Deg/HtrA proteases of the cyanobacterium

Deg/HtrA proteases

of the cyanobacterium

Synechocystis sp. PCC 6803

from biochemical characterization to their physiological functions

Lâm Xuân Tâm

Department of Chemistry

Umeå University

901 87 Umeå

Sweden 2015

Copyright©Lâm Xuân Tâm

ISBN: 978-91-7601-223-9

Front cover: "Colonies of the cyanobacterium Synechocystis sp. PCC 6803"

Printed by: KBC, Umeå University

Umeå, Sweden 2015

To my family

v

Table of content

General introduction 1

I. Synechocystis sp. PCC 6803 as a model cyanobacterium 1 I.1. Cyanobacteria 1 I.2. Synechocystis sp. PCC 6803 2

II. Proteases and Protein quality control 4 II.1. Proteases 4 II.2. Protein quality control 5

III. The family of Deg proteases 6 III.1. Diverse biological functions 7 III.2. Conserved structural architecture 7 III.3. Flexible regulation of activity 9

Aim of this thesis 11

The three Synechocystis Deg proteases 12

I. Biological characteristic 12 I.1. Sequence analysis and Tertiary Structures 12 I.2. Oligomerization states 13 I.3. Proteolytic activity 15

II. Intracellular features 16 II.1. Localization 16 II.2. Proteases expression and regulation 18

III. Physiological importance 20 III.1. Impact on metabolic pathways 20 III.2. Impact on transportation/secretion 21 III.3. Impact on cell envelope biosynthesis 22 III.4. Impact on protein quality control network 22 III.5. Impact on cellular motility 23

Overview and Outlook 24

Acknowledgment 26

References 27

vi

Abstract

The family of Deg/HtrA proteases is present in a wide range of organisms

from bacteria, archaea to eukaryota. These ATP-independent serine

endopeptidases play key roles in the cellular protein quality control. The

cyanobacterium Synechocystis sp. PCC 6803, a model organism for studies

on photosynthesis, metabolism and renewable energy, contains three Deg

proteases known as HhoA, HhoB and HtrA. The three proteases are

important for survival in stress conditions, such as high light or temperature.

In my work the biochemical characteristics of each protease were revealed in

vitro and in vivo. In vitro studies performed using recombinant

Synechocystis Deg proteases allowed conclusions about their

oligomerization states, proteolytic activities and tertiary structure. The in

vivo studies addressed their sub-cellular localization, expression and

physiological importance by comparing wild-type Synechocystis cells with

the three single mutants lacking one of the Deg proteases.

HhoA seems to be involved in the cytoplasmic protein quality control. This

protease is regulated post-transcriptional and post-translational:

oligomerization, pH and/or cation-binding are some of the important factors

to stimulate its proteolytic activity. Instead HhoB acts on periplasmic

proteins and seems to be important for the transportation/secretion of

proteins. While it has low proteolytic capacity, it may act as a chaperone. The

stress-induced HtrA functions in the cellular tolerance against

photosynthetic stress; additionally it might act as a protease partner of

HhoB, generating a protease/chaperone complex.

The results presented in this thesis lay the foundation for a better

understanding of the dynamic protein quality control in cyanobacteria,

which is undoubtedly important for various cellular metabolic pathways.

vii

List of paper

This thesis in based on the following papers that will be referred in the text

by the corresponding Roman numerals

I. Huesgen, P. F., Miranda, H., Lam, X. T., Perthold, M., Schuhmann, H., Adamska, I. and Funk, C. (2011)

Recombinant Deg/HtrA proteases from Synechocystis sp. PCC 6803 differ in substrate specificity, biochemical characteristics and mechanism.

Biochem. J. 435, 733–42.

II. Roberts, I. N., Lam, X. T., Miranda, H., Kieselbach, T. and Funk, C. (2012)

Degradation of PsbO by the Deg protease HhoA Is thioredoxin dependent.

PLoS One 7, e45713.

III. Lam, X. T., Aigner, H., Timmerman, E., Gevaert, K. and Funk, C. Proteomic approaches to identify substrates of the three Deg/HtrA

proteases of the cyanobacterium Synechocystis sp. PCC 6803. Submitted

IV. Hall, M., Lam, X. T., Funk, C. and Persson, K. Structure of the HhoA protease from Synechocystis sp. PCC 6803. Manuscript

viii

Abbreviations

2D-DIGE Two-dimensional difference gel electrophoresis

ADP/ATP Adenosine diphosphate / Adenosine triphosphate

CBB cycle Calvin-Benson-Bassham cycle

COFRADIC Combined fractional diagonal chromatography

GAP Glyceraldehyde 3-phosphate

NADP+/NADPH Nicotinamide adenine dinucleotide phosphate

PSI Photosystem I

PSII Photosystem II

Syn_Degs Synechocystis sp. PCC 6803 Deg proteases

WT Wild-type

GENERAL INTRODUCTION

1

General introduction

I. Synechocystis sp. PCC 6803 as a model

cyanobacterium

I.1. Cyanobacteria, commonly known as blue-green algae, are a

diverse group of photosynthetic prokaryotes that can be found in a wide

variety of habitats. Their name derives from their appearance due to their

green photosynthetic pigment chlorophyll a and the blue phycobilin pigment

phycocyanin. Cyanobacteria are evolutionary the first organisms with a

photosynthetic apparatus containing two photosystems (PS I and PS II)

connected via the cytochrome b6f complex [1] (Figure 1).

Figure 1 - General mechanism of photosynthesis. The absorption of light energy (light

reaction) and the assimilation of CO2 (dark reaction) take place in two separated compartments.

In the light reaction, photons are absorbed by pigments associated with the photosystems (PS I

and PS II). The photosystems as well as the cytochrome b6f complex are membrane-bound

protein complexes located in a specialised membrane system called the thylakoid. Absorbed

photons are funnelled to the reaction centre and trigger an electron flow within and between the

two photosystems. Water is oxidized within PS II producing O2, while NADP+ is reduced to

NADPH within PS I. Water oxidation and electron transport create a proton gradient across the

thylakoid membrane that powers the production of ATP by the enzyme ATP synthase. In the

dark reaction, also known as CBB cycle, ATP and NADPH provide energy for carbon fixation

producing glyceraldehyde 3-phosphate, which can be used for the biosynthesis of sugars, starch

and other organic compounds.

2

Electron transport via two photosystems generates a redox potential strong

enough to use water as an electron donor. The oxygen produced as a

byproduct accumulated and changed the Earth's atmosphere, providing the

basis for evolution of aerobic respiration and with it of almost all live forms

on Earth today. The chloroplast of eukaryotic photosynthetic organisms

(higher plants and algae) has evolved from free living cyanobacteria by

endosymbiosis [2]. Studies on cyanobacteria, therefore, allow insight into an

ancient form of photosynthesis and at the same time provide information for

agriculture and forestry.

Recently, interest in cyanobacteria for their great potential as cell factories

has increased tremendously. Many scientists believe that this is a promising

approach to solve some of the most urgent problems humanity is facing in

the 21th century. Fossil fuels are not just causing problems like pollution and

global warming, these non-renewable sources of energy are also limited in

amount. Conversion of agricultural crops into bio-fuels is less efficient as this

process directly competes to food production. Cyanobacteria, on the other

hands, have high growth rate and can be grown on non-arable land. Even

though cyanobacteria do not naturally produce bio-fuels in usable amounts,

they are easily accessible for genetic engineering. Genetically modified

cyanobacteria were shown to be promising candidates for the production of

bio-fuels and other high value compounds like 1-butanol, ethanol, ethylene,

isobutylaldehyde, isoprene, lactic acid, etc. [3][4][5][6][7]. Thorough

understanding of cyanobacterial proteomes and metabolism, however, is

crucial for further efficient genetic engineering.

I.2. Synechocystis sp. PCC 6803, hereafter: Synechocystis, is the

first cyanobacterium and also the first photosynthetic organism, of which the

entire genome has been completely sequenced [8]. Its known genome,

together with its natural competence and the ability to take up foreign DNA

from the environment, render Synechocystis an excellent model organism

for scientific research. Accordingly, molecular biological modifications

within the genome of Synechocystis, such as inactivation of a specific gene,

can be easily achieved, because the bacterium performs homologous

recombination [9].

Even though Synechocystis was first isolated from a fresh water lake, it also

survives in marine media [10]. Under laboratory maintenance, several sub-

strains of Synechocystis have been derived from the first isolated one [11].


3

Glucose-tolerance was bestowed on this organism by unidentified mutations

[9]. In addition to photoautotrophic growth, the glucose-tolerance strain can

grow under mixotrophic, photoheterotrophic or heterotrophic conditions

with glucose as a carbon source [12][13]. The Kazusa strain, of which the

entire genome has been sequenced, is glucose-tolerant.

Figure 2 - Model of a Synechocystis cell with distinct membrane layers and

separated compartments. Pairs of thylakoid membranes enclosing the thylakoid lumen

form several sheet layers and occupy the outer space of the cytoplasm. These membranes

converge at several sites adjacent to the plasma membrane. Carboxysomes usually located in the

central cytoplasm are protein-enclosed micro-compartments containing enzymes of the CBB

cycle.

Synechocystis cells are globular ( 1.5 µm in diameter) and pose the typical

structure of gram-negative cyanobacteria (Figure 2). The outer border

consists of a cytoplasmic membrane (plasma membrane), a mesh-like

peptidoglycan layer, an outer membrane and a lattice surface layer (S-layer)

[14]. Synechocystis also has surface pili supporting twitching motility [15]

and, therefore, is able to move towards a light source (phototactic

movement) [16]. Like all prokaryotes, Synechocystis does not contain a

nucleus or other membrane bound organelles. In order to perform

photosynthesis, Synechocystis has a system of thylakoid membranes that

contain the photosystems and the respiratory complexes. Biogenesis of the

thylakoid membrane in cyanobacteria is not yet understood, they might have

derived from the plasma membrane [17]. However, electron microscopic

examination of Synechocystis cells showed no evidence of association

between the two membranes [18][14].

4

II. Proteases and Protein quality control

II.1. Proteases or peptidases are proteolytic enzymes hydrolyzing

covalent peptide bonds between amino acids of proteins. Exo-proteases act

on the terminal peptide bonds (amino-proteases or carboxy-proteases),

while endo-proteases cleave proteins internally. The general mechanism of

proteolysis involves a nucleophilic attack on the peptide carboxyl group.

Depending on the nucleophile, proteases are traditionally classified into six

groups: Serine, Cysteine, Threonine, Aspartic acid, Glutamic acid and

Metallo proteases. Serine, cysteine and threonine proteases use a functional

group within the enzyme itself, while aspartic acid, glutamic acid and metallo

proteases activate a water molecule to serve as a nucleophile (Figure 3).

Modern classification (MEROPS database [19]) divides proteases into

specific clans (super-families), families and subfamilies according to their

evolutionary origin. Proteases of the same super-family share the same

tertiary structure, but can show divergent evolution to different nucleophiles.

Figure 3 - Proteolytic mechanisms catalysed by different protease groups.

Proteases can be highly specific and only cleave the peptide bond between

certain amino acids. This normally results in limited proteolysis, creating

two or several polypeptides. Limited proteolysis is important in cellular

regulation involving activation or deactivation of certain proteins, removal of

signal peptides, etc. Conversely, other proteases perform unlimited


5

proteolysis and totally degrade a protein. Unlimited proteolysis is necessary

to remove malfunctional, misfolded or unwanted proteins and to release

amino acids for new protein synthesis. Proteases performing this

"housekeeping" task play a key role in the cellular protein quality control.

II.2. Protein quality control is a mechanism maintaining

functional proteins for a vast array of cellular activities. Proteins have limited

lifetime, constant synthesis and degradation of proteins are therefore crucial

for cell viability. Newly synthesized polypeptides need to pass

unfolded/partially-folded states before gaining their native three-

dimensional structures and, in many cases, being secreted into specific

intracellular compartments or assembled into specific oligomeric protein

complexes. Under certain (e.g. stress) conditions native proteins can lose

their structures and become misfolded. These non-native protein structures

are kinetically unstable and an efficient network of molecular chaperones

and proteases has evolved to prevent their aggregation [20] (Figure 4).

Figure 4 - Chaperone and protease activities in protein quality control. Transition

between different folding states can occur naturally (gray arrows) or with the support of

chaperones/proteases (dark arrows) (Adapted from [21])

Chaperones are a diverse group of proteins that assist the non-covalent

folding or unfolding of proteins and the assembly or disassembly of protein

complexes without becoming part of the final structure [22]. Chaperones,

therefore, can either assist refolding of damaged proteins or their total

unfolding for efficient degradation. In protein quality control, proteolysis

6

and chaperone activities are often closely connected. Proteases often contain

special subunits or domains for chaperone activity. The proteolytic core

(ClpP) of the Clp protease associates with particular chaperone subunits

(ClpA or ClpX) [23], while FtsH proteases contain AAA domains that unfold

the substrates [24]. Some Deg proteases have been shown to contain dual

protease-chaperone functions [25][26][27].

III. The family of Deg proteases

Deg proteases, formerly DegP (degradation of periplasmic proteins) or HtrA

(high temperature requirement A), are a subfamily of homologous proteases

within the family of chymotrypsin-like serine endopeptidases (S1) according

to current MEROPS classification [19]. In general the structure of Deg

proteases consist of an N-terminal proteolytic domain containing the His–

Asp–Ser catalytic triad followed by one or two C-terminal PDZ domains

mediating a wide range of protein–protein interactions. Unlike other

protease machineries (ClpAP, ClpXP, Lon, FtsH, etc.), Deg proteases contain

no ATP-binding domain and therefore function in an ATP-independent

manner [28].

Figure 5 - Phylogenetic tree of some Deg protease members from A. thaliana

(Ath), E. coli (Eco), human (Hsa) and Synechocystis (Syn). Shown is a simplified

version of a phylogenetic tree constructed around the gene encoding Synechocystis HhoB

(sll1427) by PhylomedDB [29].


7

III.1. Diverse biological functions

Deg protease members were first described in E. coli [30][31] and later

detected in other bacteria, archaea and eukaryota. They are involved in

various cellular pathways related to stress tolerance or protein folding during

stress [32]. Some Deg proteases of bacteria (E. coli), mammals (human) and

higher plants (A. thaliana) have been carefully characterized. The E. coli

genome encodes 3 Deg proteases known as DegP, DegQ and DegS (Figure 5).

Periplasmic DegP of E. coli, functioning as a chaperone at normal

temperature and as a protease at high temperature [25], is important for

protein folding and degradation in the cell envelope. The membrane bound

DegS is a regulating protease that can sense misfolded proteins and release

the envelope-stress response signal σE by degrading its inhibition factor [33].

The physiological function of E. coli DegQ is not yet known, however, its

activity appears to be mediated by pH [34]. In humans there are 4 Deg

protease orthologues (HtrA1, HtrA2, HtrA3 and HtrA4); of those best

characterized are HtrA1, acting as tumour suppressor [35], and HtrA2,

involved in mitochondrial protein quality control and regulation of

programmed cell death [36].

In A. thaliana, there are in total sixteen Deg protease members detected in

various organelles (the nucleus, the chloroplast, the mitochondria and the

peroxisome) [37]. Some A. thaliana Deg proteases can have multiple-

localization and at least seven proteases are present in the plastid. So far,

best characterized are the lumenal Deg proteases (Deg1, Deg5 and Deg8)

involved in maintenance of PS II; Deg1 was shown to support the assembly

of Photosystem II [38], while Deg5 and Deg8 can form hetero-complexes

and have synergistic functions in the primary cleavage of the PSII reaction

center protein D1 [39]. These data raise questions concerning the general

role of Deg proteases in photosynthetic organisms [40].

III.2. Conserved structural architecture

Tertiary structures have been completely or partially determined of Deg

proteases in E. coli (DegP, DegQ, DegS) [41][34][33], human (HtrA1, HtrA2,

Htra3) [42][43][44], A. thaliana (Deg1, Deg2, Deg5, Deg8) [45][26][46], T.

maritima (HtrA) [27], L. fallonii (DegQ) [47] and S. pneumoniae (HtrA)

[48]. The protease domains are highly conserved in these organisms

displaying a PA clan (MEROPS classification) with two β-barrels

8

asymmetrically arranged beside each other bringing together the catalytic

triad (His and Asp on the N-terminal barrel and the nucleophilic Ser on the

C-terminal one) [49][50] (Figure 6). The small PDZ domain is less conserved

and consists of two β-sheets with two α-helices on its edges. The two

domains are connected by a flexible linker allowing en bloc mobility and

different arrangements of the PDZ domain.

Figure 6 - Tertiary structure of E. coli DegS, human HtrA2, A. thaliana Deg1 and

Synechocystis HhoA. The protease domains shown in gray (red for HhoA) with two

conserved β-barrels (C-terminal β-barrel in darker colours) containing the regulation loops (LA,

LD, L1, L2 and L3) are arranged in the same angle. The relative positions of the PDZ domains

shown in yellow (orange for HhoA) containing the interaction clamps (IC1 and IC2) compared

to the proteases domain are different in these structures.


9

Via hydrophobic interactions between the protease domains Deg proteases

form stable homo-trimeric complexes. PDZ domains and/or the protruding

LA loop of the protease domain facilitate further oligomerisation of these

basic trimer units into bigger complexes such as hexamers, 12-mers or 24-

mers. These cage-like oligomeric structures is suggested to trap substrate

within the inner cavity until the protein processing is complete. Formation of

high-order oligomers can be triggered by the presence of substrate or change

of pH [51] [45] and normally involves an activation mechanism of the

protease.

III.3. Flexible regulation of activity

Under normal condition, Deg proteases are present in a "resting" state which

can reversibly transform to the active state upon specific signals such as the

presence of misfolded polypeptide or ions, changes in pH or temperature,

etc. The activation mechanism can vary considerably in different Deg

proteases (Figure 7).

Activation by formation of higher-order oligomers is reported for E. coli

DegP and A. thaliana Deg1. In the "resting" state, E.coli DegP assembles into

hexamers, in which the active sites are blocked and distorted by the

protruding LA loop of the subunit within the opposite trimer [52]. Substrate-

binding to the PDZ1 domain destabilizes the DegP hexamer leading to

formation of 12- or 24-mers. In the multi-oligomers, the LA loop is

withdrawn from the active site and the PDZ1 domain is fixed to be able to

interact with the L3 loop, which in turn interacts with the loops LD, L1 and

L2 of the adjacent subunit [53]. Consequently, the active site is restructured

into proper conformation for proteolysis. Oligomerization of A. thaliana

Deg1 into the active hexameric state, in which the PDZ domain is fixed for

the PDZ→L3→LD/L1/L2 activation, is triggered by low pH [45].

E. coli DegS, which is normally present in the cell as membrane bound

trimer with disordered active site, is not activated by oligomer conversion.

The PDZ domain of DegS is kept in position by additional interactions with

the protease domain. It captures the L3 loop and therefore prevents

interaction with the loops LD/L1/L2 of the adjacent subunit. Binding of

allosteric activators to the PDZ domain, such as misfolded outer-membrane

proteins, triggers the L3 loop to re-orient and to interact with the loops

LD/L1/L2, and thus, restructures the active site [33].

10

Figure 7 - Schematic demonstration of the activation mechanisms of proteolytic

Deg proteases in E. coli (DegP and DegS) and human (HtrA1 and HtrA2). Trimers of

the Deg proteases are shown with the loops L1, L2, L3, LA and LD important for activity

regulation.

Different to E. coli Deg proteases, in human HtrA1 the PDZ domain is not

involved in the activation mechanism, its removal therefore has no effect on

the proteolytic activity of HtrA1 [42]. The disordered active site of the HtrA1

trimer can be oriented and stabilized by substrate binding via an induced-fit

mechanism. Binding of polypeptides to the active site and the L3 loop is

sufficient to activate HtrA1. To regulate activity of HtrA1 therefore likely

requires additional control mechanisms.

Finally, the membrane bound human HtrA2 trimer maintains its properly

oriented active site in the "resting" state. Access to the active site, however, is

restricted by the PDZ domain [44]. Certain conditions (e.g. substrate

binding) can induce relocation of the PDZ domain and allow proteolysis.

HtrA2 depleted of the PDZ domain was shown to have high proteolytic

activity.

AIM OF THIS THESIS

11

Aim of this thesis

Three members of the Deg protease family have been detected in

Synechocystis (Syn_Degs), known as HhoA (sll1679), HhoB (sll1427) and

HtrA (slr1204), which protect Synechocystis against high light or heat stress

[54]. Little is known about their course of action (e.g. their substrates,

regulation mechanisms, etc.).

My studies aimed for a more detailed view on the cyanobacterial Deg

proteases. In vitro and in vivo analyses were performed on recombinant

truncated Syn_Degs (rHhoA, rHhoB and rHtrA) and deletion mutants with

each of the three deg genes inactivated (ΔhhoA, ΔhhoB and ΔhtrA) (Figure

8). The results highlighted the function of each protease, which showed to be

individual for each protease.

Figure 8 - Summary of scientific research on Synechocystis Deg proteases carried

out in this thesis. Soluble recombinant Syn_Degs were constructed as His-tagged, truncated

versions depleted of the predicted signal peptide (for rHhoA and rHhoB) or transmembrane

domain (for rHtrA) (PAPER I). Additionally stable inactive versions were constructed of the

recombinant Deg proteases, in which the catalytic Ser residue was mutated to Ala. Single deg

mutants (ΔhhoA, ΔhhoB and ΔhtrA) were generated by replacing a fragment in the middle of

each individual deg gene (hhoA, hhoB or htrA) with a kanamycin-resistance marker, thus,

preventing the expression of a functional protease (PAPER III).

12

The three Synechocystis Deg proteases

I. Biological characteristic

I.1. Sequence analysis and Tertiary Structures

The three Synechocystis Deg proteases share around 50 % mutual sequence

identity and are evolutionary more closely related to each other than to Deg

proteases found in other organisms [55].

Figure 9 - Multiple sequence alignment of the protease domain and the first PDZ

domain belonging to Deg proteases of Synechocystis (HhoA, HhoB and HtrA), E.

coli (DegP, DegQ and DegS), human (HtrA1 and HtrA2) and A. thaliana (Deg1,

Deg5 and Deg8). Positions of the catalytic triads (His–Asp–Ser) and the regulation loops (LA,

LD, L1, L2 and L3) on the protease domain as well as the carboxylate-binding loop and the

interaction clamps (IC1 and IC2) on the PDZ domain are marked above the sequences.

Alignment were generated using Clustal Omega [56] and coloured by conservation.

THE THREE SYNECHOCYSTIS DEG PROTEASES

13

The protease domains of Syn_Degs are highly conserved to other Deg

proteases, while the PDZ domains are variable. The tertiary structure of

hexameric HhoA was solved by X-ray crystallography (PAPER IV) (Figure 6).

In this "resting" state, the loops L1, L2 and L3 are disordered, thus, a

functional catalytic triad, oxyanion hole and S1 binding pocket are lacking.

Furthermore, no electron density of the LA loop could be observed indicating

that this loop is disordered in the current HhoA structure and not required

for hexamer formation. All three Syn_Degs contain relatively long LA loop

and IC1 compared to other Deg proteases (Figure 9). Dimerization of HhoA

trimers relies solely on the PDZ domains; IC1 of a subunit within trimer 1

interacts with IC2 of the opposite subunit belonging to trimer 2.

In the PDZ domain, the carboxylate-binding loop, which is required for

recognition of substrate C-termini, has a YIGV motif in HhoA and a YLGI

motif in HhoB or HtrA. Interestingly, the Arg residue placed in front of the

carboxylate-binding loop, which is highly conserved in Deg proteases of

other organisms, is exchanged for a His in all three Syn_Degs (Figure 9). In

resolved structure of HhoA, this His286 residue is rotated and pointing away

from the carboxylate-binding pocket. By forming a salt bridge with the

conserved Asp154 of the protease domain, it determines the relative

orientation of the two domains and accordingly could affect the formation of

higher-order oligomers or PDZ–L3 interaction. In E. coli DegS, an

equivalent salt bridge between Arg and Asp residues is important for the

PDZ→L3 activation [33]. It should be noted that the proteolytic activity of

the three Syn_Degs is very low at pH ≤ 5,0 (PAPER I), at conditions favoring

a charged histidine side chain and salt bridge formation with Asp.

A Mg2+ hexa-hydrate ion was observed to be bound by the conserved Asp229

residue of the L1 loop at the central channel of the HhoA trimer unit. Binding

of Ca2+ to that position is known for the lumenal A. thaliana Deg5 [46].

Binding of divalent cations may regulate protease activity allosteric as the

proteolytic activity of the three Syn_Degs highly increased in the presence of

Ca2+ or Mg2+ (PAPER I).

I.2. Oligomerization states

The oligomerization states of recombinant Syn_Degs with different

molecular weights were determined by gel filtration (PAPER I). Among the

three recombinant proteases, rHhoA had the highest ability to form

14

oligomers, it was present hexameric, whereas rHhoB and rHtrA only formed

trimers. The presence of the substrate β-casein induced the formation of

large oligomers, rHhoA formed 12-24-mers, rHtrA 9-12-mers, however,

rHhoB remained as trimer.

It should be noted that native HtrA is expected to anchor into the cellular

membrane, and therefore aggregation into larger homo-oligomers might be

prevented in vivo. Based on immuno stains (PAPER III) its transmembrane

domain, which has been omitted for construction of the recombinant

protease (PAPER I), seems to be intact in vivo. It therefore is unlikely that

HtrA is released from the membrane by N-terminal cleavage and may

function as trimer or hetero-oligomer. Other membrane-bound Deg

proteases, like E. coli DegS or human HtrA2, have been shown to function as

trimer [36][57].

Figure 10 - Structure of the HhoA hexamer. Trimer formation is mediated by the

proteases domains (cyan), while dimerization of the two trimers relies on the PDZ domains

(blue). (Adapted from PAPER IV).

Hexameric rHhoA was analysed by X-ray crystallography (PAPER IV). The

cage-like structure forms three entrance pores allowing only small or

unfolded polypeptides to access the catalytic inner cavity (Figure 10). The

PDZ domains form three pillars within the hexameric structure and are

responsible for trimer interaction. Accordingly, rHhoA, and also rHtrA,

depleted of the PDZ domains only will be present in trimeric state and lose

the ability to form larger oligomers, even in the presence of substrate

(PAPER I).


15

I.3. Proteolytic activity

The proteolytic activity of recombinant Syn_Degs were examined against

different model substrates (PAPER I) or the whole Synechocystis proteome

(PAPER III). In general, rHhoA was shown to be the most active among the

three proteases, while rHhoB was almost proteolytic inactive (Figure 11).

Deletion of the PDZ domain, strongly impaired, but did not abolish the

proteolytic activity of rHhoA or rHtrA.

Figure 11 - In vitro activity of recombinant Synechocystis Deg proteases against

different purified proteins. All three proteases can process naturally unfolded β-casein, but

not natively folded BSA, lysozyme or PsbO from either Synechocystis or the higher plant

spinach. Reduced lysozyme or PsbO, however, could be degraded significantly by rHhoA.

(Adapted from PAPER I and PAPER II)

Using model substrates, the proteolytic activity of the three proteases

differed between the three proteases and was dependent on pH, temperature

(Figure 12) or the presence of Ca2+/Mg2+. rHhoA was highly active over a

broad pH (from pH 5 to pH 8) and temperature range. Its activity increased

tremendously with raised temperature up to 55oC, then dropped rapidly

(probably due to protease denaturation). rHhoB, on the other hand, showed

no significant proteolytic activity. Instead chaperone activity was detected

for rHhoB, but not for rHhoA or rHtrA using the refolding assay of

denatured E. coli α-amylase MalS (PAPER III). rHtrA was proteolytic active

at relative acidic pH (pH 5 - 6,5) and within a narrow range of temperature.

In contrast to rHhoA, its activity was highest at the optimal growth

temperature of Synechocystis (33 oC) [58] and decreased slowly with raised

temperature.

16

Figure 12 - pH/Temperature profile of recombinant Synechocystis Deg proteases

activity without Ca2+/Mg2+. The proteolytic activity of the three proteases was examined

using casein substrate labelled with pH-insensitive fluorescent dyes. While the fluorescent

signal is quenched in intact substrate, it will be released during degradation into smaller

fragments. (Adapted from PAPER I)

Addition of Ca2+ or Mg2+ activated rHhoB, and strongly enhanced the

proteolytic activity of rHhoA and rHtrA. The common effect of divalent ions

on the three proteases suggests a general activation mechanism. Ca2+/Mg2+-

binding may trigger structural changes of either the proteases or the

substrates and favour proteolysis. Binding of Ca2+ or Mg2+ has been reported

for rHhoA (PAPER IV) as well as for A. thaliana Deg5 [46]. In the presence

of Ca2+/Mg2+, the pH profiles of recombinant Syn_Degs was broadened

towards basic pH (PAPER I).

In total Synechocytis protein extracts, recombinant Syn_Degs tend to

degrade the same substrates, but prefer different cleavage sites (PAPER III).

rHhoA and rHtrA also were shown to degrade the model substrate β-casein

at different sites (PAPER I). All three Syn_Degs preferred hydrophobic

residues in positions P4 to P1'. High preference for small amino acids like

Val or Al were observed at P1 and Ala or Ser at P1' positions indicating small

or narrow substrate binding pockets at the active sites of the three proteases.

II. Intracellular features

II.1. Localization

Syn_Degs have been detected in multiple subcellular compartments of

Synechocystis isolated by combined method including gradient


17

centrifugation and two-phase partitioning. HhoA was detected as soluble

protein in the periplasm [59] and also seems to be bound to the plasma

membrane and thylakoid membrane [60] (PAPER II). HhoB was only found

to be associated with the plasma membrane, while HtrA was detected in

outer membrane [61], inner membrane and thylakoid membrane fractions

(PAPER II). These biochemical data, however, do not reveal, whether the

Deg proteases are located on the periplasmic, cytoplasmic or lumenal side of

the membrane.

Specific polyclonal antisera directed against recombinant Syn_Degs allowed

the localization of HhoA in the cyanobacterial cell using gold immuno-

labelling. HhoA was predominantly detected on the cytoplasmic side of the

thylakoid membranes and also concentrated on the septum of dividing cells

(Figure 13). This result is consistent to the proteomic detection of HhoA in

various samples and further shows the cellular ability to distribute HhoA to

different compartments. Apparently, the presence of HhoA in the periplasm

is only important during the division stage of non-stressed cells. Septal

localization of Deg proteases has also been reported for S. pneumoniae [62].

Figure 13 - Immuno-gold labelling of HhoA in the triple mutant depleted of all

three Deg proteases (Δ3), wild-type (WT) and the ΔhhoB-htrA double mutant (Δ2)

[54] (Bar = 0,5 µm). HhoA molecules were visualized by gold particles and appeared as small

dark spots. The immuno response to HhoA was much stronger in Δ2 mutant than in WT

suggesting an up-regulation of HhoA in the absence of other Deg proteases.

18

II.2. Proteases expression and regulation

The protein expression of Syn_Degs in WT and single deg mutants grown at

normal conditions or after exposure to high light or heat stress were

examined by immunoblotting using inactive recombinant Syn_Degs as

quantitative standards (PAPER III). The three Syn_Degs showed distinct

regulation under light/heat stress treatment. Their expression was strongly

affected by the absence of other Deg proteases, especially HhoB and HtrA

seemed to be regulated inversely correlated (Figure 14).

Figure 14 - Protein expression of

Synechocystis Deg proteases in

wild-type and the three single deg

mutants grown at normal conditions

and in response to light or heat

stress. The two forms of HhoA with

higher and lower molecular weight are

presented as upper and lower bars of the

stacked columns. (PAPER III)

Figure 15 - Activity-based profiling of

all serine hydrolases in

Synechocystis. Labelling reaction with

biotinylated fluorophosphonate was

carried out on wild-type and single deg

mutants either on total extract (TE),

loaded on equal Chl a concentration, or

the soluble fraction (Sol), isolated from the

same amount of wild-type TE. The

expected Syn_Degs positions are marked

with arrowheads (red for HhoA, green for

HhoB and blue for HtrA).

The antibody directed against HhoA immuno-stained two protein bands

differing in molecular weight with about 2 kDa, these peptides might be the

result of divergent post-translational modifications in vivo. Expression of


19

HhoA appeared to be stable, a significant increase in the total amount of

HhoA was only observed in ΔhhoB or ΔhtrA cells exposed to light stress.

While there was no obvious change in the total amount of HhoA exposed to

heat stress, the ratio of the two forms changed; the higher molecular weight

band accumulated, while the one with lower molecular weight decreased.

Both these HhoA forms appear to be proteolytic active, as they are labelled

by activity-based profiling (Figure 15); however, only the band with higher

molecular mass was detected to be soluble. In agreement to this finding,

proteomic studies revealed HhoA in the soluble fraction of Synechocystis to

accumulate under heat shock [63].

Furthermore, both forms of HhoA seemed to be activity-labelled 2-3 times

stronger in ΔhhoB and ΔhtrA background than in WT under normal growth

conditions (Figure 15), while there was no obvious difference in the total

amount of that protease (Figure 14). In WT therefore a large portion of HhoA

seems to be proteolytic inactive in a "resting" state. It could be activated

under stress conditions, which would work much faster than up-regulated

transcription of hhoA. It is known that the proteolytic activity of rHhoA

increases strongly at elevated temperature (PAPER I).

The protein amount of HhoB seems to be stable and independent of light

stress, even in the absence of HhoA or HtrA. Under heat stress, however,

HhoB amount decreased significantly, especially in ΔhtrA. Compared to

HhoA or HhoB, HtrA is most stress-induced; its usually low expression

increased considerably in response to high light, heat stress, or to the

absence of HhoB. Gene transcription of htrA also differs to hhoA and hhoB.

After shifting dark adapted cells to light, htrA is transiently expressed; its

mRNA can reach maximal level within 5 min, but then declines to the

amount of normal growth condition, while hhoA transcription is maximal

after 15 min and hhoB transcripts still accumulate slowly after 4 hours [64].

The inverse correlated regulation of HhoB or HtrA in ΔhtrA or ΔhhoB,

respectively, was drastically enhanced during high light or thermal stress,

indicating a connection between their activity. In addition, a similar, but less

obvious negative correlation was observed between HhoA and the

HhoB/HtrA pair (accumulation of HhoA in ΔhhoB/ΔhtrA exposed to light

stress and decrease of HhoB/HtrA in ΔhhoA under heat stress compared to

WT under the same conditions).

20

III. Physiological importance

Differences in the proteomes of single deg mutants compared to WT allowed

conclusions on the biological roles of each protease. The proteomes were

examined based on their differences in expressed proteins (2D-DIGE

method) (Figure 16) and based on naturally occurring N-termini in the

proteome (N-terminal COFRADIC method) (PAPER III). Deletion of the Deg

proteases had impact on various cellular activities; the main metabolic

pathways and the transportation/secretion systems were most strongly

affected.

Figure 16 - Differently regulated proteins in each single deg mutant compared to

Synechocystis wild-type. Proteins, which were detected to accumulate (black letters) in one

mutant were never found to be decreased (white letters) in other mutants and vice versa.

III.1. Impact on metabolic pathways

Deletion of either one of the Syn_Degs lead to changed accumulation of two

important enzymes (Gap2 amount decreased and amount of Pyk2 increased)

controlling the carbon flow from the CBB cycle toward GAP or pyruvate

formation (PAPER III). There are two Gap homologs in Synechocystis; Gap1

only is important in glycolysis and its absence does not affect the


21

photosynthetic growth of Synechocystis, even under light stress; Gap2 on the

other hand is functioning in CBB cycle and gluconeogenesis and is crucial for

photosynthetic growth [65]. Transcription of gap2 also is dependent on the

photosynthetic electron transport and active carbon metabolism [66]. The

decreased amount of Gap2 therefore may indicate impaired photosynthetic

activity in the single deg mutants. The amount of Sll1358, an enzyme

involved in photorespiration, was also diminished in all three mutants.

Interestingly, the amount of Gap2 and Pyk2 varies much stronger from WT

in ΔhhoA than in the other two mutants. Additionally Sll0529, a

transketolase homolog, accumulated only in ΔhhoA, but not the other two

mutants. On the other hand, FbaII, mainly responsible for Fba activity in

Synechocystis [67], and CcmA, essential for carboxysome formation in

Synechocystis [68], were both up-regulated in ΔhhoB or ΔhtrA, but not in

ΔhhoA. Furthermore, deletion of HhoA affected the neo-N-termini of

proteins involved in carbon metabolism (SucD, CP12 and RbcS), while the

neo-N-termini of proteins forming the photosystems or carboxysome were

affected in ΔhhoB (PsbO and PsaC) and ΔhtrA (PsbO, CcmM and PsaD). A

combined in vivo and in vitro COFRADIC approach identified the

cytoplasmic RbcS as a natural substrate for HhoA and the lumenal PsbO as

substrate for both HhoB and HtrA (PAPER III). HhoA may be important for

a functional CBB cycle, while the HhoB/HtrA pair regulates photosynthetic

activity.

III.2. Impact on transportation/secretion

Deletion of single Syn_Degs lead to decreased accumulation of some ABC

transporters (urea precursor UrtA and bicarbonate precursor CmpA) and to

accumulation of neo-N-termini belonging to Sll0141, a HlyD family secretion

protein. Distinct differences in the regulation of the phosphate permease

system was observed only in ΔhhoB, but not the other two mutants (Figure

17). PstS1 obviously was absent in ΔhhoB leading to strong accumulation of

SphX, PstB1' and possibly PstS2 (PAPER III). Furthermore, the COFRADIC

experiment revealed in ΔhhoB regulation of many other transportation

proteins (Sll0180, HlyD, PilJ, Ycf60, Sll7070, Cph1, FutA2, Slr0615, Slr0977

and TonB) based on the accumulation of their natural-N-termini. HhoB

activity therefore might be important for the proper function of several

transportation systems, especially the type I secretion mechanism.

22

Figure 17 - Significant differences of some phosphate import proteins in ΔhhoB

compared to WT detected by 2D-DIGE. PstS1 and SphX are two periplasmic phosphate

precursor homologs, while PstB1' is an ATP-binding subunit of the phosphate permease. (3D

mapping of the 2D-DIGE fluorescent signals generated by SameSpots software).

III.3. Impact on cell envelope biosynthesis

The neo-N-termini of some structural proteins (HlyA and Slr1908) or

enzymes (MurA, MurC and LpxA) involved in cell structure were

significantly differentially regulated upon HhoB deletion (PAPER III).

Changes in HlyA, a Synechocystis S-layer protein [69], and Slr1908

containing porin and S-layer homology domains may be the direct result of

impaired secretion systems in ΔhhoB. MurA and MurC are suggested to

catalyse the cytoplasmic reactions of the peptidoglycan biosynthesis [70],

while LpxA functions in the metabolism of lipopolysaccharides, a major

component of the outer membrane [71].

HtrA deletion lead to the complete disappearance of two neo-N-termini

belonging to the penicillin-binding protein Pbp8. Penicillin-binding proteins

are enzymes catalysing the periplasmic reactions of peptidoglycan

biosynthesis, maturation or recycling [72]. Pbp8, involved in the final step of

Synechocystis cell division [73], was also identified as natural substrate of

HtrA using N-terminal COFRADIC.

III.4. Impact on protein quality control network

Deletion of HhoA lead to accumulation of the neo-N-terminus of the

substrate-binding adaptor ClpS, which can target destabilized N-terminal

residues of substrate polypeptides to the ClpAP/ClpXP proteolytic


23

machinery. The ATP-dependent chaperone ClpC was up-regulated in ΔhhoB

and ΔhtrA. Further on, the natural N-termini belonging to several proteases

(Lons, HhoA, HtrA, Slr1288, TldD and CtpC) strongly accumulated in

ΔhhoB. It is likely that these proteases/chaperones are up-regulated to

compensate for the loss of the specific Deg protease activity in the protein

quality control network.

III.5. Impact on cellular motility

Deletion of the individual Syn_Degs affected several proteins related to

chemotaxis. The type 4 pilin polypeptides PilA1 and PilO decreased in single

deg mutants, in ΔhtrA additionally a neo-N-terminus of the pilus assembly

protein PilQ was diminished (PAPER III). PilA1, PilO and PilQ proteins are

crucial for phototactic motility of Synechocystis [74]. Furthermore, the

amount of Sll0414 was decreased in ΔhhoA and ΔhhoB and Sll0301 had neo-

N-termini accumulating in ΔhhoB. While the function of these two

pentapeptide repeat proteins is still unknown, interruption of either sll0414

or sll0301 will lead to reduced or absent cellular motility [75]. Accordingly,

absence of motility and abnormal pilus assembly in a triple deg mutant

lacking all three Syn_Degs has been observed, but not the double deg

mutants [54].

24

Overview and Outlook

The three Deg proteases of Synechocystis are distinctly different in their

biochemical characteristics. Their suspected physiological functions are

different from each other and seem to be in consistence with previous co-

expression data showing HhoA to be involved in general house-keeping

processes, HhoB co-expressing with periplasmic proteins and HtrA with

electron transport chain proteins [76].

Synechocystis HhoA appears to be a highly active protease functioning over

a wide range of pH and temperature. Unlike HhoB and HtrA, which usually

generate large proteolytic fragments, HhoA can degrade misfolded proteins

completely due to its ability to form higher order oligomers trapping

substrate within its inner cavity. In vivo, in cells growing in normal

condition, a large portion of HhoA is inactivated in a "resting" state. Instead

of being regulated on transcriptional level, several factors such as pH,

temperature influence its activity and with that its molecular weight

(distribution of higher or lower molecular weight form) or even localization.

In non-dividing cells, HhoA seems to be a cytoplasmic protease that mainly

functions on the cytoplasmic site of the thylakoid membrane. However, it

can be found also in the periplasm and in the septum during cell division,

maybe stress conditions induce various locations. Absence of HhoA affects

mainly cytoplasmic proteins, and the cytoplasmic RbcS is proposed as a

natural proteolytic substrate of HhoA. My current results point to the

physiological role of HhoA as a general cytoplasmic "house cleaning"

protease.

Further experiments based on our HhoA crystal structure are needed to

understand the regulation mechanism of its activity. While proteolytic

activation upon binding of substrate polypeptides are well studied in E. coli

Deg proteases, little is known about how other signals, such as pH, ions or

post-translational modifications can regulate proteolysis. Based on the

current HhoA structure, the Ca2+/Mg2+ binding site or the pH-dependent salt

bridge between the protease and PDZ domain might possibly function as

allosteric regulators. Furthermore, the in vivo regulation of the two forms of

HhoA with slightly different molecular weights would be interesting. Such

post-translational modifications have not been observed in other Deg

OVERVIEW AND OUTLOOK

25

proteases. Finally the localization of HhoA at multiple sites within the cell

and even its controlled secretion are interesting topics for further studies.

Synechocystis HhoB has very low proteolytic activity compared to HhoA.

HhoB may be a highly specific protease only degrading certain substrates

under specific conditions, such as E. coli DegS. We cannot exclude the low

activity of recombinant HhoB to be caused by poor construction. Based on

the molecular weight native HhoB seems to contain an intact signal peptide,

which is not present in rHhoB. However, this truncation makes rHhoB more

similar to the highly proteolytic active rHhoA. A more tempting speculation

therefore is that HhoB may function as molecular chaperone. Chaperone

activity was detected for HhoB in vitro. Furthermore, I propose the presence

of HhoB to be crucial, but not sufficient, for the degradation of PsbO in vivo;

additionally the proteolytic active HtrA is required. Absence of HhoB lead to

different expression of several periplasmic proteins, especially the secretion

systems. Depletion of the major phosphate precursor and sensor PstS1 is

unlikely the result of slower protein turnover upon protease deletion. These

changes rather might be triggered by lacking chaperone activity of HhoB,

which is important for protein secretion. My results suggest that while HhoA

or HtrA activity is important under stress conditions, HhoB is required for

normal cellular activity. During stress HhoB activity seems to be tightly

connected to HtrA activity.

Further experiments are needed to verify the hypothetical chaperone activity

of HhoB in vivo. Future studies should explore an involvement of HhoB in

cellular transportation/secretion, in cell envelope biogenesis as well as in

PSII maintenance. The exact localization of HhoB and its co-

regulation/complex formation with HtrA are key questions.

Synechocystis HtrA is a membrane bound protease, with narrow pH- and

temperature-optima. Accordingly, regulation of HtrA activity seems to be

based on rapid expression of the protease. HtrA can degrade Pbp8

independently, but needs HhoB to degrade PsbO. It will be interesting to

investigate if HhoB and HtrA degrade PsbO step-wise or if both proteases

form a hetero-oligomeric complex, similar to A. thaliana Deg5 and Deg8 for

D1 degradation. Further experiment are needed to examine the

oligomerization states and their influence on proteolytic activity of a

HhoB/HtrA mixture.

26

Acknowledgment

My doctoral studies have received financial support from the Swedish

Energy Agency, the Umeå University, the European Union 7th Framework

Program (through the Prime-XS proteomic project) and the Lawsky

foundation.

First of all, I would like to thank my supervisor, Prof. Christiane Funk,

who has given me opportunities, conditions, guidance and a lot of

encouragements for my professional development. Thank you for your

understanding, patience and generosity, Doktormutter!

I would also like to thank all my colleagues; those who have involved in my

work directly (Harald Aigner, Helder Miranda, Irma Roberts, Pitter

Huesgen, Michael Hall, Evy Timmerman, Prof. Kris Gevaert,

Samantha Bryan and Prof. Conrad Mullineaux) and some who have

given me valuable instructions, advices and opinions (Otilia Cheregi,

Tania Tibiletti, Giada Marino, Prof. Wolfgang Schröder, Patrick

Storm, Miguel Hernandez, Raik Wagner, Marcus Wallgren, Assoc

Prof. Anders Hofer, Aaron Edwin, Prof. Uwe Sauer and many others

that I don't have the ability to list here).

Lastly but above all, I want to show my deepest gratitude to my family

especially my wonderful parents (Lâm Văn Thuận and Phạm Thị Xuân),

who have devoted their life to my development in every possible ways and

my husband Harald Aigner, who has walked into my life at the beginning

of my doctoral studies and ever since then, support me and keep me from

being lost both in my professional and personal life. Thank you for always

believing in me more than myself and for maintaining my belief in good

people and good scientists!

Special thanks to my friends

in Umeå especially the

delightful Vietnamese

community and the late

beloved Lars Lundmark.

You all have been a warm

part of my Swedish life!

REFERENCES

27

References

[1] R. E. Blankenship, “Origin and early evolution of photosynthesis,”

Photosynth. Res., vol. 33, no. 2, pp. 91–111, 1992.

[2] J. A. Raven and J. F. Allen, “Genomics and chloroplast evolution: what

did cyanobacteria do for plants?,” Genome Biol., vol. 4, no. 3, p. 209,

2003.

[3] E. I. Lan and J. C. Liao, “ATP drives direct photosynthetic production

of 1-butanol in cyanobacteria,” Proc. Natl. Acad. Sci. U. S. A., vol. 109,

no. 16, pp. 6018–23, 2012.

[4] M. D. Deng and J. R. Coleman, “Ethanol synthesis by genetic

engineering in cyanobacteria,” Appl. Environ. Microbiol., vol. 65, no.

2, pp. 523–8, 1999.

[5] K. Takahama, M. Matsuoka, K. Nagahama, and T. Ogawa,

“Construction and analysis of a recombinant cyanobacterium

expressing a chromosomally inserted gene for an ethylene-forming

enzyme at the psbAI locus,” J. Biosci. Bioeng., vol. 95, no. 3, pp. 302–

5, 2003.

[6] S. Atsumi, W. Higashide, and J. C. Liao, “Direct photosynthetic

recycling of carbon dioxide to isobutyraldehyde,” Nat. Biotechnol., vol.

27, no. 12, pp. 1177–80, 2009.

[7] S. A. Angermayr, M. Paszota, and K. J. Hellingwerf, “Engineering a

cyanobacterial cell factory for production of lactic acid,” Appl.

Environ. Microbiol., vol. 78, no. 19, pp. 7098–106, 2012.

[8] T. Kaneko, S. Sato, H. Kotani, a. Tanaka, E. Asamizu, Y. Nakamura, N.

Miyajima, M. Hirosawa, M. Sugiura, S. Sasamoto, T. Kimura, T.

Hosouchi, a. Matsuno, a. Muraki, N. Nakazaki, K. Naruo, S. Okumura,

S. Shimpo, C. Takeuchi, T. Wada, a. Watanabe, M. Yamada, M.

Yasuda, and S. Tabata, “Sequence Analysis of the Genome of the

Unicellular Cyanobacterium Synechocystis sp. Strain PCC6803. II.

Sequence Determination of the Entire Genome and Assignment of

Potential Protein-coding Regions,” DNA Res., vol. 3, no. 3, pp. 109–

136, 1996.

28

[9] J. G. K. Williams, “Construction of specific mutations in photosystem

II photosynthetic reaction center by genetic engineering methods in

Synechocystis 6803,” Methods Enzymol., vol. 167, pp. 766–778, 1988.

[10] D. L. Richardson, R. H. Reed, and W. D. P. Stewart, “Synechocystis

PCC6803: a euryhaline cyanobacterium,” FEMS Microbiol. Lett., vol.

18, no. 1–2, pp. 99–102, 1983.

[11] M. Ikeuchi and S. Tabata, “Synechocystis sp. PCC 6803 - a useful tool

in the study of the genetics of cyanobacteria,” Photosynth. Res., vol.

70, no. 1, pp. 73–83, 2001.

[12] T. Nakajima, S. Kajihata, K. Yoshikawa, F. Matsuda, C. Furusawa, T.

Hirasawa, and H. Shimizu, “Integrated metabolic flux and omics

analysis of Synechocystis sp. PCC 6803 under mixotrophic and

photoheterotrophic conditions,” Plant Cell Physiol., vol. 55, no. 9, pp.

1605–12, 2014.

[13] S. L. Anderson and L. Mcintosh, “Light-activated heterotrophic

growth of the cyanobacterium Synechocystis sp . strain PCC 6803: a

Blue-Light-Requiring Process,” J. Bacteriol., vol. 173, no. 9, p. 2761,

1991.

[14] A. M. L. van de Meene, M. F. Hohmann-Marriott, W. F. J. Vermaas,

and R. W. Roberson, “The three-dimensional structure of the

cyanobacterium Synechocystis sp. PCC 6803,” Arch. Microbiol., vol.

184, no. 5, pp. 259–70, 2006.

[15] D. Bhaya, N. R. Bianco, and D. Bryant, “Type IV pilus biogenesis and

motility in the cyanobacterium Synechocystis sp . PCC6803,” Mol.

Microbiol., vol. 37, no. 4, pp. 941–951, 2000.

[16] D. Bhaya, a Takahashi, and a R. Grossman, “Light regulation of type

IV pilus-dependent motility by chemosensor-like elements in

Synechocystis PCC6803,” Proc. Natl. Acad. Sci. U. S. A., vol. 98, no.

13, pp. 7540–5, 2001.

[17] C. Mullineaux, “The thylakoid membranes of cyanobacteria: structure,

dynamics and function,” Aust. J. Plant Physiol., vol. 26, no. 7, p. 671,

1999.

REFERENCES

29

[18] M. Liberton, R. Howard Berg, J. Heuser, R. Roth, and H. B. Pakrasi,

“Ultrastructure of the membrane systems in the unicellular

cyanobacterium Synechocystis sp. strain PCC 6803,” Protoplasma,

vol. 227, no. 2–4, pp. 129–38, 2006.

[19] N. D. Rawlings, A. J. Barrett, and A. Bateman, “MEROPS: the

database of proteolytic enzymes, their substrates and inhibitors,”

Nucleic Acids Res., vol. 40, no. Database issue, pp. D343–50, 2012.

[20] S. Wickner, “Posttranslational quality control: folding, refolding, and

degrading proteins,” Science, vol. 286, no. 5446, pp. 1888–1893, 1999.

[21] F. U. Hartl, A. Bracher, and M. Hayer-Hartl, “Molecular chaperones in

protein folding and proteostasis,” Nature, vol. 475, no. 7356, pp. 324–

32, 2011.

[22] R. J. Ellis, Molecular Chaperones and Cell Signalling. Cambridge:

Cambridge University Press, 2005.

[23] J. Porankiewicz, J. Wang, and A. K. Clarke, “New insights into the

ATP-dependent Clp protease: Escherichia coli and beyond,” Mol.


[24] S. Langklotz, U. Baumann, and F. Narberhaus, “Structure and

function of the bacterial AAA protease FtsH,” Biochim. Biophys. Acta,

vol. 1823, no. 1, pp. 40–8, 2012.

[25] C. Spiess, a Beil, and M. Ehrmann, “A temperature-dependent switch

from chaperone to protease in a widely conserved heat shock protein,”

Cell, vol. 97, no. 3, pp. 339–47, 1999.

[26] R. Sun, H. Fan, F. Gao, Y. Lin, L. Zhang, W. Gong, and L. Liu, “Crystal

structure of Arabidopsis Deg2 protein reveals an internal PDZ ligand

locking the hexameric resting state,” J. Biol. Chem., vol. 287, no. 44,

pp. 37564–9, 2012.

[27] D. Y. R. Kim, S. C. Ha, N. K. Lokanath, C. J. Lee, H.-Y. Hwang, and K.

K. Kim, “Crystal structure of the protease domain of a heat-shock

protein HtrA from Thermotoga maritima,” J. Biol. Chem., vol. 278,

no. 8, pp. 6543–51, 2003.

30

[28] D. Y. Kim and K. K. Kim, “Structure and function of HtrA family

proteins, the key players in protein quality control,” J. Biochem. Mol.

Biol., vol. 38, no. 3, pp. 266–74, 2005.

[29] J. Huerta-Cepas, S. Capella-Gutiérrez, L. P. Pryszcz, M. Marcet-

Houben, and T. Gabaldón, “PhylomeDB v4: zooming into the plurality

of evolutionary histories of a genome,” Nucleic Acids Res., vol. 42, no.

Database issue, pp. D897–902, 2014.

[30] B. Lipinska, O. Fayet, L. Baird, and C. Georgopoulosl, “Identification,

characterization, and mapping of the Escherichia coli htrA gene,

whose product is essential for bacterial growth only at elevated

temperatures,” J. Bacteriol., vol. 171, no. 3, pp. 1574–1584, 1989.

[31] K. L. Strauch, K. I. T. Johnson, and J. O. N. Beckwith,

“Characterization of degP, a gene required for proteolysis in the cell

envelope and essential for growth of Escherichia coli at high

temperature,” J. Bacteriol., vol. 171, no. 5, pp. 2689–2696, 1989.

[32] T. Clausen, C. Southan, M. Ehrmann, and M. Park, “The HtrA family

of proteases: implications for protein composition and cell fate,” Mol.

Cell, vol. 10, no. 3, pp. 443–455, 2002.

[33] H. Hasselblatt, R. Kurzbauer, C. Wilken, T. Krojer, J. Sawa, J. Kurt, R.

Kirk, S. Hasenbein, M. Ehrmann, and T. Clausen, “Regulation of the

sigmaE stress response by DegS: how the PDZ domain keeps the

protease inactive in the resting state and allows integration of

different OMP-derived stress signals upon folding stress,” Genes Dev.,

vol. 21, no. 20, pp. 2659–70, 2007.

[34] J. Sawa, H. Malet, T. Krojer, F. Canellas, M. Ehrmann, and T. Clausen,

“Molecular adaptation of the DegQ protease to exert protein quality

control in the bacterial cell envelope,” J. Biol. Chem., vol. 286, no. 35,

pp. 30680–90, 2011.

[35] J. Chien, J. Staub, S.-I. Hu, M. R. Erickson-Johnson, F. J. Couch, D. I.

Smith, R. M. Crowl, S. H. Kaufmann, and V. Shridhar, “A candidate

tumor suppressor HtrA1 is downregulated in ovarian cancer,”

Oncogene, vol. 23, no. 8, pp. 1636–44, 2004.

REFERENCES

31

[36] L. Vande Walle, M. Lamkanfi, and P. Vandenabeele, “The

mitochondrial serine protease HtrA2/Omi: an overview,” Cell Death

Differ., vol. 15, no. 3, pp. 453–60, 2008.

[37] S. K. Tanz, I. Castleden, C. M. Hooper, I. Small, and A. H. Millar,

“Using the SUBcellular database for Arabidopsis proteins to localize

the Deg protease family,” Front. Plant Sci., vol. 5, p. 396, 2014.

[38] X. Sun, M. Ouyang, J. Guo, J. Ma, C. Lu, Z. Adam, and L. Zhang, “The

thylakoid protease Deg1 is involved in photosystem-II assembly in

Arabidopsis thaliana,” Plant J., vol. 62, no. 2, pp. 240–9, 2010.

[39] X. Sun, L. Peng, J. Guo, W. Chi, J. Ma, C. Lu, and L. Zhang,

“Formation of Deg5 and Deg8 complexes and their involvement in the

degradation of photodamaged photosystem II reaction center D1

protein in Arabidopsis,” Plant Cell, vol. 19, no. 4, pp. 1347–61, 2007.

[40] P. F. Huesgen, H. Schuhmann, and I. Adamska, “The family of Deg

proteases in cyanobacteria and chloroplasts of higher plants,” Physiol.

Plant., vol. 123, no. 4, pp. 413–420, 2005.

[41] S. Kim, R. A. Grant, and R. T. Sauer, “Covalent linkage of distinct

substrate degrons controls assembly and disassembly of DegP

proteolytic cages,” Cell, vol. 145, no. 1, pp. 67–78, 2012.

[42] L. Truebestein, A. Tennstaedt, T. Mönig, T. Krojer, F. Canellas, M.

Kaiser, T. Clausen, and M. Ehrmann, “Substrate-induced remodeling

of the active site regulates human HTRA1 activity,” Nat. Struct. Mol.

Biol., vol. 18, no. 3, pp. 386–8, 2011.

[43] S. T. Runyon, Y. Zhang, B. A. Appleton, S. L. Sazinsky, P. Wu, B. Pan,

C. Wiesmann, N. J. Skelton, and S. S. Sidhu, “Structural and

functional analysis of the PDZ domains of human HtrA1 and HtrA3,”

Protein Sci., vol. 16, pp. 2454–2471, 2007.

[44] W. Li, S. M. Srinivasula, J. Chai, P. Li, J.-W. Wu, Z. Zhang, E. S.

Alnemri, and Y. Shi, “Structural insights into the pro-apoptotic

function of mitochondrial serine protease HtrA2/Omi,” Nat. Struct.

Biol., vol. 9, no. 6, pp. 436–41, 2002.

[45] J. Kley, B. Schmidt, B. Boyanov, P. C. Stolt-Bergner, R. Kirk, M.

Ehrmann, R. R. Knopf, L. Naveh, Z. Adam, and T. Clausen, “Structural

32

adaptation of the plant protease Deg1 to repair photosystem II during

light exposure,” Nat. Struct. Mol. Biol., vol. 18, no. 6, pp. 728–31,

2011.

[46] W. Sun, F. Gao, H. Fan, X. Shan, R. Sun, L. Liu, and W. Gong, “The

structures of Arabidopsis Deg5 and Deg8 reveal new insights into

HtrA proteases,” Acta Crystallogr. D. Biol. Crystallogr., vol. 69, no. Pt

5, pp. 830–7, 2013.

[47] R. Wrase, H. Scott, R. Hilgenfeld, and G. Hansen, “The Legionella

HtrA homologue DegQ is a self-compartmentizing protease that forms

large 12-meric assemblies,” Proc. Natl. Acad. Sci. U. S. A., vol. 108, no.

26, pp. 10490–5, 2011.

[48] K. Fan, J. Zhang, X. Zhang, and X. Tu, “Solution structure of HtrA

PDZ domain from Streptococcus pneumoniae and its interaction with

YYF-COOH containing peptides,” J. Struct. Biol., vol. 176, no. 1, pp.

16–23, 2011.

[49] M. J. Page and E. Di Cera, “Serine peptidases: classification, structure

and function,” Cell. Mol. Life Sci., vol. 65, no. 7–8, pp. 1220–36, 2008.

[50] H. Czapinska and J. Otlewski, “Structural and energetic determinants

of the S 1 -site specificity in serine proteases,” Eur. J. Biochem., vol.

260, no. 3, pp. 571–595, 1999.

[51] T. Krojer, J. Sawa, E. Schäfer, H. R. Saibil, M. Ehrmann, and T.

Clausen, “Structural basis for the regulated protease and chaperone

function of DegP,” Nature, vol. 453, no. 7197, pp. 885–90, 2008.

[52] T. Krojer, M. Garrido-Franco, R. Huber, M. Ehrmann, and T. Clausen,

“Crystal structure of DegP (HtrA) reveals a new protease-chaperone

machine,” Nature, vol. 416, no. 6879, pp. 455–9, 2002.

[53] T. Krojer, J. Sawa, R. Huber, and T. Clausen, “HtrA proteases have a

conserved activation mechanism that can be triggered by distinct

molecular cues,” Nat. Struct. Mol. Biol., vol. 17, no. 7, pp. 844–52,

2010.

[54] M. Barker, R. de Vries, J. Nield, J. Komenda, and P. J. Nixon, “The

Deg proteases protect Synechocystis sp. PCC 6803 during heat and

light stresses but are not essential for removal of damaged D1 protein

REFERENCES

33

during the photosystem two repair cycle,” J. Biol. Chem., vol. 281, no.

41, pp. 30347–55, 2006.

[55] T. Kieselbach and C. Funk, “The family of Deg/HtrA proteases: from

Escherichia coli to Arabidopsis,” Physiol. Plant., vol. 119, no. 3, pp.

337–346, 2003.

[56] F. Sievers, A. Wilm, D. Dineen, T. J. Gibson, K. Karplus, W. Li, R.

Lopez, H. McWilliam, M. Remmert, J. Söding, J. D. Thompson, and D.

G. Higgins, “Fast, scalable generation of high-quality protein multiple

sequence alignments using Clustal Omega,” Mol. Syst. Biol., vol. 7, p.

539, 2011.

[57] J. Sohn, R. A. Grant, and R. T. Sauer, “OMP peptides activate the DegS

stress-sensor protease by a relief of inhibition mechanism,” Structure,

vol. 17, no. 10, pp. 1411–1421, 2009.

[58] M. Lopo, A. Montagud, E. Navarro, I. Cunha, A. Zille, P. F. de

Córdoba, P. Moradas-Ferreira, P. Tamagnini, and J. F. Urchueguía,

“Experimental and modeling analysis of Synechocystis sp. PCC 6803

growth,” J. Mol. Microbiol. Biotechnol., vol. 22, no. 2, pp. 71–82,

2012.

[59] S. Fulda, F. Huang, F. Nilsson, M. Hagemann, and B. Norling,

“Proteomics of Synechocystis sp. strain PCC 6803,” Eur. J. Biochem.,

vol. 267, no. 19, pp. 5900–5907, 2000.

[60] F. Huang, S. Fulda, M. Hagemann, and B. Norling, “Proteomic

screening of salt-stress-induced changes in plasma membranes of

Synechocystis sp. strain PCC 6803,” Proteomics, vol. 6, no. 3, pp.

910–20, 2006.

[61] F. Huang, E. Hedman, C. Funk, T. Kieselbach, W. P. Schröder, and B.

Norling, “Isolation of outer membrane of Synechocystis sp. PCC 6803

and its proteomic characterization,” Mol. Cell. Proteomics, vol. 3, no.

6, pp. 586–95, 2004.

[62] H. T. Tsui, S. K. Keen, L.-T. Sham, K. J. Wayne, and M. E. Winkler,

“Dynamic distribution of the SecA and SecY translocase subunits and

septal localization of the HtrA surface chaperone/protease during

Streptococcus pneumoniae D39 cell division,” MBio, vol. 2, no. 5, pp.

e00202–11, 2011.

34

[63] A. R. Slabas, I. Suzuki, N. Murata, W. J. Simon, and J. J. Hall,

“Proteomic analysis of the heat shock response in Synechocystis

PCC6803 and a thermally tolerant knockout strain lacking the

histidine kinase 34 gene,” Proteomics, vol. 6, no. 3, pp. 845–64, 2006.

[64] T. Jansén, H. Kidron, H. Taipaleenmäki, T. Salminen, and P.

Mäenpää, “Transcriptional profiles and structural models of the

Synechocystis sp. PCC 6803 Deg proteases,” Photosynth. Res., vol. 84,

no. 1–3, pp. 57–63, 2005.

[65] O. Koksharova, M. Schubert, S. Shestakov, and R. Cerff, “Genetic and

biochemical evidence for distinct key functions of two highly divergent

GAPDH genes in catabolic and anabolic carbon flow of the

cyanobacterium Synechocystis sp. PCC 6803,” Plant Mol. Biol., vol.

36, no. 1, pp. 183–194, 1998.

[66] R. M. Figge, C. Cassier-Chauvat, F. Chauvat, and R. Cerff, “The carbon

metabolism-controlled Synechocystis gap2 gene harbours a conserved

enhancer element and a Gram-positive-like -16 promoter box retained

in some chloroplast genes,” Mol. Microbiol., vol. 36, no. 1, pp. 44–54,

2000.

[67] K. Nakahara, H. Yamamoto, C. Miyake, and A. Yokota, “Purification

and characterization of class-I and class-II fructose-1,6-bisphosphate

aldolases from the cyanobacterium Synechocystis sp. PCC 6803,”

Plant Cell Physiol., vol. 44, no. 3, pp. 326–33, 2003.

[68] T. Ogawa, E. Marco, and M. I. Orus, “A gene (ccmA) required for

carboxysome formation in the cyanobacterium Synechocystis sp.

strain PCC6803,” J. Bacteriol., vol. 176, no. 8, pp. 2374–8, 1994.

[69] C. Trautner and W. F. J. Vermaas, “The sll1951 gene encodes the

surface layer protein of Synechocystis sp. strain PCC 6803,” J.

Bacteriol., vol. 195, no. 23, pp. 5370–80, 2013.

[70] A. El Zoeiby, F. Sanschagrin, and R. C. Levesque, “Structure and

function of the Mur enzymes: development of novel inhibitors,” Mol.


[71] C. R. Sweet, A. H. Williams, M. J. Karbarz, C. Werts, S. R. Kalb, R. J.

Cotter, and C. R. H. Raetz, “Enzymatic synthesis of lipid A molecules

with four amide-linked acyl chains. LpxA acyltransferases selective for

REFERENCES

35

an analog of UDP-N-acetylglucosamine in which an amine replaces

the 3"-hydroxyl group,” J. Biol. Chem., vol. 279, no. 24, pp. 25411–9,

2004.

[72] E. Sauvage, F. Kerff, M. Terrak, J. a Ayala, and P. Charlier, “The

penicillin-binding proteins: structure and role in peptidoglycan

biosynthesis,” FEMS Microbiol. Rev., vol. 32, no. 2, pp. 234–58, 2008.

[73] M. Marbouty, K. Mazouni, C. Saguez, C. Cassier-Chauvat, and F.

Chauvat, “Characterization of the Synechocystis strain PCC 6803

penicillin-binding proteins and cytokinetic proteins FtsQ and FtsW

and their network of interactions with ZipN,” J. Bacteriol., vol. 191,

no. 16, pp. 5123–33, 2009.

[74] S. Yoshihara, X. Geng, S. Okamoto, K. Yura, T. Murata, M. Go, M.

Ohmori, and M. Ikeuchi, “Mutational analysis of genes involved in

pilus structure, motility and transformation competency in the

unicellular motile cyanobacterium Synechocystis sp. PCC 6803,”

Plant Cell Physiol., vol. 42, no. 1, pp. 63–73, 2001.

[75] D. Bhaya, A. Takahashi, P. Shahi, and R. G. Arthur, “Novel motility

mutants of Synechocystis Strain PCC 6803 generated by in vitro

transposon mutagenesis,” J. Bacteriol., vol. 183, no. 20, pp. 6140–

6143, 2001.

[76] H. Miranda, O. Cheregi, S. Netotea, T. R. Hvidsten, T. Moritz, and C.

Funk, “Co-expression analysis, proteomic and metabolomic study on

the impact of a Deg/HtrA protease triple mutant in Synechocystis sp.

PCC 6803 exposed to temperature and high light stress,” J.

Proteomics, vol. 78, pp. 294–311, 2013.

Documents

Deg/HtrA proteases of the cyanobacterium