Creation of Versatile Cloning Platforms for Transgene Expression and Epigenome

Editing and Their Application to Pancreatic Islet Biology

by

Jonathan Mark Haldeman

Department of Pharmacology and Cancer Biology Duke University

Date:______Approved:

______Christopher Newgard, Supervisor

______Deborah Muoio

______Donald McDonnell

______Qi-Jing Li

______Praveen Sethupathy

Dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Pharmacology and Cancer Biology in the Graduate School of Duke University

2018

i

v

ABSTRACT

Creation of Versatile Cloning Platforms for Transgene Expression and Epigenome

Editing and Their Application to Pancreatic Islet Biology

by

Jonathan Mark Haldeman

Department of Pharmacology and Cancer Biology Duke University

Date:______Approved:

______Christopher Newgard, Supervisor

______Deborah Muoio

______Donald McDonnell

______Qi-Jing Li

______Praveen Sethupathy

An abstract of a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Pharmacology and Cancer Biology in the Graduate School of Duke University

2018

Copyright by Jonathan Mark Haldeman 2018

Abstract

Insulin secreting β-cells within the pancreatic islets of Langerhans are vital to maintaining glycemic control. β-cell functional mass is lost during the progression to both Type 1 and Type 2 diabetes mellitus, resulting in hyperglycemia. Therefore, a major goal of diabetes research is to uncover pathways that can be exploited to induce β-cell replication while simultaneously maintaining β-cell function.

We previously reported that adenovirus-mediated overexpression of the transcription factor PDX1 is sufficient to induce β-cell replication, but underlying mechanisms remain to be resolved. Using statistical modeling, we identified the miR-17 family, a member of the miR17~92 miRNA cluster, as a candidate regulator of the PDX1-gene network. We show that PDX1 can directly regulate the MIR17HG promoter, the first example of β-cell specific regulation for this important miRNA cluster. Furthermore, the miR17~92 target

PTEN is reduced in PDX1-overexpressing β-cells, and chemical inhibition of PTEN potentiates PDX1-mediated β-cell replication, supportive of the presence of a

PDX1/miR17~92/PTEN regulatory node.

Recombinant adenovirus approaches pioneered by our laboratory have been the main method of genetic manipulation of primary islets in culture since 1994. Whereas adenovirus vectors have proved useful in an otherwise difficult model system, virus construction, especially for cell-type specific applications, is still laborious and time-

iv

consuming. To overcome this, we have created a new modular cloning system (pMVP) that allows a gene of interest to be rapidly recombined in the context of an array of promoters, N- or C-terminal epitope tags, inducible gene expression modalities, and/or fluorescent reporters, into 18 custom destination vectors, including adenovirus, expression plasmid, lentivirus, and Sleeping Beauty transposon, thus, permitting the creation of > 8000 unique vector permutations. Multiple features of this new vector platform, including cell type-specific and inducible control of gene expression, were validated in the setting of pancreatic islets and other cellular contexts. Furthermore, using pMVP as a foundation, we also developed an S. aureus dCas9 epigenetic engineering platform, pMAGIC, that enables the packaging of 3 guide RNAs with Sa- dCas9 fused to one of five epigenetic modifiers into a single vector. Using pMAGIC- derived adenoviruses, we functionally validated the regulation of PDX1 by Area IV, a cross-species conserved enhancer, through LSD1-mediated epigenetic modification in both INS1 832/13 cells and primary rat pancreatic islets.

In sum, my work has uncovered novel information about the role of PDX1 in regulation of the miR17~92 miRNA cluster in pancreatic islet cells. In an effort to contribute more broadly to our laboratory’s pancreatic islet research efforts, I also designed and built the pMVP and pMAGIC systems for efficient generation of purpose- built, customized vectors for manipulation of gene expression in islets and other cell types, including via targeted epigenetic modification of putative regulatory elements

v

within their native chromatin context. Development of this novel vector platform facilitated additional discoveries about the role of Area IV in control of PDX1 expression in islet β-cells.

vi

Dedication

To my late grandfather Albert, who was a continual example of self-motivation,

perseverance, unconditional love, and the pursuit of knowledge.

vii

Contents

Abstract ...... iv

List of Tables ...... xii

List of Figures ...... xiii

List of Abbreviations ...... xvi

Acknowledgements ...... xix

1. Introduction ...... 1

1.1 The Pancreatic Islets of Langerhans ...... 2

1.2 Pancreatic Islet Dysfunction in Diabetes ...... 5

1.3 PDX1 (MODY4) is a Master Transcription Factor Required for β-Cell Development and Identity ...... 7

1.4 PDX1 Overexpression is Sufficient to Induce β-cell Replication ...... 15

1.5 Recombinant Adenovirus Vectors Permit Genetic Manipulation of Primary Islets ...... 17

1.6 Cas9-Based Epigenetic Engineering Permits Interrogation of Disease-Associated Regulatory Elements Within Native Chromatin Contexts ...... 19

1.7 Goals of the Dissertation ...... 23

2. Methods ...... 27

3. A PDX1/miR17~92/PTEN Axis Regulates β-Cell Replication ...... 46

3.1 Introduction ...... 46

3.2 Results ...... 49

3.2.1 miR-17 Target Sites are Enriched in the PDX1 Gene Network ...... 49

3.2.2 PDX1 Directly Regulates the Human MIR17HG Promoter ...... 54

viii

3.2.3 The PDX1 DNA Binding Motif in the MIR17HG Promoter is Conserved in the Rat Genome ...... 58

3.2.4 Identification of Putative Human Islet MIR17HG Enhancers...... 60

3.2.5 miR17~92 miRNAs are Induced by PDX1 Overexpression ...... 67

3.2.6 Peptide-Nucleic Acids as a Tool for Islet Research ...... 70

3.2.7 Regulation of PDX1-Mediated Replication by miR17~92 miRNAs ...... 73

3.2.8 miR17~92 miRNAs are Not Sufficient to Drive β-cell Replication ...... 74

3.2.9 The miR-17~92 Target PTEN Regulates PDX1-Mediated Replication ...... 78

3.3 Discussion ...... 80

4. Creation of pMVP, a Versatile Cloning Platform for Transgene Expression ...... 85

4.1 Introduction ...... 85

4.2 Results ...... 88

4.2.1 Application of MultiSite Gateway® Pro to Enable Rapid Assembly of Multicomponent Vectors ...... 88

4.2.2 Generation of pMVP Entry Plasmids ...... 91

4.2.2.1 Promoters ...... 91

4.2.2.2 Genes of Interest and Control Genes ...... 94

4.2.2.3 shRNA Expression ...... 95

4.2.2.4 Epitope Tags ...... 97

4.2.2.5 Fluorescent Reporters ...... 99

4.2.2.6 Degron Technologies ...... 100

4.2.2.7 Tools for Protein Interaction Partner Identification ...... 104

4.2.3 Generation of Custom Gateway Destination Vectors ...... 106

ix

4.2.3.1 Expanded Capacity Adenovirus Vector ...... 107

4.2.3.2 Fiber-Modified Adenovirus Vectors ...... 110

4.2.3.3 Expression Plasmid Vectors ...... 113

4.2.3.4 Lentivirus Vectors ...... 115

4.2.3.5 Sleeping Beauty Transposon Vectors ...... 118

4.2.4 pMVP Validation ...... 121

4.2.4.1 Epitope Tags and Fluorescent Reporters ...... 121

4.2.4.2 Modes of Inducible Transgene Regulation ...... 123

4.2.4.3 Fiber-Modified Adenoviruses ...... 126

4.3 Discussion ...... 128

5. Identification of PDX1-Regulated Secreted Factors Using pMVP-Derived Vectors ... 130

5.1 Introduction ...... 130

5.2 Results ...... 132

5.2.1 pMVP Vectors Enable the Isolation of Transduced β-Cells ...... 132

5.2.2 Identification of Secreted Factors Specifically Regulated by PDX1 ...... 135

5.3 Discussion ...... 139

6. Creation of pMAGIC, a Modular Cloning Platform for Epigenetic Engineering ...... 141

6.1 Introduction ...... 141

6.2 Results ...... 143

6.2.1 Generation of Sa-Cas9 Entry Plasmids ...... 143

6.2.2 Generation of gRNA Expressing Entry Plasmids ...... 145

6.2.3 pMAGIC Validation ...... 147

x

6.3 Discussion ...... 149

7. Utilization of pMAGIC to Functionally Validate the Contribution of Area IV in the Regulation of PDX1 ...... 152

7.1 Introduction ...... 152

7.2 Results ...... 153

7.2.1 pMAGIC-Derived Vectors ...... 153

7.2.2 pMAGIC Enables Epigenetic Manipulation of Area IV ...... 154

7.2.3 Area IV is the Major Contributor to PDX1 Transactivation in INS1 832/13 Cells ...... 157

7.2.4 Area IV Regulates PDX1 in Primary Rat Islets ...... 158

7.3 Discussion ...... 160

8. Conclusions and Perspectives ...... 163

8.1 β-Cell Replication: A Cure for Type-2 Diabetes? ...... 164

8.2 PDX1 Dysregulation: Driving Disease? ...... 166

8.3 pMVP and pMAGIC: Unmasking the Etiology of Disease ...... 168

Appendix A...... 170

Appendix B ...... 172

Appendix C ...... 178

Appendix D...... 187

References ...... 193

Biography ...... 229

xi

List of Tables

Table 1: MIR17HG Enhancer Luciferase Plasmids...... 32

Table 2: EMSA Probes...... 34

Table 3: gRNA Protospacer Motif Oligonucleotide Sequences...... 37

Table 4: RCA Detection Assay qRT-PCR Primers...... 41

Table 5: TaqMan Probe qRT-PCR Primers...... 42

Table 6: ChIP qRT-PCR Primers...... 44

Table 7: Most Abundant miRNA Targets Identified in the PDX1 and NKX6.1 Transcription Networks...... 51

Table 8: PDX1 Gene Network Components Targeted by miR-17...... 54

xii

List of Figures

Figure 1: Pancreatic Islets of Langerhans are Comprised of Several Distinct Endocrine Cell-Types...... 4

Figure 2: The Area I-IV Regulatory Elements are Highly Conserved...... 12

Figure 3: Fasting Glucose-Associated SNPs Overlap with Putative PDX1 Regulatory Elements...... 15

Figure 4: Cas9 Enables Sequence Specific Genome Targeting...... 21

Figure 5: dCas9 Fusion to Transcriptional and Epigenetic Modifiers Enables Locus Specific Targeting...... 23

Figure 6: PDX1-Bound miRNAs Target the PDX1 Gene Network...... 53

Figure 7: PDX1 Directly Regulates the MIR17HG Promoter...... 57

Figure 8: Cross-Species Alignment of the MIR17HG Promoter...... 59

Figure 9: Functional Validation of ChromHMM Identified Islet Enhancers...... 64

Figure 10: SNP rs9583907 Does Not Alter Enhancer Activity in INS1 832/13 Cells...... 66

Figure 11: miR17~92 is Induced by PDX1 Overexpression...... 68

Figure 12: PDX1-Mediated Increases in miR17HG are not Blunted by Myc Inhibition. .. 70

Figure 13: Validation of Peptide-Nucleic Acids as a Tool for miRNA Research in Primary Islets...... 72

Figure 14: miR17~92 Inhibition by PNAs Does Not Alter PDX1-Mediated Replication. . 74

Figure 15: miR17~92 is Not Sufficient to Drive β-Cell Replication...... 77

Figure 16: PTEN Status Regulates PDX1-Mediated Replication...... 80

Figure 17: Proposed Model of β-Cell Replication Driven by the PDX1/miR17~92/PTEN Axis ...... 84

Figure 18: Overview of MultiSite Gateway Pro Utilization in pMVP...... 89

xiii

Figure 19: pMVP Enables Rapid, High Fidelity Assembly of Custom Multicomponent Transgene Vectors...... 90

Figure 20: pMVP Promoter Entry Plasmids...... 93

Figure 21: pMVP Genes of Interest...... 95

Figure 22: pMVP RNAi...... 97

Figure 23: pMVP Epitope Tags...... 99

Figure 24: pMVP Fluorescent Protein Fusions and Reporters...... 100

Figure 25: pMVP Inducible Systems...... 104

Figure 26: pMVP Components for Interaction Partner Identification...... 105

Figure 27: pMVP Custom Destination Vectors...... 107

Figure 28: Creation of pMVP/Ad-DEST...... 109

Figure 29: Fiber Gene Modification Alters Adenovirus Tropism...... 111

Figure 30: Pipeline for Adenovirus Fiber Modification...... 112

Figure 31: Generation of pMVP-DEST Vectors...... 114

Figure 32: Creation of pMVP/Lenti-DEST Vectors...... 117

Figure 33: Creation of pMVP Compatible Sleeping Beauty Transposon Vectors...... 120

Figure 34: Utilization of pMVP for the Creation of Unique PDX1 Vectors...... 122

Figure 35: Creation of a Sleeping Beauty-Derived eGFP-PDX1 Stable Cell Line...... 123

Figure 36: Demonstration of pMVP Inducible Systems...... 126

Figure 37: RGD-Modified Ad5 Fiber Enables Efficient Transduction of C2C12 Myoblasts...... 127

Figure 38: pMVP Vectors Enable Isolation of β-Cells...... 133

xiv

Figure 39: A pMVP-Derived Bicistronic PDX1/eGFP Adenovirus Permits Selective Isolation of PDX1-Expressing β-Cells...... 135

Figure 40: Identification and Characterization of PDX1-Induced Secreted Factors...... 138

Figure 41: pMAGIC Enables the Generation of Multicomponent Sa-Cas9 Expression Vectors...... 145

Figure 42: pMAGIC Components for gRNA Expression...... 147

Figure 43: Activation of a Synthetic Promoter by Sa-dCas9-Mediated Recruitment of VPR...... 148

Figure 44: Generation of pMAGIC-Derived Adenoviruses Targeting Area IV...... 154

Figure 45: Alteration of the Area IV Epigenetic State by LSD1 Recruitment...... 156

Figure 46: Attenuation of Area IV is Sufficient to Decrease PDX1 and Disrupt the PDX1 Transcriptional Network...... 158

Figure 47: Area IV Regulates PDX1 Expression in Primary Rat Islets...... 160

xv

List of Abbreviations

Ad recombinant adenovirus vector

Ad5 human adenovirus serotype 5

ALB chimeric albumin enhancer/promoter

βGAL Beta-galactosidase

BP recombination reaction between attB and attP sites bp base pair (a single nucleotide unit of DNA) bps multiple base pairs

Cas9 CRISPR associated protein 9 cDNA complementary DNA

ChIP chromatin immunoprecipitation

ChromHMM chromatin state identification using hidden Markov modeling

CMV cytomegalovirus cTnT cardiac troponin T dCas9 nuclease-dead Cas9

DNA deoxyribonucleic acid

EF1a elongation factor 1-alpha eCFP enhanced cyan fluorescent protein eGFP enhanced green fluorescent protein eYFP enhanced yellow fluorescent protein

xvi

ESC embryonic stem cell gDNA genomic DNA gRNA Cas9 guide RNA

GWAS genome-wide association study

H3K27ac histone H3 lysine 27 acetylation kb kilobase (1,000 nucleotides of DNA)

LR recombination reaction between attL and attR sites

MACS model-based analysis of ChIPseq

Mb megabase (1,000,000 nucleotides of DNA) miR17~92 miRNA cluster encoding miRs-17, 18a, 19a, 20a, 19b, and 92a

MIR17HG host gene for the miR17~92 cluster miRNA microRNA

MODY maturity-onset diabetes of the young

NKX6.1 NK6 Homeobox 1

PAM protospacer adjacent motif

PCR polymerase chain reaction pDEST Gateway destination vector

PDX1 Pancreatic and Duodenal Homeobox 1 pENTR Gateway entry plasmid

xvii

pMAGIC plasmid-based modular adenovirus for genomic interrogation

with Cas9 pMVP plasmid-based modular vector platform

PolII RNA polymerase II

PolIII RNA polymerase III qRT-PCR quantitative real-time polymerase chain reaction

RefSeq NCBI reference sequence database

RIP rat insulin 1 promoter

RNAi RNA interference

SB Sleeping Beauty transposon system shRNA short hairpin RNA

SNP single nucleotide polymorphism

T1D Type 1 Diabetes

T2D Type 2 Diabetes

TBG thyroxine binding globulin

TOPO DNA topoisomerase I-based cloning

TRE tet-responsive element

TSS transcription start site

UTR untranslated region

xviii

Acknowledgements

The efforts described hereafter are indebted to the collaborative spirit of the

Newgard laboratory as well as our external collaborators. Specific gratitude is owed to my mentor, Chris Newgard, for fostering an environment of intellectual curiosity that enables the freedom to explore new scientific paradigms. This environment has yielded many tools and methods (i.e. adenoviruses, INS1 832/13 cells, quantitative metabolomics) applied ubiquitously throughout the islet biology and metabolism fields.

In addition, Chris serves as a model for the embodiment of the spirit of collegial scientific pursuit through his emphasis on the formation of teams to collaboratively scale nature’s daunting scientific peaks.

Secondly, I am also indebted to the Newgard lab family who have been instrumental in shaping both my personal and professional life. Tom, a fellow

Millersville alumnus and resident gene jock, introduced me to the Newgard lab, β-cell biology, miRNAs, and adenoviruses thru a summer research stint during my undergraduate years. More recently, he has been instrumental in the creation of pMVP and pMAGIC. Mette was my first supervisor when I returned to the Newgard lab as a technician. She taught me the ins and outs of β-cell function, replication, and the application of adenoviruses to islet biology. Her willingness to beta-test new viral and cell line approaches produced by pMVP has been crucial in validating this platform.

Personally, Mette been welcoming from the moment I arrived and has been a friend ever

xix

since. Through the organization of poker nights, dinner parties, celebratory parties, meal trains, Danish Dirty Santa, and other lab outings, she has had an outsized role in nurturing our tight-knit lab family. In a similar manner, Michelle has also been a source of enormous encouragement and help both in and out of the lab. Whether it is doing experiments in the lab, proofing graduate school essays, or hosting celebratory events for departures as well as new beginnings, Michelle has always been quick to lend a helping hand. She also has been an indispensable source of encouragement as I have navigated this process. Hans has also been an important mentor and friend, who is responsible for introducing me to PDX1 biology. I would be remiss if I did not acknowledge others who over the years have been part of my lab family and have made

North Carolina feel like home, especially: Bob and Jane; Hans, Joni and Contessa; the

Arlotto family; Pat; Lauren; Brett; Sarah; Angie; Phill; Brett; Sam; Jeff; Lu; Danhong;

Helena; Taylor; as well as Chris and Patricia.

My external collaborators have largely shaped my current interests in epigenomic dysfunction in disease. A special thank you must be given to Praveen

Sethupathy and the members his lab. Through participation in their lab meetings and collaboration, I was exposed to the world of genomics and computational biology as well as their utilization to inform wet lab experiments. My interactions with the

Sethupathy lab have also led to a collaboration with the labs of Stephen Parker, Michael

xx

Stitzel, and Francis Collins that has further spurred my interests in the intersection of epigenomics and disease.

Finally, I have been blessed beyond measure to have a family who has been a source of continual love and encouragement, not only through graduate school, but throughout my entire life. Even with his recent passing, my grandfather Albert still remains a source of inspiration. Having left the education system in the eighth grade to work on the family farm, he exemplified that the pursuit of knowledge is not driven by academic achievement, but rather academic achievement is merely a byproduct of the pursuit of knowledge. Up until his passing, he voraciously consumed non-fiction books and newspapers in a continual effort to gain more thorough insight into the world around him. I will never forget the pride and joy he had when gathered with the sixty members of his family every holiday, and the gratitude he had when receiving a stack of new books as presents. I must also thank my parents who helped shape me in my formidable years and have provided steadfast support throughout my pursuits, despite their wishes that I would pursue them a little closer to home.

A year and half ago, a new source of inspiration entered my life. Through her smiles, energy, and laughter, my daughter Raedynn has been a daily source of joy. It is my hope that the sacrifices made by individuals in pursuit of scientific knowledge will lead to a life with fewer maladies for her and her peers. Lastly, my wife Amanda, whom

I had the great fortune to meet during my graduate school recruitment weekend at

xxi

Duke, is directly responsible for my perseverance throughout this endeavor. She has been patient and understanding of the late nights spent at the bench as well as the regular rantings and ravings that were part of this process. In addition, these efforts have been greatly aided by her scientific insights and collaboration. Lastly, her unwavering love and support has been essential to the completion of this chapter of my life, and I look forward to seeing where our journey next leads.

xxii

1. Introduction

In the year 1889, removal of the pancreas was found to be sufficient to cause the onset of severe diabetes mellitus, a disease characterized by chronic hyperglycemia.

However, it was not until thirty-three years later, in 1922, that loss of the pancreatic

“islands of Langerhans”, clusters of endocrine cells comprising 1-2% of the pancreas, was identified as the root cause. In a series of landmark studies, Frederick Banting and

Charles Best demonstrated that pancreatic extracts highly enriched for these endocrine cells were sufficient to control glycemia in pancreatectomized dogs (Banting and Best

1922), normal rabbits (Banting et al. 1922b), and most importantly, in diabetic humans

(Banting et al. 1922a). Further efforts by John Abel resulted in the crystallization of the protein sufficient for glycemic regulation, insulin (Abel 1926). These discoveries have ushered us into a modern era in which diabetes is no longer a terminal diagnosis, but instead, through chronic administration of insulin, can be successfully managed for decades.

Nevertheless, despite the successes of the pharmacologic utilization of insulin for the treatment of diabetes, it is imperative to remember the final remarks of Frederic

Banting in his Nobel Lecture (Banting 1925).

“Insulin is not a cure for diabetes; it is a treatment. It enables the diabetic to burn sufficient carbohydrates, so that proteins and fats may be added to the diet in sufficient quantities to provide energy for the economic burdens of life.”

1

To this point, insulin therapy has significant, and potentially fatal side effects, most notably severe hypoglycemia, but also including weight gain and potentially an increased risk for cancer, although the latter is still controversial (Bowker et al. 2006,

Lebovitz 2011, Carstensen et al. 2016, Klil-Drori et al. 2017). While we have made significant progress over the past century in our quest to understand the etiology of diabetes and, in turn, the pancreatic islet β-cell, which is solely responsible for the production of insulin and thereby maintenance of glycemia, a cure for this disease has still not been found. Furthermore, the fact that over one-third of the population of the

United States is either prediabetic or diabetic (Centers for Disease Control and

Prevention 2017), demands continuing our pursuit of a cure. In the following pages, we discuss identification of pathways sufficient to induce β-cell replication as well as the development of innovative platforms that permit rapid generation of highly-specialized vectors for genetic manipulation of the islets of Langerhans that together contribute to our understanding of islet biology and potential now therapeutic modalities.

1.1 The Pancreatic Islets of Langerhans

The pancreatic islands/islets of Langerhans, hereafter referred to as “islets”, were first observed in the rabbit pancreas by Paul Langerhans in 1869 (Barach 1952), but the importance of these cells would not be truly appreciated for over fifty years, when

Banting and Best discovered insulin (Banting and Best 1922). Islets are highly- vascularized, complex, ~1,000 cell micro-organs comprising 1%-2% of pancreatic mass

2

and containing multiple distinct endocrine cell types that participate in metabolic fuel homeostasis (Jain and Lammert 2009). The majority of islet cell mass consists of β-cells, which produce and secrete insulin in a regulated manner in response to elevated circulating glucose levels to prevent hyperglycemia (Lacy and Davies 1957, Jensen et al.

2008). Conversely, the α-cells function to prevent hypoglycemia through the production and secretion of glucagon (Baum et al. 1962, Campbell and Drucker 2015). Additionally, the islet contains endocrine cells that produce somatostatin (δ-cells) (Goldsmith et al.

1975), pancreatic polypeptide (γ-cells) (Kimmel et al. 1975), and ghrelin (ε-cells) (Prado et al. 2004). Cross-species examination of islet composition has shown that islets uniformly contain these distinct endocrine cell-types, however, the relative ratios and placement within the islet are varied (Steiner et al. 2010). The most striking of these differences is the restriction of non-β-cells to the islet mantle in rodent islets (Figure 1A), whereas in human islets the cell types are intermixed throughout the islet and α-cells are present in almost equal numbers to β-cells (Figures 1B-C). In general, islet cell architecture is through to be related to inter-islet regulation pathways that allow the distinct cell-types to modulate the function and quantity of other cell types to maintain organismal metabolic homeostasis, but this is still an area of active research (Jain and

Lammert 2009, Caicedo 2013, Smith et al. 2014, Aamodt and Powers 2017). The presence of diverse cell types within islets presents experimental challenges when manipulation of a single cell type rather than all cell types is desired. As such, the field is continually

3

seeking experimental approaches that enable cell-type specific manipulation in intact islets.

Figure 1: Pancreatic Islets of Langerhans are Comprised of Several Distinct Endocrine Cell-Types. Schematic depictions of a representative cross-section of a (A) normal rat and (B) normal human islet of Langerhans. The presence of distinct endocrine cell-types is conserved, but the ratio and inter-islet organization are not. Most notably, (A) in normal rodents, the islet core is comprised almost solely of β-cells, with the other cell- types being largely restricted to the islet mantle; however, (B) in human islets this level of organization is not observed. (C) Legend for the endocrine cell-types, the major hormone produced, as well as the average contribution of that cell-type to the islet composition in rats and humans.

4

1.2 Pancreatic Islet Dysfunction in Diabetes

The root cause of diabetes is a failure of the pancreatic islet to produce sufficient insulin to cope with daily metabolic demands, leading to the chronic elevation of circulating glucose levels, referred to as hyperglycemia. The two main forms of diabetes are Type-1 (T1D), an autoimmune disease in which β-cells are selectively destroyed, and

Type-2 (T2D), a disease of progressive β-cell dysfunction and death (Cnop et al. 2005).

Of these, T2D is overwhelmingly more abundant and estimated to represent 90%-95% of all instances of diabetes (Centers for Disease Control and Prevention 2017). Both of these forms are caused by an unidentified intersection of genetic predisposition and environmental risk factors. For T1D, heritable predisposition has implicated 50 candidate genes, and recent studies are suggestive that β-cells may be especially susceptible to coxsackievirus infections, but there is still no rationale for why < 1% of coxsackievirus infected children are stricken with T1D (Marroqui et al. 2015, Nyalwidhe et al. 2017, Eizirik and Op de Beeck 2018). While there is a strong correlation between obesity and T2D, the phenotype is only partially penetrant, suggesting that the pro- inflammatory, insulin resistant state associated with obesity is not sufficient to cause

T2D. Additionally, despite the genome-wide association of 113 loci with T2D traits, these loci individually only account for a 5%-20% increase in T2D risk, indicative of a polygenic disease with a strong environmental component (Eckel et al. 2011, Kahn et al.

2014, Bonnefond and Froguel 2015, Prasad and Groop 2015, McCarthy 2017, Scott et al.

5

2017). For T2D treatment, oral agents and injectable biologics directly or indirectly targeting insulin resistance pathways and islet function can be sufficient for most diabetics to maintain long-term glycemic control (Lebovitz and Feinglos 1978, Foretz et al. 2014, Drucker et al. 2017). However, in a seemingly variable manner, T2D can progress to a point where even pharmaceutical intervention to aid the β-cell is insufficient, necessitating the use of escalating doses of insulin to maintain glycemic control. For both T1D and T2D, poor long-term glycemic management can lead to extensive complications, including nephropathy, retinopathy, neuropathy, coronary artery disease, peripheral artery disease, and stroke (Pinhas-Hamiel and Zeitler 2007,

Forbes and Cooper 2013). These are in addition to the ever-present acute risk of hypoglycemia (Lebovitz 2011). Therefore, a major goal of diabetes research is to uncover pathways that can be exploited to either induce β-cell replication or prevent β-cell apoptosis while simultaneously maintaining β-cell function.

In addition to T1D and T2D, monogenic forms of diabetes afflict approximately

2% of diabetic patients. Similar to other rare diseases (Conway et al. 2015), research on genes implicated in monogenic forms of diabetes has led to novel insights into β-cell biology and cellular functions that may go awry in the more prevalent forms of diabetes.

Monogenic diabetes can be further divided into neonatal diabetes and maturity-onset diabetes of the young (MODY). Neonatal diabetes is defined as persistent hyperglycemia requiring insulin management with onset within the first 6 months of

6

life. Due to recent advances in sequencing technology, researchers have been able to uncover 23 distinct genetic causes of neonatal diabetes, including rare recessive mutations in the insulin gene (Stoy et al. 2007, Hattersley and Patel 2017).

The 11 MODY gene mutations (HNF4α, MODY1; GCK, MODY2; HNF1α,

MODY3; PDX1, MODY4; HNF1β, MODY5; NEUROD1, MODY6; KLF11, MODY7; CEL,

MODY8; PAX4, MODY9; INS, MODY10; BLK, MODY11) are inherited in an autosomal dominant manner causing early-onset (< 25 years old) diabetes, and with the exception of age, the symptoms associated with MODY mirror T2D. While once thought to be exceedingly rare, genome sequencing has revealed that MODY forms of diabetes were often mischaracterized as T1D or T2D, and they are now thought to constitute 1%-2% of diabetes patients (Shields et al. 2010, Fajans and Bell 2011, Hattersley and Patel 2017).

1.3 PDX1 (MODY4) is a Master Transcription Factor Required for β-Cell Development and Identity

The parahox homeodomain transcription factor PDX1 (also known as MODY4,

IPF1, STF1, and IDX1) wields profound control over β-cell development, function, and maintenance in both rodent and humans. PDX1 is the earliest marker for pancreatic cell fate, first appearing at E8.5 during mouse development (Guz et al. 1995), and complete loss of PDX1 function results in pancreatic agenesis in both mouse models (Offield et al.

1996) and humans (Stoffers et al. 1997b, Fajans et al. 2010). Correlating with these observations, PDX1-ChIPseq experiments in embryonic stem cell-derived pancreatic progenitors have demonstrated that PDX1 drives pancreatic commitment via repression 7

of hepatic genes (Teo et al. 2015), and by serving as a pioneer factor to activate islet- specific enhancers during development (Wang et al. 2015).

Beyond early development, human mutations of PDX1 are sufficient to cause a spectrum of monogenic diabetes that can be characterized as permanent neonatal

(Nicolino et al. 2010), early-onset (MODY4) (Stoffers et al. 1997a, Fajans et al. 2010), as well as late-onset diabetes that presents as T2D (Hani et al. 1999). While MODY4 patients may exhibit clinical deficiencies of the exocrine pancreas, the impact of these mutations is largely restricted to the islet, and more specifically β-cell function (Nicolino et al. 2010). In agreement with these human pathologies, PDX1 heterozygosity in mice leads to decreased glucose tolerance and a predisposition for diabetes largely attributed to increased β-cell apoptosis (Dutta et al. 1998, Gannon et al. 2001, Johnson et al. 2003,

Sachdeva et al. 2009, Fujimoto et al. 2010). Removal of PDX1 during late-gestation causes overt neonatal diabetes corresponding with a reduction in β-cell number and replication

(Gannon et al. 2008). These observations correlate with known PDX1 functions to directly regulate the proapoptotic genes Puma (BBC3) and Noxa (PMAIP1) (Khoo et al.

2012), as well as genes crucial to β-cell function, most notably insulin (INS) (Glick et al.

2000) and the glucose transporter GLUT2 (SLC2A2) (Waeber et al. 1996). Of note, similar to insulin (Stoy et al. 2007), SLC2A2 mutations have also been associated with a monogenic form of neonatal diabetes (Sansbury et al. 2012). Finally, PDX1 is also required to maintain β-cell identity in adult mice, as shown by inducible β-cell-specific

8

deletion of PDX1 resulting in abrupt downregulation of vital β-cell genes, such as INS and SLC2A2, and a concomitant activation of an α-cell transcriptional program (Ahlgren et al. 1998, Gao et al. 2014). In total, these observations demonstrate that PDX1 is crucial not only for β-cell development, but also to maintain adequate functional β-cell mass for glycemic control. Furthermore, deficient PDX1 activity is sufficient to cause a spectrum of monogenic diabetic disorders in humans, ranging in severity from permanent neonatal to late-onset diabetes.

The essential nature of PDX1 during mouse development has driven many investigations into the regulatory elements responsible for timely PDX1 activation during this period. Transgenic reporter mice expressing βGAL demonstrated that the regulatory elements contained in the region 6.5 kb upstream of the PDX1 transcription start site are sufficient to yield islet-specific βGAL expression (Sharma et al. 1996).

Furthermore, the same 6.5 kb region used for transgenic expression of PDX1 within

PDX1 null mice was sufficient to rescue all developmental and glycemic defects associated with PDX1 deletion (Dutta et al. 2001). Subsequent efforts identified four cross-species conserved enhancers that are located 1.6 kb to 2.7 kb (Areas I-III) and 6 kb to 6.5 kb (Area IV) upstream of the PDX1 transcription start site (Figure 2) (Gerrish et al.

2000, Gannon et al. 2001, Gerrish et al. 2004). Reporter gene expression in transgenic mice has shown that Areas I-III promote expression in the pancreas and duodenum; conversely, inclusion of only Areas I-II is sufficient for islet-specific expression (Gannon

9

et al. 2001). However, while deletion of Areas I-III severely impacted pancreas morphogenesis as expected, rescue attempts utilizing transgene expression of PDX1 driven by Areas I-III was insufficient to completely restore developmental timing and islet function, resulting in glucose intolerance (Boyer et al. 2006, Fujitani et al. 2006). This is in line with early observations that the region encompassing Area IV is required for maximal reporter activity and β-cell specificity (Sharma et al. 1996, Gerrish et al. 2004).

More recently, mice containing deletions for either Area II or Area IV have been reported (Spaeth et al. 2017, Yang et al. 2017). Surprisingly, despite the suggestion that

Area II is crucial for islet PDX1 expression (Gannon et al. 2001), homozygous deletion of

Area II did not result in any gross abnormalities, although the animals did become hyperglycemic after 6 months. However, deletion of Area II in a PDX1 heterozygous background selectively disrupted development of the endocrine pancreas resulting in a

70% reduction in endocrine cell number with no obvious abnormalities of overall pancreas size or exocrine function (Yang et al. 2017). In addition, Area IV deletion was also not sufficient to perturb glucose tolerance, even in aged animals, but studies of other developmental enhancer clusters suggest that compensation from neighboring elements may mask an underlying phenotype (Spaeth et al. 2017, Osterwalder et al.

2018). In contrast to Area II, deletion of Area IV in a PDX1 heterozygous background did not immediately present with an overt phenotype; however, the animals quickly became glucose intolerant post-weaning, correlating with a significant decrease in PDX1

10

expression, decreased β-cell function, and decreased β-cell replication (Spaeth et al.

2017). Curiously, these phenotypes were only evident in male animals, with no apparent explanation for the dichotomous results (Spaeth et al. 2017). These observations are suggestive of an underappreciated role of PDX1, and Area IV, in regulating the post- weaning transition to fully functional, mature β-cells. PDX binds to Area I to form an autoregulatory loop in islets from both pre- and post-weaned animals (Gerrish et al.

2001). Interestingly, PDX1 also binds Area IV, but only in islets harvested post-weaning

(Spaeth et al. 2017). Taken together, these data demonstrate that while Areas I-III regulate PDX1 expression throughout development and maturation, Area IV is selectively involved in PDX1 regulation in adult animals. Furthermore, in a manner similar to the spectrum of disease observed with PDX1 mutations, these results demonstrate a range of diabetic phenotypes that appear upon manipulation of PDX1 expression indirectly via regulatory element function.

11

Figure 2: The Area I-IV Regulatory Elements are Highly Conserved. UCSC genome browser screenshot of (A) published regions corresponding to Areas I-IV in the mouse genome (mm9) as well as lift-over of the sequences into the (B) human (hg19) and (C) rat (rn5) genomes. (C) Note that the PDX1 locus is inverted in the rat genome and also contains a sequence gap overlapping with Area I. In addition, the PLUT LncRNA corresponding with this locus in (A) mouse and (B) human has not yet been identified/annotated in (C) rat.

Despite the overwhelming evidence that (1) PDX1 can control multiple facets of

β-cell biology, ranging from development to death; (2) mutation of PDX1 causes a spectrum of monogenic diabetic disorders; (3) and perturbation of PDX1 regulatory elements is sufficient to disrupt glycemia, we still do not have a clear picture of the role

12

of PDX1 in pathogenesis of T1D or T2D. Whereas the etiology of cancer can be studied via harvesting of tumor cells, in pancreatic islet research, researchers are challenged to infer the etiology of disease from animal and cell culture models that may, or may not, translate to human disease. With that said, in the non-obese diabetic mouse (NOD) T1D model, there is evidence of PDX1 repression prior to the onset of disease coinciding with increased ER stress and exposure to pro-inflammatory cytokines during immune attack

(Tersey et al. 2012, Rui et al. 2017). In the T2D db/db mouse model, there are conflicting reports about whether PDX1 is decreased (Talchai et al. 2012, Guo et al. 2013). However, cell line experiments have found that the ability of PDX1 to bind to chromatin is impaired upon induction of either chronic exposure to supraphysiologic glucose levels or induction of oxidative stress, both of which are thought to occur during the progression of T2D (Tanaka et al. 1999, Guo et al. 2013). Furthermore, PDX1 expression decreases with age, correlating with decreased replication capacity and increased β-cell susceptibility to apoptosis-inducing agents (Maedler et al. 2006). Interestingly, human genome-wide association studies (GWAS) for T2D related traits have identified single nucleotide polymorphisms (SNPs) associated with fasting glucose levels clustered in the

40 kb region surrounding Areas I-IV, but the causal SNP has yet to be identified (Figure

3) (Scott et al. 2012).

Of note, these SNPs also coincide with a 10 kb stretch enhancer in human islets that overlaps with Area IV (Parker et al. 2013). Stretch enhancers, also referred to as

13

super enhancers (Hnisz et al. 2013) or enhancer clusters (Pasquali et al. 2014), are defined as contiguous chromatin regions > 3 kb with epigenetic marks characteristic of enhancer activity. These elements are postulated to be regulators of cellular identity, can be dysregulated and/or hijacked in disease, and have been found to be enriched for

SNPs associated with diseases emanating from that cell-type (Hnisz et al. 2013, Parker et al. 2013, Northcott et al. 2014). For example, stretch enhancers found in human islets are enriched for SNPs associated with T2D traits (Parker et al. 2013, Pasquali et al. 2014). Of note, these regions in human islets also are enriched for PDX1 DNA binding motifs and

PDX1 protein, suggestive that PDX1 may play a role in regulating these elements

(Pasquali et al. 2014). Finally, examination of PDX1 levels in human islets isolated from

T2D donors has revealed that PDX1 expression levels are indeed diminished (Guo et al.

2013, Akerman et al. 2017). These observations are supportive of PDX1 having a role in polygenic diabetes disorders, either through molecular etiologies governing PDX1 function during ER, oxidative, or glucose-induced stress or potentially through genetic predisposition to altered function of PDX1 regulatory elements.

14

Figure 3: Fasting Glucose-Associated SNPs Overlap with Putative PDX1 Regulatory Elements. UCSC genome browser screenshot of the human (hg19) PDX1 locus. Tracks are included for Areas I-IV, human islet ChromHMM and RNAseq data (Parker et al. 2013), consensus peaks from PDX1 ChIPseq in human islets (reanalyzed data from Khoo et al. 2012), and conservation. In addition, fasting glucose-associated SNPs (p- value = 1.1x10-6 to 5.1x10-8) identified by the MAGIC consortium (Scott et al. 2012) are highlighted in red.

1.4 PDX1 Overexpression is Sufficient to Induce β-cell Replication

As previously mentioned, a major goal of diabetes research is to discover pathways sufficient to induce β-cell replication to expand β-cell mass either ex vivo for transplantation (T1D), or in vivo as a T2D therapeutic. Importantly, in the context of

T2D, discovery of such a pathway would have the potential to actually cure a T2D patient through induced expansion of functional β-cell mass, minimizing the need for

15

insulin therapy or other therapeutic intervention. Despite the dogma of β-cells as terminally differentiated and non-dividing cells, accumulated evidence has suggested that β-cells do have the capacity to replicate (albeit at a low rate that decreases with age), especially in response to physiological conditions of increased metabolic demand, such as in pregnancy and obesity (Dor et al. 2004, Georgia and Bhushan 2004, Teta et al. 2005,

Maedler et al. 2006, Meier et al. 2008, Perl et al. 2010, Rieck and Kaestner 2010, Salpeter et al. 2010, Porat et al. 2011, Stolovich-Rain et al. 2012, Saisho et al. 2013, Stolovich-Rain et al. 2015). One of the pathways that has emerged as a modulator of β-cell replication is the Cyclin D/AKT/CDK4 nexus. Interestingly, studies in rodents have shown that the β- cell is one of the few cell-types reliant on CDK4/CDK6 activity for replication and that abolishment of CDK4/CDK6 activity either by genetic deletion or overexpression of the

CDK4/CDK6 inhibitor p16INK4a (CDKN2A) results in β-cell loss and diabetes (Rane et al.

1999, Krishnamurthy et al. 2006). Subsequent studies have shown that β-cell specific enhancement of AKT signaling, via expression of constitutively active AKT or PTEN deletion, protects against T2D by increasing β-cell mass through increases in Cyclin D1

(CCND1), Cyclin D2 (CCND2) and CDK4 activity (Fatrai et al. 2006, Stiles et al. 2006,

Tong et al. 2009, Wang et al. 2010). In addition, GWAS studies demonstrate that SNPs within the CDKN2A gene locus that are associated with T2D correlate with altered gene expression and β-cell replication (Diabetes Genetics Initiative of Broad Institute of

Harvard and MIT et al. 2007, Kong et al. 2018). Taken together, these data highlight the

16

Cyclin D/AKT/CDK4 axis for regulation of β-cell replication and, as a result, protection against diabetes.

Over the past decade, one of the key projects in our laboratory has focused on identification of pathways capable of activating β-cell replication. We have shown that adenovirus-mediated overexpression of either PDX1 or NKX6.1, another homeodomain transcription factor vital for β-cell development, is sufficient to induce β-cell replication in rat and human islets. Importantly, these factors induce cell cycle progression through distinct pathways that are either CDK4-dependent (PDX1) or independent (NKX6.1), while either maintaining (PDX1) or enhancing (NKX6.1) β-cell function (Schisler et al.

2008, Stephens et al. 2012, Hayes et al. 2013, Tessem et al. 2014, Hobson et al. 2015, Hayes et al. 2016, Ray et al. 2016). Furthermore, we have found that PDX1 is potentially activating β-cell replication through both an intrinsic pathway, hypothesized to be mediated in part by upregulation of CCND1 and CCND2, and an extrinsic pathway mediated by a soluble factor or factors (Hayes et al. 2013, Hayes et al. 2016). My work has contributed to identification of both the intrinsic mechanisms and soluble factor(s) through which PDX1 induces replication.

1.5 Recombinant Adenovirus Vectors Permit Genetic Manipulation of Primary Islets

Viral-mediated gene transfer is an essential tool of current biomedical research.

The recombinant adenovirus vector field arose from efforts to determine whether the 36 kb human serotype 5 adenovirus could transform human cells. This involved 17

transfection of sheared adenovirus genomic DNA into human embryonic kidney cells, which yielded the HEK293 cell line (Graham et al. 1977). The transformative adenovirus genome fragment encodes the E1A gene, which is essential for adenovirus production.

Subsequently, it was determined that trans expression of E1A from HEK293 cells is sufficient to permit production of E1A-deleted recombinant adenoviral vectors that are otherwise unable to replicate outside of E1A expressing HEK293 cells (Berkner and

Sharp 1983). Removal of the E3 adenovirus gene, also dispensable for recombinant adenovirus production, permits these vectors to deliver of up to 7 kb of foreign DNA into a target cell (Ghosh-Choudhury et al. 1987). To combat issues of wild-type adenovirus contamination, the plasmid pJM17, encoding an E1A-altered adenovirus genome that exceeds packaging limits, was ultimately created to enable efficient production of recombinant adenoviruses (McGrory et al. 1988).

Our group was the first to demonstrate that cultured pancreatic islets could be efficiently transduced with recombinant serotype 5 adenoviruses (Ad5) constituting the first method for genetic manipulation of intact primary islets (Becker et al. 1994). We also have developed Ad5 vectors that permit use of RNAi in islets (Bain et al. 2004) and enable inducible, β-cell-specific transgene expression (Hayes et al. 2016). Our group, and the islet biology field in general, has subsequently used Ad5 vectors to study the impact of manipulation of specific genes on pancreatic islet cell function, replication, and survival (Becker et al. 1994, Bain et al. 2004, Schisler et al. 2005, Jensen et al. 2006, Joseph

18

et al. 2006, Ronnebaum et al. 2006, Fueger et al. 2008, Ronnebaum et al. 2008a,

Ronnebaum et al. 2008b, Schisler et al. 2008, Odegaard et al. 2010, Stephens et al. 2012,

Hayes et al. 2013, Jensen et al. 2013, Tessem et al. 2014, Hayes et al. 2016, Hayes et al.

2017). Of note, the general physiology and metabolism field has also utilized Ad5 for in vivo manipulation of the liver (Koyama et al. 1997, Gasa et al. 2002, An et al. 2004).

Whereas adenoviral vectors have proven to be an important tool to gain insights into an otherwise difficult model system, virus construction, especially for cell-type specific applications, is still laborious and time-consuming. In addition, the difficulty in engineering new Ad5 vectors interferes with rapid adoption of new technologies and approaches, such as the recent advances in Cas9-mediated epigenetic engineering. Given that Ad5 vectors still remain among the most efficient viral vectors for transgene delivery into islets, these shortcomings restrict our ability to identify potential cures for diabetes.

1.6 Cas9-Based Epigenetic Engineering Permits Interrogation of Disease-Associated Regulatory Elements Within Native Chromatin Contexts

As discussed previously within the context of PDX1 and CDKN2A, the past decade has brought remarkable progress in our ability to uncover disease-associated genomic loci. Surprisingly, a large portion of these genomic elements localize outside of protein coding regions, often in areas of active epigenetic modification, suggestive of a regulatory role (Maurano et al. 2012, Hnisz et al. 2013, Parker et al. 2013, Pasquali et al.

19

2014). These findings create a new challenge of ascribing a function to non-protein coding regulatory regions within their native cellular contexts. While the direct manipulation of native genomic elements is possible using zinc finger proteins

(ZFP)(Beerli et al. 1998) and transcriptional activator-like effectors (TALE) (Miller et al.

2011), the field has been limited by the arduous and low-throughput manner in which these molecules are generated. Thus, interrogation of disease-associated elements has largely been relegated to inferring that the SNP-associated gene is the closest neighbor and utilizing in vitro luciferase activity assays in relevant cell-types to ascertain whether

SNP-induced sequence alterations modulate the function of the corresponding putative regulatory element.

The realm of possibilities for evaluating SNP-encoding regions altered dramatically upon the discovery that the bacterial Cas9-family of endonucleases can be utilized in mammalian cells to edit discrete loci in a highly specific manner (Cho et al. 2013, Cong et al. 2013, Esvelt et al. 2013, Jinek et al. 2013, Mali et al. 2013, Ran et al. 2015). Importantly,

Cas9 endonucleases can be easily retargeted to a new locus simply by substitution of the

~20 bp protospacer motif of the guide RNA (gRNA) (Figure 4); in comparison, ZFP and

TALE technologies require redesign of the entire protein. This unique flexibility has further been harnessed for purposes outside of DNA editing through alanine substitution of crucial catalytic residues within the RuvC (D10A) and HNH (H841A) nuclease domains, creating a nuclease-dead Cas9 (dCas9) that can be used as a scaffold

20

(Jinek et al. 2012, Qi et al. 2013). Applications based on dCas9 have expanded rapidly to enable chromatin visualization (Chen et al. 2013a, Ma et al. 2016, Gu et al. 2018), isolation of a targeted locus and associated proteins (Fujita and Fujii 2013, Liu et al.

2017), chromosome-conformation capture (Liu et al. 2017), transcriptional modulation

(Gilbert et al. 2013, Qi et al. 2013), and epigenome editing/epigenetic engineering (Hilton et al. 2015, Kearns et al. 2015).

Figure 4: Cas9 Enables Sequence Specific Genome Targeting. (A) The RNA- guided endonuclease Cas9 is specifically recruited to target loci upon addition of a complimentary gRNA, causing double-stranded breaks. Cas9 orthologs have different targeting requirements, as demonstrated in this example, Streptococcus pyogenes Cas9 requires the protospacer adjacent sequence to be “NGG”. (B) Nuclease-dead Cas9 (dCas9) still maintains specific genomic recruitment but does not cause DNA cleavage.

21

To this point, recent publications have demonstrated that dCas9-mediated recruitment of modifiers, for instance VP64 or the Krűppel associated box transcriptional repressor domain (KRAB) to a promoter, are sufficient to alter endogenous gene expression (Figures 5A, B, D, E) (Gilbert et al. 2013, Maeder et al. 2013), providing a viable alternative to transgene overexpression and RNAi. Current approaches utilize epigenetic modulators, such as the H3K27 acetyltransferase domain of p300 (p300core) or the H3K4me1/2 demethylase LSD1, to alter the activity of a targeted enhancer element through epigenetic modification (Figures 5A, C, D, F) (Hilton et al. 2015, Kearns et al.

2015, Thakore et al. 2015). However, these approaches are limited by the fact that the relevant epigenetic engineering tools are spread across various cloning and vector platforms (i.e. lentivirus, AAV, expression plasmid) rather than being aggregated in a manner that maximizes their efficient application. Furthermore, fusion of dCas9 to large effectors, such as the 2.5 kb histone demethylase LSD1, results in inserts that are too bulky to be packaged into most viral vectors, thus restricting their use to cell line models

(Kearns et al. 2015). In addition, there are no adenovirus-based epigenetic engineering tools, therefore prohibiting the application of these technologies to examine diabetes- associated elements within their native chromatin context in primary islets.

22

Figure 5: dCas9 Fusion to Transcriptional and Epigenetic Modifiers Enables Locus Specific Targeting. Nuclease-dead Cas9 (dCas9) can be utilized as a scaffold to enable (A-C) activation or (D-F) repression of gene expression through locus-specific recruitment of modifiers. (A) A lowly transcribed gene, Gene X, can be activated by targeting dCas9 fused to (B) a transcriptional activator, i.e. VP64, to the Gene X promoter or (C) the acetyltransferase domain of p300 to a Gene X enhancer. (D) Conversely, a highly transcribed gene, Gene Z, can be repressed via targeting of dCas9 fused to (E) the KRAB transcriptional repression domain to the Gene Z promoter or (F) the histone demethylase LSD1 to a Gene Z enhancer.

1.7 Goals of the Dissertation

Given that adenovirus vectors still remain among the most efficient viral vectors for transgene delivery into islets, the inability to rapidly generate customized, cell-type specific vectors slows progress in gaining new insights into islet biology and potential new therapeutic strategies for diabetes. To address this, I sought to invent a platform that (1) requires minimal molecular biology expertise, (2) is rapid and highly efficient,

(3) is modular in nature to enable a high degree of customization and adaptability, and 23

(4) can be used universally across all vectors types for application to multiple models.

Additionally, while the advent of dCas9-based epigenetic engineering approaches permit the direct interrogation of regulatory elements in their native chromatin context, I sought to also include these capabilities in the new platform to allow for the interrogation of regulatory elements containing SNPs associated with diabetes to be manipulated in primary rat and human islets.

In previous studies we showed that adenovirus-mediated overexpression of the transcription factor PDX1 is sufficient to induce β-cell replication (Hayes et al. 2013,

Hayes et al. 2016), but the pathway is incompletely understood. Recently, we have shown that PDX1 induces expression of a soluble factor or factors capable of inducing β- cell replication. Although others in the lab had performed microarray analysis on intact islets with CMV-driven PDX1 overexpression compared to controls (Hayes et al. 2013), these data did not allow us to discern “activating” versus “response” transcriptional networks. To circumvent this issue, I sought to apply our new modular cloning platform to generate specialized bicistronic vectors that allow for β-cell specific expression of

PDX1 and an eGFP reporter to facilitate isolation. This enabled our group to define transcriptomic changes specifically in PDX1-overexpressing β-cells to identify PDX1- modulated soluble factors.

Although a role for a soluble factor has clearly emerged from foregoing efforts, we have yet to determine if a β-cell intrinsic pathway is also involved. Given that

24

CCND1 and CCND2 are known to be sufficient to promote β-cell replication, we hypothesized that their induction by PDX1 might be due to an intrinsic replication pathway at work. Our initial attempts utilizing microarray analysis were unable to find the pathway upstream of CDK4 that explained PDX1-replication. Therefore, we hypothesized that the intrinsic replication pathway may not be mediated by a single, highly PDX1-responsive gene, but was instead mediated through the subtle modulation of a gene network by miRNAs. Therefore, a component of this dissertation research focused on determining whether any PDX1/miRNA regulatory hubs exist in islets, and if so, if these contribute to PDX1-mediated islet cell replication.

Finally, after the successful creation of our new modular epigenetic engineering system, we set out to validate the utility of these vectors to perform first-of-kind experiments to manipulate an endogenous enhancer in primary rat islet β-cells. We chose to focus on a highly conserved enhancer proposed to regulate PDX1, Area IV.

Beyond being proof-of-concept experiments for our platform, these experiments were designed to determine whether Area IV was a bonafide regulator of PDX1 expression, and if so, to define its contribution. Of note, in humans this enhancer is encompassed by

SNPs that have been associated with fasting glucose levels, a T2D trait. Therefore, if

Area IV is a significant regulator of PDX1, it may shed light on a potential mechanism of disease. Finally, a long-term objective of ours is to determine whether induction of endogenous PDX1 is also capable of inducing β-cell replication, in the same manner as

25

transgene overexpression. Therefore, we felt that data generated from these initial experiments could be applicable to future efforts to develop vectors capable of simultaneous activation of multiple PDX1 regulatory elements.

26

2. Methods

pMVP and pMAGIC A comprehensive list of the 108 plasmids that currently comprise pMVP and pMAGIC are listed in Appendix B. The sequences for oligonucleotides and gBlock double-stranded DNA fragments utilized to create these vectors are recorded in

Appendix C and Appendix D, respectively.

Reagents Cell culture reagents were purchased from ThermoFisher Scientific unless otherwise specified. Chemical reagents were from Sigma-Aldrich unless specified otherwise. Restriction enzymes, NEBuilder HiFi DNA Assembly mix, and T4 DNA ligase were purchased from NEB with the exception of SgrD1 (ThermoFisher) and BarI

(SibEnzyme). Depending on the application, we utilized competent cells purchased from

ThermoFisher (ccdB Survival, Mach1, Stbl3, TOP10, Subcloning Efficiency DH5α) or

NEB (NEB10-beta, NEB 5-alpha subcloning efficiency). Oligonucleotides (Appendix C) and gBlocks (Appendix D) utilized for the creation of pMVP, pMAGIC, and other plasmid/vectors discussed in this dissertation were purchased from IDT. The MultiSite

Gateway Pro Plus kit, pAd/PL-DEST, pEF-DEST51, and pLenti6.1/V5-DEST were purchased from ThermoFisher Scientific. pBlueScript2SK(-) was purchased from

Agilent. PCR templates used for plasmid creation either originated in our lab or were received from the following: DNASU (Seiler et al. 2014) HsCD00082591 (PDX1,

PDX1open); Addgene 68495 (Kiani et al. 2015) (Sa-dCas9, Sa-dCas9-VPR); Addgene 60954

27

(Gilbert et al. 2014) (KRAB, mCherry-WPRE, P2A-mCherry-WPRE); Addgene 49042

(Mendenhall et al. 2013) (LSD1); Addgene 61357 (Hilton et al. 2015) (p300core); Addgene

61591 (Ran et al. 2015) (Sa-Cas9, 3x HA-bGH PolyA, 3x HA-bGH PolyA + hU6-Sa-Cas9 gRNA); Addgene 49386 (Lam et al. 2015) (APEX2); Addgene 47328 (Holland et al. 2012)

(osTir1); Promega pGL4.23 (Firefly Luciferase); ThermoFisher pEF-DEST51 (EF1a promoter), cTnT promoter was a gift from Brent French (Prasad et al. 2011), Addene

55759 (Heo et al. 2006) (ALB promoter), Addgene 61593 (Ran et al. 2015) (TBG promoter).

Cell Culture INS1-derived 832/13 rat insulinoma cells and any derivatives were cultured as previously described (Hohmeier et al. 2000). For Sleeping Beauty-mediated cell line creation, TransIT-LT1 (Mirus) was used for co-transfection of INS1 832/13 with a 1:10 ratio of expression plasmid for the transposase SB100X (Addgene 34879 (Mates et al.

2009)) and the appropriate pMVP-derived SB transposon vector. After 48h, cells were subjected to blasticidin selection and maintained as the polyclones 1353/pc (eGFP-PDX1) and 1367/pc (AID-PDX1). Auxin (Indole-3-acetic acid) was added to 1367/pc cells and harvested as indicated. For induction studies, doxycycline (dox, Sigma), trimethoprim

(TMP, Sigma), or Shld1 (Clontech) were added to cells 24h after adenoviral transduction and harvested at indicated timepoints. Pancreatic islets were isolated from male Wistar rats weighing ~300 g under a protocol approved by the Duke University Institutional

Animal Care and Use Committee and cultured as previously described (Schisler et al. 28

2008). For PNA studies, 2 μM of the indicated PNAs (PNAbio) were added directly into rat islet medium and incubated overnight. Cell images were captured using a Zeiss

Axiovert S100. Mouse C2C12 cells (ATCC) were grown in DMEM (low glucose, 10% heat inactivated FBS [Hyclone], 50ug/ml Gentamycin), trypsinized with 0.05% Trypsin and plated on Collagen I coated multi-well plates (Corning). When cells were 95% confluent, medium was replaced with Differentiation media (DMEM, 2% heat inactivated horse serum [Hyclone], 50ug/ml gentamycin, 100uM Carnitine). Cell images were captured using a Zeiss Axiovert S100. Media was changed daily and cells were transduced with adenovirus on differentiation day 3. Cell images were captured using a

Zeiss Axiovert S100. HEK293 cells were maintained in DMEM high glucose with 10%

FBS. For adenoviral production, 100 units/mL of penicillin and streptomycin were added. For expression plasmid studies, the indicated plasmids were transfected into

HEK293 cells using TransIT-LT1 (Mirus). HepG2 cells (ATCC) were maintained in

DMEM high glucose with 10% FBS, as recommended by the ENCODE project.

Flow Cytometry A single cell suspension of INS1 832/13 cells or 1353/pc cells in FACS buffer (PBS pH 7.4, 1% FBS, 20mM HEPES, 2.5 mM glucose, 100 units/mL penicillin and streptomycin) were evaluated for eGFP expression by live-cell flow cytometry performed on a FACSCanto II (BD Biosciences) and analyzed using the FACSdiva software (BD Biosciences). For FACS isolation of primary β-cells, intact rat islets were dispersed using StemPro Accutase and resuspended in FACS buffer. FACS was 29

performed on a MoFlo Astrios Cell Sorter (Beckman Coulter) by the Duke Flow

Cytometry Core. Cells were sorted into ice-cold rat islet medium and subsequently harvested.

[3H]-Thymidine Assay [3H]‑thymidine (Perkin Elmer) was added to fresh islet medium at a final concentration of 1 μCi/ml to pools of islets for the final 24h of cell culture. Three groups of 20 size-matched islets were hand-picked, washed in PBS, and centrifuged twice at 300

× g for 3 minutes. DNA was precipitated by adding 500 μl of ice-cold 10% trichloroacetic acid for 20 minutes on ice followed by centrifugation at 13,000 rpm at 4°C for 20 minutes. The pellet was solubilized with 80 μl of 0.3 N NaOH. The amount of [3H]- thymidine incorporated into DNA was measured by liquid scintillation counting and normalized to total cellular protein, as determined by BCA assay.

Luciferase Activity Assay A MIR17HG minimal promoter (-92 to +8 bp relative to TSS) was PCR amplified from HEK293 cell gDNA using primer pairs JH50/52 (wild-type) or JH49/52 (PDX1 binding site mutant). The PCR product was digested with SacI/MslI and ligated into

SacI/EcoRV digested pGL4.10 (Promega).

To enable rapid generation of enhancer luciferase constructs we adapted

Promega’s pGL4.23 minimal promoter luciferase vector by ligating a Gateway Cloning

Cassette downstream of the polyadenylation signal (SalI restriction site w/ overhangs filled in using T4 DNA polymerase) in the “enhancer” position. Since this was a blunt

30

end ligation, we were able to select clones in which the Gateway Cassette was inserted in either a “forward” or “reverse” orientation (HO901, HP001), allowing the same enhancer entry plasmid to be used to generate luciferase constructs in either orientation.

Putative MIR17HG enhancers were PCR amplified from HEK293 cell gDNA using the primers listed in Table 1 and then TOPO cloned into pCR8/GW/TOPO to create an entry plasmid. Control_3.1kb (lacZ) was generated by performing a BP reaction with pAd/CMV/V5-GW/lacZ (ThermoFisher) to transfer lacZ into pDONR221. HR503 and

HR606 were used as templates for PCR-mediated site-directed mutagenesis using the listed primers to create HW901 and HX002, respectively. A LR recombination reaction was performed to insert elements into both HO901 and HP001.

To perform luciferase activity assays, INS1 832/13 or HepG2 cells were split into

96-well clear-bottom, white cell culture plates (Corning) at a density of 40,000 cells per well. Plasmids for Firefly luciferase and a Renilla luciferase control (pGL4.54, Promega) were co-transfected at least in triplicate into cells using TransIT-LT1 (Mirus). 24 hours later, luciferase activity was determined using the Dual-Luciferase Reporter Assay

System (Promega) and measured on a SpectraMax L. Firefly activity measurements where normalized to Renilla activity measurements to control for transfection efficiency.

31

Table 1: MIR17HG Enhancer Luciferase Plasmids. PCR primers used to amplify putative MIR17HG enhancers to generate entry plasmids. For Ref->Minor and Minor->Ref elements the site-directed mutagenesis primers are listed. Elements were recombined into HO901 and HP001 to generate luciferase activity plasmids in both orientations.

Entry PCR Primer Luc Plasmid ID Name Plasmid Forward Reverse HO901 HP001 ID Control 3.1kb HI501 N/A N/A HP201 HQ801 A_1.3kb HS601 JH70 JH71 HU601 HW601 B_1.1 kb HR701 JH54 JH55 HT701 HV701 C_2.9 kb HR801 JH56 JH57 HT801 HV801 D_8.6kb HR304 JH26 JH25 HT301 HV301 D_3.1kb HR101 JH31 JH32 HT101 HV101 D_1.2kb HQ903 JH29 JH30 HS901 HU901 E_3.5kb HR901 JH58 JH59 HT901 HV901 F_4.8kb HS001 JH60 JH61 HU001 HW001 F_2.2kb HS101 JH62 JH61 HU101 HW101 G_1.1kb HS201 JH63 JH64 HU201 HW201 H_2.9kb HS305 JH65 JH66 HU301 HW301 I_2.1kb (Ref) HR503 JH194 JH196 HT501 HV501 Minor Allele HR606 JH194 JH196 HT601 HV601 Ref -> Minor HW901 JH212 JH213 HX601 HX801 Minor -> Ref HX002 JH214 JH215 HX701 HX901 J_6.0kb HR403 JH27 JH28 HT401 HV401 J_3.2kb HR205 JH33 JH34 HT201 HV201 J_1.2kb HR003 JH35 JH36 HT001 HV001

Electrophoretic Mobility Shift Assay To prepare INS1 832/13 cell nuclear lysates (Holden and Tacon 2011), cells from a confluent 10-cm dish were scraped on ice and transferred to pre-chilled tubes. Cells were pelleted by centrifugation at 5,000 x g, 4°C for 5 minutes and supernatant discarded. Pellet was resuspended in 250 μl Gough I buffer (10 mM Tris-HCl pH 7.5/

32

0.15 M NaCl, 1.5 mM MgCl2,0.65% NP-40, 1X HALT Protease and Phosphatase Inhibitor

Cocktail [ThermoFisher]), vortexed, and incubated on ice for 30 minutes, until the plasma membrane is dissolved but the nuclear membrane remains intact. The nuclear fraction was then pelleted by centrifugation at 12,000 x g, 4°C for 2 minutes. Supernatant

(cytoplasmic fraction) was discarded and the pellet was resuspended in 75 μl Buffer C

(20 mM HEPES pH 7.9, 25% glycerol, 0.4 M NaCl, 1.5 mM MgCl2, 0.2 mM EDTA pH 8.0,

HALT Protease and Phosphatase Inhibitor Cocktail). Resuspended pellet was agitated and incubated on ice for 2 hours, with agitation every 15 minutes. Nuclear debris was removed by centrifugation at 12,000 x g, 4°C for 10 minutes. Supernatant was transferred to a new pre-chilled tube, mixed with Buffer D (20 mM HEPES pH 7.9, 20% glycerol, 50 mM KCl, 0.2 M EDTA pH 8.0, HALT Protease and Phosphatase Inhibitor

Cocktail), and stored at -80°C. A 20 μl binding reaction was performed by incubation of

7 μg nuclear lysates and 100 nM fluorescently labeled oligonucleotide probes (IDT) in addition to 2 μl 10x Odyssey Infrared EMSA kit (Li-Cor) Binding buffer (100 mM Tris,

500 mM KCl, 10 mM DTT pH 7.5); 2 μl of 25 mM DTT/2.5% Tween 20; 1 μl of 1 μg/μl

Poly (dl.dC); and water, followed by incubation at room temperature for 20-30 minutes.

As indicated, 20 μM (200x) of unlabeled competitor probe or 2 μg of antibody (PDX1,

SC-14664x, Santa Crux or IgG, SC-2028, Santa Cruz) were also included in the reaction.

After incubation, reaction was mixed with Orange loading dye (Li-Cor) and resolved on

Novex 6% DNA Retardation TBE gels (ThermoFisher) and imaged using a Li-Cor

33

Odyssey CLx. EMSA conditions were optimized using a previously described probe for the PDX1 binding site within the insulin promoter (Le Lay et al. 2004). Two different mutations of the miR17HG PDX1 binding site were tested and showed similar results.

Table 2: EMSA Probes. Oligonucleotides probes based on human sequences utilized for EMSAs. Labeled versions of JH152, JH153, JH154, and JH155 were also ordered and synthesized with end-labeled IRDYE 700.

ID Sequence Description

INS A1 JH152 CCCAGGCCCTAATGGGCCAGGCGG (PDX1 validated)

INS A1 JH153 CCGCCTGGCCCATTAGGGCCTGGG (PDX1 validated)

MIR17HG promoter JH154 TTTATGCTAATGAGGGAGTGGG PDX1 motif

MIR17HG promoter JH155 CCCACTCCCTCATTAGCATAAA PDX1 motif

MIR17HG promoter JH156 TTTATGCGCCGGAGGGAGTGGG mutated PDX1 motif

MIR17HG promoter JH157 CCCACTCCCTCCGGCGCATAAA mutated PDX1 motif

MIR17HG promoter JH180 TTTATGCTGGGAAGGGAGTGGG mutated PDX1 motif

MIR17HG promoter JH181 CCCACTCCCTTCCCAGCATAAA mutated PDX1 motif

Gateway Cloning To generate pENTR plasmids, sequences of interest flanked by the appropriate

Gateway attB sites, as listed in the MultiSite Gateway Pro manual, were either

34

synthesized as gBlocks or PCR amplified by Platinum Taq HiFi DNA polymerase

(ThermoFisher) and subsequently purified using the QIAquick PCR Purification kit

(QIAGEN). BP reactions containing 15-75 ng of purified PCR product, 75 ng of the appropriate MultiSite Gateway Pro pDONR221 plasmid, TE added to 4 μl, and 1 μl of

BP Clonase II (ThermoFisher) were incubated at room temperature for 1h-16h. BP reactions were stopped by incubation with Proteinase K (ThermoFisher) and used to transform Mach1, TOP10, or NEB10-beta competent cells. Single colonies were selected, cultured overnight at 37°C, and plasmid DNA was isolated using the QIAprep Miniprep kit (QIAGEN). Entry plasmids were verified by restriction digest and inserts were sequence verified (Genewiz).

To generate pMVP- and pMAGIC-derived vectors, LR recombination reactions containing 5 fmol of each pENTR plasmid, 10 fmol of pDEST vector, TE added to 4 μl, and 1 μl LR Clonase II Plus (ThermoFisher), were incubated for > 16h at 25°C. LR reactions were stopped by incubation with Proteinase K (ThermoFisher), and used to transform TOP10, NEB10-beta, or Stbl3 competent cells. Transformations were streaked on LB plates and incubated at 30°C-37°C overnight, taking care to avoid colony overgrowth. Single colonies were selected, cultured for 12h-16h at 28°C-37°C, and plasmid DNA was isolated using the QIAprep Miniprep kit (QIAGEN). Adenoviral vectors were verified using diagnostic restriction digests for HindIII, BsrGI, and PacI, other vectors were verified with restriction enzymes relevant to the underlying sequence

35

as determined by in silico recombination using DNASTAR Lasergene (DNASTAR). In our experience, a 5 mL bacterial culture yielded a sufficient quantity of vector for adenovirus production or most other downstream applications.

To generate PDX1 overexpression vectors, the indicated pMVP elements were recombined into either pAd/PL-DEST (adenoviruses), pMVPBS-DEST (expression plasmids), or pMVP/SB/Blast-DEST (Sleeping Beauty transposon). Sa-dCas9/VPR and

Sa-dCas9/p300core expression plasmids were made in pEF-DEST51 by recombination of: a pENTR L1-L4 plasmid containing eGFP-P2A (II601); Sa-dCas9/VPR or Sa-dCas9/p300core, and 3x HA-bGH polyA. The TRE-mCherry plasmid was made by recombination of the

TRE promoter, mCherry, and hGH polyA signal into pMVPBS-DEST. pMAGIC adenoviruses were generated by recombination of the RIP, Sa-dCas9/LSD1, and either 3x

HA-bGH polyA+gRNAArea IV or 3x HA-bGH polyA into pAd/PL-DEST or pMVP/Ad-

DEST, respectively. eGFP expression plasmids were generated by recombination of the

CMV promoter, eGFPopen, and 3x HA-bGH polyA into pAd/PL-DEST or pMVP/Ad/RGD-DEST.

gRNA Cloning Sa-Cas9 protospacer sequences targeting Area IV and the TRE promoter were selected and screened for potential off-target activity using the resources at www.Deskgen.com (Hough et al. 2017). Oligonucleotides containing protospacer sequences with appropriate overhangs were annealed and subsequently ligated for 20 minutes at room temperature with T4 DNA ligase into a BsaI-digested pENTR L3-L2 36

entry plasmid containing 3x HA-bGH polyA and a human U6-driven Sa-Cas9 gRNA expression cassette derived from Addgene 61591(Ran et al. 2015). Ligation reactions were re-digested with BsaI for 15 minutes to reduce background and transformed into

Subcloning Efficiency DH5α competent cells, resulting plasmids were sequence verified before use (Genewiz).

Table 3: gRNA Protospacer Motif Oligonucleotide Sequences. Oligonucleotide sequences for gRNA protospacer motifs targeting Area IV or the TRE promoter. Target- specific sequence is underlined, required “G” for efficient transcription is highlighted in bold. Oligonucleotides were annealed before ligation into a gRNA expression plasmid.

ID Sequence Description JH416 CACCGTTAACAACATCAGGCTGACGG Area IV sense JH417 AAACCCGTCAGCCTGATGTTGTTAAC Area IV antisense JH514 CACCGATCAGTGATAGAGAACGTATG TRE sense JH515 AAACCATACGTTCTCTATCACTGATC TRE antisense

pMVP/Ad-DEST Vectors Custom adenovirus vectors were created as described above. NEBuilder HiFi

DNA assembly was utilized to replace the SrfI/BarI fragment of pAd/PL-DEST with a gBlock containing an E3 deletion mirroring pAdEasy-1 as well as a unique SgrDI site to create pMVP/Ad-DEST. For fiber modification, published sequences for modifications were synthesized as gBlocks and assembled into pBS/Fiber digested with BglII/BstBI

(RGD, pK, RGD pK) or Age/BstBI (Ad5/35). Modified pBS/Fiber plasmids were digested with BsmBI and inserted into BarI/BstBI digested pMVP/Ad-DEST.

37

pMVP-DEST Vectors A Gateway cloning cassette (ThermoFisher) was ligated into EcoRV digested pBlueScript2SK(-) (creating pMVPBS-DEST) and we subsequently removed all non- essential plasmid components by replacing the XhoI/PsiI and SapI/SacI fragments with annealed oligonucleotides to create pMVP-DEST. We then utilized NEBuilder HiFi DNA

Assembly to introduce an SV40 promoter, TK polyA signal, and selection markers for blasticidin (pMVP/Blast-DEST), puromycin (pMVP/Puro-DEST), eGFP (pMVP/eGFP-

DEST), or mCherry (pMVP/mCherry-DEST) into XhoI digested pMVP-DEST.

pMVP/SB-DEST Vectors NEBuilder HiFi DNA assembly was used to insert gBlocks containing SB ITR elements designed from Addgene 26557 (Courtesy of Perry Hackett) into the SpeI and

BsaXI restriction sites located on either side of the pMVP-DEST Gateway cloning cassette to create pMVP/SB-DEST. In the same manner as for expression vectors, the selection markers blasticidin (pMVP/SB/Blast-DEST), puromycin (pMVP/SB/Puro-

DEST), eGFP (pMVP/SB/eGFP-DEST), mCherry (pMVP/SB/mCherry-DEST) were inserted in the XhoI restriction site located between the attR2 Gateway site and adjacent

3’ SB ITR of pMVP/SB-DEST.

pMVP/Lenti-DEST Vectors Using pLenti6.3/V5-DEST as a template, a third-generation lentivirus vector devoid of a selection marker was generated by sequential replacement of the ClaI/SpeI fragment (CMV promoter) and XhoI/BlpI fragment (V5-WPRE-Blasticidin expression cassette) with annealed oligonucleotides to create pMVP/Lenti-DEST. A lentivirus vector 38

conferring resistance to blasticidin (pMVP/Lenti/Blast-DEST) was generated by sequential replacement of the ClaI/SpeI fragment (CMV promoter) and the XhoI/SpeI fragment (V5 epitope tag and WPRE element) with annealed oligonucleotides. Next,

NEBuilder HiFi DNA assembly was used replace the PmlI/BlpI fragment containing the blasticidin resistance gene with a neomycin resistance gene (pMVP/Lenti/Neo-DEST.)

Ad-miR17~92 The miR17~92 cluster was PCR amplified from HEK293 gDNA using primers

JH6 and JH7. PCR product was digested with BglII/SpeI and subsequently ligated into

BglII/SpeI digested GS001, an AdShuttle plasmid for β-cell specific, Dox-inducible expression created by Dr. Tom Becker (Hayes et al. 2016). Dr. Becker used the resulting plasmid, GU601, and the GS001 control plasmid to generate a pAdEasy-1 adenovirus as previously described (He et al. 1998, Hayes et al. 2016).

Recombinant Adenovirus Production To produce recombinant adenoviruses, 1 μg of PacI digested vector was transfected with 1.5 μl of TransIT-LT1 (Mirus) into one well of a 24-well plate containing

HEK293 cells that were 80% confluent. After primary lysis (7-10 days), crude adenovirus lysates were subject to freeze/thaw and further propagated in a confluent T25 flask of

HEK293 cells to create a secondary lysate. For purification, six confluent 15 cm plates of

HEK293 cells were transduced with secondary lysate and harvested after 40h-60h, at a point when 50%-75% of cells were floating. Floating and attached cells were collected by gentle scraping, pelleted by centrifugation, resuspended in freeze/thaw buffer (FT

39

buffer; 10 mM Tris/HCl pH 8.0; 1 mM MgCl2), and subsequently lysed by addition of

NP40 to a final concentration of 0.5%. Lysates were cleared by centrifugation (10 minutes at 10,000 x g and 4°C), the supernatant was loaded on a pre-formed CsCl gradient (1.45 g/mL, 1.33 g/mL, 1.2 g/mL prepared in FT buffer), and centrifuged (1-3h at

191,000 x g and 4°C) in a S100-AT6 rotor (ThermoFisher). The band containing mature adenoviral particles was isolated, desalted with Zeba Desalt spin columns (Pierce), diluted 1:1 with sterile glycerol, filter sterilized, aliquoted, and stored at -80°C. Viral particles (vp) were determined by the A260 method (Maizel et al. 1968) before addition of glycerol. Working stocks of purified adenoviruses were stored at -20°C for up to 6 months after first use. For experiments with crude viral lysates, the minimum dose sufficient for maximal transduction efficiency, determined by visual observation of eGFP expression, was used. For purified adenovirus experiments, cells were incubated with

~5x108 vp/cm2 overnight.

All recombinant adenoviruses were tested for the presence of replication competent adenoviruses (RCA) that can arise from recombination with the Ad5 E1A gene present in the HEK293 genome (Graham et al. 1977, Lavine et al. 2010). This was performed using a new qRT-PCR assay we recently developed. In brief, crude and purified adenoviruses were first subjected to DNase digest (QIAGEN) to remove any traces of contaminating HEK293 gDNA. Next, samples were protease treated and recombinant adenovirus gDNA isolated using QiaAmp MinElute Virus Spin Kit

40

(QIAGEN). Each sample is then analyzed by qRT-PCR using primers for the Ad5 E1A gene (present only in RCA), Ad5 Hexon (L3) gene (present in all Ad), and human PPIB gene (signifies HEK293 gDNA contamination). These samples are analyzed alongside a standard curve generated using a plasmid (Newgard Plasmid ID JI401) that contains the

PCR amplicons for Ad5 E1A and Ad5 Hexon. Dividing E1A values by Hexon values yields the concentration of E1A present in the sample. Adenovirus preparations with a concentration < 1 ppm were considered to be RCA-negative.

Table 4: RCA Detection Assay qRT-PCR Primers.

Primer Sequence E1A-F TATGCCAAACCTTGTACCGGAGGT E1A-R CCGGGGTGCTCCACATAATCT Hexon-F TTGGCGCATCCCATTCTCCAGTAA Hexon-R ATAAAGAAGGGTGGGCTCGTCCAT hPPIB-F CGCTTCCCCGATGAGAACTTC hPPIB-R AGGCTGTCTTGACTGTCGTGATG

Measurement of RNA Levels Total RNA was isolated using the RNeasy Mini (INS1 832/13) or Micro (islets) kit

(QIAGEN), and cDNA was synthesized using iScript (Bio-Rad). Total RNA for miRNA quantification was were isolated using miRNeasy Micro kit (QIAGEN), and miRNA cDNAs were synthesized using TaqMan MicroRNA Reverse Transcription kit

(ThermoFisher) Quantitative real-time PCR (qRT-PCR) assays were performed using the

Viia7 or QuantStudio 6 FLEX systems and software (ThermoFisher).

41

Table 5: TaqMan Probe qRT-PCR Primers.

Gene TaqMan ID Gene TaqMan ID ARX Rn01419077_m1 MYC Rn00561507_m1 BNIP3L Rn00821653_g1 NEUROD1 Rn0128117_m1 CCND1 Rn00432360_m1 NKX6.1 Rn01450076_m1 CCND2 Rn01492401_m1 PDX1 Rn00755591_m1 CDKN1A Rn00589996_m1 PPIB Rn03302274_m1 CNOT6L Rn01443816_m1 PPY Rn01496899_g1 DSTYK Rn01490357_m1 PTEN Rn00477208_m1 ELAVL4 Rn01416883_m1 RBP4 Rn01451318_m1 GCG Rn00562293_m1 RORA Rn01173769_m1 GCG Rn00562293_m1 SLC2A2 Rn00563565_m1 GHRL Rn00572319_m1 SQSTM1 Rn00709977_m1 INS1 Rn02121433_g1 SST Rn00561967_m1 INS2 Rn01774648_g1 TP53INP1 Rn00710369_m1 MAFB Rn00709456_m1 TSPAN8 Rn00591756_m1 miR-106b 000442 TXNIP Rn01533891_g1 miR-17 002308 UCN3 Rn02091611_s1 MIR17HG Rn03465733_pri VEGFA Rn01511602_m1 miR-18a 002422 VLDLR Rn01498163_m1 miR-19a 000395 WFS1 Rn00582735_m1 miR-19b 000396 ZFP367 Rn01759530_m1 miR-20a 000580 ZFYVE26 Rn01469096_m1 miR-92a 000430 ZNFX1 Rn01423205_m1

Immunoblot analysis Lysates used for immunoblotting were prepared in ice-cold RIPA Buffer containing HALT protease inhibitors (ThermoFisher). Clarified cell lysates were resolved on 4-15% mini-PROTEAN TGX pre-cast gels (Bio-Rad), and transferred to nitrocellulose membranes using a TransBlot Turbo (Bio-Rad). Membranes were blocked with Odyssey TBS Blocking Buffer (Li-Cor) and then probed with the appropriate

42

antibodies. Primary antibodies were diluted in TBS + 1% PVP as follows: anti-PDX1, ab47267, Abcam, 1:5000; anti-Tubulin, T5326, Sigma, 1:20000; anti-eGFP, ab290, Abcam,

1:5000; anti-HA, ab9110, Abcam, 1:5000; anti-OLLAS; NBP1-06713, Novus, 1:1000; anti-

FLAG, F3165-.2MG, Sigma, 1:400; anti-His, 12698, CST, 1:1000; anti-myc, 2278, CST,

1:1000. Membranes were and incubated with appropriate antibodies overnight at 4°C.

Secondary antibodies (Li-Cor) were diluted 1:10000 in Odyssey TBS Blocking Buffer (Li-

Cor) and incubated with membranes for 1h at room temperature. Immunoblots were developed using a Li-Cor Odyssey CLx.

Chromatin Immunoprecipitation INS1 832/13 (1x107) cells were crosslinked with 1% formaldehyde for 8 min, and the reaction was quenched by the addition of glycine at a final concentration of 125 mM for 5 min. Cell pellets were washed twice with PBS and resuspended in 1 ml of cold lysis buffer A (50 mM HEPES pH 7.5, 140 mM NaCl, 1 mM EDTA, 10% glycerol, 0.5%

IGEPAL CA-630, 0.25% Triton X-100, 1x protease inhibitor cocktail (PI; Roche)). After 10 min on ice, cells were pelleted and resuspended in 1 ml of cold lysis buffer B (10 mM

Tris-HCl pH 8, 200 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 1x PI) for an additional 10 min. Cells were sonicated in cold lysis buffer C (10 mM Tris-HCl pH 8, 100 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 0.5% N-lauroylsarcosine, 1x PI) using Covaris S220 ultrasonicator to an average fragment size of 500 bp. Immunoprecipitation was performed with rabbit anti-HA antibody (C29F4, CST) or mouse anti-H3K27ac (#39685,

Active Motif) overnight at 4°C. Protein G Dynabeads (ThermoFisher) were added and 43

rocked at 4°C for an additional 4 hr. Beads were washed with cold buffer TSE 150 (10 mM Tris-HCl pH 8, 150 mM NaCl, 2 mM EDTA, 1% Triton-X, 0.1% SDS), TSE 500 (10 mM Tris-HCl pH 8, 500 mM NaCl, 2 mM EDTA, 1% Triton-X, 0.1% SDS) and wash buffer (10 mM Tris-HCl pH 8, 1 mM EDTA, 1% 0.25 M LiCl, 0.5% NP-40), and twice in cold TE (20 mM Tris-HCl pH 8, 2 mM EDTA), before being eluted in elution buffer (1%

SDS, 0.1 M NaHCO3) at 65°C. Following RNAse A and proteinase K treatment, input and ChIP DNA were purified and analyzed using qRT-PCR. Amplification values were normalized to input.

Table 6: ChIP qRT-PCR Primers.

ID Sequence Description JH178 GCACAACAAAGAAACGGGCT Chr15 (F) - GPC5 locus JH179 ATTGTCAGCTCCACCAGCTC Chr15 (R) - GPC5 locus JH418 CAGGGCTACTCTCCCTTGGA Area IV #2 (F) JH419 CCACGTTAATGGTCCTGCCT Area IV #2 (R) JH420 AGGCAGGACCATTAACGTGG Area IV #3 (F) JH421 ACATGCTCATGTGGGCAGAAT Area IV #3 (R) JH422 ACCCCCAACCCTGACATAGAT Area IV #1 (F) JH423 CCAAGGGAGAGTAGCCCTGT Area IV #1 (R) JH424 GCCAGCATTCGCTTTCCAAG Area IV #4 (F) JH425 GTGCCCTTTACTCAGGAGTGG Area IV #4 (R) JH466 GGCCTTAGAGGGACCGTTTC ChrX (F) – ARX locus JH467 AAGAGTCCCACCCCTCAAGT ChrX (R) – ARX locus

44

Statistical Analysis Statistical significance was determined by unpaired, equal variance Student’s t- test. A p-value of < 0.05 was considered significant. All data is expressed as mean ± standard error of the mean (S.E.M.) of at least 3 independent experiments.

45

3. A PDX1/miR17~92/PTEN Axis Regulates β-Cell Replication 3.1 Introduction

In the field of pancreatic islet biology, key objectives are to identify pathways capable of increasing β-cell function, survival, or replication. To this end, our lab has found that adenovirus-mediated overexpression of the β-cell homeodomain transcription factors PDX1 or NKX6.1 activates islet cell replication (Schisler et al. 2008,

Hayes et al. 2013) while maintaining (PDX1) or enhancing (NKX6.1) β-cell function

(Schisler et al. 2008, Stephens et al. 2012). Furthermore, we have discovered that PDX1-, but not NKX6.1-, mediated replication is dependent on CDK4, demonstrating that these factors induce replication through independent pathways (Hayes et al. 2013).

Preliminary data from our lab have also shown that overexpression of PDX1 is sufficient to reduce ER stress-induced apoptosis in primary rat islets and the INS1 832/13 β-cell line. PDX1 mutations are sufficient to induce a monogenic form of diabetes in humans

(Stoffers et al. 1997a), and PDX1 haploinsufficient mice have an increase in β-cell apoptosis when challenged with a high fat diet or ER stress induced by injection of tunicamycin (Johnson et al. 2003, Sachdeva et al. 2009). Also, mice with decreased PDX1 expression during weaning have significantly lower levels of β-cell replication (Spaeth et al. 2017). Whereas our lab and Dr. Jeffrey Tessem, a former postdoctoral fellow in our lab, have made significant progress in defining the pathway(s) through which NKX6.1 drives β-cell replication (Schisler et al. 2008, Tessem et al. 2014, Hobson et al. 2015, Ray 46

et al. 2016), the pathways upstream of CDK4 that explain the effects of PDX1 overexpression remain unresolved (Hayes et al. 2013, Hayes et al. 2016). We hypothesize this is due in part to the fact that unlike NKX6.1, which appears to activate -cell replication strictly in a cell autonomous manner, PDX1 activates CDK4-dependent replication via both induction of soluble factors that activate both - and -cell proliferation in addition to a cell autonomous pathway mediated by upregulation of

CCND1 and CCND2.

The involvement of microRNAs (miRNA) in the development and maintenance of the β-cell in health and disease has been an area of interest since the seminal 2004 paper implicating the islet-enriched miR-375 as a regulator of β-cell function (Poy et al.

2004). Subsequent studies have utilized mouse models deficient for dicer to demonstrate that miRNAs are required for islet development (Lynn et al. 2007), maintenance of functional β-cell mass (Melkman-Zehavi et al. 2011, Mandelbaum et al. 2012), and β-cell identity (Martinez-Sanchez et al. 2015). In agreement with this, others have found that

Argonaut2, a required member of the RNA-induced silencing complex, mediates the compensatory β-cell replication observed in models of T2D (Tattikota et al. 2014).

Subsequent studies have identified miRNAs capable of regulating β-cell function (miR-

7, miR-204, miR-17, miR-29a/b), replication (mir-7, miR17~92 cluster), and survival (mir-

200, miR-29a/b, miR17~92 cluster) and demonstrated that dysregulation of these and other miRNAs contributes to development of diabetes (Melkman-Zehavi et al. 2011,

47

Pullen et al. 2011, Jacovetti et al. 2012, Wang et al. 2013, Xu et al. 2013, Kameswaran et al.

2014, Latreille et al. 2014, Tattikota et al. 2014, Belgardt et al. 2015, Jacovetti et al. 2015,

Chen et al. 2016). In spite of these efforts to identify miRNA regulators of β-cell biology, key information is lacking about how these miRNAs are actually regulated within the β- cell to modify transcriptional networks in concert with master transcription factors such as PDX1 and NKX6.1 to maintain β-cell identity.

In these studies, we sought to determine whether any PDX1/miRNA regulatory hubs exist in islets and contribute to PDX1-mediated replication. In collaboration with the laboratory of Praveen Sethupathy at the University of North Carolina at Chapel Hill, we identified a candidate PDX1/miR-17 regulatory hub in which PDX1 binds the promoter of the miR17~92 miRNA cluster and target sites for the miR-17 family members are significantly enriched in the 3’-UTRs of PDX1-bound genes. Importantly, gel-shift and luciferase assays demonstrated that PDX1 can directly regulate the miR17~92 promoter through a conserved PDX1 DNA binding motif. In support of this, we also discovered that the components of this miRNA cluster are all significantly induced by ~2-fold in response to PDX1 overexpression in primary rat islets.

Unfortunately, due to the shortcomings of the current tools available for miRNA inhibition in islets, we were unsuccessful in directly testing the contribution of miR17~92 to PDX1-mediated replication. However, in support of the involvement of miR17~92 we did find a correlation between PDX1 overexpression and repression of the well

48

characterized target of the miR17~92 cluster, PTEN. Furthermore, we show that β-cell replication mediated by PDX1, and not NKX6.1, is enhanced by small molecule inhibition of PTEN, supportive of the presence of a PDX1/miR17~92/PTEN regulatory node

3.2 Results

3.2.1 miR-17 Target Sites are Enriched in the PDX1 Gene Network

As a first approach to identify candidate miRNA regulators of the PDX1 network, we developed an in silico screening strategy based on the microRNA.org database. Here we took advantage of a microarray study performed before I joined the laboratory in which rat islets were treated with a recombinant adenovirus overexpressing PDX1 under control of the CMV promoter (Ad-CMV-PDX1) or a control virus expressing the bacterial -galactosidase gene (Ad-CMV-GAL). Putative miRNA binding sites were identified in in genes modulated at least 2-fold by PDX1 overexpression (583 increased genes, 93 decreased genes) (Hayes et al. 2013). To reduce false-positives, stringent screening parameters were employed, based on the score of the previously reported miR-375/MTPN interaction (Poy et al. 2004), which required that the miRNA be conserved, contain at least 6-mer complementarity, a mirSVR ≤ -0.1, and a

PhastCost ≥ 0.5. Additionally, to ensure that any resulting candidate miRNAs were not solely due to transcriptional changes coinciding with β-cell replication, we repeated the screen using a different microarray data set with genes modulated at least 2-fold by

49

NKX6.1 overexpression (545 increased genes, 371 decreased genes) (Tessem, 2014).

These data were then sorted by the number of target sites in each miRNA, and the top 20 miRNAs from each group were identified (Table 7). Of note, this approach identified the miR-17 family, a widely studied group of miRNAs known to enhance replication and survival in many cell-types (He et al. 2005, Olive et al. 2009), including the β-cell (Chen et al. 2016), as candidate regulators specifically for genes that decreased in expression in response to PDX1 overexpression.

50

Table 7: Most Abundant miRNA Targets Identified in the PDX1 and NKX6.1 Transcription Networks. Rank order list of the top 20 miRNA target sites identified in the microRNA.org analysis performed on genes significantly increased or decreased by overexpression of either PDX1 or NKX6.1 in rat islets for 48 hours. miRNAs representing the miR-17 family are highlighted.

PDX1 NKX6.1 Decreased Genes Increased Genes Decreased Genes Increased Genes hsa-miR-590-3p hsa-miR-590-3p hsa-miR-590-3p hsa-miR-590-3p hsa-miR-340 hsa-miR-186 hsa-miR-186 hsa-miR-485-5p hsa-miR-186 hsa-miR-495 hsa-miR-495 hsa-miR-186 hsa-miR-181b hsa-miR-340 hsa-miR-374a hsa-miR-181d hsa-miR-181d hsa-miR-374a hsa-miR-340 hsa-miR-330-5p hsa-miR-185 hsa-miR-410 hsa-miR-203 hsa-miR-495 hsa-miR-495 hsa-miR-203 hsa-miR-300 hsa-miR-326 hsa-miR-539 hsa-miR-494 hsa-miR-381 hsa-miR-539 hsa-miR-181c hsa-miR-181d hsa-miR-181d hsa-miR-181b hsa-miR-20b hsa-miR-181b hsa-miR-410 hsa-miR-214 hsa-miR-381 hsa-miR-200b hsa-miR-181b hsa-miR-340 hsa-miR-93 hsa-miR-429 hsa-miR-129-5p hsa-miR-203 hsa-miR-211 hsa-miR-23b hsa-miR-494 hsa-miR-494 hsa-miR-300 hsa-miR-200c hsa-miR-181a hsa-miR-204 hsa-miR-106b hsa-miR-23a hsa-miR-539 hsa-miR-211 hsa-miR-20a hsa-miR-181a hsa-miR-23b hsa-miR-185 hsa-miR-29b hsa-miR-539 hsa-miR-23a hsa-miR-15b hsa-miR-30c hsa-miR-144 hsa-miR-144 hsa-miR-320a hsa-miR-374a hsa-miR-181c hsa-miR-30e hsa-miR-320b hsa-miR-106a hsa-miR -129-5p hsa-miR -181c hsa-miR-181a hsa-miR-132 hsa-miR -374a hsa-miR-17 hsa-miR-181a hsa-miR-214 hsa-miR-30a hsa-miR-30b hsa-miR-519d

51

We sought to validate the results with a more statistically rigorous approach that reflected the entire PDX1 gene network instead of only those genes significantly modulated by PDX1 overexpression. To achieve this, we collaborated with the

Sethupathy lab, which had developed a methodology (miRHub) (Baran-Gale et al. 2013) that utilizes Monte Carlo-based statistical simulation analysis to identify miRNAs that target a gene network, while also determining if the network contains more miRNA/gene interactions than expected by chance. To focus the analysis on

PDX1/miRNA hubs, we limited the scope to include only miRNAs with promoters bound by PDX1. We provided the Sethupathy lab with previously published human islet PDX1 ChIPseq data (Khoo et al. 2012) that were reanalyzed using MACS (p < 0.05) and further refined to include only consensus peaks present in 2 of the 3 independent human islet samples. RefSeq genes containing a PDX1 consensus peak within -5 kb of the transcription start site (TSS) were identified as PDX1-bound genes, representing a total of 2,388 genes and hereinafter referred to as the PDX1 gene network. Epigenetic- based annotations of human islet miRNA promoters (Sethupathy, unpublished) coinciding with PDX1 consensus peaks (94 miRNAs) were identified as PDX1-bound miRNAs. Surprisingly, miRHub analysis revealed the miR-17 family of miRNAs as the sole PDX1-bound miRNA family with target sites significantly overrepresented in the

PDX1 gene network (Figure 6), corroborating our earlier analysis. Within the PDX1 gene network, miR17 is predicted to target 168 genes (Table 8). For our purposes, the most

52

notable of these genes are CCND1, a gene we previously identified to be upregulated by

PDX1 overexpression (Hayes et al. 2013, Hayes et al. 2016), PTEN, a gene previously implicated in regulation of CDK4-dependent β-cell replication (Fatrai et al. 2006, Wang et al. 2010), and BCL2L11, a pro-apoptotic gene also known as Bim (McKenzie et al.

2010). These data implicate PDX1-mediated regulation of the miR-17 family of miRNAs as a potential feedback mechanism on the PDX1 gene network serving to either amplify or dampen the magnitude of PDX1-mediated transcriptional changes.

Figure 6: PDX1-Bound miRNAs Target the PDX1 Gene Network. miRHUB analysis of human islet PDX1-bound miRNAs and the human islet PDX1 gene network. Each data point represents a PDX1-bound miRNA, and the y-axis displays the negative Log10 of the adjusted p-value of the predicted miRNA targeting score among genes bound by PDX. The red line denotes the significant threshold (adjusted p = 0.05). (Jeanette Baran-Gale and Praveen Sethupathy, unpublished)

53

Table 8: PDX1 Gene Network Components Targeted by miR-17. PDX1-bound genes in human islets identified by miRHub to be targets of the miR-17 family of miRNAs. Genes previously implicated in regulating islet-cell function, survival, and replication are highlighted. (Jeanette Baran-Gale and Praveen Sethupathy, unpublished)

PDX1-bound genes targeted by the miR-17 family ABCG4 CABLES1 EIF2S1 IQSEC1 MARS PHF6 SERP1 TNRC6B ABHD5 CAMK2N1 EIF4A2 IRF2BP2 MBD5 PHTF2 SESN2 TRIM66 ACIN1 CASC4 EPB41 IRF2BPL MFAP3L PKD2 SHANK2 TRIP10 AFF4 CCDC50 EPB41L5 ISM2 MIDN PKN2 SHOC2 TRPC4 ALG13 CCND1 FAM117A ITGB8 MYNN PLEKHM3 SLC25A27 TSHZ2 ANKIB1 CCNG2 FAM134A KIAA0232 MYT1L POU2F1 SLC25A36 TSPAN9 ANKRD12 CD69 FBXO21 KIAA1147 NEK9 PPP6R3 SLC25A44 U2SURP AP2B1 CEP97 FCHO2 KIF3B NFAT5 PTBP1 SLC39A9 UBAP1 ASF1A CFL2 FGD4 KLF11 NPAS2 PTEN SLC8A1 UBC ASH1L CIC FJX1 KLHL20 NPAS3 PTPN12 SLITRK3 UBFD1 ATG16L1 COL19A1 FOXJ3 LAMA3 NUCKS1 PXDN SMAD7 UNK ATP2B2 CORO2B FZD1 LAMC1 OCRL RAB5B SOCS7 USP3 ATXN1L CPEB3 FZD3 LASP1 OTUD3 RACGAP1 SOX4 VLDLR B3GNT1 CROT GOSR1 LCOR PAFAH1B2 RAPGEF2 SRCIN1 VPS25 B4GALT6 CRY2 GPR158 LDLR PANX2 RARB ST3GAL1 WEE1 BCL2L11 CSNK1G1 GPR6 LSAMP PARD6B RASD1 ST8SIA3 WNK1 BHLHE41 CSRNP3 HDAC4 LYPD6 PCBP2 RHOQ SYBU YPEL2 BMPR2 CYLD HLF MAB21L1 PCNP RSRC2 SYNE1 ZDHHC1 BTBD7 DLG2 HMBOX1 MAGI3 PDGFRA SCD5 TBL1X ZDHHC21 C11orf30 DYNC1LI2 IGFBP7 MAPK4 PDK4 SCN8A TBX3 ZFP91 C16orf72 E2F3 INO80D MAPRE3 PFKFB2 SEMA4B TLE4 ZFYVE9

3.2.2 PDX1 Directly Regulates the Human MIR17HG Promoter

The miR-17 family members exist in 3 paralogous miRNA clusters, miR17~92, miR106b~25, and miR106a~363 (Ventura et al. 2008); however, only the promoter for the miR17~92 host gene (MIR17HG) was found to be bound by PDX1 in human islets. Upon examination of the DNA sequence underlying the site of PDX1 enrichment, we identified a conserved PDX1 DNA binding motif (Pasquali et al. 2014) (CTAATGAGGG)

54

containing the core TAAT homeodomain motif (Figure 7A). In agreement with the human islet data, reanalysis of published PDX1 ChIPseq data in embryonic stem cell- derived pancreatic endoderm cells (Wang et al. 2015) also identified significant PDX1 enrichment corresponding with the MIR17HG promoter and the putative PDX1 binding motif (Figure 7A). To confirm that PDX1 can directly bind to this motif, electrophoretic mobility shift assays (EMSA) were performed with fluorescently labeled oligonucleotides corresponding to the 22 bps encompassing the putative PDX1 binding motif and a nuclear extract prepared from the INS1 832/13 rat insulinoma cell line that has robust PDX1 expression. Incubation of the labeled probe with nuclear extract revealed a clear shift of 2 distinct bands that disappeared upon addition of an excess of unlabeled competitor oligonucleotides of the same sequence. Importantly, upon addition of excess unlabeled competitor oligonucleotides containing a mutation in the

TAAT motif, only the uppermost band remained. Conversely, upon addition of a PDX1 antibody, only the uppermost band disappeared (Figure 7B), confirming that the uppermost band is the result of PDX1 binding to the probe. In order to functionally validate the involvement of the PDX1 DNA binding motif in the transactivation of the

MIR17HG promoter in β-cells, we cloned either a wild-type or TAAT mutant minimal human MIR17HG promoter (-92 to +8 bp relative to the transcription start site, shown in

Figure 7A) into a promoterless firefly luciferase construct. Co-transfection of these constructs with a renilla luciferase normalization plasmid into INS1 832/13 cells revealed

55

a 10-fold response in relative luciferase activity by the minimal miR17HG promoter that was reduced by > 65% when the core TAAT sequence of the PDX1 motif was mutated

(Figure 7C). In total, these data demonstrate that PDX1 directly binds to the human miR17HG promoter and that the PDX1 DNA binding motif is required for maximal transactivation of miR17HG in β-cells.

56

Figure 7: PDX1 Directly Regulates the MIR17HG Promoter. (A) UCSC genome browser screenshot displaying consensus peaks of enrichment from human islet PDX1 ChIPseq, significant PDX1 ChIPseq peaks from hESC-derived pancreatic endoderm, and the minimal miR17HG promoter region that was cloned for luciferase assays. The putative PDX1 DNA binding motif is highlighted. (B) Representative blot of an EMSA assay for PDX1 DNA binding motif found in the MIR17HG promoter using nuclear extract prepared from INS1 832/13 cells. Competitor is unlabeled oligonucleotides identical to probe sequence, mutant competitor is unlabeled oligonucleotides with PDX1 motif mutated. Anti-PDX1 or IgG antibodies were incubated with probe/lysate mixture prior to gel separation. The band representing the PDX1-mediated shift is highlighted. (C) Luciferase assay performed in INS1 832/13 cells for the minimal MIR17HG promoter or minimal MIR17HG promoter in which the PDX1 motif is mutated. Data represent mean ± S.E.M. of 3 independent experiments.

57

3.2.3 The PDX1 DNA Binding Motif in the MIR17HG Promoter is Conserved in the Rat Genome

Despite evidence that PDX1 is able to impart β-cell specific regulation of miR17~92 in humans, inspection of the rn5 and rn6 versions of the rat genome revealed the ~800 bp region of the MIR17HG promoter surrounding the PDX1 binding motif were not conserved compared to the human and mouse sequences. Surprisingly, when we evaluated an older version of the rat reference genome, rn4, we found a miR17~92 promoter sequence that is not in concordance with the more recent rn5 and rn6 genomes, but instead has a high degree of conservation with other species as one would anticipate for an essential miRNA cluster. To reconcile the discordance between the

MIR17HG promoter sequence in the rat reference genomes, we designed a strategy to

PCR amplify the ~2 kb region encompassing this region from INS1 832/13 rat cells. The

2 PCR products (primers JH190/JH199 = 1,234 bp; primers JH191/200 = 1,131 bp) were inserted by TOPO cloning into plasmids and 3 individual clones of each were subsequently submitted for sanger sequencing. The resulting 12 sequences (forward and reverse) were used to generate a consensus sequence for the disputed rat MIR17HG promoter region. Sequence alignment of the INS1 832/13 sequence, hg19 (human), rn4, and rn5 reference genomes revealed that the INS1 832/13 sequence is in almost complete agreement with rn4 and highly conserved with hg19, including our region of interest, the PDX1 DNA binding motif (Figure 8 and Appendix A).

58

Figure 8: Cross-Species Alignment of the MIR17HG Promoter. Sequence alignment of a fragment of the hg19 (human), INS1 832/13 (rat), rn4 (rat), and rn5 (rat) MIR17HG promoters containing the conserved PDX1 DNA binding motif (highlighted sequence). 59

3.2.4 Identification of Putative Human Islet MIR17HG Enhancers

During our analysis and visualization of the genomic data underlying PDX1 occupancy of the MIR17HG promoter, we noted a cluster of PDX1 peaks ~200 kb upstream of MIR17HG in the data from both human islets and human ESC-derived pancreatic endodermal cells. Since RNAseq analysis suggested the closest transcribed gene upstream of mMIR17HG is ~1 Mb away, and PDX1 is known to be enriched at enhancers (Pasquali et al. 2014, Wang et al. 2015), we postulated that these sites of PDX1 enrichment may correspond with islet enhancers. Indeed, when we compared these peaks with recently published ChromHMM data (Parker et al. 2013), it revealed that

PDX1 enrichment corresponded with a 13.5 kb islet-specific stretch enhancer (Figure

9A). While this region overlaps with LINC00379, a study characterizing human islet lncRNAs found no evidence of its expression (Moran et al. 2012). Stretch enhancers have been recently associated with driving cell-type specific gene expression, leading us to hypothesize that this upstream genomic region may be a novel control region for

MIR17HG.

We next sought to confirm that this upstream region is capable of interacting with the MIR17HG promoter. Since there is no public data for 3D chromatin confirmation in islets or other β-cell models, we focused on K562 cells, a CML leukemia line that is reliant on upregulation of miR17~92 members for survival (Venturini et al.

2007, Zhou et al. 2017) that has been extensively characterized both epigenetically (Ernst

60

et al. 2011, Encode Project Consortium 2012) and by 3-dimensional chromatin conformation techniques (Rao et al. 2014). Initial examination of published ChromHMM chromatin state data (Ernst et al. 2011, Parker et al. 2013) revealed a striking amount of enhancer activity unique to K562 cells in the ~2 Mbp encompassing MIR17HG, corresponding almost completely with the topological domain containing MIR17HG in

ESCs (Dixon et al. 2012). The WashU Epigenome Browser was used to visualize published HI-C and RNA PolII ChIA-PET data for K562 cells and showed that robust interactions occur between the MIR17HG promoter and the distal enhancers, confirming, that for K562 cells, these distal enhancers are novel regulatory elements for

MIR17HG. In summary, analysis of these data show that within K562 cells the topological domain containing miR17HG is littered with active distal enhancers that can interact, and presumably regulate, the miR17HG promoter, confirming that we have identified a novel control region for miR17HG.

Having confirmed that in K562 cells the MIR17HG promoter can interact with distal enhancers in the proximity of the PDX1-bound, islet-specific stretch enhancer, we set out to functionally validate the activity of this element as well as several other regions also characterized as enhancers in either human islets, the human fetal pancreatic bud-derived EndoC human β-cell line (Ravassard et al. 2011), or GM12878 lymphoblast cells by ChromHMM (Ernst et al. 2011, Parker et al. 2013). PCR was used to amplify 15 enhancer fragments, representing 10 non-overlapping elements (A-J in Figure

61

9A) from HEK293s that were subsequently TOPO cloned into a Gateway entry vector to allow the elements to be rapidly recombined into different vectors for functional tests.

We adapted Promega’s pGL4.23 minimal promoter luciferase vector by inserting a

Gateway Cloning Cassette downstream of the polyadenylation signal (SalI restriction site) in the “enhancer” position (pGL4.23/GW/B). To enable screening of enhancer elements in either orientation, we generated 2 versions of pGL4.23/GW/B (HO901,

HP001) in which the Gateway Cloning Cassette is inserted in opposite orientations, allowing the same enhancer entry plasmid to be used to generate luciferase constructs in either orientation. Putative enhancer elements along with a 3.1 kb (βgal) control DNA fragment were then recombined into both HO901 and HP001 and screened for activity in

INS1 832/13 cells and HepG2 cells. Both orientations of all the elements responded in a similar manner, so the data from both orientations were averaged together. As seen in

Figures 9B-C, the regions associated with either GM12878 (A, J) or human islet enhancers not bound by PDX1 (E, G) caused between 1.5- and 2.5-fold increases in luciferase activity in both INS1 832/13 and HepG2 cells. When we examined the PDX1- bound elements (B, C, D, F, J) two of the elements caused 3-fold increases in luciferase activity, with activation by the 8.6 kb islet stretch enhancer fragment (D) being largely specific for the INS1 832/13 β-cell line (Figure 9D). In an attempt to dissect the components of the stretch enhancer necessary for β-cell-specific activation, we created 2 smaller fragments (D_3.1kb, and D_1.2kb) nested within each other and focused on a

62

PDX1 ChIPseq peak found in both human islets and hESC-derived pancreatic endodermal cells. Surprisingly, these stretch enhancer fragments caused a further doubling of luciferase activity over the already elevated activation levels of the larger 8.6 kb region, while still maintaining β-cell specificity (Figure 9E). These results support the idea that a PDX1-bound element is crucial for enhancer activity and specificity, but a more thorough dissection of the region involving identification and mutation of the

PDX1 binding site is needed for full validation.

63

Figure 9: Functional Validation of ChromHMM Identified Islet Enhancers. (A) UCSC genome browser screenshot of the miR17HG locus and ChromHMM identified chromatin states (Stephen Parker, unpublished). It should be noted that ChromHMM analysis in this study identified some strong enhancer states as promoters. Cloned regions A-J are highlighted. Luciferase activity for ChromHMM identified enhancers that are (B) non-islet, (C) islet without PDX1 enrichment, (D) PDX1-bound islet, and (E) fragments of a PDX1-bound islet stretch enhancer in INS1 832/13 or HepG2 cells. Data represent mean of both orientations ± S.E.M. from 7-8 independent experiments. * p < 0.05 compared to HepG2 cells.

64

Finally, we focused on a ChromHMM enhancer identified in the EndoC β-cell line designated as region I in Figures 9A and 10A (EndoC ChromHMM analysis courtesy of Stephen Parker, unpublished). This region appears to be part of a larger element that has significant PDX1 enrichment in hESC-derived models of pancreatic development. Of interest, this region is flanked by a group of T2D-associated SNPs with p-values ranging from 1.0x10-5 to 3.1x10-5 in the DIAGRAMv3 GWAS study (Morris et al.

2012). We focused our subsequent efforts on one of these SNPs, rs9583907 (p-value =

1.00x10-5), that was found be protective against T2D, with the reference allele having an odds ratio of 1.1. In addition, only this SNP has a trend of allelic imbalance in EndoC cells for ATACseq (reference allele) and H3K27ac ChIPseq (minor allele) reads indicative of a transcription factor preferentially binding the reference allele and excluding a nucleosome (Brooke Wolford, Stephen Parker, unpublished). To determine if this SNP is sufficient to alter enhancer activity, we cloned a 2.1 kb region of this element containing either the reference (I_2.1kb) or minor allele (Minor Allele). In addition, we performed site-directed mutagenesis on each construct to change the reference allele to the minor allele (Ref->Minor) and vice versa (Minor->Ref). All four of the elements were then recombined into both of our luciferase vectors and transfected into INS1 832/13 or

HepG2 cells. Interestingly, all of these elements exhibited a strong (~4.5-fold), and cell- type specific activation; however, we could find no significant difference between those elements containing the reference or minor alleles (Figure 10B). Taken together, these

65

data suggest a previously undescribed enhancer field capable of regulating MIR17HG that contains both an islet-specific, PDX1-bound stretch enhancer as well as a potential developmental pancreatic enhancer containing SNPs associated with type-2 diabetes.

Figure 10: SNP rs9583907 Does Not Alter Enhancer Activity in INS1 832/13 Cells. (A) USCS genome browser screenshot of the putative MIR17HG enhancer field that contains SNPs associated with T2D (p < 3.1x10-5) in the DIAGRAMv3 analysis. (B) Luciferase activity for region I_2.1kb containing either the reference allele for SNP rs9583907, the minor allele, the reference allele changed to the minor allele, or the minor allele changed to the reference allele. Data represent mean of both orientations ± S.E.M. from 7-8 independent experiments. * p < 0.05 compared to HepG2 cells.

66

3.2.5 miR17~92 miRNAs are Induced by PDX1 Overexpression

The prospect of a PDX1/miR-17 regulatory system caught our attention due to our observations that PDX1 overexpression simultaneously increases β-cell replication and survival, similar to miR17~92 miRNA phenotypes reported in other cell types. In addition, recent work has demonstrated that this cluster of miRNAs is required to maintain β-cell mass in a high-fat fed mouse (Chen et al. 2016). Having established the cross-species conservation of the PDX1 DNA binding motif in addition to potential

PDX1-bound distal enhancers, we sought to directly investigate whether miR-17 is regulated by PDX1 overexpression in rat islets. We transduced primary rat islets with adenoviruses expressing either βGAL, PDX1 or NKX6.1, a transcription factor that increases β-cell replication through a PDX1-independent pathway, from the CMV promoter. The islets were harvested for RNA isolation 72h after transduction and qRT-

PCR was performed for the primary miR-17~92 transcript, MIR17HG, as well as the 6 mature miRNAs co-localized in the cluster. As shown in Figure 11, islets treated with

PDX1, but not NKX6.1, exhibited a significant 2.5-fold increase in MIR17HG transcript levels, accompanied by similar increases in the 6 mature miRNAs co-localized in the cluster, miR-17 (1.8-fold), miR-18a (2.7-fold), miR-19a (1.8-fold), miR-20a (1.6-fold), miR-

19b (1.7-fold), and miR-92a (1.5-fold), all compared to levels measured in the Ad-CMV-

βGAL-treated control cells.

67

Figure 11: miR17~92 is Induced by PDX1 Overexpression. qRT-PCR analysis for (A) PDX1 and NKX6.1; (B) MIR17HG; (C) miR17~92 mature miRNAs. Rat islets were harvested for analysis 72h after being transduced with adenoviruses expressing either βGAL, PDX1, or NKX6.1. Data represent mean ± S.E.M. from at least 3 independent experiments. * p < 0.05 compared to Ad-βGAL.

Although these data seem to corroborate data summarized earlier (e.g. section

2.2.2), it remains possible that PDX1 overexpression induces MIR17HG expression via an indirect mechanism. Indeed, microarray gene analysis and qRT-PCR validation of rat islets overexpressing PDX1 revealed a modest 1.5-fold upregulation of MYC (Figure

68

12B), a transcription factor that has been widely reported to regulate MIR17HG expression (O'Donnell et al. 2005, Venturini et al. 2007, Mu et al. 2009, Li et al. 2014). To directly test this alternative hypothesis, we incubated Ad-PDX1 transduced rat islets with the MYC/MAX inhibitor 1RH (Yin et al. 2003) for the final 24h before harvesting of

RNA. As shown in Figure 12, inhibition of MYC did not blunt PDX1 overexpression or

PDX1-mediated induction of MIR17HG. As confirmation of target engagement by 1RH, the previously reported MYC-target genes MYC and CCND2 were analyzed and found to have trends to increase 2.7-fold (MYC, p = 0.07) and decrease 20% (CCND2, p = 0.16) when compared to Ad-PDX1 alone, demonstrating successful MYC inhibition with 1RH

(Figure 12B). These results suggest that the induction of the miR17~92 cluster by PDX1 overexpression is direct, although additional experiments will be required to determine whether this is through direct transactivation of the MIR17HG promoter and/or potentiation of distal enhancer activity.

69

Figure 12: PDX1-Mediated Increases in miR17HG are not Blunted by Myc Inhibition. qRT-PCR analysis of (A) PDX1 and (B) MIR17HG, MYC, CDKN1A, and CCND2 expression. Rat islets were untreated or transduced with an adenovirus expressing PDX1 and cultured for 72h. The MYC inhibitor 1RH (40uM) was added for the final 24h of culture. Data represent mean ± S.E.M. from 3 independent experiments. * p < 0.05 compared to untreated.

3.2.6 Peptide-Nucleic Acids as a Tool for Islet Research

The islet is a sphere of ~1,000 non-replicating cells which makes transfection efficiencies very low (Bain et al. 2004). This inability to transfect islets means that β-cell miRNA studies to date have been largely restricted to cell culture models, knockout/overexpression mouse models, or use of viral vectors to overexpress miRNAs.

In order to elucidate the role of miR17~92 miRNAs in PDX1-mediated replication we had to implement a new approach. The development of peptide nucleic acids (PNA) to inhibit miRNAs was recently reported (Oh et al. 2010). These molecules contain a synthetic nucleic acid fused to a cell-penetrating peptide, allowing them to be delivered into cells without the use of transfection reagents. PNA use has not been reported in

70

islets. To test this technology in islets, we targeted miR-375, a well-characterized islet miRNA (Poy et al. 2004, Poy et al. 2009), with a FAM-labeled PNA. Rat islets were incubated with 2 µM FAM-PNA-375 for 48h, washed, and then cultured for an additional 48h. As seen in Figure 13A, examination of the islets with a fluorescent microscope revealed all islets had taken up the probe. In addition, flow cytometry analysis showed that > 40% of the total islet cells contained FAM (Figures 13B-C).

Finally, qRT-PCR analysis of rat islets treated with either PNA-Scr, a non-targeted control PNA, or PNA-375 demonstrated a significant ~1.5-fold increase of the miR-375 target gene ELAVL4 (HuD) in response to PNA-375 treatment (Figure 13D). Altogether, these data demonstrate that PNA technologies can be successfully used to inhibit miRNAs in primary islets.

71

Figure 13: Validation of Peptide-Nucleic Acids as a Tool for miRNA Research in Primary Islets. (A) Fluorescent microscopy image of rat islets treated with a FAM- labeled PNA targeting miR-375. (B) Untreated and (C) FAM-PNA-375 treated rat islets were analyzed by flow cytometry. (D) qRT-PCR analysis of ELAVL4 expression 72h after incubation with 2 uM of PNA-375. Data represent mean ± S.E.M. from 4 independent experiments. * p < 0.05 compared to PNA-Scr.

72

3.2.7 Regulation of PDX1-Mediated Replication by miR17~92 miRNAs

Having established that the miR-17 family targets the PDX1 network, miR17~92 is induced in rat islets overexpressing PDX1, and that PNAs can be utilized to inhibit miRNAs in primary rat islets, we attempted to directly test the contribution of miR17~92 miRNAs to PDX1-mediated replication. To do this we transduced islets with Ad-PDX1 and 24h later added either a control PNA with a non-targeting scrambled sequence

(PNA-Scr) or combinations of PNAs targeting the most abundant miR17~92 miRNAs based on our qRT-PCR analysis (miR-17, miR-19b, miR-20a, and miR-92a).

Unfortunately, this perturbation did not yield any significant changes to PDX1-mediated islet replication, although, we do have a trend to decrease replication by ~15% upon addition of PNAs-17, 19b, 20a, and 92a (p = 0.10; Figure 14A)). However, the changes, or lack thereof, were difficult to interpret due to an inability to find any miRNA target genes that were responsive to PNA treatment, as shown for miR-17, 20a, and 92a inhibition in Figure 14. As discussed below, the inability to detect changes in miRNA target gene expression is likely due to the heterogenous cell composition of the islet, as well as incomplete transduction efficiencies of both our adenoviral and PNA tools.

73

Figure 14: miR17~92 Inhibition by PNAs Does Not Alter PDX1-Mediated Replication. (A) [3H]-thymidine incorporation performed on rat islets treated with either Ad-βGAL or Ad-PDX1 with combinations of PNAs targeting the miR17~92 miRNAs. qRT-PCR analysis of (B) PDX1 and (C) miR17~92 target gene expression levels in rat islets treated with either Ad-βGAL, Ad-PDX1 + PNA-Scr, or Ad-PDX1 + PNAs-17, 20a, and 92a. Data represent mean ± S.E.M. from at least 3 independent experiments.

3.2.8 miR17~92 miRNAs are Not Sufficient to Drive β-cell Replication

Within the cancer biology field, miR17~92 miRNAs have been found to be sufficient in some contexts to drive cellular replication (Sandhu et al. 2013). Consistent 74

with this idea, our lab has previously shown that RNAi-mediated silencing of the cell- cycle inhibitor CDKN1A (p21), a validated target of the miR-17 family, is sufficient to drive rat islet β-cell replication (Tessem et al. 2014). To test if miR17~92 is sufficient to drive islet cell replication, we developed a novel adenovirus encoding the miR17~92 cluster in a synthetic intron as part of a larger PolII transcript expressed in a Dox- inducible, β-cell specific manner (Hayes et al. 2016). Upon addition of Dox to transduced cells, the primary transcript is expressed, and upon splicing of the synthetic intron, the miR17~92 cluster is released from the transcript in a manner suitable for downstream processing by DROSHA. Transduction of primary rat islets with this vector demonstrated successful overexpression of all six of the miR17~92 mature miRNAs ranging from 8- to 128-fold upon addition of Dox, although it should be noted that vector alone caused a 3- to 8-fold increase in miR17~92, presumably due to leakiness of the Tet-On system (Figure 15A). As a control, we performed qRT-PCR for another miR-

17 family member, miR-106b, whose expression levels were not altered by the virus.

However, despite robust overexpression of the mature miR17~92 miRNAs, no increase in islet cell replication was observed (Figure 15B). While overexpression of the miR17~92 miRNAs are not sufficient, we postulated that further overexpression of these miRNAs may enhance PDX1-mediated replication. However, as seen in Figure 15C, we also found no significant difference in PDX1-mediated islet replication in the presence or absence of miR17~92 overexpression. One possible explanation for these results is that

75

the overexpressed miRNAs were unable to engage with their target mRNAs. To test this possibility, we utilized the eGFP reporter built into both the control and miR17~92 vectors to utilize fluorescent-assisted cell sorting (FACS, validation discussed in Chapter

5.2.1) to separate the transduced β-cells from rat islets. qRT-PCR of the sorted cells confirmed target engagement and the expected inhibition of miR17~92 target genes

(Figure 15D), including CDKN1A. However, coinciding with this decrease we also observed a decrease in expression of the D-type cyclins CCND1 and CCND2. Therefore, we hypothesize that the failure of miR17~92 to induce β-cell replication is either that the level of CDKN1A suppression is insufficient as compared to RNAi-mediated silencing; or that the overall replication competency of miR17~92 overexpressing β-cells is reduced by decreased expression in D-type cyclins.

76

Figure 15: miR17~92 is Not Sufficient to Drive β-Cell Replication. (A) qRT-PCR analysis for mature miRNAs in untreated rat islets or islets transduced with either a RIP- Tet-On control (Ad-Control) or RIP-Tet-On miR-17~92 (Ad-miR17~92) adenovirus with or without Dox added. [3H]-thymidine incorporation in rat islets transduced with (B) Ad- Control, Ad-miR17~92, or Ad-PDX1, ± Dox or (C) combinations of Ad-Control or Ad- miR17~92 with Ad-βGAL or Ad-PDX1, ± Dox. (D) qRT-PCR gene expression analysis of predicted miR-17~92 targets in FACS isolated primary rat β-cells transduced with Ad- miR17~92. [3H]-thymidine incorporation data represent mean ± S.E.M. from at least 2 independent experiments. * p < 0.05 compared to untreated Ad-Control. 77

3.2.9 The miR-17~92 Target PTEN Regulates PDX1-Mediated Replication

Knowing that our average transduction efficiency for any particular islet experiment ranges from 40% - 80%, and that overall only ~5% of the islet cells are replicating after PDX1 overexpression, we hypothesized that one technical reason for our inability to detect changes in miR17~92 miRNA targets in PDX1-overexpressing cells is that the modest changes expected from miRNA regulation are diluted out by the presence of other cell types and non-transduced β-cells. To test this, we transduced rat islets with an adenovirus that is capable of co-expressing PDX1 and eGFP specifically in the β-cell (discussed in more detail in chapter 5) and then used FACS to isolate the PDX1 overexpressing β-cells for qRT-PCR analysis. As observed in intact islets, in the isolated cells we were able to detect induction of both CCND1 (3.4-fold, p = 0.09) and CCND2

(4.8-fold, p = 0.04) (Figure 16A). Importantly, under these conditions we were also able to observe a significant 33% decrease in the expression of the miR17~92 target PTEN

(Figure 16B); deletion of this gene in vivo activates AKT signaling, leading to CDK4- dependent β-cell replication (Fatrai et al. 2006, Wang et al. 2010). As previously discussed, CDK4-dependence is one of the hallmarks of PDX1-mediated β-cell replication (Hayes et al. 2013). This convergence of phenotypes mediated by PDX1 overexpression and PTEN knockout led us to postulate that these pathways may be interconnected. We tested this hypothesis by treating PDX1-overexpressing rat islets with the PTEN inhibitor VO-OHpic (Mak et al. 2010). Thymidine incorporation revealed

78

that PTEN inhibition potentiated PDX1-mediated islet replication by an additional 1.6- fold. Importantly, this effect was specific to the PDX1-driven replication pathway as we observed no increase in NKX6.1-mediated islet replication upon PTEN inhibition (Figure

16C). Taken together, these data implicate the miR17~92 target PTEN as a regulator of

PDX1-mediated islet replication.

79

Figure 16: PTEN Status Regulates PDX1-Mediated Replication. qRT-PCR analysis of (A) CCND1 and CCND2, or (B) PTEN in FACS isolated rat β-cells overexpressing GFP alone or PDX1/GFP. Data represent mean ± S.E.M. from 3 independent experiments. * p < 0.05 compared to Ad-RIP-GFP. (C) [3H]-thymidine incorporation in rat islets transduced with either Ad-βGAL, Ad-PDX1, or Ad-NKX6.1, ± the PTEN inhibitor VO-OHpic. Data represent mean ± S.E.M. from 2 independent experiments. * p < 0.05 compared to Ad-PDX1 + DMSO.

3.3 Discussion

The transcriptional network through which PDX1 drives β-cell replication has remained elusive, in part due to the activation of replication through multiple pathways and cell types (Hayes et al. 2013, Hayes et al. 2016). In this work, we used statistical 80

modeling to identify the PDX1/miR17~92/PTEN axis as one of these potential pathways, through demonstration of β-cell specific regulation of MIR17HG by PDX1, and enhancement of PDX1-mediated β-cell replication via chemical inhibition of PTEN. In addition to direct transactivation of the MIR17HG promoter, we also identified a previously unappreciated MIR17HG enhancer field that contains a PDX1-bound islet stretch enhancer that is significantly more active in INS1 832/13 β-cells than HepG2 cells.

Interestingly, this novel MIR17HG enhancer field also contains a cluster of T2D- associated SNPs (p = 1.0-3.1x105) (Morris et al. 2012) that overlap with potential pancreatic developmental enhancers identified in EndoC β-cells and bound by PDX1 in human ESC-derived pancreatic endoderm. These regions are not bound by PDX1 in adult, post-mitotic human islets. Characterization of one of these SNPs, rs9583907, and the corresponding enhancer in INS1 832/13 and HepG2 cells demonstrated β-cell specific activity but no difference in activity between the reference and minor alleles.

The regulation of miR17~92 up to this point has exclusively focused on direct promoter transactivation as well as an intronic enhancer (O'Donnell et al. 2005, Woods et al. 2007, Ji et al. 2011, Mogilyansky and Rigoutsos 2013). Most of this literature implicates the MYC family of transcription factors as being the main regulators of miR17HG, however, other work has also identified regulation by E2F transcription factors during the cell cycle, as well as TP53 (p53). Currently, there is little known about the regulation of miR17HG by cell-type specific factors and, to our knowledge, no

81

reports describing control via distal enhancer elements. Interestingly, we discovered two

β-cell specific modes of MIR17HG regulation involving the direct transactivation of the promoter by PDX1 as well as distal enhancer elements located in a putative MIR17HG enhancer field. Demonstrating specificity of these mechanisms, we showed that PDX1- mediated induction of MIR17HG was not blunted by chemical inhibition of MYC or induced by NKX6.1. The regulation of MIR17HG by PDX1 will need to be examined within other physiologic/pathophysiologic contexts such as β-cell development and β- cell compensation during high fat feeding. In this vein, we have discovered that, contrary to published reports (Lerner et al. 2012, Upton et al. 2012), the miR17~92 miRNAs are induced > 2.5 fold by induction of ER stress in isolated rat islets, but the potential role of PDX1 in this response has not been examined. In addition, future efforts expanding beyond the MIR17HG promoter to elucidate the contribution of the putative MIR17HG enhancer field in regulation of miR17~92 miRNAs under different biological conditions will be of particular interest. A prime example of this is the K562 model where the literature has attributed MIR17HG dysregulation solely to increased

MYC activity (Venturini et al. 2007), despite our observations using published epigenetic and chromatin conformation analyses that are highly suggestive of altered chromatin architecture of this locus to drive hyperactivation of multiple enhancers within the topological domain. The role of aberrant enhancer activity and chromatin architecture in the etiology of disease is currently an area of intense interest, especially in the context of

82

cancer, and future studies should focus on the MIR17HG locus as well. Fortunately, with the development of epigenetic engineering platforms, such as pMAGIC discussed in

Chapter 6, we will finally be able to interrogate enhancer elements in their native chromatin contexts to determine their contribution to gene expression and physiological phenotypes.

The ability of the miR17~92 cluster miRNAs to enable cell replication and survival is well documented in the cancer field, and recent reports have found these miRNAs to play a similar role in the β-cell as well (Jacovetti et al. 2015, Chen et al. 2016).

However, in contrast to the leukemia field, we found that overexpression of miR17~92 miRNAs are not sufficient to induce β-cell replication. A likely explanation for this lies in our observation that concerted action of the miRNA cluster balanced out repression of cell cycle inhibitors such as CDKN1A and PTEN with repression of cell cycle activating genes such as CCND1 and CCND2. Conversely, in the context of PDX1, miR17~92 miRNAs are still able to target CCND1 and CCND2 to dampen the PDX1-mediated inductions, but the end result is still increased expression, whereas miR17~92 miRNA targeting of PTEN functions as an amplifying loop ensuring PTEN repression and progression of β-cell replication as modeled in Figure 17. Our data also indicates that the suppression of PTEN within this context is incomplete, such that additional inhibition by a small molecule is able to further enhance PDX1-mediated replication. Our inability to validate target engagement of PNAs targeting the miR17~92 cluster requires that

83

further experiments be performed using alternative approaches. Taken together, these data are suggestive of a PDX1/miR17~92/PTEN axis as one of the potential pathways involved in regulating β-cell replication, although additional work with improved tools for miRNA manipulation within primary islets, such as those described in Chapter 6, will be required to confirm this.

Figure 17: Proposed Model of β-Cell Replication Driven by the PDX1/miR17~92/PTEN Axis

84

4. Creation of pMVP, a Versatile Cloning Platform for Transgene Expression 4.1 Introduction

A core strategy for biomedical research is to genetically manipulate specific components of complex physiological systems to define regulatory mechanisms and disease-causing pathways. Such strategies can be hampered by limitations imposed by current systems for delivery of transgenes or gene suppressors into specific cell-types central to disease etiology. The pancreatic islets of Langerhans serve as a case in point.

Loss of islet mass and function is central to the development of both major forms of diabetes mellitus. Islets are complex micro-organs comprised of several distinct endocrine cell types that participate in metabolic fuel homeostasis, mainly via the production and secretion of insulin (β-cells), glucagon (α-cells), and somatostatin (δ- cells) (Jain and Lammert 2009, Steiner et al. 2010, Caicedo 2013). Over the past two decades, our group has used recombinant serotype 5 adenoviruses (Ad5) to study the impact of manipulation of specific genes on pancreatic islet cell function, replication, and survival (Becker et al. 1994, Bain et al. 2004, Schisler et al. 2005, Ronnebaum et al.

2006, Schisler et al. 2008, Stephens et al. 2012, Hayes et al. 2013, Jensen et al. 2013,

Tessem et al. 2014, Hayes et al. 2016, Hayes et al. 2017). Whereas adenoviral vectors have proven to be an important tool to gain insights into an otherwise difficult model system, virus construction remains laborious and time-consuming. Furthermore, the current adenovirus platforms are almost entirely based around a single modality of 85

CMV promoter-driven transgene expression, unnecessarily restricting experimental design. Our recent development of a novel adenovirus vector that utilizes a self- contained Tet-On system driven by the rat insulin promoter (RIP) is a start to vector diversification for studies of islet biology (Hayes et al. 2016); however, adenovirus construction remains dependent upon slow and laborious endonuclease-based cloning to generate single-modality vectors. Surprisingly, these problems are not just limited to adenoviral vectors, but are universal limitations of all current vector platforms.

In this work, we sought to invent a platform that (1) requires minimal molecular biology expertise, (2) is rapid and highly efficient, (3) is modular in nature to enable a high degree of customization and adaptability, and (4) can be used universally across all vectors types. These efforts yielded pMVP, a plasmid-based modular vector platform, that utilizes Multisite Gateway® Pro cloning in lieu of traditional restriction endonuclease cloning to enable rapid, high-fidelity assembly of unique transgene expression and RNAi vectors. Gateway cloning is an in vitro recombinational cloning method based on the site-specific recombination reactions mediated by the λ integrase family recombinases (Hartley et al. 2000) that allow for sequence conserved, high fidelity, and orientation specific recombination reactions. The first phase of this process involves a “BP” reaction in which a recombination occurs between “attB” (located on the

DNA insert) and “attP” (located in the “pDONR” plasmid) sequences. This mediates the transfer of a sequence of interest into the pDONR plasmid, resulting in the creation

86

of a “pENTR” entry plasmid (also known as a Gateway shuttle plasmid) that now contains the sequence of interest flanked by “attL” recombination sites. The resultant pENTR plasmid can then be utilized in a “LR” recombination that recombines the attL sequences with “attR” sequences located in a “pDEST” vector referred to as a Gateway

Destination vector. The LR reaction yields a completed vector that is ready for use.

We included several user-selectable options for pMVP vector design, including: ubiquitous or cell-type specific promoters; conditional transgene regulation; shRNA expression; different epitope tags; and/or fluorescent reporters for tracking transduced cells. Importantly, by basing pMVP on Gateway cloning technology, the resulting expression cassettes are universally compatible with any vector containing the Gateway cloning cassette. Therefore, the utility of pMVP was expanded beyond adenovirus vectors by engineering an assortment of custom Gateway destination vectors that enable the desired components to be readily incorporated into a fiber-modified adenovirus, lentivirus, gene expression vector, or Sleeping Beauty transposon vector. As such, the current form of pMVP allows a researcher to generate > 8,000 unique transgene expression vectors for any gene inserted into the platform. Furthermore, pMVP is highly adaptable and can be readily expanded to incorporate elements or vectors of interest via the methods described below. Thus, in total, pMVP establishes a system that allows the researcher to rapidly create a myriad of purpose-built vectors with unique functional properties to match experimental goals.

87

4.2 Results

4.2.1 Application of MultiSite Gateway® Pro to Enable Rapid Assembly of Multicomponent Vectors

To enable rapid, high-fidelity assembly of unique vectors in a user-friendly manner, we utilized Multisite Gateway® Pro technology (Katzen 2007) (Figure 18) as a foundation to generate a suite of ~90 custom components that we term globally as

“pMVP”, through which a gene of interest can be paired with combinations of promoter/polyA (Figure 18B) or promoter/N-terminal fusion/polyA (Figure 18C) elements and recombined into a promoterless Gateway destination vector, “pDEST”, via an overnight LR recombination reaction. Thus, in as little as five days, a gene of interest can be added to the pMVP platform and recombined with pMVP Entry plasmids to form a unique vector suited for a specific experimental need (Figure 19A). Importantly, the modular nature of pMVP means that the same pENTR R4-R3 plasmid containing a gene of interest is capable of being recombined with any variation of the pMVP elements to create 3 or 4 component expression vectors or even co-expressed with an shRNA

(Figures 18B-C, 19B). Of note, these additional unique vectors begin with the overnight

LR reaction and can be generated within three days with < 2 hours of benchwork (Figure

19A).

88

Figure 18: Overview of MultiSite Gateway Pro Utilization in pMVP. (A) pMVP plasmids are based on five types of MultiSite Gateway Pro pENTR plasmids. Inserting a gene of interest into the pENTR R4-R3 entry plasmid allows it to be utilized in (B) 3- component or (C) 4-component attL/attR (LR) recombination reactions into a pDEST vector containing the attR1 and attR2 sites.

89

Figure 19: pMVP Enables Rapid, High Fidelity Assembly of Custom Multicomponent Transgene Vectors. (A) pMVP allows a gene of interest to be incorporated and utilized to create a custom vector in only 5 days. The process commences with a BP recombination reaction of either a PCR product or synthesized DNA fragment containing the appropriate attB sites into the attP sites of a pDONR plasmid to create an Entry plasmid. A verified entry plasmid along with entry plasmid for the desired pMVP components can then be recombined into a custom Gateway destination vector via an overnight LR recombination reaction. Transformation of this reaction into E. coli yields colonies containing the newly created vector. Additional vector cloning commences with the overnight LR recombination reaction (designated by **) and can be completed within 3 days. (B) A schematic representation of the components utilized in the different types of vectors that can be created with pMVP.

90

4.2.2 Generation of pMVP Entry Plasmids

4.2.2.1 Promoters

We designed pMVP to generate vectors capable of ubiquitous expression as well as cell type-specific expression in models of interest such as the pancreatic islet β-cell, hepatocytes, and cardiomyocytes. Therefore, to drive gene expression, we incorporated various ubiquitous (CMV, EF1a) and cell-type specific (RIP, β-cell (Hayes et al. 2016);

TBG, hepatocyte (Ran et al. 2015); ALB, hepatocyte (Heo et al. 2006); cTnT, cardiomyocyte (Prasad et al. 2011)) promoters, as well as our recently-published self- contained Tet-On inducible expression system (driven by RIP, CMV, or EF1a) that allows for transgene expression to be controlled by the presence or absence of the antibiotic Doxycycline (Dox) (Figure 20) (Hayes et al. 2016). Entry plasmids for these promoters were created through the addition of attB sites to the CMV, EF1a, RIP, TBG,

ALB, and cTnT promoters during PCR amplification from the plasmid of origin, and subsequently used in a BP recombination to insert them into pDONR221 1-4 (Figure

20A), pDONR221 1-5r (CMV, EF1a, RIP; Figure 20B), or pDONR221 5-4 (CMV, EF1a,

RIP; Figure 20C).

The self-contained Tet-On system requires two separate entry plasmids, the upstream PolII TRE (tet-response element) promoter and downstream expression cassette for the Tet-On transactivator (TETa). In the same manner described above, we generated TRE entry plasmids using pDONR221 1-5r, pDONR221 1-4, and pDONR221

5-4 (Figure 20A-C). For the downstream TETa expression cassette, we used our recently 91

published RIP-TETa containing vector (Hayes et al. 2016) as the PCR template and added the attB3 and attB2 sequences to the PolyA- RIP-eGFP-P2A-TETa-PolyA cassette.

Critically, the P2A “self-cleavage” motif in this construct enables bicistronic expression of both eGFP and TETa. The PCR product was then used in a BP recombination to insert it into the pDONR221 3-2 plasmid to generate the final entry plasmid IG806 (Figure

20D). To generate CMV (IR904) and EF1a (JN511) versions of the Tet-On system, IG806 was digested with XbaI/EcoRI to excise the RIP sequence and NEBuilder HiFi DNA

Assembly was used to insert a PCR product containing either the CMV or EF1a promoters. To enable the TETa system to be used with “open” genes cloned without a stop codon, we ligated oligonucleotides encoding a flexible GS linker and HA epitope tag into HindIII/SwaI digested IG806, IR904, and JN511 to create plasmids KA701 (RIP),

KA801 (CMV), and KA901 (EF1a) (Figure 20E). Finally, to accommodate experiments where eGFP expression is not desired, we excised eGFP from IG806, IR904, and JN511 using a EcoRI/XcmI digest and ligated the annealed oligonucleotides JH653/JH654 into the void, yielding plasmids KA401, KA501, and KA601 (Figure 20F).

92

Figure 20: pMVP Promoter Entry Plasmids. Ubiquitous (CMV, EF1a), cell-type specific (RIP, ALB, TBG, cTnT), and inducible (TRE) promoter pMVP entry plasmids were created in (A) pENTR221 1-4, (B) pENTR221 1-5r, and (C) pENTR221 5-4 variants. (D-F) The TETa expression cassette component of the Tet-On system with or without eGFP and a HA epitope tag were generated in pENTR221 3-2 plasmids. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct.

93

4.2.2.2 Genes of Interest and Control Genes

For any given experiment, having a proper control vector is as vital as the experimental vector. The current accepted practice for transgene overexpression studies are to include a control vector expressing either βGAL or a fluorescent protein. Thus, in the pMVP toolbox we included GAL in addition to eGFP, eYFP, eCFP, and mCherry cloned both with and without (open) a stop codon (Figure 21A). In addition, we included Cre recombinase to enable creation of vectors for excision of “floxed” genetic elements as well as firefly luciferase for generation of vectors that can be used in luciferase activity assays (Figure 21B). Finally, we included several genes that our lab is currently investigating, namely mouse PDX1, human PDX1, human PDX1open, and human NKX6.1 (Figure 21C). As described previously (Figure 19A), pMVP entry plasmids for these genes were inserted via gene synthesis (human NKX6.1) or by PCR addition of attB4r and attB3r sequences (all others) and subsequent BP recombination reaction to insert them into pDONR221 P4r-P3r. As an alternative approach for introducing genes of interest into pMVP, plasmid IO102 (eGFPopen) can be digested with

NcoI/BsrGI to excise eGFP and replaced with a PCR product containing a gene of interest by NEBuilder HiFi DNA Assembly or any other Gibson isothermal DNA assembly variant. It is important to note that when genes of interest are cloned it is vital to evaluate: (1) the presence of a strong kozak surrounding the ATG start site,

(G/A)NNATGG, to enable proper initiation of translation; (2) presence or absence of a

94

stop codon depending on whether a C-terminal fusion is desired; (3) incorporation of the proper number of nucleotides to maintain frame with N- and C-terminal pMVP components (Figure 21D).

Figure 21: pMVP Genes of Interest. Available pMVP pENTR R4-R3 plasmids containing (A) control genes and (B-C) genes of interest. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct. (D) Recommended flanking primer sequences containing attB sites to incorporate additional genes of interest into pMVP by PCR amplification and BP recombination into pDONR221 P4r-P3r and retain frame with the rest of the pMVP components.

4.2.2.3 shRNA Expression

Our lab pioneered the utilization of shRNA expressing adenoviral vectors for gene inhibition in primary islets (Bain et al. 2004). However, this vector is still based on an early adenovirus platform, pJM17, that is dependent on the occurrence of rare

95

homologous recombination events in HEK293 cells. To take advantage of the much greater efficiency of Gateway-based adenovirus construction, we created a pDONR221- derived entry plasmid (HD601) containing the H1 promoter and our published shRNA cloning region (Bain et al. 2004) that can be directly inserted into a promoterless

Gateway destination vector. A desirable experimental paradigm is to be able to simultaneously overexpress a factor, such as PDX1, while using RNAi to inhibit a downstream component of the PDX1 pathway. Our current system requires two individual adenoviruses for this experiment, adding unnecessary technical complexity to the model. As such, we sought to incorporate the ability for pMVP to generate vectors capable of simultaneous shRNA and transgene expression. As represented in Figure

19B, we achieved this by creating a pENTR L1-R5 plasmid containing the H1-driven shRNA expression cassette (HE401) that can be recombined with the elements required for transgene expression. Further refinement of this approach has led to the realization that it is more economical to de novo synthesize the entire shRNA expression cassette with attL1 and attR5 sites for each shRNA (Figure 22). Importantly, we have discovered that the resulting synthesized sequence can be used directly in the LR recombination reaction, saving several days of effort normally required to create an entry plasmid.

96

Figure 22: pMVP RNAi.(A) Schematic for a complete and recombination-ready RNAi synthetized cassette. Random bases (black boxes) and 5’ and 3’ recombination sites (attL1 and attR5, respectively) flank the RNAi expression cassette, comprised of the H1 RNA promoter, a target-specific stem-loop (formed from an inverted repeat) and a T5 terminator. To overcome difficulties in stem-loop DNA synthesis, targets (typically 20mers) that start with a G and end in CNC, CNNC or CNCN are selected, and the two 3’ Cs are mutated to T in the forward (i.e. identical sequence to the target sequence) repeat only (shown in red). Care must be taken to avoid formation of a premature T5 terminator, but since U and G can hybridize in a wobble, this maneuver blocks formation of a perfect hairpin longer than 18-20 bp without affecting target specificity, while simultaneously encouraging uptake of the silencing (reverse repeat) strand by the RISC silencing complex. (B) DNA synthesis template for a RNAi cassette designed to express a random control shRNA. Recombination sites are shown in green and flanked by random DNA sequence. The H1 RNA promoter is shown in bold, the inverted repeats are underlined, and the 2 C/T mutations at the 3’ end of the forward repeat are shown in red. (DNA synthesis template designed by Thomas Becker, unpublished)

4.2.2.4 Epitope Tags

For transgene detection, an assortment of epitope tags (HA, FLAG, OLLAS, V5, myc, 6x-His) were generated for fusion to either the N-terminus or C-terminus of the

97

recombinant protein of interest (Figures 23). N-terminal epitope tag constructs contain the Xenopus 5’ beta-globin leader to enhance translation and are followed by an in- frame GS flexible linker to help prevent epitope masking. These sequences were either synthesized as a gBlock (HA, OLLAS, 3x-HA, 3x-FLAG) and recombined into pDONR221 5-4 or synthesized as oligonucleotides (myc, V5) and ligated into

BglII/BamHI digested IH001.

For C-terminal epitope tags containing a polyA signal (non-lentiviral) we generated a pENTR L3-L2 plasmid containing a 3x-HA-bGH polyA (bovine growth hormone poly A signal, IA001) sequence amplified from Addgene plasmid 61591 (cite

Zhang lab) that was used as a framework for all C-terminal epitope tags. Similar to N- terminal epitope tags, all C-terminal epitope tags are preceded by a GS flexible linker to help prevent epitope masking. Entry plasmids for additional epitope tags (FLAG, IX601;

OLLAS, IX503; myc, IX301; 6x-His, IX401) were subsequently created by ligation of annealed oligonucleotides encoding epitope tags of interest into BamHI/EcoRI digested

IA001. It should be noted, pMVP also includes a pENTR L3-L2 plasmid containing only the human growth hormone polyA signal (HB901) if a C-terminal epitope tag is not required.

As RNA viruses, lentiviruses do not tolerate polyA signals and therefore recombinant lentivirus vectors utilize a Woodchuck Hepatitis Virus Posttranscriptional

Regulatory Element (WPRE) to aid in transgene expression. Thus, to enable pMVP to be

98

utilized for the creation of lentiviruses, we generated several pENTR L3-L2 plasmids containing a WPRE in replace of a polyA signal. Since the underlying sequence of the

WPRE can interfere with PCR reactions, we digested IA001 (3x HA-bGH polyA) with

SacI/KpnI to excise the polyA sequence and then used NEBuilder HiFi DNA Assembly and two linking oligonucleotides (JH543, JH557) to insert the AgeI WPRE fragment from pLenti6.3/V5-DEST creating IX907. Finally, entry plasmids containing FLAG (JA001),

OLLAS (IZ901), myc (IZ702), and 6x-His (IZ801) epitope tags were created by ligating oligonucleotides encoding the tags into BamHI/EcoRI digested IX907.

Figure 23: pMVP Epitope Tags. (A) pENTR L5-L4 plasmids containing epitope tags for N-terminal fusions. (B-C) pENTR L3-L2 plasmids containing epitope tags for C- terminal fusions in (B) DNA based and (C) lentivirus vectors. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct.

4.2.2.5 Fluorescent Reporters

It is often desirable to be able to visualize virus-transduced cells or the cellular localization of a protein of interest. For example, as discussed in Chapter 3, β-cell specific transcriptional changes mediated by PDX1 overexpression can be diluted by the other cell-types, requiring the ability to isolate and analyze only the cells of interest. 99

Therefore, we included pMVP entry plasmids enabling eGFP fusion to either terminus of a gene (IO702, IR801) or linked as a separate reporter via a P2A “self-cleaving” linker

(IF801, JP801) or an IRES (HD201, Figures 24A-B). In addition, we also ligated oligonucleotides encoding an OLLAS epitope tag prior to the IRES sequence (AfeI/XhoI sites of HD201) enabling a C-terminal OLLAS fusion as well as an IRES-eGFP reporter

(IX702). For lentiviruses, mCherry-WPRE (JG202) or P2A-mCherry-WPRE (JG102) entry plasmids were generated for C-terminal fusions (Figure 24C).

Figure 24: pMVP Fluorescent Protein Fusions and Reporters. (A) pENTR L5-L4 plasmids containing eGFP or eGFP-P2A for N-terminal fusions. (B) pENTR L3-L2 plasmids containing eGFP or eGFP-P2A for C-terminal fusions as well as IRES-eGFP. pENTR L3-L2 plasmids for C-terminal P2A-mCherry or mCherry fusion from lentiviral vectors. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct.

4.2.2.6 Degron Technologies

Temporal control of transgene expression can be an elegant way to manipulate an experimental model. We therefore included Tet-On (Hayes et al. 2016), TMP

(Iwamoto et al. 2010) and SHLD1 (Banaszynski et al. 2006) degron domain stabilization, and AID (Holland et al. 2012, Natsume et al. 2016) technologies (see overview in Figure

25A) for temporal transgene control in pMVP. The widely implemented Tet-On system,

100

discussed in section 3.2.2.1, utilizes a constitutive promoter to drive expression of the

Tet-On transactivator (TETa), which upon doxycycline (Dox) addition, binds to and activates expression from TRE (Tet-response element) promoters. Conversely, degron domain technology requires fusion of a synthetic degron domain derived from either

FKBP12 (108 amino acids) (Banaszynski et al. 2006) or ecDHFR (159 amino acids)

(Iwamoto et al. 2010) to cause rapid protein degradation. Addition of a small molecule ligand (Shld1 or TMP respectively) stabilizes the degron domain and induces transgene protein accumulation. The AID system requires fusion of an AID tag (45-229 amino acid) in addition to constitutive overexpression of an auxin perceptive F-box plant protein, osTIR1, that interacts with the SKP1 protein of the SCF ubiquitin ligase complex

(Holland et al. 2012). Under basal conditions, AID-tagged transgenes are overexpressed; however, addition of the plant phytohormone auxin triggers rapid transgene degradation through interaction with osTIR1 and subsequent ubiquitination by the SCF complex.

We synthesized the FKBP and ecDHFR degrons based on published sequences

(Banaszynski et al. 2006, Iwamoto et al. 2010) preceded by the Xenopus 5’-beta-globin leader and followed by a GGGGS flexible linker, a maneuver that can prevent the degron from interfering with the function of the gene of interest (Chen et al. 2013b). We then subsequently generated pENTR L5-L4 plasmids II901 (FKBP) and II801 (ecDHFR) to enable N-terminal fusions (Figure 25B). The fusion of these degrons and intervening

101

Gateway “scar” sequence to a gene will cause an increase in the overall protein size by

153 amino acids for FKBP and 204 amino acids for ecDHFR.

For incorporation of the AID system we took a different approach. The current systems utilizing AID require the generation of a osTIR1 expressing stable cell line that is subsequently used as the foundation to introduce AID-fusion proteins and perform experiments (Holland et al. 2012, Natsume et al. 2016). For our approach, we engineered an all-in-one bicistronic vector capable of simultaneous delivery of both osTIR1 and an

AID-fused gene of interest separated by a P2A “self-cleavage” sequence to enable simultaneous expression of independent proteins. In addition, several publications have sought to define and improve the minimal AID sequence, ranging in size from 45-229 amino acids, so it was important to design an approach that was permissive to easy adaptation of the AID sequence. To accomplish this, we first generated a pENTR L5-L4 plasmid (JD904) containing osTIR1, a nuclear export signal to enable efficient degradation of nuclear proteins, a C-terminal 9x myc epitope tag, and a unique KpnI restriction site at the 3’ end of the sequence. This was performed by PCR amplification using Addgene plasmid 47328 (Holland et al. 2012) as a template for the 2.2 kb product.

We then synthesized a recently reported 68 amino acid minimal AID sequence that maintained robust activity in mammalian cells (Natsume et al. 2016) and used

NEBuilder HiFi DNA Assembly to insert it into KpnI digested JD904 to create JE301

(Figure 25B). In total, utilization of this construct to create an AID fusion will increase

102

the overall size of the protein of interest by 82 amino acids. As a final note, while the

AID technology has proven to be robust, private discussions with the creator, Dr.

Masato Kanemaki, National Institute of Genetics Japan, has revealed that the system is sensitive to the level of AID-fused transgene expression, such that excessive cellular expression often seen in transient transfections or adenovirus transduction can overwhelm the system to yield unsatisfactory results. Therefore, this system is best utilized through the creation of a stable cell line with a minimal number of integration events to limit transgene expression.

103

Figure 25: pMVP Inducible Systems. (A) Graphical depiction of the mechanisms for the Tet-On, FKBP/ecDHFR degron domain, and AID systems. The Tet-On system has a transcriptional mechanism whereby upon Dox addition the TETa binds and activates the TRE promoter. The FKBP and ecDHFR degrons stimulate transgene protein turnover, this is blocked by addition of the ligands Shld1 and TMP, respectively. Conversely, transgene protein is overexpressed in the basal condition of the AID system, but upon addition of auxin the transgene protein is ubiquitinated through an association with osTIR1 and the SCF complex resulting in rapid transgene degradation. (B) pENTR L5-L4 plasmids containing components of the FKBP, ecDHFR, and AID degron technologies available for N-terminal fusion to a gene of interest. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct.

4.2.2.7 Tools for Protein Interaction Partner Identification

The functions of many proteins are modulated by interactions with other proteins. In recent years, technologies have been developed that utilize a promiscuous 104

biotin ligase (BioID2) (Kim et al. 2016) or an engineered peroxidase (APEX2) (Lam et al.

2015) to cause the biotinylation of proteins that interact with, or are in close proximity to, a protein of interest in living cells. This biotinylation allows the interacting proteins to be isolated using streptavidin and identified using protein mass spectrometry. We added this capability to pMVP through the addition of the BioID2 and APEX2 tools as either N- or C-terminal fusion elements (Figures 26). The BioID2 sequence from Addgene plasmid

74223 (Kim et al. 2016) was synthesized with a Xenopus 5’ beta-globin leader and followed by a 3x GGGGS flexible linker and recombined into pDONR221 P5-P4 to create

(IO901). Importantly, the 3x GGGGS flexible linker adds approximately 5 nm to the

“reach” of BioID2. A pENTR L5-L4 plasmid for APEX2 (IF701) was generated by PCR amplification using Addgene plasmid 49386 (Lam et al. 2015) as a template. IO901 and

IF701 were then used as templates to create PCR products for insertion into

BamHI/EcoRI digested IA001 or IX907 to create polyA or WPRE containing pENTR L3-

L2 plasmids for fusion to the C-terminus of a gene.

Figure 26: pMVP Components for Interaction Partner Identification. The promiscuous biotin ligase BioID2 and the engineered peroxidase APEX2 are available in (A) pENTR L5-L4 plasmids for N-terminal fusions and pENTR L3-L2 plasmids for C- terminal fusions in (B) DNA based vectors and (C) lentiviral vectors. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct.

105

4.2.3 Generation of Custom Gateway Destination Vectors

Unfortunately, there are very few promoterless Gateway destination vectors that are compatible with pMVP. Therefore, to expand the utility of pMVP, we created a set of

18 custom promoterless destination vectors for transient (adenovirus, expression plasmid) and stable (lentivirus, Sleeping Beauty transposon) transgene expression

(Figures 27). This wide range of pMVP-compatible destination vectors enables the same pMVP components to be used to generate vectors for multiple experimental paradigms through the rapid 3-day process outlined above (Figure 19A). While we have filled the current platform with a robust collection of vector options, it is important to note that pMVP can be expanded to include additional destination vectors via strategies outlined below.

106

Figure 27: pMVP Custom Destination Vectors. Custom Gateway destination vectors compatible with pMVP were created to enable recombination into (A) adenoviral, (B) expression, (C) lentiviral, and (D) Sleeping Beauty transposon destination vectors. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct.

4.2.3.1 Expanded Capacity Adenovirus Vector

Experimental design can be limited by vector insert limits. Despite the relatively permissive insert size limits for adenovirus (currently 7.5 kb), several of the multi- component options described above demanded additional cloning space. We solved this problem by engineering of an expanded capacity adenovirus vector (pMVP/Ad-

DEST) capable of accommodating an insert up to 8.4 kb by deleting ~750 bp of extraneous Ad5 E3 sequence from pAd/PL-DEST to mirror the E3 deletion found in pAdEasy-1 vectors (He et al. 1998). Manipulation of the large 35.4 kb adenovirus genome is challenging using traditional cloning methods and has been a major limiting factor in the generation of new vectors. Therefore, we designed a novel approach for

107

adenoviral vector manipulation based around the utilization of the new cloning technique NEBuilder HiFi DNA Assembly, a derivative of the recently developed

Gibson Assembly (Gibson et al. 2009). As depicted in Figure 28, this method was utilized to replace the SrfI/BarI fragment of pAd/PL-DEST with a custom designed sequence containing the desired E3 deletion. In addition, the replacement sequence was engineered to contain a unique SgrDI restriction site, thus creating a straightforward scheme for insertion of transgenes and other elements into the former E3 region using

DNA assembly (Figure 28). To put this novel adenoviral vector in perspective, the theoretical maximum gene insert size for the pJM17 adenovirus platform is 3.8 kb and for the pAdEasy-1 system is ~5 kb gene, noting that both platforms already contain a promoter and polyA signal. In contrast, after accounting for a promoter and polyA signal, our pMVP/Ad-DEST vector can readily accommodate a gene up to 7.5 kb in size representing a 100% and 50% increase over the gene insert capacity of pJM17 and pAdEasy-1, respectively.

108

Figure 28: Creation of pMVP/Ad-DEST. Cloning strategy for deletion of ~750 bp from pAd/PL-DEST Ad5 E3 region to create pMVP/Ad-DEST. (A) LR Clonase II was used to recombine pAd/PL-DEST with a eGFP entry plasmid (HJ201) to remove a SrfI restriction site. (B) The subsequent plasmid (IP002) was digested with SrfI/BarI and NEBuilder HiFi DNA assembly was used to insert a modified Ad5 sequence (JH465) replacing non-essential E3 sequences with a unique SgrDI restriction site creating IP901. (C) BP Clonase II was used to reintroduce the Gateway cloning cassette from pDONR221 into IP901, generating (D) pMVP/Ad-DEST (IQ013). 109

4.2.3.2 Fiber-Modified Adenovirus Vectors

One shortcoming of Ad5 vectors is their reliance on the coxsackievirus and adenovirus receptor (CAR), which limits their ability to transduce cell types with low or absent CAR expression. Other groups have identified that inclusion of a RGD motif and/or polylysine (pK) modifications in the H1 loop of the Ad5 fiber knob region or creating chimeric fiber proteins containing components of multiple adenovirus serotypes are sufficient to enhance CAR-independent transduction (Figure 29) (Reynolds et al. 1999, Shayakhmetov et al. 2000, Wu et al. 2002, Gaggar et al. 2003, Bramson et al.

2004). To expand the utility of pMVP to models relevant to metabolic disease but resistant to transduction with Ad5, such as C2C12 myoblasts and 3T3-L1 cells, we sought to incorporate fiber-modified adenovirus vectors. Due to the complexity of manipulating large 35 kb adenovirus vectors and based on our successful implementation of DNA assembly in lieu of traditional cloning to engineer pMVP/Ad-

DEST, we developed a novel cloning pipeline that allows for fiber gene modification and subsequent reinsertion back into the adenoviral vector using DNA assembly (Figure 30).

In brief, NEBuilder HiFi DNA Assembly was used to subclone the 1.8 kb BarI/BstBI Ad5 fiber gene fragment and 2 linker gBlocks (JH512/JH513) into a KpnI/SacI digested pBlueScript2SK(-) shuttle plasmid, simplifying the cloning process for any subsequent fiber modifications. The resulting shuttle plasmid, pBS/Fiber, was engineered to contain restriction sites for the Type IIS enzyme BsmBI, so that upon digestion the modified fiber

110

gene fragment is released with sufficient flanking Ad5 sequence for scarless reassembly into a BarI/BstBI digested pMVP/Ad-DEST vector. This pipeline was used to generate adenovirus vectors containing the RGD, pK, or RGD/pK fiber modifications in addition to a chimeric Ad5/35 fiber gene (Figure 27).

Figure 29: Fiber Gene Modification Alters Adenovirus Tropism. The Ad5 fiber is comprised of a shaft and knob region that normally binds to the CAR receptor. Insertion of a RGD motif into the H1 loop of the knob region serves to broaden the tropism of the adenovirus such that it is able to bind to integrins, in addition to the CAR, enabling efficient transduction of CAR-negative cells. The Ad5/35 modification is a chimeric fiber protein comprised of the Ad5 base, for incorporation in the virion, attached to the Ad serotype 35 shaft and knob region resulting in an adenovirus capable of transducing only CD46-positive cells.

111

Figure 30: Pipeline for Adenovirus Fiber Modification. Overview of the pipeline for engineering of a shuttle vector for the Ad5 fiber gene (L5) to enable fiber modification and subsequent reinsertion into an adenoviral vector. (A) NEBuilder HiFi DNA Assembly was used to subclone the 1.8 kb BarI/BstBI Ad5 fiber gene fragment from IQ013 and 2 linker gBlocks (JH512/JH513) into a KpnI/SacI digested pBlueScript2SK(-) shuttle plasmid. (B) The resulting shuttle plasmid, pBS/Fiber, contains restriction sites for the Type IIS enzyme BsmBI, (C) so that upon BsmBI digestion, the modified fiber gene fragment is released with sufficient flanking Ad5 sequence for scarless reassembly into BarI/BstBI digested IQ013, (D) regenerating the adenoviral vector that now contains a modified fiber gene. 112

4.2.3.3 Expression Plasmid Vectors

Despite the expansion of vector technologies over the past 30 years, a workhorse of molecular biology research remains the simple transient overexpression of a gene of interest from an expression plasmid vector. However, many expression plasmid vectors contain unnecessary components that increase overall vector size and can lead to decreased transfection efficiency. To engineer a compact expression plasmid vector compatible with pMVP, we ligated a Gateway cloning cassette into EcoRV digested pBlueScript2SK(-) (pMVPBS-DEST, IR301) and subsequently removed all non-essential plasmid components by annealing oligonucleotides JH597/JH598 and JH595/JH596 into

XhoI/PsiI/SapI/SacI digested IR301 to create pMVP-DEST (JK104). We then utilized

NEBuilder HiFi DNA Assembly to introduce an SV40 promoter, TK polyA signla, and selection markers for blasticidin (JK305), puromycin (JL302), eGFP (JL002), or mCherry

(JL202) into XhoI digested JK104 to facilitate multiple use modalities (Figure 31).

113

Figure 31: Generation of pMVP-DEST Vectors. Schematic representation of the process used to generate pMVP-DEST expression plasmid vectors. (A) A commercial Gateway Cloning Cassette was ligated into the EcoRV restriction site of pBluescript. (B) This generated that Gateway destination vector pMVPBS-DEST (IR301). Extraneous sequences were removed from IR301 by replacement of the PsiI/XhoI and SapI/SacI fragments with the annealed oligonucleotides JH597/JH598 and JH595/JH596, respectively. (C) NEBuilder HiFi DNA Assembly was used to insert selection markers into the XhoI restriction site of resulting plasmid, pMVP-DEST (JK104) to create (D) multiple variants of this vector. 114

4.2.3.4 Lentivirus Vectors

A third-generation lentivirus vector conferring resistance to blasticidin (IY601) was generated by sequential replacement of the ClaI/SpeI fragment (CMV promoter) with oligonucleotides JH547/JH548 (creating plasmid IY503) and the XhoI/SpeI fragment

(V5 epitope tag and WPRE element) with JH545/JH546 from pLenti6.3/V5-DEST.

Additionally, we used DNA assembly to replace the PmlI/BlpI fragment containing the blasticidin resistance gene with a neomycin resistance gene (IZ401) as well as ligated oligonucleotides JH549/JH556 into XhoI/BlpI digested IY503 to remove the eukaryotic selection cassette entirely (IZ501) (Figure 32). While we were able to successfully create active lentiviruses using these vectors (data not shown), we observed a significant number of unwanted recombination events during propagation of the final vector.

Lentiviral vectors are notoriously prone to homology-dependent recombination due to the long terminal repeats (LTRs) required for integration, likely amplified here by the inclusion of multiple Gateway sites that exhibit moderate homology to one another.

Others have recently published a Multisite Gateway Pro-based lentiviral system for creation of multigene Cas9-targeting vectors and demonstrate that they exhibit no increased propensity for recombination (pMuLE) (Albers et al. 2015) suggesting that modification of the E. coli strain or culture conditions may be sufficient to overcome this issue. To this point, we were able to increase our success in isolating intact lentiviral vectors through the use of the Stbl3 E. coli strain as well as minimizing incubation time

115

for transformations and subsequent cultures to approximately 12 hours. Another common approach is to lower the culture temperature from 37°C to 28°C-32°C but we did not find that this alteration alone was sufficient to prevent the unwanted recombination events in lentiviral vectors.

116

Figure 32: Creation of pMVP/Lenti-DEST Vectors. (A) To create a promoterless, pMVP-compatible lentivirus vector, the ClaI/SpeI fragment of pLenti6.3/V5-DEST was replaced with the annealed oligonucleotides JH547/JH548. (B) The resulting IY503 vector was digested with XhoI/BlpI or XhoI/AgeI and ligated with oligonucleotides to replace (C) the entire selection marker cassette (IZ501) or (D) the V5-WPRE sequence (IY601). IY601 can subsequently be digested with BlpI/PmlI to excise the blasticidin resistance gene and insert an alternative selection maker. 117

4.2.3.5 Sleeping Beauty Transposon Vectors

For many decades a standard approach for creation of stable cell lines was by transfection of an expression vector containing a gene of interest linked to a eukaryotic selection marker to screen for rare recombination events. After weeks of selection and outgrowth, colonies were selected and screened for transgene expression. This process is highly inefficient due to the large number of false-positive clones that integrate the selection cassette but not a functional transgene expression cassette and is further complicated by unstable long-term expression due to the integration of multimers of the transgene cassette that are prone to epigenetic silencing. In most research settings, this approach has been replaced by the use of retroviruses (e.g. lentiviruses) that are able to faithfully integrate the entirety of the transgene expression cassette in the majority of transformed cells, precluding the necessity for creation of a clonal population. However, lentiviruses and other retroviruses do have significant drawbacks, including preferential insertion into highly active gene bodies (Schroder et al. 2002, Wu et al. 2003) as well as the biosafety concerns that arise from handling active viral particles that exhibit a wide tropism and result in stable transgene integration. To circumvent these issues, researchers have started to more widely adopt transposon technologies as a method for transgene delivery. This change has been partially spurred on by the invention of a hyperactive Sleeping Beauty (SB) synthetic transposase that is able to mediate transposition of foreign DNA into a host genome with an efficiency approaching that of

118

lentiviruses with none of the biosafety concerns of a virus-based vector (Mates et al.

2009). Furthermore, the SB transposon system results in the stable incorporation of foreign DNA into a host genome in a random manner dictated only by the presence of a

TA dinucleotide as illustrated in Figure 35A, with no preference for active genomic elements (Huang et al. 2010). Critically, despite being plasmid-based, SB faithfully transfers the entirety of the gene expression cassette into the host genome, unlike plasmid integration techniques.

We engineered a pMVP-compatible SB destination vector (pMVP/SB-DEST,

JN607) by using DNA assembly to insert synthesized SB ITR elements (gBlocks JH620,

JH621) into the SpeI and BsaXI restriction sites located on either side of the pMVP-DEST

Gateway cloning cassette. In the same manner as for expression vectors, the selection markers (blasticidin (JN202), puromycin (JL302), eGFP (JL002), mCherry (JL202)) were then inserted in the XhoI restriction site located between the attR2 site and adjacent 3’ SB

ITR (Figure 33) of pMVP/SB-DEST. Critically, an expression cassette recombined into these vectors can be readily transposed into genomic DNA of a host cell via co- transfection with the SB100X transposase expression plasmid (Addgene plasmid 34879

(Mates et al. 2009)).

119

Figure 33: Creation of pMVP Compatible Sleeping Beauty Transposon Vectors. (A) NEBuilder HiFi DNA Assembly was utilized to insert custom gBlocks containing the Sleeping Beauty inverted terminal repeats (ITRs) into BsaXI/SpeI digested pMVP-DEST (JK104). (B) Selection markers were inserted into the XhoI site of pMVP/SB-DEST (JN607) to create (C) selectable vectors.

120

4.2.4 pMVP Validation

4.2.4.1 Epitope Tags and Fluorescent Reporters

To demonstrate the broad flexibility of pMVP, we inserted the cDNA encoding the human homeobox transcription factor PDX1 into the platform via PCR as described in Chapter 4.2.2.2. Using pMVP, we generated 9 different PDX1 vectors containing N- terminal epitope tags combined with eGFP reporters, or C-terminal eGFP/epitope tag fusions, recombined into adenovirus or expression vectors, respectively (Figures 34A-B).

Immunoblot analysis confirmed overexpression of the desired PDX1 variants upon adenoviral transduction of INS1 832/13 cells (Figure 34C) or expression plasmid transfection into HEK293 cells (Figures 34D-F). Additionally, a Sleeping Beauty transposon vector containing eGFP fused to the N-terminus of PDX1 was co-transfected into INS1 832/13 cells with the SB100x Sleeping Beauty transposase plasmid. Flow cytometric analysis of the polyclonal cells cultured for an additional 6 weeks after selection revealed > 95% of the population was eGFP positive (Figure 35).

121

Figure 34: Utilization of pMVP for the Creation of Unique PDX1 Vectors. A pENTR plasmid containing a cDNA for human PDX1 was recombined with pMVP components to generate an assortment of PDX1-expressing (A) adenovirus and (B) expression vectors. (C) Immunoblot blot analysis of INS1 832/13 cell lysates harvested 48h after treatment with crude adenovirus lysates. (D-F) Immunoblot analysis of HEK293 cell lysates 24h after transfection of expression vectors.

122

Figure 35: Creation of a Sleeping Beauty-Derived eGFP-PDX1 Stable Cell Line. (A) An overview of Sleeping Beauty (SB)-mediated genomic integration at a TA dinucleotide and the vector utilized to create an INS1 832/13 cell line stably expressing an eGFP-PDX1 fusion protein. Six weeks after removal of blasticidin selection, live-cell flow cytometry was performed on the resulting (B) polyclonal cell line (INS1 1353/pc) to measure eGFP expression compared to the (C) parental INS1 832/13 cell line.

4.2.4.2 Modes of Inducible Transgene Regulation

As described in Chapter 4.2.2.6, pMVP includes an array of technologies that enable controlled accumulation or degradation of a target protein by addition of exogenous factors. To demonstrate the utility of these tools, we generated adenoviruses that directed β-cell specific, inducible PDX1 expression in response to Dox, TMP, or

Shld1. We also generated a Sleeping Beauty vector to create an INS1 832/13-derived cell

123

line stably expressing TIR1-P2A-AID-PDX1 (Figure 36A). As expected, INS1 832/13 cells transduced with the adenoviral constructs induced expression of PDX1 upon addition of

Dox, TMP, or Shld1 and this expression was reversed upon removal of the respective inducers (Figure 36A). Additionally, the level of transgene expression could be controlled by varying the dose of inducer added to the system (Figure 36C). Finally, we evaluated the dynamics of the AID system through addition of auxin to our AID-PDX1 expressing cells. Remarkably, overexpressed human PDX1 protein was almost completely degraded within 15 minutes of auxin addition, whereas endogenous rat

PDX1 levels remained unchanged. (Figure 36D).

124

125

Figure 36: Demonstration of pMVP Inducible Systems. (A) Graphical depiction of the pMVP vectors utilized. (C) Immunoblot analysis of INS1 832/13 cell lysates transduced with crude lysates for adenoviruses expressing PDX1 using either the Tet-On, ecDHFR, or FKBP systems and harvested at the indicated time intervals after addition (induction) and removal (washout) of the respective Dox, TMP, or Shld1 inducers. Endogenous PDX1 and PDX1-fusion bands are labeled accordingly. (D) Immunoblot analysis of INS1 832/13 cell lysates transduced with crude lysates for adenoviruses expressing PDX1 using either the Tet-On or ecDHFR systems and harvested 24h after addition of a dose curve of Dox (0.5ng/ul-200ng/ul) or TMP (1nM-5uM), respectively. Endogenous PDX1 and PDX1-fusion bands are labeled accordingly. (E) Immunoblot analysis of cell lysates from an INS1 832/13-derived cell line made by Sleeping Beauty- mediated integration of Tir1-P2A-AID-PDX1. Lysates were harvested at the indicated time intervals after treatment with 500uM auxin. Bands representing endogenous PDX1 and AID-PDX1 are labeled accordingly.

4.2.4.3 Fiber-Modified Adenoviruses

To validate that our newly created fiber-modified adenoviral vectors were functional as well as able to transduce cells resistant to Ad5 transduction, we recombined the necessary elements for eGFP expression into a wild-type fiber vector as well as vectors containing the RGD, pK, RGD/pK, or Ad5/35 modified fiber gene.

Importantly, we were able to successfully create active adenoviral lysates from all of these vectors. However, while we were able to subsequently purify a highly concentrated RGD-modified adenovirus, we could not do the same for the other modifications despite the HEK293 cells exhibiting every hallmark consistent with successful adenovirus amplification. Private discussions with Dr. Peter Holst, University of Copenhagen, revealed that this is likely due to a propensity of these fiber modifications to cause precipitation of the virus when concentrated. His lab has

126

successfully overcome this issue by altering the buffers used during purification but we have not yet attempted to purify these vectors using this new approach. As such, we only moved forward with testing the transduction of C2C12 cells using equivalent viral particles of purified IW902 (wild-type fiber) and JC701 (RGD modified fiber). As seen in

Figure 37, this experiment clearly demonstrated that the RGD fiber modification dramatically enhanced transduction of these cells compared to the wild-type fiber vector.

Figure 37: RGD-Modified Ad5 Fiber Enables Efficient Transduction of C2C12 Myoblasts. (A) eGFP expressing adenoviral vectors were generated in either Ad5 fiber or Ad5/RGD fiber vectors. Equivalent doses of purified (B) wild-type fiber and (C) RGD- modified fiber adenovirus were added to confluent C2C12 cells and imaged after 48h. (C2C12 cell experiments were performed by Dorothy Slentz, unpublished).

127

4.3 Discussion

This work has conceived, generated, and implemented a modular cloning platform, pMVP, capable of rapidly producing a broad array of unique adenoviral, lentiviral, expression, and Sleeping Beauty transposon vectors for the purposes of transgene expression and RNAi. In total, this system offers ~90 unique components that provide researchers with broad flexibility to create > 8,000 novel vectors for any given gene to fulfill a host of experimental demands with markedly less effort than traditional cloning methods. As a demonstration, pMVP was utilized to generate an assortment of unique vectors for PDX1 overexpression, featuring different combinations of promoters, epitope tags, fluorescent reporters, and modes of transgene inducibility expressed from recombinant adenoviruses, expression plasmids, and Sleeping Beauty-derived stable cell lines.

We are not the first to envision the benefits of a modular cloning platform. In fact, the original iteration of Gateway cloning (Hartley et al. 2000), in which a single element can be rapidly recombined into a vector, has been widely utilized as a modular system over the past 15 years. As such, the creation of pENTR221 plasmids containing the cDNAs representing the entire collection of protein coding genes in the human genome has been a goal of the ORFeome collaboration (Lamesch et al. 2007), resulting in

> 10,000 cDNAs being available in pENTR221 plasmids through the DNASU (Arizona

State University) and PlasmID (Harvard University) repositories. Unfortunately, a

128

shortcoming of pMVP is that neither these repositories, nor any others, currently utilize the pENTR R4-R3 plasmid as ORFeome framework, thus, for the moment any genes of interest need to be incorporated into pMVP in the manner discussed in Chapter 4.2.2.2.

Conversely, the pENTR221-based ORFeome repositories are limited in their application by the lack of availability of Gateway Destination vector options containing the cell-type specific promoters, epitope tags, inducible systems, fluorescent reporters, and other technologies that we have incorporated into pMVP. For example, an open cDNA cloned into pENTR221 would require the creation of approximately 6,000 destination vectors in order to have the same flexibility for transgene expression provided from the 90 components of pMVP, excluding RNAi. This would represent an inordinate amount of time and effort, as seen by the effort required to generate only 18 custom destination vectors detailed in Chapter 4.2.3. As our understanding of the etiology of disease continues to progress, researchers will continue to be forced to create and implement increasingly complex genetic engineering strategies to better understand the nuances of cellular dysfunction. With the creation of pMVP, we have presented the modern researcher with an arsenal of tools that enable them to rapidly generate innovative vectors for exploration of their experimental model, allowing them to focus on unraveling the complex questions they seek to answer instead of being forced into an experimental tunnel by vector inflexibility and mired in the woes of vector construction.

129

5. Identification of PDX1-Regulated Secreted Factors Using pMVP-Derived Vectors 5.1 Introduction

As discussed in Chapter 3, a major goal of diabetes research is the identification of pathways sufficient to drive pancreatic islet β-cell replication to compensate for the β- cell loss observed in the progression of both major forms of diabetes. In our lab, we have discovered that overexpression of either of the homeodomain transcription factors PDX1 or NKX6.1 is sufficient to significantly increase β-cell replication in primary rat islets, albeit through independent pathways (Schisler et al. 2008, Hayes et al. 2013). While further studies have been able to identity the transcriptional network responsible for

NKX6.1-mediated replication (Schisler et al. 2008, Tessem et al. 2014, Hobson et al. 2015,

Ray et al. 2016), PDX1 has proven to be more challenging. As previously discussed, this is likely due to the multifaceted modes of cellular replication being induced by PDX1 that result in both β-cell and α-cell replication, potentially involving both cell intrinsic and extrinsic pathways. To this point, recent experiments have discovered that PDX1- driven rat β-cell replication is mediated by a secreted factor(s) (Hayes et al. 2016) and, importantly, that co-culture of PDX1-expressing islets from young rats with untreated human islets is sufficient to induce human β-cell replication (Lu Zhang, unpublished).

From a therapeutic standpoint, secreted factors could be harnessed as new pharmaceutical agents for controlled expansion of functional -cell mass in diabetes.

However, the identification of secreted factors mediating a phenotype remains a 130

challenging prospect for which the available tools and strategies are still being developed. Although others in the lab had performed microarray analysis with CMV- driven PDX1 overexpression prior to me joining the group (Hayes et al. 2013), these left us unable to discern “activating” versus “response” transcriptional networks in whole islet cultures.

In an effort to identify the factor(s) responsible for PDX1-mediated human β-cell replication, we developed a new strategy based on isolation of the PDX1-overexpressing cells for RNAseq, coupled with a bioinformatics strategy to identify all putative secreted factors regulated by PDX1. To focus on the β-cell transcriptional network modulated by

PDX1, we utilized pMVP to generate novel β-cell specific bicistronic vectors containing

PDX1 and an eGFP fluorescent tracker under control of the rat insulin promoter (RIP), enabling the isolation of rat β-cells overexpressing PDX1 from islets for RNAseq analysis. To identify PDX1-regulated secreted genes, we culled the RNAseq list by using publicly available databases of secreted proteins generated by experimental observation and computational prediction. Importantly, this approach yielded 107 candidate secreted factors that were significantly induced in β-cells overexpressing PDX1 versus β- cells overexpressing either βGAL or NKX6.1.

Contributions: The work to identify PDX1-regulated secreted factors represents a collaborative effort within the Newgard Laboratory. Study design, adenoviral vector creation, FACS-mediated isolation of transduced rat islet β-cells, and secondary

131

bioinformatic analysis of RNAseq data to identify secreted factors were performed by

Jonathan Haldeman. RNAseq was performed in collaboration with Jason Gibson and Dr.

Simon Gregory. The primary RNAseq analysis was performed by Dr. David Corcoran

(Duke Bioinformatics Core). Rat/human islet co-culture studies and on-going phenotypic analysis of identified secreted factors were performed by Dr. Lu Zhang.

5.2 Results

5.2.1 pMVP Vectors Enable the Isolation of Transduced β-Cells

Due to the seemingly multilayered biology underlying PDX1-mediated replication, we desired to limit the variables in our model by selectively overexpressing

PDX1 in β-cells. To drive β-cell specific expression, we utilized the previously characterized rat insulin promoter (RIP) (cite). In order to validate the specificity of this element as cloned in pMVP and the feasibility of the overall approach before moving forward, we generated adenoviruses containing either a RIP-mCherry or CMV-eCFP expression cassette and utilized fluorescent activated cell sorting (FACS) to isolate mCherry or eCFP expressing cells from rat islets. As seen in Figure 38, the isolated cells expressing mCherry exhibited a 2-fold enrichment of the β-cell genes INS1 and INS1, and also exhibited > 95% depletion of α-, γ-, and ε-cell associated genes when compared to the eCFP positive cells. While we observed only a partial depletion representing 75% of the δ-cell gene SST, this has also been seen in mouse models utilizing the mouse insulin promoter to drive eGFP (Dorrell et al. 2011). This is likely due at least in part to

132

PDX1-driven transactivation of the insulin promoter since PDX1 is also expressed in δ- cells (DiGruccio et al. 2016). Since δ-cells represent only 5%-10% of the overall islet cell population, we decided that these results were acceptable to move forward.

Importantly, these experiments also demonstrated that we were able to isolate sufficient quantities of primary rat β-cells for RNAseq analysis (data not shown).

Figure 38: pMVP Vectors Enable Isolation of β-Cells. qRT-PCR gene expression analysis performed on RNA from FACS isolated rat islet cells. The genes represent major transcripts for β- (INS1, INS2), α- (GCG, RBP4, ARX), δ- (SST), γ- (PPY), and ε-cells (GHRL) which, account for the endocrine cell population of the islet. Data represent mean ± S.E.M. of relative expression versus Ad-CMV-eCFP from 3 independent experiments.

Having verified the specificity of the RIP element, we utilized pMVP to generate bicistronic adenoviruses expressing either βGAL, mouse PDX1, or hamster NKX6.1, all in the context of an eGFP reporter gene linked via an IRES. In addition, the incorporated

PDX1 and NKX6.1 cDNAs included an HA epitope tag, as this vector was created with an early iteration of pMVP that did not yet incorporate epitope tags. While not used in the studies described hereafter, the inclusion of an HA epitope tag will allow future

133

ChIPseq studies to identify genes to which overexpressed PDX1 or NKX6.1 are localizing to drive their respective phenotypes, while removing any signal interference from their endogenous counterparts.

To validate these vectors, we transduced primary rat islets with either the βGAL or PDX1-expressing adenoviruses and isolated the eGFP positive cells using FACS.

Confirming our Ad-RIP-mCherry results, we observed a depletion of the α-cell gene

GCG in our eGFP-positive population when compared to the eGFP-negative cells

(Figure 39A). As expected, PDX1 was specifically overexpressed in the eGFP-positive population and this correlated with trends of increased CCND1 and CCND2, matching what we observe in intact islets (Figures 39B-D).

134

Figure 39: A pMVP-Derived Bicistronic PDX1/eGFP Adenovirus Permits Selective Isolation of PDX1-Expressing β-Cells. FACS was performed 48h after rat islet transduction with bicistronic adenoviruses using the β-cell specific RIP to express either βGAL/eGFP or PDX1/eGFP. RNA from both the positive (+) and negative (-) eGFP cell populations were utilized for qRT-PCR analysis of (A) GCG, (B) PDX1, (C) CCND1, (D) CCND2. Data represent mean ± S.E.M. from 2 independent experiments.

5.2.2 Identification of Secreted Factors Specifically Regulated by PDX1

Having validated the vectors chosen for study, we next isolated rat β-cells overexpressing either βGAL, PDX1 or NKX6.1 and subsequently performed RNAseq on each population from 3 independent experiments. To identify all the putative secreted factors regulated by either PDX1 or NKX6.1, we developed a bioinformatics approach in

135

which the 16,505 genes identified by RNAseq were filtered to identify gene(s) predicted to be secreted by secretome databases. For this purpose, we utilized MetazsecKB, a secretome database generated through the integration of multiple different bioinformatics approaches to predict subcellular localization (Meinken et al. 2015), as well as a database developed by the Human Protein Atlas (proteinatlas.org) (Uhlen et al.

2015). These filters allowed us to identify 1,951 putative secreted factors that were detected by RNAseq. Additional criteria were sequentially applied to restrict the list of putative secreted factors to those significantly modulated by PDX1 versus βGAL (615 genes) and induced at least 1.5-fold by PDX1 overexpression versus both βGAL and

NKX6.1 (218 genes). Finally, since we know that PDX1 and NKX6.1 drive β-cell replication through independent pathways (Hayes et al. 2013) and that NKX6.1 overexpression exerts its effects on replication via a -cell intrinsic rather than soluble factor mechanism (Lu Zhang, unpublished), we filtered these 218 genes to exclude genes that are also significantly increased by > 1.5-fold by NKX6.1 versus βGAL. This approach yielded 107 putative secreted factors that are significantly induced by PDX1 versus the other treatments (Figure 40A). Importantly, only 10 of these factors were previously identified by our microarray approach (Hayes et al. 2016), demonstrating the new depth of data achieved with our method. These putative secreted factors have been further characterized based on whether they have any other distinguishing characteristics (proteases, ECM components, present in exosomes, etc) or are unlikely to

136

be secreted (other) (Figure 40B). We used this information, known functions described in the literature, and qRT-PCR analysis of expression in islets to prioritize these candidate genes. At that point, a post-doctoral fellow in our lab, Dr. Lu Zhang, has begun follow-up functional experiments with a subset of 16 factors yielding two candidate secreted factors involve in -cell (BMP6) and β-cell (PDGF-A ligand) replication. These studies are being finalized for publication, and I will be a co-author of this work.

137

Figure 40: Identification and Characterization of PDX1-Induced Secreted Factors. (A) The total genes detected by RNAseq were sequentially filtered to identify genes encoding putative secreted factors that are uniquely significantly induced by PDX1 overexpression. The resulting 107 genes were (B) broadly classified based on the literature and additional protein localization tools as either a transmembrane protein, receptor, exosome component, protease, extracellular matrix component, a general secreted protein, or as “other” for proteins unlikely to be secreted.

138

5.3 Discussion

The premise of utilization of an endogenous soluble factor in a pharmacologic manner first emerged in the diabetes field in 1922 when Banting and Best discovered that injection of pancreatic extracts, containing what would later be named “insulin”, could control glycemia in diabetic human patients (Banting et al. 1922a). Discoveries in the more recent past have also ushered in new frontline therapies such as GLP-1 mimetics that exert their benefits through enhancement of β-cell function and other effects (Li et al. 2003). Recently, several others in the field have identified soluble factors emanating from the liver (El Ouaamari et al. 2016) as well as from young animals in parabiosis experiments (Salpeter et al. 2013) that are sufficient to drive β-cell replication.

However, a looming issue in the β-cell replication field is the lack of translatability of factors and pathways identified in rodent models to replication in human islets

(Kulkarni et al. 2012). Importantly, our recent work has discovered that the secreted factor(s) produced by PDX1 is capable of driving human β-cell proliferation (Lu Zhang, unpublished). To enable isolation of the relevant factors, we developed a new methodology for secreted factor identification based on computational prediction of secreted proteins to filter the transcriptional network of PDX1. There are limitations to this approach, in that unlike with mass spectrometry methods, we are not able to discern whether the modulated soluble factors are actually secreted.

139

In summary, we have applied pMVP to generate novel vectors explicitly designed for cell type-specific expression and identification of PDX1-induced soluble factors. This was coupled with a bioinformatics approach to identify soluble factors in an unbiased manner, yielding 107 putative secreted factors regulated specifically by

PDX1 overexpression. Experiments to examine the role of these factors in PDX1- mediated human - and β-cell replication are ongoing and have the potential to open the door to new biologics for the treatment of diabetes.

140

6. Creation of pMAGIC, a Modular Cloning Platform for Epigenetic Engineering 6.1 Introduction

The past decade has brought remarkable progress in our ability to uncover disease-associated genomic loci. Surprisingly, these efforts have revealed that a large portion of these genomic elements localize outside of protein coding regions, often in areas of active epigenetic modification, suggestive of a regulatory role (Maurano et al.

2012, Hnisz et al. 2013, Parker et al. 2013, Pasquali et al. 2014). These findings create a new challenge of ascribing a function to non-protein coding regulatory regions within their native cellular contexts. Recent publications have demonstrated that nuclease- dead Cas9 (dCas9) can be utilized as a scaffold, whereby addition of a guide RNA

(gRNA) enables promoter-specific recruitment of transcriptional and epigenetic modifiers. This recruitment has been shown to be sufficient to alter endogenous gene expression (Gilbert et al. 2013, Chavez et al. 2015), thereby providing a viable alternative to transgene overexpression and RNAi. Furthermore, these approaches can be applied to target epigenetic modulators to discrete endogenous regulatory elements, such as enhancers, resulting in an alteration of the epigenetic status and activity of the targeted element (Hilton et al. 2015, Kearns et al. 2015, Thakore et al. 2015). However, these approaches are limited by the fact that the relevant epigenetic engineering tools are spread across various cloning and vector platforms rather than being aggregated in a manner that maximizes their efficient application and enables delivery to disease- 141

relevant primary tissues, such as the pancreatic islets of Langerhans. Furthermore, fusion of dCas9 to large effectors, such as the 2.5 kb histone demethylase LSD1, results in genes that are too bulky to be packaged into most viral vectors, thus, further restricting their use.

To address these technological shortcomings, we used the framework established for pMVP to create a plasmid-based modular adenovirus for genomic interrogation with

Cas9 (pMAGIC). This platform was designed to contain 5 unique epigenetic engineering modalities [VPR (Chavez et al. 2015), KRAB (Gilbert et al. 2013), p300core (Hilton et al.

2015), LSD1 (Kearns et al. 2015), Dnmt3a-3L (Bernstein et al. 2015)] fused to nuclease- dead Staphylococcus aureus Cas9 (Sa-dCas9), a recently described Streptococcus pyogenes Cas9 (Sp-Cas9) ortholog that is significantly smaller than Sp-dCas9, but that still utilizes a permissive PAM sequence (NNGRRT) for targeting (Ran et al. 2015). The compact nature of Sa-Cas9 provides extra space for addition of additional gRNAs, large effectors, or other elements within a single vector. Furthermore, the purposeful use of pMVP as a foundation also confers robust versatility to pMAGIC, enabling the design of vectors specific to desired experimental goals. Altogether, the utilization of Sa-dCas9, our new expanded-capacity and fiber-modified adenovirus vectors, as well as the pMVP foundation allows pMAGIC to enable the creation of customized vectors for efficient delivery of Sa-dCas9 fused to a < 3 kb effector and up to three gRNAs, or other elements, from a single vector into a broad array of cell types.

142

6.2 Results

6.2.1 Generation of Sa-Cas9 Entry Plasmids

To develop our modular epigenetic engineering platform, pMAGIC, we used pMVP (described in Chapter 4) as a foundation to enable a promoter of choice, Sa-dCas9 fused to an effector, and up to three gRNAs or other elements, to be recombined into any of the pMVP adenovirus vectors, or other vectors with sufficient insert capacity

(Figure 41A). While the vast majority of epigenetic engineering applications to date have utilized Sp-dCas9, for size considerations we decided to instead utilize the Sa-dCas9

(Ran et al. 2015) ortholog that, to our knowledge, has only been previously used for

VPR-based transcriptional modulation (Kiani et al. 2015) and no other modalities of epigenetic engineering. Using Addgene 68495 (Kiani et al. 2015) as a PCR template, we generated pENTR R4-R3 plasmids containing either humanized Sa-dCas9 with N- and

C-terminal nuclear localization sequences (NLS) (plasmid IA201; Figure 41B), or 2x-NLS

Sa-dCas9 fused to VPR, a tripartite transcriptional activator composed of VP64-p65-Rta

(Chavez et al. 2015). To enable fusion of additional effectors to 2xNLS Sa-dCas9, we included unique restriction sites for NcoI (5’) and BamHI (3’). Subsequently, we utilized

BamHI digested IA201 and NEBuilder HiFi DNA Assembly to create C-terminal fusions for the effectors: KRAB, the Krűppel associated box transcriptional repressor domain which recruits H3K9 methyltransferases (Gilbert et al. 2013, Thakore et al. 2015); p300core, the H3K27 acetyltransferase domain of p300 (Hilton et al. 2015); LSD1, a H3K4me1/2

143

demethylase (Mendenhall et al. 2013, Kearns et al. 2015); or Dnmt3a-3L, a synthetic DNA methyltransferase formed by fusion of the C-terminal domains of Dnmt3a and Dnmt3L

(Bernstein et al. 2015) (Figure 41C). For those researchers who desire to use pMAGIC for gene editing, we also included an entry plasmid for nuclease-active Sa-Cas9 (Figure

41D). All of these Sa-Cas9 plasmids were cloned in an open format, devoid of a stop codon, to permit C-terminal fusion of an epitope tag of choice for detection. Importantly, pMAGIC is not a static, terminal entity, in that as new Cas9 approaches are developed they can readily be inserted into the system by the methods described here to be used alongside the 7 Sa-Cas9 elements already included.

144

Figure 41: pMAGIC Enables the Generation of Multicomponent Sa-Cas9 Expression Vectors. (A) A schematic representation of the structure of 3- or 4-component, pMAGIC-derived, single, double, and triple gRNA expressing vectors. pMVP components, such as an eGFP reporter, can be utilized in replace of a gRNA expression cassette. pMAGIC contains pENTR R4-R3 plasmids for (B) Sa-dCas9, (C) Sa-dCas9 fused to effector molecules, (D) as well as nuclease-active Sa-Cas9 for gene editing. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct.

6.2.2 Generation of gRNA Expressing Entry Plasmids

As we gain a deeper understanding of the parameters that mediate Cas9-based epigenetic approaches, it has become apparent that modulation of an element, especially activation, often requires simultaneous targeting with multiple gRNAs (Chavez et al.

2015, Hilton et al. 2015). Moreover, for loci and approaches where a single gRNA is

145

sufficient, there is an intense interest in being able to utilize a single vector to simultaneously engineer multiple loci. To enable multiplexed gRNA expression with pMAGIC, we used isothermal DNA assembly to create entry plasmids containing solely the human U6 promoter and a SaCas9 gRNA sequence modified to enhance expression

(in pENTR L1-R5), or the gRNA expression cassette in combination with either a promoter (RIP, CMV, EF1a in pENTR L1-R5 and pENTR L5-L4), or 3xHA bGH polyA

(pENTR L3-L2). All gRNA expression plasmids were designed such that upon BsaI digestion, protospacer oligonucleotides corresponding to a specific target sequence can be ligated into the pENTR plasmids to create a functional gRNA expression cassette

(Figure 42). Importantly, the modularity of pMAGIC allows gRNA expression plasmids to be mixed and matched with each other, as well as the various Sa-dCas9-effectors, thereby enabling the creation of complex vectors targeting up to 3 distinct genomic loci with a targeted epigenetic modifier. These gRNA entry plasmids can also be readily replaced with plasmids containing alternative gRNA sequences or promoters, such as the recently described compact tRNA promoters (Mefferd et al. 2015) that are significantly smaller than the human U6 promoter, to create room for other vector elements. Furthermore, since pMVP is the foundation of pMAGIC, gRNA expression cassettes can be replaced with any of the pMVP epitope tags, fluorescent reporters, transgene inducibility technologies, or protein identification tools (Figures 23-26), thus, adding another layer of flexibility to experimental design.

146

Figure 42: pMAGIC Components for gRNA Expression. (A) pENTR221 L1-R5 plasmids containing either a gRNA expression cassette or gRNA expression cassette and RIP, CMV, or EF1a promoter. (B) pENTR L5-L4 plasmids containing a gRNA expression cassette and RIP, CMV, or EF1a promoter. (C) pENTR L3-L2 plasmid containing a 3x HA epitope tag for C-terminal fusion to Sa-Cas9 proteins and a gRNA expression cassette. (A- C) Annealed gRNA protospacer oligonucleotides can be ligated into BsaI digested human U6-driven gRNA expression cassettes. (D) Protospacer oligonucleotide template, N represents 21-23mer protospacer sequence, highlighted sequences are required either for PolIII initiation (G) or ligation into BsaI digested gRNA expression cassette. If protospacer sequence does not begin with “G”, one must be added. Unique plasmid IDs based on the Newgard lab’s nomenclature are listed adjacent to each construct.

6.2.3 pMAGIC Validation

As a visual demonstration of Cas9-mediated targeting of effectors to a specific

DNA sequence, we performed an experiment utilizing a synthetic circuit in which mCherry is only expressed upon gRNA addition to the system. To do this, we generated individual expression vectors for: constitutively expressed Sa-dCas9/VPR; mCherry

147

expressed from the inducible TRE promoter; and expression of a gRNA targeting the

TRE promoter. As seen in Figures 43A-B, mCherry is not expressed when only Sa- dCas9/VPR and TRE-mCherry plasmid are co-transfected into HEK293 cells. In contrast, inclusion of the TRE gRNA allows for Sa-dCas9/VPR recruitment to the TRE promoter resulting in robust TRE activation and mCherry expression (Figures 43C-D). Similar results were also achieved using Sa-dCas9/p300core (data not shown).

Figure 43: Activation of a Synthetic Promoter by Sa-dCas9-Mediated Recruitment of VPR. Schematic representation and corresponding fluorescent microscopy of HEK293 cells 48h after co-transfection with plasmids for (A-B) constitutively expressed Sa-dCas9/VPR and mCherry expressed from the inducible TRE promoter (C-D) or both of these plasmids co-transfected along with a TRE gRNA expression plasmid.

148

6.3 Discussion

Although the first experiment demonstrating the potential of direct manipulation of native genomic elements was performed over 20 years ago using zinc finger proteins

(ZFP) (Beerli et al. 1998), the field has been limited by the arduous and low-throughput manner in which ZFPs and the more recently developed transcriptional activator-like effectors (TALE) are generated, although new advances are improving the process

(Bernstein et al. 2015). The advent of Cas9-based technologies has rapidly brought epigenetic engineering into the mainstream of biomedical research (Thakore et al. 2016).

This is due to the unique flexibility of Cas9, which can be retargeted to a new locus simply by inclusion of a ~20 bp substitution in the gRNA protospacer motif rather than redesign of the entire protein as required for ZFPs and TALEs. However, despite the growing number of effectors within the Cas9 epigenetic engineering toolbox, vector options that are accessible in a uniformly efficient manner continue to be lacking.

With pMAGIC, we assembled 5 different modalities of Cas9-based epigenetic engineering into a singular platform, facilitated by the versatility of the pMVP system, that allowing direct comparison the operative elements while all other parts of the delivery system are kept constant. Furthermore, while pMAGIC was designed with adenoviral delivery in mind, it is compatible with all pMVP destination vectors, limited only by vector insert capacity. In order to accommodate promoters, multiple gRNAs, and large Cas9 effectors in our vectors, pMAGIC is based on Sa-Cas9, which is ~1 kb

149

smaller than Sp-Cas9. While the PAM sequence of SaCas9 (NNGRRT) is more restrictive than Sp-Cas9 (NGG), the Sa-Cas9 PAM sequence still occurs once every 32 bp of random

DNA, allowing for ample targeting locations throughout the genome (Kleinstiver et al.

2015). The space saved by use of Sa-Cas9 permits large effectors, up to 3 kb, to be fused to Sa-dCas9 and delivered alongside 3 gRNAs. If only a single gRNA is required, pMAGIC is capable of delivery of Sa-dCas9 fused to an even larger effector, up to ~4 kb.

Also of note, Sa-dCas9 has only been utilized previously for transcriptional activation.

Therefore, the new fusions generated in this work represent new modalities for Cas9 epigenetic engineering. The modularity of pMAGIC also permits future improvements and discoveries in Cas9 orthologs and effectors to be readily incorporated. To this end, it would be appealing to modify pMAGIC to include multiple Cas9 orthologs, thus, significantly expanding the ability of pMAGIC to target almost indiscriminately across the genome due to the variety of PAM sequences represented in a multi-ortholog platform. Additionally, this envisioned version of pMAGIC would enable simultaneous engineering of multiple loci using different Cas9-modulators since each Cas9 ortholog is reliant on a unique gRNA, as recently demonstrated (Gao et al. 2016).

The cross-compatibility of pMAGIC with pMVP greatly expands the utility of pMAGIC to applications beyond epigenetic engineering. Recently it has been demonstrated that dCas9 can also be utilized for: chromosomal painting and visualization through fluorescent protein recruitment; identifying distant chromosomal

150

interactions through fusion of epitope tags or biotinylation of Cas9; and, unbiased determination of chromosomal interaction partners via fusion of epitope tags or Apex2

(Chen et al. 2013a, Fujita and Fujii 2013, Ma et al. 2016, Liu et al. 2017, Gu et al. 2018).

Importantly, pMAGIC, via pMVP, is compatible with all of these modalities and applications and, furthermore, these platforms can be readily adapted, as described in

Chapter 4, to permit use of any future technologies. Thus, in total, we have successfully created an innovative, highly versatile, modular epigenetic engineering platform, pMAGIC, that enables the locus-specific recruitment of a variety of effectors. Moreover, its compatibility with pMVP provides a myriad of options to improve experimental design and outcomes.

151

7. Utilization of pMAGIC to Functionally Validate the Contribution of Area IV in the Regulation of PDX1 7.1 Introduction

To demonstrate the power of our new pMAGIC system (see Chapter 6) for biological discovery, we targeted Sa-dCas9/LSD1 to a conserved enhancer element (Area

IV) that is proposed to modulate expression of PDX1, a homeobox transcription factor that has been shown to be critical in islet β-cell development and function in rodents. In humans, mutations of PDX1 cause a monogenic form of diabetes (MODY4) (Stoffers et al. 1997a) as well as permanent neonatal diabetes (Nicolino et al. 2010), and complete loss of PDX1 function results in pancreas agenesis (Stoffers et al. 1997b). Additionally, single-nucleotide polymorphisms (SNPs) within a 40 kb region encompassing Area IV have a significant association with fasting glucose levels (Scott et al. 2012). However, despite the sufficiency of PDX1 mutations to cause diabetes, the role of modification of

PDX1-regulating enhancer elements caused by SNPs or environmental influences in development of type 2 diabetes remains unknown.

Previous work using luciferase assays and in vivo reporters have characterized a series of cross-species conserved enhancer elements that control PDX1 expression through development (I-III) and maturation (I-IV) of the β-cell (Figure 44A) (Gerrish et al. 2000, Gannon et al. 2001). Area IV has been shown to be activated specifically during the β-cell maturation process, but its contribution to maintaining PDX1 expression in mature β-cells has not been reported (Gerrish et al. 2004). Here, we utilize a pMAGIC- 152

derived adenovirus to demonstrate that Sa-dCas9/LSD1-mediated attenuation of the

Area IV enhancer causes a significant inhibition of endogenous PDX1 expression in both

INS1 832/13 rat insulinoma cells and primary rat islets. Furthermore, this decrease in

PDX1 expression is sufficient to alter expression of classical PDX1 target genes, demonstrating an important role of the chromatin state of Area IV in maintaining PDX1 gene expression and β-cell identity.

7.2 Results

7.2.1 pMAGIC-Derived Vectors

To investigate the effect of epigenetic engineering in Area IV, we used the

Deskgen.com resource (Hough et al. 2017) to identify a Sa-Cas9 protospacer motif

(GTTAACAACATCAGGCTGACGG), hereafter referred to as gRNAAreaIV, within this region (Figure 44A). We ligated annealed oligonucleotides containing gRNAAreaIV into a

BsaI digested pENTR L3-L2 entry plasmid containing 3xHA-bGH polyA as well as a human U6-driven Sa-Cas9 gRNA cassette. The gRNAAreaIV entry plasmid, or 3xHA bGH polyA entry plasmid lacking a gRNA, were subsequently recombined with plasmids containing RIP and Sa-dCas9/LSD1open into an Ad5 destination vector. The resulting vectors were used to generate recombinant adenoviruses (Figures 44B-C) that were purified on CsCl gradients before further use.

153

Figure 44: Generation of pMAGIC-Derived Adenoviruses Targeting Area IV. (A) UCSC genome browser screenshot of the Pdx1 locus in the rat genome (rn5). Highlighted areas demark conserved Area I-IV regions, the gRNA targeting Area IV, and amplicons of four ChIP-PCR primer sets. β-cell specific, pMAGIC-derived adenoviruses expressing Sa-dCas9/LSD1 (B) without or (C) with a gRNAArea IV.

7.2.2 pMAGIC Enables Epigenetic Manipulation of Area IV

For validation of this vector, we utilized the rat INS 832/13 β-cell line for which we have unpublished epigenetic profiling demonstrating that Area IV is enriched for

H3K27ac and other histone modifications, consistent with it being an active enhancer element in these cells. Transduction of these cells followed by chromatin immunoprecipitation (ChIP) for the HA epitope and qRT-PCR for Area IV demonstrated that Sa-dCas9/LSD1 is successfully recruited to Area IV (Figure 45A). To determine if

154

this recruitment leads to modification of the epigenetic status of Area IV, we performed

ChIP-PCR for H3K27ac, a previously reported marker of LSD1 activity. As expected, recruitment of LSD1 via Sa-dCas9 to the Area IV locus caused a significant ~70% loss of

H3K27ac enrichment (Primer sets #3 and #4, Figure 45B), indicative of enhancer attenuation.

155

Figure 45: Alteration of the Area IV Epigenetic State by LSD1 Recruitment. ChIP-PCR analysis for the HA epitope tag (A) and H3K27ac (B) performed with primers for control regions (ChrX, Chr15), or Area IV (#1-#4, corresponding with Figure 44A) on cross-linked chromatin from INS1 832/13 cells 48h after transduction with either (-) gRNA control or gRNAArea IV purified adenoviruses. Data represent mean ± S.E.M. of 3 independent experiments. # p < 0.05 compared to (-) gRNA control adenovirus.

156

7.2.3 Area IV is the Major Contributor to PDX1 Transactivation in INS1 832/13 Cells

Within the cell, the genome is elegantly packaged and divided into subdomains that mediate inter-chromosomal interactions (Dixon et al. 2012). As such, linear DNA distance is not always indicative of the true distance in three-dimensional space.

Therefore, we performed qRT-PCR to determine whether epigenetic perturbation in

Area IV had any functional consequences for its putative target gene, PDX1 and the

PDX1 transcriptional network. Indeed, qRT-PCR analysis of INS1 832/13 cells transduced with these vectors exhibited a robust 56% decrease in PDX1 expression

(Figure 46A). This decrease in PDX1 was sufficient to cause significant alterations in the previously identified PDX1 target genes GCG, MAFB, TSPAN8, SLC2A2 (GLUT2), and

UCN3 (Figure 46B). These results implicate that epigenetic modification of Area IV as a significant modulatory event for control of PDX1 expression in this model.

157

Figure 46: Attenuation of Area IV is Sufficient to Decrease PDX1 and Disrupt the PDX1 Transcriptional Network. qRT-PCR analysis of (A) PDX1 and (B) PDX1 target genes in INS1 832/13 cells 48h after transduction with either (-) gRNA control or gRNAArea IV purified β-cell specific Sa-dCas9/LSD1 adenoviruses. Data represent mean ± S.E.M. of 3 independent experiments. # p < 0.05 compared to (-) gRNA control adenovirus.

7.2.4 Area IV Regulates PDX1 in Primary Rat Islets

One of the unique abilities of pMAGIC is the creation of adenoviral vectors that can efficiently deliver large Cas9 fusions into primary cell-types, such as the pancreatic islet β-cell. This is crucial for research examining the role of enhancer elements in gene

158

regulation, given that the activity and function of these elements can be highly context specific, as seen with the selective activation of Area IV in mature β-cells (Gerrish et al.

2004). Therefore, having successfully demonstrated the effectiveness of gRNAAreaIV to target Sa-dCas9/LSD1 to Area IV in a rat cell line, we proceeded to functionally validate the role of Area IV in the regulation of PDX1 in primary rat islets. Confirming our cell line results, PDX1 was significantly diminished by 30% upon targeting of PDX1 Area IV with Sa-dCas9/LSD1 (Figure 47A). Furthermore, the PDX1 targeted gene SLC2A2

(GLUT2), chosen because its expression is confined to the β-cell, was also significantly decreased (Figure 47B), albeit to a lesser extent, 12%, than the 37% decrease observed in

INS1 832/13 cells. At this juncture, we are not able to discern whether the lessened impact of Area IV attenuation in rat islets is caused by it contributing a smaller component to overall PDX1 transactivation than in insulinoma cells, or if this is due to lesser adenoviral transduction efficiency in primary islets versus INS1 832/13 cells.

Nevertheless, these experiments validate Area IV as a regulator of PDX1 expression and represent the first example of LSD1-based epigenetic engineering in a primary tissue.

159

Figure 47: Area IV Regulates PDX1 Expression in Primary Rat Islets. qRT-PCR analysis of (A) PDX1 and the PDX1 target gene (B) SLC2A2 (GLUT2) in intact primary rat islets 48h after transduction with either (-) gRNA control or gRNAArea IV purified β-cell specific Sa-dCas9/LSD1 adenoviruses. Data represent mean ± S.E.M. of 3 independent experiments. # p < 0.05 compared to (-) gRNA control adenovirus.

7.3 Discussion

Targeting of Area IV was of interest to us in part because of the specificity of its activity in mature β-cells, and its inclusion in a larger region implicated in regulating glycemia, although the causal SNP(s) and corresponding elements have not yet been identified (Gerrish et al. 2004, Scott et al. 2012). Upon β-cell specific delivery of Sa- dCas9/LSD1 targeted to Area IV, we observed epigenetic alterations indicative of enhancer attenuation in INS1 832/13 cells. These Area IV epigenetic alterations corresponded with a significant decrease in PDX1 expression and significant modulation of members of the PDX1 transcriptional network in both INS1 832/13 cells and primary rat islets. We note that our studies did not determine if downregulation of PDX1 expression in these systems was solely due to targeting of Area IV, or whether

160

disruption of a PDX1 autoregulatory loop through decreased PDX1 expression also contributed (Spaeth et al. 2017). Of note, in INS1 832/13 cells, attenuation of Area IV resulted in a significant induction of the α-cell associated genes GCG and MAFB. These observations are in agreement with a recent publication showing that inducible deletion of PDX1 in adult mouse β-cells resulted in the loss of β-cell identity and upregulation of an α-cell transcriptional network (Gao et al. 2014). Thus, our data suggests that within this model, the enhancer activity of Area IV is required to maintain PDX1 transactivation at a level sufficient to maintain β-cell maturity and identity.

In the past few months, Spaeth, et al, reported a mouse model in which Area IV was genetically deleted. In agreement with our findings, Area IV deletion corresponded with a significant decrease in PDX1 expression, correlating with decreased β-cell function and replication. However, these phenotypes were only observed upon deletion of Area IV in the context of PDX1 heterozygosity, and then only in male mice (Spaeth et al. 2017). Our data implicates the genomic region containing Area IV as having a more essential role in maintaining proper PDX1 expression levels, even in the context of two functional PDX1 alleles. One explanation for the apparent differences in results considers the complete deletion of Area IV in the Spaeth, et al study, versus our study involving targeting of an epigenomic modifier to the same region. Epigenetic profiling in human islets has found that the ~0.5 kb Area IV element is part of a > 10 kb stretch enhancer (Parker et al. 2013), and unpublished data from our group has found this

161

stretch enhancer element is conserved in primary rat islets and INS1 832/13 cells.

Therefore, the difference in these models may be a direct consequence of whether the function or higher order chromatin structure of the stretch enhancer encompassing Area

IV has been perturbed. To this point, the manner in which the composition of individual regulatory elements contribute to overall stretch enhancer function is still unknown. In extension, it will be of particular interest in the future to apply the pMAGIC toolbox to the PDX1 locus in human islets, where SNPs encompassing this stretch enhancer have been associated with glycemic control (Scott et al. 2012). Our results suggest that alteration of the highly conserved Area IV is sufficient to have significant impact on

PDX1 expression and β-cell biology, but it should be noted that the cross-species conservation of the rest of the region upstream of PDX1, especially the SNP-associated regions and additional enhancers, is still undetermined. In summary, we used pMAGIC to rapidly assemble an adenoviral vector capable of targeting and modifying a discrete genomic locus in both INS1 832/13 cells and, importantly, primary rat islets, resulting in a significant decrease of PDX1 expression and perturbation of PDX1 target genes. These results provide a functional validation of the interaction of Area IV and the PDX1 promoter and demonstrate the ability of pMAGIC-derived vectors to enable epigenetic engineering for discrete disease-associated genomic elements in primary cell-types relevant to disease pathogenesis.

162

8. Conclusions and Perspectives

Manipulation of cell-types implicated in the etiology of disease is a critical component biomedical research. Examples of such manipulations from a century ago include the pancreatic ductal ligation techniques used by Banting and Best to induce atrophy of the pancreatic exocrine tissue, permitting the discovery of insulin (Banting and Best 1922). As our knowledge of the molecular underpinnings of cellular function has expanded, our toolbox has also flourished. While techniques akin to ductal ligation are still utilized (Cavelti-Weder et al. 2013), they are now complemented by the ability to manipulate primary islets via miRNA modulation (Chapter 3), transgene expression

(Chapter 5), and even perturbation of discrete regulatory elements within a native chromatin context (Chapter 7). While expansion the experimental toolbox is rarely the main objective of biological research, it is intimately intertwined with scientific progress.

It is through this lens that the research endeavors described here have been undertaken. While our laboratory has found PDX1 overexpression to be sufficient for induction of human β-cell replication, the blunt instruments available to the field of islet biology have stymied progress in identifying and validating the molecular components of this pathway. Therefore, out of necessity, we developed two new platforms, pMVP

(Chapter 4) and pMAGIC (Chapter 6), that enable the rapid production of vectors tailored for testing of our hypotheses. The application of these vectors enabled us to create new experimental paradigms to identify soluble factors regulated by PDX1

163

(Chapter 5) as well as to define the contribution of Area IV to PDX1 expression and β- cell homeostasis (Chapter 7). We believe that with these new vectors in hand, PDX1- induced pathways and secreted factor(s) underlying β-cell replication will finally be revealed.

8.1 β-Cell Replication: A Cure for Type-2 Diabetes?

As discussed in Chapter 1, progression of T2D is marked by the incremental loss of functional β-cell mass resulting in an inability of the pancreas produce sufficient insulin for glycemic control. It seems reasonable to conclude that a therapeutic capable of expanding β-cell mass to combat these losses would be curative of the disease; but significant concerns remain about activation of β-cell replication as a therapeutic strategy. One critique is that the best-case scenario for induction of human β-cell replication is to cause 1%-2% of β-cells to undergo cell cycle progression within a 24- hour window, an amount that may be inadequate to influence glycemic control in vivo.

However, viewed from a different perspective, these results may actually be encouraging in that one would presume that β-cell function is compromised during the cell cycle process, making it undesirable to induce the majority of β-cells to replicate.

These considerations suggest that any course of therapeutic intervention would likely need to occur over the span of months to have achieve desirable outcomes.

This leads into a second concern, which is that drugs, biologic modifiers, or genes designed to induce β-cell replication may also activate cell cycle progression in

164

other tissues as well, leading to potential metabolic abnormalities and increased risk of cancer. These risks must be considered in the context of incomplete efficacy, but remarkable safety profile of the current frontline treatments for diabetes. For example, the almost universally prescribed anti-diabetes drug metformin is also being studied for its beneficial impacts on longevity, cancer risk, and cardiovascular disease (Bowker et al.

2006, Foretz et al. 2014). Therefore, any proposed future treatment will have to meet or exceed the high bar set by current interventions and possibly be coupled with pre- intervention cancer screening. One potential target that encapsulates these concerns is

PTEN inhibition, which has been shown to help prevent T2D onset in mouse models and, based on our observations in Chapter 3, can enhance specific modes of β-cell replication. While PTEN mutations occur frequently in cancers and are sufficient for oncogenesis (Wang et al. 2003), little is known about the impact of acute PTEN inhibition in vivo. From a cellular transformation perspective, there is a vast difference between acute and chronic inactivation of a pathway, but are these differences great enough to enable acute PTEN inhibition to be utilized? Furthermore, would any increases of β-cell mass be maintained post-treatment, or would other regulatory mechanisms force atrophy of β-cell mass as seen in postpartum rodent models (Scaglia et al. 1995)? Absent in vivo studies for specific interventions, these questions will remain viable concerns.

However, they also raise important questions about whether certain targets should be immediately dismissed based on an association with oncogenesis or allowed to proceed

165

with the hope that a therapeutic window capable of balancing the therapeutic needs with potential risks can be achieved. From our perspective, given a long enough time horizon, the latter is achievable through a combination of the advancement of β-cell targeting technologies being currently developed as well as a deeper understanding of the pathways that are sufficient to regulate β-cell mass

8.2 PDX1 Dysregulation: Driving Disease?

As discussed in Chapter 1.3, mutations of PDX1 in humans cause a spectrum of diabetic maladies that correlate with the level of remaining PDX1 activity. Furthermore, mouse models that induce PDX1 dysregulation through deletion of regulatory elements also result perturbed glucose homeostasis. The agreement of these results is indicative that deficient PDX1 activity, regardless of whether driven directly by PDX1 mutation or indirectly through regulatory element deletion, can be a causative factor of diabetes.

Given the crucial nature of PDX1 status for maintaining β-cell health, has the field failed to notice the contribution of enhancer-mediated dysregulation of PDX1 to the etiology of

T2D? To this point, an early study showed that Area IV activity was decreased by glucocorticoid treatment (Sharma et al. 1997). Here we report that altering the epigenetic status of Area IV is sufficient to induce significant changes in expression of PDX1 and downstream PDX1-target genes (Chapter 7). The fact that glucocorticoid treatment is a known risk factor for diabetes (Ogawa et al. 1992), leads one to speculate that Area IV dysfunction could be a driving factor for glucocorticoid-induced diabetes. It is also

166

possible SNP-mediated impairment of PDX1 enhancer function could serve as a genetic underpinning for susceptibility to glucocorticoid-induced dysfunction.

Despite the dogma of β-cell fragility that pervades the diabetes field, we have found that insulinoma cell lines are highly susceptible to ER and apoptotic stressors, whereas primary β-cells are remarkably more resistant to these same agents (Hayes et al.

2017). This suggests that the apoptosis-mediated decline in β-cell mass observed in T2D could be the result of loss of mechanisms that provide resistance to stressors. While

PDX1 heterozygosity has been shown to sensitize the β-cell (Sachdeva et al. 2009,

Fujimoto et al. 2010), would enhancer-mediated PDX1 dysregulation have a similar phenotype? It would be of great interest to utilize pMAGIC to directly query the necessity of these distal-regulatory elements to maintain β-cell health under these conditions.

Finally, while decades of studies have yielded remarkable insight into the role of

PDX1 as a master regulator, we still know little about how PDX1 chromatin occupancy is altered in disease states. For example, despite knowledge that many PDX1 mutations are sufficient to cause neonatal, MODY, and late-onset forms of disease, studies to determine if these mutations alter genome-wide chromatin occupancy have not been performed. Furthermore, while models of PDX1 haploinsufficiency are widely utilized, we do not know how PDX1 genome-wide occupancy is altered under these conditions.

Even in cell culture models, the conclusion that oxidative and glucose-induced stress

167

causes PDX1 dysfunction via loss of chromatin occupancy is based only on gel-shift assays and ChIP-PCR for a single locus (Tanaka et al. 1999, Guo et al. 2013). Therefore, despite knowing that PDX1 dysfunction is correlated with disease, we know little about how this links to PDX1 chromatin occupancy. With the advent of Cas9-based gene editing, it is now feasible to generate cell line and animal models that faithfully reproduce the PDX1 mutations associated with disease, thereby enabling the pursuit of these questions to gain true understanding of the contributions of PDX1 dysfunction to diabetes development.

8.3 pMVP and pMAGIC: Unmasking the Etiology of Disease

As our understanding of the etiology of disease continues to progress, researchers are forced to implement increasingly complex genetic engineering strategies to better understand the nuances of cellular dysfunction. With the creation of pMVP and pMAGIC, we have presented the modern researcher with an arsenal of tools that enable them to rapidly generate innovative vectors for exploration of their experimental model, allowing better focus on unraveling the complex questions they seek to answer instead of being forced into an experimental tunnel by vector inflexibility.

As discussed in Chapter 3, interrogation of the contribution of the miR17~92 cluster to PDX1-mediate replication was thwarted by lack of available islet-compatible tools for miRNA inhibition. However, using pMAGIC, we will able to directly inhibit the miR17~92 cluster by dCas9-KRAB-mediated transcriptional repression. Furthermore,

168

modulation of the epigenetic status of the putative miR17HG stretch enhancer will allow us to determine whether it regulates miR17HG in the β-cell, while also defining the contribution of these enhancers to PDX1-mediated induction of miR17HG. We also can utilize our pMVP-derived epitope tagged forms of PDX1 for ChIP experiments to determine the regulatory elements that are engaged by PDX1.

More broadly pMAGIC will enable the interrogation of enhancers containing

T2D-associated SNPs in a manner similar to PDX1 Area IV. One example is the putative developmental enhancer found in EndoC β-cells that contains T2D-associated SNPs

(Chapter 3). Not only can we utilize pMAGIC to develop adenoviruses targeting this region in EndoC β-cells, but we can also generate Sleeping Beauty vectors to create an embryonic stem cell model to examine the role of this element in pancreatic development. Therefore, due to the unique ability of pMAGIC to deliver dCas9-based tools into islets, it is uniquely positioned to answer some of the questions that have vexed the T2D GWAS field regarding regulation of specific genes by SNP-associated elements, and their potential contribution to disease pathogenesis.

169

Appendix A

Entire validated sequence (2,066 bps) for the portion of the miR17HG promoter overlapping with the PDX1 DNA binding motif as well as the transcription start site in the INS1 832/13 rat insulinoma cell line.

GGTTTCTGAGAATTCTGGAATTTCCTGAACCACAGCGTGTAGCAGGGAGTGGGC TTATCTCAGCTCTGTGGTTGTTTACAATGCTCTTTCCGAGCTGCTGTGTGCTCTCTT TCTCTCTCTCTTTCTCTCTACCGGAGAGCCAAGAGTTCTGGGTGCACTGGGCGAG GGAGGCGACTCAGGTACTGGCCGCGAAGCCCTACCCAGCACGCGCTCTGTCGCA CGCGGCCAGGTTAAAGAGACCTTTCCAGGATGAGCCGGATCCCAAAGGATCCGT GATCCGCGGAGACCGCGCTGGGGGCGACGGTAGGCGAGTGGATCTGACTTGGCT CCCACTCGCGTGGGACCCCGATCCCGGGGCGAGCAGAAAAGCTGTGGTCACCC CGCGCCTCCGTCGAGTCCCCGGCCAGCGCGGCTGGCGGGAAGCCTGAGCCGCGT GCGCTCGCAGGAGGGTGTGCGGCGCGGGGTCCGCACGCGCGGGGCTGGCTCGG GAGGGCGTGCCGGGGGCGTGGCGCGAGGTGTCCCTGCCCGGAGCCCTGCGCGG AGGGCTGCGGGCGGCGGTGGGCGGACCCCGGGGGACGCGCGGCCAGGGAAGG CGGCGCGGCGCCACCAGCGGCTCCCCGGCTCCCGCTCGCCCGGCCGCCAAGAA CGAGCCGCCGTGGGCGGGGCCTGCGGTGATTGGCGGGCGGGCGGGGAGGTCGG AAGTACTTTGTTTTTTATGCTAATGAGGGAGTGGGGCTTGTCCGTATTTACGTTGA GGCGGGAGCCGCCGCCCTTCATTCACCCACATGGTCCTTCGAGGTGCCGCCGCC GCCGCCCCGGCCCGCGCCTTCGCGCCACTTCGCGCCCTGCGGCGAGGCCGAGGG GTGGGGACGGTCCCCACGCCGCCCGCGCCCCTCCGCGGAGCCCGGAGCGCAGG CCGGGCGGGCCACGCCGCACCCCCGGCCTGGGGCCTCGGGCCGGAGTGGCGCG GAGGCTGGCGAAGATGGTGGCGGCTCCTGCGGTGAGTGCGCCCGCCCCGCCGCC TGCTCAGGGAAACCTGTTGTGCGCTAAGCGGAGCGGCGGCGGGGCGGGCGGGA CGGCCGGGGCCGCGGGCCGGGTGGGTCTCGGGCCGTTGGGCGCCGGCATGGGG CGGCCGTCCTGGCGCGCACGCGGGCCGGAGGGGGCCCGAGGGGGCGGCGGCCC GAACCCCGGCGGGGCCTGCCCGGCGCACACAATGGCCCTCGGGAGGCTTCGCA CGAGGCCGGTGGCGGTCCCGGGAGCAGCGGGAGCCAGGCGGGGCGCGCGGCC GCCGTGCGGGGAAAGTTCTCCCGGGGGCGAGAGTTAAAGCGCCTCCAGAACAA AGCGGCGGCGGCGGCACATGGGGCAGGCCGCGGGCCGGGAGGGGGCGCGCCC ACGAGGAACCTGCGCGCGGGCGGGCGGGAGGCGCGGCGTGGGCGGGAGTCGG CGTCTCCCAAACTTTGTGCGCGCCCGAGCGGGCGCCGGGACGCAGGAGCTGCGG GGCGACGGGGCCCGGGGCGCCACCCGCGCTCCGCGTGGGCTTTGTGTGCGGCCG CGTGGGCAGCTCCCCTCGTGGGCGAGGCGACTTGGCGGCCTGCGGGTGGCCCGG GACGCCTGAGCCCGCCTTCGGGACGGCGCGGAGCCGGGACGGGCGCTGGCTAC GAGCGCGGCGCGGGAGGGCTGGGCGGTCTCGGCAGGAGCGGCGGCCCCGCCGC CATGTTCCTGCGGGGCGGGCTGCGCGCGCCGAGGGCGGGGGACATGGCGGCGA CTGCGCGCCGCCGCCGATTGTTCCCGGCTTAGGCCTCGGGCCGCGTGCGGCGAG

170

CCCCCTGGCCGCGCCCCCCCTTTCCCCGGAGCCTGGCCTGGGGGCGTGCGGGAC ACAAAGGACGGCCGGCGTGTCCGGAGCCACGCTCTGCTCGGGCCTCGGCGCCG CCGCTCGCCGCGCAGCCGCCCAGAAACGGGCGGGGGGAGTCGCCGCGTCCGGC GCGGCCCTGCTCTGACCTGCCGCCCCCTGGCGGCCGCGCAGGGACCCGCAGGGC CGCCTGCCCCCTTGTGCGACATGTGCTGCCGGC

171

Appendix B

Comprehensive list of the 108 plasmids that currently comprise pMVP and pMAGIC. For each plasmid, information pertaining to the plasmid type (pENTR, pDEST, MISC), laboratory ID (XX###) element cloned, overall size of the plasmid, as well as plasmid mass equivalent to 10 fmol to aid LR recombination reaction calculations is provided. Sequences for oligonucleotides and gBlock DNA fragments utilized to construct these plasmids and vectors are provided in Appendix C and Appendix D, respectively.

Plasmid Size 10 fmol ID Element Type (bps) = X ng pENTR HB401 CMV promoter 3321 21.9 L1-L4 pENTR IF101 EF1a promoter 3796 25.1 L1-L4 pENTR HB501 Rat Ins1 promoter (RIP; β-cell specific) 2992 19.7 L1-L4 pENTR HI901 ALB promoter (Liver specific) 4894 32.3 L1-L4 pENTR IG001 TBG promoter (Liver specific) 3020 19.9 L1-L4 pENTR II701 cTnT promoter (Cardiac specific) 2972 19.6 L1-L4 pENTR IG603 TRE promoter 2893 19.1 L1-L4 pENTR IG301 CMV promoter 3370 22.2 L1-R5 pENTR IY704 EF1a promoter 3845 25.4 L1-R5 pENTR IG401 RIP 3041 20.1 L1-R5 pENTR IG703 TRE promoter 2942 19.4 L1-R5 pENTR HE401 H1-driven shRNA Expression Cassette 2893 19.1 L1-R5 pENTR JJ802 hU6-SaCas9 gRNA 2983 19.7 L1-R5

172

pENTR JI601 hU6-SaCas9 gRNA + RIP 3444 22.7 L1-R5 pENTR KB501 hU6-SaCas9 gRNA + CMV 3772 24.9 L1-R5 pENTR KB401 hU6-SaCas9 gRNA + EF1a 4234 27.9 L1-R5 pENTR IH001 HA epitope tagopen 2665 17.6 L5-L4 pENTR IH201 3x HA epitope tagopen 2719 17.9 L5-L4 pENTR IG901 FLAG epitope tagopen 2662 17.6 L5-L4 pENTR IO801 3x FLAG epitope tagopen 2704 17.8 L5-L4 pENTR IH101 OLLAS epitope tagopen 2680 17.7 L5-L4 pENTR IX104 myc epitope tagopen 2668 17.6 L5-L4 pENTR IX203 V5 epitope tagopen 2679 17.7 L5-L4 pENTR II901 FKBP Degron Domainopen 2962 19.5 L5-L4 pENTR II801 ecDHFR Degron Domainopen 3115 20.6 L5-L4 pENTR JE301 osTir1-P2A-AIDopen 5005 33.0 L5-L4 pENTR IO702 eGFPopen 3289 21.7 L5-L4 pENTR IF801 GFP-P2Aopen 3349 22.1 L5-L4 pENTR IO901 BioID2open 3370 22.2 L5-L4 pENTR IF701 APEX2open 3319 21.9 L5-L4 pENTR IH401 CMV promoter 3318 21.9 L5-L4 pENTR JZ901 Ef1a promoter 3795 25.0 L5-L4 pENTR HE302 RIP 2989 19.7 L5-L4

173

pENTR IG101 TRE promoter 2890 19.1 L5-L4 pENTR JI702 hU6-SaCas9 gRNA + RIP 3389 22.4 L5-L4 pENTR KB701 hU6-SaCas9 gRNA + CMV 3716 24.5 L5-L4 pENTR KB601 hU6-SaCas9 gRNA + EF1a 4178 27.6 L5-L4 pENTR IO201 mCherry 3378 22.3 R4-R3 pENTR HJ601 mCherryopen 3378 22.3 R4-R3 pENTR IO001 eGFP 3387 22.4 R4-R3 pENTR IO102 eGFPopen 3384 22.3 R4-R3 pENTR IO301 eYFP 3387 22.4 R4-R3 pENTR IO401 eYFPopen 3384 22.3 R4-R3 pENTR IO502 eCFP 3387 22.4 R4-R3 pENTR IO601 eCFPopen 3384 22.3 R4-R3 pENTR HD301 βgal 5722 37.8 R4-R3 pENTR HM001 Cre 3702 24.4 R4-R3 pENTR JN402 Firefly Luciferase 4320 28.5 R4-R3 pENTR IG510 Human PDX1 3522 23.2 R4-R3 pENTR JH204 Human PDX1open 3519 23.2 R4-R3 pENTR IA201 Sa-dCas9open 5931 39.1 R4-R3 pENTR IA304 Sa-dCas9/VPRopen 7527 49.7 R4-R3 pENTR IB801 Sa-dCas9/KRABopen 6168 40.7 R4-R3

174

pENTR IF311 Sa-dCas9/LSD1open 8493 56.1 R4-R3 pENTR IF405 Sa-dCas9/p300core open 7791 51.4 R4-R3 pENTR IP803 Sa-dCas9/Dnmt3a-3Lopen 7569 50.0 R4-R3 pENTR IH601 SaCas9open 5931 39.1 R4-R3 pENTR HB901 Polyadenylation signal (PolyA) 3151 20.8 L3-L2 pENTR IA001 3x HA epitope tag + PolyA 2895 19.1 L3-L2 pENTR IX601 FLAG epitope tag + PolyA 2838 18.7 L3-L2 pENTR IX503 OLLAS epitope tag + PolyA 2856 18.8 L3-L2 pENTR IX301 myc epitope tag + PolyA 2844 18.8 L3-L2 pENTR IX401 6x His epitope tag + PolyA 2832 18.7 L3-L2 pENTR IY001 BioID2 + PolyA 3540 23.4 L3-L2 pENTR IY202 APEX2 + PolyA 3573 23.6 L3-L2 pENTR IR801 eGFP + PolyA 3573 23.6 L3-L2 pENTR JP801 P2A-eGFP + PolyA 3600 23.8 L3-L2 pENTR HD201 IRES2-eGFP + PolyA 4161 27.5 L3-L2 pENTR OLLAS epitope tag-IRES2-eGFP + IX702 4193 27.7 L3-L2 PolyA pENTR IX907 3xHA epitope tag + WPRE 3310 21.8 L3-L2 pENTR JA001 FLAG epitope tag + WPRE 3253 21.5 L3-L2 pENTR IZ901 OLLAS epitope tag + WPRE 3271 21.6 L3-L2 pENTR IZ702 myc epitope tag + WPRE 3259 21.5 L3-L2

175

pENTR IZ801 6x His epitope tag + WPRE 3247 21.4 L3-L2 pENTR JD404 BioID2 + WPRE 3955 26.1 L3-L2 pENTR JD602 APEX2 + WPRE 3988 26.3 L3-L2 pENTR JG102 P2A-mCherry + WPRE 3980 26.3 L3-L2 pENTR JG202 mCherry + WPRE 3920 25.9 L3-L2 pENTR IG806 PolyA-RIP-eGFP-P2A-TETa-PolyA 5290 34.9 L3-L2 pENTR JN511 PolyA-EF1a-eGFP-P2A-TETa-PolyA 6099 40.3 L3-L2 pENTR IR904 PolyA-CMV-eGFP-P2A-TETa-PolyA 5466 36.1 L3-L2 pENTR HA epitope tag + PolyA-RIP-eGFP- KA701 5315 35.1 L3-L2 P2A-TETa-PolyA pENTR HA epitope tag + PolyA-EF1a-eGFP- KA901 6124 40.4 L3-L2 P2A-TETa-PolyA pENTR HA epitope tag + PolyA-CMV-eGFP- KA801 5491 36.2 L3-L2 P2A-TETa-PolyA pENTR KA401 PolyA-RIP-TETa-PolyA 4495 29.7 L3-L2 pENTR KA601 PolyA-EF1a-TETa-PolyA 5304 35.0 L3-L2 pENTR KA501 PolyA-CMV-TETa-PolyA 4671 30.8 L3-L2 pENTR 3x HA epitope tag + PolyA + hU6- JI501 3264 21.5 L3-L2 SaCas9 gRNA pDEST IQ013 pMVP/Ad-DEST 34625 228.5 pDEST JC001 pMVP/Ad/RGD-DEST 34658 228.7 pDEST JC101 pMVP/Ad/pK-DEST 34667 228.8 pDEST JH401 pMVP/Ad/RGD-pK-DEST 34709 229.1 pDEST JC201 pMVP/Ad/5-35-DEST 33863 223.5 pDEST JK104 pMVP-DEST 4132 27.3 pDEST JK305 pMVP/Blast-DEST 5261 34.7 pDEST JL302 pMVP/Puro-DEST 5462 36.0 176

pDEST JL002 pMVP/eGFP-DEST 5582 36.8 pDEST JL202 pMVP/mCherry-DEST 5575 36.8 pDEST JN607 pMVP/SB-DEST 4587 30.3 pDEST JN202 pMVP/SB/Blast-DEST 5716 37.7 pDEST JN302 pMVP/SB/Puro-DEST 5917 39.1 pDEST JT504 pMVP/SB/eGFP-DEST 6037 39.8 pDEST JT601 pMVP/SB/mCherry-DEST 6030 39.8 pDEST IZ501 pMVP/Lenti-DEST 7297 48.2 pDEST IY601 pMVP/Lenti/Blast-DEST 8131 53.7 pDEST IZ401 pMVP/Lenti/Neo-DEST 8462 55.8 Misc IX001 pBS/Fiber 5055 Misc JI401 RCA Assay Standard 3325

177

Appendix C

Oligonucleotide sequences utilized for PCR amplification, ligation, and isothermal DNA assembly to create components of pMVP and pMAGIC (listed in Appendix B) in addition to other plasmids described in this dissertation. For each sequence, the laboratory ID (JHXXX or TBXXXX), orientation, and application are listed. In addition, plasmids and vectors that resulted from these applications are recorded.

ID F/R App Sequence (5' to 3') Used in: JH006 Forward PCR TTTCAGATCTAGCAGGAAAAAAGAG Ad-miR17~92 GAAGTCTAGATCTTCAGTAATATTTAATTTT Reverse PCR JH007 CA Ad-miR17~92 JH025 Reverse PCR ATCACCTTGGCATCTTGTCTCACG HR304 JH026 Forward PCR AGGGTCGTTTGAATGGGCTGATAA HR304 JH027 Forward PCR TGCGAACCACCGTACTTGAGGA HR403 JH028 Reverse PCR TTGAAATGCTGCTTAATGGCTGC HR403 JH029 Forward PCR AAATTTCCTTTACCACTTGCTTACCTA HQ903 JH030 Reverse PCR TTTGAAGGCATGAAGGGCACTA HQ903 JH031 Forward PCR CCCTGAATTCTCTTTTATACCCTCTGA HR101 JH032 Reverse PCR TCACCCCCAAACTAAGCACACT HR101 JH033 Forward PCR CAGAGGGAGCAGACGAAAGAAG HR205 JH034 Reverse PCR TTGCCCAGAATGATCAGAACTAAT HR205 JH035 Forward PCR GCCTGCTATTTTTCCCGAGTCAA HR003 JH036 Reverse PCR GAAGCATCTGCAGAAACCAAAGG HR003

CGGGCGGGGAGcTCGGAAGTACTTTGTTTTT MIR17HG Mut JH049 Forward PCR TATGCTgggaAGGGAG Prom

CGGGCGGGGAGcTCGGAAGTACTTTGTTTTT MIR17HG JH050 Forward PCR TATGCTAATGAGGGAG Prom

GCGCTTTACTACGACCGGAGGCtCgAGGCC MIR17HG JH052 Reverse PCR GGGGG Prom JH054 Forward PCR TTAAAGAAACGAAAACAGGGAAACCA HR701 JH055 Reverse PCR GAATCCTTAATTCCCAAGCCAGTGC HR701 JH056 Forward PCR AATGCACTGCCCCAGAGATGTTC HR801 JH057 Reverse PCR TGAAACCAAGTGGCATGAAATAGGC HR801 JH058 Forward PCR TATCTCCAAAGAACACCACCACTACTCTGC HR901

178

JH059 Reverse PCR TCCCATGACAGACATAAGAGCAGATTCTG HR901 AAAAGGTAGGGAGAAAAATAAGGGAAGG Forward PCR JH060 TT HS001 CAGTTTCTTATATTCTTTGGCTCTTTGTTTCA Reverse PCR JH061 C HS001, HS101 JH062 Forward PCR CACACCCCAACAGCCCTTTCTAAG HS101 JH063 Forward PCR GCAACGTCAGGTGGGCTTTTACA HS201 JH064 Reverse PCR TCGCCCAGGGTTGATGTTTCTC HS201 JH065 Forward PCR CACCAGGGCAAAGCGAGAGG HS305 JH066 Reverse PCR TTTAGGGAGGTTCCTAGAGGACAAGTGA HS305 CCAAACTGATTAAATGAGAGCAGTGATAT Forward PCR JH070 GAAATG HS601 JH071 Reverse PCR TTTCCTTTCCTTTCCTTTCCTTTCCCTT HS601

GGGGACAACTTTTGTATACAAAGTTGTAGC Reverse PCR JH078 TGGTCACTTAGGGTTGG IG401

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR JH084 TAGGCCATGGATCCGAATTC HE401, HD601

GGGGACCACTTTGTACAAGAAAGCTGGGTT Reverse PCR JH085 CTCAGCACCTTCCCTCTAG HD601

GGGGACAACTTTTCTATACAAAGTTGTATT Forward PCR JH090 AACGCCATGTACCCATACGAT HD501, HP402

GGGGACAACTTTATTATACAAAGTTGTTCA Reverse PCR JH091 CCGGGGTTCCTGCG HD501, HO201

GGGGACAACTTTTCTATACAAAGTTGTAAC Forward PCR JH093 CATGGTCGTTTTACAAC HD301

GGGGACAACTTTATTATACAAAGTTGTCTT Reverse PCR JH094 ATTTTTGACACCAGACC HD301

GGGGACAACTTTGTATAATAAAGTTGTAGC Forward PCR JH095 TAGCGCTACCGGACTC HD201

GGGGACCACTTTGTACAAGAAAGCTGGGTT Reverse PCR JH096 CAAACCACAACTAGAATGCAGTG HD201

GGGGACAACTTTATTATACAAAGTTGTTCA Reverse PCR JH099 GGACGAGCCCTCGGC HP402

GGGGACAACTTTGTATACAAAAGTTGTAAG Forward PCR JH106 ACTGAGCTAAGAATCCAG HE302

GGGGACAACTTTTGTATACAAAGTTGTCTC Reverse PCR JH107 AGCACCTTCCCTCTAG HE401

179

GGGGACAACTTTGTATAGAAAAGTTGGGT Reverse PCR JH120 GGGGGTTGATAGGAAAGGTG HI901

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR JH121 TACTAGCTTCCTTAGCATGACG HI901

GGGGACAACTTTGTATACAAAAGTTGTACG Forward PCR JH122 ATCTATACATTGAATCAATATTG IH401

GGGGACAACTTTTCTATACAAAGTTGTTAC Forward PCR JH128 CATGGTGAGCAAG HJ601

GGGGACAACTTTATTATACAAAGTTGTCTT HJ601, IO102, JH129 Reverse PCR GTACAGCTCGTCCA IO401, IO601

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR JH137 TTACCATGGTGAGCAAG HJ201, II601

GGGGACCACTTTGTACAAGAAAGCTGGGTT Reverse PCR JH139 TACTTGTACAGCTCGTCCA HJ201

GGGGACAACTTTTCTATACAAAGTTGTCGA Forward PCR JH143 AATGTCCAATTTACTGACCGT HM001

GGGGACAACTTTATTATACAAAGTTGTCTA Reverse PCR JH144 ATCGCCATCTTCCAGCAG HM001

GGGGACAACTTTGTATACAAAAGTTGTCGC Forward PCR JH160 CACCATGGCATTACAA JD904

GGTTTCTGAGAATTCTGGAATTTCCTGAAC Forward PCR JH190 CACAGC HY001 JH191 Reverse PCR GCCGGCAGCACATGTCGCACAAG HY101 JH194 Forward PCR ATGTCTCCTGTCCCAAGATTCCCCA HR503, HR606 JH196 Reverse PCR TTGCACCATCATGCCCTCCTAAGC HR503, HR606

GGGGACAACTTTTCTATACAAAGTTGTTGT Forward PCR JH197 TATGAACAGTGAGGAGCA HO201

JH199 Reverse PCR GCGAAGCCTCCCGAGGGCCATT HY001 JH200 Forward PCR TGGGCGGGGCCTGCGGTGATT HY101

GTATGTACATTGGTACATATGTTGGTTGCTC SDM JH212 CTTACCTTC HW901

GAAGGTAAGGAGCAACCAACATATGTACC SDM JH213 AATGTACATAC HW901

GTATGTACATTGGTACATACGTTGGTTGCTC SDM JH214 CTTACCTTC HX002

GAAGGTAAGGAGCAACCAACGTATGTACC SDM JH215 AATGTACATAC HX002

180

GGGGACAACTTTTCTATACAAAGTTGCCAC Forward PCR JH235 CATGAACGGCGAGGAGCAG IG510, JH201

GGGGACAACTTTATTATACAAAGTTGCTCA Reverse PCR JH236 TCGTGGTTCCTGCGGCC IG510

GGGGACAACTTTGTATAATAAAGTTGGTGG Forward PCR JH250 CGGATCCTACCCATACGATGTTC IA001, IA102

GGGGACCACTTTGTACAAGAAAGCTGGGTT Reverse PCR JH251 CCTGCGGCCGCAAA IA102

GGGGACCACTTTGTACAAGAAAGCTGGGT Reverse PCR JH252 CAGGTACCTCCCCAGCA IA001

GGGGACAACTTTTCTATACAAAGTTGCCAC IA201, IA304, JH253 Forward PCR CATGGCCCCAAAG IH601

GGGGACAACTTTATTATACAAAGTTGCGGA Reverse PCR JH254 TCCCTTTTTCTTTTTTGC IA201, IH601

GGGGACAACTTTATTATACAAAGTTGCTCC Reverse PCR JH255 AAACAGAGATGTGTCGAAGATG IA304

cggccaggcaaaaaagaaaaagggcggaagcGATGCTAA Forward PCR JH282 GTCACTAACTGCCTG IB801

ACTTTATTATACAAAGTTGCGCTCCCTCCG Reverse PCR JH283 CCGGAACC IB801

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR JH284 CGTGAGGCTCCGGTGCCCG IF101, IY704

GGGGACAACTTTGTATAGAAAAGTTGGGT Reverse PCR JH285 ACCTAGCCAGCTTGGGTCTCC IF101, JZ901

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR JH286 AGGTACCTGAGGGCCTATTTCC JJ802

aggcaaaaaagaaaaaggGAGGAGGATCTTTATCT Forward PCR JH294 GGG IF311

ctttattatacaaagttgcgccCATGCTTGGGGACTGCT Reverse PCR JH295 G IF311

aggcaaaaaagaaaaagggaggcgggagcATTTTCAAA Forward PCR JH296 CCAGAAGAACTACGACAG IF405

ctttattatacaaagttgcgccGTCCTGGCTCTGCGTGT Reverse PCR JH297 G IF405

GGGGACAACTTTGTATACAAAAGTTGTGAT Forward PCR JH299 GGGAAAGTCTTACCCAACTGTGAG IF701

GGGGACAACTTTGTATAGAAAAGTTGGGT JH300 Reverse PCR GGCTGCCTCCACCGGCATCAGCAAACCCA IF701 AG

181

GGGGACAACTTTGTATACAAAAGTTGGCGC Forward PCR JH301 CATGGTGAGCAAG IF801, IO702

GGGGACAACTTTGTATAGAAAAGTTGGGT Reverse PCR JH302 GAGGACCTGGATTCTCCTCC IF801, II601

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR JH305 GGGCTGGAAGCTACCTTTGACA IG001

GGGGACAACTTTGTATAGAAAAGTTGGGT Reverse PCR JH306 GGTATCTTTCCATTTTTATAGCATGTCCTG IG001

GGGGACAACTTTTGTATACAAAGTTGGCCG Reverse PCR JH314 GATCTGACGGTTCACTAAACG IG301

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR JH315 GCCTCGAGTTTACTCCCTATCAGTG IG603, IG703

GGGGACAACTTTGTATAGAAAAGTTGGGTT Reverse PCR JH316 GGATCCAATTCTCCAGGCGATCTGAC IG101, IG603

GGGGACAACTTTGTATACAAAAGTTGGCCT Forward PCR JH317 CGAGTTTACTCCCTATCAGTG IG101

GGGGACAACTTTTGTATACAAAGTTGTGGA Reverse PCR JH318 TCCAATTCTCCAGGCGATCTGAC IG703

GGGGACAACTTTGTATAATAAAGTTGAGAT Forward PCR JH319 CAACAAAGCTTCCTAGCTCGA IG806

GGGGACCACTTTGTACAAGAAAGCTGGGTT Reverse PCR JH320 CTTCGAGTCGACAAGCCTGG IG806

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR JH343 GCAGTCTGGGCTTTCACAA II701

GGGGACAACTTTGTATAGAAAAGTTGGGT Reverse PCR JH344 GCAAGCTTAGGTCCCACGG II701

GGGGACAACTTTGTATAGAAAAGTTGGGT Reverse PCR JH403 GTGATCCGGCGCGCTTG IO702

IO001, IO102, GGGGACAACTTTTCTATACAAAGTTGCCAT IO201, IO301, JH404 Forward PCR GGTGAGCAAGG IO401, IO502, IO601

GGGGACAACTTTATTATACAAAGTTGTTTA IO001, IO201, JH405 Reverse PCR CTTGTACAGCTCGTCCA IO301, IO502

GGGGACAACTTTGTATAATAAAGTTGGAAT Forward PCR JH502 TCTGCAGTCGACGGTACC IR801

182

GGGGACCACTTTGTACAAGAAAGCTGGGT Reverse PCR JH503 CGCCTTAAGATACATTGATGAGTTTGGAC IR801

gcatttttttcactgcctcgagtctagaATAGTAATCAATT Forward PCR JH504 ACGGGGTC IR904

tcaccatggcgcgccagtcgaattcGAGATCTGAGTCCG Reverse PCR JH505 GTAG IR904

gatctgaacatgGAACAAAAACTCATCTCAGAA sense Ligation JH517 GAGGATCTGggaGGCG IX104

GATCCGCCtccCAGATCCTCTTCTGAGATGA antisense Ligation JH518 GTTTTTGTTCcatgttca IX104

gatctgacatgGGTAAGCCTATCCCTAACCCTCT sense Ligation JH519 CCTCGGTCTCGATTCTACGggaGGCG IX203

GATCCGCCtccCGTAGAATCGAGACCGAGG antisense Ligation JH520 AGAGGGTTAGGGATAGGCTTACCcatgtca IX203

gatccGAACAAAAACTCATCTCAGAAGAGGA sense Ligation JH521 TCTGtaag IX301, IZ702 aattcttaCAGATCCTCTTCTGAGATGAGTTTTT antisense Ligation JH522 GTTCg IX301, IZ702 JH523 sense Ligation gatccCACCATCATCACCATCACtaag IX401, IZ801 JH524 antisense Ligation aattcttaGTGATGGTGATGATGGTGg IX401, IZ801

gatccagcggcttcgccaacgagctgggacccaggctgatgggaa sense Ligation JH525 agtaag IX503, IZ901

aattcttactttcccatcagcctgggtcccagctcgttggcgaagccg antisense Ligation JH526 ctg IX503, IZ901 JH527 sense Ligation gatccgactacaaggatgacgacgataagtaag IX601, JA001 JH528 antisense Ligation aattcttacttatcgtcgtcatccttgtagtcg IX601, JA001

GTTGGTGGCggatccGGCGGAGGTGGATCTGG Forward PCR JH529 AGGTGGAGGCTCCTTCAAGAACCTGATC IY001, JD404

JH530 Reverse PCR gctctaggaattcttaGCTTCTTCTCAGGCTGAAC IY001, JD404

GTTGGTGGCggatccGGCGGAGGTGGATCTAA Forward PCR JH533 GTCTTACCCAACTGTGAGTG IY202, JD602

JH534 Reverse PCR gctctaggaattcttaGGCATCAGCAAACCCAAGC IY202, JD602

GGAagcggcttcgccaacgagctgggacccaggctgatgggaa sense Ligation JH535 agTAAC IX702

TCGAGTTActttcccatcagcctgggtcccagctcgttggcga antisense Ligation JH536 agccgctTCC IX702

ttacgcttaagaattcctagagctcctatagtgagACCGGTTA Forward NEBuilder JH543 GTAATGATCGACAATCA IX907

183

JH545 sense Ligation TCGAGTCTAGAGGTGGCGATGGTt IY601 JH546 antisense Ligation CCGGaACCATCGCCACCTCTAGAC IY601 JH547 sense Ligation CGATAAGCTTGTCTAGAGGATCCA IY503 JH548 antisense Ligation CTAGTGGATCCTCTAGACAAGCTTAT IY503 JH549 sense Ligation TCGAGTCTAGAGGAGtGC IZ501

GGGGACAACTTTTGTATACAAAGTTGACCT Reverse PCR JH551 AGCCAGCTTGGGTCTCC IY704

ATCCATTTTCGGATCTGATCAGCACCGCAT Forward PCR JH554 GATTGAACAAGATGGATTGC IZ401

TAAAGGTACCGAGCTCGAATTGTGCTTAGA Reverse PCR JH555 AGAACTCGTCAAGAAGGCGATAG IZ401

JH556 antisense Ligation TTAGCaCTCCTCTAGAC IZ501

GCCGCCTCCCCGCCTGGCGATGGTACtgtccta Forward NEBuilder JH557 gtacctgaCCCAGCTTTcttgtacaa IX907

GGGGACAACTTTGTATAGAAAAGTTGGGTT Reverse PCR JH567 ggtaccGCTAGTGGATCCGTTCAAGTC JD904

GGGGACAACTTTGTATAATAAAGTTGGATC Forward PCR JH569 CGGCGCTACTAACTTCAGCCTGCTGAAGc JG102

GGGGACCACTTTGTACAAGAAAGCTGGGT Reverse PCR JH570 ATGCGGGGAGGCGGCC JG102, JG202

GGGGACAACTTTGTATAATAAAGTTGgatcca Forward PCR JH573 tggtgagcaagggcgagg JG202

GGGGACAACTTTATTATACAAAGTTGATCG Reverse PCR JH574 TGGTTCCTGCGGCCG JH204

GGGGACAACTTTTGTATACAAAGTTGCAGC Reverse PCR JH580 GGCCGCAAAAATCTCGCC JJ802

JH595 Forward Ligation agcTGTAAAACGACGGCCAGTagct JK104 JH596 Reverse Ligation ACTGGCCGTCGTTTTACA JK104 JH597 Forward Ligation tcgaCATGGTCATAGCTGTTgacctcgagt JK104 JH598 Reverse Ligation actcgaggtcAACAGCTATGACCATG JK104

JK305, JL002, acatggtcatagctgttgacGGTGTGGAAAGTCCCCA JL202, JL302, JH599 Forward PCR G JN202, JN302, JT504, JT601

184

JK305, JL002, JL202, JL302, JH600 Reverse PCR GGTTTAGTTCCTCACCTTGTC JN202, JN302, JT504, JT601

JK305, JL002, JL202, JL302, JH601 Forward PCR CGTACCGGTTAGTAATGAGTTTAAAC JN202, JN302, JT504, JT601

acaaggtgaggaactaaaccATGGCCAAGCCTTTGTC Forward PCR JH603 TC JK305, JN202 actcattactaaccggtacgTTAGCCCTCCCACACAT Reverse PCR JH604 AAC JK305, JN202 acaaggtgaggaactaaaccATGGTGAGCAAGGGCG Forward PCR JH607 AG JL002, JT504 actcattactaaccggtacgTTACTTGTACAGCTCGTC Reverse PCR JH608 CATG JL002, JT504 JH613 Forward PCR acaaggtgaggaactaaaccatggtgagcaagggcgag JL202, JT601

actcattactaaccggtacgGTTTACTTGTACAGCTCG Reverse PCR JH614 TCCATG JL202, JT601

JK305, JL002, GTCTATTCTTTTGATTTAacCGGATCTGCTAT JL202, JL302, JH615 Reverse PCR GGCAGG JN202, JN302, JT504, JT601

JH616 Forward PCR acaaggtgaggaactaaaccatgaccgagtacaagcctac JL302, JN302 JH617 Reverse PCR actcattactaaccggtacgTTAGGCGCCAGGTTTTCG JL302, JN302

gcatttttttcactgcctcgagtctagaCGTGAGGCTCCGG Forward PCR JH623 TGCCCG JN511

tcaccatggcgcgccagtcgaattcACCTAGCCAGCTTG Reverse PCR JH624 GGTCTCC JN511

AGCTggatccTACCCATACGATGTGCCAGATT HKA701, JH625 Forward Ligation ACGCTtgaTTT KA801, KA901

AAAtcaAGCGTAATCTGGCACATCGTATGGG HKA701, JH626 Reverse Ligation TAggatcc KA801, KA901

GGGGACAACTTTTCTATACAAAGTTGccatgg Forward PCR JH628 aagatgccaaaaacattaagaagg JN402

GGGGACAACTTTATTATACAAAGTTGtttacac Reverse PCR JH629 ggcgatcttgccgc JN402

185

KA401, KA501, JH653 Forward Ligation AATTCGACTGGCGCGCCACCATGTCC K601 KA401, KA501, JH654 Reverse Ligation GACATGGTGGCGCGCCAGTCG K601

GGGGACAACTTTGTATACAAAAGTTGTACG Forward PCR JH697 TGAGGCTCCGGTGCCCG JZ901

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR TB1523 TACGATCTATACATTGAATCAATATTG HB401, IG301

GGGGACAACTTTGTATAGAAAAGTTGGGT Reverse PCR TB1524 GCCGGATCTGACGGTTCACTAAACG HB401, IH401

GGGGACAAGTTTGTACAAAAAAGCAGGCT Forward PCR TB1525 TAAGACTGAGCTAAGAATCCAG HB501, IG401

GGGGACAACTTTGTATAGAAAAGTTGGGT Reverse PCR TB1526 GAGCTGGTCACTTAGGGTTGG HB501, HE302

GGGGACAACTTTGTATAATAAAGTTGTACA Forward PCR TB1533 GTGCCTCTCCTGGCCTTGG HB901

GGGGACCACTTTGTACAAGAAAGCTGGGTT Reverse PCR TB1534 CAACAGGCATCTACTGAGTGG HB901

186

Appendix D

Sequences for synthesized double-stranded DNA gBlock elements utilized during the creation of the pMVP and pMAGIC components listed in Appendix B. For each sequence, the laboratory ID (JHXXX) in addition to the plasmids and vectors that resulted from these elements are recorded.

ID Sequence Used in: AATTAATACGACTCACTATAGGGAGACCCAAGCTGGCTAGTTAAGCTAT CGACAACTTTGTATACAAAAGTTGTAgaatacaagcttcttgttctttttgcagaagctcagaat JH321 aaacgctcaactttggcagatctgaacatggactacaaggatgacgacgataagggaGGCGGATCCCAC IG901 CCAACTTTTCTATACAAAGTTGTCGATCTAGAGGGCCCGCGGTTCGAAGG TAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATTCTACGCG

AATTAATACGACTCACTATAGGGAGACCCAAGCTGGCTAGTTAAGCTAT CGACAACTTTGTATACAAAAGTTGTAgaatacaagcttcttgttctttttgcagaagctcagaat aaacgctcaactttggcagatctgaacatgTACCCATACGATGTGCCAGATTACGCTggaG JH322 GCGGATCCCACCCAACTTTTCTATACAAAGTTGTCGATCTAGAGGGCCCG IH001 CGGTTCGAAGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATTCTAC GCG

AATTAATACGACTCACTATAGGGAGACCCAAGCTGGCTAGTTAAGCTAT CGACAACTTTGTATACAAAAGTTGTAgaatacaagcttcttgttctttttgcagaagctcagaat aaacgctcaactttggcagatctgaacatgagcggcttcgccaacgagctgggacccaggctgatgggaaaggga JH323 GGCGGATCCCACCCAACTTTTCTATACAAAGTTGTCGATCTAGAGGGCCC IH101 GCGGTTCGAAGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATTCTA CGCG AATTAATACGACTCACTATAGGGAGACCCAAGCTGGCTAGTTAAGCTAT CGACAACTTTGTATACAAAAGTTGTAgaatacaagcttcttgttctttttgcagaagctcagaat aaacgctcaactttggcagatctgaacatgTATCCATACGACGTGCCAGATTATGCTTACC JH324 CTTACGATGTCCCAGACTACGCCTACCCATATGATGTGCCCGATTACGCT IH201 ggaGGCGGATCCCACCCAACTTTTCTATACAAAGTTGTCGATCTAGAGGG CCCGCGGTTCGAAGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATT CTACGCG AATTAATACGACTCACTATAGGGAGACCCAAGCTGGCTAGTTAAGCTAT CGACAACTTTGTATACAAAAGTTGTAgaatacaagcttcttgttctttttgcagaagctcagaat aaacgctcaactttggcagatctgaacATGATCAGTCTGATTGCGGCGTTAGCGGTAGA TTACGTTATCGGCATGGAAAACGCCATGCCGTGGAACCTGCCTGCCGAT CTCGCCTGGTTTAAACGCAACACCTTAAATAAACCCGTGATTATGGGCC GCCATACCTGGGAATCAATCGGTCGTCCGTTGCCAGGACGCAAAAATAT TATCCTCAGCAGTCAACCGAGTACGGACGATCGCGTAACGTGGGTGAAG JH345 TCGGTGGATGAAGCCATCGCGGCGTGTGGTGACGTACCAGAAATCATGG II801 TGATTGGCGGCGGTCGCGTTATTGAACAGTTCTTGCCAAAAGCGCAAAA ACTGTATCTGACGCATATCGACGCAGAAGTGGAAGGCGACACCCATTTC CCGGATTACGAGCCGGATGACTGGGAATCGGTATTCAGCGAATTCCACG ATGCTGATGCGCAGAACTCTCACAGCTATTGCTTTGAGATTCTGGAGCGG CGAGGCGGAGGCGGATCCCACCCAACTTTTCTATACAAAGTTGTCGATCT AGAGGGCCCGCGGTTCGAAGGTAAGCCTATCCCTAACCCTCTCCTCGGT CTCGATTCTACGCG

187

AATTAATACGACTCACTATAGGGAGACCCAAGCTGGCTAGTTAAGCTAT CGACAACTTTGTATACAAAAGTTGTAgaatacaagcttcttgttctttttgcagaagctcagaat aaacgctcaactttggcagatctgaacatgggagtgcaggtggaaaccatctccccaggagacgggcgcaccttcc ccaagcgcggccagacctgtgtggtgcactacaccgggatgcttgaagatggaaagaaagtcgattcctcccggg JH346 acagaaacaagccctttaagtttatgctaggcaagcaggaggtgatccgaggctgggaagaaggggttgcccag II901 atgagtgtgggtcagagagccaaactgactatatctccagattatgcctatggtgccactgggcacccaggcatcat cccaccacatgccactctcgtcttcgatgtggagcttctaaaaccggaaGGCGGAGGCGGATCCCAC CCAACTTTTCTATACAAAGTTGTCGATCTAGAGGGCCCGCGGTTCGAAGG TAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATTCTACGCG

AATTAATACGACTCACTATAGGGAGACCCAAGCTGGCTAGTTAAGCTAT CGACAACTTTGTATACAAAAGTTGTAgaatacaagcttcttgttctttttgcagaagctcagaat aaacgctcaactttggcagatctgaacatggactacaaagaccatgacggcgattataaagatcatgacatcgatta JH461 caaggatgacgatgacaagggaGGCGGATCCCACCCAACTTTTCTATACAAAGTTGT IO801 CGATCTAGAGGGCCCGCGGTTCGAAGGTAAGCCTATCCCTAACCCTCTC CTCGGTCTCGATTCTACGCG

AATTAATACGACTCACTATAGGGAGACCCAAGCTaGCTAGTTAAGCTATC GACAACTTTGTATACAAAAGTTGTAgaatacaagcttcttgttctttttgcagaagctcagaata aacgctcaactttggcagatctgaacatggccTTCAAGAACCTGATCTGGCTGAAGGAGGT GGACAGCACCCAGGAGAGACTcAAGGAGTGGAACGTGAGCTACGGCAC CGCCCTtGTGGCCGACAGACAGACCAAGGGCAGAGGCGGCCTGGGCAG AAAGTGGCTtAGCCAGGAGGGCGGCCTGTACTTCAGCTTCCTGCTcAACC CCAAGGAGTTCGAGAACCTGCTtCAGCTGCCCCTcGTGCTtGGCCTGAGCG TGAGCGAGGCCCTGGAGGAGATCACCGAGATCCCCTTCAGCCTcAAGTG GCCCAACGACGTGTACTTCCAGGAGAAGAAGGTGAGCGGCGTGCTGTGC GAGCTtAGCAAGGACAAGCTGATCGTGGGCATCGGCATCAACGTGAACC JH462 AGAGAGAGATCCCCGAGGAGATCAAGGACAGAGCCACCACCCTcTACG IO901 AGATCACCGGCAAGGACTGGGACAGAAAGGAGGTGCTcCTtAAGGTGCT GAAGAGAATCAGCGAGAACCTcAAGAAGTTCAAGGAGAAGAGCTTCAA GGAGTTCAAGGGCAAGATCGAGAGCAAGATGCTGTACCTGGGCGAGGA GGTGAAGCTtCTGGGCGAGGGCAAGATCACCGGCAAGCTGGTGGGCCTt AGCGAGAAGGGCGGCGCCCTGATCCTGACCGAGGAGGGCATCAAGGAG ATCCTtAGCGGCGAGTTCAGCCTGAGAAGAAGCGGCGGAGGCGGATCTG GGGGTGGAGGCTCTGGTGGAGGGGGATCCCACCCAACTTTTCTATACAA AGTTGTCGATCTAGAGGGCCCGCGGTTCGAAGGTAAGCCTATCCCTAAC CCTCTCCTCGGTCTCGATTCTACGCG

188

gccggccaggcaaaaaagaaaaaggGAGGGGGAGGCAGCAACCATGACCAGGAATT TGACCCtCCAAAGGTTTACCCACCTGTGCCAGCTGAGAAGAGGAAGCCC ATCCGCGTGCTGTCTCTCTTTGATGGGATTGCTACAGGGCTCCTGGTGCTG AAGGACCTGGGCATCCAAGTGGACCGCTACATTGCCTCCGAGGTGTGTG AGGACTCCATCACaGTGGGCATGGTGCGGCACCAGGGAAAGATCATGTA CGTCGGGGACGTCCGCAGCGTCACACAGAAGCATATCCAGGAGTGGGG CCCATTCGACCTGGTGATTGGAGGCAGTCCCTGCAATGACCTCTCCATTG TCAACCCTGCCCGCAAGGGACTTTATGAGGGTACTGGCCGCCTCTTCTTT GAGTTCTACCGCCTCCTGCATGATGCtCGGCCCAAGGAGGGAGATGATCG CCCCTTCTTCTGGCTCTTTGAGAATGTGGTGGCCATGGGCGTTAGTGACA AGAGGGACATCTCcCGATTTCTTGAGTCTAACCCCGTGATGATTGACGCC AAAGAAGTGTCTGCTGCACACAGGGCCaGaTACTTCTGGGGTAACCTTCC TGGCATGAACAGGCCTTTGGCATCCACTGTGAATGATAAGCTGGAGCTG CAAGAGTGTCTGGAGCACGGCAGAATAGCCAAGTTCAGCAAAGTGAGG ACCATTACCACCAGGTCAAACTCTATAAAGCAGGGCAAAGACCAGCATT TCCCCGTCTTCATGAACGAGAAGGAGGACATCCTGTGGTGCACTGAAAT GGAAAGGGTGTTTGGCTTCCCCGTCCACTACACAGACGTCTCCAACATG JH463 AGCCGCTTGGCcAGGCAGAGACTGCTGGGCCGATCcTGGAGCGTGCCGGT IP803 CATCCGCCACCTCTTCGCTCCGCTGAAGGAATATTTTGCTTGTGTGAGCA GCGGCAACAGCAACGCCAACAGCCGCGGCCCCAGCTTCAGCAGCGGCC TGGTGCCCCTGAGCCTGCGCGGCAGCCACATGGGCCCTATGGAGATATA CAAGACAGTGTCTGCATGGAAGAGACAGCCAGTGCGGGTgCTGAGCCTg TTcAGAAAcATcGATAAAGTgCTgAAGAGTTTGGGCTTcTTGGAAAGCGGT TCTGGTTCTGGGGGAGGAACcCTGAAGTACGTGGAAGATGTCACAAATG TCGTGAGGAGAGACGTGGAGAAATGGGGCCCCTTTGACCTGGTGTACGG CTCcACtCAGCCCCTgGGCAGCTCTTGTGATCGCTGTCCCGGCTGGTACAT GTTCCAGTTCCACCGAATCCTGCAGTATGCcCTGCCTCGCCAGGAGAGTC AGCGGCCCTTCTTCTGGATATTCATGGACAATCTGCTGCTGACTGAGGAT GACCAAGAGACAACTACCCGCTTCCTTCAGACAGAGGCTGTGACCCTCC AGGATGTCaGaGGCAGAGACTACCAGAATGCTATGCGGGTGTGGAGCAA CATTCCAGGGCTGAAGAGCAAGCATGCtCCCCTGACCCCAAAGGAAGAA GAGTATCTGCAAGCCCAAGTCAGAAGCAGGAGCAAGCTGGACGCCCCG AAAGTTGACCTCCTGGTGAAGAACTGCCTTCTCCCGCTGAGAGAGTACTT CAAGTATTTTTCTCAAAACTCACTTCCTCTcgCAACTTTgtataataaagttg

189

ACGACTCACTATAGGGAGACCCAAGCTGGCTAGTTAAGCTATCGGACAA CTTTTCTATACAAAGTTGGAAGCATGTTGGCGGTCGGAGCTATGGAAGGT ACACGACAATCAGCCTTCTTGCTCAGCTCACCTCCCCTTGCTGCGCTTCA TTCAATGGCGGAAATGAAAACACCTTTGTACCCAGCCGCCTACCCGCCA CTCCCGGCTGGACCTCCCTCTAGTAGCTCCAGTTCATCTTCATCCAGTTCT CCATCACCTCCGCTGGGAACACACAACCCAGGTGGCCTCAAACCACCTG CTACAGGAGGATTGTCCTCACTGGGTTCACCACCACAACAACTTAGTGC AGCAACTCCTCATGGTATTAACGACATACTGTCCCGGCCGTCAATGCCTG TCGCGTCCGGTGCTGCGCTCCCGTCCGCGTCCCCATCTGGGTCCTCAAGT TCTTCATCTAGCAGTGCAAGTGCaTCtTCAGCCTCaGCAGCCGCtGCCGCTG CTGCaGCAGCCGCCGCAGCCGCATCtAGTCCaGCaGGaCTCCTTGCAGGaC TCCCaaGATTCTCTTCCCTCAGTCCCCCTCCCCCACCACCGGGACTCTACT TCTCCCCGAGTGCTGCCGCTGTTGCTGCGGTAGGCCGCTACCCTAAACCC JH464 CTTGCGGAGCTCCCCGGGCGGACACCCATCTTCTGGCCAGGCGTTATGC IH301 AGTCCCCACCGTGGAGAGATGCACGGCTGGCATGCACTCCTCACCAAGG TTCAATATTGCTGGATAAGGATGGGAAAAGGAAACACACTCGCCCAAC ATTTTCCGGTCAACAAATCTTTGCCTTGGAAAAGACCTTCGAACAAACCA AGTACCTGGCCGGCCCCGAAAGGGCGAGGCTGGCGTACTCTCTTGGAAT GACCGAAAGCCAGGTTAAGGTCTGGTTTCAAAATAGACGCACAAAATG GCGCAAAAAACATGCAGCTGAAATGGCAACCGCAAAGAAAAAACAGG ACTCAGAGACCGAACGGCTGAAAGGCGCGTCAGAGAACGAAGAGGAG GATGACGACTACAATAAGCCACTGGATCCTAACTCAGATGACGAGAAG ATTACTCAACTCCTCAAGAAGCACAAAAGTTCTAGTGGTGGTGGCGGCG GCCTTCTTCTTCATGCAAGCGAACCGGAGTCCTCTAGCTAGGCAACTTTG TATAATAAAGTTGTCCGATCTAGAGGGCCCGCGGTTCGAAGGTAAGCCT ATCCCTAACCCTCT

CTTTCGTCACAGGGTGCGGTCGCCCGGGCAGGGTATAACTCACCTGACA ATCAGAGGGCGAGGTATTCAGCTCAACGACGAGTCGGTGAGCTCCTCGC TTGGTCTCCGTCCGGACGGGACATTTCAGATCGGCGGCGCCGGCCGTCCT TCATTCACGCCTCGTCAGGCAATCCTAACTCTGCAGACCTCGTCCTCTGA GCCGCGCTCTGGAGGCATTGGAACTCTGCAATTTATTGAGGAGTTTGTGC CATCGGTCTACTTTAACCCCTTCTCGGGACCTCCCGGCCACTATCCGGAT CAATTTATTCCTAACTTTGACGCGGTAAAGGACTCGGCGGACGGCTACG ACTGAATGTTAAGTGGAGAGGCAGAGCAACTGCGCCTGAAACACCTGGT JH465 CCACTGTCGCCGCCACAAGTGCTTTGCCCGCGACTCCGGTGAGTTTTGCT IP901, IQ013 ACTTTGAATTGCCCGAGGATCATATCGAGGGCCCGGCGCACGGCGTCCG GCTTACCGCCCAGGGAGAGCTTGCCCGTAGCCTGATTCGGGAGTTTACCC AGCGCCCCCTGCTAGTTGAGCGGGACAGGGGACCCTGTGTTCTCACTGT GATTTGCAACTGTCCTAACCTTGGATTACATCAAGATCTTATTCCCTTTAA CTAATAAAAAAAAATAATAAAGCATCACTTACTTAAAATCAGTTAGCAA ATTTCTGTCCAGTTTATTCAGCAGCACCTCCTTGCCCTCCTCCCAcgtcgacg GCTCTGGTATTGCAGCTTCCTCCTGGCTGCAAACTTTCTCCACAATCTAA ATG

ACTCACTATAGGGCGAATTGGGTACGACCCAAGCaGGCTAGTTACCCAT ACGATGTGCCAGATTACGCTggaGGCaGATCaCACCTAAGCaCGCAAATGG JH512 GCGGTAGGCGTGCGTCTCACAGCAGCACCTCCTTGCCCTCCTCCCAcgtcga IX001 cgGCTCTGGTATTGCAGCTTCCTCCTGGCTGCAAACTTTCTCCACAATCTA AATGGAA

TTTATTTTTCAATTGCAGAAAATTTCGAATCATTTTTCATTCAGTAGTATA GCCCCAGAGACGGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATTG JH513 TAgaatacgagcttcttgttctttttgcagaagcacagaataaacgctcaactttggcagctgtgaacatTCTAC IX001 GAGCTCCAGCTTTTGTTCCCTTTAGT

190

GGACCCAGAATATTGGAACTTTAGAAATGGAGATCTTACTGAAGGCACA GCCTATACAAACGCTGTTGGATTTATGCCTAACCTATCAGCTTATCCAAA ATCTCACGGTAAAACTGCCAAAAGTAACATTGTCAGTCAAGTTTACTTA AACGGAGACAAAACTAAACCTGTAACACTAACCATTACACTAAACGGT ACACAGGAAACAGGAGACACAACTGCCTGCGACTGCAGAGGCGACTGC JH558 TTCTGCGGCCCAAGTGCATACTCTATGTCATTTTCATGGGACTGGTCTGG JC001 CCACAACTACATTAATGAAATATTTGCCACATCCTCTTACACTTTTTCATA CATTGCCCAAGAATAAAGAATCGTTTGTGTTATGTTTCAACGTGTTTATTT TTCAATTGCAGAAAATTTCGAATCATTTTTCATTCAGTAGTATAGCCCCA GAG GGACCCAGAATATTGGAACTTTAGAAATGGAGATCTTACTGAAGGCACA GCCTATACAAACGCTGTTGGATTTATGCCTAACCTATCAGCTTATCCAAA ATCTCACGGTAAAACTGCCAAAAGTAACATTGTCAGTCAAGTTTACTTA AACGGAGACAAAACTAAACCTGTAACACTAACCATTACACTAAACGGT ACACAGGAAACAGGAGACACAACTAGCTTCGGCAAGAAAAAGAAGAA JH559 AAAGAAAGACCCCGAGGCCCCAAGTGCATACTCTATGTCATTTTCATGG JC101 GACTGGTCTGGCCACAACTACATTAATGAAATATTTGCCACATCCTCTTA CACTTTTTCATACATTGCCCAAGAATAAAGAATCGTTTGTGTTATGTTTCA ACGTGTTTATTTTTCAATTGCAGAAAATTTCGAATCATTTTTCATTCAGTA GTATAGCCCCAGAG AACCCCGTGTATCCATATGACACGGAAACCGGTCCTCCAACTGTGCCTTT TCTTACTCCTCCCTTTGTATCCCCCAATGGGTTTCAAGAGAGTCCCCCTGG GGTACTCggagttcttactttaaaatgtttaaccccactaacaaccacaggcggatctctacagctaaaagtgg gagggggacttacagtggatgacactgatggtaccttacaagaaaacatacgtgctacagcacccattactaaaaat aatcactctgtagaactatccattggaaatggattagaaactcaaaacaataaactatgtgccaaattgggaaatgg gttaaaatttaacaacggtgacatttgtataaaggatagtattaacaccttatggactggaataaaccctccacctaac tgtcaaattgtggaaaacactaatacaaatgatggcaaacttactttagtattagtaaaaaatggagggcttgttaat JH561 ggctacgtgtctctagttggtgtatcagacactgtgaaccaaatgttcacacaaaagacagcaaacatccaattaaga JC201 ttatattttgactcttctggaaatctattaactgaggaatcagacttaaaaattccacttaaaaataaatcttctacagcg accagtgaaactgtagccagcagcaaagcctttatgccaagtactacagcttatcccttcaacaccactactagggat agtgaaaactacattcatggaatatgttactacatgactagttatgatagaagtctatttcccttgaacatttctataatg ctaaacagccgtatgatttcttccaatgttgcctatgccatacaatttgaatggaatctaaatgcaagtgaatctccaga aagcaacatagctacgctgaccacatccccctttttcttttcttacattacagaagacgacaacTAAAGAATCG TTTGTGTTATGTTTCAACGTGTTTATTTTTCAATTGCAGAAAATTTCGAAT CATTTTTCATTCAGTAGTATAGCCCCAGAG

GACTTGAACGGATCCACTAGCggtacaGCCGGATCAGGAGCTACTAATTTC TCTCTGTTAAAGCAGGCAGGAGATGTGGAGGAGAATCCAGGTCCTAAGG AGAAGAGTGCTTGTCCTAAAGATCCAGCCAAACCTCCGGCCAAGGCACA JH566 AGTTGTGGGATGGCCACCGGTGAGATCATACCGGAAGAACGTGATGGTT JE301 TCCTGCCAAAAATCAAGCGGTGGCCCGGAGGCGGCGGCGTTCGTGAAG GTATCAATGGACGGAGCACCGTACTTGAGGAAAATCGATTTGAGGATGT ATAAAggcagtaccaACCCAACTTTtctataca

ACCCAGAATATTGGAACTTTAGAAATGGAGATCTTACTGAAGGCACAGC CTATACAAACGCTGTTGGATTTATGCCTAACCTATCAGCTTATCCAAAAT CTCACGGTAAAACTGCCAAAAGTAACATTGTCAGTCAAGTTTACTTAAA CGGAGACAAAACTAAACCTGTAACACTAACCATTACACTAAACGGTAC ACAGGAAACAGGAGACACAACTGCCTGCGACTGCAGAGGCGACTGCTT JH575 CTGCGGCCCAAGTGCATACTCTATGTCATTTTCATGGGACTGGTCTGGCC JH401 ACAACTACATTAATGAAATATTTGCCACATCCTCTTACACTTTTTCATACA TTGCCCAAGAAGGCAGCGGCAGCGGTTCGGGCAGTGGTAGCAAAAAAA AGAAAAAGAAAAAATAAAGAATCGTTTGTGTTATGTTTCAACGTGTTTAT TTTTCAATTGCAGAAAATTTCGAATCATTTTTCATTCAGTAGTATAGCCCC A

191

CTCGAGGTCGACGGTATCGATAAGCCCGAGCAGCCGGAGCAGAGAGCC TTGGGTCCGGTTTCTATGCCAAACCTTGTACCGGAGGTGATCGATCTTAC CTGCCACGAGGCTGGCTTTCCACCCAGTGACGACGAGGATGAAGAGGGT GAGGAGTTTGTGTTAGATTATGTGGAGCACCCCGGGCACGGTTGCAGGT CTTGTCATTATCACCGGAGGAATACGCAGAAAAAGTTTCTTTGCGATCGC JH576 ACCCTTTGGCGCATCCCATTCTCCAGTAACTTTATGTCCATGGGCGCACT JI401 CACAGACCTGGGCCAAAACCTTCTCTACGCCAACTCCGCCCACGCGCTA GACATGACTTTTGAGGTGGATCCCATGGACGAGCCCACCCTTCTTTATGT TTTGTTTGAAGTCTTTGACGTGGTCCGCTAGTTCTAGAGCGGCCGCCACC GC tatatatcttGTGGAAAGGACGAAACACCgGAGACCacggcaGGTCTCagtttAagtac JH578 tctgTGCTGgaaaCAGCAcagaatctactaaacaaggcaaaatgccgtgtttatctcgtcaacttgttggcga JI501 gatttttgcggccgcaggaaCCCAGCTTTcttgtacaaagttggcattataagaaagcatt

tatatatcttGTGGAAAGGACGAAACACCgGAGACCacggcaGGTCTCagtttAagtac JI601, JI702, JH579 tctgTGCTGgaaaCAGCAcagaatctactaaacaaggcaaaatgccgtgtttatctcgtcaacttgttggcga KB401, KB501, gatttttgcggccgctgtacaTACGACTCTagAAAGActgagctaagaatccagcta KB601, KB701

GCGGTGGCGGCCGCTCTAGAACTAGtgtacaccctatacagttgaagtcggaagtttacataca cttaagttggagtcattaaaactcgtttttcaactactccacaaatttcttgttaacaaacaatagttttggcaagtcagtt JH620 aggacatctactttgtgcatgacacaagtcatttttccaacaattgtttacagacagattatttcacttataattcactgtat JN607 cacaattccagtgggtcagaagtttacatacactaagttGGGCTGCAGGAATTCGATATC

GTTCCAGTTTGGAACAAGAGTCCACtgtacattgagtgtatgtaaacttctgacccactgggaa tgtgatgaaagaaataaaagctgaaatgaatcattctctctactattattctgatatttcacattcttaaaataaagtggt JH621 gatcctaactgacctaagacagggaatttttactaggattaaatgtcaggaattgtgaaaaagtgagtttaaatgtatt JN607 tggctaaggtgtatgtaaacttccgacttcaactgtatagggattgtacaAAGGGCGAAAAACCGTCT ATCAGGG ataatgccaactttgtataatAAAGTTGGAGGATCCGGCGCTACTAACTTCAGCCTGC JH627 TGAAGCAGGCTGGAGACGTGGAGGAGAACCCTGGACCTGTCGCCACCA JP801 TGGTGAGCAAGGGCGAGGAG

192

References

Aamodt, K. I., and A. C. Powers. 2017. "Signals in the pancreatic islet microenvironment influence beta-cell proliferation." Diabetes Obes Metab 19 Suppl 1:124-136. doi: 10.1111/dom.13031.

Abel, J. J. 1926. "Crystalline Insulin." Proc Natl Acad Sci U S A 12 (2):132-6.

Ahlgren, U., J. Jonsson, L. Jonsson, K. Simu, and H. Edlund. 1998. "beta-cell-specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the beta-cell phenotype and maturity onset diabetes." Genes Dev 12 (12):1763-8.

Akerman, I., Z. Tu, A. Beucher, D. M. Y. Rolando, C. Sauty-Colace, M. Benazra, N. Nakic, J. Yang, H. Wang, L. Pasquali, I. Moran, J. Garcia-Hurtado, N. Castro, R. Gonzalez-Franco, A. F. Stewart, C. Bonner, L. Piemonti, T. Berney, L. Groop, J. Kerr-Conte, F. Pattou, C. Argmann, E. Schadt, P. Ravassard, and J. Ferrer. 2017. "Human Pancreatic beta Cell lncRNAs Control Cell-Specific Regulatory Networks." Cell Metab 25 (2):400-411. doi: 10.1016/j.cmet.2016.11.016.

Albers, J., C. Danzer, M. Rechsteiner, H. Lehmann, L. P. Brandt, T. Hejhal, A. Catalano, P. Busenhart, A. F. Goncalves, S. Brandt, P. K. Bode, B. Bode-Lesniewska, P. J. Wild, and I. J. Frew. 2015. "A versatile modular vector system for rapid combinatorial mammalian genetics." J Clin Invest 125 (4):1603-19. doi: 10.1172/JCI79743.

An, J., D. M. Muoio, M. Shiota, Y. Fujimoto, G. W. Cline, G. I. Shulman, T. R. Koves, R. Stevens, D. Millington, and C. B. Newgard. 2004. "Hepatic expression of malonyl-CoA decarboxylase reverses muscle, liver and whole-animal insulin resistance." Nat Med 10 (3):268-74. doi: 10.1038/nm995.

An, J., L. Wang, M. L. Patnode, V. K. Ridaura, J. M. Haldeman, R. D. Stevens, O. Ilkayeva, J. R. Bain, M. J. Muehlbauer, E. L. Glynn, S. Thomas, D. Muoio, S. A. Summers, J. E. Vath, T. E. Hughes, J. I. Gordon, and C. B. Newgard. 2018. "Physiological mechanisms of sustained fumagillin-induced weight loss." JCI Insight 3 (5). doi: 10.1172/jci.insight.99453.

193

Bain, J. R., J. C. Schisler, K. Takeuchi, C. B. Newgard, and T. C. Becker. 2004. "An adenovirus vector for efficient RNA interference-mediated suppression of target genes in insulinoma cells and pancreatic islets of langerhans." Diabetes 53 (9):2190-4.

Banaszynski, L. A., L. C. Chen, L. A. Maynard-Smith, A. G. Ooi, and T. J. Wandless. 2006. "A rapid, reversible, and tunable method to regulate protein function in living cells using synthetic small molecules." Cell 126 (5):995-1004. doi: 10.1016/j.cell.2006.07.025.

Banting, F. G. 1925. "Diabetes and Insulin." Nobel Lecture: Physiology or Medicine 1923, Stockholm, Sweden.

Banting, F.G., and C.H. Best. 1922. "The internal secretion of the pancreas." The Journal of Laboratory and Clinical Medicine 7 (5):251-66.

Banting, F.G., C.H. Best, J.B. Collip, W.R. Campbell, and A.A. Fletcher. 1922a. "Pancreatic Extracts in the Treatment of Diabetes Mellitus." Can Med Assoc J 12 (3):141-6.

Banting, F.G., C.H. Best, J.B. Collip, J.J.R. Macleod, and E.C. Noble. 1922b. "The effect of pancreatic extract (insulin) on normal rabbits." The American Journal of Physiology 62 (1):162-76.

Barach, J. H. 1952. "Paul Langerhans, 1847-1888." Diabetes 1 (5):411-3.

Baran-Gale, J., E. E. Fannin, C. L. Kurtz, and P. Sethupathy. 2013. "Beta cell 5'-shifted isomiRs are candidate regulatory hubs in type 2 diabetes." PLoS One 8 (9):e73240. doi: 10.1371/journal.pone.0073240.

Baum, J., B. E. Simons, Jr., R. H. Unger, and L. L. Madison. 1962. "Localization of glucagon in the alpha cells in the pancreatic islet by immunofluorescent technics." Diabetes 11:371-4.

194

Becker, T. C., H. BeltrandelRio, R. J. Noel, J. H. Johnson, and C. B. Newgard. 1994. "Overexpression of hexokinase I in isolated islets of Langerhans via recombinant adenovirus. Enhancement of glucose metabolism and insulin secretion at basal but not stimulatory glucose levels." J Biol Chem 269 (33):21234-8.

Beerli, R. R., D. J. Segal, B. Dreier, and C. F. Barbas, 3rd. 1998. "Toward controlling gene expression at will: specific regulation of the erbB-2/HER-2 promoter by using polydactyl zinc finger proteins constructed from modular building blocks." Proc Natl Acad Sci U S A 95 (25):14628-33.

Belgardt, B. F., K. Ahmed, M. Spranger, M. Latreille, R. Denzler, N. Kondratiuk, F. von Meyenn, F. N. Villena, K. Herrmanns, D. Bosco, J. Kerr-Conte, F. Pattou, T. Rulicke, and M. Stoffel. 2015. "The microRNA-200 family regulates pancreatic beta cell survival in type 2 diabetes." Nat Med 21 (6):619-27. doi: 10.1038/nm.3862.

Berkner, K. L., and P. A. Sharp. 1983. "Generation of adenovirus by transfection of plasmids." Nucleic Acids Res 11 (17):6003-20.

Bernstein, D. L., J. E. Le Lay, E. G. Ruano, and K. H. Kaestner. 2015. "TALE-mediated epigenetic suppression of CDKN2A increases replication in human fibroblasts." J Clin Invest 125 (5):1998-2006. doi: 10.1172/JCI77321.

Bonnefond, A., and P. Froguel. 2015. "Rare and common genetic events in type 2 diabetes: what should biologists know?" Cell Metab 21 (3):357-68. doi: 10.1016/j.cmet.2014.12.020.

Bowker, S. L., S. R. Majumdar, P. Veugelers, and J. A. Johnson. 2006. "Increased cancer- related mortality for patients with type 2 diabetes who use sulfonylureas or insulin." Diabetes Care 29 (2):254-8.

Boyer, D. F., Y. Fujitani, M. Gannon, A. C. Powers, R. W. Stein, and C. V. Wright. 2006. "Complementation rescue of Pdx1 null phenotype demonstrates distinct roles of proximal and distal cis-regulatory sequences in pancreatic and duodenal expression." Dev Biol 298 (2):616-31. doi: 10.1016/j.ydbio.2006.07.020.

195

Bramson, J. L., N. Grinshtein, R. A. Meulenbroek, J. Lunde, D. Kottachchi, I. A. Lorimer, B. J. Jasmin, and R. J. Parks. 2004. "Helper-dependent adenoviral vectors containing modified fiber for improved transduction of developing and mature muscle cells." Hum Gene Ther 15 (2):179-88. doi: 10.1089/104303404772679986.

Caicedo, A. 2013. "Paracrine and autocrine interactions in the human islet: more than meets the eye." Semin Cell Dev Biol 24 (1):11-21. doi: 10.1016/j.semcdb.2012.09.007.

Campbell, J. E., and D. J. Drucker. 2015. "Islet alpha cells and glucagon--critical regulators of energy homeostasis." Nat Rev Endocrinol 11 (6):329-38. doi: 10.1038/nrendo.2015.51.

Carstensen, B., S. H. Read, S. Friis, R. Sund, I. Keskimaki, A. M. Svensson, R. Ljung, S. H. Wild, J. J. Kerssens, J. L. Harding, D. J. Magliano, S. Gudbjornsdottir, Diabetes, and Consortium Cancer Research. 2016. "Cancer incidence in persons with type 1 diabetes: a five-country study of 9,000 cancers in type 1 diabetic individuals." Diabetologia 59 (5):980-8. doi: 10.1007/s00125-016-3884-9.

Cavelti-Weder, C., M. Shtessel, J. E. Reuss, A. Jermendy, T. Yamada, F. Caballero, S. Bonner-Weir, and G. C. Weir. 2013. "Pancreatic duct ligation after almost complete beta-cell loss: exocrine regeneration but no evidence of beta-cell regeneration." Endocrinology 154 (12):4493-502. doi: 10.1210/en.2013-1463.

Centers for Disease Control and Prevention. 2017. National Diabetes Statistics Report, 2017. Atlanta, GA: Centers for Disease Control and Prevention, U.S. Dept of Health and Human Services.

Chavez, A., J. Scheiman, S. Vora, B. W. Pruitt, M. Tuttle, P. R. Iyer E, S. Lin, S. Kiani, C. D. Guzman, D. J. Wiegand, D. Ter-Ovanesyan, J. L. Braff, N. Davidsohn, B. E. Housden, N. Perrimon, R. Weiss, J. Aach, J. J. Collins, and G. M. Church. 2015. "Highly efficient Cas9-mediated transcriptional programming." Nat Methods 12 (4):326-8. doi: 10.1038/nmeth.3312.

Chen, B., L. A. Gilbert, B. A. Cimini, J. Schnitzbauer, W. Zhang, G. W. Li, J. Park, E. H. Blackburn, J. S. Weissman, L. S. Qi, and B. Huang. 2013a. "Dynamic imaging of

196

genomic loci in living human cells by an optimized CRISPR/Cas system." Cell 155 (7):1479-91. doi: 10.1016/j.cell.2013.12.001.

Chen, X., J. L. Zaro, and W. C. Shen. 2013b. "Fusion protein linkers: property, design and functionality." Adv Drug Deliv Rev 65 (10):1357-69. doi: 10.1016/j.addr.2012.09.039.

Chen, Y., L. Tian, S. Wan, Y. Xie, X. Chen, X. Ji, Q. Zhao, C. Wang, K. Zhang, J. M. Hock, H. Tian, and X. Yu. 2016. "MicroRNA-17-92 cluster regulates pancreatic beta-cell proliferation and adaptation." Mol Cell Endocrinol 437:213-223. doi: 10.1016/j.mce.2016.08.037.

Cho, S. W., S. Kim, J. M. Kim, and J. S. Kim. 2013. "Targeted genome engineering in human cells with the Cas9 RNA-guided endonuclease." Nat Biotechnol 31 (3):230- 2. doi: 10.1038/nbt.2507.

Cnop, M., N. Welsh, J. C. Jonas, A. Jorns, S. Lenzen, and D. L. Eizirik. 2005. "Mechanisms of pancreatic beta-cell death in type 1 and type 2 diabetes: many differences, few similarities." Diabetes 54 Suppl 2:S97-107.

Cong, L., F. A. Ran, D. Cox, S. Lin, R. Barretto, N. Habib, P. D. Hsu, X. Wu, W. Jiang, L. A. Marraffini, and F. Zhang. 2013. "Multiplex genome engineering using CRISPR/Cas systems." Science 339 (6121):819-23. doi: 10.1126/science.1231143.

Conway, A. E., J. M. Haldeman, D. S. Wechsler, and C. P. Lavau. 2015. "A critical role for CRM1 in regulating HOXA gene transcription in CALM-AF10 leukemias." Leukemia 29 (2):423-32. doi: 10.1038/leu.2014.221.

Diabetes Genetics Initiative of Broad Institute of Harvard and MIT, Lund University, and Novartis Institutes of BioMedical Research,, R. Saxena, B. F. Voight, V. Lyssenko, N. P. Burtt, P. I. de Bakker, H. Chen, J. J. Roix, S. Kathiresan, J. N. Hirschhorn, M. J. Daly, T. E. Hughes, L. Groop, D. Altshuler, P. Almgren, J. C. Florez, J. Meyer, K. Ardlie, K. Bengtsson Bostrom, B. Isomaa, G. Lettre, U. Lindblad, H. N. Lyon, O. Melander, C. Newton-Cheh, P. Nilsson, M. Orho- Melander, L. Rastam, E. K. Speliotes, M. R. Taskinen, T. Tuomi, C. Guiducci, A. Berglund, J. Carlson, L. Gianniny, R. Hackett, L. Hall, J. Holmkvist, E. Laurila, M. 197

Sjogren, M. Sterner, A. Surti, M. Svensson, M. Svensson, R. Tewhey, B. Blumenstiel, M. Parkin, M. Defelice, R. Barry, W. Brodeur, J. Camarata, N. Chia, M. Fava, J. Gibbons, B. Handsaker, C. Healy, K. Nguyen, C. Gates, C. Sougnez, D. Gage, M. Nizzari, S. B. Gabriel, G. W. Chirn, Q. Ma, H. Parikh, D. Richardson, D. Ricke, and S. Purcell. 2007. "Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels." Science 316 (5829):1331-6. doi: 10.1126/science.1142358.

DiGruccio, M. R., A. M. Mawla, C. J. Donaldson, G. M. Noguchi, J. Vaughan, C. Cowing- Zitron, T. van der Meulen, and M. O. Huising. 2016. "Comprehensive alpha, beta and delta cell transcriptomes reveal that ghrelin selectively activates delta cells and promotes somatostatin release from pancreatic islets." Mol Metab 5 (7):449- 458. doi: 10.1016/j.molmet.2016.04.007.

Dixon, J. R., S. Selvaraj, F. Yue, A. Kim, Y. Li, Y. Shen, M. Hu, J. S. Liu, and B. Ren. 2012. "Topological domains in mammalian genomes identified by analysis of chromatin interactions." Nature 485 (7398):376-80. doi: 10.1038/nature11082.

Dor, Y., J. Brown, O. I. Martinez, and D. A. Melton. 2004. "Adult pancreatic beta-cells are formed by self-duplication rather than stem-cell differentiation." Nature 429 (6987):41-6. doi: 10.1038/nature02520.

Dorrell, C., M. T. Grompe, F. C. Pan, Y. Zhong, P. S. Canaday, L. D. Shultz, D. L. Greiner, C. V. Wright, P. R. Streeter, and M. Grompe. 2011. "Isolation of mouse pancreatic alpha, beta, duct and acinar populations with cell surface markers." Mol Cell Endocrinol 339 (1-2):144-50. doi: 10.1016/j.mce.2011.04.008.

Drucker, D. J., J. F. Habener, and J. J. Holst. 2017. "Discovery, characterization, and clinical development of the glucagon-like peptides." J Clin Invest 127 (12):4217- 4227. doi: 10.1172/JCI97233.

Dutta, S., S. Bonner-Weir, M. Montminy, and C. Wright. 1998. "Regulatory factor linked to late-onset diabetes?" Nature 392 (6676):560. doi: 10.1038/33311.

Dutta, S., M. Gannon, B. Peers, C. Wright, S. Bonner-Weir, and M. Montminy. 2001. "PDX:PBX complexes are required for normal proliferation of pancreatic cells 198

during development." Proc Natl Acad Sci U S A 98 (3):1065-70. doi: 10.1073/pnas.031561298.

Eckel, R. H., S. E. Kahn, E. Ferrannini, A. B. Goldfine, D. M. Nathan, M. W. Schwartz, R. J. Smith, and S. R. Smith. 2011. "Obesity and type 2 diabetes: what can be unified and what needs to be individualized?" J Clin Endocrinol Metab 96 (6):1654-63. doi: 10.1210/jc.2011-0585.

Eizirik, D. L., and A. Op de Beeck. 2018. "Coxsackievirus and Type 1 Diabetes Mellitus: The Wolf's Footprints." Trends Endocrinol Metab 29 (3):137-139. doi: 10.1016/j.tem.2017.12.002.

El Ouaamari, A., E. Dirice, N. Gedeon, J. Hu, J. Y. Zhou, J. Shirakawa, L. Hou, J. Goodman, C. Karampelias, G. Qiang, J. Boucher, R. Martinez, M. A. Gritsenko, D. F. De Jesus, S. Kahraman, S. Bhatt, R. D. Smith, H. D. Beer, P. Jungtrakoon, Y. Gong, A. B. Goldfine, C. W. Liew, A. Doria, O. Andersson, W. J. Qian, E. Remold- O'Donnell, and R. N. Kulkarni. 2016. "SerpinB1 Promotes Pancreatic beta Cell Proliferation." Cell Metab 23 (1):194-205. doi: 10.1016/j.cmet.2015.12.001.

Encode Project Consortium. 2012. "An integrated encyclopedia of DNA elements in the human genome." Nature 489 (7414):57-74. doi: 10.1038/nature11247.

Ernst, J., P. Kheradpour, T. S. Mikkelsen, N. Shoresh, L. D. Ward, C. B. Epstein, X. Zhang, L. Wang, R. Issner, M. Coyne, M. Ku, T. Durham, M. Kellis, and B. E. Bernstein. 2011. "Mapping and analysis of chromatin state dynamics in nine human cell types." Nature 473 (7345):43-9. doi: 10.1038/nature09906.

Esvelt, K. M., P. Mali, J. L. Braff, M. Moosburner, S. J. Yaung, and G. M. Church. 2013. "Orthogonal Cas9 proteins for RNA-guided gene regulation and editing." Nat Methods 10 (11):1116-21. doi: 10.1038/nmeth.2681.

Fajans, S. S., and G. I. Bell. 2011. "MODY: history, genetics, pathophysiology, and clinical decision making." Diabetes Care 34 (8):1878-84. doi: 10.2337/dc11-0035.

199

Fajans, S. S., G. I. Bell, V. P. Paz, J. E. Below, N. J. Cox, C. Martin, I. H. Thomas, and M. Chen. 2010. "Obesity and hyperinsulinemia in a family with pancreatic agenesis and MODY caused by the IPF1 mutation Pro63fsX60." Transl Res 156 (1):7-14. doi: 10.1016/j.trsl.2010.03.003.

Fatrai, S., L. Elghazi, N. Balcazar, C. Cras-Meneur, I. Krits, H. Kiyokawa, and E. Bernal- Mizrachi. 2006. "Akt induces beta-cell proliferation by regulating cyclin D1, cyclin D2, and p21 levels and cyclin-dependent kinase-4 activity." Diabetes 55 (2):318-25.

Forbes, J. M., and M. E. Cooper. 2013. "Mechanisms of diabetic complications." Physiol Rev 93 (1):137-88. doi: 10.1152/physrev.00045.2011.

Foretz, M., B. Guigas, L. Bertrand, M. Pollak, and B. Viollet. 2014. "Metformin: from mechanisms of action to therapies." Cell Metab 20 (6):953-66. doi: 10.1016/j.cmet.2014.09.018.

Fueger, P. T., J. C. Schisler, D. Lu, D. A. Babu, R. G. Mirmira, C. B. Newgard, and H. E. Hohmeier. 2008. "Trefoil factor 3 stimulates human and rodent pancreatic islet beta-cell replication with retention of function." Mol Endocrinol 22 (5):1251-9. doi: 10.1210/me.2007-0500.

Fujimoto, K., E. L. Ford, H. Tran, B. M. Wice, S. D. Crosby, G. W. Dorn, 2nd, and K. S. Polonsky. 2010. "Loss of Nix in Pdx1-deficient mice prevents apoptotic and necrotic beta cell death and diabetes." J Clin Invest 120 (11):4031-9. doi: 10.1172/JCI44011.

Fujita, T., and H. Fujii. 2013. "Efficient isolation of specific genomic regions and identification of associated proteins by engineered DNA-binding molecule- mediated chromatin immunoprecipitation (enChIP) using CRISPR." Biochem Biophys Res Commun 439 (1):132-6. doi: 10.1016/j.bbrc.2013.08.013.

Fujitani, Y., S. Fujitani, D. F. Boyer, M. Gannon, Y. Kawaguchi, M. Ray, M. Shiota, R. W. Stein, M. A. Magnuson, and C. V. Wright. 2006. "Targeted deletion of a cis- regulatory region reveals differential gene dosage requirements for Pdx1 in

200

foregut organ differentiation and pancreas formation." Genes Dev 20 (2):253-66. doi: 10.1101/gad.1360106.

Gaggar, A., D. M. Shayakhmetov, and A. Lieber. 2003. "CD46 is a cellular receptor for group B adenoviruses." Nat Med 9 (11):1408-12. doi: 10.1038/nm952.

Gannon, M., E. T. Ables, L. Crawford, D. Lowe, M. F. Offield, M. A. Magnuson, and C. V. Wright. 2008. "pdx-1 function is specifically required in embryonic beta cells to generate appropriate numbers of endocrine cell types and maintain glucose homeostasis." Dev Biol 314 (2):406-17. doi: 10.1016/j.ydbio.2007.10.038.

Gannon, M., L. W. Gamer, and C. V. Wright. 2001. "Regulatory regions driving developmental and tissue-specific expression of the essential pancreatic gene pdx1." Dev Biol 238 (1):185-201. doi: 10.1006/dbio.2001.0359.

Gao, T., B. McKenna, C. Li, M. Reichert, J. Nguyen, T. Singh, C. Yang, A. Pannikar, N. Doliba, T. Zhang, D. A. Stoffers, H. Edlund, F. Matschinsky, R. Stein, and B. Z. Stanger. 2014. "Pdx1 maintains beta cell identity and function by repressing an alpha cell program." Cell Metab 19 (2):259-71. doi: 10.1016/j.cmet.2013.12.002.

Gao, Y., X. Xiong, S. Wong, E. J. Charles, W. A. Lim, and L. S. Qi. 2016. "Complex transcriptional modulation with orthogonal and inducible dCas9 regulators." Nat Methods 13 (12):1043-1049. doi: 10.1038/nmeth.4042.

Gasa, R., C. Clark, R. Yang, A. A. DePaoli-Roach, and C. B. Newgard. 2002. "Reversal of diet-induced glucose intolerance by hepatic expression of a variant glycogen- targeting subunit of protein phosphatase-1." J Biol Chem 277 (2):1524-30. doi: 10.1074/jbc.M107744200.

Georgia, S., and A. Bhushan. 2004. "Beta cell replication is the primary mechanism for maintaining postnatal beta cell mass." J Clin Invest 114 (7):963-8. doi: 10.1172/JCI22098.

201

Gerrish, K., M. A. Cissell, and R. Stein. 2001. "The role of hepatic nuclear factor 1 alpha and PDX-1 in transcriptional regulation of the pdx-1 gene." J Biol Chem 276 (51):47775-84. doi: 10.1074/jbc.M109244200.

Gerrish, K., M. Gannon, D. Shih, E. Henderson, M. Stoffel, C. V. Wright, and R. Stein. 2000. "Pancreatic beta cell-specific transcription of the pdx-1 gene. The role of conserved upstream control regions and their hepatic nuclear factor 3beta sites." J Biol Chem 275 (5):3485-92.

Gerrish, K., J. C. Van Velkinburgh, and R. Stein. 2004. "Conserved transcriptional regulatory domains of the pdx-1 gene." Mol Endocrinol 18 (3):533-48. doi: 10.1210/me.2003-0371.

Ghosh-Choudhury, G., Y. Haj-Ahmad, and F. L. Graham. 1987. "Protein IX, a minor component of the human adenovirus capsid, is essential for the packaging of full length genomes." EMBO J 6 (6):1733-9.

Gibson, D. G., L. Young, R. Y. Chuang, J. C. Venter, C. A. Hutchison, 3rd, and H. O. Smith. 2009. "Enzymatic assembly of DNA molecules up to several hundred kilobases." Nat Methods 6 (5):343-5. doi: 10.1038/nmeth.1318.

Gilbert, L. A., M. A. Horlbeck, B. Adamson, J. E. Villalta, Y. Chen, E. H. Whitehead, C. Guimaraes, B. Panning, H. L. Ploegh, M. C. Bassik, L. S. Qi, M. Kampmann, and J. S. Weissman. 2014. "Genome-Scale CRISPR-Mediated Control of Gene Repression and Activation." Cell 159 (3):647-61. doi: 10.1016/j.cell.2014.09.029.

Gilbert, L. A., M. H. Larson, L. Morsut, Z. Liu, G. A. Brar, S. E. Torres, N. Stern-Ginossar, O. Brandman, E. H. Whitehead, J. A. Doudna, W. A. Lim, J. S. Weissman, and L. S. Qi. 2013. "CRISPR-mediated modular RNA-guided regulation of transcription in eukaryotes." Cell 154 (2):442-51. doi: 10.1016/j.cell.2013.06.044.

Glick, E., D. Leshkowitz, and M. D. Walker. 2000. "Transcription factor BETA2 acts cooperatively with E2A and PDX1 to activate the insulin gene promoter." J Biol Chem 275 (3):2199-204.

202

Goldsmith, P. C., J. C. Rose, A. Arimura, and W. F. Ganong. 1975. "Ultrastructural localization of somatostatin in pancreatic islets of the rat." Endocrinology 97 (4):1061-4. doi: 10.1210/endo-97-4-1061.

Graham, F. L., J. Smiley, W. C. Russell, and R. Nairn. 1977. "Characteristics of a human cell line transformed by DNA from human adenovirus type 5." J Gen Virol 36 (1):59-74. doi: 10.1099/0022-1317-36-1-59.

Gu, B., T. Swigut, A. Spencley, M. R. Bauer, M. Chung, T. Meyer, and J. Wysocka. 2018. "Transcription-coupled changes in nuclear mobility of mammalian cis-regulatory elements." Science 359 (6379):1050-1055. doi: 10.1126/science.aao3136.

Guo, S., C. Dai, M. Guo, B. Taylor, J. S. Harmon, M. Sander, R. P. Robertson, A. C. Powers, and R. Stein. 2013. "Inactivation of specific beta cell transcription factors in type 2 diabetes." J Clin Invest 123 (8):3305-16. doi: 10.1172/JCI65390.

Guz, Y., M. R. Montminy, R. Stein, J. Leonard, L. W. Gamer, C. V. Wright, and G. Teitelman. 1995. "Expression of murine STF-1, a putative insulin gene transcription factor, in beta cells of pancreas, duodenal epithelium and pancreatic exocrine and endocrine progenitors during ontogeny." Development 121 (1):11-8.

Hani, E. H., D. A. Stoffers, J. C. Chevre, E. Durand, V. Stanojevic, C. Dina, J. F. Habener, and P. Froguel. 1999. "Defective mutations in the insulin promoter factor-1 (IPF- 1) gene in late-onset type 2 diabetes mellitus." J Clin Invest 104 (9):R41-8. doi: 10.1172/JCI7469.

Hartley, J. L., G. F. Temple, and M. A. Brasch. 2000. "DNA cloning using in vitro site- specific recombination." Genome Res 10 (11):1788-95.

Hattersley, A. T., and K. A. Patel. 2017. "Precision diabetes: learning from monogenic diabetes." Diabetologia 60 (5):769-777. doi: 10.1007/s00125-017-4226-2.

Hayes, H. L., L. G. Moss, J. C. Schisler, J. M. Haldeman, Z. Zhang, P. B. Rosenberg, C. B. Newgard, and H. E. Hohmeier. 2013. "Pdx-1 activates islet alpha- and beta-cell

203

proliferation via a mechanism regulated by transient receptor potential cation channels 3 and 6 and extracellular signal-regulated kinases 1 and 2." Mol Cell Biol 33 (20):4017-29. doi: 10.1128/MCB.00469-13.

Hayes, H. L., B. S. Peterson, J. M. Haldeman, C. B. Newgard, H. E. Hohmeier, and S. B. Stephens. 2017. "Delayed apoptosis allows islet beta-cells to implement an autophagic mechanism to promote cell survival." PLoS One 12 (2):e0172567. doi: 10.1371/journal.pone.0172567.

Hayes, H. L., L. Zhang, T. C. Becker, J. M. Haldeman, S. B. Stephens, M. Arlotto, L. G. Moss, C. B. Newgard, and H. E. Hohmeier. 2016. "A Pdx-1-Regulated Soluble Factor Activates Rat and Human Islet Cell Proliferation." Mol Cell Biol 36 (23):2918-2930. doi: 10.1128/MCB.00103-16.

He, L., J. M. Thomson, M. T. Hemann, E. Hernando-Monge, D. Mu, S. Goodson, S. Powers, C. Cordon-Cardo, S. W. Lowe, G. J. Hannon, and S. M. Hammond. 2005. "A microRNA polycistron as a potential human oncogene." Nature 435 (7043):828-33. doi: 10.1038/nature03552.

He, T. C., S. Zhou, L. T. da Costa, J. Yu, K. W. Kinzler, and B. Vogelstein. 1998. "A simplified system for generating recombinant adenoviruses." Proc Natl Acad Sci U S A 95 (5):2509-14.

Heo, J., V. M. Factor, T. Uren, Y. Takahama, J. S. Lee, M. Major, S. M. Feinstone, and S. S. Thorgeirsson. 2006. "Hepatic precursors derived from murine embryonic stem cells contribute to regeneration of injured liver." Hepatology 44 (6):1478-86. doi: 10.1002/hep.21441.

Hilton, I. B., A. M. D'Ippolito, C. M. Vockley, P. I. Thakore, G. E. Crawford, T. E. Reddy, and C. A. Gersbach. 2015. "Epigenome editing by a CRISPR-Cas9-based acetyltransferase activates genes from promoters and enhancers." Nat Biotechnol 33 (5):510-7. doi: 10.1038/nbt.3199.

Hnisz, D., B. J. Abraham, T. I. Lee, A. Lau, V. Saint-Andre, A. A. Sigova, H. A. Hoke, and R. A. Young. 2013. "Super-enhancers in the control of cell identity and disease." Cell 155 (4):934-47. doi: 10.1016/j.cell.2013.09.053. 204

Hobson, A., C. Draney, A. Stratford, T. C. Becker, D. Lu, M. Arlotto, and J. S. Tessem. 2015. "Aurora Kinase A is critical for the Nkx6.1 mediated beta-cell proliferation pathway." Islets 7 (1):e1027854. doi: 10.1080/19382014.2015.1027854.

Hohmeier, H. E., H. Mulder, G. Chen, R. Henkel-Rieger, M. Prentki, and C. B. Newgard. 2000. "Isolation of INS-1-derived cell lines with robust ATP-sensitive K+ channel- dependent and -independent glucose-stimulated insulin secretion." Diabetes 49 (3):424-30.

Holden, N. S., and C. E. Tacon. 2011. "Principles and problems of the electrophoretic mobility shift assay." J Pharmacol Toxicol Methods 63 (1):7-14. doi: 10.1016/j.vascn.2010.03.002.

Holland, A. J., D. Fachinetti, J. S. Han, and D. W. Cleveland. 2012. "Inducible, reversible system for the rapid and complete degradation of proteins in mammalian cells." Proc Natl Acad Sci U S A 109 (49):E3350-7. doi: 10.1073/pnas.1216880109.

Hough, S. H., K. Kancleris, L. Brody, N. Humphryes-Kirilov, J. Wolanski, K. Dunaway, A. Ajetunmobi, and V. Dillard. 2017. "Guide Picker is a comprehensive design tool for visualizing and selecting guides for CRISPR experiments." BMC Bioinformatics 18 (1):167. doi: 10.1186/s12859-017-1581-4.

Huang, X., H. Guo, S. Tammana, Y. C. Jung, E. Mellgren, P. Bassi, Q. Cao, Z. J. Tu, Y. C. Kim, S. C. Ekker, X. Wu, S. M. Wang, and X. Zhou. 2010. "Gene transfer efficiency and genome-wide integration profiling of Sleeping Beauty, Tol2, and piggyBac transposons in human primary T cells." Mol Ther 18 (10):1803-13. doi: 10.1038/mt.2010.141.

Iwamoto, M., T. Bjorklund, C. Lundberg, D. Kirik, and T. J. Wandless. 2010. "A general chemical method to regulate protein stability in the mammalian central nervous system." Chem Biol 17 (9):981-8. doi: 10.1016/j.chembiol.2010.07.009.

Jacovetti, C., A. Abderrahmani, G. Parnaud, J. C. Jonas, M. L. Peyot, M. Cornu, R. Laybutt, E. Meugnier, S. Rome, B. Thorens, M. Prentki, D. Bosco, and R. Regazzi. 2012. "MicroRNAs contribute to compensatory beta cell expansion during pregnancy and obesity." J Clin Invest 122 (10):3541-51. doi: 10.1172/JCI64151. 205

Jacovetti, C., S. J. Matkovich, A. Rodriguez-Trejo, C. Guay, and R. Regazzi. 2015. "Postnatal beta-cell maturation is associated with islet-specific microRNA changes induced by nutrient shifts at weaning." Nat Commun 6:8084. doi: 10.1038/ncomms9084.

Jain, R., and E. Lammert. 2009. "Cell-cell interactions in the endocrine pancreas." Diabetes Obes Metab 11 Suppl 4:159-67. doi: 10.1111/j.1463-1326.2009.01102.x.

Jensen, M. V., J. M. Haldeman, H. Zhang, D. Lu, M. O. Huising, W. W. Vale, H. E. Hohmeier, P. Rosenberg, and C. B. Newgard. 2013. "Control of voltage-gated potassium channel Kv2.2 expression by pyruvate-isocitrate cycling regulates glucose-stimulated insulin secretion." J Biol Chem 288 (32):23128-40. doi: 10.1074/jbc.M113.491654.

Jensen, M. V., J. W. Joseph, O. Ilkayeva, S. Burgess, D. Lu, S. M. Ronnebaum, M. Odegaard, T. C. Becker, A. D. Sherry, and C. B. Newgard. 2006. "Compensatory responses to pyruvate carboxylase suppression in islet beta-cells. Preservation of glucose-stimulated insulin secretion." J Biol Chem 281 (31):22342-51. doi: 10.1074/jbc.M604350200.

Jensen, M. V., J. W. Joseph, S. M. Ronnebaum, S. C. Burgess, A. D. Sherry, and C. B. Newgard. 2008. "Metabolic cycling in control of glucose-stimulated insulin secretion." Am J Physiol Endocrinol Metab 295 (6):E1287-97. doi: 10.1152/ajpendo.90604.2008.

Ji, M., E. Rao, H. Ramachandrareddy, Y. Shen, C. Jiang, J. Chen, Y. Hu, A. Rizzino, W. C. Chan, K. Fu, and T. W. McKeithan. 2011. "The miR-17-92 microRNA cluster is regulated by multiple mechanisms in B-cell malignancies." Am J Pathol 179 (4):1645-56. doi: 10.1016/j.ajpath.2011.06.008.

Jinek, M., K. Chylinski, I. Fonfara, M. Hauer, J. A. Doudna, and E. Charpentier. 2012. "A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity." Science 337 (6096):816-21. doi: 10.1126/science.1225829.

Jinek, M., A. East, A. Cheng, S. Lin, E. Ma, and J. Doudna. 2013. "RNA-programmed genome editing in human cells." Elife 2:e00471. doi: 10.7554/eLife.00471. 206

Johnson, J. D., N. T. Ahmed, D. S. Luciani, Z. Han, H. Tran, J. Fujita, S. Misler, H. Edlund, and K. S. Polonsky. 2003. "Increased islet apoptosis in Pdx1+/- mice." J Clin Invest 111 (8):1147-60. doi: 10.1172/JCI16537.

Joseph, J. W., M. V. Jensen, O. Ilkayeva, F. Palmieri, C. Alarcon, C. J. Rhodes, and C. B. Newgard. 2006. "The mitochondrial citrate/isocitrate carrier plays a regulatory role in glucose-stimulated insulin secretion." J Biol Chem 281 (47):35624-32. doi: 10.1074/jbc.M602606200.

Kahn, S. E., M. E. Cooper, and S. Del Prato. 2014. "Pathophysiology and treatment of type 2 diabetes: perspectives on the past, present, and future." Lancet 383 (9922):1068-83. doi: 10.1016/S0140-6736(13)62154-6.

Kameswaran, V., N. C. Bramswig, L. B. McKenna, M. Penn, J. Schug, N. J. Hand, Y. Chen, I. Choi, A. Vourekas, K. J. Won, C. Liu, K. Vivek, A. Naji, J. R. Friedman, and K. H. Kaestner. 2014. "Epigenetic regulation of the DLK1-MEG3 microRNA cluster in human type 2 diabetic islets." Cell Metab 19 (1):135-45. doi: 10.1016/j.cmet.2013.11.016.

Katzen, F. 2007. "Gateway((R)) recombinational cloning: a biological operating system." Expert Opin Drug Discov 2 (4):571-89. doi: 10.1517/17460441.2.4.571.

Kearns, N. A., H. Pham, B. Tabak, R. M. Genga, N. J. Silverstein, M. Garber, and R. Maehr. 2015. "Functional annotation of native enhancers with a Cas9-histone demethylase fusion." Nat Methods 12 (5):401-403. doi: 10.1038/nmeth.3325.

Khoo, C., J. Yang, S. A. Weinrott, K. H. Kaestner, A. Naji, J. Schug, and D. A. Stoffers. 2012. "Research resource: the pdx1 cistrome of pancreatic islets." Mol Endocrinol 26 (3):521-33. doi: 10.1210/me.2011-1231.

Kiani, S., A. Chavez, M. Tuttle, R. N. Hall, R. Chari, D. Ter-Ovanesyan, J. Qian, B. W. Pruitt, J. Beal, S. Vora, J. Buchthal, E. J. Kowal, M. R. Ebrahimkhani, J. J. Collins, R. Weiss, and G. Church. 2015. "Cas9 gRNA engineering for genome editing, activation and repression." Nat Methods 12 (11):1051-4. doi: 10.1038/nmeth.3580.

207

Kim, D. I., S. C. Jensen, K. A. Noble, B. Kc, K. H. Roux, K. Motamedchaboki, and K. J. Roux. 2016. "An improved smaller biotin ligase for BioID proximity labeling." Mol Biol Cell 27 (8):1188-96. doi: 10.1091/mbc.E15-12-0844.

Kimmel, J. R., L. J. Hayden, and H. G. Pollock. 1975. "Isolation and characterization of a new pancreatic polypeptide hormone." J Biol Chem 250 (24):9369-76.

Kleinstiver, B. P., M. S. Prew, S. Q. Tsai, N. T. Nguyen, V. V. Topkar, Z. Zheng, and J. K. Joung. 2015. "Broadening the targeting range of Staphylococcus aureus CRISPR- Cas9 by modifying PAM recognition." Nat Biotechnol 33 (12):1293-1298. doi: 10.1038/nbt.3404.

Klil-Drori, A. J., L. Azoulay, and M. N. Pollak. 2017. "Cancer, obesity, diabetes, and antidiabetic drugs: is the fog clearing?" Nat Rev Clin Oncol 14 (2):85-99. doi: 10.1038/nrclinonc.2016.120.

Kong, Y., R. B. Sharma, S. Ly, R. E. Stamateris, W. M. Jesdale, and L. C. Alonso. 2018. "CDKN2A/B T2D GWAS Risk-SNPs Impact Locus Gene Expression and Proliferation in Human Islets." Diabetes. doi: 10.2337/db17-1055.

Koyama, K., G. Chen, M. Y. Wang, Y. Lee, M. Shimabukuro, C. B. Newgard, and R. H. Unger. 1997. "beta-cell function in normal rats made chronically hyperleptinemic by adenovirus-leptin gene therapy." Diabetes 46 (8):1276-80.

Krishnamurthy, J., M. R. Ramsey, K. L. Ligon, C. Torrice, A. Koh, S. Bonner-Weir, and N. E. Sharpless. 2006. "p16INK4a induces an age-dependent decline in islet regenerative potential." Nature 443 (7110):453-7. doi: 10.1038/nature05092.

Kulkarni, R. N., E. B. Mizrachi, A. G. Ocana, and A. F. Stewart. 2012. "Human beta-cell proliferation and intracellular signaling: driving in the dark without a road map." Diabetes 61 (9):2205-13. doi: 10.2337/db12-0018.

Lacy, P. E., and J. Davies. 1957. "Preliminary studies on the demonstration of insulin in the islets by the fluorescent antibody technic." Diabetes 6 (4):354-7.

208

Lam, S. S., J. D. Martell, K. J. Kamer, T. J. Deerinck, M. H. Ellisman, V. K. Mootha, and A. Y. Ting. 2015. "Directed evolution of APEX2 for electron microscopy and proximity labeling." Nat Methods 12 (1):51-4. doi: 10.1038/nmeth.3179.

Lamesch, P., N. Li, S. Milstein, C. Fan, T. Hao, G. Szabo, Z. Hu, K. Venkatesan, G. Bethel, P. Martin, J. Rogers, S. Lawlor, S. McLaren, A. Dricot, H. Borick, M. E. Cusick, J. Vandenhaute, I. Dunham, D. E. Hill, and M. Vidal. 2007. "hORFeome v3.1: a resource of human open reading frames representing over 10,000 human genes." Genomics 89 (3):307-15. doi: 10.1016/j.ygeno.2006.11.012.

Latreille, M., J. Hausser, I. Stutzer, Q. Zhang, B. Hastoy, S. Gargani, J. Kerr-Conte, F. Pattou, M. Zavolan, J. L. Esguerra, L. Eliasson, T. Rulicke, P. Rorsman, and M. Stoffel. 2014. "MicroRNA-7a regulates pancreatic beta cell function." J Clin Invest 124 (6):2722-35. doi: 10.1172/JCI73066.

Lavine, J. A., P. W. Raess, D. B. Davis, M. E. Rabaglia, B. K. Presley, M. P. Keller, M. C. Beinfeld, A. S. Kopin, C. B. Newgard, and A. D. Attie. 2010. "Contamination with E1A-positive wild-type adenovirus accounts for species-specific stimulation of islet cell proliferation by CCK: a cautionary note." Mol Endocrinol 24 (2):464-7. doi: 10.1210/me.2009-0384.

Le Lay, J., T. A. Matsuoka, E. Henderson, and R. Stein. 2004. "Identification of a novel PDX-1 binding site in the human insulin gene enhancer." J Biol Chem 279 (21):22228-35. doi: 10.1074/jbc.M312673200.

Lebovitz, H. E. 2011. "Insulin: potential negative consequences of early routine use in patients with type 2 diabetes." Diabetes Care 34 Suppl 2:S225-30. doi: 10.2337/dc11-s225.

Lebovitz, H. E., and M. N. Feinglos. 1978. "Sulfonylurea drugs: mechanism of antidiabetic action and therapeutic usefulness." Diabetes Care 1 (3):189-98.

Lerner, A. G., J. P. Upton, P. V. Praveen, R. Ghosh, Y. Nakagawa, A. Igbaria, S. Shen, V. Nguyen, B. J. Backes, M. Heiman, N. Heintz, P. Greengard, S. Hui, Q. Tang, A. Trusina, S. A. Oakes, and F. R. Papa. 2012. "IRE1alpha induces thioredoxin- interacting protein to activate the NLRP3 inflammasome and promote 209

programmed cell death under irremediable ER stress." Cell Metab 16 (2):250-64. doi: 10.1016/j.cmet.2012.07.007.

Li, Y., P. S. Choi, S. C. Casey, D. L. Dill, and D. W. Felsher. 2014. "MYC through miR-17- 92 suppresses specific target genes to maintain survival, autonomous proliferation, and a neoplastic state." Cancer Cell 26 (2):262-72. doi: 10.1016/j.ccr.2014.06.014.

Li, Y., T. Hansotia, B. Yusta, F. Ris, P. A. Halban, and D. J. Drucker. 2003. "Glucagon-like peptide-1 receptor signaling modulates beta cell apoptosis." J Biol Chem 278 (1):471-8. doi: 10.1074/jbc.M209423200.

Liu, X., Y. Zhang, Y. Chen, M. Li, F. Zhou, K. Li, H. Cao, M. Ni, Y. Liu, Z. Gu, K. E. Dickerson, S. Xie, G. C. Hon, Z. Xuan, M. Q. Zhang, Z. Shao, and J. Xu. 2017. "In Situ Capture of Chromatin Interactions by Biotinylated dCas9." Cell 170 (5):1028- 1043 e19. doi: 10.1016/j.cell.2017.08.003.

Lynn, F. C., P. Skewes-Cox, Y. Kosaka, M. T. McManus, B. D. Harfe, and M. S. German. 2007. "MicroRNA expression is required for pancreatic islet cell genesis in the mouse." Diabetes 56 (12):2938-45. doi: 10.2337/db07-0175.

Ma, H., L. C. Tu, A. Naseri, M. Huisman, S. Zhang, D. Grunwald, and T. Pederson. 2016. "Multiplexed labeling of genomic loci with dCas9 and engineered sgRNAs using CRISPRainbow." Nat Biotechnol 34 (5):528-30. doi: 10.1038/nbt.3526.

Maeder, M. L., S. J. Linder, V. M. Cascio, Y. Fu, Q. H. Ho, and J. K. Joung. 2013. "CRISPR RNA-guided activation of endogenous human genes." Nat Methods 10 (10):977-9. doi: 10.1038/nmeth.2598.

Maedler, K., D. M. Schumann, F. Schulthess, J. Oberholzer, D. Bosco, T. Berney, and M. Y. Donath. 2006. "Aging correlates with decreased beta-cell proliferative capacity and enhanced sensitivity to apoptosis: a potential role for Fas and pancreatic duodenal homeobox-1." Diabetes 55 (9):2455-62. doi: 10.2337/db05-1586.

210

Maizel, J. V., Jr., D. O. White, and M. D. Scharff. 1968. "The polypeptides of adenovirus. I. Evidence for multiple protein components in the virion and a comparison of types 2, 7A, and 12." Virology 36 (1):115-25.

Mak, L. H., R. Vilar, and R. Woscholski. 2010. "Characterisation of the PTEN inhibitor VO-OHpic." J Chem Biol 3 (4):157-63. doi: 10.1007/s12154-010-0041-7.

Mali, P., L. Yang, K. M. Esvelt, J. Aach, M. Guell, J. E. DiCarlo, J. E. Norville, and G. M. Church. 2013. "RNA-guided human genome engineering via Cas9." Science 339 (6121):823-6. doi: 10.1126/science.1232033.

Mandelbaum, A. D., T. Melkman-Zehavi, R. Oren, S. Kredo-Russo, T. Nir, Y. Dor, and E. Hornstein. 2012. "Dysregulation of Dicer1 in beta cells impairs islet architecture and glucose metabolism." Exp Diabetes Res 2012:470302. doi: 10.1155/2012/470302.

Marroqui, L., M. Lopes, R. S. dos Santos, F. A. Grieco, M. Roivainen, S. J. Richardson, N. G. Morgan, A. Op de Beeck, and D. L. Eizirik. 2015. "Differential cell autonomous responses determine the outcome of coxsackievirus infections in murine pancreatic alpha and beta cells." Elife 4:e06990. doi: 10.7554/eLife.06990.

Martinez-Sanchez, A., M. S. Nguyen-Tu, and G. A. Rutter. 2015. "DICER Inactivation Identifies Pancreatic beta-Cell "Disallowed" Genes Targeted by MicroRNAs." Mol Endocrinol 29 (7):1067-79. doi: 10.1210/me.2015-1059.

Mates, L., M. K. Chuah, E. Belay, B. Jerchow, N. Manoj, A. Acosta-Sanchez, D. P. Grzela, A. Schmitt, K. Becker, J. Matrai, L. Ma, E. Samara-Kuko, C. Gysemans, D. Pryputniewicz, C. Miskey, B. Fletcher, T. VandenDriessche, Z. Ivics, and Z. Izsvak. 2009. "Molecular evolution of a novel hyperactive Sleeping Beauty transposase enables robust stable gene transfer in vertebrates." Nat Genet 41 (6):753-61. doi: 10.1038/ng.343.

Maurano, M. T., R. Humbert, E. Rynes, R. E. Thurman, E. Haugen, H. Wang, A. P. Reynolds, R. Sandstrom, H. Qu, J. Brody, A. Shafer, F. Neri, K. Lee, T. Kutyavin, S. Stehling-Sun, A. K. Johnson, T. K. Canfield, E. Giste, M. Diegel, D. Bates, R. S. Hansen, S. Neph, P. J. Sabo, S. Heimfeld, A. Raubitschek, S. Ziegler, C. Cotsapas, N. Sotoodehnia, I. Glass, S. R. Sunyaev, R. Kaul, and J. A. Stamatoyannopoulos. 211

2012. "Systematic localization of common disease-associated variation in regulatory DNA." Science 337 (6099):1190-5. doi: 10.1126/science.1222794.

McCarthy, M. I. 2017. "Painting a new picture of personalised medicine for diabetes." Diabetologia 60 (5):793-799. doi: 10.1007/s00125-017-4210-x.

McGrory, W. J., D. S. Bautista, and F. L. Graham. 1988. "A simple technique for the rescue of early region I mutations into infectious human adenovirus type 5." Virology 163 (2):614-7.

McKenzie, M. D., E. Jamieson, E. S. Jansen, C. L. Scott, D. C. Huang, P. Bouillet, J. Allison, T. W. Kay, A. Strasser, and H. E. Thomas. 2010. "Glucose induces pancreatic islet cell apoptosis that requires the BH3-only proteins Bim and Puma and multi-BH domain protein Bax." Diabetes 59 (3):644-52. doi: 10.2337/db09- 1151.

Mefferd, A. L., A. V. Kornepati, H. P. Bogerd, E. M. Kennedy, and B. R. Cullen. 2015. "Expression of CRISPR/Cas single guide RNAs using small tRNA promoters." RNA 21 (9):1683-9. doi: 10.1261/rna.051631.115.

Meier, J. J., A. E. Butler, Y. Saisho, T. Monchamp, R. Galasso, A. Bhushan, R. A. Rizza, and P. C. Butler. 2008. "Beta-cell replication is the primary mechanism subserving the postnatal expansion of beta-cell mass in humans." Diabetes 57 (6):1584-94. doi: 10.2337/db07-1369.

Meinken, J., G. Walker, C. R. Cooper, and X. J. Min. 2015. "MetazSecKB: the human and animal secretome and subcellular proteome knowledgebase." Database (Oxford) 2015. doi: 10.1093/database/bav077.

Melkman-Zehavi, T., R. Oren, S. Kredo-Russo, T. Shapira, A. D. Mandelbaum, N. Rivkin, T. Nir, K. A. Lennox, M. A. Behlke, Y. Dor, and E. Hornstein. 2011. "miRNAs control insulin content in pancreatic beta-cells via downregulation of transcriptional repressors." EMBO J 30 (5):835-45. doi: 10.1038/emboj.2010.361.

212

Mendenhall, E. M., K. E. Williamson, D. Reyon, J. Y. Zou, O. Ram, J. K. Joung, and B. E. Bernstein. 2013. "Locus-specific editing of histone modifications at endogenous enhancers." Nat Biotechnol 31 (12):1133-6. doi: 10.1038/nbt.2701.

Miller, J. C., S. Tan, G. Qiao, K. A. Barlow, J. Wang, D. F. Xia, X. Meng, D. E. Paschon, E. Leung, S. J. Hinkley, G. P. Dulay, K. L. Hua, I. Ankoudinova, G. J. Cost, F. D. Urnov, H. S. Zhang, M. C. Holmes, L. Zhang, P. D. Gregory, and E. J. Rebar. 2011. "A TALE nuclease architecture for efficient genome editing." Nat Biotechnol 29 (2):143-8. doi: 10.1038/nbt.1755.

Mogilyansky, E., and I. Rigoutsos. 2013. "The miR-17/92 cluster: a comprehensive update on its genomics, genetics, functions and increasingly important and numerous roles in health and disease." Cell Death Differ 20 (12):1603-14. doi: 10.1038/cdd.2013.125.

Moran, I., I. Akerman, M. van de Bunt, R. Xie, M. Benazra, T. Nammo, L. Arnes, N. Nakic, J. Garcia-Hurtado, S. Rodriguez-Segui, L. Pasquali, C. Sauty-Colace, A. Beucher, R. Scharfmann, J. van Arensbergen, P. R. Johnson, A. Berry, C. Lee, T. Harkins, V. Gmyr, F. Pattou, J. Kerr-Conte, L. Piemonti, T. Berney, N. Hanley, A. L. Gloyn, L. Sussel, L. Langman, K. L. Brayman, M. Sander, M. I. McCarthy, P. Ravassard, and J. Ferrer. 2012. "Human beta cell transcriptome analysis uncovers lncRNAs that are tissue-specific, dynamically regulated, and abnormally expressed in type 2 diabetes." Cell Metab 16 (4):435-48. doi: 10.1016/j.cmet.2012.08.010.

Morris, A. P., B. F. Voight, T. M. Teslovich, T. Ferreira, A. V. Segre, V. Steinthorsdottir, R. J. Strawbridge, H. Khan, H. Grallert, A. Mahajan, I. Prokopenko, H. M. Kang, C. Dina, T. Esko, R. M. Fraser, S. Kanoni, A. Kumar, V. Lagou, C. Langenberg, J. Luan, C. M. Lindgren, M. Muller-Nurasyid, S. Pechlivanis, N. W. Rayner, L. J. Scott, S. Wiltshire, L. Yengo, L. Kinnunen, E. J. Rossin, S. Raychaudhuri, A. D. Johnson, A. S. Dimas, R. J. Loos, S. Vedantam, H. Chen, J. C. Florez, C. Fox, C. T. Liu, D. Rybin, D. J. Couper, W. H. Kao, M. Li, M. C. Cornelis, P. Kraft, Q. Sun, R. M. van Dam, H. M. Stringham, P. S. Chines, K. Fischer, P. Fontanillas, O. L. Holmen, S. E. Hunt, A. U. Jackson, A. Kong, R. Lawrence, J. Meyer, J. R. Perry, C. G. Platou, S. Potter, E. Rehnberg, N. Robertson, S. Sivapalaratnam, A. Stancakova, K. Stirrups, G. Thorleifsson, E. Tikkanen, A. R. Wood, P. Almgren, M. Atalay, R. Benediktsson, L. L. Bonnycastle, N. Burtt, J. Carey, G. Charpentier, A. T. Crenshaw, A. S. Doney, M. Dorkhan, S. Edkins, V. Emilsson, E. Eury, T. 213

Forsen, K. Gertow, B. Gigante, G. B. Grant, C. J. Groves, C. Guiducci, C. Herder, A. B. Hreidarsson, J. Hui, A. James, A. Jonsson, W. Rathmann, N. Klopp, J. Kravic, K. Krjutskov, C. Langford, K. Leander, E. Lindholm, S. Lobbens, S. Mannisto, G. Mirza, T. W. Muhleisen, B. Musk, M. Parkin, L. Rallidis, J. Saramies, B. Sennblad, S. Shah, G. Sigurethsson, A. Silveira, G. Steinbach, B. Thorand, J. Trakalo, F. Veglia, R. Wennauer, W. Winckler, D. Zabaneh, H. Campbell, C. van Duijn, A. G. Uitterlinden, A. Hofman, E. Sijbrands, G. R. Abecasis, K. R. Owen, E. Zeggini, M. D. Trip, N. G. Forouhi, A. C. Syvanen, J. G. Eriksson, L. Peltonen, M. M. Nothen, B. Balkau, C. N. Palmer, V. Lyssenko, T. Tuomi, B. Isomaa, D. J. Hunter, L. Qi, Consortium Wellcome Trust Case Control, Glucose Meta-Analyses of, Investigators Insulin-related traits Consortium, ANthropometric Traits Consortium Genetic Investigation of, Consortium Asian Genetic Epidemiology Network-Type 2 Diabetes, Consortium South Asian Type 2 Diabetes, A. R. Shuldiner, M. Roden, I. Barroso, T. Wilsgaard, J. Beilby, K. Hovingh, J. F. Price, J. F. Wilson, R. Rauramaa, T. A. Lakka, L. Lind, G. Dedoussis, I. Njolstad, N. L. Pedersen, K. T. Khaw, N. J. Wareham, S. M. Keinanen-Kiukaanniemi, T. E. Saaristo, E. Korpi-Hyovalti, J. Saltevo, M. Laakso, J. Kuusisto, A. Metspalu, F. S. Collins, K. L. Mohlke, R. N. Bergman, J. Tuomilehto, B. O. Boehm, C. Gieger, K. Hveem, S. Cauchi, P. Froguel, D. Baldassarre, E. Tremoli, S. E. Humphries, D. Saleheen, J. Danesh, E. Ingelsson, S. Ripatti, V. Salomaa, R. Erbel, K. H. Jockel, S. Moebus, A. Peters, T. Illig, U. de Faire, A. Hamsten, A. D. Morris, P. J. Donnelly, T. M. Frayling, A. T. Hattersley, E. Boerwinkle, O. Melander, S. Kathiresan, P. M. Nilsson, P. Deloukas, U. Thorsteinsdottir, L. C. Groop, K. Stefansson, F. Hu, J. S. Pankow, J. Dupuis, J. B. Meigs, D. Altshuler, M. Boehnke, M. I. McCarthy, D. IAbetes Genetics Replication, and Consortium Meta-analysis. 2012. "Large-scale association analysis provides insights into the genetic architecture and pathophysiology of type 2 diabetes." Nat Genet 44 (9):981-90. doi: 10.1038/ng.2383.

Mu, P., Y. C. Han, D. Betel, E. Yao, M. Squatrito, P. Ogrodowski, E. de Stanchina, A. D'Andrea, C. Sander, and A. Ventura. 2009. "Genetic dissection of the miR-17~92 cluster of microRNAs in Myc-induced B-cell lymphomas." Genes Dev 23 (24):2806-11. doi: 10.1101/gad.1872909.

Natsume, T., T. Kiyomitsu, Y. Saga, and M. T. Kanemaki. 2016. "Rapid Protein Depletion in Human Cells by Auxin-Inducible Degron Tagging with Short Homology Donors." Cell Rep 15 (1):210-8. doi: 10.1016/j.celrep.2016.03.001.

214

Nicolino, M., K. C. Claiborn, V. Senee, A. Boland, D. A. Stoffers, and C. Julier. 2010. "A novel hypomorphic PDX1 mutation responsible for permanent neonatal diabetes with subclinical exocrine deficiency." Diabetes 59 (3):733-40. doi: 10.2337/db09- 1284.

Northcott, P. A., C. Lee, T. Zichner, A. M. Stutz, S. Erkek, D. Kawauchi, D. J. Shih, V. Hovestadt, M. Zapatka, D. Sturm, D. T. Jones, M. Kool, M. Remke, F. M. Cavalli, S. Zuyderduyn, G. D. Bader, S. VandenBerg, L. A. Esparza, M. Ryzhova, W. Wang, A. Wittmann, S. Stark, L. Sieber, H. Seker-Cin, L. Linke, F. Kratochwil, N. Jager, I. Buchhalter, C. D. Imbusch, G. Zipprich, B. Raeder, S. Schmidt, N. Diessl, S. Wolf, S. Wiemann, B. Brors, C. Lawerenz, J. Eils, H. J. Warnatz, T. Risch, M. L. Yaspo, U. D. Weber, C. C. Bartholomae, C. von Kalle, E. Turanyi, P. Hauser, E. Sanden, A. Darabi, P. Siesjo, J. Sterba, K. Zitterbart, D. Sumerauer, P. van Sluis, R. Versteeg, R. Volckmann, J. Koster, M. U. Schuhmann, M. Ebinger, H. L. Grimes, G. W. Robinson, A. Gajjar, M. Mynarek, K. von Hoff, S. Rutkowski, T. Pietsch, W. Scheurlen, J. Felsberg, G. Reifenberger, A. E. Kulozik, A. von Deimling, O. Witt, R. Eils, R. J. Gilbertson, A. Korshunov, M. D. Taylor, P. Lichter, J. O. Korbel, R. J. Wechsler-Reya, and S. M. Pfister. 2014. "Enhancer hijacking activates GFI1 family oncogenes in medulloblastoma." Nature 511 (7510):428-34. doi: 10.1038/nature13379.

Nyalwidhe, J. O., G. R. Gallagher, L. M. Glenn, M. A. Morris, P. Vangala, A. Jurczyk, R. Bortell, D. M. Harlan, J. P. Wang, and J. L. Nadler. 2017. "Coxsackievirus-Induced Proteomic Alterations in Primary Human Islets Provide Insights for the Etiology of Diabetes." J Endocr Soc 1 (10):1272-1286. doi: 10.1210/js.2017-00278.

O'Donnell, K. A., E. A. Wentzel, K. I. Zeller, C. V. Dang, and J. T. Mendell. 2005. "c-Myc- regulated microRNAs modulate E2F1 expression." Nature 435 (7043):839-43. doi: 10.1038/nature03677.

Odegaard, M. L., J. W. Joseph, M. V. Jensen, D. Lu, O. Ilkayeva, S. M. Ronnebaum, T. C. Becker, and C. B. Newgard. 2010. "The mitochondrial 2-oxoglutarate carrier is part of a metabolic pathway that mediates glucose- and glutamine-stimulated insulin secretion." J Biol Chem 285 (22):16530-7. doi: 10.1074/jbc.M109.092593.

215

Offield, M. F., T. L. Jetton, P. A. Labosky, M. Ray, R. W. Stein, M. A. Magnuson, B. L. Hogan, and C. V. Wright. 1996. "PDX-1 is required for pancreatic outgrowth and differentiation of the rostral duodenum." Development 122 (3):983-95.

Ogawa, A., J. H. Johnson, M. Ohneda, C. T. McAllister, L. Inman, T. Alam, and R. H. Unger. 1992. "Roles of insulin resistance and beta-cell dysfunction in dexamethasone-induced diabetes." J Clin Invest 90 (2):497-504. doi: 10.1172/JCI115886.

Oh, S. Y., Y. Ju, S. Kim, and H. Park. 2010. "PNA-based antisense oligonucleotides for micrornas inhibition in the absence of a transfection reagent." Oligonucleotides 20 (5):225-30. doi: 10.1089/oli.2010.0238.

Olive, V., M. J. Bennett, J. C. Walker, C. Ma, I. Jiang, C. Cordon-Cardo, Q. J. Li, S. W. Lowe, G. J. Hannon, and L. He. 2009. "miR-19 is a key oncogenic component of mir-17-92." Genes Dev 23 (24):2839-49. doi: 10.1101/gad.1861409.

Osterwalder, M., I. Barozzi, V. Tissieres, Y. Fukuda-Yuzawa, B. J. Mannion, S. Y. Afzal, E. A. Lee, Y. Zhu, I. Plajzer-Frick, C. S. Pickle, M. Kato, T. H. Garvin, Q. T. Pham, A. N. Harrington, J. A. Akiyama, V. Afzal, J. Lopez-Rios, D. E. Dickel, A. Visel, and L. A. Pennacchio. 2018. "Enhancer redundancy provides phenotypic robustness in mammalian development." Nature 554 (7691):239-243. doi: 10.1038/nature25461.

Parker, S. C., M. L. Stitzel, D. L. Taylor, J. M. Orozco, M. R. Erdos, J. A. Akiyama, K. L. van Bueren, P. S. Chines, N. Narisu, Nisc Comparative Sequencing Program, B. L. Black, A. Visel, L. A. Pennacchio, F. S. Collins, Authors National Institutes of Health Intramural Sequencing Center Comparative Sequencing Program, and Nisc Comparative Sequencing Program Authors. 2013. "Chromatin stretch enhancer states drive cell-specific gene regulation and harbor human disease risk variants." Proc Natl Acad Sci U S A 110 (44):17921-6. doi: 10.1073/pnas.1317023110.

Pasquali, L., K. J. Gaulton, S. A. Rodriguez-Segui, L. Mularoni, I. Miguel-Escalada, I. Akerman, J. J. Tena, I. Moran, C. Gomez-Marin, M. van de Bunt, J. Ponsa-Cobas, N. Castro, T. Nammo, I. Cebola, J. Garcia-Hurtado, M. A. Maestro, F. Pattou, L.

216

Piemonti, T. Berney, A. L. Gloyn, P. Ravassard, J. L. G. Skarmeta, F. Muller, M. I. McCarthy, and J. Ferrer. 2014. "Pancreatic islet enhancer clusters enriched in type 2 diabetes risk-associated variants." Nat Genet 46 (2):136-143. doi: 10.1038/ng.2870.

Perl, S., J. A. Kushner, B. A. Buchholz, A. K. Meeker, G. M. Stein, M. Hsieh, M. Kirby, S. Pechhold, E. H. Liu, D. M. Harlan, and J. F. Tisdale. 2010. "Significant human beta-cell turnover is limited to the first three decades of life as determined by in vivo thymidine analog incorporation and radiocarbon dating." J Clin Endocrinol Metab 95 (10):E234-9. doi: 10.1210/jc.2010-0932.

Pinhas-Hamiel, O., and P. Zeitler. 2007. "Acute and chronic complications of type 2 diabetes mellitus in children and adolescents." Lancet 369 (9575):1823-31. doi: 10.1016/S0140-6736(07)60821-6.

Porat, S., N. Weinberg-Corem, S. Tornovsky-Babaey, R. Schyr-Ben-Haroush, A. Hija, M. Stolovich-Rain, D. Dadon, Z. Granot, V. Ben-Hur, P. White, C. A. Girard, R. Karni, K. H. Kaestner, F. M. Ashcroft, M. A. Magnuson, A. Saada, J. Grimsby, B. Glaser, and Y. Dor. 2011. "Control of pancreatic beta cell regeneration by glucose metabolism." Cell Metab 13 (4):440-449. doi: 10.1016/j.cmet.2011.02.012.

Poy, M. N., L. Eliasson, J. Krutzfeldt, S. Kuwajima, X. Ma, P. E. Macdonald, S. Pfeffer, T. Tuschl, N. Rajewsky, P. Rorsman, and M. Stoffel. 2004. "A pancreatic islet- specific microRNA regulates insulin secretion." Nature 432 (7014):226-30. doi: 10.1038/nature03076.

Poy, M. N., J. Hausser, M. Trajkovski, M. Braun, S. Collins, P. Rorsman, M. Zavolan, and M. Stoffel. 2009. "miR-375 maintains normal pancreatic alpha- and beta-cell mass." Proc Natl Acad Sci U S A 106 (14):5813-8. doi: 10.1073/pnas.0810550106.

Prado, C. L., A. E. Pugh-Bernard, L. Elghazi, B. Sosa-Pineda, and L. Sussel. 2004. "Ghrelin cells replace insulin-producing beta cells in two mouse models of pancreas development." Proc Natl Acad Sci U S A 101 (9):2924-9. doi: 10.1073/pnas.0308604100.

217

Prasad, K. M., Y. Xu, Z. Yang, S. T. Acton, and B. A. French. 2011. "Robust cardiomyocyte-specific gene expression following systemic injection of AAV: in vivo gene delivery follows a Poisson distribution." Gene Ther 18 (1):43-52. doi: 10.1038/gt.2010.105.

Prasad, R. B., and L. Groop. 2015. "Genetics of type 2 diabetes-pitfalls and possibilities." Genes (Basel) 6 (1):87-123. doi: 10.3390/genes6010087.

Pullen, T. J., G. da Silva Xavier, G. Kelsey, and G. A. Rutter. 2011. "miR-29a and miR-29b contribute to pancreatic beta-cell-specific silencing of monocarboxylate transporter 1 (Mct1)." Mol Cell Biol 31 (15):3182-94. doi: 10.1128/MCB.01433-10.

Qi, L. S., M. H. Larson, L. A. Gilbert, J. A. Doudna, J. S. Weissman, A. P. Arkin, and W. A. Lim. 2013. "Repurposing CRISPR as an RNA-guided platform for sequence- specific control of gene expression." Cell 152 (5):1173-83. doi: 10.1016/j.cell.2013.02.022.

Ran, F. A., L. Cong, W. X. Yan, D. A. Scott, J. S. Gootenberg, A. J. Kriz, B. Zetsche, O. Shalem, X. Wu, K. S. Makarova, E. V. Koonin, P. A. Sharp, and F. Zhang. 2015. "In vivo genome editing using Staphylococcus aureus Cas9." Nature 520 (7546):186-91. doi: 10.1038/nature14299.

Rane, S. G., P. Dubus, R. V. Mettus, E. J. Galbreath, G. Boden, E. P. Reddy, and M. Barbacid. 1999. "Loss of Cdk4 expression causes insulin-deficient diabetes and Cdk4 activation results in beta-islet cell hyperplasia." Nat Genet 22 (1):44-52. doi: 10.1038/8751.

Rao, S. S., M. H. Huntley, N. C. Durand, E. K. Stamenova, I. D. Bochkov, J. T. Robinson, A. L. Sanborn, I. Machol, A. D. Omer, E. S. Lander, and E. L. Aiden. 2014. "A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping." Cell 159 (7):1665-80. doi: 10.1016/j.cell.2014.11.021.

Ravassard, P., Y. Hazhouz, S. Pechberty, E. Bricout-Neveu, M. Armanet, P. Czernichow, and R. Scharfmann. 2011. "A genetically engineered human pancreatic beta cell line exhibiting glucose-inducible insulin secretion." J Clin Invest 121 (9):3589-97. doi: 10.1172/JCI58447. 218

Ray, J. D., K. B. Kener, B. F. Bitner, B. J. Wright, M. S. Ballard, E. J. Barrett, J. T. Hill, L. G. Moss, and J. S. Tessem. 2016. "Nkx6.1-mediated insulin secretion and beta-cell proliferation is dependent on upregulation of c-Fos." FEBS Lett 590 (12):1791-803. doi: 10.1002/1873-3468.12208.

Reynolds, P., I. Dmitriev, and D. Curiel. 1999. "Insertion of an RGD motif into the HI loop of adenovirus fiber protein alters the distribution of transgene expression of the systemically administered vector." Gene Ther 6 (7):1336-9. doi: 10.1038/sj.gt.3300941.

Rieck, S., and K. H. Kaestner. 2010. "Expansion of beta-cell mass in response to pregnancy." Trends Endocrinol Metab 21 (3):151-8. doi: 10.1016/j.tem.2009.11.001.

Ronnebaum, S. M., O. Ilkayeva, S. C. Burgess, J. W. Joseph, D. Lu, R. D. Stevens, T. C. Becker, A. D. Sherry, C. B. Newgard, and M. V. Jensen. 2006. "A pyruvate cycling pathway involving cytosolic NADP-dependent isocitrate dehydrogenase regulates glucose-stimulated insulin secretion." J Biol Chem 281 (41):30593-602. doi: 10.1074/jbc.M511908200.

Ronnebaum, S. M., M. V. Jensen, H. E. Hohmeier, S. C. Burgess, Y. P. Zhou, S. Qian, D. MacNeil, A. Howard, N. Thornberry, O. Ilkayeva, D. Lu, A. D. Sherry, and C. B. Newgard. 2008a. "Silencing of cytosolic or mitochondrial isoforms of malic enzyme has no effect on glucose-stimulated insulin secretion from rodent islets." J Biol Chem 283 (43):28909-17. doi: 10.1074/jbc.M804665200.

Ronnebaum, S. M., J. W. Joseph, O. Ilkayeva, S. C. Burgess, D. Lu, T. C. Becker, A. D. Sherry, and C. B. Newgard. 2008b. "Chronic suppression of acetyl-CoA carboxylase 1 in beta-cells impairs insulin secretion via inhibition of glucose rather than lipid metabolism." J Biol Chem 283 (21):14248-56. doi: 10.1074/jbc.M800119200.

Rui, J., S. Deng, A. Arazi, A. L. Perdigoto, Z. Liu, and K. C. Herold. 2017. "beta Cells that Resist Immunological Attack Develop during Progression of Autoimmune Diabetes in NOD Mice." Cell Metab 25 (3):727-738. doi: 10.1016/j.cmet.2017.01.005.

219

Sachdeva, M. M., K. C. Claiborn, C. Khoo, J. Yang, D. N. Groff, R. G. Mirmira, and D. A. Stoffers. 2009. "Pdx1 (MODY4) regulates pancreatic beta cell susceptibility to ER stress." Proc Natl Acad Sci U S A 106 (45):19090-5. doi: 10.1073/pnas.0904849106.

Saisho, Y., A. E. Butler, E. Manesso, D. Elashoff, R. A. Rizza, and P. C. Butler. 2013. "beta- cell mass and turnover in humans: effects of obesity and aging." Diabetes Care 36 (1):111-7. doi: 10.2337/dc12-0421.

Salpeter, S. J., A. Khalaileh, N. Weinberg-Corem, O. Ziv, B. Glaser, and Y. Dor. 2013. "Systemic regulation of the age-related decline of pancreatic beta-cell replication." Diabetes 62 (8):2843-8. doi: 10.2337/db13-0160.

Salpeter, S. J., A. M. Klein, D. Huangfu, J. Grimsby, and Y. Dor. 2010. "Glucose and aging control the quiescence period that follows pancreatic beta cell replication." Development 137 (19):3205-13. doi: 10.1242/dev.054304.

Sandhu, S. K., M. Fassan, S. Volinia, F. Lovat, V. Balatti, Y. Pekarsky, and C. M. Croce. 2013. "B-cell malignancies in microRNA Emu-miR-17~92 transgenic mice." Proc Natl Acad Sci U S A 110 (45):18208-13. doi: 10.1073/pnas.1315365110.

Sansbury, F. H., S. E. Flanagan, J. A. Houghton, F. L. Shuixian Shen, A. M. Al-Senani, A. M. Habeb, M. Abdullah, A. Kariminejad, S. Ellard, and A. T. Hattersley. 2012. "SLC2A2 mutations can cause neonatal diabetes, suggesting GLUT2 may have a role in human insulin secretion." Diabetologia 55 (9):2381-5. doi: 10.1007/s00125- 012-2595-0.

Scaglia, L., F. E. Smith, and S. Bonner-Weir. 1995. "Apoptosis contributes to the involution of beta cell mass in the post partum rat pancreas." Endocrinology 136 (12):5461-8. doi: 10.1210/endo.136.12.7588296.

Schisler, J. C., P. T. Fueger, D. A. Babu, H. E. Hohmeier, J. S. Tessem, D. Lu, T. C. Becker, B. Naziruddin, M. Levy, R. G. Mirmira, and C. B. Newgard. 2008. "Stimulation of human and rat islet beta-cell proliferation with retention of function by the homeodomain transcription factor Nkx6.1." Mol Cell Biol 28 (10):3465-76. doi: 10.1128/MCB.01791-07.

220

Schisler, J. C., P. B. Jensen, D. G. Taylor, T. C. Becker, F. K. Knop, S. Takekawa, M. German, G. C. Weir, D. Lu, R. G. Mirmira, and C. B. Newgard. 2005. "The Nkx6.1 homeodomain transcription factor suppresses glucagon expression and regulates glucose-stimulated insulin secretion in islet beta cells." Proc Natl Acad Sci U S A 102 (20):7297-302. doi: 10.1073/pnas.0502168102.

Schroder, A. R., P. Shinn, H. Chen, C. Berry, J. R. Ecker, and F. Bushman. 2002. "HIV-1 integration in the human genome favors active genes and local hotspots." Cell 110 (4):521-9.

Scott, R. A., V. Lagou, R. P. Welch, E. Wheeler, M. E. Montasser, J. Luan, R. Magi, R. J. Strawbridge, E. Rehnberg, S. Gustafsson, S. Kanoni, L. J. Rasmussen-Torvik, L. Yengo, C. Lecoeur, D. Shungin, S. Sanna, C. Sidore, P. C. Johnson, J. W. Jukema, T. Johnson, A. Mahajan, N. Verweij, G. Thorleifsson, J. J. Hottenga, S. Shah, A. V. Smith, B. Sennblad, C. Gieger, P. Salo, M. Perola, N. J. Timpson, D. M. Evans, B. S. Pourcain, Y. Wu, J. S. Andrews, J. Hui, L. F. Bielak, W. Zhao, M. Horikoshi, P. Navarro, A. Isaacs, J. R. O'Connell, K. Stirrups, V. Vitart, C. Hayward, T. Esko, E. Mihailov, R. M. Fraser, T. Fall, B. F. Voight, S. Raychaudhuri, H. Chen, C. M. Lindgren, A. P. Morris, N. W. Rayner, N. Robertson, D. Rybin, C. T. Liu, J. S. Beckmann, S. M. Willems, P. S. Chines, A. U. Jackson, H. M. Kang, H. M. Stringham, K. Song, T. Tanaka, J. F. Peden, A. Goel, A. A. Hicks, P. An, M. Muller-Nurasyid, A. Franco-Cereceda, L. Folkersen, L. Marullo, H. Jansen, A. J. Oldehinkel, M. Bruinenberg, J. S. Pankow, K. E. North, N. G. Forouhi, R. J. Loos, S. Edkins, T. V. Varga, G. Hallmans, H. Oksa, M. Antonella, R. Nagaraja, S. Trompet, I. Ford, S. J. Bakker, A. Kong, M. Kumari, B. Gigante, C. Herder, P. B. Munroe, M. Caulfield, J. Antti, M. Mangino, K. Small, I. Miljkovic, Y. Liu, M. Atalay, W. Kiess, A. L. James, F. Rivadeneira, A. G. Uitterlinden, C. N. Palmer, A. S. Doney, G. Willemsen, J. H. Smit, S. Campbell, O. Polasek, L. L. Bonnycastle, S. Hercberg, M. Dimitriou, J. L. Bolton, G. R. Fowkes, P. Kovacs, J. Lindstrom, T. Zemunik, S. Bandinelli, S. H. Wild, H. V. Basart, W. Rathmann, H. Grallert, D. IAbetes Genetics Replication, Consortium Meta-analysis, W. Maerz, M. E. Kleber, B. O. Boehm, A. Peters, P. P. Pramstaller, M. A. Province, I. B. Borecki, N. D. Hastie, I. Rudan, H. Campbell, H. Watkins, M. Farrall, M. Stumvoll, L. Ferrucci, D. M. Waterworth, R. N. Bergman, F. S. Collins, J. Tuomilehto, R. M. Watanabe, E. J. de Geus, B. W. Penninx, A. Hofman, B. A. Oostra, B. M. Psaty, P. Vollenweider, J. F. Wilson, A. F. Wright, G. K. Hovingh, A. Metspalu, M. Uusitupa, P. K. Magnusson, K. O. Kyvik, J. Kaprio, J. F. Price, G. V. Dedoussis, P. Deloukas, P. Meneton, L. Lind, M. Boehnke, A. R. Shuldiner, C. M. van Duijn, A. D. Morris, A. Toenjes, P. A. Peyser, J. P. Beilby, A. Korner, J. Kuusisto, M. Laakso, 221

S. R. Bornstein, P. E. Schwarz, T. A. Lakka, R. Rauramaa, L. S. Adair, G. D. Smith, T. D. Spector, T. Illig, U. de Faire, A. Hamsten, V. Gudnason, M. Kivimaki, A. Hingorani, S. M. Keinanen-Kiukaanniemi, T. E. Saaristo, D. I. Boomsma, K. Stefansson, P. van der Harst, J. Dupuis, N. L. Pedersen, N. Sattar, T. B. Harris, F. Cucca, S. Ripatti, V. Salomaa, K. L. Mohlke, B. Balkau, P. Froguel, A. Pouta, M. R. Jarvelin, N. J. Wareham, N. Bouatia-Naji, M. I. McCarthy, P. W. Franks, J. B. Meigs, T. M. Teslovich, J. C. Florez, C. Langenberg, E. Ingelsson, I. Prokopenko, and I. Barroso. 2012. "Large-scale association analyses identify new loci influencing glycemic traits and provide insight into the underlying biological pathways." Nat Genet 44 (9):991-1005. doi: 10.1038/ng.2385.

Scott, R. A., L. J. Scott, R. Magi, L. Marullo, K. J. Gaulton, M. Kaakinen, N. Pervjakova, T. H. Pers, A. D. Johnson, J. D. Eicher, A. U. Jackson, T. Ferreira, Y. Lee, C. Ma, V. Steinthorsdottir, G. Thorleifsson, L. Qi, N. R. Van Zuydam, A. Mahajan, H. Chen, P. Almgren, B. F. Voight, H. Grallert, M. Muller-Nurasyid, J. S. Ried, N. W. Rayner, N. Robertson, L. C. Karssen, E. M. van Leeuwen, S. M. Willems, C. Fuchsberger, P. Kwan, T. M. Teslovich, P. Chanda, M. Li, Y. Lu, C. Dina, D. Thuillier, L. Yengo, L. Jiang, T. Sparso, H. A. Kestler, H. Chheda, L. Eisele, S. Gustafsson, M. Franberg, R. J. Strawbridge, R. Benediktsson, A. B. Hreidarsson, A. Kong, G. Sigurethsson, N. D. Kerrison, J. Luan, L. Liang, T. Meitinger, M. Roden, B. Thorand, T. Esko, E. Mihailov, C. Fox, C. T. Liu, D. Rybin, B. Isomaa, V. Lyssenko, T. Tuomi, D. J. Couper, J. S. Pankow, N. Grarup, C. T. Have, M. E. Jorgensen, T. Jorgensen, A. Linneberg, M. C. Cornelis, R. M. van Dam, D. J. Hunter, P. Kraft, Q. Sun, S. Edkins, K. R. Owen, J. R. B. Perry, A. R. Wood, E. Zeggini, J. Tajes-Fernandes, G. R. Abecasis, L. L. Bonnycastle, P. S. Chines, H. M. Stringham, H. A. Koistinen, L. Kinnunen, B. Sennblad, T. W. Muhleisen, M. M. Nothen, S. Pechlivanis, D. Baldassarre, K. Gertow, S. E. Humphries, E. Tremoli, N. Klopp, J. Meyer, G. Steinbach, R. Wennauer, J. G. Eriksson, S. Mnnisto, L. Peltonen, E. Tikkanen, G. Charpentier, E. Eury, S. Lobbens, B. Gigante, K. Leander, O. McLeod, E. P. Bottinger, O. Gottesman, D. Ruderfer, M. Bluher, P. Kovacs, A. Tonjes, N. M. Maruthur, C. Scapoli, R. Erbel, K. H. Jockel, S. Moebus, U. de Faire, A. Hamsten, M. Stumvoll, P. Deloukas, P. J. Donnelly, T. M. Frayling, A. T. Hattersley, S. Ripatti, V. Salomaa, N. L. Pedersen, B. O. Boehm, R. N. Bergman, F. S. Collins, K. L. Mohlke, J. Tuomilehto, T. Hansen, O. Pedersen, I. Barroso, L. Lannfelt, E. Ingelsson, L. Lind, C. M. Lindgren, S. Cauchi, P. Froguel, R. J. F. Loos, B. Balkau, H. Boeing, P. W. Franks, A. Barricarte Gurrea, D. Palli, Y. T. van der Schouw, D. Altshuler, L. C. Groop, C. Langenberg, N. J. Wareham, E. Sijbrands, C. M. van Duijn, J. C. Florez, J. B. Meigs, E. Boerwinkle, C. Gieger, K. Strauch, A. Metspalu, A. D. Morris, C. N. A. Palmer, F. B. Hu, U. Thorsteinsdottir, K. Stefansson, J. Dupuis, A. P. Morris, M. Boehnke, M. I. 222

McCarthy, I. Prokopenko, D. IAbetes Genetics Replication, and Consortium Meta-analysis. 2017. "An Expanded Genome-Wide Association Study of Type 2 Diabetes in Europeans." Diabetes 66 (11):2888-2902. doi: 10.2337/db16-1253.

Seiler, C. Y., J. G. Park, A. Sharma, P. Hunter, P. Surapaneni, C. Sedillo, J. Field, R. Algar, A. Price, J. Steel, A. Throop, M. Fiacco, and J. LaBaer. 2014. "DNASU plasmid and PSI:Biology-Materials repositories: resources to accelerate biological research." Nucleic Acids Res 42 (Database issue):D1253-60. doi: 10.1093/nar/gkt1060.

Sharma, S., U. S. Jhala, T. Johnson, K. Ferreri, J. Leonard, and M. Montminy. 1997. "Hormonal regulation of an islet-specific enhancer in the pancreatic homeobox gene STF-1." Mol Cell Biol 17 (5):2598-604.

Sharma, S., J. Leonard, S. Lee, H. D. Chapman, E. H. Leiter, and M. R. Montminy. 1996. "Pancreatic islet expression of the homeobox factor STF-1 relies on an E-box motif that binds USF." J Biol Chem 271 (4):2294-9.

Shayakhmetov, D. M., T. Papayannopoulou, G. Stamatoyannopoulos, and A. Lieber. 2000. "Efficient gene transfer into human CD34(+) cells by a retargeted adenovirus vector." J Virol 74 (6):2567-83.

Shields, B. M., S. Hicks, M. H. Shepherd, K. Colclough, A. T. Hattersley, and S. Ellard. 2010. "Maturity-onset diabetes of the young (MODY): how many cases are we missing?" Diabetologia 53 (12):2504-8. doi: 10.1007/s00125-010-1799-4.

Smith, E. P., Z. An, C. Wagner, A. G. Lewis, E. B. Cohen, B. Li, P. Mahbod, D. Sandoval, D. Perez-Tilve, N. Tamarina, L. H. Philipson, D. A. Stoffers, R. J. Seeley, and D. A. D'Alessio. 2014. "The role of beta cell glucagon-like peptide-1 signaling in glucose regulation and response to diabetes drugs." Cell Metab 19 (6):1050-7. doi: 10.1016/j.cmet.2014.04.005.

Spaeth, J. M., M. Gupte, M. Perelis, Y. P. Yang, H. Cyphert, S. Guo, J. H. Liu, M. Guo, J. Bass, M. A. Magnuson, C. Wright, and R. Stein. 2017. "Defining a Novel Role for the Pdx1 Transcription Factor in Islet beta-Cell Maturation and Proliferation During Weaning." Diabetes 66 (11):2830-2839. doi: 10.2337/db16-1516.

223

Steiner, D. J., A. Kim, K. Miller, and M. Hara. 2010. "Pancreatic islet plasticity: interspecies comparison of islet architecture and composition." Islets 2 (3):135-45.

Stephens, S. B., J. C. Schisler, H. E. Hohmeier, J. An, A. Y. Sun, G. S. Pitt, and C. B. Newgard. 2012. "A VGF-derived peptide attenuates development of type 2 diabetes via enhancement of islet beta-cell survival and function." Cell Metab 16 (1):33-43. doi: 10.1016/j.cmet.2012.05.011.

Stiles, B. L., C. Kuralwalla-Martinez, W. Guo, C. Gregorian, Y. Wang, J. Tian, M. A. Magnuson, and H. Wu. 2006. "Selective deletion of Pten in pancreatic beta cells leads to increased islet mass and resistance to STZ-induced diabetes." Mol Cell Biol 26 (7):2772-81. doi: 10.1128/MCB.26.7.2772-2781.2006.

Stoffers, D. A., J. Ferrer, W. L. Clarke, and J. F. Habener. 1997a. "Early-onset type-II diabetes mellitus (MODY4) linked to IPF1." Nat Genet 17 (2):138-9. doi: 10.1038/ng1097-138.

Stoffers, D. A., N. T. Zinkin, V. Stanojevic, W. L. Clarke, and J. F. Habener. 1997b. "Pancreatic agenesis attributable to a single nucleotide deletion in the human IPF1 gene coding sequence." Nat Genet 15 (1):106-10. doi: 10.1038/ng0197-106.

Stolovich-Rain, M., J. Enk, J. Vikesa, F. C. Nielsen, A. Saada, B. Glaser, and Y. Dor. 2015. "Weaning triggers a maturation step of pancreatic beta cells." Dev Cell 32 (5):535- 45. doi: 10.1016/j.devcel.2015.01.002.

Stolovich-Rain, M., A. Hija, J. Grimsby, B. Glaser, and Y. Dor. 2012. "Pancreatic beta cells in very old mice retain capacity for compensatory proliferation." J Biol Chem 287 (33):27407-14. doi: 10.1074/jbc.M112.350736.

Stoy, J., E. L. Edghill, S. E. Flanagan, H. Ye, V. P. Paz, A. Pluzhnikov, J. E. Below, M. G. Hayes, N. J. Cox, G. M. Lipkind, R. B. Lipton, S. A. Greeley, A. M. Patch, S. Ellard, D. F. Steiner, A. T. Hattersley, L. H. Philipson, G. I. Bell, and Group Neonatal Diabetes International Collaborative. 2007. "Insulin gene mutations as a cause of permanent neonatal diabetes." Proc Natl Acad Sci U S A 104 (38):15040-4. doi: 10.1073/pnas.0707291104.

224

Talchai, C., S. Xuan, H. V. Lin, L. Sussel, and D. Accili. 2012. "Pancreatic beta cell dedifferentiation as a mechanism of diabetic beta cell failure." Cell 150 (6):1223- 34. doi: 10.1016/j.cell.2012.07.029.

Tanaka, Y., C. E. Gleason, P. O. Tran, J. S. Harmon, and R. P. Robertson. 1999. "Prevention of glucose toxicity in HIT-T15 cells and Zucker diabetic fatty rats by antioxidants." Proc Natl Acad Sci U S A 96 (19):10857-62.

Tattikota, S. G., T. Rathjen, S. J. McAnulty, H. H. Wessels, I. Akerman, M. van de Bunt, J. Hausser, J. L. Esguerra, A. Musahl, A. K. Pandey, X. You, W. Chen, P. L. Herrera, P. R. Johnson, D. O'Carroll, L. Eliasson, M. Zavolan, A. L. Gloyn, J. Ferrer, R. Shalom-Feuerstein, D. Aberdam, and M. N. Poy. 2014. "Argonaute2 mediates compensatory expansion of the pancreatic beta cell." Cell Metab 19 (1):122-34. doi: 10.1016/j.cmet.2013.11.015.

Teo, A. K., N. Tsuneyoshi, S. Hoon, E. K. Tan, L. W. Stanton, C. V. Wright, and N. R. Dunn. 2015. "PDX1 binds and represses hepatic genes to ensure robust pancreatic commitment in differentiating human embryonic stem cells." Stem Cell Reports 4 (4):578-90. doi: 10.1016/j.stemcr.2015.02.015.

Tersey, S. A., Y. Nishiki, A. T. Templin, S. M. Cabrera, N. D. Stull, S. C. Colvin, C. Evans- Molina, J. L. Rickus, B. Maier, and R. G. Mirmira. 2012. "Islet beta-cell endoplasmic reticulum stress precedes the onset of type 1 diabetes in the nonobese diabetic mouse model." Diabetes 61 (4):818-27. doi: 10.2337/db11-1293.

Tessem, J. S., L. G. Moss, L. C. Chao, M. Arlotto, D. Lu, M. V. Jensen, S. B. Stephens, P. Tontonoz, H. E. Hohmeier, and C. B. Newgard. 2014. "Nkx6.1 regulates islet beta- cell proliferation via Nr4a1 and Nr4a3 nuclear receptors." Proc Natl Acad Sci U S A 111 (14):5242-7. doi: 10.1073/pnas.1320953111.

Teta, M., S. Y. Long, L. M. Wartschow, M. M. Rankin, and J. A. Kushner. 2005. "Very slow turnover of beta-cells in aged adult mice." Diabetes 54 (9):2557-67.

Thakore, P. I., J. B. Black, I. B. Hilton, and C. A. Gersbach. 2016. "Editing the epigenome: technologies for programmable transcription and epigenetic modulation." Nat Methods 13 (2):127-37. doi: 10.1038/nmeth.3733. 225

Thakore, P. I., A. M. D'Ippolito, L. Song, A. Safi, N. K. Shivakumar, A. M. Kabadi, T. E. Reddy, G. E. Crawford, and C. A. Gersbach. 2015. "Highly specific epigenome editing by CRISPR-Cas9 repressors for silencing of distal regulatory elements." Nat Methods 12 (12):1143-9. doi: 10.1038/nmeth.3630.

Tong, Z., Y. Fan, W. Zhang, J. Xu, J. Cheng, M. Ding, and H. Deng. 2009. "Pancreas- specific Pten deficiency causes partial resistance to diabetes and elevated hepatic AKT signaling." Cell Res 19 (6):710-9. doi: 10.1038/cr.2009.42.

Uhlen, M., L. Fagerberg, B. M. Hallstrom, C. Lindskog, P. Oksvold, A. Mardinoglu, A. Sivertsson, C. Kampf, E. Sjostedt, A. Asplund, I. Olsson, K. Edlund, E. Lundberg, S. Navani, C. A. Szigyarto, J. Odeberg, D. Djureinovic, J. O. Takanen, S. Hober, T. Alm, P. H. Edqvist, H. Berling, H. Tegel, J. Mulder, J. Rockberg, P. Nilsson, J. M. Schwenk, M. Hamsten, K. von Feilitzen, M. Forsberg, L. Persson, F. Johansson, M. Zwahlen, G. von Heijne, J. Nielsen, and F. Ponten. 2015. "Proteomics. Tissue- based map of the human proteome." Science 347 (6220):1260419. doi: 10.1126/science.1260419.

Upton, J. P., L. Wang, D. Han, E. S. Wang, N. E. Huskey, L. Lim, M. Truitt, M. T. McManus, D. Ruggero, A. Goga, F. R. Papa, and S. A. Oakes. 2012. "IRE1alpha cleaves select microRNAs during ER stress to derepress translation of proapoptotic Caspase-2." Science 338 (6108):818-22. doi: 10.1126/science.1226191.

Ventura, A., A. G. Young, M. M. Winslow, L. Lintault, A. Meissner, S. J. Erkeland, J. Newman, R. T. Bronson, D. Crowley, J. R. Stone, R. Jaenisch, P. A. Sharp, and T. Jacks. 2008. "Targeted deletion reveals essential and overlapping functions of the miR-17 through 92 family of miRNA clusters." Cell 132 (5):875-86. doi: 10.1016/j.cell.2008.02.019.

Venturini, L., K. Battmer, M. Castoldi, B. Schultheis, A. Hochhaus, M. U. Muckenthaler, A. Ganser, M. Eder, and M. Scherr. 2007. "Expression of the miR-17-92 polycistron in chronic myeloid leukemia (CML) CD34+ cells." Blood 109 (10):4399-405. doi: 10.1182/blood-2006-09-045104.

226

Waeber, G., N. Thompson, P. Nicod, and C. Bonny. 1996. "Transcriptional activation of the GLUT2 gene by the IPF-1/STF-1/IDX-1 homeobox factor." Mol Endocrinol 10 (11):1327-34. doi: 10.1210/mend.10.11.8923459.

Wang, A., F. Yue, Y. Li, R. Xie, T. Harper, N. A. Patel, K. Muth, J. Palmer, Y. Qiu, J. Wang, D. K. Lam, J. C. Raum, D. A. Stoffers, B. Ren, and M. Sander. 2015. "Epigenetic priming of enhancers predicts developmental competence of hESC- derived endodermal lineage intermediates." Cell Stem Cell 16 (4):386-99. doi: 10.1016/j.stem.2015.02.013.

Wang, L., Y. Liu, S. Yan Lu, K. T. Nguyen, S. A. Schroer, A. Suzuki, T. W. Mak, H. Gaisano, and M. Woo. 2010. "Deletion of Pten in pancreatic ss-cells protects against deficient ss-cell mass and function in mouse models of type 2 diabetes." Diabetes 59 (12):3117-26. doi: 10.2337/db09-1805.

Wang, S., J. Gao, Q. Lei, N. Rozengurt, C. Pritchard, J. Jiao, G. V. Thomas, G. Li, P. Roy- Burman, P. S. Nelson, X. Liu, and H. Wu. 2003. "Prostate-specific deletion of the murine Pten tumor suppressor gene leads to metastatic prostate cancer." Cancer Cell 4 (3):209-21.

Wang, Y., J. Liu, C. Liu, A. Naji, and D. A. Stoffers. 2013. "MicroRNA-7 regulates the mTOR pathway and proliferation in adult pancreatic beta-cells." Diabetes 62 (3):887-95. doi: 10.2337/db12-0451.

Woods, K., J. M. Thomson, and S. M. Hammond. 2007. "Direct regulation of an oncogenic micro-RNA cluster by E2F transcription factors." J Biol Chem 282 (4):2130-4. doi: 10.1074/jbc.C600252200.

Wu, H., T. Seki, I. Dmitriev, T. Uil, E. Kashentseva, T. Han, and D. T. Curiel. 2002. "Double modification of adenovirus fiber with RGD and polylysine motifs improves coxsackievirus-adenovirus receptor-independent gene transfer efficiency." Hum Gene Ther 13 (13):1647-53. doi: 10.1089/10430340260201734.

Wu, X., Y. Li, B. Crise, and S. M. Burgess. 2003. "Transcription start regions in the human genome are favored targets for MLV integration." Science 300 (5626):1749-51. doi: 10.1126/science.1083413. 227

Xu, G., J. Chen, G. Jing, and A. Shalev. 2013. "Thioredoxin-interacting protein regulates insulin transcription through microRNA-204." Nat Med 19 (9):1141-6. doi: 10.1038/nm.3287.

Yang, Y. P., M. A. Magnuson, R. Stein, and C. V. Wright. 2017. "The mammal-specific Pdx1 Area II enhancer has multiple essential functions in early endocrine cell specification and postnatal beta-cell maturation." Development 144 (2):248-257. doi: 10.1242/dev.143123.

Yin, X., C. Giap, J. S. Lazo, and E. V. Prochownik. 2003. "Low molecular weight inhibitors of Myc-Max interaction and function." Oncogene 22 (40):6151-9. doi: 10.1038/sj.onc.1206641.

Zhou, Y., D. He, J. Zeng, S. Bao, J. Lai, Y. Weng, and S. Chen. 2017. "The Effects of Antisense miRNA-20a Alone or in Combination with Imatinib on K562 Cell Proliferation." Front Pharmacol 8:127. doi: 10.3389/fphar.2017.00127.

228

Biography

Jonathan M. Haldeman was born in Manheim, Pennsylvania, USA. He earned dual bachelor’s degrees in Biology and Chemistry from Millersville University,

Millersville, Pennsylvania, USA. The last summer of his undergraduate education was spent at Duke University under the direction of Dr. Thomas Becker, a fellow Millersville

University alumnus, focused on the role miRNAs in β-cell function which was presented by Dr. Becker at a Keystone Islet Biology meeting. Jonathan completed his PhD in 2018 in Pharmacology with a certificate in Cellular and Molecular Biology at Duke

University, Durham, North Carolina, USA. During graduate school he served as the Cell and Molecular Biology program’s representative to the Graduate and Professional

Student Council (GPSC), participated in the Duke Scholars of Molecular Medicine program (DSMM), and mentored high school and undergraduate students who came to the lab through Project SEED and other avenues. During his time at Duke, he was a co- author on six publications (Hayes et al. 2013, Jensen et al. 2013, Conway et al. 2015,

Hayes et al. 2016, Hayes et al. 2017, An et al. 2018), in addition to a first-author manuscript describing the pMVP and pMAGIC platforms that is currently under review.

229