Embed Size (px)
Citation preview
The Role of Smaug in Post-transcriptional Regulation
by
Linan Emily Chen
A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy
Molecular Genetics University of Toronto
© Copyright by Linan Emily Chen 2015
ii
The Role of Smaug in Post-transcriptional Regulation
Linan Emily Chen
Doctor of Philosophy
Molecular Genetics
University of Toronto
2015
Abstract
Smaug is a sequence-specific RNA-binding protein (RBP) and a multifunctional post-
transcriptional regulator. In this thesis, I have used genome level analysis to gain a panoramic
view of regulation by Smaug during early Drosophila embryogenesis. I show that Smaug plays a
direct and major role in the translation and stability of a large set of maternal mRNAs. Smaug’s
target transcripts function in a diverse array of processes including metabolism, lipid droplet
function, protein folding and protein stability, suggesting previously uncharacterized functions
for Smaug. I also performed detailed analysis of Smaug’s regulation in the embryo’s germ plasm
and found that Smaug sits at the top of a posttranscriptional regulatory cascade that controls
primordial germ-cell number by attenuating synthesis of the embryonic germ plasm. In the germ
plasm, Smaug binds to arrest mRNA to repress the translation of Bruno (BRU) protein. BRU
potentiates the production of Oskar (OSK) protein and germ plasm, Thus, repression of
Arrest/Bruno (Bru) by Smaug is required for the establishment of the correct number of germ
cells during early embryogenesis.
iii
Acknowledgments
I offer my sincere gratitude to my supervisors Dr. Howard Lipshitz and Dr. Craig Smibert for
their guidance and encouragement throughout of the course of my graduate career. The
completion of this thesis would not have been possible without their patience, kindness and
support for me. I would also like to thank my committee members Dr. Brenda Andrews and Dr.
Benjamin Blencowe for providing me with guidance and valuable insights on the project, and Dr.
Tim Westwood and Dr. Quaid Morris for providing me with vaulable suggestions and great
expertise on the project.
I would like to thank all the members of the Lipshitz and Smibert lab for their helpful technical
suggestions and delightful scientific discussion, and most of all, for their friendship and support.
Special thanks to Wael Tadros and Claudia Walser for their help, support and mentorship
through the most challenging parts of my graduate study. Special thanks to Ben Pinder, Angelo
Karaiskakis, and Hua Luo for their technical help and suggestions. Big thanks to Najeeb
Siddiqui, Jason Dumelie, Xiao Li, John Laver and Zhiyong Yang for working with me on this
project and contributing to the completion of this project.
Finally, big thanks to all the family and friends who have supported and encouraged me
throughout this long journey. To my father Ye Chen and my mother Shaoling Huang, thank you
for providing me love, support and motivation throughout the course of my education. To my
husband Eric Feng, thank you for your love, support and understanding, this is not possible
without you. To aunt Shaohua, uncle Carl and cousin Chriss, thank you for your support and
loving care. To my friends Frankie, Michael, Jen and Mengshu, thank you for your support and
friendship.
iv
Table of Contents
Acknowledgments .......................................................................................................................... iii
Table of Contents ........................................................................................................................... iv
List of Tables .................................................................................................................................. x
List of Figures ............................................................................................................................... xii
Chapter 1 Introduction .................................................................................................................... 1
1.1 Post-transcriptional regulation ............................................................................................ 1
1.1.1 Pre-mRNA processing ............................................................................................ 1
1.1.2 Transport and Localization of mRNAs ................................................................... 1
1.1.2.1 Mechanisms of mRNA localization ......................................................... 2
1.1.2.1.1 Localization by active transport ............................................................... 2
1.1.2.1.2 Localization by diffusion and entrapment ................................................ 3
1.1.2.1.3 Localization by degradation and protection ............................................. 4
1.1.2.2 Role of mRNA localization in oocytes and early embryos ...................... 4
1.1.2.3 Role of mRNA localization in the nervous system and migrating cells ... 5
1.1.3 Translational regulation of mRNAs ........................................................................ 6
1.1.3.1 Mechanisms of translational regulation .................................................... 6
1.1.3.1.1 Cap-dependent translation initiation ......................................................... 7
1.1.3.1.2 Regulation of cap-dependent translation initiation ................................... 7
1.1.3.1.2.1 Translational repression by 4E-BPs ......................................................... 7
1.1.3.1.2.2 Translational repression by 4EHPs .......................................................... 8
1.1.3.1.3 Role of poly(A) and PABP in translation initiation ............................... 10
1.1.3.1.4 Translational regulation by modulation of poly(A) tail length............... 10
v
1.1.3.1.5 Translational repression post-initiation .................................................. 11
1.1.3.2 Coupling of translational regulation with mRNA localization ............... 11
1.1.3.3 Role of translational regulation in the nervous system........................... 11
1.1.3.4 Role of translational regulation in cell division and embryonic
development ........................................................................................... 12
1.1.4 Regulation of mRNA stability .............................................................................. 12
1.1.4.1 Mechanisms of mRNA decay ................................................................. 13
1.1.4.1.1 Deadenylation-dependent mRNA decay ................................................ 13
1.1.4.1.2 Deadenylation-independent mRNA decay ............................................. 14
1.1.4.1.3 Cis-acting elements that function in mRNA degradation ....................... 14
1.1.4.1.3.1 ARE-mediated decay .............................................................................. 14
1.1.4.1.3.2 GRE-mediated decay .............................................................................. 15
1.1.4.1.3.3 IRE-mediated decay ............................................................................... 15
1.1.4.1.3.4 miRNA-mediated decay ......................................................................... 16
1.1.4.2 Role of mRNA turnover in early development ...................................... 16
1.1.5 Global analyses of post-transcriptional regulation ............................................... 17
1.1.5.1 Systematic identification of RNA-protein interactions .......................... 17
1.1.5.2 Global regulation of mRNA localization................................................ 18
1.1.5.3 Global regulation of translational control ............................................... 18
1.1.5.4 Global studies of mRNA decay .............................................................. 19
1.2 Smaug, a multifunctional post-transcriptional regulator of maternal RNAs .................... 20
1.2.1 Mechanisms of Smaug regulation ......................................................................... 20
1.2.1.1 RNA-binding SAM domain ................................................................... 22
1.2.1.2 Smaug recognition elements (SREs) ...................................................... 22
vi
1.2.1.3 Regulation of nos mRNA and translational repression .......................... 22
1.2.1.4 Regulation of Hsp83 mRNA and transcript decay ................................. 23
1.2.2 Temporal control of Smaug .................................................................................. 25
1.2.3 Role of Smaug in early Drosophila embryogenesis ............................................. 25
1.2.3.1 Role of Smaug in cleavage divisions ...................................................... 25
1.2.3.2 Role of Smaug in the MZT ..................................................................... 26
1.2.3.3 Role of Smaug in PGCs .......................................................................... 26
1.2.4 Role of Smaug in mammals .................................................................................. 26
1.3 Thesis rationale ................................................................................................................. 27
1.3.1 Global analysis of Smaug’s mRNA targets .......................................................... 27
1.3.2 The role of Smaug in the germ plasm ................................................................... 28
Chapter 2 Global regulation of mRNA translation and stability in the early Drosophila
embryo by the Smaug RNA-binding protein ........................................................................... 29
2.1 Abstract ............................................................................................................................. 30
2.2 Introduction ....................................................................................................................... 30
2.3 Results ............................................................................................................................... 33
2.3.1 The mRNAs encoded by 339 genes associate with Smaug .................................. 33
2.3.2 The mRNAs encoded by 342 genes are translationally repressed by Smaug ....... 37
2.3.3 Targets of Smaug-mediated translation are recruited to polysomes in a smaug
mutant ................................................................................................................... 47
2.3.4 Smaug is likely to repress the translation of ~3000 mRNA targets ...................... 47
2.3.5 SRE stem-loops are highly enriched in Smaug’s target mRNAs ......................... 54
2.3.6 Smaug co-regulates translational repression and degradation of a large fraction
of its target mRNAs .............................................................................................. 57
2.3.7 Subcellular localization of Smaug’s target mRNAs ............................................. 65
2.3.8 Functional analysis of Smaug-regulated mRNAs ................................................. 66
2.3.9 Validation of new Smaug targets .......................................................................... 87
vii
2.4 Discussion ......................................................................................................................... 87
2.4.1 Translational repression versus mRNA decay ...................................................... 88
2.4.2 Smaug’s role in the regulation of posterior-localized mRNAs ............................. 88
2.4.3 Identification of new biological functions for Smaug .......................................... 89
2.4.4 Biological implications of the large number of Smaug-target mRNAs ................ 91
2.5 Material and methods ........................................................................................................ 92
2.5.1 Drosophila Stocks ................................................................................................. 92
2.5.2 RNA co-immunoprecipitations ............................................................................. 93
2.5.3 Polysome gradients ............................................................................................... 93
2.5.4 Microarrays ........................................................................................................... 96
2.5.5 RT-qPCR ............................................................................................................... 97
2.5.6 Estimating the number of genes that are translationally repressed by Smaug ...... 97
2.5.7 SRE searching ....................................................................................................... 98
2.5.8 Localization pattern enrichment analysis .............................................................. 99
2.5.9 Western blots ........................................................................................................ 99
2.5.10 Glycolytic enzyme assays ..................................................................................... 99
Chapter 3 Smaug regulates primordial germ cell number by repressing synthesis of Bruno in
the germ plasm of Drosophila embryos ................................................................................. 101
3.1 Abstract ........................................................................................................................... 102
3.2 Introduction ..................................................................................................................... 102
3.3 Results ............................................................................................................................. 104
3.3.1 Smaug accumulates in the germ plasm of early embryos ................................... 104
3.3.2 Smaug and VAS are dynamic components of the polar granules ....................... 104
3.3.3 Excess OSK protein is produced in the germ plasm of smaug mutant embryos 110
3.3.4 smaug mutants produce excess primordial germ cells ........................................ 115
3.3.5 The arrest mRNA co-purifies with the Smaug protein ...................................... 118
viii
3.3.6 Smaug represses translation of the arrest mRNA in the germ plasm ................. 123
3.3.7 Over-expression of BRU in the embryo leads to excess OSK, germ plasm and
primordial germ cells .......................................................................................... 123
3.4 Discussion ....................................................................................................................... 129
3.4.1 Mechanisms of posttranscriptional regulation by Smaug ................................... 129
3.4.2 Regulation of primordial germ cell number and embryonic pattern ................... 131
3.5 Experimental procedures ................................................................................................ 132
3.5.1 Drosophila culture and mutants .......................................................................... 132
3.5.2 Generation and molecular analysis of new smaug alleles .................................. 133
3.5.3 Construction of transgenes and production of transgenic flies ........................... 133
3.5.4 Confocal microscopy and live imaging .............................................................. 134
3.5.5 Fluorescence recovery after photobleaching (FRAP) ......................................... 134
3.5.6 Analysis of Smaug particle movement ............................................................... 135
3.5.7 Cryo-immunogold electron microscopy ............................................................. 135
3.5.8 Three-dimensional reconstruction and counts of primordial germ cells and
PH3-labeled nuclei .............................................................................................. 136
3.5.9 Immunoprecipitation of Smaug together with its bound mRNAs ...................... 137
3.5.10 Reverse transcription-quantitative PCR (RT-qPCR) .......................................... 137
3.5.11 Western blotting .................................................................................................. 137
Chapter 4 Conclusions and Future Directions ............................................................................ 139
4.1 Conclusions ..................................................................................................................... 139
4.2 Future Directions ............................................................................................................ 139
4.2.1 A general strategy for the investigation of de novo Smaug mRNA targets ........ 140
4.2.1.1 Construction of SRE mutants ............................................................... 140
4.2.1.2 Validation of Smaug mRNA targets ..................................................... 140
4.2.2 What are the mechanisms of Smaug’s regulation? ............................................. 141
4.2.3 Smaug’s mRNA targets and their biological roles ............................................. 143
ix
4.2.3.1 Role of Smaug in embryonic body patterning and germ line
specification .......................................................................................... 144
4.2.3.1.1 Smaug’s role in pole plasm and pole cell regulation ............................ 145
4.2.3.1.2 Smaug’s role in embryonic body patterning ........................................ 145
4.2.3.2 The role of Smaug in ubiquitin-mediated proteolysis .......................... 148
4.2.3.2.1 The role of Smaug in regulating proteasome activity .......................... 148
4.2.3.2.2 Role of Smaug in modulating ubiquitin-mediated proteolysis and its
impact in cell cycle regulation ....................................................................................... 150
4.2.3.3 Role of Smaug in modulating embryonic metabolism ......................... 151
4.2.3.3.1 Role of Smaug in the regulation of glycolysis ..................................... 151
4.2.3.3.2 Role of Smaug in energy state switches ............................................... 155
4.2.4 Closing statement ................................................................................................ 156
Appendix A: Supplemental data files ......................................................................................... 157
Appendix B: List of Abbreviations ............................................................................................. 158
References ................................................................................................................................... 160
x
List of Tables
Table 1 Replicate-to-replicate comparisons of transcript microarray signal intensities from RIP-
Chip experiments* ........................................................................................................................ 35
Table 2 Replicate-to-replicate comparisons of transcript microarray signal intensities from wild-
type polysome gradients* ............................................................................................................. 44
Table 3 Replicate-to-replicate comparisons of transcript microarray signal intensities from smaug
mutant polysome gradients* ......................................................................................................... 45
Table 4 Replicate-to-replicate comparisons of transcript microarray signal intensities from smaug
mutant polysome gradients +/- puromycin* ................................................................................. 46
Table 5 Fly-FISH localization patterns and degradation categories enriched among Smaug-bound
mRNAs ......................................................................................................................................... 69
Table 6 Smaug-bound mRNAs that are localized to the posterior of the embryo ........................ 71
Table 7 Smaug-bound mRNAs annotated with roles in cell cycle, checkpoint response and/or
response to DNA damage ............................................................................................................. 73
Table 8 Smaug-bound mRNAs annotated with roles in transcription and/or chromatin .............. 74
Table 9 Gene set annotation enrichment analysis results for Smaug-bound mRNAs .................. 75
Table 10 Smaug-bound mRNAs encode proteins in the Interpro Chaperonin Cpn60/TCP-1
family ............................................................................................................................................ 76
Table 11 Status of other components of the TRiC/CCT complex ................................................ 77
Table 12 Smaug-bound mRNAs that encode proteins found in the proteasome regulatory particle
and the ubiquitin proteasome pathway ......................................................................................... 78
Table 13 Status of other components of the proteasome regulatory particle ................................ 79
Table 14 Smaug-bound mRNAs that encode proteins associated with lipid droplets .................. 80
xi
Table 15 Smaug-bound mRNAs that encode metabolic enzymes ................................................ 81
Table 16 Smaug-bound mRNAs that encode metabolic enzymes involved in glycolysis and
related pathways ............................................................................................................................ 83
Table 17 Status of other enzymes in the glycolytic pathway ........................................................ 84
Table 18 MuD-PIT normalized spectral counts for the wild-type and smaug-mutant primordial
germ cells. ................................................................................................................................... 111
Table 19 Primordial germ cell number in wild type and in smaug mutants. .............................. 114
xii
List of Figures
Figure 1 Translation initiation. ....................................................................................................... 9
Figure 2 RNA-binding SAM domain and Smaug recognition element. ....................................... 21
Figure 3 Smaug is a major regulator of maternal transcript degradation. .................................... 24
Figure 4 Identification of Smaug-bound mRNAs. ........................................................................ 34
Figure 5 Validation of Smaug-bound mRNAs. ............................................................................ 36
Figure 6 Polysome gradient characterization. ............................................................................... 38
Figure 7 Inclusion of Pool 2 in the denominator when calculating the TI does not significantly
alter the values calculated without Pool 2. .................................................................................... 39
Figure 8 Validation of polysome gradient/microarrays. ............................................................... 40
Figure 9 Inclusion of Pool 2 in the denominator when calculating the change in TI in smaug
mutants versus wild type does not significantly alter the values calculated without Pool 2. ....... 41
Figure 10 Identification of the targets of Smaug-mediated translational repression. ................... 43
Figure 11 Smaug represses the translation of thousands of mRNAs in the early embryo. ........... 49
Figure 12 Comparison of the TIs in wild-type versus smaug-mutant embryos for mRNAs that are
unlikely to be bound by Smaug. ................................................................................................... 51
Figure 13 Kernel density plots comparing the change in TI in smaug-mutant versus wild-type
embryos for the top and bottom Smaug binders. .......................................................................... 53
Figure 14 SREs are enriched in Smaug-bound mRNAs and those that are translationally
repressed by Smaug. ..................................................................................................................... 56
Figure 15 Comparisons of Smaug-bound, repressed and degraded data sets. .............................. 60
Figure 16 FDR-based rank of genes from polysome gradient microarrays. ................................. 61
xiii
Figure 17 Overlaps between Smaug-bound genes and Smaug- regulated genes. ......................... 62
Figure 18 Smaug degraded and Smaug repressed mRNAs are enriched for SREs. ..................... 63
Figure 19 SRE scores for the 5’UTR, open reading frame and 3’UTR of Smaug-regulated
mRNAs. ........................................................................................................................................ 64
Figure 20 Fly-FISH degradation categories and localization patterns enriched among Smaug-
bound mRNAs. ............................................................................................................................. 70
Figure 21 Validation of new Smaug targets. ................................................................................ 85
Figure 22 Glycolytic enzymes are overexpressed in smaug mutant embryos. ............................. 86
Figure 23 Assessing the integrity of polysome fractionated mRNA. ........................................... 95
Figure 24 Smaug protein is enriched in the germ plasm of early embryos. ............................... 106
Figure 25 Smaug is a component of the polar granules. ............................................................. 107
Figure 26 FRAP analysis of Smaug and VAS protein in the polar granules. ............................. 109
Figure 27 Excess OSK protein is produced in the germ plasm of smaug mutants. .................... 112
Figure 28 Excess OSK protein is produced in the germ plasm of smaug mutants. .................... 113
Figure 29 Excess primordial germ cells form in smaug mutants. ............................................... 116
Figure 30 Smaug represses translation of the arrest mRNA in the germ plasm. ........................ 120
Figure 31 arrest mRNA is stabilized in the bulk cytoplasm of smaug mutants. ........................ 121
Figure 32 Excess BRU protein is produced in the germ plasm of smaug mutants. .................... 122
Figure 33 Predicted SREs in the arrest mRNA and the mutations introduced into the SREs for
construction of the S-A(5xSRE-)-S transgene. ............................................................................ 125
Figure 34 Over-expression of BRU protein results in synthesis of excess OSK and production of
extra primordial germ cells. ........................................................................................................ 127
xiv
Figure 35 Pathway for regulation of primordial germ cell number by Smaug. .......................... 128
Figure 36 Proposed new pathway for regulation of primordial germ cell number by Smaug. ... 147
Figure 37 Glycolysis ................................................................................................................... 154
1
Chapter 1 Introduction
Post-transcriptional regulation plays an important role in regulating gene expression. In
particular, cytoplasmic mRNA regulatory processes including transcript localization, translation
and transcript degradation are essential for many biological functions. In this chapter, I will
review the mechanisms and functions of post-transcriptional regulation.
1.1 Post-transcriptional regulation
1.1.1 Pre-mRNA processing
In the nucleus, a nascent transcript undergoes several steps of processing to become a mature
mRNA, including: 5’-capping, splicing, 3’-end cleavage and polyadenylation (Moore &
Proudfoot, 2009; reviewed in Shatkin & Manley, 2000). First, a pre-mRNA is capped at the
5’end with a 7-methylguanosine cap. Capping protects a transcript from 5’-3’ exonucleolytic
degradation and this modification provides a binding site for the cap-binding complex (CBC)
and eIF4E. Next, a large ribonucleoprotein (RNP) complex known as the spliceosome, which
catalyzes reactions that join adjacent exons while removing introns, splices the pre-mRNA.
Finally, a transcript is cleaved and polyadenylated at its 3’end. This involves a cis-acting
polyadenylation signal which is located just upstream of the site where the RNA is cleaved and
polyadenylated. The poly(A) tail is added by a poly(A) polymerase and bound by poly(A)-
binding protein (PABP). Processed, mature mRNAs will then be exported from the nucleus to
the cytoplasm where additional post-transcriptional processes can act to regulate mRNA’s
subcellular localization, translation and stability.
In the following sections, I will explore the mechanisms and biological significance of
cytoplasmic post-transcriptional regulation, as well as their interplay in the post-transcriptional
regulatory networks.
1.1.2 Transport and Localization of mRNAs
30 years ago, the first examples of mRNA localization were found in mammalian glial cells
(Colman et al., 1982) and ascidian (Jeffery et al., 1983), Xenopus (Rebagliati et al., 1985) and
Drosophila eggs and/or embryos (Berleth et al., 1988; Frigerio et al., 1986). Messenger RNA
localization is now known to be a very common mechanism found in many cell types in a wide
2
variety of organisms (reviewed in Martin & Ephrussi, 2009). A more recent large-scale study has
also shown that a large proportion of cellular mRNAs assume discrete sub-cellular localizations
(Lecuyer et al., 2007). Thus, mRNA localization is a wide spread post-transcriptional mechanism
used to organize cells into discrete compartments.
Localization of mRNAs serves diverse biological functions. First, localization of mRNAs allows
localized translation and the generation of high local protein concentrations. Second, transporting
mRNAs can prevent the expression of proteins in locations where they can cause deleterious
effects. These two functions are best illustrated by the examples of localized maternal
determinants involved in the specification of embryonic body axes and the germline (reviewed in
King et al., 2005; Kugler & Lasko, 2009; Medioni et al., 2012; Sardet et al., 2005). Third,
localization of mRNA can fine tune protein expression and is especially important for localized
translation in response to environmental stimuli. This function of mRNA localization has best
demonstrated by examples from neurons and migrating fibroblasts (reviewed in Condeelis &
Singer, 2005; Liao et al., 2015).
In this section, I will first introduce the different mechanisms of mRNA localization in
eukaryotes, followed by a detailed review of the biological implications of mRNA localization
and localized translation, with an emphasis on the role of mRNA localization in early Drosophila
development.
1.1.2.1 Mechanisms of mRNA localization
There are three distinct but not mutually exclusive mechanisms for the asymmetric localization
mRNAs in the cell cytoplasm: 1) active transport along a polarized cytoskeleton network; 2)
entrapment of diffusing transcripts by a localized anchor; and 3) spatially regulated mRNA
stability where transcripts found in one region of the cell are degraded while transcripts in other
regions are protected from degradation.
1.1.2.1.1 Localization by active transport
Messenger RNAs can be actively transported by microtubule- or actin-associated motors. In
S.cerevisiae, at least 30 mRNAs are transported and localized to the bud tip (Andoh et al., 2006;
Aronov et al., 2007; Shepard et al., 2003; Takizawa et al., 2000). One of the best-studied
examples, Ash1 mRNA, is transported along the actin-based cytoskeleton (reviewed in
3
Gonsalvez et al., 2005; Paquin & Chartrand, 2008). Localization of Ash1 mRNA requires She2,
which is an RNA binding protein (RBP) that interacts with Ash1 mRNA through elements within
the transcript’s open reading frame (ORF) and 3’ untranslated region (UTR) (Chartrand et al.,
1999). She2 also interacts with the She3 protein, which in turn interacts with the myosin motor
protein, Myo4 (Bohl et al., 2000; Long et al., 2000; Takizawa & Vale, 2000). Myo4 thus is able
to transport Ash1 mRNA along actin microfilaments.
In Xenopus oocytes, a group of vegetally localized mRNAs is transported by microtubule-based
mechanisms (King et al., 2005). For example, Vg1 mRNA is transported to the vegetal cortex by
Kinesin-1 (Messitt et al., 2008) and Kinesin-2 (Betley et al., 2004) on specialized populations of
microtubules. Kinesin-2 has been shown to co-localize with Vg1 mRNA in the vegetal cortex,
and injection of antibodies that block either Kinesin-1 or Kinesin-2 function disrupts Vg1 mRNA
localization (Messitt et al., 2008). A more recent study has also shown a role for dynein in Vg1
mRNA transport. Dynein directs unidirectional transport of RNA towards the vegetal cortex
while Kinesin-1 promotes bidirectional transport (Gagnon et al., 2013).
Another interesting case is the transport of β-actin mRNA. Zipcode binding protein 1 (ZBP1)
interacts with β-actin mRNA to target it for localization in a variety of cell types (reviewed in
Condeelis & Singer, 2005). In fibroblasts, β-actin mRNA transport is actin-dependent (Sundell
& Singer, 1991). In neurons, β-actin mRNA is transported via microtubules (Zhang et al., 1999).
In Drosophila ooyctes and embryos, a large number of maternal mRNAs is localized (reviewed
in Kugler & Lasko, 2009; Lecuyer et al., 2007). Among these, are transcripts that are involved in
embryonic axis determination and/or germline specification (reviewed in Kugler & Lasko,
2009). The localization of bicoid (bcd) and oskar (osk) mRNAs has been particularly well
characterized. The osk mRNA is translocated to the posterior of the Drosophila embryo via a
microtubule-based mechanism driven by the plus-end directed microtubule motor, Kinesin-1
(Brendza et al., 2000). The delivery of bcd transcripts to the anterior of the oocyte is driven by
the minus-end directed motor, Dynein (Weil et al., 2010).
1.1.2.1.2 Localization by diffusion and entrapment
Localization of Drosophila nanos (nos) mRNA during oogenesis is achieved through a
mechanism whereby transcripts are trapped by an actin-dependent anchor localized at the
4
posterior of the oocyte (Forrest & Gavis, 2003). This trapping is mediated by elements located in
the nos mRNA 3’UTR and is very inefficient as only 4% of nos mRNA is localized to the
posterior pole (Bergsten & Gavis, 1999; Gavis et al., 1996). In early Xenopus oocytes,
localization of the germ plasm mRNAs, nanos1 and Xdazl, to the vegetally localized
mitochondrial cloud has also been proposed to use a diffusion and entrapment mechanism
(Chang et al., 2004).
1.1.2.1.3 Localization by degradation and protection
Drosophila Hsp83 mRNA is localized through a mechanism involving generalized degradation
of the transcript in the bulk of the embryo, while transcripts present in the germ plasm and germ
cells of the embryo are protected from degradation (Ding et al., 1993). Hsp83 mRNA is degraded
in the bulk cytoplasm during the general wave of maternal transcript degradation that is mediated
by the Smaug RBP (Semotok et al., 2008). Regulation of Hsp83 mRNA by Smaug will be
discussed in more detail below. Degradation/protection mechanisms also participate in the
localization of other mRNAs to the posterior of the Drosophila embryo (Lecuyer et al., 2007)
and germ plasm of the zebrafish embryo (Koprunner et al., 2001; Wolke et al., 2002). The
mechanisms that protect transcripts from decay in particular sub-cellular regions remain unclear.
1.1.2.2 Role of mRNA localization in oocytes and early embryos
During development, localization of maternal mRNAs in oocytes and embryos of several
organisms, including ascidians, Drosophila and Xenopus, plays important roles in the
establishment of embryonic body patterning and specification of the germ cell fate (reviewed in
Kugler & Lasko, 2009; Medioni et al., 2012).
In Xenopus oocytes, many maternal mRNAs are localized to the vegetal cortex during oogenesis.
Vegetal localized mRNAs follow two distinct localization pathways (King et al., 2005).
Messenger RNAs encoding the RBPs Xdazl and Nanos1, which play important roles in germ cell
determination and germ cell migration, are localized initially to the mitochondrial cloud and,
later, to the vegetal cortical region during early oogenesis (Chang et al., 2004). During mid
oogenesis, Vg1 and VgT mRNAs, which are required for endoderm and mesoderm induction,
become restricted to the vegetal cortex of the oocyte, and are inherited by the most vegetal cells
of the embryo (Birsoy et al., 2006; Zhang et al., 1998).
5
The Drosophila oocyte provides one of the best systems to study mRNA localization and
localized translation. During oogenesis, localization of gurken (grk), bcd, osk and nos mRNAs is
required for the establishment of embryonic axes (Johnstone & Lasko, 2001; reviewed in Lasko,
1999). The initial establishment of the oocyte antero-posterior (AP) axis requires localization of
grk mRNA and protein to the posterior pole (Gonzalez-Reyes et al., 1995; Roth et al., 1995).
Localization of Grk protein, a TGFα-like ligand, induces the adjacent follicle cells to take on a
posterior fate. These cells, in turn, signal back to the oocyte to specify its posterior (Gonzalez-
Reyes et al., 1995; Roth et al., 1995). Localization of osk mRNA coupled with translational
regulation is also essential for embryonic patterning. Ectopic expression of OSK results in lethal
patterning defects (Ephrussi & Lehmann, 1992; Smith et al., 1992). At Stage 8 of oogenesis, osk
mRNA is localized to the posterior of the embryo, and its localization requires many trans-acting
factors (reviewed in St Johnston, 2005). For example, Staufen (STAU) protein co-localizes with
osk mRNA throughout oogenesis, and might be a trans-factor that targets osk mRNA (Kim-Ha et
al., 1991; St Johnston, 2005; St Johnston et al., 1991). NOS is another translational regulator that
is essential for posterior embryonic patterning (reviewed in Johnstone & Lasko, 2001; Kugler &
Lasko, 2009). Once osk mRNA is localized to the posterior pole, it is translated, and in turn,
recruits other factors to the posterior, including Vasa (VAS) protein. VAS, another core posterior
determinant, is a DEAD-box RNA helicase that can bind to and regulate the translation of
mRNAs (Ephrussi & Lehmann, 1992; Liang et al., 1994). Together, localized OSK and VAS are
required for the recruitment of nos mRNAs to the posterior of the oocyte later in oogenesis
(Ephrussi et al., 1991; Ephrussi & Lehmann, 1992).
The pathway described above, which is involved in the specification of the embryo AP axis, also
functions in germ cell specification (reviewed in Mahowald, 2001). The mRNAs and proteins
localized to the posterior generate a specialized cytoplasm at the posterior that is known as germ
plasm (a.k.a. the pole plasm) (reviewed in Mahowald, 2001). This germ plasm is incorporated
into the pole cells that form at the embryo posterior and directs them to take on the primordial
germ cell fate (reviewed in Mahowald, 2001).
1.1.2.3 Role of mRNA localization in the nervous system and migrating cells
In the nervous system, localization of mRNA plays important roles in axon steering and neuronal
development. In Xenopus and murine axonal processes, localized translation in response to
6
external signals is essential for the growth cone turning response. For example, β-actin mRNA is
localized to the side of the growth cone that is exposed to attractive cues, which leads to a local
increase in β-actin protein production, in turn triggering growth cone turning (Leung et al., 2006;
Welshhans & Bassell, 2011; Yao et al., 2006). During development, mRNA localization is also
responsible for the remodeling of dendritic trees. For example, studies have shown that the
transport of nos mRNA to dendrites is required for proper branching of peripheral neurons in
Drosophila larvae (Brechbiel & Gavis, 2008). In young mouse hippocampal neurons, the RBP,
Staufen-1, is required for the localization of β-actin mRNA to dendrites (Vessey et al., 2008).
Inactivation of Staufen-1 leads to a reduction in dendritic length and branching (Vessey et al.,
2008).
In migrating cells, asymmetric localization of mRNA is also responsible for establishing and
maintaining cell polarity and directionality. In chicken fibroblasts, mRNAs encoding β-actin and
subunits of the actin-polymerization nucleating Arp2/3 complex are all localized behind the
leading edge, which suggests that, in cell protrusions, local translation allows efficient local
complex assembly to regulate actin dynamics and directionality (Lawrence & Singer, 1986;
Mingle et al., 2005). In mouse fibroblasts, at least 50 transcript species are enriched in
pseudopodial protrusions produced in response to migratory stimuli (Mili et al., 2008).
1.1.3 Translational regulation of mRNAs
Translational regulation is also essential in many cellular processes. During early development,
translational regulation is a major mechanism used to modulate gene expression prior to zygotic
genome activation. It is required for proper temporal and spatial expression of proteins by
ensuring the repression of localizing mRNAs until they reach their destination as well as
repression of unlocalized mRNAs in the event that localization is inefficient. Coupled to mRNA
localization, translational regulation also provides a quick and efficient means of localized
protein production in response to environmental stimuli.
1.1.3.1 Mechanisms of translational regulation
Translation of an mRNA involves three steps: initiation, elongation and termination. In general,
translation is regulated at initiation, the rate-limiting step of translation, thus allowing rapid and
reversible control of protein synthesis. In the following section, I will first introduce the
7
mechanism of cap-dependent translation initiation, followed by a discussion of the factors and
mechanisms that regulate initiation. Next, I will review the role of the poly (A) tail in initiating
translation, as well as mechanisms of translational regulation involving the modulation of
poly(A) tail length.
1.1.3.1.1 Cap-dependent translation initiation
The 5’ cap of eukaryotic mRNAs plays a crucial role in initiating translation. Cap-dependent
translation initiation can be divided into several important steps including: binding of the eIF4F
complex to the mRNA; formation of the 43S pre-initiation complex and its recruitment to the 5’
cap of the mRNA; scanning of the 5’UTR in a 5’-to-3’ direction by the 43S pre-initiation
complex; recognition of the initiation codon and 48S initiation complex formation; joining of the
60S subunit to the 48S complex; and assembly and elongation of competent 80S ribosomes
(reviewed in Jackson et al., 2010).
1.1.3.1.2 Regulation of cap-dependent translation initiation
Formation of the eIF4F complex is an important step of translation initiation that is often subject
to regulation (reviewed in Jackson et al., 2010). The eIF4F complex contains eIF4E, eIF4G and
eIF4A (Figure 1). eIF4E is a cap-binding protein that initiates the assembly of the eIF4F
complex. eIF4G is a scaffold protein that can bind to eIF4E through its eIF4E-binding motif,
which has the consensus YXXXXLɸ (where ɸ is a hydrophobic residue and X is any amino acid)
(Mader et al., 1995). Interaction between eIF4E and eIF4G recruits eIF4G to the cap and eIF4G,
in turn, recruits eIF3. eIF3 is a multiprotein complex that interacts with and recruits the 40S
ribosome to the 5’end of the mRNA. Because the interactions of eIF4E with the 5’cap and eIF4E
with eIF4G are crucial to translation initiation, they are the targets for multiple modes of
translational regulation.
1.1.3.1.2.1 Translational repression by 4E-BPs
A well-characterized mechanism of translational repression is the inhibition of eIF4F complex
formation by a number of non-homologous proteins that carry eIF4E-binding motifs. The
presence of eIF4E-binding motifs within these proteins, which are known as eIF4E-binding
proteins (4E-BPs), allows them to inhibit translation by blocking the eIF4E-eIF4G interaction
(Marcotrigiano et al., 1999). In mammals, there are three low-molecular-weight 4E-BPs: 4E-
8
BP1, 4E-BP2 and 4E-BP3 (Pause et al., 1994; Poulin et al., 1998). The interactions of these 4E-
BPs with eIF4E are regulated by phosphorylation, with the hypophosphorylated forms able to
bind eIF4E (Gingras et al., 2001; Haghighat et al., 1995; Pause et al., 1994). In yeast, two 4E-
BPs are found: Caf20p and Eap1p (Altmann et al., 1997; Cosentino et al., 2000). Both of these
4E-BPs can cause translational repression of reporter mRNAs through eIF4E interaction
(Altmann et al., 1997; Cosentino et al., 2000). A recent study had shown that yeast 4E-BPs can
modulate the translation of more than 1000 transcripts, and they can form complexes with PUF
proteins (Cridge et al., 2010). In Xenopus, Maskin is a 4E-BP that is known to disrupt eIF4F
formation(Stebbins-Boaz et al., 1999). In immature oocytes, Maskin is recruited to specific
mRNAs through its interaction with the RBP, Cytoplasmic Polyadenylation Element-Binding
protein (CPEB) (Cao & Richter, 2002; Stebbins-Boaz et al., 1999). Thus, the CPEB-Maskin
complex represses the translation of mRNAs that contain binding sites for CPEB, which are
known as cytoplasmic polyadenylation elements (CPEs) (Cao & Richter, 2002; Hake & Richter,
1994). In Drosophila, an example of a 4E-BP is the Cup protein. Cup mediates translational
repression of unlocalized osk and nos mRNAs during oogenesis and early embryogenesis,
respectively (Nakamura et al., 2004; Nelson et al., 2004). The RBP BRU recruits Cup to osk
mRNA through BRU-binding sites (a.k.a. BRU Response Elements or BREs) in the transcript
3’UTR, thereby repressing its translation (Nakamura et al., 2004). Translational repression of nos
mRNA involves recruitment of Cup by the Smaug RBP through Smaug-binding sites (a.k.a.
Smaug Recognition Elements or SREs) in the nos 3’UTR (Nelson et al., 2004).
1.1.3.1.2.2 Translational repression by 4EHPs
Another mechanism of translational repression involves the eIF4E homologous protein (4EHP).
Like eIF4E, 4EHP, binds directly to the cap structure, but 4EHP cannot interact with eIF4G
(Joshi et al., 2004; Rom et al., 1998). Thus, 4EHP prevents translation by blocking assembly of
the eIF4F complex. In Drosophila, 4EHP is required to establish the AP axis of the embryo
through its repression of the caudal (cad) and hunchback (hb) mRNAs (Cho et al., 2005). Bicoid
protein binds to 4EHP and recruits it to cad mRNA through Bicoid-binding sites within the cad
transcript (Cho et al., 2005). 4EHP is recruited to hb mRNA through the PUM/NOS/BRAT
protein complex bound to the transcript (Cho et al., 2006). In mammals, there is also evidence of
4EHP-dependent translational repression. For example, the mRNA encoding a transcriptional
9
regulator important for stem cell renewal, Homeobox B4 (HOXB4), is repressed by 4EHP
(Villaescusa et al., 2009).
The translation initiation complex eIF4F contains eIF4E, eIF4G and eIF4A. eIF4G functions as a
scaffold protein to assemble the initiation complex onto mRNAs. eIF4G also binds eIF3. eIF3
recruits 40S ribosome to the 5’UTR of the mRNA. Poly(A)-binding protein, PABP, can also
interact with eIF4G, and this interation results in a closed loop formation that stimulates
translation initiation.
Figure 1 Translation initiation.
10
1.1.3.1.3 Role of poly(A) and PABP in translation initiation
The poly(A) tail of mRNAs plays a major role in translation initiation (Figure 1). In vitro
experiments have shown that adding a poly(A) tail to a reporter mRNA can result in a modest
increase in translation (Munroe & Jacobson, 1990; Wickens, 1990). RNA electroporation
experiments using animal, plant, and yeast cells showed that the 5’cap or the 3’ poly(A) tail can
each individually stimulate translation, while together they synergistically stimulate translation
(Gallie, 1991). In subsequent yeast biochemical and genetic studies, poly(A)-binding protein
(PABP), a protein that binds to the poly(A) tail, was found to physically interact with eIF4G,
which was bound at the 5’ end of the mRNA through its interaction with eIF4E (Tarun & Sachs,
1995, 1996; Tarun et al., 1997). This series of interactions results in the formation of a closed
loop mRNA structure and underlies the ability of the cap and poly(A) tail to synergistically
stimulate translation initiation (Wells et al., 1998).
1.1.3.1.4 Translational regulation by modulation of poly(A) tail length
Maternal mRNAs during oogenesis and early embryogenesis provide some of the best examples
of translational regulation via modulation of poly(A) tail length. In the maturing Xenopus oocyte,
a population of maternal mRNAs is stored in a dormant form with short poly(A) tails.
Translation remains repressed until their poly(A) tails are elongated in a mechanism that requires
CPEB (Hake & Richter, 1994; McGrew et al., 1989). Thus, CPEB is responsible for both
translational repression and activation of maternal mRNAs. As described above, Maskin
interacts with eIF4E and CPEB to disrupt eIF4G recruitment, and maintain mRNA in a
translational repressed state (Stebbins-Boaz et al., 1999). Upon CPEB-mediated polyadenylation,
PABP is recruited to the poly(A) tail. The binding of PABP in turn recruits eIF4G, which
displaces Maskin thereby permitting translation initiation (Cao & Richter, 2002; reviewed in
Richter & Sonenberg, 2005; Wakiyama et al., 2000). CPEB also regulates translation by
assembling a complex that includes a deadenylase called poly(A) ribonuclease (PARN) and a
poly(A) polymerase known as Germline-development factor 2 (Gld2) (Barnard et al., 2004; Kim
& Richter, 2006). Within this complex the deadenylase activity of PARN is more active than the
poly(A) polymerase activity of Gld2, so CPEB-bound mRNAs have short poly(A) tails and are
translationally repressed (Kim & Richter, 2006). Phosphorylation of CPEB causes expulsion of
PARN from the complex allowing Gld2 to elongate the poly(A) tail, thereby activating
translation (Sarkissian et al., 2004).
11
1.1.3.1.5 Translational repression post-initiation
Translational repression can also occur post-initiation. An example that acts at the step of
elongation is phosphorylation of the translation elongation factor, eEF2. Phosphorylation of
eEF2 prevents it from binding to ribosomes and results in reduced overall protein synthesis in
eukaryotes (reviewed in Kaul et al., 2011). Another example is the translational repression of nos
mRNA. In Drosophila, translation of unlocalized nos during oogenesis is repressed post-
initiation by the RNA binding protein, Glorund (Andrews et al., 2011; Kalifa et al., 2006).
1.1.3.2 Coupling of translational regulation with mRNA localization
Translational regulation is often coupled with mRNA localization to localize proteins to specific
subcellular compartments. Messenger RNAs being transported are generally translationally
repressed until they reach their final destination. Alternatively, specific subcellular distributions
of translational repressors and activators can function to provide localized protein production.
Localized mRNAs are typically repressed during transport to prevent production of ectopic
protein. For example, Drosophila osk mRNA is localized to the posterior pole and, while
undergoing transport, is translationally repressed by the RBPs, BRU and Apontic (Kim-Ha et al.,
1995; Lie & Macdonald, 1999). Disruption of these regulatory mechanisms results in ectopic
OSK protein, which results in lethal developmental defects in the embryo (Kim-Ha et al., 1995).
It has also been shown that, during transport of CPE-containing mRNAs, CPEB colocalizes with
Maskin in transport particles to repress target mRNAs being transported to dendrites (Huang et
al., 2003).
1.1.3.3 Role of translational regulation in the nervous system
As mentioned in the previous section, localization of mRNAs is important for the cellular
response to environmental stimuli. Coupled to localization, translational regulation also serves a
similar purpose. Localized translation is essential for neuronal growth cone guidance and
dendritic branching. For example, Zbp1/Vg1 RBP-directed repression of β-actin translation
during β-actin mRNA transport is required for the proper localization of β-actin protein in
neuronal processes (Leung et al., 2006; Yao et al., 2006). Another example is the regulation of
synaptic plasticity and memory formation. Localized translation of mRNAs plays an important
12
role in eliciting and maintaining long-term potentiation (LTP) or long-term depression (LTD),
and to stabilize long-term memory (reviewed in Bramham, 2008; Bramham et al., 2008).
1.1.3.4 Role of translational regulation in cell division and embryonic development
Translational regulation provides an efficient and rapid way to control gene expression.
Translational efficiency can be adjusted through various mechanisms to meet specific demands
for protein production within a cell. In Xenopus, for example, translational regulation of cyclin B
mRNA is critical for meiotic divisions during oogenesis and mitotic divisions in early embryos
(reviewed in Radford et al., 2008). The cyclin B1 mRNA is polyadenylated at oocyte maturation,
which activates its translation. Upon egg activation, further increase in poly(A) tail length results
in higher Cyclin B protein synthesis.
Translational control is also well characterized during Drosophila development. In oocytes and
embryos, translational regulation of maternal mRNAs is essential for embryonic patterning and
germ cell formation. For example, translational repression of cad, osk, and nos mRNAs in the
embryo anterior is required for posterior embryonic patterning (reviewed in Kong & Lasko,
2012) while translational repression of hb mRNA in the posterior of the embryo is essential for
anterior embryonic patterning (Murata & Wharton, 1995).
1.1.4 Regulation of mRNA stability
Modulation of mRNA stability is essential for coordinating global changes in gene expression
networks. Some of the best-characterized examples of degradation pathways and cis-acting
elements that function to coordinate gene expression networks can be found in early embryonic
development. In metazoans, zygotic transcription is silent during the initial stages of
development and, thus, early embryogenesis is programmed by maternally provided mRNAs
synthesized during oogenesis (reviewed in Tadros & Lipshitz, 2009; Walser & Lipshitz, 2011).
Regulation of maternal mRNA stability plays a prominent role in controlling their expression.
First, stability is coupled with mRNA localization and local translation. Second, during the
maternal-to-zygotic transition (MZT), the increase in zygotic synthesis of mRNAs is
accompanied by a degradation of many maternally deposited transcripts, thus enabling the
former to take over control of embryonic development. Therefore, early embryonic development
provides an excellent system for deciphering the mechanisms and functions of mRNA stability
13
regulation. In addition to its well-studied role in development, regulation of mRNA stability is
also essential for many other cellular functions (review in Balagopal et al., 2012).
In this section, I will first concentrate on the mechanisms of mRNA decay with a focus on
deadenylation-dependent mRNA decay. Next, I will review in more detail the functional role of
mRNA stability regulation during early embryonic development.
1.1.4.1 Mechanisms of mRNA decay
There are several major pathways that mediate mRNA degradation (reviewed in Garneau et al.,
2007). Deadenylation of an mRNA can trigger DCP1/DCP2-mediated removal of the 5’ cap
followed by 5’-3’ exonucleolytic decay mediated by XRN1. Deadenylated transcripts can also be
degraded in the 3’-5’ direction by a complex known as the exosome. In addition, there are
several deadenylation-independent mechanisms of mRNA decay, including nonsense-mediated
decay (NMD) and endonucleolytic cleavage-mediated decay.
1.1.4.1.1 Deadenylation-dependent mRNA decay
In most eukaryotes, the major pathways that mediate mRNA decay are deadenylation-dependent
and poly(A) tail removal by deadenylase complexes is often the rate-limiting step in decay
(reviewed in Wiederhold & Passmore, 2010). There are three major deadenylase complexes
known to function in eukaryotes: 1) the CCR4-NOT complex; 2) the PAN2/PAN3 complex; and
3) PARN.
The CCR4-NOT complex is conserved throughout eukaryotes, and is the main deadenylase in S.
cerevisiae (reviewed in Wiederhold & Passmore, 2010). Genetic studies in yeast have provided
much insight into the mechanisms and functions of this deadenylase complex. In yeast, the
CCR4-NOT complex is composed of nine subunits, two of which, CCR4 and POP2, are
homologous to unrelated deadenylase enzymes (Daugeron et al., 2001; reviewed in Garneau et
al., 2007; Tucker et al., 2002). Deletion of either the ccr4 or pop2 genes results in defective
deadenlyation (Chen et al., 2002; Daugeron et al., 2001; Tucker et al., 2002; Tucker et al., 2001).
In human cells, there are two isoforms of both CCR4 and POP2, which allows formation of a
variety of complexes (reviewed in Wiederhold & Passmore, 2010). The Drosophila CCR4/NOT
complex also contains two deadenylases (CCR4 and POP2), and mutations in CCR4, which is
14
encoded by the twin gene, have been reported to cause defects in germ cells (Morris et al., 2005;
Temme et al., 2004; Zaessinger et al., 2006).
The PAN2/PAN3 complex is also present in all eukaryotes. In yeast, the Pan2/Pan3 complex is
responsible for deadenylation in ccr4Δ mutants (Tucker et al., 2001). In this complex, Pan2 is a
deadenylase, and Pan3 is a regulator that binds to both Pan2 and PABP (Brown et al., 1996).
Deletion of the pan2 or pan3 genes results in minor deadenylation defects and longer poly(A)
tails, which suggests that the Pan2/Pan3 complex functions to trim poly(A) tails (Brown &
Sachs, 1998). Studies in mammals also suggest that the PAN complex carries out the initial
shortening of poly(A) tails to a specific length, followed by CCR4/NOT mediated completion of
deadenylation (Yamashita et al., 2005).
PARN is a cap-dependent deadenylase. Its deadenylase activity is stimulated by binding to the
5’cap of the mRNA, and inhibited by poly(A)- or cap-binding protein (reviewed in Parker &
Song, 2004). PARN has been characterized in Xenopus ooctyes and mammalian cells, and is
found in many higher eukaryotes, but not in Drosophila and S. cerevisiae.
1.1.4.1.2 Deadenylation-independent mRNA decay
Messenger RNA decay also occurs independent of deadenylation. The endonucleolytic cleavage-
mediated decay pathway involves the internal cleavage of mRNA to create unprotected 5’ and 3’
fragments, which are then used as substrates for exoribonucleolytic decay (reviewed in Beelman
& Parker, 1995). Another pathway of degradation is NMD. NMD targets bypass deadenylation
but do undergo 5’ cap removal by the decapping complex followed by 5′ to 3′ degradation by
XRN1 (reviewed in Chang et al., 2007).
1.1.4.1.3 Cis-acting elements that function in mRNA degradation
Deadenlyation of specific transcripts can be mediated by cis-acting elements present within the
transcript, which act as binding sites for trans-acting factors that function to recruit deadenylases
to the target mRNA.
1.1.4.1.3.1 ARE-mediated decay
AU-rich elements (AREs) are among the best-characterized cis-acting elements that
mediate mRNA decay in mammals (Bakheet et al., 2006; reviewed in Schoenberg & Maquat,
15
2012; Wu & Brewer, 2012). They are also found in yeast and Drosophila (Cairrao et al., 2009).
AREs function, at least in part, through deadenylation-dependent degradation and are often
located within the 3’UTRs of unstable mRNAs but can also be found within open reading
frames. AREs represent binding sites for a number of trans-acting factors, collectively known as
ARE-binding proteins (AUBPs). AUBPs, such as TTP and BRF1, mediate degradation through
association with decay factors such deadenlyases, decapping factors and exonucleases (reviewed
in Sanduja et al., 2011). While AREs were initially characterized as instability elements, they can
also serve to antagonize decay and stabilize mRNAs. The Hu/ELAV family AUBPs can function
to stabilize mRNAs and up-regulate translation (reviewed in Brennan & Steitz, 2001).
1.1.4.1.3.2 GRE-mediated decay
Another important cis-acting element involved in the regulation of mRNA stability in the
mammalian system is the GU-rich element (GRE). GREs have been found to be enriched in
unstable transcripts in muscle cells and T-cells (reviewed in Lee et al., 2010; Vlasova &
Bohjanen, 2008). GRE-containing transcripts encode proteins that are important for a wide
variety of cellular functions, including transcription, apoptosis, RNA processing, cell division,
signaling and metabolism (Rattenbacher et al., 2010). GREs can be bound by the human protein
CUGBP-1, known in frogs as EDEN-BP, and this binding promotes mRNA deadenylation and
degradation (Vlasova et al., 2008).
1.1.4.1.3.3 IRE-mediated decay
In higher eukaryotes, the iron-regulated degradation of transferrin receptor (TfR) mRNA is
mediated by another well-characterized class of mRNA decay cis-elements, the Iron Regulatory
Elements (IREs) (Binder et al., 1994). IREs are highly conserved RNA hairpins that can function
to regulate mRNA translation or stability (Binder et al., 1994; Caughman et al., 1988; reviewed
in Thomson et al., 1999). When intracellular iron concentrations are low, IRE-binding protein
(IRE-BP) binds to IREs to stabilize the mRNAs (Binder et al., 1994). When intracellular iron
concentration increases, the IRE-BP-IRE interaction weakens, permitting cleavage by
endonuclease (Binder et al., 1994).
16
1.1.4.1.3.4 miRNA-mediated decay
In addition to RBPs, small non-coding RNAs are also known to play an important role in post-
transcriptional regulation (reviewed in Fabian et al., 2010). microRNAs (miRs), a class of non-
coding RNAs 21-24 nucleotides long, function as part of the RNA-induced silencing complex
(RISC) to mediate translational repression and/or degradation of target mRNAs (reviewed in
Pratt & MacRae, 2009). miRs do not bind target transcripts alone but, instead, do so as part of a
miRNA-Argonaute (AGO) protein complex. In metazoans, miRs bind mRNAs with imperfect
complementarity, their binding specificity residing in the 2-8 nucleotide “seed” region of the
miRNA (Ameres et al., 2007; Haley et al., 2003; Lewis et al., 2005). While the mechanisms that
underlie the ability of miRs to regulate target transcripts remain controversial, perhaps the best
characterized involves the ability of AGOs to recruit GW182 to target transcripts (Behm-
Ansmant et al., 2006). GW182 can, in turn, recruit the CCR4/NOT deadenylase, which triggers
transcript deadenylation to produce translational repression and/or transcript decay (Behm-
Ansmant et al., 2006; Braun et al., 2011). miRs have been shown to mediate target-transcript
decay across animal species, including mammals, C. elegans, Drosophila, and zebrafish (Bagga
et al., 2005; Behm-Ansmant et al., 2006; Bushati et al., 2008; Guo et al., 2010). For example,
zebrafish miR-430 and Drosophila mir-309 miRs are responsible for the deadenylation and
degradation of hundreds of maternal mRNAs during early embryogenesis (Bushati et al., 2008;
Giraldez et al., 2006).
1.1.4.2 Role of mRNA turnover in early development
In addition to localization and translational control, turnover of mRNA is a highly regulated
process during early development. In early Drosophila embryos, a large wave of maternal
mRNA degradation occurs during the MZT (reviewed in Walser & Lipshitz, 2011). There are
two general types of pathways involved in the degradation of maternal mRNAs in embryos.
‘Maternal’ degradation pathways, triggered by egg activation, are mediated, in part, by the
maternally loaded RBP, Smaug, which binds to specific sequences known as Smaug Recognition
Elements (SREs) and is required for clearance of two-thirds of the unstable maternal mRNAs
(Tadros et al., 2007). Smaug recruits the CCR4/NOT deadenylase complex to mRNAs to cause
shortening of poly(A) tails and, thus, degradation of its targets (Semotok et al., 2005). Details of
Smaug will be discussed below. ‘Zygotic’ degradation pathways require zygotic transcription to
17
synthesize and/or activate degradation factors. In Drosophila, the miR-309 family of miRs serves
in one of the zygotic pathways (Bushati et al., 2008).
1.1.5 Global analyses of post-transcriptional regulation
Over the past few decades, genetics and biochemistry have provided with researchers useful
tools to study the mechanisms and functions of post-transcriptional regulation with a focus on
individual protein-RNA interactions. In recent years, the rapid development of genome-wide
tools, such as microarrays and next-generation sequencing, initiated an era of large scale “-omic”
studies. These technologies, in combination with biochemical methods, have allowed the global
mapping of RNA-protein interactions, which comprise the “ribonome”.
1.1.5.1 Systematic identification of RNA-protein interactions
RBP immunoprecipitation followed by microarray/sequencing of associated mRNAs is a robust
method that has been used to systematically identify RBP-RNA interactions. It involves co-
immunoprecipitation of an endogenous or tagged RBP followed by DNA microarray analysis
(RIP-Chip) or next-generation sequencing (RIP-Seq) to identify co-purifying RNAs (Keene et
al., 2006; Tenenbaum et al., 2000). RIP-Chip was first used in embryonal carcinoma stem cells
to identify targets of endogenous HuR, PABP and eIF4E (Tenenbaum et al., 2000).
Subsequently, it has been used to screen for binding targets for more than a hundred RBPs,
including examples in yeast, worms, flies and mammals (reviewed in Morris et al., 2010). Some
technical issues must be considered when analyzing the results of RIP-Chip/Seq experiments.
For example, mRNAs that co-purify with an RBP in RIPs can be either directly or indirectly
associated with that protein. One must also consider that re-association of protein-RNA
complexes can occur after cells are lysed, which might result in identification of non-
physiological RNA targets (Mili & Steitz, 2004). Several other techniques, including CLIP-Chip,
HITS-CLIP, and PAR-CLIP have been developed, aimed at the detection of direct, physiological
protein-RNA interactions (Hafner et al., 2010a, b; Licatalosi et al., 2008; Ule et al., 2003). In
general, they involve steps that crosslink proteins to directly bound transcripts prior to cell lysis.
With the accumulation of RIP-Chip/Seq studies, the idea of a eukaryotic RNA regulon was
proposed by drawing an analogy to DNA operons in prokaryotes (Keene, 2007). In eukaryotes,
transcription and protein production is uncoupled, and complex mechanisms are used to
18
coordinate gene expression. For example, at the DNA level, the chromatin landscape can be
modulated to coordinate transcription and gene expression. However, protein production is
highly modulated at the post-transcriptional level, and the expression of groups of functionally
related transcripts may be co-regulated by trans-acting factors, forming RNA regulons (Keene,
2007).
1.1.5.2 Global regulation of mRNA localization
In recent years, large-scale studies of mRNA localization have provided evidence that the
regulation of mRNA localization in the cell is much more prevalent than previously thought.
Several techniques have been developed for mapping the global distribution of localized
mRNAs. For example, RIP-Chip can be used to study mRNA localization. RIP-Chip of the yeast
She mRNP identified a group of transcripts localized to the bud-tip of dividing cells (Shepard et
al., 2003; Takizawa et al., 2000). Microscopy-based RNA in situ hybridization (ISH) has also
been used to detect localization patterns of mRNAs (Lein et al., 2007; Tomancak et al., 2002).
To examine mRNA localization at the subcellular level, high-throughput fluorescence ISH
(FISH) was developed using tyramide to greatly increase resolution. This study in Drosophila
embryos using FISH annotated 3300 mRNAs in embryos, and showed that 71% of these mRNAs
are localized (Lecuyer et al., 2007). Combining tissue sectioning with next-generation
sequencing of the RNAs in each section has also been used to assay transcript localization on a
global scale (Combs & Eisen, 2013).
1.1.5.3 Global regulation of translational control
Modulation of mRNA translation can be accomplished globally by targeting general mechanisms
that control translation, or for specific groups of mRNAs through cis-elements. In recent years,
mechanistic and functional insights have emerged from genome-wide studies of translational
control.
One of the most common experimental approaches to study translation regulation is polysome
profiling. Translationally active mRNAs are found on polysomes although not all polysome-
associated mRNAs are translated. In this method, fractionation of extracts using sucrose density
gradients is used to separate polysome-bound mRNAs from mRNAs that are not polysome-
bound (and thus cannot be translated). Traditionally, to study the translational regulation of a
19
specific mRNA, total RNA can be isolated from the fractions of a sucrose gradient and detected
using northern blot or RT-PCR analysis. To extend this method to the genomic scale, it can also
be combined with microarray or next-generation sequencing analysis. For example, to study the
dynamics of translational regulation during Drosophila embryonic development, polysome-
associated mRNAs were compared across several developmental stages using polysome
gradients and microarray analysis to determine the translational profile of each mRNA (Qin et
al., 2007). In yeast, the same method has been used to profile global translational regulation
(Arava et al., 2003). Ultimately, these methods can be combined with other studies such as RIP-
Chip/Seq, expression data and proteomics to provide a global view of gene expression regulation
(reviewed in Kuersten et al., 2013).
1.1.5.4 Global studies of mRNA decay
Large-scale studies of mRNAs decay have been established throughout a wide range of systems.
Initial studies in many model systems examined the rate of mRNA decay by detecting changes in
mRNA levels upon transcription inhibition (Grigull et al., 2004; Wang et al., 2002; Yang et al.,
2003). It was found that the mRNA stability did not correlate with ORF length, mRNA
abundance or ribosome density; however, rates of decay clustered among groups of transcripts
that encode similar cellular components or biological functions (Wang et al., 2002). Early
embryonic development also provides an excellent system to study mRNA decay, as early
embryos are typically transcriptionally silent, thus allowing one to measure the stability of
maternally provided mRNAs without having to use experimental means to inhibit transcription.
Global studies among a variety of eukaryotic species have shown that 30-40% of maternally
loaded transcripts are eliminated during the MZT in early embryos (reviewed in Walser &
Lipshitz, 2011).
20
1.2 Smaug, a multifunctional post-transcriptional regulator of maternal RNAs
Smaug is a multifunctional RBP that regulates the expression of maternally loaded mRNAs in
early Drosophila embryos. It was the first identified as a translational repressor of unlocalized
nos mRNA in the embryo’s bulk cytoplasm (Dahanukar et al., 1999; Dahanukar & Wharton,
1996; Smibert et al., 1999; Smibert et al., 1996). Subsequent studies identified Smaug as a
member of a conserved class of RBPs that is characterized by an RNA-binding SAM domain
(Aviv et al., 2003; Green et al., 2003). The SAM domain of Smaug and its homologues can
interact with Smaug recognition elements (SREs) to recruit other cellular protein or complexes to
mediate translational repression and/or mRNA degradation (Aviv et al., 2006; Aviv et al., 2003;
Edwards et al., 2006; Green et al., 2003; Oberstrass et al., 2006). Genetic and biochemical
analyses have also identified another mRNA targeted by Smaug for decay: Hsp83 (Semotok et
al., 2005; Semotok et al., 2008). More recent studies have shown that Smaug plays a major role
in coordinating maternal transcript degradation during the MZT in both the bulk cytoplasm and
the primordial germ cells of early embryos (Siddiqui et al., 2012; Tadros et al., 2007). Embryos
produced by smaug mutant mothers display a wide array of phenotypes, including nuclear
dropout, failure to slow and terminate the syncytial nuclear cleavage cycles, failure to activate
the DNA replication checkpoint, and failure to cellularize (Benoit et al., 2009; Dahanukar et al.,
1999). These phenotypic defects cannot be fully explained by the misregulation of nos and
Hsp83 alone but are likely to be caused by the misregulation of additional targets. Below, I will
provide a brief review of Smaug and its homologues, including studies aimed at deciphering the
mechanisms that underlie regulation by these proteins and their biological roles.
1.2.1 Mechanisms of Smaug regulation
Drosophila Smaug and its budding yeast homologue, Vts1, have been closely examined through
a series of biochemical and genetic experiments which have identified the mechanism that
underlies RNA binding and the physical interactions with regulatory complexes that trigger
transcript degradation or translational repression.
21
Figure 2 RNA-binding SAM domain and Smaug recognition element.
(A) Cladogram representing overall sequence similarity and domain architecture of the Smaug
homologs. Species abbreviations: dm, Drosophila melanogaster; ag, Anopheles gambiae; hs,
Homo sapiens; mm, Mus musculus; ce, Caenorhabditis elegans; ca, Candida albicans; sp,
Schizosaccharomyces pombe; sc, Saccharomyces cerevisiae. SAM, sterile-α motif; SSR1, Smaug
similarity region 1; SSR2 Smaug similarity region 2; Zif, CCHC zinc-finger domain. Adapted
from (Aviv et al., 2003). (B) Schematic of a model SRE: a hair pin structure with non-specific
stem and a loop consensus of CNGGN(0-3). N represents any nucleotide.
22
1.2.1.1 RNA-binding SAM domain
Sterile Alpha Motif (SAM) domains are known to participate in a diverse set of interactions.
They can interact with other SAM domains, non-SAM-containing proteins and lipids (reviewed
in Qiao & Bowie, 2005). In Drosophila Smaug, the SAM domain includes basic amino acid
residues and binds to RNA (Aviv et al., 2003; Green et al., 2003). Using sequence alignments, a
conserved family of Smaug homologs was found in budding yeast, C. elegans, mouse and
humans, which defines a class of post-transcriptional regulators that is characterized by an RNA-
binding SAM domain (Figure 2A) (Aviv et al., 2003). The Smaug SAM domain binds RNA with
high affinity, and mutations of a conserved surface of the Smaug SAM domain abolish RNA
binding (Aviv et al., 2003). The SAM domain of Vts1 binds RNA with the same specificity as
Drosophila Smaug (Aviv et al., 2003). In addition to the SAM domain, Smaug also has a Smaug
Similarity Region 1 (SSR1) domain, which functions as a dimerization domain and is conserved
in yeast, mouse and human Smaug homologues (Figure 2A) (Aviv et al., 2003). Another
conserved domain, SSR2, whose function is unknown, is present in its mouse and human
homologues (Figure 2A) (Aviv et al., 2003).
1.2.1.2 Smaug recognition elements (SREs)
The SAM domain of Smaug and Vts1 can specifically recognize RNA-hairpin structures termed
Smaug Recognition Elements (SREs). The first SRE hairpin containing a penta-loop of CUGGC
was found in the nos mRNA 3’UTR (Smibert et al., 1996). Using the nos SRE as a prototype,
subsequent in vitro and structural analyses identified an SRE consensus of a four-or-more base-
pair non-specific stem and a specific loop sequence of CNGGN(0-3) (Figure 2B) (Smibert et al.,
1999; Smibert et al., 1996). Biochemical and structural studies of Vts1’s SAM-SRE complex
revealed that the RNA interaction surface of the SAM domain predominantly recognizes the
shape of the SRE, and the only base-specific contact made between SAM and an SRE is at the
G-3 position of the SRE loop (Aviv et al., 2006; Edwards et al., 2006; Johnson & Donaldson,
2006; Oberstrass et al., 2006).
1.2.1.3 Regulation of nos mRNA and translational repression
The first identified mRNA target of Smaug was the nos mRNA. In the early embryo, nos mRNA
is inefficiently localized to the posterior of the embryo, where it is translated, while the 96% of
nos mRNA that escapes the localization machinery is translationally repressed by Smaug in the
23
bulk of the embryo (Dahanukar et al., 1999; Dahanukar & Wharton, 1996; Smibert et al., 1999;
Smibert et al., 1996). One mechanism responsible for the translational repression of unlocalized
nos involves Smaug-mediated recruitment of the eIF4E-binding protein, Cup, to the transcript
(Nelson et al., 2004). The presence of Cup blocks the interaction between eIF4E and eIF4G,
thereby preventing translation initiation (Nelson et al., 2004). More recently, it has been reported
that Smaug repression of nos translation is mediated by AGO1 (Pinder & Smibert, 2013). Smaug
associates with AGO1 and recruits it to nos mRNAs in a miR-independent manner (Pinder &
Smibert, 2013).
1.2.1.4 Regulation of Hsp83 mRNA and transcript decay
Another major role of Smaug and its homologues is to initiate transcript degradation. Using a
reporter construct bearing three SREs, an earlier study showed that the present of SREs is
sufficient to cause degradation, and that this degradation is mediated by the CCR4 deadenylase
(Aviv et al., 2003). Vts1 can also initiate deadenylation through recruitment of the CCR4-NOT
deadenylase complex (Rendl et al., 2008). In Drosophila, Smaug is a major regulator of maternal
mRNA decay in the early Drosophila embryo, with two-thirds of unstable maternal transcripts
cleared in a Smaug-dependent manner (Figure 3) (Tadros et al., 2007). A known decay target of
Smaug is Hsp83 mRNA (Semotok et al., 2005). Hsp83 mRNA has eight predicted SREs in its
ORF, and mutation of all of these SREs stabilizes the transcript (Semotok et al., 2008). Smaug
initiates Hsp83 mRNA degradation through the recruitment of CCR4-NOT deadenylase to the
transcript, causing the removal of the poly(A) tail and degradation (Semotok et al., 2005;
Semotok et al., 2008).
24
Figure 3 Smaug is a major regulator of maternal transcript degradation.
Microarray-based gene expression profiling of maternal transcript stability in activated
unfertilized eggs from wild-type and smaug mutant females. 1069 destabilized maternal
transcripts are analyzed; many are stabilized in a smaug mutant. Y-axis: log base 2 of ratio. X-
axis: hours after egg laid (AEL). Adapted from (Tadros et al., 2007).
25
1.2.2 Temporal control of Smaug
The smaug mRNA is also a target of post-transcriptional regulation. In Drosophila, smaug
transcripts are present in the oocytes, but expression of the Smaug protein is repressed
throughout oogenesis. Ectopic expression of Smaug in ovaries disrupts oogenesis (Semotok et
al., 2005). Smaug translation after egg activation requires the PAN GU kinase (Tadros et al.,
2007). Smaug is present throughout the first three hours of embryonic development (Smibert et
al., 1999), peaking at 1-2h, during the MZT (Benoit et al., 2009). Smaug is actively eliminated
from the bulk cytoplasm after the first three hours of embryogenesis but it persists in the germ
plasm and the primordial germ cells (Siddiqui et al., 2012; Smibert et al., 1999).
1.2.3 Role of Smaug in early Drosophila embryogenesis
Embryos produced by homozygous smaug mothers display lethal defects. The smaug mutant
phenotype includes defects in nuclear division, defects in maternal mRNA degradation and
zygotic transcription, cell cycle and cell cycle checkpoint activation defects, and failure to
cellularize and gastrulate (Benoit et al., 2009; Dahanukar et al., 1999). The developmental
phenotypes of smaug mutants start to emerge at embryonic cell cycle 10-11, corresponding to the
time that zygotic genome activation would normally occur (Benoit et al., 2009; Dahanukar et al.,
1999).
1.2.3.1 Role of Smaug in cleavage divisions
In Drosophila, early embryonic development proceeds through several stages: syncytial
blastoderm, cellular blastoderm and gastrulation. Prior to and during the syncytial blastoderm
stage, the embryo undergoes rapid, synchronous, syncytial nuclear divisions. The first eight
cycles average 8 minutes and occur within the centre of the yolk. During the next two divisions,
nuclei migrate to the periphery of the embryo, following which the replication checkpoint
becomes active and the cell cycle slows down. During the interphase of nuclear cycle 14,
cellularization occurs via membrane invagination, also marking the midblastula transition
(MBT). Cellularization is the first event in embryogenesis that requires zygotic transcription. In
smaug mutants, both the gradual slowing of the cell cycle and replication checkpoint activation
fail (Benoit et al., 2009).
26
1.2.3.2 Role of Smaug in the MZT
As described above, Smaug plays a major role in the MZT through its ability to induce the
degradation of a large fraction of maternally loaded mRNAs. In addition, Smaug is required for
high-level transcription of most zygotic mRNAs at the onset of zygotic genome activation
(Benoit et al., 2009). Indeed, in smaug mutants, RNA Polymerase II fails to achieve Serine-2
phosphorylation of its C-terminal domain, a hallmark of active transcription (Benoit et al., 2009).
Also, high-level expression of the miR-309 cluster is activated 2 hours post-fertilization, and is
essential for the degradation for a subset of maternal transcripts (Bushati et al., 2008).
Transcription of this family of miRs requires Smaug (Benoit et al., 2009). Thus, Smaug is also
indirectly involved in the zygotic degradation pathway.
1.2.3.3 Role of Smaug in PGCs
In addition to its role in the soma, gene expression profiling on sorted wild-type and smaug-
mutant primordial germ cells (PGCs) indicated that Smaug plays a major role in the regulation of
the PGC MZT. That study showed that 34% of unstable PGC transcripts are Smaug-dependent,
and that these transcripts are enriched for ones encoding posttranscriptional regulators of germ
plasm assembly or developmental proteins, some of which control stem cell division or
transcriptional regulation (Siddiqui et al., 2012). In Drosophila embryos, zygotic genome
activation is delayed in the PGCs compared to the soma (Siddiqui et al., 2012). When examining
zygotic gene expression in smaug mutants, it was found that 36% of zygotically expressed
transcripts are expressed at significantly lower levels in the mutant PGCs (Siddiqui et al., 2012).
Thus, Smaug is a major regulator of the MZT in both the somatic and germ-line lineages.
1.2.4 Role of Smaug in mammals
To this point I have focused on Drosophila Smaug and budding yeast Vts1. There are two
homologues of Smaug in mammals. A human homologue of Smaug, hSmaug1, can repress
translation of SRE-containing messengers in fibroblast cell lines (Baez & Boccaccio, 2005).
Recently, hSmaug1 has also been reported to play a role in modulating CUG-induced toxicity in
human Myotonic Dystrophy Type 1 (MD1) (de Haro et al., 2013). In that study, hSmaug1 was
shown to up-regulate the activity of the CUGBP1-containing translational complexes in DM1 (de
Haro et al., 2013). A murine protein, mSmaug 1, is expressed in the central nervous system and
27
may have a role in RNA granule formation and translational regulation in neurons (Baez &
Boccaccio, 2005).
1.3 Thesis rationale
1.3.1 Global analysis of Smaug’s mRNA targets
Smaug has two well-characterized target mRNAs, nos and Hsp83. Genetic and biochemical
studies of these two target mRNAs have provided insights into the mechanisms and functions of
Smaug’s role in post-transcriptional regulation. Embryos produced by homozygous smaug-
mutant mothers show defects in nuclear division, cell cycle and cell cycle checkpoint activation,
and they fail to undergo zygotic genome activation and cellularization (Benoit et al., 2009).
These phenotypes cannot be fully explained by the misregulation of nos or Hsp83 alone (Benoit
et al., 2009; Tadros et al., 2007), implying that Smaug has additional direct targets misregulation
of which leads to these additional phenotypes. The first goal of my project (Chapter 2) was to
identify and characterize these additional mRNA targets. In order to achieve this goal, I used a
genome-wide strategy, RIP-Chip, to recover endogenous Smaug mRNPs and identify mRNAs
that are associated with Smaug. RIP-Chip allowed me to identify mRNAs that are present in the
Smaug mRNP and highly likely to be directly bound and regulated by Smaug. These studies
revealed an important role for Smaug in regulating cellular processes including metabolism,
protein turnover and protein folding, opening the way to a more detailed understanding of
Smaug’s functions in early embryos.
Hsp83 mRNA is destabilized but not repressed by Smaug whereas nanos mRNA is repressed but
not destabilized (Semotok & Lipshitz, 2007). A previous global expression study has shown that
over two-thirds of unstable maternal transcripts are stabilized in smaug-mutant embryos (Tadros
et al., 2007) but global analysis of mRNAs that are translationally repressed by Smaug had not
been conducted. Thus, it was not possible to address whether Smaug regulates the stability or
repression of its target mRNAs but not both. My second goal (Chapter 2), in collaboration with
Jason Dumelie, was to globally identify mRNAs that are translationally repressed by Smaug.
When this list was compared to those previously identified as requiring Smaug for degradation
(Tadros et al., 2007), we found that a large proportion of Smaug targets are regulated by Smaug
for both translational repression and transcript degradation. Likewise, the majority of the direct
targets of Smaug that I identified by RIP-Chip were both repressed and destabilized; thus Hsp83
28
and nanos mRNAs are the exception rather than the rule. Future studies will be needed to
understand the mechanistic basis for individual versus co-regulation of these two post-
transcriptional processes.
1.3.2 The role of Smaug in the germ plasm
Smaug protein is present uniformly at a high level in the early embryo (Benoit et al., 2009;
Dahanukar et al., 1999; Smibert et al., 1999). During interphase of nuclear cycle 14, Smaug
protein level declines rapidly. Although Smaug protein is eliminated from the bulk cytoplasm,
studies have shown that Smaug persists in the germ plasm and primordial germ cells of the
embryo (Siddiqui et al., 2012; Smibert et al., 1999). Smaug’s role in the germ plasm is unknown.
Chapter 3 focuses on the biological role of Smaug in the germ plasm and provides a molecular
basis for a novel smaug phenotype discovered by my collaborator, Najeeb Siddiqui – namely,
that embryos from smaug-mutant females produce more primordial germ cells that wild-type.
Bruno is a known post-transcriptional regulator of osk mRNA in the oocyte, serving as a
repressor of unlocalized osk mRNA and, potentially, as an activator of osk translation in the
oocyte’s germ plasm (Chekulaeva et al., 2006; Kim-Ha et al., 1995; Reveal et al., 2010; Webster
et al., 1997). Bruno is encoded by the arrest mRNA, which contains multiple potential SREs,
and was high on the list of Smaug targets that I identified in Chapter 2. We hypothesized that
Smaug attenuates arrest/Bruno mRNA translation in the germ plasm of embryos, which in turn
down-regulates Bruno-dependent activation of osk mRNA translation, and thus germ plasm
synthesis and primordial germ cell number in the embryo. To examine Smaug’s regulation of
arrest mRNA, I first confirmed that arrest mRNA co-purifies with Smaug. I also evaluated the
level of Bruno protein in smaug mutants, and found that protein levels are elevated in smaug
mutants. To further test the hypothesis, I asked whether over-expression of Bruno during
embryogenesis could result in excess OSK production and formation of extra germ cells. I
expressed smaug 5’UTR-arrest ORF(5xSRE-)-smaug 3’UTR transgenic mRNA in the embryo,
which resulted in over-expression of the Bruno protein, particularly in the germ plasm, and an
increased number of primordial germ cells in the embryo.
29
Chapter 2 Global regulation of mRNA translation and stability in the early Drosophila embryo by the Smaug RNA-binding protein
The work described in this chapter was published in:
Chen L, Dumelie JG, Li X, Cheng MH, Yang Z, Laver JD, Siddiqui NU, Westwood JT, Morris
Q, Lipshitz HD, Smibert CA: Global regulation of mRNA translation and stability in the
early Drosophila embryo by the Smaug RNA-binding protein. Genome Biol 2014, 15:R4
Contributions:
I performed the following experiments and analysis: RIP-Chip and subsequent analysis, RT-
qPCR validation, GO term analysis, and protein expression assessment of selected Smaug
targets. Jason G. Dumelie performed the following experiments and analysis: polysome
gradient/microarray and subsequent data analyses; SRE-enrichment analyses. Jason G. Dumelie
and I together carried out comparisons of binding, translation and stability data sets. Xiao Li
performed normalization of polysome gradient/microarray data. Zhiyong Yang performed the
enzyme assays. Matthew H.K. Cheng generated the smaug30
and smaug47
alleles. Najeeb U.
Siddiqui made smaug rescue transgene constructs. John D. Laver performed mRNA localization
enrichment analysis.
30
2.1 Abstract
Smaug is the founding member of a conserved class of RBPs that induces the degradation and/or
represses the translation of mRNAs in the early Drosophila embryo. Smaug has two identified
direct target mRNAs that it differentially regulates: Smaug represses the translation of nos
mRNA but has only a modest effect on its stability, while Smaug destabilizes Hsp83 mRNA but
has no detectable effect on Hsp83 translation. Smaug is required to destabilize well over a
thousand mRNAs in the early embryo but whether these transcripts represent direct targets of
Smaug is unclear and the extent of Smaug-mediated translational repression is unknown. To gain
a panoramic view of Smaug function in the early embryo I identified mRNAs that are bound to
Smaug using RNA co-immunoprecipitation followed by hybridization to DNA microarrays.
With my collaborators, I also identified the mRNAs that are translationally repressed by Smaug
using polysome gradients and microarrays. Comparison of the bound mRNAs to those that are
translationally repressed by Smaug and those that require Smaug for their degradation suggests
that a large fraction of Smaug’s target mRNAs are both translationally repressed and degraded
by Smaug. Smaug directly regulates components of the TRiC/CCT chaperonin, the proteasome
regulatory particle and lipid droplets as well as many metabolic enzymes, including several
glycolytic enzymes. I concluded that Smaug plays a direct and global role in regulating the
translation and stability of a large fraction of the mRNAs in the early Drosophila embryo and has
unanticipated functions in control of protein folding and degradation, lipid droplet function and
metabolism.
2.2 Introduction
Post-transcriptional regulatory mechanisms that function in the cytoplasm to control mRNA
translation, stability and subcellular localization play essential roles in a wide variety of
biological processes. While these types of controls function in a variety of cell types, they are
particularly prevalent during early metazoan development where mRNAs synthesized from the
mother’s genome direct the early stages of embryogenesis (Tadros & Lipshitz, 2009). Indeed,
genome-wide studies in Drosophila, C. elegans, zebrafish and mouse embryos have highlighted
the substantial role that cytoplasmic post-transcriptional regulation plays in early embryos
(Bazzini et al., 2012; Benoit et al., 2009; Bushati et al., 2008; De Renzis et al., 2007; Ferg et al.,
31
2007; Giraldez et al., 2006; Hamatani et al., 2004; Lecuyer et al., 2007; Mathavan et al., 2005;
Qin et al., 2007; Tadros et al., 2007; Tadros & Lipshitz, 2009; Thomsen et al., 2010).
During early embryogenesis regulation of specific transcripts is achieved through cis-acting
elements that represent binding sites for miRNAs or RBPs. For example, miRNAs induce
degradation of specific transcripts in both zebrafish and Drosophila (Bushati et al., 2008;
Giraldez et al., 2006). Similarly the RBP Smaug plays a major role in mRNA destabilization in
the early Drosophila embryo (Tadros et al., 2007). Smaug is the founding member of a
conserved family of post-transcriptional regulators that bind target mRNAs through stem-loop
structures, known as SREs (Aviv et al., 2003; Baez & Boccaccio, 2005; Dahanukar et al., 1999;
Smibert et al., 1999; Smibert et al., 1996). SRE recognition by Smaug family members is
mediated by a SAM domain, which contains a cluster of conserved basic residues that function
as an RNA-binding surface (Aviv et al., 2006; Aviv et al., 2003; Green et al., 2003; Johnson &
Donaldson, 2006; Oberstrass et al., 2006).
Upon binding to target mRNAs Smaug family members repress translation and/or induce
transcript decay through their ability to recruit various factors to a transcript (Aviv et al., 2003;
Baez & Boccaccio, 2005; Dahanukar et al., 1999; Rendl et al., 2008; Rendl et al., 2012; Smibert
et al., 1999; Smibert et al., 1996). For example, Drosophila Smaug can recruit the Cup protein to
an mRNA and Cup in turn interacts with the cap-binding protein, eIF4E (Nelson et al., 2004).
The Cup/eIF4E interaction inhibits translation by blocking eIF4E-mediated 40S ribosomal
subunit recruitment. Smaug can also recruit AGO1 to an mRNA, thereby repressing translation
(Pinder & Smibert, 2013). Typically, Ago proteins are bound to small RNAs, such as miRNAs,
that function to target the AGO1 protein to transcripts (Hammell, 2008). In contrast, Smaug can
recruit AGO1 in a miRNA-independent manner (Pinder & Smibert, 2013).
Smaug can also remove an mRNA’s poly(A) tail through its ability to recruit the CCR4/NOT
deadenylase (Jeske et al., 2006; Semotok et al., 2005; Semotok et al., 2008; Zaessinger et al.,
2006). In the case of at least one target mRNA this recruitment is thought to involve a complex
containing Smaug and the Piwi-type AGO proteins, Aubergine and AGO3 (Rouget et al., 2010).
This complex has been proposed to bind this target transcript through SREs (bound by Smaug)
together with sites complementary to piwi-RNAs (piRNAs) that are bound to AGO3 and/or
Aubergine. Since the poly(A) tail plays a role in both initiating translation and stabilizing an
32
mRNA, deadenylase recruitment can, in principle, both block translation and/or induce transcript
decay.
Smaug has two well-characterized target mRNAs, nos and Hsp83. Smaug represses nos
translation through two SREs in the nos 3’UTR whereas loss of Smaug has only a modest effect
on nos mRNA stability (Dahanukar et al., 1999; Semotok et al., 2005; Semotok & Lipshitz,
2007; Smibert et al., 1999; Smibert et al., 1996). In contrast, Smaug induces the degradation of
Hsp83 mRNA through eight SREs in the Hsp83 open reading frame, while having no detectable
effect on Hsp83 translation (Semotok et al., 2005; Semotok et al., 2008). Thus, Smaug can
differentially regulate the expression of its target mRNAs.
Messenger RNAs of nos and Hsp83 are localized to the posterior of the embryo and Smaug’s
regulation of these two transcripts is intimately associated with their localization. nos mRNA is
inefficiently localized to the posterior and nos mRNA that escapes the localization machinery is
found distributed throughout the bulk of the embryo where it is translationally repressed by
Smaug (Bergsten & Gavis, 1999; Dahanukar et al., 1999; Smibert et al., 1999; Smibert et al.,
1996; Wang & Lehmann, 1991). nos mRNA localized to the posterior is not repressed by Smaug
and thus, NOS protein expression is restricted to the posterior of the embryo. Hsp83 mRNA is
uniformly distributed in early embryos and, as embryogenesis proceeds, Smaug degrades Hsp83
mRNA in the bulk cytoplasm of the embryo while transcripts at the posterior of the embryo are
protected (Bashirullah et al., 1999; Ding et al., 1993; Semotok et al., 2005; Semotok et al., 2008).
This degradation/protection mechanism thus results in the localization of Hsp83 mRNA to the
posterior of the embryo.
In addition to nos and Hsp83 mRNA, Smaug is likely to regulate the expression of a large
number of mRNAs in the early embryo through direct binding. For example, genome-wide
experiments have shown that embryos collected from homozygous-mutant smaug females show
stabilization of ~1,800 transcripts (Tadros et al., 2007). In addition, smaug mutant embryos also
show cell-cycle defects associated with a failure of DNA replication checkpoint activation and
they also fail to undergo zygotic genome activation (Benoit et al., 2009; Dahanukar et al., 1999).
As neither of these phenotypes can be explained by a defect in Smaug’s regulation of nos or
Hsp83, this is consistent with a role for Smaug in regulation of the expression of additional
mRNAs.
33
To elucidate the global functions of Smaug in early embryos my collaborators and I employed
two genome-wide approaches: 1) RIP-Chip to identify mRNAs that are bound by Smaug and 2)
polysome gradients coupled to microarrays to identify targets of Smaug-mediated translational
repression. Our data suggest that Smaug directly regulates the expression of a large number of
mRNAs in the early embryo. Comparison of Smaug-bound mRNAs to those that are
translationally repressed by Smaug (identified in this study), and those that are degraded in a
Smaug-dependent manner (Tadros et al., 2007) suggest that two-thirds to three-quarters of
Smaug’s target mRNAs are either translationally repressed or degraded by Smaug. We also find
that Smaug regulates the expression of multiple mRNAs that are localized to the posterior of the
embryo. Gene set annotation enrichment analysis of the mRNAs directly bound by Smaug
suggests that it regulates a diverse array of processes in the early embryo, including protein
folding and degradation as well as metabolism. We present data indicating that Smaug regulates
the expression of several of the mRNAs that encode glycolytic enzymes.
2.3 Results
2.3.1 The mRNAs encoded by 339 genes associate with Smaug
To identify Smaug’s target mRNAs on a genome-wide scale I used RIP-Chip. Extracts, prepared
from 0-3 hour old wild-type embryos, were immunoprecipitated with an anti-Smaug antibody
(hereafter denoted as Smaug RIPs) while immunoprecipitations using non-immune serum served
as a negative control (hereafter denoted as control RIPs). Genes that were not expressed or were
expressed at low levels in starting crude extracts were removed from further analysis and
Significance Analysis of Microarrays (SAM) (Tusher et al., 2001) was then used to identify 339
genes whose mRNAs were significantly enriched in Smaug RIPs compared to control RIPs at a
False Discovery Rate (FDR) of <5% (Figure 4, Table 1 and Table S1). Importantly, this list
contains both of the well-characterized Smaug-target mRNAs, nos and Hsp83.
To verify the quality of my microarray data I used reverse transcription followed by RT-qPCR to
assay the enrichment of specific mRNAs in Smaug RIPs compared to control RIPs. Twelve
selected mRNAs from the RIP-Chip target list with FDRs <5%, including nos and Hsp83, were
enriched in Smaug RIPs compared to control RIPs. In contrast, four mRNAs that, based on my
RIP-Chip data, are not bound by Smaug, showed little or no enrichment (Figure 5).
34
Figure 4 Identification of Smaug-bound mRNAs.
The average, across three biological replicates and one technical replicate, of the microarray
signal intensities of each expressed transcript in the Smaug and control RIPs divided by the
signal intensities of each transcript in the immunoprecipitation inputs, were plotted against one
another. SAM analysis allowed for the identification of 384 transcripts (blue dots) representing
339 genes that are enriched in the Smaug RIPs versus control RIPs at an FDR of <5%. The dots
representing Smaug’s two known target mRNAs, nos and Hsp83, are indicated. The dark dashed
line represents no enrichment and the light dashed diagonal lines represent 2-fold enrichment or
depletion.
35
Table 1 Replicate-to-replicate comparisons of transcript microarray signal intensities from
RIP-Chip experiments*
input Smaug RIP control RIP
replicate 1a v. replicate 1b N/A 0.873 0.865
replicate 1a v. replicate 2 0.960 0.822 0.796
replicate 1a v. replicate 3 0.962 0.776 0.697
replicate 1b v. replicate 2 N/A 0.844 0.810
replicate 1b v. replicate 3 N/A 0.802 0.702
replicate 2 v. replicate 3 0.973 0.802 0.725
*The Pearson correlation coefficient was calculated for each pair of replicates and was used to
determine the degree of similarity between them. Replicates 1a and 1b are technical replicates of
one another, while replicates 1, 2 and 3 are biological replicates.
36
Figure 5 Validation of Smaug-bound mRNAs.
The fold enrichment of mRNAs in Smaug RIPs versus control RIPs was determined via RT-
qPCR and normalized to the levels of RpL32 mRNA in the immunoprecipitated material. The red
line indicates one-fold (i.e. no) enrichment. Results are the average of three independent
experiments and error bars indicate standard error of the mean.
37
2.3.2 The mRNAs encoded by 342 genes are translationally repressed by Smaug
Smaug is a multifunctional regulator that is capable of both repressing translation and inducing
the degradation of target mRNAs. To complement identification of the targets of Smaug-
mediated mRNA decay (Tadros et al., 2007) and identification of Smaug-bound mRNAs
described above, my collaborators and I employed polysome gradients coupled with microarrays
to identify targets of Smaug-mediated translational repression. This approach relies on the fact
that the position of an mRNA in a polysome gradient is related to the number of ribosomes
associated with that mRNA and can be employed to identify mRNAs that are regulated at the
level of translation initiation (Grolleau et al., 2002; Johannes et al., 1999; Kuhn et al., 2001). As
a first step towards applying this method we assessed the position of polysome-bound and free
ribosomes in our gradients. Extracts prepared from 0-2 hour old wild-type embryos were applied
to polysome gradients, in the absence or presence of EDTA. After centrifugation, gradients were
separated into 12 equal fractions and the level of 18S rRNA in these fractions was determined
via northern blot (Figure 6). In the absence of EDTA, rRNA is distributed throughout the
gradient, consistent with the presence of both free and polysome-associated ribosomes. In
contrast, treatment with EDTA, which disrupts polysomes, resulted in a shift of 18S rRNA to the
top fractions of the gradient. From these analyses we concluded that fractions 7-12 are
exclusively polysomal, while fractions 5-6 are a mix of polysome and non-polysomal material
and fractions 1-4 are non-polysomal fractions. Subsequent gradients were, therefore, divided into
4 unequal pooled fractions, which, from the top to the bottom of the gradient were: pool 1
(fractions 1-4) containing free mRNAs; pool 2 (fractions 5-6) containing a mix of free and
polysome bound mRNAs; pool 3 (fractions 7-9) and pool 4 (fractions 10-12), which both contain
polysome-associated mRNAs.
38
Figure 6 Polysome gradient characterization.
Polysome gradients were used to fractionate embryo extract collected 0-2 hours post-
egglaying in the presence or absence of EDTA. Gradients were fractionated into 12 equal
fractions, with fraction 1 representing the top of the gradient. RNA harvested from each
fraction was analyzed via northern blot using a probe to detect 18S ribosomal RNA.
39
Figure 7 Inclusion of Pool 2 in the denominator when calculating the TI does not
significantly alter the values calculated without Pool 2.
To assess whether including Pool 2 in the calculation had a significant effect on the calculated
TIs for mRNAs in wild type embryos, we compared TI = [3 + 4]/[1] on the y-axis to TI = [3 +
4]/[1 + 2] on the x-axis. The fact that the Pearson R2 = 0.92 shows that the inclusion of Pool 2
does not significantly alter the TI calculation used for our analyses, which excluded Pool 2 since
it was likely to contain a mix of free and polysome-bound mRNAs.
40
Figure 8 Validation of polysome gradient/microarrays.
TIs calculated in this study were used to generate box plots to compare the range of TIs for genes
that were previously categorized in Qin et al., (Qin et al., 2007) as ‘translationally active’ or
‘translationally inactive’ in embryos from the same developmental stage.
41
Figure 9 Inclusion of Pool 2 in the denominator when calculating the change in TI in smaug
mutants versus wild type does not significantly alter the values calculated without Pool 2.
To assess whether including Pool 2 in the calculation had a significant effect on the calculated
TIs for mRNAs in wild type embryos, we compared TI = [3 + 4]/[1] on the y-axis to TI = [3 +
4]/[1 + 2] on the x-axis. The fact that the Pearson R2 = 0.92 shows that the inclusion of Pool 2
does not significantly alter the TI calculation used for our analyses, which excluded Pool 2 since
it was likely to contain a mix of free and polysome-bound mRNAs.
42
RNA from the resulting pools was extracted and used to probe microarrays to assess the
distribution of transcripts within the gradient. To quantify the level of translation for each gene
we divided the average amount of the corresponding mRNA in pools 3 and 4 by the amount of
mRNA in pool 1; and we define the translation index (TI) as the log2-transformed version of this
ratio. We removed genes from the polysome data that were not expressed or were expressed at
only low levels. We also omitted the data from pool 2 in the TI calculation as it represents a
mixed population of translated and translationally repressed mRNAs. We note that inclusion of
pool 2 in the TI calculation has little effect on the calculated TI (Figure 7).
We then compared the TI for each gene in wild-type embryos to previous published
polysome/microarray data from similarly staged wild-type embryos (Qin et al., 2007). In this
previous study mRNA levels were assayed across polysome gradients divided into 12 fractions
and genes whose mRNAs were preferentially translated or preferentially untranslated were
identified. Figure 8 shows that the TI calculated from our data is significantly higher for the
preferentially translated group of mRNAs compared to preferentially untranslated group
(Wilcoxon rank-sum test, P < 3x10-16
) indicating an excellent correlation between the two data
sets.
To identify mRNAs that are translationally repressed by Smaug, we fractionated extracts from
embryos collected from 0-2 hour old homozygous mutant smaug mothers (hereafter denoted as
‘smaug mutant embryos’). We then compared the TI for each expressed gene in wild-type and
smaug-mutant embryos (Figure 10A); as above, we note that inclusion of pool 2 in the TI
calculation has little effect on the calculated TI (Figure 9). We expected the mRNA targets of
Smaug-mediated translational repression to shift their distribution from pool 1 in wild-type
embryos to pools 3 and 4 in smaug mutant embryos, thus resulting in an increase in the gene’s
TI. Using SAM we identified 342 genes, with an FDR of less than 5%, where the TI increased in
smaug-mutant embryos versus wild-type (Figure 10A, Table 2, Table 3, Table 4 and Table S2).
These genes represent a high-confidence list of Smaug-mediated translational repression targets.
As expected, neither Hsp83 nor nos mRNA was present in this high-confidence list: first, using
metabolic labeling, we previously showed that Smaug has no effect on Hsp83 translation
(Semotok et al., 2005); second, Clark et al. (Clark et al., 2000) have shown that a substantial
fraction of translationally repressed nos mRNA is associated with polysomes, consistent with our
observation that ~54% of nos mRNA is polysome-associated in wild-type embryos.
43
Figure 10 Identification of the targets of Smaug-mediated translational repression.
(A) The averages, across three biological replicates, of the TI in smaug mutant and wild-type
embryos were plotted against one another. SAM analysis allowed for the identification of 359
transcripts (blue dots) representing 342 genes that show an increase in TI in smaug mutant versus
wild-type at an FDR of <5%, while the red dots represent genes with FDRs of >5%. The solid
diagonal line represents no enrichment and the dotted diagonal lines represent 2-fold enrichment
or depletion. (B) Polysome gradients from smaug-mutant embryos were performed with or
without puromycin treatment and the average, across two biological replicates, of the TI for each
gene was calculated. Box plots show the range of TIs for genes where the TI increased in smaug-
mutant embryos versus wild-type with an FDR<5% and those with an FDR>5%, as defined in
(A).
44
Table 2 Replicate-to-replicate comparisons of transcript microarray signal intensities from
wild-type polysome gradients*
fraction 1 fraction 2 fraction 3 fraction 4
replicate 1 v.replicate 2 0.51 0.69 0.69 0.83
replicate 1 v.replicate 3 0.57 0.86 0.87 0.95
replicate 2 v.replicate 3 0.86 0.68 0.77 0.87
*The Pearson correlation coefficient was calculated for each pair of replicates and was used to
determine the degree of similarity between them. All replicates are biological replicates.
45
Table 3 Replicate-to-replicate comparisons of transcript microarray signal intensities from
smaug mutant polysome gradients*
fraction 1 fraction 2 fraction 3 fraction 4
replicate 1 v.replicate 2 0.67 0.77 0.79 0.7
replicate 1 v.replicate 3 0.63 0.69 0.84 0.82
replicate 2 v.replicate 3 0.6 0.7 0.84 0.82
*The Pearson correlation coefficient was calculated for each pair of replicates and was used to
determine the degree of similarity between them. All replicates are biological replicates.
46
Table 4 Replicate-to-replicate comparisons of transcript microarray signal intensities from
smaug mutant polysome gradients +/- puromycin*
fraction 1 fraction 2 fraction 3 fraction 4
replicate 1 v.replicate 2(- puromycin) 0.68 0.95 0.95 0.96
replicate 1 v.replicate 2(+ puromycin) 0.9 0.96 0.96 0.74
*The Pearson correlation coefficient was calculated for each pair of replicates and was used to
determine the degree of similarity between them. All replicates are biological replicates.
47
2.3.3 Targets of Smaug-mediated translation are recruited to polysomes in a smaug mutant
To confirm that the increase in TI was indeed the result of the recruitment of mRNAs onto
polysomes, smaug-mutant extracts were treated with puromycin, applied to polysome gradients
and the resulting fractions were then analyzed via microarray. Puromycin is a translational
inhibitor that causes premature chain termination during translation, thereby releasing mRNAs
from polysomes. Figure 10B shows that puromycin causes a significant decrease in the TI
(Fisher’s exact test, P < 3x10-16
), for the bulk of mRNAs present in smaug-mutant embryos (i.e.
those genes whose mRNAs show a FDR>5%), consistent with the fact that the majority of the
mRNAs that are present in pools 3 and 4 of our gradients are indeed polysome-associated.
Similarly, we also saw a significant decrease in the TI (Fisher’s exact test, P < 3x10-16
) for the
342 genes that are targets of Smaug translational repression (FDR<5%), consistent with the fact
that, in smaug-mutant embryos, these mRNAs are highly associated with polysomes.
2.3.4 Smaug is likely to repress the translation of ~3000 mRNA targets
In addition to those genes that meet an FDR of <5% (shown in blue in Figure 10A) the TI of a
large number of additional genes increased in smaug mutants. This suggests that a substantial
subset of the genes with >5% FDR are potential targets of Smaug-mediated translational
repression. Since SAM corrects for an average change in TI, if a large proportion of transcripts
were in fact translationally repressed by Smaug, SAM would over-correct, thereby increasing the
number of false negatives. To further evaluate the extent of Smaug-mediated translational
repression we generated lists of genes that encode mRNAs that are unlikely to be bound by
Smaug and are, therefore, unlikely to be targets of Smaug-mediated translational repression and
then assessed their behavior in the polysome-gradient microarray experiments. We did this by
identifying the 250, 500 and 1000 genes whose mRNAs showed the lowest fold-enrichment in
Smaug RIPs versus control RIPs. A comparison of the TI for each of these genes in wild-type
and smaug-mutant embryos shows a distribution with little bias towards an increase in TI in the
smaug mutant, confirming that few are likely to be targets of Smaug-mediated translational
repression (Figure 11A; Figure 12). In general, most genes not bound by Smaug have TI changes
below the median of the smaug mutant (see Figure 11B) where genes were ranked based on the
extent of the increase in TI in smaug mutant versus wild type, with the gene having the highest
48
increase being ranked number one). This trend is highly significant (e.g., 350 of the 500
‘unbound’ list are below the median and the distributions of the bottom 250, 500 and 1000 genes
are all significantly different from the distribution for all genes, Fisher’s exact test, P < 3x10-16
).
Finally, we performed a kernel density estimation of the change in TI for the genes whose
mRNAs fell into the top 250, 500 and 1000 Smaug-bound transcripts (i.e., those mRNAs with
the highest fold-enrichment in Smaug RIPs compared to control RIPs) as compared with the 250,
500 and 1000 genes whose mRNAs were unlikely to be bound by Smaug (i.e., with the lowest
fold-enrichment in Smaug RIPs versus control RIPs). This analysis showed a peak fold-change
of TI in smaug-mutant embryos versus wild type of 2.97, 2.80 and 2.80, respectively, for each of
the top three sets of bound transcripts but a fold-change of only 0.99, 1.07, and 1.09,
respectively, for each of these unbound sets (Figure 11C and Figure 13). The fact that transcripts
not bound by Smaug have no change in TI on average (i.e. the TI ratio is 1), suggests that our TI
estimates are directly comparable between the smaug-mutant and wild-type datasets. As such,
the distribution of TI changes for all genes is consistent with Smaug repressing the translation of
a large number of mRNAs in the early Drosophila embryo.
To estimate the actual number of genes that are translationally repressed by Smaug, we
deconvolved the distribution of TI changes for all genes (Figure 11 and Figure 13) to estimate
the relative contributions of genes whose TI changes are distributed according to the top N and
bottom N Smaug-binders (for N = 250, 500, and 1000), respectively. Based on this analysis, we
estimated that 3135, 3094, or 2728 are likely to be translationally repressed by Smaug using the
distributions for N=250, 500, or 1000, respectively (for details see Materials and Methods). We
conclude that Smaug represses the translation of approximately 3000 mRNAs in early embryos,
representing about half of the 5886 genes whose expression we detected in the polysome-
microarray data set.
49
Figure 11 Smaug represses the translation of thousands of mRNAs in the early embryo.
(A) The 500 bottom Smaug binders are the 500 genes whose mRNAs show the lowest fold
enrichment in Smaug RIPs versus control RIPs and they were plotted as in Figure 4A. The solid
diagonal lines represent no enrichment and the dotted diagonal lines represent 2-fold enrichment
or depletion. (B) Genes were ranked based on the extent of the increase in TI in smaug mutant
versus wild type, with the gene having the highest increase being ranked number one. Box plots
were then used to show the range of ranks for all genes, and the bottom 250, 500 and 1000
50
Smaug binders as defined in (A). (C) Kernel density plot shows the change in TI in smaug
mutant versus wild type for the bottom 500 Smaug binders as defined in (A) compared to the top
500 Smaug binders and all genes in the data set.
51
Figure 12 Comparison of the TIs in wild-type versus smaug-mutant embryos for mRNAs
that are unlikely to be bound by Smaug.
The 250 (A), 500 (B), and 1000 (C) bottom Smaug binders are the genes whose mRNAs show
the lowest fold-enrichment in Smaug RIPs versus control RIPs. The averages, across three
biological replicates, of the TI in smaug mutants and wild type for each of these groups of genes
were plotted against one another. The blue dots represent transcripts that show an increase in TI
in smaug mutants versus wild type at an FDR of <5%.
52
A
B
C
53
Figure 13 Kernel density plots comparing the change in TI in smaug-mutant versus wild-
type embryos for the top and bottom Smaug binders.
Kernel density plots show the change in TI in smaug-mutant versus wild-type embryos for the
top and bottom 250 (A), 500 (B), and 1000 (C) Smaug binders. These top and bottom binders are
the genes whose mRNAs show the highest and lowest fold enrichment in Smaug RIPs versus
control RIPs, respectively. Each graph also contains a plot of all genes in the data set and a light
blue vertical line that indicates the position of genes showing no change in their mRNAs’
distribution in polysome gradients in smaug-mutant compared to wild-type embryos.
54
2.3.5 SRE stem-loops are highly enriched in Smaug’s target mRNAs
Smaug binds to and regulates its target mRNAs through SRE stem-loop structures and, as such,
we would expect that mRNAs bound by Smaug as well as mRNAs translationally repressed by
Smaug would be enriched for these stem-loops. The consensus sequence for the SRE loop is
CNGGN0-3 (where N is any base) (Aviv et al., 2006; Aviv et al., 2003). The variability in the
number of nucleotides at the 3’-end of the loop derives from structural studies showing that
while the RNA-binding domain of the yeast Smaug homolog, Vts1p, interacts with the loop and
stem 5’ to the loop, it does not make contact with the 3’ region of the loop (Aviv et al., 2006;
Oberstrass et al., 2006). Thus, loop sequences where N is greater than 3 at this position are also
expected to be Smaug-binding sites.
To ask whether SREs are predictive of Smaug binding and translational repression we searched
all expressed genes in the RIP-Chip and polysome-microarray datasets for stem-loops with the
loop sequence CNGGN0-4 (see Materials and Methods for details). Our method assigned a
probability for each potential SRE within a transcript based on the likelihood that it would fold
into a stem-loop structure where the loop matches the CNGGN0-4 consensus. For each mRNA, an
SRE score was then calculated as the sum of the probabilities for each SRE within that mRNA
(Li et al., 2010). Strikingly, for the RIP-Chip experiment, bound mRNAs (FDR <5%) had a
median SRE score of 25.9 whereas unbound mRNAs (FDR >5%) had a 10-fold lower SRE score
(2.4). Likewise, for the polysome-microarray experiment, repressed mRNAs (FDR <5%) had a
median SRE score of 36.2 whereas unrepressed mRNAs (FDR >5%) had a median SRE score of
only 3.9. Within each of the regulated sets, however, the mRNAs nearer the top of the list (top 50
or top 100 as defined by fold-enrichment in Smaug RIPs versus control RIPs for binding or the
change in TI between smaug mutant and wild type for translational repression) did not have
higher SRE scores than the median for the bound or repressed mRNAs with FDR <5%.
Next, again using fold-enrichment and change in TI as metrics for binding and translational
repression, respectively, we employed multiple linear regression to simultaneously assess the
possible contributions of stem-loops carrying CNGGN0-4 loops along with six altered stem-loops.
The altered structures contain changes in the invariant nucleotides in the CNGGN0-4 loop that are
predicted to lower their affinity for the Smaug RNA-binding domain. We found that the bona
fide SRE was a significantly better predictor of both Smaug binding and Smaug-mediated
55
translational repression than any of the altered stem-loops (Figure 14A). These results are
consistent with positive correlations between the presence of sequences matching the SRE
consensus within mRNAs that are translationally repressed and/or degraded in wild-type
Drosophila embryos (Foat & Stormo, 2009).
We next used these data sets to explore the predictive power of other SRE features using the
same approach. We first tested SRE variants carrying different nucleotides in the N2 position of
the loop and found that CUGG performed better than CGGG, CAGG and CCGG loops, which
were similarly predictive of both Smaug binding and translational repression (Figure 14B).
These data are largely consistent with work suggesting that the yeast and human Smaug
homologs have binding preferences for SREs bearing CUGG and CGGG loops over CAGG and
CCGG (Li et al., 2010; Ray et al., 2013). We next tested the preference for the nucleotide
immediately 5’ to the loop and found that, while A, C and U performed similarly, G performed
better (Figure 14C). This result is consistent with the binding specificity determined for the yeast
and human Smaug homologs (Kazan et al., 2010; Ray et al., 2009; Ray et al., 2013; Riordan et
al., 2011). Finally, we tested the effect of varying the SRE loop size and found that loops of five
nucleotides performed best of all, with a gradual decrease in the predictive value of shorter or
longer loops (Figure 14D).
56
Figure 14 SREs are enriched in Smaug-bound mRNAs and those that are translationally
repressed by Smaug.
Multiple linear regression was used to simultaneously assess the contribution of various stem-
loop structures to Smaug binding and Smaug-mediated translational repression. Smaug binding
was quantified using the fold enrichment in Smaug RIPs compared to control RIPs and Smaug-
mediated translational repression was quantified by comparing the TI in smaug mutant versus
wild-type embryos. The structures tested in (A) include a consensus SRE with the loop sequence
CNGGN0-4 while the other sequences tested carry the indicated changes in invariant positions of
the loop that are predicted to reduce or block Smaug binding. The structures tested in (B)
included all possible nucleotides in the second position of the loop. The structures tested in (C)
57
2.3.6 Smaug co-regulates translational repression and degradation of a large fraction of its target mRNAs
Smaug employs different mechanisms to regulate the expression of its two characterized target
mRNAs, nos and Hsp83 (Dahanukar et al., 1999; Semotok et al., 2005; Semotok & Lipshitz,
2007; Semotok et al., 2008; Smibert et al., 1999; Smibert et al., 1996). To gain a panoramic view
of how Smaug regulates its target transcripts we compared the data for Smaug binding and
translational repression from the current study to the data from our previous, genome-wide
analyses of Smaug-induced transcript decay (Tadros et al., 2007). For the first set of comparisons
the fold-enrichment of an mRNA in Smaug RIPs versus control RIPs was used as a metric for
Smaug binding and the change in TI between smaug mutant and wild type was used as a metric
for translational regulation. We found that mRNAs requiring Smaug for their degradation
showed significantly higher levels of both Smaug binding (Figure 15A) (Wilcoxon rank-sum
test, P < 3x10-16
) and Smaug-mediated translational repression (Figure 15B) (Wilcoxon rank-
sum test, P < 3x10-16
) than mRNAs whose decay is not regulated by Smaug. Using these two
measures we also found a genome-wide correlation between Smaug binding and Smaug-
mediated translational repression (Spearman’s rho = 0.43, Fisher’s exact test P < 3x10-16
, Figure
15C).
We then compared the lists of genes whose mRNAs are bound by Smaug to those that are
degraded or translationally repressed by Smaug (Figure 15D). As described above, our data
suggest that several thousand mRNAs are translationally repressed by Smaug and that the
calculated FDR overestimates the true FDR (Storey & Tibshirani, 2003). Thus, for all
comparisons involving polysome data we used a list of genes whose mRNAs show an increase in
TI in smaug mutant embryos versus wild type at an FDR <10% rather than at <5%. This cut-off,
often used in place of 5%, is near an inflection point in the plot of gene number versus FDR
(Figure 16), indicating that there is a much higher, and fairly consistent, enrichment for true
positives up until that point.
included all possible nucleotides in the position that immediately proceeds the loop. The
structures tested in (D) included loops matching the CNGGN0-4 consensus where the loop size
varied from 4-8 nucleotides.
58
We found that at least 67% of the mRNAs bound by Smaug are targets of Smaug-mediated
decay, while at least 74% of the mRNAs bound by Smaug are translationally repressed by
Smaug (Figure 15D). We also found a substantial and significant overlap between the lists of
genes that encode mRNAs that are translationally repressed by Smaug and those that require
Smaug for their degradation (i.e., 71% of the mRNAs that are degraded by Smaug are also
translationally repressed by Smaug while 46% of mRNAs that are translationally repressed by
Smaug are targets of Smaug-mediated mRNA decay; Figure 15D). A comparison of all three
data sets can be viewed in Figure 17. Taken together, these data indicate that a large fraction of
Smaug’s targets are both translationally repressed and degraded by Smaug.
The comparisons from Figure 15D identified a substantial number of genes that require Smaug
for their degradation or translational repression but do not appear to be bound by Smaug. These
transcripts may require Smaug indirectly for their regulation or they may represent false
negatives from the RIP-Chip experiments. To assess the latter possibility, we grouped mRNAs
into four different classes where Smaug binders were defined as having an FDR in RIP-Chip of
<5% and the targets of Smaug-mediated decay are based on the results of Tadros et al., (Tadros
& Lipshitz, 2009). The four classes were: 1) those mRNAs that were bound by Smaug and
required Smaug for their degradation (‘bound+degraded’, Figure 18A); 2) those that were neither
bound nor degraded by Smaug (‘unbound+not degraded’); 3) those that were bound by Smaug
and did not require Smaug for their degradation (‘only bound’); and 4) those that were not bound
by Smaug and did require Smaug for their degradation (‘only degraded’). We then assessed the
SRE scores for the mRNAs in each of these groups and found a substantially higher SRE
enrichment for the mRNAs in the ‘only degraded’ class compared to the ‘unbound+not
degraded’ class (Figure 18A) (Wilcoxon rank-sum test, P < 3x10-16
). Similar results were
obtained for Smaug-mediated translational repression (i.e. a significantly higher SRE enrichment
for the ‘only repressed’ class of mRNAs compared to the ‘unbound+not repressed’ class of
mRNAs (Figure 18B) (Wilcoxon rank-sum test, P < 3x10-16
). Together these data suggest that a
large fraction of the mRNAs that require Smaug for their degradation and/or translational
repression that were scored as unbound in the RIP-Chip experiments, are nonetheless directly
bound by Smaug.
The nos mRNA’s SREs are found in the 3’UTR (Dahanukar et al., 1999; Smibert et al., 1999;
Smibert et al., 1996) and the Hsp83 mRNA’s SREs are found in the open reading frame
59
(Semotok et al., 2005; Semotok et al., 2008), raising the possibility that the differential
regulation of these transcripts relates to SRE position. To assess this possibility we compared the
SRE scores for the 5’UTR, open reading frame and 3’UTR of genes that encode mRNAs that are
translationally repressed but not degraded by Smaug, degraded by Smaug but not translationally
repressed, and both repressed and degraded by Smaug (Figure 19). These results indicate that the
vast majority of SREs are localized within target transcripts’ open reading frames and that SRE
location within target mRNAs does not explain their differential regulation by Smaug.
60
Figure 15 Comparisons of Smaug-bound, repressed and degraded data sets.
(A) Smaug binding was assessed using the fold enrichment in Smaug RIPs compared to control
RIPs and box plots show the range of these enrichments for the targets of Smaug-mediated
mRNA decay and for non-targets. (B) Smaug-mediated translational repression was assessed
using the change in TI in smaug mutant compared to wild-type and box plots show the range of
these changes for the targets of Smaug-mediated mRNA decay and for non-targets. (C) Smaug
binding and translational repression were quantitated as described in (A) and (B), respectively,
and these values were plotted against one another. The dashed vertical and horizontal lines
represent the median values for Smaug binding and Smaug-mediated translational repression,
respectively. (D) Venn diagrams show the overlap between the genes whose mRNAs are bound
by Smaug, those that are degraded by Smaug and those that are translationally repressed by
Smaug (FDR<10%). Note that for each comparison only genes scored as expressed in both data
sets are included in the comparison.
61
Figure 16 FDR-based rank of genes from polysome gradient microarrays.
The FDR of each gene based on its change in TI in smaug-mutant versus wild-type embryos was
used to rank genes, with the lowest ranking gene having an FDR of 0%. Only genes with an
FDR<25% were used in this plot.
62
Figure 17 Overlaps between Smaug-bound genes and Smaug- regulated genes.
Areas of the overlapping regions are approximately proportional to the degree of overlap of the
indicated data sets. Note that for this comparison any gene that was not scored as expressed in all
three data sets is not included.
63
Figure 18 Smaug degraded and Smaug repressed mRNAs are enriched for SREs.
(A) Genes were divided into one of four classes (see the main text for more details): 1) bound and
degraded by Smaug; 2) neither bound nor degraded by Smaug; 3) only bound by Smaug; and 4)
only degraded by Smaug. The range of SRE scores for these classes are shown in the box plots.
(B) Genes were divided into one of four classes (see the main text for more details): 1) bound and
translationally repressed by Smaug; 2) neither bound nor translationally repressed by Smaug; 3)
only bound by Smaug; and 4) only translationally repressed by Smaug. The range of SRE scores
for these classes are shown in the box plots. “All mRNAs” shows the range of SREs scores for all
of the transcripts represented in (A) or (B), respectively.
64
Figure 19 SRE scores for the 5’UTR, open reading frame and 3’UTR of Smaug-regulated
mRNAs.
Cumulative distribution plots show the SRE scores for the indicated region of the mRNA for
Smaug-target transcripts that are translationally repressed and degraded by Smaug, those that are
just translationally repressed, those that are just degraded, and those that are neither repressed
nor degraded by Smaug. (B) Shows the results when SRE scores are divided by the length of the
relevant region of the transcripts in nucleotides while (A) does not take into account feature
length.
65
2.3.7 Subcellular localization of Smaug’s target mRNAs
Given Smaug’s role in controlling the expression of localized mRNAs, we analyzed the list of
Smaug-bound mRNAs for subcellular localization patterns reported by the Fly-FISH database
(http://flyfish.ccbr.utoronto.ca/) (Lecuyer et al., 2007). We searched for enrichment of the Fly-
FISH database categories defined in embryonic stages 1-3 and 4-5, representing the stages from
which the Smaug-regulated mRNAs were identified (
Table 5 and Figure 20). The Fly-FISH database not only categorizes sub-cellular localization
patterns but also reports whether an mRNA is degraded. Consistent with Smaug’s role in
transcript degradation, Smaug-bound mRNAs were enriched for the Fly-FISH category
‘degraded’. Additional highly enriched categories were those that describe mRNAs that are
localized to the posterior of the embryo (e.g. ‘posterior localization’, ‘pole cell enrichment’ and
‘pole cell localization’). Together the Smaug-bound mRNAs that fall into these categories
produce a collection of 44 genes, including nos and Hsp83, whose mRNAs are localized to the
posterior. Of these 44 genes, 38 are regulated by Smaug at the level of mRNA stability and/or
translation (
66
Table 6).
2.3.8 Functional analysis of Smaug-regulated mRNAs
To gain insights into Smaug’s biological functions in early embryos I searched the list of Smaug-
bound mRNAs for encoded proteins with functions related to known aspects of the smaug
mutant phenotype. Embryos that lack Smaug show defects in the cell cycle that are associated
with a failure in DNA replication checkpoint activation (Benoit et al., 2009; Dahanukar et al.,
1999) suggesting that Smaug might regulate the expression of genes involved in these processes.
Thus, I searched the list of Smaug-bound mRNAs for genes that are annotated to play roles in the
cell cycle, checkpoint response and/or response to DNA damage. I found a total of 32 such genes
and enrichment for the GO term ‘cellular response to DNA damage’. This list of genes included
cdc2c, mitotic 15 (mit(1)15), Replication Protein A 70 (RpA-70), Regulator of cyclin A1 (Rca1),
Cyclin E (CycE), Minichromosome maintenance 3 (Mcm3), CDC45L, mutagen-sensitive 201
(mus201) and Msh6. Of these 32 genes, 29 are regulated by Smaug at the level of mRNA
stability and/or translation (Table 7).
Smaug also plays a prominent role in activating the transcription of the zygotic genome in the
early embryo (Benoit et al., 2009). I thus searched the list of Smaug-bound mRNAs for genes
that are annotated to have roles in transcription and/or chromatin and found a total of 25 genes
including dre4, Polycomblike (Pcl), Nucleosome assembly protein 1 (Nap1), Nucleosome
remodeling factor - 38kD (Nurf-38), anti-silencing factor 1 (asf1), Caf1-180, Caf1-105, and vig2.
Of these 25 genes, 24 are regulated by Smaug at the level of mRNA stability and/or translation
(Table 8).
I also searched for novel functions of Smaug by analyzing the Smaug-bound mRNAs via gene
set annotation enrichment analysis using the DAVID annotation tool (Huang et al., 2009)
applying two stringencies to the analysis: the standard DAVID FDR cut-off of <10% and the
more stringent Benjamini-Hochberg FDR (P-value of <0.1). These analyses suggest several
previously unrecognized roles for Smaug in the early embryo (Table 9).
First, Smaug may play a role in regulation of protein folding. For example, Smaug-bound
mRNAs encode five proteins (Hsp60, T-cp1ζ, CG5525, CG8258 and CG7033) that are members
of the Chaperonin Cpn60/TCP-1 family as defined by the Interpro database and are involved in
67
protein folding. The last four of these proteins are subunits of the eukaryotic chaperonin TCP1-
ring complex (TRiC), also known as the chaperonin containing TCP-1 (CCT), which consists of
two rings comprised of eight different subunits (Hartl et al., 2011). Consistent with a role for
Smaug in regulating protein folding, all five of these genes are regulated by Smaug at the level of
translation and/or mRNA stability. (Table 10).
Second, Smaug-associated mRNAs are enriched for the related GO terms ‘proteasome regulatory
particle’ and ‘proteasome complex’ as well as the Protein Analysis Through Evolutionary
Relationships (Panther) term ‘Ubiquitin proteasome pathway’. The ubiquitin proteasome system
plays a vital part in a variety of cellular processes through its role in the degradation of target
proteins. This mechanism involves the post-translational addition of multiple ubiquitin moieties
onto a protein, which, in turn, target the protein for proteasomal degradation (Saeki & Tanaka,
2012). The 26S proteasome consists of a 20S core particle, which carries the proteasome’s
proteolytic activity, and either one or two 19S regulatory particles, which are necessary for
proteasome activity and are composed of 19 subunits (Saeki & Tanaka, 2012). Strikingly, Smaug
associates with nine of the mRNAs that encode the regulatory subunits (Regulatory particle
triple-A ATPase 3 (Rpt3), Regulatory particle triple-A ATPase 5 (Rpt5), Regulatory particle non-
ATPase 1 (Rpn1), Regulatory particle non-ATPase 2 (Rpn2), Regulatory particle non-ATPase 7
(Rpn7), Regulatory particle non-ATPase 9 (Rpn9), Regulatory particle non-ATPase 10 (Rpn10),
Regulatory particle non-ATPase 11 (Rpn11) and Regulatory particle non-ATPase 13 (Rpn13)).
In contrast, Smaug does not interact with any of the mRNAs that encode the 20S core particle
proteins. In addition, Smaug interacts with mRNAs that encode proteins involved in other
aspects of the ubiquitin-proteasome system (Ubiquitin activating enzyme 1 (Uba1), Ubiquitin
fusion-degradation 1-like (Ufd1-like), TER94 and CG9588). Consistent with a role for Smaug in
control of the ubiquitin-proteasome system, 12 out of these 13 mRNAs (Table 12), including all
of the transcripts that encode regulatory subunit proteins, are regulated by Smaug at the level of
translation and/or mRNA stability.
Third, Smaug might play a role in regulating lipid storage and/or mobilization since the GO term
‘lipid droplet’ is enriched in the Smaug-bound mRNAs. Lipid droplets are ubiquitous organelles
that are found in a wide range of organisms from bacteria to humans. They consist of a neutral-
lipid core composed of triacylglycerols and sterol esters surrounded by a phospholipid
monolayer, and they serve as storage sites for energy, sterols and membrane precursors (Farese
68
& Walther, 2009). Several studies have used proteomic approaches to identify lipid-droplet
associated proteins, including two studies that purified lipid droplets from Drosophila fat-body
tissue or from Drosophila embryos (Beller et al., 2006; Cermelli et al., 2006). Comparison of
those lists with our data identified 33 Smaug-bound mRNAs that encode lipid-droplet associated
proteins. In addition, our data indicate that 29 of these 33 mRNAs are destabilized and/or
translationally repressed by Smaug (Table 14). Taken together these data suggest that Smaug
may control aspects of lipid droplet function through its regulation of these mRNAs.
Fourth, a direct role for Smaug in regulation of metabolism is suggested by the enrichment for
terms such as the SwissProt keywords ‘oxidoreductase’ and ‘NAD’ and the GO terms ‘oxidation
reduction’ and ‘cofactor binding’ within Smaug-bound mRNAs. Together these lists comprise a
total of 37 metabolic enzymes that function in a wide variety of pathways including fatty acid
metabolism, pyruvate metabolism, amino acid metabolism, the citric acid cycle and oxidative
phosphorylation. Our data suggest that 28 out of 37 of these genes are regulated by Smaug at the
level of mRNA stability and/or translation (Table 15). In addition, I found enrichment for the GO
term ‘glucose metabolic process’ and the Kyoto Encyclopedia of Genes and Genomes (KEGG)
pathway ‘Glycolysis/Gluconeogenesis’. These lists contain nine genes including four enzymes of
the glycolytic pathway (Hexokinase A (Hex-A), Phosphoglycerate kinase (Pgk), Phosphoglucose
isomerase (Pgi) and both genes encoding Glyceraldehyde 3 phosphate dehydrogenase (GAPDH1
and GAPDH2)) and our data indicate that all nine of these genes are regulated by Smaug at the
level of stability and/or translation repression (Table 16). Furthermore, our data suggest that
mRNAs encoding four additional glycolytic enzymes may be regulated by Smaug.
Phosphofructokinase (Pfk )and Triose phosphate isomerase (Tpi), have FDRs in the RIP-Chip
data of 5.15% and 6.08%, respectively and both are targets of Smaug-mediated transcript
degradation and translational repression (Table 17). Also, Enolase (Eno) and Pyruvate kinase
(Pyk) are regulated by Smaug at the level of stability and/or translation. In summary, our data
suggest that 8 of the 10 glycolytic enzymes may be regulated by Smaug.
69
Table 5 Fly-FISH localization patterns and degradation categories enriched among Smaug-
bound mRNAs
Embryonic stages Localization or degradation category FDR (%)
Stages 1-3 Degraded 0.0585
Degraded partially stage 3 0.1553
Degradation maternal transcripts 0.1553
Posterior localization 3.9050
Stages 4-5 Pole cell localization 0.0002
Degraded 0.0032
Pole cell enrichment 0.0032
Posterior localization 0.0118
Degraded partially stage 4 0.3571
Degradation maternal transcripts 1.1039
Degraded partially stage 5 2.4097
Yolk plasm localization 5.3805
70
Figure 20 Fly-FISH degradation categories and localization patterns enriched among
Smaug-bound mRNAs.
A search of the Fly-FISH database was performed to identify degradation categories (A) and
localization patterns (B) that are enriched among Smaug-bound mRNAs for embryonic stages 1–
3 and 4–5. For both (A) and (B), categories in closed boxes were tested for enrichment, and the
blue shading indicates the degradation categories and localization patterns that were enriched
among the Smaug-bound mRNAs. Detailed results for highlighted categories are presented in
Table 5 .
71
Table 6 Smaug-bound mRNAs that are localized to the posterior of the embryo
Flybase number
Annotation
symbol Name Symbol
Smaug
mRNA
decay
target
Smaug
translational
repression
target
(FDR<10%)
FBgn0035720 CG10077 - CG10077 no yes
FBgn0260634 CG10192
eukaryotic
translation initiation
factor 4G2 eIF4G2 N/A yes
FBgn0028691 CG10230
Regulatory particle
non-ATPase 9 Rpn9 yes yes
FBgn0027512 CG10254 - CG10254 yes yes
FBgn0028684 CG10370
Regulatory particle
triple-A ATPase 5 Rpt5 yes yes
FBgn0039929 CG11076 - CG11076 yes yes
FBgn0037207 CG11100
Mesoderm-
expressed 2 Mes2 yes yes
FBgn0004913 CG1119
Germ line
transcription factor
1 Gnf1 yes yes
FBgn0046214 CG11844 vig2 vig2 yes yes
FBgn0005630 CG12052
longitudinals
lacking lola no yes
FBgn0001091 CG12055
Glyceraldehyde 3
phosphate
dehydrogenase 1 Gapdh1 yes yes
FBgn0001233 CG1242
Heat shock protein
83 Hsp83 yes yes
FBgn0011692 CG1258 pavarotti pav yes No
FBgn0037810 CG12819 slender lobes sle yes yes
FBgn0023215 CG13316 Mnt Mnt yes yes
FBgn0263594 CG14648 lost lost N/A N/A
FBgn0037728 CG16817 - CG16817 yes yes
FBgn0262735 CG1691
IGF-II mRNA-
binding protein Imp N/A No
FBgn0002183 CG1828 dre4 dre4 yes yes
FBgn0020622 CG2699 Pi3K21B Pi3K21B yes N/A
FBgn0034753 CG2852 - CG2852 yes yes
FBgn0001186 CG3001 Hexokinase A Hex-A no yes
FBgn0000114 CG31762 arrest aret yes yes
FBgn0020513 CG3989 ade5 ade5 no No
FBgn0024332 CG4206
Minichromosome
maintenance 3 Mcm3 yes N/A
FBgn0035978 CG4347 UGP UGP no yes
FBgn0263755 CG43664
Suppressor of
variegation 3-9 Su(var)3-9 N/A N/A
FBgn0264785 CG44015
HIF prolyl
hydroxylase Hph N/A N/A
FBgn0004868 CG4422 GDP dissociation Gdi yes yes
72
inhibitor
FBgn0029687 CG5014 Vap-33-1 Vap-33-1 no yes
FBgn0015268 CG5330
Nucleosome
assembly protein 1 Nap1 yes yes
FBgn0002962 CG5637 nanos nos N/A yes
FBgn0086687 CG5887 desat1 desat1 yes yes
FBgn0027615 CG6404 - CG6404 yes yes
FBgn0004363 CG6647 porin porin yes N/A
FBgn0032400 CG6770 - CG6770 N/A yes
FBgn0036486 CG7003 Msh6 Msh6 yes yes
FBgn0003041 CG8114 pebble pbl yes yes
FBgn0027329 CG8231 T-cp1zeta Tcp-1zeta yes N/A
FBgn0000556 CG8280
Elongation factor
1alpha48D Ef1alpha48D no yes
FBgn0037756 CG8507 - CG8507 no No
FBgn0000615 CG8994 exuperantia exu no yes
FBgn0031769 CG9135 - CG9135 yes yes
FBgn0011826 CG9842
Protein phosphatase
2B at 14D Pp2B-14D N/A yes
FBgn0033886 CG13349
Regulatory particle
non-ATPase 13 Rpn13 yes yes
73
Table 7 Smaug-bound mRNAs annotated with roles in cell cycle, checkpoint response
and/or response to DNA damage
Flybase number
Annotation
symbol Name Symbol
Smaug
mRNA
decay target
Smaug translational
repression target
(FDR<10%)
FBgn0000541 CG32346 Enhancer of bithorax E(bx) yes Yes
FBgn0000826 CG11420 pan gu png no Yes
FBgn0001120 CG5272 giant nuclei gnu yes No
FBgn0002622 CG6779 Ribosomal protein S3 RpS3 no N/A
FBgn0004107 CG10498 cdc2c cdc2c yes No
FBgn0004432 CG9916 Cyclophilin 1 Cyp1 yes Yes
FBgn0010173 CG9633
Replication Protein A
70 RpA-70 yes Yes
FBgn0010382 CG3938 Cyclin E CycE yes No
FBgn0010431 CG18543 matrimony mtrm yes Yes
FBgn0011020 CG10061 Sas-4 Sas-4 yes Yes
FBgn0011739 CG12072 warts wts N/A Yes
FBgn0017551 CG10800
Regulator of cyclin
A1 Rca1 no Yes
FBgn0020653 CG2151
Thioredoxin
reductase-1 Trxr-1 yes Yes
FBgn0023215 CG13316 Mnt Mnt yes Yes
FBgn0024332 CG4206
Minichromosome
maintenance 3 Mcm3 yes N/A
FBgn0026143 CG3658 CDC45L CDC45L yes Yes
FBgn0026430 CG3917
gamma-tubulin ring
protein 84 Grip84 yes Yes
FBgn0031006 CG8002
rapamycin-insensitive
companion of Tor rictor N/A Yes
FBgn0035334 CG8993 - CG8993 N/A Yes
FBgn0035631 CG5495 Thioredoxin-like Txl N/A no
FBgn0036486 CG7003 Msh6 Msh6 yes yes
FBgn0040070 CG31884 thioredoxin-2 Trx-2 yes yes
FBgn0041781 CG4636 SCAR SCAR yes yes
FBgn0042134 CG18811 Caprin Capr no yes
FBgn0052251 CG32251 Claspin Claspin yes yes
FBgn0053193 CG33193 salvador sav yes yes
FBgn0061515 CG6513 endosulfine endos yes yes
FBgn0086695 CG2669 humpty dumpty hd no yes
FBgn0260634 CG10192
eukaryotic translation
initiation factor 4G2 eIF4G2 no yes
FBgn0262699 CG3000
retina aberrant in
pattern rap N/A N/A
FBgn0004643 CG9900 mitotic 15 mit(1)15 yes yes
FBgn0002887 CG10890 mutagen-sensitive 201 mus201 N/A yes
74
Table 8 Smaug-bound mRNAs annotated with roles in transcription and/or chromatin
Flybase number
Annotation
symbol Name Symbol
Smaug
mRNA
decay target
Smaug translational
repression target
(FDR<10%)
FBgn0000413 CG5102 daughterless da N/A yes
FBgn0000541 CG32346 Enhancer of bithorax E(bx) yes yes
FBgn0000927 CG2707
female sterile (1)
Young arrest fs(1)Ya yes yes
FBgn0002183 CG1828 dre4 dre4 yes yes
FBgn0002781 CG32491 modifier of mdg4 mod(mdg4) no N/A
FBgn0002962 CG5637 nanos nos N/A yes
FBgn0003044 CG5109 Polycomblike Pcl yes yes
FBgn0004432 CG9916 Cyclophilin 1 Cyp1 yes yes
FBgn0004913 CG1119
Germ line
transcription factor 1 Gnf1 yes yes
FBgn0005630 CG12052 longitudinals lacking lola no yes
FBgn0015268 CG5330
Nucleosome
assembly protein 1 Nap1 yes yes
FBgn0015299 CG8396
Single stranded-
binding protein c31A Ssb-c31a N/A yes
FBgn0016687 CG4634
Nucleosome
remodeling factor -
38kD Nurf-38 yes yes
FBgn0020616 CG3423 Stromalin SA yes yes
FBgn0023215 CG13316 Mnt Mnt yes yes
FBgn0024332 CG4206
Minichromosome
maintenance 3 Mcm3 yes N/A
FBgn0026143 CG3658 CDC45L CDC45L yes yes
FBgn0029094 CG9383
anti-silencing factor
1 asf1 yes yes
FBgn0030054 CG12109 Caf1-180 Caf1-180 no yes
FBgn0030990 CG7556 - CG7556 N/A yes
FBgn0033526 CG12892 Caf1-105 Caf1-105 yes yes
FBgn0037760 CG9461 FBX011 ortholog FBX011 yes yes
FBgn0040477 CG13329 centromere identifier cid no yes
FBgn0046214 CG11844 vig2 vig2 yes yes
FBgn0086757 CG4840
centrosomin's
beautiful sister cbs no yes
75
Table 9 Gene set annotation enrichment analysis results for Smaug-bound mRNAs
76
Table 10 Smaug-bound mRNAs encode proteins in the Interpro Chaperonin Cpn60/TCP-1
family
Flybase number
Annotation
symbol Name Symbol
Smaug
mRNA
decay
target
Smaug
translational
repression target
(FDR<10%)
FBgn0015245 CG12101 Heat shock protein 60 Hsp60 yes no
FBgn0033342* CG8258 - CG8258 yes yes
FBgn0032444* CG5525 - CG5525 no yes
FBgn0027329* CG8231 T-cp1ζ Tcp-1ζ yes no
FBgn0030086* CG7033 - CG7033 yes yes
*TRiC/CCT complex components
77
Table 11 Status of other components of the TRiC/CCT complex
Flybase
number
Annotation
symbol Name Symbol
% FDR
in
RIP/Chip
Smaug
mRNA
decay
target
Smaug
translational
repression
target
(FDR<10%)
FBgn0003676 CG5374 Tcp1-like T-cp1 13.2 no yes
FBgn0010621 CG8439 T-complex Chaperonin 5 Cct5 31.5 no N/A
FBgn0037632 CG8351 Tcp-1η Tcp-1η 31 yes yes
FBgn0015019 CG8977 Cctγ Cctγ 52.4 yes no
78
Table 12 Smaug-bound mRNAs that encode proteins found in the proteasome regulatory
particle and the ubiquitin proteasome pathway
Flybase number
Annotation
symbol Name Symbol
Smaug
mRNA
decay
target
Smaug
translational
repression
target
(FDR<10%)
FBgn0028686* CG16916 Regulatory particle triple-A ATPase 3 Rpt3 yes no
FBgn0028684* CG10370 Regulatory particle triple-A ATPase 5 Rpt5 yes yes
FBgn0028695* CG7762 Regulatory particle non-ATPase 1 Rpn1 yes yes
FBgn0028692* CG11888 Regulatory particle non-ATPase 2 Rpn2 yes no
FBgn0028688* CG5378 Regulatory particle non-ATPase 7 Rpn7 yes yes
FBgn0028691* CG10230 Regulatory particle non-ATPase 9 Rpn9 yes yes
FBgn0015283* CG7619 Regulatory particle non-ATPase 10 Rpn10 yes yes
FBgn0028694* CG18174 Regulatory particle non-ATPase 11 Rpn11 yes yes
FBgn0033886* CG13349 Regulatory particle non-ATPase 13 Rpn13 yes yes
FBgn0038166 CG9588 - CG9588 N/A no
FBgn0036136 CG6233 Ubiquitin fusion-degradation 1-like Ufd1-like yes yes
FBgn0261014 CG2331 TER94 TER94 N/A yes
FBgn0023143 CG1782 Ubiquitin activating enzyme 1 Uba1 yes yes
*proteasome regulatory particle components
79
Table 13 Status of other components of the proteasome regulatory particle
Flybase
number
Annotation
symbol Name Symbol
% FDR
in
RIP/Chip
Smaug
mRNA
decay
target
Smaug
translational
repression
target
(FDR<10%)
FBgn0028687 CG1341
Regulatory particle triple-A
ATPase 1 Rpt1 NaN yes yes
FBgn0015282 CG5289
Regulatory particle triple-A
ATPase 2 Rpt2 N/A yes no
FBgn0028685 CG3455
Regulatory particle triple-A
ATPase 4 Rpt4 N/A no N/A
FBgn0020369 CG1489
Regulatory particle triple-A
ATPase 6 Rpt6 N/A yes yes
FBgn0261396 CG42641
Regulatory particle non-
ATPase 3 Rpn3 N/A N/A N/A
FBgn0028690 CG1100
Regulatory particle non-
ATPase 5 Rpn5 71.4 yes yes
FBgn0028689 CG10149
Regulatory particle non-
ATPase 6 Rpn6 35.6 yes no
FBgn0002787 CG3416
Regulatory particle non-
ATPase 8 Rpn8 N/A yes yes
FBgn0028693 CG4157
Regulatory particle non-
ATPase 12 Rpn12 -64.2 yes no
FBgn0040954 CG13779 - CG13779 N/A yes no
80
Table 14 Smaug-bound mRNAs that encode proteins associated with lipid droplets
Flybase number
Annotation
symbol Name Symbol
Smaug
mRNA
decay target
Smaug
translational
repression target
(FDR<10%)
FBgn0250814 CG4169 - CG4169 no no
FBgn0012036 CG3752 Aldehyde dehydrogenase Aldh yes yes
FBgn0039635 CG11876 - CG11876 N/A yes
FBgn0001091 CG12055
Glyceraldehyde 3
phosphate dehydrogenase 1 Gapdh1 yes yes
FBgn0001092 CG8893
Glyceraldehyde 3
phosphate dehydrogenase 2 Gapdh2 yes yes
FBgn0001233 CG1242 Heat shock protein 83 Hsp83 yes yes
FBgn0038271 CG3731 - CG3731 yes no
FBgn0038947 CG7073
Sar1 ortholog (S.
cerevisiae) Sar1 yes yes
FBgn0021795 CG9035
Translocon-associated
protein delta Tapdelta yes yes
FBgn0053129 CG33129 - CG33129 yes yes
FBgn0023529 CG2918 - CG2918 yes yes
FBgn0043884 CG33106
multiple ankyrin repeats
single KH domain mask no yes
FBgn0010217 CG11154 ATP synthase-beta ATPsyn-beta yes yes
FBgn0000556 CG8280
Elongation factor
1alpha48D Ef1alpha48D no yes
FBgn0086687 CG5887 desat1 desat1 yes yes
FBgn0032444 CG5525 - CG5525 no yes
FBgn0002622 CG6779 Ribosomal protein S3 RpS3 no no
FBgn0003074 CG8251 Phosphoglucose isomerase Pgi yes yes
FBgn0030086 CG7033 - CG7033 yes yes
FBgn0250848 CG8947 26-29kD-proteinase 26-29-p no no
FBgn0000055 CG3481 Alcohol dehydrogenase Adh N/A yes
FBgn0001125 CG4233
Glutamate oxaloacetate
transaminase 2 Got2 no N/A
FBgn0029687 CG5014 Vap-33-1 Vap-33-1 no yes
FBgn0027580 CG1516 - CG1516 yes no
FBgn0027291 CG12233 lethal (1) G0156 l(1)G0156 yes no
FBgn0015795 CG5915 Rab7 Rab7 yes yes
FBgn0027329 CG8231 T-cp1zeta Tcp-1zeta yes no
FBgn0034753 CG2852 - CG2852 yes yes
FBgn0042134 CG18811 Caprin Capr no yes
FBgn0011211 CG3612 bellwether blw N/A yes
FBgn0015778 CG9412 rasputin rin yes N/A
FBgn0015245 CG12101 Heat shock protein 60 Hsp60 yes no
FBgn0004363 CG6647 porin porin yes no
81
Table 15 Smaug-bound mRNAs that encode metabolic enzymes
Flybase number
Annotation
symbol Name Symbol
Smaug
mRNA
decay target
Smaug
translational
repression
target
(FDR<10%)
FBgn0020653 CG2151
Thioredoxin reductase-
1 Trxr-1 yes yes
FBgn0250814 CG4169 - CG4169 no no
FBgn0012036 CG3752
Aldehyde
dehydrogenase Aldh yes yes
FBgn0039635 CG11876 - CG11876 N/A yes
FBgn0001091 CG12055
Glyceraldehyde 3
phosphate
dehydrogenase 1 Gapdh1 yes yes
FBgn0001092 CG8893
Glyceraldehyde 3
phosphate
dehydrogenase 2 Gapdh2 yes yes
FBgn0027597 CG17712 - CG17712 no yes
FBgn0029648 CG3603 - CG3603 N/A no
FBgn0036290 CG10638 - CG10638 N/A yes
FBgn0038271 CG3731 - CG3731 yes no
FBgn0040529 CG9603 - CG9603 yes no
FBgn0086687 CG5887 desat1 desat1 yes yes
FBgn0031824 CG9547 - CG9547 N/A yes
FBgn0004087 CG14887
Dihydrofolate
reductase Dhfr N/A no
FBgn0264785 CG44015
HIF prolyl
hydroxylase Hph yes N/A
FBgn0033979 CG10243 Cyp6a19 Cyp6a19 yes yes
FBgn0032910 CG9265 - CG9265 yes yes
FBgn0036824 CG3902 - CG3902 no no
FBgn0086254 CG6084 - CG6084 yes yes
FBgn0000055 CG3481
Alcohol
dehydrogenase Adh N/A yes
FBgn0038742 CG4703 Arc42 Arc42 no yes
FBgn0051548 CG31548 - CG31548 N/A yes
FBgn0027291 CG12233 lethal (1) G0156 l(1)G0156 yes no
FBgn0035911 CG6638 - CG6638 no yes
FBgn0030758 CG9819 Calcineurin A at 14F CanA-14F no no
FBgn0000565 CG7266
Ecdysone-induced
protein 28/29kD Eip71CD no no
FBgn0004654 CG3724
Phosphogluconate
dehydrogenase Pgd yes yes
FBgn0030305 CG1749 - CG1749 N/A yes
FBgn0003204 CG1799 raspberry ras yes yes
FBgn0034245 CG14482 - CG14482 N/A no
FBgn0039358 CG5028 - CG5028 N/A yes
82
FBgn0023537 CG17896 - CG17896 N/A yes
FBgn0040070 CG31884 thioredoxin-2 Trx-2 yes yes
FBgn0001125 CG4233
Glutamate
oxaloacetate
transaminase 2 Got2 no N/A
FBgn0034488 CG11208 - CG11208 yes no
FBgn0032393 CG12264 - CG12264 no yes
FBgn0025885 CG11143 Inos Inos no no
83
Table 16 Smaug-bound mRNAs that encode metabolic enzymes involved in glycolysis and
related pathways
Flybase number
Annotation
symbol Name Symbol
Smaug
mRNA
decay
target
Smaug
translational
repression
target
(FDR<10%)
FBgn0001186* CG3001 Hexokinase A Hex-A no yes
FBgn0003074* CG8251 Phosphoglucose isomerase Pgi yes yes
FBgn0001091* CG12055
Glyceraldehyde 3 phosphate
dehydrogenase 1 Gapdh1 yes yes
FBgn0001092* CG8893
Glyceraldehyde 3 phosphate
dehydrogenase 2 Gapdh2 yes yes
FBgn0250906* CG3127 Phosphoglycerate kinase Pgk yes no
FBgn0023477 CG2827 Tal Tal no yes
FBgn0004654 CG3724 Phosphogluconate dehydrogenase Pgd yes yes
FBgn0027580 CG1516 - CG1516 yes no
FBgn0000055 CG3481 Alcohol dehydrogenase Adh N/A yes
84
Table 17 Status of other enzymes in the glycolytic pathway
Flybase
number
Annotation
symbol Name Symbol
% FDR in
RIP/Chip
Smaug
mRNA
decay
target
Smaug
translational
repression
target
(FDR<10%)
FBgn0003071 CG4001 Phosphofructokinase Pfk 5.2 yes yes
FBgn0000064 CG6058 Aldolase Ald 58.3 no no
FBgn0086355 CG2171
Triose phosphate
isomerase Tpi 6.1 yes yes
FBgn0014869 CG1721 Phosphoglyceromutase Pglym78 -60.2 no no
FBgn0000579 CG17654 Enolase Eno 11.8 yes yes
FBgn0003178 CG7070 Pyruvate kinase Pyk 43.7 yes no
85
Figure 21 Validation of new Smaug targets.
Extracts were prepared from 0-1, 1-2 and 2-3 hour old wild-type and smaug mutant embryos and
assayed for the levels of Rpn7 (A), Su(z)12 (B) and Bicaudal C (C) proteins via western blots.
86
Figure 22 Glycolytic enzymes are overexpressed in smaug mutant embryos.
Extracts were prepared from 0-1, 1-2 and 2-3 hour old wild-type and smaug mutant embryos and
assayed for Hexokinase activity (A) or Phosphofructokinase activity (B). Activities are shown
relative to the wild-type 0-1 hour time point in each case. Results are the average of three
independent experiments and error bars indicate standard deviation. Data were analyzed using a
Student’s t test (** P < 0.05, * 0.05 < P <0.1).
87
2.3.9 Validation of new Smaug targets
To assess the role of Smaug in regulating the expression of the new target mRNAs, I selected
five for further analysis: Rpn7, Hexokinase, Phosphofructokinase, Su(z)12, and Bicaudal C
(BicC). Rpn7 is a proteasome regulatory particle subunit and was selected because of the
observed enrichment for GO terms related to ‘proteasome regulatory particle’. Likewise, because
of enrichment for the GO term ‘glucose metabolic process’ and the KEGG pathway
‘Glycolysis/Gluconeogenesis’, I assayed Hexokinase, the first enzyme in glycolysis, and
Phosphofructokinase, which represents a critical point of regulation (Pilkis et al., 1995; Pilkis et
al., 1988) and catalyzes the committed step of glycolysis (i.e., the product of this reaction serves
solely as a precursor to the final product of the glycolytic pathway). Polycomb repressive
complex 2 (PRC2) trimethylates histone H3 on lysine 27, a mark that is associated with
transcriptional silencing (Simon & Kingston, 2013). Thus, Su(z)12, a component of PRC2, was
of interest in light of the failure to induce zygotic transcription in smaug mutant embryos (Benoit
et al., 2009). Bicaudal C (BICC) is an RBP that represses the translation of target mRNAs during
Drosophila oogenesis (Gamberi & Lasko, 2012). Thus, Bicaudal C overexpression in smaug-
mutant embryos could disrupt normal patterns of post-transcriptional regulation.
Western blots (Rpn7, Su(z)12, Bicaudal C; Figure 21) or enzyme activity assays (Hexokinase,
Phosphofructokinase; Figure 22) showed that, in all cases, there was an increase in expression in
smaug-mutant embryos versus wild-type ones (Figure 21 and Figure 22), consistent with a role
for Smaug in down-regulation of its new target transcripts.
2.4 Discussion
Here I have used genome-wide approaches to identify mRNAs that are bound by Smaug and we
also defined those that are translationally repressed by Smaug. Our results show that the presence
of SREs is predictive of both binding and translational repression and, consistent with previous
work on the yeast and human Smaug homologs (Kazan et al., 2010; Li et al., 2010; Ray et al.,
2009; Ray et al., 2013; Riordan et al., 2011), indicate that the Drosophila SRE consensus is more
restricted than previously thought (Aviv et al., 2003). Integration of these new results with earlier
ones on Smaug’s global role in mRNA decay (Tadros et al., 2007) has led to the following
88
conclusions: 1) Smaug directly regulates the expression of a large number of mRNAs; 2) a large
fraction of Smaug-regulated transcripts are both destabilized and translationally repressed; and 3)
Smaug plays a key role in controlling the expression of mRNAs localized to the posterior of the
embryo. In addition, I have uncovered new and unanticipated roles for Smaug in regulation of
protein folding and decay, as well as in metabolism.
2.4.1 Translational repression versus mRNA decay
Previous work has firmly established that Smaug can both repress translation and induce
degradation of target mRNAs. However, Smaug’s two well-characterized target transcripts, nos
and Hsp83, are differentially regulated by Smaug (Dahanukar et al., 1999; Semotok et al., 2005;
Semotok & Lipshitz, 2007; Semotok et al., 2008; Smibert et al., 1999; Smibert et al., 1996). The
work presented here suggests that, unlike nos and Hsp83, Smaug both translationally represses
and degrades a large fraction of its target mRNAs. I hypothesize that the extent to which Smaug
regulates the translational repression and/or destabilization of its targets is likely to be a
consequence of additional cis-elements within target mRNAs. For example, the Hsp83 3’UTR
contains a translational enhancer that may mitigate Smaug-mediated translational repression
(Nelson et al., 2007). Similarly, the modest stabilization of nos mRNA observed in the absence
of Smaug suggests that additional cis-elements within the nos transcript function in its
destabilization.
2.4.2 Smaug’s role in the regulation of posterior-localized mRNAs
Smaug functions in the localization and regulation of its target mRNAs at the posterior pole of
the embryo (Bashirullah et al., 1999; Bergsten & Gavis, 1999; Dahanukar et al., 1999; Ding et
al., 1993; Semotok et al., 2005; Semotok et al., 2008; Smibert et al., 1999; Smibert et al., 1996;
Wang & Lehmann, 1991). This is a consequence of Smaug’s ability to induce transcript decay
and to repress translation in the bulk cytoplasm of the embryo combined with mechanisms that
inactivate Smaug function in the germ plasm at the posterior. Indeed, I have found that 38 of the
44 posterior-localized mRNAs that are bound to Smaug are regulated by Smaug at the level of
stability and/or translation.
A critical aspect of Smaug’s role in the regulation of nos and Hsp83 mRNA is the fact that
transcripts found at the posterior of the embryo escape Smaug regulation. The molecular
89
mechanisms that underlie this spatial regulation of Smaug function are not understood, but OSK
protein has been implicated in blocking Smaug function at the posterior and OSK protein has
been shown to physically interact with Smaug (Dahanukar et al., 1999; Ding et al., 1993; Gavis
& Lehmann, 1994; Smibert et al., 1996). Indeed, it has been shown that OSK’s interaction with
Smaug blocks Smaug’s ability to bind to its target mRNAs and it has therefore been proposed the
that OSK-Smaug interaction blocks Smaug function by preventing its interaction with target
transcripts (Jeske et al., 2011; Zaessinger et al., 2006). This simple model, however, is not
consistent with work showing that a torso mRNA carrying the first 96 nucleotides of the nos
mRNA’s 3’UTR, which carries one of the nos SREs, is repressed at both the anterior and
posterior of the embryo (Smibert et al., 1996). In addition, a torso mRNA carrying the first 185
nucleotides of the nos 3’UTR, which contains both nos SREs, is repressed at the anterior but is
expressed at the posterior (Dahanukar & Wharton, 1996). Taken together these data suggest the
existence of one or more cis-elements mapping within nucleotides 97-185 of the nos 3’UTR that
localize nos transcripts to the germ plasm (Gavis et al., 1996) and/or abrogate Smaug’s ability to
repress nos mRNA expression in the germ plasm. Our identification of several dozen posterior-
localized, Smaug-bound transcripts should facilitate identification of any additional cis-elements.
2.4.3 Identification of new biological functions for Smaug
My analysis of the mRNAs that are bound by Smaug has identified a number of mRNAs that
encode proteins that are involved in cell-cycle control and transcriptional regulation. Mis-
regulation of one or more of these mRNAs could underlie the cell-cycle and transcriptional
defects that occur in the absence of Smaug. My data also suggest that Smaug has several new
and unanticipated biological functions, including control of protein folding and degradation, lipid
droplet function and basic metabolism.
Protein folding and stability. My data suggest that Smaug downregulates the expression of nine
of the 19 subunits of the proteasome regulatory particle and four out of the eight that encode the
TRiC/CCT complex. In addition, three of the four remaining TRiC/CCT mRNAs and eight of the
remaining 10 proteasome regulatory particle mRNAs require Smaug for their degradation and/or
translational repression (Table 11and Table 13). It is unclear at this time whether these additional
mRNAs represent false negatives in the RIP-Chip experiments or whether Smaug regulates their
expression indirectly. Nonetheless, my data indicate that Smaug regulates the expression of
90
almost all of the components of these two protein complexes. Previous work has shown that
proteasome levels are repressed in early embryos (Klein et al., 1990) and my data suggest that
Smaug plays a major role in this repression. Given the role of the proteasome in cell-cycle
regulation (Reed, 2006), Smaug-mediated regulation of the proteasome may underlie some or all
of the cell-cycle defects observed in smaug mutants.
Lipid droplets. Previous experiments to characterize lipid droplet-associated proteins in embryos
employed six independent purifications and grouped the identified proteins based on the number
of purifications in which they were detected (Cermelli et al., 2006). They found 127 that were
identified in at least three purifications and 453 that were identified in one or two runs. Of the 28
Smaug-bound mRNAs that encode lipid-droplet proteins, 22 were identified in three or more
runs, suggesting that Smaug regulates mRNAs that encode proteins abundant in and/or tightly
associated with lipid droplets.
Lipid droplets are storage sites of triacylglycerols, hydrolysis of which yields fatty acids that can
be metabolized for energy or serve as a source of membrane precursors. Thus, lipid droplets
could function as the source of membrane precursors that are required during blastoderm
cellularization, a process during which plasma membrane invaginates around the syncytial nuclei
that are found at the embryo’s periphery. A role for Smaug in regulating lipid droplet function is
intriguing as smaug mutant embryos show defects in cellularization. In addition, given the
possible use of fatty acids as an energy source, Smaug’s regulation of lipid droplet function
could also reflect Smaug’s more general role in control of metabolic processes (see below).
Metabolism. My data also suggest a widespread role for Smaug in regulating metabolism in the
early embryo including a role for Smaug in down-regulation of glycolysis. Previous work has
suggested that maternal mRNAs encoding the glycolytic enzymes are present in early
Drosophila embryos but are rapidly degraded (Currie & Sullivan, 1994a, b; Roselli-Rehfuss et
al., 1992; Shaw-Lee et al., 1992; Shaw-Lee et al., 1991; Sun et al., 1988; Tennessen et al., 2011).
Glycolysis is down-regulated, not only in Drosophila, but also in frog and mammalian early
embryos (Dumollard et al., 2009; Dworkin & Dworkin-Rastl, 1991) but the molecular
mechanisms involved are unknown. My data implicate Smaug in the degradation and/or
translational repression of many of the glycolytic mRNAs. It will be interesting to test whether
91
post-transcriptional regulation of these mRNAs by Smaug’s homologs plays a role in the early
embryos of all animals.
2.4.4 Biological implications of the large number of Smaug-target mRNAs
My data are consistent with Smaug directly regulating a large number of mRNAs in early
embryos through translational repression and/or transcript degradation. This raises the question
as to whether all of these repressive interactions are biologically important.
In one model only a subset of Smaug’s targets are biologically relevant because the extent of
downregulation by Smaug varies in a target-dependent manner. For the biologically relevant
target transcripts, Smaug would effectively turn off their expression while for the others, Smaug
would reduce their expression insufficiently to have an effect on their biological function. A
similar type of model has been suggested for repression mediated by individual miRNAs, which,
as in the case of Smaug, regulate the expression of a large number of transcripts (Bartel & Chen,
2004). Given the low complexity of the binding sites of most RBPs it is likely that many of the
trans-acting factors that control mRNA translation and/or stability will regulate a large number
of transcripts and, as such, the same concepts should apply.
An alternative, but not mutually exclusive, model is that factors like Smaug, which repress the
expression of a large number of mRNAs, do so in order to limit the total levels of available
mRNA within a cell. This reduction could result from both Smaug-directed degradation of
transcripts and/or Smaug-mediated translational repression, the former eliminating the mRNAs
and the latter removing them from the pool of available mRNAs. In this model, Smaug would
function to control the competition among transcripts for limiting cellular components, such as
the translation machinery. I note, however, that my data do not support this model – at least in
regard to the translation machinery – as I fail to see a decrease in the translation of mRNAs that
are not bound by Smaug in smaug mutant embryos.
A third model to explain the biological significance of the regulation of a large number of
mRNAs by a single factor relates to a requirement for large-scale changes in a cell’s function.
Under such a circumstance one might expect that the expression of a large number of mRNAs
must be translationally repressed and/or degraded while a new group of genes is activated. For
example, during the first 2 to 3 hours of Drosophila embryogenesis, nuclei are transcriptionally
92
silent and development is driven by mRNAs synthesized by the mother and deposited into the
egg during oogenesis. Subsequently, one- to two-thirds of these maternal mRNAs are degraded
(De Renzis et al., 2007; Tadros et al., 2007; Thomsen et al., 2010) – the majority in a Smaug-
dependent manner – concurrent with activation of transcription in embryonic nuclei. In the early
embryo this widespread degradation appears to serve at least two purposes. The first involves
clearing the embryo of mRNAs that are no longer required. In the second, ubiquitously
distributed mRNAs are degraded but locally protected from decay (Bashirullah et al., 1999;
Semotok et al., 2005) or are degraded everywhere and then subsequently re-expressed in
spatially restricted patterns through transcriptional activation in select embryonic nuclei (De
Renzis et al., 2007). Thus, Smaug, through its regulation of a large number of mRNAs, may play
a major role in producing spatial precision in gene expression during the maternal-to-zygotic
transition in early embryos.
2.5 Material and methods
2.5.1 Drosophila Stocks
Wild-type flies consisted of the w1118
stock maintained in a large-scale Drosophila culture.
smaug mutant alleles included smaug1 (Dahanukar et al., 1999) and smaug
47. The smaug
47 allele
was generated via imprecise excision of a P-element (GE21229) using standard methods
(Hummel & Klambt, 2008). GE21229 is inserted 2499 base pairs 5` of the smaug start codon
and 20 base pairs down steam of the transcriptional start site of the smaug-RB isoform. All
isoforms are defined as described at http://flybase.org/. The original smaug1 allele showed
homozygous maternal effect lethality (Dahanukar et al., 1999) and we recovered six excision
lines demonstrating this phenotype. The extent of the deletion in these six lines was determined
via PCR analysis of genomic DNA. Two of the lines, smaug30
and smaug47
, showed deletions
removing large portions of the smaug gene, but not effecting the neighbouring upstream and
downstream genes– CG5087 and CG5280, respectively. Sequencing revealed that the smaug30
allele is a 4514 base pair deletion of the smaug gene beginning 2480 base pairs 5` of and ending
2034 base pairs 3` of the smaug start codon. Sequencing also showed that this allele retains 933
base pairs of the P-element. This deletion removes 2020 of 2997 base pairs of the open reading
frame of smaug RA, RB, RC, and RE isoforms. The smaug47
allele is a 5542 base pair deletion
beginning 2483 base pairs 5` of and ending 3059 base pairs 3` of the smaug start codon. This
93
deletion leaves 39 base pair of the open reading frame in the smaug RA, RB, RC, and RE
isoforms.
2.5.2 RNA co-immunoprecipitations
Embryos collected at 0-3 hours post-egglaying were dechorionated with 50% bleach and
homogenized in a minimal volume of RIP lysis buffer (150 mM KCl, 20 mM HEPES pH 7.4, 1
mM MgCl2, 1 mM DTT, 1x protease inhibitor cocktail (Bioshop)). Extracts were centrifuged for
10 minutes at 4°C, and the supernatant was supplemented with 9 M urea to a final concentration
of 2 M. Protein A beads were pre-incubated with either guinea pig anti-Smaug antibody (Tadros
et al., 2007) or normal guinea pig serum followed by four washes with RIP lysis buffer
supplemented with urea. These beads were then incubated with embryo extract for 2 h at 4°C
followed by four washes with RIP lysis buffer supplemented with urea and RNA was extracted
from the beads using the Trizol reagent (Invitrogen).
2.5.3 Polysome gradients
Embryos laid by wild-type or smaug1 homozygous mothers were collected 0-2 hours post-
egglaying, dechorionated with 100% bleach and lysed in an equal volume of polysome lysis
buffer (7.5 mM MgCl2, 500 mM NaCl, 25 mM Tris pH 7.5, 2 mg/mL heparin, 0.5 mg/mL
cycloheximide, 1 mM DTT, 50 U/mL RNasin, 1 mM 4-(2-Aminoethyl) benzenesulfonyl fluoride
hydrochloride (AEBSF), 2 g/mL leupeptin, 2 mM benzamidine, 2 g/mL pepstatin A). Lysed
samples were diluted 1 in 12.5 in polysome lysis buffer and 30% Triton was added to a final
concentration of 1% and then spun at 6000xg for 10 minutes and the resulting supernatant was
diluted in polysome lysis buffer supplemented with 1% Triton to an A260 of 12.5.
A 12 mL 15% to 45% linear sucrose gradient in 7.5 mM MgCl2, 500 mM NaCl, 50 mM Tris pH
7.5 was created using a BioComp Model 117 Gradient Mate gradient maker using a rotation
angle of 80.5o and a rotation speed of 18 rpm for 1 minute and 58 seconds. After chilling the
polysome gradient on ice, 400 L of diluted embryo extract was loaded onto the top of the
gradient which was then spun at 36,000 rpm in a Beckman SW 41 Ti rotor for 2.5 hours. The
gradients were then separated into four pools (pool 1 contained the top 4 mL, pool 2 contained
the next 2 mL, pool 3 contained the next 3 mL and pool 4 contained the last 3 mL and the pellet).
A fixed amount of exogenous in vitro transcribed Arabidopsis spike-in RNAs were then added to
94
each pool. Our microarrays contain probes that allow for the detection of these RNAs allowing
for subsequent data normalization. 20% SDS, 0.5 M EDTA and 20 mg/mL proteinase K were
added to each fraction to final concentrations of 0.8%, 0.01 M and 0.128 mg/mL, respectively
and then incubated for 30 minutes at room temperature. Glycogen was then added to a final
concentration of 80 g/mL and samples were ethanol precipitated overnight and the resulting
pellet was washed with 75% ethanol and resuspended in phenol-saturated water. Following two
phenol-chloroform extractions samples were precipitated by the addition of 7.5 M LiCl to a final
concentration of 1.5 M and an overnight incubation at 4oC. The resulting pellet was washed with
75% ethanol, resuspended in water and ethanol precipitated in the presence of 80 g/mL of
glycogen and 0.3 M sodium acetate. The precipitate was then washed with 75% ethanol and
resuspended in water. The integrity of RNA in each pool was confirmed via northern blots which
were probed for nos mRNA (Figure 23).
Experiments that utilized EDTA treatment involved lysis of embryos in polysome lysis buffer
and the resulting sample was split in two and the polysome gradient experiment proceeded as
described above with the following changes. One sample was diluted into polysome lysis buffer
and fractionated as normal, while the other was diluted in polysome lysis buffer lacking MgCl2
and containing 25mM EDTA and fractionated on gradients containing 25mM EDTA and lacking
MgCl2. After centrifugation these gradients were divided into 12 1 mL fractions and RNA was
extracted from each fraction and analyzed via northern blot.
For experiments that utilized puromycin embryos were lysed in puromycin lysis buffer (50 mM
Tris pH 7.5, 2 mM MgCl2, 500 mM KCl, 100 M GTP, 1 mM DTT, 50 U/ml RNasin, 1 mM
AEBSF, 2 g/mL leupeptin, 2 mM benzamidine, 2 g/mL pepstatin A). The lysed samples were
split in half and cycloheximide was added to one sample to a final concentration of 0.5 mg/mL
and puromycin was added to the other sample to a final concentration of 2 mM. Samples were
left on ice for 20 minutes and then incubated at 30oC for 10 minutes. Both samples were then
diluted 1 in 12.5 with polysome lysis buffer supplemented with either puromycin or
cycloheximide and 30% triton was added to a final concentration of 1%. The samples were then
spun at 6000xg for 10 minutes and the supernatant was diluted with polysome lysis buffer
supplemented with either puromycin or cycloheximide to give an A260 of 12.5 and these diluted
samples were then fractionated as described above.
95
Figure 23 Assessing the integrity of polysome fractionated mRNA.
Aliquots of RNA from polysome gradient pools 1, 2, 3 and 4 that were analyzed by microarray
were subjected to northern blot analysis by probing for nos mRNA.
96
2.5.4 Microarrays
RNA samples from RIP experiments were used to prepare single stranded cDNA using anchored
oligo(dT) primers and the Canadian Drosophila Microarray Centre indirect labeling protocol
which can be viewed at http://www.flyarrays.com/. Anchored oligo(dT) primers consist of 20 T
residues and end in an A, C or G residue followed by an A, C, G or T. Thus priming occurs only
at the 5’ end of the poly(A) tail and transcripts with short tails will be primed with equal
efficiency to those that have long tails. RNA samples from polysome experiments were used to
generate double stranded cDNA following the protocol described in the NimbleGen Array User’s
Guide (Gene Expression Arrays, version 5.0) using all reagents at half the normal amount and a
primer mixture of random hexamer primers and anchored-oligo-dT primers. Cy3 or Cy5-tagged
random nonamers were then used to labeled cDNAs using the Roche NimbleGen protocol. The
cDNA resulting from RIP experiments were used to probe Nimblegen 4x72K arrays (GEO
platform number: GPL13782), while the cDNA from polysome gradients was used to probe a
custom-designed Drosophila 4x72K NimbleGen arrays (GEO platform number: GPL10539) that
contain probes for Arabidopsis spike-in RNAs (see below). Microarrays were scanned using
Genepix Pro software on a Molecular Devices GenePix 4000B or 4300A scanner and quantified
using Nimblescan.
RIP microarrays were normalized using the Robust Multi-array Average (RMA) quantile method
and transcripts that were expressed at levels significantly above background in total RNA
collected 0-3 hours post-egglaying were determined using ‘one class unpaired analysis’ in SAM
and transcripts with an FDR>5% were excluded from further analysis of the RIP data. mRNAs
that were reproducibly enriched in Smaug RIPs versus control RIPs were then identified by
comparing the log2(Smaug IP/total RNA) and the log2(mock IP/total RNA) using ‘two class
unpaired analysis’ in SAM (FDR<5%).
Polysome microarrays were normalized using the RMA quantile method. We further normalized
the data using Arabidopsis spike-in RNAs. The hybridization signals from the spike-in RNAs
were utilized by applying a linear transformation to each sample with the parameters, a and b,
determined by fitting the linear function Y = aX+b using the spike-in signal, where X is the
expression level of the spike-in RNAs in a specific sample, and Y is the mean expression level of
the spike-in RNAs across all the samples. The genes significantly expressed in wild-type or
97
smaug mutant embryos in each of pools 1, 2, 3 and 4 were separately determined using ‘one
class unpaired analysis’ in SAM (FDR<5%). We defined the genes significantly expressed in the
wild-type and smaug mutant embryos as the union of the significantly expressed genes from the
four fractions derived from that genotype. We then compared these two lists and defined their
intersection as the list of genes significantly expressed in both wild-type and smaug mutant
embryos, and restricted all the following analysis to the genes in this list. To determine the list of
genes with different polysome association in wild-type and smaug mutants, we compared the
geometric mean of the expression level in pools 3 and 4 (normalized to the levels in pool 1) in
wild-type and smaug mutant embryos, using ‘two class unpaired analysis’ in SAM.
2.5.5 RT-qPCR
cDNA was synthesized using SuperScript II reverse transcriptase (Invitrogen) and random
primers according to the manufactures instructions. Quantitative PCR reactions were carried out
using the BioRad Real-time PCR system as per the manufactures instructions. Levels of RpL32
mRNA in each immunoprecipitated sample were used to normalize the levels of the
experimental mRNA in that sample.
2.5.6 Estimating the number of genes that are translationally repressed by Smaug
The fraction of genes expected to have changed in TI in smaug and wild type embryo samples
for the top N and bottom N Smaug-binders (for N = 250, 500, and 1000) was calculated using the
R (version 2.14.1) algorithm sm.density() in the sm package (version 2.2-4.1). The sm.density()
algorithm provided smoothed density estimates for 100 values of change in TI for the top and
bottom N binders, with the 100 values calculated by the sm.density() algorithm with each
smoothed density estimate.
For every gene expressed in our polysome gradient experiments, the probability that it was a
positive target (i.e., a target of Smaug-mediated repression) was estimated using the top N and
bottom N Smaug-binders (for N = 250, 500, and 1000) . First, for each gene, the density of its
change in TI under the positive and negative distributions as defined by N top and bottom
binders, respectively, was set to be equal to that of the closest grid point higher than the change
in TI. We then estimated the probability that a gene was a positive by taking the ratio of its
density under the positive distribution and the sum of its densities under the positive and
98
negative distributions This procedure was repeated for each of our three sets of positive and
negative distributions to give us three different sets of probabilities. For each of these three sets
of probabilities, we estimated the expected number of Smaug targets for that set by summing the
“positive probabilities” for all genes.
2.5.7 SRE searching
We used a two-step procedure to computationally predict SRE stem/loops carrying the loop
sequence CNGGN0-4 on a non-specific stem. First, we performed an initial scan using
RNAplfold (version 2.0.7) (Bernhart et al., 2006) with the parameters '-W', '170', '-L', '120', '-T',
'25' choosing parameters as they were within the range suggested by Lange et al (Lange et al.,
2012). Potential SREs for further analysis were identified as CNGG sequences where the base
immediately 5’ to the CNGG sequence was involved in a canonical base pair with one of five
nucleotides immediately 3’ to the CNGG sequence with probability > 0.01. We estimated the
probability of formation of an actual SRE (i.e., CNGG at the 5’ end of the hairpin loop and a
loop of length 4 to 8 nucleotides) at each candidate site using the RNAsubopt (Wuchty et al.,
1999) routine from the Vienna RNA package. In particular, we sampled 3,000 structures for each
of a series of windows overlapping the candidate site (from the Boltzmann ensemble using the ‘-
p’ option), computed the empirical probability of SRE formation in each window, and set the
SRE probability for a site to be the average of these probabilities. The most 5’ of the sequence
windows spanned 75 nucleotides upstream of the candidate site, the site itself, and the 40
nucleotides downstream of the site. The most 3’ of the windows spanned 40 nucleotides
upstream of the site to 75 nucleotides downstream. Between these two, all of the other windows
were offset by a single nucleotide. These site probabilities were then summarized at the
transcript level. The initial SRE score for each transcript was the sum of the SRE probability
values at each candidate site within the entire transcript. The same procedure was used to search
for CNGG sequence variants and calculate a variant score for each transcript. Once obtained,
SRE scores and the scores of sequence variants were compared with polysome and RIP data
using standard R packages. Spearman's correlation values across all of the expressed genes were
determined using the cor.test() algorithm with default parameters and the Spearman method.
Linear models were created using the lm() algorithm with default parameters.
99
2.5.8 Localization pattern enrichment analysis
These analyses were carried out as described in Laver et al., (Laver et al., 2013).
2.5.9 Western blots
Antibodies against Rpn7 (Santa Cruz Biotechnology, catalogue #SC-65750), Su(z)12 (Tie et al.,
2003) and Bicaudal C (Chicoine et al., 2007) were used in standard western blot assays.
2.5.10 Glycolytic enzyme assays
For enzyme assays smaug mutant embryos were collected from females homozygous for the
smaug47
allele, while wild-type embryos were collected from females homozygous for the
smaug47
allele that were also homozygous for a genomic smaug rescue transgene that was
inserted at the attP40 site on the second chromosome by Genetic Services (Cambridge, MA)
using PhiC31 integrase-mediated transgenesis (Bischof et al., 2007). The smaug transgene,
which rescues the smaug mutant phenotype, is a modified version of a previously generated
smaug rescue construct (Dahanukar et al., 1999) that expresses a version of Smaug that is tagged
at its N-terminus with FLAG and p53 epitope tags.
For the Hexokinase assay, embryos were homogenized in extraction buffer (0.05 M Tris-HCl,
pH 8.0 with 13.3 mM MgCl2) and assayed in extraction buffer supplemented with 16.5 mM
ATP, 20 mM beta-NADP and 0.67 M glucose. Hexokinase catalytic activity was measured by
adding Leuconostoc mesenteroides glucose-6-phosphate dehydrogenase (Sigma-Aldrich
Chemicals) (Worthington Code: ZF or ZFL) dissolved at a concentration of 300 IU/mL in
extraction buffer. The production of beta-NADPH was monitored at 340 nm in a Thermo
SPECTRONIC spectrophotometer. Experiments were conducted with an amount of embryo
extract that was in the linear range of the assay and enzyme activity was normalized to protein
concentrations in each homogenate using the Bradford assay (Bio-Rad). Enzyme activity was
calculated using the formula: Units/mg protein = A340/min ÷ 6.22 x mg enzyme/mL reaction
mixture, as described by Worthington (http://www.worthington-biochem.com/HK/assay.html).
For Phosphofructokinase assay, we used the Phosphofructokinase activity colorimetric assay kit
(BioVision, Milpitas, CA, USA), which converts fructose-6-phosphate and ATP to fructose-
diphosphate and ADP. The final product, NADH, reduces a colorless probe to a colored product
100
with strong absorbance at 450 nm. The absorbance was measured with a TECAN INFINITE
m200 microplate reader. Experiments were conducted with an amount of embryo extract that
was in the linear range of the assay and enzyme activity was normalized to protein concentration.
101
Chapter 3 Smaug regulates primordial germ cell number by repressing synthesis of Bruno in the germ plasm of Drosophila
embryos
The work described in this chapter was modified from the following manuscript (in preparation):
Najeeb U. Siddiqui, Linan Chen, Angelo Karaiskakis, Aaron L. Goldman, Hua Luo, Xiao Li,
Zhiyong Yang, Thomas Kislinger, Quaid Morris, Craig A. Smibert, Howard D. Lipshitz. Smaug
regulates primordial germ cell number by repressing synthesis of BRU in the germ plasm of
Drosophila embryos.
Contributions:
I performed the following experiments and analysis: Co-IP of Smaug and arrest mRNAs and
qRT-PCR analysis; Assessment of Arrest/Bru’s mRNA and protein level in whole embryo
extract; generation of the S-A(5xSRE-)-S transgenic construct and flies. Najeeb U. Siddiqui
performed all the live imaging experiments and analysis of Smaug particle movement. Angelo
Karaiskakis performed all confocal imaging experiments including three-dimensional
reconstruction and counts of PGC numbers. Aaron L. Goldman performed electron microscopy
experiments.
102
3.1 Abstract
The amount of germ plasm produced during Drosophila oogenesis determines the number of
primordial germ cells that form in the embryo. I show that the Smaug RBP sits at the top of a
posttranscriptional cascade that regulates primordial germ cell number by attenuating synthesis
of germ plasm in the embryo. Smaug is transported into the germ plasm of the early embryo
where it represses translation of the arrest mRNA, which encodes the BRU RBP. Since BRU is a
positive regulator of osk mRNA translation, repression of BRU translation prevents OSK
synthesis in the embryo, thus attenuating production of germ plasm. Eliminating the Smaug
protein results in production of excess BRU, excess OSK, excess germ plasm, and extra
primordial germ cells. Over-expression of BRU in an otherwise wild-type embryo produces the
same results. Smaug and BRU thus regulate primordial germ cell number.
3.2 Introduction
In Drosophila, C. elegans, Xenopus and many other animals, a specialized region of the egg
cytoplasm known as the germ plasm (in Drosophila, the ‘pole’ plasm) directs the formation and
fate of the primordial germ cells of the early embryo (Houston & King, 2000; Mahowald, 2001).
Within the germ plasm reside electron-dense, non-membrane-bound organelles known as polar
granules (in Drosophila, these are referred to as ‘polar granules’ and in C. elegans as ‘P
granules’), which contain specific molecular determinants essential for primordial germ cell fate.
In Drosophila, a genetic pathway has been defined for the assembly of germ plasm and polar
granules (Mahowald, 2001; Rongo et al., 1997). During oogenesis, the osk mRNA is transported
to the posterior pole of the oocyte where it is translated and nucleates the assembly of the germ
plasm. OSK is a component of the polar granules that is essential for the recruitment and
translation of the other components, including VAS and NOS protein.
Translation of the osk mRNA is kept repressed during its transport to the posterior pole of the
oocyte (Kim-Ha et al., 1995). Furthermore, localization of osk transcripts to the posterior is
inefficient, with over 85% of the mRNA remaining unlocalized, necessitating mechanisms to
prevent this pool of mRNA from producing OSK protein (Bergsten & Gavis, 1999). A major
repressor of osk mRNA translation is the BRU protein, which is encoded by the arrest gene
(Chekulaeva et al., 2006; Kim-Ha et al., 1995; Reveal et al., 2010; Webster et al., 1997). BRU is
103
an RBP that recognizes cis-elements (BRU response elements or BREs) in the osk 3’UTR to
establish repression. Relief of repression in the germ plasm is accomplished by three types of cis-
elements: a 5’UTR-located activating element that recruits Hrp48 (Yano et al., 2004); IMP/Zip-
code binding elements in the 3’UTR (IBEs) (Munro et al., 2006); and, unexpectedly, one of the
clusters of BREs (the ‘C’ cluster) in the 3’UTR (Reveal et al., 2010). BRU itself is a strong
candidate to be the factor that binds the C cluster to relieve repression but the early oogenesis
arrest of mutants has precluded proof that BRU is required for osk translation in the germ plasm.
BRU protein disappears from late stage oocytes and has been reported to be undetectable in
embryos by Western blot (Webster et al., 1997). Furthermore, immunostains have suggested that
BRU protein is absent from the germ plasm and the primordial germ cells when they bud
(Rangan et al., 2009).
The primordial germ cells are the first cells to form in the early Drosophila embryo. They bud
from the posterior pole of the embryo 90 minutes after fertilization and then undergo one or two
rounds of division to give the stereotypical number of ~30 primordial germ cells (Rongo et al.,
1997). These then cease division and subsequently migrate to the somatic component of the
gonad. The amount of germ plasm correlates with the number of primordial germ cells that form.
Loss of germ plasm, as in mutants of osk and other key players, results in failure to form
primordial germ cells (Boswell & Mahowald, 1985; Lehmann & Nusslein-Volhard, 1986;
Schupbach & Wieschaus, 1986) while reduction of osk gene dose from two copies to one results
in about a 50% decrease in primordial germ cell number (Ephrussi & Lehmann, 1992). In
contrast, over-expression of OSK protein using four or six copies of an osk transgene results in
about a 50% increase in primordial germ cell number (Ephrussi & Lehmann, 1992; Smith et al.,
1992).
We previously reported that the multifunctional RBP, Smaug, is enriched in the Drosophila germ
plasm and persists in the primordial germ cells after it is eliminated from the somatic component
of the embryo (Siddiqui et al., 2012; Smibert et al., 1999). Here we show that Smaug protein
accumulates in the germ plasm of early embryos where it is integrated into the polar granules.
Mutations in smaug result in a 30% increase in primordial germ cell number as a consequence of
production of excess OSK in the germ plasm of the early embryo. We show that the arrest
mRNA co-purifies with Smaug protein, that Smaug represses translation of the arrest mRNA in
the germ plasm of the embryo, and that over-expression of BRU in a wild-type embryo results in
104
excess OSK and extra primordial germ cells. Thus, Smaug regulates primordial germ cell
number by controlling the amount of BRU and OSK in the germ plasm of early embryos. An
implication of these results is that germ plasm assembly is not restricted to oogenesis but can
also occur in embryos.
3.3 Results
3.3.1 Smaug accumulates in the germ plasm of early embryos
Smaug protein is translated in early embryos but is rapidly eliminated during interphase of
nuclear cycle 14 (Benoit et al., 2009; Smibert et al., 1999; Tadros et al., 2007). An exception
occurs in the germ plasm and primordial germ cells, where Smaug persists (Siddiqui et al., 2012;
Smibert et al., 1999). To examine Smaug protein accumulation in the germ plasm of live
embryos, we produced transgenic flies that express either mCherry-Smaug or Venus-Smaug
under control of the smaug gene’s endogenous promoter, 5’- and 3’UTRs.
Smaug protein could be seen to be enriched in the germ plasm as early as nuclear cycle 5 (~30
minute after fertilization: Venus-Smaug; Figure 24A; Movie S1), to become associated with the
spindle microtubules and then to be taken up into the primordial germ cells when they bud (~90
minutes: mCherry-Smaug; Figure 24B; Movie S2). Smaug-containing particles showed net
movement towards the germ plasm, with a greater mean velocity towards (0.47 +/- 0.04 SEM
µm/second; n = 29) than away from the germ plasm (0.26 +/- 0.02 SEM µm/second; n = 19).
Since such movement is towards the minus end of microtubules, we assayed Smaug movement
in a dynein mutant (dhc6-10
/dhc6-12
) and found Smaug accumulation to be severely compromised
(Movies S3 and S4), suggesting a role for microtubules in the germ-plasm enrichment of Smaug.
We conclude that Smaug protein accumulates in the germ plasm of early embryos and that a
subset is actively captured by, or transported into, the germ plasm.
3.3.2 Smaug and VAS are dynamic components of the polar granules
Several known components of the germ plasm are localized in subcellular organelles known as
germinal (‘polar’) granules. It has been shown that the VAS protein and nos mRNA are captured
by the astral microtubules of arriving nuclei and become concentrated in the polar granules about
105
an hour after fertilization (Lerit & Gavis, 2011). To assess whether Smaug is also a component
of the polar granules we simultaneously imaged mCherry-Smaug and VAS-GFP in live embryos.
We found that these proteins co-localize prior to, during, and after budding of the primordial
germ cells (Figure 25A; Movie S5). Immuno-electron microscopy confirmed that Smaug and
VAS co-localize in the polar granules (Figure 25B, left and center panels).
During these analyses we noted that a subset of Smaug remains in the posterior cytoplasm, apical
to the somatic nuclei well after budding of the primordial germ cells is complete (Figure 25C).
That this represents germ plasm that is left behind after formation of the primordial germ cells,
was confirmed by its electron-dense, non-membrane-bound nature (Figure 25B, right panel) and
colocalization of Smaug protein and VAS protein (Figure 25C). These data are consistent with
histological analyses 50 years ago reporting that some of the polar granules remain behind after
the primordial germ cells bud (Counce, 1963).
To assess whether VAS and Smaug are dynamic components of the polar granules, we carried
out fluorescence recovery after photobleaching (FRAP) analysis using VAS-GFP and mCherry-
Smaug (Figure 26). VAS-GFP rapidly recovered in the polar granules in the primordial germ
cells (Figure 26A) as well as in the left-behind granules in the cytoplasm apical to the somatic
nuclei (Figure 26B): The time to half-recovery for VAS in the primordial germ cell-located
granules was ~10 seconds and in the granules left behind in the apical cytoplasm was <5
seconds. In contrast, while mCherry-Smaug did recover, the rate was about ten-times slower
(Figure 26C, D): 75 seconds and 45 seconds for primordial germ cell granules and left-behind
granules, respectively.
We conclude that both Smaug and VAS are dynamic components of polar granules but differ in
their mobility and the kinetics of their movement into the granules.
106
Figure 24 Smaug protein is enriched in the germ plasm of early embryos.
(A) Live imaging of Venus-Smaug shows that Smaug protein is enriched in the germ plasm as
early as nuclear cycle 5 (~30 minute after fertilization). Still images were taken from Movie S1.
The time after the commencement of live imaging is indicated at the top right of each still image
in minutes and seconds. (B) In live embryos, mCherry-Smaug (red) becomes associated with the
spindle microtubules (green; imaged with Jupiter-GFP) upon arrival of nuclei at the posterior
(~60 minutes of embryogenesis) and is taken up into the primordial germ (‘pole’) cells when
they bud (~90 minutes of embryogenesis). Still images were taken from Movie S2. See also
associated Movies S3 and S4.
107
Figure 25 Smaug is a component of the polar granules.
(A) Smaug co-localizes with VAS, a known component of polar granules. Panels are still images
taken from live-imaged movies of mCherry-Smaug (red) and VAS-GFP (green) transgenic
108
embryos. See associated Movie S5. (B) Immuno-electron microscopy shows that, in the pole
cells, Smaug (10 nm gold: small particles) is found in polar granules along with VAS (15 nm
gold: large particles, indicated with arrowheads) (left and center panels). In the posterior somatic
cells (right panel), Smaug (10 nm gold) is enriched in electron-dense, non-membrane-bound
organelles. (C) Smaug and VAS co-localize in the apical region of the posterior soma,
representing polar granules that are not taken up into the primordial germ cells when they bud.
See also associated Figure 26.
109
Figure 26 FRAP analysis of Smaug and VAS protein in the polar granules.
(A) FRAP of VAS-GFP in polar granules in primordial germ cells. (B) FRAP of VAS-GFP in
polar granules that are left behind in soma. (C) FRAP of mCherry-Smaug in polar granules in
primordial germ cells. (D) FRAP of mCherry- Smaug in polar granules that are left behind in
soma. The mobile fractions were 55% (Smaug in primordial germ cell granules), 70% (VAS in
primordial germ cell and leftbehind granules), and 90% (Smaug in left-behind granules).
110
3.3.3 Excess OSK protein is produced in the germ plasm of smaug mutant embryos
We recently used MuDPIT mass spectrometry to define the proteome of newly formed
Drosophila primordial germ cells purified by flow cytometry from 1-to-3 hour-old wild-type
embryos (Siddiqui et al., 2012). Several germ plasm proteins are highly enriched in these cells,
most notably OSK, Tudor, VAS, Piwi and Aubergine.
Here we report MuDPIT analysis of the proteome of primordial germ cells from embryos
produced by smaug mutant females (henceforth referred to as ‘smaug mutant embryos’) that
were sorted in parallel with those wild-type primordial germ cells. Four of the five proteins that
are highly enriched in wild-type primordial germ cells (OSK, Tudor, Aubergine, VAS) are
present at even higher levels in smaug-mutant primordial germ cells, with OSK the most over-
expressed (~2.5-fold higher levels than in wild type) (Table 18). The fifth, Piwi, was over-
expressed in one smaug replicate but not detected in the other, therefore it did not meet our
criteria for being ‘present’ (see Materials and methods in Siddiqui et al., 2012).
To assess whether an increase in OSK levels in the germ plasm preceded budding of the
primordial germ cells, we immunostained embryos and found that, in smaug mutants, excess
OSK is present in the germ plasm prior to, as well as during, budding of the primordial germ
cells. This observation was confirmed both in embryos expressing VAS-GFP (Figure 27 A cf. B)
and in ones in which only endogenous VAS was present (Figure 28,Table 19). Since Smaug
protein has no function during oogenesis, is absent from oocytes, and is not synthesized until
after egg activation (Dahanukar et al., 1999; Smibert et al., 1999; Tadros et al., 2007), we
conclude that, directly or indirectly, Smaug attenuates production of OSK in the germ plasm of
the early embryo.
111
Table 18 MuD-PIT normalized spectral counts for the wild-type and smaug-mutant
primordial germ cells.
The fold change was calculated using the geometric mean of the MuD-PIT reads from the two
replicates of smaug mutants divided by the geometric mean of the MuD-PIT reads from the two
replicates of wild type. The normalization was performed using the trimmed mean method as
previously described (Siddiqui et al., 2012).
112
Figure 27 Excess OSK protein is produced in the germ plasm of smaug mutants.
(A) Wild-type embryos immunostained to visualize VAS-GFP (left) and endogenous OSK
(right) (B) Embryos from smaug mutant mothers immunostained to visualize VAS-GFP (left)
and endogenous OSK (right). The genotype of the smaug mutant females from which the
embryos were obtained was smg1/Df(3L)Scf-R6. In both (A) and (B) the upper panels show the
posterior pole of an embryo prior to budding of the primordial germ cells and the lower panels
show the posterior pole of an embryo after budding of the primordial germ cells. See also
associated Figure 28 and Table 18.
113
Figure 28 Excess OSK protein is produced in the germ plasm of smaug mutants.
(A) Wild-type embryos immunostained to visualize OSK. (B) Embryos from smaug mutant mothers
(genotype: smg1/Df(3L)Scf-R6) immunostained to visualize OSK. In both (A) and (B) the upper two
sets of panels show the posterior pole of an embryo prior to budding of the primordial germ cells and
the lower panels show the posterior pole of an embryo after budding of the primordial germ cells.
Scale bar in the low magnification image is 50 μm; that in the high magnification image is 25 μm.
A B
114
Table 19 Primordial germ cell number in wild type and in smaug mutants.
115
3.3.4 smaug mutants produce excess primordial germ cells
It has been shown that over-expression of OSK protein leads to the production of extra
primordial germ cells (Ephrussi & Lehmann, 1992; Smith et al., 1992). We, therefore, next
assessed whether more primordial germ cells form in smaug mutants than in wild type. We
examined the original smaug allele (smg1), which introduces a stop codon before the RNA-
binding domain, resulting in synthesis of a truncated protein (Benoit et al., 2009), as well as a
new allele (smg47
), which was generated by imprecise P-element excision and is a protein-null
allele (see Experimental Procedures). Extra primordial germ cells form in smaug mutant
embryos: mean numbers ranged from 37.5 to 45.5 in smaug mutants versus 29.9 to 31.6 in wild
type (Figure 29A, B, Table 19). The overall mean for wild type was 30.5 and for smaug mutants
it was 39.7. There are significantly more primordial germ cells in smaug mutants than in wild
type (Wilcoxon-Mann-Whitney [WMW] rank sum P < 1.3 x 10-14
). The increase in germ cell
number in smaug mutants (an increase of between 25 and 44% depending on the allele
combination, with an overall average of increase of 30%) is similar to that previously observed
in “4 x” and “6 x” osk embryos (Ephrussi & Lehmann, 1992; Smith et al., 1992), consistent with
the extra germ cells resulting from the increased OSK levels that we observe in the germ plasm
of smaug mutants.
There are two possible reasons for appearance of extra primordial germ cells in smaug mutants:
more primordial germ cells may bud from the posterior tip of the embryo and/or the normal
number may bud but then undergo more cell divisions than in wild type. To assess these
possibilities we co-stained wild-type and smaug-mutant embryos with VAS, which marks the
primordial germ cells, and phosphohistone H3 (PH3), which marks cells in mitosis (Hendzel et
al., 1997). The mean number of PH3-positive primordial germ cells ranged from 0.7 to 2.3 in
smaug mutants versus 1.1 to 2.2 in wild type (Figure 29C-K and Table 19). The overall mean for
wild type was 1.96 and for smaug mutants was 1.12. There are, thus, fewer – rather than more –
dividing primordial germ cells in smaug mutants than in wild type (WMW rank sum P =
0.0022).
We conclude that more primordial germ cells bud from the posterior pole of smaug mutant
embryos than wild type. This conclusion is consistent with the known role of OSK levels in
specifying the number of primordial germ cells that form.
116
Figure 29 Excess primordial germ cells form in smaug mutants.
117
(A), (B) Histograms showing the number of primordial germ cells (PGCs) in embryos from a
variety of wild-type (A) or smaug mutant (B) mothers. (C, E, G, I, J) Confocal images showing
Vasa (green) and Phospho-histone H3 (PH3; red) in wild type (w1118
). (D, F, H, K) Confocal
images showing Vasa (green) and Phospho-histone H3 (PH3; red) in smaug mutants. The
genotype of the smaug mutant females from which the embryos shown in (D), (F), (H) and (K)
were obtained was smg1/Df(3L)Scf-R6. See also associated Table 19.
118
3.3.5 The arrest mRNA co-purifies with the Smaug protein
BRU is an RBP that is encoded by the arrest gene (Chekulaeva et al., 2006; Kim-Ha et al., 1995;
Webster et al., 1997). While BRU was initially identified as a translational repressor, it has been
shown more recently that positive regulation of OSK translation in the oocyte’s germ plasm is
accomplished through a cluster of BREs in the osk 3’UTR (Reveal et al., 2010). Smaug regulates
its target mRNAs by binding to cis-elements known as Smaug recognition elements (SREs)
(Dahanukar & Wharton, 1996; Smibert et al., 1996). The arrest mRNA contains five predicted
SREs, all in its ORF (Figure 33) and therefore is a candidate for regulation by Smaug.
To assess whether the arrest mRNA co-purifies with Smaug, I immunoprecipitated the Smaug
protein from early embryos and carried out RT-qPCR of the co-purifying mRNAs. The arrest
mRNA was highly co-enriched with Smaug (Figure 30A): After normalization to levels of a
negative control mRNA (RpL32), arrest mRNA was enriched 11.9-fold in the Smaug
immunoprecipitate relative to the mock immunoprecipitate. Two previously known Smaug-
associated mRNAs, nos and Hsp83, were enriched 9.8- and 6.4-fold, respectively, consistent
with our previous analyses (Semotok et al., 2005). I conclude that arrest mRNA is associated
with Smaug protein in early embryos.
119
120
Figure 30 Smaug represses translation of the arrest mRNA in the germ plasm.
(A) The arrest mRNA is associated with Smaug protein. Histogram showing fold-enrichment
assayed by RT-qPCR after immunoprecipitation of Smaug protein. The numbers on the Y-axis
are fold-enrichment values of Smaug immunoprecipitate relative to mock immunoprecipitate
normalized to a negative control mRNA (RpL32). These are shown for arrest mRNA and two
known targets of Smaug (Hsp83 and nos). N = 3. The standard error of the mean is indicated.
See also associated Figure 33 and Figure 31, the latter of which shows that Smaug is essential for
degradation of arrest mRNA in the bulk cytoplasm. (B) Western blot showing that, in wild-type
embryos, BRU is weakly detectable at 0-to-1 hr and undetectable thereafter whereas, in smaug
mutants, BRU is strongly detectable at all three time points assayed and levels increase with
time. (C) Embryos from wild-type mothers immunostained to visualize VAS-GFP (left) and
endogenous BRU (right). (D) Embryos from smaug mutant mothers immunostained to visualize
VAS-GFP (left) and endogenous BRU (right). The genotype of the smaug mutant females from
which the embryos were obtained for (A) and (C) was smg1/Df(3L)Scf-R6. In both (C) and (D)
the upper panels show the posterior pole of an embryo prior to budding of the primordial germ
cells and the lower panels show the posterior pole of an embryo after budding of the primordial
germ cells. See also associated Figure 32.
121
Figure 31 arrest mRNA is stabilized in the bulk cytoplasm of smaug mutants.
Endogenous arrest mRNA levels were measured in wild-type (black) or smaug-mutant (red)
embryos that were 0-to-1 hr, 1-to-2 hr, 2-to-3 hr old. The amount of RNA was quantified by RT-
qPCR. Endogenous RpL32 mRNA levels were used for normalization and the 0-to-1 hr amount
was set to 100% for each genotype. It can be seen that arrest transcripts are strongly stabilized in
smaug embryos. The standard error of the mean is indicated.
122
Figure 32 Excess BRU protein is produced in the germ plasm of smaug mutants.
(A) Wild-type embryos immunostained to visualize BRU. (B) Embryos from smaug mutant
mothers (genotype: smg1/Df(3L)Scf-R6) immunostained to visualize BRU. In both (A) and (B)
the upper two sets of panels show the posterior pole of an embryo prior to budding of the
primordial germ cells and the lower panels show the posterior pole of an embryo after budding of
the primordial germ cells. Scale bar in the low magnification image is 50 μm; that in the high
magnification image is 25 μm.
A B
123
3.3.6 Smaug represses translation of the arrest mRNA in the germ plasm
In wild-type embryos, arrest mRNA is highly enriched in the germ plasm and is taken up into
the primordial germ cells when they bud (Rangan et al., 2009; Siddiqui et al., 2012). We showed
previously (Siddiqui et al., 2012) that the amount of arrest mRNA in the germ plasm of smaug
mutant embryos is similar to that in wild type (as assayed by fluorescent in situ hybridization)
and that primordial germ cells purified from 1-to-3 hr old smaug mutant embryos have very
similar levels of arrest mRNA to wild-type primordial germ cells of the same age (as assayed by
gene expression profiling using microarrays).
To assess whether Smaug represses translation of the arrest mRNA in the germ plasm of early
embryos I assayed BRU protein using Western blots (Figure 30B) and immunostains (Figure
30C,D). In wild type, BRU was barely detectable on Western blots in 0-to-1 hour old embryos
and was undetectable at later time points (Figure 30B). Likewise, by immunostain BRU was
detectable weakly in the very early embryonic germ plasm and was almost completely absent by
the time the primordial germ cells budded from the posterior pole (Figure 30B and Figure 32). In
contrast, in smaug mutants, BRU was readily detectable on Western blots throughout the first
three hours of embryogenesis (Figure 30B) and BRU could be seen in the germ plasm of
embryos both prior to and during budding of the primordial germ cells (Figure 30D and Figure
32). These observations were confirmed both in embryos expressing VAS-GFP (Figure 30) and
in ones in which only endogenous VAS was present (Figure 32).
I conclude that Smaug represses translation of the arrest mRNA in the germ plasm.
3.3.7 Over-expression of BRU in the embryo leads to excess OSK, germ plasm and primordial germ cells
Finally, I asked whether over-expression of BRU in a wild-type embryo led to increased OSK,
germ plasm and primordial germ cells. To ensure that the excess BRU was produced exclusively
in the early embryo, I placed the arrest ORF under the control of the smaug promoter, 5’ and
3’UTRs, which we have previously shown serve to repress translation in the oocyte but promote
translation in the embryo (Tadros et al., 2007). To ensure that translation and/or stability of the
transgenic mRNA was not repressed by the Smaug protein, I mutated the five predicted SREs in
the arrest ORF without affecting the amino acid sequence (see Experimental Procedures and
Figure 33).
124
The transgenic mRNA, referred to as smaug 5’UTR-arrest ORF(5xSRE-)-smaug 3’UTR or S-
A(5xSRE-)-S, was translated at high levels throughout the embryo, including the germ plasm
(Figure 34A) and resulted in excess OSK protein in the germ plasm (Figure 34B). Strikingly,
there was a concomitant increase in primordial germ cell number (Figure 34C) similar to that
observed in smaug mutants. The average number of primordial germ cells produced by embryos
from females expressing the S-A(5xSRE-)-S transgene was 43.8+6.8; in smaug mutants the
average was 39.7 and ranged from 37.5 and 45.5.
I conclude that over-expression of BRU in the embryo is sufficient to induce extra primordial
germ cells in numbers very similar to those seen in smaug mutants.
125
Figure 33 Predicted SREs in the arrest mRNA and the mutations introduced into the SREs
for construction of the S-A(5xSRE-)-S transgene.
Five predicted SREs in the arrest ORF (blue) are shown. Nucleotides predicted to form the SRE
stems are underlined. Boxed nucleotides were mutated (red) to disrupt the formation of stem
126
loop structures in SRE1 and SRE4, or to disrupt Smaug binding in SRE2, SRE3 and SRE5. In
SRE2, SRE3, SRE4 and SRE5, the encoded amino acids were unchanged. In SRE1, the mutation
results in a conserved L107F substitution. This substitution is not present in any functional
domain of BRU.
127
Figure 34 Over-expression of BRU protein results in synthesis of excess OSK and
production of extra primordial germ cells.
(A) The S-A(5xSRE-)-S transgenic mRNA results in over-expression of the BRU protein
throughout the embryo with particularly high levels in the germ plasm. Left panels show BRU
(red) and right panels show BRU and nuclei (green). Nuclei have not yet migrated into the germ
plasm of the upper embryo whereas the primordial germ cells are beginning to bud from the
lower embryo. (B) The S-A(5xSRE-)-S transgenic mRNA results in over-expression of the OSK
protein in the germ plasm. Left panels show OSK (red) and right panels show OSK and nuclei
(green). Nuclei have not yet migrated into the germ plasm of the upper embryo whereas the
primordial germ cells are to budding from the lower embryo. (C) Histogram showing the number
of PGCs in embryos expressing the S-A(5xSRE-)-S transgenic mRNA.
128
Figure 35 Pathway for regulation of primordial germ cell number by Smaug.
Smaug protein is synthesized and transported into the germ plasm of the early embryo. There it
represses translation of the arrest mRNA into the BRU protein. BRU in turn functions to
potentiate translation of the osk mRNA, which promotes germ plasm assembly and primordial
germ cell formation. In wild-type embryos, the role of Smaug is to repress germ plasm synthesis
whereas in embryos from smaug mutant females, repression is abrogated, excess OSK and germ
plasm are synthesized, and extra primordial germ cells form.
129
3.4 Discussion
Here we have reported that the Smaug RBP sits at the top of a posttranscriptional cascade that
regulates germ plasm synthesis in the embryo and thus primordial germ cell number. I have
shown that BRU is present transiently in wild-type early embryonic germ plasm whereas, in
smaug mutants, BRU is present at high levels in the germ plasm and persists there until after the
primordial germ cells bud from the posterior pole. Smaug is associated with, and represses
translation of the arrest mRNA, thus down-regulating synthesis of BRU protein in the embryo’s
germ plasm. Furthermore, over-expression of BRU in a wild-type embryo (i.e., in the absence of
a smaug mutation) is sufficient to induce extra germ plasm and primordial germ cells. Since
BRU is a positive regulator of OSK translation, Smaug’s normal role is, therefore, to attenuate
production of germ plasm in the embryo. The amount of germ plasm determines the number of
primordial germ cells that bud from the posterior pole. Thus, Smaug sits at the top of a
posttranscriptional cascade that is active in the embryonic germ plasm and modulates primordial
germ cell number (Figure 35).
3.4.1 Mechanisms of posttranscriptional regulation by Smaug
Smaug was first identified as a translational repressor of nos mRNA in the bulk cytoplasm of
early Drosophila embryos (Dahanukar & Wharton, 1996; Smibert et al., 1996). Smaug acts by
recruiting the eIF4E-binding protein, Cup, to repress translation initiation (Nelson et al., 2004),
as well as the Argonaute 1 protein, which is a component of the microRNA-mediated machinery
but is recruited to nos mRNA in a microRNA-independent manner (Pinder & Smibert, 2013).
Smaug also regulates mRNA stability: by recruiting the CCR4-POP2-NOT deadenylase complex
to the Hsp83 mRNA, Smaug triggers elimination of maternal Hsp83 transcripts from the early
embryo (Semotok et al., 2005; Semotok et al., 2008). While Smaug directs degradation of the
Hsp83 mRNA, it does so without repressing translation of these transcripts (Semotok et al.,
2005). In contrast, while Smaug is essential for translational repression of nos transcripts in early
embryos, its role in elimination of these transcripts is more limited: Hsp83 transcripts are
completely stabilized in smaug mutant embryos but nos transcripts are only partially stabilized
(Semotok et al., 2005; Semotok & Lipshitz, 2007). For both Hsp83 and nos mRNAs, Smaug’s
role is restricted to the bulk cytoplasm. Indeed, mechanisms exist to protect these transcripts
from degradation in the germ plasm as well as, in the case of nos, to relieve translational
130
repression therein (Bashirullah et al., 2001; Bashirullah et al., 1999; Gavis & Lehmann, 1994;
Smibert et al., 1999; Zaessinger et al., 2006).
Here I have confirmed that the arrest mRNA is a direct target of Smaug, consistent with my data
reported in Chapter 2. Posttranscriptional regulation of the arrest mRNA exhibits similarities to
and differences from the two previously known targets, Hsp83 and nos. Maternally provided
arrest transcripts are present throughout the early embryo but are enriched in the germ plasm,
resembling nos in this regard. Elimination of arrest transcripts from the bulk cytoplasm at 2-to-3
hours of embryogenesis is absolutely dependent on Smaug (see Figure 31), in this regard
resembling Hsp83 but not nos. As is the case for nos, translation of the arrest mRNA in the bulk
cytoplasm is repressed. However, unlike nos, Smaug is not essential for this repression, which
must require one or more as-yet-unknown factors.
Similar to both Hsp83 and nos, arrest transcripts are protected from degradation in the germ
plasm and are taken up into the primordial germ cells when they bud (Rangan et al., 2009;
Siddiqui et al., 2012). I have shown here that Smaug represses translation of arrest mRNA in the
germ plasm. In contrast, Smaug does not repress translation of nos mRNA in the germ plasm;
indeed, although both Smaug protein and nos mRNA are present in the germ plasm, it has been
shown that OSK protein abrogates Smaug’s ability to repress nos (Dahanukar et al., 1999; Gavis
& Lehmann, 1994; Smibert et al., 1996; Zaessinger et al., 2006). Taken together, and consistent
with previously published work (Dahanukar et al 1996; Smibert et al 1996; Bergsten and Gavis
1999), our results suggest that, in addition to its SREs the nos mRNA is likely to contain one or
more cis-elements that direct relief of translational repression in the germ plasm. This model
contrasts with previously proposed models suggesting that no other cis-acting elements are
required to activate nos translation in the germ plasm (Jeske et al., 2011; Zaessinger et al., 2006).
I speculate that the arrest mRNA lacks these additional cis-elements and, therefore, translation
of the arrest mRNA remains repressed in the germ plasm.
131
3.4.2 Regulation of primordial germ cell number and embryonic pattern
During the past several decades, the molecular hierarchy that directs assembly of the germ plasm
at the posterior pole of the developing Drosophila oocyte has been defined in detail (Mahowald,
2001; Rongo et al., 1997). Briefly, osk mRNA is transported to the posterior pole of the oocyte
by a microtubule-based mechanism. During transport, osk translation is repressed. After
localization, this repressive mechanism is relieved and local synthesis of OSK protein occurs.
This, in turn, nucleates production of additional germ-plasm components (VAS, Tudor, etc.) and
assembly of functional germ plasm. Not only is osk mRNA translation repressed during its
transport to the posterior pole of the oocyte, but localization is inefficient with 85% of osk
mRNA remaining in the bulk cytoplasm of early embryos (Bergsten & Gavis, 1999); thus
translational repression is required during transport of the transcripts that become localized as
well as for the transcripts that remain unlocalized.
One of the best-characterized repressors of osk translation is the BRU RBP (Chekulaeva et al.,
2006; Kim-Ha et al., 1995; Webster et al., 1997). BRU binds cis-elements known as BRU
response elements (BREs) in the osk 3’UTR. These reside in three clusters, referred to as A, B
and C. More recently, it has been shown that cluster C also participates in translational activation
of osk mRNA in the germ plasm at the posterior of the late-stage oocyte (Reveal et al., 2010). In
that study, however, it was not possible to assess whether the activation of osk was directed by
BRU itself since mutants arrest development earlier in oogenesis.
Here I have shown that BRU is indeed a positive regulator of osk mRNA translation and that it is
capable of potentiating osk translation in the embryo, not just the oocyte. Whereas BRU is
detectable in the germ plasm at the posterior of the very early embryo it rapidly disappears
therefrom. Synthesis of BRU protein in the germ plasm of the embryo is attenuated by the
Smaug RBP, which is enriched in the germ plasm of the early embryo. This, in turn, results in
production of little or no additional OSK protein or additional germ plasm in embryos. Thus,
Smaug sits at the top of a posttranscriptional hierarchy that regulates germ plasm in the embryo
(Figure 35).
It has previously been shown that reducing the dose of the osk gene during oogenesis leads to
underproduction of germ plasm and, subsequently, budding of a reduced number of primordial
132
germ cells in the embryo, whereas overexpression of OSK during oogenesis results in excess
primordial germ cells (Ephrussi & Lehmann, 1992; Smith et al., 1992). Our data represent the
first evidence that posttranscriptional regulation of BRU and OSK synthesis in the early embryo,
rather than just during oogenesis, is required to regulate the amount of germ plasm and, thus, the
number of primordial germ cells that are produced by the embryo.
During gonad development, germline-soma interactions regulate the number of cell divisions
undergone by the primordial germ cells during the larval period; if the number of primordial
germ cells that form in the embryo is smaller than usual, then extra divisions occur (Gilboa &
Lehmann, 2006). This feedback mechanism ensures that germline stem cells are present in
sufficient numbers by the end of the larval period to populate the stem-cell niches that form at
that time. Although those experiments did not analyze whether, if extra primordial germ cells
form in the early embryo, they undergo fewer divisions during the larval stages, this seems
highly probable.
Why, then, does the regulatory pathway I have defined here, exist? The germ plasm controls, not
only primordial germ cell formation but also anterior-posterior axis formation in the embryo.
Thus, severe body pattern defects result when the dose of the osk gene is increased (Ephrussi &
Lehmann, 1992; Smith et al., 1992). I speculate that the Smaug-BRU-OSK pathway has evolved,
not specifically to attenuate the number of primordial germ cells that form but to ensure correct
body pattern. An implication is that, during the evolution of organisms such as Drosophila –
where the germ plasm mechanistically links primordial germ cell formation and embryonic
patterning – a balance has been achieved that optimizes the amount of germ plasm to ensure that
both developmental processes can proceed unhindered. By regulating the amount of germ plasm
that is synthesized in the early embryo, Smaug achieves this balance.
3.5 Experimental procedures
3.5.1 Drosophila culture and mutants
All stocks were maintained at 25oC. The following fly strains were used: w
1118, Vasa-
GFP/CyO ; PrDr/TM3,Sb (Johnstone and Lasko 2004), Vasa-GFP/CyO ; smg1/TM3,Sb and
Df(3L)Scf-R6/TM3,Sb (Bloomington) ; Dhc6-6
, Dhc6-10
, Dhc6-12
(Gepner et al., 1996); Jupiter-
GFP/TM3, Sb (Bloomington); Histone-RFP/CyO (Bloomington); mCherry-Smaug7/FM7 ;
133
Venus-SMG6/CyO ; mCherry-Smaug8/TM3, Sb. The mCherry-Smaug7, mCherry-Smaug8, and
Venus-Smaug6 transgenes completely rescue the smg1 mutation.
3.5.2 Generation and molecular analysis of new smaug alleles
The smg30
and smg47
alleles were generated via imprecise excision of a P element using standard
methods (Salz et al., 1987). The P element (GE21229) is inserted 2,499 base pairs 5’ of the
smaug translation start codon (20 base pairs downstream of the transcriptional start site of the
smaug-RB isoform). All isoforms are defined as described at http://flybase.org/. The original smg
allele showed homozygous maternal effect lethality (Dahanukar et al., 1999) and we recovered
six excision lines demonstrating this phenotype. The extent of the deletion in these six lines was
determined via PCR analysis of genomic DNA. Two of the lines, smg30
and smg47
, showed
deletions removing large portions of the smg gene, but not effecting the neighboring upstream
and downstream genes, CG5087 and CG5280, respectively. Sequencing revealed that the smg30
allele is a 4,528 base pair deletion of the smg gene beginning 2,494 base pairs 5’ of and ending
2,034 base pairs 3’ of the translation start codon. The excision event that created the smg30
allele
left behind a 1,140 base pair fragment of the P element. The smg30
deletion removes 2,034 of
2,997 base pair of the open reading frame in four of five smg isoforms (RA, RB, RC, RE) and
2,034 of 3,327 base pairs in the RD isoform. The smg47
allele is a 5,542 base pair deletion
beginning 2,483 base pairs 5’ of and ending 3,059 base pairs 3’ of the smaug start codon. This
deletion leaves 325 base pairs of the open reading frame for the smg-RD isoform and leaves 39
base pairs of the open reading frame in the other four smg transcript isoforms.
3.5.3 Construction of transgenes and production of transgenic flies
The N-terminal mCherry-Smaug and Venus-Smaug constructs were generated by
replacing the N-terminal TAP-tag from the TAP-tagged Smaug plasmid (Semotok et al., 2005)
with either mCherry or Venus tags. The resulting plasmid was digested with NotI and the smaug
genomic fragment with its regulatory regions and an N-terminal mCherry- or Venus-tag was
subcloned into pCaSperR4 (Thummel & Pirrotta, 1991). Transgenic flies were generated by
random insertion of the transgene in the germ line.
The Smaug genomic construct, smaug5’UTR-smaugORF-smaug3’UTR plasmid with
attB, referred to as S-S-S, was constructed by using the S-BsiWI-S plasmid (Tadros et al., 2007).
134
A polylinker containing p53 and FLAG tags and unique restriction sites (AscI and PmeI) was
inserted using the BsiWI site. The Smaug ORF was PCR amplified and was inserted as an AscI-
PmeI fragment. The resulting plasmid was digested with NotI and the complete smaug5’UTR-
smaugORF-smaug3’UTR with p53 and FLAG tags was cloned in pCaSpeR4 (Thummel &
Pirrotta, 1991) plasmid with attB sites.
The S-A(5xSRE-)-S constructs were generated by replacing the Smaug ORF in the S-S-S
construct in pCaSpeR4 with the 5x SRE- arrest ORF sequence (Figure 33). The SRE mutations
were introduced into the arrest ORF by PCR using overlapping primers that carried the
mutations with the arrest cDNA as a template. Transgenic flies were generated by insertion of
the transgene into the phiC31 attP3 landing site (Markstein et al., 2008).
3.5.4 Confocal microscopy and live imaging
For live imaging, flies were grown in small cages made from 100 ml plastic beakers fitted with a
4 cm petri-dish agar plate on the open side of the beaker. Embryos were collected for 15 minutes
to 1 h, depending on the stage to be imaged, dechorionated using 50% bleach, washed and
mounted as a monolayer of embryos in a drop of halocarboon oil on the outer surface of the
petriPERM dishes (Sigma-Aldrich) and covered with a coverslip before imaging. All movies
were captured using a 40x Planapochromat objective (NA 1.5) on a Nikon Eclipse TE2000-E
inverted microscope with a Perkin-Elmer spinning disk and a Hamamatsu Orca-EM CCD
camera. Movies were processed using the Metamorph software package (Universal Imaging,
Downington, Pennsylvania).
Antibodies used for immunostains were: guinea-pig anti-Smaug (1:500) (Tadros et al., 2007);
rabbit anti-GFP (1:500) (Abcam, Inc.); rabbit anti-VAS (1:100) (a gift from P. Lasko, McGill
University); rabbit anti-OSK (1:250) (Kim-Ha et al., 1995); rabbit anti-NOS (1:1000, newly
made using a Nanos antigen as previously described by Smibert et al., 1999).
3.5.5 Fluorescence recovery after photobleaching (FRAP)
Embryos were mounted on petriPERM dishes as described above. Samples were photobleached
using an argon laser attached to the WAVEFX spinning disk (Quorum Technologies). A small
circular area approximately the size of the polar granule to be bleached was selected around a
single polar granule for photo bleaching. Images were captured using the 40X/1.4 oil immersion
135
lens (Carl Zeiss, Inc.), an EM charge-coupled camera (Hamamatsu Photonics), and Volocity
Imaging software (Perkin Elmer). Samples were photobleached for 2 seconds and images were
collected at two frames/second for a total of four minutes. Images were collected in a single Z
plane before, during and after photobleaching. Signal recovery was measured using ImageJ
(National Institutes of Health) for polar granules that stayed in the focal plane for the duration of
the image acquisition. These values were corrected for background by subtracting the mean
fluorescence of a similar size area outside of the embryo (Pi). To correct for general bleaching of
the embryo from imaging, fluorescence was measured for a large area covering the entire pole
cell corrected for the background as above (Bi). The corrected fluorescence intensity was
measured by dividing Pi / Bi and normalized to the time point just before the bleaching and
plotted using Excel (Microsoft). Recovery rates were calculated from the slopes of the best-fit
lines for the first 30 sec after photobleaching.
3.5.6 Analysis of Smaug particle movement
Smaug particle movement was measured using the MetaMorph software. Briefly, individual
particles that remained visible for four consecutive frames were tracked using MetaMorph
software package. The particle velocity was calculated by measuring the distance traveled by the
particle divided by the transit time. The directionality of each particle was noted as posterior if
the particle moved towards the pole cells. Among the 47 particles that were tracked, 28 particles
were directed towards the posterior while 19 particles moved towards the anterior of the embryo.
3.5.7 Cryo-immunogold electron microscopy
Embryos were collected 140-160 min after egg-laying, dechorionated in 50% bleach, rinsed in
dH2O, and then shaken for 30 min in a biphasic solution containing 8 mL of heptane and 2.75
mL of “fix mix” (4% paraformaldehyde in 0.1M Sorenson’s phosphate buffer). Fixed embryos
were then transferred to a 1:1 heptane:methanol solution, shaken for 20 sec, rinsed five times in
fix mix, then fixed for another 90 min in fix mix. Embryos were then stored in 1%
paraformaldehyde in 0.1M Sorenson’s phosphate buffer overnight.
The following steps for cryo-sectioning and labeling were modified from a published protocol
(Peters et al., 2006). After fixation embryos were washed twice with 0.15 M glycine in PBS,
then incubated in 1%, 5% and 12% gelatin in PBS at 37o C for 20 min each, and embedded in
136
12% gelatin. Embedded embryos were then cryo-protected in 2.3 M sucrose in PBS overnight.
Embryos were then frozen on aluminum pins by plunging into liquid nitrogen. Sections were cut
at –100oC at a thickness of 75 nm using a Leica Ultracut UCT ultramicrotome with a cryo-
chamber attachment. Sections were picked up with a loop using a 1:1 mixture of 2%
methylcellulose:2.3 M sucrose. Sections were placed on Formvar coated nickel grids.
For immunolabeling, grids were floated on drops of PBS for 10 min to rehydrate sections, and
then floated on drops of 0.15 M glycine in PBS for 10 min to quench aldehydes. Sections were
blocked with 5% cold-water-fish gelatin in PBS for 30 min, then incubated in primary antibody,
guinea pig anti-SMG (1:40), rabbit anti-VAS diluted (1:20) in 1% cold-water-fish gelatin
containing 0.1% Tween in PBS for 1 hr. Sections were then washed five times for 2 min each in
PBS, followed by incubation with secondary antibody gold conjugate (goat anti-guinea pig, 10
nm gold, goat anti-rabbit 15 nm gold) diluted 1 in 10 in 1% cold-water-fish gelatin, 0.1% Tween
in PBS for 1 hr. Sections were then washed five times for 2 min each in PBS, and fixed for 10
min with 2% glutaraldehyde in phosphate buffer. Sections were then washed twice for 2 min
each in PBS, followed by five times 2 min each in dH2O. Sections were then incubated on drops
of 2% methylcellulose, 0.4% uranyl acetate for 10 min. Grids were picked up in loops and the
excess methylcellulose was wicked off. Sections were examined using an FEI Tecnai 20
transmission electron microscope. Images were captured using a Gatan Dualview digital camera.
3.5.8 Three-dimensional reconstruction and counts of primordial germ cells and PH3-labeled nuclei
Series of confocal sections through the primordial germ cells of Stage 4 to 5 (nuclear cycle 12-
14) embryos were collected using an Olympus IX81 inverted fluorescence microscope equipped
with a Hamamatsu Back-Thinned EM-CCD camera (9100-13) and Quorum spinning disk
confocal scan head. The unit is equipped with 4 separate diode-pumped solid-state laser lines
(Spectral Applied Research: 405nm, 491nm, 561nm, 642nm), an ASI motorized XY stage, an
Improvision Piezo Focus Drive and a 1.5X magnification lens (Spectral Applied Research). The
equipment is driven by Volocity 5.5.0 acquisition software (Perkin Elmer) and runs on a Dell
PC. It is equipped with 10x/.4NA, 20x/.75NA, 40x/.95NA, 60x/1.2NA(w) and 60x/1.35NA
objectives. It has the following emission filters; 460/50, 525/50, 593/40, 620/60, 700/75.
137
In order to count the total number of primordial germ cells and the number with PH3-positive
nuclei, the confocal stacks were imported into Volocity 5.5.0 software (PerkinElmer), for three-
dimensional reconstruction. Embryos were then viewed in the ‘3D Opacity’ mode allowing one
to freely turn, rotate and spin individual embryos in each plane in order to facilitate counting
primordial germ cell number, PH3-positive primordial germ cell nuclei, and to distinguish which
PH3-positive signal was in the primordial germ cell nuclei versus in somatic nuclei (in smaug
mutants, where somatic nuclei continue to divide at these stages).
3.5.9 Immunoprecipitation of Smaug together with its bound mRNAs
Immunoprecipitation of Smaug followed by RNA extraction was performed as previously
described (Semotok et al., 2005) with minor modifications. Embryos collected at 0-3 hr after egg
lay were dechorionated in 50% bleach and homogenized in lysis buffer (150 mM KCl, 20 mM
HEPES pH 7.4, 1 mM MgCl2, 1 mM DTT, 1x protease inhibitor cocktail (Bioshop)). Extracts
were centrifuged for 10 min at 4°C, and the supernatant was supplemented with urea to a 2 M
final concentration. Protein A beads were pre-incubated with either guinea pig anti-Smaug
antibody or normal guinea pig serum followed by four washes. Pre-cleared Beads were then
incubated for 2 h at 4°C followed by four washed. RNA was extracted using Trizol reagent
(Invitrogen).
3.5.10 Reverse transcription-quantitative PCR (RT-qPCR)
cDNAs used for quantitative PCR were synthesized using SuperScript II reverse transcriptase
(Invitrogen). Quantitative PCR reactions were carried out using a BioRad C1000 Touch Thermal
Cycler with a CFX384 Real-time PCR system. Total transcript levels of arrest and Smaug-bound
arrest were determined using the arrest-F primer (TGAACGCAAACTCTTTGTGG) and arrest-
R primer (GGCTCCGTGGACTTCAAATA). RpL32 levels used for normalization were
determined by rp49-F primers (AGTCGGATCGATATGCTAAGCTG) and rp49-R primers
(CGATGTTGGGCATCAGATACTG).
3.5.11 Western blotting
Embryos were dechorionated in 50% bleach, and homogenized in lysis buffer (150 mM NaCl, 50
mM Tris pH 7.5, 1 mM DTT, 1x protease inhibitor cocktail (Bioshop)). Extracts were boiled for
5 minutes in 6xSDS loading buffer, and run on 10% SDS-PAGE. Proteins were transfered to
138
PVDF membranes. Blots were blocked in 2% skim milk. Primary antibodies used were rabbit
anti-BRU polyclonal (1:5000; a gift of Paul Macdonald, Austin, Texas), rabbit anti-OSK
(1:2000; a gift of Anne Ephrussi, Heidelberg), and mouse anti-Tubulin (1:3000).
Chemiluminescence was detected using the Pierce ECL Western blotting reagent and a BioRad
ChemiDoc system.
139
Chapter 4 Conclusions and Future Directions
4
4.1 Conclusions
In this thesis, using genome wide approaches, I have found that Smaug is a direct regulator of a
large set of maternal transcripts during early Drosophila embryogenesis, and that a significant
fraction of these are targeted by Smaug for both degradation and translational repression. I have
demonstrated that a large subset of Smaug targets is transcripts localized to the posterior of the
embryo, and I discovered novel functions of Smaug including roles in regulation of metabolism,
lipid droplet function, protein folding and protein stability.
In addition, I have performed a detailed characterization of one of these newly discovered Smaug
targets: the arrest mRNA, which encodes the BRU RBP. Here I have shown BRU protein is only
present transiently in the germ plasm of the early embryo, but in smaug mutants, it is expressed
at high levels in the embryo’s germ plasm and primordial germ cells. I have demonstrated that
arrest mRNA co-purifies with Smaug protein and that Smaug down-regulates the synthesis of
BRU protein in the embryo’s germ plasm. Furthermore, over-expression of Arrest/Bru in a wild-
type embryo is sufficient to induce formation of extra primordial germ cells. This result suggests
that, in addition to its role in oogenesis, BRU is able to potentiate OSK protein production during
embryogenesis. In conclusion, Smaug sits at the top of a post-transcriptional pathway that
modulates germ plasm and germ cell formation in the embryo.
4.2 Future Directions
In future studies, I would follow up on several sets of newly identified Smaug mRNA targets,
and further study the mechanisms of action and biological functions of Smaug. In the following
sections, I propose a general strategy to validate and characterize novel Smaug mRNA targets. I
then focus on selected Smaug mRNA targets and outline experiments to study these targets to
gain more information regarding the mechanisms of Smaug-dependent regulation as well as the
functions of Smaug during early embryogenesis.
140
4.2.1 A general strategy for the investigation of de novo Smaug mRNA targets
4.2.1.1 Construction of SRE mutants
Smaug had previously been shown to bind to, and directly regulate, two maternal mRNAs: nos
and Hsp83. In both cases, mutation of the SREs within these transcripts provided evidence that
they were direct Smaug targets. Therefore, I propose to mutate SREs within new Smaug mRNA
targets to validate them as direct targets and to explore the biological significance of Smaug’s
ability to regulate them.
Here I will describe a general approach for creating SRE mutant constructs. First, I would
generate a transgenic construct encoding the wild-type version of the RNA of interest, which I
will refer to as X, expressed using the female germline-specific Gal4/UAS system (Rorth, 1998)
and I will refer to such a construct as X-SRE+ hereafter. An X-SRE
+ construct would be used as a
template for mutating transcript X’s SREs, and serve as a control in later experiments. X-SRE-
constructs would produce transcripts with mutations that either disrupt formation of the SRE
stem or change nucleotides within loop, which are required for Smaug binding. When the SREs
reside in the ORF, best efforts would be made to ensure that X-SRE- mRNA encodes the wild-
type protein sequence. All constructs will carry a tag, which would allow me to distinguish the
transgenic products from the endogenous ones in RT-qPCR assays, Northern and Western blots
and in immunostaining.
One problem with the approach outlined above is that expression of X-SRE+ or X-SRE
- using the
Gal4/UAS system might disturb oogenesis. In this case, I could restrict transgenic expression to
early embryogenesis using Smaug regulatory elements, as was done for Arrest in Chapter 3.
4.2.1.2 Validation of Smaug mRNA targets
To validate that an mRNA of interest is directly bound by Smaug through SRE recognition, I
would perform RIP using anti-Smaug antibodies and embryo extracts expressing either X-SRE+
or X-SRE- transgenes, followed by RT-qPCR. If Smaug directly interacts with transcript X, and
the interaction is through recognition of X’s SREs, then the X-SRE+ transcript should be enriched
in Smaug immunoprecipitates while X-SRE- should not.
141
In addition to validating direct interactions, I would also explore Smaug-mediated regulation of
mRNA X. Smaug is expected to be a repressor of gene expression, through either transcript
degradation and/or translational repression. Therefore, in embryos expressing an X-SRE- mRNA
I would expect to find over-expression of X’s mRNA and/or protein. To evaluate the level of
transgenic transcripts, I could use Northern blot analysis or RT-qPCR. To measure and compare
transgenic X protein level, I could use Western blot analysis. To assess whether Smaug might
spatially regulate X’s expression I would use FISH to compare localization of the X-SRE+ and X-
SRE- mRNAs and immunostaining to compare the localization of the proteins encoded by the X-
SRE+ or X-SRE
- mRNAs.
4.2.2 What are the mechanisms of Smaug’s regulation?
In Chapter 2, I have shown that a large fraction of Smaug-bound mRNAs are subject to Smaug-
dependent transcript degradation and translational repression, which suggests that Smaug’s
default function is to both repress translation and induce degradation of its targets. Regulation of
Hsp83 and nos, the two well-characterized targets of Smaug, both involve the CCR4/POP/NOT
deadenylase complex. For Hsp83 mRNA, Smaug recruits the CCR4/POP/NOT deadenylase
complex to induce transcript degradation but not translational repression (Semotok et al., 2005;
Semotok et al., 2008). In the case of nos CCR4/POP/NOT-mediated deadenylation contributes to
translational repression (Jeske et al., 2006; Jeske et al., 2011; Zaessinger et al., 2006) but Smaug
plays only a limited role in transcript decay (Semotok et al., 2005; Tadros et al., 2007). Inferring
from these two examples, I propose one of Smaug’s default mechanisms would be to recruit the
CCR4/POP/NOT deadenylase complex to induce transcript deadenylation, which in principle
could initiate both transcript degradation and translational repression.
In addition, considering that a fraction of Smaug’s targets depend on Smaug for either
translational repression or transcript degradation, but not both, it is likely that the differential
regulation of these transcripts involves additional co-factors or cis-acting elements. As
mentioned above, it is known that Hsp83 and nos mRNAs are differentially regulated despite
both suffering deadenylation. For Hsp83, Smaug binding triggers its mRNA decay, but its
translation is not repressed (Semotok et al., 2005). This may be a consequence of a translational
enhancer element, which has been mapped in its 3’UTR, and is known to bind three RBPs - DP1,
HRP48 and PABP (Nelson et al., 2007). For nos mRNA, Smaug is required to repress its
142
translation, but its mRNA is only partially stabilized in a smaug mutant (Semotok et al., 2005;
Semotok et al., 2008), indicating the other, Smaug-independent cis-acting elements may
destabilize nos mRNA. It has been reported that piRNAs bind to an elemant in the nos 3’UTR to
destabilize the transcripts in a Smaug-independent but CCR4-NOT-deadenylase-dependent
manner (Rouget et al., 2010). These two examples are consistent with the idea that for transcripts
categorized as “only degraded by Smaug” or “only repressed by Smaug”, additional cis-acting
elements are likely to be involved in their regulation.
Taken together, a plausible model is that Smaug recruits the CCR4/POP/NOT deadenylase
complex to induce deadenylation, and initiates both translational repression and degradation of
its target transcripts; the presence of additional co-factors and cis-acting elements could act in
concert with Smaug or overcome Smaug’s repression, and allow more specific regulation of
Smaug’s targets.
This model predicts that, in the absence of the additional cis-elements, Smaug binding should be
sufficient to trigger both translational repression and transcript degradation. To test this
hypothesis, I would utilize reporters (e.g., luciferase, GFP) containing an array of SREs, and
assess whether insertion of SREs into a model, stable and actively translated reporter mRNA is
sufficient to target it for both translational repression and transcript degradation.
Luciferase reporters carrying SREs inserted in their 3’UTR have been previously constructed to
examine Smaug-mediated regulation (Semotok et al., 2005; Semotok et al., 2008; Smibert et al.,
1999). However, in Chapter 2, I have found that the majority of predicted SREs are present
within ORFs, suggesting that the SRE-array should be inserted into the reporter transgene’s
ORF. Therefore, in addition to 3’UTR reporters, I would construct reporters that include SREs in
their ORF and express them under a stable mRNA’s 5’ and 3’UTRs (e.g., αtub84B). My
hypothesis predicts that these ORF SRE+
reporters are both unstable and translationaly repressed.
Since previous study showed that inserting SREs into the tubulin 3’UTR induced decay, I would
also test the existing 3’UTR SRE+
reporters for translational repression (Semotok et al., 2005).
To test my hypothesis, I would use qRT-PCR and Western blot analysis to measure the modified
reporters’ mRNA levels and protein levels compared to controls. Although there is a concern
regarding SRE insertion in an ORF, because this might disrupt folding or activity of the reporter
protein, by using Western blot analysis to measure protein level, I could neglect this problem. In
143
addition, I would use polysome profiling to determine the reporter mRNA’s translational status.
Note that, even though the reporter may be translated during oogenesis since it is under tubulin
UTR control, the polysome experiments will be carried out in embryos, thus assaying repression
by Smaug at this stage. This experiment will be carried out in a wild-type embryo background,
and if the results are as I predict, I would expect to find reduce mRNA and protein levels in SRE
containing reporter mRNA. I would also expect to find reduced association of SRE+ containing
reporter mRNAs with polysomes compared to SRE- controls, indicating translational repression.
I also hypothesize that the dual regulation of stability and translation by Smaug is mediated by
the CCR4/POP/NOT deadenylase complex. To test this prediction, I would examine the
modified reporters in a CCR4 mutant background and perform the same experiments described
above.
The second part of the model predicts the “degradation only” and “translational repression only”
targets contain additional cis-elements that are responsible for their differential regulation. Such
elements might be contained anywhere in a transcript. For simplicity my efforts would search for
them in 3’UTR sequences since this is where they map in Hsp83 and nos (Nelson et al., 2007;
Rouget et al., 2010). To do this, I would exchange the previously proposed SRE+
reporters’
3’UTR with ones that are selected from a list of “only degraded by Smaug” or “only repressed
by Smaug” mRNAs and ask if these new reporters display differential regulation. For those that
do I would generate a series of mutant constructs to identify small regions of the 3’UTR that are
necessary and/or sufficient for differential regulation. Subsequent experiments would focus on
identifying the trans-acting factor(s) that act through any such a region to modulate Smaug
function and then study the molecular mechanisms that are involved. The design of constructs to
search for cis-acting elements that modulate Smaug function found within 5’UTR or ORF
sequences would be considerably more complicated. As such, it would be unlikely that I would
search for such elements.
4.2.3 Smaug’s mRNA targets and their biological roles
In this section, I will explore some of the newly predicted functions of Smaug, which are derived
from the functional annotation of Smaug’s direct mRNA targets. I propose experiments to verify
and study Smaug’s novel functions in: 1) embryonic body patterning and germ line specification;
2) protein degradation and cell cycle progression; and 3) metabolism.
144
Previous studies of Smaug’s biological function have been focused on examining the phenotype
of smaug mutants. However, I have shown that Smaug targets a large set of transcripts for
regulation, with some of them encoding post-transcriptional regulators. Therefore, it is difficult
to distinguish direct functional defects from secondary effects in smaug mutants. One of the best
solutions to address this concern is to mutate all of the SREs within a specific transcript and then
assess the effects that expression of such a transgene has on the embryo. Therefore, in the
following sections, I propose to use both smaug mutants and SRE- mutant transgenes to dissect
each distinct function of Smaug. One of the drawbacks of using SRE- transgenes is that some of
the Smaug bound mRNAs encode parts of a complex or members of various pathways. For
example, Smaug regulates multiple mRNAs encoding the proteasome regulatory complex and
the glycolytic pathway. In both of these examples, SRE mutations in a single target transcript
might not be sufficient to induce a phenotype. If I encounter this problem, I would express
multiple functionally related transcripts with all SREs mutated in the same embryo.
4.2.3.1 Role of Smaug in embryonic body patterning and germ line specification
Many of Smaug’s target mRNAs are known to play important roles in embryonic body
patterning and germline specification. For example, in Chapter 3, it is shown that Smaug targets
maternal arrest mRNA in the early embryo for translational repression and transcript
degradation, and in so doing regulates germ plasm assembly and germ cell formation. In
addition, the Smaug-BRU-OSK pathway may regulate, not only germ cell number, but also
embryonic body pattern.
Another mRNA target of Smaug that plays a role in embryonic body patterning is Drosophila
Bicaudal C. BICC is a KH domain containing RBP required for anterior-posterior patterning
during oogenesis (Mahone et al., 1995; Mohler & Wieschaus, 1985). In Drosophila,
heterozygous BicC mothers produce embryos with defects that vary in severity, and include
bicaudal embryos in which anterior structures are replaced by a mirror-image duplication of
posterior structures (Mohler & Wieschaus, 1985). Such phenotypes can result from inappropriate
OSK expression and it is, indeed, known that BICC represses osk translation at the anterior of the
oocyte (Saffman et al., 1998). In the early embryo, BicC mRNA is present throughout the
embryo and is more concentrated at the posterior pole (Mahone et al., 1995). However, due to
145
the strong phenotype produced by complete loss of BicC function during oogenesis, BicC’s
function during embryogenesis remains unclear.
4.2.3.1.1 Smaug’s role in pole plasm and pole cell regulation
I hypothesize that Smaug directly binds to Arrest/Bru and BicC transcripts at the posterior pole
of the early embryo and represses their expression. In a wild-type embryo, repression of BicC by
Smaug permits the activation of osk at the posterior pole, and repression of Arrest/Bru prevents
over-activation of osk. Together, Smaug, BicC and Arrest/Bru would, thus, constitute a pathway
to fine tune and balance osk expression in the pole plasm of the early embryo (Figure 36).
To examine Smaug’s regulation of BicC in the embryo posterior, and its impact on germ cell
formation, I would carry out similar experiments to those performed to examine Smaug’s
regulation of Arrest/Bru and osk in Chapter 3. If my hypothesis is correct, I would expect a
BicC-SRE- embryo to have increased level of BICC protein and reduced levels of OSK protein at
the posterior; I would also expect a BicC-SRE- embryo to have a reduced number of pole cells.
Another possible outcome of the experiment is that Smaug regulates BicC, but BicC is not
involved in the regulation of osk at the posterior of the embryo. If this is the case, I would screen
for the mis-regulation of other BicC targets in BicC-SRE- embryos, and look for functional
defects. A list of 89 BICC target mRNAs has been defined through RIP-Chip experiments
(Chicoine et al., 2007). The stability and translation of specific targets could be assessed via RT-
qPCR, Northern and Western blots and polysome gradients, while genome-wide surveys could
be conducted using microarrays or next-generation sequencing on their own or in combination
with polysome gradients.
I would also assess osk expression in embryos carrying both the S-A(5xSRE-)-S and BicC-SRE
-
mRNAs. If BicC indeed represses osk expression at the posterior of the embryo, I would predict
that this would result in a decrease in pole cell number relative to the number seen in the
prescence of the S-A(5xSRE-)-S transgene on its own.
4.2.3.1.2 Smaug’s role in embryonic body patterning
So far, I have focused on examining the role of Smaug in regulating germ cell establishment. In
the future, I would further examine the role of Smaug’s regulation pathways in the establishment
of embryonic body patterning. To test if Smaug’s regulation of Arrest/Bru and BicC during
146
embryogenesis is required for proper embryonic body patterning, I would analyze the cuticles
produced by the transgenic embryos. Based on my model I would expect that embryos
expressing either S-A(5xSRE-)-S would show body patterning defects that are associated with
OSK overexpression compared to the SRE+ versions of these transgenes. In contrast, the BicC-
SRE- transgene would show phenotypes similar to those associated with reduced OSK
expression, while the expression of both S-A(5xSRE-)-S and BicC-SRE
- transgenes together
would be predicted to mitigate the phenotype associated with each on their own.
147
Figure 36 Proposed new pathway for regulation of primordial germ cell number by Smaug.
In the germ plasm, Smaug protein represses translation of arrest mRNA and BicC mRNA into
BRU and BICC protein. In term, BRU protein activates the translation of osk mRNA into OSK
protein; BICC protein functions to repress the osk translation. In concert, Smaug’s regulation
fine-tunes OSK protein production; hence, germ plasm synthesis and primordial germ cell
formation.
148
4.2.3.2 The role of Smaug in ubiquitin-mediated proteolysis
In Chapter 2, I have shown that almost all of mRNAs encoding components of the 19S
regulatory particle (RP) are targeted by Smaug for regulation. The 26S proteasome is a highly
conserved protein complex that is responsible for the proteolysis of polyubiquitinated proteins
and the elimination of damage and misfolded protein, as well as proteins that are involved in cell
cycling, DNA damage and metabolism (Forster et al., 2010; Saeki & Tanaka, 2012). The 19S RP
is one of the two complexes that make up the 26S proteasome; it enhances the peptidase activity
of core proteolytic particle when they are in a complex. Therefore, I hypothesize that Smaug co-
coordinates the repression of 19S RP subunits expression and, by doing so, regulates ubiquitin-
dependent proteolysis and targeted protein turnover during early embryogenesis.
4.2.3.2.1 The role of Smaug in regulating proteasome activity
To examine the role of Smaug in ubiquitin-mediated proteolysis, I would measure the activity of
the 26S proteasome in early wild-type and smaug mutant embryos. This could be done by
assaying the chymotrypsin-like activity of the 26S proteasome using Suc-LLVY-AMC substrate
in a microtitre plate-based assay (Bader et al., 2011; Holzl et al., 2000). In addition, I could use
an in-gel assay where extracts are electrophoretically separated using non-denaturing gels, and
proteasome activity is detected by overlaying the substrate (Holzl et al., 2000). The in-gel assay
separates the 20S and 26S proteasome complexes, hence allows for the detection of the activities
of these complexes separately. I would ensure that any activity detected is proteasome-specific
through the addition of proteasome inhibitors such as MG132 or lactacystin (Fenteany et al.,
1995; Tsubuki et al., 1996). If Smaug co-represses proteasome RP subunits expression and thus
down-regulates overall proteasome activity, I would expect to find increased proteasome-specific
proteolytic activity in smaug mutant embryos.
Next, to assess the function of Smaug in modulating ubiquitin-dependent protein degradation, the
level of total polyubiquitinated protein in embryo extracts would be examined (Fredriksson et al.,
2012). I would compare the total level of ubiquitinated protein in wild-type, and smaug-mutant
embryos using Western blot analysis. If Smaug is involved in repressing ubiquitin-dependent
protein degradation, I would expect to find an increase in polyubiquitinated protein in the smaug-
mutant embryos.
149
If smaug mutants show upregulated proteasome activity and/or increases ubiquitinated protein
levels, I would perform these assays on embryos expressing SRE- versions of RP component
transgenes. This would allow me to confirm these effects are due to direct regulation of
proteasome expression and not due to an indirect consequence of the smaug mutant phenotype. It
would also allow me to ask if there are biological consequences to specifically disrupting
Smaug’s ability to regulate proteasome function. Embryonic hatching would assess any lethality
associated with mutant transgenes while examination of embryonic cuticles will uncover
patterning defects. These transgenes might also disrupt the cell cycle and assays to detect such
defects are discussed below.
Since Smaug appears to co-regulate multiple subunits of the proteasome regulatory complex, it is
possible that increased proteasome activity and/or increases in levels of ubiquitinated proteins
will require co-expression of SRE- versions of many or all of the Smaug RP component targets.
Indeed, while it would be laborious, it should be possible to co-express SRE- versions of all nine
of the RP mRNAs that my data indicate are direct Smaug targets. Another caveat is that eight of
the remaining 10 RP components that appear not to be direct Smaug targets are upregulated in
smaug mutant embryos at the level of translation and/or stability. Whether or not these mRNAs
represent false negatives from the RIP-Chip experiments is unclear. Nonetheless, the
upregulation of these RP components might be required to see upregulated proteasome activity
and/or increases ubiquitinated proteins levels and failure to detect any changes in upon co-
expression of the SRE- version of the nine direct Smaug RP targets might be explained by such a
complication. Also note that there is a limit of 10 to the number of transgenes that can readily be
combined in the same embryo using the phiC31 site-specific recombination system.
It is also possible that the up-regulation of individual proteasome subunits would not result in an
overall increase of ubiquitin-dependent protein degradation. For example, Rpn10, a Smaug
target, has been shown to be present extra-proteasomally in flies and yeast, and to bind and
sequester polyubiquitinylated proteins, preventing their degradation by the proteasome (Haracska
& Udvardy, 1995; van Nocker et al., 1996). In this case, Smaug might be responsible for
maintaining the proper ratio of 19S RP subunits and the balance between proteasomal and extra-
proteasomal Rpn10s. In other words, if Smaug regulation of Rpn10 is abrogated, then there
might be an increase in extra-proteasomal Rpn10 with consequent sequestration of
polyubiquitinated proteins from the proteasome; thus, I might detect a decrease in overall of
150
ubiquitin-dependent protein degradation. I would, therefore, assay the effects of single RP
subunit SRE mutants before combining them into the same embryo.
4.2.3.2.2 Role of Smaug in modulating ubiquitin-mediated proteolysis and its impact in cell cycle regulation
In addition to the proteasome, Smaug targets other parts of the ubiquitin proteasome system for
regulation. One Smaug-bound mRNA, Apc7, encodes a subunit of the anaphase promoting
complex (APC/C), which is one of the two E3-ligase complexes that are involved in the
ubiquitin-proteasome-dependent degradation of key cell cycle regulatory proteins (Harper et al.,
2002; Kashevsky et al., 2002; Peters, 2006). Thus, the smaug cell cycle defects and nuclear
phenotype could be partially due to the mis-regulation of cell cycle regulatory proteins that are
ubiquitin-proteasome-dependent for degradation. To test this hypothesis, I would measure cell
cycle progression in individual/combination proteasome RP subunit SRE+ and SRE
- embryos or
Apc7 SRE+ and SRE
- embryos using time-lapse imagining of live embryos (Benoit et al., 2009).
The detection of cell cycle defects in embryos expressing proteasome RP SRE- mRNAs and/or
Apc7 SRE- mRNA would suggest that these defects result from defects in the degradation of one
or more proteins. In Drosophila, the early embryo displays rapid S-M cell cycles, with interphase
being gradually lengthened at later cycles. Cyclin B is an important regulator of cell cycle
progression and is usually required for a cell to enter and exit M phase. During S-M cycling, exit
from M phase requires APC/C-ubiquitin-proteasome mediated degradation of Cyclin B (Lee &
Orr-Weaver, 2003). Therefore, Cyclin B protein is an attractive candidate whose stability may be
misregulated when Smaug fails to properly regulate expression of the proteasome RP and/or
Apc7.
To test this model I would compare overall Cyclin B protein levels in embryos. In wild-type
blastoderm-stage embryos, Cyclin B levels oscillate during cycle 8-13, so it is important to
measure Cyclin B level at each phase of the cell cycle for each nuclear division, and this can be
done by staging methanol-fixed embryos prior to Western blot analysis (Edgar et al., 1994). I
would also compare Cyclin B degradation in wild-type, smaug mutants and embryos expressing
SRE- transgenes. It has been reported that the degradation of Cyclin B gradually increases at late
cycles, and it may be the limiting factor for cell cycle progression (Edgar et al., 1994). To
measure Cyclin B turnover at each nuclear cycle, I would treat permeablized embryos with
151
Cyloheximide for 0, 10, 20, 40 minutes, fix them with methanol for staging, and then pool
embryos that are arrested at the interphase of the same cycle together for Western blot analysis
(Edgar et al., 1994).
Smaug is known to repress the expression of its targets, so I expect to find overexpression of
Apc7 protein in embryos expressing Apc7 SRE- transgenes. If Apc7 overexpression results in
activation of APC/C, I might find a higher degree of Cyclin B degradation during each cell cycle
in embryos expressing Apc7 SRE- transgenes. Because the level of Cyclin B is inversely
correlated with the embryo’s ability to initiate cell cycle delay, I would expect to also find earlier
onset of cell cycle delay, or longer nuclear cycle progression at late cycles (Crest et al., 2007).
Similar results will be expected in embryos expressing proteasome RP subunit SRE- transgenes
if Smaug’s regulation increases ubiquitin-dependent protein degradation. Alternatively, if Apc7
overexpression results in inactivation of APC/C, I might find overexpression of Cyclin B in
embryos expressing Apc7 SRE- transgenes due to decreased degradation, which could result in
premature M phase, abnormal nuclear division, and nuclear defects. The same would be expected
in embryos expressing proteasome RP subunit SRE- transgenes if Smaug’s regulation reduces
ubiquitin-dependent protein degradation.
In addition, to complement the analyses proposed above, I could look at genetic interactions
among members of the proposed pathway, and test whether smaug’s cell cycle defects could
partially be rescued by decreasing the gene dose of Apc7or proteasome RP subunits.
4.2.3.3 Role of Smaug in modulating embryonic metabolism
Maintaining the appropriate energy metabolism is essential for embryonic development
(Dworkin & Dworkin-Rastl, 1991; Johnson et al., 2003; Leese et al., 2007; Tennessen et al.,
2011; Tennessen et al., 2014; Vastag et al., 2011; Vleck & Vleck, 1987), yet energy metabolism
during early Drosophila embryogenesis remains underexplored. In Chapter 2, functional analysis
of Smaug’s targets suggests that Smaug may mediate the degradation and/or translational
repression of mRNAs encoding many metabolic enzymes.
4.2.3.3.1 Role of Smaug in the regulation of glycolysis
It is known that maternal mRNAs encoding enzymes of the glycolytic pathway degrade rapidly
after the first 2 hours of Drosophila embryonic development (Tennessen et al., 2011). Of the ten
152
glycolytic enzymes (Figure 37), eight are likely to be down-regulated by Smaug. Therefore,
Smaug might play an important role in regulating glycolysis during early embryogenesis. To
assess this, I would use radiolabeled compounds to measure and compare the glycolytic flux of
wild-type embryos, smaug-mutant embryos, and embryos that express glycolytic enzyme SRE-
transgenes.
During glycolysis, [5-3H]-glucose is released as radiolabeled
3H2O at the enolase step, and thus
3H2O levels are a measure of the rate of glycolysis. To carry out this experiment, I would
incubate embryos with [5-3H]-glucose and measure the rate at which
3H2O is produced. A vapor-
phase equilibration step will be used to distinguish the 3H2O produced from the [5-
3H]-glucose
that has not be used (Hughes et al., 1993; Zhang et al., 2011). Since Smaug is likely to down-
regulate the expression of glycolytic enzymes and the activity of HK and PFK are up-regulated
in smaug mutants (Chapter 2), I expect to observe an increase in the rate of glycolysis in smaug
mutants and in embryos that express glycolytic enzyme SRE- transgenes compared to control.
There are many caveats to this experiment. First, even if Smaug regulates the expression of
glycolytic enzymes, changes in an enzyme’s expression does not always correlate with changes
in its activity, as enzyme activity can also be regulated through other mechanisms including
substrate induction, changes in inhibitor levels, and post-translational modifications which might
compensate for changes enzyme amount. If I do not obtain the anticipated outcome, this could
also be one of the explanations. Second, Smaug regulates eight of the mRNAs encoding
glycolytic pathway enzymes; thus, expressing a single glycolytic enzyme SRE- transgene might
not be sufficient to induce a phenotype. If this problem arises, I would express multiple
glycolytic SRE- transgenes in the same embryo.
More importantly, there are many alternative pathways for metabolizing carbohydrate (e.g.,
gluconeogenesis, pentose phosphate pathway, glycogenic pathway), and Smaug’s regulation
might be important for controlling the direction of glycolytic carbon flux during early
embryogenesis. In early frog embryos, the flow of glycolytic carbon was directed toward
glycogen synthesis, suggesting glycolysis was running in reverse (Dworkin & Dworkin-Rastl,
1989, 1991). In early mammalian embryo, metabolism goes through a proposed “quiet” stage
(Leese et al., 2007). Glycolysis is blocked and pyruvate metabolism is preferred over glucose’s
(Barbehenn et al., 1974; Johnson et al., 2003). In several insect species, including Drosophila,
glycogen is the primary energy store and glucose content is low during early embryogenesis
153
(Fraga et al., 2013; Moraes et al., 2007; Vital et al., 2010; Yamazaki & Nusse, 2002). Thus,
glycolysis in early Drosophila embryo could also be running in reverse. To determine the
direction of the glycolytic flux, I would use 32
p-labeled glycolytic intermediates such as glucose-
6-phosphate or phosphoenolpyruvate (PEP), which was used to trace carbon flux in frog embryos
(Dworkin & Dworkin-Rastl, 1989). The results of this experiment would allow me to monitor the
direction of glycolytic flow in wild-type embryos and to investigate Smaug’s regulatory role.
154
Figure 37 Glycolysis
Glycolysis is a process of carbohydrate metabolism that produces energy molecules and
intermediary metabolites for other metabolic pathways. Enzymes that are bound and/or regulated
by Smaug are marked with a yellow star. This figure is made by JohnyAbb and liscenced under
Creative Commons Attribution-Share Alike 3.0 Unported.
155
4.2.3.3.2 Role of Smaug in energy state switches
In frog embryos, a metabolomic study reported major remodeling of metabolic pathways during
cleavage cycles and the MBT (Vastag et al., 2011). In several insect species, changes in glucose
metabolism are also correlated with morphological changes during early embryogenesis (Fraga
et al., 2013; Moraes et al., 2007; Vital et al., 2010). Therefore, a switch of energy state is often
coupled with change in developmental stages.
Recent studies reported that Drosophila embryos undergo a metabolic switch during mid-
embryogenesis, directing metabolism towards aerobic glycolysis, in order to prepare for the
energy needs of newly hatched larvae (Tennessen et al., 2011; Tennessen et al., 2014). A time
course metabolomic study of Drosophila embryos also shows that metabolites change drastically
from early stage to late stage embryogenesis (An et al., 2014). Therefore, in the Drosophila
embryo, energy programs, especially glycolytic enzymes, also change in response to
developmental demands.
The Drosophila embryo undergoes rapid cell cycle divisions during early stages. During the
MBT, the cell cycle slows, and zygotic program is initiated, so there might be a process of global
energy remodeling in Drosophila embryo similar to the one found in early frog embryos. In
Chapter 2, I have reported Smaug’s mRNA targets encode metabolic enzymes from various
metabolic pathways, suggesting that Smaug could be one of players that directs the energy state
switch proposed above. All together, I hypothesize that a switch in energy state in Drosophila
embryos during the MBT occurs in response to changes in energy requirements, and that Smaug
is involved in this global metabolic remodeling by co-coordinating the down-regulation of
metabolic enzymes.
To test this hypothesis, I would perform metabolomic analysis to examine the global energy state
in 0-1, 1-2 and 2-3 hour wild-type and smaug-mutant embryos. I would extract metabolites, and
perform comprehensive metabolomic analysis using Gas chromatography/mass spectrometry
(GC/MS) (An et al., 2014; Tennessen et al., 2011; Tennessen et al., 2014). I expect to find global
remodeling of metabolic pathways during the MBT in wild-type embryos, and defects in this
process in the smaug embryos.
156
Because Smaug is expected to repress target-mRNA expression, I would expect to find de-
repression of many metabolic pathways in a smaug mutants compared to wild type. However,
due to complexity of metabolic networks, this might not always be the case. For example, during
mid-embryogenesis, almost every gene encoding a glycolytic enzyme is up-regulated resulting in
the up-regulation of glycolysis (Tennessen et al., 2014). However, genes that are involved in the
TCA cycle and electron transport chains are also up-regulated, yet their activities are blocked
(Tennessen et al., 2014).
The results of these metabolomic analyses will be compared to the lists of Smaug-regulated
mRNAs and this may suggest Smaug targets whose misregulation is likely to have a dramatic
impact on the metabolic state of the early embryo. Such targets would be particularly attractive
candidates for generation of SRE- transgenes.
4.2.4 Closing statement
In summary, Smaug is a multifunctional post-transcriptional regulator that controls several key
cellular processes during early development. The experiments outlined above will begin to
uncover the molecular mechanisms that underlie Smaug’s ability to differentially regulate target
mRNAs. Finally, by assessing the biological consequences of disruption of Smaug-directed
regulation of specific mRNAs, this work would generate a more complete picture of Smaug’s
roles during early embryogenesis.
157
Appendix A: Supplemental data files
Table S1 and S2: Chen_Linan_E_201506_PhD_datatables.exl
Movie S1, related to Figure 24A. Smaug protein is enriched in the germ plasm of early
embryos. Live imaging of Venus-Smaug shows that Smaug protein is enriched in the germ
plasm as early as nuclear cycle 5 (~30 minute after fertilization).
Movie S2, related to Figure 24B. Smaug protein is enriched in the germ plasm of early
embryos. In live embryos, mCherry-Smaug (red) becomes associated with the spindle
microtubules (green; imaged with Jupiter-GFP) upon arrival of nuclei at the posterior (~60
minutes of embryogenesis) and is taken up into the primordial germ (‘pole’) cells when they bud
(~90 minutes of embryogenesis).
Movie S3, related to Figure 24. mCherry-Smaug movement is severely compromised in a
dynein mutant (dhc6-10/dhc6-12). This is a movie in the wild-type control that shows transport into
the germ plasm.
Movie S4, related to Figure 24. mCherry-Smaug movement is severely compromised in a
dynein mutant (dhc6-10/dhc6-12). This is a movie in the dynein mutant that shows lack of transport
into the germ plasm.
Movie S5, related to Figure 25. Smaug is a component of the polar granules. Smaug co-
localizes with Vasa, a known component of polar granules. Live imaging of mCherry-Smaug
(red) and Vasa-GFP (green) transgenic embryos. Smaug and Vasa co-localize in the pole cells as
well as in the apical region of the posterior soma, representing polar granules that are not taken
up into the primordial germ cells when they bud.
158
Appendix B: List of Abbreviations
4E-BPs eIF4E-binding proteins
AEL after egg laid
AGO Argonaute
AP antero-posterior
AREs AU-rich elements
AUBPs ARE-binding proteins
bcd bicoid
BicC Bicaudal C
BICC Bicaudal C
BREs BRU Response Elements
BRU Bruno
Bru Bruno
cad caudal
CBC cap-binding complex
CPEB Cytoplasmic Polyadenylation Element-Binding protein
FDR False Discovery Rate
FISH high-throughput fluorescence RNA in situ hybridization
FRAP fluorescence recovery after photobleaching
GO Gene Ontology
GRE GU-rich element
grk gurken
hb hunchback
HOXB4 Homeobox B4
IRE-BP IRE-binding protein
ISH RNA in situ hybridization
LTD long-term depression
LTP long-term potentiation
MBT midblastula transition
miRs microRNAs
159
MZT maternal-to-zygotic transition
NMD nonsense-mediated decay
nos nanos
NOS Nanos
ORF open-reading frame
OSK Oskar
osk oskar
PABP poly A-binding protein
PARN poly A ribonuclease
PEP phosphoenolpyruvate
PGCs primordial germ cells
RBP RNA-binding protein
RIP-Chip RNA co-immunoprecipitations-microarray analysis
RIP-Seq RNA co-immunoprecipitations-next-generation sequencing
RISC RNA-induced silencing complex
RMA Robust Multi-array Average
RNP Ribonucleoprotein
RP regulatory particle
RT-qPCR quantitative polymerase chain reaction
SAM Significance Analysis of Microarrays
SAM Sterile Alpha Motif
SREs Smaug Recognition Elements
SSR1 Smaug Similarity Region 1
STAU Staufen
TfR transferrin receptor
TI translation index
UTR untranslated region
VAS Vasa
160
References
Altmann, M., Schmitz, N., Berset, C., and Trachsel, H. (1997). A novel inhibitor of cap-
dependent translation initiation in yeast: p20 competes with eIF4G for binding to eIF4E. The
EMBO journal 16, 1114-1121.
Ameres, S.L., Martinez, J., and Schroeder, R. (2007). Molecular basis for target RNA
recognition and cleavage by human RISC. Cell 130, 101-112.
An, P.N., Yamaguchi, M., Bamba, T., and Fukusaki, E. (2014). Metabolome analysis of
Drosophila melanogaster during embryogenesis. PLoS One 9, e99519.
Andoh, T., Oshiro, Y., Hayashi, S., Takeo, H., and Tani, T. (2006). Visual screening for
localized RNAs in yeast revealed novel RNAs at the bud-tip. Biochem Biophys Res Commun
351, 999-1004.
Andrews, S., Snowflack, D.R., Clark, I.E., and Gavis, E.R. (2011). Multiple mechanisms
collaborate to repress nanos translation in the Drosophila ovary and embryo. RNA 17, 967-977.
Arava, Y., Wang, Y., Storey, J.D., Liu, C.L., Brown, P.O., and Herschlag, D. (2003). Genome-
wide analysis of mRNA translation profiles in Saccharomyces cerevisiae. Proc Natl Acad Sci U
S A 100, 3889-3894.
Aronov, S., Gelin-Licht, R., Zipor, G., Haim, L., Safran, E., and Gerst, J.E. (2007). mRNAs
encoding polarity and exocytosis factors are cotransported with the cortical endoplasmic
reticulum to the incipient bud in Saccharomyces cerevisiae. Mol Cell Biol 27, 3441-3455.
Aviv, T., Lin, Z., Ben-Ari, G., Smibert, C.A., and Sicheri, F. (2006). Sequence-specific
recognition of RNA hairpins by the SAM domain of Vts1p. Nature structural & molecular
biology 13, 168-176.
Aviv, T., Lin, Z., Lau, S., Rendl, L.M., Sicheri, F., and Smibert, C.A. (2003). The RNA-binding
SAM domain of Smaug defines a new family of post-transcriptional regulators. Nat Struct Biol
10, 614-621.
Bader, M., Benjamin, S., Wapinski, O.L., Smith, D.M., Goldberg, A.L., and Steller, H. (2011). A
conserved F box regulatory complex controls proteasome activity in Drosophila. Cell 145, 371-
382.
Baez, M.V., and Boccaccio, G.L. (2005). Mammalian Smaug is a translational repressor that
forms cytoplasmic foci similar to stress granules. J Biol Chem 280, 43131-43140.
Bagga, S., Bracht, J., Hunter, S., Massirer, K., Holtz, J., Eachus, R., and Pasquinelli, A.E.
(2005). Regulation by let-7 and lin-4 miRNAs results in target mRNA degradation. Cell 122,
553-563.
161
Bakheet, T., Williams, B.R., and Khabar, K.S. (2006). ARED 3.0: the large and diverse AU-rich
transcriptome. Nucleic Acids Res 34, D111-114.
Balagopal, V., Fluch, L., and Nissan, T. (2012). Ways and means of eukaryotic mRNA decay.
Biochimica et biophysica acta 1819, 593-603.
Barbehenn, E.K., Wales, R.G., and Lowry, O.H. (1974). The explanation for the blockade of
glycolysis in early mouse embryos. Proc Natl Acad Sci U S A 71, 1056-1060.
Barnard, D.C., Ryan, K., Manley, J.L., and Richter, J.D. (2004). Symplekin and xGLD-2 are
required for CPEB-mediated cytoplasmic polyadenylation. Cell 119, 641-651.
Bartel, D.P., and Chen, C.Z. (2004). Micromanagers of gene expression: the potentially
widespread influence of metazoan microRNAs. Nature reviews. Genetics 5, 396-400.
Bashirullah, A., Cooperstock, R.L., and Lipshitz, H.D. (2001). Spatial and temporal control of
RNA stability. Proc Natl Acad Sci U S A 98, 7025-7028.
Bashirullah, A., Halsell, S.R., Cooperstock, R.L., Kloc, M., Karaiskakis, A., Fisher, W.W., Fu,
W., Hamilton, J.K., Etkin, L.D., and Lipshitz, H.D. (1999). Joint action of two RNA degradation
pathways controls the timing of maternal transcript elimination at the midblastula transition in
Drosophila melanogaster. EMBO J. 18, 2610-2620.
Bazzini, A.A., Lee, M.T., and Giraldez, A.J. (2012). Ribosome profiling shows that miR-430
reduces translation before causing mRNA decay in zebrafish. Science 336, 233-237.
Beelman, C.A., and Parker, R. (1995). Degradation of mRNA in eukaryotes. Cell 81, 179-183.
Behm-Ansmant, I., Rehwinkel, J., Doerks, T., Stark, A., Bork, P., and Izaurralde, E. (2006).
mRNA degradation by miRNAs and GW182 requires both CCR4:NOT deadenylase and
DCP1:DCP2 decapping complexes. Genes Dev 20, 1885-1898.
Beller, M., Riedel, D., Jansch, L., Dieterich, G., Wehland, J., Jackle, H., and Kuhnlein, R.P.
(2006). Characterization of the Drosophila lipid droplet subproteome. Molecular & cellular
proteomics : MCP 5, 1082-1094.
Benoit, B., He, C.H., Zhang, F., Votruba, S.M., Tadros, W., Westwood, J.T., Smibert, C.A.,
Lipshitz, H.D., and Theurkauf, W.E. (2009). An essential role for the RNA-binding protein
Smaug during the Drosophila maternal-to-zygotic transition. Development 136, 923-932.
Bergsten, S., and Gavis, E. (1999). Role for mRNA localization in translational activation but not
spatial restriction of nanos RNA. Development 126, 659-669.
Berleth, T., Burri, M., Thoma, G., Bopp, D., Richstein, S., Frigerio, G., Noll, M., and Nusslein-
Volhard, C. (1988). The role of localization of bicoid RNA in organizing the anterior pattern of
the Drosophila embryo. The EMBO journal 7, 1749-1756.
Bernhart, S.H., Hofacker, I.L., and Stadler, P.F. (2006). Local RNA base pairing probabilities in
large sequences. Bioinformatics 22, 614-615.
162
Betley, J.N., Heinrich, B., Vernos, I., Sardet, C., Prodon, F., and Deshler, J.O. (2004). Kinesin II
mediates Vg1 mRNA transport in Xenopus oocytes. Curr Biol 14, 219-224.
Binder, R., Horowitz, J.A., Basilion, J.P., Koeller, D.M., Klausner, R.D., and Harford, J.B.
(1994). Evidence that the pathway of transferrin receptor mRNA degradation involves an
endonucleolytic cleavage within the 3' UTR and does not involve poly(A) tail shortening. The
EMBO journal 13, 1969-1980.
Birsoy, B., Kofron, M., Schaible, K., Wylie, C., and Heasman, J. (2006). Vg 1 is an essential
signaling molecule in Xenopus development. Development 133, 15-20.
Bischof, J., Maeda, R.K., Hediger, M., Karch, F., and Basler, K. (2007). An optimized
transgenesis system for Drosophila using germ-line-specific phiC31 integrases. Proc Natl Acad
Sci U S A 104, 3312-3317.
Bohl, F., Kruse, C., Frank, A., Ferring, D., and Jansen, R.P. (2000). She2p, a novel RNA-binding
protein tethers ASH1 mRNA to the Myo4p myosin motor via She3p. The EMBO journal 19,
5514-5524.
Boswell, R.E., and Mahowald, A.P. (1985). tudor, a gene required for assembly of the germ
plasm in Drosophila melanogaster. Cell 43, 97-104.
Bramham, C.R. (2008). Local protein synthesis, actin dynamics, and LTP consolidation. Current
opinion in neurobiology 18, 524-531.
Bramham, C.R., Worley, P.F., Moore, M.J., and Guzowski, J.F. (2008). The immediate early
gene arc/arg3.1: regulation, mechanisms, and function. J Neurosci 28, 11760-11767.
Braun, J.E., Huntzinger, E., Fauser, M., and Izaurralde, E. (2011). GW182 proteins directly
recruit cytoplasmic deadenylase complexes to miRNA targets. Mol Cell 44, 120-133.
Brechbiel, J.L., and Gavis, E.R. (2008). Spatial regulation of nanos is required for its function in
dendrite morphogenesis. Curr Biol 18, 745-750.
Brendza, R.P., Serbus, L.R., Duffy, J.B., and Saxton, W.M. (2000). A function for kinesin I in
the posterior transport of oskar mRNA and Staufen protein. Science 289, 2120-2122.
Brennan, C.M., and Steitz, J.A. (2001). HuR and mRNA stability. Cellular and molecular life
sciences : CMLS 58, 266-277.
Brown, C.E., and Sachs, A.B. (1998). Poly(A) tail length control in Saccharomyces cerevisiae
occurs by message-specific deadenylation. Mol Cell Biol 18, 6548-6559.
Brown, C.E., Tarun, S.Z., Jr., Boeck, R., and Sachs, A.B. (1996). PAN3 encodes a subunit of the
Pab1p-dependent poly(A) nuclease in Saccharomyces cerevisiae. Mol Cell Biol 16, 5744-5753.
Bushati, N., Stark, A., Brennecke, J., and Cohen, S.M. (2008). Temporal reciprocity of miRNAs
and their targets during the maternal-to-zygotic transition in Drosophila. Curr Biol 18, 501-506.
163
Cairrao, F., Halees, A.S., Khabar, K.S., Morello, D., and Vanzo, N. (2009). AU-rich elements
regulate Drosophila gene expression. Mol Cell Biol 29, 2636-2643.
Cao, Q., and Richter, J.D. (2002). Dissolution of the maskin-eIF4E complex by cytoplasmic
polyadenylation and poly(A)-binding protein controls cyclin B1 mRNA translation and oocyte
maturation. The EMBO journal 21, 3852-3862.
Caughman, S.W., Hentze, M.W., Rouault, T.A., Harford, J.B., and Klausner, R.D. (1988). The
iron-responsive element is the single element responsible for iron-dependent translational
regulation of ferritin biosynthesis. Evidence for function as the binding site for a translational
repressor. J Biol Chem 263, 19048-19052.
Cermelli, S., Guo, Y., Gross, S.P., and Welte, M.A. (2006). The lipid-droplet proteome reveals
that droplets are a protein-storage depot. Curr Biol 16, 1783-1795.
Chang, P., Torres, J., Lewis, R.A., Mowry, K.L., Houliston, E., and King, M.L. (2004).
Localization of RNAs to the mitochondrial cloud in Xenopus oocytes through entrapment and
association with endoplasmic reticulum. Mol Biol Cell 15, 4669-4681.
Chang, Y.F., Imam, J.S., and Wilkinson, M.F. (2007). The nonsense-mediated decay RNA
surveillance pathway. Annual review of biochemistry 76, 51-74.
Chartrand, P., Meng, X.H., Singer, R.H., and Long, R.M. (1999). Structural elements required
for the localization of ASH1 mRNA and of a green fluorescent protein reporter particle in vivo.
Curr Biol 9, 333-336.
Chekulaeva, M., Hentze, M.W., and Ephrussi, A. (2006). Bruno acts as a dual repressor of oskar
translation, promoting mRNA oligomerization and formation of silencing particles. Cell 124,
521-533.
Chen, J., Chiang, Y.C., and Denis, C.L. (2002). CCR4, a 3'-5' poly(A) RNA and ssDNA
exonuclease, is the catalytic component of the cytoplasmic deadenylase. The EMBO journal 21,
1414-1426.
Chicoine, J., Benoit, P., Gamberi, C., Paliouras, M., Simonelig, M., and Lasko, P. (2007).
Bicaudal-C recruits CCR4-NOT deadenylase to target mRNAs and regulates oogenesis,
cytoskeletal organization, and its own expression. Dev Cell 13, 691-704.
Cho, P.F., Gamberi, C., Cho-Park, Y.A., Cho-Park, I.B., Lasko, P., and Sonenberg, N. (2006).
Cap-dependent translational inhibition establishes two opposing morphogen gradients in
Drosophila embryos. Curr Biol 16, 2035-2041.
Cho, P.F., Poulin, F., Cho-Park, Y.A., Cho-Park, I.B., Chicoine, J.D., Lasko, P., and Sonenberg,
N. (2005). A new paradigm for translational control: inhibition via 5'-3' mRNA tethering by
Bicoid and the eIF4E cognate 4EHP. Cell 121, 411-423.
Clark, I.E., Wyckoff, D., and Gavis, E.R. (2000). Synthesis of the posterior determinant Nanos is
spatially restricted by a novel cotranslational regulatory mechanism. Curr Biol 10, 1311-1314.
164
Colman, D.R., Kreibich, G., Frey, A.B., and Sabatini, D.D. (1982). Synthesis and incorporation
of myelin polypeptides into CNS myelin. J Cell Biol 95, 598-608.
Combs, P.A., and Eisen, M.B. (2013). Sequencing mRNA from cryo-sliced Drosophila embryos
to determine genome-wide spatial patterns of gene expression. PLoS One 8, e71820.
Condeelis, J., and Singer, R.H. (2005). How and why does beta-actin mRNA target? Biol Cell
97, 97-110.
Cosentino, G.P., Schmelzle, T., Haghighat, A., Helliwell, S.B., Hall, M.N., and Sonenberg, N.
(2000). Eap1p, a novel eukaryotic translation initiation factor 4E-associated protein in
Saccharomyces cerevisiae. Mol Cell Biol 20, 4604-4613.
Counce, S.J. (1963). Developmental morphology of polar granules in Drosophila. Journal of
Morphology 112, 129-145.
Crest, J., Oxnard, N., Ji, J.Y., and Schubiger, G. (2007). Onset of the DNA replication
checkpoint in the early Drosophila embryo. Genetics 175, 567-584.
Cridge, A.G., Castelli, L.M., Smirnova, J.B., Selley, J.N., Rowe, W., Hubbard, S.J., McCarthy,
J.E., Ashe, M.P., Grant, C.M., and Pavitt, G.D. (2010). Identifying eIF4E-binding protein
translationally-controlled transcripts reveals links to mRNAs bound by specific PUF proteins.
Nucleic Acids Res 38, 8039-8050.
Currie, P.D., and Sullivan, D.T. (1994a). Structure and expression of the gene encoding
phosphofructokinase (PFK) in Drosophila melanogaster. J Biol Chem 269, 24679-24687.
Currie, P.D., and Sullivan, D.T. (1994b). Structure, expression and duplication of genes which
encode phosphoglyceromutase of Drosophila melanogaster. Genetics 138, 352-363.
Dahanukar, A., Walker, J.A., and Wharton, R.P. (1999). Smaug, a novel RNA-binding protein
that operates a translational switch in Drosophila. Mol Cell 4, 209-218.
Dahanukar, A., and Wharton, R.P. (1996). The nanos gradient in Drosophila embryos is
generated by translational regulation. Genes & Dev. 10, 2610-2620.
Daugeron, M.C., Mauxion, F., and Seraphin, B. (2001). The yeast POP2 gene encodes a nuclease
involved in mRNA deadenylation. Nucleic Acids Res 29, 2448-2455.
de Haro, M., Al-Ramahi, I., Jones, K.R., Holth, J.K., Timchenko, L.T., and Botas, J. (2013).
Smaug/SAMD4A restores translational activity of CUGBP1 and suppresses CUG-induced
myopathy. PLoS genetics 9, e1003445.
De Renzis, S., Elemento, O., Tavazoie, S., and Wieschaus, E.F. (2007). Unmasking activation of
the zygotic genome using chromosomal deletions in the Drosophila embryo. PLoS Biol 5, e117.
Ding, D., Parkhurst, S.M., Halsell, S.R., and Lipshitz, H.D. (1993). Dynamic Hsp83 RNA
localization during Drosophila oogenesis and embryogenesis. Mol Cell Biol 13, 3773-3781.
165
Dumollard, R., Carroll, J., Duchen, M.R., Campbell, K., and Swann, K. (2009). Mitochondrial
function and redox state in mammalian embryos. Semin Cell Dev Biol 20, 346-353.
Dworkin, M.B., and Dworkin-Rastl, E. (1989). Metabolic regulation during early frog
development: flow of glycolytic carbon into phospholipids in Xenopus oocytes and fertilized
eggs. Dev Biol 132, 524-528.
Dworkin, M.B., and Dworkin-Rastl, E. (1991). Carbon metabolism in early amphibian embryos.
Trends Biochem Sci 16, 229-234.
Edgar, B.A., Sprenger, F., Duronio, R.J., Leopold, P., and O'Farrell, P.H. (1994). Distinct
molecular mechanism regulate cell cycle timing at successive stages of Drosophila
embryogenesis. Genes Dev 8, 440-452.
Edwards, T.A., Butterwick, J.A., Zeng, L., Gupta, Y.K., Wang, X., Wharton, R.P., Palmer, A.G.,
3rd, and Aggarwal, A.K. (2006). Solution structure of the Vts1 SAM domain in the presence of
RNA. Journal of molecular biology 356, 1065-1072.
Ephrussi, A., Dickinson, L.K., and Lehmann, R. (1991). Oskar organizes the germ plasm and
directs localization of the posterior determinant nanos. Cell 66, 37-50.
Ephrussi, A., and Lehmann, R. (1992). Induction of germ cell formation by oskar. Nature 358,
387-392.
Fabian, M.R., Sonenberg, N., and Filipowicz, W. (2010). Regulation of mRNA translation and
stability by microRNAs. Annual review of biochemistry 79, 351-379.
Farese, R.V., Jr., and Walther, T.C. (2009). Lipid droplets finally get a little R-E-S-P-E-C-T.
Cell 139, 855-860.
Fenteany, G., Standaert, R.F., Lane, W.S., Choi, S., Corey, E.J., and Schreiber, S.L. (1995).
Inhibition of proteasome activities and subunit-specific amino-terminal threonine modification
by lactacystin. Science 268, 726-731.
Ferg, M., Sanges, R., Gehrig, J., Kiss, J., Bauer, M., Lovas, A., Szabo, M., Yang, L., Straehle,
U., Pankratz, M.J., et al. (2007). The TATA-binding protein regulates maternal mRNA
degradation and differential zygotic transcription in zebrafish. Embo J 26, 3945-3956.
Foat, B.C., and Stormo, G.D. (2009). Discovering structural cis-regulatory elements by modeling
the behaviors of mRNAs. Mol Syst Biol 5.
Forrest, K.M., and Gavis, E.R. (2003). Live imaging of endogenous RNA reveals a diffusion and
entrapment mechanism for nanos mRNA localization in Drosophila. Curr Biol 13, 1159-1168.
Forster, F., Lasker, K., Nickell, S., Sali, A., and Baumeister, W. (2010). Toward an integrated
structural model of the 26S proteasome. Molecular & cellular proteomics : MCP 9, 1666-1677.
Fraga, A., Ribeiro, L., Lobato, M., Santos, V., Silva, J.R., Gomes, H., da Cunha Moraes, J.L., de
Souza Menezes, J., de Oliveira, C.J., Campos, E., et al. (2013). Glycogen and glucose
166
metabolism are essential for early embryonic development of the red flour beetle Tribolium
castaneum. PLoS One 8, e65125.
Fredriksson, A., Johansson Krogh, E., Hernebring, M., Pettersson, E., Javadi, A., Almstedt, A.,
and Nystrom, T. (2012). Effects of aging and reproduction on protein quality control in soma and
gametes of Drosophila melanogaster. Aging cell 11, 634-643.
Frigerio, G., Burri, M., Bopp, D., Baumgartner, S., and Noll, M. (1986). Structure of the
segmentation gene paired and the Drosophila PRD gene set as part of a gene network. Cell 47,
735-746.
Gagnon, J.A., Kreiling, J.A., Powrie, E.A., Wood, T.R., and Mowry, K.L. (2013). Directional
transport is mediated by a Dynein-dependent step in an RNA localization pathway. PLoS Biol
11, e1001551.
Gallie, D.R. (1991). The cap and poly(A) tail function synergistically to regulate mRNA
translational efficiency. Genes Dev 5, 2108-2116.
Gamberi, C., and Lasko, P. (2012). The bic-C family of developmental translational regulators.
Comparative and functional genomics 2012, 141386.
Garneau, N.L., Wilusz, J., and Wilusz, C.J. (2007). The highways and byways of mRNA decay.
Nature reviews. Molecular cell biology 8, 113-126.
Gavis, E.R., Curtis, D., and Lehmann, R. (1996). Identification of cis-acting sequences that
control nanos RNA localization. Dev Biol 176, 36-50.
Gavis, E.R., and Lehmann, R. (1994). Translational regulation of nanos by RNA localization.
Nature 369, 315-318.
Gepner, J., Li, M., Ludmann, S., Kortas, C., Boylan, K., Iyadurai, S.J., McGrail, M., and Hays,
T.S. (1996). Cytoplasmic dynein function is essential in Drosophila melanogaster. Genetics 142,
865-878.
Gilboa, L., and Lehmann, R. (2006). Soma-germline interactions coordinate homeostasis and
growth in the Drosophila gonad. Nature 443, 97-100.
Gingras, A.C., Raught, B., Gygi, S.P., Niedzwiecka, A., Miron, M., Burley, S.K., Polakiewicz,
R.D., Wyslouch-Cieszynska, A., Aebersold, R., and Sonenberg, N. (2001). Hierarchical
phosphorylation of the translation inhibitor 4E-BP1. Genes Dev 15, 2852-2864.
Giraldez, A.J., Mishima, Y., Rihel, J., Grocock, R.J., Van Dongen, S., Inoue, K., Enright, A.J.,
and Schier, A.F. (2006). Zebrafish MiR-430 promotes deadenylation and clearance of maternal
mRNAs. Science 312, 75-79.
Gonsalvez, G.B., Urbinati, C.R., and Long, R.M. (2005). RNA localization in yeast: moving
towards a mechanism. Biol Cell 97, 75-86.
167
Gonzalez-Reyes, A., Elliott, H., and St Johnston, D. (1995). Polarization of both major body
axes in Drosophila by gurken-torpedo signalling. Nature 375, 654-658.
Green, J.B., Gardner, C.D., Wharton, R.P., and Aggarwal, A.K. (2003). RNA recognition via the
SAM domain of Smaug. Mol Cell 11, 1537-1548.
Grigull, J., Mnaimneh, S., Pootoolal, J., Robinson, M.D., and Hughes, T.R. (2004). Genome-
wide analysis of mRNA stability using transcription inhibitors and microarrays reveals
posttranscriptional control of ribosome biogenesis factors. Mol Cell Biol 24, 5534-5547.
Grolleau, A., Bowman, J., Pradet-Balade, B., Puravs, E., Hanash, S., Garcia-Sanz, J.A., and
Beretta, L. (2002). Global and specific translational control by rapamycin in T cells uncovered
by microarrays and proteomics. J Biol Chem 277, 22175-22184.
Guo, H., Ingolia, N.T., Weissman, J.S., and Bartel, D.P. (2010). Mammalian microRNAs
predominantly act to decrease target mRNA levels. Nature 466, 835-840.
Hafner, M., Landthaler, M., Burger, L., Khorshid, M., Hausser, J., Berninger, P., Rothballer, A.,
Ascano, M., Jr., Jungkamp, A.C., Munschauer, M., et al. (2010a). Transcriptome-wide
identification of RNA-binding protein and microRNA target sites by PAR-CLIP. Cell 141, 129-
141.
Hafner, M., Landthaler, M., Burger, L., Khorshid, M., Hausser, J., Berninger, P., Rothballer, A.,
Ascano, M., Jungkamp, A.C., Munschauer, M., et al. (2010b). PAR-CliP--a method to identify
transcriptome-wide the binding sites of RNA binding proteins. Journal of visualized experiments
: JoVE.
Haghighat, A., Mader, S., Pause, A., and Sonenberg, N. (1995). Repression of cap-dependent
translation by 4E-binding protein 1: competition with p220 for binding to eukaryotic initiation
factor-4E. The EMBO journal 14, 5701-5709.
Hake, L.E., and Richter, J.D. (1994). CPEB is a specificity factor that mediates cytoplasmic
polyadenylation during Xenopus oocyte maturation. Cell 79, 617-627.
Haley, B., Tang, G., and Zamore, P.D. (2003). In vitro analysis of RNA interference in
Drosophila melanogaster. Methods 30, 330-336.
Hamatani, T., Carter, M.G., Sharov, A.A., and Ko, M.S. (2004). Dynamics of global gene
expression changes during mouse preimplantation development. Dev Cell 6, 117-131.
Hammell, C.M. (2008). The microRNA-argonaute complex: a platform for mRNA modulation.
RNA biology 5, 123-127.
Haracska, L., and Udvardy, A. (1995). Cloning and sequencing a non-ATPase subunit of the
regulatory complex of the Drosophila 26S protease. Eur J Biochem 231, 720-725.
Harper, J.W., Burton, J.L., and Solomon, M.J. (2002). The anaphase-promoting complex: it's not
just for mitosis any more. Genes Dev 16, 2179-2206.
168
Hartl, F.U., Bracher, A., and Hayer-Hartl, M. (2011). Molecular chaperones in protein folding
and proteostasis. Nature 475, 324-332.
Hendzel, M.J., Wei, Y., Mancini, M.A., Van Hooser, A., Ranalli, T., Brinkley, B.R., Bazett-
Jones, D.P., and Allis, C.D. (1997). Mitosis-specific phosphorylation of histone H3 initiates
primarily within pericentromeric heterochromatin during G2 and spreads in an ordered fashion
coincident with mitotic chromosome condensation. Chromosoma 106, 348-360.
Holzl, H., Kapelari, B., Kellermann, J., Seemuller, E., Sumegi, M., Udvardy, A., Medalia, O.,
Sperling, J., Muller, S.A., Engel, A., et al. (2000). The regulatory complex of Drosophila
melanogaster 26S proteasomes. Subunit composition and localization of a deubiquitylating
enzyme. J Cell Biol 150, 119-130.
Houston, D.W., and King, M.L. (2000). Germ plasm and molecular determinants of germ cell
fate. Curr Top Dev Biol 50, 155-181.
Huang, D.W., Sherman, B.T., and Lempicki, R.A. (2009). Bioinformatics enrichment tools:
paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 37, 1-
13.
Huang, Y.S., Carson, J.H., Barbarese, E., and Richter, J.D. (2003). Facilitation of dendritic
mRNA transport by CPEB. Genes Dev 17, 638-653.
Hughes, S.D., Quaade, C., Johnson, J.H., Ferber, S., and Newgard, C.B. (1993). Transfection of
AtT-20ins cells with GLUT-2 but not GLUT-1 confers glucose-stimulated insulin secretion.
Relationship to glucose metabolism. J Biol Chem 268, 15205-15212.
Hummel, T., and Klambt, C. (2008). P-element mutagenesis. Methods Mol Biol 420, 97-117.
Jackson, R.J., Hellen, C.U., and Pestova, T.V. (2010). The mechanism of eukaryotic translation
initiation and principles of its regulation. Nature reviews. Molecular cell biology 11, 113-127.
Jeffery, W.R., Tomlinson, C.R., and Brodeur, R.D. (1983). Localization of actin messenger RNA
during early ascidian development. Dev Biol 99, 408-417.
Jeske, M., Meyer, S., Temme, C., Freudenreich, D., and Wahle, E. (2006). Rapid ATP-dependent
deadenylation of nanos mRNA in a cell-free system from Drosophila embryos. J Biol Chem 281,
25124-25133.
Jeske, M., Moritz, B., Anders, A., and Wahle, E. (2011). Smaug assembles an ATP-dependent
stable complex repressing nanos mRNA translation at multiple levels. Embo J 30, 90-103.
Johannes, G., Carter, M.S., Eisen, M.B., Brown, P.O., and Sarnow, P. (1999). Identification of
eukaryotic mRNAs that are translated at reduced cap binding complex eIF4F concentrations
using a cDNA microarray. Proc Natl Acad Sci U S A 96, 13118-13123.
Johnson, M.T., Mahmood, S., and Patel, M.S. (2003). Intermediary metabolism and energetics
during murine early embryogenesis. J Biol Chem 278, 31457-31460.
169
Johnson, P.E., and Donaldson, L.W. (2006). RNA recognition by the Vts1p SAM domain.
Nature structural & molecular biology 13, 177-178.
Johnstone, O., and Lasko, P. (2001). Translational regulation and RNA localization in
Drosophila oocytes and embryos. Annu Rev Genet 35, 365-406.
Joshi, B., Cameron, A., and Jagus, R. (2004). Characterization of mammalian eIF4E-family
members. Eur J Biochem 271, 2189-2203.
Kalifa, Y., Huang, T., Rosen, L.N., Chatterjee, S., and Gavis, E.R. (2006). Glorund, a Drosophila
hnRNP F/H homolog, is an ovarian repressor of nanos translation. Dev Cell 10, 291-301.
Kashevsky, H., Wallace, J.A., Reed, B.H., Lai, C., Hayashi-Hagihara, A., and Orr-Weaver, T.L.
(2002). The anaphase promoting complex/cyclosome is required during development for
modified cell cycles. Proc Natl Acad Sci U S A 99, 11217-11222.
Kaul, G., Pattan, G., and Rafeequi, T. (2011). Eukaryotic elongation factor-2 (eEF2): its
regulation and peptide chain elongation. Cell biochemistry and function 29, 227-234.
Kazan, H., Ray, D., Chan, E.T., Hughes, T.R., and Morris, Q. (2010). RNAcontext: A New
Method for Learning the Sequence and Structure Binding Preferences of RNA-Binding Proteins.
Plos Comput Biol 6.
Keene, J.D. (2007). RNA regulons: coordination of post-transcriptional events. Nature reviews.
Genetics 8, 533-543.
Keene, J.D., Komisarow, J.M., and Friedersdorf, M.B. (2006). RIP-Chip: the isolation and
identification of mRNAs, microRNAs and protein components of ribonucleoprotein complexes
from cell extracts. Nature protocols 1, 302-307.
Kim-Ha, J., Kerr, K., and Macdonald, P.M. (1995). Translational regulation of oskar mRNA by
bruno, an ovarian RNA-binding protein, is essential. Cell 81, 403-412.
Kim-Ha, J., Smith, J.L., and Macdonald, P.M. (1991). oskar mRNA is localized to the posterior
pole of the Drosophila oocyte. Cell 66, 23-35.
Kim, J.H., and Richter, J.D. (2006). Opposing polymerase-deadenylase activities regulate
cytoplasmic polyadenylation. Mol Cell 24, 173-183.
King, M.L., Messitt, T.J., and Mowry, K.L. (2005). Putting RNAs in the right place at the right
time: RNA localization in the frog oocyte. Biol Cell 97, 19-33.
Klein, U., Gernold, M., and Kloetzel, P.M. (1990). Cell-specific accumulation of Drosophila
proteasomes (MCP) during early development. J Cell Biol 111, 2275-2282.
Kong, J., and Lasko, P. (2012). Translational control in cellular and developmental processes.
Nature reviews. Genetics 13, 383-394.
170
Koprunner, M., Thisse, C., Thisse, B., and Raz, E. (2001). A zebrafish nanos-related gene is
essential for the development of primordial germ cells. Genes Dev 15, 2877-2885.
Kuersten, S., Radek, A., Vogel, C., and Penalva, L.O. (2013). Translation regulation gets its
'omics' moment. Wiley interdisciplinary reviews. RNA 4, 617-630.
Kugler, J.M., and Lasko, P. (2009). Localization, anchoring and translational control of oskar,
gurken, bicoid and nanos mRNA during Drosophila oogenesis. Fly 3, 15-28.
Kuhn, K.M., DeRisi, J.L., Brown, P.O., and Sarnow, P. (2001). Global and specific translational
regulation in the genomic response of Saccharomyces cerevisiae to a rapid transfer from a
fermentable to a nonfermentable carbon source. Mol Cell Biol 21, 916-927.
Lange, S.J., Maticzka, D., Mohl, M., Gagnon, J.N., Brown, C.M., and Backofen, R. (2012).
Global or local? Predicting secondary structure and accessibility in mRNAs. Nucleic Acids Res
40, 5215-5226.
Lasko, P. (1999). RNA sorting in Drosophila oocytes and embryos. Faseb J 13, 421-433.
Laver, J.D., Li, X., Ancevicius, K., Westwood, J.T., Smibert, C.A., Morris, Q.D., and Lipshitz,
H.D. (2013). Genome-wide analysis of Staufen-associated mRNAs identifies secondary
structures that confer target specificity. Nucleic Acids Res 41, 9438-9460.
Lawrence, J.B., and Singer, R.H. (1986). Intracellular localization of messenger RNAs for
cytoskeletal proteins. Cell 45, 407-415.
Lecuyer, E., Yoshida, H., Parthasarathy, N., Alm, C., Babak, T., Cerovina, T., Hughes, T.R.,
Tomancak, P., and Krause, H.M. (2007). Global analysis of mRNA localization reveals a
prominent role in organizing cellular architecture and function. Cell 131, 174-187.
Lee, J.E., Lee, J.Y., Wilusz, J., Tian, B., and Wilusz, C.J. (2010). Systematic analysis of cis-
elements in unstable mRNAs demonstrates that CUGBP1 is a key regulator of mRNA decay in
muscle cells. PLoS One 5, e11201.
Lee, L.A., and Orr-Weaver, T.L. (2003). Regulation of cell cycles in Drosophila development:
intrinsic and extrinsic cues. Annu Rev Genet 37, 545-578.
Leese, H.J., Sturmey, R.G., Baumann, C.G., and McEvoy, T.G. (2007). Embryo viability and
metabolism: obeying the quiet rules. Human reproduction 22, 3047-3050.
Lehmann, R., and Nusslein-Volhard, C. (1986). Abdominal segmentation, pole cell formation,
and embryonic polarity require the localized activity of oskar, a maternal gene in Drosophila.
Cell 47, 141-152.
Lein, E.S., Hawrylycz, M.J., Ao, N., Ayres, M., Bensinger, A., Bernard, A., Boe, A.F., Boguski,
M.S., Brockway, K.S., Byrnes, E.J., et al. (2007). Genome-wide atlas of gene expression in the
adult mouse brain. Nature 445, 168-176.
171
Lerit, D.A., and Gavis, E.R. (2011). Transport of germ plasm on astral microtubules directs germ
cell development in Drosophila. Curr Biol 21, 439-448.
Leung, K.M., van Horck, F.P., Lin, A.C., Allison, R., Standart, N., and Holt, C.E. (2006).
Asymmetrical beta-actin mRNA translation in growth cones mediates attractive turning to netrin-
1. Nat Neurosci 9, 1247-1256.
Lewis, B.P., Burge, C.B., and Bartel, D.P. (2005). Conserved seed pairing, often flanked by
adenosines, indicates that thousands of human genes are microRNA targets. Cell 120, 15-20.
Li, X., Quon, G., Lipshitz, H.D., and Morris, Q. (2010). Predicting in vivo binding sites of RNA-
binding proteins using mRNA secondary structure. RNA 16, 1096-1107.
Liang, L., Diehl-Jones, W., and Lasko, P. (1994). Localization of vasa protein to the Drosophila
pole plasm is independent of its RNA-binding and helicase activities. Development 120, 1201-
1211.
Liao, G., Mingle, L., Van De Water, L., and Liu, G. (2015). Control of cell migration through
mRNA localization and local translation. Wiley interdisciplinary reviews. RNA 6, 1-15.
Licatalosi, D.D., Mele, A., Fak, J.J., Ule, J., Kayikci, M., Chi, S.W., Clark, T.A., Schweitzer,
A.C., Blume, J.E., Wang, X., et al. (2008). HITS-CLIP yields genome-wide insights into brain
alternative RNA processing. Nature 456, 464-469.
Lie, Y.S., and Macdonald, P.M. (1999). Apontic binds the translational repressor Bruno and is
implicated in regulation of oskar mRNA translation. Development 126, 1129-1138.
Long, R.M., Gu, W., Lorimer, E., Singer, R.H., and Chartrand, P. (2000). She2p is a novel RNA-
binding protein that recruits the Myo4p-She3p complex to ASH1 mRNA. The EMBO journal 19,
6592-6601.
Mader, S., Lee, H., Pause, A., and Sonenberg, N. (1995). The translation initiation factor eIF-4E
binds to a common motif shared by the translation factor eIF-4 gamma and the translational
repressors 4E-binding proteins. Mol Cell Biol 15, 4990-4997.
Mahone, M., Saffman, E.E., and Lasko, P.F. (1995). Localized Bicaudal-C RNA encodes a
protein containing a KH domain, the RNA binding motif of FMR1. Embo J 14, 2043-2055.
Mahowald, A.P. (2001). Assembly of the Drosophila germ plasm. Int Rev Cytol 203, 187-213.
Marcotrigiano, J., Gingras, A.C., Sonenberg, N., and Burley, S.K. (1999). Cap-dependent
translation initiation in eukaryotes is regulated by a molecular mimic of eIF4G. Mol Cell 3, 707-
716.
Markstein, M., Pitsouli, C., Villalta, C., Celniker, S.E., and Perrimon, N. (2008). Exploiting
position effects and the gypsy retrovirus insulator to engineer precisely expressed transgenes.
Nat Genet 40, 476-483.
172
Martin, K.C., and Ephrussi, A. (2009). mRNA localization: gene expression in the spatial
dimension. Cell 136, 719-730.
Mathavan, S., Lee, S.G.P., Mak, A., Miller, L.D., Murthy, K.R.K., Govindarajan, K.R., Tong,
Y., Wu, Y.L., Lam, S.H., Yang, H., et al. (2005). Transcriptome analysis of zebrafish
embryogenesis using microarrays. PLoS genetics 1, 260-276.
McGrew, L.L., Dworkin-Rastl, E., Dworkin, M.B., and Richter, J.D. (1989). Poly(A) elongation
during Xenopus oocyte maturation is required for translational recruitment and is mediated by a
short sequence element. Genes Dev 3, 803-815.
Medioni, C., Mowry, K., and Besse, F. (2012). Principles and roles of mRNA localization in
animal development. Development 139, 3263-3276.
Messitt, T.J., Gagnon, J.A., Kreiling, J.A., Pratt, C.A., Yoon, Y.J., and Mowry, K.L. (2008).
Multiple kinesin motors coordinate cytoplasmic RNA transport on a subpopulation of
microtubules in Xenopus oocytes. Dev Cell 15, 426-436.
Mili, S., Moissoglu, K., and Macara, I.G. (2008). Genome-wide screen reveals APC-associated
RNAs enriched in cell protrusions. Nature 453, 115-119.
Mili, S., and Steitz, J.A. (2004). Evidence for reassociation of RNA-binding proteins after cell
lysis: implications for the interpretation of immunoprecipitation analyses. Rna 10, 1692-1694.
Mingle, L.A., Okuhama, N.N., Shi, J., Singer, R.H., Condeelis, J., and Liu, G. (2005).
Localization of all seven messenger RNAs for the actin-polymerization nucleator Arp2/3
complex in the protrusions of fibroblasts. J Cell Sci 118, 2425-2433.
Mohler, J., and Wieschaus, E.F. (1985). Bicaudal mutations of Drosophila melanogaster:
alteration of blastoderm cell fate. Cold Spring Harb Symp Quant Biol 50, 105-111.
Moore, M.J., and Proudfoot, N.J. (2009). Pre-mRNA Processing Reaches Back toTranscription
and Ahead to Translation. Cell 136, 688-700.
Moraes, J., Galina, A., Alvarenga, P.H., Rezende, G.L., Masuda, A., da Silva Vaz, I., Jr., and
Logullo, C. (2007). Glucose metabolism during embryogenesis of the hard tick Boophilus
microplus. Comparative biochemistry and physiology. Part A, Molecular & integrative
physiology 146, 528-533.
Morris, A.R., Mukherjee, N., and Keene, J.D. (2010). Systematic analysis of posttranscriptional
gene expression. Wiley interdisciplinary reviews. Systems biology and medicine 2, 162-180.
Morris, J.Z., Hong, A., Lilly, M.A., and Lehmann, R. (2005). twin, a CCR4 homolog, regulates
cyclin poly(A) tail length to permit Drosophila oogenesis. Development 132, 1165-1174.
Munro, T.P., Kwon, S., Schnapp, B.J., and St Johnston, D. (2006). A repeated IMP-binding
motif controls oskar mRNA translation and anchoring independently of Drosophila melanogaster
IMP. J Cell Biol 172, 577-588.
173
Munroe, D., and Jacobson, A. (1990). mRNA poly(A) tail, a 3' enhancer of translational
initiation. Mol Cell Biol 10, 3441-3455.
Murata, Y., and Wharton, R.P. (1995). Binding of pumilio to maternal hunchback mRNA is
required for posterior patterning in Drosophila embryos. Cell 80, 747-756.
Nakamura, A., Sato, K., and Hanyu-Nakamura, K. (2004). Drosophila cup is an eIF4E binding
protein that associates with Bruno and regulates oskar mRNA translation in oogenesis. Dev Cell
6, 69-78.
Nelson, M.R., Leidal, A.M., and Smibert, C.A. (2004). Drosophila Cup is an eIF4E-binding
protein that functions in Smaug-mediated translational repression. The EMBO journal 23, 150-
159.
Nelson, M.R., Luo, H., Vari, H.K., Cox, B.J., Simmonds, A.J., Krause, H.M., Lipshitz, H.D., and
Smibert, C.A. (2007). A multiprotein complex that mediates translational enhancement in
Drosophila. J Biol Chem 282, 34031-34038.
Oberstrass, F.C., Lee, A., Stefl, R., Janis, M., Chanfreau, G., and Allain, F.H.T. (2006). Shape-
specific recognition in the structure of the Vts1p SAM domain with RNA. Nature structural &
molecular biology 13, 160-167.
Paquin, N., and Chartrand, P. (2008). Local regulation of mRNA translation: new insights from
the bud. Trends Cell Biol 18, 105-111.
Parker, R., and Song, H. (2004). The enzymes and control of eukaryotic mRNA turnover. Nature
structural & molecular biology 11, 121-127.
Pause, A., Methot, N., Svitkin, Y., Merrick, W.C., and Sonenberg, N. (1994). Dominant negative
mutants of mammalian translation initiation factor eIF-4A define a critical role for eIF-4F in cap-
dependent and cap-independent initiation of translation. The EMBO journal 13, 1205-1215.
Peters, J.M. (2006). The anaphase promoting complex/cyclosome: a machine designed to
destroy. Nat Rev Mol Cell Biol 7, 644-656.
Peters, P.J., Bos, E., and Griekspoor, A. (2006). Cryo-immunogold electron microscopy. Curr
Protoc Cell Biol Chapter 4, Unit 4 7.
Pilkis, S.J., Claus, T.H., Kurland, I.J., and Lange, A.J. (1995). 6-Phosphofructo-2-
kinase/fructose-2,6-bisphosphatase: a metabolic signaling enzyme. Annual review of
biochemistry 64, 799-835.
Pilkis, S.J., el-Maghrabi, M.R., and Claus, T.H. (1988). Hormonal regulation of hepatic
gluconeogenesis and glycolysis. Annual review of biochemistry 57, 755-783.
Pinder, B.D., and Smibert, C.A. (2013). microRNA-independent recruitment of Argonaute 1 to
nanos mRNA through the Smaug RNA-binding protein. EMBO Reports 14, 80-86.
174
Poulin, F., Gingras, A.C., Olsen, H., Chevalier, S., and Sonenberg, N. (1998). 4E-BP3, a new
member of the eukaryotic initiation factor 4E-binding protein family. J Biol Chem 273, 14002-
14007.
Pratt, A.J., and MacRae, I.J. (2009). The RNA-induced silencing complex: a versatile gene-
silencing machine. J Biol Chem 284, 17897-17901.
Qiao, F., and Bowie, J.U. (2005). The many faces of SAM. Science's STKE : signal transduction
knowledge environment 2005, re7.
Qin, X., Ahn, S., Speed, T.P., and Rubin, G.M. (2007). Global analyses of mRNA translational
control during early Drosophila embryogenesis. Genome biology 8, R63.
Radford, H.E., Meijer, H.A., and de Moor, C.H. (2008). Translational control by cytoplasmic
polyadenylation in Xenopus oocytes. Biochimica et biophysica acta 1779, 217-229.
Rangan, P., DeGennaro, M., Jaime-Bustamante, K., Coux, R.X., Martinho, R.G., and Lehmann,
R. (2009). Temporal and spatial control of germ-plasm RNAs. Curr Biol 19, 72-77.
Rattenbacher, B., Beisang, D., Wiesner, D.L., Jeschke, J.C., von Hohenberg, M., St Louis-
Vlasova, I.A., and Bohjanen, P.R. (2010). Analysis of CUGBP1 targets identifies GU-repeat
sequences that mediate rapid mRNA decay. Mol Cell Biol 30, 3970-3980.
Ray, D., Kazan, H., Chan, E.T., Pena Castillo, L., Chaudhry, S., Talukder, S., Blencowe, B.J.,
Morris, Q., and Hughes, T.R. (2009). Rapid and systematic analysis of the RNA recognition
specificities of RNA-binding proteins. Nature biotechnology 27, 667-670.
Ray, D., Kazan, H., Cook, K.B., Weirauch, M.T., Najafabadi, H.S., Li, X., Gueroussov, S., Albu,
M., Zheng, H., Yang, A., et al. (2013). A compendium of RNA-binding motifs for decoding
gene regulation. Nature 499, 172-177.
Rebagliati, M.R., Weeks, D.L., Harvey, R.P., and Melton, D.A. (1985). Identification and
cloning of localized maternal RNAs from Xenopus eggs. Cell 42, 769-777.
Reed, S.I. (2006). The ubiquitin-proteasome pathway in cell cycle control. Results and problems
in cell differentiation 42, 147-181.
Rendl, L.M., Bieman, M.A., and Smibert, C.A. (2008). S. cerevisiae Vts1p induces
deadenylation-dependent transcript degradation and interacts with the Ccr4p-Pop2p-Not
deadenylase complex. RNA 14, 1328-1336.
Rendl, L.M., Bieman, M.A., Vari, H.K., and Smibert, C.A. (2012). The eIF4E-binding protein
Eap1p functions in Vts1p-mediated transcript decay. PLoS One 7, e47121.
Reveal, B., Yan, N., Snee, M.J., Pai, C.I., Gim, Y., and Macdonald, P.M. (2010). BREs mediate
both repression and activation of oskar mRNA translation and act in trans. Dev Cell 18, 496-502.
Richter, J.D., and Sonenberg, N. (2005). Regulation of cap-dependent translation by eIF4E
inhibitory proteins. Nature 433, 477-480.
175
Riordan, D.P., Herschlag, D., and Brown, P.O. (2011). Identification of RNA recognition
elements in the Saccharomyces cerevisiae transcriptome. Nucleic Acids Res 39, 1501-1509.
Rom, E., Kim, H.C., Gingras, A.C., Marcotrigiano, J., Favre, D., Olsen, H., Burley, S.K., and
Sonenberg, N. (1998). Cloning and characterization of 4EHP, a novel mammalian eIF4E-related
cap-binding protein. J Biol Chem 273, 13104-13109.
Rongo, C., Broihier, H.T., Moore, L., Van Doren, M., Forbes, A., and Lehmann, R. (1997).
Germ plasm assembly and germ cell migration in Drosophila. Cold Spring Harb Symp Quant
Biol 62, 1-11.
Rorth, P. (1998). Gal4 in the Drosophila female germline. Mechanisms of development 78, 113-
118.
Roselli-Rehfuss, L., Ye, F., Lissemore, J.L., and Sullivan, D.T. (1992). Structure and expression
of the phosphoglycerate kinase (Pgk) gene of Drosophila melanogaster. Molecular & general
genetics : MGG 235, 213-220.
Roth, S., Neuman-Silberberg, F.S., Barcelo, G., and Schupbach, T. (1995). cornichon and the
EGF receptor signaling process are necessary for both anterior-posterior and dorsal-ventral
pattern formation in Drosophila. Cell 81, 967-978.
Rouget, C., Papin, C., Boureux, A., Meunier, A.C., Franco, B., Robine, N., Lai, E.C., Pelisson,
A., and Simonelig, M. (2010). Maternal mRNA deadenylation and decay by the piRNA pathway
in the early Drosophila embryo. Nature 467, 1128-1132.
Saeki, Y., and Tanaka, K. (2012). Assembly and function of the proteasome. Methods Mol Biol
832, 315-337.
Saffman, E.E., Styhler, S., Rother, K., Li, W., Richard, S., and Lasko, P. (1998). Premature
translation of oskar in oocytes lacking the RNA-binding protein bicaudal-C. Mol Cell Biol 18,
4855-4862.
Salz, H.K., Cline, T.W., and Schedl, P. (1987). Functional changes associated with structural
alterations induced by mobilization of a P element inserted in the Sex-lethal gene of Drosophila.
Genetics 117, 221-231.
Sanduja, S., Blanco, F.F., and Dixon, D.A. (2011). The roles of TTP and BRF proteins in
regulated mRNA decay. Wiley interdisciplinary reviews. RNA 2, 42-57.
Sardet, C., Dru, P., and Prodon, F. (2005). Maternal determinants and mRNAs in the cortex of
ascidian oocytes, zygotes and embryos. Biol Cell 97, 35-49.
Sarkissian, M., Mendez, R., and Richter, J.D. (2004). Progesterone and insulin stimulation of
CPEB-dependent polyadenylation is regulated by Aurora A and glycogen synthase kinase-3.
Genes Dev 18, 48-61.
Schoenberg, D.R., and Maquat, L.E. (2012). Regulation of cytoplasmic mRNA decay. Nature
reviews. Genetics 13, 246-259.
176
Schupbach, T., and Wieschaus, E. (1986). Germline autonomy of maternal-effect mutations
altering the embryonic body pattern of Drosophila. Dev Biol 113, 443-448.
Semotok, J.L., Cooperstock, R.L., Pinder, B.D., Vari, H.K., Lipshitz, H.D., and Smibert, C.A.
(2005). Smaug recruits the CCR4/POP2/NOT deadenylase complex to trigger maternal transcript
localization in the early Drosophila embryo. Curr Biol 15, 284-294.
Semotok, J.L., and Lipshitz, H.D. (2007). Regulation and function of maternal mRNA
destabilization during early Drosophila development. Differentiation 75, 482-506.
Semotok, J.L., Luo, H., Cooperstock, R.L., Karaiskakis, A., Vari, H.K., Smibert, C.A., and
Lipshitz, H.D. (2008). Drosophila maternal Hsp83 mRNA destabilization is directed by multiple
SMAUG recognition elements in the open reading frame. Mol Cell Biol 28, 6757-6772.
Shatkin, A.J., and Manley, J.L. (2000). The ends of the affair: Capping and polyadenylation.
Nature Structural Biology 7, 838-842.
Shaw-Lee, R., Lissemore, J.L., Sullivan, D.T., and Tolan, D.R. (1992). Alternative splicing of
fructose 1,6-bisphosphate aldolase transcripts in Drosophila melanogaster predicts three
isozymes. J Biol Chem 267, 3959-3967.
Shaw-Lee, R.L., Lissemore, J.L., and Sullivan, D.T. (1991). Structure and expression of the
triose phosphate isomerase (Tpi) gene of Drosophila melanogaster. Molecular & general genetics
: MGG 230, 225-229.
Shepard, K.A., Gerber, A.P., Jambhekar, A., Takizawa, P.A., Brown, P.O., Herschlag, D.,
DeRisi, J.L., and Vale, R.D. (2003). Widespread cytoplasmic mRNA transport in yeast:
identification of 22 bud-localized transcripts using DNA microarray analysis. Proc Natl Acad Sci
U S A 100, 11429-11434.
Siddiqui, N., Li, X., Luo, H., Karaiskakis, A., Hou, H., Kislinger, T., Westwood, J.T., Morris, Q.,
and Lipshitz, H.D. (2012). Genome-wide analysis of the maternal-to-zygotic transition in
Drosophila primordial germ cells. Genome biology 13, R11.
Simon, J.A., and Kingston, R.E. (2013). Occupying chromatin: Polycomb mechanisms for
getting to genomic targets, stopping transcriptional traffic, and staying put. Mol Cell 49, 808-
824.
Smibert, C.A., Lie, Y.S., Shillinglaw, W., Henzel, W.J., and Macdonald, P.M. (1999). Smaug, a
novel and conserved protein, contributes to repression of nanos mRNA translation in vitro. Rna
5, 1535-1547.
Smibert, C.A., Wilson, J.E., Kerr, K., and Macdonald, P.M. (1996). smaug protein represses
translation of unlocalized nanos mRNA in the Drosophila embryo. Genes & Dev. 10, 2600-
2609.
Smith, J.L., Wilson, J.E., and Macdonald, P.M. (1992). Overexpression of oskar directs ectopic
activation of nanos and presumptive pole cell formation in Drosophila embryos. Cell 70, 849-
859.
177
St Johnston, D. (2005). Moving messages: the intracellular localization of mRNAs. Nature
reviews. Molecular cell biology 6, 363-375.
St Johnston, D., Beuchle, D., and Nusslein-Volhard, C. (1991). Staufen, a gene required to
localize maternal RNAs in the Drosophila egg. Cell 66, 51-63.
Stebbins-Boaz, B., Cao, Q., de Moor, C.H., Mendez, R., and Richter, J.D. (1999). Maskin is a
CPEB-associated factor that transiently interacts with elF-4E. Mol Cell 4, 1017-1027.
Storey, J.D., and Tibshirani, R. (2003). Statistical significance for genomewide studies. Proc Natl
Acad Sci U S A 100, 9440-9445.
Sun, X.H., Tso, J.Y., Lis, J., and Wu, R. (1988). Differential regulation of the two
glyceraldehyde-3-phosphate dehydrogenase genes during Drosophila development. Mol Cell
Biol 8, 5200-5205.
Sundell, C.L., and Singer, R.H. (1991). Requirement of microfilaments in sorting of actin
messenger RNA. Science 253, 1275-1277.
Tadros, W., Goldman, A.L., Babak, T., Menzies, F., Vardy, L., Orr-Weaver, T., Hughes, T.R.,
Westwood, J.T., Smibert, C.A., and Lipshitz, H.D. (2007). SMAUG is a major regulator of
maternal mRNA destabilization in Drosophila and its translation is activated by the PAN GU
kinase. Dev Cell 12, 143-155.
Tadros, W., and Lipshitz, H.D. (2009). The maternal-to-zygotic transition: a play in two acts.
Development 136, 3033-3042.
Takizawa, P.A., DeRisi, J.L., Wilhelm, J.E., and Vale, R.D. (2000). Plasma membrane
compartmentalization in yeast by messenger RNA transport and a septin diffusion barrier.
Science 290, 341-344.
Takizawa, P.A., and Vale, R.D. (2000). The myosin motor, Myo4p, binds Ash1 mRNA via the
adapter protein, She3p. Proc Natl Acad Sci U S A 97, 5273-5278.
Tarun, S.Z., Jr., and Sachs, A.B. (1995). A common function for mRNA 5' and 3' ends in
translation initiation in yeast. Genes Dev 9, 2997-3007.
Tarun, S.Z., Jr., and Sachs, A.B. (1996). Association of the yeast poly(A) tail binding protein
with translation initiation factor eIF-4G. The EMBO journal 15, 7168-7177.
Tarun, S.Z., Jr., Wells, S.E., Deardorff, J.A., and Sachs, A.B. (1997). Translation initiation factor
eIF4G mediates in vitro poly(A) tail-dependent translation. Proc Natl Acad Sci U S A 94, 9046-
9051.
Temme, C., Zaessinger, S., Meyer, S., Simonelig, M., and Wahle, E. (2004). A complex
containing the CCR4 and CAF1 proteins is involved in mRNA deadenylation in Drosophila. The
EMBO journal 23, 2862-2871.
178
Tenenbaum, S.A., Carson, C.C., Lager, P.J., and Keene, J.D. (2000). Identifying mRNA subsets
in messenger ribonucleoprotein complexes by using cDNA arrays. Proc Natl Acad Sci U S A 97,
14085-14090.
Tennessen, J.M., Baker, K.D., Lam, G., Evans, J., and Thummel, C.S. (2011). The Drosophila
estrogen-related receptor directs a metabolic switch that supports developmental growth. Cell
metabolism 13, 139-148.
Tennessen, J.M., Bertagnolli, N.M., Evans, J., Sieber, M.H., Cox, J., and Thummel, C.S. (2014).
Coordinated Metabolic Transitions During Drosophila Embryogenesis and the Onset of Aerobic
Glycolysis. G3.
Thomsen, S., Anders, S., Janga, S.C., Huber, W., and Alonso, C.R. (2010). Genome-wide
analysis of mRNA decay patterns during early Drosophila development. Genome biology 11,
R93.
Thomson, A.M., Rogers, J.T., and Leedman, P.J. (1999). Iron-regulatory proteins, iron-
responsive elements and ferritin mRNA translation. The international journal of biochemistry &
cell biology 31, 1139-1152.
Thummel, C.S., and Pirrotta, V. (1991). New pCaSpeR P element vectors. Drosophila
Information Newsletter 2, 19.
Tie, F., Prasad-Sinha, J., Birve, A., Rasmuson-Lestander, A., and Harte, P.J. (2003). A 1-
megadalton ESC/E(Z) complex from Drosophila that contains polycomblike and RPD3. Mol
Cell Biol 23, 3352-3362.
Tomancak, P., Beaton, A., Weiszmann, R., Kwan, E., Shu, S., Lewis, S.E., Richards, S.,
Ashburner, M., Hartenstein, V., Celniker, S.E., et al. (2002). Systematic determination of
patterns of gene expression during Drosophila embryogenesis. Genome biology 3,
RESEARCH0088.
Tsubuki, S., Saito, Y., Tomioka, M., Ito, H., and Kawashima, S. (1996). Differential inhibition of
calpain and proteasome activities by peptidyl aldehydes of di-leucine and tri-leucine. Journal of
biochemistry 119, 572-576.
Tucker, M., Staples, R.R., Valencia-Sanchez, M.A., Muhlrad, D., and Parker, R. (2002). Ccr4p is
the catalytic subunit of a Ccr4p/Pop2p/Notp mRNA deadenylase complex in Saccharomyces
cerevisiae. The EMBO journal 21, 1427-1436.
Tucker, M., Valencia-Sanchez, M.A., Staples, R.R., Chen, J., Denis, C.L., and Parker, R. (2001).
The transcription factor associated Ccr4 and Caf1 proteins are components of the major
cytoplasmic mRNA deadenylase in Saccharomyces cerevisiae. Cell 104, 377-386.
Tusher, V.G., Tibshirani, R., and Chu, G. (2001). Significance analysis of microarrays applied to
the ionizing radiation response. Proc Natl Acad Sci U S A 98, 5116-5121.
Ule, J., Jensen, K.B., Ruggiu, M., Mele, A., Ule, A., and Darnell, R.B. (2003). CLIP identifies
Nova-regulated RNA networks in the brain. Science 302, 1212-1215.
179
van Nocker, S., Sadis, S., Rubin, D.M., Glickman, M., Fu, H., Coux, O., Wefes, I., Finley, D.,
and Vierstra, R.D. (1996). The multiubiquitin-chain-binding protein Mcb1 is a component of the
26S proteasome in Saccharomyces cerevisiae and plays a nonessential, substrate-specific role in
protein turnover. Mol Cell Biol 16, 6020-6028.
Vastag, L., Jorgensen, P., Peshkin, L., Wei, R., Rabinowitz, J.D., and Kirschner, M.W. (2011).
Remodeling of the metabolome during early frog development. PLoS One 6, e16881.
Vessey, J.P., Macchi, P., Stein, J.M., Mikl, M., Hawker, K.N., Vogelsang, P., Wieczorek, K.,
Vendra, G., Riefler, J., Tubing, F., et al. (2008). A loss of function allele for murine Staufen1
leads to impairment of dendritic Staufen1-RNP delivery and dendritic spine morphogenesis. Proc
Natl Acad Sci U S A 105, 16374-16379.
Villaescusa, J.C., Buratti, C., Penkov, D., Mathiasen, L., Planaguma, J., Ferretti, E., and Blasi, F.
(2009). Cytoplasmic Prep1 interacts with 4EHP inhibiting Hoxb4 translation. PLoS One 4,
e5213.
Vital, W., Rezende, G.L., Abreu, L., Moraes, J., Lemos, F.J., Vaz Ida, S., Jr., and Logullo, C.
(2010). Germ band retraction as a landmark in glucose metabolism during Aedes aegypti
embryogenesis. BMC developmental biology 10, 25.
Vlasova, I.A., and Bohjanen, P.R. (2008). Posttranscriptional regulation of gene networks by
GU-rich elements and CELF proteins. RNA biology 5, 201-207.
Vlasova, I.A., Tahoe, N.M., Fan, D., Larsson, O., Rattenbacher, B., Sternjohn, J.R., Vasdewani,
J., Karypis, G., Reilly, C.S., Bitterman, P.B., et al. (2008). Conserved GU-rich elements mediate
mRNA decay by binding to CUG-binding protein 1. Mol Cell 29, 263-270.
Vleck, C.M., and Vleck, D. (1987). Metabolism and energetics of avian embryos. The Journal of
experimental zoology. Supplement : published under auspices of the American Society of
Zoologists and the Division of Comparative Physiology and Biochemistry / the Wistar Institute
of Anatomy and Biology 1, 111-125.
Wakiyama, M., Imataka, H., and Sonenberg, N. (2000). Interaction of eIF4G with poly(A)-
binding protein stimulates translation and is critical for Xenopus oocyte maturation. Curr Biol
10, 1147-1150.
Walser, C.B., and Lipshitz, H.D. (2011). Transcript clearance during the maternal-to-zygotic
transition. Curr Opin Genet Dev 21, 431-443.
Wang, C., and Lehmann, R. (1991). Nanos is the localized posterior determinant in Drosophila.
Cell 66, 637-647.
Wang, Y., Liu, C.L., Storey, J.D., Tibshirani, R.J., Herschlag, D., and Brown, P.O. (2002).
Precision and functional specificity in mRNA decay. Proc Natl Acad Sci U S A 99, 5860-5865.
Webster, P.J., Liang, L., Berg, C.A., Lasko, P., and Macdonald, P.M. (1997). Translational
repressor bruno plays multiple roles in development and is widely conserved. Genes Dev 11,
2510-2521.
180
Weil, T.T., Xanthakis, D., Parton, R., Dobbie, I., Rabouille, C., Gavis, E.R., and Davis, I. (2010).
Distinguishing direct from indirect roles for bicoid mRNA localization factors. Development
137, 169-176.
Wells, S.E., Hillner, P.E., Vale, R.D., and Sachs, A.B. (1998). Circularization of mRNA by
eukaryotic translation initiation factors. Mol Cell 2, 135-140.
Welshhans, K., and Bassell, G.J. (2011). Netrin-1-induced local beta-actin synthesis and growth
cone guidance requires zipcode binding protein 1. J Neurosci 31, 9800-9813.
Wickens, M. (1990). In the beginning is the end: regulation of poly(A) addition and removal
during early development. Trends Biochem Sci 15, 320-324.
Wiederhold, K., and Passmore, L.A. (2010). Cytoplasmic deadenylation: regulation of mRNA
fate. Biochemical Society transactions 38, 1531-1536.
Wolke, U., Weidinger, G., Koprunner, M., and Raz, E. (2002). Multiple levels of
posttranscriptional control lead to germ line-specific gene expression in the zebrafish. Curr Biol
12, 289-294.
Wu, X., and Brewer, G. (2012). The regulation of mRNA stability in mammalian cells: 2.0. Gene
500, 10-21.
Wuchty, S., Fontana, W., Hofacker, I.L., and Schuster, P. (1999). Complete suboptimal folding
of RNA and the stability of secondary structures. Biopolymers 49, 145-165.
Yamashita, A., Chang, T.C., Yamashita, Y., Zhu, W., Zhong, Z., Chen, C.Y., and Shyu, A.B.
(2005). Concerted action of poly(A) nucleases and decapping enzyme in mammalian mRNA
turnover. Nature structural & molecular biology 12, 1054-1063.
Yamazaki, H., and Nusse, R. (2002). Identification of DCAP, a drosophila homolog of a glucose
transport regulatory complex. Mechanisms of development 119, 115-119.
Yang, E., van Nimwegen, E., Zavolan, M., Rajewsky, N., Schroeder, M., Magnasco, M., and
Darnell, J.E., Jr. (2003). Decay rates of human mRNAs: correlation with functional
characteristics and sequence attributes. Genome research 13, 1863-1872.
Yano, T., Lopez de Quinto, S., Matsui, Y., Shevchenko, A., and Ephrussi, A. (2004). Hrp48, a
Drosophila hnRNPA/B homolog, binds and regulates translation of oskar mRNA. Dev Cell 6,
637-648.
Yao, J., Sasaki, Y., Wen, Z., Bassell, G.J., and Zheng, J.Q. (2006). An essential role for beta-
actin mRNA localization and translation in Ca2+-dependent growth cone guidance. Nat Neurosci
9, 1265-1273.
Zaessinger, S., Busseau, I., and Simonelig, M. (2006). Oskar allows nanos mRNA translation in
Drosophila embryos by preventing its deadenylation by Smaug/CCR4. Development 133, 4573-
4583.
181
Zhang, C., Lin, M., Wu, R., Wang, X., Yang, B., Levine, A.J., Hu, W., and Feng, Z. (2011).
Parkin, a p53 target gene, mediates the role of p53 in glucose metabolism and the Warburg
effect. Proc Natl Acad Sci U S A 108, 16259-16264.
Zhang, H.L., Singer, R.H., and Bassell, G.J. (1999). Neurotrophin regulation of beta-actin
mRNA and protein localization within growth cones. J Cell Biol 147, 59-70.
Zhang, J., Houston, D.W., King, M.L., Payne, C., Wylie, C., and Heasman, J. (1998). The role of
maternal VegT in establishing the primary germ layers in Xenopus embryos. Cell 94, 515-524.
