Role of Rna Processing Machinery and Signaling Pathways in Regulating the Dynamic Epigenetic Landscape Through Heterochromatin Assembly

ROLE OF RNA PROCESSING MACHINERY AND SIGNALING PATHWAYS IN REGULATING THE DYNAMIC EPIGENETIC LANDSCAPE THROUGH HETEROCHROMATIN ASSEMBLY

by Nathan N. Lee

A dissertation submitted to Johns Hopkins University in conformity with the requirements for the degree of Doctor of Philosophy

Baltimore, Maryland October, 2016

ABSTRACT

The regulation of protein coding and noncoding RNAs is linked to nuclear processes including chromatin modifications and gene silencing. Heterochromatin is the major form of chromatin in higher eukaryotes that impacts various chromosomal processes including genomic stability and global gene expression patterns. In the fission yeast Schizosaccharomyces pombe heterochromatin is enriched across large chromosomal domains such as centromeres, telomeres and the mating-type region. In addition, recent work has revealed the existence of small blocks of facultative heterochromatin across the genome. Importantly, the assembly of facultative heterochromatin is mediated by RNA-based mechanisms and dynamically regulated in response to environmental and developmental cues. However, the mechanisms that distinguish RNAs for assembly of heterochromatin at different regions of the genome and how signals trigger changes at the chromatin are poorly understood.

I describe the discovery of a nuclear RNA processing network in fission yeast with a core module comprising the Mtr4-like protein named Mtl1 and the zinc finger protein, Red1. The Mtl1-Red1 core (MTREC) promotes degradation of mRNAs and noncoding RNAs, and associates with different proteins to assemble heterochromatin via distinct mechanisms. Mtl1 also forms Red1-independent interactions with evolutionarily conserved proteins named Nrl1 and Ctr1, which associate with splicing factors. Whereas

Nrl1 targets transcripts with cryptic introns to form heterochromatin at developmental genes and retrotransposons, Ctr1 functions in processing intron-containing telomerase

RNA. Together with our discovery of widespread cryptic introns, including in noncoding

RNAs, these findings reveal unique cellular strategies for recognizing regulatory RNAs.

ii Furthermore, I have found that Tor2, the yeast homolog of mTOR, functionally connects environmental and developmental signaling cues with remodeling of facultative heterochromatin by regulating cellular protein levels of the MTREC-associated factor

Pir1. This process involves Cul4, an E3 ubiquitin ligase in the ClrC complex that also includes the methyltransferase Clr4, Pyp1, a tyrosine phosphatase implicated in TOR pathway, and Swi6/HP1, the heterochromatin binding protein. The mechanism that regulates Pir1 also provides a feedback loop for maintaining the proper level of facultative heterochromatin. These findings reveal signaling pathways and mechanisms that are involved in the dynamic regulation of facultative heterochromatin in response to environmental and developmental signals.

Name of Readers/Advisors: Shiv Grewal, Ph.D. (Thesis Advisor) Karen Beemon, Ph.D. Xin Chen, Ph.D. Michael Lichten, Ph.D.

iii PREFACE

All of the work presented henceforth was conducted in the Chromosome Biology

Section of the Laboratory of Biochemistry and Molecular Biology at the Center for

Cancer Research of the National Cancer Institute.

A version of Chapters 1, 2, and 3 has been published (Lee NN, Chalamcharla VR,

Reyes-Turcu F, Mehta S, Zofall M, Balachandran V, Dhakshnamoorthy J, Taneja N,

Yamanaka S, Zhou M, Grewal SI. Mtr4-like protein coordinates nuclear RNA processing for heterochromatin assembly and for telomere maintenance. Cell. 2013 Nov

21;155(5):1061-74). I was the lead author of the published work, responsible for areas of hypothesis, data collection, analysis, and manuscript composition while there were a number of co-authors who contributed significantly to the published work. Venkata R.

Chalamcharla was involved in purification of proteins and small RNA sequencing.

Francisca Reyes-Turcu was involved in the initial concept formation and experiments including protein purification, co-IP, and ChIP-chip. Sameet Mehta helped with bioinformatics analysis. Martin Zofall contributed to some protein purifications, RNA sequencing, and Northern blot analysis. Vanivilasini Balachandran and Jothy

Dhakshnamoorthy helped in generating strains. Nitika Taneja contributed to immunofluorescence experiments. Soichiro Yamanaka contributed to one ChIP-chip experiment. Ming Zhou helped with mass-spectrometry analysis. Shiv I.S. Grewal was the supervisory author on this project and was involved throughout the project in concept formation and manuscript composition.

A version of Chapters 4, 5, and 6 is being prepared for publication. I am the lead author of the work, responsible for areas of hypothesis, initial concept formation, data

iv collection, analysis, and manuscript composition while a number of co-authors also contributed to the project. Gobi Thillainadesan was involved in bioinformatics analysis.

David Wang and Avindra Nath provided help with human neural stem cell culture. Shiv

Grewal is the supervisory author on this project and was involved throughout the project in concept formation and manuscript editing.

v ACKNOWLEDGEMENT

I thank everyone who has worked and exchanged ideas with me over this project.

I thank every member of the Shiv Grewal laboratory (Chromosome Biology Section of the Laboratory of Biochemistry and Molecular Biology, Center for Cancer Research,

National Cancer Institute) during my time for valuable contributions and help. I thank my thesis advisor, Shiv Grewal, for support and guidance. And I thank my thesis committee members, Karen Beemon, Xin Chen, and Michael Lichten for advice and support.

I thank Peter FitzGerald for intron analysis, Robin Allshire for the cwf10-1 mutant, Masayuki Yamamoto for the tor2-ts6 mutant, Jemima Barrowman for her valuable help in editing the manuscripts, and Michael Lichten for comments.

This research was supported by the Intramural Research Program of the National

Institutes of Health (NIH), National Cancer Institute, and utilized the Helix Systems and the Biowulf Linux cluster at the NIH.

vi TABLE OF CONTENTS

Page

ABSTRACT ...... ii

PREFACE ...... iv

ACKNOWLEDGEMENT ...... vi

TABLE OF CONTENTS ...... vii

LIST OF TABLES ...... ix

LIST OF FIGURES ...... x

INTRODUCTION ...... 1

RESULTS ...... 11

Chapter 1. Purification of Red1 and Identification of Its Associated Proteins ...... 11 1.1. Red1 Interacts with Various Factors ...... 11 1.2. Red1 and Mtl1 Form a Core Module that Interacts with Different Nuclear Proteins ...... 14 1.3. Red1- and Mtl1-Associated Factors Differentially Affect Heterochromatin Domains ...... 18

Chapter 2. MTREC Regulates Gene Expression and Targets Noncoding RNAs and Pre-mRNAs Degraded by the Exosome ...... 29 2.1. Mtl1 Regulates Expression of Genes Involved in Sexual Differentiation, Stress Response and Membrane Transport ...... 29 2.2. MTREC Targets Regulatory Noncoding RNA and Pre-mRNA Degraded by the Exosome ...... 31 2.3. Noncoding RNA Regulates Gene Expression in Response to Environmental Changes ...... 33

Chapter 3. Mtl1 Forms Red1-Independent Interactions with Nrl1 and Ctr1 for Regulation of Splicing Associated with Heterochromatin Assembly and Telomere Maintenance ...... 38 3.1. Mtl1 Associates with Nrl1 and Ctr1 without Red1 ...... 38 3.2. Nrl1 Promotes Assembly of HOODs at Genes and Retrotransposons ...40 3.3. Nrl1 Interacts with Splicing Factors to Assemble HOODs via Cryptic Introns ...... 42 3.4. Noncoding RNAs and Read-through Transcripts Contain Introns ...... 51 3.5. Mtl1 and Ctr1 Promote Telomerase RNA Biogenesis and Telomere Maintenance ...... 54

vii Chapter 4. TOR Signaling Pathway Regulates Facultative Heterochromatin ...... 58 4.1. TOR Signaling Pathway Regulates Heterochromatin Islands ...... 58 4.2. Tor2 Regulates MTREC-dependent Heterochromatin through Pir1 ...... 62 4.3. Tor2 and Pir1 Regulate Environmentally-sensitive and Disease- associated Genes ...... 67

Chapter 5. Factors Involved in the Regulation of Pir1 by Tor2 ...... 71 5.1. Cul4 is Required for Pir1 Degradation upon Signaling ...... 71 5.2. Pir1 and Cul4/ClrC Regulate Facultative Heterochromatin through a Feedback Loop ...... 76 5.3. Tor2 Regulates Pir1 through Pyp1 ...... 80

Chapter 6. Tor2 and Pir1 Regulate Formation of Latent Heterochromatin Islands at Genes Associated with Human Diseases ...... 82 6.1. MTREC-Pir1 is Involved in Formation of Latent Heterochromatin at Environmentally-sensitive Genes ...... 82 6.2. Tor2 and Pir1 Regulate Latent Heterochromatin Islands ...... 87 6.3. Latent Heterochromatin Island Formation in Human Cells ...... 89

Chapter 7. Discussion: MTREC Forms a Nuclear RNA Processing Network that Silences Genes and Retrotransposons and Also Maintains Telomeres ...... 94 7.1. Mtl1 and Red1 Exist in Multiple Protein Assemblies With Distinct Functions ...... 95 7.2. Noncoding RNAs and Environmental Gene Control ...... 97 7.3. Introns, Splicing and Epigenetic Genome Control ...... 98

Chapter 8. Discussion: Regulation of Facultative Heterochromatin by the TOR Pathway and MTREC RNA Processing Machinery in Response to Signals ...... 102 8.1. TOR Pathway Regulates Formation of Facultative Heterochromatin through RNA Processing Machinery ...... 104 8.2. Facultative Heterochromatin is Regulated through a Feedback Loop ..105 8.3. Latent Heterochromatin Formation at Disease-associated Genes ...... 107

Chapter 9. Experimental Procedures ...... 109

REFERENCES ...... 118 1. Appendices ...... 118 2. Bibliography ...... 119

CURRICULUM VITAE ...... 136

viii LIST OF TABLES

Page

Table 1.1 Table summarizing results of ChIP-chip analysis for Red1 and Mtl1 and Pir1 localizations in wild-type cells, and H3K9me2 enrichment in wild- type and mutant strains at Red1-dependent and -independent islands ...... 21

Table 1.2 Rmn1 facilitates small RNA production and H3K9me at genes and retrotransposons ...... 27

Table 3.1 Rmn1, Nrl1 and Cwf10 facilitate small RNA production and H3K9me at genes and retrotransposons ...... 48

Table 4.1 List of S. pombe genes whose human orthologs are associated with diseases and are upregulated in pir1∆, and tor2-ts6 cells ...... 69

Table 6.1 List of latent heterochromatin islands in S. pombe ...... 84

Table 6.2 List of S. pombe genes whose human orthologs are associated with diseases and are targets of latent heterochromatin island domain assembly ...... 89

Table 6.3 List of human ortholog genes associated with diseases and are targets of heterochromatin formation in specific cell lines or tissues ...... 90

ix LIST OF FIGURES

Page

Figure 1.1 Expression of tagged proteins ...... 11

Figure 1.2 Proteins copurified with Red1, Mtl1, Pir1, and Rmn1 ...... 11

Figure 1.3 Proteins associated with Red1, Mtl1, Pir1, and Rmn1 that were identified from the indicated purifications ...... 12

Figure 1.4 Predicted protein domains of Mtl1, Ctr1 and their human orthologs ...... 13

Figure 1.5 Co-IP of associated proteins from strains expressing tagged proteins ...... 14

Figure 1.6 Immunofluorescence analysis of MYC-tagged proteins ...... 15

Figure 1.7 Co-IP analysis of Pla1 and Mtl1 interaction ...... 16

Figure 1.8 Red1 and Mtl1 co-localize in the nucleus ...... 16

Figure 1.9 Protein interaction network of MTREC and Mtl1 ...... 17

Figure 1.10 Localization of Red1 and Mtl1 at heterochromatin islands ...... 18

Figure 1.11 Role of MTREC factors in the assembly of heterochromatin islands ...... 19

Figure 1.12 Average distribution of Red1 and Mtl1 across Red1-dependent and – independent islands, as determined by ChIP-chip analysis ...... 20

Figure 1.13 H3K9me2 enrichment at Red1-independent islands and a pericentromeric region in indicated strains ...... 22

Figure 1.14 Pir1 is localized at heterochromatin islands ...... 23

Figure 1.15 Effect of Rmn1 and Pir1 at HOODs ...... 24

Figure 1.16 Quantitative analysis of small RNAs at HOODs in rrp6∆ and rrp6∆ rmn1∆ strains ...... 27

Figure 2.1 Effect of MTREC on genes and ncRNAs ...... 29

Figure 2.2 RT-PCR analysis of Red1- and Mtl1-bound loci in indicated strains ...... 30

Figure 2.3 Effect of MTREC on snoRNAs ...... 31

x Figure 2.4 Effect of MTREC on ribosomal RNA ...... 32

Figure 2.5 ChIP-chip analysis of Red1-MYC and Mtl1- MYC distribution at the pho1 locus ...... 33

Figure 2.6 ncRNA regulates gene expression in response to environmental changes ...... 34

Figure 2.7 Detection of ncRNA at pho1 locus ...... 35

Figure 2.8 ChIP analysis of Red1-MYC and Mtl1-FLAG enrichment at the pho1 locus ...... 36 Figure 2.9 Role of ncRNA in regulation of pho1 locus ...... 36

Figure 3.1 Protein purification of Nrl1 and Ctr1 ...... 38

Figure 3.2 Proteins associated with Nrl1 and Ctr1 that were identified from the indicated purifications ...... 39

Figure 3.3 Nrl1 and Ctr1 interact with Mtl1 ...... 40

Figure 3.4 Role of Nrl1 at HOODs loci ...... 40

Figure 3.5 Effect of nrl1∆ at different heterochromatin domains ...... 41

Figure 3.6 Nrl1 interacts with splicing factors ...... 42

Figure 3.7 Schematic of cryptic introns detected at HOODs by RNA-seq in the indicated strains ...... 43

Figure 3.8 Schematic of RNA-Seq reads that map to Tf2 and SPCC1442.04c in rrp6∆, which contain cryptic introns ...... 43

Figure 3.9 Role of SPCC1442.04c intron 2 ...... 44

Figure 3.10 Normalized number of small RNA reads at HOODs in indicated strains .48

Figure 3.11 Quantitative analysis of small RNAs at HOODs in indicated samples .....49

Figure 3.12 Role of Cwf10 at HOODs loci ...... 49

Figure 3.13 Effect of cwf10-1 at different Loci ...... 50

Figure 3.14 Analysis of introns in different mutant strains ...... 51

Figure 3.15 Cryptic introns are found in the pericentromeric region ...... 52

Figure 3.16 Examples of cryptic introns in different mutants ...... 53

Figure 3.17 Telomerase RNA TER1 is regulated by factors associated with MTREC 55

Figure 3.18 Mtl1 and Ctr1 regulate TER1 and telomere maintenance ...... 56

Figure 3.19 Nrl1 is not involved in the regulation of TER1 and telomere maintenance57

Figure 4.1 Facultative heterochromatin domains respond to environmental signals ..58

Figure 4.2 ChIP qPCR analysis of H3K9me2 enrichment at mei4 locus in indicated strains ...... 59

Figure 4.3 Tor2 regulates heterochromatin islands ...... 60

Figure 4.4 ChIP qPCR analysis of H3K9me2 enrichment at the dg locus in the indicated strains ...... 60

Figure 4.5 MTREC-independent heterochromatin islands are not significantly affected by the TOR pathway ...... 61

Figure 4.6 Effect of Tor2 on MTREC-associated factors ...... 62

Figure 4.7 Tor2 regulates Pir1 enrichment at the heterochromatin island ...... 63

Figure 4.8 Genome-wide analysis of ChIP-chip for H3K9me2 at facultative heterochromatin domains ...... 64

Figure 4.9 RNA expression analysis at heterochromatin islands in different mutants65

Figure 4.10 Analysis of RNA-seq results in pir1∆ and tor2-ts6 cells ...... 67

Figure 4.11 Effect of pir1∆ and tor2-ts6 at various genes associated with diseases ....70

Figure 5.1 Co-IP analysis of the interaction between Pir1 and Clr4 ...... 71

Figure 5.2 Pir1 is regulated by Cul4 in response to signaling ...... 72

Figure 5.3 Western blot analysis of Pir1-MYC in the indicated strains grown with or without nitrogen source (+N/-N) ...... 73

Figure 5.4 Western blot analysis of Pir1-MYC in the indicated strains grown with or without nitrogen source (+N/-N) ...... 73

xii Figure 5.5 Western blot analysis of Pir1-MYC in the indicated strains grown with or without nitrogen source (+N/-N) ...... 74

Figure 5.6 Western blot analysis of Pir1-MYC in the indicated strains grown with or without nitrogen source (+N/-N) ...... 75

Figure 5.7 ChIP qPCR analysis of H3K9me2 enrichment at the mei4 locus in the indicated strains ...... 76

Figure 5.8 ChIP qPCR analysis of Pir1-MYC enrichment at the mei4 locus in the indicated strains ...... 77

Figure 5.9 ChIP qPCR analysis of Pir1-MYC enrichment at the mei4 locus in the indicated strains and conditions ...... 78

Figure 5.10 Figure 5.10. Western blot analysis of Pir1-MYC in the indicated strains grown with or without a nitrogen source (+N/-N) ...... 78

Figure 5.11 Effect of swi6∆ on the heterochromatin island ...... 79

Figure 5.12 ChIP qPCR analysis of H3K9me2 enrichment at the mei4 locus in the indicated strains ...... 80

Figure 5.13 Pyp1 is involved in the regulation of Pir1 by Tor2 ...... 81

Figure 6.1 ChIP-chip analysis of MTREC factors and H3K9me2 ...... 82

Figure 6.2 Latent heterochromatin islands form at genes sensitive to environmental signals ...... 85

Figure 6.3 Effect of swi6∆ on the latent heterochromatin island ...... 86

Figure 6.4 ChIP-chip analysis of H3K9me2 distribution at latent heterochromatin islands domain loci in the indicated conditions ...... 86

Figure 6.5 Latent heterochromatin domains are regulated by the Tor2 pathway ...... 87

Figure 6.6 Genome-wide analysis of ChIP-chip data ...... 88

Figure 6.7 ChIP-Seq analysis of H3K9me3 in different human cell lines and tissues for latent island loci (ENCODE) ...... 91

Figure 6.8 ChIP qPCR analysis of H3K9me3 enrichment in neuronal stem cells ...... 92

xiii Figure 7.1 Schematic illustrating the role of MTREC and Mtl1-Nrl1 protein assemblies in the degradation of various RNA species and the formation of heterochromatin at islands or HOODs ...... 95

Figure 7.2 Schematics illustrating the role of intron and splicing factors in targeting MTREC-associated machinery to HOODs loci ...... 99

Figure 8.1 Schematic illustrating a model of how Tor2 signaling regulates facultative heterochromatin formation at genes sensitive to environmental signals .102

Figure 8.2 Schematic illustrating how the regulation of Pir1 is utilized to form and maintain the proper level of facultative heterochromatin through a feedback loop ...... 105

xiv INTRODUCTION

The availability of the DNA sequence of many model organisms, as well as the human genome, has allowed the development of novel approaches to explore the complexities of eukaryotic genomes. A surprising finding is that the majority of the genome is transcribed, often on both DNA strands (Dutrow et al., 2008; Gingeras, 2007;

Grewal, 2010; Wilhelm et al., 2008). In addition to protein coding RNAs, the occurrence of so-called cryptic transcripts, antisense transcripts, and non-coding RNAs (ncRNA) is widespread. Repetitive elements such as transposons provide an abundant source of noncoding RNAs (Grewal, 2010; Orom and Shiekhattar, 2011; Struhl, 2007). In several cases, expression of long non-coding RNAs (lncRNA) is regulated and occurs under specific growth or developmental conditions. lncRNAs are emerging as critical components of epigenetic regulatory mechanisms that direct chromatin modifications

(Batista and Chang, 2013; Feng and Jacobsen, 2011; Lee, 2012). However, the mechanisms by which cells recognize these RNAs and mediate their effects are poorly understood.

Cellular RNA levels are tightly controlled at both transcriptional and posttranscriptional levels at regions across the genome, including those containing repetitive

DNA and disease-associated genes, to avoid genome instability and cellular imbalance that can lead to diseases such as cancer (Beisel and Paro, 2011; Cooper et al., 2009;

Grewal, 2010; Peng and Karpen, 2008). RNA processing activities, such as the exosome and RNAi machinery, control the steady-state levels of diverse RNA species (Doma and

Parker, 2007; Houseley et al., 2006; Reyes-Turcu and Grewal, 2012; Schmid and Jensen,

1 2008) in conjunction with 3’ end formation mechanisms that determine the fate of various RNAs (Tuck and Tollervey, 2013). The exosome processes and degrades RNA substrates in the cytoplasm and the nucleus (Houseley et al., 2006). The nuclear exosome contains the 3’à 5’ exonuclease Rrp6, which processes various RNAs to their mature forms and degrades ncRNA, antisense RNA, and transcripts produced from repeat elements (Houseley et al., 2006; Reyes-Turcu and Grewal, 2012). Cofactors such as

TRAMP (Trf4-Air2-Mtr4 polyadenylation), which contains the RNA helicase Mtr4 and polyadenylates RNA substrates through its non-canonical poly (A) polymerase Trf4/5

(Cid14 in fission yeast Schizosaccharomyces pombe), stimulate exosome activity

(Houseley et al., 2006). In S. pombe, TRAMP also activates RNAi to degrade transcripts from repeat elements and antisense RNA (Zhang et al., 2011).

The regulated degradation of RNAs also impacts gene control during differentiation. Meiosis is the most dramatic differentiation program, and in S. pombe, meiotic induction is accompanied by upregulation of meiosis-specific genes, which are silenced during vegetative growth (Mata et al., 2007) by an RNA elimination system involving the exosome (Yamamoto, 2010). Polyadenylation of meiotic RNAs is essential for their elimination, and requires the canonical poly(A) polymerase Pla1, which acts together with Mmi1, a protein that binds RNA containing determinant of selective removal (DSR) elements (Harigaya et al., 2006; Sugiyama and Sugioka-Sugiyama, 2011;

Yamanaka et al., 2010). TRAMP is dispensable for this process (McPheeters et al., 2009;

St-Andre et al., 2010; Yamanaka et al., 2010). Mmi1 recruits the Zn-finger protein Red1, which associates with Pla1 and the exosome to degrade RNAs. The nuclear poly(A) binding protein Pab2, implicated in processing snoRNAs and in turnover of pre-mRNAs

2 by the exosome (Lemay et al., 2010; Lemieux et al., 2011), is also required (St-Andre et al., 2010; Sugiyama and Sugioka-Sugiyama, 2011; Yamanaka et al., 2010).

Recent work has demonstrated a functional link between widespread transcription and the formation of heterochromatin, characterized by methylation of histone H3 at lysine 9 and the presence of HP1 proteins, at various sites throughout the genome (Reyes-

Turcu and Grewal, 2012; Zhang et al., 2011; Zofall et al., 2009). H3K9 is methylated by

Clr4 – a homolog of Drosophila SU(VAR)3-9 and human Suv39h (Zhang et al., 2008).

Clr4 forms a multisubunit E3 ubiquitin ligase complex (ClrC) that contains Cul4, Rik1,

Raf1, and Raf2 (Zhang et al., 2008). Cul4 is a member of the cullin family of proteins that are components of E3 ubiquitin ligases involved in protein ubiquitination. Previous studies have shown that Cul4 regulates heterochromatin assembly through its interaction with Clr4 (Horn et al., 2005; Jia et al., 2005). H3K9 methylation by ClrC also recruits the HP1 family protein Swi6, which provides a dynamic platform for recruitment of various regulatory proteins (Fischer et al., 2009; Reyes-Turcu and Grewal, 2012). To impose efficient repression at various regions of the genome, eukaryotic genomes utilize multiple mechanisms involving RNAi machinery, chromatin-modifying factors, and

RNA-processing factors (Grewal and Elgin, 2007; Matzke and Birchler, 2005; Reyes-

Turcu and Grewal, 2012; Reyes-Turcu et al., 2011; Zhang et al., 2011; Zofall et al., 2009;

Zofall et al., 2012).

S. pombe contains three distinct types of heterochromatin, which differ in their dependency on trans-acting factors. The first type corresponds to major H3K9me peaks at centromeres, telomeres and the mating-type locus. RNAi proteins Argonaute (Ago1),

Dicer (Dcr1) and RNA-dependent RNA polymerase (Rdp1) target transcripts produced

3 by dg/dh repeats in these regions to generate siRNAs that facilitate loading of the Clr4

(Reyes-Turcu and Grewal, 2012). An Ago1-containing RNA-induced transcriptional silencing complex (RITS) binds siRNAs that provide specificity for loading of this complex onto nascent repeat transcripts. RITS also associates with Clr4, which is believed to be critical for binding of this complex to loci and the nucleation of heterochromatin structures (Cam et al., 2005; Hall et al., 2002; Motamedi et al., 2004;

Sugiyama et al., 2005; Verdel et al., 2004). Interestingly, the formation of heterochromatin by RNAi requires transcription of the repressed regions to generate siRNA precursors. This RNAPII transcription is facilitated by a JmjC domain-containing anti-silencing factor Epe1 (Zofall and Grewal, 2006).

The second type of heterochromatin includes small blocks of heterochromatin islands, which include meiotic genes and other loci (Cam et al., 2005; Zofall et al., 2012).

The majority of heterochromatin islands are composed of meiotic genes that are tightly repressed during vegetative growth (Zofall et al., 2012). Repression at the islands is achieved by mechanisms involving the RNA binding protein Mmi1, which recognizes and binds determinant of selective removal (DSR) domains present in RNAs of meiotic genes. Mmi1 is believed to recruit the RNA processing factor Red1 and its associated 3’-

5’ exonuclease exosome component, Rrp6, to meiotic genes (Harigaya et al., 2006;

Sugiyama and Sugioka-Sugiyama, 2011; Zofall et al., 2012). Red1 forms a complex with

Clr4, whose activity is critical for RNAi-independent assembly of heterochromatin islands (Zofall et al., 2012). Evidence suggests that heterochromatin islands are akin to facultative heterochromatin that is dynamically regulated by cellular and environmental signals. In particular, nutritional signals such as nitrogen starvation trigger meiotic

4 induction, causing disassembly of heterochromatin islands (Zofall et al., 2012).

However, the cascade of events that connect nutritional signals to remodeling of heterochromatin islands is not understood.

The third type of heterochromatin consists of domains named HOODs, which are detected under specific growth conditions and are dynamically regulated (Yamanaka et al., 2013). HOODs preferentially assemble at sexual differentiation genes and retrotransposons, and require RNAi as well as elimination factors (Yamanaka et al.,

2013). Red1 and its partners Pla1 and Pab2 are cofactors for both RNAi and the exosome, which act in parallel to degrade transcripts from loci within these domains (Yamanaka et al., 2013).

In addition to heterochromatin, there are three major chromatin modifying complexes that are critical for transcriptional repression in different parts of the genome: the Clr6 histone deacetylase (HDAC) complex, the Snf2-HDAC repressor complex

(SHREC), and the Asf1-histone interacting protein A (HIRA) histone chaperone (Nicolas et al., 2007; Sugiyama et al., 2007; Yamada et al., 2005; Yamane et al., 2011). These complexes are recruited by different mechanisms to diverse locations throughout the genome where they enforce silencing of the genome at specific loci (Cam et al., 2008;

Nicolas et al., 2007; Sugiyama et al., 2007; Yamada et al., 2005; Yamane et al., 2011;

Zhang et al., 2008). For example, the Clr6 HDAC complex and Clr3-containing SHREC complex are recruited to transposable elements and their remnants, such as 13 Tf2 retrotransposons and 249 solo long terminal repeats (LTRs), by transposon-derived centromeric CENP-B proteins (Cam et al., 2008). On the other hand, Clr6, SHREC

HDAC complexes, and Asf1-HIRA histone chaperones are recruited to pericentromeric

5 regions, which contain a high-density of repetitive DNA elements, by the HP1 proteins

Chp2 and Swi6 bound to histone H3 methylated at lysine 9 (H3K9me) (Fischer et al.,

2009; Nicolas et al., 2007; Yamane et al., 2011). Once recruited to genomic loci, chromatin-modifying factors promote the assembly and spreading of heterochromatin

(Cam et al., 2005; Hall et al., 2002).

The dynamic control of pathways that mediate the repression of various genetic elements is a conserved feature of eukaryotic genomes (Reyes-Turcu and Grewal, 2012).

Changes in nutritional status, or cellular stresses such as a temperature shift, can have long-lasting effects on metabolism, development, and cellular homeostasis (Feil and

Fraga, 2011). The S. pombe basic leucine zipper (bZIP) family transcription factors, Atf1 and Pcr1, have been shown to regulate chromatin architecture in response to stress (Jia et al., 2004; Lawrence et al., 2007; Wilkinson et al., 1996). In plants, siRNAs have been shown to contribute to the regulation of the epigenome at repeated DNA sequences in response to heat-stress (Ito et al., 2011). In Drosophila melanogaster, heterochromatin formation at certain genomic regions is affected in response to heat and osmotic stress

(Seong et al., 2011). Moreover, in mammals, dietary compounds have been shown to influence the activity of chromatin-modifying factors such as histone deacetylases

(Dashwood and Ho, 2007).

A cascade of signaling pathways is involved in the transmission of environmental and cellular signals to bring about changes in the expression of target sequences. The major signaling pathways that respond to environmental signals are conserved from S. pombe to human (Chen et al., 2003; Perez and Cansado, 2010; Yamamoto, 2010). These pathways include phosphatidylinositol 3-kinase related protein kinases (PIKKs),

6 mitogen-activated protein kinases (MAPKs), and the cAMP-dependent protein kinase

(PKA) pathways (Alvarez and Moreno, 2006; Chen et al., 2003; Matsuo et al., 2007;

Perez and Cansado, 2010; Samejima et al., 1997; Sanso et al., 2011; Schonbrun et al.,

2009; Versele et al., 2001; Yamamoto, 2010). Central components of the PIKK- dependent pathways include Rad3 (ATR), Tel1 (ATM), Tor1 (TORC2 complex), and

Tor2 (TORC1 complex), and all play important roles in a variety of cellular functions:

Rad3 and Tel1 pathways respond to genotoxic stress and are essential for DNA damage responses, while Tor1 and Tor2 pathways respond to nutrient, energy, and stress, playing a central role in cellular growth (Alvarez and Moreno, 2006; Matsuo et al., 2007;

Schonbrun et al., 2009; Yamamoto, 2010). The MAPK pathways include the central components Sty1 and Pmk1 (Samejima et al., 1997; Sanso et al., 2011; Yamamoto,

2010). Finally, Pka1 of PKA pathway is essential for responding to cellular glucose levels (Versele et al., 2001; Yamamoto, 2010).

Among the signaling pathways listed above, mutations in Sty1 (P38 MAPK),

Tor1 (PIKK pathway), and Pka1 (PKA pathway) have been shown to influence silencing at specific loci (Lawrence et al., 2007; Murayama et al., 2008; Schonbrun et al., 2009).

Stress induces the phosphorylation and activation of Sty1, and it in turn phosphorylates and stabilizes its target substrate, the transcription factor Atf1, which binds to the promoters of stress-response genes (Jia et al., 2004; Lawrence et al., 2007). The binding of Atf1 to cAMP response element sites (CRE) at promoter regions induces the expression of stress-response genes by remodeling the heterochromatin (Jia et al., 2004).

Tor1 is implicated in sensing nutrients and stress, and may be involved in gene silencing, telomere length maintenance, and DNA damage response (Schonbrun et al., 2009). In

7 mammals, PKA senses glucose deprivation and signals to eNoSC (energy-dependent nucleolar silencing complex), which contains silent information regulator 1 (SIRT1),

SUV39H1, and Nucleomethylin (NML) (Murayama et al., 2008). Under low glucose conditions, the eNoSC acts to repress the rDNA locus, reducing ribosome biosynthesis, which is the most energy-consuming process. Heterochromatin formation at the rDNA locus is promoted by SIRT1 activity, which deacetylates the H3 histone, and SUV39H1 activity, which methylates Lys9 of H3 (Murayama et al., 2008).

Tor2 (TORC1) is an essential protein and regulates cell growth, sexual development, and protein synthesis in response to nutrients (Alvarez and Moreno, 2006;

Matsuo et al., 2007). Mammalian TOR (mTOR) pathway is well conserved from S. pombe, and mTORC1 (mammalian homologue of S. pombe Tor2) in particular plays a central role as a regulator of energy homeostasis, metabolism, and growth that responds to the level of nutrients and cellular energy (Dazert and Hall, 2011; Sengupta et al., 2010;

Wullschleger et al., 2006; Zoncu et al., 2011). Interestingly, recent findings have shown that inhibition of the TOR pathway in yeast, flies, worms, and mice extends lifespan, potentially mimicking the outcome of dietary restriction (Dazert and Hall, 2011).

Moreover, the TOR pathway is implicated in various diseases such as cancer, metabolic diseases, neurological conditions, and aging, where cellular homeostasis is compromised and growth is deregulated (Dazert and Hall, 2011; Zoncu et al., 2011). In S. pombe, two different TOR complexes, Tor1 (TORC2) and Tor2 (TORC1) play opposite roles in response to developmental signals such as nitrogen starvation (Weisman et al., 2007).

Unlike Tor1, Tor2 is critical for regulating the expression of meiotic genes that are activated upon nitrogen starvation (Matsuo et al., 2007), a condition that has recently

8 been shown to trigger disassembly of heterochromatin islands containing meiotic genes

(Zofall et al., 2012).

Previous works have shown how facultative heterochromatin domains in

S.pombe, comprised of islands and HOODs, are regulated through RNAi-independent and –dependent mechanisms, respectively. However, the mechanisms by which Red1 promotes assembly of RNAi-independent and -dependent heterochromatin, and by which

HOOD formation is specified, were not understood. Also, how environmental and developmental signals are linked to regulation of facultative heterochromatin, especially those at heterochromatin islands that are dynamically regulated in response to nitrogen starvation, was unknown.

Here we showed that Red1 interacts with Mtl1 (Mtr4-like protein 1) to form a core module, MTREC (Mtl1-Red1 Core), which associates with other factors to assemble islands or HOODs. We found that MTREC coordinates degradation of meiotic mRNAs, and localizes to ncRNA loci that regulate adjacent gene expression in response to environmental changes. Mtl1 also associates with the putative C. elegans NRDE-2

(Guang et al., 2010) homolog Nrl1, an evolutionarily conserved novel protein named

Ctr1, and splicing factors. We also discovered numerous cryptic introns in the S. pombe genome. Splicing machinery and Nrl1 act on cryptic introns that specify assembly of

HOODs, whereas Ctr1 facilitates processing of intron-containing telomerase RNA to maintain telomeres. These results uncovered an RNA processing network that targets various RNA substrates and promotes heterochromatin assembly.

Furthermore, we showed that the TOR pathway component Tor2, which is the homologue of mTOR, is involved in regulating the assembly of facultative

9 heterochromatin structures in S. pombe. We found that Tor2 regulates the cellular protein levels of Pir1, an MTREC-associated factor involved in RNA processing and heterochromatin assembly. This process requires Cullin4 (Cul4), an E3 ubiquitin ligase, which among other factors is a component of the ClrC complex containing Clr4. The regulation of Pir1 through Tor2 also involves other factors including Pyp1, a tyrosine phosphatase implicated in the TOR pathway (Petersen and Nurse, 2007), and also

Swi6/HP1 that plays a role in providing a platform and a feedback loop for facultative heterochromatin assembly. These results revealed a signaling pathway and factors that are involved in the dynamic regulation of facultative heterochromatin in response to environmental and developmental signals.

10 RESULTS

Chapter 1. Purification of Red1 and Identification of Its Associated Proteins

1.1.Red1 Interacts with Various Factors

Figure 1.1. Expression of tagged proteins. Extracts from tagged and untagged strains were analyzed by western blot. Others contributed: Francisca Reyes-Turcu and Martin Zofall.

To investigate the role of Red1 in meiotic RNA processing and in RNAi- dependent and –independent heterochromatin formation, we expressed Red1 with

3×FLAG-tag at its C-terminus under its native promoter (Figure 1.1).

Figure 1.2. Proteins copurified with Red1, Mtl1, Pir1, and Rmn1. Proteins highly associated with each other are shaded in blue. Others contributed: Francisca Reyes- Turcu, Martin Zofall, and Ming Zhou.

Purification of Red1-FLAG followed by mass spectrometry analysis identified a number of proteins specific to the Red1-FLAG purified fraction, including Red1 and its known interaction partners Mmi1 and Pla1 (Figure 1.2) (Sugiyama and Sugioka-Sugiyama,

2011).

Red1-FLAG Mtl1-FLAG Pir1-Myc

Gene name Coverage Gene name Coverage Gene name Coverage Red1 66.1 % Mtl1 71.0 % Pir1 83.5 % Mtl1 62.1 % Red1 78.2 % Mtl1 33.8 % Pla1 44.0 % Pir1 60.1 % Red1 45.0 % Pir1 26.0 % Mmi1 35.9 % Dis3 8.1% Mmi1 10.2 % Pla1 18.0 % Pir2/C725.08 6.7 % Rmn1 4.6 % Rmn1 13.2 % Brr2 13.0 % Ctr1 20.7 % Splicing Rmn1-FLAG Cwf25 3.2 % Nrl1 13.2 % Rrp42 11.7 % Pab2 Gene name Coverage Exosome 7.2 % Dis3 6.0 % Rmn1 Splicing Cwf25 3.2 % 43.3 % Cft1 5.7 % Brr2 1.6 % Red1 13.1 % RNA Seb1 5.0 % RNA Seb1 3.1 % Mtl1 9.1 % Processing Rrp5 3.5 % Processing Mtr4 2.0 % Mmi1 5.9 % Mtr4 2.0 % Pir2/C725.08 7.1 % Pab2 16.9 % RNA Pol I Rpa1 9.0 % Rpa2 6.2 % Pir2/SPBC725.08 7.1 % Spt6 6.1 %

Figure 1.3. Proteins associated with Red1, Mtl1, Pir1, and Rmn1 that were identified from the indicated purifications. Others contributed: Francisca Reyes-Turcu, Martin Zofall, and Ming Zhou.

The purified fraction also contained components of the exosome (Figure 1.3), consistent with the role of Red1 in facilitating RNA degradation by the exosome (Sugiyama and

Sugioka-Sugiyama, 2011).

We also identified new Red1 interacting partners (Figure 1.2 and 1.3), including the uncharacterized proteins SPAC17H9.02 and SPAC7D4.14c as well as Rmn1, which

12 contains an RNA recognition motif (Cho et al., 2012). SPAC17H9.02 shows significant homology to human hMtr4/SKIV2L2, which belongs to an RNA helicase family related to S. cerevisiae Ski2 (Lubas et al., 2011).

Mutations in mtl1-1: I522M, L543P, Y551H, L557P, D793G, A998V

Mtl1 1 141 272 358 522 637 765 851 1030 S. pombe DEXDc HELICc KOW_Mtr4 DSHCT 118.46 kDa hMtr4/ 1 155 286 374 521 653 770 1042 SKIV2L2 DEXDc HELICc KOW_Mtr4 DSHCT H. sapiens 117.82 kDa 1 82 117 196 279 295 Ctr1 S. pombe CC DUF 35.32 kDa 1 229 303 467 CCDC174 63 89 CC DUF 53.97 kDa H. sapeins

Figure 1.4. Predicted protein domains of Mtl1, Ctr1 and their human orthologs.

The hMTR4 domains, including DEXDc, HELICc, KOW Mtr4, and DSHCT, are conserved in SPAC17H9.02 (Figure 1.4). We named this novel protein Mtl1 (Mtr4-like protein 1). Notably, Mtl1 is distinct from the previously described Mtr4 component of

TRAMP (Buhler et al., 2007; Zhang et al., 2011). We named SPAC7D4.14c, a serine/proline rich protein, Pir1 (Protein interacting with Red1-1). In human, Red1 and

Pir1 become one protein, ZFC3H1 that contains Zn finger and Ser/Pro-rich domains.

Additional reproducibly detected proteins included the zf-C2H2 type zinc finger protein

SPBC725.08 (named Pir2, Protein Interacting with Red1-2), RNA polymerase I subunits

(Rpa1 and Rpa2), Spt6 involved in chromatin and RNA processing (Kiely et al., 2011), and an assortment of proteins involved in RNA metabolism (Figure 1.3).

13 1.2. Red1 and Mtl1 Form a Core Module that Interacts with Different Nuclear

Proteins

To explore the interactions among Red1-associated proteins, we appended MYC or FLAG epitope tags to Mtl1, Pir1 and Rmn1 for immuno-affinity purification (Figure

1.1). Mass spectrometry analyses of purified fractions revealed specific networks of interactions among Red1-associated proteins. Mtl1 purification yielded all of the major proteins identified in Red1 purifications, as well as additional proteins (Figure 1.2 and

1.3). In particular, Rmn1 and proteins that specifically associated with Mtl1 did not purify with Pir1 (Figures 1.2 and 1.3). Conversely, purification of Rmn1 yielded Red1,

Mtl1 and Pab2, but not Pir1 (Figures 1.2 and 1.3). We conclude that Red1 and Mtl1, which were detected in all four purifications, form the common core module of distinct protein assemblies.

Į)/$*,3 Į0<&,3 0WO)/$* 3LU0<& 5HG)/$* Į)/$* Į0<& Į0<& Į)/$* Į)/$*,3 Į)/$*,3 5PQ)/$* 3DE)/$* Į)/$* Į)/$*

Į0<& Į0<&

Figure 1.5. Co-IP of associated proteins from strains expressing tagged proteins. Others contributed: Francisca Reyes-Turcu.

14 α-MYC DAPI Merged

Red1

Mtl1

Pir1

Rmn1

Figure 1.6. Immunofluorescence analysis of MYC-tagged proteins. Others contributed: Nitika Taneja.

To confirm interactions among Red1-associated factors, we performed co-IPs

(Figure 1.5) and determined their subcellular localization by immunofluorescence (Figure

1.6). We found that Pir1 and Rmn1 co-IP with Red1 (Figure 1.5). However, Pir1 and

Rmn1 did not co-purify (Figures 1.2 and 1.3), consistent with their affiliation with distinct Red1-containing modules. Our analyses also confirmed the Rmn1 and Pab2 interaction (Figure 1.5).

15 Į)/$*,3 3OD)/$* Į)/$*

Į0<&

Figure 1.7. Co-IP analysis of Pla1 and Mtl1 interaction. Others contributed: Francisca Reyes-Turcu.

In addition to the previously established interaction between Pla1 and Red1 (Sugiyama and Sugioka-Sugiyama, 2011), we found that Pla1 and Mtl1 also co-IP (Figure 1.7).

Mtl1, Rmn1 and Pir1 localized to discrete foci in the nucleus (Figure 1.6), similar to

Red1 (Sugiyama and Sugioka-Sugiyama, 2011).

Figure 1.8. Red1 and Mtl1 co-localize in the nucleus. (A) Immunofluorescence analysis of Red1 and Mtl1 localization. (B) Venn diagram depicting the overlap in Mtl1- and Red1-binding sites across the genome, as determined by ChIP-chip analysis. Others contributed: Nitika Taneja and Sameet Mehta.

Furthermore, Mtl1 co-localized with Red1 (Figure 1.8A), and ChIP-chip analyses revealed shared binding sites across the genome (Figure 1.8B), suggesting that Mtl1 and

Red1 exist together in the nucleus and share many targets forming a core complex for various roles.

Pab2 Pir1 Rmn1 Nrl1 Ctr1 Pla1

Red1 Mtl1 Mtl1 MTREC

Figure 1.9. Protein interaction network of MTREC and Mtl1. (Left) MTREC, comprising Mtl1 and Red1, is the core module that interacts with Pir1, Rmn1-Pab2, or Pla1 to form different functional modules. (Right) Mtl1 also forms Red1-independent interactions with Nrl1 and Ctr1.

Based on these results, we propose that a core comprising Red1 and Mtl1 (named

MTREC, Mtl1-Red1 core complex) exists in multiple protein assemblies (Figure 1.9).

One assembly contains Pir1, and another is formed with Rmn1 and Pab2. A third includes

Pla1, which is absent from Pir1 and Rmn1 purifications (Figures 1.2 and 1.3). Mtl1 also engages in additional Red1-independent interactions (Fig. 1.9; see below).

17 1.3. Red1- and Mtl1-Associated Factors Differentially Affect Heterochromatin

Domains

A Island 1 Island 4 Island 6 Island 9 mcp7 ssm4 C8C9.04 mei4 4.5 3 3 8.5

1 1 Red1 ChIP 1 1 8.5 3 5.5 15

1 1 1 2.5 Mtl1 ChIP

UntaggedRed1-MYCMtl1-MYCPir1-MYC mei4 MYC leu1 ChIP 0.9 2.5 3.5 1.5

WCE

Figure 1.10. Localization of Red1 and Mtl1 at heterochromatin islands. (A) ChIP- chip analysis of Red1-MYC and Mtl1- MYC distribution at heterochromatin island loci. (B) ChIP analysis of Red1-MYC, Mtl1-MYC, and Pir1-MYC enrichment at the mei4 island. Numbers shown below ChIP lanes represent fold enrichments. Others contributed: Francisca Reyes-Turcu.

Red1 localizes to meiotic loci and facilitates formation of heterochromatin islands

(Tashiro et al., 2013; Zofall et al., 2012). Since Mtl1 is similarly enriched at meiotic islands (Figures 1.10A and 1.10B), we examined whether loss of Mtl1 affects

18 heterochromatin islands. Because Mtl1 is an essential protein (Kim et al., 2010), we generated a partial loss of function mutant allele, mtl1-1 (Figure 1.4).

A Island 1 Island 4 Island 6 Island 9 mcp7 ssm4 C8C9.04 mei4 WT red1∆ 10 4.5 15 30

1 1 1 5 10 mtl1-1 4.5 mtl1-1 10 mtl1-1 30 mtl1-1

1 1 1 5 10 pir1∆ 4.5 pir1∆ 10 pir1∆ 30 pir1∆ H3K9me2 ChIP 1 1 1 5 20 rmn1∆ 15 rmn1∆ 30 rmn1∆ 40 rmn1∆

1 1 5 5

WT red1∆ WT pir1∆ mtl1-1 mei4 H3K9me2 leu1 ChIP 7.7 1.0 7.7 1.1 1.2

WCE

Figure 1.11. Role of MTREC factors in the assembly of heterochromatin islands. (A) ChIP-chip analysis of H3K9me2 distribution at heterochromatin island loci in indicated strains. (B) ChIP analysis of H3K9me2 enrichment at mei4 in indicated strains. Others contributed: Francisca Reyes-Turcu and Soichiro Yamanaka.

The mtl1-1 mutant showed severe defects in H3K9me at heterochromatin islands affected in red1∆ (Figures 1.11A and 1.11B), suggesting that Mtl1 works in the same pathway

19 with Red1 to regulate formation of heterochromatin islands where they co-localize together.

# # #

UHG¨

!# #$ "# ""

Figure 1.12. Average distribution of Red1 and Mtl1 across Red1-dependent and – independent islands, as determined by ChIP-chip analysis. Also shown is average H3K9me2 enrichment in wild-type and indicated mutants. Others contributed: Sameet Mehta.

UHG¨ SLU¨ !

20 Red1-dependent Islands Protein localization H3K9me2

Island Gene Red1 Mtl1 Pir1 red1∆ mtl1-1 pir1∆ IS 1 mcp7 + + + - - - IS 2 mug8 + + + - - - IS 4 C8c9.04 + + + - - - IS 6 ssm4 + + + - - - IS 8 mcp5 + + + - - - IS 9 mei4 + + + - - - IS 16 mbx2 + + + - - - IS 17 mug45 + + + - - - IS 20 mug9 + + + - - -

Red1-independent Islands Protein localization H3K9me2

Island Gene Red1 Mtl1 Pir1 red1∆ mtl1-1 pir1∆ IS 3 C23H3.14 - - - + + + IS 7 C144.02 - - - + + + IS 10 isp4 - - - + + + IS 11 cdc28 - - - + + + IS 12 sre1 - - - + + + IS 13 C3H7.08c - - - + + + IS 14 rna.394 - - - + + + IS 15 C24C6.09c - - - + + + IS 19 C1259.02c - - - + + + IS 21 C569.06 + + + - - -

Table 1.1. Table summarizing results of ChIP-chip analysis for Red1 and Mtl1 and Pir1 localizations in wild-type cells, and H3K9me2 enrichment in wild-type and mutant strains at Red1-dependent and -independent islands. Others contributed: Francisca Reyes-Turcu.

We further analyzed the ChIP-chip results for genome-wide role of Red1 and Mtl1 in facultative heterochromatin assembly. Both Red1 and Mtl1 were specifically enriched at heterochromatin islands that require these factors for H3K9me (Figure 1.12 and Table

1.1), and Red1-dependent islands also showed strong correlation between levels of

H3K9me and Mtl1 binding (correlation coefficient 0.90, p-value < 0.0001).

21 Island 14 Island 15 Centromere 1 ncRNA LTR IRC otr1L imr1L 17G9.13c 24C6.09c 4 WT 10 WT 140 WT

1 1 20 4 red1∆ 10 red1∆ 180 red1∆

1 1 20 18 mtl1-1 14 mtl1-1 800 mtl1-1

H3K9me2 ChIP 2 2 100 3.5 pir1∆ 18 pir1∆ 1000 pir1∆

1 2 100 6 rmn1∆ 8 rmn1∆ 300 rmn1∆

1 1 50

Figure 1.13. H3K9me2 enrichment at Red1-independent islands and a pericentromeric region in indicated strains. Others contributed: Francisca Reyes- Turcu and Soichiro Yamanaka.

In addition to Red1-independent heterochromatin islands that do not show Red1 and Mtl1 enrichment and that are not affected in mtl1-1 (Figures 1.12 and 1.13 and Table 1.1), constitutive heterochromatin formed at pericentromeric region is also not affected in mtl1-1 and red1∆. These results suggest that Mtl1 and Red1 act together as components of MTREC to specifically assemble facultative heterochromatin at islands.

We expanded our analyses to include factors associated with MTREC.

Interestingly, loss of Pir1, but not Rmn1, led to defects in H3K9me at Red1-dependent heterochromatin islands (Figures 1.11A and 1.11B and Table 1.1).

22 Island 6 ssm4 2 WT red1∆ pir1∆mtl1-1 1 ssm4 Pir1 ChIP Island 9 mei4

+RT act1 3 mei4 - RT 1

Pir1 ChIP

Figure 1.14. Pir1 is localized at heterochromatin islands. Pir1-MYC distribution at heterochromatin islands (left) and RT-PCR analysis of ssm4 and mei4 transcript levels in wild-type and indicated mutants (right). The act1 locus was used as a control. +RT and – RT indicate the presence or absence of reverse transcriptase. Others contributed: Francisca Reyes-Turcu.

A low level of Pir1 enrichment was observed at heterochromatin islands, consistent with its direct involvement, and its loss affected silencing of target loci (Figure 1.14). No significant changes in H3K9me could be detected in cells carrying a mutation in Pla1, another MTREC-associated factor.

Red1 also affects RNAi-dependent assembly of HOODs (Yamanaka et al., 2013).

Pab2 interacts with Rmn1 (Figure 1.5), and is required along with Pla1 for H3K9me at all known HOODs (Yamanaka et al., 2013). We investigated the role of MTREC factors in

HOOD assembly by creating deletions in cells that lack Rrp6. Rrp6 acts parallel to RNAi to silence loci within HOODs and consequently, rrp6∆ cells show a significant increase in siRNA-dependent heterochromatin modifications under lab growth conditions

(Yamanaka et al., 2013).

23 HOOD 3 HOOD 4 HOOD 10 HOOD 24 myp2 Tf2-3 Tf2-5 LTR C1235.01 rrp6∆ rrp6∆ rmn1∆ 12 10 12 3.5

2 2 2 1 rrp6∆ pir1∆ rrp6∆ pir1∆ rrp6∆ pir1∆ rrp6∆ pir1∆ 12 12 12 8 H3K9me2 ChIP 2 2 2 1

20 3 3 10

rrp6∆ 20 3 3 10 20 3 3 10 Small RNA rrp6∆

rmn1∆ 20 3 3 10

(normalized reads)

Figure 1.15. Effect of Rmn1 and Pir1 at HOODs. (Top) ChIP-chip analysis of the effect of pir1∆ and rmn1∆ on H3K9me2 distribution at HOODs in rrp6∆. (Bottom) Normalized number of small RNA reads plotted in alignment with HOOD loci. The signals above and below the line represent small RNAs that map to the top and bottom DNA strands, respectively. Others contributed: Francisca Reyes-Turcu and Venkata Chalamcharla.

To investigate the role of MTREC-associated factors, we performed ChIP-chip for H3K9me2 with mutants of Rmn1 and Pir1 in Rrp6 deletion background. Remarkably, loss of Rmn1, but not Pir1, abolished RNAi-dependent H3K9me at HOODs (Figure

1.15). Defects in H3K9me correlated with defects in siRNA production at affected loci

(Figure 1.15). These results show that Rmn1, but not Pir1, is involved in RNAi- dependent mechanism for heterochromatin assembly at HOODs.

24 Table S1. Rmn1 facilitates small RNA production and H3K9me at genes and retrotransposons

Genomic Location Genomic Small RNAs H3K9me2 regions features Chr Start End rrp6∆ rrp6∆ rrp6∆ rrp6∆ rmn1∆ rmn1∆ 1 1 9289 tlh1 + + + + 1 17000 19000 SPAC212.06c + + + + 1 21377 22666 SPAC212.04c meiosis↑ + + + +/– 1 28579 32648 SPAC212.01c meiosis↑ SPAC977.01 memb + + + +/–

1 59551 60745 SPAC977.14c memb + – + – 1 5564170 5570038 SPAC750.04c memb SPAC750.05c meiosis↑, memb SPAC750.06c meiosis↑, memb (Repeat region 5204, + + + +/– Repeat region 5205, Repeat region 5206, Repeat region 5207) 1 5572541 5574430 (Repeat region 5207, Repeat region 5212) + + + + Subtelom- 2 5244 9353 SPBC1348.01 meiosis↑, memb eric SPBC1348.02 meiosis↑, memb regions (Repeat region 1a, + + + + Repeat region 8b) 2 4463576 4466315 SPBPB2B2.03c memb SPBPB2B2.04 meiosis↑ + + + + 2 4501916 4507716 SPBPB2B2.17c memb SPBPB2B2.18 SPBPB2B2.19c meiosis↑, memb SPBCPT2R1.01 meiosis↑, memb + + + + (Repeat region 4340, Repeat region 4352) 2 4512962 4514857 SPBCPT2R1.04c meiosis↑, memb (Repeat region 4358) + + NP NP 2 4515585 4515702 (intergenic region) + + NP NP 2 4517451 4519112 SPBCPT2R1.09c + + NP NP 2 4523251 4532665 SPBCPT2R1.07c SPBCPT2R1.10 + + NP NP tlh2 HOOD-1 1 1465847 1469848 tf2-1 + – + – HOOD-2 1 1564163 1568414 tf2-2 + – + – HOOD-3 1 2544835 2561773 isp6 meiosis↑ myp2 meiosis↑ SPAC4A8.06c + – + – SPAC4A8.07c vas1 HOOD-4 1 2927156 2941954 tf2-3 tf2-ORF(truncated) + – + – HOOD-5 1 2977013 2988899 mfc1 memb SPAPB1A11.02 SPAPB1A11.03 – – + – mca1 meiosis↑ HOOD-6 1 2994807 3009469 lys12 SPAC31G5.05c rgg8 dni1 meiosis↑, memb ups1 – – + – spk1 meiosis↑ eta2 meiosis↑ pac2 meiosis↑ maf1 HOOD-7 1 3361499 3365727 tf2-4 + – + – HOOD-8 1 3736902 3743575 SPAP7G5.03 meiosis↑, memb + + +* +* lys1

25 HOOD-9 1 3791825 3796114 rad50 + + +* +* HOOD-10 1 3996353 4000615 tf2-5 + – + – HOOD-11 1 4022326 4026545 tf2-6 + – + – HOOD-12 1 5069824 5083205 SPAC1006.02 meiosis↑ SPAC1006.03c meiosis↑ mcp3 meiosis↑ + – + – och1 meiosis↑ rgf2 meiosis↑ HOOD-13 1 5191103 5195325 tf2-7 + – + – HOOD-14 1 5195656 5199909 tf2-8 + – + –

HOOD-15 1 5234000 5250176 SPAC14C4.04 meiosis↑ man1 meiosis↑, memb SPAC14C4.06c SPAC14C4.07 memb mug5 meiosis↑ + – + – agn1 meiosis↑ SPAC14C4.10c meiosis↑, memb SPAC14C4.11 memb HOOD-16 2 91600 101684 SPBPB10D8.04c memb SPBPB10D8.05c memb + + + +/– SPBPB10D8.06c memb SPBPB10D8.07c memb HOOD-17 2 347505 354250 SPBC1271.09 meiosis↑, memb SPBC1271.10c memb + – + – (Intergenic region) HOOD-18 2 898107 902233 mcp5 meiosis↑ + + + – HOOD-19 2 1812684 1816937 tf2-9 + – + –

HOOD-20 2 1965175 1969519 tf2-10 pseudo + – + – HOOD-21 2 2126590 2128479 SPBC23G7.14 meiosis↑ + + +** +** rpp202 HOOD-22 2 4414469 4418768 tf2-11 + – + – HOOD-23 2 4442538 4449562 SPBC8E4.02c meiosis↑ SPBC8E4.01c memb + + + + pho1 meiosis↑ HOOD-24 3 173841 176400 SPCC1235.01 (internal repeats) + – + – HOOD-25 3 254411 256353 tf2-ORF(truncated) + – + – HOOD-26 3 778123 782331 tf2-12 + – + –

HOOD-27 3 1047657 1056145 SPCC1259.08 meiosis↑ SPCC1259.09c pgp1 + +/– +* +* gyp2 HOOD-28 3 1168500 1176000 nte1 SPCC4B3.03c meiosis↑, memb + – +* +* SPNCRNA.120 HOOD-29 3 1179500 1182650 rhp26 + + +* +* HOOD-30 3 1196050 1196500 tf2-ORF(truncated) + +/– + + HOOD-31 3 1763512 1775613 SPCC1450.16c memb ste6 meiosis↑ SPCC1442.02 meiosis↑ + + + + SPCC1442.03 SPCC1442.04c meiosis↑ HOOD-32 3 2320230 2324503 tf2-13-pseudo + – + –

Genomic Location Genomic Small RNAs regions features Chr Start End rrp6∆ rrp6∆ rmn1∆ 1 1511797 1512189 rps2202 + + 1 5090060 5094840 win1 Small RNA + – 1 5274775 5281509 SPAC2H10.01 meiosis↑ Clusters that do not show SPAC2H10.02c meiosis↑ + – H3K9me swr1 SPAC11E3.02c meiosis↑ 2 1923486 1925465 erg6 + +

– – not detected +/– – major reduction +* – near centromeric region +** – near mat locus NP – no probe on microarray meiosis↑ – genes upregulated during mating or meiosis memb – genes encoding transmembrane domain proteins underlined – genes that overlap with annotated non-coding RNA Tf2 – considering high similarity among tf2 ORFs, it was not possible to define the exact origin of the signal

Table 1.2. Rmn1 facilitates small RNA production and H3K9me at genes and retrotransposons. Others contributed: Francisca Reyes-Turcu and Venkata Chalamcharla.

UUS¨ UUS¨ UPQ¨

Figure 1.16. Quantitative analysis of small RNAs at HOODs in rrp6∆ and rrp6∆ rmn1∆ strains. Staples indicate 5 percentile (lower) and 95 percentile (upper). The lower and upper bounds of the boxes denote 25 and 75 percentile, respectively. Others contributed: Sameet Mehta.

27 We further analyzed the ChIP-chip data for the genome-wide role of Rmn1 in regulation of siRNA and heterochromatin formation at HOODs. Table for each HOODs locus (Table 1.2) and the genome-wide quantitative analysis (Figure 1.16) show that

Rmn1 plays a critical role in facilitating small RNA production and H3K9me at genes and retrotransposons across the genome.

Results showed that Mtl1 and Red1 form a core, MTREC, that associates with other factors such as Pir1 and Rmn1 to differentially regulate formation of heterochromatin islands and HOODs. MTREC assembly that contains Pir1 regulates

RNAi-independent heterochromatin assembly at meiotic genes while the assembly with

Rmn1 regulates RNAi-dependent heterochromatin assembly at genes and retrotransposons.

28 Chapter 2. MTREC Regulates Gene Expression and Targets Noncoding RNAs and

Pre-mRNAs Degraded by the Exosome

2.1. Mtl1 Regulates Expression of Genes Involved in Sexual Differentiation, Stress

Response and Membrane Transport

Red1 coordinates exosome-dependent repression of meiotic genes in vegetative cells (Sugiyama and Sugioka-Sugiyama, 2011; Zofall et al., 2012). Mtl1 and Red1 interaction and co-localization at several genomic sites (Figures 1.5, 1.8B and 1.10A) prompted us to investigate if they collaborate to regulate gene expression. We detected increased expression of various mRNAs in mtl1-1, as compared to wild type (Appendix

A). A majority of genes affected in mtl1-1 were also upregulated in red1∆ and rrp6∆

(correlation coefficient 0.63 and 0.65, respectively, p-value < 0.0001) (Appendix A).

mcp5 SPNCRNA.103 (sme2) ncRNA 3 2.5

Red1 ChIP 1 1 3 2.5

1 Mtl1 ChIP 1

1000 WT

red1∆

mtl1-1 Expression

(normalized reads) rrp6∆

Figure 2.1. Effect of MTREC on genes and ncRNAs. (Top) ChIP-chip analysis of Red1-MYC and Mtl1-MYC distribution at indicated loci. (Bottom) RNA-seq analysis of Red1- and Mtl1-bound genes in red1∆, mtl1-1, and rrp6∆ strains. Others contributed: Francisca Reyes-Turcu and Martin Zofall.

29 WT red1∆ mtl1-1 rrp6∆ mcp5 rec10 mei4 ssm4

sme2 ncRNA.488 ncRNA.405 ncRNA.247

+ RT act1 - RT

Figure 2.2. RT-PCR analysis of Red1- and Mtl1-bound loci in indicated strains. The transcript level of act1 was used as a control. +RT and –RT indicate presence or absence of reverse transcriptase.

For example, the mcp5 gene, which is bound by Red1 and Mtl1, was upregulated in mtl1-

1 as well as red1∆ and rrp6∆ (Figure 2.1). We confirmed that loci affected in red1∆ and rrp6∆ were also upregulated in mtl1-1 by RT-PCR (Figure 2.2).

The upregulated loci belong to three distinct groups (Appendix A). The first group contains genes involved in meiosis (e.g. crs1, mug1, ssm4 and mei4) or that show an increase in expression during meiotic induction (e.g. mug8, mug9 and meu10). The second group includes genes implicated in cellular stress response (Appendix A). The third group consists of genes that encode transmembrane proteins involved in transport of amino acids, ions and nutrients (Appendix A). Other loci affected in all three mutants are metabolic genes and pseudogenes. These results indicate that the interaction between

Red1 and Mtl1 is important to suppress mRNAs targeted by the exosome.

30 2.2. MTREC Targets Regulatory Noncoding RNA and Pre-mRNA Degraded by the

Exosome

1672.09 snR99 1672.10 mim1 c713.09 snR3 2.5 2

1 1 Red1 ChIP 2.5 2.5

1 1 Mtl1 ChIP

WT red1∆ mtl1-1 rrp6∆ WT red1∆ mtl1-1 rrp6∆ dT - + - + - + - + dT - + - + - + - + + A+ A

A-

A- snR99 snR3

snR99 snR3 rRNA rRNA

Figure 2.3. Effect of MTREC on snoRNAs. (Top) ChIP-chip analysis of Red1-MYC and Mtl1-MYC distribution at snoRNAs. (Bottom) Northern analysis of snR99 and snR3 processing in mutant strains. –/+ dT denotes RNase H treatment in the absence or presence of oligo dT. A–/A+ indicates nonpolyadenylated and poly- adenylated species. Lighter exposures of mature snR99 and snR3 are shown below. rRNA was used as a loading control. Others contributed: Francisca Reyes-Turcu.

The S. pombe genome encodes a large number of ncRNAs (Bitton et al., 2011; Dutrow et al., 2008; Rhind et al., 2011; Wilhelm et al., 2008). Mtl1 and Red1 localized to ribosomal genes and loci encoding noncoding RNAs such as lncRNAs, tRNAs and snoRNAs

(Figure 2.3). Significantly, mtl1-1, red1∆ and rrp6∆ all showed extensive changes in

31 ncRNA levels (Appendix A). A ncRNA produced from the sme2 locus was among a cohort of upregulated transcripts including ~300 previously identified ncRNAs and several unannotated transcripts (Appendix A). The sme2 ncRNA accumulates during meiotic induction and promotes pairing of homologous chromosomes (Ding et al., 2012), but is not detected in vegetative cells. Red1 and Mtl1 enrichment at sme2 prevented the accumulation of ncRNA in vegetative cells, similar to Rrp6 (Figure 2.1) (Sugiyama and

Sugioka-Sugiyama, 2011). We confirmed the increase in sme2 and other ncRNAs in mtl1-1, red1∆ and rrp6∆ mutants by RT-PCR (Figure 2.2). Thus, MTREC regulates the abundance of ncRNAs that are substrates of the nuclear exosome.

1006.09 1250.04c

rpl30-2 WT red1∆mtl1-1rrp6∆ 2 1250.07

pre-mRNA 1 Red1 ChIP rpl30-2 1.5 mRNA

rRNA 1 Mtl1 ChIP

Figure 2.4. Effect of MTREC on ribosomal RNA. (Left) Red1-MYC and Mtl1-MYC distribution at rpl30-2. (Right) Northern blot analysis of rpl30-2 pre-mRNA processing in the indicated strains. Others contributed: Francisca Reyes-Turcu.

Red1 and Mtl1 are enriched at majority of snoRNA loci, including snR3 and snR99 (Figure 2.3) and ribosomal genes (Figure 2.4), the processing of which requires

Rrp6 and Pab2 (Lemay et al., 2010; Lemieux et al., 2011). Northern blot analysis using snoRNA probes showed accumulation of a large heterogeneous population of transcripts in rrp6∆ (Figure 2.3). The larger species represent hyperadenylated transcripts that appeared as discrete bands upon oligo dT-directed RNase H cleavage (Figure 2.3). The

32 mtl1-1 and red1∆ mutants accumulated 3’-extended snoRNAs, but polyadenylation levels in these mutants were lower than in rrp6∆ (Figure 2.3). Whereas RNase H cleavage caused 3’-extended snR99 to collapse to a single discrete product, cleavage of snR3 yielded several bands, possibly signifying transcript termination at multiple sites in mtl1-

1 and red1∆ (Figure 2.3). We propose that Mtl1 and Red1 cooperate with polyadenylation and exosome machinery to process and degrade pre-snoRNAs, and may also play a role in transcription termination.

MTREC was also required for degradation of pre-mRNA from ribosomal gene rpl30-2, which contains a 246 nt intron. Similar to rrp6∆, mtl1-1 and red1∆ accumulated unspliced pre-mRNAs with a slight increase in spliced rpl30-2 transcript (Figure 2.4).

Thus, MTREC functions in the turnover of specific intron-containing pre-mRNAs, in addition to meiotic mRNA and ncRNA decay.

2.3. Noncoding RNA Regulates Gene Expression in Response to Environmental

Changes

8E4.01c pho1 4G3.03

ncRNA 12 8

4 Red1 ChIP

10 Mtl1 ChIP

Figure 2.5. ChIP-chip analysis of Red1-MYC and Mtl1- MYC distribution at the pho1 locus. Others contributed: Francisca Reyes-Turcu.

33 A C 11E10.01 1271.09 ncRNA ncRNA 10 15

Red1 ChIP 2 15 40

5 Mtl1 ChIP 5 35 70

Rrp6 ChIP 5 5 200 500 WT 200 500 red1∆ 200 500 mtl1-1 Expression 200 500

(normalized reads) rrp6∆

B D ncRNA pho1 Loci showing prominent Red1 and Mtl1 peaks 15 corresponding to upstream ncRNA ags1 5

Rrp6 ChIP bgs4 but2 200 byr2 WT pma1 200 tea1 red1∆ SPAC631.02 200 SPAPB15E09.01c mtl1-1 SPBC1652.01 Expression 200 SPBPJ4664.02

(normalized reads) rrp6∆ SPCC132.04c

Figure 2.6. ncRNA regulates gene expression in response to environmental changes (A) ChIP-chip analysis of Red1-MYC, Mtl1-MYC, and Rrp6-MYC enrichment at an unannotated ncRNA upstream of the SPBC1271.09 gene (top). RNA- Seq analysis is shown for wild-type, red1∆, mtl1-1, and rrp6∆ (bottom). (B) ChIP-chip analysis of Rrp6 enrichment upstream of the pho1 locus (top). RNA-Seq analysis is shown for wild-type, red1∆, mtl1-1, and rrp6∆ (bottom). (C) ChIP-chip analysis of

34 Red1-MYC, Mtl1-MYC, and Rrp6-MYC enrichment at an unannotated ncRNA upstream of the SPCC11E10.01 gene (top). RNA-Seq analysis is shown for wild-type, red1∆, mtl1- 1, and rrp6∆ (bottom). (D) Additional examples of genes that contain upstream ncRNA with Red1 and Mtl1 enrichment. Others contributed: Francisca Reyes-Turcu and Martin Zofall.

The biological function of widespread ncRNA production is unknown. We noted prominent Red1, Mtl1 and Rrp6 peaks upstream of several protein-coding genes, such as pho1, SPBC1271.09 and SPCC11E10.01 (Figures 2.5 and 2.6A-D). These peaks correlate with ncRNAs that accumulate in red1∆, mtl1-1 and rrp6∆ cells (Figures 2.6A-C). We hypothesized that upstream ncRNAs recruit RNA processing factors, and might regulate adjacent genes.

WT rrp6∆ ncRNA - - + RNase H mRNA kb 3.0 ncRNA pho1 1.5 ncRNA*

rRNA

Figure 2.7. Detection of ncRNA at pho1 locus. (Left) Schematic depicting transcription of ncRNA and mRNA at the pho1 locus. The location of the northern blot probe is indicated (thick wavy line). The dotted line indicates the site cut by RNase H. (Right) Northern blot of ncRNA in wild- type and rrp6∆strains. The asterisk denotes ncRNA cleaved by oligo-directed RNase H treatment. Lanes shown are from the same northern blot with additional control lanes removed. Others contributed: Martin Zofall.

To test this, we focused on the pho1 locus. We detected upstream ncRNA accumulation in rrp6∆ (Figures 2.7 and 2.6B). Results from oligo-directed RNase H cleavage were consistent with upstream initiation and continuation of transcription through pho1 in rrp6∆ cells (Figure 2.7).

35 Expression of pho1 occurs only in phosphate-limiting conditions. To explore the effect of the ncRNA on pho1, we replaced the -400 to -1200 bp region upstream of pho1 with ura4+ to generate ncRNA∆.

No Tag Red1-MYC No Tag Mtl1-FLAG + + ∆ ncRNA + + ∆ ncRNA pho1 MYC FLAG leu1 ChIP ChIP 0.8 7.4 0.7 0.9 2.8 0.5

WCE WCE

Figure 2.8. ChIP analysis of Red1-MYC and Mtl1-FLAG enrichment at the pho1 locus. Numbers shown below ChIP lanes represent fold enrichments.

A B C

WT ncRNA∆ WT WT ncRNA∆ WT rrp6∆clr4∆rrp6∆ clr4∆ - + + Phosphate - + + Phosphate - + + + + Phosphate kb kb pho1 H3K9me2 3.0 3.0 leu1 ChIP ncRNA 1.5 pho1 1.3 2.4 1.1 1.5 pho1 mRNA mRNA WCE rRNA rRNA

Figure 2.9. Role of ncRNA in regulation of pho1 locus. (A) Northern blot analysis of pho1 mRNA expression in ncRNA∆. (B) ChIP analysis of H3K9me2 enrichment at pho1 in wild-type and ncRNA∆. Numbers shown below ChIP lanes represent fold enrichments. (C) Northern blot analysis of ncRNA and pho1 mRNA expression in the indicated strains. Others contributed: Martin Zofall.

Deletion of ncRNA abolished Red1 and Mtl1 recruitment to pho1 (Figure 2.8), and de- repressed pho1 mRNA in the presence of phosphate (Figure 2.9A). Moreover, in the presence of phosphate H3K9me could be detected at pho1 in wild-type, but not in ncRNA∆ cells (Figure 2.9B). Loss of the sole H3K9 methyltransferase, Clr4, had no major impact on pho1 expression (Figure 2.9C). However, when clr4∆ was combined with rrp6∆, synergistic accumulation of pho1 mRNA was observed (Figure 2.9C).

Together, these results implicate ncRNA in the regulation of gene expression. To perform this function, ncRNA utilizes not only heterochromatin machinery, but also RNA processing factors to degrade gene transcripts.

37 Chatper 3. Mtl1 Forms Red1-Independent Interactions with Nrl1 and Ctr1 for

Regulation of Splicing Associated with Heterochromatin Assembly and Telomere

Maintenance.

3.1. Mtl1 Associates with Nrl1 and Ctr1 without Red1

Among the proteins specific to the Mtl1 purification were the putative homolog of C. elegans NRDE-2, Nrl1, and the coiled-coil and DUF4078 domain containing protein

SPAC140.04, which we named Ctr1 (Coiled-coil domain telomerase regulatory protein)

(Figures 1.2 and 1.9). Ctr1 is evolutionarily conserved and shares homology with human

CCDC174 protein of unknown function (Figure 1.4).

A B

UntaggedNrl1-MYC UntaggedCtr1-MYC Gene Mtl1-FLAG Nrl1-MYC Ctr1-MYC 191 name Coverage Coverage Coverage 191 97 97 Mtl1 71.0 % 45.7 % 35.8 % 64 Red1 78.2 % 0 % 0 % 64 51 Pir1 60.1 % 0 % 0 % 51 39 Ctr1 20.7 % 82.7 % 68.1 % 39 28 Nrl1 13.2 % 33.5 % 20.6 % 28 14 Rmn1 13.2 % 0 % 0 % 14

Figure 3.1. Protein purification of Nrl1 and Ctr1. (A) Expression of Nrl1-MYC and Ctr1-MYC. The arrow indicates MYC-tagged Nrl1 or Ctr1. (B) Proteins associated with Mtl1, Nrl1, or Ctr1. Proteins highly associated with each other are shaded in blue. Others contributed: Venkata Chalamcharla and Ming Zhou.

38 Nrl1-MYC Nrl1-FLAG Ctr1-MYC Gene name Coverage Coverage Gene name Coverage Ctr1 82.7 % 80.7 % Ctr1 68.1 % Mtl1 45.7 % 61.3 % Mtl1 35.8 % Nrl1 33.5 % 44.3 % Nrl1 20.6 % Prp43 71.0 % 41.3 % Cwf10 23.8 % Ntr2 30.0 % 34.9 % Smd2 17.4 % C1486.03c 24.6 % 33.2 % Tgs1 15.1 % Cwf10 23.7 % 37.6 % Smd1 11.1 % Cwf17 23.5 % 62.3% Cwf11 9.8 % C20H4.06c 17.4 % 41.9 % Prp19 9.2 % Prp5 13.3 % 51.4 % Smb1 8.2 % Cwf11 4.9 % 30.4 % Spp42 6.1 %

Figure 3.2. Proteins associated with Nrl1 and Ctr1 that were identified from the indicated purifications. Components of the Sm protein complex identified in the Ctr1 purification are underlined. Others contributed: Venkata Chalamcharla and Ming Zhou.

We tagged Nrl1 and Ctr1 with the MYC epitope (Figure 3.1A). Nrl1-MYC purification identified Mtl1 and Ctr1 (Figure 3.1B), as well as several splicing factors (Figure 3.2, see below). Red1, Pir1, Rmn1 and the exosome were not present (Figures 3.1B and 3.2).

Ctr1-MYC purification also identified Nrl1, Mtl1 and splicing factors, as well as the core spliceosomal Sm proteins and the Tgs1 RNA methyltransferase (Figure 3.2), suggesting a possible role in splicing related processes.

39 A B α-MYC DAPI Merged Input α-FLAG IP Mtl1-FLAG - + - + Nrl1-MYC + + + + Nrl1 α-FLAG α-MYC

Input α-MYC IP Ctr1-MYC - + - + + + + + Ctr1 Mtl1-FLAG α-MYC α-FLAG

Figure 3.3. Nrl1 and Ctr1 interacts with Mtl1. (A) Immunofluorescence analysis of MYC-tagged Nrl1 and Ctr1. (B) Co-IP analysis of Mtl1 with Nrl1 and Ctr1. Others contributed: Nitika Taneja and Francisca Reyes-Turcu.

Nrl1 and Ctr1 localized to the nucleus, similar to Mtl1 (Figure 3.3A). Interactions between these factors were confirmed by co-IP (Figure 3.3B).

3.2. Nrl1 Promotes Assembly of HOODs at Genes and Retrotransposons

HOOD 10 HOOD 31 ste6 LTR TF2-5 1442.04c LTR LTR 1442.02 3 20

rrp6∆ 3 20 3 20 Small RNA nrl1∆ rrp6∆

(normalized reads) 3 20 rrp6∆ rrp6∆ nrl1∆ 10 35

2 5

H3K9me2 ChIP

Figure 3.4. Role of Nrl1 at HOODs loci. (Top) Normalized number of small RNA reads plotted in alignment with HOOD 10 and HOOD 31 loci in rrp6∆ and rrp6∆ nrl1∆ strains. (Bottom) ChIP-chip analysis of H3K9me2 at HOODs. Others contributed: Venkata Chalamcharla and Francisca Reyes-Turcu.

40 A B HOOD 3 HOOD 4 Island 6 Island 9 myp2 Tf2-3 ssm4 LTR mei4 15 WT 30 WT 10 3

rrp6∆ 10 3 5 10 3 1

Small RNA 15 nrl1∆ 30 nrl1∆ nrl1∆ rrp6∆ 10 3 (normalized reads)

rrp6∆ rrp6∆ nrl1∆ 1 5 H3K9me2 ChIP 10 10 18 ctr1∆ 30 ctr1∆

ChIP 2 2 5 H3K9me2 2

Figure 3.5. Effect of nrl1∆ at different heterochromatin domains. (A) Normalized small RNA reads in rrp6∆ and rrp6∆ nrl1∆ plotted in alignment with the loci map of HOOD 3 and HOOD 4 (top). The signals above and below the line represent small RNAs that map to the top and bottom DNA strands, respectively. ChIP-chip analysis of H3K9me2 in rrp6∆ and rrp6∆ nrl1∆ (bottom). (B) ChIP-chip analysis of H3K9me2 enrichment at heterochromatin islands in wild-type, nrl1∆ and ctr1∆. Others contributed: Venkata Chalamcharla and Francisca Reyes-Turcu.

C. elegans NRDE-2 is involved in the nuclear RNAi pathway and mediates changes in chromatin structure (Guang et al., 2010). Production of small RNAs and heterochromatin formation at HOODs are severely affected in nrl1∆ rrp6∆ cells (Figures 3.4 and 3.5A;

Table 1.2). However, nrl1∆ had no major impact on siRNAs and H3K9me at centromeres

(data not shown) or assembly of heterochromatin islands formed by RNAi-independent mechanisms (Figure 3.5B). Thus, Nrl1 selectively affects RNAi-dependent assembly of heterochromatin at retrotransposons and developmental genes, and is dispensable for heterochromatin formation at centromeres and meiotic islands.

41 3.3. Nrl1 Interacts with Splicing Factors to Assemble HOODs via Cryptic Introns

A B

Gene Nrl1-MYC Nrl1-FLAG Input α-FLAG IP name Coverage Coverage Cwf10-FLAG - + - + Prp43 71.0 % 41.3 % Nrl1-MYC + + + + Splicing Ntr2 30.0 % 34.9 % α-FLAG Factors Cwf10 23.7 % 37.6 % Cwf17 23.5 % 62.3% Prp5 13.3 % 51.4 % α-MYC * Cwf11 4.9 % 30.4 %

Figure 3.6. Nrl1 interacts with splicing factors. (A) Splicing factors identified in Nrl1 purifications. (B) Co-IP analysis of the interaction between Cwf10 and Nrl1. Asterisk indicates the correct size of Nrl1. Multiple Nrl1 bands may indicate protein modifications. Others contributed: Francisca Reyes-Turcu and Ming Zhou.

As described above, Nrl1 co-purified with factors involved in pre-mRNA splicing.

Independent purifications of Nrl1-MYC and Nrl1-FLAG contained components of U2 and U5 small nuclear ribonucleic particles (snRNPs), and other splicing factors (Figure

3.6A). Among these proteins was Cwf10, a homolog of S. cerevisiae U5 snRNP Snu114 and of human EFTUD2 (Jurica and Moore, 2003), which has been shown to interact with

Cid12, a subunit of the RNA-dependent RNA polymerase involved in RNAi (Bayne et al., 2008). Co-IP analysis confirmed that Nrl1 interacts with Cwf10 (Figure 3.6B).

42 HOOD 31 HOOD 10 SPCC1442.04c TF2-5

LTR LTR int2 int1 WT 4 3016 WT 1 12107

nrl1∆ 0 6610 nrl1∆ 0 28750

cwf10-1 0 2111 cwf10-1 0 32666

rrp6∆ 57 12085 rrp6∆ 6 12412

Intron junction reads red1∆ 111 21212 red1∆ 16 75711

mtl1-1 72 32133 mtl1-1 11 103524

Figure 3.7. Schematic of cryptic introns detected at HOODs by RNA-seq in the indicated strains. Tf2 and SPCC1442.04c are shown with cryptic introns (brown). The arcs below the line represent intron junction reads that map to the bottom DNA strands. The thickness of the arc corresponds to the number of reads detected. The total unnormalized read count for each locus, which is indicative of sequencing depth, is shown. The gray line denotes antisense RNA. Others contributed: Francisca Reyes- Turcu and Martin Zofall.

LTR Tf2-5 SPCC1442.04c LTR LTR

Intron Intron 2 Intron 1 rrp6∆

Intron: I:3998944-3999006 Intron 1: III:1775136-1775205 AAGTATGTCA...... AATACAAGGT ACGTATGACA...... TATTAGAGAG Intron 2: III:1774146-1775011 AGGTACGACG...... TGAGCCAGGA

Figure 3.8. Schematic of RNA-Seq reads that map to Tf2 and SPCC1442.04c in rrp6∆, which contain cryptic introns. Genomic location of the introns and sequences at intron junctions are shown below. 30 and 50 splice sites are indicated with bold blue lettering. Others contributed: Martin Zofall.

We hypothesized that Nrl1 might cooperate with splicing factors to promote siRNA production at retrotransposons and genes. Surprisingly, analyses of RNA-Seq data from wild-type and mutants revealed unannotated cryptic introns in Tf2 and

SPCC1442.04c loci showing siRNA clusters (Figures 3.7 and 3.8). The red1∆, mtl1-1 and rrp6∆ mutants that were defective in silencing of SPCC1442.04c and Tf2 showed higher levels of spliced reads, as compared to wild-type. However, splicing was not detected in nrl1∆, even amongst a greater number of Tf2 and SPCC1442.04c sequencing reads as compared to wild type (Figure 3.7).

ste6 1442.04c 1442.02 20

rrp6∆ 20 20 Small RNA

rrp6∆ 20 04c-int2∆ (normalized reads)

rrp6∆rrp6∆04c-int2∆ rrp6∆ rrp6∆04c-int2∆ 1442 .04c leu1 6.0 1.2 H3K9me2 WCE ChIP

Figure 3.9. Role of SPCC1442.04c intron 2. (Top) Normalized number of small RNA reads at HOOD 31 in rrp6∆ and rrp6∆ 04c-int2∆. (Bottom) ChIP analysis of H3K9me2 enrichment. Numbers shown below ChIP lanes represent fold enrichments. Others contributed: Venkata Chalamcharla.

We asked whether the introns in RNAi-targeted loci are involved in generation of small RNAs and H3K9me. Deletion of an intron in SPCC1442.04c severely affected

44 production of siRNAs and H3K9me at this locus (Figure 3.9), and also caused defects in production of siRNAs at surrounding loci including ste6, which is involved in sexual differentiation (Figure 3.9). siRNA production at centromeric repeats and other HOODs was not affected. These results are consistent with our previous finding that deletion of

RNAi-dependent heterochromatin nucleation sites affects expression of neighboring loci within HOODs (Yamanaka et al., 2013). Our results suggest that a cryptic intron, which is differentially spliced in the absence of Nrl1, serves to recruit RNAi and target

H3K9me.

45 Table S1. Rmn1, Nrl1 and Cwf10 facilitate small RNA production and H3K9me at genes and retrotransposons, Related to Figures 2, 5, and 6.

Genomic Location Genomic Small RNAs H3K9me2 regions features Chr Start End rrp6∆ rrp6∆ rrp6∆ rrp6∆ rrp6∆ rrp6∆ rrp6∆ rrp6∆ rmn1∆ nrl1∆ cwf10-1 rmn1∆ nrl1∆ cwf10-1 1 1 9289 tlh1 + + + + + + + + 1 17000 19000 SPAC212.06c + + +/– +/– + + + + 1 21377 22666 SPAC212.04c meiosis↑ + + – – + +/– + + 1 28579 32648 SPAC212.01c meiosis↑ SPAC977.01 memb + + – – + +/– + +

1 59551 60745 SPAC977.14c memb + – – + + – – + 1 5564170 5570038 SPAC750.04c memb SPAC750.05c meiosis↑, memb SPAC750.06c meiosis↑, memb (Repeat region 5204, + + – – + +/– + + Repeat region 5205, Repeat region 5206, Repeat region 5207) 1 5572541 5574430 (Repeat region 5207, Repeat region 5212) + + +/– +/– + + + + Subtelom- 2 5244 9353 SPBC1348.01 meiosis↑, memb eric SPBC1348.02 meiosis↑, memb regions (Repeat region 1a, + + – – + + + + Repeat region 8b) 2 4463576 4466315 SPBPB2B2.03c memb SPBPB2B2.04 meiosis↑ + + – +/– + + + + 2 4501916 4507716 SPBPB2B2.17c memb SPBPB2B2.18 SPBPB2B2.19c meiosis↑, memb SPBCPT2R1.01 meiosis↑, memb + + – – + + + + (Repeat region 4340, Repeat region 4352) 2 4512962 4514857 SPBCPT2R1.04c meiosis↑, memb (Repeat region 4358) + + – – NP NP NP NP 2 4515585 4515702 (intergenic region) + + – – NP NP NP NP 2 4517451 4519112 SPBCPT2R1.09c + + +/– +/– NP NP NP NP 2 4523251 4532665 SPBCPT2R1.07c SPBCPT2R1.10 + + + + NP NP NP NP tlh2 HOOD-1 1 1465847 1469848 tf2-1 + – – – + – – +/– HOOD-2 1 1564163 1568414 tf2-2 + – – – + – – – HOOD-3 1 2544835 2561773 isp6 meiosis↑ myp2 meiosis↑ SPAC4A8.06c + – – – + – – – SPAC4A8.07c vas1 HOOD-4 1 2927156 2941954 tf2-3 tf2-ORF(truncated) + – – – + – – – HOOD-5 1 2977013 2988899 mfc1 memb SPAPB1A11.02 SPAPB1A11.03 – – – – + – – + mca1 meiosis↑ HOOD-6 1 2994807 3009469 lys12 SPAC31G5.05c rgg8 dni1 meiosis↑, memb ups1 – – – – + – – – spk1 meiosis↑ eta2 meiosis↑ pac2 meiosis↑ maf1 HOOD-7 1 3361499 3365727 tf2-4 + – – – + – – –

HOOD-8 1 3736902 3743575 SPAP7G5.03 meiosis↑, memb lys1 + + – + +* +* – +* HOOD-9 1 3791825 3796114 rad50 + + – + +* +* +* +* HOOD-10 1 3996353 4000615 tf2-5 + – – – + – – – HOOD-11 1 4022326 4026545 tf2-6 + – – – + – – – HOOD-12 1 5069824 5083205 SPAC1006.02 meiosis↑ SPAC1006.03c meiosis↑ mcp3 meiosis↑ + – – + + – – + och1 meiosis↑ rgf2 meiosis↑ HOOD-13 1 5191103 5195325 tf2-7 + – – – + – – – HOOD-14 1 5195656 5199909 tf2-8 + – – – + – – – HOOD-15 1 5234000 5250176 SPAC14C4.04 meiosis↑ man1 meiosis↑, memb SPAC14C4.06c SPAC14C4.07 memb mug5 meiosis↑ + – +/– – + – + – agn1 meiosis↑ SPAC14C4.10c meiosis↑, memb SPAC14C4.11 memb HOOD-16 2 91600 101684 SPBPB10D8.04c memb SPBPB10D8.05c memb SPBPB10D8.06c memb + + – – + +/– – – SPBPB10D8.07c memb HOOD-17 2 347505 354250 SPBC1271.09 meiosis↑, memb SPBC1271.10c memb + – – – + – – – (Intergenic region) HOOD-18 2 898107 902233 mcp5 meiosis↑ + + – – + – – – HOOD-19 2 1812684 1816937 tf2-9 + – – – + – – – HOOD-20 2 1965175 1969519 tf2-10 pseudo + – – – + – – – HOOD-21 2 2126590 2128479 SPBC23G7.14 meiosis↑ rpp202 + + – + +** +** +** +** HOOD-22 2 4414469 4418768 tf2-11 + – – – + – – – HOOD-23 2 4442538 4449562 SPBC8E4.02c meiosis↑ SPBC8E4.01c memb + + + – + + + – pho1 meiosis↑ HOOD-24 3 173841 176400 SPCC1235.01 (internal repeats) + – – – + – – – HOOD-25 3 254411 256353 tf2-ORF(truncated) + – – – + – – – HOOD-26 3 778123 782331 tf2-12 + – – – + – – – HOOD-27 3 1047657 1056145 SPCC1259.08 meiosis↑ SPCC1259.09c pgp1 + +/– – +/– +* +* +*/– +*/– gyp2 HOOD-28 3 1168500 1176000 nte1 SPCC4B3.03c meiosis↑, memb + – – – +* +* +*/– +*/– SPNCRNA.120 HOOD-29 3 1179500 1182650 rhp26 + + – – +* +* +*/– +*/– HOOD-30 3 1196050 1196500 tf2-ORF(truncated) + +/– +/– +/– + + +/– +/– HOOD-31 3 1763512 1775613 SPCC1450.16c memb ste6 meiosis↑ SPCC1442.02 meiosis↑ + + – – + + – – SPCC1442.03 SPCC1442.04c meiosis↑ HOOD-32 3 2320230 2324503 tf2-13-pseudo + – – – + – – –

Genomic Location Genomic Small RNAs regions features Chr Start End rrp6∆ rrp6∆ rrp6∆ rrp6∆ rmn1∆ nrl1∆ cwf10-1 1 1511797 1512189 rps2202 + + – +/– 1 5090060 5094840 win1 Small RNA + – – + Clusters 1 5274775 5281509 SPAC2H10.01 meiosis↑ that do not show SPAC2H10.02c meiosis↑ + – – – H3K9me swr1 SPAC11E3.02c meiosis↑ 2 1923486 1925465 erg6 + + – +

Table 3.1. Rmn1, Nrl1 and Cwf10 facilitate small RNA production and H3K9me at genes and retrotransposons. Others contributed: Francisca Reyes-Turcu and Venkata Chalamcharla.

HOOD 3 HOOD 17 HOOD 18 mcp5 myp2 4A8.06c vrs2 1271.09 isp6 ncRNA ncRNA 4A8.07c 10 2.5 2.5

rrp6∆ 10 2.5 2.5 10 2.5 2.5 Small RNA nrl1∆ rrp6∆ 10 2.5 2.5 (normalized reads)

20 10 10

rrp6∆ 20 10 10 20 10 10 Small RNA rrp6∆

cwf10-1 20 10 10

(normalized reads)

Figure 3.10. Normalized number of small RNA reads at HOODs in indicated strains. The signals above and below the line represent small RNAs that map to the top and bottom DNA strands, respectively. Small RNA-Seq data for two independent rrp6D samples and corresponding double mutants are shown. Others contributed: Venkata Chalamcharla.

UUS¨ UUS¨ UUS¨ QUO¨

Figure 3.11. Quantitative analysis of small RNAs at HOODs in indicated samples. Staples indicate 5 percentile (lower) and 95 percentile (upper). The lower and upper bounds of the boxes denote 25 and 75 percentile, respectively. Others contributed: Sameet Mehta.

UUS¨

UUS¨ UUS¨ UUS¨FZI

Figure 3.12. Role of Cwf10 at HOODs loci. (Top) Normalized number of small RNA reads at HOOD 31 and HOOD 10 in rrp6D and rrp6D cwf10-1. (Bottom) ChIP-chip analysis of H3K9me2. Others contributed: Venkata Chalamcharla and Francisca Reyes- Turcu.

We next explored the role of splicing machinery in the formation of HOODs.

Analysis of rrp6∆ and rrp6∆ cwf10-1 strains grown at permissive temperature (26oC) revealed cwf10-1 caused major reductions in siRNAs and H3K9me at a majority of

49 HOODs (Table 3.1 and Figures 3.10 and 3.11), including Tf2 and SPCC1442.04c loci

(Figure 3.12).

A HOOD 12 B psp3 rgf2 tim16 mcp3

20 WT

rrp6∆ 20 cwf10-1 20 Small RNA rrp6∆

cwf10-1 20

(normalized reads) int6 rrp6∆ rrp6∆ cwf10-1 80 WT

cwf10-1 10 H3K9me2 ChIP

Figure 3.13. Effect of cwf10-1 at different loci. (A) The normalized number of small RNA reads at the mcp3 HOOD 12 locus in rrp6∆ and rrp6∆ cwf10-1 are plotted (top). The signals above and below the line represent small RNAs that map to the top and bottom DNA strands, respectively. ChIP-chip analysis of H3K9me2 is also plotted (bottom). (B) Splicing of annotated introns in wild-type and cwf10-1. The arcs above (red) and below (blue) the line represent intron junction reads that map to the top and bottom DNA strands, respectively. The thickness of the arc corresponds to the number of reads. Others contributed: Venkata Chalamcharla and Francisca Reyes-Turcu.

Importantly, higher levels of siRNA and H3K9me could be detected at HOOD-12, which contains mcp3 (Figure 3.13A), where the RNA binding protein Mmi1 directs RNAi

(Yamanaka et al., 2013). Thus, the observed changes were not due to general defects in splicing of genes involved in RNAi. Indeed, whereas splicing of cryptic introns, such as in Tf2 and SPCC1442.04c, was not detected in cwf10-1 mutant (Figure 3.7), splicing of annotated introns was not affected (Figure 3.13B). These results suggest that factors

50 involved in splicing of introns embedded in HOOD loci direct RNAi to assemble heterochromatin domains.

3.4. Noncoding RNAs and Read-through Transcripts Contain Introns

We expanded our analyses of the RNA-Seq data to search for additional introns, and discovered 3218 unannotated cryptic introns. Some introns were detected only in red1∆, mtl1-1, rrp6∆ and/or nrl1∆ mutants, while others were also found in wild-type (Appendix

B).

30 WT red1∆ mtl1-1

Counts 5

0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000 30 Intron length (bp) Intron length (bp) nrl1∆ Annotated Unannotated 10 Splice Site WT WT red1∆ mtl1-1 nrl1∆

Counts 5 GT...AG 99.72% 95.71% 98.35% 94.97% 94.79% GC...AG 0.13% 2.47% 1.01% 2.70% 1.78% AT...AC 0% 1.82% 0.64% 2.33% 3.43% 0 200 400 600 800 1000 Intron length (bp)

Figure 3.14. Analysis of introns in different mutant strains. Graphs show the comparison of the intron length distribution for annotated introns (red) to the length distribution for previously unannotated cryptic introns (black) in wild-type, red1∆, mtl1-1 and nrl1∆ strains. The table shows splice site utilization for annotated introns compared to cryptic introns. Others contributed: Peter FitzGerald.

51 These introns showed a broader length distribution, and a small fraction of them contained variant splice junctions (Figure 3.14). An intron in the centromeric dg repeat, but not the dh repeat, was previously reported (Chinen et al., 2010).

imr1L imr1R IRC1-L dh dg cnt1 dg dh IRC1-R

ago1∆ Intron Intron

30 WT Small RNA 30 (normalized reads) I:3755483-3755536 I:3763794-3763853 I:3778419-3778478 I:3786794-3786847 Depth = 20 Depth = 31 Depth = 31 Depth = 20 TTGTAAT....AATAGAA GAGTAAG....TCTAGCT GAGTAAG....TCTAGCT TTGTAAT....AATAGAA

I:3763794-3763839 I:3778433-3778478 Depth = 23 Depth = 23 ATGTAGT....TCTAGCT ATGTAGT....TCTAGCT

Figure 3.15. Cryptic introns are found in the pericentromeric region. Schematic of cryptic introns detected by RNA-Seq in the pericentromeric region of ago1∆ (top). The arcs above (red) and below (blue) the line represent intron junction reads that map to the top and bottom DNA strands, respectively. The thickness of the arc corresponds to the number of reads. Small RNA reads at the pericentromeric region in wild-type are plotted (middle). The signals above and below the line represent small RNAs that map to the top and bottom DNA strands, respectively. Schematic of RNA-Seq reads detected at the pericentromeric region that contain cryptic introns (bottom). Genomic location of the introns and sequences at intron junctions are shown. 30 and 50 splice sites are indicated with bold blue lettering. The coordinates shown correspond to the February 2013 genome assembly. Others contributed: Martin Zofall.

52 We detected spliced RNA products from both dg and dh repeats in ago1∆ cells, and found that these introns mapped to regions containing siRNA clusters (Figure 3.15).

A SPNCRNA.1080 SPNCRNA.1696 SPNCRNA.1249

WT WT WT

red1∆ red1∆ red1∆

mtl1-1 mtl1-1 mtl1-1

nrl1∆ nrl1∆ nrl1∆ Intron junction reads B meu1-1 crs1 D apr6 dfp1

WT WT WT

red1∆ red1∆ red1∆

mtl1-1 mtl1-1 mtl1-1

rrp6∆ rrp6∆ rrp6∆

Intron junction reads nrl1∆ nrl1∆ nrl1∆ E C SPAPB24d3.01 SPAC23D3.16 rec25

WT WT WT

mtl1-1 red1∆ nrl1∆

nrl1∆ rrp6∆ mtl1-1 Intron junction reads Transcription SPAC2F3.16 500 500 Transcription WT WT WT

red1∆ mtl1-1 ∆ red1

Expression red1∆

(normalized reads) rrp6∆ mtl1-1

Figure 3.16. Examples of cryptic introns in different mutants. (A) Cryptic introns detected by RNA-Seq at ncRNA loci in indicated strains. The arcs above (red) and below (blue) the line represent intron junction reads that map to the top and bottom DNA strands, respectively. The thickness of the arc corresponds to the number of reads. (B Splicing of annotated introns in meiotic genes in indicated strains during vegetative

53 growth. (C) Cryptic introns detected by RNA-Seq at the extended 30 end of loci with convergent genes in indicated strains (top). Normalized RNA-Seq reads are shown for wild-type, mtl1-1 and rrp6∆ (bottom). (D) Example of a cryptic intron detected in the UTR region. (E) Alternative splicing of annotated introns detected by RNA-Seq in nrl1∆ (top) and red1∆ (bottom). Others contributed: Francisca Reyes-Turcu and Martin Zofall.

Interestingly, a substantial portion of the new introns was located within ncRNAs, or in regions that contain no annotated features (Appendix B and Figure 3.16A). In many instances, intron-containing transcripts were upregulated in red1∆, mtl1-1 and/or rrp6∆

(Appendix A). Moreover, meiotic mRNAs that are normally spliced specifically during meiosis were spliced during vegetative growth in mutants (Figure 3.16B).

We also found introns located within UTRs of protein-coding genes. Many transcripts containing introns in their 3’ UTR are produced from convergent genes while in other cases, introns are within unannotated transcripts near the 5’ or 3’ region of genes

(Figures 3.16C and 3.16D). We observed elevated levels of read-through transcripts, normally suppressed by Rrp6 (Zhang et al., 2011; Zofall et al., 2009), at several loci in cells defective in Mtl1 or its associated factors (Figure 3.16C). Cells lacking Nrl1 or

Red1 showed changes in splice site utilization at several sites across the genome (Figure

3.16E and Appendix B), indicating that these factors affect splicing of pre-mRNAs.

3.5. Mtl1 and Ctr1 Promote Telomerase RNA Biogenesis and Telomere

Maintenance

Spliceosome Sm proteins and the Tgs1 RNA methyltransferase were detected in the Ctr1 purification (Figure 3.2), and are required for processing of intron-containing precursor telomerase RNA (TER1) into a mature form (Box et al., 2008; Tang et al.,

54 2012). Core spliceosomal Sm proteins generate the mature 3’ end of TER1, releasing the active form of RNA without exon ligation (Box et al., 2008). Blocking the first step or permitting the completion of splicing generates inactive forms of TER1 (Box et al.,

2008).

A B Exon 1 Intron Exon 2 TER1 Precursor

WT mtl1-1 ctr1∆ rrp6∆ red1∆ Mature TER1 Spliced Precursor Red1 Spliced Sm Mtl1 Ctr1 Mtl1 Rrp6 % Spliced 7 23 1 39 32 + RT act1 - RT

Figure 3.17. Telomerase RNA TER1 is regulated by factors associated with MTREC. (A) (Top) Schematic of TER1 RNA showing the RT- PCR amplification region. (Bottom) RT-PCR analysis of splicing in indicated strains. (B) Proposed roles for different factors in TER1 RNA processing. The Sm and Ctr1-Mtl1 proteins process TER1 to the mature form. Red1-Mtl1 targets spliced TER1 for degradation by the exosome but might also inhibit complete splicing.

We tested if Mtl1-associated factors affect TER1 processing. Spliced product accumulated in mtl1-1, red1∆ and rrp6∆. However, ctr1∆ caused defective splicing of

TER1 (Figure 3.17A). One explanation is that Red1- and Ctr1-containing protein assemblies carry out distinct functions. Mtl1-Red1 and Rrp6 could prevent splicing or promote degradation of spliced product, while Mtl1-Ctr1 could cooperate with the spliceosome to generate mature TER1 (Figure 3.17B).

55 A Exon 1 Intron Exon 2 B TER1

WT mtl1-1 ctr1∆ rrp6∆ red1∆ WTmtl1-1ctr1∆rrp6∆red1∆ Precursor (kb) Telomeres Spliced 1.2

Mature 1.0

Mature

1 0.66 0.33 1.34 1.01 rRNA

Figure 3.18. Mtl1 and Ctr1 regulate TER1 and telomere maintenance. (A) (Top) Schematic of TER1 RNA showing the probe used (thick wavy line) for northern analysis. Dotted lines indicate regions cut by oligo-directed RNase H. (Bottom) Northern blot analysis of TER1. A lighter exposure of mature TER1 is shown below. Relative quantitation of mature TER1 to wild-type is indicated. (B) Southern blot analysis of telomere length.

In this case, Mtl1 would perform dual functions by preventing accumulation of spliced product and generating mature TER1. Supporting this idea, mtl1-1 or ctr1∆ reduced the level of mature TER1, although a more severe phenotype was observed in ctr1∆ compared to the partial loss of function mtl1-1 mutant (Figure 3.18A). No major reduction in mature TER1 levels was observed in red1∆ and rrp6∆, which caused accumulation of spliced product (Figure 3.18A). These results extend previous studies and suggest that Ctr1 and Mtl1, which associate with splicing factors, are required for production of mature TER1.

56 A reduction in mature TER1 is expected to cause shortening of telomeres. As expected, both mtl1-1 and ctr1∆ caused significant reduction in telomere length, with ctr1∆ producing a stronger phenotype, whereas red1∆ did not cause reduction (Figure

3.18B). A B C

WT red1∆ nrl1∆ WT nrl1∆ WT nrl1∆ ctr1∆ (kb) Telomeres Precursor 1.2 Spliced Mature TER1 1.0 + RT act1 rRNA - RT

Figure 3.19. Nrl1 is not involved in the regulation of TER1 and telomere maintenance. (A) RT-PCR analysis of splicing in indicated strains. (B) Northern blot analysis of TER1. rRNA used as a loading control. (C) Southern blot analysis of telomere length.

Loss of Nrl1, which did not co-purify with Sm proteins, had no major effect on processing of TER1 and telomere length (Figure 3.19). Interestingly, telomere shortening was evident in rrp6∆ cells, despite normal levels of mature TER1. The exact cause remains to be investigated, but it is possible that the exosome is required for maturation of TER1 or indirectly interferes with telomere maintenance.

57 Chapter 4. TOR Signaling Pathway Regulates Facultative Heterochromatin

4.1. TOR Signaling Pathway Regulates Heterochromatin Islands

A B

1.0 1.0

0.6 0.6

0.2 0.2 Average H3K9me2 Average H3K9me2

Relative Fold Enrichment WT + N WT - N WT WT Relative Fold Enrichment 2% Glc 0.1% Glc C Island 4 Island 6 Island 9 ssm4 mei4 C8C9.04 2% 0.5% 0.1% Glc 4 18 25

H3K9me2 1 2 5

Figure 4.1. Facultative heterochromatin domains respond to environmental signals. (A) Average H3K9me2 fold enrichment at facultative heterochromatin domains across the genome in ChIP-chip analysis of cells grown in the presence or absence of nitrogen source (+N/-N). (B) Average H3K9me2 fold enrichment at facultative heterochromatin domains across the genome determined from ChIP-chip analysis of cells grown in different amounts of glucose (C) ChIP-chip analysis of H3K9me2 distribution at heterochromatin island loci in indicated conditions. Others contributed: Chanan Rubin and Gobi Thillainadesan.

Facultative heterochromatin islands formed across the S. pombe genome are dynamically regulated by environmental and developmental signals. It has been shown that environmental and developmental signals such as nitrogen starvation abolishes heterochromatin islands (Zofall et al., 2012). In addition to nitrogen starvation, low glucose conditions also abolish MTREC-dependent heterochromatin islands, as revealed by ChIP-chip analysis of H3K9me2 (Figures 4.1A-4.1C).

58 We hypothesized that a signaling pathway plays a critical role in the detection of environmental conditions and the dynamic regulation of facultative heterochromatin. To investigate the role of signaling pathways, we tested candidate signaling pathway factors that are involved in nutrient signaling in S. pombe. Our candidate factors included TOR pathway factors Tor1 and Tor2, which are involved in stress and nitrogen signaling, and

Pka1, which is a cAMP-dependent kinase. Tor2 was a particularly good candidate as a temperature-sensitive mutant was shown to mimic the phenotypes of nitrogen starvation, including upregulating those meiotic genes that are normally repressed by mechanisms that form heterochromatin islands (Matsuo et al., 2007).

2.0 mei4

1.5

1.0

H3K9me2 ChIP 0.5 % Input Recovered

WT pka1∆ tor1∆ tor2-ts6

Figure 4.2. ChIP qPCR analysis of H3K9me2 enrichment at mei4 locus in the indicated strains.

We performed ChIP of H3K9me2 at the heterochromatin island locus mei4 in

WT, pka1, tor1∆, and tor2-ts6 strains. Intriguingly, we found that loss of Tor2 affects the formation of facultative heterochromatin while Pka1 and Tor1 do not (Figure 4.2). Tor2

(TORC1) is an essential protein and regulates cell growth, sexual development, and protein synthesis in response to nutrients. Importantly, the TOR pathway is well conserved from S. pombe to mammals (mTOR) (Alvarez and Moreno, 2006; Matsuo et al., 2007).

A B Island 4 Island 6 Island 9 ssm4 1.0 C8C9.04 mei4 WT tor2-ts6 14 30 0.6 4.5

0.2 Average H3K9me2 Fold Relatve to WT H3K9me2 1 2 5 WT tor2-ts6

Figure 4.3. Tor2 regulates heterochromatin islands. (A) Average H3K9me2 fold enrichment at facultative heterochromatin domains across the genome from ChIP-chip analysis in the indicated strains. (B) ChIP-chip analysis of H3K9me2 distribution at heterochromatin island loci in the indicated strains. Others contributed: Gobi Thillainadesan.

To further investigate the role of Tor2 in the assembly of heterochromatin islands, we performed ChIP-chip for H3K9me2 in tor2-ts6 cells grown at 30˚C, which is a semi- permissive temperature. We observed a significant reduction of H3K9me2 at MTREC- dependent heterochromatin islands in tor2-ts6 cells compared to wild type cells (Figures

4.3A and 4.3B).

4 WT tor2-ts6

2 H3K9me2 ChIP % Input Recovered

leu1 dg

Figure 4.4. ChIP qPCR analysis of H3K9me2 enrichment at the dg locus in the indicated strains.

60 A Island 3 Island 13 Island 14

C23H3.14 spbc3h7.08c C17G9.13c WT pir1∆ 3 3 4

H3K9me2 1 1 0.5

tor2-ts6 tor2-ts6 tor2-ts6 3 3 4

H3K9me2 1 1 0.5

B Island 3 Island 13 Island 14

C23H3.14 spbc3h7.08c C17G9.13c

2% 0.1% Glc 8 2 3.5

H3K9me2 1 2 0.5

Figure 4.5. MTREC-independent heterochromatin islands are not significantly affected by the TOR pathway. (A) ChIP-chip analysis of H3K9me2 distribution at heterochromatin island loci in the indicated strains. (B) ChIP-chip analysis of H3K9me2 distribution at heterochromatin island loci in cells grown with different amounts of glucose. Others contributed: Francisca Reyes-Turcu.

By contrast, constitutive heterochromatin at the centromeric region (dg) and MTREC- independent heterochromatin islands were not abolished in tor2-ts6 cells (Figures 4.4 and

4.5A), suggesting that the role of Tor2 in the regulation of heterochromatin formation is specific for MTREC dependent facultative heterochromatin. Consistently, low glucose conditions also do not abolish MTREC-independent heterochromatin islands (Figure

4.5B), which are formed by a distinct mechanism involving Taz1 protein, a component of the Shelterin complex (Zofall et al., 2016). These results reveal that the essential

61 signaling factor Tor2 is involved in the dynamic regulation of facultative heterochromatin islands in response to environmental signals such as glucose and nitrogen starvation

(Zofall et al., 2012), suggesting that Tor2 may play a key role in connecting environmental signals to chromatin.

4.2. Tor2 Regulates MTREC-dependent Heterochromatin through Pir1

A C pir1 WT tor2-ts6 Pir1-MYC 5 WT Ponceau S Expression tor2-ts6 (normalized reads)

WT tor2-ts6 WT tor2-ts6 WT tor2-ts6 Red1-MYC Mtl1-MYC Swi6

Ponceau S

Figure 4.6. Effect of Tor2 on MTREC-associated factors. (A) Western blot analysis of Pir1-MYC from extracts prepared from the indicated strains. Ponceau S staining is used as a control. (B) Western blot analysis of Red1-MYC, Mtl1-MYC, and Swi6 in the indicated strains. (C) RNA-Seq analysis of the Pir1 gene in the indicated strains.

Since the effect of Tor2 on heterochromatin islands is specific to MTREC- dependent islands, we hypothesized that Tor2 may regulate heterochromatin formation through MTREC components. To test whetherTor2 regulates the stability or modification

62 of any MTREC factors, we analyzed protein levels by Western blotting. Surprisingly, tor2-ts6 cells showed a dramatic loss of Pir1 protein (Figure 4.6A), while the other

MTREC factors, Red1 and Mtl1, and another nuclear factor Swi6 were not significantly affected (Figures 4.6A and 4.6B). The pir1 RNA transcript level is unchanged in tor2-ts6 cells, suggesting that Tor2 likely regulates the stability of the Pir1 protein (Figure 4.6C).

A 1.5 mei4

1.0 B mei4 3 0.5 WT Pir1-MYC ChIP tor2-ts6 % Input Recovered 2

No Tag WT tor2-ts6

C MYC ChIP 1 1.2 mei4 % Input Recovered

0.8 Pir1-MYC Red1-MYC Mtl1-MYC

0.4 Pir1-MYC ChIP % Input Recovered

No Tag WT + N WT - N

Figure 4.7. Tor2 regulates Pir1 enrichment at the heterochromatin island. (A) ChIP qPCR analysis of Pir1-MYC enrichment at the mei4 locus in the indicated strains. An untagged strain is used as a negative control. (B) ChIP qPCR analysis of protein enrichment at the mei4 locus for Pir1-MYC, Red1-MYC, and Mtl1-MYC in the indicated strains. (C) ChIP qPCR analysis of Pir1-MYC enrichment at the mei4 locus in cells grown with or without a nitrogen source (+N/-N).

Consistent with regulation by Tor2, we found that Pir1 localization to heterochromatin islands is lost in tor2-ts6, while Red1 and Mtl1 are still localized at the islands (Figures 4.7A and 4.7B). Nitrogen starvation also caused loss of Pir1, as in tor2- ts6 (Figure 4.7C). These results are consistent with the Western Blot results showing that

Pir1, but not Red1 or Mtl1, is significantly affected by the loss of Tor2 (Figures 4.6A and

63 4.6B). Taken together, Tor2 appears to regulate genes at heterochromatin islands through targeting a specific MTREC-associated factor, Pir1.

A B 1.2 1.2 0.8 0.8

0.4 0.4 Average H3K9me2 Average H3K9me2 Fold Relatve to WT Fold Relatve to WT

WT tor2-ts6 pir1∆ WT tor2-ts6 pir1∆

Figure 4.8. Genome-wide analysis of ChIP-chip for H3K9me2 at facultative heterochromatin domains. (A) Average H3K9me2 enrichment at Pir1 bound regions in the indicated strains. (B) Average H3K9me2 enrichment at regions without Pir1 localization in the indicated strains. Others contributed: Gobi Thillainadesan.

Next, we performed a genome-wide analysis of H3K9me2 ChIP-chip for wild type, tor2-ts6, and pir1∆. We found that only the Pir1 enriched regions showed reduced

H3K9me2 in tor2-ts6 and pir1∆ (Figure 4.8A). Regions that do not show Pir1 localization were not affected (Figure 4.8B). This result suggests that the effect of Tor2 on heterochromatin is specific for Pir1 bound regions, which is further supported by site- specific ChIP results (Figures 4.3B and 4.5A).

64 A

WT tor2-ts6 pir1∆ red1∆ mtl1-1

mei4

+RT act1 - RT

B Island 4 Island 6 Island 9 ssm4 C8C9.04 mei4 60 10 5 WT

pir1∆ Expression

(Normalized reads) tor2-ts6

Figure 4.9. RNA expression analysis at heterochromatin islands in different mutants. (A) RT-PCR analysis of the mei4 locus in the indicated strains. The transcript level of act1 was used as a control. +RT and –RT indicate the presence or absence of reverse transcriptase. (B) RNA-Seq analysis of Pir1 bound genes in the indicated strains.

Pir1 and MTREC factors have been shown to regulate the expression of meiotic genes at heterochromatin islands to different degrees (Egan et al., 2014; Lee et al., 2013;

Yamashita et al., 2013). The specific targeting of Pir1 suggested that the loss of Tor2 would mimic the loss of Pir1 rather than Red1 or Mtl1. Indeed, the degree of derepression of the meiotic gene mei4 in tor2-ts6 is more similar to pir1∆ than red1∆ or mtl1-1, which show considerably severe derepression of MTREC target loci as compared to cells lacking Pir1 (Figure 4.9A). We performed RNA-seq to further investigate the effects of Tor2 and Pir1 on gene expression across the genome, and found that other heterochromatin islands are upregulated in a similar manner in tor2-ts6 and pir1∆ cells

65 (Figure 4.9B). The finding that Tor2 targets Pir1 for regulation of heterochromatin islands is consistent with the fact that Pir1 is the only component of MTREC that exclusively affects heterochromatin assembly at islands (Lee et al., 2013).

66 4.3. Tor2 and Pir1 Regulate Environmentally-sensitive and Disease-associated

Genes

-1 ./&'2*"1&% ./&'2*"1&%#5 ./&'2*"1&%#5SLU¨ ./&'2*"1&%#5 SLU¨

)',"*),' /",0+&+#/",&/",0.-/1

. 4 . 4 &)-1)$2$*&"/)3)0)-, &0)$*&+&%)"1&%/",0.-/1

.4 . 4 )1-$(-,%/)-,/'",)6"1)-, ,$0

. 4 .4

)/( ! ¨ 4./&00)-,

,-/+"*)6&%/&"%0

Figure 4.10. Analysis of RNA-seq results in pir1∆ and tor2-ts6 cells. (A) Pie of pie chart analysis of genes upregulated by tor2-ts6, pir1∆ or both for different S. pombe GO slim terms. P value indicates significance of overlap between genes upregulated in pir1∆ and tor2-ts6. (B) ChIP-chip analysis of Pir1-MYC distribution at the indicated loci (top). RNA-Seq analysis of Pir1 bound genes in WT, pir1∆, and tor2-ts6 strains (bottom). Others contributed: Francisca Reyes-Turcu and Gobi Thillainadesan.

Since Pir1 is involved in gene regulation as a part of the nuclear RNA processing machinery MTREC and is a downstream target of Tor2, we investigated how the loss of

Tor2 or Pir1 affects the regulation of genes across the genome. RNA-Seq analysis of pir1∆ and tor2-ts6 revealed that in addition to derepression of genes at heterochromatin islands (Figure 4.9B), various genes across the genome were also upregulated (Figure

4.11A). Strikingly, a common set of upregulated genes was observed for pir1∆ and tor2- ts6. These genes belong to specific categories that are sensitive to environmental cues, including signaling, transmembrane transporter, meiotic, vesicle-mediated transport, mitochondria organization, and non-coding RNA (Figure 4.11A). Many of these genes showed Pir1 localization (Figure 4.11B).

68 Gene Human Orthologs Pathway invovled Disease associated Pir1

scs22 VAPB/VAPA Metabolism Amyotrophic Lateral Sclerosis +

hem3 HMBS Metabolism Porphyria +

wsp1 WAS/WASL Cytoskeleton Wiskott-Aldrich Syndrome +

npc2 NPC2/LY96 Transporter Niemann-Pick Disease +

SPAPJ691.02 YPEL4/3/1/2 Proliferation Breast Cancer -

vrp1 WIPF1/WIPF2 Signaling Wiskott-Aldrich Syndrome -

cut2 PTTG3P/2/1 Cell Cycle Thyroid Cancer -

scs2 VAPB/VAPA Metabolism Amyotrophic Lateral Sclerosis +

mbx1 MEF2C/B/D/A Signaling Leukemia, Epilepsy -

bhd1 FLCN Signaling Birt-Hogg-Dube Syndrome -

ell1 ELL Transcription Leukemia +

fum1 FH Metabolism Renal Cell Cancer +

sdh3 SDHC Metabolism Gastric Stromal Sarcoma +

zwf2 H6PD Metabolism Hyperandrogenism -

Table 4.1. List of S. pombe genes whose human orthologs are associated with diseases and upregulated in pir1∆ and tor2-ts6 cells. +/- in the Pir1 column indicates the presence or absence of Pir1 at the gene. Others contributed: Gobi Thillainadesan.

Moreover, since Tor2 and Pir1 are involved in regulation of environmentally controlled genes that are conserved in higher eukaryotes, we wondered whether the genes affected by tor2-ts6 and pir1∆ would include those known to be associated with diseases in human. When we examined published works and databases including www.genecards.org and www.pombase.org for S. pombe genes that are upregulated in tor2-ts6 and pir1∆ and also conserved in human, we found that many of these genes are indeed associated with human diseases including various cancers, neurodegenerative diseases and metabolic diseases (Table 4.1) (Rappaport et al., 2013; Wood et al., 2012).

wsp1 fum1

2 2 1.5 sdh3

1 1 1 Pir1 ChIP

10 25 20 WT 10 25 20 pir1∆ 10 25 20 Expression tor2-ts6

(normalized reads)

Figure 4.11. Effect of pir1∆ and tor2-ts6 at various genes associated with diseases. ChIP-chip analysis of Pir1-MYC distribution at the indicated loci (top). RNA-Seq analysis of Pir1 bound genes in WT, pir1∆, and tor2-ts6 strains (bottom). Others contributed: Francisca Reyes-Turcu.

Interestingly, most of these disease-associated genes are involved in signaling and metabolism, which are sensitive to environmental cues, and many were bound by Pir1 in

S. pombe orthologs (Table 4.1 and Figure 4.12). These findings suggest that Tor2 regulates environmentally sensitive and disease-associated genes through its downstream effector Pir1, and that a similar mechanism may be conserved in humans.

70 Chapter 5. Factors Involved in the Regulation of Pir1 by Tor2

5.1. Cul4 is Required for Pir1 Degradation upon Signaling

Input α-MYC IP Pir1-MYC - + - + FLAG-Clr4 + + + + α-MYC α-FLAG

Figure 5.1. Co-IP analysis of the interaction between Pir1 and Clr4. Others contributed: Francisca Reyes-Turcu.

We wondered what other regulatory factors are connected to Tor2 and Pir1. In a previous study, Tor2 has been shown to regulate Mei2, an RNA-binding protein that regulates meiosis, by destabilizing it through the ubiquitin-proteasome pathway(Otsubo et al., 2014). However, unlike the regulation of Mei2, Tor2 appears to regulate Pir1 by stabilizing the protein. Thus, different downstream factors are likely to be involved in the regulation of Pir1 and Mei2.

Since Pir1 is degraded in response to signaling, we speculated that a mechanism that involves factors such as a ubiquitin ligase may be involved. Given the effect of pir1∆ and tor2-ts6 on heterochromatin, one likely candidate is the ubiquitin ligase complex

ClrC, which assembles heterochromatin at islands through its histone H3 lysine-9 methyltransferase component Clr4. Intriguingly, our co-immunoprecipitation (co-IP) experiment result showed that Pir1 interacts with Clr4 (Figure 5.1). The ClrC complex includes the E3 ubiquitin ligase Cullin4 (Cul4) and other subunits including Clr4, Rik1,

Raf1, and Raf2 (Jia et al., 2005; Zhang et al., 2008). Thus we investigated whether Cul4 is involved in regulating the stability of the Pir1 protein. Since Cul4 is involved in both

71 protein degradation and also formation of heterochromatin (Horn et al., 2005; Jia et al.,

2005), this unique combination of properties made it a good candidate for regulating Pir1.

A B

WT tor2-ts6 tor2-ts6 cul4∆ tor2-ts6 WT tor2-ts6 cul4∆ Pir1-MYC

Ponceau S Pir1-MYC

C D +N -N

WT WT cul4∆ WT (+N) WT (-N) cul4∆ (-N) Pir1-MYC

Ponceau S Pir1-MYC

Figure 5.2. Pir1 is regulated by Cul4 in response to signaling. (A) Western blot analysis of Pir1-MYC in the indicated strains. (B) Immunofluorescence analysis of Pir1- MYC in the indicated strains. (C) Western blot analysis of Pir1-MYC in the indicated strains grown with or without a nitrogen source (+N/-N). (D) Immunofluorescence analysis of Pir1-MYC in the indicated strains grown with or without a nitrogen source (+N/-N).

We first tested whether the loss of Cul4 would rescue Pir1 degradation in the tor2-ts6 mutant. Our Western Blot and IF analysis results showed that Pir1 was indeed stabilized in the tor2-ts6 cul4∆ mutant (Figures 5.2A and 5.2B), suggesting that Tor2 is

72 affecting Pir1 through Cul4. Furthermore, when cells were subjected to nitrogen starvation, we found that the loss of Cul4 led to stabilization of Pir1. Thus, Cul4 is important for the degradation of Pir1 in response to the environmental signal and in Tor2 signaling pathway (Figures 5.2C and 5.2D).

+N -N

WT WT cul3∆ Pir1-MYC

Ponceau S

Figure 5.3. Western blot analysis of Pir1-MYC in the indicated strains grown with or without a nitrogen source (+N/-N).

Whereas Cul4 is involved in the regulation of Pir1, another ubiquitin ligase cullin family protein, Cul3, does not seem to be involved (Figure 5.3), indicating that the regulation of

Pir1 is specifically through Cul4.

+N -N

WT WT clr4∆ rik1∆ raf1∆ raf2∆ Pir1-MYC

Ponceau S

Figure 5.4. Western blot analysis of Pir1-MYC in the indicated strains grown with or without a nitrogen source (+N/-N).

73 Since Cul4 is a part of the ClrC complex, we wondered whether other components of ClrC would also have an effect on Pir1 by aiding Cul4 within the complex. Western

Blot analysis showed the loss of other ClrC components, Clr4, Rik1, Raf1, or Raf2 also led to stabilization of Pir1 under nitrogen starvation condition (Figure 5.4).

+N -N

WT WT cul4∆ clr4∆ rik1∆ raf1∆ raf2∆

Pir1-MYC

Ponceau S

Figure 5.5. Western blot analysis of Pir1-MYC in the indicated strains grown with or without a nitrogen source (+N/-N).

We compared the degree of Pir1 stabilization under nitrogen starvation in the cul4∆ mutant to the other ClrC mutants (Figure 5.5). While all components of the ClrC complex contribute to the degradation of Pir1, Cul4 seems to be the major ClrC component for the regulation of Pir1. The other components of ClrC, including Clr4,

Rik1, Raf1, and Raf2, may promote the efficiency or the recruitment of Cul4 for the regulation of Pir1. Since Cul4 also exists in other complexes, it is also possible that other mechanisms are involved.

74 +N -N

WT WT ddb1∆

Pir1-MYC

Ponceau S

Figure 5.6. Western blot analysis of Pir1-MYC in the indicated strains grown with or without nitrogen source (+N/-N).

In addition to ClrC, Cul4 is also a component of a distinct complex containing

Ddb1. Interestingly, the loss of Ddb1 leads to a high level of Pir1 stabilization under nitrogen starvation, similar to the level observed without nitrogen starvation (Figure 5.6) or with the loss of Cul4 (Figure 5.2C and 5.5). This result suggests that Ddb1 may work closely with Cul4 in the complex to regulate Pir1. However, Cul4 can only interact with either Rik1 or Ddb1 in the complex (Braun et al., 2011; Jackson and Xiong, 2009). The fact that Rik1, but not Ddb1, is part of the ClrC complex that includes both Cul4 and Clr4 for heterochromatin formation and silencing (Braun et al., 2011; Jackson and Xiong,

2009), suggests that the Cul4/ClrC population is likely to be the major complex responsible for the regulation of Pir1 at heterochromatin loci, whereas the Cul4 population that interacts with Ddb1 may regulate a pool of Pir1 localized at other parts of the genome and be involved in functions other than heterochromatin assembly.

These results suggest that Cul4 plays a critical role in the regulation of Pir1 protein levels in response to environmental cues such as nitrogen starvation and signaling through the Tor2 pathway. This is a particularly intriguing result given that the targets of

Cul4 that regulate heterochromatin formation have remained poorly understood. Thus far,

75 the only documented regulatory role for Cul4 has been the regulation of the anti-silencing factor Epe1 (Braun et al., 2011).

5.2. Pir1 and Cul4/ClrC Regulate Facultative Heterochromatin through a Feedback

Loop

It was previously shown that Cul4 brings Clr4 to form constitutive heterochromatin (Jia et al., 2005). However, how Cul4 is brought to target loci and the role of Cul4 at facultative heterochromatin still remain unclear. To test whether Cul4 plays a role in bringing Clr4 to facultative heterochromatin, I performed ChIP for

H3K9me2 in cul4∆.

0.5 WT cul4∆

0.3 H3K9me2 ChIP

% Input Recovered 0.1

leu1 mei4

Figure 5.7. ChIP qPCR analysis of H3K9me2 enrichment at the mei4 locus in the indicated strains. The leu1 locus is used as a negative control.

We found that the loss of Cul4 leads to loss of facultative heterochromatin at the heterochromatin island locus, even though the loss of Cul4 stabilizes Pir1 (Figure 5.7).

This observation led us to hypothesize that Cul4 is brought to the heterochromatin islands

76 by targeting Pir1 for degradation, and it brings Clr4 along with it in the complex to form heterochromatin. Cul4 degrades Pir1 as a negative feedback loop that maintains the proper level of Pir1 and heterochromatin as well as the dynamic nature of the facultative heterochromatin domains. This would explain the relatively low enrichment of Pir1 at these loci compared to other MTREC factors, and allow for fine-tuning in response to environmental cues and signaling through Tor2.

0.6 mei4

0.4

0.2 Pir1-MYC ChIP % Input Recovered

No Tag WT cul4∆

Figure 5.8. ChIP qPCR analysis of Pir1-MYC enrichment at the mei4 locus in the indicated strains. An untagged strain is used as a negative control.

To test whether Pir1 is indeed an upstream factor of Cul4/ClrC, we looked for

Pir1 localization at the heterochromatin island in the absence of Cul4/ClrC. ChIP of Pir1-

MYC in cul4∆ cells showed that Pir1 is still present at a significant level (Figure 5.8).

This suggests that while Cul4/ClrC or heterochromatin may help in retaining Pir1 at the heterochromatin loci in a feedback loop, Pir1 is first localized without them and plays a role in recruiting Cul4 and heterochromatin. Because Pir1 is still localized in the absence of ClrC (Figure 5.8), this is also consistent with the result that the loss of heterochromatin assembly factor Clr4 alone does not increase the expression of island genes (Zofall et al.,

2012).

77 A B mei4 mei4 0.8 1.2

0.6 0.8 0.4 0.4 Pir1-MYC ChIP

Pir1-MYC ChIP 0.2 % Input Recovered % Input Recovered

WT tor2-ts6 tor2-ts6 WT cul4∆ cul4∆ cul4∆ -N

Figure 5.9. ChIP qPCR analysis of Pir1-MYC enrichment at the mei4 locus in the indicated strains and conditions.

Furthermore, consistent with the result that the loss of Cul4 restores the Pir1 protein level in tor2-ts6 cells (Figures 5.2A and 5.2B), the localization of Pir1 at the heterochromatin island is also enhanced (Figure 5.9A). In addition, no further loss of

Pir1 localization occurs when cul4∆ is subjected to nitrogen starvation, suggesting that

Cul4 and nitrogen starvation are in the same Pir1 regulatory pathway (Figure 5.9B).

+N -N

WT WT swi6∆

Pir1-MYC

Ponceau S

Figure 5.10. Western blot analysis of Pir1-MYC in the indicated strains grown with or without a nitrogen source (+N/-N).

78 The potential regulation of Pir1 by a feedback loop at facultative heterochromatin loci led us to ask whether the heterochromatin binding protein Swi6/HP1 is also involved in Pir1 regulation. To test this, we looked for degradation of Pir1 under nitrogen starvation in swi6∆ cells. Strikingly, we found that Pir1 was stabilized in swi6∆ cells under nitrogen starvation, suggesting that Swi6 is involved in the regulation of Pir1, potentially by serving as a platformA to bring in other regulatoryB factors (Figure 5.10).

Island 9 LID 6 tgp1 mei4 WT swi6∆ WT swi6∆ 200 3

H3K9me2 1 20

Figure 5.11. Effect of swi6∆ on the heterochromatin island. ChIP-chip analysis of H3K9me2 at a heterochromatin island locus, mei4. Others contributed: Chanan Rubin.

Interestingly, we observed increased heterochromatin at islands in swi6∆, in which Pir1 is stable, compared to wild type (Figure 5.11). This may be due to the role of Swi6 in regulating Pir1 (Figure 5.10) or recruiting Epe1 (Zofall and Grewal, 2006), or both.

These results suggest that facultative heterochromatin is regulated by a feedback loop that utilizes the RNA-processing machinery factor Pir1, the ubiquitin ligase Cul4, and the heterochromatin binding protein Swi6. One possible mechanism is that once Pir1 brings Cul4/ClrC to the target site, Cul4/ClrC forms heterochromatin, which is bound by

Swi6. Cul4 and Swi6 then maintain heterochromatin at the proper level by destabilizing

Pir1, especially in response to signaling such as nitrogen starvation, keeping the sites

79 dynamic and facultative. By such a mechanism, the sites could rapidly lose heterochromatin by degrading Pir1 or gain heterochromatin by stabilizing Pir1.

5.3. Tor2 Regulates Pir1 through Pyp1

We next wondered whether there are any other factors involved in regulation of

Pir1 by Tor2. Since previous studies have suggested a link between the Pyp1/Pyp2 tyrosine phosphatase pathway and the TOR pathway (Petersen and Nurse, 2007), we investigated whether Pyp1 and Pyp2 are involved in regulation of Pir1 by Tor2.

0.4 mei4

0.3

0.2 H3K9me2 ChIP

% Input Recovered 0.1

WT tor2-ts6 tor2-ts6 tor2-ts6 pyp1∆ pyp2∆

Figure 5.12. ChIP qPCR Analysis of H3K9me2 Enrichment at the mei4 Locus in the Indicated Strains.

Interestingly, we found that the loss of Pyp1, but not Pyp2, rescues H3K9me2 in tor2-ts6 cells (Figure 5.12), suggesting that Pyp1 may contribute to Tor2-mediated regulation of

Pir1 to modulate assembly of heterochromatin islands.

80 A B

tor2-ts6 WT tor2-ts6 tor2-ts6 pyp1∆ WT tor2-ts6 pyp1∆ Pir1-MYC

Ponceau S Pir1-MYC

C 0.8 mei4

0.6

0.4 Pir1-MYC ChIP % Input Recovered 0.2

WT tor2-ts6 tor2-ts6 pyp1∆

Figure 5.13. Pyp1 is involved in the regulation of Pir1 by Tor2. (A) Western blot analysis of Pir1-MYC in the indicated strains. (B) Immunofluorescence analysis of Pir1- MYC in the indicated strains. (C) ChIP qPCR analysis of Pir1-MYC enrichment at the mei4 locus in the indicated strains.

Since the loss of Pyp1 restored heterochromatin at islands in tor2-ts6 cells, we next asked whether Pir1 protein levels are also restored. Our results showed that the Pir1 loss observed in tor2-ts6 cells is also rescued in tor2-ts6 pyp1∆ double mutant cells

(Figures 5.13A and 5.13B). Consistently, Pir1 localization that is lost in tor2-ts6 cells is also rescued in tor2-ts6 pyp1∆ double mutant cells (Figure 5.13C). These results suggest that Tor2 regulation of Pir1 for mediating facultative heterochromatin formation involves

Pyp1, however, there may be additional factors involved.

81 Chapter 6. Tor2 and Pir1 Regulate Formation of Latent Heterochromatin Islands

6.1. MTREC-Pir1 is Involved in Formation of Latent Heterochromatin at

Environmentally-sensitive Genes

A c11E10.01 tgp1 mcp2 ggc1 3.5 2.5 6

Pir1-MYC 1 1 1 10 5 16

Red1-MYC 2 1 2 14 7 40

Mtl1-MYC 2 1 5

B Latent Island 6 Latent Island 17 Latent Island 18 c11E10.01 tgp1 mcp2 ggc1 3.5 14 8 WT WT WT

H3K9me2 0.5 2 1 C 3.5 epe1∆ 14 epe1∆ 8 epe1∆

H3K9me2 0.5 2 1

Figure 6.1. ChIP-chip analysis of MTREC factors and H3K9me2. (A) ChIP-chip analysis of Pir1-MYC, Red1-MYC, and Mtl1-MYC distribution at different loci. (B, C) ChIP-chip analysis of H3K9me2 at loci with Pir1-MYC, Red1-MYC, and Mtl1-MYC localization in the indicated strains. Orange lines indicate ncRNAs. Others contributed: Francisca Reyes-Turcu and Chanan Rubin.

We have shown that Pir1 along with MTREC localizes at heterochromatin islands to regulate formation of heterochromatin. In addition to the heterochromatin islands, Pir1 and MTREC also localize at other regions across the genome (Figure 6.1A). However, those loci are not targets of heterochromatin (Figure 6.1B), leading to a question of why

82 Pir1 would localize to those loci. Our previous work has shown that a JmjC domain- containing anti-silencing factor Epe1 negatively affects assembly of heterochromatin islands (Zofall et al., 2012). Thus, we hypothesized that a balance between Pir1 and Epe1 activity in normal growth conditions prevents heterochromatin at these genes.

A prediction of this hypothesis is that loss of Epe1 would disrupt this balance and lead to formation of heterochromatin at Pir1 bound loci. Indeed, our previous work has shown that the loss of Epe1 leads to an appearance of new heterochromatin domains throughout the genome (Zofall et al., 2012). Intriguingly, we discovered that a majority of these new heterochromatin domains coincide with localization of MTREC factors

(Figures 6.1A and 6.1C).

83 Latent&Island Chr Genomic+Features Category 1 1 mfs2 Transmembrane*Transporter

2 1 chs1 Metabolic*/*Biosynthesis SPAC23C4.05c 3 1 Response*to*nutrient*&*stress tht2 4 1 mug8 Response*to*nutrient*&*stress

5 1 pyk1 Metabolic*/*Biosynthesis

6 2 tgp1 Transmembrane*Transporter

7 2 pfl3 Response*to*nutrient*&*stress

8 2 SPBC36.02c Transmembrane*Transporter

9 2 mfs3 Transmembrane*Transporter

10 2 SPBC337.02c Response*to*nutrient*&*stress

11 2 pfk1 Metabolic*/*Biosynthesis exg3 12 2 Response*to*nutrient*&*stress rep1

13 2 mug45 Response*to*nutrient*&*stress

14 2 SPBC1652.01 Response*to*nutrient*&*stress gmh6 15 2 Response*to*nutrient*&*stress SPBC1289.14

16 2 pho1 Metabolic*/*Biosynthesis mcp2 17 3 Transmembrane*Transporter ggc1 18 3 SPCC11E10.01 Metabolic*/*Biosynthesis

19 3 mug1 Response*to*nutrient*&*stress SPCTRNASER.11 SPCTRNAMET.07 SPRRNA.05 SPCC417.09c 20 3 dal51 Transmembrane*Transporter SPCC417.11c SPCC417.12 SPCTRNAASN.06 SPRRNA.06 SPCC70.08c 21 3 Metabolic*/*Biosynthesis mug9 SPCC569.06 22 3 Transmembrane*Transporter SPCC569.05c

Table 6.1. List of latent heterochromatin islands in S. pombe.

84 A B

Latent Island 6 Response to Nutrient & Stress tgp1 27% (6) 0.08 41% (9) Transmembrane Transporter

Metabolic & Biosynthesis Pathway 0.04 32% (7)

H3K9me2 ChIP 0.02 % Input Recovered

2% Glc 8% Glc Latent Heterochromatin Islands

Figure 6.2. Latent heterochromatin islands form at genes sensitive to environmental signals. (A) Pie chart showing different categories of genes for latent heterochromatin islands. (B) ChIP qPCR analysis of wild type H3K9me2 enrichment at the tgp1 locus in cells grown in different amounts of glucose.

We named these MTREC-dependent new heterochromatin domains that appear in epe1∆ (Table 6.1), Latent Heterochromatin Islands (Latent Islands). Interestingly, latent islands target genes that belong to distinct categories including transmembrane transporters, metabolic and biosynthetic pathways, and response to nutrients and stress

(Figure 6.2A). They share the common characteristic that they are sensitive to environmental signals, suggesting that the latent state of heterochromatin at these loci is linked to cells’ response to different environmental cues. For instance, heterochromatin formation could be detected at Latent Island 6 locus under specific conditions, such as high glucose (Figure 6.2B).

85 Latent Island 6 tgp1

WT swi6∆ 3

H3K9me2 1

Figure 6.3. Effect of swi6∆ on the latent heterochromatin island. ChIP-chip analysis of H3K9me2 at a latent heterochromatin island locus, tgp1. Others contributed: Chanan Rubin.

Moreover, disrupting the balance at the latent heterochromatin island loci through the loss of a heterochromatin binding protein, Swi6/HP1, also leads to the formation of heterochromatin (Figure 6.3). This may be due to the role of Swi6 in recruiting Epe1

(Zofall and Grewal, 2006) or regulating Pir1 (Figure 5.10), or both.

Latent Island 6 Latent Island 8 tgp1 c36.02c

3.5 epe1∆ 35 epe1∆

H3K9me2 0.5 1

3.5 WT 3 WT

H3K9me2 0.5 1 WT (H O ) WT (+Methionine) 3.5 2 2 3

H3K9me2 0.5 1

Figure 6.4. ChIP-chip analysis of H3K9me2 distribution at latent heterochromatin islands domain loci in the indicated conditions. Others contributed: Chanan Rubin.

86 We could also detect formation of heterochromatin at latent island loci when cells were grown in different growth conditions such as methionine and H2O2 (Figure 6.4).

We speculate that there would be additional conditions that trigger latent island formation, and that different loci would be sensitive particular conditions or triggers. Our results suggest that Pir1 and MTREC localize at those loci for epigenetic regulation of genes through heterochromatin formation, but Epe1 keeps the region free of heterochromatin under normal growth conditions, making the region latent or active for facultative heterochromatin formation depending on cellular needs.

6.2. Tor2 and Pir1 Regulate Latent Heterochromatin Islands

A Latent Island 6 Latent Island 17 Latent Island 18 c11E10.01 tgp1 mcp2 ggc1 70 50 30 WT

Expression tor2-ts6 (Normalized reads)

B Latent Island 6 Latent Island 17 Latent Island 18 c11E10.01 tgp1 mcp2 ggc1 3.5 epe1∆ 14 epe1∆ 8 epe1∆

H3K9me2 0.5 2 1 3.5 epe1∆ tor2-ts6 14 epe1∆ tor2-ts6 8 epe1∆ tor2-ts6 H3K9me2 0.5 2 1 C 3.5 epe1∆ pir1∆ 14 epe1∆ pir1∆ 8 epe1∆ pir1∆

H3K9me2 0.5 2 1

Figure 6.5. Latent heterochromatin domains are regulated by the Tor2 pathway. (A) RNA-Seq analysis of latent island genes in the indicated strains. (B, C) ChIP-chip analysis of H3K9me2 at latent island loci in the indicated strains. Others contributed: Chanan Rubin.

87 Since latent island genes belong to specific categories that are sensitive to nutritional and environmental signals (Figure 6.2A), we hypothesized that Tor2, which regulates meiotic genes at the canonical heterochromatin islands, would also play a role in regulating these genes. Indeed, RNA-seq analysis showed that latent island genes are upregulated in tor2-ts6 cells (Figure 6.5A). Thus, we also tested whether Tor2 plays a role in regulating heterochromatin formation at these latent island loci. Intriguingly, we found that loss of Tor2 in epe1∆ cells led to loss of latent islands, suggesting that Tor2 is required for formation of latent islands (Figure 6.5B).

We next tested whether Tor2 regulates latent island formation through Pir1, which is enriched at latent island loci (Figure 6.1A). We found that latent islands are indeed abolished with the loss of Pir1 (Figure 6.5C), suggesting that Tor2 and Pir1 play an essential role in formation of latent islands, similar to heterochromatin islands.

A B 4 3 3 2 2 1 1 Fold Relatve to WT Average H3K9me2

WT epe1∆ epe1∆ epe1∆ Average Protein Enrichment tor2-ts6 pir1∆ Pir1 Red1

Figure 6.6. Genome-wide analysis of ChIP-chip data. (A) Average H3K9me2 fold enrichment at facultative heterochromatin domains from ChIP-chip analysis of the indicated strains. (B) Average protein fold enrichment at facultative heterochromatin domains for Pir1-MYC and Red1-MYC tagged strains. Others contributed: Gobi Thillainadesan.

88 Genome-wide analysis of H3K9me2 level at locations with Pir1 localization showed that loss of Pir1 or tor2-ts6 leads to significant global reduction of heterochromatin at latent islands across the genome (Figures 6.6A and 6.6B).

These results suggest that Tor2 and Pir1-MTREC act together to form heterochromatin at genes sensitive to environmental signals, while Epe1 acts as an anti- silencing factor to keep the genes free of heterochromatin, maintaining the balance in normal growth conditions.

6.3. Latent Heterochromatin Islands Formation in Human Cells

LI Gene Human Orthologs Pathway invovled Disease associated

1 mfs2 SPNS1-SPNS3 Signaling Chronic Maxillary Sinusitis

5 pyk1 PKLR/PKM Signaling Pyruvate Kinase Deficiency

Glycogen Storage Disease 11 pfk1 PFKL/PFKM/PFKP Metabolism Diabetes, Whitaker Syndrome

Hypertension, Gastroschisis 15 SPBC1289.14 ADD1/ADD3 Signaling Spastic Tetraplegia

Thyroid Carcinoma 16 pho1 MINPP1 Metabolism Bannayan-Riley-Ruvalcaba Syndrome

Schmid–Fraccaro Syndrome 17 ggc1 SLC25A18/SLC25A22 Transporter Epileptic Encephalopathy

18 SPCC11E10.01 CTH Metabolism Cystathioninuria

Cardiovascular disease, Diabetes 20 dal51 SLC17A1-SLC17A8 Transporter Schizophrenia, Gout, Osteoporosis

Gastrointestinal Stromal Tumor 22 SPCC569.05c SV2A/SV2B/SV2C Transporter Prostate Small Cell Carcinoma

Table 6.2. List of S. pombe genes whose human orthologs are associated with diseases and are targets of latent heterochromatin island domain assembly. The GeneCard database was used to collect information. Latent islands are numbered as LI.

Since latent islands form at genes that are sensitive to environmental cues and are regulated by the TOR pathway, a conserved pathway implicated in various human diseases, we explored human orthologues of S. pombe genes at latent islands. We found that out of 22 latent islands in S. pombe, genes at 9 latent islands are conserved and have human orthologues,. All are involved in specific pathways similar to those of the S. pombe genes, including signaling, metabolism, and transporter (Table 6.2; GeneCard)

(Rappaport et al., 2014). Interestingly, all of the 9 latent islands are associated with diseases including cancers, neurodegenerative diseases, and metabolic diseases (Table

6.2; GeneCard) (Rappaport et al., 2014).

LI Human Ortholog Disease associated Cells with H3K9me3

1 SPNS1 Chronic Maxillary Sinusitis NT2-D1, HepG2

5 PKLR Pyruvate Kinase Deficiency Neuronal Stem Cell, Neuron

11 PFKM Glycogen Storage Disease NT2-D1

15 ADD1 Hypertension, Gastroschisis NT2-D1, Neurosphere

Thyroid Carcinoma NT2-D1, GM12878, HepG2, Neuron 16 MINPP1 Bannayan-Riley-Ruvalcaba Syndrome Neuronal Stem Cell, Neurosphere

17 SLC25A22 Epileptic Encephalopathy Brain tissue

18 CTH Cystathioninuria GM12878, Neurosphere, B Cell

Neuronal Stem Cell, Neuron 20 SLC17A7 Schizophrenia Neurosphere

22 SV2B Prostate Small Cell Carcinoma GM12878

Table 6.3. List of human ortholog genes associated with diseases and are targets of heterochromatin formation in specific cell lines or tissues. The GeneCard database was used to collect information. Latent islands are numbered as LI.

Latent Island 16 Latent Island 17 MINPP1 SLC25A22

30 15 NT2-D1 Brain

Neuronal Stem Cell HepG2 H3K9me3 GM12878 NT2-D1

HMEC HMEC

Latent Island 18 Latent Island 20 CTH SLC17A7

15 30 Neuronal GM12878 Stem Cell

Neurosphere Neuron

H3K9me3 B Cell HepG2

HMEC HMEC

Figure 6.7. ChIP-Seq analysis of H3K9me3 in different human cell lines and tissues for latent island loci (ENCODE).

Based on this observation, we hypothesized that some of these genes would form heterochromatin under certain conditions or in particular cell types, just as in S. pombe.

Thus, we examined the ENCODE database for disease conditions and cell types that form

H3K9me3, and intriguingly, we found that human orthologous genes of all latent islands are targets of heterochromatin in specific cell types and disease conditions (Table 6.3 and

Figure 6.7; ENCODE).

Intriguingly, there seems to be a strong correlation between diseases associated with genes and the cell types in which the facultative heterochromatin is formed. For

91 instance, Latent Islands 17 and 20 are associated with diseases specific to brain such as epilepsy and schizophrenia (Table 6.3; GeneCard) (Epi et al., 2013; Gururajan and van den Buuse, 2014), and they showed formation of heterochromatin only in brain tissue, neuron, neuronal stem cell, and neurosphere, and not in other cell types (Table 6.3 and

Figure 6.7). On the other hand, genes associated with metabolic diseases such as in

Latent Islands 16 and 18 (Table 6.3; GeneCard), showed heterochromatin formation in various cell types including NT2-D1, GM12878, HepG2, Neurosphere, and B cell (Table

6.3. and Figure 6.7).

14 12 DMSO H3K9me3 10 DMSO igG 8 Rapa H3K9me3 6 Rapa igG

H3K9me3 ChIP 4 % Input Recovered 2

CYLC2 MINPP1 SLC17A7

Figure 6.8. ChIP qPCR analysis of H3K9me3 enrichment in neuronal stem cells. Latent islands ortholog loci, MINPP1 and SLC17A7, after treating with DMSO control or Rapamycin (200nM) for 1 day. The CYLC2 locus was used as a negative control. Others contributed: David Wang.

To further investigate whether the heterochromatin formed at the human ortholog of latent island genes is regulated by the mTOR pathway as in S. pombe, we treated human neuronal stem cells with rapamycin, an inhibitor of mTOR, and performed ChIP for H3K9me3. Strikingly, results showed that cells treated with rapamycin have a reduced amount of H3K9me3 at latent island loci that are associated with neuronal

92 diseases such as schizophrenia and Bannayan–Riley–Ruvalcaba syndrome (Figure 6.8), both of which are linked to abnormal mTOR pathway (Crino, 2016; Gururajan and van den Buuse, 2014; Inoki et al., 2005). These results suggest that the Tor2 pathway plays a critical role in mediating formation of latent heterochromatin islands for proper response to environment cues, and that this mechanism may be conserved in human and associated with various diseases.

93 Chapter 7. Discussion: MTREC Forms a Nuclear RNA Processing Network that

Silences Genes and Retrotransposons and Also Maintains Telomeres.

The diverse RNA species produced by eukaryotic genomes provides the basis for biological diversity and differentiation into various cell types (Licatalosi and Darnell,

2010; Sharp, 2009). RNAs are not only messengers of genetic information for protein synthesis but also regulatory hubs that govern expression of genetic information. These

RNA functions are enacted through associated proteins, but how regulatory RNAs are recognized remains largely unknown. We utilized the S. pombe system, which contains conserved RNAi and heterochromatin assembly mechanisms, and discovered a nuclear

RNA processing network that silences genes and retrotransposons, and is required for adaptive regulation of the genome.

MTREC, which is composed of conserved proteins, Mtl1 and Red1, forms multiple protein assemblies involved in regulation of various RNAs including mRNAs, snoRNA and telomerase RNA. We also detected cryptic introns in ncRNA and gene transcripts. MTREC associates with conserved proteins and splicing factors that act through these introns to promote RNAi-dependent assembly of heterochromatin at developmental genes and retrotransposons. Below we discuss potential advantages of ncRNA- and intron-based heterochromatin assembly pathways for reprogramming the genome under different growth conditions and during differentiation.

94 7.1. Mtl1 and Red1 Exist in Multiple Protein Assemblies With Distinct Functions

Red1 was identified as a factor that promotes degradation of meiotic mRNAs via mechanisms that require Pla1, Pab2 and Rrp6 (Sugiyama and Sugioka-Sugiyama, 2011).

It was later reported that Red1, but not Pab2 or Pla1, is involved in the assembly of heterochromatin islands (Hiriart et al., 2012; Tashiro et al., 2013; Zofall et al., 2012). In contrast, Red1 as well as Pla1 and Pab2 are required for RNAi-dependent assembly of

HOODs (Yamashita et al., 2013). The molecular basis of these distinct functions of Red1 was not known.

Figure 7.1. Schematic illustrating the role of MTREC and Mtl1-Nrl1 protein assemblies in the degradation of various RNA species and the formation of heterochromatin at islands or HOODs. Mtl1 in association with Ctr1 also promotes processing of telomerase RNA.

We show that Red1 functions with Mtl1 as part of the MTREC module (Figure

7.1). Similar to Red1, Mtl1 is required for meiotic gene silencing, and regulates stress response genes, genes encoding membrane transporters, and ncRNAs. In addition, both

95 Red1 and Mtl1 are critical for polyadenylation and degradation of pre-mRNAs and snoRNAs, the processing of which also requires Pab2 and Rrp6 (Lemay et al., 2010;

Lemieux et al., 2011; St-Andre et al., 2010). MTREC associated factors coordinate the distinct functions attributed to Red1 (Figure 7.1). For example, Pir1 assembles heterochromatin islands, whereas Rmn1 directs HOOD formation with Pab2 and Pla1

(Figures 1.11A and 1.15) (Yamanaka et al., 2013). Current evidence suggests two modes for MTREC involvement in RNA degradation and heterochromatin formation. First,

MTREC could act as a scaffold that directly engages ribonucleolytic activities, including the exosome that interacts with Red1(Sugiyama and Sugioka-Sugiyama, 2011) or its associated factors. Second, MTREC likely promotes polyadenylation of RNA substrates by Pla1, which in turn recruits Pab2 and its associated proteins to deliver RNA to the exosome and/or trigger RNAi to assemble HOODs. The latter is analogous to functions performed by TRAMP (Houseley et al., 2006), and the putative RNA helicase Mtl1 might unwind RNA substrates to aid their processing.

The Mtl1-Red1 interaction network is reminiscent of the human NEXT complex, which includes hMtr4 among other factors (Lubas et al., 2011). hMtr4 couples the exosome to its various cofactors including ZFC3H1, the Zn-knuckle and Ser/Pro-rich protein that appears to be related to Red1 and Pir1 containing Zn finger and Ser/Pro-rich domains, respectively. However, whether hMtr4 also interacts with factors that trigger

RNAi or guide epigenetic chromatin modifications directed by lncRNAs is unknown.

96 7.2. Noncoding RNAs and Environmental Gene Control

Many conditionally expressed ncRNAs, whose roles are not understood, are derived from regions containing genes and intergenic regions (Dutrow et al., 2008; Rhind et al., 2011; Wilhelm et al., 2008). We demonstrate that ncRNA targeted by the nuclear exosome regulates pho1 expression in response to environmental cues. Deletion of ncRNA decreases H3K9me at pho1 and causes its de-repression in the presence of phosphate. Loss of heterochromatin machinery alone does not de-repress pho1. We find that heterochromatin and the exosome act in parallel to fully repress pho1, as clr4∆ rrp6∆ shows a synergistic increase in pho1 mRNA expression (Figure 2.9C).

How ncRNA facilitates heterochromatin nucleation and RNA processing is an important question. ncRNA could recruit Red1 and Mtl1, which are cofactors for the exosome and participate in RNAi-dependent and –independent assembly of heterochromatin. In this case, ncRNA would provide high affinity binding sites for RNA processing factors that regulate gene expression through mRNA turnover and heterochromatin formation. Environmental changes could modulate RNA processing factors, which regulate levels of ncRNA and/or their entry into pathways that nucleate heterochromatin. Transcription of upstream RNA could potentially also affect gene expression by interfering with binding of activators to the gene promoter, as observed in

S. cerevisiae (Martens et al., 2004).

The regulation of gene expression by ncRNA in response to the environment is a conserved feature in eukaryotes. In S. cerevisiae, which lacks RNAi, ncRNAs modulate gene expression in response to growth conditions by targeting a histone deacetylase

97 (Camblong et al., 2007; Kim et al., 2012; van Werven et al., 2012). Similarly, the antisense RNA COOLAIR regulates flowering time by inducing chromatin modifications in Arabidopsis (Ietswaart et al., 2012). Further analyses may uncover conceptual parallels and highlight the roles of RNA processing factors in directing chromatin changes in response to developmental and environmental signals. We note that an elaborate array of ncRNAs is produced during sexual differentiation in S. pombe, which can impact gene function (Bitton et al., 2011). It is therefore conceivable that targeting of ncRNAs by

MTREC is a fundamental component of gene expression reprogramming.

7.3. Introns, Splicing and Epigenetic Genome Control

Mmi1 binds specific RNAs and directs RNAi to assemble HOODs at several locations (Hiriart et al., 2012; Yamanaka et al., 2013). However, the specificity for assembly of HOODs at other loci has remained unclear. We show that within certain

HOODs, RNAi targets contain cryptic introns. Deletion of such an intron in the

SPCC1442.04c locus abolished siRNA production and H3K9me across the entire heterochromatin domain. Moreover, cwf10-1 severely affects siRNA production and

H3K9me at a majority of HOODs. We also find that Nrl1 associates with splicing factors and facilitates formation of HOODs.

98 RNAiRNAi ExosomeExosome machinery machinery

Mtl1 Nrl1 Red1 Splicing Factors Rmn1 Transcript Pab2 AA..A Intron

Figure 7.2. Schematics illustrating the role of intron and splicing factors in targeting MTREC-associated machinery to HOODs loci. Intron-containing RNA species can be targeted by the Mtl1-Nrl1 complex in association with splicing factors or by MTREC bound to Pab2-Rmn1 and Pla1 (data not shown) to degrade RNAs and assemble HOODs through recruitment of RNAi machinery or promote RNA decay by the exosome.

Together, these analyses significantly extend previous work and reveal that cryptic introns and the spliceosome, which acts co-transcriptionally, play an important role in defining the targets of RNAi-mediated heterochromatin assembly at various loci, including developmental genes and Tf2 retrotransposons (Figure 7.2).

How might the spliceosome and Nrl1 trigger RNAi? Nrl1 could associate with

RNAi machinery, as C. elegans NRDE-2 does (Guang et al., 2010). However, no RNAi factors were identified in our Nrl1 purification. Moreover, cwf10-1 has broad effects on

HOOD assembly, including loci not affected by Nrl1 (Table 3.1). Therefore, it is possible that the spliceosome itself recruits RNAi proteins, and Nrl1 helps in this process by engaging the spliceosome to cryptic introns. Indeed, Nrl1 affects both the splicing efficiency and alternative splicing of various RNA substrates, including RNAi targets

(Figures 3.7, 3.14, and 3.16; Appendix B). Supporting a direct role in triggering RNAi, splicing machinery interacts with components of the RNA-dependent RNA polymerase complex (Bayne et al., 2008), which is involved in production of siRNA and H3K9me at

99 HOODs (Yamanaka et al., 2013). In Cryptococcus neoformans, stalling of the spliceosome is required for synthesis of siRNAs by a spliceosome-associated RNAi complex, although it is not known if this process facilitates formation of heterochromatin

(Dumesic et al., 2013). In addition to engaging RNAi factors, the spliceosome may prevent the release of target transcripts from chromatin and make them available for double strand RNA production through hybridization with opposite strand RNA. Indeed, transcription from the opposite strand occurs at most loci associated with siRNA clusters, including SPCC1442.04c and Tf2 (Figure 3.7) (Yamanaka et al., 2013).

The fact that cryptic introns and the spliceosome affect HOOD assembly is a highly significant finding. Splicing factors have been identified in screens for RNAi components in other systems (Ausin et al., 2012; Tabach et al., 2013; Zhang et al., 2013) and human lncRNAs contain inefficiently spliced introns (Tilgner et al., 2012), suggesting this mechanism is likely conserved in higher eukaryotes. Since splicing can be affected by environmental and developmental signals (Averbeck et al., 2005; Keren et al.,

2010; McPheeters et al., 2009; Pleiss et al., 2007), a regulatory cascade involving cryptic introns might be responsible for the assembly/disassembly of HOODs observed under specific growth conditions (Yamanaka et al., 2013). Introns may serve as “sensors” that can reprogram gene expression by degrading mRNAs and forming heterochromatin.

An intron-based mechanism might also facilitate RNAi-independent degradation of RNAs, including certain meiotic genes and read-through transcripts targeted by the exosome (Harigaya et al., 2006; Zhang et al., 2011; Zofall et al., 2009). This is supported by the observation that RNA elimination factor Red1 also associates with splicing machinery (Figure 7.2), and that various RNAs containing cryptic introns accumulate in

100 cells lacking Red1 and Mtl1, as well as Rrp6 (Tables 2.1 and 3.2 and Figure 3.16C). In these cases, the degradation of RNA by factors such as MTREC and Rrp6 might be functionally coupled to regulation of their splicing. Notable in this regard, Mmi1 is involved in both meiotic gene silencing and regulation of intron splicing (Chen et al.,

2011b; McPheeters et al., 2009). Moreover, splicing of conserved introns in the 3’ UTR of transcripts regulates gene expression in mammals by targeting mRNA for degradation

(McGlincy and Smith, 2008).

Processing of TER1 precursor RNA, which undergoes maturation by splicing- related mechanisms (Box et al., 2008), by Mtl1 protein assemblies uncovers an important aspect of proper telomere maintenance. Whereas MTREC and Rrp6 prevent formation or accumulation of inactive spliced product, Mtl1-Ctr1 and splicing factors generate mature

TER1. The spliceosome may function as part of a complex in which splicing factors join

Ctr1 and Mtl1 to process TER1. The exact roles of Ctr1 and Mtl1 are not clear, but they could be involved in spliceosomal cleavage, 3’-end formation and/or protection of mature

TER1 from degradation by ribonucleases. Ctr1 and Mtl1 are highly conserved, and could be required for telomere maintenance in higher eukaryotes.

101 Chapter 8. Discussion: Regulation of Facultative Heterochromatin by the TOR

Pathway and MTREC RNA Processing Machinery in Response to Signals.

Formation of facultative heterochromatin in eukaryotic cells is an intricate process that is dynamically regulated in response to environmental and developmental cues, and it is associated with development, differentiation, and diseases (Trojer and Reinberg,

2007). Signaling pathways that transmit environmental or developmental signals to bring about changes in the epigenetic chromatin landscape of cells have remained largely unknown.

Nutrient & Stress

Tor2

Pyp1

ClrC Epe1 Clr4 Cul4 MTREC Swi6 Pir1 Red1 Mtl1

Facultative Heterochromatin

Genes sensitive to environmental signals

Figure 8.1. Schematic illustrating a model of how Tor2 signaling regulates facultative heterochromatin formation at genes sensitive to environmental signals. Upon environmental cues such as nutrient and stress, Tor2 regulates Pir1 through the ClrC component Cul4 and Pyp1. While Pir1 and MTREC assemble heterochromatin at these loci, Epe1 opposes the formation of heterochromatin as an anti-silencing factor in normal growth conditions.

102 We utilized S. pombe as a model system, which has conserved signaling pathways, including the TOR pathway, as well as conserved heterochromatin and RNA processing factors to explore regulation of facultative heterochromatin. We uncovered a pathway that includes Tor2, a TORC1 factor, Cul4, an E3 ligase complex component, and Pir1, a component of MTREC nuclear RNA processing network (Figure 8.1). Tor2 is a conserved protein homologous to mammalian TOR protein, mTOR. Tor2 is the core catalytic protein of the TORC1 complex, which is homologous to mTORC1 (Alvarez and

Moreno, 2006; Matsuo et al., 2007). Similar to mTORC1, Tor2 is involved in various essential cellular functions in response to environmental signals (Alvarez and Moreno,

2006; Matsuo et al., 2007; Sengupta et al., 2010; Wullschleger et al., 2006).

Pir1 is a component of the MTREC network containing Mtl1 and Red1 core components (Lee et al., 2013; Zhou et al., 2015). MTREC is conserved with high homology to the human NEXT complex, and it is involved in various nuclear mechanisms including regulation of various nuclear RNA species and formation of facultative heterochromatin (Kilchert et al., 2016; Lee et al., 2013; Zhou et al., 2015).

Pir1 shares high homology with the N-terminal region of ZFC3H1, and the serine/proline rich domain of the protein makes it a nice effector in signaling pathway as we have shown with TOR pathway. We showed that Tor2 regulates Pir1 for dynamic formation of facultative heterochromatin in response to signaling cues, and this regulation involves other factors such as Pyp1, Cul4, and Swi6 (Figure 8.1).

103 8.1. TOR Pathway Regulates Formation of Facultative Heterochromatin through

RNA Processing Machinery

In response to various environmental cues, cellular RNA levels are dynamically controlled at both transcriptional and posttranscriptional levels. RNA-processing machinery, such as MTREC and the exosome, control the steady-state levels of diverse

RNA species along with formation of heterochromatin at many target loci (Doma and

Parker, 2007; Houseley et al., 2006; Lee et al., 2013; Reyes-Turcu and Grewal, 2012;

Schmid and Jensen, 2008). The regulation of RNAs in response to signals also impacts gene control during developmental processes such as meiosis. In S. pombe, meiosis is induced upon nitrogen starvation and accompanied by upregulation of meiosis-specific genes, which are silenced during vegetative growth by an RNA elimination system involving the exosome, and MTREC, which also bring heterochromatin to meiotic genes

(Lee et al., 2013; Yamamoto, 2010; Zofall et al., 2012).

Our results showed that the TOR pathway and Pir1 are involved in the regulation of facultative heterochromatin in response to environmental and developmental signals

(Figure 8.1). For sensitive and dynamic regulation of genes, cells may have developed the mechanism of heterochromatin formation through utilizing a RNA processing machinery such as MTREC (Lee et al., 2013). Importantly, we discovered a link between signaling and nuclear RNA-processing machinery that is required for adaptive regulation of the genome through heterochromatin formation.

104 8.2. Facultative Heterochromatin is Regulated through a Feedback Loop

Feedback Loop ClrC

Clr4 Cul4

Pir1 Swi6 MTREC Facultative Heterochromatin

Genes sensitive to environmental signals

Figure 8.2. Schematic illustrating how the regulation of Pir1 is utilized to form and maintain the proper level of facultative heterochromatin through a feedback loop. Clr4/ClrC is recruited to loci with genes sensitive to environmental signals by Cul4 targeting of Pir1, to form heterochromatin at the loci. As heterochromatin forms, Cul4/ClrC and Swi6/HP that are bound to heterochromatin destabilize Pir1 to maintain the proper amount of H3K9me2 and maintain the dynamic nature of facultative heterochromatin for remodeling in response to environmental cues.

Facultative heterochromatin is dynamically regulated in response to environmental and developmental signals (Narita et al., 2003; Zhu et al., 2013; Zofall et al., 2012). To achieve such dynamic regulation, S. pombe cells utilize an intelligent feedback loop system where the very factors that are used to form heterochromatin are in turn used to counteract the mechanism that brings the heterochromatin. We found that

Pir1 brings the ClrC complex, which includes the ubiquitin ligase Cul4 and the methyltransferase Clr4, to form heterochromatin at target loci. Intriguingly, Cul4 and other components of ClrC, which are recruited by Pir1, in turn degrade Pir1 to keep the facultative heterochromatin at a proper level. This mechanism ensures rapid removal of heterochromatin under environmental and developmental signals such as nitrogen starvation (Figure 8.2).

105 Interestingly, the heterochromatin binding protein Swi6/HP1 is also involved in the feedback loop. We found that Swi6, which binds to the heterochromatin mark

H3K9me2 to form a platform for recruiting various factors, contributes to degradation of

Pir1 (Figure 5.10), in addition to recruiting an anti-silencing factor Epe1 (Zofall and

Grewal, 2006). Thus, formation of larger H3K9me2 domains occur in the absence of

Swi6 (Figure 5.11). These results suggest that once enough heterochromatin is formed at the target loci, Swi6 regulates Pir1 in a feedback loop to prevent Pir1 from forming an excessive amount of heterochromatin. This feedback loop enables dynamic remodeling of heterochromatin in response to developmental and environmental signals (Figure 8.2).

These results suggest another functional role of heterochromatin. When heterochromatin at the island is abolished by loss of ClrC component such as Clr4, RNA silencing is still intact (Zofall et al., 2012). This is because RNA processing factors such as MTREC components that bring exosome still remain at the region for silencing of the transcripts even in the absence of heterochromatin (Figure 5.8). Our results suggest that heterochromatin plays a functional role in fine-tuning the activity of RNA processing machinery in different conditions, as demonstrated by the roles of Cul4 and Swi6 in regulating Pir1 under nitrogen starvation (Figures 5.2C and 5.10).

Cells seem to have developed a check-point that recognizes sufficient heterochromatin and limits further accumulation by regulating a targeting factor.

Considering MTREC, Cul4, Clr4, Swi6, and H3K9me are all conserved in human, it is possible that a similar mechanism exists in human for dynamic regulation of facultative heterochromatin and gene expression in cells to respond to developmental and environmental cues.

106 8.3. Latent Heterochromatin Formation at Disease-associated Genes

Facultative heterochromatin formation is associated with the regulation of developmental programming and cell fate. It is responsive to environmental and developmental cues and contributes to epigenetic silencing of genes in a specific cell type or tissue (Narita et al., 2003; Zhu et al., 2013). The deregulation of epigenetic control on gene expression is frequently linked to various human diseases (Hendrich and Bickmore,

2001; Zardo et al., 2005). Data from the ENCODE Project shows a correlation between diseases associated with genes and the specific cell types that show facultative heterochromatin formation at orthologous S. pombe latent island genes that are regulated by the TOR pathway (Figure 6.7 and Table 6.3; GeneCard) (Rappaport et al., 2014).

A variety of human diseases including cancers, neurological disorders, metabolic diseases, and immune disorders are associated with mTOR (Dazert and Hall, 2011;

Zoncu et al., 2011). Importantly, mTOR is implicated in diseases associated with latent islands such as thyroid cancer, prostate cancer, Bannayan-Riley-Ruvalcaba Syndrome

(BRRS), schizophrenia, and epileptic encephalopathy (Crino, 2016; Epi et al., 2013;

Gururajan and van den Buuse, 2014; Inoki et al., 2005; Morgan et al., 2009; Petrulea et al., 2015).

For instance, a recent study found a de novo mutation in mTOR in patients with epileptic encephalopathies, which is associated with Latent Island 17 (Epi et al., 2013).

Also, the activity of the mTOR cascade was shown to be influenced by several extracellular and environmental factors that have been implicated in schizophrenia, which is associated with Latent Island 20 (Gururajan and van den Buuse, 2014). Studies have

107 shown that an administration of mTOR inhibitors rescues the cellular abnormalities and seizures associated with mutations in mTOR pathway that cause Bannayan-Riley-

Ruvalcaba Syndrome (BRRS) (Crino, 2016; Inoki et al., 2005). Furthermore, there are many studies that link epigenetic regulation such as heterochromatin to various neurological conditions (Jakovcevski and Akbarian, 2012). mTOR signaling is upregulated in 30-50% of prostate cancers (Morgan et al., 2009), which is associated with

Latent Island 16. Moreover, mTOR pathway is a key player in pathogenesis of thyroid cancer, which is associated with Latent Island 22, and activation of mTOR promotes tumor cell proliferation and dedifferentiation (Petrulea et al., 2015).

Since facultative heterochromatin is formed at the conserved orthologous genes in humans and is regulated by a conserved signaling factor, Tor2, as well as the RNA processing machinery, Pir1-MTREC, it is possible that the mechanism of regulating facultative heterochromatin at such genes may be conserved in human as well. The human ortholog of Pir1, ZFC3H1, is expressed in various tissues similarly to mTOR

(GeneCard) (Shmueli et al., 2003). Thus, the roles of mTOR, ZFC3H1, and the NEXT complex in regulating such facultative heterochromatin at disease-associated genes in specific cell types or diseases would be an intriguing target for future studies.

108 Chapter 9. Experimental Procedures

Strains and Media. Standard cell culture methods and media were used. Deletion strains were made using a standard PCR based gene targeting strategy (Bahler et al., 1998). To construct a temperature-sensitive mutation in mtl1 (mtl1-1), error prone PCR was used to amplify a cassette containing a FLAG-tagged mtl1 and a Kanamycin selection marker. The PCR product was integrated into the mtl1 locus of a wild type strain, and transformants from G418 selection plates were tested for temperature sensitivity (ts). Candidate ts colonies were confirmed. Candidates were crossed with wild type, analyzed for marker co-segregation by tetrad analysis, and sequenced for confirmation. To construct a strain with a deletion of intron 2 at SPCC1442.04c, the ura4 gene was inserted and then replaced by a fragment lacking intron 2. The mtl1-1 strain was grown at 26°C on plates, and liquid cultures were grown overnight at 30°C or 33°C to log phase for all experiments. The rrp6∆ cwf10-1 strain was grown overnight at 26°C. For +/- phosphate media, EMM was prepared with or without 15.5 mM sodium phosphate and 20 mM potassium phosphate. tor2-ts6 strain was provided by Masayuki Yamammoto Lab in Japan. tor2- ts6 cells were grown at 26˚C on plates and grown in liquid culture by shifting to 30˚C for experiments. For +/- nitrogen media, EMM was prepared with or without 93.5 mM NH4Cl. For different glucose media, 0.1%, 0.5%, 2%, or 8% of glucose was added to EMM media. For media with methionine, 200 mg/ml of L-Methionine was added to standard EMM media. For cells treated with H2O2, 0.5mM H2O2 was added to YEA media and cells were grown for 2 hours.

109 Purification and Co-Immunoprecipitation. Cells were harvested from overnight 2L cultures

grown to OD595 1.5, washed and flash frozen in liquid nitrogen. Thawed cell pellets were ground with glass beads in a Pulverisette 6 system (Labsynergy) in buffer containing complete protease inhibitors (Roche) and 1mM PMSF. Lysate was cleared by centrifugation at 27,000 × g for 1h, and the supernatant was incubated with anti-FLAG M2 affinity gel (Sigma) or anti-c-MYC agarose affinity gel (Sigma) for 2h. Beads were washed extensively, and precipitated proteins were eluted with 200µl 1mg/ml FLAG or MYC peptides (Sigma). Proteins were precipitated with 10% TCA and resuspended in sample buffer. Samples were separated in 4-12% Bis-Tris

Gel (Invitrogen) for either Western blot analysis or mass spectrometry analysis.

Mass Spectrometry Analysis. The sample preparation and mass spectrometry analysis were performed as described (Zofall et al., 2009). In brief, the gel bands stained with Coommassie

Blue were subjected to in-gel tryptic digestion to extract the peptides. After desalting, each sample was loaded on an Agilent 1200 nano-capillary HPLC system (Agilent Technologies) with a 10 cm integrated µRPLC-electrospray ionization (ESI) emitter columns, coupled online with an

Orbitrap velos mass spectrometer (Thermo Fisher Scientific) for LC-MS analysis. Peptides were eluted using a linear gradient of 2% mobile phase B (acetonitrile with 0.1% formiac acid) to 42% mobile phase B within 40 min at a constant flow rate of 250 nL/min. The fourteen most intense molecular ions in the MS scan were sequentially selected for collision-induced dissociation

(CID) using a normalized collision energy of 35%. The mass spectra were acquired at the mass range of m/z 380-2000, with ion source capillary voltage settting at 1.7 kV, S-Lens RF level set at 69% and ion source temperature setting at 250 °C. The MS/MS data were searched against

UniProt Schizosaccharomyces pombe database from the European Bioinformatics Institute

110 (http://www.ebi.ac.uk/integr8) using BioWorks interfaced SEQUEST (Thermo Fisher

Scientific). Up to two missed cleavage sites was allowed during the database search. The cut-off for legitimate identifications were: charge state dependent cross correlation (Xcorr) ≥ 2.0 for

1+ 2+ 3+ [M+H] , ≥ 2.5 for [M+2H] and ≥ 3.0 for [M+3H] with delta correlation (ΔCn) ≥ 0.10.

Immunofluorescence. Paraformaldehyde fixed cells were incubated with the primary antibodies mouse anti-MYC (9E10, Covance) or mouse anti-FLAG (F7425, Sigma), at 1:1000 and 1:500 dilution, respectively. Cells were subsequently incubated with the appropriate secondary antibody, Alexa Flour 488 anti-mouse IgG (Molecular Probes, Invitrogen) or Alexa Flour 647 anti-rabbit IgG (Molecular Probes, Invitrogen) at 1:2000 dilutions. Immunostained cells were analyzed using a Zeiss Axioplan 2 fluorescence microscope with oil immersion objective lens

(Plan Apochromat, 100×, NA 1.4, Zeiss).

Northern Analysis. Total RNA samples were treated with RNase H in the presence or absence of target-specific oligonucleotides to generate shorter fragments for better resolution and analysis. For analysis of snR99, snR3, Rpl30-2, and TER1 RNAs, 10 g of RNA samples were resolved on a 6% polyacrylamide TBE-Urea gel (Invitrogen). Samples were transferred to

Hybond N+ nylon membrane (GE Healthcare) in 0.5X TBE buffer. For analysis of pho1 mRNA and its upstream ncRNA, RNA samples were separated in a 1% agarose gel. RNA was transferred to BrightStar-Plus membrane (Ambion) using Nothern Max transfer buffer (Ambion).

For both polyacrylamide and agarose gel Northern blots, RNA was UV cross-linked using a

Stratalinker (Stratagene). The 32P-labelled single-stranded RNA probes were generated using the

111 MAXIscript T7 kit (Invitrogen). The membrane was hybridized overnight with the probe in

ULTRAhyp buffer at 65°C (Ambion).

Southern Analysis. Genomic DNA was extracted from cells that were propagated for 90 generations. The DNA was digested with EcoRI for analysis of telomere length. Southern blot analysis was done following standard procedure. The probe used is same as described previously

(Hall et al., 2003).

ChIP and ChIP-chip in S.pombe. ChIP and ChIP-chip experiments were performed as previously described (Cam et al., 2005). Briefly, cells from exponentially growing cultures were fixed with 3% paraformaldehyde, and for epitope-tagged protein ChIP chromatin was cross- linked with 10mM dimethyl adipimidate. Cells were washed in PBS buffer, resuspended in lysis buffer (50 mM HEPES/KOH, pH 7.5, 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1%

DOC), and disrupted by glass bead for lysis. The cell lysate was sonicated to shear chromatin into fragments of 500–1,000 bp, and the resulting homogenate was precleared with Protein A/G beads prior to immunoprecipitation with antibody against dimethylated H3K9 (Abcam), anti-

FLAG M2 affinity gel (Sigma), or anti-MYC antibodies (3µl 9E10, Covance and 3µl A14, Santa

Cruz). Antibodies were recovered with Protein A or Protein G bead slurry. Beads were washed extensively, and cross-linking was reversed by incubation at 65 °C. Immunoprecipitated DNA and DNA from whole cell crude extract (WCE) were analyzed by qPCR or labeled with

Cy5/Cy3 and used for microarray-based ChIP-chip analysis by hybridization to a custom 4X44K oligonucleotide array according to Agilent’s recommended procedure. Custom python scripts were used to call peaks in ChIP-chip data. Briefly, the stretches of probes as arranged on the

112 genome were identified such that a minimum 5 consecutive probes had a mean protein enrichment of 85 percentile or above as compared to all the enrichment values in the ChIP-chip array.

RT-PCR. Total RNA was isolated using the MasterPure™ Yeast RNA Purification Kit

(Epicentre) according to the manufacturer’s instructions. Reverse transcription was performed using the One-step RT-PCR kit (Qiagen) with 100 ng of DNase treated total RNA.

Small RNA and RNA-Seq Library Preparation. RNA was purified for small RNA and RNA-

Seq libraries from 4 OD595 units of log phase cells (0.5) using the MasterPure™ Yeast RNA

Purification Kit (Epicentre) according to the manufacturer’s instructions. For RNA-Seq libraries, ribosomal RNA was first removed using the Ribo-Zero™ rRNA Removal Magnetic Kit

(human/mouse/rat) (Epicentre), and the library was constructed using the ScriptSeq v2 RNA-Seq

Library Preparation Kit (Epicentre). For small RNA library preparation, small RNAs (21-25nt) were excised following denaturing electrophoresis (Urea-PAGE, 17.5%), by using SybrGold nucleic acid stain (Life Technologies) and UV transillumination. Gel-eluted RNAs were recovered with a cellulose acetate Corning® Costar® Spin-X® centrifuge tube filter (Sigma

Aldrich), ethanol precipitated overnight, and resuspended in DEPC treated water. For library construction, 3’ and 5’ adaptor ligation and PCR amplification were performed using the

NEBNext® Small RNA Library Prep Set for Illumina® (NEB) according to the manufacturer’s instructions. Amplified DNA was purified using the Qiagen MinElute PCR Purification Kit

(Qiagen). DNA fragments of 140-150bp were excised from a 6% PAGE gel, eluted overnight, and recovered using a cellulose acetate Corning® Costar® Spin-X® centrifuge tube filter (Sigma

113 Aldrich). Library DNA was ethanol precipitated overnight and resuspended in TE buffer. The final library preparation was analyzed using an Agilent 2100 BioAnalyzer (Agilent).

Small-RNA Sequencing. The libraries were sequenced on the Illumina MiSeq platform. The sequencer-generated files were aligned using the commercial aligner Novoalign. The aligned files were further processed with custom Python scripts to generate files for visualization. The data was first normalized to account for multiply mapping reads, and was further normalized to million mapped reads.

RNA-Seq Analysis. The libraries were sequenced on the Illumina MiSeq Platform. The libraries generated were stranded and were sequenced as single end and paired end. The sequenced files were aligned using the TopHat (Kim et al., 2013) aligning program. The transcripts were assembled using the Cufflinks program and the Ensemble Feb 2013 version of the genome was used for aligning the reads. Annotation for the same version of the genome was used as reference annotation for Cufflinks to assemble the transcripts and calculate abundance of transcripts.

Cuffmerge was used to pool all the assembled transcripts. The junctions.bed files generated during the tophat alignment of the RNA-Seq data were used to obtain the locations and abundances of the introns detected in each sequencing run. Custom Python scripts were used to analyze, collate and tabulate the junction data.

Western Blot Analysis.

For Western analyses, samples were separated on a denaturing 4–12% Bis-Tris Gel

(Invitrogen). Protein from the denaturing gel was transferred onto a nitrocellulose

114 membrane (Bio-Rad) and probed with anti–c-MYC 9E10 (MMS-150R; Covance), anti–c-

MYC A-14 (Sc-789; Santa Cruz) primary antibody. For ECL, anti-rabbit or anti-mouse

HRP-conjugated secondary antibody was used as a secondary antibody.

ChIP-qPCR. Real-time RT-qPCR reactions were performed using the QuantiTect SYBR

Green PCR Kit (Qiagen) on ChIP DNA or input DNA using gene specific oligonucleotides. ChIP-qPCR results were represented as percentage (%) of IP/input signal (% input). ChIP-qPCR values represent at least two independent experiments, and each was evaluated in duplicate or triplicate by qPCR. Error bars indicate the SD from the duplicate or triplicate experiments.

Human Cell Culture. Culture media and components were purchased from Invitrogen

(Carlsbad, CA) and growth factors and cytokines were purchased from PeproTech

(Rocky Hill, NJ) and chemicals were purchased from Sigma (St. Louis MO) if not specified. ES cell line WA09-FI was obtained from WiCell (Madison, WI) and cultured in E8 medium on matrigel coated plates in a 5% CO2 incubator and subcultured using

EDTA method following published protocols (Beers et al., 2012; Chen et al., 2011a).

Neural stem cells were induced using PSC neural induction medium (Invitrogen) following the manufacturer’s instruction. Briefly, WA09-FI cells were replated on a

Matrigel-coated 6 well plate in E8 medium for 24 hours. The medium was then replaced with neural induction medium. After 7 days of induction, the resulting neural stem cells were collected after dissociation with accutase and replated on poly-D-lysine (PDL)- coated 100mm dishes in StemPro neural stem cell serum free medium for expansion. The

115 cells were ready for further subculture or treatment when reach 60% confluence. The neural stem cells were characterized by nestin immunostaining and could be further differentiated to neurons when incubated with neuronal differentiation medium

(DMEM/F12 containing 1xN2 supplement, 1xB27 supplement, 300 ng/ml cAMP and 0.2 mM vitamin C, 10 ng/ml brain derived neurotrophic factor (BDNF) and 10 ng/ml glial- derived neurotrophic factor (GDNF) for 14 days. Blood from adult healthy donors was collected at Transfusion Medicine Blood Bank of the National Institutes of Health and signed informed consents were obtained in accordance with the NIH Institutional Review

Board. Induced neural stem cells (iNSC) were then derived from purified human adult hematopoietic progenitor cells (CD34 cells) and further characterized as shown in the previous publications (Wang et al., 2013; Wang et al., 2015). The iNSC were cultured on

PDL-coated 100mm dishes in StemPro neural stem cell serum free medium for expansion and were ready for treatment when reach 60% confluence. Rapamycin treatment was done by adding 200nM of Rapamycin (EMD Millipore) to the culture while adding the same amount of DMSO to a control sample culture. Fresh media was added every 24 hours until harvest.

ChIP in Human Cell Line. Cells from cultures were fixed with 3% paraformaldehyde and washed in PBS buffer. Cells are resuspended in lysis buffer (50 mM HEPES/KOH, pH 7.5, 140 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% DOC, protease inhibitor) for lysis. The cell lysate was sonicated to shear chromatin into fragments of 500–1,000 bp by BioRuptor (Diagenode), and the resulting homogenate was precleared with Protein A

Sepharose beads (Life Technologies) prior to immunoprecipitation with antibody against

116 trimethylated H3K9 (Abcam). Antibodies were recovered with Protein A Sepharose bead slurry. Beads were washed extensively, and cross-linking was reversed by incubation at

65 °C. Immunoprecipitated DNA and input DNA were analyzed by qPCR.

117 REFERENCES

1. Appendices

Appendix A: Genes and ncRNA upregulated in red1Δ, mtl1-1, and rrp6Δ mutants. http://www.cell.com/cms/attachment/2010703077/2032746546/mmc2.xlsx

Appendix B: Previously unannotated introns identified in wild-type and different mutants. https://ccrod.cancer.gov/confluence/display/CLGP/CCR+LBMB+Grewal+Pub+Home

118 2. Bibliography

Alvarez, B., and Moreno, S. (2006). Fission yeast Tor2 promotes cell growth and represses cell differentiation. J Cell Sci 119, 4475-4485.

Ausin, I., Greenberg, M.V., Li, C.F., and Jacobsen, S.E. (2012). The splicing factor SR45 affects the RNA-directed DNA methylation pathway in Arabidopsis. Epigenetics : official journal of the DNA Methylation Society 7, 29-33.

Averbeck, N., Sunder, S., Sample, N., Wise, J.A., and Leatherwood, J. (2005). Negative control contributes to an extensive program of meiotic splicing in fission yeast.

Molecular cell 18, 491-498.

Bahler, J., Wu, J.Q., Longtine, M.S., Shah, N.G., McKenzie, A., 3rd, Steever, A.B.,

Wach, A., Philippsen, P., and Pringle, J.R. (1998). Heterologous modules for efficient and versatile PCR-based gene targeting in Schizosaccharomyces pombe. Yeast 14, 943-

951.

Batista, P.J., and Chang, H.Y. (2013). Long noncoding RNAs: cellular address codes in development and disease. Cell 152, 1298-1307.

Bayne, E.H., Portoso, M., Kagansky, A., Kos-Braun, I.C., Urano, T., Ekwall, K., Alves,

F., Rappsilber, J., and Allshire, R.C. (2008). Splicing factors facilitate RNAi-directed silencing in fission yeast. Science 322, 602-606.

Beers, J., Gulbranson, D.R., George, N., Siniscalchi, L.I., Jones, J., Thomson, J.A., and

Chen, G. (2012). Passaging and colony expansion of human pluripotent stem cells by enzyme-free dissociation in chemically defined culture conditions. Nat Protoc 7, 2029-

2040.

119 Beisel, C., and Paro, R. (2011). Silencing chromatin: comparing modes and mechanisms.

Nature reviews Genetics 12, 123-135.

Bitton, D.A., Grallert, A., Scutt, P.J., Yates, T., Li, Y., Bradford, J.R., Hey, Y., Pepper,

S.D., Hagan, I.M., and Miller, C.J. (2011). Programmed fluctuations in sense/antisense transcript ratios drive sexual differentiation in S. pombe. Mol Syst Biol 7, 559.

Box, J.A., Bunch, J.T., Tang, W., and Baumann, P. (2008). Spliceosomal cleavage generates the 3' end of telomerase RNA. Nature 456, 910-914.

Braun, S., Garcia, J.F., Rowley, M., Rougemaille, M., Shankar, S., and Madhani, H.D.

(2011). The Cul4-Ddb1(Cdt)(2) ubiquitin ligase inhibits invasion of a boundary- associated antisilencing factor into heterochromatin. Cell 144, 41-54.

Buhler, M., Haas, W., Gygi, S.P., and Moazed, D. (2007). RNAi-dependent and - independent RNA turnover mechanisms contribute to heterochromatic gene silencing.

Cell 129, 707-721.

Cam, H.P., Noma, K., Ebina, H., Levin, H.L., and Grewal, S.I. (2008). Host genome surveillance for retrotransposons by transposon-derived proteins. Nature 451, 431-436.

Cam, H.P., Sugiyama, T., Chen, E.S., Chen, X., FitzGerald, P.C., and Grewal, S.I.

(2005). Comprehensive analysis of heterochromatin- and RNAi-mediated epigenetic control of the fission yeast genome. Nature genetics 37, 809-819.

Camblong, J., Iglesias, N., Fickentscher, C., Dieppois, G., and Stutz, F. (2007). Antisense

RNA stabilization induces transcriptional gene silencing via histone deacetylation in S. cerevisiae. Cell 131, 706-717.

120 Chen, D., Toone, W.M., Mata, J., Lyne, R., Burns, G., Kivinen, K., Brazma, A., Jones,

N., and Bahler, J. (2003). Global transcriptional responses of fission yeast to environmental stress. Molecular biology of the cell 14, 214-229.

Chen, G., Gulbranson, D.R., Hou, Z., Bolin, J.M., Ruotti, V., Probasco, M.D., Smuga-

Otto, K., Howden, S.E., Diol, N.R., Propson, N.E., et al. (2011a). Chemically defined conditions for human iPSC derivation and culture. Nat Methods 8, 424-429.

Chen, H.M., Futcher, B., and Leatherwood, J. (2011b). The fission yeast RNA binding protein Mmi1 regulates meiotic genes by controlling intron specific splicing and polyadenylation coupled RNA turnover. PloS one 6, e26804.

Chinen, M., Morita, M., Fukumura, K., and Tani, T. (2010). Involvement of the spliceosomal U4 small nuclear RNA in heterochromatic gene silencing at fission yeast centromeres. The Journal of biological chemistry 285, 5630-5638.

Cho, Y.S., Jang, S., and Yoon, J.H. (2012). Isolation of a novel rmn1 gene genetically linked to spnab2 with respect to mRNA export in fission yeast. Molecules and cells 34,

315-321.

Cooper, T.A., Wan, L., and Dreyfuss, G. (2009). RNA and disease. Cell 136, 777-793.

Crino, P.B. (2016). The mTOR signalling cascade: paving new roads to cure neurological disease. Nat Rev Neurol 12, 379-392.

Dashwood, R.H., and Ho, E. (2007). Dietary histone deacetylase inhibitors: from cells to mice to man. Semin Cancer Biol 17, 363-369.

Dazert, E., and Hall, M.N. (2011). mTOR signaling in disease. Current opinion in cell biology 23, 744-755.

121 Ding, D.Q., Okamasa, K., Yamane, M., Tsutsumi, C., Haraguchi, T., Yamamoto, M., and

Hiraoka, Y. (2012). Meiosis-specific noncoding RNA mediates robust pairing of homologous chromosomes in meiosis. Science 336, 732-736.

Doma, M.K., and Parker, R. (2007). RNA quality control in eukaryotes. Cell 131, 660-

668.

Dumesic, P.A., Natarajan, P., Chen, C., Drinnenberg, I.A., Schiller, B.J., Thompson, J.,

Moresco, J.J., Yates, J.R., 3rd, Bartel, D.P., and Madhani, H.D. (2013). Stalled spliceosomes are a signal for RNAi-mediated genome defense. Cell 152, 957-968.

Dutrow, N., Nix, D.A., Holt, D., Milash, B., Dalley, B., Westbroek, E., Parnell, T.J., and

Cairns, B.R. (2008). Dynamic transcriptome of Schizosaccharomyces pombe shown by

RNA-DNA hybrid mapping. Nature genetics 40, 977-986.

Egan, E.D., Braun, C.R., Gygi, S.P., and Moazed, D. (2014). Post-transcriptional regulation of meiotic genes by a nuclear RNA silencing complex. Rna 20, 867-881.

Epi, K.C., Epilepsy Phenome/Genome, P., Allen, A.S., Berkovic, S.F., Cossette, P.,

Delanty, N., Dlugos, D., Eichler, E.E., Epstein, M.P., Glauser, T., et al. (2013). De novo mutations in epileptic encephalopathies. Nature 501, 217-221.

Feil, R., and Fraga, M.F. (2011). Epigenetics and the environment: emerging patterns and implications. Nature reviews Genetics 13, 97-109.

Feng, S., and Jacobsen, S.E. (2011). Epigenetic modifications in plants: an evolutionary perspective. Curr Opin Plant Biol 14, 179-186.

Fischer, T., Cui, B., Dhakshnamoorthy, J., Zhou, M., Rubin, C., Zofall, M., Veenstra,

T.D., and Grewal, S.I. (2009). Diverse roles of HP1 proteins in heterochromatin assembly

122 and functions in fission yeast. Proceedings of the National Academy of Sciences of the

United States of America 106, 8998-9003.

Gingeras, T.R. (2007). Origin of phenotypes: genes and transcripts. Genome research 17,

682-690.

Grewal, S.I. (2010). RNAi-dependent formation of heterochromatin and its diverse functions. Current opinion in genetics & development 20, 134-141.

Grewal, S.I., and Elgin, S.C. (2007). Transcription and RNA interference in the formation of heterochromatin. Nature 447, 399-406.

Guang, S., Bochner, A.F., Burkhart, K.B., Burton, N., Pavelec, D.M., and Kennedy, S.

(2010). Small regulatory RNAs inhibit RNA polymerase II during the elongation phase of transcription. Nature 465, 1097-1101.

Gururajan, A., and van den Buuse, M. (2014). Is the mTOR-signalling cascade disrupted in Schizophrenia? J Neurochem 129, 377-387.

Hall, I.M., Noma, K., and Grewal, S.I. (2003). RNA interference machinery regulates chromosome dynamics during mitosis and meiosis in fission yeast. Proceedings of the

National Academy of Sciences of the United States of America 100, 193-198.

Hall, I.M., Shankaranarayana, G.D., Noma, K., Ayoub, N., Cohen, A., and Grewal, S.I.

(2002). Establishment and maintenance of a heterochromatin domain. Science 297, 2232-

2237.

Harigaya, Y., Tanaka, H., Yamanaka, S., Tanaka, K., Watanabe, Y., Tsutsumi, C.,

Chikashige, Y., Hiraoka, Y., Yamashita, A., and Yamamoto, M. (2006). Selective elimination of messenger RNA prevents an incidence of untimely meiosis. Nature 442,

45-50.

123 Hendrich, B., and Bickmore, W. (2001). Human diseases with underlying defects in chromatin structure and modification. Hum Mol Genet 10, 2233-2242.

Hiriart, E., Vavasseur, A., Touat-Todeschini, L., Yamashita, A., Gilquin, B., Lambert, E.,

Perot, J., Shichino, Y., Nazaret, N., Boyault, C., et al. (2012). Mmi1 RNA surveillance machinery directs RNAi complex RITS to specific meiotic genes in fission yeast. The

EMBO journal 31, 2296-2308.

Horn, P.J., Bastie, J.N., and Peterson, C.L. (2005). A Rik1-associated, cullin-dependent

E3 ubiquitin ligase is essential for heterochromatin formation. Genes & development 19,

1705-1714.

Houseley, J., LaCava, J., and Tollervey, D. (2006). RNA-quality control by the exosome.

Nat Rev Mol Cell Biol 7, 529-539.

Ietswaart, R., Wu, Z., and Dean, C. (2012). Flowering time control: another window to the connection between antisense RNA and chromatin. Trends Genet 28, 445-453.

Inoki, K., Corradetti, M.N., and Guan, K.L. (2005). Dysregulation of the TSC-mTOR pathway in human disease. Nature genetics 37, 19-24.

Ito, H., Gaubert, H., Bucher, E., Mirouze, M., Vaillant, I., and Paszkowski, J. (2011). An siRNA pathway prevents transgenerational retrotransposition in plants subjected to stress.

Nature 472, 115-119.

Jackson, S., and Xiong, Y. (2009). CRL4s: the CUL4-RING E3 ubiquitin ligases. Trends

Biochem Sci 34, 562-570.

Jakovcevski, M., and Akbarian, S. (2012). Epigenetic mechanisms in neurological disease. Nat Med 18, 1194-1204.

124 Jia, S., Kobayashi, R., and Grewal, S.I. (2005). Ubiquitin ligase component Cul4 associates with Clr4 histone methyltransferase to assemble heterochromatin. Nat Cell

Biol 7, 1007-1013.

Jia, S., Noma, K., and Grewal, S.I. (2004). RNAi-independent heterochromatin nucleation by the stress-activated ATF/CREB family proteins. Science 304, 1971-1976.

Jurica, M.S., and Moore, M.J. (2003). Pre-mRNA splicing: awash in a sea of proteins.

Molecular cell 12, 5-14.

Keren, H., Lev-Maor, G., and Ast, G. (2010). Alternative splicing and evolution: diversification, exon definition and function. Nature reviews Genetics 11, 345-355.

Kiely, C.M., Marguerat, S., Garcia, J.F., Madhani, H.D., Bahler, J., and Winston, F.

(2011). Spt6 is required for heterochromatic silencing in the fission yeast

Schizosaccharomyces pombe. Mol Cell Biol 31, 4193-4204.

Kilchert, C., Wittmann, S., and Vasiljeva, L. (2016). The regulation and functions of the nuclear RNA exosome complex. Nature reviews Molecular cell biology 17, 227-239.

Kim, D., Pertea, G., Trapnell, C., Pimentel, H., Kelley, R., and Salzberg, S.L. (2013).

TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome biology 14, R36.

Kim, D.U., Hayles, J., Kim, D., Wood, V., Park, H.O., Won, M., Yoo, H.S., Duhig, T.,

Nam, M., Palmer, G., et al. (2010). Analysis of a genome-wide set of gene deletions in the fission yeast Schizosaccharomyces pombe. Nature biotechnology 28, 617-623.

Kim, T., Xu, Z., Clauder-Munster, S., Steinmetz, L.M., and Buratowski, S. (2012). Set3

HDAC mediates effects of overlapping noncoding transcription on gene induction kinetics. Cell 150, 1158-1169.

125 Lawrence, C.L., Maekawa, H., Worthington, J.L., Reiter, W., Wilkinson, C.R., and Jones,

N. (2007). Regulation of Schizosaccharomyces pombe Atf1 protein levels by Sty1- mediated phosphorylation and heterodimerization with Pcr1. The Journal of biological chemistry 282, 5160-5170.

Lee, J.T. (2012). Epigenetic regulation by long noncoding RNAs. Science 338, 1435-

1439.

Lee, N.N., Chalamcharla, V.R., Reyes-Turcu, F., Mehta, S., Zofall, M., Balachandran,

V., Dhakshnamoorthy, J., Taneja, N., Yamanaka, S., Zhou, M., et al. (2013). Mtr4-like protein coordinates nuclear RNA processing for heterochromatin assembly and for telomere maintenance. Cell 155, 1061-1074.

Lemay, J.F., D'Amours, A., Lemieux, C., Lackner, D.H., St-Sauveur, V.G., Bahler, J., and Bachand, F. (2010). The nuclear poly(A)-binding protein interacts with the exosome to promote synthesis of noncoding small nucleolar RNAs. Molecular cell 37, 34-45.

Lemieux, C., Marguerat, S., Lafontaine, J., Barbezier, N., Bahler, J., and Bachand, F.

(2011). A Pre-mRNA degradation pathway that selectively targets intron-containing genes requires the nuclear poly(A)-binding protein. Molecular cell 44, 108-119.

Licatalosi, D.D., and Darnell, R.B. (2010). RNA processing and its regulation: global insights into biological networks. Nature reviews Genetics 11, 75-87.

Lubas, M., Christensen, M.S., Kristiansen, M.S., Domanski, M., Falkenby, L.G., Lykke-

Andersen, S., Andersen, J.S., Dziembowski, A., and Jensen, T.H. (2011). Interaction profiling identifies the human nuclear exosome targeting complex. Molecular cell 43,

624-637.

126 Martens, J.A., Laprade, L., and Winston, F. (2004). Intergenic transcription is required to repress the Saccharomyces cerevisiae SER3 gene. Nature 429, 571-574.

Mata, J., Wilbrey, A., and Bahler, J. (2007). Transcriptional regulatory network for sexual differentiation in fission yeast. Genome biology 8, R217.

Matsuo, T., Otsubo, Y., Urano, J., Tamanoi, F., and Yamamoto, M. (2007). Loss of the

TOR kinase Tor2 mimics nitrogen starvation and activates the sexual development pathway in fission yeast. Mol Cell Biol 27, 3154-3164.

Matzke, M.A., and Birchler, J.A. (2005). RNAi-mediated pathways in the nucleus.

Nature reviews Genetics 6, 24-35.

McGlincy, N.J., and Smith, C.W. (2008). Alternative splicing resulting in nonsense- mediated mRNA decay: what is the meaning of nonsense? Trends Biochem Sci 33, 385-

393.

McPheeters, D.S., Cremona, N., Sunder, S., Chen, H.M., Averbeck, N., Leatherwood, J., and Wise, J.A. (2009). A complex gene regulatory mechanism that operates at the nexus of multiple RNA processing decisions. Nature structural & molecular biology 16, 255-

264.

Morgan, T.M., Koreckij, T.D., and Corey, E. (2009). Targeted therapy for advanced prostate cancer: inhibition of the PI3K/Akt/mTOR pathway. Curr Cancer Drug Targets 9,

237-249.

Motamedi, M.R., Verdel, A., Colmenares, S.U., Gerber, S.A., Gygi, S.P., and Moazed, D.

(2004). Two RNAi complexes, RITS and RDRC, physically interact and localize to noncoding centromeric RNAs. Cell 119, 789-802.

127 Murayama, A., Ohmori, K., Fujimura, A., Minami, H., Yasuzawa-Tanaka, K., Kuroda,

T., Oie, S., Daitoku, H., Okuwaki, M., Nagata, K., et al. (2008). Epigenetic control of rDNA loci in response to intracellular energy status. Cell 133, 627-639.

Narita, M., Nunez, S., Heard, E., Narita, M., Lin, A.W., Hearn, S.A., Spector, D.L.,

Hannon, G.J., and Lowe, S.W. (2003). Rb-mediated heterochromatin formation and silencing of E2F target genes during cellular senescence. Cell 113, 703-716.

Nicolas, E., Yamada, T., Cam, H.P., Fitzgerald, P.C., Kobayashi, R., and Grewal, S.I.

(2007). Distinct roles of HDAC complexes in promoter silencing, antisense suppression and DNA damage protection. Nature structural & molecular biology 14, 372-380.

Orom, U.A., and Shiekhattar, R. (2011). Noncoding RNAs and enhancers: complications of a long-distance relationship. Trends Genet 27, 433-439.

Otsubo, Y., Yamashita, A., Ohno, H., and Yamamoto, M. (2014). S. pombe TORC1 activates the ubiquitin-proteasomal degradation of the meiotic regulator Mei2 in cooperation with Pat1 kinase. J Cell Sci 127, 2639-2646.

Peng, J.C., and Karpen, G.H. (2008). Epigenetic regulation of heterochromatic DNA stability. Current opinion in genetics & development 18, 204-211.

Perez, P., and Cansado, J. (2010). Cell integrity signaling and response to stress in fission yeast. Curr Protein Pept Sci 11, 680-692.

Petersen, J., and Nurse, P. (2007). TOR signalling regulates mitotic commitment through the stress MAP kinase pathway and the Polo and Cdc2 kinases. Nat Cell Biol 9, 1263-

1272.

128 Petrulea, M.S., Plantinga, T.S., Smit, J.W., Georgescu, C.E., and Netea-Maier, R.T.

(2015). PI3K/Akt/mTOR: A promising therapeutic target for non-medullary thyroid carcinoma. Cancer Treat Rev 41, 707-713.

Pleiss, J.A., Whitworth, G.B., Bergkessel, M., and Guthrie, C. (2007). Rapid, transcript- specific changes in splicing in response to environmental stress. Molecular cell 27, 928-

937.

Rappaport, N., Nativ, N., Stelzer, G., Twik, M., Guan-Golan, Y., Stein, T.I., Bahir, I.,

Belinky, F., Morrey, C.P., Safran, M., et al. (2013). MalaCards: an integrated compendium for diseases and their annotation. Database (Oxford) 2013, bat018.

Rappaport, N., Twik, M., Nativ, N., Stelzer, G., Bahir, I., Stein, T.I., Safran, M., and

Lancet, D. (2014). MalaCards: A Comprehensive Automatically-Mined Database of

Human Diseases. Curr Protoc Bioinformatics 47, 1 24 21-19.

Reyes-Turcu, F.E., and Grewal, S.I. (2012). Different means, same end-heterochromatin formation by RNAi and RNAi-independent RNA processing factors in fission yeast.

Current opinion in genetics & development 22, 156-163.

Reyes-Turcu, F.E., Zhang, K., Zofall, M., Chen, E., and Grewal, S.I. (2011). Defects in

RNA quality control factors reveal RNAi-independent nucleation of heterochromatin.

Nature structural & molecular biology 18, 1132-1138.

Rhind, N., Chen, Z., Yassour, M., Thompson, D.A., Haas, B.J., Habib, N., Wapinski, I.,

Roy, S., Lin, M.F., Heiman, D.I., et al. (2011). Comparative functional genomics of the fission yeasts. Science 332, 930-936.

129 Samejima, I., Mackie, S., and Fantes, P.A. (1997). Multiple modes of activation of the stress-responsive MAP kinase pathway in fission yeast. The EMBO journal 16, 6162-

6170.

Sanso, M., Vargas-Perez, I., Garcia, P., Ayte, J., and Hidalgo, E. (2011). Nuclear roles and regulation of chromatin structure by the stress-dependent MAP kinase Sty1 of

Schizosaccharomyces pombe. Mol Microbiol 82, 542-554.

Schmid, M., and Jensen, T.H. (2008). The exosome: a multipurpose RNA-decay machine. Trends in biochemical sciences 33, 501-510.

Schonbrun, M., Laor, D., Lopez-Maury, L., Bahler, J., Kupiec, M., and Weisman, R.

(2009). TOR complex 2 controls gene silencing, telomere length maintenance, and survival under DNA-damaging conditions. Mol Cell Biol 29, 4584-4594.

Sengupta, S., Peterson, T.R., and Sabatini, D.M. (2010). Regulation of the mTOR complex 1 pathway by nutrients, growth factors, and stress. Molecular cell 40, 310-322.

Seong, K.H., Li, D., Shimizu, H., Nakamura, R., and Ishii, S. (2011). Inheritance of stress-induced, ATF-2-dependent epigenetic change. Cell 145, 1049-1061.

Sharp, P.A. (2009). The centrality of RNA. Cell 136, 577-580.

Shmueli, O., Horn-Saban, S., Chalifa-Caspi, V., Shmoish, M., Ophir, R., Benjamin-

Rodrig, H., Safran, M., Domany, E., and Lancet, D. (2003). GeneNote: whole genome expression profiles in normal human tissues. C R Biol 326, 1067-1072.

St-Andre, O., Lemieux, C., Perreault, A., Lackner, D.H., Bahler, J., and Bachand, F.

(2010). Negative regulation of meiotic gene expression by the nuclear poly(a)-binding protein in fission yeast. The Journal of biological chemistry 285, 27859-27868.

130 Struhl, K. (2007). Transcriptional noise and the fidelity of initiation by RNA polymerase

II. Nature structural & molecular biology 14, 103-105.

Sugiyama, T., Cam, H., Verdel, A., Moazed, D., and Grewal, S.I. (2005). RNA- dependent RNA polymerase is an essential component of a self-enforcing loop coupling heterochromatin assembly to siRNA production. Proceedings of the National Academy of

Sciences of the United States of America 102, 152-157.

Sugiyama, T., Cam, H.P., Sugiyama, R., Noma, K., Zofall, M., Kobayashi, R., and

Grewal, S.I. (2007). SHREC, an effector complex for heterochromatic transcriptional silencing. Cell 128, 491-504.

Sugiyama, T., and Sugioka-Sugiyama, R. (2011). Red1 promotes the elimination of meiosis-specific mRNAs in vegetatively growing fission yeast. The EMBO journal 30,

1027-1039.

Tabach, Y., Billi, A.C., Hayes, G.D., Newman, M.A., Zuk, O., Gabel, H., Kamath, R.,

Yacoby, K., Chapman, B., Garcia, S.M., et al. (2013). Identification of small RNA pathway genes using patterns of phylogenetic conservation and divergence. Nature 493,

694-698.

Tang, W., Kannan, R., Blanchette, M., and Baumann, P. (2012). Telomerase RNA biogenesis involves sequential binding by Sm and Lsm complexes. Nature 484, 260-264.

Tashiro, S., Asano, T., Kanoh, J., and Ishikawa, F. (2013). Transcription-induced chromatin association of RNA surveillance factors mediates facultative heterochromatin formation in fission yeast. Genes to cells : devoted to molecular & cellular mechanisms

18, 327-339.

131 Tilgner, H., Knowles, D.G., Johnson, R., Davis, C.A., Chakrabortty, S., Djebali, S.,

Curado, J., Snyder, M., Gingeras, T.R., and Guigo, R. (2012). Deep sequencing of subcellular RNA fractions shows splicing to be predominantly co-transcriptional in the human genome but inefficient for lncRNAs. Genome research 22, 1616-1625.

Trojer, P., and Reinberg, D. (2007). Facultative heterochromatin: is there a distinctive molecular signature? Molecular cell 28, 1-13.

Tuck, A.C., and Tollervey, D. (2013). A transcriptome-wide atlas of RNP composition reveals diverse classes of mRNAs and lncRNAs. Cell 154, 996-1009. van Werven, F.J., Neuert, G., Hendrick, N., Lardenois, A., Buratowski, S., van

Oudenaarden, A., Primig, M., and Amon, A. (2012). Transcription of two long noncoding

RNAs mediates mating-type control of gametogenesis in budding yeast. Cell 150, 1170-

1181.

Verdel, A., Jia, S., Gerber, S., Sugiyama, T., Gygi, S., Grewal, S.I., and Moazed, D.

(2004). RNAi-mediated targeting of heterochromatin by the RITS complex. Science 303,

672-676.

Versele, M., Lemaire, K., and Thevelein, J.M. (2001). Sex and sugar in yeast: two distinct GPCR systems. EMBO reports 2, 574-579.

Wang, T., Choi, E., Monaco, M.C., Campanac, E., Medynets, M., Do, T., Rao, P.,

Johnson, K.R., Elkahloun, A.G., Von Geldern, G., et al. (2013). Derivation of neural stem cells from human adult peripheral CD34+ cells for an autologous model of neuroinflammation. PloS one 8, e81720.

132 Wang, T., Choi, E., Monaco, M.C., Major, E.O., Medynets, M., and Nath, A. (2015).

Direct induction of human neural stem cells from peripheral blood hematopoietic progenitor cells. J Vis Exp, 52298.

Weisman, R., Roitburg, I., Schonbrun, M., Harari, R., and Kupiec, M. (2007). Opposite effects of tor1 and tor2 on nitrogen starvation responses in fission yeast. Genetics 175,

1153-1162.

Wilhelm, B.T., Marguerat, S., Watt, S., Schubert, F., Wood, V., Goodhead, I., Penkett,

C.J., Rogers, J., and Bahler, J. (2008). Dynamic repertoire of a eukaryotic transcriptome surveyed at single-nucleotide resolution. Nature 453, 1239-1243.

Wilkinson, M.G., Samuels, M., Takeda, T., Toone, W.M., Shieh, J.C., Toda, T., Millar,

J.B., and Jones, N. (1996). The Atf1 transcription factor is a target for the Sty1 stress- activated MAP kinase pathway in fission yeast. Genes & development 10, 2289-2301.

Wood, V., Harris, M.A., McDowall, M.D., Rutherford, K., Vaughan, B.W., Staines,

D.M., Aslett, M., Lock, A., Bahler, J., Kersey, P.J., et al. (2012). PomBase: a comprehensive online resource for fission yeast. Nucleic acids research 40, D695-699.

Wullschleger, S., Loewith, R., and Hall, M.N. (2006). TOR signaling in growth and metabolism. Cell 124, 471-484.

Yamada, T., Fischle, W., Sugiyama, T., Allis, C.D., and Grewal, S.I. (2005). The nucleation and maintenance of heterochromatin by a histone deacetylase in fission yeast.

Molecular cell 20, 173-185.

Yamamoto, M. (2010). The selective elimination of messenger RNA underlies the mitosis-meiosis switch in fission yeast. Proceedings of the Japan Academy Series B,

Physical and biological sciences 86, 788-797.

133 Yamanaka, S., Mehta, S., Reyes-Turcu, F.E., Zhuang, F., Fuchs, R.T., Rong, Y., Robb,

G.B., and Grewal, S.I. (2013). RNAi triggered by specialized machinery silences developmental genes and retrotransposons. Nature 493, 557-560.

Yamanaka, S., Yamashita, A., Harigaya, Y., Iwata, R., and Yamamoto, M. (2010).

Importance of polyadenylation in the selective elimination of meiotic mRNAs in growing

S. pombe cells. The EMBO journal 29, 2173-2181.

Yamane, K., Mizuguchi, T., Cui, B., Zofall, M., Noma, K., and Grewal, S.I. (2011).

Asf1/HIRA facilitate global histone deacetylation and associate with HP1 to promote nucleosome occupancy at heterochromatic loci. Molecular cell 41, 56-66.

Yamashita, A., Takayama, T., Iwata, R., and Yamamoto, M. (2013). A novel factor Iss10 regulates Mmi1-mediated selective elimination of meiotic transcripts. Nucleic acids research 41, 9680-9687.

Zardo, G., Fazi, F., Travaglini, L., and Nervi, C. (2005). Dynamic and reversibility of heterochromatic gene silencing in human disease. Cell Res 15, 679-690.

Zhang, C.J., Zhou, J.X., Liu, J., Ma, Z.Y., Zhang, S.W., Dou, K., Huang, H.W., Cai, T.,

Liu, R., Zhu, J.K., et al. (2013). The splicing machinery promotes RNA-directed DNA methylation and transcriptional silencing in Arabidopsis. The EMBO journal 32, 1128-

1140.

Zhang, K., Fischer, T., Porter, R.L., Dhakshnamoorthy, J., Zofall, M., Zhou, M.,

Veenstra, T., and Grewal, S.I. (2011). Clr4/Suv39 and RNA quality control factors cooperate to trigger RNAi and suppress antisense RNA. Science 331, 1624-1627.

134 Zhang, K., Mosch, K., Fischle, W., and Grewal, S.I. (2008). Roles of the Clr4 methyltransferase complex in nucleation, spreading and maintenance of heterochromatin.

Nature structural & molecular biology 15, 381-388.

Zhou, Y., Zhu, J., Schermann, G., Ohle, C., Bendrin, K., Sugioka-Sugiyama, R.,

Sugiyama, T., and Fischer, T. (2015). The fission yeast MTREC complex targets CUTs and unspliced pre-mRNAs to the nuclear exosome. Nat Commun 6, 7050.

Zhu, J., Adli, M., Zou, J.Y., Verstappen, G., Coyne, M., Zhang, X., Durham, T., Miri, M.,

Deshpande, V., De Jager, P.L., et al. (2013). Genome-wide chromatin state transitions associated with developmental and environmental cues. Cell 152, 642-654.

Zofall, M., Fischer, T., Zhang, K., Zhou, M., Cui, B., Veenstra, T.D., and Grewal, S.I.

(2009). Histone H2A.Z cooperates with RNAi and heterochromatin factors to suppress antisense RNAs. Nature 461, 419-422.

Zofall, M., and Grewal, S.I. (2006). Swi6/HP1 recruits a JmjC domain protein to facilitate transcription of heterochromatic repeats. Molecular cell 22, 681-692.

Zofall, M., Smith, D.R., Mizuguchi, T., Dhakshnamoorthy, J., and Grewal, S.I. (2016).

Taz1-Shelterin Promotes Facultative Heterochromatin Assembly at Chromosome-Internal

Sites Containing Late Replication Origins. Molecular cell 62, 862-874.

Zofall, M., Yamanaka, S., Reyes-Turcu, F.E., Zhang, K., Rubin, C., and Grewal, S.I.

(2012). RNA elimination machinery targeting meiotic mRNAs promotes facultative heterochromatin formation. Science 335, 96-100.

Zoncu, R., Efeyan, A., and Sabatini, D.M. (2011). mTOR: from growth signal integration to cancer, diabetes and ageing. Nature reviews Molecular cell biology 12, 21-35.

135 CURRICULUM VITAE

Birth Place: Seoul, South Korea

Citizenship: United States

Education

2010 – Present The Johns Hopkins University and National Institutes of Health Partnership Doctorate of philosophy in Biomedical Science Ph.D. – Anticipated 2016 Fall

2008 – 2010 The Johns Hopkins University Zanvyl Krieger School of Arts and Sciences Master of Science in Biotechnology

2005 – 2008 The University of Iowa BS, Biochemistry BS, Chemistry with Honor

2002 – 2005 Valley High School, West Des Moines, Iowa

2001 – 2002 Myungduk High School, Seoul, South Korea

Research Experience

2011 Fall – Present JHU-NIH Fellow Cancer Research Training Award Fellow National Cancer Institute, National Institutes of Health Dr. Shiv Grewal, NIH Distinguished Investigator

2011 Spring JHU-NIH Fellow Investigation of a universal conserved core model of telomerase RNA in yeast Dr. David Zappulla, Assistant Professor

2010 Fall JHU-NIH Fellow Anteriror Follicle Cells Function as Corpus Luteum-like Cells in Controlling Egg Maturation and Ovulation in Drosophila Melanogaster Dr. Allan Spradling, Professor

136 2008 – 2010 Intramural Research Training Award (IRTA) Fellow National Cancer Institute, National Institutes of Health Molecular mechanisms regulating MHC Class I gene expression Dr. Dinah S. Singer, Senior Investigator, Acting Deputy Director

2005 – 2008 Research Internship. Iowa Bioscience Advantage (IBA) Scholar University of Iowa, IA Functionalization of zeolite nanoparticles for applications in gene therapy. Dr. Sarah C. Larsen, Professor, Department of Chemistry Dr. Aliasger K. Salem, Professor, Department of Pharmaceutics, College of Pharmacy

2007 Summer Yale University Research Fellowship. Leadership Alliance Scholar Yale University, CT Activation of T cells by using single wall carbon nanotubes. Dr. Tarek Fahmy, Associate Professor, Biomedical Engineering and Chemical Engineering

Honors and Awards

2009 – Present Cancer Research Training Award (CRTA) at NCI/NIH 2009 Presidential Inaugural Scholar 2008 Intramural Research Training Award (IRTA) at NCI/NIH 2008 The American Institute of Chemists Award 2008 The University of Iowa Department of Chemistry Award 2008 Russell K. Simms Scholarship 2007 Honors Commendation Award 2007 Alpha Chi Sigma Professional Chemistry Fraternity Scholarship 2007 Leadership Alliance Scholar at Yale University 2007 Outstanding Poster Presentation Award at Honors Symposium at the University of Iowa 2007 Scholar at Center for Research on Interface Structure and Phenomena 2006 Presenter Award at SACNAS National Conference 2006 – 2008 IMSD Scholar – Iowa Bioscience Advantage (IBA) Scholar 2005 – 2008 University of Iowa Scholars Scholarship 2005 – 2008 Dean’s List at the University of Iowa

Memberships

Member, American Chemical Society (ACS) Member, American Institute of Chemists Member, National Scholars Honor Society Member, National Society of Collegiate Scholars (NSCS) Member, Phi Eta Sigma, Honors Society

137 Teaching

2010 Fall Biochemistry, The Johns Hopkins University 2011 Spring Cell Biology, The Johns Hopkins University

Presentations

2015 The 8th International Fission Yeast Meeting, Kobe, Japan 2014 Keystone Symposia on Molecular and Cellular Biology: Transcription 2014 Center of Excellence in Chromosome Biology Seminar, NCI, NIH 2014 Washington Area Yeast Club, NCI, NIH 2014 Annual NIH Graduate Student Research Symposia 2008 The University of Iowa Department of Chemistry Undergraduate Award Poster Presentation 2007 Oral Presentation at Leadership Alliance National Conference 2007 Poster Presentation at Center for Biocatalysis and Bioprocessing Annual Conference 2007 Poster Presentation at Tau Beta Pi's Annual Scholz Symposium titled "Frontiers & Horizons: The Future of Biotechnology and Engineering 2006 Poster Presentation at IBA Annual Symposium 2006 Poster Presentation at SACNAS National Conference

Publications

1. Lee NN, Chalamcharla VR, Reyes-Turcu F, Mehta S, Zofall M, Balachandran V, Dhakshnamoorthy J, Taneja N, Yamanaka S, Zhou M, Grewal SI. Mtr4-like protein coordinates nuclear RNA processing for heterochromatin assembly and for telomere maintenance. Cell. 2013 Nov 21;155(5):1061-74

2. Lee N, Iyer SS, Mu J, Weissman JD, Ohali A, Howcroft TK, Lewis BA, Singer DS. Three novel downstream promoter elements regulate MHC class I promoter activity in mammalian cells. PLoS One. 2010 Dec 13;5(12)

3. Pearce ME, Mai HQ, Lee N, Larsen SC, Salem AK. Silicalite nanoparticles that promote transgene expression. Nanotechnology. 2008 Apr 30;19(17):175103.

138