THE REQUIREMENT OF FACULTATIVE HETEROCHROMATIN IN
MAINTAINING DROSOPHILA FEMALE GERM CELL IDENTITY
by
ANNE ELIZABETH SMOLKO
Submitted in partial fulfillment of the requirements for the degree of Doctor of
Philosophy
Dissertation Advisor:
Helen K. Salz, PhD
Department of Genetics and Genome Sciences
CASE WESTERN RESERVE UNIVERSITY
August, 2019
CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the dissertation of
Anne E. Smolko
Candidate for the degree of Doctor of Philosophy *
Thesis Advisor:
Helen Salz, PhD
Committee Chair:
Peter Harte, PhD
Committee Member:
Heather Broihier, PhD
Committee Member:
Michelle Longworth, PhD
Committee Member:
Peter Scacheri, PhD
Date of Defense:
May 10, 2019
*We also certify that written approval has been obtained for any proprietary material contained therein
DEDICATION
To my parents. For imparting to me the wisdom that if something is hard work, it
is likely worth having.
i TABLE OF CONTENTS
DEDICATION……………………………………………………………………………..i
TABLE OF CONTENTS…………………………………………………………………ii
LIST OF TABLES………………………………………………………………………..v
LIST OF FIGURES……………………………………………………………………...vi
ABSTRACT………………………………………………………………………………1
CHAPTER 1: INTRODUCTION………………………………………………………..3
1.1 Introduction…………………………………………………………………………..4
1.2 Sexual fate instability increases germ cell tumor risk…………………………….6
1.3 Sexual stability beyond the germline………………………………………………8
1.4 Drosophila gametogenesis…………………………………………………………9
1.5 Drosophila sex determination in the soma and germline……………………….11
1.6 SXL maintains female germ cell identity…………………………………………15
1.7 The H3K9methyltransferase, SETDB1, in regulating female identity…………19
1.8 SETDB1 is required in the female Drosophila germline………………………..22
1.9 Histone 3 Lysine 9 methylation restricts cellular identity……………………….24
1.10 Focus of this dissertation………………………………………………………...28
CHAPTER 2: THE H3K9 METHYLTRANSFERASE SETDB1 MAINTAINS
FEMALE IDENTITY IN DROSOPHILA FEMALE GERM CELLS………………...30
2.1 Abstract……………………………………………………………………………..31
ii 2.2 Introduction…………………………………………………………………………32
2.3 Methods…………………………………………………………………………….34
2.4 Results……………………………………………………………………………...42
2.4.1 SETDB1, WDE, and HP1a loss blocks germ cell
differentiation…………………………………………………………………...42
2.4.2 SETDB1, WDE, and HP1a mutant ovaries express testis genes…...46
2.4.3 H3K9me3 islands correlate with sex-specific gene expression …….52
2.4.4 SXL loss in germ cells interferes with H3K9me3 accumulation……..64
2.5 Discussion………………………………………………………………………….69
CHAPTER 3 FAILURE TO REPRESS PHF7 REPROGRAMS SEXUAL FATE BY
ACTIVATING H3K9ME3 SILENCED TESTIS GENES IN DROSOPHILA
OVARIES……………………………………………………………………………….75
3.1 Abstract……………………………………………………………………………..76
3.2 Introduction…………………………………………………………………………77
3.3 Methods…………………………………………………………………………….79
3.4 Results……………………………………………………………………………...86
3.4.1 phf7 expression in female germ cells is sufficient for
tumorigenesis…………………………………………………………………..86
3.4.2 phf7 is sufficient to activate testis gene expression…………………..88
3.4.3 PHF7 autoregulates its expression by decreasing H3K9me3
enrichment at its testis-specific region……………………………………….93
iii 3.4.4 PHF7 decreases H3K9me3 at additional SETDB1-regulated
genes………………………………………………………………………….101
3.5 Discussion………………………………………………………………………...108
CHAPTER 4 CONCLUDING REMARKS………………………………………….113
4.1 Sex-specific recruitment of SETDB1 in Drosophila female germ cells………115
4.2 PHF7 activation of the testis gene program in females……………………….117
4.2.1 Mechanism of phf7 autoregulation in female germ cells……………117
4.2.2 Activation of the spermatogenesis pathway in germ cell tumors…..124
4.3 Maintenance of sex identity in male germ cells………………………………..127
4.4 Maintenance of sex identity in mammalian germ cells………………………..130
4.4.1 SETDB1 requirement in mammalian germ cells…………………….131
4.4.2 PHF7 expression in mammals………………………………………..133
4.5 Conclusion………………………………………………………………………..136
APPENDIX
Extended Table 1 Genes significantly upregulated in phf7ey/+;HA-phf7,nos-gal4 tumors relative to wild type (yw/yw) ovaries………………………………………..138
Extended Table 2 Genes significantly downregulated in phf7ey/+;HA-phf7,nos-gal4 tumors relative to wild type (yw/yw) ovaries………………………………………..155
REFERENCES……………………………………………………………………….173
iv LIST OF TABLES
Table Title Page 2.1 SETDB1/H3K9me3 regulated genes in ovaries 55
2.2 Location of SETDB1/H3K9me3 regulated genes 59
3.1 Genes with decreased H3K9me3 and ectopic gene 104 expression in phf7ey and snf148 ovaries
Ext. Table 1 Genes significantly upregulated in phf7ey/+;HA-phf7,nos- 138 gal4 tumors relative to wild type (yw/yw) ovaries
Ext Table 2 Genes significantly downregulated in phf7ey/+;HA-phf7,nos- 155 gal4 tumors relative to wild type (yw/yw) ovaries
v LIST OF FIGURES
Figure Title Page
1.1 Diagram of a Drosophila ovariole 10
1.2 Diagram of Drosophila testis 11
1.3 Sex determination in Drosophila somatic cells 12
1.4 Drosophila germ cell sex determination 14
1.5 phf7 is sex-specifically transcribed in Drosophila gonads 17
1.6 SXL facilitates the GSC to CB cell fate switch and represses 18 testis enriched genes
1.7 Screen to identify genes involved in maintaining female identity 19 through germ cell-specific knockdown
1.8 Does SXL regulate sex-specific H3K9me3 deposition via 24 SETDB1 to maintain female germ cell identity?
1.9 H3K9me3 acts as a barrier to cell fate changes 28
2.1 Loss of rhino in germ cells does not lead to global changes in 43 gene expression
2.2 Reduced H3K9me3 staining in germ cells upon SETDB1 and 45 WDE depletion
2.3 Undifferentiated germ cells accumulate in setdb1 GLKD, wde 46 GLKD, and hp1a GLKD mutant germaria
2.4 SETDB1, WDE, and HP1a depletion leads to female-to-male 48 reprogramming at phf7
2.5 Depletion of H3K9me3 pathway members leads to similar 49 differential gene expression
2.6 Depletion of H3K9me3 pathway members results in expression 50 of a similar set of genes not normally expressed in ovaries
2.7 Aside from testis, no predominant tissue-specific signature was 52 identified amongst the genes repressed by SETDB1, WDE, and HP1a
vi 2.8 SETDB1 germline knockdown alters H3K9me3 at a set of 53 protein-encoding genes
2.9 The H3K9me3 peak over the testis-specific phf7-RC TSS is 54 decreased in setdb1 GLKD and snf148 mutant ovaries
2.10 Ectopic phf7 is require for setdb1 GLKD tumor growth 57
2.11 Distribution of the 21 SETDB1/H3K9me3 regulated genes in 58 the genome
2.12 SETDB1 regulates H3K9me3 at CG34434 60
2.13 SETDB1 regulates H3K9me3 at 12477 61
2.14 SETDB1 regulates H3K9me3 at skpE 61
2.15 SETDB1 regulates H3K9me3 at CG32679 61
2.16 SETDB1 regulates H3K9me3 at CG42299 62
2.17 SETDB1 regulates H3K9me3 at CG13423 62
2.18 SETDB1 regulates H3K9me3 at CG15818 62
2.19 SETDB1 regulates H3K9me3 at CG17636 63
2.20 H3K9me3 peaks at the 21 SETDB1-regulated genes are 63 localized over the gene body
2.21 H3K9me3 distribution if affected in snf148 mutant ovaries 65
2.22 SETDB1 protein expression in not altered in snf148 mutants 66
2.23 Diagram of the genetic pathway controlling H3K9me3 67 accumulation in female germ cells
2.24 H3K9me3 is reduced at a set of genes in snf148 mutants 68
2.25 setdb1 GLKD and snf148 influence H3K9me3 accumulation on 68 a similar set of genes
vii 2.26 Schematic summary of discrete facultative heterochromatin 70 island assembly at phf7
3.1 phf7 expression is forced through the insertion of an Upstream 87 Activating Sequence (UAS)
3.2 Forced expression of phf7 in female germ cells results in tumor 87 formation
3.3 A subset of phf7ey differentially expressed genes are 89 “ectopically expressed” or “turned off”
3.4 A subset of snf148 differentially expressed genes are 90 “ectopically expressed” or “turned off”
3.5 Overlap of genes that are turned off and ectopically expressed 91 in phf7ey and snf148 ovaries
3.6 Ovary-specific genes are not turned off in phf7ey and snf148 91 ovaries
3.7 Genes ectopically expressed in phf7ey and snf148 ovaries are 92 testis-specific
3.8 Over half of phf7ey and snf148 shared ectopically expressed 93 genes are testis enriched
3.9 Schematic for a possible genetic feedback mechanism 94 activated with ectopic phf7 in female germ cells
3.10 Ectopic phf7 results in autoregulation 95
3.11 RNAseq and H3K9me3 ChIPseq reads at the phf7 locus 95
3.12 CRISPR/Cas9 gene editing to replace the phf7 coding region 97 with gfp
3.13 phf7 sex-specific transcription is regulated by its non-coding 97 regions
3.14 H3K9me3 enrichment in control (HA-phf7,nos-gal4) and wild 99 type ovaries (yw/yw) ovaries is highly correlated
viii 3.15 Dissolution of H3K9me3 occurs at genes activated in PHF7’s 99 genetic feedback loop
3.16 Forced expression of phf7 does not alter Sxl splicing 100
3.17 Forced phf7 expression alters SXL expression patterns 101
3.18 Forced expression of phf7 does not significantly alter H3K9me3 102 levels
3.19 RNAseq and H3K9me3 ChIPseq reads at the skpE locus 102
3.20 RNAseq and H3K9me3 ChIPseq reads at the CG42299 locus 103
3.21 RNAseq and H3K9me3 ChIPseq reads at the CG34434 locus 103
3.22 Forced phf7 expression alters H3K9me3 enrichment at a 105 subset of gene bodies
3.23 CG6324, CR42767, and CG32706 display reduced H3K9me3 106 with forced phf7 expression
3.24 RNAseq and H3K9me3 ChIPseq reads at the CG6324 locus 106
3.25 RNAseq and H3K9me3 ChIPseq reads at the CR42767 locus 107
3.26 RNAseq and H3K9me3 ChIPseq reads at the CG32706 locus 107
3.27 PHF7 autoregulated its own expression and activates a genetic 108 feedback loop
3.28 Hypothesis for how inappropriate PHF7 may alter facultative 110 heterochromatin formation
4.1 Model for how SXL and SETDB1 sex-specifically deposit 116 H3K9me3
4.2 Deletion of the phf7 histone binding region 119
4.3 Tumor formation require phf7’s PHD domain 119
4.4 RNA and H3K4me2 profiles for phf7 in ovaries and testis 121
ix 4.5 RNA and H3K4me2 profiles for CG34434 in ovaries and testis 122
4.6 RNA and H3K4me2 profiles for CG42299 in ovaries and testis 122
4.7 RNA and H3K4me2 profiles for skpE in ovaries and testis 123
4.8 Prediction for PHF7 autoregulation through chromatin 123 interactions
4.9 Transcripts specific to male meiosis and translation are 126 upregulated in female germ cell tumors
4.10 PHF7 activates expression of DANY in a temporal fashion 127
4.11 Loss of phf7 in bam mutant testis shows minimal changes in 129 gene expression
4.12 Setdb1 knockout in mouse female germ cells upregulates 133 genes normally expressed in testis
4.13 PHF7 protein in humans is only expressed in testis 134
4.14 PHF7 protein expression in human cancers 134
4.15 Annotated phf7 gene regions in Drosophila, mouse, and human 135
x The Requirement Of Facultative Heterochromatin In Maintaining Drosophila
Female Germ Cell Identity
ABSTRACT
by
ANNE E. SMOLKO
Germ cell development and gametogenesis relies on sex-specific signaling and genetic pathways in the developing and adult gonad. Dysregulation of these processes contributes to infertility and germ cell tumor development. Using
Drosophila melanogaster as a model for female germ cell development, I uncover a requirement for facultative heterochromatin in repressing spermatogenesis gene expression and maintaining female identity.
First, I find that deposition of H3K9me3 is required to repress a set of key spermatogenesis genes. Germline knockdown of H3K9me3 pathway members setdb1, wde, and hp1a results in tumor formation and derepression of testis- enriched genes. This regulation occurs independently from H3K9me3-dependent transposable element repression. Genes directly regulated by SETDB1 exist in euchromatin and reside in discrete facultative heterochromatin islands.
Additionally, I determine that H3K9me3 deposition is dependent on the female sex- determining gene, Sxl. SETDB1 and SXL are both required to recruit H3K9me3 at the testis-specific region of PHD finger protein 7 (phf7), a regulator of male identity.
Loss of H3K9me3 enrichment results in testis-specific isoform expression and
1 protein production, contributing to the tumor phenotype. These findings identify
H3K9me3 as a mechanism by which females germ cell retain their sex-identity.
I extend these findings to focus on the molecular consequences of ectopic phf7 expression in female germ cells. Forcing phf7 expression is sufficient for tumor formation and testis gene expression. I further discover that PHF7 in females possesses autoregulatory activity. In addition to activating its own expression, phf7 turns on a genetic feedback mechanism that turns on a set of SETDB1-regulated testis genes through dissolution of H3K9me3. This work underscores the essentiality of maintaining mechanisms that actively repress inappropriate gene expression programs. Overall, I provide a means by which Drosophila female germ cells secure their sex-identity to complete oogenesis. I expect this work to inform our understanding of normal germ cell development, germ cell tumor formation, and infertility in higher organisms.
2 CHAPTER 1
INTRODUCTION
3 1.1 Introduction
Reproduction and the passage of genetic material from one generation to the next is dependent on the production of functional gametes. In both sexes, gametes are produced from germline stem cells (GSCs) that are set aside early in development. Sex-specific development begins as GSCs enter the differentiation pathway and undergo meiosis to produce either sperm or egg. Work in the germline across model organisms has highlighted the importance of maintaining the integrity of these sex-specific mechanisms. The inability to do so often results in developmental and fertility defects. However, how germ cells preserve their sex- identity was unknown.
In this thesis, I utilize the Drosophila germline to further understand how sex-specific genetic programming is maintained. The conservation in biological processes and the wide array of genetic tools available make Drosophila an efficient and advantageous model. Furthermore, Drosophila germ cells can be studied in their natural environment, allowing for in vivo analysis of germ cell development. I entered the Salz lab just as we identified Sex-lethal (Sxl), the female sex-determining gene, as being required for maintaining female germ cell identity through adulthood (Shapiro-Kulnane, Smolko, & Salz, 2015). Loss of germline Sxl results in germ cell tumor formation and a derepression of testis- enriched genes. How Sxl acts to repress testis genes remained an open question.
Information that emerged from genetic screens early in my thesis work helped to generate the hypothesis that repressive Histone 3 Lysine 9 trimethylation
(H3K9me3) maintains Drosophila female germ cell identity. H3K9me3 has
4 emerged as not only a marker of heterochromatin, but also as a barrier to cell fate changes by regulating the expression of tissue-specific genes [Reviewed by
Becker, 2016]. My thesis work identified the H3K9 methyltransferase, SETDB1 as being required to repress spermatogenesis gene expression and maintain female identity. I further identify PHD finger protein 7 (PHF7) as a key testis-specific gene that must actively be repressed by H3K9me3. When ectopically expressed, PHF7 autoregulates its expression and activates expression of additional testis-specific genes. The work presented here furthers our knowledge of the molecular mechanisms underlying maintenance of sex-identity and will inform our understanding of human germ cell development, infertility, and cancers arising from germ cell populations.
In this introduction, I review work that provides the basis for the questions I focus on in this thesis. First, I discuss why sex matters in germ cells and how its disruption results in defects in germ cell development. The existence of sex- specific genetic programming in somatic cells will also be addressed. I next describe Drosophila gametogenesis and the known roles of SETDB1 in the germline. Last, the role of H3K9me3 and SETDB1 in regulating euchromatic gene expression to maintain cellular identity in mammalian systems will be reviewed.
Overall, the studies discussed here provide the background for my exploration of
H3K9me3 in maintaining Drosophila female germ cell identity.
5 1.2 Sexual fate instability increases germ cell tumor risk
The sex-identity of germ cells is critical for gametogenesis and it is important to understand how it is regulated and maintained. Studies in both flies and mice implicate maintenance of sex-specific programming in successful germ cell differentiation and prevention of tumor formation [Reviewed by Salz, 2017]. In flies, misregulation of the sex-determining gene, Sex-lethal (Sxl), in adult female germ cells leads to a germ cell tumor (GCT) phenotype and an upregulation of spermatogenesis genes [Chau, 2009; Shapiro-Kulnane, 2015]. GCTs in mouse testis arise from defects in male-specific developmental timing and the failure to turn on male differentiation genetic pathways [Reviewed by Salz, 2017]. In mice, the timing of meiotic entry differs between males and females. Female germ cells enter meiosis in the embryonic stages. In males, germ cells undergo mitotic arrest and meiosis occurs postnatally. Mouse strains susceptible to germ cell tumors express genes normally only present in premeiotic female germ cells and present delayed mitotic arrest [Heaney, 2012]. Additionally, genes necessary to protect against mitotic entry in males were found to be downregulated [Dawson, 2018].
This failure in the male genetic program causes germ cells to retain pluripotency, contributing to tumor formation.
Misregulation of sex-specific pathways in germ cell tumors in flies and mice extends to the human germline. Testicular germ cells tumors (TGCTs) are the most common solid tumor in men ages 15 to 34 with almost 10,000 new cases a year [Reviewed by Gonzalez-Exposito, 2015]. TGCT risk has been shown to be heritable based on family and twin studies, suggesting a genetic
6 component to tumor susceptibility [Swerdlow, 1997; Hemminki, 2004; Kharazmi et al, 2014; Reviewed by Litchfield, 2016]. In support of this, Genome Wide
Association Studies (GWAS) focusing on testicular cancer have identified mutations in loci containing genes involved in spermatogenesis and male germ cell development as correlating with increased risk for TGCTs [Turnbull, 2010;
Chung, 2013]. One identified susceptibility locus harbors only one protein- encoding gene, Dmrt [Turnbull, 2010]. Dmrt is required for male gonadal sex- identity as its loss reprograms male germ cells to a female fate and increases tumor susceptibility in mice [Raymond, 2000; Krentz, 2009; Matson, 2011]. In an independent GWAS study, Dmrt was also identified as well as other candidates involved in spermatogenesis and male germ cell development [Chung, 2013].
Here, genes located in one linkage disequilibrium (LD) block have all been implicated as being involved in spermatogenesis. While the implications of these mutations in TGCTs have not been fully validated, these GWAS studies support dysregulation of sex-specific genetic pathways in tumor formation.
GCT risk is also increased in individuals with disorders of sex development
(DSD), which are classified by the atypical development of the gonad, sexual organs, or secondary sex characteristics [Cools, 2009]. It has been reported that up to 60% of DSD patients are at risk for developing germ cell cancer. Histological examination of tissues from these patients report that malignant cells can be characterized by the presence of the pluripotency factor, OCT4, which is normally only present in embryonic stem cells [Juniarto, 2012]. Therefore, it may be possible that disorders in germ cell sex development are connected with blocks in GSC
7 differentiation, ultimately leading to tumor susceptibility. Overall, it is clear from these studies in flies, mice, and humans that sexual stability in germ cells is critical for regulation of gene expression, correct meiotic timing, and the ultimate differentiation to either sperm or egg.
1.3 Sexual stability beyond the germline
While work presented in this thesis focuses on sex-identity in germ cell development, it is important to note that sexual dimorphism in somatic adult tissues has also been reported. One example is the gut, where cells age differently in males and females. In flies, female intestinal stem cells (ISCs) proliferate faster relative to males. This allows enhanced growth and repair in younger flies but results in the formation of small tumors with age [Regan, 2016]. This has been attributed to sex-dimporphic transcriptional profiles, where the female sex determining gene, Sxl, regulates differences in gene expression [Hudry, 2016]. In these studies, genetic intestinal feminization in male flies and vice versa resulted in a switch in sex-dimorphic phenotypes, such as intestinal size and tumor susceptibility [Regan, 2016; Hudry, 2016]. This work implicates that the sex of intestinal cells must be actively specified and maintained.
Sex-specificity must also be maintained in the nervous system. Male and female differences in Drosophila neuronal circuitry have been attributed to sex- specific expression of fruitless (fru). Fru has been shown to physically interact with chromatin modifying proteins in a subset of male neurons. This likely regulates sex-specific gene expression to facilitate sex-dimorphic behavior [Ito, 2012]. In
8 human brains sex-specific variances have commonly been attributed to differences in hormone secretion. However, while studies are still limited, evidence of genetic and epigenetic factors contributing to male and female brain differences are emerging. Genes involved in neuronal development and function are prone to escape X chromosome inactivation, leading to sex-specific differences in X-linked gene expression. Female brains have also been found to have higher levels of
DNA methylation [Reviewed by Ratnu, 2016]. Overall, these studies highlight that, even in somatic cells, sex must be actively specified and maintained. Further understanding of how sex-identity is maintained and the consequences of its loss will inform our understanding of sex-biased diseases in both germline and somatic tissues.
1.4 Drosophila gametogenesis
In this work, I utilize the Drosophila germline to further understand how sex- specific genetic programming is maintained. The Drosophila germline has been used as a powerful model to study many of the biological processes involved in germ cell behavior. Adult females possess two ovaries that consist of developing strands of egg chambers termed ovarioles. At one end of each ovariole resides the germarium, where two to three GSCs reside. GSCs are maintained through signals from the surrounding somatic niche, preventing differentiation. As GSCs asymmetrically divide, the cell furthest from the niche will begin to differentiate to form a cystoblast (CB). This cell will undergo 4 mitotic divisions and incomplete cytokinesis to create a 16-cell cyst. One of these cells is destined to become the
9 oocyte. The other 15 will form supportive, polyploid, nurse cells. Follicle cells will form around the 16-cell cyst to form an egg chamber that will eventually end in the generation of a mature oocyte [Reviewed by McLaughlin and Bratu, 2015] (Fig.
1.1).
Figure 1.1 Diagram of a Drosophila ovariole. Drosophila ovarioles consist of a germarium region at the anterior end. Here, 2-3 germline stem cells (GSCs) (magenta) reside. GSCs asymmetrically divide with one cell differentiating to a cystoblast (CB) (green). CB’s mitotically divide to form an interconnected 16-cell cyst. These cells eventually form 15 polyploid nurse cells (light red) that support the fated oocyte (dark red) in developing egg chambers surrounded by follicle cells (gray).
In testis, GSCs also asymmetrically divide to produce a daughter cell that remains in contact with the somatic hub and a differentiating gonialblast (GB). As in females, GBs undergo 4 mitotic divisions to form a spermatogonial cyst.
However, unlike in oogenesis, all 16 of these cells will undergo meiosis to eventually differentiate into sperm [Reviewed in Demarco, 2014] (Fig. 1.2).
Maintenance and differentiation of GSCs in testis and ovaries requires sex-specific signaling to develop and sustain a population of gametes throughout adulthood.
10
Figure 1.2 Diagram of Drosophila testis. Image adapted from Gleason, 2018. Drosophila testis contain a population of germline stem cells (GSCs) (blue) in contact with somatic hub cells (green). GSCs asymmetrically divide with one daughter cell forming the gonialblast (GB) (purple). GBs mitotically divide to form a 16-cell spermatogonial cyst. Each of these cells undergoes meiosis and matures to form functional sperm.
1.5 Drosophila sex determination in the soma and germline
Drosophila somatic sex determination has been extensively studied, with
Sex-lethal (Sxl), an RNA binding protein, being at the top of the hierarchy
[Reviewed by Salz, 2010 and Moschall, 2017]. Sexual fate is initially determined by the number of X chromosomes. The presence of two X chromosomes in females turns on Sxl whereas expression is kept off in males harboring only one.
Once the sex fate decision is initiated, SXL maintains female identity by autoregulating its expression at the level of splicing. SXL then regulates a set of downstream targets to turn on female somatic differentiation via sex-specific splicing of transformer (tra) and double-sex (dsx). It also prevents male dosage compensation through translational repression of the male dosage compensation
11 complex. In males, the absence of SXL prevents functional tra expression and turns on the male form of dsx. It also allows for the activation of the male dosage compensation pathway (Fig. 1.3).
Figure 1.3 Sex determination in Drosophila somatic cells. Drosophila somatic sex is determined by the number of X chromosomes present. In females carrying two copies of the X chromosome, Sxl expression is turned on and is maintained through splicing autoregulation. It then turns of female somatic differentiation through female splicing of transformer (tra), which produces the female form of double-sex (dsx). SXL also prevents male dosage compensation through translational repression. In males with one X chromosome, Sxl expression is kept off and male dosage compensation occurs. No functional tra is produced and the male dsx isoform allows for male somatic differentiation.
In germ cells, genetic pathways different from those utilized in the soma determine and maintain sex-fate decisions. This was illustrated through experiments where pole cells (precursor germ cells) mutant for tra and dsx were transplanted into wild type male or female gonads [March and Wieschaus, 1978;
Schupbach, 1982]. Although components of the sex-determination pathway were no longer active, transplanted germ cells still developed normally based on their chromosomal sex. Later studies ascertained that germ cell sex is initially
12 established early in embryogenesis by a non-cell-autonomous mechanism
[Waterbury, 2000; Casper,2009]. At this time, germ cells initially take on the sex of the surrounding gonad independent of their chromosome constitution (Fig. 1.4).
These conclusions arose by taking advantage of the fact that germ cells do not use the somatic sex-determination pathway. Here, germ cells of one genotype were placed in the somatic environment of another. In these studies, XY germ cells residing in a feminized somatic environment exhibited repressed male gene expression with upregulation of female-specific markers [Waterbury, 2000].
Conversely, XX germ cells in contact with a masculinized somatic environment increased expression of male-specific genes [Casper,2006].
While germ cell sex is initially determined by the soma, its maintenance through adulthood requires autonomous, sex-specific, signals (Fig. 1.4). This was indicated by germ cell transplantation experiments where XY germ cells placed in females and vice versa did not lead to the production of sperm or eggs [Casper,
2006]. In testis, germ cells intrinsically utilize the JAK/STAT pathway to regulate male development [Wawersik, 2005; Casper,2006]. Female germ cells do not exhibit any JAK/STAT activity [Wawersick, 2005]. Rather, work from our lab identified the need for SXL in autonomously regulating maintenance of the female fate in adult stages [Shapiro-Kulnane, 2015]. In females, SXL is not autonomously required for initial germline sex determination or development. Larval gonads lacking Sxl develop normally and ectopic expression does not impact male development. However, in adult stages, loss of Sxl in the germline blocks germ cell differentiation, leading to tumor formation [Steinmann-Zwicky, 1994; Chau,
13 2009]. Additionally, loss of Sxl in female germ cells is accompanied by inappropriate expression of testis genes [Chau, 2009; Shapiro-Kulane, Smolko,
Salz, 2015].
Figure 1.4 Drosophila germ cell sex determination. The sex of the germline is initially determined by that of the surrounding gonad via the somatic sex-determination pathway. In adults, germ cell sex is maintained cell autonomously through JAK/STAT in males and Sxl in females. However, the somatic sex-determination pathway is not utilized.
14 1.6 SXL maintains female germ cell identity
How Sxl functions in the germline was unclear as its downstream somatic sex-determination targets are not utilized in germ cells. This prompted further investigation into what Sxl’s role in female germ cells may be. Work from our lab determined its requirement in the GSC to CB cell fate switch and in the preservation of female identity [Chau, 2009; Chau, 2012; Shapiro-Kulnane, 2015].
In wild type ovaries, SXL protein is highly expressed in the cytoplasm of undifferentiated germ cells [Bopp, 1993; Chau, 2009]. As cells enter the differentiation pathway, cytoplasmic SXL expression decreases and becomes more nuclear. Genetic loss of Sxl in germ cells results in tumors comprised of cells arrested in between the GSC and CB cell fate [Chau, 2009].
As the somatic sex-determination pathway does not function in germ cells, targets of Sxl in the germline remained unknown. Additional work identified nanos
(nos) as a Sxl target [Chau, 2012]. nos is required in the germline to repress differentiation genes and maintain the GSC population [Wang, 2004]. SXL binds nos RNA to downregulate its expression and facilitate differentiation [Chau, 2012].
Genetic epistasis experiments where nos is mutated in germ cells lacking Sxl demonstrated that nos is required for tumor proliferation. However, double mutant ovarioles still resembled tumors even though they do not grow over time.
Furthermore, ectopic expression of nos in a wild type background had no impact on female germ cell differentiation. Therefore, other factors downstream of SXL contribute to tumorigenesis and remained to be identified.
15 While additional direct targets of Sxl in adult germ cells remains unknown, we have shown that it is required for maintaining female identity [Shapiro-Kulnane,
2015]. Loss of germline Sxl is accompanied by inappropriate expression of genes normally restricted to testis [Chau, 2009; Shapiro-Kulnane, 2015]. In this work, it was found that tumors lacking germline Sxl ectopically expressed genes that have are normally expressed in testis with known roles in male meiosis [Shapiro-
Kulnane, 2015].
One of the key testis genes identified was PHD finger protein 7 (phf7)
[Shapiro-Kulnane, 2015], which has been characterized as a regulator of male germ cell identity [Yang, 2012]. Protein expression is normally restricted to undifferentiated male germ cells and its loss has been shown to reduce male fecundity. Genetic mosaic clonal analysis removing phf7 from a subset of cells suggests that it may be involved in male GSC maintenance. Humans express a homologue of phf7 that is also expressed specifically in testis and its expression is able to rescue the fecundity defects observed in phf7 mutant flies. Additionally, in vitro experiments indicate PHF7 can bind histones, specifically H3K4me2, predicting a role in transcriptional regulation. Gene expression analysis of phf7 mutant testis indicated that PHF7 may act as a transcriptional repressor to allow for proper timing of gene expression in the male germline [Yang, 2017].
Inappropriate expression of phf7 contributes to the tumor phenotypes observed [Shapiro-Kulnane, 2015]. Knockdown of phf7 in ovaries lacking germline
Sxl partially rescues tumor formation and restores fertility. Reciprocally, germline- specific overexpression of phf7 results in sterility and tumor formation. It has also
16 been shown that forced phf7 expression in the germline of XX flies with a masculinized soma resulted in sperm production in a small percentage of gonads
[Yang, 2012].
Interestingly, while PHF7 protein is restricted to testis in wild type animals, mRNA is transcribed in both sexes [Shapiro-Kulnane, 2015]. Through alternate transcription start site (TSS) usage, phf7 is sex-dimorphically transcribed to produce an isoform with a longer 5’ Untranslated Region (UTR) in testis (phf7-RC) relative to ovaries (phf7-RA). Sex-specific 5’ UTR generation correlates with translational capacity, where only expression from phf7-RC results in detectable protein (Fig. 1.5). In germ cell tumors where Sxl expression is absent, a switch from phf7-RA to phf7-RC transcription occurs, resulting in detectable protein levels. From these studies, our lab concluded that Sxl acts not only in the GSC to
CB cell fate switch, but is also necessary to repress phf7 and other downstream testis-enriched genes (Fig. 1.6).
Figure 1.5 phf7 is sex-specifically transcribed in Drosophila gonads. Diagram depicted the alternative TSS start site usage that produces sex-specific mRNAs. In testis, a longer 5’ UTR is generated. In ovaries, a more downstream TSS is utilized, producing a shorter 5’ UTR. While both isoforms possess identical coding regions. However, protein is only produced when phf7-RC is expressed.
17 Figure 1.6 SXL facilitates the GSC to CB cell fate switch and represses testis enriched genes. Image adapted from Shapiro-Kulnane et. al, 2015. In female germ cells, SXL and BAM are required to repress nanos to allow GSCs to transition to a cystoblast fate. They are also required to repress a testis gene network to maintain female identity.
As Sxl is at the top of the hierarchy in the sex-determination pathway in somatic cells, its importance in maintaining germ cell female identity was unsurprising. However, how Sxl works to suppress phf7-RC expression was still unclear. SXL is an RNA binding protein that preferentially binds U-rich tracts
[Reviewed by Salz, 2010 and Moschall, 2017]. phf7 contains no predicted SXL binding sites, making direct regulation unlikely [Shapiro-Kulnane, 2015]. During my tenure in the lab, I was tasked with identifying processes downstream of Sxl that regulate sex-specific gene expression in the Drosophila female germline. In the following sections, I describe the generation of my hypothesis that female germ cell identity is maintained through repressive histone 3 lysine 9 trimethylation
(H3K9me3).
18 1.7 The H3K9 methyltransferase, SETDB1, in regulating female identity
To identify additional factors in maintaining Drosophila female germ cell identity, we screened for genes required for germ cell differentiation and repression of phf7-RC. Here, we utilized data from a genome wide genetic screen identifying genes involved in female germ cell function [Yan, 2014]. This screen identified 37 genes that, when knocked down, formed germ cell tumors. Using germ cell-specific knockdown, we rescreened these genes and assessed ovaries for tumor formation and testis-specific phf7-RC expression (Fig. 1.7). Doing so identified the H3K9 methyltransferase, setdb1, as possibly being involved in maintaining female germ cell identity.
Figure 1.7 Screen to identify genes involved in maintaining female identity through germ cell-specific knockdown. To identify additional factors required for female germ cell identity, genes were knocked down in the germline. Ovaries were assessed for both tumor formation and phf7-RC expression. Our screen identified the H3K9 methyltransferase, setdb1 as a candidate.
19 The SETDB1 protein is conserved between flies, worms, and mammals and is necessary for H3K9me3 deposition. H3K9me3 is an established feature of heterochromatin and has roles in chromosome structure and gene silencing
[Reviewed by Elgin, 2013]. Two additional H3K9 methyltransferases exist in flies and are encoded by Su(var)3-9, and G9a [Schultz, 2002; Tachibana, 2002;
Eskeland, 2004]. Of the three known methyltransferases, only germline knockdown of setdb1 produces a tumor phenotype [Yan, 2014]. SETDB1 has known roles in silencing endogenous retroviruses (ERVs) to maintain genome stability. It has also been postulated that SETDB1 influences gene expression by regulating higher order chromatin structure, guiding inactive regions to the nuclear periphery [Reviewed by Mozzetta, 2015].
Work discussed in this thesis additionally focuses on H3K9me3 pathway members windei/atf7IP (wde) and heterochromatin protein 1a (hp1a). We rationalized that if SETDB1-mediated H3K9me3 is involved in maintaining female identity, knockdown of wde and hp1a would yield similar phenotypes. In both flies and humans, WDE/ATP7IP has been identified as an essential cofactor of
SETDB1 [Koch, 2009; Timms, 2016]. WDE is required for SETDB1 nuclear localization and its loss decreases H3K9me3 in both HeLa cells and Drosophila female germ cells. Interestingly, GWAS studies identified variants in the wde/atf7IP locus that were associated with a higher risk of testicular germ cell tumor development [Turnbull, 2010]. HP1a recognizes and binds H3K9me3, recruiting additional enzymes that facilitate the spreading of heterochromatin, including lysine deacetylases and methyltransferases [Reviewed by Eissenberg and Elgin,
20 2010]. The concerted activities of these proteins are required for the formation of higher order heterochromatin structures that promote genome integrity and regulation of gene expression.
While research in sex-specific gene regulation in adult tissues is relatively limited, work in Drosophila somatic cells implicates repressive chromatin landscapes in regulating sex-specific gene expression. In S2 cells, clusters of testis-specific transcripts are embedded in repressed chromatin domains that associate with the nuclear envelope [Shevelyov, 2009]. Detachment from the nuclear envelope results in increased expression of these genes. In the brain, fruitless (fru) is the master regulator of male specific behaviors [Ito, 2012]. fru interacts with either HDAC1, a histone lysine deacetylase, or HP1a. These complexes have been shown to feminize or masculinze neurons, with HP1a causing feminized phenotypes and HDAC1 masculinizing.
This idea extends to mammalian models. In mice, ATAC-seq in gonadal somatic support cells shows that while the open chromatin landscape between males and females is similar very early in development, sex-specific landscapes are eventually developed and maintained [Garcia-Morena, 2018]. In these later stages of development, altered regions of open chromatin are associated with sex- specific gonadal genes. Here, it was found that open chromatin regions diminished at genes promoting male gonad development in females and vice versa.
Additionally, data supporting H3K9me3 as a barrier to cell fate changes in embryonic development and cellular reprogramming has emerged [Reviewed by
Becker, 2016].
21 1.8 SETDB1 is required in the female Drosophila germline
The results of our genetic screen support previous work implicating the requirement of setdb1 and H3K9me3 in Drosophila germ cells. setdb1/eggless was identified through an EMS mutagenic screen to uncover female sterile mutants
[Clough, 2007]. Mutants were found to be homologs of mammalian setdb1 and additional studies confirmed its H3K9me3 enzymatic activity [Clough, 2007; Seum,
2007; Tzeng, 2007]. These studies found that SETDB1 is expressed in early germ cells and its loss decreases H3K9me3 [Clough, 2007; Wang, 2011]. While it is agreed that SETDB1 is required in the germline, different ovarian phenotypes have been observed based on the methods used to eliminate its activity [Clough, 2007;
Wang, 2011; Rangan, 2011; Clough, 2014]. Mutant setdb1 alleles result in oogenesis arrest with ovarioles containing undifferentiated cells [Clough, 2007;
Wang, 2011]. However, this does not address whether the phenotype is due to
SETDB1 activity in ovarian somatic or germ cells. Additional experiments aimed to address the role of SETDB1 specifically in germ cells, yielding conflicting results and conclusions [Wang, 2011; Rangan, 2011; Clough, 2014]. In mosaic clonal analysis, the subset of GSCs harboring setdb1 mutations were not maintained, indicating its need for survival and proliferation [Wang, 2011; Clough, 2014].
However, knockdown in germ cells through RNAi resulted in tumorous germaria with GSC-like cells, pointing to a role for SETDB1 in facilitating differentiation
[Rangan, 2011; Wang, 2011]. It is possible that these contradictory results are due to differences in the experimental techniques. This includes cell competition in
22 mosaic clones where mutant cells are outcompeted by their wild type neighbors and residual SETDB1 activity with RNAi methods.
SETDB1 is thought to enable germ cell development by repressing transposable elements through regulating the production of Piwi-interacting small
RNAs (piRNAs) in the female germline [Rangan, 2011]. piRNA’s are transcribed from heterochromatin clusters to target transposable elements (TEs) for degradation to preserve genome integrity. Loss of setdb1 resulted in decreased piRNA transcription and increased TE levels. However, its impact on protein- encoding genes was not explored. Furthermore, knockdown of rhino, an hp1-like gene specifically required for germline piRNA production, yielded no fertility defects [Mohn, 2014]. This left open the possibility that SETDB1-mediated
H3K9me3 may have an additional role in preserving female identity. In this work, I postulate that H3K9me3 pathway members regulate sex-specific H3K9me3 deposition, possibly with SXL. I ask if their presence represses expression of the testis-specific phf7 isoform and additional testis genes to ensure female fate and prevent tumor formation (Fig. 1.8).
23
Figure 1.8 Does SXL regulate sex-specific H3K9me3 deposition via SETDB1 to maintain female germ cell identity? Possible genetic pathway hypothesizing that H3K9me3 is deposited in a sex-specific manner in the female germline to prevent a testis gene expression program
1.9 Histone 3 Lysine 9 methylation restricts cellular identity
In flies, H3K9me3 has mostly been studied in the context of constitutive heterochromatin that spans large, repeat rich regions of the genome [Reviewed by
Corradini, 2007 and Elgin 2013]. These works have focused on H3K9me3 involvement at pericentric chromosome regions, transposable element repression, and gene regulation on the highly condensed 4th chromosome [Seum, 2007;
Brower-Toland, 2009; Rangan, 2011; Mohn, 2014; Penke, 2016]. Little attention
24 has been given to whether or not it is involved in regulating euchromatic regions of the genome.
Recent work in organisms outside of Drosophila has implicated H3K9me3 in facultative heterochromatin formation and euchromatic, lineage-specific, gene regulation [Reviewed by Becker, 2016]. In fission yeast, H3K9me3 was found to be located at discrete regions of the genome [Zofall, 2012]. These facultative heterochromatin islands were associated with meiotic genes that are repressed when cells are in a vegetative state. Studies such as these challenge the accepted belief that facultative heterochromatin is generally mediated by H3K27me3
[Reviewed by Becker, 2016]. However, not all lineage-inappropriate and dynamic gene repression can be explained by H3K27me3. Regions marked by H3K27me3 are still accessible to general transcription factor binding and RNA polymerase
[Breiling, 2001; Dellino, 2004]. This in in contrast to H3K9me3, where binding of these factors is prevented [Soufi, 2012]. Furthermore, work in oligodendrocyte precursor cells (OPCs) and differentiated oligodendrocytes (OLs) identifies
H3K9me3, rather than H3K27me3, in facilitating differentiation [Liu, 2015].
Examination of H3K9me3 and H3K27me3 enrichment showed an increase only in
H3K9me3 in later OL stages. H3K9me3 was observed mainly over gene bodies and corresponded with repression of genes regulating properties specific to OPCs.
Furthermore, inhibition of lysine 9 methyltransferases, not lysine 27, impeded differentiation. This implicates H3K9me3, in addition to H3K27me3, as an important aspect of lineage restriction in dynamically regulating cell type specific gene expression.
25 H3K9me3 enrichment is both dynamic and transient during multicellular development [Wang, 2018; Nicetto, 2019]. During mouse embryogenesis, dynamic
H3K9me3 patterns exist from the oocyte, through fertilization, into the early embryonic stages [Wang, 2018]. Loss and gain of H3K9me3 enrichment occurs in a developmentally appropriate manner. In these studies, H3K9me3 domains could be classified in a stage-specific manner and were transient on stage-specific genes. H3K9me3 domains marked “oocyte-specific” were lost soon after fertilization. These domains were identified on the promoters of genes involved in zygotic activation, which would need to remain off until the appropriate developmental timepoint. In another study, H3K9me3 loss at protein-encoding genes was observed to allow the expression of tissue-specific genes in adult cell types [Nicetto, 2019]. Here, open chromatin and H3K9me3 profiles were compared between cultured pluripotent germ layered-derived cells and differentiated hepatic and pancreatic cells. These experiments showed a loss of compaction and activation of gene expression at adult function genes of the more mature cell types.
These findings were extended in vivo, where knockdown of setdb1 in the mouse endoderm resulted in failed induction of hepatic markers in liver cells.
Derepression of non-hepatic genes was also observed.
SETDB1-mediated H3K9me3 has also been implicated in facilitating adult cell type stability. In the immune system, naïve CD4+ T cells are capable of differentiating into Th1 or Th2 cells and requires SETDB1 to prevent fate plasticity
[Adoue, 2019]. In this context, it does so by repressing endogenous retroviruses
(ERVs) located in enhancers of Th1-specific genes. Loss of H3K9me3 over these
26 ERVs increased chromatin accessibility, resulting in enhancer activation and gene expression, converting Th-1 cells to a Th-2 fate.
The role of H3K9me3 as a barrier to changes in cell fate extends to the production of induced pluripotent stem cells (iPSCs) [Reviewed by Becker, 2016].
The process of reprogramming somatic cells is inefficient and residual H3K9me3 has been implicated as a barrier to inducing a fully pluripotent state [Chen, 2013].
Loss of H3K9 methyltransferases removes remaining enrichment and greatly improves reprogramming efficiency [Sridharan, 2013]. This work supports studies indicating that open chromatin is a hallmark of the pluripotent state [Ahmed, 2010].
Through electron microscopy, it has been observed that one cell embryos have relatively dispersed chromatin that becomes more compact as cells divide and differentiate. Furthermore, the embryonic stem cell (ESC) genome has been shown to be transcriptionally hyperactive with decreased activity as cells become more lineage-restricted [Efroni, 2008]. Overall, these studies highlight the role of
SETDB1 and H3K9me3 in regulating developmental stages and preventing aberrant adult cell type plasticity (Fig. 1.9). It is likely that H3K9me3’s involvement in cell lineage restriction in embryonic differentiation and cell reprogramming extends to adult stem cell lineages, including germ cells.
27
Figure 1.9 H3K9me3 acts a barrier to cell fate changes. Sourced from Becker, 2016. H3K9 methylation has been shown to be a source of lineage restriction has pluripotent cells terminally differentiate during embryogenesis and prevent reprogramming differentiated cells to a more pluripotent state.
1.10 Focus of this dissertation
The establishment of sex-identity in germ cells is crucial to the completion of gametogenesis. Failure to regulate the genetic programs required to maintain sex-identity increases the risk of GCT formation. Work in Drosophila has established that the female sex-determining gene, Sxl, is required in the germline to maintain female identity by repressing testis-specific genes. phf7, a regulator of male germ cell identity, was identified as a key gene that must remain repressed to prevent tumor formation. A genetic screen in our lab presented the possibility that H3K9me3 pathway members are involved in repression of the testis-specific phf7 isoform and maintenance of female identity. This hypothesis is supported by
28 work outside of Drosophila showing that facultative heterochromatin marked by
H3K9me3 regulates lineage-specific, euchromatic genes. This dissertation aims to explore if H3K9me3 maintains Drosophila female identity and the key spermatogenesis genes that it may repress.
In chapter 2, I determine that Sxl and setdb1 act in concert to sex- specifically deposit discrete H3K9me3 heterochromatin islands on testis-specific genes. I show that this is accomplished independently of the piRNA pathway and
TE silencing. Loss of germline Sxl or setdb1 decreases H3K9me3 at the testis- specific region of phf7 and other key spermatogenesis genes. These results implicate H3K9me3 in maintaining Drosophila female germ cell identity.
Chapter 3 further explores the sufficiency of ectopic phf7 in females to disrupt sex-specific transcription. I show that forced phf7 turns on a testis gene program in female germ cells. Furthermore, phf7 triggers a genetic feedback loop, where it autoregulates itself and turns on transcription genes normally repressed by H3K9me3. While the direct molecular mechanism remains unknown, I show that genetic feedback activation by PHF7 is correlated with H3K9me3 dissolution.
Concluding remarks and future directions for this work are addressed in chapter 4.
29 CHAPTER 2
THE H3K9 METHYLTRANSFERASE SETDB1 MAINTAINS FEMALE
IDENTITY IN DROSOPHILA GERM CELLS
Anne E. Smolko, Laura Shapiro-Kulnane, and Helen K. Salz
A version of this chapter has been published as:
Smolko, A. E., Shapiro-Kulnane, L., & Salz, H. K. (2018). The H3K9 methyltransferase SETDB1 maintains female identity in Drosophila germ cells. Nature communications, 9(1), 4155.
Author Contribution Statement:
A.S and H.S conceived and designed the entire project; L.S.K performed western blot experiments and contributed to confocal image acquisition for H3K9me3 immunostaining; A.S performed all other experiments and analyzed all RNAseq and ChIPseq datasets
30 2.1 Abstract
The preservation of germ cell sexual identity is essential for gametogenesis.
Here we show that H3K9me3-mediated gene silencing is integral to female fate maintenance in Drosophila germ cells. Germ cell specific loss of the H3K9me3 pathway members, the H3K9 methyltransferase SETDB1, WDE, and HP1a, leads to ectopic expression of genes, many of which are normally expressed in testis.
SETDB1 controls the accumulation of H3K9me3 over a subset of these genes without spreading into neighboring loci. At phf7, a regulator of male germ cell sexual fate, the H3K9me3 peak falls over the silenced testis-specific transcription start site. Furthermore, H3K9me3 recruitment to phf7 and repression of testis- specific transcription is dependent on the female sex determination gene Sxl.
Thus, female identity is secured by an H3K9me3 epigenetic pathway in which Sxl is the upstream female-specific regulator, SETDB1 is the required chromatin writer, and phf7 is one of the critical SETDB1 target genes.
31 2.2 Introduction
In metazoans, germ cell development begins early in embryogenesis when the primordial germ cells are specified as distinct from somatic cells. Specified primordial germ cells then migrate into the embryonic gonad, where they begin to exhibit sex-specific division rates and gene expression programs, ultimately leading to meiosis and differentiation into either eggs or sperm. Defects in sex- specific programming interferes with germ cell differentiation leading to infertility and germ cell tumors [Reviewed by Lesch, 2012; Reviewed by Salz, 2017].
Successful reproduction, therefore, depends on the capacity of germ cells to maintain their sexual identity in the form of sex-specific regulation of gene expression.
In Drosophila melanogaster, germ cell sexual identity is specified in embryogenesis by the sex of the developing somatic gonad [Murray, 2010].
However, extrinsic control is lost after embryogenesis and sexual identity is preserved by a cell-intrinsic mechanism. The SEX-LETHAL (SXL) female-specific
RNA binding protein is an integral component of the cell-intrinsic mechanism, as loss of SXL specifically in germ cells leads to a global upregulation of spermatogenesis genes and a germ cell tumor phenotype [Shapiro-Kulnane,
2015]. Remarkably, sex-inappropriate transcription of a single gene, PHD finger protein 7 (phf7), a key regulator of male identity [Yang, 2012], is largely responsible for the tumor phenotype [Shapiro-Kulnane, 2015]. Depletion of phf7 in mutants lacking germline SXL suppresses the tumor phenotype and restores oogenesis.
Moreover, forcing PHF7 protein expression in ovarian germ cells is sufficient to
32 disrupt female fate and give rise to a germ cell tumor. Interestingly, sex-specific regulation of phf7 is achieved by a mechanism that relies primarily on alternative promoter choice and transcription start site (TSS) selection. Sex-specific transcription produces mRNA isoforms with different 5’ untranslated regions that affect translation efficiency, such that PHF7 protein is only detectable in the male germline [Shapiro-Kulnane, 2015; Yang, 2012; Yang, 2017]. Although the SXL protein is known to control expression post-transcriptionally in other contexts
[Reviewed by Salz, 2010], the observation that germ cells lacking SXL protein show defects in phf7 transcription argues that Sxl is likely to indirectly control phf7 promoter choice. Thus, how this sex-specific gene expression program is stably maintained remains to be determined.
Here, we report our discovery that female germ cell fate is maintained by an epigenetic regulatory pathway in which SETDB1 (aka EGGLESS, KMT1E and
ESET) is the required chromatin writer and phf7 is one of the critical SETDB1 target genes. SETDB1 trimethylates H3K9 (H3K9me3), a feature of heterochromatin
[Brower-Toland, 2009; Reviewed by Elgin, 2013]. Using tissue-specific knockdown approaches we establish that germ cell specific loss of SETDB1, its protein partner
WINDEI (WDE, aka ATF7IP, MCAF1 and hAM [Koch, 2009], and the H3K9me3 reader, Heterochromatin binding protein 1a (HP1a, encoded by the Su(var)205 locus [Eissenberg, 2014], leads to ectopic expression of euchromatic protein- encoding genes, many of which are normally expressed only in testis. We further find that H3K9me3 repressive marks accumulate in a SETDB1 dependent manner at 21 of these ectopically expressed genes, including phf7. Remarkably, SETDB1
33 dependent H3K9me3 deposition is highly localized and does not spread into neighboring loci. Regional deposition is especially striking at the phf7 locus, where
H3K9me3 accumulation is restricted to the region surrounding the silent testis- specific TSS. Lastly, we find that H3K9me3 accumulation at many of these genes, including phf7, is dependent on Sxl. Collectively our findings support a model in which female fate is preserved by deposition of H3K9me3 repressive marks on key spermatogenesis genes.
2.3 Methods
Drosophila stocks and culture conditions
Fly strains were kept on standard medium at 25oC unless otherwise noted.
Knockdown studies were carried out with the following lines generated by the
Drosophila Transgenic RNAi Project [Hu, 2017]: setdb1-P{TRiP.HMS00112}
(BDSC #34803, RRID: BDSC_34803), Su(var)205/HP1a-P{TRiP.GL00531}
(BDSC #36792, RRID: BDSC_36792), and wde-P{TRiP.HMS00205} (BDSC
#33339, BDSC_33339). Different conditions were used to maximize the penetrance of the germ cell tumor phenotype. For knockdown of setdb1 and HP1a the nos-Gal4;bam-Gal80 driver was used [Matias, 2015], crosses were set up at
29oC and adults were aged 3-5 days prior to gonad dissection. For wde knockdown the nos-Gal4 driver was used (BDSC #4937, RRID: BDSC_4937) [van Doren,
1998], crosses were set up at 18oC and adults were transferred to 29oC for 2 days prior to gonad dissection. The following Drosophila stocks were also used in this
34 study: snf148 (BDSC #7398, RRID: BDSC_7398) [Nagengast, 2003], HA-setdb1
[Seum, 2007] and phf7DN18 and PBac(3XHA-Phf7) [Yang, 2012]. Wild-type ovaries are either sibling controls, or y1 w1 (BDSC #1495, RRID: BDSC_1495).
For each experiment described below, sample sizes were not predetermined by statistical calculations, but were based on the standard of the field. In a pool of control or experimental animals, specimens of the correct age and genotype were selected randomly and independently from different vials/bottles. Data acquisition and analysis were not performed blindly.
Immunofluorescence and image analysis
Drosophila gonads were fixed and stained according to standard procedures with the following primary antibodies: mouse a-Spectrin (1:100,
Developmental Studies Hybridoma Bank [DSHB] 3A9 RRID: AB_528473], rabbit a-H3K9me3 (1:1,000, Active Motif Cat# 39162, RRID: AB_2532132), rat a-HA
(1:500, Roche Cat# 11867423001, RRID: AB_390919), mouse a-Sxl (1:100,
DSHB M18 RRID: AB_528464), rabbit a-Vasa (1:2,000, a gift from the Rangan lab), and rat a-Vasa (1:100, DSHB RRID: AB_760351). Staining was detected by
FITC (1:200, Jackson ImmnoResearch Labs) or Alexa Fluor 555 (1:200, Life
Technologies) conjugated species appropriate secondary antibodies. TO-PRO-3
Iodide (Fisher, Cat# T3605) was used to stain DNA. Images were taken on a Leica
SP8 confocal with 1024x1024 pixel dimensions, a scan speed of 600 Hz, and a frame average of 3. Sequential scanning was done for each channel and three Z- stacks were combined for each image. Processed images were compiled with Gnu
35 Image Manipulation Program (GIMP) and Microsoft PowerPoint. The GIMP Hue- saturation tool was used to assign appropriate colors to merged panels. All staining experiments were replicated at least two times. The “n” in the figure legends represents the number of germaria scored from a single staining experiment.
Fluorescence intensity quantification of H3K9me3 was measured in 5 wild- type and mutant germaria using GIMP and normalized to fluorescence intensity of
DNA to control for the number of cells. Each image was obtained under identical conditions and consisted of three Z-stacks, with the germline stem cells in the plane of focus.
Western analysis
Ovary extracts for Western blots were prepared from hand-dissected ovaries from 100 females homogenized in 2X sample buffer (100 mM TRIS, pH
6.8, 10% B-Mercaptoethanol, 4% SDS, 20% glycerol, 0.1% Bromophenol Blue).
Westerns were performed according to standard procedures with the following primary antibodies: rat a-HA high affinity (1:500, Roche # 11867423001, RRID:
AB_390919), mouse a alpha-tubulin (1:500, DSHB #AA4.3, RRID: AB_579793),
Enhanced chemiluminescence (ECL) was used for detection, with the following secondary antibodies: ECL goat anti-rat IgG HRP (1:2,000, Fisher # 45-001-200,
RRID: AB_772207), and ECL sheep anti-mouse IgG HRP (1:2,000, Fisher # 45-
000-679, RRID: AB_772210). For Western Blot Quantitation: Three replicates were used and the relative densities for each band were calculated with ImageJ
[(https://imagej.nih.gov/ij/)]. Adjusted densities were determined by dividing the
36 H3K9me3 relative densities by that of their corresponding loading controls.
Uncropped scans are presented in Supplementary Fig. 6.
qRT-PCR and data analysis
RNA was extracted from dissected ovaries using TRIzol (Invitrogen, Cat#
15596026) and DNase (Promega, Cat# M6101). Quantity and quality were measured using a NanoDrop spectrophotometer. cDNA was generated by reverse transcription using the SuperScript First-Strand Synthesis System Kit
(Invitrogen, Cat# 11904018) using random hexamers. Quantatative real-time
PCR was performed using Power SYBR Green PCR Master Mix (ThermoFisher,
Cat# 4367659) with the Applied Biosystems 7300 Real Time PCR system. PCR steps were as follows: 95oC for 10 minutes followed by 40 cycles of 95oC for 15 seconds and 60oC for 1 minute. Melt curves were generated with the following parameters: 95oC for 15 seconds, 60oC for 1 minutes, 95oC for 15 seconds, and
60oC for 15 seconds. Measurements were taken in biological triplicate with two technical replicates. The phf7-RC levels were normalized to the total amount phf7. Relative transcript amounts were calculated using the 2-DDCt method
[Livak, 2001] . Primer sequences for measuring the total phf7 and phf7-RC levels were: for total phf7, forward GAGCTGATCTTCGGCACTGT and reverse
GCTTCGATGTCCTCCTTGAG; for phf7-RC forward
AGTTCGGGAATTCAACGCTT and reverse GAGATAGCCCTGCAGCCA.
37 RNA-seq and data analysis
For wild-type, setdb1 GLKD, wde GLKD, and HP1a GLKD ovaries: Total
RNA was extracted from dissected ovaries using standard TRIzol (Invitrogen, Cat#
15596026) methods. RNA quality was assessed with Qubit and Agilent
Bioanalyzer. Libraries were generated using the Illumina TruSeq Stranded Total
RNA kit (Cat# 20020599). Sequencing was completed on 2 biological replicates of each genotype with the Illumina HiSeq 2500 v2 with 100bp paired end reads.
Sequencing reads were aligned to the Drosophila genome (UCSC dm6) using
TopHat (2.1.0) [Trapnell, 2009]. Differential analysis was completed using CuffDiff
(2.2.1) [Trapnell, 2012]. Genes were considering differentially expressed if they exhibited a two-fold or higher change relative to wild-type with a False Discovery
Rate (FDR) <0.05. Heat maps were generated using the heatmap.2 function of the gplots R package. Scatter plots were generated using ggplot function in R.
Genes that were expressed in mutant (FPKM ≥ 1) but not expressed in wild- type ovaries (FPKM<1) were called ectopic. Genes normally expressed in testes were identified by interrogating the published mRNA-seq data sets GSE86974
[Shan, 2017] [(http://www.ncbi.nlm.nih.gov/geo/)], as above. Genes whose expression levels were two-fold or higher in bgcn mutant testis relative to wild-type testis were called “early-stage testis genes”. Genes whose expression levels were two-fold or higher in wild-type testis relative to bgcn mutant testis were called “late- stage testis genes”. Genes that were not differentially expressed but expressed in testis (FPKM ≥ 1) were called simply “testis genes”. The RNA-seq data on Fly Atlas
[Leader, 2018] was used to identify genes with testis-specific isoforms, i.e. no
38 expression in any other adult tissue.
Tissue expression clustering of the ectopically expressed genes not normally expressed in testis was performed to identify tissue-specific signatures.
Expression values normalized to the whole fly were extracted from FlyAtlas.
Heatmaps to compare the tissue expression profile of these genes per tissue were generated in R with heatmap.2 (gplots). Genes were clustered and normalized per row.
Screen shots are from Integrated Genome Viewer (IGV). To account for the differences in sequencing depth when creating IGV screenshots, the processed
RNAseq alignment files were scaled to the number of reads in the wild-type file.
This was done with Deeptools bigwigCompare using the scale Factors parameter with a bin size of 5.
For rhino mutant ovaries, differentially expressed genes were identified utilizing published mRNA-seq datasets available from the National Center for
Biotechnology Information's GEO database [(http://www.ncbi.nlm.nih.gov/geo/)] under accession number GSE55824 [Mohn, 2014] as described above.
ChIP-seq and data analysis
For each chromatin immunoprecipitation, 400 pairs of Drosophila ovaries were dissected in PBS plus protease inhibitors. Tissues were fixed in 1.8% formaldehyde for 10 minutes at room temperature and quenched with 225 mM glycine for 5 minutes. Tissues were washed twice and stored at -80oC for downstream applications. Samples were lysed using the protocol from [Lee, 2006].
39 Tissue was placed in lysis buffer 1 (140 mM HEPES pH 7.5, 200 mM NaCl, 1 mM
EDTA, 10% glycerol, 0.5% NP-40, 0.25% Triton X-100), homogenized using sterile beads, and rocked at 4oC for 10 minutes. Tissue was then washed 10 minutes at
4oC in lysis buffer 2 (10 mM Tris pH 8, 200 mM NaCl, 1 mM EDTA, 0.5 mM EGTA).
Tissues were then placed in 1.5 mL lysis buffer 3 (10 mM Tris pH 8, 100 mM NaCl,
1 mM EDTA, 0.5 mM EGTA, 0.1% Na-deoxycholate, 0.5% N-lauroylsarcosine). All buffers were supplemented with protease inhibitors. Chromatin was sheared to
200-700 base pairs using the QSONICA sonicator (Q800R). The chromatin lysate was incubated overnight at 4oC with H3K9me3 antibodies pre-bound to magnetic beads. The beads were prepared as follows: 25 ul Protein A and 25 ul Protein G
Dynabeads (Invitrogen, Cat# 10002D and 10004D) per sample were washed twice with ChIP blocking buffer (0.5% Tween 20, 5 mg/mL BSA), then blocked by rocking at 4oC for 1 hour in ChIP blocking buffer, and then conjugated to 5 ug H3K9me3 antibody (Abcam Cat# 8898 RRID: AB_306848) by rocking in new ChIP blocking buffer at 4oC for 1 hour. Following immunoprecipitation, the samples were washed
6 times with ChIP-RIPA buffer (10 mM Tris-HCl pH 8, 1 mM EDTA, 140 mM NaCl,
1% Triton X-100, 0.1% SDS, 0.1% Na-Deoxycholate), 2 times with ChIP-RIPA/500 buffer (ChIP-RIPA + 500 mM NaCl), 2 times with ChIP-LiCl buffer (10 mM Tris-HCl pH 8, 1 mM EDTA, 250 mM LiCl, 0.5% NP-40, 0.5% Na-deoxycholate), and twice with TE buffer. DNA was eluted from beads with 50 ul elution buffer (10 mM Tris-
HCl pH 8, 5 mM EDTA, 300 mM NaCl, 0.1% SDS) and reverse crosslinked at 65oC for 6 hours. Beads were spun down and eluted DNA was transferred to a new tube and extracted using phenol-chloroform extraction. All buffers were supplemented
40 with protease inhibitors.
ChIP sequencing libraries were prepared using the Rubicon Genomics
Library Prep Kit (Cat# R440406) with 16 amplification cycles. DNA was cleaned and assessed for quality with Qubit and Agilent Bioanalyzer. Sequencing was completed on 2 biological replicates of each genotype with the Illumina HiSeq 2500 v2 with 50 bp single end reads.
H3K9me3 reads were aligned to the Drosophila genome (UCSC dm6) using bowtie2 (2.2.6) [Langmead, 2012], and duplicate reads were removed with samtools (1.3) [Li, 2009]. Peaks were called with MACS (2.1.20150731) using the broadpeaks option with all other paramaters set to default [Zhang, 2008].
Differential peak analysis on all replicates was completed with the DIFFBIND program (2.4.8), using summits=500 and the DESeq2 package
[http://bioconductor.org/packages/release/bioc/vignettes/DiffBind/inst/doc/DiffBin d.pdf]. The mean peak concentration (log2) was calculated by normalizing reads to the total library size and subtracting the corresponding input reads. Differential peak fold changes were calculated by subtracting wild-type mean concentrations from mutant mean concentrations. Mutant peaks were considered significantly altered relative to wild-type if they had a False Discovery Rate (FDR) < 5%. The average H3K9me3 deposition on genes in wild-type ovaries was generated with deeptools (2.5.3), using normalized ChIP reads from wild-type ovaries [Ramirez,
2016]. Screen shots are from Integrated Genome Viewer (IGV).
41 Data availability
RNA-seq and ChIP-seq data sets generated during the course of this study are available from the National Center for Biotechnology Information's GEO database [(http://www.ncbi.nlm.nih.gov/geo/)] under accession number
GSE109852. RNAseq datasets from testis [Shan, 2017] and rhino mutant ovaries
[Mohn, 2014] are available from the National Center for Biotechnology
Information’s GEO database [(http://www.ncbi.nlm.nih.gov/geo/)] under accession numbers GSE86974 and GSE55824. Tissue enrichment datasets are available from the FlyAtlas [(http://flyatlas.org/atlas.cgi)] and FlyAtlas.2
[(http://flyatlas.gla.ac.uk/FlyAtlas2/index.html)].
2.4 Results
2.4.1 SETDB1, WDE, and HP1a loss blocks germ cell differentiation
Of the three Drosophila enzymes known to methylate H3K9, only SETDB1 is required for germline development [Reviewed by Elgin, 2013]. Several studies reported that loss of SETDB1 caused a block in germ cell differentiation, characteristic of a germ cell tumor phenotype [Clough, 2007; Clough, 2014;
Rangan, 2011; Wang, 2011]. Because of the known connection between the germ cell tumor phenotype and ectopic testis gene transcription, we wondered whether
SETDB1 played a role in silencing the expression of testis genes in female germ cells. Previous studies established that SETDB1 is important for Piwi-interacting small RNA (piRNA) biogenesis and transposable element (TE) silencing in germ cells [Rangan, 2011; Siensei, 2015; Yu, 2015]. However, mutations that
42 specifically interfere with piRNA production, such as rhino, complete oogenesis
[Volpe, 2001; Klattenhoff, 2009; Zhang, 2014; Mohn, 2014]. Furthermore, our analysis of published RNA-sequencing (RNA-seq) data from rhino mutant ovaries
[Mohn, 2014] revealed only very minor effects on gene expression (Fig. 2.1).
Together these observations suggest that SETDB1 has a role in germ cell development that is unrelated to its canonical role in piRNA biogenesis and TE silencing.
Figure 2.1 Loss of rhino in germ cells does not lead to global changes in gene expression. (A) Scatter plot comparing gene expression in rhino
mutant ovaries (log2, FPKM) relative to wild-type (WT). The 28 genes with at least a two-fold change in expression (FDR < 0.05) are highlighted in red. (B) rhino does not control sex-specific phf7 transcription. Genome browser view of the phf7 locus. Tracks show RNA-seq reads aligned to the Drosophila genome (UCSC dm6). All tracks are viewed at the same scale. The screen shot is reversed so that the 5’ end of the gene is on the left. The two phf7 transcripts, phf7-RA and phf7-RC, are indicated. phf7-RA is normally expressed in ovaries, and is expressed in both wild-type and rhino mutant ovaries. phf7-RC is normally only expressed in testis, and is not expressed in rhino mutant ovaries.
We first confirmed that the loss of SETDB1 and its binding partner WDE specifically in germ cells was the cause of the germ cell tumor phenotype. To achieve Germ Line specific Knock Down (GLKD), we expressed an inducible RNA
43 interference (RNAi) transgene with nos-Gal4, which is specifically expressed in germ cells. We demonstrated RNAi efficiency by showing that setdb1 GLKD and wde GLKD abolished the intense H3K9me3 staining foci observed in wild-type germ cells (Fig. 2.2). Furthermore, we found the oogenesis defects elicited by setdb1 and wde GLKD to be similar, as judged by the number of round spectrosome like structures present in the germarium. The spectrosome is a spherical alpha-spectrin-containing structure that is normally found only in germline stem cells (GSCs) and its differentiating daughter cell, the cystoblast (~5 per germarium) (Fig. 2.3 A). As differentiation proceeds, the round spectrosome elongates and branches out to form a fusome. We found that the majority of setdb1 and wde GLKD mutant germaria contain 6 or more spectrosome containing germ cells (Fig. 2.3 B,C,I). In wild-type, fusomes degenerate as the 16-cell germ cell cyst, consisting of an oocyte and 15 nurse cells, are enveloped by somatic follicle cells forming an egg chamber (Fig. 2.3 E). In mutants, however, we observed spectrosome containing germ cells enclosed by follicle cells (Fig. 2.3 F,G). This indicates that loss of SETDB1 and WDE in germ cells blocks differentiation, giving rise to a tumor phenotype.
44 Figure 2.2 Reduced H3K9me3 staining in germ cells upon SETDB1 and WDE depletion. Representative confocal images of a (A) wild-type (WT), (B) setdb1 GLKD, and (C) wde GLKD germarium stained for H3K9me3 (green, white in A’’-C’’). Germ cells were identified by a-VASA staining (magenta). Scale bar, 25 mm. Insets show a higher magnification of a single germ cell, in which the nucleus is outlined by a white dashed line in A’’-C’’. (D) Quantification of H3K9me3 intensity, normalized to DNA, in 5 wild-type and mutant germaria. Error bars indicate standard deviation. Significance of the difference between wild-type and mutant was determined by calculating p-values by Student’s T-test from 5 replicates. **p<0.01.
Recently, a large-scale RNAi screen identified a role for H3K9me3 binding protein HP1a in germ cell differentiation [Yan, 2014]. In agreement with their findings, we observed that loss of HP1a in germ cells gave rise to germ cell tumors
(Fig. 2.3 D,H,I). HP1a is required for gene silencing in other contexts [Reviewed by Eissenberg, 2014]. This suggests that HP1a may act in a common pathway with
SETDB1 and WDE in female germ cells.
45 Figure 2.3 Undifferentiated germ cells accumulate in setdb1 GLKD, wde GLKD, and hp1a GLKD mutant germaria. (A-D) Representative confocal images of wild- type and mutant germaria stained for DNA (magenta) and a-spectrin (cyan) to visualize spectrosomes (sp), fusomes (fu), and somatic cell membranes. Scale bar, 25 mm. (E-H) Representative confocal images of wild-type (WT) and mutant ovarioles stained for DNA (magenta) and a-spectrin (cyan, white in E’-H’) to visualize spectrosomes, fusomes, and somatic cell membranes. Egg chambers in mutant ovarioles contain germ cells that retain spectrosomes (yellow arrow head). (I) Quantification of mutant germaria with 0, 1-5, and >5 round spectrosome-containing germ cells. The number of scored germaria (n) is indicated
2.4.2 SETDB1, WDE, and HP1a mutant ovaries express testis genes
To investigate the possibility that the loss of H3K9me3 pathway members in female germ cells leads to ectopic testis gene expression, we first used RT- qPCR to assay for the presence of the testis-specific phf7-RC isoform in mutant ovaries. Using primer pairs capable of detecting phf7-RC, we found that phf7-RC is ectopically expressed in setdb1, wde, and hp1a mutant ovaries (Fig. 2.4 A).
46 Next, we asked whether ectopic phf7-RC expression correlates with ectopic PHF7 protein expression. Previous work using an HA-tagged phf7 locus in the context of an 20 kb BAC rescue construct showed that PHF7 protein is normally restricted to testis [Shapiro-Kulnane, 2015; Yang, 2012]. We found that in contrast to wild-type ovaries, HA-PHF7 protein is detectable in the cytoplasm and in the nucleus of setdb1, wde, and hp1a mutant ovaries (Fig. 2.4 B-E). We therefore conclude that the H3K9me3 pathway members are essential for suppression of testis-specific phf7 transcription and PHF7 protein expression in female germ cells.
To gain a genome-wide view of the expression changes associated with the loss of H3K9me3 pathway members in germ cells, we used RNA-seq to compare the transcriptomes of GLKD mutant ovaries with wild-type ovaries from newborn
(0-24 hour) females. In agreement with our RT-qPCR analysis, we find that the testis-specific phf7 transcript, phf7-RC, is ectopically expressed in setdb1, wde, and hp1a GLKD mutant ovaries (Fig. 2.4 F). In addition to phf7, our differential analysis identified 1191 genes in setdb1 GLKD mutant ovaries, 904 in wde GLKD ovaries, and 1352 in hp1a GLKD ovaries that are upregulated at least 2-fold relative to wild-type (FDR<0.05). Additionally, 657 genes in setdb1 GLKD mutant ovaries, 756 in wde GLKD ovaries, and 877 in hp1a GLKD ovaries are downregulated. Comparison of the differential gene expression profiles of setdb1
GLKD mutant ovaries with wde and hpa1 GLKD mutant ovaries revealed extensive similarities, as expected for genes functioning in the same pathway (Fig. 2.5).
47
Figure 2.4 SETDB1, WDE, and HP1a depletion leads to female-to-male reprogramming at phf7 (A) Depletion of H3K9me3 pathway members leads to ectopic expression of the testis-specific phf7-RC isoform. RT-qPCR analysis of the testis phf7- RC transcript in wild-type testis, wild-type and mutant ovaries. Expression, normalized to the total level of phf7, is shown as fold change relative to testis. Primers are shown in panel F. Error bars indicate standard deviation (s.d.) of three biological replicates. (B-E) Depletion of H3K9me3 pathway members leads to ectopic expression of the testis- specific PHF7 protein. Ovaries from animals carrying an HA-PHF7 transgene stained for HA (green, white in B’-E’). Germ cells were identified by a-VASA staining (magenta). Scale bar, 25 mm. (F) RNA-seq data confirms ectopic expression of the testis-specific phf7-RC isoform. Genome browser view of the phf7 locus. Tracks show RNA-seq reads aligned to the Drosophila genome (UCSC dm6). All tracks are viewed at the same scale. The screen shot is reversed so that the 5’ end of the gene is on the left. The reads that are unique to the mutant ovaries are highlighted in gray. The two phf7 transcripts, phf7- RA and phf7-RC, are indicated. phf7-RC is normally only expressed in testis (blue arrow). phf7-RA is normally expressed in ovaries (pink arrow). Primers for RT-qPCR are indicated by arrowheads: gray for phf7-RC, red for total phf7
48 Figure 2.5 Depletion of H3K9me3 pathway members leads to similar differential gene expression Comparisons of the deregulated gene expression profiles of setdb1, wde and hp1a GLKD ovaries reveals extensive similarities. Heat map comparing changes in gene expression in setdb1, wde, and hp1a GLKD ovaries compared to wild-type (WT) ovaries. Each row depicts a gene whose express is deregulated at least 2-fold (FDR<0.05) in all mutants when compared to wild-type.
Interestingly, all mutants express a set of upregulated genes that are normally not expressed in wild-type ovaries (FPKM<1 in wild-type ovaries Fig. 2.6
A,B). While we found that there was a significant overlap between the genes ectopically expressed in all three mutant backgrounds (Fig. 2.6 B), we did not find that they were enriched for specific gene ontology terms. However, the pivotal role of phf7 in controlling germ cell sex identity suggested to us that many of the ectopically expressed genes might be normally expressed in testis. To test this idea, we compared our data with published RNA-seq analysis of wild-type testis and bgcn mutant testis [Shan, 2017]. In spermatogenesis, bgcn is required for the undifferentiated spermatogonia to stop mitosis and transition into the spermatocyte
49 stage. In bgcn mutants this transition is blocked and the testis are enriched for dividing spermatogonial cells. The comparison of the wild-type and mutant expression profiles can therefore be used to identify genes preferentially expressed in early-stage spermatogonia (>2-fold increase in bgcn compared to wild-type, in blue) and in late-stage spermatocytes (>2-fold decrease in bgcn compared to wild type, in green). We also identified genes that are normally expressed in testis but are not differentially expressed (FPKM > 1 in both samples, in red) and genes that are not detectable in either sample (FPKM <1, in gray). This analysis revealed that 63%-67% of the ectopically expressed genes are expressed in testis (Fig. 2.6C). Because both early and late stage testis genes are ectopically expressed in the mutant ovaries, we conclude that the H3K9me3 pathway genes
SETDB1, WDE and HP1a are required to repress spermatogenesis transcription in female germ cells.
50 Figure 2.6 Depletion of H3K9me3 pathway members results in expression of a similar set of genes not normally expressed in ovaries. (A) Depletion of pathway members leads to ectopic expression of genes which are not normally expressed in ovaries. Scatter plots of significantly upregulated genes in setdb1,
wde, and hp1a GLKD ovaries. The log2 fold change in gene expression is
plotted against the log2 of the FPKM values in wild-type ovaries. Colored points
indicate ectopically expressed genes (log2 FPKM < 0 in WT ovaries). (B) Venn diagram showing overlap of ectopically expressed genes in setdb1, wde, and hp1a GLKD ovaries. The amount of overlap is significantly higher than expected -6 stochastically (P<10 ). The significance of each two-way overlap was assessed -6 using Fisher’s exact test performed in R, in each case yielding P<10 . The significance of the three-way overlap was assessed by a Monte Carlo -6 simulation, yielding P<10 . For each genotype, the number (n) of ectopically expressed genes is indicated. (C) A majority of ectopically expressed genes are normally expressed in testis. Bar chart showing the percentage of ectopically expressed genes in setdb1, wde, and hp1a GLKD ovaries that are normally expressed in wild-type testis. Genes are assigned into groups based on expression in wild-type and bgcn mutant testis (see text for details): early stage spermatogonia genes (>2-fold increase in bgcn testis compared to wild- type testis, in blue), late-stage spermatocyte genes (>2-fold decrease in bgcn testis compared to wild-type testis, in green) and genes detected during both stages (FPKM>1 in both samples, in red). Genes which are not expressed in either testis sample (FPKM<1) are in gray. For each genotype, the number (n) of ectopically expressed genes is indicated.
This analysis also shows that 33%-37% of the ectopically expressed genes are not normally expressed in gonads (gray, Fig. 2.6C), suggesting that the
H3K9me3 pathway also represses somatic gene transcription. However, we did not identify a predominant tissue-specific signature amongst the remaining ectopically expressed genes (Fig. 2.7). Furthermore, and despite ectopic expression of somatic genes, the mutant germ cells retained their germ cell identity, as evidenced by the presence of spectrosomes and fusomes, germ-cell specific organelles, as well as expression of the germ cell marker VASA (Fig.
2.2B,C; 2.3 A-H). These results indicate that function of the H3K9me3 pathway in
51 germ cells is not restricted to repressing the spermatogenesis gene expression program.
Fig. 2.7 Aside from the testis, no predominant tissue-specific signature was identified amongst the genes repressed by SETDB1, WDE and HP1a. Hierarchical clustering of the tissue expression profiles of the ectopically expressed genes not normally expressed in testis genes. Gene expression per tissue (normalized to fly average) is shown as a z-score heatmap.
2.4.3 H3K9me3 islands correlate with sex-specific gene expression
Our studies raise the possibility that SETDB1 prevents linage-inappropriate gene transcription by mediating the deposition of H3K9me3 on its target loci. To test this idea directly, we performed H3K9me3 chromatin immunoprecipitation followed by sequencing (ChIP-seq) on wild-type and setdb1 GLKD ovaries. By
52 limiting the differential peak analysis to euchromatin genes, we identified 746
H3K9me3 enrichment peaks in wild-type that were significantly altered in setdb1
GLKD ovaries (Fig. 2.8). Whereas a majority of the gene associated peaks show the expected decrease in H3K9me3 enrichment (84%, 630/746), we also observed regions with an increase in H3K9me3 enrichment (15%, 116/746). How loss of
SETDB1 might lead to an increase in H3K9me3 is not known, but the effect is most likely indirect.
At the phf7 locus, H3K9me3 is concentrated over the silenced testis-specific
TSS in wild-type ovaries (Fig. 2.9). Reduction of this peak in setdb1 GLKD ovaries correlates with aberrant testis-specific transcription (Fig. 2.4F). These data suggest a functional link between the presence of repressive H3K9me3 chromatin and the silencing of phf7 testis-specific isoform transcription.
Figure 2.8 SETDB1 germline knockdown alters H3K9me3 at a set of protein- encoding genes. Differential analysis of paired H3K9me3 ChIP-seq data sets identifies SETDB1-dependent H3K9me3 peaks. Scatter plot showing the significantly altered (FDR<0.05) H3K9me3 peaks in setdb1 GLKD ovaries relative
to wild-type (WT) ovaries. The x-axis is the log2 H3K9me3 enrichment of wild-type
peaks subtracted from the log2 H3K9me3 enrichment of setdb1 GLKD peaks. Negative values indicate a reduction in H3K9me3 in mutant ovaries.
53 Figure 2.9 The H3K9me3 peak over the testis-specific phf7-RC TSS is 148 decreased in setdb1 GLKD and snf mutant ovaries. Genome browser view of phf7 and neighboring genes CG9577 and rab35. Tracks are shown on the same scale with the 5’ end of phf7 on the left
We identified an additional 24 normally silenced euchromatic genes at which reduction of the H3K9me3 peak in setdb1 GLKD ovaries correlates with ectopic expression. Of these genes, 4 contain transposable element (TE) sequences. Prior studies have shown that H3K9me3 is enriched around euchromatic TE insertion sites, suggesting the possibility that transcriptional repression might result from spreading of H3K9me3 from a silenced TE [Sienski,
2012; Lee, 2015; Lee, 2017]. The absence of TE sequences at phf7 and the other
20 genes suggests H3K9me3 deposition is controlled by a different mechanism.
Like phf7, the majority of these 20 SETDB1-regulated genes are normally expressed in spermatogenesis (Table 2.1). Furthermore, examination of their expression pattern in adult tissues, as reported in FlyAtlas [Leader, 2018], indicates that 8 of these genes express at least one testis-specific isoform. Of the
54 remaining genes, 7 express isoforms in the testis and other tissues, and 5 are not normally expressed in the adult testis. Further studies are needed to determine whether repression of these genes is as important for female germ cell development as phf7.
Table 2.1 SETDB1/H3K9me3 regulated genes in ovaries
GENE EXPRESSION PATTERN PREDICTED FUNCTION IN ADULTS (SEE TEXT) genes normally expressed only in testis, or genes with testis-specific transcripts phf7 early stage testis specific H3K4me2 binding transcript, phf7-RC skpE early-stage spermatogonia SKP1 gene family CG12477 early-stage spermatogonia Ring finger domain/E3 ligase CG42299 both early and late Zinc finger MIZ-type/E3 ligase CG12061 both early and late Sodium/calcium exchanger CG13423 late-stage spermatocyte Peptidase CG15172 late-stage spermatocyte Unknown CG34434 late-stage spermatocyte Unknown CR43299 late-stage spermatocyte ncRNA genes normally expressed in testis and other tissues Lim1 early-stage Homeobox transcription factor spermatogonia; brain/CNS CG32506 early-stage Rab GTPase activating proteins spermatogonia; brain/CNS CG42613 early-stage CUB domain spermatogonia; brain/CNS CG10440 late-stage spermatocyte; BTB/POZ domain brain/CNS CG10483 late-stage spermatocyte; SPX and EXS domains brain/CNS CG17636 late-stage spermatocyte; Hydrolase gut CG31202 late-stage spermatocyte; Alpha-mannosidase class I crop genes normally not expressed in testis CG12607 no expression Unknown CG32679 no expression Secretory protein CG15818 gut Carbohydrate binding MsR1 brain/CNS Transmembrane G protein coupled receptor Rab3-GEF Brain/CNS GDP-GTP exchange factor
55 Table 2.1 SETDB1-regulated genes in ovaries. Table of genes ectopically expressed with decreased H3K9me3 enrichment in setdb1 GLKD ovaries. Table lists testis expression based on RNAseq analysis and tissue enrichment according to FlyAtlas. Predicted gene function is described based on information provided by FlyBase.
Previous studies have shown that ectopic PHF7 protein expression is sufficient to disrupt female fate and give rise to a germ cell tumor [Shapiro-
Kulnane, 2015]. We therefore asked if ectopic PHF7 contributes to the setdb1
GLKD mutant phenotype by generating double mutant females. We found that while loss of phf7 did not restore oogenesis, there was a shift in the distribution of mutant phenotypes such that the majority of phf7DN18 ; setdb1 GLKD double mutant ovarioles contained no germ cells (Fig. 2.10). While these data indicate that phf7 is a critical target of SETDB1 silencing, our finding that the phenotype was not fully rescued suggests that ectopic expression of one or more of the other SETDB1 target genes we identified in this study also contribute to the tumor phenotype.
56 Figure 2.10 Ectopic phf7 is required for setdb1 GLKD tumor growth. (A-C) Germaria from wild-type, setdb1GLKD, or double mutant D18 D18 phf7 /phf7 ; setdb1 GLKD females stained for DNA (magenta) and a- spectrin (cyan) to visualize spectrosomes, fusomes, and somatic cell membranes. Scale bar, 25 mm. (D) Quantification of mutant germaria with 0, 1-5, and >5 round spectrosome-containing germ cells. The number of scored germaria (n) is indicated
Co-regulated genes are often clustered together in the genome. In
Drosophila, about a third of the testis-specific genes are found in groups of three or more genes [Parisi, 2004; Shevelyov, 2009]. However, the 21 SETDB1- regulated genes we have identified do not fall within the previously identified testis gene clusters, nor are they clustered together in the genome (Fig. 2.11, Table 2.2).
Even on the X chromosome, where 11 of the 21 genes are located, the closest two genes are 100 kb away from each other. Thus, the SETDB1 regulated genes are not located within co-expression domains.
57
Fig. 2.11 Distribution of the 21 SETDB1/H3K9me3 regulated genes in the genome. The 21 genes are not clustered together in the genome. Gene positions are shown on the six major chromosome arms. Chromosome length is indicated in megabase pairs (Mb). See Supplementary Table 1 for the exact position of each gene. The normal expression pattern of these genes is indicated as follows: genes with testis-specific isoform (dark blue), genes expressed in testis and other tissues (turquoise) and genes not normally expressed in adult testis (red). See Table 1 for details.
58
Table 2.2
Location of SETDB1/H3K9me3 regulated genes
Distance from Chromosome Cytogenetic nearest gene & sequence location map neighbor (Mb) CG17636 X:124,370..126,714 [-] 1A -- CG34434 X:5,614,554..5,616,370 [-] 5A8 5.5 Lim1 X:8,756,539..8,805,804 [-] 8A5-8B2 3.1 CG32679 X:10,366,402..10,367,472 [-] 9B12 1.6 Rab3-GEF X:15,089,682..15,113,762 [+] 13A10-12 4.7 CG42299 X:15,574,947..15,575,683 [-] 13D 0.5 CR43299 X:15,662,289..15,665,707 [-] 13E 0.1 SkpE X:19,823,368..19,824,080 [-] 18F 4.2 Phf7 X:20,160,355..20,165,830 [-] 19B3-19C1 0.3 CG32506 X:20,449,331..20,467,065 [+] 19D1 0.3 CG12061 X:23,036,066..23,047,678 [+] 20F4 2.6 CG15818 2L:7,410,909..7,412,229 [+] 27F3 -- CG15172 2L:19,022,417..19,023,646 [+] 37B9 11.6 CG13423 2R:20,536,578..20,538,264 [+] 57A5 -- CG10440 2R:21,560,246..21,576,035 [-] 57F1-2 1.0 MsR1 3L:2,322,693..2,345,799 [+] 62D6 -- CG12607 3L:4,446,573..4,448,072 [+] 64B6 2.1 CG10483 3L:5,917,905..5,921,316 [-] 64F5 1.5 CG12477 3L:18,721,893..18,723,272 [-] 75D7 12.8 CG42613 3R:18,929,196..18,978,088 [+] 91D5-91E1 -- CG31202 3R:29,928,701..29,930,386 [-] 99C7 10.9
Table 2.2 Genomic locations of SETDB1-regulated genes. Table indicating the chromosomal locations and cytogenic map coordinates of SETDB1-regulated genes. Their distance to their closest SETDB1-regulated neighbor is reported
59 In fact, the H3K9me3 peaks at the 21 SETDB1-regulated genes are highly localized and do not spread into the neighboring genes (Fig. 2.12-12.19).
Averaging the H3K9me3 distribution over these 21 genes, scaled to 1.5 kb and aligned at their 5’ and 3’ ends, demonstrates a prominent enrichment over the gene body (Fig. 2.20, in blue). In contrast, the average enrichment profile over all euchromatic genes with H3K9me3 peaks showed a broader pattern extending both upstream and downstream of the gene body (Fig. 2.20, in red). Together these results clearly show that silencing involves formation of gene-specific blocks of
H3K9me3 islands at a select set of testis genes.
Figure 2.12 SETDB1 and SXL regulate H3K9me3 at CG34434. Genome browser view of CG34434 illustrates the H3K9me3 islands present in WT 148 ovaries that is decreased in setdb1 GLKD and snf ovaries. H3K9me3 148 reads are shown for WT (black), setdb1 GLKD (green), and snf (red) ovaries. RNAseq reads are shown for WT (blue) and setdb1 GLKD (orange) ovaries. Tracks are shown at the same scale with the 5’ end on the left. Shading highlights the gene region of interest.
60
Figure 2.13 SETDB1 regulates H3K9me3 at CG12477. Genome browser view of CG12477 illustrates the H3K9me3 islands present in WT ovaries that is decreased in setdb1 GLKD. H3K9me3 reads are shown for WT (black), setdb1 GLKD (green) ovaries. RNAseq reads are shown for WT (blue) and setdb1 GLKD (orange) ovaries. Tracks are shown at the same scale with the 5’ end on the left. Shading highlights the gene region of interest.
Figure 2.14 SETDB1 regulates H3K9me3 at skpE. Genome browser view of skpE illustrates the H3K9me3 islands present in WT ovaries that is decreased in setdb1 GLKD. H3K9me3 reads are shown for WT (black), setdb1 GLKD (green) ovaries. RNAseq reads are shown for WT (blue) and setdb1 GLKD (orange) ovaries. Tracks are shown at the same scale with the 5’ end on the left. Shading highlights the gene region of interest.
Figure 2.15 SETDB1 regulates H3K9me3 at CG32679. Genome browser view of CG32679 illustrates the H3K9me3 islands present in WT ovaries that is decreased in setdb1 GLKD. H3K9me3 reads are shown for WT (black), setdb1 GLKD (green) ovaries. RNAseq reads are shown for WT (blue) and setdb1 GLKD (orange) ovaries. Tracks are shown at the same scale with the 5’ end on the left. Shading highlights the gene region of interest.
61
Figure 2.16 SETDB1 regulates H3K9me3 at CG42299. Genome browser view of CG42299 illustrates the H3K9me3 islands present in WT ovaries that is decreased in setdb1 GLKD. H3K9me3 reads are shown for WT (black), setdb1 GLKD (green) ovaries. RNAseq reads are shown for WT (blue) and setdb1 GLKD (orange) ovaries. Tracks are shown at the same scale with the 5’ end on the left. Shading highlights the gene region of interest.
Figure 2.17 SETDB1 regulates H3K9me3 at CG13423. Genome browser view of CG13423 illustrates the H3K9me3 islands present in WT ovaries that is decreased in setdb1 GLKD. H3K9me3 reads are shown for WT (black), setdb1 GLKD (green) ovaries. RNAseq reads are shown for WT (blue) and setdb1 GLKD (orange) ovaries. Tracks are shown at the same scale with the 5’ end on the left. Shading highlights the gene region of interest.
Figure 2.18 SETDB1 regulates H3K9me3 at CG15818. Genome browser view of CG15818 illustrates the H3K9me3 islands present in WT ovaries that is decreased in setdb1 GLKD. H3K9me3 reads are shown for WT (black), setdb1 GLKD (green) ovaries. RNAseq reads are shown for WT (blue) and setdb1 GLKD (orange) ovaries. Tracks are shown at the same scale with the 5’ end on the left. Shading highlights the gene region of interest.
62
Figure 2.19 SETDB1 regulates H3K9me3 at CG17636. Genome browser view of CG17636 illustrates the H3K9me3 islands present in WT ovaries that is decreased in setdb1 GLKD. H3K9me3 reads are shown for WT (black), setdb1 GLKD (green) ovaries. RNAseq reads are shown for WT (blue) and setdb1 GLKD (orange) ovaries. Tracks are shown at the same scale with the 5’ end on the left. Shading highlights the gene region of interest.
Figure 2.20 H3K9me3 peaks at the 21 SETDB1-regulated genes are localized over the gene body. The average H3K9me3 enrichment profile on the average gene body (transcription start site to transcription end site) scaled to 1500 base pairs, ± 3 kb. In red, the average enrichment profile of euchromatic genes which display an H3K9me3 peak in wild-type ovaries. In blue, the average H3K9me3 profile of the 21 genes which are both ectopically expressed and display a loss of H3K9me3 enrichment in setdb1 GLKD ovaries.
63 2.4.4 SXL loss in germ cells interferes with H3K9me3 accumulation
Previous studies have established that sex-specific phf7 transcription is controlled by the female sex determination gene Sxl [Shapiro-Kulnane, 2015].
Together with studies demonstrating that SXL protein is expressed in setdb1 mutant germ cells [Clough, 2014], these data suggest that SXL acts upstream, or in parallel to SETDB1 to control phf7 transcription. To assess the potential of a Sxl- mediated mechanism, we asked whether the loss of SXL in germ cells affects
H3K9me3 accumulation. As with our earlier studies, we take advantage of the viable sans-fille148 (snf148) allele to selectively eliminate SXL protein in germ cells without disrupting function in the surrounding somatic cells. SNF, a general splicing factor, is essential for Sxl autoregulatory splicing [Reviewed by Salz, 2010]. The viable snf148 allele disrupts the Sxl autoregulatory splicing loop in female germ cells, leading to a failure in SXL protein accumulation, masculinization of the gene expression program (including phf7), and a germ cell tumor phenotype [Shapiro-
Kulnane, 2015; Nagengast, 2003; Chau, 2009; Chau, 2012]. All aspects of the snf148 mutant phenotype described to date are restored by germ cell-specific expression of a Sxl cDNA. Therefore, the analysis of snf148 mutant ovaries directly informs us of Sxl function in germ cells. Interestingly, we found that that the intense
H3K9me3 staining foci observed in wild-type germ cells was reduced in snf148 mutants (Fig. 2.2, 2.21 A,B). However, two lines of evidence indicate that SETDB1 protein expression, measured by an HA tag knocked into the endogenous locus33, was not disrupted in snf148 mutant germ cells. First, in whole mount immunostaining of both wild-type and snf148 mutant germ cells, SETDB1 showed
64 diffuse cytoplasmic staining and punctate nuclear staining (Fig. 2.22 A,B). Second,
Western blot analysis of ovarian extracts showed that snf148 tumors and wild-type ovaries have a similar level of SETDB1 protein (Fig. 2.22 C). Our finding that
H3K9me3 staining is disrupted, even though SETDB1 protein accumulation appears normal in snf148 mutant ovaries, leads us to conclude that SXL and
SETDB1 collaboratively promote H3K9me3-mediated silencing (Fig. 2.23).
148 Figure 2.21 H3K9me3 distribution is affected in snf mutant ovaries. 148 Confocal images of (A) wild-type (WT) and (B) snf germaria stained for H3K9me3 (green, white in inset). Germ cells were identified by a-VASA staining (magenta). Scale bar, 25 mm. Insets show a higher magnification of a single germ cell, outlined by a dashed line.
65
148 Figure 2.22 SETDB1 protein expression is not altered in snf mutants. 148 Confocal images of a germaria from (A) wild-type and (B) snf females carrying a copy an endogenously HA-tagged allele of setdb1 stained for HA (green, white in inset). Germ cells were identified by a-VASA staining (magenta). Scale bar, 12.5 mm. Insets show a higher magnification of a single germ cell, outlined by a dashed 148 line. (C) Western blot of ovarian lysates from wild-type and snf females carrying a copy an endogenously HA-tagged allele of setdb1 probed with an antibody against the HA tag. a-tubulin the loading control. (D) Quantitation comparing wild-type (WT) and mutant levels of HA normalized to a-tubulin. Error bars indicate standard deviation (s.d). The significance of the differences between wild-type and mutant was determined by calculating p-values by Student’s T-test from 3 replicates.
66
Figure 2.23 Diagram of the genetic pathway controlling H3K9me3 accumulation in female germ cells. In WT, SXL collaborates with SETDB1 to 148 regulate H3K9me3 accumulation. snf , by virtue of the fact that it interferes with Sxl splicing, leads to germ cells without SXL protein. This in turn leads to a defect in H3K9me3 accumulation without interfering with SETDB1 protein accumulation.
To directly test whether Sxl plays a role in controlling H3K9me3 deposition, we profiled the distribution of H3K9me3 by ChIP-seq in snf148 mutant ovaries and compared it to the distribution in wild-type ovaries. By limiting the differential peak analysis to within 1 kb of euchromatic genes, we identified 1,039 enrichment peaks in wild-type that were significantly altered in snf148 mutant ovaries, 91% of which show the expected decrease in H3K9me3 enrichment (Fig. 2.24). When we compared the changes in snf148 and setdb1 GLKD mutants, we found a close correlation (R2=0.6; Fig. 2.25). The strong overlap between the regions displaying decreased H3K9me3 enrichment in snf148 and setdb1 GLKD mutant ovaries suggests that SXL and SETDB1 influence H3K9me3 accumulation on the same set of genes (Fig. 2.25), including phf7 (Fig. 2.9) and CG34434 (Fig. 2.12). Based
67 on these studies we conclude that SXL functions with SETDB1 in the assembly of
H3K9me3 silencing islands in germ cells.
148 Figure 2.24 H3K9me3 is reduced at a set of genes in snf mutants. Differential analysis of paired H3K9me3 ChIP-seq data sets identifies peak 148 changes in snf mutant ovaries. Scatter plot of significantly altered 148 (FDR<0.05) H3K9me3 peaks in snf ovaries relative to wild-type (WT) ovaries.
The x-axis is the log2 H3K9me3 enrichment of wild type peaks subtracted from 148 the log2 H3K9me3 enrichment of snf peaks. Negative values indicate a decreased H3K9me3 peak in mutant ovaries.
68 148 Figure 2.25 setdb1 GLKD and snf influence H3K9me3 accumulation on a similar set of genes, including. Plot comparing the significantly altered 148 H3K9me3 peaks observed in snf ovaries to setdb1 GLKD ovaries. Genes that are not normally expressed in ovaries but are ectopically expressed in setdb1 GLKD are labeled in red or in blue. Blue indicates genes which are normally expressed in testis.
2.5 Discussion
This study reveals a previously undescribed role for H3K9me3 chromatin, operationally defined as facultative heterochromatin, in securing female identity by silencing lineage-inappropriate transcription. We show that H3K9me3 pathway members, the H3K9 methyltransferase SETDB1, its binding partner WDE, and the
H3K9 binding protein HP1a, are required for silencing testis gene transcription in female germ cells. Our studies further suggest a mechanism in which SETDB1, in conjunction with the female fate determinant SXL, controls transcription through deposition of highly localized H3K9me3 islands on a select subset of these genes.
The male germ cell sexual identity gene phf7 is one of the key downstream
SETDB1 target genes. H3K9me3 deposition on the region surrounding the testis- specific TSS guaranties that no PHF7 protein is produced in female germ cells. In this model, failure to establish silencing leads to ectopic PHF7 protein expression, which in turn drives aberrant expression of testis genes and a tumor phenotype
(Fig. 2.26).
Prior studies have established a role for SETDB1 in germline Piwi- interacting small RNA (piRNA) biogenesis and TE silencing [Rangan, 2011;
Sienski, 2015; Yu, 2015]. However, piRNAs are unlikely to contribute to sexual identity maintenance as mutations that specifically interfere with piRNA production,
69 such as rhino, do not cause defects in germ cell differentiation [Volpe, 2001;
Klattenhoff, 2009; Zhang, 2014; Mohn, 2014] or lead to global changes in gene expression. These findings, together with our observation that rhino does not control sex-specific phf7 transcription, suggests that the means by which SETDB1 methylates chromatin at testis genes is likely to be mechanistically different from what has been described for piRNA-guided H3K9me3 deposition on TEs.
Figure 2.26 Schematic summary of discrete facultative heterochromatin island assembly at phf7. In female germ cells, SETDB1, together with SXL, directs assembly of a highly localized H3K9me3 domain around the testis- specific TSS. In germ cells lacking SETDB1 or SXL protein, the dissolution of the H3K9me3 domain correlates with ectopic testis-specific phf7-RC transcription and PHF7 protein expression. Ectopic PHF7 protein activity leads to activation of downstream testis genes and a tumor phenotype.
We find that H3K9me3 accumulation at many of these genes, including phf7, is dependent on the presence of SXL protein. Thus, our studies suggest that
SXL is required for female-specific SETDB1 function. SXL encodes an RNA binding protein known to regulate its target genes at the posttranscriptional levels
70 [Reviewed by Salz, 2010]. SXL control may therefore be indirect. However, studies in mammalian cells suggest that proteins with RNA binding motifs are important for H3K9me3 repression [Becker,2017; Thompson, 2015], raising the tantalizing possibility that SXL might play a more direct role in governing testis gene silencing.
Further studies will be necessary to clarify how the sex determination pathway feeds into the heterochromatin pathway.
phf7 stands out among the cohort of genes regulated by facultative heterochromatin because of its pivotal role in controlling germ cell sexual identity
[Shapiro-Kulnane, 2015; Yang, 2012]. Because ectopic protein expression is sufficient to disrupt female fate, tight control of phf7 expression is essential. phf7 regulation is complex, employing a mechanism that includes alternative promoter usage and TSS selection. We report here that SETDB1/H3K9me3 plays a critical role in controlling phf7 transcription. In female germ cells, H3K9me3 accumulation is restricted to the region surrounding the silent testis-specific transcription start site. Dissolution of the H3K9me3 marks via loss of SXL or SETDB1 protein is correlated with transcription from the upstream testis-specific site and ectopic protein expression, demonstrating the functional importance of this histone modification. Together, these studies suggest that maintaining the testis phf7 promoter region in an inaccessible state is integral to securing female germ cell fate.
Although the loss of H3K9me3 pathway members in female germ cells leads to the ectopic, lineage-inappropriate transcription of hundreds of genes, our integrative analysis identified only 21 SETDB1/H3K9me3 regulated genes. Given
71 that one of these genes is phf7 and that ectopic PHF7 is sufficient to destabilize female fate [Shapiro-Kulane, 2015; Yang, 2012], it is likely that inappropriate activation of a substantial number of testis genes is a direct consequence of ectopic PHF7 protein expression. How PHF7 is able to promote testis gene transcription is not yet clear. PHF7 is a PHD-finger protein that preferentially binds to H3K4me2 [Yang, 2012], a mark associated with poised, but inactive genes and linked to epigenetic memory [Pekowska, 2010; Zhang, 2012; Pinskaya, 2009].
Thus, one simple model is that ectopic PHF7 binds to H3K4me2 marked testis genes to tag them for transcriptional activation.
It will be interesting to explore whether any of the other 20
SETDB1/H3K9me3 regulated genes also have reprogramming activity. In fact, ectopic fate-changing activity has already been described for the homeobox transcription factor Lim1 in the eye-antenna imaginal disc [Roignant, 2010].
However, whether Lim1 has a similar function in germ cells is not yet known.
Intriguingly, protein prediction programs identify three of the uncharacterized testis-specific genes as E3 ligases ([http:www.gene2f.org]) [Hu, 2017]. SkpE is a member of the SKP1 gene family, which are components of the Skp1-Cullin-F-box type ubiquitin ligase. CG12477 is a RING finger domain protein, most of which are believed to have ubiquitin E3 ligases activity. CG42299 is closely related to the human small ubiquitin-like modifier (SUMO) E3 ligase NSMCE2. Given studies that have linked E3 ligases to the regulation of chromatin remodeling [Dubiel, 2018;
Wotton, 2017], it is tempting to speculate that ectopic expression of one or more of these E3 ligases will be sufficient to alter cell fate. Future studies focused on
72 this diverse group of SETDB1/H3K9me3 regulated genes and their role in reprogramming may reveal the multiple layers of regulation required to secure cell fate.
The SETDB1-mediated mechanism for maintaining sexual identity we have uncovered may not be restricted to germ cells. Recent studies have established that the preservation of sexual identity is essential in the adult somatic gut and gonadal cells for tissue homeostasis [Hudry, 2016; Regan, 2016; Ma, 2016; Grmai,
2018]. Furthermore, the finding that loss of HP1a in adult neurons leads to masculinization of the neural circuitry and male specific behaviors [Ito, 2012] suggests a connection between female identity maintenance and H3K9me3 chromatin. Thus, we speculate that SETDB1 is more broadly involved in maintaining female identity.
Our studies highlight an emerging role for H3K9me3 chromatin in cell fate maintenance [Reviewed by Becker, 2016]. In the fission yeast S. pombe, discrete facultative heterochromatin islands assemble at meiotic genes that are maintained in a silent state during vegetative growth [Zofall, 2012; Nakayama, 2001]. Although less well understood, examples in mammalian cells indicate a role for SETDB1 in lineage-specific gene silencing [Jiang, 2017; Du, 2018; Koide, 2016; Schultz, 2002;
Tan, 2012]. Thus, silencing by SETDB1 controlled H3K9 methylation may be a widespread strategy for securing cell fate. Interestingly, H3K9me3 chromatin impedes the reprogramming of somatic cells into pluripotent stem cells (iPSCs).
Conversion efficiency is improved by depletion of SETDB1 [Soufi, 2012; Sridharan,
2013; Chen, 2013]. If erasure of H3K9me3 via depletion of SETDB1 alters the
73 sexually dimorphic gene expression profile in reprogrammed cells, as it does in
Drosophila germ cells, the resulting gene expression differences may cause stem cell dysfunction, limiting their therapeutic utility.
74 CHAPTER 3
FAILURE TO REPRESS PHF7 REPROGRAMS SEXUAL FATE BY
ACTIVATING H3K9ME3 SILENCED TESTIS GENES IN DROSOPHILA
OVARIES
Anne E. Smolko and Helen K. Salz
This work is currently being prepared for publication.
Author Contribution Statement:
A.S and H.S conceived and designed the entire project; A.S performed experiments and analyzed all RNAseq and ChIPseq datasets
75 3.1 Abstract
In Drosophila, preservation of female germ cell identity is critical for the completion of oogenesis. Female germ cells utilize H3K9me3 facultative heterochromatin to repress a subset of spermatogenesis genes through adulthood. The female sex-determining gene, Sxl, facilitates deposition of
H3K9me3 and prevent the expression of testis-enriched genes. This includes phf7, a regulator of male germ cell identity. Loss of Sxl in germ cells results in decreased
H3K9me3 at the testis-specific region of the phf7 locus. This results in male isoform transcription and protein expression. Here, we report that failure to actively repress PHF7 production in female germ cells is detrimental to the preservation of female identity. Forcing phf7 expression results in tumor formation and a derepression of testis-specific genes. We determine that phf7 autoregulates itself and activates a genetic feedback loop by reprogramming the chromatin landscape at a small set of spermatogenesis genes. This occurs through dissolution of
H3K9me3 facultative heterochromatin. This work characterizes PHF7 as a pioneer factor in female germ cells that can reprogram cell fate through chromatin remodeling.
76 3.2 Introduction
Drosophila female fertility is dependent on the ability of germ cells to maintain their sex-identity during oogenesis. Female germ cells preserve their sex- identity through deposition of H3K9me3 at a select set of testis-enriched genes
[Smolko, 2018]. This is facilitated Sex-lethal (Sxl), the female sex-determining gene. Germline specific loss of Sxl results in tumors with decreased H3K9me3 and ectopic expression of a testis gene program. Transcriptional repression of a subset of these directly require deposition of discrete H3K9me3 “islands” in wild type ovaries.
A key gene that must be repressed in female germ cells is PHD Finger
Protein 7 (phf7), a regulator of male germ cell identity [Yang, 2012; Shapiro-
Kulnane, 2015; Smolko, 2018]. In wild type animals, PHF7 protein expression is restricted to undifferentiated male germ cells [Yang, 2012; Yang, 2017]. phf7 mRNA is transcribed in both testis and ovaries with two sex-specific isoforms
[Shapiro-Kulnane, 2015]. In testis, transcription occurs from an upstream TSS, generating a longer 5’ UTR (phf7-RC). In ovaries, phf7-RC expression is repressed by an H3K9me3 island over the testis-specific region, where SXL is required for deposition [Smolko, 2018]. This produces a shorter 5’ UTR generated from a downstream TSS (phf7-RA). Only expression of phf7-RC results in detectable protein. Repression of phf7-RC by H3K9me3 is key to female germ cell differentiation as its expression contributes to tumor formation [Shapiro-Kulnane;
Smolko, 2018]. However, the sufficiency of PHF7 to alter gene expression to a male-like fate and the mechanism by which is may do so is not known.
77 In males, PHF7 contributes to fecundity and germ cell maintenance [Yang,
2012]. It is capable of binding histones, specifically H3K4me2, and has been postulated to regulate male germ cell gene expression [Yang, 2012; Yang, 2017].
Furthermore, female germ cells residing in a masculinized somatic environment that ectopically express phf7 are able to produce sperm [Yang, 2012]. Therefore, we hypothesized that failure to prevent PHF7 expression in females is sufficient to alter transcriptional regulation and compromise female germ cell fate.
Here, we find that forced germline expression of phf7 in females is sufficient to form tumors and upregulate testis gene transcription. Many of these genes are also ectopically expressed in ovaries lacking germline Sxl, consistent with it’s role in repressing phf7 in female germ cells. We make the surprising observation that a set of these genes, including phf7 itself, are normally repressed by H3K9me3.
As phf7 functions downstream of Sxl and H3K9me3 deposition, expression of these genes places phf7 in a genetic feedback loop. Analysis of H3K9me3 enrichment at these loci reveals that ovaries ectopically expressing phf7 have decreased enrichment.
Overall, we demonstrate that failure to prevent PHF7 protein expression in female germ cells compromises sex-identity by activating a genetic feedback mechanism. Here, PHF7 alters H3K9me3 at a subset of testis-enriched genes to activate their expression. This work implicates PHF7 as a possible pioneer factor that can misread the genome, alter chromatin architecture, and misroute female germ cells to a male-like fate
78 3.3 Methods
Fly Stocks and Husbandry
Fly strains were kept on standard medium at 25oC unless otherwise noted.
Transgenic, UAS, and Gal4 drivers used include phf7ey03023 (Bloomington
Drosophila Stock Center/BDSC #15894), HA-phf7 [Yang, 2012], and nanos-gal4
(VP16) (BDSC #4937). The HA-phf7 transgene was recombined to segregate with nanos-gal4 (HA-phf7,nos-gal4/Tm3-Sb). To drive phf7ey03023 expression, HA- phf7,nos-gal4/Tm3-Sb females were crossed to phf7ey03023 males at 18oC. The phf7+/phf7ey;HA-phf7,nos-gal4/+ daughters were selected and moved to 29oC for
10 days prior to analysis.
Immunofluorescence and Microscopy
Drosophila gonads were fixed and stained according to standard procedures with the following primary antibodies: mouse a-Spectrin (1:100,
Developmental Studies Hybridoma Bank [DSHB] 3A9 RRID: AB_528473], rabbit a-H3K9me3 (1:1,000, Active Motif Cat# 39162, RRID: AB_2532132), rat a-HA
(1:500, Roche Cat# 11867423001, RRID: AB_390919), mouse a-Sxl (1:100,
DSHB M18 RRID: AB_528464), and rat a-Vasa (1:100, DSHB RRID: AB_760351).
Staining was detected by FITC (1:200, Jackson ImmnoResearch Labs) or Alexa
Fluor 555 (1:200, Life Technologies) conjugated species appropriate secondary antibodies. TO-PRO-3 Iodide (Fisher, Cat# T3605) was used to stain DNA. Images were taken on a Leica SP8 confocal with 1024x1024 pixel dimensions, a scan speed of 600 Hz, and a frame average of 3. Sequential scanning was done for
79 each channel and three Z-stacks were combined for each image. Processed images were compiled with Gnu Image Manipulation Program (GIMP) and
Microsoft PowerPoint. All staining experiments were replicated at least two times.
The “n” in the figure legends represents the number of germaria scored from a single staining experiment.
Fluorescence intensity quantification of H3K9me3 was measured in 5 wild- type and mutant germaria using GIMP and normalized to fluorescence intensity of
DNA to control for cell number and DNA visibility. Each image was obtained under identical conditions and consisted of three Z-stacks, with the germline stem cells in the plane of focus.
qRT-PCR and data analysis
RNA was extracted from dissected ovaries using TRIzol (Invitrogen, Cat#
15596026) and DNase (Promega, Cat# M6101). Quantity and quality were measured using a NanoDrop spectrophotometer. cDNA was generated by reverse transcription using the SuperScript First-Strand Synthesis System Kit
(Invitrogen, Cat# 11904018) using random hexamers. Quantatative real-time
PCR was performed using Power SYBR Green PCR Master Mix (ThermoFisher,
Cat# 4367659) with the Applied Biosystems 7300 Real Time PCR system. PCR steps were as follows: 95oC for 10 minutes followed by 40 cycles of 95oC for 15 seconds and 60oC for 1 minute. Melt curves were generated with the following parameters: 95oC for 15 seconds, 60oC for 1 minutes, 95oC for 15 seconds, and
60oC for 15 seconds. Measurements were taken in biological triplicate with two
80 technical replicates. Transcript levels were normalized to rp49. To measure
transcription from the phf7-RC TSS in phf7 ΔCDS-GFP flies, transcript levels were normalized to gfp. Relative transcript amount were calculated using the 2-DDCt method [Livak, 2001]. Primer sequences for measuring the total phf7 and gfp were: for total phf7, forward GAGCTGATCTTCGGCACTGT and reverse
GCTTCGATGTCCTCCTTGAG; for phf7-RC forward
AGTTCGGGAATTCAACGCTT and reverse GAGATAGCCCTGCAGCCA; for gfp forward ACGTAAACGGCCACAAGTTC and reverse
AAGTCGTGCTGCTTCATGTG
RNAseq and data analysis
For wild type (yw/yw) and mutant (HA-phf7,nos>phf7ey): Total RNA was extracted from dissected ovaries using standard TRIzol (Invitrogen, Cat# 15596026) methods. RNA quality was assessed with Qubit and Agilent Bioanalyzer. Libraries were generated using the Illumina TruSeq Stranded Total RNA kit (Cat#
20020599). Sequencing was completed on 2 biological replicates of each genotype with the Illumina HiSeq 2500 v2 with 100bp paired end reads.
Sequencing reads were aligned to the Drosophila genome (UCSC dm6) using
TopHat (2.1.0) [Trapnell, 2009]. Differential analysis was completed using CuffDiff
(2.2.1) [Trapnell, 2012]. Genes were considering differentially expressed if they exhibited a two-fold or higher change relative to wild-type with a False Discovery
Rate (FDR) <0.05. Scatter plots were generated using ggplot function in R. Genes that were expressed in mutant (FPKM ≥ 1) but not expressed in wild type ovaries
81 (FPKM<1) were termed as ectopic. Genes not expressed in mutant (FPKM < 1) but expressed in wild type ovaries (FPKM ≥ 1) were termed as turned off.
Tissue expression clustering of the ectopically expressed and turned off genes were performed to identify tissue-specific signatures. Expression values normalized to the whole fly were extracted from FlyAtlas (versions 1 and 2)
[Chintapalli, 2007; Leader, 2018]. Heatmaps to compare the tissue expression profile of these genes per tissue were generated in R with heatmap.2 (gplots).
Genes were clustered and normalized per row.
Screen shots are from Integrated Genome Viewer (IGV). To account for the differences in sequencing depth when creating IGV screenshots, the processed
RNAseq alignment files were scaled to the number of reads in the wild type file.
This was done with Deeptools bigwigCompare using the scale Factors parameter with a bin size of 5.
CRISPR
CRISPR/Cas9 was utilized to replace the coding region of phf7 with the sequence encoding for gfp. Two gRNAs targeting the phf7 locus on either side of the coding region were synthesized and individually cloned into the pU6-BbsI- chiRNA vector (addgene #45946). For gRNA’s:
5’ sense CTTCGGTCACCGGAAACGCATCCA and antisense AAACTGGATGCGTTTCCGGTGACC;
3’ sense CTTCGAATCCTTGCGGCTGGCCATG and antisense AAACCATGGCCAGCCGCAAGGATTC.
82 Flanking phf7 homology arms were generated through PCR of genomic DNA was cloned into the dsRed vector, containing an attP site [addgene #51019]. Vectors were injected into transgenic flies expressing Cas9 through the vasa promoter to generate phf7 ΔCDS-dsRed animals (vas-Cas9; BDSC #51324). To insert the gfp sequence, gfp was cloned into the RIV-white vector containing an attB site [Baena-
Lopez, 2013]. The RIV-white-gfp vector was injected into phf7 ΔCDS-dsRed with phiC enzyme to generate phf7 ΔCDS-GFP. The dsRed and white+ markers remained intact.
All injections were completed by Rainbow Transgenic Flies Inc.
ChIPseq and data analysis
For each chromatin immunoprecipitation, 150 pairs of Drosophila ovaries were dissected in PBS plus protease inhibitors. Tissues were fixed in 1.8% formaldehyde for 10 minutes at room temperature and quenched with 225 mM glycine for 5 minutes. Tissues were washed twice and stored at -80oC for downstream applications. Samples were lysed using the protocol from [Lee et al,
2006]. Tissue was placed in lysis buffer 1 (140 mM HEPES pH 7.5, 200 mM NaCl,
1 mM EDTA, 10% glycerol, 0.5% NP-40, 0.25% Triton X-100), homogenized using sterile beads, and rocked at 4oC for 10 minutes. Tissue was then washed 10 minutes at 4oC in lysis buffer 2 (10 mM Tris pH 8, 200 mM NaCl, 1 mM EDTA, 0.5 mM EGTA). Tissues were then placed in 1.5 mL lysis buffer 3 (10 mM Tris pH 8,
100 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 0.1% Na-deoxycholate, 0.5% N- lauroylsarcosine). All buffers were supplemented with protease inhibitors.
Chromatin was sheared to 200-700 base pairs using the Diagenode Bioruptor Pico
83 (Cat #B010600100). The chromatin lysate was incubated overnight at 4oC with
H3K9me3 antibodies pre-bound to magnetic beads. The beads were prepared as follows: 25 ul Protein A and 25 ul Protein G Dynabeads (Invitrogen, Cat# 10002D and 10004D) per sample were washed twice with ChIP blocking buffer (0.5%
Tween 20, 5 mg/mL BSA), then blocked by rocking at 4oC for 1 hour in ChIP blocking buffer, and then conjugated to 5 ug H3K9me3 antibody (Abcam Cat#
8898 RRID: AB_306848) by rocking in new ChIP blocking buffer at 4oC for 1 hour.
Following immunoprecipitation, the samples were washed 6 times with ChIP-RIPA buffer (10 mM Tris-HCl pH 8, 1 mM EDTA, 140 mM NaCl, 1% Triton X-100, 0.1%
SDS, 0.1% Na-Deoxycholate), 2 times with ChIP-RIPA/500 buffer (ChIP-RIPA +
500 mM NaCl), 2 times with ChIP-LiCl buffer (10 mM Tris-HCl pH 8, 1 mM EDTA,
250 mM LiCl, 0.5% NP-40, 0.5% Na-deoxycholate), and twice with TE buffer. DNA was eluted from beads with 50 ul elution buffer (10 mM Tris-HCl pH 8, 5 mM EDTA,
300 mM NaCl, 0.1% SDS) and reverse crosslinked at 65oC for 6 hours. Beads were spun down and eluted DNA was transferred to a new tube and extracted using phenol-chloroform extraction. All buffers were supplemented with protease inhibitors.
ChIP sequencing libraries were prepared using the Rubicon Genomics
Library Prep Kit (Cat# R440406) with 15 amplification cycles. DNA was cleaned and assessed for quality with Qubit and Agilent Bioanalyzer. Sequencing was completed on 2 biological replicates of each genotype with the Illumina NextSeq
550 High Output Flowcell with 75 bp single end reads.
Sequencing reads were filtered for quality. Those having a quality score
84 greater than 27 in at least 80% of bases were kept and adapter sequences were removed. H3K9me3 reads were aligned to the Drosophila genome (UCSC dm6 using bowtie2 (2.2.6) [Langmead, 2012], and duplicate reads were removed with samtools (1.3) [Li, 2009].
HA-phf,nos>phf7ey H3K9me3 enrichment over gene bodies was compared to wildtype/yw/yw (n=2) and HA-phf7,nos-gal4 (n=1). Correlation for control replicates was calculated with deeptools multiBamSummary. Differential enrichment analysis completed with the DIFFBIND program (2.4.8). Enrichment was measured using genomic coordinates for annotated gene bodies, using summits=500 and the DESeq2 package
[http://bioconductor.org/packages/release/bioc/vignettes/DiffBind/inst/doc/DiffBin d.pdf]. The mean peak concentration was calculated by normalizing reads to the total library size and subtracting the corresponding input reads. Differential peak fold changes were calculated by subtracting log2 wild-type mean concentrations from mutant mean concentrations. Enrichment was considered significantly altered relative to controls if they had a False Discovery Rate (FDR) < 5%. For
ChIPseq IGV tracks, HA-phf7,nos>phf7ey ovaries are compared to HA-phf7,nos- gal4 ovaries to control for phf7 copy number. Genome browser tracks are normalized to their input and scaled to the sequencing depth of control ovaries
(HA-phf7,nos-gal4). Boxplots and scatterplots were generated using ggplot function in R.
85 3.4 Results
3.4.1 phf7 expression in female germ cells is sufficient for tumorigenesis
We had previously reported that phf7 expression in female germ cells is sufficient for tumorigenesis by driving expression of phf7EY03023, and EP insertion line with nanos-Gal4 (nos-Gal4>phf7ey) [Bellen, 2004; Yang, 2012; Shapiro-
Kulnane, 2015] (Fig. 3.1). However, tumors were only present in 18% of germaria
(Fig. 3.2A) [Shapiro-Kulnane, 2015]. We noted that providing an additional copy of phf7 in females via the HA-phf7 transgene [Yang, 2012] (HA-phf7, nos>phf7EY03023; referred to as phf7ey henceforth) to monitor protein expression dramatically increased tumor frequency. The HA-phf7 transgene is located within a 20 kb BAC rescue construct appropriately regulated, with protein expression only detectable in testis [Yang, 2012; Shapiro-Kulnane, 2015; Smolko, 2018].
Here, as in earlier studies [Shapiro-Kulanane, 2015], we reared phf7ey flies at a restrictive temperature (18oC) to avoid germ cell lethality. Adults were then shifted to the permissive temperature (29oC) for 10 days to allow for sufficient phf7 expression. Tumor frequency was quantitated by assessing spectrosome content.
Spectrosomes are spherical, alpha-spectrin containing structures that associate with germline stem cells. As differentiation proceeds, spectrosomes branch out to form fusomes (Fig. 3.2B). In phf7ey ovaries containing a wild type, a transgenic, and the phf7ey allele, 70% (n=88/126) of germaria contained 6 or more spectrosomes, indicative of tumor formation (Fig. 3.2A,C). These experiments support previous data showing that phf7 expression in the female germline is sufficient to prevent oogenesis [Shapiro-Kulnane et al, 2015]. Here, providing
86 additional copies of phf7 exacerbates tumor formation and provides a more homogenous phenotype to complete studies at the molecular level.
Figure 3.1: phf7 expression is forced through the insertion of an Upstream Activating Sequence (UAS). Schematic of the phf7 coding region with the EY03023 phf7 insertion. The 5’ and 3’ untranslated regions are denoted in gray with the coding region in black. The blue triangle indicates the location of the UAS sequence insertion. Gal4 is tissue-specifically expressed in germ cells through the nanos (nos) promoter. Gal4 protein (orange ovals) binds the UAS sequence, driving gene expression.
87 Figure 3.2: Forced expression of phf7 in female germ cells results in tumor formation. (A) Quantification of germaria with 0, 1-5, and >5 round spectrosome- containing germ cells. The number of germaria scored (n) and genotypes are ey indicated (B) Wild type (yw/yw) and (C) HA-phf7,nos>phf7 representative germaria immunostained for alpha-spectrin (magenta) and DNA (blue). Yellow arrows indicate spectrosomes (sp) that associate with germline stem cells and fusomes (fu) that connect the differentiating cystoblast cells. Scale bar indicates 25 um
3.4.2 phf7 is sufficient to activate testis gene expression
Female germ cells that have lost their sex-identity mis-express a large set of testis-enriched genes [Shapiro-Kulnane, 2015; Smolko, 2018].. Common to these is the inappropriate expression of phf7-RC, the testis-specific isoform. PHF7 has been proposed to be a chromatin reader with a role in transcriptional regulation in male germ cells [Yang, 2012; Yang, 2017]. Therefore, we hypothesized that
PHF7 might be sufficient to activate a testis gene expression program in female germ cells. Differential RNAseq analysis identified 1,634 significantly altered genes (Absolute Fold Change ≥ 2 and FDR < 0.05). Of these, 51% are downregulated (n= 835/1634) and 49% are upregulated (n=799/1634) (Ext. Tables
1 and 2, Fig. 3.3).
PHF7 protein expression is normally restricted to the undifferentiated germ cells of adult testis and is thought to regulate male germ cell fate (Yang, 2012;
Yang, 2017). This pointed to the possibility that its expression in females may allow for genes to be either “turned on/ectopically expressed” or “turned off” rather than simply being up- or down-regulated. Further interrogation of differentially expressed genes revealed a set of 286 ectopically expressed genes that normally
88 show no expression in wild type ovaries (log2 WT FPKM < 0 and log2 HA-
ey phf7,nos>phf7 FPKM > 0) with 340 downregulated genes being turned off (log2
ey HA-phf7,nos>phf7 FPKM < 0 and log2 WT FPKM < 0) (Fig. 3.3 red (ectopic) and blue (turned off) points). These data suggest that phf7 may be transcriptionally altering germ cell identity by turning off ovary-specific genes or turning on genes specific to another cell type.
ey Figure 3.3: A subset of phf7 differentially expressed genes are “ectopically expressed” or “turned off”. Scatter plot of differentially expressed genes ey comparing expression in wild (yw/yw) (x-axis) and HA-phf7,nos>phf7 ovaries (y- axis). Values plotted indicate the log (FPKM) in each sample. Dashed lines 2 separate upregulated and downregulated genes relative to wild type. Red points ey indicate genes ectopically expressed in phf7 ovaries(WT log FPKM < 0). Blue 2 ey ey points indicate genes turned off in phf7 ovaries (phf7 log FPKM < 0). 2
Our previous work identified phf7 as being repressed by SXL to maintain female germ cell identity [Shapiro-Kulnane, 2015]. As Sxl is genetically upstream of phf7, we were interested in how the transcriptional profiles of each mutant compared to each other. We utilized differential analysis from our previously
89 published RNA-seq data [Shapiro-Kulnane, 2015], where the sans-fille148 (snf148) allele was utilized to eliminate SXL in germ cells while retaining its expression in somatic cells [Nagengast, 2003]. Using the expression values reported in this study [Shapiro-Kulnane, 2015], we find that 207 genes are ectopically expressed
(Fig. 3.4 orange points) and 47 are turned off (Fig. 3.4 purple points) in snf148 ovaries.
148 Figure 3.4: A subset of snf differentially expressed genes are “ectopically expressed” or “turned off”. Scatter plot of differentially 148 expressed genes comparing expression in wild (yw/yw) (x-axis) and snf ovaries (y-axis). Values plotted indicate the log (FPKM) in each sample. 2 Dashed lines separate upregulated and downregulated genes relative to wild 148 type. Orange points indicate genes ectopically expressed in snf ovaries (WT 148 148 log FPKM < 0). Purple points indicate genes turned off in snf ovaries (snf 2 log FPKM < 0). 2
Comparison of differentially expressed genes in phf7ey and snf148 mutants reveals that only 10 genes are turned off in both (Fig. 3.5A). Furthermore, analysis of adult tissue expression patterns as reported by FlyAtlas [Chintapalli, 2007;
90 Leader, 2018]. showed that turned off genes in either mutant do not show any clear tissue enrichment (Fig. 3.6A,B). We note that there is a particular lack of genes having ovary-specific enrichment. This disagrees with previous work suggesting that PHF7 functions in male germ cells to repress female-specific gene expression
[Yang, 2017]. This may also indicate that the function of PHF7 is not the same in a male versus female environment.
Figure 3.5: Overlap of genes that are turned off and ectopically expressed in ey 148 phf7 and snf ovaries. Venn diagrams of genes that are either (A) turned off (log FPKM < 1 in mutant) or (B) ectopically expressed (log FPKM < 1 in wild 2 2 ey 148 type) in both phf7 and snf ovaries.
ey 148 Figure 3.6 Ovary-specific genes are not turned off in phf7 and snf ovaries. Heatmaps of the relative enrichment scores (FlyAtlas) of turned off genes in (A) ey 148 phf7 and (B) snf ovaries. Enrichment was calculated based on gene expression relative to the whole fly. Each column is an adult tissue. Rows are individual genes normalized based on their Z-score. Positive values indicate a higher relative enrichment score.
91 For the ectopically expressed genes, relative tissue enrichment in each mutant was much more striking. Here, prominent testis-specific expression is observed in both mutants. Furthermore, genes that are not enriched in testis do not show obvious enrichment in any other individual tissue (Fig. 3.7A,B). This pattern was even more apparent when we focused on the tissue-enrichment patterns of the 59 shared ectopically expressed genes, where 52% (n=31/59) are highly testis-enriched (Fig. 3.5, 3.8). Overall, this analysis shows that snf148 and phf7ey defects are similar, as one would expect as Sxl acts upstream of phf7
[Shapiro-Kulnane, 2015]. They also indicate the phf7 is sufficient to activate a testis gene program in female germ cells.
ey 148 Figure 3.7 Genes ectopically expressed in phf7 and snf ovaries are testis- enriched. Heatmaps of the relative enrichment scores (FlyAtlas) of ectopically ey 148 expressed genes in (A) phf7 and (B) snf ovaries. Enrichment was calculated based on gene expression relative to the whole fly. Each column is an adult tissue. Rows are individual genes normalized based on their Z-score. Positive values indicate a higher relative enrichment score.
92
Figure 3.8 Over half of shared phf7ey and snf148 ectopically expressed genes are testis enriched. Heatmap of the relative enrichment scores (FlyAtlas) ey 148 for ectopically expressed genes shared between phf7 and snf ovaries. Enrichment was calculated based on gene expression relative to the whole fly. Each column is an adult tissue. Rows are individual genes normalized based on their Z-score. Positive values indicate a higher relative enrichment score. Testis- enriched genes are listed on the right with those regulated by SETDB1-mediated H3K9me3 highlighted in red text
3.4.3 PHF7 autoregulates its expression by decreasing H3K9me3 enrichment at its testis-specific region
Prior observations showed that SXL facilitates H3K9me3 deposition at a small set of spermatogenesis genes that are directly repressed by H3K9me3 enrichment [Smolko, 2018]. We make the surprising observation that 5 of these genes are ectopically expressed in both phf7ey and snf148 ovaries (Fig. 3.8, red gene names). In a linear genetic pathway where SXL influences H3K9me3
93 deposition upstream of phf7, we would not expect their ectopic expression in phf7ey mutants. This indicates that PHF7 activates a genetic feedback mechanism in female germ cells (Fig. 3.9).
Figure 3.9: Schematic for a possible genetic feedback mechanism activated with ectopic phf7 in female germ cells.
The presence of genetic feedback and the increased tumor incidence brought on by an additional copy of phf7 (Fig. 3.2C) suggested that PHF7 may autoregulate its expression. If so, forced transcription of from the phf7ey allele should result in HA-PHF7 transgene protein production (Fig 3.10A).
Immunostaining for HA-PHF7 in phf7ey ovaries shows expression of the transgene with no expression in control ovaries (Fig. 3.10B,C). As PHF7 protein expression correlates with transcription from the testis-specific, phf7-RC, isoform, its expression in phf7ey ovaries should no longer be only occurring from phf7-RA.
To determine phf7 isoform expression, we visualized our aligned RNA-seq reads at the phf7 locus. Consistent with the longer 5’ UTR correlating with translation, we found that forced phf7 expression switched TSS usage to that of phf7-RC
94 (Fig. 3.11, green RNAseq track). These results indicate that phf7 autoregulates its expression in female germ cells by regulating its TSS choice.
Figure 3.10: Ectopic phf7 results in autoregulation (A) Schematic for genetic ey scheme to drive phf7 from its endogenous locus (nos>phf7 ) to determine if it can turn on expression from the HA transgene. (B,C) Representative immunostained images from control germaria with no gal4-induced phf7 expression (B) and that where phf7 is overexpressed (C). Tissue are stained for DNA (magenta) and HA (green). Scale bar represents 25 um
95 Figure 3.11 RNAseq and H3K9me3 ChIPseq reads at the phf7 locus. Genome browser view of the phf7 locus. mRNA expression in Wild type ey 148 (pink), phf7 (green), and snf (orange) ovaries and H3K9me3 in Control ey 148 (HA-phf7,nos; in gray), phf7 (blue), and snf (red) ovaries. The testis- specific region is outlined. Tracks are viewed on the same scale with the 5’ end of the gene on the left.
In wild type females, phf7-RC is repressed by H3K9me3 deposition surrounding the testis TSS (Smolko, 2018) (Fig. 3.11, gray ChIPseq track).
Therefore, we hypothesized that PHF7 autoregulates in females by impacting heterochromatin at this region. To test this, we first wanted to confirm that phf7 regulation occurs in its non-coding regions. We generated a null reporter allele by replacing the phf7 coding region with gfp (phf7ΔCDS-GFP) through CRISPR/Cas9 editing (Fig. 3.12) [Baena-Lopez, 2013; Gratz, 2014]. In both testis and ovaries, gfp mRNA is transcribed from the phf7 locus at similar levels (Fig 3.13A). However,
GFP protein is only produced in testis, consistent with normal PHF7 expression
(Fig. 3.13B,C). This striking observation led us to ask if sex-specific TSS usage is preserved. Indeed, qrtPCR analysis on phf7ΔCDS-GFP testis and ovaries shows that phf7-RC is still only transcribed in testis (Fig 3.13D). These data show that the cis- regulatory regions that dictate sex-specific transcription lie within phf7’s noncoding regions.
96
Figure 3.12: CRISPR/Cas9 gene editing to replace the phf7 coding region with gfp. Schematic of CRISPR/Cas9 removal of of the phf7 coding region with gRNAs flanking the coding region. Homology directed repair was used to insert an attP site and dsRed marker. The RIV-white vector containing an attB site and GFP was then inserted through ⍦C31 integration. The mini-white and dsRed markers are flanked by loxP sites. Blue and green triangles indicate qPCR primers used to measure gfp and phf7-RC transcription. The TSS’s for testis phf7-RC and ovary phf7-RA isoforms are marked by arrows and indicate the direction of transcription.
97 Figure 3.13: phf7 sex-specific transcription is regulated by its non-coding ΔCDS-GFP regions. (A) RT-qPCR of GFP mRNA in phf7 testis and ovaries. Expression is normalized to rp49 and shown as the fold change relative to testis. Error bars indicate standard deviation (s.d) of three biological replicates. Student T- ΔCDS- test shows no significance (N.S) with p > 0.05 (B-C) Confocal images of phf7 GFP testis and ovaries stained for GFP (green, white in B’,C’) and DNA (magenta). ΔCDS-GFP Scale bar, 50 um. (D) RT-qPCR of of the testis phf7-RC transcript in phf7 testis and ovaries. Expression is normalized to gfp mRNA and shown as fold change relative to testis. Error bars indicate standard deviation (s.d) of three biological replicates with a p-value < 0.001.
We next wished to directly test if ectopic PHF7 alters the H3K9me3 landscape at its testis-specific region. We performed Chromatin
Immunoprecipitation followed by sequencing (ChIPseq) on HA-phf7,nos-gal4 and phf7ey ovaries. We find that aligned reads in control and our previously published
H3K9me3 ChIPseq on two wild type (yw/yw) samples are highly correlated (Fig.
3.14). Therefore, our analyses were completed with three “control” samples (two wild type and one HA-phf7,nos-gal4). At the phf7 locus, quantification of the average H3K9me3 reads showed a significant decrease in phf7ey ovaries relative to our control replicates, similar to what we observe in snf148 ovaries (Fig. 3.11,
3.15). These results indicate that PHF7 is capable of altering its own TSS usage by decreasing H3K9me3 enrichment at its locus, ultimately resulting in protein expression.
98
Figure 3.14 H3K9me3 enrichment in control (HA-phf7,nos-gal4) and wild type (yw/yw) ovaries is highly correlated. Scatter plots of aligned H3K9me3 ChIPseq reads in control (HA-phf7,nos-gal4) and previously published wild type (yw/yw) samples. Axis represent the number of DNA fragments aligned to each bin in the genome. Pearson correlation coefficients for each sample are indicated.
Figure 3.15 Dissolution of H3K9me3 occurs at genes activated in PHF7’s genetic feedback loop. Plot of the H3K9me3 enrichment control (gray) and HA- ey phf7,nos>phf7 (blue) ovaries. The y-axis represents the calculated log2 concentration of H3K9me3 (normalized to input) over each gene body, labeled on the x-axis. Each dot represents one replicate. Black points represent average intensities. Error bars indicate standard deviation. Significance was calculated by Student’s T-test. p < 0.05 (*) and p < 0.01 (**).
99 In female germ cells, SXL maintains female identity through H3K9me3 deposition over the phf7-RC region [ Smolko, 2018]. Therefore, we asked if ectopic phf7 decreases H3K9me3 deposition by affecting Sxl expression. In females, SXL protein is produced due to the exclusion of the male-specific third exon, which contains a premature transcription stop site [Reviewed by Salz,
2010]. The snf148 allele causes a Sxl splicing defect, resulting in inclusion of the third exon and preventing translation [Nagengast, 2003]. We find that ectopic expression of phf7 does not impact female-specific Sxl splicing (Fig. 3.16). In wild type germaria, SXL is expressed at high levels in the cytoplasm of GSCs and decreases as cells enter the differentiation pathway (3.17A). While SXL protein is still expressed in phf7ey mutants, its expression appears decreased and more uniform throughout the germarium region (3.17B). These data suggest that ectopic phf7 may impact production of SXL protein, possibly hindering its function.
Figure 3.16: Forced expression of phf7 does not alter Sxl splicing. Genome browser view of a portion of the Sxl locus with mRNA ey 148 expression in Wild type (yw/yw), phf7 and snf ovaries. The male specific exon 3 is highlighted by the gray box. Tracks are viewed on the same scale with the 5’ end of the gene on the left.
100
Figure 3.17 Forced phf7 expression alters SXL expression patterns. Representative confocal images of (A) control (HA-phf7,nos) and (B) HA- ey phf7,nos>phf7 SXL. Scale bar is 25 um.
3.4.4 PHF7 decreases H3K9me3 at additional SETDB1-regulated genes
Although not statistically significant, we observed a small decrease in overall H3K9me3 levels in phf7ey ovaries relative to control (HA-phf7,nos-gal4)
(Fig. 3.18). This and the decreased SXL expression we observed suggested that ectopic phf7 may affect H3K9me3 at additional loci. First, we focused on the genes we identified as part of our genetic feedback loop. We measured the average
H3K9me3 enrichment over each gene body in control (yw/yw and HA-phf7,nos) relative to phf7ey ovaries and found that CG34434, CG42299, and skpE also had significantly decreased enrichment (Fig. 3.15, 3.19-3.21). We find it striking that each of these genes shows tissue enrichment exclusively in testis, with no measurable expression in other adult tissues [FlyAtlas2, Leader, 2018]. While the function of CG34434 is unknown, both skpE and CG42299 are predicted to act as ubiquitin and SUMO E3 ligases respectively (FlyBase) (Table 3.1). These data place PHF7 in a genetic feedback mechanism that alters gene expression through modulation of H3K9me3 at a set of genes likely involved in spermatogenesis.
101
Figure 3.18 Forced expression of phf7 does not significantly alter H3K9me3 levels Representative confocal images of (A) HA-phf7,nos-gal4 ey (control) and (B) HA-phf7,nos>phf7 ovaries stained for the germ cell marker, vasa (magenta) and H3K9me3). Scale bar represents 25 um. Inset shows a higher magnification of a single germ cell, in which the nucleus is outlined by a dashed yellow box in A and B. (C) Quantification of H3K9me3 intensity in 5 ey control (pink) and phf7 (green) germaria. Signal was normalized to DNA to control for cell number and overall DNA visibility. Black points represent average intensities. Error bars indicate standard deviation. Significance was calculated by Student’s T-test. (N.S) p > 0.05
Figure 3.19 RNAseq and H3K9me3 ChIPseq reads at the skpE locus. Genome browser view of the skpE locus. mRNA expression in Wild type (yw/yw in pink), ey 148 phf7 (green), and snf (orange) ovaries and H3K9me3 in Contol (HA-phf7,nos in ey 148 gray), phf7 (blue), and snf (red) ovaries. The gene body of interest is outlined. Tracks are viewed on the same scale with the 5’ end of the gene on the left.
102
Figure 3.20 RNAseq and H3K9me3 ChIPseq reads at the CG42299 locus. Genome browser view of the CG42299 locus. mRNA expression in Wild type ey 148 (yw/yw in pink), phf7 (green), and snf (orange) ovaries and H3K9me3 in ey 148 Control (HA-phf7,nos in gray), phf7 (blue), and snf (red) ovaries. The gene body of interest is outlined. Tracks are viewed on the same scale with the 5’ end of the gene on the left.
Figure 3.21 RNAseq and H3K9me3 ChIPseq reads at the CG34434 locus. Genome browser view of the CG34434 locus. mRNA expression in Wild type ey 148 (yw/yw in pink), phf7 (green), and snf (orange) ovaries and H3K9me3 in ey 148 Control (HA-phf7,nos ingray), phf7 (blue), and snf (red) ovaries. The gene body of interest is outlined. Tracks are viewed on the same scale with the 5’ end of the gene on the left.
103 Table 3.1 Genes with decreased H3K9me3 and ectopic gene expression in phfey and snf148 ovaries. List of genes ectopically genes with decreased H3K9me3 and their relative tissue enrichment (FlyAtlas) and predicted functions (FlyBase). * indicates genes identified to be regulated by SETDB1.
We extended our analysis globally to identify any additional transcriptional changes correlating with phf7 overexpression and H3K9me3 alteration. Average
H3K9me3 enrichment over all annotated gene bodies in control and phf7ey replicates were measured. Doing so revealed 994 genes with significantly altered
H3K9me3 enrichment, with a decrease at 588 and an increase at 406 regions (Fig.
3.22). To identify possible biologically significant regions, we integrated our
ChIPseq and RNAseq data. As turned off genes did not yield any tissue-specific enrichment, we focused on genes that are ectopically expressed and have decreased H3K9me3 in phf7ey mutants. We further limited our analysis to ectopically expressed genes with an FPKM > 2.5 to isolate more highly expressed transcripts. This revealed an additional 7 genes with ectopic gene expression and decreased H3K9me3. Strikingly, 3 of these genes are also ectopically expressed in snf148 tumors and do not contain any annotated TE insertions. These include
CG6324, CR42767, and CG32706 (Fig. 3.23, Fig. 3.24-3.26). H3K9me3 enrichment in snf148 ovaries is also visibly decreased over these gene bodies (Fig.
104 3.24-3.26). The roles of these genes are currently uncategorized with CG32706 being predicted to function as an RNA-binding protein (Table 3.1). Overall, these data show that PHF7 activates genetic feedback to turn on a small number of testis-enriched genes by impacting H3K9me3. We note that all genes in this feedback pathway are located on the X chromosome. This significance of this and whether or not H3K9me3 deposition at any these loci are directly regulated by
PHF7 is unknown. It will also be important to determine what contributions these genes may have to tumor formation and global testis gene activation.
Figure 3.22 Forced phf7 expression alters H3K9me3 enrichment at a subset of gene bodies. Scatterplot of significantly altered H3K9me3 regions associated with an annotated gene body (UCSC dm6). The x-axis represents ey the log fold change of the average H3K9me3 read concentrations in phf7 2 ovaries relative to controls. The y-axis is the –log of the FDR. Regions were considered significantly altered if the FDR < 0.05.
105
Figure 3.23 CG6324, CR42767, and CG32706 display reduced H3K9me3 with forced phf7 expression. Plot plot of the H3K9me3 enrichment control ey (gray) and HA-phf7,nos>phf7 (blue) ovaries. The y-axis represents the calculated log concentration of H3K9me3 (normalized to input) over each 2 gene body, labeled on the x-axis. Each dot represents one replicate. Average and standard deviation are shown with black lines. Black points represent average intensities. Error bars indicate standard deviation. Significance was calculated by Student’s T-test. p < 0.05 (*) and p < 0.01 (**).
Figure 3.24 RNAseq and H3K9me3 ChIPseq reads at the CG6324 locus. Genome browser view of the CG6324 locus. mRNA expression in Wild type ey 148 (yw/yw in pink), phf7 (green), and snf (orange) ovaries and H3K9me3 in ey 148 Control (HA-phf7,nos in gray), phf7 (blue), and snf (red) ovaries. The gene body of interest is outlined. Tracks are viewed on the same scale with the ey exception of the phf7 ovary RNAseq track. The 5’ end of the gene is shown on the left.
106
Figure 3.25 RNAseq and H3K9me3 ChIPseq reads at the CR42767 locus. Genome browser view of the CR42767 locus. mRNA expression in Wild type ey 148 (yw/yw in pink), phf7 (green), and snf (orange) ovaries and H3K9me3 in Wild ey 148 type (HA-phf7,nos in gray), phf7 (blue), and snf (red) ovaries. The gene body of interest is outlined. Tracks are viewed on the same scale with the 5’ end of the gene on the left.
Figure 3.26 RNAseq and H3K9me3 ChIPseq reads at the CG32706 locus. Genome browser view of the CG32706 locus. mRNA expression in ey 148 Wild type (yw/yw in pink), phf7 (green), and snf (orange) ovaries and ey 148 H3K9me3 in Control (HA-phf7,nos in gray), phf7 (blue), and snf (red) ovaries. The gene body of interest is outlined. Tracks are viewed on the same scale with the 5’ end of the gene on the left.
107 3.5 Discussion
Drosophila female germ cells must actively repress testis gene expression to maintain their female identity. Failure to do so results in germ cell tumor formation and infertility. phf7 has been identified as a key gene that must maintain sex-specific isoform expression [Shapiro-Kulnane, 2015; Smolko,2018]. Sex- specific transcription in females is regulated by H3K9me3 deposition over its testis- specific region, preventing protein expression [Smolko,2018]. Here we present data showing that failure to repress phf7 is sufficient to activate a testis transcriptional program and compromise female identity. Surprisingly, we find that
PHF7 participates in a genetic feedback mechanism, where it autoregulates itself and turns on expression of a subset of H3K9me3-regulated genes normally restricted to testis. Forcing phf7 in female germ cells decreases H3K9me3 at these loci, correlating with inappropriate gene expression. Derepression of this set of genes may facilitate a more global activation of testis genes (Fig. 3.27).
108 Figure 3.27 PHF7 autoregultes its own expression and activates a genetic feedback loop. Model of how failure to repress PHF7 expression in female germ cells results in ectopic expression of itself and other genes normally repressed by H3K9me3. This likely leads to the derepression of a widespread testis gene program, resulting in a failure to maintain female identity and germ cell tumor formation.
Our finding that PHF7 positively autoregulates itself in the female germline highlights the importance of actively regulating its sex-specific transcription.
Forced phf7 expression was sufficient to reduce H3K9me3 enrichment in the testis-specific first intron and switch TSS usage to transcribe the testis-specific phf7-RC. These results were also observed at CG34434, CG42299, skpE,
CG6324, CR42767, and CG32706. However, the mechanism by which this occurs remains unknown. Our previous work shows that setdb1 and Sxl regulate
H3K9me3 enrichment over the testis specific region to ensure sex-specific transcription. Sxl and setdb1 act upstream of phf7 to deposit H3K9me3. Therefore, it is possible that PHF7 may compromise facultative heterochromatin at these regions through recruitment of demethylases to dissolve H3K9me3 or act to prevent its deposition altogether. Alternatively, PHF7 may affect H3K9me3 spreading and prevent formation of the heterochromatin “islands” normally seen in female ovaries.
We also make the interesting observation that all H3K9me3-regulated genes in PHF7’s genetic feedback loop are located on the X chromosome.
Previous work suggests that the X-chromosome is relatively depleted for testis- specific genes [Parisi, 2003 and 2004]. However, some of the top testis-enriched genes are encoded on the X-chromosome [Reviewed by White-Cooper, 2010].
109 This opens up the intriguing question of whether or not the X-chromosome has evolved a chromatin-based mechanism in female germ cells to ensure sex-specific transcriptional regulation.
The directness of PHF7’s effect on H3K9me3 remains unknown. In vitro studies indicate that PHF7 is able to bind histones, specifically H3K4me2, a mark of active transcription [Yang, 2012]. While these experiments did not show binding to repressive H3K27me3, PHF7’s capacity to bind H3K9me3 histones was not tested. Additional biochemical studies should be completed to determine if PHF7 can directly bind H3K9me3 or other regulators of heterochromatin formation. The autoregulatory effects we see may also be indirect, where chromatin remodelers must erase H3K9me3 and possibly replace it with activating histone modifications
(Fig. 3.28).
Figure 3.28 Hypotheses for how inappropriate PHF7 may alter facultative heterochromatin formation.
110 We find it striking that our autoregulated genes have testis-specific expression patterns in adults [Smolko, 2018]. While the role of CG34434 is unknown, skpE and CG42299 are predicted to function as SUMO and ubiquitin E3 ligases. Studies have linked E3 ligases to chromatin remodeling and have been associated with repressive states [Dubiel, 2017; Wotton, 2017]. Recent work implicates the SUMO E3 ligase, Su(var)2-10, in regulating SETDB1 recruitment in
Drosophila ovaries to regulate H3K9me3 deposition [Ninova, 2019a]. It will be interesting to explore if the predicted E3 ligase genes reported here promote germ cell tumor formation and testis gene expression through remodeling the chromatin landscape.
In addition to autoregulating itself and activating a small number of
H3K9me3-regulated genes through genetic feedback, we find that ectopic phf7 is sufficient to more globally alter transcription to a more male-like fate. However, we find that its expression in females does not turn off ovary specific genes. This is in opposition to its proposed role as a repressor of female-specific genes in male germ cells [Yang, 2017]. Instead, phf7 expression in female germ cells is sufficient to turn on a set of testis-enriched genes. However, PHF7 function in female germ cells may not necessarily be equivalent to that in males. Discerning the normal function of PHF7 in male germ cells has been relatively difficult, as its loss in testis does not produce a robust phenotype [Yang, 2012]. Furthermore, it is normally only expressed in a small number of cells in the testis, posing a challenge for molecular studies. Differential RNAseq analysis utilizing double mutants that block germ cell differentiation to enrich for cells expressing PHF7 revealed very few
111 changes in gene expression [Yang, 2017]. However, arresting differentiation may have unintended genetic consequences that perturb the PHF7’s normal functions.
Modified studies to determine the molecular consequences of phf7 loss in male germ cells will help in assessing whether or not it acts in the same manner in female germ cells.
Overall, our studies uncover a genetic feedback loop that is activated when sex-specific transcriptional regulation of phf7 is lost. This genetic feedback results in widespread changes in gene transcription, leading to loss of female identity and tumor formation. In humans, an ortholog of phf7 with multiple isoforms exists and expression is also restricted to testis [Fagerberg, 2014]. Its most conserved region exists in its PHD domain, which are commonly found on chromatin associating proteins [Yang, 2017; Reviewed by Sanchez and Zhou, 2011]. Proteins containing
PHD fingers have been implicated in leukemia, melanoma, and myeloid tumors
[Gatchalian, 2013; Soto-Feliciano, 2017]. Additionally, individuals with Disorders of Sex Development (DSD) are more susceptible to germ cell tumor development
[Cools, 2009]. It would be interesting to determine if phf7 or other sex-specific gene candidates are disrupted in DSD patients and if their disease phenotypes, including germ cell tumors, are attributable to a mechanism similar to what we report in flies.
112 Chapter 4
Concluding Remarks
113 Gametogenesis relies on the activation and maintenance of sex-specific genetic programs to create egg or sperm from a pool of germline stem cells. In
Drosophila females, failure to repress spermatogenesis genes through adulthood results in a germ cell tumor phenotype and infertility [Shapiro-Kulnane, 2015]. In this dissertation, I provided a mechanism by which Drosophila female germ cell identity is actively maintained. I described how facultative heterochromatin islands marked by H3K9me3 are deposited by SETDB1 and SXL. Germline depletion of either of these genes results in tumor formation and testis gene expression independent of transposable element repression. PHF7, a regulator of male germ cell identity was identified as a key gene repressed by H3K9me3 in female germ cells. Failure to prevent transcription of the testis-specific, phf7-RC, isoform results in ectopic protein expression. This activates a genetic feedback mechanism, where
PHF7 turns on expression of itself and a subset of SETDB1-regulated genes through loss of H3K9me3. This work provides novel insights towards how germ cell sex identity is preserved throughout adulthood to allow for the faithful production of gametes.
While this work provides a means by which female identity is maintained, a number of intriguing questions remain. First, how is SETDB1, a ubiquitously expressed protein, sex-specifically recruited to its targets in female germ cells to deposit H3K9me3? It also remains to be determined how ectopic PHF7 disrupts
H3K9me3 and whether or not it directly activates a global testis gene program.
Furthermore, does ectopic PHF7 in female germ cells target the same genes by the same mechanism it normally employs in males? Lastly, is maintenance of sex-
114 identity required in mammalian germ cells and is maintenance achieved through a mechanism similar to that in flies? In this chapter, I will discuss hypotheses and future directions and towards answering these questions.
4.1 Sex-specific recruitment of SETDB1 in Drosophila female germ cells
While I identified SETDB1 and SXL as depositing H3K9me3 on a small number of testis genes, how they sex-specifically target these loci has not been interrogated. SXL is an RNA binding protein that regulates sex-specific processes through splicing and translational regulation [Reviewed by Moschall, 2017]. Its targets in adult germ cells are not fully known. However, RNA immunoprecipitation coupled to sequencing (RIPseq) in Drosophila primordial germ cells identified
Su(var)2-10 as a potential SXL target [Ota, 2017]. Germline knockdown of
Su(var)2-10 resulted in a tumor phenotype similar to the germ cell tumors discussed in this thesis. This led the authors to hypothesize that SXL may promote its expression. Su(var)2-10 is a SUMO E3 ligase whose activity induces H3K9me3 deposition [Ninova, 2019a]. Interestingly, loss Su(var)2-10 also leads to the derepression of a small number of euchromatic testis-enriched genes [Ninova,
2019b]. One of these is CG34434, which I identified in chapter 3 as being repressed by H3K9me3 and part of the genetic feedback mechanism activated by
PHF7.
Data from from Ota et al, 2017 and Ninova et al, 2019 allows for the prediction that SXL functions in two distinct pathways in female germ cells. In one,
SXL represses nanos translation to facilitate expression of differentiation factors
115 and allow for the GSC/CB fate switch [Chau, 2012]. In the other, it activates
Su(var)2-10 translation. Studies show that Su(var)2-10, SETDB1, and WDE contain SUMO-interaction motifs and can themselves be SUMOylated [Uchimura,
2006; Koch, 2009; Ninova, 2019]. Furthermore, SUMOylation has been implicated in the assembly of protein complexes that allow for the modification of transcription factors and histone tails during heterochromatin formation [Reviewed by Gill,
2005]. One can imagine that SXL may regulate Su(var)2-10 by promoting expression of a female-specific isoform. This could allow for SUMOylation of a yet unidentified protein complex to recruit SETDB1 and repress spermatogenesis gene expression (Fig. 4.1). In the future, it would be interesting to combine proteomic and genetic studies to identify SUMOylated SETDB1-interacting proteins that facilitate preservation of female germ cell identity.
116 Figure 4.1: Model for how SXL and SETDB1 sex-specifically deposit H3K9me3. Diagram depicting SXL as functioning in two distinct pathways. In one, SXL represses nanos translation to translate differentiation mRNAs and allow for the GSC to CB fate switch. In the other, SXL activates Su(var)2-10 translation. Su(var)2- 10 allows SUMOylation of unidentified complex members that sex-specifically recruit SETDB1 to its targets
4.2 PHF7 activation of the testis gene program in females
Data presented in chapter 3 shows that PHF7 activates a genetic feedback loop through dissolution of H3K9me3 at itself and an additional 6 genes. However, the directness of this regulation and how a more global testis gene program is activated with ectopic PHF7 are not known. Here, I discuss hypothesis for how
PHF7 might alter the chromatin landscape to shuttle female germ cells into the spermatogenesis pathway.
4.2.1 Mechanism of phf7 autoregulation in female germ cells
PHF7 contains a PHD domain [Yang, 2012], which are frequently found in chromatin associating proteins [Reviewed by Sanchez and Zhou, 2011]. Previous work shows that PHF7 preferentially binds H3K4me2 [Yang, 2012], a mark of active transcription [Reviewed by Pinskaya, 2009], and may act in testis to read and interpret the chromatin landscape to regulate sex specific gene expression
[Yang, 2017]. However, this proposed mechanism has not been fully characterized. Here, co-immunoprecipitation assays were completed in vitro where PHF7 was mixed with K4me2, K4me3, and K27me3 modified H3 peptides
[Yang, 2012]. Furthermore, the ability to ChIP for PHF7 has been a significant technical roadblock for myself and others [Yang, 2017]. This has prevented
117 analysis of the genes PHF7 directly regulates and if it correlates with H3K4me2 in germ cells.
While it is not definitively clear that PHF7 is a chromatin “reader”, preliminary observations indicate that, in female germ cells, the PHD domain is required for germ cell tumor development. Here, I generated a GFP-tagged cDNA transgene of the phf7 coding region under the control of the pUASP promoter with random insertion in the genome [Rørth, 1998]. The coding region contains either the wild type sequence or an in-frame deletion of its annotated PHD/histone binding region [Marchler-Bauer, 2017] (Fig. 4.2). My preliminary data shows that forcing phf7 from its endogenous locus (nos>GFP-phf7WT) results in tumorous ovarioles (56% tumor, n=31/55). This was determined by qualitatively assessing overall germaria morphology (Fig. 4.3A). Removal of the predicted histone binding region (nos>GFP-phf7DPHD) had almost no phenotypic impact (93% normal, n=38/41) (Fig. 4.3 B). These results support that the PHD domain is required for
PHF7 function and that it may act through interpreting the chromatin landscape.
However, these experiments were done with untargeted insertions and nos>GFP- phf7WT were still able to lay eggs. Our lab is currently working towards repeating and improving these experiments in constructs with stronger germline expression and targeted insertions [DeLuca, 2018]. Additional biochemical analyses to determine the post-translational histone modifications PHF7 interacts with in germ cells should also be completed.
118
Figure 4.2: Deletion of the phf7 histone binding region. The coding region cDNA of phf7 was cloned and N-terminally tagged with GFP. The phf7 coding region is either (A) wild type or (B) contains an in-frame deletion of its PHD region. The GFP-tagged coding regions are present under the control of the pUASP promoter to allow for expression in the germline.
Figure 4.3: Tumor formation requires PHF7’s PHD domain. Representative ΔPHD immunostaining for (A) a normal nos>GFP-phf7 and (B) tumorous WT nos>GFP-phf7 germaria stained for GFP(green, white in A’ and B’) and DNA (magenta, white in A” and B”). Scale bar indicates 25 um. The percentage of tumor formation was qualitatively assessed through observation of germarium ΔPHD morphology. For nos>GFP-phf7 , the number of ovarioles with normal WT morphology are reported. For nos>GFP-phf7 , the number of tumorous ovarioles are reported
119 The necessity of the PHD domain allows us to predict that PHF7 regulates transcription through chromatin binding. However, of the H3K9me3-regulated genes identified in chapter 3, only phf7 possesses H3K4me2 in wild type ovaries
(Fig. 4.4-4.7). Interestingly, H3K4me2 enrichment is restricted to the phf7-RA TSS that is normally expressed in ovaries (Fig. 4.4). If PHF7 binds H3K4me2 as predicted, it is likely that it only directly regulates itself. In this model, PHF7 would bind where H3K4me2 is already present over the phf7-RA TSS in ovaries. PHF7 may then recruit histone modifiers to alter the chromatin landscape, allowing for testis-specific TSS usage (phf7-RC). In this scenario, genetic feedback and
H3K9me3 dissolution at the remaining 6 genes would likely be indirect.
Alternatively, PHF7 may interact with H3K9me3. PHD fingers are not thought to be capable of directly binding H3K9me3 [Reviewed by Sanchez and
Zhou, 2011]. However, this does not exclude the possibility that H3K9me3- interacting proteins indirectly recruit PHF7 to H3K9me3-marked heterochromatin.
Previous studies have shown that chromatin associating proteins containing PHD domains can be recruited to H3K9me3 through interactions with HP1a [Borgel,
2016]. Aside from its PHD domain, the majority of PHF7’s protein composition remains uncharacterized [Yang, 2012]. This leaves open the possibility that additional protein interaction domains exist. One could speculate that PHF7 binds
HP1a, disrupting heterochromatin formation and facilitating expression at its target genes. To discern these two scenarios, it will be important to determine the histone modifications and genomic regions PHF7 binds to in germ cells.
120 Both of these predictions present the intriguing possibility that inappropriate
PHF7 acts as a pioneer factor in females to remodel the chromatin landscape. In either case, targeting of PHF7 would likely be accompanied by recruitment of additional chromatin remodeling proteins. For example, dissolution of H3K9me3 may occur through an H3K9 demethylase. These models predict that loss of the
H3K9 demethylases, KDM4A and KDM4B, would rescue tumor phenotypes resulting from misexpression of phf7. Knockout alleles of the Drosophila demethylases are available and have been provided to the lab to test this hypothesis [Shalaby, 2017]. In parallel, it would be interesting to determine if loss of H3K9me3 is accompanied by its replacement with activating H3K4me2 to match what is observed in male germ cells (Fig. 4.4-4.7). If so, PHF7 would also recruit
H3K4 methyltransferases and their loss might also ameliorate the tumor phenotypes we observe. Overall, I predict that ectopic PHF7 modifies sex-specific gene expression by recruiting histone modifiers to its targets to create a testis-like chromatin landscape (Fig. 4.8).
Figure 4.4: RNA and H3K4me2 profiles for phf7 in ovaries and testis. Genome browser view of the phf7 locus with aligned RNAseq reads in wild type ovaries (magenta) and testis (green) as well as H3K4me2 reads in wild type ovaries (light pink) and bam mutant testis (purple). The image is shown so that the 5’ end of the gene is on the left. The gray box indicates the testis- specific genomic region. Scales for each track are indicated on the right.
121
Figure 4.5: RNA and H3K4me2 profiles for CG34434 in ovaries and testis. Genome browser view of the CG34434 locus with aligned RNAseq reads in wild type ovaries (magenta) and testis (green) as well as H3K4me2 reads in wild type ovaries (light pink) and bam mutant testis (purple). The image is shown so that the 5’ end of the gene is on the left. The gray box indicates the annotated gene region. Scales for each track are indicated on the right.
Figure 4.6: RNA and H3K4me2 profiles for CG42299 in ovaries and testis. Genome browser view of the CG42299 locus with aligned RNAseq reads in wild type ovaries (magenta) and testis (green) as well as H3K4me2 reads in wild type ovaries (light pink) and bam mutant testis (purple). The image is shown so that the 5’ end of the gene is on the left. The gray box indicates the annotated gene region. Scales for each track are indicated on the right.
122 Figure 4.7: RNA and H3K4me2 profiles for skpE in ovaries and testis. Genome browser view of the skpE locus with aligned RNAseq reads in wild type ovaries (magenta) and testis (green) as well as H3K4me2 reads in wild type ovaries (light pink) and bam mutant testis (purple). The image is shown so that the 5’ end of the gene is on the left. The gray box indicates the annotated gene region. Scales for each track are indicated on the right.
Figure 4.8: Prediction for PHF7 autoregulation through chromatin interactions. Schematic representing how PHF7 may autoregulate gene expression through binding histone modifications using the phf7 locus as an example. In females, PHF7 may bind H3K9me3 indirectly through interactions with a heterochromatin regulator such as HP1a. Alternatively, it may bind H3K4me2 already present. In both scenarios, PHF7 contributes to ablation of H3K9me3 and the possible spreading of an activating mark such as H3K4me2.
123 4.2.2 Activation of the spermatogenesis pathway in germ cell tumors
In testis, a class of “meiotic arrest” mutants were identified and classified as necessary for spermatocytes to differentiate and enter meiosis [Reviewed by
White-Cooper, 2010]. These genes comprise members of either the tTAF or tMAC complexes and contribute to the activation of genes required for spermatogenesis
[Reviewed by White-Cooper, 2010]. In all tumors discussed in this work (phf7ey, snf 148, setdb1 GLKD, wde GLKD, and hp1a GLKD), a number of these components were reported as significantly upregulated (Fold Change > 2 and FDR
< 0.05) (Fig. 4.9). This suggests that ectopic PHF7 may cause female germ cells to attempt to undergo spermatogenesis. This is supported by previous studies that show female germ cells overexpressing phf7 in a male somatic environment are able to undergo spermatogenesis [Yang, 2012]. My own RNAseq data suggests that PHF7 may attempt to turn on spermatogenesis in a cell autonomous manner.
All tumors misexpress the testis-specific dany transcript (Fig. 4.9). While dany is thought to function independently of tMAC and tTAF, it is still required for spermatocyte differentiation [Trost, 2016]. We were able to determine if dany mRNA is translated into protein in female snf148 and phf7ey tumors using a functional gfp-dany transgene [Trost, 2016]. In testis, DANY protein is present throughout spermatocyte stages and does not turn on until PHF7 expression turns off (Fig. 4.10A). No protein expression is observed in control ovaries (Fig. 4.10 C).
In snf148 and phf7ey tumors, DANY protein is expressed (Fig 4.10 B,D).
Furthermore, PHF7 and DANY protein expression appear to be mutually exclusive in phf7ey ovarioles (Fig. 4.10B). These exciting preliminary results indicate that
124 ectopic phf7 may turn on a male meiosis gene program in a temporal manner.
Additionally taf12L, a member of the tTAF complex, is ectopically expressed in all tumors (Fig. 4.9). However, tools to study its protein expression are currently unavailable. It would be interesting to determine if this protein is misexpressed in the same temporal fashion as DANY.
We also noted a number of testis-specific proteins involved in translation as being significantly upregulated in our tumors (Fig. 4.9). These ribosomal components (rpl22-like and rpL37b) and translation initiation factors (eIF4E-3 and eIF4E-6) are exclusive to testis and are required for spermatogenesis [Kai, 2005;
Kearse, 2011; Hernandez 2012; Ghosh, 2015]. It is my hypothesis that these ribosomal proteins and translation initiation factors contribute to activation of male meiosis and tumor formation by facilitating the translation of misexpressed spermatogenesis genes. This prediction is especially intriguing as tTAF/tMAC genes such as achi, vis, and mia normally have low levels of RNA accumulation
(Fig. 4.9). The presence of spermatogenesis translation components may allow for translation of these mRNAs.
In the future, it would be interesting to determine how these male “meiotic arrest” and translational regulators are normally repressed and if their ectopic expression contributes to tumor formation. Are these genes regulated directly or indirectly by PHF7? What spermatogenesis genes are turned on in females as a consequence of tTAF and tMAC gene expression? Furthermore, would loss testis- specific translation factors in tumors prevent protein expression of spermatogenesis genes and rescue the phenotype?
125
Figure 4.9: Transcripts specific to male meiosis and translation are upregulated in female germ cell tumors. Heat map of log FPKM values 2 ey 148 from RNAseq analysis in wild type, phf7 , snf , setdb1 GLKD, wde GLKD, and hp1a GLKD ovaries. Genes are grouped by their complex/function. Black boxes indicate genes that are significantly upregulated in female germ cell tumors (log fold change > 2 and FDR < 0.05). 2
126 Figure 4.10: PHF7 activates expression of DANY in a temporal fashion. A,B Representative confocal images of (A) control testis (dany- gfp;HA-phf7,nos) and (B) ovarian tumors ectopically expressing phf7 (dany-gfp; HA-phf7,nos>phf7ey). Flies contain the HA-phf7 and GFP-dany transgenes. Tissue is stained for HA (magenta) and GFP (green). C,D 148 Representative confocal images of (C) control (snf ;Fm7;dany-gfp) and 148 148 (D) snf ;snf ;dany-gfp. Images are stained for DNA (magenta) and GFP(green). Scale bar represents 50 um.
4.3 Maintenance of sex identity in male germ cells
Work in this thesis focused on how Drosophila female germ cells maintain their sex-identity. However, maintenance of sex-identity in adult male germ cells remains to be thoroughly explored. Germline knockdown of setdb1 did not appear to have any effect on male fecundity (data not shown). This suggests that
H3K9me3 is dispensable for male germline stem cell differentiation. However, this does not preclude an alternative, chromatin-based, mechanism for maintenance
127 of male germ cell fate. One could also speculate the male genetic program is the default in Drosophila germ cells. If so, female germ cells actively prevent its activation through H3K9me3, allowing for a female fate.
In this work, I show that failure to prevent PHF7 protein expression in female germ cells is critical for maintaining sex-identity. While PHF7 is thought to function as a regulator of male germ cell identity, my work did not address whether it has the same genetic targets in both sexes. It is worthwhile to note that genes discussed in chapter 3 that are part of PHF7’s genetic feedback loop possess H3K4me2 enrichment in testis in a mutant background that enriches the tissue for undifferentiated cells (bam-/-) [Yang, 2017] (Fig. 4.4-4.7). However, PHF7 binding to H3K4me2 in germ cells and if the overlap in gene targets in males and females remains to be confirmed.
If PHF7 regulates the same genes in both sexes, one would hypothesize that those ectopically expressed in ovaries with forced phf7 would be downregulated in phf7 loss of function testis. In an attempt to determine if this is the case, I reanalyzed published RNAseq datasets on control and phf7 mutant testis [Yang, 2017]. In this work, all experiments were completed in a bag of marbles (bam) mutant background. Loss of bam blocks germ cell differentiation, enriching testis for cells that normally express phf7. Differential expression analysis revealed minimal changes in gene expression (Fig. 4.11). This would indicate that PHF7 does not play a major role in male germ cell transcriptional regulation. However, I feel that the use of double mutants in this study may have prevented the identifying PHF7’s transcriptional role in male germ cells. In wild
128 type testis, PHF7 is highly expressed in both GSCs and spermatogonial cells that express BAM [Yang, 2017]. By eliminating bam and blocking differentiation,
PHF7’s activity in male germ cells may be hindered. Furthermore, PHF7 may target genes that are poised for transcriptional activation as differentiation proceeds. By blocking differentiation via bam mutations, transcriptional defects caused by loss of phf7 would be lost.
Figure 4.11: Loss of phf7 in bam mutant testis shows minimal changes in -/- gene expression. Scatter plot of transcript levels in control (bam ) and phf7 null -/- mutant (phf7 ;bam ) tesits. Gene expression is plotted as the log of the FPKM 2 value. Genes significantly altered in phf7 mutants relative to controls (fold change > or < 2 and FDR < 0.05) are indicated by red points.
Determining the molecular role of PHF7 in testis is technically challenging as it is only expressed in a small number of cells relative to the entire tissue. To perform molecular studies, it is necessary to enrich for cells that actually express protein. In the future, it will be important to improve experimental designs to study the roles of proteins expressed exclusively in undifferentiated germ cells.
129 Advances in genetic tools and sequencing technologies will hopefully make this possible in the coming years. Developments in single cell RNAseq (scRNAseq) have allowed for transcriptome studies in rare cell populations and have shed light on gene expression through stem cell lineages [Reviewed by Kumar, 2017]. Usage of this technology in Drosophila gonads would help to shed light on gene expression changes as GSCs differentiate. This knowledge could be applied to how proteins such as PHF7 are regulating transcription.
In addition to innovations in RNA sequencing technologies, the ability to profile binding of chromatin associated proteins normally expressed in a small number of cells is improving. For example, Targeted DamID (TaDa) allows for tissue-specific overexpression of chromatin associated proteins fused to a Dam methylase [Reviewed by Ameele, 2019]. Here, methylation of regions surrounding binding sites can be measured without the use of fixation or immunoprecipitation.
Additionally, single cell ChIPseq experiments have been successful in mammalian cell lines [Rotem, 2015]. It would be exciting to answer how cell-type specific genes, such as phf7, are acting as these technologies further develop and become more widely accessible to the scientific community.
4.4 Maintenance of sex-identity in mammalian germ cells
While it is clear that maintenance of germ cell sex-identity is crucial in
Drosophila females, its extension to mammalian models remains relatively unexplored. In humans, both ovary and testis can form GCTs, with tumors occurring more frequently in testis [Reviewed by Jorgensen, 2015 and Gonzalez-
130 Exposito, 2016]. Genome wide association studies have consistently linked mutations in germ cell development genes with an increased risk of testicular GCT development [Turnbull, 2010; Chung, 2013; Ruark, 2013; Koster, 2014]. For example, variants in the DMRT1 locus were identified in each of these studies.
DMRT1 in humans and mice exhibits sexually dimorphic timing and expression patterns that control the mitosis to meiosis switch in germ cells [Jorgensen, 2012].
Identification of these genetic risk factors points to a loss of sex-identity as possibly driving GCT development. In addition to gene variants involved in germ cell development, ATF7IP, the wde homologue, was identified as increasing testicular cancer risk [Turnbull, 2010]. ATF7IP/wde has been shown to be required for
SETDB1 function in human cells [Timms, 2016], supporting a role for heterochromatin in preventing human GCT formation. Therefore, it is possible that the results presented in this thesis can be extended to mammalian models.
4.4.1 SETDB1 requirement in mammalian germ cells
In mouse primordial germ cells (PGCs), sex-specific transcriptome heterogeneity has been reported [Sakashita, 2015]. However, how transcription of male and female specific transcripts are regulated remained unanswered.
SETDB1 has been shown to be required for mouse germ cell development [Liu,
2014; Mochizuki, 2018]. Embryonic knock out of setdb1 in PGCs results in fewer germ cells being formed and maintained through adulthood. As in flies, SETDB1- mediated H3K9me3 represses transposable elements in mouse germ cells [Liu,
2014]. Interestingly, the authors of this study reported that the types of TEs
131 repressed were sex-specific. Furthermore, changes in protein-encoding genes due to setdb1 knockdown showed minimal overlap between males and females. These data support SETDB1’s involvement in also regulating sex-specific transcription in mammalian germ cells.
Datasets provided in Liu, 2014 allowed us to ask if genes upregulated in female germ cells are normally expressed in testis (Analysis by Dr. Ron Conlon,
CWRU Department of Genetics and Genome Sciences). Of the 157 genes reported to be significantly upregulated, 15% are expressed in testis or are involved in male fertility. Furthermore, only 3 of these genes are driven by a TE insertion (Fig. 4.12). These preliminary data support a role for SETDB1-mediated heterochromatin in also maintaining female germ cell fate in mice independently of TE repression. Future work could extend these analyses by knocking out setdb1 at different stages of germ cell development. One might predict that it’s loss at different developmental timepoints would result in germ cell tumors rather than loss. It would also be important to further examine if upregulation of spermatogenesis genes contributes to female germ cell defects and if they are directly regulated by heterochromatin formation.
132
Figure 4.12: setdb1 knockout in mouse female germ cells upregulates genes normally expressed in testis. Pie chart representing significantly upregulated genes in females where setdb1 was knocked out in primordial germ cells (Liu, 2014). 15% of these normally show expression in testis or are involved in male fertility (in blue); data provided by Dr. Ron Conlon). The majority of these loci do not contain a derepressed transposable element (in green).
4.4.2 PHF7 expression in mammals
In this work, I identified repression of PHF7 protein as being critical for maintaining Drosophila female germ cell identity. phf7 was originally identified in a screen for sex-specific gene expression in embryonic gonads [Yang, 2012]. It was given its name due to its homology with human phf7, where protein is also exclusively expressed in testis (The Human Protein Atlas) (Fig. 4.13). Here, expression is reported as being highest in differentiating spermatocytes and maturing sperm rather than undifferentiated germ cells. Very little functional analysis on phf7 has been completed in humans or mouse models. However, analysis of mRNA expression in testicular samples revealed decreased expression in patients exhibiting spermatogenesis arrest [Xiao, 2002]. It was also shown that expression of human phf7 cDNA in Drosophila male germ cells rescued fertility
133 defects [Yang, 2012]. It is possible that phf7 regulates germ cell development similarly to what is reported in flies. Additionally, information provided by The
Cancer Genome Atlas (TCGA) reports expression of PHF7 in tumors originating from tissues outside of the testis. These include colorectal, endometrial, or pancreatic cancers (Fig. 4.14). Future work exploring how phf7 is tissue- specifically regulated in mammals may be important in furthering our understanding of germ cell development and tumor formation.
Figure 4.13: PHF7 protein in humans is only expressed in testis. Protein expression scores in human tissues for PHF7 provided by The Human Protein Atlas
Figure 4.14: PHF7 protein expression in human cancers. The percentage of patients expressing PHF7 in tumor samples derived from different tissues. Data provided by The Cancer Genome Atlas
134 One interesting question is whether or not phf7 is transcriptionally regulated in mammals similarly to what we report in flies. In this thesis, I have shown that phf7 is sex-dimorphically transcribed with two TSS’s in flies (Fig. 4.15A). In mice, only one annotated isoform of phf7 exists (Fig. 4.15B). However, in humans, isoforms with alternative TSS’s have been reported (Fig. 4.15C). Is the upstream
TSS in humans exclusively used in testis and does only this isoform produce a functional protein? Additionally, the phf7 genomic region in all three species contains long introns that take up at least 40% of the gene region. In Drosophila ovaries, this intron is normally enriched for H3K9me3. Is H3K9me3 present over these regions in mouse and human ovaries? If so, is it regulated by SETDB1 (Fig.
4.15)?
Figure 4.15: Annotated phf7 gene region in Drosophila, mouse, and human. Schematic of the phf7 locus in (A) Drosophila, (B) mice, and (C) humans. Possible sex -specific TSS’s and H3K9me3 deposition by SETDB1 in mice and humans is depicted.
135 Currently, a limited number of datasets exist in mouse and human germ cells to answer any of these questions. RNAseq and H3K9me3 ChIPseq information in control and SETDB1 knockout male and female mouse PGCs is currently available [Liu, 2014]. In humans, H3K9me3 ChIP seq in human testis and ovaries is available through the ENCODE consortium [Consortium, 2012].
However, these datasets were generated from single individuals and tissue samples contain many different cell types. This greatly limits data quality and data interpretation. Recent advances in testicular organoid development and in vitro culturing of ovarian cell types may make human germ cell experiments more attainable in the future [Baert, 2017; Jung, 2017; Pendergraft, 2017].
4.5 Conclusion
The work presented in this dissertation highlights the requirement for
Drosophila female germ cells to actively maintain their sex-identity through gametogenesis. I identify H3K9me3 marked facultative heterochromatin as a mechanism by which spermatogenesis genes are repressed. Here, the testis specific isoform of phf7 was identified as a key H3K9me3 repressed gene. Ectopic expression of PHF7 in female germ cells is sufficient to form tumors and upregulate a testis gene program, including a subset that function in a genetic feedback loop.
Future work will hopefully determine how H3K9me3 is sex-specifically recruited in female germ cells and if this mechanism is conserved in mammalian systems. It will also be important to further explore the mechanism by which PHF7 alters female fate and if it functions in a similar manner in male germ cells. Overall, this
136 work provides insight into normal germ cell development and how sex-specific genetic programs are regulated to ensure gamete production.
137 APPENDIX
EXTENDED TABLES
Extended Table 1: Genes significantly upregulated in phf7ey/+;HA-phf7,nos- gal4 tumors relative to wild type (yw/yw) ovaries
gene WT (yw/yw) phf7ey log2 fold p-value q-value FPKM FPKM change Oli 0.0000 1.09511 inf 0.00005 0.00012399 CG32263 0.0000 1.09769 inf 0.00005 0.00012399 CG14069 0.0000 1.09806 inf 0.00005 0.00012399 CG15200 0.0000 1.10957 inf 0.00005 0.00012399 CR45695 0.0000 1.11695 inf 0.0174 0.0306537 CG7211 0.0000 1.118 inf 0.00005 0.00012399 CG13428 0.0000 1.12547 inf 0.00005 0.00012399 Cpr49Ag 0.0000 1.17867 inf 0.00005 0.00012399 CG18258 0.0000 1.2078 inf 0.00005 0.00012399 Or9a 0.0000 1.2146 inf 0.00005 0.00012399 CG31639 0.0000 1.26547 inf 0.00005 0.00012399 CG5255 0.0000 1.33299 inf 0.00005 0.00012399 CG1394 0.0000 1.35411 inf 0.00005 0.00012399 CR45503 0.0000 1.38043 inf 0.00175 0.00363453 CG17376 0.0000 1.44119 inf 0.00005 0.00012399 CG18278 0.0000 1.54532 inf 0.00005 0.00012399 CG31926 0.0000 1.65793 inf 0.00005 0.00012399 CR45755 0.0000 1.6996 inf 0.00015 0.00035649 Mst84Db 0.0000 1.7906 inf 0.00005 0.00012399 CR45625 0.0000 1.94109 inf 0.00005 0.00012399 CG3698 0.0000 1.97346 inf 0.00005 0.00012399 CG2121 0.0000 2.04018 inf 0.00005 0.00012399 Peritrophin- 0.0000 2.04026 inf 0.00005 0.00012399 15b CG7330 0.0000 2.30028 inf 0.00005 0.00012399 CR44826 0.0000 2.53073 inf 0.00005 0.00012399 CG17377 0.0000 2.65368 inf 0.00005 0.00012399 CG32642 0.0000 2.72635 inf 0.00005 0.00012399 CR44103 0.0000 2.89819 inf 0.00005 0.00012399 CR44285 0.0000 2.9515 inf 0.0006 0.00132671 CG31226 0.0000 3.02217 inf 0.00005 0.00012399 ocn 0.0000 3.0429 inf 0.00005 0.00012399 CG34168 0.0000 3.82292 inf 0.00005 0.00012399 CG34170 0.0000 5.39187 inf 0.00005 0.00012399 CG31928 0.0000 7.00938 inf 0.00005 0.00012399 Cp16 0.0000 8.43307 inf 0.00005 0.00012399 CG15571 0.0000 9.09362 inf 0.00005 0.00012399
138 CR45453 0.0000 9.15379 inf 0.00005 0.00012399 CG10163 0.0000 9.1769 inf 0.00005 0.00012399 CG42821 0.0000 10.9909 inf 0.00005 0.00012399 CG1077 0.0000 13.118 inf 0.00005 0.00012399 CG14187 0.0000 22.8188 inf 0.00005 0.00012399 Cp15 0.4796 783.568 10.6739 0.0033 0.00657108 CG14309 0.0402 35.7067 9.79599 0.00615 0.0117217 CG15721 0.1064 67.4892 9.30854 0.00005 0.00012399 CG15570 0.0473 19.2688 8.67137 0.0013 0.00275014 Cp38 0.8220 316.239 8.58764 0.00005 0.00012399 yellow-g2 0.1272 45.9093 8.49538 0.00355 0.00702743 Cp36 1.2023 423.385 8.46006 0.00005 0.00012399 CG4009 0.1293 33.7139 8.02697 0.00005 0.00012399 Muc12Ea 0.0088 1.2547 7.14919 0.028 0.0474175 Osi20 0.1503 21.2336 7.14261 0.00355 0.00702743 Cp7Fa 0.3427 36.0284 6.71616 0.0051 0.00985593 yellow-g 0.3752 38.5173 6.6817 0.00005 0.00012399 CG13936 0.1680 12.8219 6.25441 0.00025 0.00058003 Ir7c 0.2413 17.6669 6.19424 0.00005 0.00012399 CG13299 1.2152 84.765 6.12415 0.00005 0.00012399 Taf12L 0.8546 59.0943 6.11156 0.00005 0.00012399 CG42704 0.7612 51.4879 6.07977 0.00065 0.00143132 CG34434 0.1977 13.098 6.04955 0.00005 0.00012399 Cp7Fb 0.6888 43.4185 5.97812 0.00005 0.00012399 Syt4 0.0803 4.92918 5.93916 0.00005 0.00012399 CG42822 0.3364 16.7257 5.63582 0.00325 0.00648062 CR45631 0.4160 20.0638 5.59193 0.0013 0.00275014 Rpt6R 0.1375 6.22378 5.50014 0.0013 0.00275014 Cp7Fc 0.7538 32.7606 5.44161 0.00005 0.00012399 Femcoat 0.1447 6.02778 5.38056 0.01425 0.0255289 Vml 9.7533 375.067 5.26512 0.00005 0.00012399 RpL37b 0.7064 26.9965 5.25622 0.00125 0.00265143 CG13577 0.0941 3.55727 5.23969 0.00195 0.00402261 Obp56d 0.1929 7.00541 5.18255 0.02425 0.0415146 CR43615 4.8602 175.573 5.17492 0.00005 0.00012399 CG11381 0.3681 12.7492 5.11412 0.00005 0.00012399 43435 16.0134 536.638 5.0666 0.00005 0.00012399 CG10175 0.1009 3.36114 5.0585 0.00005 0.00012399 Fcp3C 0.6729 21.5691 5.00244 0.00005 0.00012399 CG6324 0.4599 14.1161 4.94004 0.00005 0.00012399 Cyp6a17 0.2969 9.01689 4.92446 0.00005 0.00012399 RpL22-like 0.4803 14.4209 4.90799 0.00005 0.00012399 Vm26Aa 48.6287 1437.76 4.88587 0.00005 0.00012399 Yp2 75.3836 2177.28 4.85213 0.00005 0.00012399 CR45601 109.9700 3096.04 4.81524 0.00005 0.00012399 kkv 0.4717 13.1257 4.79837 0.00005 0.00012399 Yp1 37.1718 1030.08 4.7924 0.00005 0.00012399 GABA-B-R1 0.1485 4.03313 4.76339 0.00005 0.00012399
139 Lip4 3.5197 91.5839 4.70156 0.00005 0.00012399 Nox 0.2858 7.21378 4.65785 0.00005 0.00012399 CG17104 0.0432 1.01339 4.55317 0.0111 0.020285 w 0.9894 23.1513 4.54845 0.0028 0.00564312 skpE 0.2812 6.55197 4.54241 0.0013 0.00275014 CG6999 0.1894 4.40034 4.53808 0.00325 0.00648062 Phf7 13.9280 319.725 4.52078 0.00005 0.00012399 GstD2 1.5929 35.5457 4.47999 0.00005 0.00012399 snoRNA:U14: 353655.0000 7480000 4.4025 0.0123 0.0222791 30Eb ktub 0.0789 1.54517 4.29175 0.01885 0.0329989 CG42705 0.7664 14.8013 4.27156 0.00015 0.00035649 CG5958 12.9865 249.305 4.26283 0.00005 0.00012399 Actbeta 0.4410 8.18463 4.21406 0.00005 0.00012399 CG32706 0.4058 7.4525 4.19874 0.00005 0.00012399 conv 0.7154 12.6327 4.14219 0.00005 0.00012399 CG43134 132.0770 2318.84 4.13395 0.00005 0.00012399 CG17625 0.3552 6.13905 4.11119 0.0169 0.0298514 Ilp8 1.8637 32.1311 4.10771 0.00005 0.00012399 St4 1.3585 22.7424 4.06525 0.00005 0.00012399 Sas 0.4959 8.23092 4.05287 0.00485 0.00940829 CG12477 0.3096 5.11506 4.04637 0.00005 0.00012399 CG42857 2.0549 33.7041 4.03577 0.00005 0.00012399 CG10005 0.6498 10.4693 4.01001 0.00005 0.00012399 psd 4.9729 80.0289 4.00838 0.00005 0.00012399 CG11327 0.1385 2.19372 3.98516 0.0001 0.0002417 CG12398 2.7540 43.1663 3.97033 0.00435 0.00849815 CG43132 0.4151 6.36737 3.93908 0.01715 0.0302457 Phlpp 0.1486 2.27378 3.93579 0.00005 0.00012399 danr 0.0861 1.29887 3.91479 0.0068 0.0128713 Vm34Ca 118.7340 1774.81 3.90186 0.00005 0.00012399 Cp18 0.3736 5.48859 3.8768 0.0038 0.00748785 TrxT 0.4556 6.59763 3.85604 0.00005 0.00012399 CG15599 0.1339 1.93856 3.85597 0.00005 0.00012399 CG44004 1.4132 20.1721 3.83533 0.00005 0.00012399 CG9975 0.1020 1.41927 3.7991 0.01675 0.0296169 CG14357 0.2106 2.90667 3.78706 0.0018 0.00373183 Dh44-R2 1.6892 22.5989 3.74185 0.00005 0.00012399 CG3280 0.0784 1.03943 3.72813 0.00005 0.00012399 Yp3 88.2353 1159.39 3.71587 0.00005 0.00012399 CG18131 0.3465 4.43418 3.67762 0.00005 0.00012399 CG1998 1.4147 17.8492 3.65729 0.00005 0.00012399 CG7888 0.8036 10.1279 3.65572 0.0001 0.0002417 CG9813 0.3407 4.27133 3.64828 0.00005 0.00012399 blanks 1.9254 23.9872 3.63906 0.00005 0.00012399 skpC 0.5494 6.75865 3.62077 0.0001 0.0002417 HP1D3csd 0.6364 7.65972 3.58935 0.00005 0.00012399 CG5853 8.4914 102.052 3.58716 0.00005 0.00012399
140 CG42613 0.2391 2.77336 3.53619 0.00005 0.00012399 CG15484 1.6557 18.8472 3.5088 0.0049 0.00949916 CG34353 0.3352 3.81468 3.50834 0.00005 0.00012399 CG32719 0.9272 10.4792 3.49849 0.00005 0.00012399 snoRNA:U14: 979590.0000 10800000 3.46883 0.00005 0.00012399 30Ea Rpt3R 1.4856 16.3747 3.46239 0.00005 0.00012399 CG15930 2.8648 31.5299 3.46021 0.00005 0.00012399 rdgC 0.2012 2.18439 3.44078 0.00005 0.00012399 CG11052 0.1261 1.35789 3.42879 0.0113 0.0206192 CG14834 2.5418 27.2507 3.42235 0.00005 0.00012399 Rpt4R 0.1432 1.51226 3.4003 0.0024 0.00488204 CG18568 0.2265 2.31685 3.35437 0.0003 0.00068977 CG3251 0.1734 1.76757 3.3497 0.00085 0.00184174 para 0.1097 1.08733 3.30877 0.00005 0.00012399 CG4221 0.3429 3.38225 3.30228 0.00005 0.00012399 CG31279 2.2260 21.8373 3.29427 0.00005 0.00012399 CG10407 4.0802 39.8252 3.28697 0.00005 0.00012399 CG8870 0.6038 5.89085 3.28631 0.002 0.00411952 Ptp52F 0.9967 9.70191 3.28307 0.00005 0.00012399 CG13998 1.1428 11.0959 3.27941 0.00175 0.00363453 CG42300 0.2488 2.4075 3.27475 0.0094 0.017385 Tbh 0.1709 1.65142 3.27261 0.00005 0.00012399 CG1941 1.1702 11.2877 3.26991 0.00005 0.00012399 gk 0.9523 9.09913 3.25622 0.00005 0.00012399 CG34244 0.2437 2.2779 3.22471 0.0151 0.0269434 fw 0.3758 3.49829 3.21877 0.00005 0.00012399 Hs3st-A 0.1963 1.82524 3.21731 0.00005 0.00012399 Mst36Fa 0.1243 1.15279 3.21355 0.00185 0.00382769 Dip-B 59.0735 546.396 3.20936 0.00005 0.00012399 FucTA 0.4316 3.98944 3.20838 0.02235 0.038546 m 0.9065 8.36329 3.20568 0.00005 0.00012399 CG15927 0.1332 1.2284 3.2055 0.02685 0.045623 CG8964 1.2323 11.3464 3.20278 0.00005 0.00012399 CG16995 0.4546 4.1804 3.20093 0.02275 0.0391592 CG4393 0.6719 6.13338 3.19026 0.00005 0.00012399 tyn 1.4309 12.904 3.17278 0.00005 0.00012399 CG12607 0.1957 1.761 3.16976 0.0118 0.021439 karr 0.4155 3.73534 3.16826 0.0076 0.0142745 bwa 4.9903 44.8528 3.16801 0.00005 0.00012399 CG16700 6.0576 54.2782 3.16355 0.00005 0.00012399 CG42299 0.2961 2.618 3.14436 0.0046 0.00895241 CG11263 4.1969 36.4974 3.12039 0.00005 0.00012399 CR44560 0.1756 1.51952 3.11313 0.025 0.0427004 CR45667 5.3115 45.5546 3.10041 0.0001 0.0002417 CG32091 4.3855 37.3287 3.08948 0.00005 0.00012399 PPO1 0.1914 1.61563 3.07754 0.0004 0.00090401 CR45665 2.7011 22.6832 3.07003 0.0006 0.00132671
141 CG34234 1.8951 15.8813 3.06699 0.0173 0.0304939 Ef1alpha100 1.0058 8.41875 3.06526 0.00005 0.00012399 E CG30345 11.1506 92.098 3.04604 0.00005 0.00012399 CG7720 0.8367 6.90124 3.04408 0.00005 0.00012399 CG10126 2.6890 22.1185 3.04012 0.00005 0.00012399 RpS19b 5.6012 46.0295 3.03876 0.00005 0.00012399 Esp 4.1402 33.9444 3.03542 0.00005 0.00012399 CR45564 0.3808 3.05738 3.00536 0.0131 0.0236251 CG31477 1.7085 13.6631 2.99945 0.00005 0.00012399 CG9766 1.3765 11.0053 2.99911 0.00005 0.00012399 CG14346 0.4029 3.21206 2.99488 0.00015 0.00035649 Sox14 2.0323 16.1719 2.99228 0.00005 0.00012399 skl 0.1607 1.26493 2.97636 0.00595 0.0113708 CG11756 0.2067 1.61185 2.9628 0.0292 0.0492707 Patsas 0.3490 2.71314 2.95851 0.00295 0.00592427 dysc 0.2346 1.80273 2.94173 0.00005 0.00012399 CG11550 0.2252 1.72476 2.937 0.0052 0.010036 PGRP-LB 1.3722 10.4063 2.92295 0.00005 0.00012399 CG5835 1.3489 10.1767 2.91537 0.00005 0.00012399 Nha1 0.3771 2.83505 2.91031 0.00005 0.00012399 CG3348 2.3451 17.409 2.89211 0.024 0.0411305 CG15673 1.2907 9.56914 2.89025 0.00005 0.00012399 mthl8 0.6543 4.80793 2.87735 0.00005 0.00012399 CG15347 13.3963 98.3413 2.87597 0.00005 0.00012399 CG4842 3.2823 23.9243 2.8657 0.00005 0.00012399 CG13003 0.3005 2.16015 2.84588 0.00005 0.00012399 CG43085 0.2148 1.54363 2.84514 0.0028 0.00564312 CG13737 0.3300 2.3639 2.84066 0.02565 0.0437228 CR45666 3.9337 28.0304 2.83303 0.00005 0.00012399 Tengl1 0.1900 1.33954 2.81787 0.0063 0.0119883 CG30401 0.7598 5.35369 2.81693 0.00005 0.00012399 Drat 13.9089 97.8905 2.81516 0.00005 0.00012399 CR44107 1.2515 8.80173 2.81416 0.01215 0.0220295 brv3 0.2020 1.3906 2.78364 0.02205 0.0380713 Ilp6 2.9162 20.0033 2.77806 0.00005 0.00012399 Osi24 0.1637 1.11822 2.77219 0.0002 0.00046869 stnA 1.0993 7.49011 2.76837 0.00005 0.00012399 CG4666 3.6064 24.4557 2.76155 0.00005 0.00012399 CG9850 0.7185 4.83215 2.74957 0.00005 0.00012399 CR45174 0.4109 2.73043 2.73225 0.00005 0.00012399 CR42767 0.4342 2.87228 2.72577 0.00075 0.00163799 yellow-k 1.8306 12.0781 2.72202 0.00005 0.00012399 bgcn 1.0365 6.82788 2.7197 0.00005 0.00012399 Rbp4 0.1899 1.24763 2.71558 0.00475 0.00922669 scb 5.3581 35.0034 2.7077 0.00005 0.00012399 Ace 0.5084 3.31894 2.70666 0.00005 0.00012399 GstE5 0.8280 5.39454 2.70386 0.0001 0.0002417
142 Cralbp 0.8666 5.60937 2.69435 0.00035 0.0007973 CG8526 0.6777 4.31688 2.67129 0.00005 0.00012399 CG7730 3.3777 21.4316 2.66563 0.00005 0.00012399 CG44532 2.3777 15.0377 2.66096 0.00005 0.00012399 CG10151 1.2455 7.86839 2.65938 0.00005 0.00012399 CG13982 0.3248 2.03479 2.64717 0.009 0.0166935 snRNA:U4:38 472.2040 2954.93 2.64564 0.00005 0.00012399 AB Atet 5.1506 31.9471 2.63287 0.00005 0.00012399 rpr 0.7385 4.575 2.63117 0.00005 0.00012399 CG43278 1.0399 6.32755 2.60522 0.00005 0.00012399 Ac3 0.6747 4.08319 2.59748 0.00005 0.00012399 CG42673 4.4771 26.989 2.59174 0.00005 0.00012399 Porin2 0.2446 1.46646 2.58407 0.00175 0.00363453 Idgf2 11.0376 65.6335 2.57201 0.00005 0.00012399 Cpr72Ec 0.1825 1.08422 2.57107 0.00075 0.00163799 knk 0.2698 1.60292 2.57091 0.00005 0.00012399 CREG 9.7919 58.0153 2.56677 0.00005 0.00012399 CG33474 1.6021 9.45881 2.56169 0.00045 0.00101056 CG4587 0.2252 1.32816 2.56007 0.00005 0.00012399 CR43626 0.9817 5.78716 2.55952 0.00005 0.00012399 Vm26Ab 11.6565 68.6271 2.55764 0.00005 0.00012399 CG8335 0.5123 3.01293 2.55604 0.00035 0.0007973 CG10730 0.4461 2.6099 2.54844 0.02575 0.0438829 CG1942 0.6155 3.57458 2.53803 0.00005 0.00012399 CG9314 0.7247 4.19209 2.53225 0.00955 0.0176412 CG32436 0.2072 1.18899 2.52036 0.00015 0.00035649 CG7900 2.1872 12.5313 2.51838 0.00005 0.00012399 CG8602 37.1969 212.28 2.51271 0.00005 0.00012399 CG42260 0.4563 2.59698 2.50865 0.00005 0.00012399 CG11378 6.7295 37.9549 2.49572 0.00005 0.00012399 Cad96Ca 0.3605 2.03246 2.49522 0.00005 0.00012399 CG15760 0.8888 4.98224 2.48686 0.00005 0.00012399 Hr46 0.4276 2.37923 2.47605 0.00005 0.00012399 GstD10 2.5003 13.8751 2.47232 0.00005 0.00012399 roX2 0.4417 2.44947 2.47121 0.00895 0.0166093 CR42874 0.3283 1.80904 2.46225 0.00135 0.00285049 Hr4 0.5749 3.16657 2.46164 0.00005 0.00012399 CG45087 16.0225 87.5779 2.45047 0.00005 0.00012399 Ntf-2r 0.4927 2.67449 2.44061 0.00145 0.00304774 CG14227 0.8242 4.44343 2.43068 0.00005 0.00012399 Mct1 5.8248 31.3805 2.42958 0.00005 0.00012399 tal-AA 3.3486 17.9753 2.42441 0.00005 0.00012399 Rcd-1r 1.2581 6.68296 2.40927 0.00005 0.00012399 CG32462 0.4190 2.187 2.38404 0.0024 0.00488204 CG33970 1.2919 6.71298 2.37751 0.00005 0.00012399 CG11966 0.6900 3.58147 2.37587 0.00005 0.00012399 CG13117 4.5676 23.6402 2.37174 0.00005 0.00012399
143 CG1732 0.2285 1.18009 2.36862 0.0002 0.00046869 CG30197 14.5903 75.3191 2.36801 0.00005 0.00012399 CR44923 0.2804 1.43508 2.35551 0.0127 0.0229507 CG18628 11.7242 59.6098 2.34606 0.00005 0.00012399 CG44250 2.2261 11.3011 2.34387 0.00005 0.00012399 CG14870 0.2820 1.42442 2.33655 0.008 0.0149709 Cyp4p2 0.5503 2.77842 2.33588 0.00065 0.00143132 CG34355 1.1592 5.78638 2.3195 0.00195 0.00402261 GstE8 7.4705 37.2022 2.3161 0.00005 0.00012399 CG18577 0.6254 3.09519 2.30722 0.00145 0.00304774 Ady43A 5.4170 26.8025 2.30681 0.00005 0.00012399 Rab3-GEF 0.2813 1.38677 2.3015 0.00005 0.00012399 CG5973 8.4561 41.4449 2.29313 0.00005 0.00012399 PGRP-LA 1.0483 5.10286 2.28329 0.00005 0.00012399 CR43637 0.8439 4.09254 2.27793 0.00575 0.0110178 CG5024 0.2211 1.07206 2.27758 0.02935 0.0495076 eIF4E-6 0.8182 3.92936 2.2637 0.0003 0.00068977 CG4496 2.1589 10.3271 2.25808 0.00005 0.00012399 pon 2.3225 10.9617 2.23875 0.00005 0.00012399 cer 25.0297 117.895 2.23579 0.00005 0.00012399 TM4SF 1.8204 8.42659 2.21071 0.00605 0.0115458 CG4822 3.8541 17.8024 2.20762 0.0149 0.0266089 PH4alphaPV 1.9096 8.7773 2.20051 0.00005 0.00012399 CG12851 1.3369 6.11252 2.19289 0.00005 0.00012399 CG3921 2.4045 10.9555 2.18783 0.00005 0.00012399 NLaz 9.6472 43.9143 2.18652 0.00005 0.00012399 CG32137 1.5813 7.14146 2.17512 0.00005 0.00012399 Sp7 3.2096 14.4676 2.17237 0.00005 0.00012399 mlt 4.2617 19.1497 2.16783 0.00005 0.00012399 Itgbetanu 1.5400 6.83015 2.14899 0.00005 0.00012399 CG7381 2.3456 10.38 2.14578 0.00005 0.00012399 CG15056 0.3116 1.3781 2.14514 0.0026 0.00525882 cln3 6.3085 27.8549 2.14257 0.00005 0.00012399 CG4586 4.5455 19.9216 2.13184 0.00005 0.00012399 crq 33.9683 148.753 2.13066 0.00005 0.00012399 CG11382 2.8344 12.3805 2.12693 0.00005 0.00012399 Myo28B1 0.3311 1.44471 2.12547 0.00005 0.00012399 Mmp2 3.2084 13.9438 2.11971 0.00005 0.00012399 CG3301 0.5056 2.19569 2.11865 0.0017 0.00353841 pim 17.5700 75.8853 2.1107 0.00005 0.00012399 Gadd45 3.5411 15.2539 2.10691 0.00005 0.00012399 CG17075 1.6363 6.99654 2.09622 0.00005 0.00012399 CG11320 0.3573 1.51239 2.0815 0.00195 0.00402261 CG15398 0.4068 1.72019 2.08016 0.02755 0.0467146 bip1 18.6473 78.446 2.07273 0.00005 0.00012399 VhaAC39-2 0.6106 2.56726 2.072 0.0004 0.00090401 CG43999 0.5463 2.28548 2.06484 0.0066 0.0125141 CG31676 0.9392 3.92888 2.06464 0.00005 0.00012399
144 CG12689 0.4905 2.04961 2.06302 0.0058 0.0111056 CG33468 0.6629 2.74714 2.05098 0.0043 0.00840501 CG14131 3.7312 15.4454 2.04945 0.00035 0.0007973 amd 0.4995 2.06663 2.0486 0.00005 0.00012399 Cyp4s3 2.9066 12.0147 2.0474 0.00005 0.00012399 CG13321 9.2773 38.3238 2.04646 0.00005 0.00012399 CG32816 10.9367 45.1565 2.04576 0.0027 0.00545162 CR44966 1.1501 4.74037 2.04325 0.00015 0.00035649 beat-Ib 0.5138 2.11773 2.04322 0.00005 0.00012399 CG4839 0.4774 1.92995 2.01517 0.00005 0.00012399 spartin 2.6888 10.8486 2.01248 0.00005 0.00012399 CG4676 0.6573 2.65103 2.01192 0.00075 0.00163799 Lapsyn 6.2196 25.0814 2.01172 0.00005 0.00012399 CG13384 22.4367 89.842 2.00153 0.00005 0.00012399 mei-P22 1.4874 5.94163 1.99803 0.00005 0.00012399 Dip3 1.3975 5.57762 1.99682 0.00005 0.00012399 nrv3 2.1102 8.39907 1.99286 0.00005 0.00012399 CR45618 0.5205 2.07053 1.99209 0.01935 0.0337956 CrebA 4.0835 16.2313 1.99092 0.00005 0.00012399 CG43066 0.8474 3.36275 1.98855 0.00005 0.00012399 wbl 4.7292 18.7535 1.98749 0.00005 0.00012399 Hsp23 48.1189 189.772 1.97959 0.00005 0.00012399 CG8097 14.2018 55.9418 1.97785 0.00005 0.00012399 CG6231 8.4981 33.3425 1.97216 0.00005 0.00012399 CG14837 0.2896 1.12951 1.96351 0.00005 0.00012399 CG34398 0.2899 1.13003 1.96291 0.00005 0.00012399 CG1265 22.2823 86.664 1.95953 0.00005 0.00012399 CG7386 1.5712 6.09637 1.95613 0.00005 0.00012399 ImpL2 5.0994 19.7788 1.95555 0.00005 0.00012399 Syn2 0.3967 1.533 1.95023 0.00005 0.00012399 CR43727 2.2154 8.5422 1.94703 0.00005 0.00012399 Irc 12.5827 48.2339 1.9386 0.00005 0.00012399 Zip99C 26.1130 99.8674 1.93524 0.01265 0.022866 CR43257 0.8742 3.34285 1.93506 0.0093 0.0172137 pip 2.6546 10.0457 1.92002 0.00005 0.00012399 bora 3.7348 14.0992 1.91653 0.00005 0.00012399 cpx 1.8079 6.80151 1.91158 0.00005 0.00012399 Drsl5 0.9412 3.53349 1.90854 0.01995 0.0347587 Ugt36Ba 0.4963 1.8525 1.90008 0.00005 0.00012399 CR30267 0.4289 1.5968 1.89654 0.0097 0.0179046 cib 153.6190 571.164 1.89455 0.00005 0.00012399 Nlg2 0.6711 2.48415 1.88814 0.00005 0.00012399 CG4415 1.8764 6.94249 1.88746 0.00005 0.00012399 ebd2 11.0498 40.797 1.88444 0.00005 0.00012399 CG9503 5.0125 18.5036 1.88421 0.0136 0.0244572 CG7675 16.0301 58.7391 1.87353 0.00005 0.00012399 sff 0.7540 2.7622 1.87323 0.00005 0.00012399 CR45108 0.3254 1.19136 1.87242 0.0008 0.00174155
145 Dtg 0.3099 1.12929 1.86567 0.00005 0.00012399 CG5281 3.8453 13.969 1.86105 0.00005 0.00012399 sev 1.7911 6.49898 1.85937 0.00005 0.00012399 Rad51D 2.0148 7.29648 1.85659 0.00005 0.00012399 CG11369 0.2966 1.07023 1.85126 0.00015 0.00035649 Kal1 6.8507 24.6323 1.84622 0.00005 0.00012399 CG34031 0.8996 3.23042 1.84433 0.00525 0.0101252 CG44098 0.6184 2.21427 1.84014 0.0087 0.0161761 snRNA:U4:25 24.6102 87.9238 1.837 0.0052 0.010036 F tun 8.0316 28.6912 1.83685 0.00005 0.00012399 CR44784 1.0381 3.69406 1.83132 0.0019 0.00392473 CG4670 20.5553 73.0597 1.82956 0.00005 0.00012399 Lsd-2 50.5742 179.656 1.82877 0.00005 0.00012399 CG11409 1.4686 5.21092 1.82709 0.00005 0.00012399 CG18343 11.3394 40.0835 1.82166 0.00005 0.00012399 GstD3 29.9001 105.47 1.81861 0.00005 0.00012399 CG31475 0.3981 1.4016 1.81581 0.00005 0.00012399 CG4570 14.4548 50.8009 1.8133 0.00005 0.00012399 ndl 11.6334 40.8074 1.81056 0.00005 0.00012399 CG12290 3.5672 12.4981 1.80886 0.00005 0.00012399 CG33169 7.6997 26.9728 1.80862 0.00005 0.00012399 Tie 0.8999 3.15041 1.80763 0.00005 0.00012399 CG12224 0.4354 1.52329 1.80673 0.00215 0.00440387 sim 5.9093 20.583 1.80041 0.00005 0.00012399 CG11241 5.2342 18.2066 1.79841 0.00005 0.00012399 Cad88C 0.7697 2.6692 1.79408 0.00005 0.00012399 CG31869 7.4491 25.6936 1.78627 0.00005 0.00012399 Lsd-1 10.9056 37.489 1.7814 0.00005 0.00012399 CG13540 2.1443 7.36243 1.77971 0.0001 0.0002417 CG4998 0.3203 1.09618 1.77486 0.00005 0.00012399 Arpc3B 2.9360 9.99104 1.76678 0.00005 0.00012399 CG5111 0.4288 1.45889 1.76639 0.00015 0.00035649 CG43295 1.8497 6.28977 1.7657 0.00005 0.00012399 CG8420 3.7435 12.7186 1.76448 0.00005 0.00012399 CG18508 7.2874 24.7364 1.76317 0.00005 0.00012399 CG1124 0.5632 1.91119 1.76272 0.00155 0.00324556 CG5245 1.9307 6.49134 1.74938 0.00005 0.00012399 CG2709 2.2048 7.40064 1.74703 0.00005 0.00012399 CG3376 3.4095 11.433 1.74557 0.00005 0.00012399 Lerp 9.0733 30.377 1.74329 0.00005 0.00012399 CG42346 0.6262 2.09307 1.74102 0.0055 0.0105729 CG32687 7.8404 26.2051 1.74086 0.00005 0.00012399 CG30285 2.9408 9.80993 1.73806 0.00005 0.00012399 ocm 15.0692 50.1518 1.7347 0.00005 0.00012399 zfh2 0.9679 3.21745 1.73295 0.0026 0.00525882 GstD9 21.8973 72.7237 1.73168 0.00005 0.00012399 CG15394 3.0484 10.1126 1.73 0.00005 0.00012399
146 CG5888 1.2804 4.23729 1.72658 0.00075 0.00163799 Sans 3.2045 10.6045 1.72649 0.00005 0.00012399 CG10013 1.6578 5.47678 1.72407 0.00005 0.00012399 CG32720 0.7956 2.61907 1.71894 0.0056 0.0107466 CG13707 0.4534 1.4888 1.71524 0.00375 0.00739578 CG34253 1.5048 4.92865 1.7116 0.00565 0.0108377 dpr18 0.7868 2.55138 1.6972 0.00005 0.00012399 CG32165 6.8116 22.0155 1.69245 0.00005 0.00012399 CG44251 0.6005 1.93728 1.68986 0.0004 0.00090401 Hexo2 3.4148 11.0157 1.6897 0.0001 0.0002417 dah 7.0752 22.7737 1.68654 0.00005 0.00012399 CG12075 1.0323 3.31016 1.68102 0.00005 0.00012399 Spn77Bc 0.6110 1.95808 1.6801 0.00015 0.00035649 CG13893 3.7339 11.9123 1.6737 0.00005 0.00012399 pie 5.4720 17.4285 1.6713 0.00005 0.00012399 Ppt2 8.1702 25.9995 1.67004 0.00005 0.00012399 Jabba 72.6072 230.569 1.66701 0.00005 0.00012399 CG6330 14.5506 46.0712 1.66279 0.00005 0.00012399 ZnT41F 8.3306 26.3574 1.66172 0.00005 0.00012399 CG14624 0.4737 1.49741 1.66044 0.0162 0.0287507 CG1637 31.4814 99.3838 1.65851 0.00005 0.00012399 GstZ2 0.4530 1.42567 1.65412 0.00425 0.00831313 CG5191 7.1873 22.5592 1.6502 0.00005 0.00012399 CR45681 0.8126 2.5375 1.64283 0.0039 0.00766901 CG8369 9.6239 30.0268 1.64156 0.00005 0.00012399 nos 67.4310 210.369 1.64144 0.00005 0.00012399 CR45195 1.7844 5.551 1.63732 0.0004 0.00090401 mfas 10.4943 32.637 1.6369 0.00005 0.00012399 CG7408 2.0928 6.50697 1.63657 0.00005 0.00012399 CG5375 0.3784 1.17375 1.63318 0.0003 0.00068977 CG11658 3.0056 9.31654 1.63214 0.00005 0.00012399 sage 0.5167 1.60132 1.63194 0.0014 0.00294985 snRNA:U1:95 710.2360 2198.85 1.63038 0.0008 0.00174155 Cc CG5144 0.7588 2.34838 1.62985 0.0001 0.0002417 dgo 1.1560 3.56225 1.62364 0.00005 0.00012399 Prosbeta2R2 0.3476 1.0708 1.62334 0.01135 0.0206938 CG9336 27.1903 83.7458 1.62292 0.00005 0.00012399 bgm 22.2576 68.3516 1.61867 0.00005 0.00012399 CG31100 0.5162 1.58242 1.61603 0.0001 0.0002417 CG33062 1.8976 5.77733 1.60626 0.00005 0.00012399 CG17658 2.6490 8.02689 1.5994 0.00035 0.0007973 Appl 0.3713 1.12438 1.59845 0.00005 0.00012399 CG14102 3.5414 10.7163 1.59741 0.022 0.0379977 CR44992 1.2081 3.65298 1.59632 0.0274 0.0464788 qjt 1.0963 3.2828 1.58234 0.00085 0.00184174 MFS16 29.8916 89.3325 1.57945 0.00005 0.00012399 CR45444 2.2936 6.84769 1.57802 0.01265 0.022866
147 CR44057 1.0090 3.00843 1.57614 0.02825 0.0478129 Pu 3.6958 11.0179 1.57589 0.00005 0.00012399 CG2065 0.4064 1.21144 1.5758 0.00545 0.010486 CG17650 1.9794 5.89454 1.57432 0.0006 0.00132671 Cp1 217.9500 645.86 1.56722 0.00005 0.00012399 CR42875 0.7827 2.30078 1.55555 0.0062 0.0118117 CG32506 0.7114 2.08891 1.55404 0.00005 0.00012399 Fer2LCH 143.6910 421.185 1.55149 0.00005 0.00012399 Cerk 15.5773 45.6163 1.5501 0.00005 0.00012399 lat 7.2782 21.3062 1.54963 0.00005 0.00012399 CG43120 0.3790 1.10789 1.5476 0.0177 0.0311397 CG6405 0.4073 1.18909 1.54565 0.0025 0.00507274 Cht2 6.2898 18.3609 1.54555 0.00005 0.00012399 cbt 17.8737 51.9058 1.53806 0.00005 0.00012399 CG8316 6.8831 19.9531 1.53548 0.00005 0.00012399 Syx4 3.3753 9.78118 1.53498 0.0083 0.0154914 IntS12 6.1713 17.8747 1.53427 0.00005 0.00012399 RhoGEF64C 5.0388 14.5625 1.5311 0.00005 0.00012399 CG4297 2.8767 8.26785 1.52311 0.00005 0.00012399 CG12159 20.0018 57.4392 1.5219 0.00005 0.00012399 W 5.8935 16.8695 1.51723 0.00005 0.00012399 DnaJ-H 14.2638 40.8251 1.51709 0.00005 0.00012399 CG9003 9.3863 26.7748 1.51225 0.00005 0.00012399 Rab7 57.7241 164.249 1.50864 0.00005 0.00012399 CG7352 1.9221 5.46299 1.50699 0.00005 0.00012399 CG32373 10.0091 28.4192 1.50555 0.01435 0.025694 Reg-5 5.7993 16.451 1.50423 0.00005 0.00012399 CG3662 55.0578 154.497 1.48856 0.00005 0.00012399 CG4267 9.8126 27.4581 1.48452 0.00005 0.00012399 CG43986 22.3687 62.4596 1.48144 0.00985 0.0181629 RhoGAP100 3.5430 9.8738 1.47865 0.00005 0.00012399 F Nhe3 17.2984 48.1145 1.47583 0.00005 0.00012399 Dmtn 14.8902 41.387 1.47482 0.00005 0.00012399 CG13771 0.4165 1.15741 1.47459 0.0067 0.0126907 snRNA:U2:14 68.2166 189.392 1.47318 0.00005 0.00012399 B CR45481 0.6709 1.85793 1.46942 0.0003 0.00068977 nmdyn-D7 6.2339 17.244 1.46789 0.00005 0.00012399 CG5337 2.1046 5.81711 1.46675 0.00055 0.00122116 Rcd4 14.1733 39.1555 1.46604 0.00005 0.00012399 CG31176 1.2634 3.46837 1.45696 0.00005 0.00012399 CG7194 17.8915 48.9214 1.45119 0.00005 0.00012399 CG6928 7.1473 19.5395 1.45091 0.00005 0.00012399 CG7997 21.8201 59.6263 1.45029 0.00005 0.00012399 Cyp6w1 0.7037 1.91885 1.44714 0.0003 0.00068977 Kul 0.6148 1.67357 1.44484 0.00005 0.00012399 CR44275 0.8030 2.18544 1.44438 0.00145 0.00304774
148 CG4797 0.9063 2.46167 1.44153 0.0069 0.0130469 mia 4.2361 11.499 1.4407 0.00005 0.00012399 CG32698 0.5007 1.3565 1.43773 0.00115 0.00245271 ord 1.4684 3.96668 1.43367 0.00005 0.00012399 GstD4 1.2172 3.28375 1.43184 0.00295 0.00592427 GlcAT-P 7.8640 21.2078 1.43125 0.00005 0.00012399 srp 3.7562 10.121 1.43002 0.00005 0.00012399 CG31955 3.2615 8.7742 1.42772 0.00005 0.00012399 insv 6.4300 17.2808 1.42628 0.00005 0.00012399 GstT4 7.3360 19.6953 1.42479 0.00005 0.00012399 CG42666 8.9792 24.1057 1.42472 0.00005 0.00012399 Cyp6a13 4.2012 11.2744 1.42419 0.00005 0.00012399 CG17129 4.6554 12.4921 1.42404 0.00005 0.00012399 magu 8.3457 22.3858 1.42348 0.00005 0.00012399 pirk 0.5243 1.40567 1.42288 0.00185 0.00382769 CG5704 3.8612 10.3015 1.41574 0.00005 0.00012399 CG7130 5.5395 14.7605 1.41392 0.00005 0.00012399 GstE7 9.0297 24.0408 1.41273 0.00005 0.00012399 CG10262 0.8859 2.35572 1.41103 0.00185 0.00382769 Reg-2 2.8010 7.4446 1.41027 0.00005 0.00012399 CG17928 2.5429 6.75727 1.41 0.00005 0.00012399 CG4061 10.5892 28.0124 1.40348 0.00005 0.00012399 CG42231 7.4446 19.6851 1.40284 0.0064 0.012162 CG7102 4.4617 11.785 1.40129 0.0242 0.0414369 CG10413 20.3977 53.6026 1.3939 0.00005 0.00012399 cu 8.0676 21.2005 1.39389 0.00005 0.00012399 CG33926 0.7637 2.00629 1.39352 0.0065 0.0123384 Tsp42Ej 1.7702 4.64511 1.39182 0.00005 0.00012399 CG11261 7.1299 18.6692 1.3887 0.00005 0.00012399 Fer1HCH 166.6660 435.873 1.38695 0.00005 0.00012399 Sodh-2 2.5727 6.7193 1.38504 0.00005 0.00012399 Gli 2.4727 6.45091 1.38344 0.00005 0.00012399 cv-d 0.9473 2.46945 1.38233 0.00005 0.00012399 rdgB 9.0791 23.6588 1.38175 0.00005 0.00012399 CG9815 3.6265 9.44248 1.38058 0.00005 0.00012399 CG32039 10.6156 27.6275 1.37993 0.00005 0.00012399 cid 7.5795 19.6912 1.37739 0.00005 0.00012399 CG7565 6.8048 17.6707 1.37673 0.00005 0.00012399 CG6414 2.4322 6.30374 1.37395 0.00005 0.00012399 CR44289 1.5520 4.02001 1.3731 0.00005 0.00012399 CG32483 0.5219 1.35188 1.37305 0.00395 0.00776078 CG31344 8.0680 20.8888 1.37245 0.00005 0.00012399 CG14741 1.4649 3.78818 1.37071 0.00005 0.00012399 CG14669 0.5435 1.40537 1.37055 0.00005 0.00012399 CG9796 86.1429 222.605 1.36968 0.00005 0.00012399 CG12194 1.5105 3.89665 1.36722 0.00005 0.00012399 KCNQ 1.7761 4.56184 1.36093 0.00065 0.00143132 hb 9.9655 25.5952 1.36086 0.00005 0.00012399
149 Trf4-2 12.2090 31.2988 1.35816 0.00005 0.00012399 CG18547 5.3715 13.7606 1.35714 0.00005 0.00012399 bib 2.5734 6.58478 1.35545 0.00005 0.00012399 CG42342 1.6944 4.32692 1.35259 0.00005 0.00012399 CG2694 7.7405 19.7643 1.3524 0.00005 0.00012399 CG42709 12.8894 32.8722 1.35068 0.00005 0.00012399 Nnf1b 6.7652 17.2013 1.34632 0.00005 0.00012399 CG6785 0.7841 1.99355 1.34625 0.00005 0.00012399 DOR 10.4580 26.5784 1.34564 0.00005 0.00012399 CG3323 0.5924 1.50322 1.34354 0.0005 0.00111605 CG1399 10.9168 27.6693 1.34173 0.00005 0.00012399 Hydr2 29.6071 74.6216 1.33365 0.00005 0.00012399 CG14906 8.2425 20.768 1.33322 0.00005 0.00012399 CG1103 28.6788 71.9662 1.32733 0.00005 0.00012399 CG31202 0.5996 1.50235 1.32516 0.00365 0.00721195 stil 16.7399 41.9307 1.32472 0.00005 0.00012399 CG34112 5.9936 14.9854 1.32208 0.0003 0.00068977 CG10178 7.5815 18.9125 1.31879 0.00005 0.00012399 CG8245 0.7069 1.76154 1.31722 0.0053 0.0102143 pinta 5.5588 13.8421 1.31622 0.00005 0.00012399 CG42694 1.5547 3.87001 1.31574 0.00005 0.00012399 CG18858,CG 3.1888 7.93431 1.31511 0.00005 0.00012399 31683 CG9902 3.4190 8.47711 1.31001 0.00005 0.00012399 hiw 0.7773 1.92243 1.30644 0.00345 0.00684617 Mes2 13.2694 32.8149 1.30624 0.00005 0.00012399 Paip2 31.9258 78.8947 1.30521 0.00005 0.00012399 inaF-D 4.9581 12.2419 1.30397 0.00005 0.00012399 CG12795 13.0413 32.169 1.30258 0.00005 0.00012399 CG3386 4.6607 11.4948 1.30237 0.00005 0.00012399 CG12112 19.0348 46.9451 1.30234 0.00005 0.00012399 Hsp27 93.0147 229.383 1.30223 0.00005 0.00012399 CG13197 10.7732 26.5651 1.30208 0.00005 0.00012399 twe 13.6685 33.6815 1.3011 0.00005 0.00012399 CG32812 0.6942 1.70889 1.2997 0.0032 0.00638939 Nc 11.8377 29.0317 1.29424 0.00005 0.00012399 CG3356 23.7922 58.3214 1.29354 0.00005 0.00012399 veil 7.0967 17.3817 1.29234 0.00005 0.00012399 CG12782 0.7526 1.84273 1.29181 0.00455 0.00886178 CG31898 5.4037 13.2118 1.28981 0.00005 0.00012399 CG8089 3.3065 8.06988 1.28724 0.00005 0.00012399 IP3K1 28.1132 68.6033 1.28703 0.00005 0.00012399 CG10505 3.0314 7.37747 1.28314 0.0004 0.00090401 nub 0.4371 1.06315 1.28232 0.0002 0.00046869 CG13148 1.5283 3.71229 1.28037 0.00005 0.00012399 CG7840 63.7730 154.781 1.27922 0.00005 0.00012399 shd 0.8043 1.95007 1.27766 0.00005 0.00012399 thr 9.6749 23.3511 1.27118 0.00005 0.00012399
150 SP1029 8.0361 19.3654 1.26892 0.00005 0.00012399 Qtzl 3.7876 9.12581 1.26865 0.00005 0.00012399 CycB3 17.3741 41.8132 1.26702 0.00005 0.00012399 spz 6.9999 16.8195 1.26473 0.00005 0.00012399 CG31370 2.6163 6.28318 1.26396 0.00005 0.00012399 Cht3 0.8548 2.05005 1.26194 0.00005 0.00012399 CR43609 1.0562 2.53033 1.26041 0.00005 0.00012399 CG11125 4.3746 10.4566 1.2572 0.00005 0.00012399 CG6808 3.6153 8.64014 1.25693 0.00105 0.00225012 Acox57D-p 6.8546 16.3197 1.25147 0.00005 0.00012399 CG18530 0.4300 1.02332 1.2507 0.02765 0.0468699 CR43967 6.6358 15.7842 1.25013 0.02405 0.0412074 CG5707 4.8020 11.4147 1.24918 0.00005 0.00012399 CR44524 2.2442 5.33202 1.24849 0.01475 0.0263592 CG14946 0.5412 1.28456 1.24703 0.0121 0.0219471 exu 137.6600 325.599 1.24198 0.00005 0.00012399 nAChRalpha 4.4743 10.5702 1.24027 0.00005 0.00012399 4 CR42862 95.7075 225.496 1.2364 0.00005 0.00012399 CG32138 15.6292 36.7993 1.23544 0.00005 0.00012399 Oatp74D 1.3386 3.14778 1.23363 0.00005 0.00012399 Pka-C3 2.2800 5.35937 1.23305 0.00005 0.00012399 CG17119 3.1280 7.33386 1.22935 0.00105 0.00225012 l(3)72Dp 4.1175 9.63978 1.22723 0.00005 0.00012399 Exn 5.5104 12.8978 1.22689 0.00005 0.00012399 CG17361 4.1768 9.7734 1.22646 0.0007 0.00153459 fog 20.1966 47.2291 1.22556 0.00005 0.00012399 CG5342 0.4382 1.02284 1.22301 0.00355 0.00702743 CG4991 73.1309 170.647 1.22246 0.00005 0.00012399 Cyp4d8 2.3288 5.42869 1.22105 0.00005 0.00012399 mthl14 3.4077 7.93207 1.2189 0.02815 0.0476559 CG10559 0.9946 2.314 1.21821 0.00125 0.00265143 Rhp 1.7409 4.04629 1.21673 0.00005 0.00012399 CG5664 6.6999 15.5587 1.21551 0.00125 0.00265143 CG18605 0.6494 1.50566 1.21333 0.0048 0.0093163 CG3077 8.7917 20.3735 1.21247 0.00005 0.00012399 CG17977 4.1011 9.49477 1.21111 0.00005 0.00012399 CG9203 4.6680 10.8017 1.21037 0.00005 0.00012399 CenB1A 22.6168 52.2112 1.20696 0.00005 0.00012399 be 10.0402 23.1646 1.20614 0.00005 0.00012399 CG14352 5.2177 12.0158 1.20344 0.0221 0.0381504 CG44774 35.8017 82.4253 1.20306 0.00005 0.00012399 CG6762 4.8698 11.1856 1.19972 0.0056 0.0107466 fab1 2.6324 6.04314 1.1989 0.00005 0.00012399 CG17636 0.6632 1.51982 1.19648 0.00095 0.00204645 RabX1 6.4133 14.6901 1.19571 0.00005 0.00012399 retm 7.8209 17.8875 1.19355 0.00005 0.00012399 M6 5.3787 12.2999 1.19333 0.0001 0.0002417
151 CR45472 2.3040 5.26442 1.19213 0.00235 0.00478707 Tsp42Ei 1.6969 3.87447 1.19108 0.00105 0.00225012 CG11880 28.1164 64.1922 1.19099 0.00005 0.00012399 Cyp18a1 8.4688 19.3167 1.18962 0.00005 0.00012399 CG14965 4.3720 9.94792 1.18611 0.00005 0.00012399 mad2 20.6438 46.9487 1.18538 0.00005 0.00012399 CG43689 1.2556 2.85515 1.18522 0.00005 0.00012399 MESR3 12.1093 27.472 1.18184 0.00005 0.00012399 Sox102F 2.7663 6.27271 1.18116 0.00005 0.00012399 CG45068 4.4589 10.1052 1.18033 0.00005 0.00012399 CG31619 0.5317 1.20382 1.17903 0.00005 0.00012399 chinmo 1.2058 2.72873 1.17822 0.00005 0.00012399 CR43144 9.5275 21.5428 1.17703 0.00005 0.00012399 Cad74A 2.3822 5.38241 1.17594 0.00005 0.00012399 CG16947 1.8101 4.0765 1.17129 0.00005 0.00012399 CG32364 2.2760 5.12481 1.17102 0.00055 0.00122116 QC 22.0744 49.7042 1.17099 0.0033 0.00657108 Elba2 13.7862 31.0375 1.17079 0.00005 0.00012399 Prx2540-2 3.6324 8.13434 1.16312 0.00005 0.00012399 Nek2 6.1061 13.6547 1.16108 0.00005 0.00012399 CG13366 21.8713 48.7152 1.15533 0.00005 0.00012399 CG31457 8.2891 18.4587 1.15501 0.00005 0.00012399 CG43064 1.0378 2.31068 1.15474 0.0007 0.00153459 Hsp26 278.9540 617.72 1.14692 0.00005 0.00012399 Vmat 4.2388 9.3837 1.14652 0.00005 0.00012399 CG1943 131.7000 291.535 1.14641 0.00005 0.00012399 Glut4EF 5.7068 12.6316 1.14627 0.0049 0.00949916 wake 9.0506 20.0276 1.1459 0.00005 0.00012399 CG2861 0.7120 1.57528 1.14559 0.00005 0.00012399 CG34149 1.9319 4.2663 1.14296 0.0002 0.00046869 neo 1.5110 3.33005 1.14003 0.00005 0.00012399 EndoB 7.6431 16.8227 1.13818 0.00005 0.00012399 CG16953 29.6591 65.2406 1.1373 0.00005 0.00012399 CG3589 6.3966 14.0575 1.13596 0.00005 0.00012399 CG13868 12.8617 28.1811 1.13165 0.00005 0.00012399 CG8303 27.1321 59.3652 1.12962 0.00005 0.00012399 Ance 5.7075 12.4833 1.12906 0.00005 0.00012399 Ten-a 2.3478 5.12713 1.12683 0.00005 0.00012399 CG6770 399.5530 872.465 1.12671 0.00005 0.00012399 CR45596 5.2721 11.5024 1.1255 0.00255 0.00516564 CG2938 19.0360 41.5048 1.12455 0.00005 0.00012399 CG11210 17.9937 39.2216 1.12416 0.0001 0.0002417 tef 14.0203 30.4815 1.12041 0.00005 0.00012399 CG12061 0.6599 1.43309 1.11874 0.01675 0.0296169 l(3)mbt 10.1769 22.0925 1.11826 0.00005 0.00012399 CG18749 1.8346 3.97711 1.11624 0.0079 0.014793 l(2)41Ab 3.0563 6.62427 1.11598 0.00005 0.00012399 CG10420 6.2471 13.5256 1.11443 0.00005 0.00012399
152 CG13907 27.8342 60.2433 1.11394 0.0001 0.0002417 CG11593 12.5836 27.2217 1.11322 0.00005 0.00012399 CG6967 13.6891 29.5836 1.11177 0.00005 0.00012399 CG14868 0.9338 2.01749 1.11142 0.0003 0.00068977 CG12179 2.9269 6.32124 1.11086 0.00005 0.00012399 gol 1.3793 2.97455 1.10874 0.00005 0.00012399 CG17839 1.8679 4.01661 1.10454 0.00005 0.00012399 Naa60 7.8930 16.9004 1.09841 0.00005 0.00012399 Nep2 5.3023 11.3418 1.09697 0.00005 0.00012399 Sse 6.1011 13.0495 1.09686 0.00005 0.00012399 Inx3 28.5590 61.0673 1.09646 0.00005 0.00012399 CG6425 4.7497 10.1547 1.09625 0.00005 0.00012399 Tre1 18.1541 38.7909 1.09542 0.00005 0.00012399 Syx17 5.7140 12.1957 1.0938 0.01925 0.0336315 CG18870 17.1859 36.6597 1.09297 0.00005 0.00012399 CR45397 6.3935 13.6114 1.09015 0.0196 0.0341893 CG11353 1.2083 2.56585 1.08644 0.00005 0.00012399 Tl 32.9470 69.8609 1.08434 0.00005 0.00012399 CR45036 1.8890 4.00218 1.08317 0.00645 0.0122512 Csat 15.3833 32.5811 1.08267 0.00005 0.00012399 wun 13.4558 28.4679 1.0811 0.00005 0.00012399 Sap-r 216.1630 457.144 1.08052 0.00005 0.00012399 CG15385 0.6927 1.46325 1.07891 0.0009 0.00194442 Oamb 1.0557 2.2283 1.07775 0.00005 0.00012399 CG13510 3.8947 8.21662 1.07704 0.02045 0.0355526 plx 4.4973 9.48491 1.07656 0.00005 0.00012399 CG30046 1.7939 3.78299 1.07642 0.00005 0.00012399 Hnf4 16.7759 35.3681 1.07606 0.00005 0.00012399 Prestin 13.1745 27.7516 1.07482 0.00005 0.00012399 mars 6.4902 13.6701 1.0747 0.00015 0.00035649 CG5381 7.8245 16.4356 1.07075 0.0224 0.0386239 mgl 13.2649 27.8354 1.06931 0.00005 0.00012399 CG17834 8.4774 17.7695 1.0677 0.00005 0.00012399 slow 1.8719 3.91909 1.06603 0.00005 0.00012399 CG3394 6.5144 13.6376 1.06588 0.00005 0.00012399 hdm 0.8932 1.86874 1.06499 0.00595 0.0113708 bif 34.1173 71.3471 1.06435 0.00005 0.00012399 Tgi 11.2942 23.6142 1.06408 0.00005 0.00012399 CG5466 3.2411 6.76154 1.06087 0.00005 0.00012399 ana3 11.7926 24.5896 1.06017 0.00005 0.00012399 CG8180 35.0250 72.9056 1.05764 0.00005 0.00012399 eap 13.6313 28.3704 1.05746 0.00515 0.00994804 RhoGAP54D 4.5814 9.53071 1.05678 0.00005 0.00012399 Ugt35a 4.8218 10.0241 1.05582 0.00005 0.00012399 CG10660 1.6171 3.35794 1.05419 0.0001 0.0002417 CG30148 0.9670 2.00795 1.05415 0.02585 0.0440429 CG12818 5.1780 10.7338 1.05169 0.00005 0.00012399 CG5326 11.4396 23.6854 1.04996 0.00005 0.00012399
153 CG12253 7.9301 16.418 1.04987 0.00005 0.00012399 CR43651 15.7800 32.6397 1.04854 0.00005 0.00012399 CG4476 23.4717 48.4858 1.04664 0.00005 0.00012399 CG5399 12.7202 26.2548 1.04546 0.00005 0.00012399 GstE11 6.9489 14.3227 1.04345 0.00005 0.00012399 CG8239 7.7155 15.8896 1.04226 0.01025 0.0188423 aph-1 17.4632 35.948 1.04159 0.00005 0.00012399 CG9601 14.6464 30.1459 1.04142 0.00005 0.00012399 CG4080 15.1617 31.1998 1.0411 0.00005 0.00012399 fz 17.7790 36.574 1.04065 0.00005 0.00012399 brn 6.2764 12.8926 1.03854 0.00005 0.00012399 CG17778 0.9662 1.98149 1.03625 0.0035 0.00693834 CkIIalpha-i3 11.5250 23.6304 1.03588 0.00005 0.00012399 cpo 2.3823 4.87879 1.0342 0.00005 0.00012399 CG9663 3.1518 6.45441 1.03412 0.00005 0.00012399 CG2909 3.1226 6.3851 1.03197 0.0199 0.0346791 bmm 15.5649 31.7928 1.0304 0.00005 0.00012399 Rab30 6.8905 14.0671 1.02964 0.00005 0.00012399 CG7852 8.1078 16.5429 1.02882 0.00005 0.00012399 CG4050 7.9368 16.1733 1.02698 0.00005 0.00012399 CG9577 18.4577 37.6113 1.02695 0.00005 0.00012399 ckd 1.6570 3.37437 1.02605 0.0008 0.00174155 rudhira 16.5721 33.7203 1.02486 0.00005 0.00012399 CG9641 16.8643 34.3013 1.02429 0.00005 0.00012399 crim 18.4185 37.4486 1.02376 0.0091 0.0168656 CG15362 6.4630 13.1366 1.02332 0.00005 0.00012399 CG13531 12.5704 25.5455 1.02304 0.00005 0.00012399 CG11247 13.0803 26.5799 1.02294 0.00005 0.00012399 ush 3.1998 6.5002 1.02251 0.00115 0.00245271 CG13905 5.1441 10.4473 1.02214 0.00005 0.00012399 CR44294 2.4851 5.0416 1.02061 0.026 0.0442807 CG3842 10.1582 20.5881 1.01917 0.00005 0.00012399 Cyp6d4 3.7584 7.61644 1.01901 0.00005 0.00012399 CG7083 9.6542 19.5595 1.01865 0.00005 0.00012399 rad50 14.4524 29.2784 1.01852 0.00005 0.00012399 Cen 17.4902 35.4228 1.01813 0.00005 0.00012399 CG8569 7.3346 14.8522 1.01788 0.00005 0.00012399 Ccm3 33.0100 66.7486 1.01583 0.00005 0.00012399 CG6805 7.6213 15.4084 1.0156 0.00005 0.00012399 plu 11.2255 22.6919 1.0154 0.00005 0.00012399 CG1492 3.3772 6.81583 1.01305 0.00005 0.00012399 cmet 10.5398 21.27 1.01298 0.00005 0.00012399 CG14314 1.2659 2.54872 1.00963 0.00115 0.00245271 CG8112 19.5818 39.3547 1.00703 0.0004 0.00090401 CG8668 11.6010 23.3135 1.00692 0.00005 0.00012399 Sirup 31.2982 62.8612 1.00609 0.00005 0.00012399 CG11357 11.8380 23.7698 1.0057 0.00005 0.00012399 unc 1.2056 2.41995 1.00517 0.00005 0.00012399
154 Klp68D 7.9669 15.983 1.00444 0.00005 0.00012399 CG7785 8.5745 17.1968 1.00401 0.00005 0.00012399 CG9636 7.6331 15.3063 1.00378 0.00005 0.00012399 CG13659 1.3358 2.67563 1.00214 0.00235 0.00478707 spo 7.6084 15.2242 1.0007 0.00005 0.00012399
Extended Table 2: Genes significantly downregulated in phf7ey/+;HA- phf7,nos-gal4 tumors relative to wild type (yw/yw) ovaries
gene WT (yw/yw) phf7ey log2 fold p-value q-value FPKM FPKM change Cht5 22.1466 0.0845277 -8.03345 0.00105 0.00225012 CG8774 6.8124 0.0427746 -7.31527 0.006 0.0114579 IM3 104.5720 0.893014 -6.8716 0.00655 0.0124245 Fbp1 20.4042 0.189133 -6.75332 0.00005 0.00012399 CG13492 7.1876 0.108755 -6.04636 0.00005 0.00012399 IM4 21.9053 0.452882 -5.596 0.00605 0.0115458 Lsp1beta 7.1838 0.156802 -5.51773 0.00005 0.00012399 CG16826 4.9924 0.109649 -5.50877 0.0051 0.00985593 Cyp6a2 18.0851 0.404048 -5.48413 0.00005 0.00012399 CG15067 23.0777 0.542509 -5.41071 0.00235 0.00478707 CG31343 14.5234 0.342666 -5.40543 0.00005 0.00012399 Obp99c 57.4125 1.44093 -5.3163 0.00005 0.00012399 e 2.1365 0.0538699 -5.30959 0.0055 0.0105729 Mal-A1 3.8522 0.0995821 -5.27363 0.00345 0.00684617 CG10073 3.3860 0.0888047 -5.25281 0.0014 0.00294985 ple 4.9320 0.137714 -5.16243 0.00005 0.00012399 Fmo-1 3.6234 0.105831 -5.0975 0.0055 0.0105729 Cyp318a1 19.5300 0.571074 -5.09587 0.00005 0.00012399 CG13067 74.3740 2.20387 -5.07668 0.00005 0.00012399 ect 3.5988 0.108338 -5.05391 0.00005 0.00012399 Vha100-5 2.4259 0.0752471 -5.01072 0.0006 0.00132671 MtnC 175.8400 5.46927 -5.00677 0.00005 0.00012399 Gasp 16.4623 0.542011 -4.9247 0.00005 0.00012399 CG33281 23.2508 0.776634 -4.9039 0.00005 0.00012399 CG2650 6.0736 0.211062 -4.84682 0.02705 0.0459271 Lsp1gamma 22.6911 0.803577 -4.81955 0.00005 0.00012399 Lsp2 17.8907 0.637976 -4.80956 0.00005 0.00012399 CG32834 2.2934 0.0822984 -4.80046 0.02045 0.0355526 CG31233 2.3957 0.0865327 -4.79105 0.0002 0.00046869 CG13075 20.4799 0.74358 -4.78358 0.00025 0.00058003 Cyp4e3 5.0715 0.195965 -4.69374 0.00005 0.00012399 CG3264 3.7721 0.157398 -4.58289 0.00005 0.00012399 vvl 1.7762 0.0753005 -4.55997 0.00005 0.00012399 CG8630 4.3130 0.18685 -4.52872 0.00005 0.00012399 CG12119 14.1270 0.650769 -4.44017 0.00005 0.00012399 CG11852 62.8022 2.89842 -4.43747 0.00005 0.00012399
155 ninaD 15.5183 0.729838 -4.41025 0.00005 0.00012399 Tsp42En 6.7373 0.325894 -4.36969 0.0001 0.0002417 Mdr50 2.0397 0.10362 -4.29899 0.00005 0.00012399 fon 11.1776 0.57045 -4.29237 0.00005 0.00012399 CG31248 13.1362 0.693377 -4.24377 0.00005 0.00012399 CG4950 13.8139 0.73932 -4.22378 0.00005 0.00012399 CG3999 1.1252 0.0618965 -4.18421 0.00115 0.00245271 Act87E 48.0620 2.65302 -4.17919 0.00005 0.00012399 vanin-like 1.8983 0.106887 -4.15058 0.0008 0.00174155 CG10680 5.8068 0.330907 -4.13324 0.00005 0.00012399 Cyp28a5 2.1786 0.124163 -4.13305 0.00105 0.00225012 CG33511 23.8904 1.37312 -4.1209 0.00005 0.00012399 Oatp58Db 3.2085 0.18472 -4.1185 0.00005 0.00012399 CG9568 7.8756 0.456868 -4.10754 0.0016 0.00334415 Npc2b 34.8978 2.05035 -4.08919 0.00005 0.00012399 Cyp313a4 13.5846 0.815977 -4.0573 0.00005 0.00012399 Vha36-3 1.8903 0.114598 -4.04398 0.0065 0.0123384 Glt 4.7515 0.302552 -3.97313 0.00005 0.00012399 Mal-A6 1.3034 0.0833105 -3.96758 0.00195 0.00402261 CG6733 35.6245 2.28405 -3.9632 0.00005 0.00012399 Gycbeta100B 1.8496 0.120517 -3.93992 0.0001 0.0002417 Scp2 46.5614 3.12774 -3.89594 0.00005 0.00012399 CG13283 2.0801 0.141649 -3.87623 0.0001 0.0002417 CG16820 5.6487 0.392778 -3.84613 0.00005 0.00012399 CG5867 54.5734 3.85104 -3.82488 0.00005 0.00012399 Hsp68 58.2405 4.21762 -3.78752 0.00005 0.00012399 Hsp70Ba 4.2298 0.30723 -3.78318 0.00005 0.00012399 Act57B 539.1640 39.4562 -3.7724 0.00005 0.00012399 CG8773 1.3545 0.101111 -3.74379 0.00015 0.00035649 CG17386 1.4107 0.105699 -3.73836 0.00005 0.00012399 CG4115 12.4066 0.936521 -3.72766 0.00005 0.00012399 CG17664 6.3741 0.485829 -3.7137 0.0002 0.00046869 Frq2 3.0389 0.233052 -3.70481 0.00005 0.00012399 CG6639 1.2970 0.0999165 -3.69832 0.0113 0.0206192 CG33282 4.8794 0.379375 -3.68501 0.00005 0.00012399 Ggamma30A 1.5107 0.117658 -3.6825 0.0008 0.00174155 wat 11.7737 0.917566 -3.68161 0.00005 0.00012399 Hsp70Bb 8.0487 0.631382 -3.67216 0.00005 0.00012399 CG8012 102.5510 8.11909 -3.65888 0.00005 0.00012399 IM1 15.9220 1.26428 -3.65463 0.00075 0.00163799 CG8785 3.2429 0.259808 -3.64175 0.00005 0.00012399 CG15293 10.4013 0.833956 -3.64065 0.00005 0.00012399 CG7497 44.0867 3.58654 -3.61968 0.00005 0.00012399 CG7881 4.3883 0.36077 -3.60451 0.00005 0.00012399 CG14523 4.9305 0.408759 -3.59241 0.00005 0.00012399 List 4.9488 0.411108 -3.5895 0.00045 0.00101056 Antp 4.7948 0.411343 -3.54307 0.00005 0.00012399 Dbi 15.2319 1.31332 -3.53581 0.00005 0.00012399
156 CG13068 7.2014 0.636404 -3.50025 0.012 0.0217772 CG11160 2.6566 0.238583 -3.47703 0.00005 0.00012399 Cyp6a16 1.7190 0.154407 -3.47677 0.0001 0.0002417 Cda4 2.5114 0.225945 -3.47443 0.00005 0.00012399 CG31097 4.4392 0.400449 -3.4706 0.00005 0.00012399 snRNA:U6ata 55704.9000 5036.16 -3.46741 0.00005 0.00012399 c:29B CG8654 3.5921 0.327924 -3.45341 0.00005 0.00012399 Tsp42Er 2.8048 0.258903 -3.4374 0.0057 0.0109272 Nplp3 4.5516 0.422324 -3.42994 0.0129 0.023287 GstE14 32.1023 2.99214 -3.42343 0.00005 0.00012399 CG10560 23.4996 2.19387 -3.42109 0.00005 0.00012399 lectin-28C 4.5435 0.427374 -3.41024 0.00055 0.00122116 Ppn 85.1714 8.10109 -3.39418 0.00005 0.00012399 Ugt37b1 1.6026 0.153859 -3.38077 0.00075 0.00163799 CngA 2.1902 0.210567 -3.3787 0.00005 0.00012399 CG12934 1.4914 0.143686 -3.37569 0.0077 0.0144499 CG42711 4.0746 0.392902 -3.37442 0.0004 0.00090401 NimC2 1.1597 0.112122 -3.37057 0.0013 0.00275014 CG17110 3.0083 0.291012 -3.36978 0.02205 0.0380713 Cyp313b1 7.1056 0.692352 -3.35937 0.00005 0.00012399 CG32024 43.8803 4.33406 -3.33978 0.00005 0.00012399 Clect27 1.4045 0.139277 -3.33399 0.00415 0.00812941 CG3290 52.6574 5.26323 -3.32262 0.00005 0.00012399 CG16712 41.7410 4.20723 -3.31052 0.00005 0.00012399 CG1213 4.7836 0.483512 -3.30648 0.00005 0.00012399 CG9512 53.0762 5.38231 -3.30177 0.00005 0.00012399 CG7059 2.0566 0.21147 -3.28174 0.005 0.00967666 CG17190 24.5922 2.54277 -3.27372 0.00005 0.00012399 CG13012 5.2515 0.548851 -3.25824 0.00005 0.00012399 Prat2 2.4611 0.262226 -3.23043 0.00005 0.00012399 Sodh-1 6.5775 0.709024 -3.21363 0.00005 0.00012399 AttB 1.3395 0.144796 -3.2096 0.0179 0.031457 CG12105 1.3135 0.142097 -3.20843 0.0008 0.00174155 Obp83cd 2.6702 0.290142 -3.20212 0.00185 0.00382769 Mhc 138.0560 15.2613 -3.17731 0.00005 0.00012399 CG17374 6.4648 0.716377 -3.17381 0.00005 0.00012399 NAAT1 2.1559 0.240203 -3.16596 0.00005 0.00012399 CG34284 96.7512 10.8958 -3.1505 0.00005 0.00012399 CG6503 34.3918 3.89043 -3.14406 0.00005 0.00012399 CG2680 25.9087 2.93706 -3.14099 0.00005 0.00012399 Hex-C 23.5040 2.71621 -3.11324 0.00005 0.00012399 Ance-4 1.0338 0.121077 -3.0939 0.00155 0.00324556 Nplp2 233.1320 27.4172 -3.08799 0.00005 0.00012399 CG3699 44.2771 5.20719 -3.08798 0.00005 0.00012399 obst-A 1.1861 0.139639 -3.08649 0.0012 0.00255302 mex1 1.6109 0.190202 -3.08222 0.0136 0.0244572 Mlc2 201.8680 23.8685 -3.08023 0.00005 0.00012399
157 CR45280 2.3647 0.280558 -3.07526 0.00005 0.00012399 Mlc1 125.6400 15.0291 -3.06347 0.00005 0.00012399 CG15556 2.2281 0.266883 -3.06152 0.00005 0.00012399 TotA 3.1348 0.378586 -3.04967 0.0083 0.0154914 CG3191 8.9218 1.09 -3.03301 0.00005 0.00012399 NimB2 4.2245 0.516233 -3.03268 0.00005 0.00012399 verm 4.6412 0.572367 -3.01949 0.00005 0.00012399 CG42492 9.7390 1.20106 -3.01947 0.00005 0.00012399 CG34423 9.5698 1.18212 -3.01711 0.00695 0.0131356 CG10553 49.7173 6.14603 -3.01602 0.00005 0.00012399 CG6142 1.4761 0.182524 -3.01561 0.0001 0.0002417 CG8483 1.5031 0.186739 -3.00883 0.00055 0.00122116 slo 2.9174 0.362453 -3.0088 0.00005 0.00012399 CG8888 8.4940 1.06017 -3.00215 0.00005 0.00012399 CG8586 2.2010 0.274746 -3.00201 0.00005 0.00012399 Cpr49Ab 1.2760 0.159719 -2.99798 0.0204 0.0354708 Unc-89 7.7460 0.980623 -2.98169 0.00005 0.00012399 CG10361 1.5887 0.201579 -2.97844 0.0019 0.00392473 Tsp29Fa 6.0414 0.769556 -2.97278 0.00005 0.00012399 NtR 6.0903 0.776749 -2.971 0.00005 0.00012399 CG4847 1.5753 0.201623 -2.96585 0.00125 0.00265143 CG9486 1.2320 0.158201 -2.96116 0.02755 0.0467146 CG30047 1.5335 0.1972 -2.95907 0.0001 0.0002417 CG32695 10.2519 1.32288 -2.95414 0.00005 0.00012399 CG2254 10.6593 1.37652 -2.95301 0.00005 0.00012399 Tm2 133.4110 17.2581 -2.95053 0.00005 0.00012399 CG14694 8.1145 1.05161 -2.9479 0.00005 0.00012399 CG9396 14.6172 1.91383 -2.93313 0.00005 0.00012399 CG7759 1.4178 0.186314 -2.9278 0.00025 0.00058003 Cpr97Eb 2.2214 0.292157 -2.92662 0.00085 0.00184174 AdoR 2.0901 0.275276 -2.92461 0.00005 0.00012399 CG7882 43.6296 5.74899 -2.92393 0.00005 0.00012399 CG9270 3.4367 0.457616 -2.90883 0.00005 0.00012399 FucTC 1.1836 0.160179 -2.88548 0.0022 0.00449953 CG9498 13.7889 1.86635 -2.88521 0.00005 0.00012399 Irk3 29.3882 4.01688 -2.87109 0.00005 0.00012399 Neto 2.4907 0.341274 -2.86752 0.00005 0.00012399 Odc1 1.4428 0.199735 -2.85273 0.00335 0.00666329 Smyd4 2.0897 0.29016 -2.84834 0.00005 0.00012399 CG3285 11.5504 1.61351 -2.83966 0.00005 0.00012399 CG4019 9.4169 1.32568 -2.82851 0.00005 0.00012399 TpnC73F 79.0423 11.1566 -2.82472 0.00005 0.00012399 CG4259 3.2293 0.456979 -2.82102 0.00005 0.00012399 CG8837 4.5092 0.643033 -2.80991 0.00005 0.00012399 CG2010 3.5840 0.515813 -2.79664 0.00005 0.00012399 Elp2 96.0561 13.8353 -2.79553 0.00005 0.00012399 CG8028 7.5687 1.09303 -2.79171 0.00005 0.00012399 CadN 10.3337 1.5088 -2.77588 0.00005 0.00012399
158 CG40472 33.4077 4.92157 -2.76299 0.00005 0.00012399 CG9914 15.0440 2.27149 -2.72748 0.0037 0.00730216 salt 40.9824 6.20741 -2.72294 0.00005 0.00012399 Obp18a 3.5810 0.542837 -2.72175 0.0011 0.00235148 Nplp4 2.3718 0.359579 -2.72158 0.0202 0.035151 CG42833 26.0677 3.96344 -2.71744 0.00005 0.00012399 St1 2.0184 0.307793 -2.71315 0.00095 0.00204645 CDase 3.1501 0.481059 -2.71111 0.00005 0.00012399 ry 2.7653 0.422671 -2.7098 0.00005 0.00012399 Dh31-R 2.3085 0.353165 -2.70851 0.00005 0.00012399 DNaseII 1.1574 0.177562 -2.70455 0.0079 0.014793 Cyp4ad1 1.2068 0.185848 -2.69904 0.0023 0.0046921 CG9691 7.1479 1.10553 -2.69279 0.00005 0.00012399 Cyp4g15 3.6611 0.566354 -2.69251 0.00005 0.00012399 CG42807 50.7804 7.86638 -2.6905 0.00005 0.00012399 CG3270 36.7501 5.70897 -2.68644 0.00005 0.00012399 Ugt36Bc 8.4886 1.32269 -2.68205 0.00005 0.00012399 CG9194 2.2193 0.346814 -2.6779 0.02275 0.0391592 mspo 3.1733 0.4985 -2.6703 0.00005 0.00012399 CG6225 2.5528 0.401468 -2.66872 0.00005 0.00012399 MtnE 25.4525 4.00869 -2.6666 0.00005 0.00012399 CG15531 3.9373 0.620576 -2.66551 0.00005 0.00012399 Adgf-C 2.9462 0.464622 -2.66475 0.00005 0.00012399 CG15043 1.4174 0.224691 -2.65724 0.01885 0.0329989 CG6337 1.5843 0.25271 -2.64827 0.00415 0.00812941 Drs 48.2844 7.72565 -2.64383 0.00005 0.00012399 nAChRbeta3 4.2866 0.685885 -2.64381 0.00005 0.00012399 Oatp58Da 5.6025 0.900211 -2.63774 0.00005 0.00012399 CG34426 20.9711 3.38761 -2.63006 0.00005 0.00012399 NimC1 4.7735 0.771475 -2.62934 0.00005 0.00012399 CG7443 180.9270 29.2862 -2.62711 0.00005 0.00012399 fusl 3.6213 0.586329 -2.62673 0.00005 0.00012399 hh 1.1272 0.182515 -2.6266 0.00055 0.00122116 NimB5 3.2937 0.53432 -2.62395 0.00005 0.00012399 CG14195 2.2278 0.362176 -2.62088 0.00005 0.00012399 Fst 3.7564 0.61148 -2.61898 0.00005 0.00012399 CCHa2 16.4969 2.68961 -2.61672 0.00005 0.00012399 CG13607 2.8312 0.462661 -2.61339 0.01115 0.0203697 CG13360 1.3704 0.224474 -2.61 0.006 0.0114579 CG11889 28.5247 4.67945 -2.6078 0.00005 0.00012399 Mlp60A 199.7320 32.8391 -2.60458 0.00005 0.00012399 CR44897 3.4042 0.560389 -2.60283 0.0008 0.00174155 Obp99a 30.5793 5.03936 -2.60124 0.00005 0.00012399 Prm 20.6803 3.41635 -2.59773 0.00005 0.00012399 SPE 2.0177 0.335869 -2.58677 0.0008 0.00174155 ap 2.0194 0.337092 -2.58271 0.00005 0.00012399 CG31198 30.9469 5.18376 -2.57772 0.00005 0.00012399 Toll-7 1.0019 0.169162 -2.56618 0.00005 0.00012399
159 shakB 2.7677 0.469507 -2.55944 0.00005 0.00012399 Cyp4e1 1.9228 0.326739 -2.55697 0.00005 0.00012399 CG11395 3.8691 0.65895 -2.55377 0.00005 0.00012399 CG34198 5.1798 0.884135 -2.55056 0.00265 0.005355 CG43163 1.7229 0.294367 -2.54911 0.00005 0.00012399 Oatp33Eb 1.9918 0.340363 -2.54894 0.00005 0.00012399 CG18095 12.2239 2.09051 -2.54778 0.00005 0.00012399 CG9090 5.0534 0.872496 -2.53402 0.00005 0.00012399 PGRP-SB1 6.6863 1.15508 -2.53322 0.00015 0.00035649 CG16857 8.7619 1.51411 -2.53278 0.00005 0.00012399 RyR 2.0102 0.348096 -2.52975 0.00005 0.00012399 CG30196 4.9410 0.860214 -2.52203 0.0032 0.00638939 CG17669 2.0788 0.362536 -2.51954 0.00005 0.00012399 CG6300 1.4418 0.251898 -2.51693 0.00165 0.00344128 CG1136 8.5385 1.49746 -2.51147 0.00005 0.00012399 CG8642 5.1927 0.910896 -2.51112 0.00075 0.00163799 Cg25C 122.6750 21.5475 -2.50925 0.00005 0.00012399 CG31517 23.7437 4.17876 -2.5064 0.00005 0.00012399 CG5955 1.8263 0.323485 -2.49716 0.0011 0.00235148 up 113.7210 20.1969 -2.49329 0.00005 0.00012399 CG5321 1.5651 0.278073 -2.49269 0.00095 0.00204645 CG8468 6.7455 1.1997 -2.49126 0.00355 0.00702743 CG42624 1.1499 0.2047 -2.48995 0.01835 0.0321962 hth 2.9051 0.523966 -2.47106 0.00005 0.00012399 CG9717 4.2865 0.773212 -2.47086 0.00005 0.00012399 Hsc70-2 2.7667 0.505294 -2.45294 0.00005 0.00012399 Tig 2.7387 0.502241 -2.44706 0.00005 0.00012399 Tsp42Eq 4.5730 0.843992 -2.43784 0.00005 0.00012399 Cpr49Ae 3.0896 0.571508 -2.43458 0.00005 0.00012399 CG42235 51.9144 9.62014 -2.432 0.00005 0.00012399 CR44363 1.8771 0.349313 -2.42594 0.0112 0.0204523 CG15695 2.2904 0.426642 -2.42449 0.00005 0.00012399 CG31140 2.9254 0.546725 -2.41976 0.00005 0.00012399 Irk2 6.0255 1.13132 -2.41306 0.00005 0.00012399 Smvt 18.6304 3.50038 -2.41207 0.00005 0.00012399 eater 1.4805 0.279296 -2.40624 0.00005 0.00012399 CG4716 8.3437 1.57418 -2.40609 0.00005 0.00012399 CG18327 7.6159 1.43888 -2.40406 0.0008 0.00174155 NimB4 1.1480 0.218096 -2.3961 0.0057 0.0109272 CR44608 2.1936 0.417476 -2.39352 0.0212 0.0367416 CR44965 3.9280 0.748045 -2.39259 0.00475 0.00922669 Tsf1 20.6588 3.93445 -2.39252 0.00005 0.00012399 ETHR 1.4793 0.281823 -2.39209 0.00005 0.00012399 Gbp 11.9014 2.26881 -2.39113 0.00005 0.00012399 serp 3.3186 0.634886 -2.38602 0.00005 0.00012399 CG17470 13.1620 2.5302 -2.37906 0.00005 0.00012399 AstA-R2 2.5650 0.493706 -2.37726 0.00005 0.00012399 per 3.0436 0.587057 -2.37419 0.00005 0.00012399
160 MtnD 44.0423 8.50673 -2.37221 0.00005 0.00012399 CG12963 9.0829 1.75981 -2.36773 0.00005 0.00012399 wupA 156.5730 30.462 -2.36175 0.00005 0.00012399 ine 2.4038 0.470077 -2.35437 0.00005 0.00012399 CG30411 1.7392 0.340296 -2.35358 0.0088 0.0163498 CG2233 29.7659 5.83025 -2.35203 0.00005 0.00012399 CG42808 7.2948 1.43299 -2.34784 0.00005 0.00012399 CG13506 22.1465 4.35995 -2.3447 0.00005 0.00012399 CG5773 1.5572 0.307225 -2.34155 0.0112 0.0204523 CG5697 13.1221 2.59056 -2.34066 0.00005 0.00012399 CG6465 8.3163 1.64271 -2.33986 0.00005 0.00012399 CG2187 17.4136 3.44053 -2.33951 0.00005 0.00012399 CG14298 6.5904 1.30759 -2.33345 0.00005 0.00012399 CG13086 3.1620 0.627681 -2.33273 0.00245 0.00497772 CG1208 3.0551 0.607593 -2.33003 0.00005 0.00012399 CG10562 4.2776 0.851446 -2.32881 0.00005 0.00012399 tn 10.2058 2.03757 -2.32447 0.00005 0.00012399 CG5177 29.8955 5.97462 -2.32301 0.00005 0.00012399 lea 2.2230 0.445993 -2.31743 0.00005 0.00012399 CG5022 1.1161 0.224815 -2.31161 0.0005 0.00111605 CG34166 17.1824 3.46199 -2.31126 0.00005 0.00012399 CG13284 4.7751 0.966287 -2.30501 0.00005 0.00012399 bt 44.0207 8.94036 -2.29978 0.00005 0.00012399 CG8665 2.0935 0.426221 -2.29621 0.00005 0.00012399 Obp44a 6.7442 1.37352 -2.29577 0.01585 0.0281759 disco-r 4.4361 0.909104 -2.28676 0.00005 0.00012399 CG6972 2.2823 0.468212 -2.28525 0.00005 0.00012399 Hsp22 59.3294 12.1823 -2.28397 0.00005 0.00012399 AkhR 1.2025 0.247218 -2.28219 0.0002 0.00046869 CG5023 44.2761 9.10802 -2.28132 0.00005 0.00012399 Oat 34.7321 7.15221 -2.27981 0.00005 0.00012399 bab2 3.1948 0.658113 -2.27932 0.00005 0.00012399 Tps1 29.1436 6.00575 -2.27876 0.00005 0.00012399 CG3597 4.5030 0.931782 -2.27283 0.00005 0.00012399 axo 1.2030 0.249032 -2.2722 0.00005 0.00012399 nAChRalpha 1.3910 0.287971 -2.27212 0.00895 0.0166093 3 CG10396 2.2810 0.472839 -2.27025 0.009 0.0166935 CG18480 1.3664 0.283543 -2.26875 0.0001 0.0002417 CG31288 15.5226 3.22219 -2.26826 0.00005 0.00012399 CG6726 135.2370 28.1258 -2.26552 0.00005 0.00012399 CG16727 10.2279 2.14331 -2.2546 0.00005 0.00012399 CG6012 2.0149 0.42509 -2.24485 0.0015 0.00314697 CG31663 6.8124 1.43978 -2.24232 0.00005 0.00012399 CG5171 13.3151 2.81767 -2.24049 0.00005 0.00012399 CG10550 5.0760 1.07435 -2.24022 0.00005 0.00012399 Gal 3.0568 0.649219 -2.23525 0.00015 0.00035649 Cyp28d1 1.6472 0.351556 -2.22818 0.00035 0.0007973
161 CG43074 12.7892 2.73177 -2.22702 0.00005 0.00012399 zormin 11.3285 2.42577 -2.22344 0.00005 0.00012399 CG32023 58.4120 12.5265 -2.22128 0.00005 0.00012399 sls 17.2488 3.7006 -2.22066 0.00005 0.00012399 Peb 1.5502 0.333148 -2.21819 0.0008 0.00174155 Blimp-1 2.6652 0.572901 -2.2179 0.00005 0.00012399 CG11380 1.1918 0.2573 -2.21167 0.00005 0.00012399 CG17027 7.9978 1.74474 -2.19658 0.00535 0.0103035 CG10132 1.0645 0.232273 -2.19629 0.02095 0.0363388 vkg 124.1000 27.1132 -2.19444 0.00005 0.00012399 CG9780 1.6905 0.369683 -2.19309 0.0295 0.0497397 CG8501 1.5236 0.333248 -2.1928 0.02275 0.0391592 NaPi-T 30.9749 6.77581 -2.19263 0.00005 0.00012399 CapaR 2.5230 0.551976 -2.19249 0.00005 0.00012399 Ance-5 3.0546 0.671753 -2.18496 0.00005 0.00012399 Mlp84B 24.0737 5.2972 -2.18415 0.0001 0.0002417 CG11659 2.0461 0.452936 -2.17551 0.00005 0.00012399 Ude 4.6428 1.03378 -2.16707 0.00005 0.00012399 Oatp58Dc 11.1055 2.48207 -2.16165 0.00005 0.00012399 CG7084 20.5354 4.59044 -2.16141 0.00005 0.00012399 CG15406 22.3902 5.01943 -2.15727 0.00005 0.00012399 CG42817 5.3899 1.20891 -2.15655 0.00005 0.00012399 Zasp66 61.6947 13.8729 -2.15287 0.00005 0.00012399 lectin-24Db 4.1614 0.936065 -2.1524 0.00005 0.00012399 CG11231 2.3401 0.530795 -2.14034 0.00005 0.00012399 CG42369 3.8583 0.875572 -2.13967 0.0002 0.00046869 CG34331 3.8578 0.877523 -2.13626 0.0014 0.00294985 GluRIID 3.1598 0.722351 -2.12905 0.0021 0.00430915 CG9119 29.6541 6.77976 -2.12893 0.00005 0.00012399 CG13793 2.1016 0.481219 -2.12672 0.00005 0.00012399 mthl2 2.3839 0.546089 -2.12612 0.0111 0.020285 CG4829 6.9351 1.59439 -2.12091 0.00005 0.00012399 CG1674 26.0433 6.01445 -2.11441 0.00005 0.00012399 CG32599 2.2794 0.52697 -2.11286 0.0003 0.00068977 Cyp4p1 4.1997 0.972023 -2.11122 0.00005 0.00012399 Gs2 27.9774 6.49391 -2.1071 0.00005 0.00012399 CG10513 68.9833 16.0323 -2.10527 0.00005 0.00012399 CR44510 2.1037 0.489593 -2.1033 0.0059 0.0112845 CG15186 1.3913 0.324441 -2.1004 0.00005 0.00012399 E(spl)malpha 3.0824 0.719588 -2.0988 0.00125 0.00265143 -BFM CG10131 1.6438 0.383993 -2.09791 0.0047 0.00913669 Mipp1 96.4194 22.5829 -2.09409 0.00005 0.00012399 Spn42Dd 11.3129 2.65243 -2.09258 0.00005 0.00012399 dsx 13.0789 3.08485 -2.08397 0.00035 0.0007973 CG10086 3.1981 0.756848 -2.07915 0.00005 0.00012399 tnc 2.3316 0.552381 -2.0776 0.00005 0.00012399 CG1607 1.0226 0.243874 -2.06808 0.00005 0.00012399
162 Pkg21D 2.4817 0.592039 -2.06758 0.00005 0.00012399 BM-40- 104.1230 24.9081 -2.0636 0.00005 0.00012399 SPARC Ir40a 1.0713 0.257215 -2.05836 0.0001 0.0002417 Rgk3 2.8451 0.685015 -2.05426 0.00005 0.00012399 Cyt-c-d 6.9146 1.67025 -2.04958 0.0004 0.00090401 CG12014 2.0726 0.502864 -2.04321 0.00005 0.00012399 CG12825 19.2595 4.68765 -2.03863 0.00005 0.00012399 CG17111 5.8182 1.41915 -2.03555 0.0003 0.00068977 CR45485 34.7765 8.53356 -2.02689 0.00005 0.00012399 Tsp29Fb 27.1321 6.67785 -2.02255 0.00005 0.00012399 E(spl)m3- 44.4592 10.9435 -2.02241 0.00005 0.00012399 HLH CG9164 1.4128 0.34787 -2.02193 0.00005 0.00012399 CG14572 1.1406 0.281377 -2.01917 0.0151 0.0269434 CG34043 16.5907 4.0965 -2.01791 0.00005 0.00012399 CG3987 2.5221 0.623788 -2.01547 0.00115 0.00245271 CG18304 3.5782 0.885716 -2.01432 0.00005 0.00012399 CG9505 8.9233 2.2114 -2.01262 0.00005 0.00012399 Fuca 13.1104 3.26817 -2.00416 0.00005 0.00012399 CG3835 25.0702 6.2502 -2.004 0.00005 0.00012399 trh 1.7108 0.426598 -2.00372 0.00005 0.00012399 CG13704 3.6013 0.899798 -2.00086 0.00105 0.00225012 CG6847 6.8361 1.71036 -1.99886 0.00005 0.00012399 CG31808 35.3694 8.87891 -1.99405 0.00005 0.00012399 CG14777 28.4456 7.16646 -1.98887 0.00005 0.00012399 Mf 145.1400 36.5676 -1.98881 0.00005 0.00012399 smp-30 15.1667 3.82236 -1.98838 0.0113 0.0206192 CG32750 1.1729 0.295766 -1.98753 0.00165 0.00344128 Ca-alpha1D 2.0377 0.516272 -1.98075 0.00005 0.00012399 Ms 1.4100 0.357816 -1.97837 0.0049 0.00949916 CG10514 96.8955 24.6226 -1.97644 0.00005 0.00012399 CG12541 5.1400 1.31246 -1.96949 0.00005 0.00012399 CG14661 1.6999 0.434744 -1.96723 0.0019 0.00392473 eya 50.6211 12.9613 -1.96553 0.00005 0.00012399 CG43156 61.7409 15.8643 -1.96044 0.00005 0.00012399 CG14606 4.6269 1.18947 -1.95973 0.00005 0.00012399 CG7724 1.2398 0.323057 -1.94029 0.00385 0.00757914 Spn31A 4.3172 1.12685 -1.93779 0.00005 0.00012399 RhoGAP102 4.1856 1.09523 -1.93421 0.00005 0.00012399 A LKR 6.4595 1.6917 -1.93294 0.00005 0.00012399 l(2)08717 29.4947 7.72508 -1.93283 0.00005 0.00012399 Fmo-2 10.4829 2.74588 -1.93269 0.00005 0.00012399 CG45076 16.9849 4.46045 -1.92899 0.00005 0.00012399 CG9760 2.2118 0.582155 -1.92576 0.00005 0.00012399 CG31102 2.1601 0.570954 -1.91966 0.00005 0.00012399 CG18324 2.1326 0.565075 -1.91612 0.0213 0.0369007
163 Rh50 1.8314 0.486986 -1.91103 0.00015 0.00035649 CG4928 32.8224 8.74655 -1.9079 0.00005 0.00012399 frma 1.7818 0.475523 -1.90573 0.00005 0.00012399 trol 139.8660 37.4601 -1.90062 0.00005 0.00012399 CG12355 27.2471 7.33909 -1.89243 0.00005 0.00012399 CR44953 5.8841 1.58597 -1.89147 0.00005 0.00012399 CG42249 1.8649 0.506585 -1.88023 0.00005 0.00012399 Epac 2.2734 0.617657 -1.88 0.00005 0.00012399 ppa 5.0441 1.37051 -1.87989 0.00005 0.00012399 CG43968 3.3618 0.917928 -1.87277 0.00005 0.00012399 CG10359 1.1947 0.327892 -1.8653 0.0016 0.00334415 CG14275 19.2169 5.28321 -1.86289 0.00005 0.00012399 CG42450 4.0934 1.12649 -1.86147 0.00005 0.00012399 CG8157 1.1348 0.313185 -1.85729 0.02605 0.0443554 SclB 44.4250 12.2925 -1.8536 0.00005 0.00012399 CG3690 19.6214 5.44135 -1.85039 0.00005 0.00012399 CG12069 1.1223 0.31132 -1.84994 0.0011 0.00235148 CG33557 3.9760 1.10389 -1.84874 0.0006 0.00132671 CG3168 96.2017 26.9025 -1.83832 0.00005 0.00012399 CG11321 1.1545 0.323121 -1.83714 0.0058 0.0111056 CG8539 9.8625 2.77462 -1.82966 0.00005 0.00012399 Apoltp 1.3682 0.384933 -1.8296 0.00005 0.00012399 CG9743 9.5513 2.69659 -1.82456 0.00005 0.00012399 Sobp 9.9875 2.82061 -1.82412 0.00005 0.00012399 Mal-A3 1.3051 0.368968 -1.82256 0.00065 0.00143132 CG9510 4.2116 1.19449 -1.81797 0.0007 0.00153459 CG18473 4.8728 1.38313 -1.8168 0.00005 0.00012399 CG2993 2.0093 0.574157 -1.80719 0.00005 0.00012399 phm 7.6477 2.18806 -1.80537 0.00005 0.00012399 Scgalpha 2.3544 0.673986 -1.80459 0.00005 0.00012399 CG4038 77.2916 22.1415 -1.80356 0.00005 0.00012399 Npc2g 4.3781 1.2566 -1.80078 0.0005 0.00111605 Nep3 2.8168 0.813107 -1.79253 0.00005 0.00012399 Pif1A 13.2693 3.83324 -1.79146 0.00005 0.00012399 tutl 4.2431 1.22659 -1.79048 0.00005 0.00012399 CG44245 9.0139 2.60794 -1.78923 0.00005 0.00012399 fau 85.1381 24.7327 -1.78339 0.00005 0.00012399 CG33521 15.0883 4.3868 -1.78219 0.00005 0.00012399 CR43242 7.0494 2.05045 -1.78156 0.00005 0.00012399 CG6602 29.3664 8.54735 -1.78062 0.00005 0.00012399 Cyt-b5-r 9.6709 2.81764 -1.77917 0.00005 0.00012399 CG15766 12.9615 3.79525 -1.77197 0.00005 0.00012399 Cap-H2 29.2357 8.58164 -1.76841 0.00005 0.00012399 drl 7.0091 2.05957 -1.76688 0.00005 0.00012399 CG2681 1.4778 0.434359 -1.76649 0.00005 0.00012399 otk 2.8891 0.84984 -1.76535 0.00005 0.00012399 Plod 165.2690 48.6857 -1.76325 0.00005 0.00012399 CG5910 4.3518 1.28676 -1.75788 0.0095 0.0175565
164 CG43729 4.6574 1.37781 -1.75715 0.00005 0.00012399 CR44281 5.1307 1.52099 -1.75416 0.01755 0.0308961 CG12814 6.2785 1.86239 -1.75327 0.00005 0.00012399 bru-3 5.1217 1.5199 -1.75264 0.00005 0.00012399 CG5758 21.0742 6.25834 -1.75162 0.00005 0.00012399 CG31690 1.7101 0.512144 -1.73946 0.00005 0.00012399 CG12974 1.9794 0.593115 -1.73865 0.00015 0.00035649 CG17752 23.6628 7.10025 -1.73668 0.00005 0.00012399 Oct-TyrR 1.5811 0.475893 -1.73218 0.00005 0.00012399 Spat 5.8846 1.77518 -1.72897 0.00005 0.00012399 dnd 4.0375 1.21837 -1.72851 0.009 0.0166935 CG43897 14.8912 4.50386 -1.72522 0.00005 0.00012399 CG4250 127.6910 38.6835 -1.72287 0.00005 0.00012399 CG4623 1.9149 0.583407 -1.71471 0.0239 0.0409688 nord 12.6244 3.84777 -1.71412 0.00005 0.00012399 Nlg1 4.3908 1.33918 -1.71313 0.00005 0.00012399 CG8051 1.4549 0.445361 -1.70789 0.00005 0.00012399 ham 3.3067 1.01223 -1.70786 0.00005 0.00012399 CG8248 2.4335 0.746267 -1.70529 0.01695 0.029928 CG30016 36.0259 11.0539 -1.70448 0.00005 0.00012399 CR43866 1.6735 0.514289 -1.70223 0.02565 0.0437228 CG13795 1.1830 0.36447 -1.69854 0.0013 0.00275014 dpp 4.8543 1.49801 -1.69621 0.00005 0.00012399 betaTub97EF 34.9650 10.7966 -1.69533 0.00005 0.00012399 CG31272 10.3149 3.18592 -1.69495 0.00005 0.00012399 tau 2.7913 0.862229 -1.69481 0.00005 0.00012399 Eaat1 2.8832 0.89241 -1.6919 0.00005 0.00012399 Ugt58Fa 36.1497 11.1989 -1.69063 0.00005 0.00012399 so 2.6500 0.822204 -1.68843 0.00005 0.00012399 CG4302 7.0839 2.20061 -1.68663 0.00015 0.00035649 CG33080 4.5635 1.42306 -1.68113 0.00005 0.00012399 loh 2.7545 0.859991 -1.67942 0.00005 0.00012399 Hf 2.2537 0.703978 -1.6787 0.00205 0.00421386 su(r) 6.0361 1.88919 -1.67585 0.00005 0.00012399 Karl 3.3111 1.03777 -1.67383 0.00005 0.00012399 comt 6.8692 2.15578 -1.67193 0.00005 0.00012399 TrissinR 2.9145 0.918531 -1.66585 0.00005 0.00012399 Uhg8 65.4204 20.7389 -1.6574 0.00005 0.00012399 CG34446 17.2469 5.47337 -1.65583 0.0002 0.00046869 AdamTS-A 22.7705 7.25248 -1.65062 0.00005 0.00012399 shf 5.7780 1.84467 -1.64721 0.0001 0.0002417 CG32669 7.6554 2.44877 -1.64442 0.01235 0.022364 Drip 47.1252 15.1186 -1.64017 0.00005 0.00012399 msta 1.6638 0.533884 -1.63985 0.0001 0.0002417 qtc 22.6614 7.2776 -1.6387 0.00005 0.00012399 CG10170 1.8897 0.607349 -1.63759 0.0001 0.0002417 CG18810 1.4261 0.458488 -1.6371 0.01175 0.02136 CG7589 2.3353 0.751519 -1.63569 0.00005 0.00012399
165 CG11594 5.9582 1.91946 -1.63418 0.00005 0.00012399 CG3961 8.2694 2.68156 -1.62471 0.00005 0.00012399 CG40002 131.3710 42.7819 -1.61858 0.00005 0.00012399 Neurochondri 15.3068 4.99237 -1.61637 0.00005 0.00012399 n CG15408 9.7660 3.20055 -1.60945 0.00005 0.00012399 CG12592 6.2944 2.0631 -1.60925 0.00005 0.00012399 Faa 2.0833 0.683308 -1.60825 0.00065 0.00143132 Ca-beta 3.1504 1.03347 -1.60806 0.00655 0.0124245 Pde1c 3.1084 1.02019 -1.60733 0.0025 0.00507274 lbm 6.4895 2.13509 -1.60381 0.00005 0.00012399 CG7564 111.5350 36.8107 -1.59929 0.00005 0.00012399 CG45078 13.8737 4.58155 -1.59844 0.00005 0.00012399 Wnt4 3.7504 1.24464 -1.59131 0.00005 0.00012399 LanB2 39.8903 13.3399 -1.58029 0.00005 0.00012399 scro 2.7132 0.909457 -1.5769 0.0002 0.00046869 Phae1 1.9324 0.64816 -1.57599 0.0024 0.00488204 CG33514 2.6139 0.87848 -1.5731 0.0083 0.0154914 CR43650 5.0321 1.69728 -1.56793 0.00005 0.00012399 St3 6.9627 2.35147 -1.56609 0.00005 0.00012399 abd-A 6.3835 2.15866 -1.56421 0.00005 0.00012399 CG17029 29.1155 9.85122 -1.56341 0.0002 0.00046869 Hsp67Bb 13.7024 4.67305 -1.55199 0.00755 0.0141891 pio 7.9317 2.71864 -1.54474 0.00005 0.00012399 caz 45.6981 15.6734 -1.54381 0.00005 0.00012399 Pxn 11.6818 4.00806 -1.54329 0.00005 0.00012399 jbug 10.9843 3.77131 -1.54231 0.00005 0.00012399 CR43820 6.9373 2.38221 -1.54207 0.0148 0.0264434 CG10650 4.0072 1.37625 -1.54186 0.00005 0.00012399 if 8.8134 3.02955 -1.5406 0.00005 0.00012399 form3 3.9671 1.36434 -1.53989 0.0001 0.0002417 CG17732 4.3776 1.50591 -1.53949 0.00005 0.00012399 CR40976 986.0630 339.31 -1.53908 0.00005 0.00012399 htl 8.7151 3.00925 -1.53412 0.00005 0.00012399 Hrb87F 65.3080 22.5646 -1.5332 0.00005 0.00012399 NfI 1.8649 0.644418 -1.53299 0.00005 0.00012399 obst-E 1.3152 0.455713 -1.52908 0.01655 0.029302 CG31323 4.1964 1.45448 -1.52864 0.01175 0.02136 CG3746 5.3549 1.86118 -1.52463 0.0016 0.00334415 CG5819 1.7376 0.60601 -1.51971 0.00005 0.00012399 Pkcdelta 3.7635 1.31353 -1.51863 0.00005 0.00012399 dpr19 1.4457 0.505569 -1.51581 0.0002 0.00046869 CG13255 4.0642 1.42293 -1.51409 0.00025 0.00058003 CG8547 11.9186 4.20744 -1.5022 0.00005 0.00012399 Gpb5 5.5758 1.96937 -1.50145 0.00015 0.00035649 alpha-Est8 3.5109 1.24107 -1.50027 0.00005 0.00012399 CG10830 6.7553 2.395 -1.496 0.00005 0.00012399 CG5162 6.3159 2.2395 -1.49582 0.00005 0.00012399
166 LanA 89.3140 31.7019 -1.49432 0.00005 0.00012399 CG44476 181.2200 64.3352 -1.49406 0.00005 0.00012399 E(spl)m7- 5.3053 1.88408 -1.49356 0.00005 0.00012399 HLH CG9338 46.4355 16.5799 -1.48579 0.00005 0.00012399 CTPsyn 125.7390 44.9083 -1.48538 0.00005 0.00012399 CG40298 3.3255 1.19032 -1.48221 0.00035 0.0007973 Pvf2 2.7019 0.969619 -1.47848 0.0001 0.0002417 Mal-A4 1.0001 0.359695 -1.47523 0.00425 0.00831313 CG11407 2.0841 0.751061 -1.47243 0.0002 0.00046869 CR45015 2.9347 1.05844 -1.47129 0.0043 0.00840501 l(1)G0004 46.2314 16.6853 -1.4703 0.00005 0.00012399 Ssk 15.6650 5.66381 -1.4677 0.0032 0.00638939 CG33978 1.6335 0.590663 -1.46756 0.00005 0.00012399 Ucrh 659.1360 239.236 -1.46214 0.00005 0.00012399 CG7781 5.0255 1.83016 -1.4573 0.00055 0.00122116 melt 4.0023 1.45813 -1.45669 0.00005 0.00012399 CG15543 1.0258 0.375823 -1.44861 0.00635 0.0120755 CR31451 2.4567 0.902266 -1.44508 0.00005 0.00012399 os 1.9586 0.719989 -1.44375 0.0001 0.0002417 CG4398 24.6005 9.05823 -1.44139 0.00005 0.00012399 5-HT2A 1.2998 0.478985 -1.4402 0.00005 0.00012399 fat-spondin 7.5080 2.76837 -1.43939 0.0028 0.00564312 nrm 12.2413 4.51447 -1.43913 0.00005 0.00012399 Uhg4 329.9060 121.735 -1.43831 0.00005 0.00012399 Nrx-1 2.0051 0.740435 -1.43726 0.00005 0.00012399 Fib 240.3890 88.8438 -1.43603 0.00005 0.00012399 bnl 1.5890 0.587481 -1.43546 0.00005 0.00012399 Uhg2 237.8620 88.1464 -1.43215 0.00005 0.00012399 CG17739 1.3614 0.504994 -1.43077 0.0002 0.00046869 CG31380 2.5904 0.962752 -1.42795 0.0001 0.0002417 CG4975 23.5154 8.74187 -1.42759 0.00005 0.00012399 klu 2.4501 0.910883 -1.42752 0.00005 0.00012399 CG43062 5.3515 1.99093 -1.42649 0.00085 0.00184174 CG32686 1.9645 0.732627 -1.42297 0.0274 0.0464788 CG42455 22.8276 8.51788 -1.42221 0.0055 0.0105729 CG31769 2.5592 0.95594 -1.4207 0.0004 0.00090401 noc 15.1562 5.66784 -1.41904 0.00005 0.00012399 CG10399 15.8860 5.96202 -1.41388 0.00725 0.0136633 CG17801 10.0972 3.78977 -1.41377 0.00005 0.00012399 CG4213 1.3925 0.523218 -1.41223 0.00005 0.00012399 Cyp6a21 1.3526 0.509606 -1.40823 0.00205 0.00421386 CG16758 2.9587 1.11601 -1.40661 0.0001 0.0002417 CG32409 101.9680 38.4669 -1.40642 0.00005 0.00012399 CG1628 36.9903 14.0142 -1.40026 0.00005 0.00012399 Uhg1 206.0500 78.2419 -1.39698 0.00005 0.00012399 CG6074 13.7594 5.22718 -1.39631 0.00005 0.00012399 CG11400 6.8262 2.59333 -1.39628 0.00005 0.00012399
167 CG9416 13.5414 5.16671 -1.39006 0.00005 0.00012399 Cyp12a4 9.9331 3.80884 -1.38289 0.00005 0.00012399 Parp 95.8306 36.8815 -1.37759 0.00005 0.00012399 CG43102 2.1751 0.837849 -1.37634 0.00005 0.00012399 stv 8.6079 3.31618 -1.37614 0.00005 0.00012399 cry 22.0123 8.52132 -1.36916 0.00005 0.00012399 Hn 3.7092 1.43692 -1.36814 0.00005 0.00012399 ci 29.5829 11.4716 -1.36669 0.00005 0.00012399 CG31826 6.9976 2.71418 -1.36634 0.00005 0.00012399 CG32174 8.1808 3.17976 -1.36332 0.0267 0.0453894 CG14655 1.5684 0.610274 -1.36174 0.00055 0.00122116 Sug 87.2047 33.9453 -1.3612 0.00005 0.00012399 CG2145 2.6830 1.04728 -1.3572 0.00005 0.00012399 CG34445 15.4020 6.01496 -1.35649 0.00005 0.00012399 Syp 34.3465 13.4299 -1.35471 0.00005 0.00012399 Tequila 9.6063 3.75659 -1.35456 0.00005 0.00012399 CG15822 10.5894 4.14471 -1.35327 0.00005 0.00012399 CG12824 2.6731 1.04942 -1.34891 0.02095 0.0363388 CG10639 15.8141 6.2314 -1.34358 0.00005 0.00012399 Msp300 25.3482 9.98971 -1.34337 0.00005 0.00012399 CG31106 23.5110 9.29684 -1.33852 0.00005 0.00012399 CG3857 7.9053 3.12844 -1.33738 0.00005 0.00012399 Msr-110 23.6116 9.37253 -1.33299 0.00005 0.00012399 CG6912 4.4771 1.77736 -1.33283 0.0003 0.00068977 Cyp6a18 9.8857 3.92498 -1.33265 0.00005 0.00012399 CG1648 37.3471 14.9179 -1.32395 0.00005 0.00012399 CG12721 15.9504 6.377 -1.32264 0.025 0.0427004 Wnt2 4.6333 1.85461 -1.32091 0.00005 0.00012399 CG6767 299.8570 120.046 -1.32068 0.00005 0.00012399 CG10479 6.7496 2.71417 -1.31429 0.00005 0.00012399 Pka-C2 2.8766 1.15759 -1.31322 0.0002 0.00046869 CG17672 1.6144 0.64978 -1.31293 0.003 0.00601599 CR44206 3.8526 1.55458 -1.3093 0.00005 0.00012399 CG8306 11.0332 4.45375 -1.30876 0.00005 0.00012399 metro 26.8077 10.8232 -1.30852 0.00005 0.00012399 CG7120 14.5838 5.89602 -1.30655 0.00005 0.00012399 Lmpt 67.8756 27.4612 -1.3055 0.00005 0.00012399 CG9380 13.8567 5.63219 -1.29882 0.00005 0.00012399 CG15027 74.5052 30.2966 -1.29819 0.00005 0.00012399 Nsun2 62.9764 25.6383 -1.29651 0.0014 0.00294985 Ir93a 1.0508 0.428115 -1.29536 0.0009 0.00194442 kcc 15.0196 6.13053 -1.29277 0.00005 0.00012399 CG17278 2.8594 1.16792 -1.29179 0.00005 0.00012399 mRpL2 38.6941 15.8725 -1.28558 0.00005 0.00012399 CG4872 5.3325 2.18846 -1.2849 0.00455 0.00886178 swi2 7.9550 3.267 -1.2839 0.00015 0.00035649 CG5080 14.8317 6.09627 -1.28268 0.00005 0.00012399 sqa 2.5162 1.03559 -1.28076 0.00005 0.00012399
168 eIF6 124.8490 51.5103 -1.27725 0.00005 0.00012399 CG13324 3.4138 1.40958 -1.27612 0.0093 0.0172137 CG1671 49.6931 20.5638 -1.27294 0.00005 0.00012399 Sur 3.1408 1.3 -1.27262 0.00005 0.00012399 Cyp6g1 14.2379 5.89902 -1.27118 0.00005 0.00012399 Pdp1 32.7443 13.5718 -1.27063 0.00005 0.00012399 sals 2.7525 1.14223 -1.26887 0.00075 0.00163799 Fili 4.3661 1.81221 -1.26857 0.00005 0.00012399 CG1698 1.6387 0.680286 -1.26832 0.0005 0.00111605 CG14687 12.0583 5.01355 -1.26612 0.00005 0.00012399 CG9286 61.8538 25.773 -1.263 0.00005 0.00012399 CG7296 2.1063 0.878832 -1.26106 0.02845 0.0481255 Tsp42Eg 28.4764 11.8914 -1.25985 0.00005 0.00012399 CG10249 24.7798 10.3478 -1.25984 0.00005 0.00012399 CG15546 1.6438 0.686578 -1.25953 0.00075 0.00163799 jp 7.9377 3.31562 -1.25944 0.00005 0.00012399 E(spl)mbeta- 15.9440 6.66798 -1.25769 0.00005 0.00012399 HLH Hsp67Bc 21.4346 8.97957 -1.25522 0.00005 0.00012399 Mdr49 4.5344 1.89967 -1.25516 0.00005 0.00012399 LanB1 76.9196 32.2288 -1.255 0.00005 0.00012399 Fas2 54.3724 22.7958 -1.25411 0.00005 0.00012399 CG6484 1.7271 0.724749 -1.2528 0.00415 0.00812941 sty 17.0357 7.15469 -1.2516 0.00005 0.00012399 comm2 4.4226 1.86019 -1.24945 0.00005 0.00012399 CG5597 11.8212 4.97971 -1.24724 0.0015 0.00314697 Gbeta5 4.3738 1.84494 -1.24532 0.00005 0.00012399 CG4045 43.5567 18.3818 -1.24461 0.00075 0.00163799 CG14292 99.1889 41.9879 -1.2402 0.00005 0.00012399 TpnC47D 10.2593 4.35165 -1.2373 0.00005 0.00012399 GstE12 124.6300 53.0578 -1.23201 0.00005 0.00012399 CG1463 36.2567 15.4364 -1.23192 0.00005 0.00012399 snoRNA:Psi1 32265.4000 13772.3 -1.22822 0.00005 0.00012399 8S-176 CG31675 2.3135 0.987581 -1.22809 0.0027 0.00545162 CG11837 50.0606 21.3781 -1.22754 0.00005 0.00012399 ptc 2.7880 1.19375 -1.22371 0.00005 0.00012399 CG34411 2.5092 1.07542 -1.22232 0.02895 0.0488923 r 68.8989 29.5397 -1.22183 0.00005 0.00012399 Oseg5 3.4346 1.47379 -1.22062 0.0005 0.00111605 mid 9.8309 4.22019 -1.22001 0.00005 0.00012399 ec 5.5723 2.39739 -1.2168 0.00005 0.00012399 CR45621 4.0610 1.74922 -1.21512 0.011 0.0201155 CG6836 2.4026 1.03538 -1.21445 0.00305 0.00610879 CG16799 1.7669 0.761459 -1.21438 0.01915 0.0334698 Ugt35b 14.9016 6.42364 -1.21401 0.00005 0.00012399 CG30195 14.3251 6.18079 -1.21268 0.00005 0.00012399 rgn 61.1833 26.404 -1.21238 0.00005 0.00012399
169 Mp20 114.9240 49.8365 -1.20541 0.00005 0.00012399 CG18003 5.2618 2.28315 -1.20453 0.00005 0.00012399 CR45736 4.6075 1.99954 -1.20433 0.0042 0.00822089 CG1750 11.2727 4.90344 -1.20097 0.00585 0.0111945 CG4749 11.6178 5.05651 -1.20012 0.00005 0.00012399 Shaw 1.0926 0.475551 -1.20008 0.0001 0.0002417 Eap 46.4221 20.279 -1.19483 0.00005 0.00012399 ade3 118.3630 51.87 -1.19025 0.00005 0.00012399 babos 82.0490 36.0147 -1.1879 0.00005 0.00012399 stumps 3.1840 1.39804 -1.18744 0.00005 0.00012399 CG11563 88.7996 39.0056 -1.18687 0.00005 0.00012399 hdly 21.2334 9.32708 -1.18684 0.00005 0.00012399 Sr-CIV 1.9848 0.873611 -1.18395 0.00415 0.00812941 CG32813 7.7620 3.41667 -1.18383 0.00005 0.00012399 Crg-1 3.4958 1.53955 -1.18312 0.0004 0.00090401 CG6329 2.1772 0.959897 -1.18153 0.0019 0.00392473 CG8005 26.4565 11.6831 -1.1792 0.00005 0.00012399 CG7702 16.0888 7.11208 -1.17771 0.0002 0.00046869 CHKov2 5.3526 2.36652 -1.17748 0.00005 0.00012399 CG13900 248.9260 110.132 -1.17648 0.00005 0.00012399 Dbp80 112.7880 49.9275 -1.17571 0.00005 0.00012399 mRpL36 62.2542 27.6461 -1.1711 0.00005 0.00012399 CG32037 2.6759 1.1895 -1.16964 0.0008 0.00174155 CG8128 19.3055 8.58416 -1.16926 0.00005 0.00012399 rau 14.1953 6.32648 -1.16593 0.00005 0.00012399 Alg-2 85.5855 38.1922 -1.16409 0.00005 0.00012399 CG15279 13.1033 5.85221 -1.16287 0.00005 0.00012399 CG3984 3.2570 1.45557 -1.16197 0.0037 0.00730216 CG8525 24.5446 10.9702 -1.16182 0.00005 0.00012399 CG14109 1.9139 0.856239 -1.16039 0.00935 0.0173001 Nop56 374.3640 167.536 -1.15997 0.00005 0.00012399 CG7637 505.2210 226.297 -1.1587 0.00005 0.00012399 Fad2 1.0256 0.459437 -1.15849 0.02775 0.0470316 mod 962.9830 431.484 -1.1582 0.00005 0.00012399 Spn42Dc 3.0883 1.39036 -1.15134 0.00185 0.00382769 CG3817 61.2787 27.6038 -1.15052 0.00005 0.00012399 CG14210 84.6697 38.2439 -1.14662 0.00005 0.00012399 elB 6.0591 2.73854 -1.1457 0.00005 0.00012399 La 261.8150 118.34 -1.14561 0.00005 0.00012399 ASPP 6.4645 2.92425 -1.14447 0.00005 0.00012399 CG15529 14.7701 6.68748 -1.14315 0.00005 0.00012399 CG6724 76.4784 34.6604 -1.14176 0.00005 0.00012399 CG10026 1.8777 0.851784 -1.14037 0.004 0.0078541 Cyp4p3 2.6242 1.19067 -1.14008 0.01045 0.0191865 NHP2 278.2180 126.242 -1.14003 0.00005 0.00012399 CG10341 35.7505 16.2445 -1.13801 0.00005 0.00012399
170 alphagamma- 15.2966 6.95193 -1.13772 0.00005 0.00012399 element:CR3 2865 CG9629 10.3113 4.69352 -1.13549 0.0016 0.00334415 CG2543 3.0204 1.37633 -1.13392 0.00015 0.00035649 CG17514 237.4360 108.203 -1.1338 0.00005 0.00012399 RpLP0-like 172.5580 78.8621 -1.12968 0.00005 0.00012399 vn 1.1310 0.517819 -1.12703 0.00005 0.00012399 CG11858 63.5114 29.1701 -1.12252 0.0245 0.0419079 Pabp2 59.2792 27.2298 -1.12234 0.00005 0.00012399 mbl 45.8397 21.127 -1.11751 0.00005 0.00012399 Nop60B 422.4130 194.993 -1.11523 0.00005 0.00012399 CG1970 139.7610 64.6968 -1.1112 0.00005 0.00012399 snRNA:U1:82 171.6350 79.4783 -1.11071 0.00035 0.0007973 Eb CG11089 150.3020 69.6316 -1.11005 0.00005 0.00012399 CHKov1 4.7622 2.20713 -1.10946 0.00005 0.00012399 l(1)1Bi 76.5674 35.5411 -1.10724 0.00005 0.00012399 by 17.6035 8.17708 -1.10621 0.00005 0.00012399 Rpp30 28.8789 13.4169 -1.10597 0.00005 0.00012399 Tal 161.1370 74.912 -1.10502 0.00005 0.00012399 Tg 18.0919 8.42307 -1.10293 0.00005 0.00012399 CR44201 9.4648 4.41306 -1.10078 0.02165 0.0374469 ligatin 21.3090 9.93829 -1.10039 0.0001 0.0002417 Clk 3.8580 1.80054 -1.09944 0.00005 0.00012399 nclb 96.0554 44.8531 -1.09866 0.00005 0.00012399 CG12744 64.5089 30.1307 -1.09826 0.00775 0.0145343 plum 15.5990 7.32215 -1.09111 0.00005 0.00012399 Ac76E 2.9585 1.3893 -1.09049 0.00005 0.00012399 CG7180 1.6691 0.784359 -1.0895 0.0001 0.0002417 qin 36.0742 16.9666 -1.08827 0.00005 0.00012399 CG8398 1.7646 0.831881 -1.0849 0.0006 0.00132671 Pdfr 1.2630 0.595488 -1.08472 0.0006 0.00132671 CG32082 2.4216 1.14204 -1.08438 0.00005 0.00012399 CG45263 3.0969 1.46155 -1.08333 0.00005 0.00012399 vito 40.7305 19.2223 -1.08333 0.00005 0.00012399 CG6574 16.6801 7.87759 -1.0823 0.00005 0.00012399 RluA-2 18.6140 8.81242 -1.07878 0.00005 0.00012399 CG10286 75.1199 35.5921 -1.07764 0.00005 0.00012399 CG1092 6.0402 2.86287 -1.07715 0.00005 0.00012399 l(1)G0020 64.6760 30.6752 -1.07616 0.00005 0.00012399 Six4 39.7076 18.8836 -1.07228 0.00005 0.00012399 CG10516 15.6163 7.43119 -1.07138 0.00005 0.00012399 CG9107 34.7649 16.5474 -1.07103 0.00005 0.00012399 CG16790 8.3707 3.98434 -1.07101 0.0002 0.00046869 CG42566 47.2912 22.5141 -1.07074 0.00005 0.00012399 Mur29B 1.2912 0.615951 -1.06785 0.00585 0.0111945 CG18522 21.3350 10.1852 -1.06675 0.00005 0.00012399
171 CG9098 4.6431 2.2183 -1.06564 0.00005 0.00012399 GluRIIA 1.0779 0.515028 -1.06554 0.0011 0.00235148 CG6426 8.6808 4.1483 -1.06531 0.0005 0.00111605 Swim 36.3605 17.4055 -1.06283 0.00195 0.00402261 CG18600 60.3709 28.9181 -1.06188 0.00005 0.00012399 CG12909 80.1636 38.4036 -1.0617 0.00005 0.00012399 CG15365 2.0147 0.967888 -1.05761 0.00005 0.00012399 CG8353 3.5253 1.69384 -1.05744 0.01395 0.0250386 CG7272 18.6194 8.98025 -1.05198 0.0001 0.0002417 CG11555 71.1441 34.3301 -1.05127 0.005 0.00967666 Eip63F-1 7.7975 3.76679 -1.04967 0.00005 0.00012399 CG8317 1.1829 0.571758 -1.04889 0.02355 0.0404207 CG34396 2.8566 1.38124 -1.04834 0.0212 0.0367416 Dsp1 122.3640 59.2649 -1.04593 0.00005 0.00012399 ct 8.4881 4.125 -1.04105 0.00005 0.00012399 Hml 3.1043 1.51115 -1.03861 0.00005 0.00012399 SC35 135.4420 66.0606 -1.03582 0.01915 0.0334698 CR45600 1.8778 0.916787 -1.03435 0.00895 0.0166093 CG5002 6.2372 3.04865 -1.03273 0.00005 0.00012399 CG12512 29.4815 14.4164 -1.0321 0.00005 0.00012399 CG3621 81.6629 39.9728 -1.03066 0.00005 0.00012399 CG15739 4.8025 2.35141 -1.03025 0.00015 0.00035649 CG6125 18.7182 9.16651 -1.02999 0.00005 0.00012399 CG30438 2.7820 1.36286 -1.02947 0.0008 0.00174155 FK506-bp1 514.2740 252.57 -1.02585 0.00005 0.00012399 CG11279 67.5334 33.1683 -1.02579 0.00005 0.00012399 bowl 13.4192 6.59217 -1.02547 0.00005 0.00012399 Rrp4 43.2778 21.3122 -1.02194 0.00005 0.00012399 CG11891 3.2646 1.60838 -1.02131 0.01055 0.0193509 CG4159 28.8658 14.2474 -1.01866 0.00025 0.00058003 CG15096 12.0685 5.97365 -1.01456 0.00005 0.00012399 Trim9 2.1388 1.0588 -1.01436 0.0001 0.0002417 Nmd3 77.1341 38.2051 -1.0136 0.00005 0.00012399 CG34134 10.3702 5.13694 -1.01347 0.00235 0.00478707 sick 7.5808 3.75524 -1.01344 0.00225 0.00459619 Cct5 299.6090 148.441 -1.01319 0.00005 0.00012399 grass 1.4056 0.69718 -1.01161 0.0188 0.0329194 qless 12.9144 6.40715 -1.01122 0.00005 0.00012399 CG12288 79.2315 39.3145 -1.01101 0.0028 0.00564312 CG15155 21.3105 10.5784 -1.01044 0.00005 0.00012399 p130CAS 14.1175 7.03025 -1.00584 0.00005 0.00012399 Rsf1 23.2925 11.6 -1.00574 0.00005 0.00012399 H15 1.7537 0.87346 -1.00555 0.0005 0.00111605 CG5527 1.0164 0.506535 -1.00474 0.0061 0.011635 MFS17 56.9366 28.3985 -1.00354 0.00005 0.00012399 CG16986 22.8115 11.4046 -1.00015 0.0034 0.00675493
172 REFERENCES
Adoue, V., Binet, B., Malbec, A., Fourquet, J., Romagnoli, P., van Meerwijk, J. P. M., … Joffre, O. P. (2019). The Histone Methyltransferase SETDB1 Controls T Helper Cell Lineage Integrity by Repressing Endogenous Retroviruses. Immunity. http://doi.org/10.1016/J.IMMUNI.2019.01.003 Ahmed, K., Dehghani, H., Rugg-Gunn, P., Fussner, E., Rossant, J., & Bazett- Jones, D. P. (2010). Global Chromatin Architecture Reflects Pluripotency and Lineage Commitment in the Early Mouse Embryo. PLoS ONE, 5(5), e10531. http://doi.org/10.1371/journal.pone.0010531 Ameele van den, K. J., & Brand, A. H. (2019). TaDa! Analysing cell type-specific chromatin in vivo with Targeted DamID. Current Opinion in Neurobiology, 56, 160–166. http://doi.org/10.1016/J.CONB.2019.01.021 Baena-Lopez Luis Alberto, Alexandre Cyrille, Mitchell Alice, P. L. and V. J.-P. (2013). Accelerated homologous recombination and subsequent genome modification in Drosophila. Development. http://doi.org/10.1242/dev.100933 Baert, Y., Rombaut, C., & Goossens, E. (2017). Scaffold-Based and Scaffold- Free Testicular Organoids from Primary Human Testicular Cells (pp. 1–8). Humana Press. http://doi.org/10.1007/7651_2017_48 Becker, J. S., McCarthy, R. L., Sidoli, S., Donahue, G., Kaeding, K. E., He, Z., … Zaret, K. S. (2017). Genomic and Proteomic Resolution of Heterochromatin and Its Restriction of Alternate Fate Genes. Molecular Cell, 68(6), 1023– 1037.e15. http://doi.org/10.1016/J.MOLCEL.2017.11.030 Becker, J. S., Nicetto, D., & Zaret, K. S. (2016). H3K9me3-Dependent Heterochromatin: Barrier to Cell Fate Changes. Trends in Genetics. http://doi.org/10.1016/j.tig.2015.11.001 Bellen, H. J., Levis, R. W., Liao, G., He, Y., Carlson, J. W., Tsang, G., … Spradling, A. C. (2004). The BDGP gene disruption project: single transposon insertions associated with 40% of Drosophila genes. Genetics, 167(2), 761–81. http://doi.org/10.1534/genetics.104.026427 Bopp, D., Horabin, J. I., Lersch, R. a, Cline, T. W., & Schedl, P. (1993). Expression of the Sex-lethal gene is controlled at multiple levels during Drosophila oogenesis. Development (Cambridge, England), 118, 797–812. Breiling, A., Turner, B. M., Bianchi, M. E., & Orlando, V. (2001). General transcription factors bind promoters repressed by Polycomb group proteins. Nature, 412(6847), 651–655. http://doi.org/10.1038/35088090 Brower-Toland, B., Riddle, N. C., Jiang, H., Huisinga, K. L., & Elgin, S. C. R. (2009). Multiple SET Methyltransferases Are Required to Maintain Normal Heterochromatin Domains in the Genome of Drosophila melanogaster. Genetics, 181, 1303–1319. http://doi.org/10.1534/genetics.108.100271 Casper, A. . and V. D. M. (2009). The establishment of sexual identity in the Drosophila germline. Development (Cambridge, England), 136(22), 3821– 3830. http://doi.org/10.1242/dev.042374 Casper, A., & Van Doren, M. (2006). The control of sexual identity in the Drosophila germline. Development (Cambridge, England), 133(15), 2783– 91. http://doi.org/10.1242/dev.02415
173 Chau, J., Kulnane, L. S., & Salz, H. K. (2009). Sex-lethal facilitates the transition from germline stem cell to committed daughter cell in the Drosophila ovary. Genetics, 182(1), 121–32. http://doi.org/10.1534/genetics.109.100693 Chau, J., Kulnane, L. S., & Salz, H. K. (2012). Sex-lethal enables germline stem cell differentiation by down-regulating Nanos protein levels during Drosophila oogenesis. Proceedings of the National Academy of Sciences of the United States of America, 109(24), 9465–70. http://doi.org/10.1073/pnas.1120473109 Chen, J., Liu, H., Liu, J., Qi, J., Wei, B., Yang, J., … Pei, D. (2013). H3K9 methylation is a barrier during somatic cell reprogramming into iPSCs. Nature Genetics, 45(1), 34–42. http://doi.org/10.1038/ng.2491 Chintapalli, V. R., Wang, J., & Dow, J. A. T. (2007). Using FlyAtlas to identify better Drosophila melanogaster models of human disease. Nature Genetics, 39(6), 715–720. http://doi.org/10.1038/ng2049 Chung, C. C., Hildebrandt, , Christian P. Kratz, C., Turnbull, , Anne C. Bakken, D. Timothy Bishop, M. B., Cook1, R. Loren Erickson, Sophie D. Fosså, Kevin B. Jacobs, L. A. K., … J. Chanock1, and K. L. N. (2013). Meta-analysis identifies four new loci associated with testicular germ cell tumor. Nature Genetics, 45(6), 1–18. http://doi.org/10.1038/ng.2634.Meta-analysis Clough, E., Moon, W., Wang, S., Smith, K., & Hazelrigg, T. (2007). Histone methylation is required for oogenesis in Drosophila. Development (Cambridge, England), 134(1), 157–165. http://doi.org/10.1242/dev.02698 Clough, E., Tedeschi, T., & Hazelrigg, T. (2014). Epigenetic regulation of oogenesis and germ stem cell maintenance by the Drosophila histone methyltransferase Eggless/dSetDB1. Developmental Biology, 388(2), 181– 191. http://doi.org/10.1016/j.ydbio.2014.01.014 Consortium, T. E. P. (2012). An integrated encyclopedia of DNA elements in the human genome. Nature, 489(7414), 57–74. http://doi.org/10.1038/nature11247 Cools, M., Looijenga, L. H. J., Wolffenbuttel, K. P., & Drop, S. L. S. (2009). Disorders of sex development: update on the genetic background, terminology and risk for the development of germ cell tumors. World Journal of Pediatrics, 5(2), 93–102. http://doi.org/10.1007/s12519-009-0020-7 Dawson, E. P., Lanza, D. G., Webster, N. J., Benton, S. M., Suetake, I., & Heaney, J. D. (2018). Delayed male germ cell sex-specification permits transition into embryonal carcinoma cells with features of primed pluripotency. Development (Cambridge, England), 145(6), dev156612. http://doi.org/10.1242/dev.156612 Dellino, G. I., Schwartz, Y. B., Farkas, G., McCabe, D., Elgin, S. C. ., & Pirrotta, V. (2004). Polycomb Silencing Blocks Transcription Initiation. Molecular Cell, 13(6), 887–893. http://doi.org/10.1016/S1097-2765(04)00128-5 DeLuca, S. Z., & Spradling, A. C. (2018). Efficient Expression of Genes in the Drosophila Germline Using a UAS Promoter Free of Interference by Hsp70 piRNAs. Genetics, 209(2), 381–387. http://doi.org/10.1534/genetics.118.300874 Demarco, R. S., Eikenes, Å. H., Haglund, K., & Jones, D. L. (2014). Investigating
174 spermatogenesis in Drosophila melanogaster. Methods (San Diego, Calif.), 68(1), 218–27. http://doi.org/10.1016/j.ymeth.2014.04.020 Du, D., Katsuno, Y., Meyer, D., Budi, E. H., Chen, S.-H., Koeppen, H., … Derynck, R. (2018). Smad3-mediated recruitment of the methyltransferase SETDB1/ESET controls Snail1 expression and epithelial-mesenchymal transition. EMBO Reports, 19(1), 135–155. http://doi.org/10.15252/embr.201744250 Dubiel, W., Dubiel, D., Wolf, D. A., & Naumann, M. (2018). Cullin 3-Based Ubiquitin Ligases as Master Regulators of Mammalian Cell Differentiation. Trends in Biochemical Sciences, 43(2), 95–107. http://doi.org/10.1016/j.tibs.2017.11.010 Efroni, S., Duttagupta, R., Cheng, J., Dehghani, H., Hoeppner, D. J., Dash, C., … Meshorer, E. (2008). Global Transcription in Pluripotent Embryonic Stem Cells. Cell Stem Cell, 2(5), 437–447. http://doi.org/10.1016/J.STEM.2008.03.021 Eissenberg, J. C., & Elgin, S. C. R. (2014). HP1a: a structural chromosomal protein regulating transcription. Trends in Genetics, 30(3), 103–110. http://doi.org/10.1016/J.TIG.2014.01.002 Elgin, S. C. R., & Reuter, G. (2013). Formation , and Gene Silencing in Drosophila. Cold Spring Harb. Perspect Biol., (5), 1–26. http://doi.org/10.1101/cshperspect.a017780 Eskeland, R., Czermin, B., Boeke, J., Bonaldi, T., Regula, J. T., & Imhof, A. (2004). The N-Terminus of Drosophila SU(VAR)3-9 Mediates Dimerization and Regulates Its Methyltransferase Activity †. http://doi.org/10.1021/bi035964s Fagerberg, L., Hallström, B. M., Oksvold, P., Kampf, C., Djureinovic, D., Odeberg, J., … Uhlén, M. (2014). Analysis of the human tissue-specific expression by genome-wide integration of transcriptomics and antibody- based proteomics. Molecular & Cellular Proteomics : MCP, 13(2), 397–406. http://doi.org/10.1074/mcp.M113.035600 Garcia-Moreno, S. A., Futtner, C. R., Salamone, I. M., Gonen, N., Lovell-Badge, R., & Maatouk, D. M. (2018). Gonadal supporting cells acquire sex-specific chromatin landscapes during mammalian sex determination. Developmental Biology. http://doi.org/10.1016/j.ydbio.2018.12.023 Gatchalian, J., Fütterer, A., Rothbart, S. B., Tong, Q., Rincon-Arano, H., Sánchez de Diego, A., … Kutateladze, T. G. (2013). Dido3 PHD Modulates Cell Differentiation and Division. Cell Reports, 4(1), 148–158. http://doi.org/10.1016/J.CELREP.2013.06.014 Ghosh, S., & Lasko, P. (2015). Loss-of-Function Analysis Reveals Distinct Requirements of the Translation Initiation Factors eIF4E, eIF4E-3, eIF4G and eIF4G2 in Drosophila Spermatogenesis. PLOS ONE, 10(4), e0122519. http://doi.org/10.1371/journal.pone.0122519 Gonzalez-Exposito, R., Merino, M., & Aguayo, C. (2016). Molecular biology of testicular germ cell tumors. Clinical and Translational Oncology, 18(6), 550– 556. http://doi.org/10.1007/s12094-015-1423-7 Gratz, S. J., Ukken, F. P., Rubinstein, C. D., Thiede, G., Donohue, L. K.,
175 Cummings, A. M., & O’Connor-Giles, K. M. (2014). Highly specific and efficient CRISPR/Cas9-catalyzed homology-directed repair in Drosophila. Genetics, 196(4), 961–71. http://doi.org/10.1534/genetics.113.160713 Grmai, L., Hudry, B., Miguel-Aliaga, I., & Bach, E. A. (2018). Chinmo prevents transformer alternative splicing to maintain male sex identity. PLOS Genetics, 14(2), e1007203. http://doi.org/10.1371/journal.pgen.1007203 Heaney, J. D., Anderson, E. L., Michelson, M. V., Zechel, J. L., Conrad, P. A., Page, D. C., & Nadeau, J. H. (2012). Germ cell pluripotency, premature differentiation and susceptibility to testicular teratomas in mice. Development, 139(9), 1577–1586. http://doi.org/10.1242/dev.076851 Hemminki, K., & Li, X. (2004). Familial risk in testicular cancer as a clue to a heritable and environmental aetiology. British Journal of Cancer, 90(9), 1765–1770. http://doi.org/10.1038/sj.bjc.6601714 Hernandez, G., Han, H., Gandin, V., Fabian, L., Ferreira, T., Zuberek, J., … Lasko, P. (2012). Eukaryotic initiation factor 4E-3 is essential for meiotic chromosome segregation, cytokinesis and male fertility in Drosophila. Development, 139(17), 3211–3220. http://doi.org/10.1242/dev.073122 Hu, Y., Comjean, A., Mohr, S. E., Consortium, T. F., & Perrimon, N. (2017). Gene2Function: An Integrated Online Resource for Gene Function Discovery. http://doi.org/10.1534/g3.117.043885 Hudry, B., Khadayate, S., & Miguel-Aliaga, I. (2016). The sexual identity of adult intestinal stem cells controls organ size and plasticity. Nature, 530(7590), 344–348. http://doi.org/10.1038/nature16953 Ito, H., Sato, K., Koganezawa, M., Ote, M., Matsumoto, K., Hama, C., & Yamamoto, D. (2012). Fruitless recruits two antagonistic chromatin factors to establish single-neuron sexual dimorphism. Cell, 149(6), 1327–1338. http://doi.org/10.1016/j.cell.2012.04.025 Jiang, Y., Loh, Y.-H. E., Rajarajan, P., Hirayama, T., Liao, W., Kassim, B. S., … Akbarian, S. (2017). The methyltransferase SETDB1 regulates a large neuron-specific topological chromatin domain. Nature Genetics. http://doi.org/10.1038/ng.3906 Jørgensen, A., Lindhardt Johansen, M., Juul, A., Skakkebaek, N. E., Main, K. M., & Rajpert-De Meyts, E. (2015). Pathogenesis of germ cell neoplasia in testicular dysgenesis and disorders of sex development. Seminars in Cell & Developmental Biology, 45, 124–137. http://doi.org/10.1016/J.SEMCDB.2015.09.013 Jorgensen, A., Nielsen, J. E., Blomberg Jensen, M., Graem, N., & Rajpert-De Meyts, E. (2012). Analysis of meiosis regulators in human gonads: a sexually dimorphic spatio-temporal expression pattern suggests involvement of DMRT1 in meiotic entry. Molecular Human Reproduction, 18(11), 523– 534. http://doi.org/10.1093/molehr/gas030 Jung, D., Xiong, J., Ye, M., Qin, X., Li, L., Cheng, S., … Kee, K. (2017). In vitro differentiation of human embryonic stem cells into ovarian follicle-like cells. Nature Communications, 8, 15680. http://doi.org/10.1038/ncomms15680 Juniarto, A. Z., Setyawati, B. a, Miranti, I. P., Santosa, A., Hersmus, R., Stoop, H., … Looijenga, L. H. J. (2013). Gonadal malignancy in 13 consecutive
176 collected patients with disorders of sex development (DSD) from Semarang (Indonesia). Journal of Clinical Pathology, 66(3), 198–204. http://doi.org/10.1136/jclinpath-2012-201062 Kai, T., Williams, D., & Spradling, A. C. (2005). The expression profile of purified Drosophila germline stem cells. Genomes & Developmental Control. http://doi.org/10.1016/j.ydbio.2005.04.018 Kearse, M. G., Chen, A. S., & Ware, V. C. (2011). Expression of ribosomal protein L22e family members in Drosophila melanogaster: rpL22-like is differentially expressed and alternatively spliced. Nucleic Acids Research, 39(7), 2701–2716. http://doi.org/10.1093/nar/gkq1218 Kharazmi, E., Hemminki, K., Pukkala, E., Sundquist, K., Tryggvadottir, L., Tretli, S., … Fallah, M. (2015). Cancer Risk in Relatives of Testicular Cancer Patients by Histology Type and Age at Diagnosis: A Joint Study from Five Nordic Countries. European Urology, 68(2), 283–289. http://doi.org/10.1016/J.EURURO.2014.12.031 Klattenhoff, C., Xi, H., Li, C., Lee, S., Xu, J., Khurana, J. S., … Theurkauf, W. E. (2009). The Drosophila HP1 Homolog Rhino Is Required for Transposon Silencing and piRNA Production by Dual-Strand Clusters. Cell, 138(6), 1137–1149. http://doi.org/10.1016/J.CELL.2009.07.014 Koch, C. M., Honemann-Capito, M., Egger-Adam, D., Wodarz, A., & Akhtar, A. (2009). Windei, the Drosophila Homolog of mAM/MCAF1, Is an Essential Cofactor of the H3K9 Methyl Transferase dSETDB1/Eggless in Germ Line Development. PLoS Genet, 5(9). http://doi.org/10.1371/journal.pgen.1000644 Koide, S., Oshima, M., Takubo, K., Yamazaki, S., Nitta, E., Saraya, A., … Iwama, A. (2016). Setdb1 maintains hematopoietic stem and progenitor cells by restricting the ectopic activation of nonhematopoietic genes. Blood, 128(5), 638–49. http://doi.org/10.1182/blood-2016-01-694810 Koster, R., Mitra, N., D’Andrea, K., Vardhanabhuti, S., Chung, C. C., Wang, Z., … Kanetsky, P. A. (2014). Pathway-based analysis of GWAs data identifies association of sex determination genes with susceptibility to testicular germ cell tumors. Human Molecular Genetics, 23(22), 6061–6068. http://doi.org/10.1093/hmg/ddu305 Krentz, A. D., Murphy, M. W., Kim, S., Cook, M. S., Capel, B., Zhu, R., … Zarkower, D. (2009). The DM domain protein DMRT1 is a dose-sensitive regulator of fetal germ cell proliferation and pluripotency. Retrieved from www.pnas.org/cgi/content/full/ Langmead, B., & Salzberg, S. L. (2012). Fast gapped-read alignment with Bowtie 2. Nature Methods, 9(4), 357–359. http://doi.org/10.1038/nmeth.1923 Leader, D. P., Krause, S. A., Pandit, A., Davies, S. A., & Dow, J. A. T. (2018). FlyAtlas 2: a new version of the Drosophila melanogaster expression atlas with RNA-Seq, miRNA-Seq and sex-specific data. Nucleic Acids Research, 46(D1), D809–D815. http://doi.org/10.1093/nar/gkx976 Lee, T. I., Johnstone, S. E., & Young, R. a. (2010). Chromatin imunoprecipitation and microarray-based analysis of protein location. Nationa Institute of Health, 1(2), 729–748. http://doi.org/10.1038/nprot.2006.98.Chromatin
177 Lesch, B. J., & Page, D. C. (2012). Genetics of germ cell development. Nature Reviews Genetics, 13(11), 781–794. http://doi.org/10.1038/nrg3294 Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., … Subgroup, 1000 Genome Project Data Processing. (2009). The Sequence Alignment/Map format and SAMtools. Bioinformatics, 25(16), 2078. http://doi.org/10.1093/BIOINFORMATICS/BTP352 Litchfield, K., Levy, M., Huddart, R. A., Shipley, J., & Turnbull, C. (2016). The genomic landscape of testicular germ cell tumours: from susceptibility to treatment. Nature Reviews Urology, 13(7), 409–419. http://doi.org/10.1038/nrurol.2016.107 Liu, J., Magri, L., Zhang, F., Marsh, N. O., Albrecht, S., Huynh, J. L., … Casaccia, P. (2015). Chromatin landscape defined by repressive histone methylation during oligodendrocyte differentiation. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 35(1), 352–65. http://doi.org/10.1523/JNEUROSCI.2606-14.2015 Liu, S., Brind’Amour, J., Karimi, M. M., Shirane, K., Bogutz, A., Lefebvre, L., … Lorincz, M. C. (2014). Setdb1 is required for germline development and silencing of H3K9me3-marked endogenous retroviruses in primordial germ cells. Genes & Development, 28(18), 2041–55. http://doi.org/10.1101/gad.244848.114 Livak, K. J., & Schmittgen, T. D. (2001). Analysis of Relative Gene Expression Data Using Real-Time Quantitative PCR and the 2 C T Method. METHODS, 25, 402–408. http://doi.org/10.1006/meth.2001.1262 Ma, Q., Wawersik, M., & Matunis, E. L. (2014). The Jak-STAT Target Chinmo Prevents Sex Transformation of Adult Stem Cells in the Drosophila Testis Niche. Developmental Cell, 31(4), 474–86. http://doi.org/10.1016/j.devcel.2014.10.004 Marchler-Bauer, A., Bo, Y., Han, L., He, J., Lanczycki, C. J., Lu, S., … Bryant, S. H. (2017). CDD/SPARCLE: functional classification of proteins via subfamily domain architectures. Nucleic Acids Research, 45(D1), D200–D203. http://doi.org/10.1093/nar/gkw1129 MARSH, J. L., & WIESCHAUS, E. (1978). Is sex determination in germ line and soma controlled by separate genetic mechanisms? Nature, 272(5650), 249– 251. http://doi.org/10.1038/272249a0 Matson, C. K., Murphy, M. W., Sarver, A. L., Griswold, M. D., Bardwell, V. J., & Zarkower, D. (2011). DMRT1 prevents female reprogramming in the postnatal mammalian testis. Nature, 476(7358), 101–4. http://doi.org/10.1038/nature10239 McLaughlin, J. M., & Bratu, D. P. (2015). Drosophila melanogaster Oogenesis: An Overview (pp. 1–20). Humana Press, New York, NY. http://doi.org/10.1007/978-1-4939-2851-4_1 Mochizuki, K., Tando, Y., Sekinaka, T., Otsuka, K., Hayashi, Y., Kobayashi, H., … Matsui, Y. (2018). SETDB1 is essential for mouse primordial germ cell fate determination by ensuring BMP signaling. Development (Cambridge, England), dev.164160. http://doi.org/10.1242/dev.164160 Mohn, F., Sienski, G., Handler, D., & Brennecke, J. (2014). The Rhino-Deadlock-
178 Cutoff complex licenses noncanonical transcription of dual-strand piRNA clusters in Drosophila. Cell, 157(6), 1364–1379. http://doi.org/10.1016/j.cell.2014.04.031 Moschall, R., Gaik, M., & Medenbach, J. (2017). Promiscuity in post- transcriptional control of gene expression: Drosophila sex-lethal and its regulatory partnerships. FEBS Letters, 591(11), 1471–1488. http://doi.org/10.1002/1873-3468.12652 Mozzetta, C., Boyarchuk, E., Pontis, J., & Ait-Si-Ali, S. (2015). Sound of silence: the properties and functions of repressive Lys methyltransferases. Nature Publishing Group, 16. http://doi.org/10.1038/nrm4029 Murray, S. M., Yang, S. Y., & Van Doren, M. (2010). Germ cell sex determination: a collaboration between soma and germline. Current Opinion in Cell Biology, 22(6), 722–729. http://doi.org/10.1016/J.CEB.2010.09.006 Nagengast, A. a, Stitzinger, S. M., Tseng, C.-H., Mount, S. M., & Salz, H. K. (2003). Sex-lethal splicing autoregulation in vivo: interactions between SEX- LETHAL, the U1 snRNP and U2AF underlie male exon skipping. Development (Cambridge, England), 130(3), 463–471. http://doi.org/10.1242/dev.00274 Nakayama Jun-ichi, Rice Judd C., Strahl Brian D., Allis David C., G. S. I. . (2001). Role of Histone H3 Lysine 9 Methylation in Epigenetic Control of Heterochromatin Assembly. Science, 292(110–113). http://doi.org/10.1126/science.1060118 Nicetto, D., Donahue, G., Jain, T., Peng, T., Sidoli, S., Sheng, L., … Zaret, K. S. (2019). H3K9me3-heterochromatin loss at protein-coding genes enables developmental lineage specification. Science (New York, N.Y.), eaau0583. http://doi.org/10.1126/science.aau0583 Ninova Maria, Ariel Chen Yung-Chia, Godneeva Baira, Rogers Alicia K., Luo Yicheng, A. A. A. and T. K. F. (2019). The SUMO ligase Su(var)2-10 links piRNA-guided target recognition to chromatin silencing. BioRxiv, 2–10. http://doi.org/http://dx.doi.org/10.1101/533091 Ninova Maria, Godneeva Baira, Chen Yung-Chia Ariel, Luo Yicheng, Prakash Sharan J., Jankovics Ferenc, Erdélyi Miklos, Tóth Katalin Fejes, and A. A. A. (2019). The SUMO ligase Su(var)2-10 controls eu- and heterochromatic gene expression via establishment of H3K9 trimethylation and negative feedback regulation. BioRxiv, 2–10. http://doi.org/http://dx.doi.org/10.1101/533232 Ota, R., Morita, S., Sato, M., Shigenobu, S., Hayashi, M., & Kobayashi, S. (2017). Transcripts immunoprecipitated with Sxl protein in primordial germ cells of Drosophila embryos. Development, Growth & Differentiation, 59(9), 713–723. http://doi.org/10.1111/dgd.12408 Parisi, M., Nuttall, R., Edwards, P., Minor, J., Naiman, D., Lü, J., … Oliver, B. (2004). A survey of ovary-, testis-, and soma-biased gene expression in Drosophila melanogaster adults. Genome Biology, 5(6), R40. http://doi.org/10.1186/gb-2004-5-6-r40 Pekowska, A., Benoukraf, T., Ferrier, P., & Spicuglia, S. (2010). A unique H3K4me2 profile marks tissue-specific gene regulation. Genome Research,
179 20(11), 1493–502. http://doi.org/10.1101/gr.109389.110 Pendergraft, S. S., Sadri-Ardekani, H., Atala, A., & Bishop, C. E. (2017). Three- dimensional testicular organoid: a novel tool for the study of human spermatogenesis and gonadotoxicity in vitro†. Biology of Reproduction, 96(3), 720–732. http://doi.org/10.1095/biolreprod.116.143446 Penke, T. J. R., McKay, D. J., Strahl, B. D., Matera, A. G., & Duronio, R. J. (2016). Direct interrogation of the role of H3K9 in metazoan heterochromatin function. Genes & Development, 30(16), 1866–80. http://doi.org/10.1101/gad.286278.116 Pinskaya, M., & Morillon, A. (2009). Histone H3 Lysine 4 di-methylation: A novel mark for transcriptional fidelity? http://doi.org/10.4161/epi.4.5.9369 Rangan, P., Malone, C. D., Navarro, C., Newbold, S. P., Hayes, P. S., Sachidanandam, R., … Lehmann, R. (2011). Report piRNA Production Requires Heterochromatin Formation in Drosophila. Current Biology, 21, 1373–1379. http://doi.org/10.1016/j.cub.2011.06.057 Ratnu, V. S., Emami, M. R., & Bredy, T. W. (2017). Genetic and epigenetic factors underlying sex differences in the regulation of gene expression in the brain. Journal of Neuroscience Research, 95(1–2), 301–310. http://doi.org/10.1002/jnr.23886 Raymond, C. S., Murphy, M. W., O’Sullivan, M. G., Bardwell, V. J., & Zarkower, D. (2000). Dmrt1, a gene related to worm and fly sexual regulators, is required for mammalian testis differentiation. Genes & Development, 14(20), 2587–2595. http://doi.org/10.1101/GAD.834100 Regan Jennifer C., Khericha Mobina , Dobson Adam J., Bolukbasi Ekin , Rattanavirotkul Nattaphong, and P. L. (2016). 1 Sex difference in pathology of the ageing gut mediates the greater response 2 of female lifespan to dietary restriction. ELife, 1–16. http://doi.org/10.7554/eLife.10956 Roignant Jean-Yves, Legent Kevin, J. F. and T. J. E. (2010). The transcriptional co-factor Chip acts with LIM-homeodomain proteins to set the boundary of the eye field in Drosophila. Development, 137, 273–281. http://doi.org/10.1242/dev.041244 Rørth, P. (1998). Gal4 in the Drosophila female germline. Mechanisms of Development, 78(1–2), 113–118. http://doi.org/10.1016/S0925- 4773(98)00157-9 Rotem, A., Ram, O., Shoresh, N., Sperling, R. A., Goren, A., Weitz, D. A., & Bernstein, B. E. (2015). Single-cell ChIP-seq reveals cell subpopulations defined by chromatin state. Nature Biotechnology, 33(11), 1165–1172. http://doi.org/10.1038/nbt.3383 Ruark, E., Seal, S., McDonald, H., Zhang, F., Elliot, A., Lau, K., … Turnbull, C. (2013). Identification of nine new susceptibility loci for testicular cancer, including variants near DAZL and PRDM14. Nature Genetics, 45(6), 686– 689. http://doi.org/10.1038/ng.2635 Sakashita, A., Kawabata, Y., Jincho, Y., Tajima, S., Kumamoto, S., Kobayashi, H., … Kono, T. (2015). Sex Specification and Heterogeneity of Primordial Germ Cells in Mice. PLOS ONE, 10(12), e0144836. http://doi.org/10.1371/journal.pone.0144836
180 Salz, H. K., Dawson, E. P., & Heaney, J. D. (2017). Germ cell tumors: Insights from the Drosophila ovary and the mouse testis. Molecular Reproduction and Development, 84(3), 200–211. http://doi.org/10.1002/mrd.22779 Salz, H. K., & Erickson, J. W. (2010). Sex determination in Drosophila: The view from the top. Fly, 4(1), 60–70. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/20160499 Sanchez, R., & Zhou, M.-M. (n.d.). The PHD Finger: A Versatile Epigenome Reader The PHD finger as a sequence-specific histone recognition protein module. http://doi.org/10.1016/j.tibs.2011.03.005 Schultz, D. C., Ayyanathan, K., Negorev, D., Maul, G. G., & Rauscher, F. J. (2002). SETDB1: a novel KAP-1-associated histone H3, lysine 9-specific methyltransferase that contributes to HP1-mediated silencing of euchromatic genes by KRAB zinc-finger proteins. Genes & Development, 16(8), 919–32. http://doi.org/10.1101/gad.973302 Schüpbach, T. (1982). Autosomal mutations that interfere with sex determination in somatic cells of Drosophila have no direct effect on the germline. Developmental Biology, 89(1), 117–127. http://doi.org/10.1016/0012- 1606(82)90300-1 Seum, C., Reo, E., Peng, H., Rauscher Iii, F. J., Spierer, P., & Bontron, S. V. (2007). Drosophila SETDB1 Is Required for Chromosome 4 Silencing. PLoS Genetics, 3(5). http://doi.org/10.1371/journal Shalaby, N. A., Sayed, R., Zhang, Q., Scoggin, S., Eliazer, S., Rothenfluh, A., & Buszczak, M. (2017). Systematic discovery of genetic modulation by Jumonji histone demethylases in Drosophila. Scientific Reports, 7(1), 5240. http://doi.org/10.1038/s41598-017-05004-w Shan, L., Wu, C., Chen, D., Hou, L., Li, X., Wang, L., … Wang, Z. (2017). Regulators of alternative polyadenylation operate at the transition from mitosis to meiosis. http://doi.org/10.1016/j.jgg.2016.12.007 Shapiro-Kulnane, L., Smolko, A. E., & Salz, H. K. (2015). Maintenance of Drosophila germline stem cell sexual identity in oogenesis and tumorigenesis. Development , 142(6), 1073–1082. http://doi.org/10.1242/dev.116590 Shevelyov, Y. Y., Lavrov, S. A., Mikhaylova, L. M., Nurminsky, I. D., Kulathinal, R. J., Egorova, K. S., … Nurminsky, D. I. (2009). The B-type lamin is required for somatic repression of testis-specific gene clusters. Proceedings of the National Academy of Sciences of the United States of America, 106(9), 3282–7. http://doi.org/10.1073/pnas.0811933106 Sienski, G., Batki, J., Senti, K.-A., Dönertas, D., Tirian, L., Meixner, K., & Brennecke, J. (2015). Silencio/CG9754 connects the Piwi-piRNA complex to the cellular heterochromatin machinery. Genes & Development, 29(21), 2258–71. http://doi.org/10.1101/gad.271908.115 Smolko, A. E., Shapiro-Kulnane, L., & Salz, H. K. (2018). The H3K9 methyltransferase SETDB1 maintains female identity in Drosophila germ cells. Nature Communications, 9(1), 4155. http://doi.org/10.1038/s41467- 018-06697-x Soto-feliciano, Y. M., Bartlebaugh, J. M. E., Liu, Y., Sánchez-rivera, F. J.,
181 Bhutkar, A., Weintraub, A. S., … Hemann, M. T. (2017). PHF6 regulates phenotypic plasticity through chromatin organization within lineage-specific genes, 1–17. http://doi.org/10.1101/gad.295857.117.6 Soufi, A., Donahue, G., & Zaret, K. S. (2012). Facilitators and Impediments of the Pluripotency Reprogramming Factors’ Initial Engagement with the Genome. Cell, 151(5), 994–1004. http://doi.org/10.1016/J.CELL.2012.09.045 Sridharan, R., Gonzales-Cope, M., Chronis, C., Bonora, G., McKee, R., Huang, C., … Plath, K. (2013). Proteomic and genomic approaches reveal critical functions of H3K9 methylation and heterochromatin protein-1γ in reprogramming to pluripotency. Nature Cell Biology, 15(7), 872–882. http://doi.org/10.1038/ncb2768 Steinmann-Zwicky, M., Niederer, E., Steinmann-Zwicky, M., Pennetta, G., & Pauli, D. (1994). Sex determination of the Drosophila germ line: tra and dsx control somatic inductive signals. Development (Cambridge, England), 120(3), 707–16. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8162863 Swerdlow, A., De Stavola, B., Swanwick, M., & Maconochie, N. (1997). Risks of breast and testicular cancers in young adult twins in England and Wales: evidence on prenatal and genetic aetiology. The Lancet, 350(9093), 1723– 1728. http://doi.org/10.1016/S0140-6736(97)05526-8 Tachibana, M., Sugimoto, K., Nozaki, M., Ueda, J., Ohta, T., Ohki, M., … Shinkai, Y. (2002). G9a histone methyltransferase plays a dominant role in euchromatic histone H3 lysine 9 methylation and is essential for early embryogenesis. Genes & Development, 16(14), 1779–91. http://doi.org/10.1101/gad.989402 Tan, S.-L., Nishi, M., Ohtsuka, T., Matsui, T., Takemoto, K., Kamio-Miura, A., … Kageyama, R. (2012). Essential roles of the histone methyltransferase ESET in the epigenetic control of neural progenitor cells during development. Development (Cambridge, England), 139(20), 3806–16. http://doi.org/10.1242/dev.082198 Thompson, P. J., Dulberg, V., Moon, K.-M., Foster, L. J., Chen, C., Karimi, M. M., & Lorincz, M. C. (2015). hnRNP K Coordinates Transcriptional Silencing by SETDB1 in Embryonic Stem Cells. PLOS Genetics, 11(1), e1004933. http://doi.org/10.1371/journal.pgen.1004933 Timms, R. T., Tchasovnikarova, I. A., Antrobus, R., Dougan, G., & Lehner Correspondence, P. J. (2016). ATF7IP-Mediated Stabilization of the Histone Methyltransferase SETDB1 Is Essential for Heterochromatin Formation by the HUSH Complex. CellReports, 17, 653–659. http://doi.org/10.1016/j.celrep.2016.09.050 Trapnell, C., Pachter, L., & Salzberg, S. L. (2009). TopHat: discovering splice junctions with RNA-Seq. Bioinformatics (Oxford, England), 25(9), 1105–11. http://doi.org/10.1093/bioinformatics/btp120 Trapnell, C., Roberts, A., Goff, L., Pertea, G., Kim, D., Kelley, D. R., … Pachter, L. (2012). Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nature Protocols, 7(3), 562–78. http://doi.org/10.1038/nprot.2012.016
182 Trost, M., Blattner, A. C., Leo, S., & Lehner, C. F. (n.d.). Drosophila dany is essential for transcriptional control and nuclear architecture in spermatocytes. http://doi.org/10.1242/dev.134759 Turnbull, C., Rapley, E. a, Seal, S., Pernet, D., Renwick, A., Hughes, D., … Rahman, N. (2010). Variants near DMRT1, TERT and ATF7IP are associated with testicular germ cell cancer. Nature Genetics, 42(7), 604–7. http://doi.org/10.1038/ng.607 Tzeng, T.-Y., Lee, C.-H., Chan, L.-W., & Shen, J. (2007). Epigenetic regulation of the Drosophila chromosome 4 by the histone H3K9 methyltransferase dSETDB1. PNAS. Retrieved from http://www.pnas.org/content/104/31/12691.full.pdf Uchimura, Y., Ichimura, T., Uwada, J., Tachibana, T., Sugahara, S., Nakao, M., & Saitoh, H. (2006). Involvement of SUMO modification in MBD1- and MCAF1-mediated heterochromatin formation. The Journal of Biological Chemistry, 281(32), 23180–90. http://doi.org/10.1074/jbc.M602280200 Volpe, A. M., Horowitz, H., Grafer, C. M., Jackson, S. M., & Berg, C. A. (n.d.). Drosophila rhino Encodes a Female-Specific Chromo-domain Protein That Affects Chromosome Structure and Egg Polarity. Retrieved from http://www.genetics.org/content/genetics/159/3/1117.full.pdf Wang, C., Liu, X., Gao, Y., Yang, L., Li, C., Liu, W., … Gao, S. (2018). Reprogramming of H3K9me3-dependent heterochromatin during mammalian embryo development. Nature Cell Biology, 20(5), 620–631. http://doi.org/10.1038/s41556-018-0093-4 Wang, X., Pan, L., Wang, S., Zhou, J., McDowell, W., Park, J., … Xie, T. (2011). Histone H3K9 trimethylase eggless controls germline stem cell maintenance and differentiation. PLoS Genetics, 181, 1303–1319. http://doi.org/10.1371/journal.pgen.1002426 Waterbury, J.A, Horabin, J.I, D. Bopp, A. S. P. (2000). Sex Determination in the Drosophila Germline Is Dictated by the Sexual Identity of the Surrounding Soma. Genetics, 155(4), 1741–56. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10430590 Wawersik, M., Milutinovich, A., Casper, A. L., Matunis, E., Williams, B., & Van Doren, M. (2005). Somatic control of germline sexual development is mediated by the JAK/STAT pathway. Nature, 436(7050), 563–567. http://doi.org/10.1038/nature03849 White-Cooper, H. (2010). Molecular mechanisms of gene regulation during Drosophila spermatogenesis. Reproduction, 139(1), 11–21. http://doi.org/10.1530/REP-09-0083 Wotton, D., Pemberton, L. F., & Merrill-Schools, J. (2017). SUMO and Chromatin Remodeling (pp. 35–50). Springer, Cham. http://doi.org/10.1007/978-3-319- 50044-7_3 Xiao, J., Xu, M., Li, J., Chang Chan, H., Lin, M., Zhu, H., … Sha, J. (2002). NYD- SP6, a Novel Gene Potentially Involved in Regulating Testicular Development/Spermatogenesis. Biochemical and Biophysical Research Communications, 291(1), 101–110. http://doi.org/10.1006/BBRC.2002.6396 Yan, D., Neumüller, R. a, Buckner, M., Ayers, K., Li, H., Hu, Y., … Perrimon, N.
183 (2014). A regulatory network of Drosophila germline stem cell self-renewal. Developmental Cell, 28(4), 459–73. http://doi.org/10.1016/j.devcel.2014.01.020 Yang, S. Y., Baxter, E. M., & Van Doren, M. (2012). Phf7 Controls Male Sex Determination in the Drosophila Germline. Developmental Cell, 22(5), 1041– 1051. http://doi.org/10.1016/j.devcel.2012.04.013 Yang, S. Y., Chang, Y.-C., Wan, Y. H., Whitworth, C., Baxter, E. M., Primus, S., … Doren, M. Van. (2017). Control of a novel spermatocyte-promoting factor by the male germline. http://doi.org/10.1534/genetics.117.199935 Yu, Y., Gu, J., Jin, Y., Luo, Y., Preall, J. B., Ma, J., … Hannon, G. J. (2015). Panoramix enforces piRNA-dependent cotranscriptional silencing. Science (New York, N.Y.), 350(6258), 339–42. http://doi.org/10.1126/science.aab0700 Zhang, J., Parvin, J., & Huang, K. (2012). Redistribution of H3K4me2 on neural tissue specific genes during mouse brain development. BMC Genomics 2012 13:8, 13(8), S5. http://doi.org/10.1186/1471-2164-13-s8-s5 Zhang, Y., Liu, T., Meyer, C. A., Eeckhoute, J., Johnson, D. S., Bernstein, B. E., … Liu, X. S. (2008). Model-based Analysis of ChIP-Seq (MACS). Genome Biology, 9(9), R137. http://doi.org/10.1186/gb-2008-9-9-r137 Zhang, Z. H., Jhaveri, D. J., Marshall, V. M., Bauer, D. C., Edson, J., Narayanan, R. K., … Zhao, Q.-Y. (2014). A Comparative Study of Techniques for Differential Expression Analysis on RNA-Seq Data. PLoS ONE, 9(8), e103207. http://doi.org/10.1371/journal.pone.0103207 Zofall Martin, Yamanaka Soichiro, Reyes-Turcu Francisca E., Zhang Ke, Rubin Chanan, G. S. I. . (2012). RNA Elimination Machinery Targeting Meiotic mRNAs Promotes Facultative Heterochromatin Formation. Science, 335, 96–99. http://doi.org/10.1126/science.1214115
184