterocyst development, possibly through inhibition of HetR [14]. Another

protein that appears to affect HetR, PatA, contains a phophoacceptor domain with homology to

the response regulator CheY, although a corresponding histidine kinase has not been isolated. It

appears to attenuate the negative regulation of heterocyst differentiation by PatS and HetN [15],

while promoting the activity, and limiting the accumulation, of HetR [16].

In addition, protein phosphorylation has also been implicated in the later stages of development,

in particular during the process of heterocyst maturation. Formation of the minor heterocyst

glycolipid involves two protein kinases of the hstK family, Pkn30 and Pkn44, which contain both

an N-terminal Ser/Thr kinase domain and a C-terminal His-kinase domain [17]. PrpJ1 has been

shown to regulate the synthesis of the major heterocyst glycolipid [18]. Alr0117 and HepK

(All4496) both encode putative two-component system sensory histidine kinases that appear to

be involved in the induction of hepA, one of the genes responsible for synthesis of the heterocyst

polysaccharide layer [19-21]. The manganese-dependent Ser/Thr protein phophatase of the PPP

family DevT (Alr4674) accumulates in mature heterocysts, is not regulated by NtcA, and appears

to have a role in the later steps of heterocyst differentiation (Espinosa et al., 2010). In this study,

the isolation and characterization of a gene encoding a putative PP2C-type protein phosphatase,

all1758 is described.


Bacterial strains and growth conditions. Strains used in this study are described in Table 1.

The growth of Escherichia coli and Anabaena sp. strain PCC 7120 and its derivatives;

concentrations of antibiotics; and induction of heterocyst formation in BG-110, which lacks a

combined-nitrogen source; regulation of PpetE expression; and conditions for photomicroscopy

were as previously described [22]. Images were processed in Adobe Photoshop CS2. To avoid

complications from the vacuolization phenotype observed with strains UHM183 and UHM184

upon growth in BG-11 liquid medium, which contains a source of combined nitrogen, these

strains were transferred from solid BG-11 medium to liquid BG-110 medium for induction of

heterocyst differentiation. Plasmids were conjugated from E. coli into Anabaena strains as

previously described [23].

Construction of plasmids used in this study. Plasmids used in this study are described in

Table 1. Oligonucleotides used in this study are described in Table 2. Constructs derived by

PCR were sequenced to verify the integrity of the sequence. Plasmid pST112 was used to delete

most of the coding region of all1758. An 852-bp region containing the first 30-bp of the all1758

183

coding region and upstream DNA was amplified from the chromosome (with primers 1758 up F

and 1758 up R) and fused to an 848-bp region containing the last 30-bp of the all1758 coding

region and downstream DNA (using primers 1758 down F and 1758 down R) via overlap

extension PCR. The 1700-bp fragment was cloned into pRL277 as a BglII-SacI fragment using

restriction sites introduced on the primers to generate pST112.


The 1700-bp fragment used for construction of pST112 was moved into the EcoRV site of

pBluescript SK+ (Stratagene) to generate pST110. The 1700-bp fragment was moved from

pST110 into pRL278 as a BlgII-SacI fragment to create pST113. To generate pST114, the 2082-

bp Ω interposon, which confers resistance to Sp and Sm [24], was introduced into pST113 at a

SmaI site generated during overlap extension PCR.

pST141 is a mobilizable shuttle vector containing the putative promoter and coding regions of

all1758. An 1866-bp fragment was amplified from the chromosome (using primers Pall 1758

BamHI F and all1758 SacI R) and cloned into pAM504 as a BamHI-SacI fragment using

restriction sites engineered on the primers to generate pST141.

The translational all1758-gfp reporter fusion under the control of the all1758 promoter was

created using primers Pall1758 BamHI F and all1758 SmaI R to amplify the 1866-bp fragment

from the chromosome. The fragment was moved as a BamHI-SmaI fragment into pSMC232, a

replicating plasmid used generate translational fusions to gfp [16], to give pST150.


amplified from the chromosome with primers Pall1758 SacI F and all1758 15 bp up SmaI R and

cloned into pAM1956 as a SacI-SmaI fragment to create pST151. To create the gfp

transcriptional reporter for all1759, a 304-bp region upstream of all1759 was amplified from the

chromosome using primers Pall1759-SacI-F and Pall1759-SmaI-R and moved into the EcoRV

site of pBluescript SK+ to generate pST252. The fragment was moved from pST252 as a SacI-

SmaI fragment into pAM1956 to create pST249.

Two plasmids carrying a PpetE-all1758 transcriptional fusion were constructed. Plasmid pST156 is

a mobilizable shuttle vector carrying a transcriptional fusion between the petE promoter and

all1758. The 1392-bp coding region of all1758 was amplified from the chromosome using

primers all1758 OE EcoRI F and all1758 OE BamHI R and ligated into the EcoRV site of

pBluescript SK+ to create pST184. The fragment was moved from pST184 as a EcoRI-BamHI

fragment into pKH256 [25], which is designed to create transcriptional fusions to the petE

184

promoter, to generate pST156. Plasmid pST367 differs from pST156 in the engineered

ribosomal binding site that was introduced by using primer all1758 EcoRI rbs F in place of all1758

OE EcoRI F to amplify all1758 from the chromosome. The PCR fragment was moved into the

EcoRV site of pBluescript SK+ to generate pST365, and subsequently moved from pST365 as a


The pPROEX-1 (Life Technologies) derivative, pST188 was used to over-express all1758 in E.

coli and facilitate purification. A 1392-bp region containing the coding region for all1758 was

amplified from the chromosome using primers all1758-NdeI-F and all1758-NH6-NotIR, and

moved into the EcoRV site of pBluescript SK+ to create pST192. To generate pST188, the

fragment was moved as a NdeI-NotI fragment into the pPROEX-1 backbone derived from

pSMC148 [16] digested with NdeI-NotI. The resulting open reading frame, designated

all1758(H6), encodes a glycine, then six histidine codons preceding sixteen codons

(DYDIPTTENLYFQGAHAH, which contains a sequence recognized by the Tobacco Etch Virus

protease) fused to all1758 followed by a stop codon after the last codon of all1758.

The transcriptional fusion Pnir-all1758 present on plasmid pST368 was created using the primers

all1758 rbs EcoRI F and all1758 OE BamHI R to amplify all1758 from the chromosome. The

fragment was ligated into the EcoRV site of pBluescript SK+ to generate pST366. To create

pST368, the fragment was moved as a EcoRI-BamHI fragment into pSMC188, a replicating

plasmid used to generate transcriptional fusions to the nirA promoter [22]

Strain construction. Deletion of the all1758 coding region was performed as previously

described [26], using plasmid pST112 or pST114 and Anabaena sp. strain PCC 7120 to yield

strains UHM183 and UHM184, respectively. The resultant strains were screened via colony PCR

with primers all1758 flank up F and all1758 flank down R, which anneal outside the chromosomal

region introduced onto pST112 and pST114 for deletion of all1758, and tested for sensitivity and

resistance to the appropriate antibiotics. Primers all1758 3’ F and all1757 5’ R were used in a

PCR reaction to test if recombination between plasmid pST141, with harbors Pall1758-all1758, and

the chromosomal locus of all1758 of strain UHM183 had occurred.

Overexpression and purification of recombinant all1758 and phosphatase assays.

Recombinant All1758 was produced from E. coli BL21 (DE3) transformed with pST188 using Ni-

nitrilotriacetic acid affinity chromatography (Qiagen) as described previously [22] with the

following exception: 4 mL of wash buffer (50 mM NNaH2PO4, 300 mM NaCl, 20 mM imidazole, 20%

glycerol) was added to the column twice, followed by four 500 µL volumes of elution buffer (50

mM NaH2PO4, 300 mM NaCl, 250 mM imidazole, 20% glycerol), each diluted four-fold directly

185

into a total volume of 1.5 mL of capture buffer (50 mM NaH2PO4, 300 mM NaCl) upon elution

from the column.

Assays for phosphatase activity were conducted as described by previously [27]. The substrate,

4-nitrophenol phosphate (pNPP), was dissolved in pNPP buffer containing 10 mM Tris-HCl, pH

8.5, 1 mM DDT, and 50 mM NaCl with addition of either 2 mM MnCl2 or 5 mM MgCl2 and

supplemented with either 10 or 100 µM cAMP or cGMP. As a test of hydrolysis of pNPP by the

putative phosphatase activity of All1758, 2 µg of the recombinant protein was added to the

cocktail at a final reaction volume of 1 mL. Absorbance at 400 nm was monitored as a

measurement of hydrolysis of pNPP at 25 °C every thirty seconds for the first thirty minutes after

addition of protein to the cocktail, then at 45 minutes, 1 hour, 2 hours, 5 hours, and 24 hours.

Measurement of cell size. Reported values for cell sizes were obtained by measuring the

“height” of one hundred randomly distributed cells of Anabaena sp. strains PCC 7120 and

UHM184. Height refers to the dimension perpendicular to the longitudinal axis of the filament

halfway between cell poles. Significance of height distributions was established using an

unpaired t-test with a two-tailed P value of less than 0.0001. Approximate cell volumes were

calculated by treating cells as spheres using the average cell height as the radius in the formula,

Volumesphere=4/3πr3.

RNA isolation and RT-PCR. Total RNA was extracted as previously described [16]. For

reverse-transcription PCR (RT-PCR), 0.5 µg of total RNA was used for the synthesis of cDNA

with reverse transcriptase and the corresponding reverse primers for subsequent use in PCR

carried out for 27 cycles using primers 23 to 32 in a manner as previously described [16].

Acetylene reduction, glycolipid and exopolysaccharide assays. Acetylene reduction assays

measured from three independent cultures were performed as previously described [26].

Glycolipid extraction for thin layer chromatography and staining of exopolysaccharides also were

conducted as previously described [28].

186

Table 1. Strains and plasmids used in this study

Strain or plasmid Relevant Characteristic(s) Source or Reference

Anabaena sp. strains

PCC 7120 Wild type Pasteur Culture Collection

UHM103 ΔhetR [29]

UHM128 Δpbp6 [30]

UHM183 Δall1758 This study

UHM184 Δall1758 with Ω cassette This study

Plasmids

pAM504 Shuttle vector for replication in E. coli and Anabaena; Km

r,, Neo

r

[31]

pAM505 Shuttle vector pAM504 with inverted multiple cloning site

[31]


pAM1956 pAM505 bearing promoterless gfp for transcriptional fusions

[32]





r

Life Technologies


r [34]



pSMC148 pPROEX-1 carrying hetR [16]

pSMC188 pAM504 bearing Pnir for transcriptional fusions [22]

pSMC232 pAM504 bearing promotorless gfp for translational fusions

[16]



This study

187

Table 1. (Continued) Strains and plasmids used in this study




pST156 pAM505 bearing PpetE-all1758 This study

pST188 pPROEX-1 carrying all1758 This study


pST367 pAM505 bearing PpetE-all1758 This study

pST368 pAM504 bearing Pnir-all1758 This study

188

Table 2. Oligonucletotides used in this study

Primer No.

Primer Name Sequence (5’ to 3’)

1 1758 up F AGATCTGGTGGTGCGATCGCTTGGAG

2 1758 up R GCCACGACTGTCCCCGGGTGCCTGTTGTTATTATCACG

3 1758 down F ACGGACAACAATAATCCCGGGCGGTGCTGACAG

4 1758 down R GAGGCTCGCAGCCGCCCAGCTGTATCTAC

5 all1758 flank up F GCCCAAGCACGAAATACAG

7 all1758 flank down R CTTCCCGGTAAACTGTTGG

8 all1758 3’ F GCTAGAACCTGGTGATACAG

9 all1757 5’ R ACAGACTGAGAGTTACGTGC

10 Pall 1758 BamHI F TATATGGATCCTAGAAGCTCTCTTGAGTGGC

11 all1758 SacI R ATATAGAGCTCCCCAGTCCCTTTTATACTCG

12 all1758 SmaI R ATATACCCGGGATTCGATCTGTAAGACCAC

13 Pall1758 SacI F TATATGAGCTCTAGAAGCTCTCTTGAGTGGC

14 all1758 15bp up SmaI R ATATACCCGGGAAAGGGGTTAATTTTAGATTGAC

15 Pall1759-SacI-F TATATGAGCTCCCAAAAAAAGGCGCTGAGTG

16 Pall1759-SmaI-R ATATACCCGGGTAGTTGTTACGAGTTATGAG

17 all1758 OE EcoRI F TATATGAATCCATGTGCCTGTGC CTCCATTTTC

18 all1758 OE BamHI R TATATGGATCCCCCAGTCCCTTTTATACTCG

19 all1758-NdeI-F ATATACATATG CCTGTGCCTCCATTTTCCTCTCAAC

20 all1758-NH6-NotIR TTACTGCGGCCGCTTTTCGATCTGTAAGACCACTAAAGTC

21 all1758 EcoRI rbs F TATATGAATTCAGGAGGTGATTGTGCCTGTGCCTCCATTTTC

22 all1758 rbs EcoRI F TATATAAGGAGGAATTCTGTGCCTGTGCCTCCATTTTC

23 asr5349 RT F AACTAAATCTAGTGAGCATGG

24 asr5349 RT R AAGCATATAAAAGGTGATGGT

25 asr5350 RT F ATGGCATTCATCAAGATACAG

26 asr5350 RT R AACGCTTGACTCTTGATATTG

189

Table 2. (Continued) Oligonucletotides used in this study

27 all1757 RT F CTGCCGTCAGTACTGTTAC

28 all1757 RT R AGTCCAGGCTTGCTCTTTAG

29 all1759 RT F GGTTCCGTCGTCAAACTAACG

30 all1759 RT R CACGCCAGTTGTTTGTACTG

31 rnpB RT F ATAGTGCCACAGAAAAATACCG

32 rnpB RT R AAGCCGGGTTCTGTTCTCTG

RESULTS

Mutation of all1758 prevents diazotrophic growth

In an effort to identify genes necessary for diazotrophic growth of Anabaena sp. strain PCC 7120,

a genetic screen using transposon mutagenesis to both increase and mark the sites of mutations

was conducted as previously described [30]. In one mutant the transposon had disrupted open

reading frame all1758 in the annotated genome sequence, suggesting that all1758 was

necessary for diazotrophic growth. To verify a cause-and-effect relationship between inactivation

of all1758 and the phenotype of the mutant, an Ω interposon conferring spectinomycin and

streptomycin resistance was used to replace nucleotides +31 to +1358 relative to the GTG

translational initiation codon of all1758 in the wild type strain. The resulting mutant, UHM184,

contained the first and last thirty nucleotides of all1758 flanking the Ω interposon and exhibited

impaired diazotrophic growth similar to that of the isolated transposon mutant. Complementation

of UHM184 with a plasmid bearing all1758 preceded by 497 bp upstream of all1758, including the

474 bp intergenic region between the end of upstream gene all1759 and the start of all1758,

restored diazotrophic growth to the mutant and suggested that inactivation of all1758 and not

polar effects of the transposon insertion was the cause of the mutant phenotype. The all1758

gene is located between two genes transcribed in the same orientation and predicted to have a

role in cell division, ftsX (all1757) and ftsY (all1759). To verify that inactivation of all1758 was

solely responsible for the phenotypes of the mutants, nucleotides +31 to +1358 were cleanly

deleted from the chromosome of the wild type. The resulting strain, UHM183, had an in-frame

deletion of all1758 and differed from UHM184 only by the absence of the Ω interposon. The

phenotype of this strain was similar to that of UHM184. Strain UHM183 was complemented in

the same manner as for UHM184, and there was no evidence for recombination between the

plasmid and the chromosomal locus of all1758 as indicated by the lack of a PCR product using

primers corresponding to the region of all1758 that is deleted from the mutant chromosome and

190

the coding region of all1757. Lastly, transcripts of all1757 and all1759 were detected by RT-PCR

in RNA isolated from strains UHM184 and the wild type (data not shown). Taken together, these

results indicated that a functional copy of all1758 was required for diazotrophic growth of

Anabaena sp. strain PCC 7120.

The predicted 463 amino acid sequence encoded by all1758 was similar to sequences of other

proteins in the Protein Database maintained by NCBI. Residues 45 to 200 appeared to comprise

a GAF domain, which is named after the types of proteins containing the domain (cGMP-specific

phosphodiesterases, adenylyl cyclases and FhlA). Many of these proteins bind cyclic nucleotides.

Residues 270 to 463 were similar in sequence to serine/threonine protein phosphatases of the

PP2C superfamily, which is comprised of metallo-phosphatases that employ two metal ions in the

catalytic site. PP2C proteins generally have a catalytic domain of approximately 290 amino acids

that includes eight absolutely conserved residues within eleven conserved motifs [17, 36].

However, the predicted protein sequence of All1758 lacked the invariant glycine present in Motif 5

and, instead, had an alanine in its place.

In an attempt to demonstrate phosphatase activity of All1758, the 52 kD protein with an N-

terminal polyhistidine epitope tag was purified from E. coli strain BL21 by affinity chromatography.

Attempts to characterize phosphatase activity using p-nitrophenol (pNPP) as a substrate were

unsuccessful even after addition of either 10 or 100 uM cAMP or cGMP separately with the PNPP

substrate, in an effort to support the potential role of the GAF domain in phosphatase function.

Pleiotropic phenotype of all1758 null mutants

Although all1758 was discovered in a genetic screen designed to isolate mutants incapable of

growth on nitrogen-deficient media, disruption of all1758 also affected viability on nitrogen-replete

media. Colonies of the mutant were pale green rather than the vibrant green of the wild type,

and after about three weeks of growth on BG-11 solid media with either 17 mM nitrate or 2 mM

ammonia as a source of fixed nitrogen, growth appeared to cease, and the strain could not be

revived by subculturing onto fresh medium. In comparison, the wild type strain remained viable

under the same conditions in excess of three months. However, the strain could apparently be

maintained indefinitely if repeatedly subcultured onto BG-11 solid medium every two weeks. To

test if loss of viability was the result of a change in the medium over time, non-viable filaments of

UHM184 were scraped from a plate culture 3 weeks after its inoculation, and the resulting

“conditioned” solid medium was streaked with PCC 7120. Growth of the wild-type strain was

indistinguishable from that on unconditioned medium. Cells of UHM184 filaments cultured on

BG-11 solid medium were markedly smaller than those of the wild type grown under the same

191

Figure 1. Phenotype of the all1758 deletion strain, UHM184 under varying conditions. The

diminutive cell size of UHM184 (B) as compared to the wild type strain Anabaena sp. strain

PCC 7120 (A) cultured on solid BG-11 medium. At 24 h after culturing in liquid BG-11 medium,

UHM184 develops intracellular vacuoles among all cells of a filament (C). At 24 h after transfer

from solid BG-11 medium to liquid BG-110 medium, which lacks a source of combined nitrogen,

heterocysts form in PCC 7120 (D) but not in UHM184 (E). After 48 h of nitrogen starvation,

UHM 184 exhibits heterocyst patterning among cells of diminutive size (F). Carets indicate

heterocysts. Bars = 10 µM.

192

conditions. The average vegetative cell height (measured perpendicular to the longitudinal axis

of the filament) of mutant cells was 2.88 ± 0.25 m, compared to 4.34 ± 0.23 m for cells of the

wild type, which corresponded to a more than 3-fold reduction in cell volume for the mutant (Fig.

1a, b). This diminutive phenotype persisted until loss of viability and when cells were transferred

to liquid medium lacking fixed nitrogen.

Growth of the mutant in liquid BG-11 containing nitrate or ammonium as a source of fixed

nitrogen was different from that on solid medium. Within 24 hours of inoculation into liquid

medium, the cytoplasm was found on one side of the cell, and the majority of the cell volume was

occupied by what appears to be a single, large vacuole (Fig. 1c). Alternatively, the plasma

membrane may have pulled away from the cell wall in a manner resembling plasmolysis in plant

cells. The process was unlikely to be attributable to the formation of a large gas vesicle because

mutant cells settled more readily than those of the wild type to the bottom of growth vessels.

Vacuole formation was accompanied by enlargement of cells and resulted in cell lysis within 48

hours. To test if cell lysis and/or vacuolization of cells was in response to a substance excreted

into the medium by the mutant, conditioned medium separated from cellular debris by

centrifugation was inoculated with PCC 7120. Growth of PCC 7120 in medium conditioned by

UHM184 was similar to that in unconditioned medium, indicating that lysis of UHM184 was not

caused by amendment of the medium by the mutant so as to make it unfit for growth of PCC

7120. The pH of medium conditioned by growth of UHM184 was similar to that conditioned by

the wild type and unconditioned medium. In addition, the inverse experiment, growing UHM184 in

spent medium of the strain PCC 7120 did not rescue any of the mutant phenotypes.

UHM184 was incapable of growth in the absence of fixed nitrogen, and morphologically distinct

heterocysts were not observed until 48 h after removal of fixed nitrogen, about twice the time for

the wild type (Fig. 1d, e, f). The periodic pattern of heterocyts along filaments was similar to that

of the wild type. Heterocysts of UHM184 had characteristic reduced levels of auto-fluorescence,

indicative of degradation of phycobiliproteins, and polar cyanophycin granules associated with

mature heterocysts. Note that all of the mutant phenotypes were complemented by introduction

of all1758 on a plasmid to strain UHM184 as described above.

193

Strain UHM184 was Fix- and lacked the minor heterocyst-specific glycolipid.

During the late stages of

heterocyst development,

deposition of heterocyst-specific

polysaccharides contributes to the

creation of the microoxic

environment necessary for the

function of nitrogenase, an oxygen-

labile enzyme. Heterocysts of

strain UHM184 were readily

stained by Alcian blue, indicative of

the presence of heterocyst

envelope exopolysaccharides (Fig.

2). Thus, polysaccharide synthesis

and deposition appears unimpaired

in the mutant.

Just interior to the layer of exopolysaccharide, a layer of heterocyst-specific glycolipid also

contributes to the microoxic interior of a heterocyst. To determine if heterocyst-specific

glycolipids (HGLs) were present in mutant strain UHM184, lipids from strains PCC 7120,

UHM184, and UHM103, a hetR mutant that served as a negative control, were extracted after 72

hours of combined nitrogen deprivation, and separated by thin-layer chromatography for

visualization (Fig. 3). Two HGLs have been characterized in the wild type [37-39]. Although the

more abundant, slower-migrating species corresponding to 1-(0-a-D-glucopyranosyl)-3,25-

hexacosanediol was present in strains UHM184 and PCC 7120 (the major HGL), the less

abundant, faster migrating glycolipid, 1-(0-a-D-glycopyranosyl)-3-keto-25-hexacosanol (the minor

HGL) was absent from UHM184. As expected, both HGLs were absent from strain UHM103,

which does not make heterocysts. The absence of the minor HGL, which is necessary for a

functional heterocyst, could account for the inability of UHM184 to grow under diazotrophic

conditions. Typically strains affected in HGL synthesis lack both the minor and major HGL, with

one exception. A double kinase mutant was shown by RT-PCR to lack expression of two

glycolipid synthesis genes, asr5349 and asr5350 [17]. Conversely, transcripts of asr5349 and

asr5350 were detected by RT-PCR in RNA isolated from strain UHM184 (data not shown).

Strains that lack one or both of the heterocyst specific glycolipids cannot fix nitrogen in the

presence of molecular oxygen (Fox- phenotype) but can in its absence. Other strains that are

Figure 2. Heterocyst-specific exopolysaccharide of

UHM184. Bright-field images of heterocyst-specific

exopolysaccharides stained by Alcian Blue in (A) PCC

7120 and (B) UHM184 at 48 h after removal of combined

nitrogen. Carets indicate heterocysts. Bars = 10 µM.

194

deficient in more than just the creation of microoxic heterocysts cannot fix under either condition

(Fix- phenotype). The nitrogenase activity of strain UHM184 at 48 h post-induction was assessed

Figure 3. Heterocyst-specific glycolipids of UHM184. 1, 2, 3 are lane numbers. Lane 1,

lipid extract from the all1758 mutant, UHM184, which lacks the minor heterocyst glycolipid

(HGL); lane 2, PCC 7120; and lane 3, the ∆hetR mutant, UHM103, which lacks both the

minor and major HGLs separated by thin-layer chromatography 72 h after removal of

combined nitrogen. The minor (1-(0-a-D-glycopyranosyl)-3-keto-25-hexacosanol), and the

major (1-(0-a-D-glucopyranosyl)-3,25-hexacosanediol) heterocyst-specific glycolipids are

indicated by the top and bottom arrows, respectively. Lipid extracts were deposited at the

origin, denoted by ‘*’.

195

by acetylene reduction assays under both oxic and anoxic conditions to determine if it was a Fox-

or Fix- strain. Under oxic conditions, rates of acetylene reduction were 0.350 ± 0.064 nmol

ethylene h-1

ml -1

(OD750 unit)-1

for PCC 7120 and undetectable for strains UHM184; UHM103, a

Fix- strain; and UHM128, a previously characterized Fox

- strain [30, 40]. Under anoxic conditions

rates of acetylene reduction were 0.073 ± 0.038 nmol ethylene h-1

ml -1

(OD750 unit)-1

for strain

UHM128 and undetectable for strains UHM184 and UHM103. Thus, the phenotype of UHM184

was Fix-, suggesting that lack of a functional copy of all1758 disrupted more than just the creation

of microoxic conditions in heterocysts.

all1758 is constitutively expressed and autoregulates its own expression.

Expression of all1758 was

assessed both temporally

and spatially using the

gene for green

fluorescence protein (GFP)

as a reporter. A uniform

level of bright florescence

was observed in

vegetative cells and

heterocysts when a

shuttle vector carrying the

putative promoter region

of all1758 fused to gfp

was introduced into the

wild-type strain. Similar

levels of expression were

seen with and without

nitrate in the medium (Fig.

4a). However, no GFP-

dependent fluorescence

was observed when the

same plasmid was

introduced into strain

UHM184, which has

all1758 deleted from the

chromosome (Fig. 4b).

Figure 4. A functional all1758 gene is required for expression of

all1758 but not for patterned expression of patS. Expression of

all1758 as visualized with a transcriptional Pall1758-gfp fusion on

plasmid pST151 in strains (A) PCC 7120 or (B) UHM184 at 24 h

after removal of combined nitrogen. Expression of patS in PCC

7120 (C) and UHM184 (D) as visualized using the transcriptional

fusion patS-gfp on plasmid pAM1951 at 14 h after induction of

heterocyst formation. Micrographs show bright-field (left panels)

and fluorescence (right panels) images recorded using identical

microscope and camera settings. Carets indicate heterocysts.

Bars = 10 µM.

196

No GFP-dependent fluorescence was observed from the wild type strain harboring a plasmid

bearing all1758 translationally fused to gfp under the control of its native promoter in the

presence or absence of nitrate (data not shown). UHM184 bearing the same plasmid also

remained dim, similar to the wild type, although the cell size had returned to the wild-type

dimensions, apparently due to complementation by the fusion protein encoded on the plasmid. In

addition, over-expression of all1758 from the copper-inducible petE promoter in the wild-type

strain had no apparent effect on heterocyst differentiation or cell size, whereas extra copies of

patS or hetN expressed from the petE promoter on a plasmid in strain UHM184 prevented

heterocyst differentiation, but the diminutive cell phenotype remained (data not shown). Over-

expression of all1758 from the nirA promoter, which is induced under both nitrate-replete

conditions or conditions of nitrogen-starvation [41], yielded the same results as those observed

with use of the petE promoter (data not shown).

all1758 is not required for the timing of pattern formation.

A null mutation in all1758 delays morphological differentiation of heterocysts. Prior to changes in

the morphology of differentiating cells, the cells of filaments that will differentiate are specified by

the generation of a pattern of gene expression. This pattern can be visualized using fusions of

the promoters of hetR or patS to gfp [32, 35]. Strain UHM184 carrying a plasmid with the

promoter of either hetR or patS fused to gfp exhibited a pattern of GFP-dependent fluorescence

that was similar to that observed in the wild type carrying the same plasmid (Fig. 4c, d). Both the

number of cells between fluorescing cells and the timing of the appearance of fluorescence were

similar. The only noticeable difference was the smaller size of the cells in filaments of UHM184.

Thus, all1758 was not required for formation of the pattern that specifies which cells will

differentiate into heterocysts.

Bypass of mutation of al1758 by extra copies of hetR

In an attempt to ascertain where in the regulatory network controlling heterocyst differentiation

all1758 may have been acting, plasmid borne copies of hetR were put into UHM184 and the

resulting strain was examined for phenotypic differences. The strain with extra copies of hetR

had multiple contiguous heterocysts and reduced spacing between groups of heterocysts in

media with or without nitrate as a fixed nitrogen source (data not shown). The same Mch

phenotype was observed when the plasmid was introduced into the wild-type strain. Extra copies

of hetR bypassed the delay in morphological differentiation of heterocysts, but the diminutive cell

phenotype and other associated phenotypes remained.

197

DISCUSSION

Phosphorylation of proteins is widely used in the regulation of various cellular processes. All1758

is predicted to be a PP2C phosphatase, which is a member of the PPM family of serine/threonine

phosphatases that are dependent on the presence of the divalent cation Mg2+

for catalytic

function. However, repeated attempts to support the role of All1758 as a phosphatase were

unsuccessful. The simplest explanation for the lack of discernable phosphatase activity relates to

the absence of one of the eight absolutely conserved residues in the protein sequence of all1758.

PP2C proteins generally have a catalytic domain of approximately 290 amino acids that includes

eight absolutely conserved residues within eleven conserved motifs [17, 36]. All1758 harbors an

alanine residue, a conservative substitution, in lieu of the glycine residue reported as invariant in

Motif 5. Alternatively, phosphatase activity of the protein encoded by all1758 may not be

measurable under the conditions of our assay due, for instance, to substrate specificity, lack of a

cofactor, or improper folding of the recombinant protein.

The minor heterocyst glycolipid, which is essential for provision of microaerobic conditions

necessary for the activity of nitrogenase within the heterocyst, was absent in a strain lacking

all1758. However, the absence of nitrogen fixation in the mutant strain even under anoxic

conditions suggests that all1758 is required for some cellular process(es) related to fixation of

nitrogen not related to diminution of molecular oxygen in heterocysts. Indeed, a diverse and

seemingly unrelated range of phenotypes resulted from mutation of all1758. For instance, the

delay in heterocyst formation in UHM184 may suggest the involvement of all1758 in the timing of

heterocyst morphology development. The diminutive cell size and associated decrease in cell

volume apparent upon mutation of all1758 may be indicative of a change in the timing of cell

division in the mutant, which would prevent the cell from attaining a normal and/or enlarged cell

size. Inhibition of cell division has been shown to prevent differentiation of heterocysts, and

heterocysts have twice the DNA content of vegetative cells [42, 43]. The increased rate of cell

division in strains lacking all1758 may account for the defect in heterocyst differentiation in the

mutant. The location of all1758 between the cell division genes ftsX and ftsY may implicate

all1758 as a cell division gene as well. In Escherichia coli, for example, the cell division gene ftsE

(b3463) lies between ftsX (b3462) and ftsY (b3464). Alternatively, a defect in a process specific to

heterocyst differentiation may influence cell division. Also, the decrease in cell volume may have

perturbed the concentration of specific intracellular components, resulting in the observed

phenotypes. The relationship between cell division and heterocyst differentiation remains

enigmatic at this time.

198

Unpatterned expression of all1758 in all cells was evident under both nitrate-replete and depleted

conditions using a GFP transcriptional reporter fusion, suggesting that the gene is constitutively

expressed. In contrast, GFP-dependent fluorescence was not observed from an all1758

translational fusion in all background strains tested. However, the fusion protein did appear to

complement strain UHM184, suggesting the fusion protein was functional. The lack of detectable

fluorescence may indicate that level of All1758 are very low in cells, and the associated

fluoresecence from the fusion protein was below the level of detection, or the GFP portion of the

protein may have been non-functional. This phenomenon is not new. In our hands, plasmid

constructs bearing hetR-gfp and hetF-gfp translational fusions in hetR and hetF backgrounds,

respectively, also complement the inactivated gene, yet visible fluorescence is lacking (Risser

and Callahan, 2008).

Alterations in the levels of expression of other genes involved in heterocyst differentiation have

been shown to affect cell size. Inactivation of all2874 encoding a diguanylate cyclase results in a

light intensity-dependent variability in heterocyst frequency and reduction in vegetative cell size

(Neunuebel and Golden, 2008). Conversely, a strain lacking hetF has enlarged cells, and extra

copies of hetF on a plasmid result in a diminutive cell phenotype [16]. In contrast with the

correlation between smaller cell size and a delay in heterocyst differentiation in the all1758

mutant, a strain with extra copies of hetF forms more heterocysts than the wild type. Over-

expression of patA also results in enlarged cells [44]. There is no evidence for the

phosphorylation of HetF or PatA, but PatA is similar in sequence to response regulators of two-

component transduction systems [45]. The growing number of phosphatases and kinases

implicated in heterocyst differentiation indicates that protein phosphorylation plays a prominent

role in the regulation of differentiation. However, understanding the extent and nature of that role

is dependent on uncovering the targets of kinase and phosphatase activity.

ACKNOWLDEMENTS

We wish to thank Kelly C. Higa for plasmid pKH256. This work was supported by grant IOS-

0919878 from the National Science Foundation.

FOOTNOTES

*For correspondence: E-mail [email protected]; Tel. (+1) 808 956 8015; Fax (+1) 808 956

5339

199

REFERENCES CITED

1. Murry, M.A. and C.P. Wolk, Evidence that the barrier to the penetration of oxygen into

heterocysts depends upon two layers of the cell envelope. Arch. Microbiol, 1989. 151: p.

469-474.


N]N2 and 13










Sci. U.S.A., 2005. 102: p. 9907-9912.







8. Olmedo-Verd, E., et al., HetR-dependent and -independent expression of heterocyst-

related genes in an Anabaena strain overproducing the NtcA transcription factor. Journal

of bacteriology, 2005. 187(6): p. 1985-91.

9. Ehira, S. and M. Ohmori, NrrA directly regulates expression of hetR during heterocyst

differentiation in the cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol., 2006.

188: p. 8520-8525.

10. Jang, J., et al., Mutual regulation of ntcA and hetR during heterocyst differentiation

requires two similar PP2C-type protein phosphatases, PrpJ1 and PrpJ2, in Anabaena sp.


11. Cheng, Y., et al., A pair of iron-responsive genes encoding protein kinases with a Ser/Thr

kinase domain and a His kinase domain are regulated by NtcA in the Cyanobacterium


12. Laurent, S., et al., Cell-type specific modification of PII is involved in the regulation of

nitrogen metabolism in the cyanobacterium Anabaena PCC 7120. FEBS letters, 2004.

576(1-2): p. 261-5.

200

13. Zhang, C.C., A. Friry, and L. Peng, Molecular and genetic analysis of two closely linked

genes that encode, respectively, a protein phosphatase 1/2A/2B homolog and a protein

kinase homolog in the cyanobacterium Anabaena sp. strain PCC 7120. Journal of

bacteriology, 1998. 180(10): p. 2616-22.








17. Shi, L., et al., Two genes encoding protein kinases of the HstK family are involved in

synthesis of the minor heterocyst-specific glycolipid in the cyanobacterium Anabaena sp.


18. Jang, J., et al., PrpJ, a PP2C-type protein phosphatase located on the plasma membrane,

is involved in heterocyst maturation in the cyanobacterium Anabaena sp. PCC 7120.

Molecular microbiology, 2007. 64(2): p. 347-58.

19. Ning, D. and X. Xu, alr0117, a two-component histidine kinase gene, is involved in

heterocyst development in Anabaena sp. PCC 7120. Microbiology, 2004. 150(Pt 2): p.

447-53.

20. Ramirez, M.E., et al., Anabaena sp. strain PCC 7120 gene devH is required for synthesis

of the heterocyst glycolipid layer. Journal of bacteriology, 2005. 187(7): p. 2326-31.

21. Xu, W.L., et al., pkn22 (alr2502) encoding a putative Ser/Thr kinase in the

cyanobacterium Anabaena sp. PCC 7120 is induced by both iron starvation and oxidative

stress and regulates the expression of isiA. FEBS letters, 2003. 553(1-2): p. 179-82.



Bacteriol., 2007. 189: p. 2460-2467.

23. Elhai, J. and C.P. Wolk, Conjugal transfer of DNA to cyanobacteria. Methods Enzymol.,

1988. 167: p. 747-754.



bacteria. Gene, 1987. 52: p. 147-154.



2010. 77: p. 562-574.

201



Microbiol., 2005. 57: p. 111-123.

27. Ruppert, U., et al., The novel protein phosphatase PphA from Synechocystis PCC 6803

controls dephosphorylation of the signalling protein PII. Molecular microbiology, 2002.

44(3): p. 855-64.










p. 4473-4482.





Acad. Sci., 2009. 106: p. 19884-19888.



77-84.


pattern in Anabaena sp. PCC 7120. Molecular microbiology, 2001. 40(4): p. 941-50.

36. Kennelly, P.J., Protein phosphatases--a phylogenetic perspective. Chemical reviews,

2001. 101(8): p. 2291-312.

37. Lambein, F. and C.P. Wolk, Structural studies on the glycolipids from the envelope of the

heterocyst of Anabaena cylindrica. Biochemistry, 1973. 12: p. 791-798.

38. Soriente, A., et al., Reinvestigation of heterocyst glycolipids from the cyanobacterium,

Anabaena cylindrica. Phytochemistry (Oxford), 1994. 38(3): p. 641-645.

39. Gambacorta, A., et al., Heterocyst glycolipids from five nitrogen-fixing cyanobacteria.

Gazz. Chim. Ital., 1996. 126: p. 653-656.

40. Leganés, F., et al., Wide variation in the cyanobacterial complement of presumptive

penicillin-binding proteins. Arch. Microbiol., 2005. 184: p. 234-248.

202

41. Frias, J.E., E. Flores, and A. Herrero, Nitrate assimilation gene cluster from the

Heterocyst-forming cyanobacterium Anabaena sp. strain PCC 7120. J. Bacteriol., 1997.

179: p. 477-486.










203

APPENDIX B. PEER REVIEWED PUBLICATION: Evidence for Direct Binding Between

HetR from Anabaena sp. PCC 7120 and PatS-5

BIOCHEMISTRY Vol. 50, p. 9212-9244

Copyright © 2011, American Chemical Society. All Rights Reserved.

Erik A. Feldmanna, Shuisong Ni

a, Indra Dev Sahu

a, Clay H. Mishler

a, Douglas D. Risser

b, Jodi L.

Murakamib, Sasa K. Tom

b, Robert M. McCarrick

a, Gary A. Lorigan

a, Blanton S. Tolbert

a, Sean M.

Callahanb and Michael A. Kennedy

a1

a Department of Chemistry and Biochemistry, Miami University, Oxford, Ohio 45056

b Department of Microbiology, University of Hawaii, Honolulu, Hawaii 96822

Received 5 August 2011/Revised 22 September 2011/Published 26 September 2011

ABSTRACT

HetR, the master regulator of heterocyst differentiation in the filamentous cyanobacterium

Anabaena sp. strain PCC 7120, stimulates heterocyst differentiation via transcriptional

autoregulation and is negatively regulated by PatS and HetN, both of which contain the active

pentapeptide RGSGR. However, the direct targets of PatS and HetN are unknown. Here we

report experimental evidence for direct binding between HetR and the RGSGR pentapeptide,

PatS-5. Strains with an allele of hetR coding for conservative substitutions at residues 250-256

had an altered pattern of heterocsyts and, in some cases, reduced sensitivity to PatS-5. Cysteine

scanning mutagenesis coupled with electron paramagnetic resonance (EPR) spectroscopy

showed quenching of spin label motion at HetR amino acid 252 upon titration with PatS-5,

indicating direct binding of PatS-5 to HetR. Gel shift assays confirmed that PatS-5 disrupted

binding of HetR to a 29 base pair inverted-repeat-containing DNA sequence upstream from hetP.

Double electron electron resonance EPR experiments confirmed that HetR existed as a dimer in

solution and indicated that PatS-5 bound to HetR without interfering with the dimer form of HetR.

Isothermal titration calorimetry experiments corroborated direct binding of PatS-5 to HetR with a

Kd of 227 nM. Taken together, these results indicate that 2 molecules of PatS-5 bind to a dimer

of HetR to prevent binding of DNA. PatS-5 appears to either bind in the vicinity of HetR amino

acid L252 or, alternately, to bind in a remote site that leads to constrained motion of this amino

acid via an allosteric effect or change in tertiary structure.

204

INTRODUCTION

Cellular differentiation and patterning are fundamental concepts in the field of developmental

biology. One of the earliest known examples of cell differentiation is that of ancient filamentous

cyanobacteria, which, under pressure of nitrogen starvation, evolved the capability to fix

atmospheric N2 using specialized terminally differentiated cells called “heterocysts” [1-12] more

than 2 billion years ago [13, 14] Heterocyst differentiation in filamentous cyanobacteria evolved

as a means of isolating oxygenic photosynthesis associated with CO2 fixation in vegetative cells

from oxygen-sensitive nitrogenases that carry out nitrogen fixation in heterocysts [15, 16]. Soluble

nitrogen-containing compounds generated as a result of nitrogen fixation in heterocysts are then

shared with neighboring vegetative cells to sustain continued growth of the organism (see

discussion by [17] regarding intercellular transport in filamentous cyanobacteria). Under

conditions of nitrogen starvation, a pattern is established in which approximately every tenth cell

along the filament is terminally differentiated into a heterocyst [18, 19]. Heterocysts support

nitrogenase-based N2 fixation by generating a microoxic environment within the cell that involves

production of two additional layers external to the outer membrane found in vegetative cells,

including a heterocyst-specific glycolipid layer and a heterocyst envelope polysaccharide layer

[20-22]. Initiation of heterocyst differentiation, pattern formation, and pattern maintenance are

regulated by small signaling molecules and a host of different genes in a process that resembles

signaling pathways of higher eukaryotic organisms [10, 11].

The hetR gene plays a central role in regulation of heterocyst differentiation [23]. The master

regulatory protein, HetR, controls heterocyst differentiation through transcriptional autoregulation

[24] and responds to two heterocyst differentiation inhibitors, PatS and HetN, both of which

contain the active pentapeptide RGSGR [24-26]. Interestingly, hetR and patS genes are

widespread throughout both non-heterocyst-forming, as well as heterocyst-forming, filamentous

cyanobacteria, suggesting that their evolutionary role may have emerged before heterocyst

differentiation appeared [27]. While much has been learned about regulation of heterocyst

differentiation in recent years, the molecular level interactions between HetR, PatS, and HetN

during regulation of heterocyst differentiation are still not well understood. There is growing

experimental evidence that PatS and HetN control pattern formation by establishing concentration

gradients along the filament that promote degradation of HetR in an activator-inhibitor type

manner [28], however, the precise mechanism for how HetR interacts with PatS and HetN

remains unknown.

HetR is a 299 residue DNA binding transcriptional regulator believed to be active as a homodimer

[29]. It has been reported that HetR dimer formation involves a disulfide bridge at position C48,

205

and that mutation of C48 to alanine abolished both dimerization and binding activity to its own

DNA promoter in vitro [29]. It has also been reported that HetR has Ser-type autoproteolytic

activity [30, 31] mediated through Ser152 [32]. However, Risser and Callahan have reported that

Cys48 and Ser152 in HetR from Anabaena are not required for proper heterocyst differentiation

[33]. The cause of the contradictory observations has not yet been resolved.

PatS is a short peptide, predicted to be 13 or 17 amino acids, that acts as a negative regulator of

heterocyst differentiation [25]. It is thought that a shorter, processed form of PatS acts as a

diffusible signal molecule [25, 34]. The RGSGR carboxyl terminus of PatS (PatS-5) inhibits

heterocyst differentiation when added to culture medium [25], and has been shown in vitro to

inhibit binding of HetR to a DNA sequence upstream from its own promoter [29, 33]. It has also

been shown that a R223W mutant of HetR was insensitive to in vivo overexpression of both PatS

and HetN [35]. However, the mechanism of PatS-5 disruption of HetR binding to DNA, and the

molecular level cause of loss of R223W HetR sensitivity to PatS remains unknown. A molecular-

level understanding of the biochemical mechanism of action of PatS-5 has not been established.

While it has been widely assumed that PatS-5 disrupts HetR binding to DNA through a direct

interaction between PatS-5 and HetR, based on observations reported in the literature [29, 33,

35], it is alternatively possible that PatS-5 disrupts the HetR-DNA complex through a direct

interaction between PatS-5 and the DNA. This possibility is worth considering given that the

amino acid sequence of PatS-5, RGSGR, with its symmetrically positioned arginines, is similar to

that of the well-established DNA-binding “AT-hook” motif [36], which has the minimal amino acid

consensus sequence PRGRP [37] with both symmetrically positioned arginine residues

conferring, and being required for, DNA binding capacity. The AT-hook is the fundamental DNA

binding motif building block of the HMG-I(Y) (a.k.a. HMGA) subfamily of non-histone high mobility

group chromatin proteins [38, 39], also known as “architectural transcription factors” [37, 40, 41],

with each HMG-I(Y) protein containing three AT-hooks. The AT-hook adopts a crescent-shaped

structure that interacts with AT-rich DNA through non-specific electrostatic interactions between

the negatively charged phosphodiester DNA backbone and the positively charged arginine side

chains that insert into the DNA minor groove [42]. Each PRGRP AT-hook sequence motif spans

5-6 DNA base pairs [36]. Given the similarity between the amino acid sequences of the AT-hook

motif and the PatS-5 sequence, and the observation that a single AT-hook confers DNA binding

capability [37], it is plausible that PatS-5 might also interact with DNA via non-specific

electrostatic interactions through its symmetrically positioned pair of arginine residues. The lack

of three-dimensional structures of most heterocyst regulatory proteins, including HetR, and any

complexes involving PatS-5, has impeded progress towards achieving a complete molecular level

understanding of regulation of heterocyst differentiation.

206

Here we report several major findings that should advance the understanding of regulation of

heterocyst pattern formation in Anabaena, including 1) discovery of a region of HetR including

amino acids 250-256 necessary for sensitivity of HetR to PatS-5; 2) evidence that PatS binds

directly to HetR from cysteine scanning mutagenesis combined with continuous wave (CW)

electron paramagnetic resonance (EPR) spectroscopy; 3) determination of the stoichiometry of

PatS-5 binding to HetR and that PatS-5 binds HetR as a dimer from double electron electron

resonance EPR experiments; and, finally, 4) corroboration that PatS-5 binds directly to HetR and

measurement of the dissociation constant for HetR binding to PatS-5 using isothermal titration

calorimetry.


Bacterial strains and growth conditions. Growth of Escherichia coli and Anabaena sp. strain

PCC 7120 and its derivatives; concentrations of antibiotics; induction of heterocyst formation;

regulation of the petE and nir promoters; and photomicroscopy were as previously described [33].

Plasmids were conjugated from E. coli to Anabaena sp. strain PCC 7120 and its derivatives as

previously described [43].

Plasmid construction for making chromosomal alleles. Strains of Anabaena and plasmids

used in this study are described in Table S1. Plasmids pJM100, pJM101, pJM102, pJM103,

pST211, pJM104 , pJM105 and pDR219 are suicide vectors used to replace the chromosomal

hetR-locus with hetR(R250K), hetR(A251G), hetR(L252V), hetR(E253D), hetR(E254D),

hetR(L255V), hetR(D256E) and hetR(E254G), respectively. Overlap extension PCR was used to

generate each of the mutant hetR alleles except hetR(E254D) and hetR(E254G) using the inner

primers listed in Table S2 with names corresponding to that of the resulting substitution with the

outer primers PhetR-BamHI-F and hetR 3’ Seq. The resulting PCR products were cloned into

plasmid pDR327 as NcoI-SpeI fragments to create plasmids pJM100, pJM101, pJM102, pJM103,

pJM104 and pJM105. Plasmids pST211 and pDR219 were generated in a similar fashion, but

outer primer hetR-SacI-SpeI-R was used in place of hetR 3’ seq for pST211 and previously

published inner primers [33] were used to create hetR(E254G) in pDR219.

Construction of strains containing mutant alleles. Strains of Anabaena with mutant alleles of

hetR in place of the wild-type hetR were created as described previously [28] using the hetR-

deletion strain UHM103 and plasmids pJM100, pJM101, pJM102, pJM103, pST211, pJM104 ,

pJM105 and pDR219 to generate strains UHM163, UHM164, UHM165, UHM166, UHM167,

UHM168, UHM169 and UHM122, respectively. Strains with hetR(E253D) and hetR(E254D) are

207

single recombinants in which the entire plasmid is in the hetR chromosomal locus, whereas the

others are the same as PCC 7120 except for the change in hetR sequence.

In vivo PatS-5 sensitivity assays. Duplicate cultures of Anabaena sp. strain PCC 7120 and the

hetR mutant strains were grown to an approximate optical density of 0.4 at 750 nm in BG-11

medium, which contains nitrate, a fixed form of nitrogen. For one set of cultures, the culture

medium was replaced with fresh BG-11, and replaced thereafter every 48 h with BG-110, which

lacks fixed nitrogen. For the other set of cultures, PatS-5 was included in the medium at a

concentration of 1 µM. The percentage of 500 cells that were heterocysts was determined

microscopically and recorded after each change of medium. Reported values are the average of

three replicates with one standard deviation. Conditions for photomicroscopy were as described

previously [33].

Cloning, overexpression and purification of recombinant soluble HetR. The hetR gene was

PCR amplified from genomic DNA of Anabaena sp. PCC 7120 using the forward primer 5’-

ATCGATCGCATATGAGTAACGACATCGATCTGATC-3’ and the reverse primer 5’-

TGACTCTCGAGCTAATCTTCTTTTCTACCAAACACC-3’, cloned into pET28b (Novagen) at Nde

I and Xho I sites with an N-terminal 6x-His tag including a thrombin cleavage site, then

transformed into competent cells of the cloning host E. coli DH5α. For preparations of hetR

mutants, refer to Table S3. All mutants were generated using the QuickChange II XL site-directed

mutagenesis kit (Stratagene) and their DNA sequences confirmed by capillary electrophoresis

based sequencing at the Miami University Center for Bioinformatics and Functional Genomics.

HetR 250-256C mutants were generated using the C48A mutant plasmid as the DNA template.

Correctly mutated 250-256C plasmids were transformed into BL21(DE3) competent cells

containing the pGroESL vector to assist in proper protein folding [44]. The plasmid construct was

isolated using a Wizard Plus Miniprep kit (Promega) and transformed into competent cells of the

expression host BL21(DE3) (Novagen). The hetR-containing E. coli clone was grown at 37°C with

250 rpm shaking to an OD600 of 0.6-0.9 in 1 L of LB-Miller broth supplemented with 30 µg/mL

kanamycin. Protein expression was induced by addition of 0.25 mM isopropyl β-D-1-

thiogalactopyranoside at 18°C overnight. Cells were harvested and stored at -80°C for later use.

Thawed cells were resuspended in 25 mL of lysis buffer [1 M NaCl, 10% (w/v) glycerol, 10 mM

Tris pH 7.8] followed by 4 passes through a French Pressure Cell Press (Thermo Fisher). Cell

lysates were centrifuged at 24,000g for 20 min. The supernatant was loaded onto a 10 mL Ni-

NTA affinity column (Qiagen) and washed with 50 mL of wash buffer [1 M NaCl, 10% (w/v)

glycerol, 10 mM Tris pH 7.8] followed by a second wash with pre-elution buffer [1 M NaCl, 10%

(w/v) glycerol, 10 mM Tris pH 7.8, 30 mM imidazole]. Soluble His tagged protein was eluted from

the Ni-NTA column with elution buffer [1 M NaCl, 10% (w/v) glycerol, 10 mM Tris pH 7.8, 300 mM

208

imidazole] and concentrated with an Amicon Ultra (Millipore) to a concentration of 15 mg/mL as

determined by the Bradford assay (Thermo Scientific). Concentrated protein solutions were

further purified and analyzed on a Pharmacia Superdex200 HiLoad size exclusion column

equilibrated with buffer containing 1 M NaCl, 10% (w/v) glycerol, 10 mM Tris pH 7.8, and 300 mM

imidazole using a flow rate of 1 mL/min. The mutants co-expressed with the GroESL chaperone

were also supplemented with 30 µg/mL chloramphenicol during expression. Circular dichroism

spectra of HetR did not change with the buffer and solution conditions used in our experiments

following incubation at 37°C for 24 hours, indicating no loss of structure or change in the

secondary structure composition of the protein due to autoproteolytic activity, which has been

reported by Zhou et al. [30, 31], in the absence of phenylmethanesulfonylfluoride protease

inhibitor. The calcium ion concentrations in these solutions were on the order of 50 µM as

determined by inductively coupled plasma atomic emission spectroscopy.

DNA binding assays. Electrophoretic mobility shift assays were performed using 1.8% agarose

gels. Agarose gel electrophoresis experiments used buffer containing 0.2 µg/mL ethidium

bromide (Fisher) run at 80 mAmp for 1 h in TAE buffer. DNA binding reactions were incubated for

10 min at 22°C prior to electrophoresis. Images were generated using an Alpha Innotech camera

and Alpha Imager software. The individual strands of the 29 base pair inverted repeat upstream

DNA fragment were synthesized (250 nmole scale synthesis for each strand) and HPLC purified

(for determining HetR binding stoichiometry), or for all other experiments, synthesized at a 25

nmol scale with standard desalting, by Integrated DNA Technologies (Coralville, Iowa). The

complementary oligonucleotides, forward 5’-GTAGGCGAGGGGTCTAACCCCTCATTACC-3’ and

reverse 5’-GGTAATGAGGGGTTAGACCCCTCGCCTAC-3’, were annealed by suspending

equivalent stoichiometric amounts at 200 µM in the same buffer used to prepare HetR solutions,

heated to 85°C, and then allowed to cool slowly to room temperature. PatS-5 (RGSGR) peptide

was custom synthesized and purified by A & A Labs LLC (San Diego, CA), and PolyG (GGGGG)

peptide was synthesized and purified by Peptide2.0 (Chantilly, VA), both on a 10 mg scale. PatS-

5 was suspended in 100% nanopure H2O to stock 10 mM concentrations. PolyG was suspended

in 100% acetonitrile to a stock 10 mM concentration.

Circular dichroism spectroscopy. Circular dichroism spectra were obtained on a Jasco model

J-810 spectropolarimeter. Measurements were obtained on recombinant HetR using 300 µL of

approximately 15 µM at 25°C using a quartz cell of 1 mm path length. All circular dichroism

spectra were the result of 10 averaged scans from 200 to 250 nm. Jasco Spectra Analysis

software was used to generate the plots of molar ellipticity vs. wavelength.

209

Site-directed spin labeling. The nitroxide spin radical (1-Oxyl-2,2,5,5-tetramethyl-pyrrolin-3-

yl)methyl methanethiosulfonate (MTSL), (Toronto Research Chemicals Inc.) was dissolved in 50%

methanol to a stock concentration of 35 mM. HetR 250-256C mutants were spin-labeled using a

2-fold molar excess of MTSL at 22°C in the dark overnight with gentle shaking. Excess label was

removed (confirmed using CW EPR) by size exclusion chromatography using an analytical grade

Pharmacia Superdex200 10/300 GL column.

Preparation of samples for electron paramagnetic resonance spectroscopy. Spin-labeled

HetR proteins were concentrated to 100 µM. For CW EPR experiments, 30 µL of protein solution

was drawn into 1.1 mm internal diameter (1.6 mm external diameter) quartz capillaries. The

capillary tubes containing the samples were then placed into 3 mm internal diameter quartz EPR

tubes and inserted into the instrument microwave cavity. For pulsed EPR DEER experiments, a

cryoprotectant was added to the samples (samples were brought to a final concentration of 30%

glycerol) and then 8 µL of the cryoprotected protein solutions were drawn into 1.1 mm internal

diameter (1.6 mm external diameter) quartz capillaries. The capillary tubes containing the

samples were frozen in liquid nitrogen and then inserted into the resonator for data collection.

For DEER experiments involving samples containing DNA, the DNA concentration was 100 µM.

Electron paramagnetic resonance spectroscopy. EPR spectra were collected at the Ohio

Advanced EPR Laboratory. CW-EPR spectra were collected at X-band on a Bruker EMX CW-

EPR spectrometer using an ER041xG microwave bridge and ER4119-HS cavity coupled with a

BVT 3000 nitrogen gas temperature controller (temperature stability ± 0.2 K). CW EPR spectra

were collected by signal averaging 15 42-s field scans with a center field of 3370 G and sweep

width of 100 G, microwave frequency of 9.5 GHz, modulation frequency of 100 kHz, modulation

amplitude of 1 G, microwave power of 1 mW at 298 K. DEER data were collected using a Bruker

ELEXSYS E580 spectrometer equipped with a SuperQ-FT pulse Q-band system and EN5107D2

resonator. DEER data were collected at Q-band with a probe frequency of 34.174 GHz and a

pump frequency of 34.235 GHz, a probe pulse width of 20/40 ns, a pump pulse width of 48 ns,

shot repetition time of 499.8 µs, 100 echoes/point, 2-step phase cycling at 80 K collected out to 2

ns.

EPR spectral simulations. Qualitative analysis of spin-label mobility was obtained from

simulations of the CW EPR line shapes to extract best-fit values of the inverse line-width of the

central resonance line (ΔH0-1

) and the components of the diffusion tensor (R) required to

reproduce the observed averaging of the hyperfine interaction tensor. The line-width of central

resonance line was calculated by measuring central peak-to-peak magnetic field from the first

derivative spectrum. Simulations were performed in Matlab using a non-linear least square data

210

analysis program developed by Budil et al. [45, 46]. The three components of electronic Zeeman

interaction tensors (gxx, gyy, and gzz) and hyperfine interaction tensors (Axx, Ayy, and Azz) were

optimized using the spectrum of HetR 256C, which was characteristic of a spin label approaching

the rigid limit. This spectrum provided well-defined features to constrain the A and g tensors to

account for the polarity of the local environment of the MTSL nitroxide spin label [47, 48]. The A

and g tensors were held constant for all remaining simulations and it was assumed that remaining

differences in spectra were due to differential motion of the MTSL spin label. The three

components of the rotational diffusion tensor (Rxx, Ryy & Rzz, units of log (sec-1

)) were varied

during fitting. The best-fit components of the tensor were averaged to report the overall rate of

diffusion (Riso = 1/3(Rxx+Ryy+Rzz)) [49]. EPR spectra of HetR 252C in the presence of PatS-5

consisted of a sum of free and Pats-5 bound species. In this case, a two-site fit was used to

account for free and bound states. Best-fit diffusion tensors for each species in the mixture were

determined using a Brownian diffusion model. The percentage contribution of each motional

component to the overall spectrum was obtained for each sample. DEER data was simulated

using DEER Data Analysis 2009 [50]. The distance distributions P(r) were obtained by Tikhonov

regularization [51] in the distance domain, incorporating the constraint P(r) > 0. The regularization

parameter was adjusted to obtain the realistic resolution.

Molecular dynamics simulations. The atomic coordinates for the HetR crystal structure

(PDB ID: 3QOE) from Fischerella were downloaded from the Protein Data Bank and used to

generate the structures of various spin-labeled HetR constructs with the Nano-scale Molecular

Dynamics (NAMD) program [51]. The C48A and 250-256C cysteine mutations were created

using the molecular graphics software VMD [52]. The nitroxide spin-probe MTSL was attached

using CHARMM force field topology files incorporated into NAMD. The modified protein

assembly was solvated into a spherical water environment and further equilibrated and

minimized by running NAMD simulations at room temperature using CHARMM force field

parameters. The distance distribution for the 252C mutants was predicted with rotamer library

modeling of MTSL conformations using Multi-scale modeling of macromolecular systems

(MMM version 2010) [53]

Isothermal titration calorimetry. ITC measurements were performed at 25°C with a VP-ITC

titration calorimeter (Microcal, Northampton, MA). Size-exclusion-chromatography purified HetR

samples were dialyzed for 18 h at 4°C against the reference buffer: 1 M NaCl, 10% glycerol,

300 mM imidazole, 10 mM Tris pH 7.8. HetR was then diluted to a final concentration of 5 µM

for experiments. PatS-5 peptide was diluted into the reference buffer to a final concentration of

100 µM. The titrations were performed in triplicate using a total of 36 injections of PatS-5

peptide for each titration as follows: one 4 µL injection followed by 35 8 µL injections. For each

211

titration 0.60 mL of 100 µM PatS-5 peptide was loaded into the injection syringe. Blank titrations

of PatS-5 solutions into reference buffer were performed to correct for the heats of dilution of

PatS-5, which were found to be insignificant. Following the blank experiments, a 1.46 mL

sample of HetR was degassed and loaded into the sample cell from the 5 µM prepared stock for

each titration. The resulting titration curves were deconvoluted and fit using a one binding site

model with the ORIGIN for ITC software package (Microcal, Piscataway, NJ).

RESULTS

Conservative substitutions at HetR residues 250-256 affect heterocyst formation and

sensitivity to PatS-5

As part of a mutagenesis study designed to identify residues of HetR required for function,

an allele of hetR coding for an E254G substitution was found to cause differentiation into

heterocysts of essentially all cells containing a multicopy plasmid carrying the mutant gene (32).

Introduction of a plasmid bearing the hetR(E254G) allele under the control of the native promoter

of hetR to PCC 7120 by conjugation resulted in no viable transconjugants. To limit expression of

the hetR(E254G) allele, the wild-type promoter region was replaced with that of the copper-

inducible petE promoter and the plasmid was introduced into PCC 7120. Limited growth of a

small number of transconjugants on solid BG-11 medium containing ammonia and lacking

copper was observed. The resulting colonies lacked the green color of colonies of the wild

type, were composed primarily of heterocysts, and showed little to no growth in liquid

culture (data not shown). By comparison, a strain with PpetE driving transcription of the wild-

type allele hetR in place of the hetR(E254G) allele under the same conditions differentiated less

than 1% heterocysts. Replacement of the wild-type promoter region with a second

promoter, that of nirA, transcription from which is repressed in ammonia and induced in nitrate

or in the absence of fixed nitrogen [54], permitted the growth of filaments on solid and liquid BG-

11 medium with ammonia replacing nitrate as the nitrogen source. Apparently, there is tighter

on-off control of transcription with the nir promoter than with petE in our hands. Transfer of

filaments to BG-11 with nitrate or lacking a fixed source of nitrogen resulted in the differentiation

of greater than 90% of cells into heterocysts. By comparison, a strain with Pnir driving

transcription of the wild-type allele of hetR in place of the hetR(E254G) allele under the same

conditions differentiated about 30% heterocysts (data not shown). When the native copy of hetR

in PCC 7120 was replaced with an allele encoding the E254G substitution, about 25% of cells in

the resulting strain were heterocysts 48 h after induction. The phenotype of this strain was

indistinguishable from that of a strain with a copy of hetR encoding the more conservative

E254D substitution, which is discussed in more detail below.

212

Figure 1. Sensitivity to PatS-5 of strains with mutant alleles of hetR. (A) Bar graph of

the percentage of cells that are heterocysts in the wild-type strain (PCC 7120) and

strains with an allele of hetR encoding the indicated amino-acid substitution 96 h after

removal of combined nitrogen with (red-striped bars) or without (solid blue bars) the

addition of PatS-5 to the medium. Values represent the average of 500 cells from

three independent cultures. Strain PCC 7120 (B and C) and UHM167, which

contained an allele of hetR encoding an E254D substitution (D and E), 96 h after

removal of combined nitrogen without (B and D) or with (C and E) the addition of

PatS-5 to the medium. Carets indicate heterocysts.

213

Differentiation of nearly all cells of a filament has been observed when both patS and hetN

are inactivated simultaneously or when an allele of hetR encoding protein less sensitive to both

inhibitors is overexpressed ectopically and the mutant strains are grown in the absence of

combined nitrogen [34, 55]. To determine if the more conservative E254D substitution also

resulted in an overactive allele of HetR and if residues in the region of E254 were involved in

the response of HetR to PatS-5, alleles of HetR encoding individual conservative

substitutions at residues R250 – D256 were used to substitute hetR by allelic replacement in

PCC 7120 (see Table S1). Filaments of strains with alleles encoding R250K, E253D,

E254D, L255V, and D256E substitutions consisted of about 14 to 48% heterocysts, and the

presence or absence of fixed nitrogen in the medium had little effect on differentiation by an

individual strain. By comparison, about 9% of cells in filaments of PCC 7120 were heterocysts

in a medium that lacked fixed nitrogen, and about 1% when fixed nitrogen was present.

Conversely, A251G and L252V substitutions prevented or reduced differentiation, respectively

(Fig. 1). Strains that differentiated an increased number of heterocysts were also less

sensitive to PatS-5. Addition of PatS-5 to the growth medium prevented differentiation of

heterocysts by the wild-type strain, PCC 7120 [25]. In contrast, 8 to 25% of cells in

filaments of strains with alleles encoding R250K, E253D, E254D, L255V, and D256E

substitutions were heterocysts in a medium that contained PatS-5 at a concentration of 1 µM

(Fig. 1). As expected, PCC 7120 lacked heteroycsts under the same conditions. Taken

together, these results suggest that residues R250, E253, E254, L255 and D256 of HetR

are involved in sensitivity to PatS-dependent signals in vivo.

Evidence for direct binding of PatS-5 to HetR

The previous experiments pointed to amino acids at the C-terminus as being important for HetR

sensitivity to PatS. Guided by these in vivo observations, we designed in vitro experiments to

test the hypothesis that PatS-5 binds to HetR in the vicinity of amino acids 250-256. Initially, we

characterized the mobility and secondary structural environment of these amino acids using a

combination of site-directed spin labeling and continuous wave (CW) EPR spectroscopy to

provide a baseline for PatS-5 binding studies and then we used these techniques to determine

how the mobility of these residues were affected after the addition of PatS-5. CW EPR spectra

of MTSL spin-labeled proteins can be affected by interactions with other proteins, peptides or

small molecules (56-58) and can be monitored to identify changes in protein conformational

dynamics that accompany substrate binding [56-60]. To enable this approach, it was necessary

to employ site-specific mutagenesis and nitroxide spin labeling [61], whereby amino acids 250-

256 were individually mutated to cysteine and then labeled with the stable nitroxide free radical

(1-oxyl-2,2,5,5-tetramethyl- pyrrolin-3-yl)methyl methanethiosulfonate (MTSL). MTSL spin

214

labeling requires

substitution of the amino

acid at the desired site

with cysteine, and to

ensure that MTSL labeling

occurs only at one

position on the protein,

this should be the only

cysteine in the protein.

Therefore, the HetR

protein used to generate

CW EPR spectra had a

C48A substitution to

remove the naturally

occurring cysteine in

addition to substituting

amino acids 250-256

individually with cysteine.

The C48A substituted

protein resembled the

unmodified protein with

respect to binding of DNA

and response to PatS-5

peptide (see below). The

CW EPR spectrum of

each MTSL-labeled HetR

mutant (Fig. 2) was

measured to establish the

degree of conformational

mobility present at each

site in the absence of

PatS-5. Acquisition of these spectra were necessary to establish a baseline prior to looking for

changes in the conformational mobility of the spin label after addition of PatS-5, which, if

observed, would indicate a direct HetR PatS binding interaction. Control CW spectra of each

MTSL spin-labeled mutant were recorded at 100 µM protein, a concentration unlikely to exist

in cells but required for CW EPR spectra. Each mutant was titrated with 100 µM, 200 µM,

500 µM, and 1 mM PatS-5 and the CW EPR spectrum recorded after each addition.

Figure 2. Room temperature X-band CW-EPR spectra showing

variations in spin label mobility for HetR mutants in the range

250−256. The amino acid position of the MTSL label is indicated

above each spectrum. The solid blue line indicates experimental

data and the red dashed line is the best-fit simulation of the data.

Simulation parameters are included in Supporting Information

Table S4.

215

Simulations enabled determination of g-, A-, and R-tensor parameters (Table S2). CW EPR

spectra of MTSL labeled at positions R250C and L252C exhibited significant conformational

averaging and corresponding conformational dynamics based on their central resonance line

widths and isotropic rotational diffusion rates (∆Ho = 2.03 and 3.45 G, and log(Riso) = 7.73

and 6.53 sec-1

, respectively) indicating these residues likely occur either on surface exposed

helices or surface exposed loops [59, 62, 63]. MTSL at the A251C, E253C and D256C positions

were less mobile with time-scales of motion too slow to cause motional averaging of the

hyperfine interaction tensor (∆Ho values in the range 5.85-7.70 G and log(Riso) values in the

range of 5.54-5.60 sec-1

) indicating that they likely resided either at helix-helix contacts or in

buried locations within the protein [59, 62, 63]. The E254C and L255C mutants were difficult

to label, exhibited intrinsic protein instability and aggregation, gradually precipitated

over time, and produced poor EPR spectra, indicating that mutations at these sites disrupted the

local protein structure. The recently released HetR crystal structure from Fischerella (PDB ID:

3QOE, [64]) was used to examine the structural context of the MTSL spin- label in each HetR

cysteine mutant. The atomic coordinates for the structure were downloaded from the Protein

Data Bank and analyzed using the program VMD [52] in order to generate the initial C48A

mutation, followed by individual R250C-E256C cysteine mutations. Following substitutions,

MTSL groups were engineered onto the side chains, attached via the disulfide linkage to the free

sulfhydryl of cysteine using NAMD [51] and then simulated to achieve a local energy

minimization. The resulting structures, shown in Fig. 3 were rendered using Pymol molecular

visualization software. The general location of amino acids 250-256 can be seen in Fig. 3A in

which these residues are rendered using stick representation. A clear correspondence was

evident between the location of each spin label and the magnitude of the conformational mobility

based on the inspection and simulation of the CW EPR spectra. For example, the 250C mutant

exhibits a highly motionally averaged EPR spectrum (Fig. 2) that is consistent with the location

of residue 250 on the surface of the protein (Fig. 3B). The 251C mutant displays an EPR

spectrum devoid of motional averaging effects and the spin label appears buried in the model

(Fig. 3C). The 252C mutant exhibits significant motional averaging, consistent with the location

of the spin label on the surface of the protein (Fig. 3D). Likewise, the EPR spectra of the 253C

and 256C mutants show no signs of motional averaging consistent with the buried nature of

the spin label in the models (Figs. 3E and 3G).

The CW EPR spectrum of HetR spin-labeled at 252C changed dramatically after addition

of PatS-5 (Fig. 4), going from a distinctly motionally averaged spectrum in the absence of PatS-5

to a spectrum that showed no evidence of motional averaging in the presence of a ten-fold

molar excess of PatS-5. This observation indicated substantial quenching of the conformational

dynamics of the MTSL nitroxide electron radical upon addition of PatS-5. These observations

216

Figure 3. (A) Crystal structure of the HetR homodimer (PDB ID: 3QOE, [43]) from

Fischerella with the general location of residues 250−256 depicted by stick

representation, which were rendered using Pymol molecular visualization

software (The PyMOL Molecular Graphics System, Version 1.3, Schrodinger,

LLC.). Space filling representation of the MTSL group in HetR mutant (B) 250C,

(C) 251C, (D) 252C, (E) 253C, (F) 254C, (G) 255C, and (H) 256C.

217

Figure 4. CW-EPR spectra indicating direct binding between HetR and PatS-5.

The MTSL-labeled 252C HetR mutant was titrated with PatS-5 and monitored by

CW X-band EPR spectroscopy at room temperature. The control spectrum is

shown at the bottom. For the remaining spectra, the ratio of PatS-5 to HetR-252C

(indicated above each spectrum) increases from bottom to top in the stack. The

solid blue line indicates experimental data and the red dashed line is the best-fit

simulation of the data. Parameters for the EPR spectral simulations are included

in Supporting Information Table S4.

218

provided clear evidence for a direct binding interaction between HetR and PatS-5. At PatS-5 to

HetR-252C ratios < 10:1, a mixture of the free and bound states in slow exchange was evident

and the EPR spectra could be explained as a sum of spectra for free HetR mutant and HetR

mutant bound to PatS-5. At a 10:1 ratio, HetR-252C appeared predominantly in the PatS-5-

bound state. Simulations of the CW EPR spectra of the MTSL-labeled HetR-252C mutant

enabled determination of the change in the overall diffusion constant for spin-label motion

after binding to PatS-5, indicating almost an order of magnitude reduction in the diffusion rate in

the bound state (log(Riso) = 6.62 sec-1

in the free state compared to an upper limit of

log(Riso) = 5.92 sec-1

in the bound state (simulation parameters are in Table S3). Quenching

of the spin-label motion indicated either a proximal interaction with PatS-5 binding nearby, but

not necessarily at, amino acid 252, or a distal interaction in which PatS-5 could bind away from

the spin label, but the spin label motion could be quenched due to a tertiary or quaternary

change in protein structure that caused restriction of the spin label motion at amino acid 252.

Evidence for PatS-5 binding to the dimer form of HetR

The pulsed EPR double

electron electron resonance

(DEER) experiment [65, 66] can

be used to measure long-range

distances (from 20-70 Å)

between spin labels in large

proteins [67, 68]. Here, DEER

EPR experiments were used 1)

to measure the intermolecular

distance across the HetR

homodimer, 2) to determine if

PatS binding caused a change

in the quaternary structure of

the HetR dimer, and 3) to

determine if PatS-5 binding

disrupted the HetR dimer.

DEER data was initially

collected on the MTSL-labeled 252C HetR mutant. The time-domain DEER signal (Fig. 5A) had

an excellent signal to noise ratio and exhibited a strong modulation indicating substantial dipolar

coupling between the two MTSL spin labels across the dimer. Fourier transform of the time

Figure 5. DEER data for 252C HetR mutant. (A) The

time-domain DEER signal; (B) the frequency spectrum

resulting from the Fourier transform of the time-domain

data; (C) the best fit of the distance distribution that

explains the DEER data using a Tikhonov

regularization fitting procedure.

219

domain DEER modulation produced

a dipolar spectrum with well-

defined features (Fig. 5B).

Simulations of the time-domain

DEER data produced a clear peak

maximum in the distance

distribution indicating that the

distance between the MTSL spin

labels in each monomer was 2.7

nm (Fig. 5C). Observation of a

DEER signal in the MTSL spin-

labeled 252C mutant (which

contained the C48A mutation)

confirmed that C48 was not

necessary for HetR dimer formation,

since no DEER signal would be

detectable if the HetR mutant

existed as a monomer in solution.

In the presence of a 10-fold excess

of PatS-5, the DEER signal was

virtually unchanged compared to

the HetR alone sample,

demonstrating that PatS-5 binds

to HetR without disrupting the

HetR dimer. Furthermore, the

distance between the spin labels,

measured to be 2.6 nm in the presence of bound PatS-5, was the same within the uncertainty of

the measurement as the distance measured for HetR in the absence of bound PatS-5,

indicating that HetR binding to PatS-5 did not cause a structural rearrangement that resulted in

a change in the distance between the two spin-labeled 252C residues in the HetR dimer.

Furthermore, a DEER distance of 2.7 nm was measured when HetR 252C was bound to the 29

base pair hetP DNA fragment, also indicating that the HetR dimer also did not undergo a

structural rearrangement that changed the distance between the two 252C residues in the HetR

dimer upon DNA binding. The atomic coordinates of the crystal structure of HetR from

Fischerella (PDB ID: 3QOE, [64]) were used to measure the distance across the HetR dimer

from residue 252 (Fig. 6). The crystal structure was modified as discussed in the methods

section to create the 252C mutant modified by the MTSL spin label. The resulting modified

Figure 6. (A) Superposition of an ensemble of

structures of MTSLmodified 252C HetR mutant

generated by molecular dynamics simulation starting

from the crystal structure of the HetR homodimer

(PDB ID: 3QOE, [43]) from Fischerella as described

in the Materials and Methods. (B) Representation of

the average distance between the nitroxide groups of

the 252C-MTSL HetR mutant revealed an average

N−N distance of 26.5 Å and an average O−O

distance of 27.9 Å.

220

structure was subjected to molecular dynamics simulation to create an ensemble of structures,

which are depicted in Fig 6A. Average distances between the spin labels were analyzed using

the distance measurement tool in Pymol. Distance measurements between the nitroxide

groups of the 252C-MTSL revealed an average N-N distance of 26.5 Å and an average O-O

distance of 27.9 Å (Fig. 6B), consistent with our observed DEER distance of 27 Å.

Characterization of binding of HetR to a 29 bp inverted repeat containing DNA sequence

upstream of hetP

HetR was recently shown to bind

tightly to a 29 bp region upstream of

the hetP gene containing the

inverted repeat sequence 5´-

GAGGGGTCTAACCCCTC-3´ [69].

Indeed, HetR appears to bind more

tightly to this upstream hetP DNA

sequence than to any previously

used DNA substrate reported in the

literature, and thus, enabled us to

investigate the stoichiometry of

HetR binding to DNA. At a HetR to

DNA ratio of 3:1, the majority of the

DNA was shifted and approximately

half of the DNA was shifted at

between 2:1 and 3:1 (Fig. 7). This

data is consistent with HetR binding

the upstream hetP DNA sequence

with a 2:1 stoichiometry, since the

extent of completion of the gel shift

depends on the dissociation

constant, Kd. All these data taken

together, along with the fact that most prokaryotic transcription factors bind inverted repeat DNA

sequences as homodimers (70-72), suggests that HetR binds this single inverted- repeat

containing upstream hetP DNA sequence as a homodimer. Interestingly, at HetR to DNA ratios

greater than 4:1, a supershifted species was also present (Fig. 7) possibly caused by two HetR

homodimers binding to a single 29 base pair DNA molecule. The amount of supershifted species

depended on the amount of NaCl in solution, and the most DNA in the supershifted band

Figure 7. EMSA showing binding of HetR to the 29

base pair inverted repeat containing upstream hetP

DNA fragment. The ratio of HetR to DNA increases

from left to right. The resulting completeness of

shifted DNA was used to assess the stoichiometry of

binding between HetR and DNA as one molecule of

HetR dimer per double stranded DNA molecule. The

box contains the supershifted bands observed at high

HetR to DNA ratios.

221

compared to the shifted band was observed when using 0.5 M NaCl in the solution (data not

shown).

Sensitivity of the HetR-DNA complex to PatS-5

The strong interaction between HetR

and DNA from the hetP promoter

region permitted assessment of the

PatS-5 to HetR ratios required to

disrupt HetR binding to DNA. In gel

shift assays it was found that PatS-5

completely disrupted the supershifted

species at a substoichiometric ratio of

0.5:1 PatS-5 to HetR monomer and

the shifted species was completely

disrupted at a 10:1 ratio of PatS-5 to

HetR (Fig. 8). When PatS-5 was

titrated into a solution of the MTSL-

labeled HetR 252C-DNA complex,

the CW EPR spectra were the same

as HetR plus PatS-5 alone (Fig. S1

and Table S3). This result is

consistent with the gel shift assay

showing that the HetR-DNA

interaction is disrupted when HetR

binds to PatS-5. In other words, in a

solution containing HetR, PatS-5 and DNA, only a complex of HetR and PatS-5 would exist and

the DNA would be free in solution.

Similarity of HetR and HetR-C48A

As mentioned above, HetR containing a C48A substitution was used in the CW EPR work that

showed binding of PatS-5. It was therefore important to examine whether or not C48A substituted

HetR behaved similarly to wild-type protein. Huang et al. [29] previously reported that HetR

residue C48 was required for both dimer formation and DNA binding. Here, size exclusion

chromatography analysis of HetR-C48A showed that it eluted identically to wild type HetR

indicating that the proposed intermolecular cysteine disulfide bond was not required for dimer

formation (Fig. S1). The DEER EPR data confirmed that the C48A HetR spin-labeled mutant

existed as a dimer in solution. Comparison of wild type HetR and HetR-C48A using PAGE (Fig.

Figure 8. EMSA showing binding of HetR to the 29

base pair inverted repeat containing upstream hetP

DNA fragment. The ratio of HetR to DNA increases

from left to right. The resulting completeness of

shifted DNA was used to assess the stoichiometry of

binding between HetR and DNA as one molecule of

HetR dimer per double stranded DNA molecule. The

box contains the supershifted bands observed at high

HetR to DNA ratios.

222

9A) showed that the wild type HetR

exhibited a strong dimer band even

under denaturing conditions (0.1%

SDS in the protein buffer) and in the

presence of 10 mM dithiothreitol,

whereas the dimer band was almost

completely gone in the C48A mutant

under the same conditions. Huang et

al. [29] reported that the dimer band

was completely gone for the C48A

HetR mutant in SDS PAGE and

concluded that HetR-C48A was

unable to form dimers in vitro, that

the HetR dimer was formed through

a disulfide bond and that Cys-48 was

required for dimerization. We were,

however, able to detect a dimer band

for HetR-C48A under milder

denaturing conditions using 0.01%

SDS-PAGE (Fig. 9B). Taken

together with the size exclusion

chromatography and DEER data,

these results indicated that C48 was

not required for dimer formation,

however the dimer may have been stabilized by an interchain disulfide bond. In addition, circular

dichroism spectra of wild type HetR and the C48A HetR mutant were identical and characteristic

of a protein with predominantly α-helical secondary structure as indicated by distinct minima in

molar ellipticity at 209 and 222 nm (wild type HetR CD spectrum shown in Fig. S2).

In addition to having the same physical characteristics as the wild-type protein, C48A substituted

HetR behaved similarly to the wild type HetR in gel shift assays. At the same concentrations as

those described above, both shifted and supershifted species were observed, however the

supershifted band tended to be less prominent compared to that generated with wild type HetR.

The HetR-252C mutant also shifted the DNA in the same manner as the C48A mutant of HetR

(data not shown). When PatS-5 was included in the binding reactions, binding of the C48A HetR

mutant to the 29 bp hetP promoter region was affected in a manner similar to that for the wild

type protein, suggesting that the C48A substitution does not affect binding of HetR to DNA or its

interaction with PatS-5 peptide.

Figure 9. Polyacrylamide gel electrophoresis

analysis of soluble-expressed purified recombinant

HetR and C48A HetR mutant. (A) At 0.1% SDS and

10 mM dithiothreitol. Molecular weight marker (Fisher

Bioreagents EZ-run Rec Protein Ladder) are

indicated in units of kilodaltons (kDa) to the left of the

lane marked M. (B) At 0.01% SDS and 10 mM

dithiothreitol. Lanes are marked “HetR” for wild-type

HetR and “C48A” for the C48A mutant of HetR.

223

Determination of the stoichiometry, dissociation constant, and thermodynamic parameters

for binding of PatS-5 to HetR

Isothermal titration calorimetry

(ITC) is a highly sensitive

biophysical technique

designed for analyzing the

thermodynamics of molecular

interactions. ITC can be used

to study protein-protein or

protein-small molecule

binding interactions and is

capable of detecting heats

of binding directly, without

the use of any molecular

labels or tags that might

interfere with or influence

binding [73]. Here, ITC was

used to confirm direct binding

of PatS-5 to HetR (Fig. 10).

The integrated binding heats

were indicative of a single

exothermic binding event. A

single binding site model was

used to fit the binding curves,

which allowed determination

of the binding stoichiometry n,

Ka (equal to 1/Kd) and ∆H.

The ∆G was determined from

Ka, and ∆S was then

calculated using the ∆G and

∆H values. The data indicated

that one PatS-5 peptide binds

to one HetR monomer, i.e.

two peptides bind per

homodimer. It has been previously shown that calculations using a model for two non-interacting

identical binding sites of a homodimeric protein binding to two identical peptides are essentially

the same as one peptide binding-site for each monomer [74]. PatS-5 was found to bind HetR

Figure 10. Polyacrylamide gel electrophoresis analysis of

soluble-expressed purified recombinant HetR and C48A

HetR mutant. (A) At 0.1% SDS and 10 mM dithiothreitol.

Molecular weight marker (Fisher Bioreagents EZ-run Rec

Protein Ladder) are indicated in units of kilodaltons (kDa) to

the left of the lane marked M. (B) At 0.01% SDS and 10 mM

dithiothreitol. Lanes are marked “HetR” for wild-type HetR

and “C48A” for the C48A mutant of HetR.

224

with a mean Kd of 227 ± 23 nM, ∆G of -9.063 ± 0.059 kcal/mol, and n of 1.04 ± 0.01 sites,

indicating a single tight exothermic binding interaction. PatS-5 binding to HetR exhibited a

relatively large negative enthalpy (-8.337 ± 0.410) and a positive entropy (2.44± 1.55

cal/(mol·deg)). Inspection of Table 1 shows that the binding affinity of HetR for PatS-5 was

dominated by the change in enthalpy, in comparison to the relatively small change (more than 10-

fold smaller) in T∆S. This indicates that binding of PatS-5 to HetR is enthalpically driven as

opposed to an entropically driven process. It is difficult to elucidate the relative contributions to

binding of PatS-5, however, the relatively large negative enthalpy of binding is consistent with the

formation of hydrogen bonds or ionic interactions, which are associated with relatively large

negative contributions to enthalpy for ligands binding to proteins [75-77]. Collectively, the

isothermal titration calorimetry data corroborate the EPR observations that PatS-5 binds directly

to HetR in the absence of DNA.

Table 1. Summary of Quantities Associated with HetR Binding to PatS-5 Peptide Derived from

Isothermal Titration Calorimetry Experiments

DISSCUSSION

Since the original report of PatS as a diffusible inhibitor of heterocyst differentiation [25], the

specific partners that interact directly in the process of PatS-dependent heterocyst regulation

have been difficult to identify, both in vitro and in vivo. While multiple groups have

demonstrated that the RGSGR carboxyl terminus of PatS is capable of inhibiting DNA binding of

HetR to a region upstream from its own promoter in vitro [29, 32], and a R223W mutant of HetR

has been shown to be insensitive to in vivo overexpression of both PatS and HetN [34], all of

these observations have provided only indirect evidence of a direct interaction between HetR and

PatS. The specific partners that interact directly during PatS disruption of HetR binding to DNA,

and the direct molecular-level interactions that cause the loss of R223W HetR sensitivity to PatS,

have not been unambiguously elucidated with support of experimental data. Specifically

with regards to the gel shift experiments, no experimental data has been reported in the literature

to distinguish between the two following possibilities: 1) that the disruption is caused by PatS-5

displacing the DNA upon binding directly to HetR, or 2) that PatS-5 displaces HetR upon binding

directly to the DNA. Resolving these issues has been impossible using traditional techniques

used to observe direct intermolecular interactions, such as gel shift assays and size exclusion

chromatography assays to detect HetR binding directly to PatS-5 in the absence of DNA,

225

because these techniques lack sufficient resolution to detect shifts in the migration of either the

70 kDa HetR dimer or the > 20 kDa DNA fragments upon binding to the PatS-5 pentapeptide,

which adds only about 0.6 kDa to the overall complex molecular weight. Moreover, PatS lacks

any naturally occurring spectroscopic probe that could be used to detect binding between HetR

and PatS-5.

In this manuscript, we report the first unequivocal experimental data that shows direct binding

between HetR and PatS-5. Two different and independent techniques corroborated that PatS-5

binds directly to HetR, namely EPR spectroscopy and isothermal titration calorimetry. The

conclusion from the EPR experiments was based on the observation that the motion of the MTSL

spin-label on the HetR 252C mutant was quenched upon titration with PatS-5 in the absence of

any DNA in solution. This observation clearly indicated a direct binding interaction between the

mutant spin-labeled HetR and PatS-5. However, these data did not unambiguously

indicate that PatS-5 bound in the immediate vicinity of residue 252, since it is possible that PatS-5

could have bound to HetR in a site remote from residue 252 and caused quenching of spin label

motion through an allosteric effect or change in tertiary structure. However, given that in vivo

experiments indicated that HetR residues 250-256 were critical for sensitivity to PatS-5, it would

not be surprising if PatS-5 bound HetR in the C-terminal domain in the vicinity of amino acids

250-256. Besides demonstrating that PatS-5 binds directly to HetR, the DEER EPR experiments

also demonstrated that HetR persisted as a dimer even after binding PatS-5. The isothermal

titration calorimetry data independently corroborated that PatS-5 binds directly to HetR, using the

native HetR protein, without the mutation or MTSL spin-label required for the EPR experiments,

and firmly established that the stoichiometry of binding is one PatS-5 peptide per HetR monomer,

or said differently, two molecules of PatS-5 bind to each HetR dimer. By employing isothermal

titration calorimetry, we were not only able to corroborate that PatS-5 binds directly to HetR in the

absence of DNA, but we were also able to measure its dissociation constant and determine

thermodynamic parameters, which indicated that binding of PatS-5 to HetR was enthalpically

driven. The high affinity of HetR for the 29 base pair inverted repeat containing upstream hetP

DNA sequence enabled a semi-quantitative analysis of the stoichiometry of binding of HetR to

this DNA fragment, which indicated HetR likely binds this DNA at a 2:1 ratio, i.e. HetR binds the

inverted repeat containing upstream hetP DNA sequence as a homodimer. Finally, based on the

determination that the shifted species was a 2:1 complex of HetR to DNA, we concluded that the

supershifted complex possibly consisted of two HetR homodimers bound to a single hetP DNA

sequence, which is consistent with the observation that the supershifted species was only

detected when the ratio of HetR to DNA was greater than 4:1. Again, despite detecting the

supershifted species, the biological relevance of this species, if any, remains to be determined.

226

The biophysical data was driven by the in vivo studies with mutant alleles of hetR. The

phenotypes of strains with alleles of hetR encoding substitutions at residues R250, E253, E254,

L255 and D256 are consistent with decreased sensitivity of the substituted HetR proteins to PatS-

and HetN-dependent inhibitory signals. First, both patS and hetN null mutants as well as the

strains with overactive alleles described here differentiate an increased number of heterocysts

relative to that of the wild type, PCC 7120 [25, 26]. Second, a patS null mutant and the strains

described here with overactive alleles differentiate heterocysts in the presence of a fixed source

of nitrogen. And third, inactivation of both patS and hetN simultaneously results in the formation

of a number of heterocysts in excess of that made by either of the individual mutants [55], similar

to the strains with alleles of hetR encoding E253D, E254D, L255V, and D256E substitutions.

ACKNOWLEDGEMENTS

The authors thank Professor Susan Barnum for providing the genomic DNA of Anabaena sp.

strain PCC 7120 and Kelly Higa for insightful discussions. Bryan J. Glaenzer, Alisha N. Jones,

and Andrea F. Schilling are acknowledged for contributing to early stages of characterization of

soluble HetR. This work was supported in part by a grant from NSF (IOS-0919878) to SMC.

FOOTNOTES

1 Corresponding Author. Michael A. Kennedy, Department of Chemistry and Biochemistry,

Hughes Laboratories, Room 106, Miami University, 701 High Street, Oxford, OH 45056. Email:

[email protected]. Phone: 513-529-8267. Fax: 513-529-5715

227

REFERENCES CITED

1. Fritsch, F. (1904) Studies on cyanophyceae, The New Phytologist 3, 85-96.

2. Fogg, G. (1944) Growth and heterocyst production in Anabaena cylindrica Lemm, New

Phytologist 43, 164-175.

3. Fogg, G. (1949) Growth and heterocyst production in Anabaena Cylindrica Lemm. II in

relation to carbon and nitrogen metabolism, Annals of Botany 13, 241-259.

4. Haselkorn, R. (1978) Heterocysts, Annual Review of Plant Physiology and Plant

Molecular Biology, 319-344.

5. Adams, D. G., and Carr, N. G. (1981) The developmental biology of heterocyst and

akinete formation in cyanobacteria, Crit Rev Microbiol 9, 45-100.

6. Wolk, C. P. (1996) Heterocyst formation, Annu Rev Genet 30, 59-78.

7. Adams, D. G. (2000) Heterocyst formation in cyanobacteria, Curr Opin Microbiol 3, 618-

624.

8. Meeks, J. C., and Elhai, J. (2002) Regulation of cellular differentiation in filamentous

cyanobacteria in free-living and plant-associated symbiotic growth states., Microbiol Mol

Biol Rev 66, 94-121; table of contents.

9. Golden, J. W., and Yoon, H. S. (2003) Heterocyst development in Anabaena., Curr Opin

Microbiol 6, 557-563.

10. El-Shehawy, R., and Kleiner, D. (2003) The mystique of irreversibility in cyanobacterial

heterocyst formation: parallels to differentiation and senescence in eukaryotic cells,

Physiologia Plantarum, 49-55.

11. Zhang, C. C., Laurent, S., Sakr, S., Peng, L., and Bédu, S. (2006) Heterocyst

differentiation and pattern formation in cyanobacteria: a chorus of signals, Mol Microbiol

59, 367-375.

12. Kumar, K., Mella-Herrera, R. A., and Golden, J. W. (2010) Cyanobacterial heterocysts,

Cold Spring Harb Perspect Biol 2, a000315.

13. Brocks, J. J., Logan, G. A., Buick, R., and Summons, R. E. (1999) Archean molecular

fossils and the early rise of eukaryotes, Science 285, 1033-1036.

228

14. Tomitani, A., Knoll, A. H., Cavanaugh, C. M., and Ohno, T. (2006) The evolutionary

diversification of cyanobacteria: molecular-phylogenetic and paleontological perspectives,

Proc Natl Acad Sci U S A 103, 5442-5447.

15. Fay, P. (1992) Oxygen relations of nitrogen fixation in cyanobacteria, Microbiol Rev 56,

340-373.

16. Gallon, J. (1992) Reconciling the incompatible: N2 fixation and O2, New Phytologist 122,

571-609.

17. Haselkorn, R. (2008) Cell-cell communication in filamentous cyanobacteria, Mol


18. Fritsch, F. (1951) The heterocyst: a botanical enigma, Proceedings of the Linnean

Society of London 162, 194-211.

19. Wolk, C. P., and Quine, M. P. (1975) Formation of one-dimensional patterns by

stochastic processes and by filamentous blue-green algae, Dev Biol 46, 370-382.

20. Lambein, F., and Wolk, C. P. (1973) Structural studies on the glycolipids from the

envelope of the heterocyst of Anabaena cylindrica, Biochemistry 12, 791-798.

21. Khudyakov, I., and Wolk, C. P. (1997) hetC, a gene coding for a protein similar to

bacterial ABC protein exporters, is involved in early regulation of heterocyst differentiation

in Anabaena sp. strain PCC 7120, J Bacteriol 179, 6971-6978.

22. Nicolaisen, K., Hahn, A., and Schleiff, E. (2009) The cell wall in heterocyst formation by

Anabaena sp.PCC 7120, J Basic Microbiol 49, 5-24.

23. Buikema, W. J., and Haselkorn, R. (1991) Characterization of a gene controlling

heterocyst differentiation in the cyanobacterium Anabaena 7120, Genes Dev 5, 321-330.

24. Black, T. A., Cai, Y., and Wolk, C. P. (1993) Spatial expression and autoregulation of

hetR, a gene involved in the control of heterocyst development in Anabaena, Mol

Microbiol 9, 77-84.

25. Yoon, H. S., and Golden, J. W. (1998) Heterocyst pattern formation controlled by a

diffusible peptide, Science 282, 935-938.

26. Callahan, S. M., and Buikema, W. J. (2001) The role of HetN in maintenance of the

heterocyst pattern in Anabaena sp. PCC 7120, Mol Microbiol 40, 941-950.

229

27. Zhang, J. Y., Chen, W. L., and Zhang, C. C. (2009) hetR and patS, two genes necessary

for heterocyst pattern formation, are widespread in filamentous nonheterocyst-forming

cyanobacteria, Microbiology 155, 1418-1426.

28. Risser, D. D., and Callahan, S. M. (2009) Genetic and cytological evidence that

heterocyst patterning is regulated by inhibitor gradients that promote activator decay,


29. Huang, X., Dong, Y., and Zhao, J. (2004) HetR homodimer is a DNA-binding protein

required for heterocyst differentiation, and the DNA-binding activity is inhibited by PatS,


30. Zhou, R., Wei, X., Jiang, N., Li, H., Dong, Y., Hsi, K. L., and Zhao, J. (1998) Evidence

that HetR protein is an unusual serine-type protease, Proc Natl Acad Sci U S A 95, 4959-

4963.

31. Zhou, R., Cao, Z., and Zhao, J. (1998) Characterization of HetR protein turnover in

Anabaena sp. PCC 7120, Arch Microbiol 169, 417-423.

32. Risser, D. D., and Callahan, S. M. (2007) Mutagenesis of hetR reveals amino acids

necessary for HetR function in the heterocystous cyanobacterium Anabaena sp. strain

PCC 7120, J Bacteriol 189, 2460-2467.

33. Wu, X., Liu, D., Lee, M. H., and Golden, J. W. (2004) patS minigenes inhibit heterocyst

development of Anabaena sp. strain PCC 7120, J Bacteriol 186, 6422-6429.

34. Khudyakov, I. Y., and Golden, J. W. (2004) Different functions of HetR, a master

regulator of heterocyst differentiation in Anabaena sp. PCC 7120, can be separated by

mutation, Proc Natl Acad Sci U S A 101, 16040-16045.

35. Reeves, R., and Nissen, M. S. (1990) The A.T-DNA-binding domain of mammalian high

mobility group I chromosomal proteins. A novel peptide motif for recognizing DNA

structure, J Biol Chem 265, 8573-8582.

36. Banks, G. C., Mohr, B., and Reeves, R. (1999) The HMG-I(Y) A.T-hook peptide motif

confers DNA- binding specificity to a structured chimeric protein, J Biol Chem 274,

16536-16544.

37. Goodwin, G. H., and Johns, E. W. (1973) Isolation and characterisation of two calf-

thymus chromatin non-histone proteins with high contents of acidic and basic amino

acids, Eur J Biochem 40, 215-219.

230

38. Goodwin, G. H., Sanders, C., and Johns, E. W. (1973) A new group of chromatin-

associated proteins with a high content of acidic and basic amino acids, Eur J Biochem

38, 14-19.

39. Grosschedl, R., Giese, K., and Pagel, J. (1994) HMG Domain Proteins - Architectural

Elements in the Assembly of Nucleoprotein Structures, Trends in Genetics, 94-100.

40. Reeves, R., Leonard, W. J., and Nissen, M. S. (2000) Binding of HMG-I(Y) imparts

architectural specificity to a positioned nucleosome on the promoter of the human

interleukin-2 receptor alpha gene, Mol Cell Biol 20, 4666-4679.

41. Huth, J. R., Bewley, C. A., Nissen, M. S., Evans, J. N., Reeves, R., Gronenborn, A. M.,

and Clore, G. M. (1997) The solution structure of an HMG-I(Y)-DNA complex defines a

new architectural minor groove binding motif, Nat Struct Biol 4, 657-665.

42. Elhai, J., and Wolk, C. P. (1988) Conjugal transfer of DNA to cyanobacteria, Methods

Enzymol 167, 747-754.

43. Limphong, P., Crowder, M. W., Bennett, B., and Makaroff, C. A. (2009) Arabidopsis

thaliana GLX2-1 contains a dinuclear metal binding site, but is not a glyoxalase 2,

Biochem J 417, 323-330.

44. Schneider, D., and Freed, J. (1989) Calculating slow motional magnetic resonance

spectra: A user's guide, In Biological Magnetic Resonance (Berliner, L., and Reuben, J.,

Eds.), pp 1-76, Plenum, New York, NY.

45. Budil, D., Lee, S., Saxena, S., and Freed, J. (1996) Nonlinear-least-squares analysis of

slow-motion EPR spectra in one and two dimensions using a modified Levenberg-

Marquardt algorithm, Journal of Magnetic Resonance Series a, 155-189.

46. Hoofnagle, A. N., Stoner, J. W., Lee, T., Eaton, S. S., and Ahn, N. G. (2004)

Phosphorylation-dependent changes in structure and dynamics in ERK2 detected by

SDSL and EPR, Biophys J 86, 395-403.

47. Zhang, Z., Fleissner, M. R., Tipikin, D. S., Liang, Z., Moscicki, J. K., Earle, K. A., Hubbell,

W. L., and Freed, J. H. (2010) Multifrequency electron spin resonance study of the

dynamics of spin labeled T4 lysozyme, J Phys Chem B 114, 5503-5521.

48. Nesmelov, Y. E., Agafonov, R. V., Burr, A. R., Weber, R. T., and Thomas, D. D. (2008)

Structure and dynamics of the force-generating domain of myosin probed by

multifrequency electron paramagnetic resonance, Biophys J 95, 247-256.

231

49. Jeschke, G., Chechik, V., Ionita, P., Godt, A., Zimmermann, H., Banham, J., Timmel, C.,

Hilger, D., and Jung, H. (2006) DeerAnalysis2006 - a comprehensive software package

for analyzing pulsed ELDOR data, Applied Magnetic Resonance, 473-498.

50. Chiang, Y. W., Borbat, P. P., and Freed, J. H. (2005) The determination of pair distance

distributions by pulsed ESR using Tikhonov regularization, J Magn Reson 172, 279-295.

51. Phillips, J. C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., Chipot, C.,

Skeel, R. D., Kalé, L., and Schulten, K. (2005) Scalable molecular dynamics with NAMD,

J Comput Chem 26, 1781-1802.

52. Humphrey, W., Dalke, A., and Schulten, K. (1996) VMD: visual molecular dynamics, J

Mol Graph 14, 33-38, 27-38.

53. Polyhach, Y., Bordignon, E., and Jeschke, G. (2011) Rotamer libraries of spin labelled

cysteines for protein studies, Phys Chem Chem Phys 13, 2356-2366.

54. Frías, J. E., Flores, E., and Herrero, A. (1997) Nitrate assimilation gene cluster from the

heterocyst- forming cyanobacterium Anabaena sp. strain PCC 7120, J Bacteriol 179,

477-486.

55. Borthakur, P. B., Orozco, C. C., Young-Robbins, S. S., Haselkorn, R., and Callahan, S. M.

(2005) Inactivation of patS and hetN causes lethal levels of heterocyst differentiation in

the filamentous cyanobacterium Anabaena sp. PCC 7120, Mol Microbiol 57, 111-123.

56. Klug, C. S., Eaton, S. S., Eaton, G. R., and Feix, J. B. (1998) Ligand-induced

conformational change in the ferric enterobactin receptor FepA as studied by site-directed

spin labeling and time-domain ESR, Biochemistry 37, 9016-9023.

57. Buchaklian, A. H., and Klug, C. S. (2005) Characterization of the Walker A motif of MsbA

using site- directed spin labeling electron paramagnetic resonance spectroscopy,

Biochemistry 44, 5503-5509.

58. Hanson, S. M., Francis, D. J., Vishnivetskiy, S. A., Kolobova, E. A., Hubbell, W. L., Klug,

C. S., and Gurevich, V. V. (2006) Differential interaction of spin-labeled arrestin with

inactive and active phosphorhodopsin, Proc Natl Acad Sci USA 103, 4900-4905.

59. Fanucci, G. E., and Cafiso, D. S. (2006) Recent advances and applications of site-

directed spin labeling, Curr Opin Struct Biol 16, 644-653.

60. Klug, C. S., and Feix, J. B. (2008) Methods and applications of site-directed spin

labeling EPR spectroscopy, Methods Cell Biol 84, 617-658.

232

61. Altenbach, C., Marti, T., Khorana, H. G., and Hubbell, W. L. (1990) Transmembrane

protein structure: spin labeling of bacteriorhodopsin mutants, Science 248, 1088-1092.

62. Mchaourab, H. S., Lietzow, M. A., Hideg, K., and Hubbell, W. L. (1996) Motion of spin-

labeled side chains in T4 lysozyme. Correlation with protein structure and dynamics,

Biochemistry 35, 7692-7704.

63. Mchaourab, H. S., Kálai, T., Hideg, K., and Hubbell, W. L. (1999) Motion of spin-labeled

side chains in T4 lysozyme: effect of side chain structure, Biochemistry 38, 2947-2955.

64. Kim, Y., Joachimiak, G., Ye, Z., Binkowski, T. A., Zhang, R., Gornicki, P., Callahan, S. M.,

Hess, W. R., Haselkorn, R., and Joachimiak, A. (2011) Structure of transcription

factor HetR required for heterocyst differentiation in cyanobacteria, Proc Natl Acad Sci

U S A 108, 10109-10114.

65. Milov, A., Ponomarev, A., and Tsvetkov, Y. (1984) Electron electron double-resonance in

electron-spin echo - model biradical systems and the sensitized photolysis of decalin,

Chemical Physics Letters, 67-72.

66. Milov, A., Salikhov, K., and Shirov, M. (1981) Application of ELDOR in electron-spin

echo for paramagnetic center space distribution in solids, Fiz Tverd Tela (Leningrad) 23,

957-982.

67. Borbat, P. P., Davis, J. H., Butcher, S. E., and Freed, J. H. (2004) Measurement of large

distances in biomolecules using double-quantum filtered refocused electron spin-echoes,

J Am Chem Soc 126, 7746-7747.

68. Borbat, P., and Freed, J. (2001) Double quantum ESR and distance measurements,

Kluwer, New York, NY.

69. Higa, K. C., and Callahan, S. M. (2010) Ectopic expression of hetP can partially bypass

the need for hetR in heterocyst differentiation by Anabaena sp. strain PCC 7120, Mol


70. Harrison, S. C. (1991) A structural taxonomy of DNA-binding domains, Nature 353, 715-

719.

71. Latchman, D. S. (1997) Transcription factors: an overview, Int J Biochem Cell Biol 29, 1

305-1312.

72. Huffman, J. L., and Brennan, R. G. (2002) Prokaryotic transcription regulators: more than

just the helix- turn-helix motif, Curr Opin Struct Biol 12, 98-106.

233

73. Feig, A. L. (2007) Applications of isothermal titration calorimetry in RNA

biochemistry and biophysics, Biopolymers 87, 293-301.

74. Streicher, W. W., Lopez, M. M., and Makhatadze, G. I. (2009) Annexin I and annexin II N-

terminal peptides binding to S100 protein family members: specificity and thermodynamic

characterization, Biochemistry 48, 2788-2798.

75. Connelly, P. R., Aldape, R. A., Bruzzese, F. J., Chambers, S. P., Fitzgibbon, M. J.,

Fleming, M. A., Itoh, S., Livingston, D. J., Navia, M. A., and Thomson, J. A. (1994)

Enthalpy of hydrogen bond formation in a protein-ligand binding reaction, Proc Natl Acad

Sci U S A 91, 1964-1968.

76. Calderone, C. T., and Williams, D. H. (2001) An enthalpic component in cooperativity: the

relationship between enthalpy, entropy, and noncovalent structure in weak associations,

J Am Chem Soc 123, 6262-6267.

77. Williams, D. H., Stephens, E., O'Brien, D. P., and Zhou, M. (2004) Understanding

noncovalent interactions: ligand binding energy and catalytic efficiency from ligand-

induced reductions in motion within receptors and enzymes, Angew Chem Int Ed Engl 43,

6596-6616.

234

APPENDIX C. PEER REVIEWED PUBLICATION: The RGSGR amino-acid motif of the

intercellular signaling protein, HetN, is required for patterning of heterocysts in Anabaena

sp. strain PCC 7120

MOLECULAR MICROBIOLOGY Vol. 83, p. 682-693

Copyright © 2012, Blackwell Publishing Ltd. All Rights Reserved.

Kelly C. Higa1,*, Ramya Rajagopalan

2,*, Douglas D. Risser

3,*, Orion S. Rivers*, Sasa K. Tom,

Patrick Videau, Sean M. Callahan#

University of Hawaii, Department of Microbiology, Honolulu, HI 96822

Accepted 8 December 2011

ABSTRACT

Nitrogen-fixing heterocysts are arranged in a periodic pattern on filaments of the cyanobacterium

Anabaena sp. strain PCC 7120 under conditions of limiting combined nitrogen. Patterning

requires two inhibitors of heterocyst differentiation, PatS and HetN, which work at different stages

of differentiation by laterally suppressing levels of an activator of differentiation, HetR, in cells

adjacent to source cells. Here we show that the RGSGR sequence in the 287 amino acid HetN

protein, which is shared by PatS, is critical for patterning. Conservative substitutions in any of the

five amino acids lowered the extent to which HetN inhibited differentiation when over-produced

and altered the pattern of heterocysts in filaments with an otherwise wild-type genetic background.

Conversely, substitution of amino acids comprising the putative catalytic triad of this predicted

reductase had no effect on inhibition or patterning. Deletion of putative domains of HetN

suggested that the RGSGR motif is the primary component of HetN required for both its inhibitory

and patterning activity, and that localization to the cell envelope is not required for patterning of

heterocysts. The intercellular signaling proteins PatS and HetN use the same amino-acid motif to

regulate different stages of heterocyst patterning.

INTRODUCTION

Intercellular signaling regulates a variety of processes in bacteria, most notably as part of

quorum-sensing systems. In quorum sensing, individual cells communicate with short peptides or

other small molecules that are secreted into the surrounding medium to coordinate the behavior

235

of multiple cells. As a counterpoint to quorum sensing, signaling between cells in direct physical

contact has the potential to employ intercellular signals that pass directly between cells, without

an intermediate extracellular stage. Examples include signaling during the formation of transient

multicellular structures, such as the development of spores in Bacillus subtilis and in the fruiting

bodies elaborated by Myxococcus xanthus, or in bacteria that grow as multicellular organisms,

such as filamentous streptomycetes and cyanobaceria. In the latter of these examples, inhibitors

of cellular differentiation in filamentous cyanobacteria are thought to move from cell to cell and

govern the periodic patterning of heterocysts that form in response to deprivation of fixed nitrogen.

Anabaena sp. strain PCC 7120 is a filamentous cyanobacterium that can be induced to form a

periodic pattern of nitrogen-fixing heterocysts from an unbranched chain of undifferentiated

vegetative cells. Heterocysts occur, on average, at 10-cell intervals, are terminally differentiated,

and differ from vegetative cells morphologically, metabolically, and genetically [1, 2]. They have a

single known function, to supply the remaining vegetative cells of the filament with a bioactive

form of nitrogen, which allows growth of the organism in nitrogen-poor environments. The

elaboration of heterocysts facilitates the spatial separation of an oxygen-labile metabolic process,

nitrogen fixation, from one that evolves molecular oxygen, photosynthesis with photosystem II.

Fixed nitrogen is supplied to vegetative cells from heterocysts, and in return, heterocysts receive

a source of carbon and reductant to compensate for their lack of PS II and the Calvin cycle [3, 4].

Experimental evidence for the phenomenon of “lateral inhibition” of differentiation to explain how

Anabaena “counts to 10” was first described by Wolk in 1967 [5, 6]. More recently, patterning of

differentiation has been shown to be dependent on an activator of differentiation, HetR, and two

inhibitors of differentiation, PatS and HetN, that have properties of intercellular signaling

molecules. HetR is at the center of a regulatory circuit that shares the properties of biological

switches, which turn graded input signals into a binary output: when the switch is “off”, the cell

remains undifferentiated, but when the switch is “turned on”, the differentiation process begins

and eventually becomes irreversible and self-sustaining [7, 8]. HetN or PatS produced in one cell

lowers levels of HetR in adjacent cells [9], and it has been proposed, but not demonstrated, that

the relative positions of cells is conveyed by concentration gradients of PatS and/or HetN

extending from source cells [10, 11].

HetR has prominent roles in both the patterning and differentiation of heterocysts. In a wild-type

genetic background, expression of hetR is both necessary and sufficient for differentiation. Its

inactivation prevents development of heterocysts, and over-expression of hetR leads to the

formation of multiple contiguous heterocysts (Mch phenotype), even in the presence of

concentrations of nitrate or ammonia that suppress differentiation in the wild type [12, 13]. HetR

236

has DNA-binding activity, which is necessary for heterocyst formation [14], suggesting that its

primary role in promoting differentiation is the regulation of transcription of other developmental

genes. The protein has also been shown to bind directly to the promoter regions of 5 genes, patS,

hepA, hetP, pknE and hetR [14-16]. Induction of hetR by HetR is an example of direct positive

auto-regulation, predicted by patterning models, and the mutual dependence of ntcA, which is

also necessary for differentiation, and hetR creates a second, indirect example of positive auto-

regulation [17]. The recent crystal structure of HetR reveals a dimer with a central DNA-binding

region, two outward facing flaps, and a hood-like domain over the core of the protein [18]. It was

suggested that HetR might serve as a scaffold for the assembly of transcription initiation

complexes.

The patS gene is predicted to encode a 13 or 17 amino acid protein that is presumably processed

to a smaller, active form, perhaps during export from the cytoplasm to the periplasm of the cell or

directly to a neighboring cell [11]. The active form of PatS has yet to be identified, and a

presumed gradient of PatS has not been demonstrated. However, when confined to the

cytoplasm of the cell that produced it via a translational fusion to GFP, PatS is incapable of

restoring a normal pattern of heterocysts to a patS-mutant strain, suggesting that it must diffuse

from cell to cell to function properly [19]. A synthetic peptide corresponding to the predicted C-

terminal 5 amino acids of PatS (PatS-5; RGSGR) prevents DNA-binding activity of HetR in vitro

[14], prevents differentiation of heterocysts when added to the medium [11], and was recently

shown to bind directly to HetR with a 1:1 stoichiometry [20]. In a genetic selection to identify

residues of the 17 amino acid peptide that are necessary for activity of PatS, mutations were

found only in the RGSGR sequence, with mutations resulting in substitution of the underlined

residues reducing activity. A patS-null mutant has reduced spacing between heterocysts, and

adjacent cells often differentiate, the hallmark of the Mch phenotype (multiple contiguous

heterocysts). Transcription of patS is initially up-regulated in contiguous groups of cells, as shown

by a patS-gfp transcriptional reporter fusion, but between 10 and 12 hours after induction, at

about the same time that cells are irreversibly committed to differentiation, fluorescence has been

resolved to single cells in a pattern that predicts the eventual pattern of heterocysts along

filaments [21].

Interactions between PatS and HetR appear to control formation of the initial pattern of

differentiation, but HetN is necessary for stabilization of the initial pattern and prevents

differentiation of cells adjacent to existing heterocysts. It seems likely that it is also involved in

maintenance of the pattern as filaments lengthen. Several cell-generation times after the

formation of an initial Mch pattern by a patS-null mutant, the pattern normalizes to a wild-type-like

pattern [21], suggesting the presence of other patterning factors. In contrast, a patS mutant

237

conditionally lacking expression of hetN forms an Mch pattern initially like that of the patS mutant,

but the pattern fails to normalize over time, and eventually most cells of the filament differentiate

to form heterocysts [22]. A hetN single mutant has a delayed Mch phenotype; the pattern is

initially wild type at 24 h after induction but is Mch after 48 h, approximately twice the time from

induction to the formation of an initial pattern of mature heterocysts [10]. The inhibitory activity of

HetN, as for that of PatS, includes blockage of the positive auto-regulation of hetR [10, 23].

When grown with combined nitrogen, all cells of filaments have a low level of HetN in thylakoid

and cytoplasmic membranes, but upon nitrogen step-down, HetN is degraded [23]. After

formation of proheterocysts, HetN protein and expression of hetN is found exclusively in

proheterocysts and heterocysts. The 287 amino-acid HetN protein is predicted to be a reductase

in the short chain alcohol dehydrogenase (ADH) family [24]. Within the sequence of HetN is the

PatS-5 sequence, RGSGR, at positions 132-136, raising the possibility that this same sequence

is responsible for the inhibitory activity of HetN. Experiments showing that the PatS- and HetN-

dependent pathways converge at or just before HetR and the identification of a point mutation in

hetR that confers resistance to suppression by PatS and HetN suggests that the RGSGR

sequence of HetN may be critical for its activity as it is in PatS [22, 25]. Here we show that the

RGSGR motif of HetN and not the predicted ADH reductase activity is necessary for patterning of

heterocysts.


Culture conditions. Anabaena sp. PCC 7120 and its derivatives were grown in BG-11 medium

as previously described [12]. Media were supplemented with neomycin at 45 µg ml-1

or

spectinomycin and streptomycin at 2.5 µg ml-1

each as appropriate. Escherichia coli strains were

grown in Luria-Bertani (LB) broth for liquid culture and LB solidified with 1.5% agar for plate

culture. For selective growth, media were supplemented with 100 µg ml-1

spectinomycin, 50 µg

ml-1

kanamycin, 100 µg ml-1

ampicillin, or 10 µg ml-1

chloramphenicol.

To induce heterocyst formation, early exponential-phase cultures grown in BG-11 were washed

three times with BG-11 medium without combined nitrogen (BG-110), resuspended in BG-110

without antibiotics, and incubated under growth conditions. To induce the copper-inducible petE

promoter, media were supplemented with 2 μM CuSO4. Heterocyst percentages were

determined by counting 300 cells or more and scoring heterocysts as previously described (Yoon

and Golden, 2001). These counts were done on three separate cultures, and the percentages

reported are an average of the three counts.

238

Plasmid construction. Tables 1 and S2 describe the plasmids and oligonucleotides,

respectively, used in this study. Details of how plasmids were constructed can be found in

Supporting Information.

Construction of strains. Table 1 lists the strains used in this study. The hetN gene was cleanly

deleted from the chromosome of PCC 7120 using allelic replacement as previously described [10]

using plasmid pCCO103 to create strain UHM150. Allelic replacement was used with the

following plasmids and strain UHM150 to create the following strains with altered alleles of hetN

at the original hetN locus: pDR387 for UHM202; pDR388 for UHM200; pDR389 for UHM203;

pDR390 for UHM201; pDR391 for UHM204; pDR392 for UHM205; pDR393 for UHM206;

pDR394 for UHM207; pCO100 for UHM192; pCO102 for UHM209; pCO103 for UHM210;

pCO104 for UHM211; pCO111 for UHM212. As a control to re-create PCC 7120, pDR382 was

introduced into the ΔhetN strain using the same technique to create strain UHM208. Strains were

confirmed as double recombinants by PCR using primers up-hetN-F and down-hetN-R.

Microscopy. Cells were routinely viewed and imaged as described previously [22]. Confocal

microscopy was performed using an Olympus Fluoview 1000 laser scanning confocal mounted

on an IX81 motorized inverted microscope. Fluorescence from Turbo-YFP was detected with an

excitation of 525 nm and an emission of 538 nm. All images were processed in Adobe

Photoshop CS2. To estimate levels of fluorescence in parts of images, a two-dimensional plot

profile of pixel intensities along a line drawn on the image was constructed in ImageJ.

RNA isolation and analysis. Wild type Anabaena sp. PCC 7120 cells grown for RNA isolation

were harvested from nitrogen replete conditions as well as 21 h after induction of heterocyst

differentiation, flash-frozen in liquid nitrogen, and stored at -80ºC until processing. Samples were

lysed and the RNA was precipitated and extracted as described previously [26] using a mini

beadbeater (Biospec) and the RNeasy Kit (Qiagen) for RNA isolation. RLM-RACE (RNA Ligase

Mediated Rapid Amplification of cDNA Ends) was conducted using the FirstChoice RLM-RACE

kit (Ambion) according to the manufacturer’s instructions. The nested primer pairs petEint and

petEext and PhetN-R-Int and PhetN-R-Ext were used for RACE to determine the transcriptional

start points of petE and hetN, respectively. The internal PCR primer was used to sequence RACE

products to identify the transcriptional start points.

239

Table 1. Strains and plasmids used in this study

Strain or plasmid


Anabaena sp. strains


UHM150 ΔhetN This study

UHM163 hetR(R250K) [27]

UHM192 hetN(Δ2-46) This study

UHM201 hetN(G133A) This study

UHM202 hetN(G135A) This study

UHM203 hetN(R132K) This study

UHM204 hetN(S134A) This study

UHM205 hetN(R136K) This study

UHM206 hetN(S142A) This study

UHM207 hetN(Y155F) This study

UHM208 hetN(K159R) This study

UHM210 UHM150 with hetN reintroduced This study





Plasmids


r Neo

r

[28]

pAM505 Same as pAM504 with multiple cloning site inverted [28]


r [29]

pRL278 Suicide vector; Kmr Nm

r [29]

pSMC115 pAM504 with PpetE-hetN [10]

pSMC187 pAM505 with the EcoR1 site removed This study

pDR320 pAM504 with PpetE-hetN [30]

pDR364 pAM504 with PpetE-hetN(R132K) This study

pDR365 pAM504 with PpetE-hetN(G133A) This study

pDR366 pAM504 with PpetE-hetN(S134A) This study

pDR367 pAM504 with PpetE-hetN(S142A) This study

pDR368 pAM504 with PpetE-hetN(Y155F) This study

pDR369 pAM504 with PpetE-hetN(K159R) This study

pDR382 pRL277 used to make UHM208 This study

pDR386 pAM504 with PpetE-hetR-yfp This study




240

Table 1. (Continued) Strains and plasmids used in this study

Strain or plasmid


Plasmids






pRR154 pAM504 with PpetE-hetN(G135A) This study

pRR155 pAM504 with PpetE-hetN(R136K) This study

pRR159 pAM504 with PhetN-hetN-yfp This study



pRR163 pAM504 with PpetE-hetN-yfp This study


pCO100 pRL277 used to make UHM192 This study




pCO105 pAM505 with PpetE-hetN(Δ2-46) This study




pCO110 pAM505 with PpetE-hetN This study







pCCO103 pRL278 used to make UHM150 This study

pPJAV155 pAM504 with PhetN-hetN(1-192)-yfp This study

241

RESULTS

Alleles of hetN encoding R132K and R136K substitutions fail to suppress heterocyst

differentiation

The wild-type allele of hetN completely suppresses the formation of heterocysts when expressed

from the copper-inducible promoter of petE [10]. To test if alleles of hetN encoding substitutions

in the putative catalytic triad of this predicted ketoacyl reductase also suppress differentiation of

heterocysts, alleles of hetN encoding S142A, Y155F, and K159R conservative substitutions were

placed behind the petE promoter on a shuttle vector and tested for their ability to suppress

differentiation of heterocysts. As expected, PCC 7120 with an empty control plasmid formed a

normal pattern and number of heterocysts, and the same plasmid with the wild-type allele of hetN

behind the petE promoter lacked heterocysts (Fig. 1A, B & C). As with the wild-type allele of hetN,

PCC 7120 with any of the three mutant alleles of hetN failed to form heterocysts, indicating that

each allele is capable of suppressing differentiation of heterocysts when expressed ectopically

from the petE promoter (Fig. 1A & D and data not shown), suggesting that the predicted

reductase activity of HetN is not required for suppression of differentiation.

Alleles of hetN encoding substitutions in the RGSGR motif of HetN were also tested in the

manner described above for their ability to suppress heterocyst differentiation. Alleles of hetN

encoding R132K, G133A, S134A, G135A, or R136K substitutions in HetN were introduced

individually behind the petE promoter on a shuttle vector to strain PCC 7120. Each of the five

substitutions reduced the ability of HetN to suppress heterocyst differentiation. Whereas strain

PCC 7120 with the wild-type allele formed no heterocysts, PCC 7120 with either the G133A or

S134A alleles formed morphologically distinct heterocysts that represented less than 1% of the

total number of cells in filaments (Fig. 1A & E). PCC 7120 with the G135A allele formed

approximately 2% heterocysts, and PCC 7120 with either the R132K or R136K alleles formed a

pattern and number of heterocysts similar to that of the same strain with the empty plasmid,

pAM504 (Fig. 1A & F). Substitution of each of the individual amino acids of the RGSGR motif

reduced the ability of HetN to suppress differentiation of heterocysts.

242

Figure 1. Alleles of hetN encoding R132K and R136K substitutions fail to

suppress heterocyst differentiation. A. Heterocyst percentages for PCC 7120

carrying control plasmid pAM504 (control), pAM504 carrying wild-type hetN

transcriptionally fused to PpetE, and pAM504 carrying the indicated alleles of hetN

transcriptionally fused to PpetE. Error bars represent standard error of the mean.

B-F. Brightfield images of PCC 7120 with pAM504 (B); pDR320, which contains

wild-type hetN (C); pDR367, which contains hetN(S142A) (D); pDR365, which

contains hetN(G133A) (E); and pDR364, which contains hetN(R132K) (F).

Micrographs were taken 24 hours after switching from BG-11 to BG-110 with 2 μM

CuSO4. Carets indicate heterocysts.

243

Each amino acid of the RGSGR motif of HetN is necessary for proper patterning of

heterocysts

Suppression of heterocyst differentiation with mutant alleles of hetN suggests that the RGSGR

motif and not the putative catalytic triad of HetN is involved in regulation of heterocyst

differentiation. However, expression of hetN from the petE promoter on a multi-copy shuttle

vector presumably results in much higher expression of hetN than is normally found in the

organism, potentially compensating for reduced activity of some of the substituted proteins. In

addition, the patterned expression of hetN from its native promoter is lost with the petE promoter.

To examine heterocyst patterning and number in strains with the mutant alleles of hetN, strains

were created that had a wild-type genotype except for a change in the chromosome of one or two

nucleotides that corresponded to one of the amino-acid substitutions mentioned previously.

These strains were created by reintroducing the alleles of hetN to a hetN-deletion strain. In strain

PCC 7120, which is the wild-type organism, and a strain with the wild-type allele of hetN

reintroduced to the deletion mutant, 7.1% and 6.9 % of cells were heterocysts, respectively, 48 h

after removal of combined nitrogen (Fig. 2A, B). These heterocysts occurred as single

heterocysts flanked on each side by groups of vegetative cells. In the hetN-deletion strain,

UHM150, about 12.2% of cells were heterocysts (Fig. 2A, C), and heterocysts were often found in

groups of 2, 3 or more contiguous heterocysts, the hallmark of the Mch phenotype. In strains with

alleles of hetN encoding one of the S142A, Y155F, and K159R conservative substitutions in the

putative catalytic triad, 7.1%, 7.3%, and 6.3% of cells were heterocysts, respectively, arranged in

a pattern similar to that of the wild type (Fig. 2A, D). Conversely, strains with alleles of hetN

encoding R132K, G133A, S134A, G135A, and R136K conservative substitutions formed an

average of 12.9%, 9.7%, 10.1%, 9.7%, and 11.9% heterocysts, respectively (Fig. 2A, E, F). For

each of these strains, heterocysts were often found in groups of 2, 3 or more, indicative of the

Mch phenotype and similar to the hetN-deletion strain. These results indicate that each of the

amino acids of the RGSGR motif and not those of the putative catalytic triad of HetN are

necessary for differentiation of the number and pattern of heterocysts found in the wild-type

organism.

244

Figure 2. Each amino acid of the RGSGR motif of HetN is necessary for proper

patterning of heterocysts. A. Heterocyst percentages for strains PCC 7120, wild-

type hetN reintroduced into a ΔhetN strain, a ΔhetN strain, and strains with the

indicated chromosomal hetN alleles. B-F. Brightfield images of PCC 7120 (B),

UHM150, which is ΔhetN (C); UHM207, which is hetN(K159R) (D); UHM203,

which is hetN(S134A) (E); and UHM201, which is hetN(G135A) (F). Micrographs

were taken 48 hours after removal of combined nitrogen. Carets indicate

heterocysts.

245

Regions of HetN necessary for heterocyst patterning and number

HetN can be divided into 4 domains

based on hydrophobicity. An N-

terminal hydrophobic domain, which

may serve as a leader peptide, was

found to consist of amino acids 1 – 46

(N-terminal hydrophobic domain); an

N-terminal hydrophilic domain

containing the RGSGR motif and

putative catalytic triad was found at

amino acids 47 – 176 (N-terminal

hydrophilic domain); a second

hydrophobic domain, which has a

length consistent with that of a central

transmembrane domain, was found at

amino acids 177 – 195 (central

hydrophobic domain); and a C-

terminal hydrophilic domain was found

at amino acids 196 – 287 (C-terminal

domain; Fig 3A). To identify the

domains of HetN that are necessary

for its suppression of heterocyst

differentiation, alleles of hetN coding

for protein that lacks one of the 4

domains were cloned behind the petE

promoter and introduced on a

replicating plasmid to PCC 7120.

Strains containing alleles encoding

deletions of the N-terminal

hydrophobic, central hydrophobic, or

C-terminal domain behaved the same

way as those containing the wild-type

allele and did not produce heterocysts

(Fig. 3B). Conversely, PCC 7120

containing an allele of hetN encoding

a deletion of the N-terminal hydrophilic

Figure 3. Regions of HetN necessary for heterocyst

patterning and number. A. Schematic of the domains

of HetN. B. Heterocyst percentages for PCC 7120

carrying control plasmid pKH256, pKH256 with wild-

type hetN, and pKH256 carrying the indicated alleles

of hetN. C. Heterocyst percentages for strains PCC

7120; strain UHM208, which has wild-type hetN

reintroduced into the ΔhetN strain UHM150; the ΔhetN

strain UHM150; and strains with the indicated hetN

alleles. Error bars represent standard error of the

mean.

246

domain formed a number and pattern of heterocysts similar to the negative control, the same

strain carrying the empty shuttle vector pAM504 (Fig. 3B). An allele of hetN encoding protein with

a partial deletion of the N-terminal hydrodophilic domain was also constructed and tested for its

ability to suppress differentiation. An allele encoding protein with deletion of residues 47 - 128, 3

amino acids before the RGSGR motif, fully suppressed differentiation of heterocysts (Fig. 3B).

Consistent with results of the amino acid substitutions, the RGSGR motif is required for

suppression of heterocyst differentiation.

To identify regions of HetN that are necessary for heterocyst patterning, strains with one of the

alleles of hetN tested above in place of the wild-type chromosomal hetN gene were created and

heterocyst patterning was examined. All of the strains formed a pattern of heterocysts similar to

that of the wild type with the exception of UHM209 and UHM212 (Fig. 3C). UHM209 has an allele

of hetN encoding protein that lacks residues 47 – 176, which includes the RGSGR motif.

Heterocyst formation in this strain was similar to that of the hetN-deletion strain. The hetN allele

of UHM212 lacks residues 47 – 128. Unexpectedly, 4.6% of cells in this strain were individual

heterocysts separated by regions of vegetative cells. This strain formed fewer heterocysts than

the wild type, indicating that this allele has an increased ability to suppress differentiation of cells

in filaments.

The 5’-ends of the hetN and petE transcripts

Immunoblots of proteins from fractionated whole filaments have shown that HetN is present in

both plasma and thylakoid cell membrane preparations [23]. Here, translational fusions of full

length HetN and several of its derivatives to YFP were used in an attempt to visualize their

location in cells. Transcription of fusion genes was driven by either the petE promoter or the

native hetN promoter. Because the start sites within the hetN and petE promoters have not been

characterized, we first determined the transcriptional start points (tsps) of hetN and petE. A single

tsp for petE was found using RNA ligase mediated rapid amplification of cDNA ends (RLM-RACE)

at -32 relative to the translational start site (Fig. S1). This putative transcription start site is well

within the regions used for the petE-promoter fusions used in this work, which begin 338 or 339

nucleotides upstream of the predicted translational start site, depending on the individual fusion.

For hetN, a single tsp at -401 relative to the annotated translational start site was found using

RLM-RACE (Fig. S1). Centered 10.5 nucleotides upstream of this putative transcription start site

is the sequence TATAAA, which is similar to the -10 sigma 70 binding site. No -35 binding site

was apparent. Correspondingly, fusions with the hetN promoter started at position –554 relative

to the annotated translational start site, upstream of a probable intrinsic transcription terminator of

hetM, which is located upstream of hetN. It is interesting to note that these genes are expressed

247

in both nitrogen replete and nitrogen limiting conditions, which, for hetN, corroborates previous

immunoblotting results [23]. Because RLM-RACE distinguishes between 5’-ends of transcripts

generated by initiation of transcription and RNA processing, it is likely that the 5’-ends of

transcripts detected here represent transcription start sites.

Localization of HetN-YFP to the cell periphery

Fluorescence from full-length HetN fused at its C-terminus to YFP and driven by the petE

promoter was observed primarily at the periphery of cells in filaments, consistent with localization

to the plasma membrane (Fig. 4A). In addition, a more diffuse fluorescence signal was observed

from the interior region of cells, consistent with previous detection of HetN in thylakoid

membranes. The fusion protein prevented the formation of heterocysts, so it was not possible to

view the location of HetN in heterocysts in a wild-type genetic background. To facilitate

visualization of HetN-YFP in heterocysts, plasmid-borne hetN-yfp fusions were put into a

derivative of PCC 7120 that continues to form heterocysts even when extra copies of hetN are

introduced [20]. This strain, UHM163, has the wild-type copy of hetR replaced by an allele that

encodes a R250K substitution. With transcription of hetN-yfp from either the native promoter or

the petE promoter, fluorescence from YFP was seen primarily at the periphery of heterocysts (Fig.

4B, C). With expression from the petE promoter, fluorescence was seen in both cell types,

whereas with the native hetN promoter fluorescence was seen only in heterocysts, consistent

with heterocyst-specific expression of hetN [10].

C-terminal translational fusions to YFP were also made with the following variants of HetN:

HetN(1-47), HetN(1-176), HetN(1-192), HetN(∆2-46), HetN(∆47-128), HetN(∆47-176), and

HetN(∆177-195), where the numbers in parentheses represent either the portion of HetN in the

fusion or, if preceded by a ∆ symbol, a protein lacking the amino acids indicated. Heterocyst

differentiation in a wild-type background was suppressed with all fusions in which the RGSGR

motif was present. However, with one exception, no fluorescence from YFP was observed. In

filaments with the construct encoding HetN(1-192)-YFP, which lacks the C-terminal domain of

HetN, YFP-dependent fluorescence was observed primarily at cell junctions (Fig. 4D). In

particular, rings of fluorescence circumscribed cell septa. Fusions were made with expression

from the native promoter of hetN or the petE promoter, but fluorescence was observed only when

the petE promoter was used. Rings of fluorescence were detected in vegetative cells but absent

from heterocysts. In addition, HetN-YFP-dependent fluorescence with both full-length HetN and

the C-terminal deletion was reduced from the cytoplasm of heterocysts relative to that from

vegetative cells (Fig. 4C, D). In a wild-type genetic background, the fusion protein inhibited

differentiation, so fluorescence could only be observed in vegetative cells (Data not shown).

248

When the background strain was UHM163, which forms heterocysts even when hetN is over-

expressed ectopically, rings of fluorescence similar to those in the wild type were observed in

vegetative cells but lacking in heterocysts (Fig. 4D).

249

Figure 4. Each amino acid of the RGSGR motif of HetN is necessary for proper

patterning of heterocysts. A. Heterocyst percentages for strains PCC 7120, wild-type

hetN reintroduced into a ΔhetN strain, a ΔhetN strain, and strains with the indicated

chromosomal hetN alleles. B-F. Brightfield images of PCC 7120 (B), UHM150, which

is ΔhetN (C); UHM207, which is hetN(K159R) (D); UHM203, which is hetN(S134A) (E);

and UHM201, which is hetN(G135A) (F). Micrographs were taken 48 hours after

removal of combined nitrogen. Carets indicate heterocysts.

250

DISCUSSION

HetN and PatS are required for different stages of heterocyst patterning. PatS is necessary for

formation of the initial pattern of cells in a filament composed exclusively of vegetative cells that

will differentiate after a transition to conditions that require diazotrophy for growth. A hetN-mutant

strain is capable of forming this de novo pattern whereas a patS mutant is not [10, 11]. HetN is

necessary for stabilization of this initial pattern. The wild-type pattern of heterocysts initially

formed in a hetN-mutant turns Mch shortly thereafter [10]. Genetic and mathematical modeling

evidence relating to placement of additional heterocysts between existing ones as vegetative

cells divide to maintain the pattern and ratio of heterocysts to vegetative cells suggests that a

HetN-dependent signal produced in mature heterocysts is responsible for specification of a region

of cells at the midpoint between two heterocysts, and that PatS resolves this region to a single

cell that will differentiate [11, 31]. Despite the different roles of PatS and HetN in patterning, the

RGSGR motif is essential to the function of both. In the over-expression studies presented here,

conservative substitutions in any of the 5 amino acids reduced the ability of HetN to suppress

differentiation. Alteration of either of the two arginines appeared to eliminate suppression by

HetN, even though the gene for HetN was expressed at a level higher than that experienced in

the wild-type organism. These results are comparable to those found with substitutions in PatS.

In this case, a genetic selection identified four of the five amino acids of the RGSGR motif as

necessary for suppression of differentiation [11]. Whether or not substitution of the fifth has the

same effect is unknown. To take the analysis of HetN further, a single nucleotide or two

corresponding to a codon in the RGSGR motif was changed in the chromosome of the wild type.

Each of the strains with an allele of hetN encoding a conservative substitution in one of the amino

acids of the RGSGR motif resembled the hetN-deletion strain more than the wild type; the

number of cells that differentiated was higher than in wild type and instead of a periodic pattern of

single heterocysts, the Mch phenotype was observed. As for the over-expression studies,

substitution of the two arginines had the most dramatic effect, producing strains with phenotypes

indistinguishable from that of the hetN-deletion strain. In contrast, conservative substitutions in

the putative catalytic triad of ADH activity had no effect on the function of HetN in suppression of

differentiation or patterning of heterocysts. The RGSGR motif of HetN and not predicted ADH

reductase activity was necessary for patterning of heterocysts.

Three previous studies have addressed the quality of HetN necessary for its ability to suppress

heterocyst differentiation, with varied interpretations. In the first, mutation of residues in the

RGSGR sequence of HetN was reported to have no effect on the ability of HetN to suppress

differentiation when the corresponding alleles of hetN were over-expressed [23]. In addition to

R132K and R136L substitutions, the report claims that G134S and S135D substitutions were

251

made and had no effect. However, this is confusing because the HetN sequence is S134 and

G135, not G134 and S135 as indicated, so what substitutions were actually made is unclear. In

the second study, substitution of one of the three amino acids in the predicted catalytic triad of

ADH activity in HetN prevented suppression of heterocyst differentiation by over-expression of

the corresponding mutant allele [32]. However, substitution of the other two had no effect. In the

most recent study, over-expression of hetN was shown to cause post-translational decay of HetR,

which presumably contributes to suppression of heterocyst formation. When the RGSGR motif

was replaced by RGDAR, over-expression of the corresponding allele of hetN did not lower levels

of HetR in filaments nor did it suppress differentiation [9], suggesting that the RGSGR motif is

necessary for suppression. The present study is the most comprehensive with respect to the

regions of HetN that are necessary for inhibition and patterning. It clearly shows that the RGSGR

motif of HetN and not the predicted ADH reductase activity is necessary for patterning of

heterocysts.

The RGSGR motif as the primary functional group of HetN and PatS in patterning is consistent

with known activities of HetN and PatS. In electrophoretic mobility shift assays the synthetic

RGSGR peptide prevents the binding of HetR to a region of hetR-promoter DNA that includes the

autoregulated transcriptional start point at position -271, and over-expression of hetN or patS

prevents transcription from this same tsp [8, 14]. Addition of RGSGR peptide to medium causes

posttranslational decay of HetR protein in whole filaments, as does over-expression of hetN or

patS in all cells of filaments [9]. Over-expression of hetN or patS in individual cells of filaments

also causes decay of HetR levels in adjacent groups of cells, suggesting that the HetN- and PatS-

dependent signals that diffuse from cell to cell contain the RGSGR sequence.

Individual deletions of the predicted signal sequence, membrane-spanning hydrophobic domain,

and C-terminal domain had no effect on suppression of differentiation or patterning. However,

two pieces of evidence suggest that it is not solely the RGSGR motif of HetN that is necessary for

normal suppression and patterning of heterocysts. First, it has been reported that a K159E

substitution prevents the ability of HetN to suppress differentiation when the gene encoding the

protein is over-expressed [32]. This non-conservative substitution of a positively charged amino

acid with a negatively charged one, unlike the conservative K159R substitution used in the work

presented here, apparently alters HetN sufficiently to prevent it from suppressing formation of

heterocysts. The second piece of evidence is the formation of fewer heterocysts by a strain with

the wild-type hetN replaced with an allele encoding a deletion of amino acids 47 – 128, which

would lack the N-terminal hydrophilic domain up to three amino acids before the RGSGR

sequence. The resulting protein appears to lead to suppression of differentiation of cells in a

larger region of a filament than does the wild-type protein. Possible explanations of this

252

enhanced range of lateral inhibition include a more active form of the protein, a longer half-life, an

increased rate of diffusion or.increased production of the mature form of HetN that presumably is

transferred between cells.

Despite sharing a functional protein motif, the forms of HetN and PatS that diffuse from cell to cell

and presumably interact with HetR may be different and remain unknown. Increased expression

of hetN in cells that will become heterocysts around the time of commitment to differentiation is

very different from that of patS, which occurs in groups of cells very soon after induction of

filaments to differentiate by growth-limiting levels of fixed nitrogen in the medium [21, 33]. The

difference in expression patterns reflects the respective roles of HetN and PatS in patterning.

The involvement of two separate proteins with a similar functional motif, rather than one protein

with two modes of expression, suggests that intrinsic differences in the mature HetN- and PatS-

dependent signals play roles in patterning. Recent mathematical modeling of heterocyst pattering

with two inhibitors having different rates of diffusion is consistent with this notion [31].

HetN has been reported to be a membrane protein that resides in both cytoplasmic and thylakoid

membranes. In this study, fluorescence from HetN-YFP was observed primarily in the cell

envelope, presumably in the cytoplasmic membrane, with a more diffuse, lower level of

fluorescence from the interior of the cell, suggesting that the concentration of HetN is higher than

that in thylakoid membranes. HetN-YFP-dependent fluorescence intensity in the membrane was

about twice that from thylakoid membranes, but given the much larger surface area of thylakoid

membranes, the majority of HetN in a cell is likely to be in thylakoid membranes, consistent with

earlier western blot analysis [23]. The only fusions to YFP in this study for which fluorescence

was observed were those where YFP either replaced or was at the end of the C-terminal domain,

suggesting that this domain of the protein is located in the cytoplasm. Folding of GFP is efficient

only in the cytoplasm of bacterial cells, which has been used to map the topology of inner-

membrane proteins [34]. The YFP used in the fusions is a derivative of GFP and would be

expected to behave in a similar fashion. If the C-terminal domain is in the cytoplasm, the N-

terminal hydrophilic domain, which contains the RGSGR motif, may be located in the periplasm if

the central hydrophobic domain at amino acids 177 – 195 spans the cytoplasmic membrane.

However, it is difficult to interpret the lack of fluorescence with fusions to other parts of HetN,

which could be explained by the creation of an unstable protein with a short half-life.

Lateral inhibition implies the intercellular transfer of a signal that suppresses differentiation.

Based on the deletion studies, the essential part of HetN that comprises the suppression signal

appears to be little more than the RGSGR motif. This raises the possibility that the signal that

diffuses from cell to cell may be a peptide processed from the full protein. Location of HetN in the

253

cytoplasmic membrane suggests three potential routes of transfer between cells. First, the

periplasm of cells in filaments is contiguous, and there is evidence both for and against the

diffusion of proteins between cells via the periplasmic space [35, 36]. Processing of the N-

terminal domain after insertion in the membrane could release a soluble fragment of HetN that

contains the RGSGR motif and diffuses through the contiguous periplasmic space. Uptake into

the cytoplasm would be necessary to allow interaction with HetR. Inhibition of heterocyst

differentiation by addition of synthetic RGSGR peptide to the medium [11] suggests that transport

into the cytoplasm is possible for such a molecule or that it can act from the periplasm. The

second potential route of transfer is direct exchange between the membranes of adjacent cells.

Although the cytoplasmic membranes at cell septa are not shared by adjacent cells and so are

not continuous between cells, they are in close proximity [37], and could be bridged by an as yet

uncharacterized intermembrane transport system. Results from this work strongly suggest that

the HetN-dependent patterning signal does not travel from cell to cell via one of these two

potential routes because localization of HetN to the membrane is not required for proper

patterning of heterocysts. Deletion of the putative transmembrane or signal sequence prevented

HetN-YFP from localizing to the membrane and had no effect on heterocyst patterning when

deleted from the chromosomal copy of hetN. In its simplest form, transfer of the RGSGR-

containing portion of HetN via the cytoplasmic membrane or the periplasmic space would likely

require localization to the membrane. The most likely route of transfer is via inter-cytoplasmic

exchange mediated by SepJ, FraC, and/or FraD, which are located at intercellular septa.

Intercellular transfer of the fluorescent molecular tracer calcein occurs in the wild type, but is

impaired in strains lacking one of the three proteins [38, 39]. Calcein has a molecular weight of

623 Da, slightly more than that of RGSGR peptide. An RGSGR-containing peptide could be

processed from a subpopulation of HetN prior to insertion into the membrane and be transferred

via inter-cytoplasmic pores. Deletion of the putative transmembrane and signal sequence

domains of HetN-YFP prevented not only localization to the membrane, but also detectable

fluorescence. The lack of fluorescence is likely attributable to proteolytic digestion of a protein

that cannot fold properly. Processing of an RGSGR-containing peptide from HetN in the

cytoplasm prior to proteolytic digestion would be consistent with retention of function of these

truncated forms of HetN even in the absence of an accumulation of HetN-YFP that can be

detected by fluorescence microscopy. In addition, preliminary results with a strain lacking SepJ,

which has been shown to be necessary for inter-cytoplasmic exchange of calcien, indicate that

SepJ is necessary for decay of HetR in cells adjacent to those overexpressing hetN in genetic

mosaics [9]. Taken together, these results suggest that an RGSGR-containing peptide derived

from HetN diffuses between cells via direct cytoplasmic exchange to direct pattering of

heterocysts.

254

ACKNOWLEDGEMENTS

We thank John Branigan, Michael Derocher and Cameron Olson for help with preliminary studies

and Tung Hoang for plasmid pUC57-PS12-yfp, the source of the gene for YFP used in this work.

This work was supported by grant number IOS-0919878 from the National Science Foundation.

FOOTNOTES

#For correspondence: E-mail [email protected]; Tel. (+1) 808 956 8015; Fax (+1) 808 956

5339

1Present address: University of Hawaii Cancer Center, Honolulu, HI 96813

2Present address: Department of Biochemistry and Molecular Biology, Michigan State University,

East Lansing, MI 48824

3Present address: Department of Microbiology, University of California, Davis, CA 95616

*These authors contributed equally to the work.

255

REFERENCES CITED

1. Golden, J.W. and H.-S. Yoon, Heterocyst development in Anabaena. Curr. Opin.

Microbiol., 2003. 6: p. 557-563.

2. Zhang, C.-C., et al., Heterocyst differentiation and pattern formation in cyanobacteria: a

chorus of signals. Mol. Microbiol., 2006. 59(2): p. 367-375.


N]N2 and 13





5. Haselkorn, R., How cyanobacteria count to 10. Science, 1998. 282: p. 891-892.

6. Wolk, C.P., Physiological basis of the pattern of vegetative growth of a blue-green alga.


7. Ferrell, J.E.J., Self-perpetuating states in signal transduction: positive feedback, double-

negative feedback and bistability. Curr. Opin. Cell Bio., 2002. 14: p. 140-148.


transcription start sites of hetR from Anabaena sp. strain PCC 7120. J. Bacteriol., 2010.

192: p. 1088-1096.



Acad. Sci., 2009. 106: p. 19884-19888.












Acad. Sci. U.S.A., 2004. 101(14): p. 4848-4853.



2010. 77: p. 562-574.

256








19. Wu, X., et al., patS minigenes inhibit heterocyst development of Anabaena sp. strain

PCC 7120. J. Bacteriol., 2004. 186: p. 6422-6429.


7120 and PatS-5. Biochemistry, 2011. 50: p. 9212-9224.





Microbiol., 2005. 57: p. 111-123.

23. Li, B., X. Huang, and J. Zhao, Expression of hetN during heterocyst differentiation and its

inhibition of hetR up-regulation in the cyanobacterium Anabaena sp. PCC 7120. FEBS

Lett., 2002. 517: p. 87-91.

24. Black, T., A. and C.P. Wolk, Analysis of a Het- mutation in Anabaena sp. strain PCC 7120

implicates a secondary metabolite in the regulation of heterocyst spacing. J. Bacteriol.,

1994. 176: p. 2282-2292.




26. Campbell, E.L., et al., Global gene expression patterns of Nostoc punctiforme in steady-

state dinitrogen-grown heterocyst-containing cultures and at single time points during the

differentiation of akinetes and hormogonia. J. Bacteriol., 2007. 189: p. 5247-5256.





p. 4473-4482.

29. Cai, Y. and C.P. Wolk, Use of a conditionally lethal gene in Anabaena sp. strain PCC

7120 to select for double recombinants and to entrap insertion sequences. J. Bacteriol.,

1990. 172: p. 3138-3145.

257



Sci U S A, 2009. 106(47): p. 19884-8.

31. Zhu, M., S.M. Callahan, and J.A. Allen, Maintenance of heterocyst patterning in a

filamentous cyanobacterium. J. Biol. Dynamics, 2010.

32. Liu, J. and W.-L. Chen, Characterization of HetN, a protein involved in heterocyst

differentiation in the cyanobacterium Anabaena sp. strain PCC 7120. FEMS Microbiol.

Lett., 2009. 297: p. 17-23.

33. Bauer, C.C., et al., Suppression of heterocyst differentiation in Anabaena sp. strain PCC

7120 by a cosmid carrying wild-type genes encoding enzymes for fatty acid synthesis.

FEMS Microbiol. Lett., 1995. 151: p. 23-30.

34. Drew, D., et al., Rapid topology mapping of Escherichia coli inner-membrane proteins by

prediction and PhoA/GFP fusion analysis. Proc. Natl. Acad. Sci., 2002. 99: p. 2690-2695.

35. Mariscal, V., A. Herrero, and E. Flores, Continuous periplasm in a filamentous,

heterocyst-forming cyanobacterium. Mol. Microbiol., 2007. 65: p. 1139-1145.

36. Zhang, L.-C., et al., Existence of periplasmic barriers preventing green fluorescent protein

diffusion from cell to cell in the cyanobacterium Anabaena sp. strain PCC 7120. Mol.

Microbiol., 2008. 70: p. 814-823.

37. Flores, E., et al., Is the periplasm continuous in filamentous multicellular cyanobacteria?

Trends Microbiol, 2005. 14: p. 439-443.

38. Merino-Puerto, V., et al., Fra proteins influencing filament integrity, diazotrophy and

localization of septal protein SepJ in the heterocyst forming cyanobacterium Anabaena

sp. Mol. Microbiol., 2010. 75: p. 1159-1170.

39. Mullineaux, C.W., et al., Mechanism of intercellular molecular exchange in heterocyst-

forming cyanobacteria. EMBO J, 2008. 27: p. 1299-1308.

Documents

ROLE OF HETF AND PATU3 IN THE REGULATION OF …DOCTOR OF PHILOSOPHY IN MICROBIOLOGY AUGUST 2014 By Sasa K. Tom Dissertation Committee: Sean M. Callahan, Chairperson Maqsudul Alam Hongwei