University of Cincinnati

Date: August 23, 2012

The Role of Sox17 in Normal and Pathological Beta Cell

A dissertation submitted to the Division of Graduate Studies and Research of the

University of Cincinnati

in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

in the Graduate Program in Molecular and Developmental Biology

of the College of Medicine

2012

by

Diva Jonatan

Bachelor of Science, Xavier University, 2006

Committee Chair: James Wells, PhD

Jeffrey Whitsett, MD, PhD

Aaron Zorn, PhD

Jonathan Katz, PhD

Gail Deutsch, MD ABSTRACT

Glucose homeostasis is a complex process involving many regulatory molecules and disruption of this process can result in diabetes. Sox17 is a transcription factor and a key regulator in various developmental and disease contexts. During endoderm development, Sox17 acts, in part, by transcriptionally regulating other important endodermal transcription factors including HNF and

Foxa2, which are also known regulators of postnatal cell function. Our data shows that Sox17 is expressed in all islet cells. Thus, we hypothesize that Sox17 is a key modulator of cell homeostasis.

In the course of this study, we discovered a novel role for Sox17 in regulating insulin trafficking and secretion in normal and pathological cells.

Loss of Sox17 throughout pancreas development in a wildtype background resulted in mice with prediabetes. These mice had higher proinsulin content in the ER of the islet cells and dilated secretory organelles in cells. In line with the prediabetes phenotype, these mice went on to develop additional symptoms of diabetes when placed on a high fat diet, including elevated fasting levels and an inability to respond to a glucose challenge. This suggested that Sox17 affects either insulin processing and/or transit through the secretory system. To more directly investigate these possibilities, we used a tetracycline regulated transgenic system to overexpress Sox17 in mature cells in wildtype background. Transgenic Sox17 expression resulted in rapid trafficking and secretion of improperly processed proinsulin. At the transcriptional level, Sox17 altered expression of involved in biological processes that regulate

iii hormone transport, secretion, and cellular localization, which led to diabetes after prolonged exposure. This demonstrates that Sox17 affects insulin trafficking throughout the secretory pathway. We therefore wanted to explore the possibility that physiologic levels of Sox17 might be used to positively impact diabetic phenotypes using a genetic model of diabetes (MODY4). We did so by overexpressing Sox17 two-fold in a MODY4 mouse model (Pdx1-tTA hemizygote mice). Increased expression of Sox17 in the -cells of MODY4 animals was sufficient to transiently normalize basal blood glucose and insulin levels as well as restore islet cell organization architecture; however, Sox17 overexpression was not able to rescue the inability of MODY4 animals to properly respond to glucose challenge.

Together, these data demonstrate new and critical role for Sox17 in regulating insulin trafficking and secretion processes in the adult pancreas that are important to ensure proper glucose homeostasis. This study also suggests that modulation of Sox17-regulated pathways can be used therapeutically to improve cell function in the context of diabetes.

iv v Table of Contents

ABSTRACT iii

TABLE OF CONTENTS 1

LIST OF FIGURES AND TABLES 4

CHAPTER 1. Introduction

Glucose homeostasis, prediabetes and the different forms of diabetes 7

MODY and the Islet Transcriptional Factor Network Involved 13

Insulin Biosynthesis, Processing, and Secretion 14

Overview: Functional Roles of Sox Family 18

Overview of Sox Family of Transcription Factors 18

Sox Proteins in Endocrine Pancreas Development and Function 19

The Roles of Sox17 in Different Biological Contexts 22

References 26

Figure Legends 34

Figures 35

CHAPTER 2. The Role of Sox17 in Insulin Processing and  Cell Secretory Pathway

Abstract 39

Introduction 40

Materials and methods 43

Results 47

1 Discussion 59

Acknowledgements 66

Sources of Funding 66

References 67

Figure Legends 73

Figures 78

Supplementary Figure Legends 86

Supplementary Figures 89

CHAPTER 3. The Partial Rescue of MODY4 Phenotypes by Sox17

Summary 115

Introduction 116

Materials and Methods 121

Results 123

Discussion 128

Acknowledgements 130

Sources of Funding 131

References 131

Figure Legends 134

Figures 137

CHAPTER 4. Summary and Discussion

Major Findings 144

2 Sox17 in normal cells 144

Sox17 in pathological cells 146

Experimental limitations and alternative approaches 147

Acknowledgements 149

Sources of funding 150

References 150

Figure Legends 152

Figures 152

3 LIST OF FIGURES AND TABLES

CHAPTER 1.

Figure 1. MODY genes islet transcriptional network 35

Figure 2. Insulin biosynthesis in pancreatic cells 36

Figure 3. Glucose sensing and glucose-stimulated insulin release pathway 37

CHAPTER 2.

Figure 1. Sox17 is not required for cell development 78

Figure 2. Sox17-paLOF results in elevated proinsulin protein in the islets 79

Figure 3. Loss of Sox17 in the pancreas causes accumulation of proinsulin in the ER and structural changes in secretory organelles 80

Figure 4. Sox17-paLOF mice are prediabetic and prone to high fat diet-induced stress of

cells 81

Figure 5. Sox17 overexpression for 24 hours is sufficient to alter proinsulin:total insulin protein ratio and proinsulin secretion in vivo, followed by accumulation of proinsulin in the plasma, leading to diabetes after prolonged exposure of Sox17 82

Figure 6. 24 hours of Sox17 overexpression alters proinsulin trafficking through the secretory organelle machinery 84

Figure 7. Sox17 regulates pathways involved in insulin transport and secretion 85

Supplementary Figure 1. Sox17 immunostaining in Sox17pa-LOF, wildtype, and Sox17-

GOF islets 89

4 Supplementary Figure 2. Percent colocalizations between proinsulin and organelle markers, and their total regional areas 89

Supplementary Figure 3. Plasma proinsulin, total insulin, and their ratio in Sox17-paLOF mice on high fat diet 90

Supplementary Figure 4. Transcriptional insulin processing level 91

Supplementary Figure 5. Plasma proinsulin and total insulin levels in Sox17-GOF mice

91

Supplementary Figure 6. Percent colocalization and total regional areas of proinsulin and various organelle markers 92

Supplementary Figure 7. Microarray validation of altered genes in Sox17-GOF islets

93

Supplementary Table 1. Primary and Secondary Antibodies 95

Supplementary Table 2. Sox17-GOF islet microarray list 97

CHAPTER 3.

Figure 1. Sox17 overepression in the context of MODY4 background rescued resting hyperglycemia 135

Figure 2. Sox17 overexpression rescued MODY4 disrupted islet architecture 138

Figure 3. Sox17 altered the distribution of islet sizes and cell-cell adhesion contacts of the MODY4 mice 140

CHAPTER 4.

Figure 1. The role of Sox17 in regulating insulin trafficking

5 Chapter 1.

Introduction

6 Glucose homeostasis, prediabetes and the different forms of diabetes.

“Disease does not occur unexpectedly, it is the result of constant violation of Nature’s laws.”- Hippocrates

Glucose is the body’s primary source of energy and maintenance of blood glucose homeostasis is achieved by a complex endocrine regulatory network.

Central to this network is the hormone insulin, which is secreted in response to elevated glucose levels and acts on peripheral tissues in several ways. In the liver, insulin promotes conversion of glucose to glycogen for storage. In other tissues like skeletal muscle, insulin activates the insulin signaling receptor, which interfaces with several downstream effector pathways, such as the insulin receptor substrates/phosphatidylinositol 3- pathway (IRS/PI3-K) and the

Ras/mitogen-activated (MAPK) pathway, to mediate glucose uptake and (reviewed in1,2).

Insulin protein is regulated at several levels: insulin biosynthesis, processing, secretion, cell uptake, and its breakdown in the body. A defect in any of these stages of insulin regulation, or loss of cell mass, can lead to elevated blood glucose levels (hyperglycemia), and over the long term, this can result in diabetes. The World Health Organization Diabetes Fact Sheet 2011 suggests that there are 346 millions of people with diabetes worldwide. According to the

2011 National Diabetes Fact Sheet, released by the American Diabetes

Association, there is a total 25.8 million of diabetic children and adults in the

United States, which is 8.3% of the population. In addition, 79 million people are considered to be prediabetic, meaning these people are exhibiting one or more

7 symptoms of glucose dysregulation, including impaired fasting blood glucose, elevated ratio of plasma proinsulin:insulin, and impaired glucose tolerance3-8.In terms of the rate of mortality, it is estimated that diabetes causes more deaths per year than breast cancer and AIDS combined.

There are several different types of diabetes, such as the polygenic forms of diabetes: type 1 diabetes and type 2 diabetes, and the monogenic forms of diabetes: mature onset diabetes of the young (MODY) and neonatal diabetes.

Type 1 diabetes (~5% of diabetes cases, according to the National Diabetes

Information Clearinghouse - NIDC) is a result of autoimmune-mediated destruction of pancreatic cells, leading to loss of insulin production. It is usually diagnosed in children and young adults. Type 2 diabetes is the most common form of diabetes (90-95% of diabetes cases) and is associated with either insufficient production of insulin or peripheral insulin insensitivity. MODY (1-2% of diabetes cases) results from mutations in single genes involved in glucose sensing and/or insulin regulation. There are 10 MODY genes discovered so far, including H, Pdx1, HNF1, NeuroD1, KLF11, CEL,

Pax4, and the Insulin gene itself. This type of diabetes is characterized by young onset and autosomal dominant inheritance. In the case of neonatal diabetes, a rare condition occurring in only one out of 100,000 – 500,000 live births, it is known that insulin regulation is disrupted in these diabetic infants. They are usually diagnosed in the first 6 months of life. In all forms of diabetes, complications including retinopathies, neuropathies, nephropathies, and

8 incidences of extreme hypo- and hyper-glycemia that can lead to coma and death are known to occur.

From studies, we know that most patients enter a prediabetic stage prior to developing full blown type 2 diabetes. Prediabetes is characterized by blood glucose levels that are higher than normal but not yet high enough to be classified as diabetes. In mouse models of diabetes, such as the NOD and Akita mice, a prediabetes stage was also observed before the mice became full diabetics3,9. Several hallmarks of prediabetes in these mice included impaired fasting blood glucose, impaired glucose tolerance, disproportionately elevated proinsulin:insulin levels, and dilated (ER)3-5,9-11. Taken together, these data suggest that defective cell function, particularly in the insulin trafficking process, are hallmarks to the prediabetic stage and may contribute to the development of Type 2 diabetes.

Our understanding of the biology of prediabetes is limited, in large part due to a lack of diverse pool of animal models that more accurately represent the genetic diversity that is found in human Type 2 diabetes. However, a prediabetes stage has been observed in mouse models of other forms of diabetes. For example, as described above, the Akita mouse. It harbors a mutation in the

Insulin-2 gene that results in improper folding of the mutant protein3. This mouse model is most similar to the permanent neonatal diabetes that is caused by human mutations in Insulin12. In the Akita mouse, misfolded proinsulin proteins were accumulated in both the ER and pre-Golgi organelles, but mainly in the pre-

Golgi13. These animals had a progressive deterioration of secretory organelle

9 structure and function that accompanied the transition from prediabetes into full diabetes. By 4 weeks of age when blood glucose began to increase and the mice were becoming glucose intolerant, a decrease in insulin secretory granules and an enrichment of distended ER-like structures were observed. By 13 weeks of age, the lumen area of the ER increased, and by 18 weeks, the cytoplasm became filled with vacuole-like structure and comprised of distended ER-like structures. Overall, this progressive organelle dysfunction in individual cells was thought to be the primary cause of diabetes in the Akita mouse3.Another mouse model is the NOD mouse that develops spontaneous autoimmune diabetes and shares many similarities to type 1 diabetics14. Specifically, the NOD mouse strain had pancreas-specific autoantibodies, autoreactive CD4+ and

CD8+ T cells, and disease genetic linkage similar to the human patients. In this model, prediabetes stage occurred when islets were infiltrated by macrophages and T cells, which led to a condition called insulitis. Previous study showed that

NOD mice progressively activated the unfolded protein response (UPR), assessed by the increased in transcript levels of ER stress-responsive genes, starting at 6-8 weeks of age until it no longer could compensate the ER stress by

10 weeks of age9. To note, there was a significant increased in the ratio of proinsulin:insulin by 10 weeks compared to CD1 and NOD-SCID control mice, suggesting that an ER folding/processing defect leads to secretion of improperly processed proinsulin. In line with this analysis, electron microscopy study also found fewer secretory granules and dilated, fragmented ER morphology in the

NOD mice. These mouse models suggest that in disease progression of both

10 Neonatal and Type 1 diabetes, defects in insulin trafficking and secretory organelle integrity are observed prior to development of full diabetes. These data further suggest that tight regulation of proinsulin biosynthesis and processing is central to normal cell function and that the regulatory molecules involved in insulin trafficking and secretion may be exploited therapeutically to prevent or delay the progression of diabetes.

In addition to mouse model studies, data from human studies have shed light on proinsulin processing as a measure of cell pathology. In the fasting circulation of human adults, only about 5-22% of proinsulin out of the total insulin is found in the plasma, suggesting an extremely efficient conversion of proinsulin to mature insulin under normal circumstances15. However, the healthy siblings of patients with type 1 diabetes showed that 30% of these siblings analyzed had significantly elevated levels of fasting proinsulin with a median range of 1.7-58 pM compared to 1.2-28 pM in the controls (p-value < 0.01). And in this study, they had normal blood glucose and fasting insulin levels16. A more recent study, with a total of 148 siblings of newly diagnosed type 1 diabetics, found that there were two different groups of siblings with elevated proinsulin level. One group had elevated proinsulin but remain unaffected and another group that had higher proinsulin and progressively became diabetics (n=12 siblings)4. These studies suggested that an increased in fasting proinsulin in the circulation is often, but not always, an indicator that a patient will go on to develop full diabetes.

Elevated ratio of plasma proinsulin:insulin is not only a characteristic of the prediabetes phase, it is also a well known feature of type 2 diabetes5,7,8,17.It was

11 known that basal proinsulin immunoreactivity in these type 2 diabetes patients is

2-3 times higher compared to the healthy subjects7,18,19. Roder and colleagues suggested an inverse correlation between fasting proinsulin and blood glucose levels; a high fasting proinsulin:insulin ratio in human type 2 diabetes patients indicates defects in the secretory pathway, resulting in higher fasting glycemia levels5. Additionally, it has been shown that the ER is distended and increased in size in type 2 diabetes patients20. The ER volume density in these patients was doubled, compared to the ER of non-diabetic controls though only minor difference in the ER stress marker genes was observed. These studies together suggested that patients of type 2 diabetes had altered efficiency of proinsulin to insulin conversion and insulin secretory pathway.

Taken together, these studies have identified several cell pathologies that are associated with development of all forms of diabetes. These include defects in insulin transport, processing and folding, and impairment of the secretory machinery, all resulting in impaired cell function and failure to maintain glucose homeostasis. Thus, the molecular process involved in insulin trafficking may be exploited therapeutically to prevent or delay the progression of diabetes.

12 MODY and the Islet Transcriptional Factor Network Involved.

“The reality check is that anyone that seems to be ‘doing it all’ is not doing it by herself. Not even close. Take a moment to look around from time to time, see who helps keep all of your balls in the air. When you figure it out, be grateful for them. If you find yourself short on help, enlist some. And, of course, be part of someone else’s village too when you can. It makes everything you accomplish together so much richer and more possible.” –Leigh Standley of Curly Girl Design

Studies of patients with maturity onset diabetes of the young (MODY) have led to the discovery of single gene mutations that disrupt a cell regulatory network that is required for cell homeostasis21-24. MODY patients are distinguished from type 1 diabetes patients by the absence of autoantibodies against pancreatic antigens and the presence of C-peptide even with hyperglycemia outside their ‘honeymoon period’ (up to 5 years after diagnosis, with the remaining cells still able to secrete insulin). Furthermore, MODY clinical characteristics are also different from type 2 diabetes. MODY patients usually have prominent family history of diabetes in two or more generations

(autosomal dominant inheritance), they are not typically obese or insulin resistant, and they present their symptoms at a young age21,24.

There are 10 MODY genes discovered so far, including HNF4,

Glucokinase, HNF1, Pdx1, HNF1, NeuroD1, KLF11, CEL, Pax4, and the

Insulin gene itself. Taken together, the studies of MODY have led to a better understanding of the genetic causes of cell dysfunction and have identified a network of transcription factors that work together to ensure proper cell function.

Components of the network and how different transcription factors functionally relate to each other is summarized in Figure 123,25-30. This present study on

13 Chapter 3 will focus on MODY4 gene, Pdx1. Detailed introduction in regards to the function of Pdx1 in mature cells is outlined in Chapter 3.

Insulin Biosynthesis, Processing, and Secretion.

“There and back again, An insulin tale. ”

Insulin is the central endocrine hormonal signal that functions to regulate blood glucose levels after eating. Insulin is produced by pancreatic cells, which along with other endocrine cells, are organized into clusters of cells known as the islets of Langerhans (100-200 μm)31. The main function of a pancreatic cell is to secrete insulin in response to glucose, so much so that half of its protein synthesis capacity is dedicated to insulin production32. An average rodent cell consists of approximately 10,000 insulin granules (300-350 nm), which makes up to 10-20% of total cell volume33,34. Each granule stores about 200,000 insulin molecules, corresponding to ~50-60% of total granule protein. Taken together, there are approximately 2 billion insulin molecules per pancreatic cell35.

Therefore, cells maintain a tight molecular regulation of insulin biosynthesis, processing, and secretion to ensure that the cell correctly functions to regulate glucose homeostasis.

This assembly of the hormone insulin is a highly complex and regulated process, and perturbation of any of these steps can result in loss of insulin activity (a schematic of this process is depicted in Figure 2). This is a good example in biology where many highly regulated steps are required for a specific function. Insulin biosynthesis starts with the translation of a 110 amino acid,

14 single chain precursor preproinsulin in the ribosome. Within seconds, during, or immediately after synthesis, signal peptide is removed proteolytically to generate proinsulin (±86 amino acids long). This signal sequence is known to be important for stabilizing the ribosome on the rough ER membrane36. Within the preproinsulin structure, the signal sequence is followed by the B-chain (30 amino acids long), then the C-peptide (30-35 amino acids long), and followed by the A- chain (21 amino acids long). The C-peptide, or connecting peptide, connects the

A- and B-chains.

Once proinsulin is transported to the ER, proinsulin folding is coupled to specific pairing of three sets of cysteine residues that form disulfide bridges: A6 –

A11, A7 – B7, and A20 – B19. Cysteine bond formation involves a large family of

ER oxidoreductases that act to reduce improper cysteine pairings as well as reoxidize substrates to make proper disulphide bonds37,38. Overall, both A6 –A11 and A19 – B20 disulphide bridges provide interior struts in hydrophobic core of the structure, while A7 – B7 disulphide bond provides an external staple. These bonds are required for the stability and bioactivity of insulin39.

After forming the cysteine bonds in the ER, proinsulin is then transported to the intermediate organelle of the ER and the Golgi, called the preGolgi. The pre-Golgi compartment is the most dynamic structure out of these three secretory organelles. It moves cargo back and forth between ER and Golgi, regulated by the anterograde and retrograde trafficking pathways40. Once proinsulin arrives in the Golgi, which takes about 20 min after synthesis, it is then stabilized by zinc and calcium binding to the B10 histidine residue41,42, forming a proinsulin-zinc-

15 calcium hexamer complex43. In addition, it has also been shown that zinc protects proinsulin from reduction in the Golgi but not in the ER41.

In the Golgi, proinsulin is then concentrated into clathrin-coated immature secretory vesicles. As the vesicles mature, small clathrin-coated vesicles bud off.

In the vesicles, proinsulin is then cleaved into mature insulin and C-peptide by

Prohormone Convertases 1/3, 2 (PC1/3, PC2) and Carboxypeptidase E (CPE)44.

C-peptide is excised at its conserved dibasic sites; the B-chain – C peptide junction (fast site) is cleaved by PC1/3, and the A chain – C peptide junction

(slow site) is cleaved by PC245. Once the amino terminal end of C-peptide is free from B-chain, it allows the protein to position itself better for binding to PC2 active site46,47. The cleavage by prohormone convertases, occurring within 1-2 hours after insulin synthesis, is highly efficient under normal conditions, and is augmented by the pH change from the Golgi region (pH 6.5) to the mature granule (pH 5.0 – 5.5)44. In addition, this transition to pH 5.3 is important for insulin aggregation and crystallization in the mature granule. In the fasting circulation of human adult, only about 5-22% of proinsulin out of the total insulin is found in the plasma, suggesting an extremely efficient conversion of proinsulin to mature insulin under normal circumstances15.

In the mature granule, mature insulin is stored as zinc-stabilized hexamer with 1 calcium and 2 zinc ions, and this granule can then be stored in the cytoplasm for several days with a half-life of ~3 days48,49. These insulin hexamers favour microcrystals formation, which is suggested to protect insulin from further proteolysis50. Upon stimulation, insulin release is triggered. On release into the

16 circulation, the increased in pH disrupts the interaction of the carboxylic acids of the six GluB13 residues that are packed together at the center of insulin hexamer.

Moreover, the dilution of zinc and calcium that are essential to stabilize the hexamer structure also cause the crystal to disintegrate rapidly into monomers of insulin in the blood48. When insulin binds to its receptor in the corresponding cell, it binds in its monomer form.

Stimulation of insulin release by glucose involves a complex regulation and convergence of different signaling pathways (a schematic of this process is depicted in Figure 3). Glucose is taken up by Type 2 Glucose Transporter (Glut2) and then phosphorylated by the rate-limiting . Glucose-6- phosphate is then rapidly metabolized to yield ATP in the mitochondria, leading to an increased in the ATP:ADP ratio51,52. This increased in the ATP:ADP ratio

+ triggers closure of ATP-sensitive K channels (KATP), preventing potassium ions from being transported across the plasma membrane. This rise in positive charge inside the cell causes depolarization, which activates voltage- gated calcium channels, resulting in an influx of calcium ions53-57. In response to this rise in the intracellular calcium levels, granules fuse with the plasma membrane and release of insulin occurs55. The initial rapid exocytosis (lasts ~5-

10 minutes upon stimulation) is referred to as the first phase of insulin secretion.

This is followed by less robust but sustained release of insulin for several hours after stimulation called the second phase secretion58-60. Therefore, glucose- stimulated insulin release is referred to as a biphasic process.

17 Overview: Functional Roles of Sox Family Proteins.

“When we seek to discover the best in others, we somehow bring out the best in ourselves.” – William Arthur Ward

Overview of Sox Family of Transcription Factors.

At some point in development, almost all cells or tissues express at least one Sox transcription factor, a family of transcription factors that share a highly conserved high-mobility group (HMG) DNA–binding domain61,62.The HMG domain performs several functions, including DNA binding, DNA bending, protein-protein interactions, and nuclear import/export. There are approximately

20 Sox transcription factors identified so far and they are divided into 8 subgroups: SoxA–H, based on sequence similarity in their HMG domain61,63-65.

These Sox proteins are involved in various biological and pathological processes, including tissue specification, organ development, stem cell homeostasis, and cancer61,66,67.

Sox proteins bind to the minor groove of DNA, causing a 70 -85 DNA bend. This allows the assembly of multiprotein complexes that regulate transcription or facilitate the recruitment of distant transcriptional activators to target promoters68. All Sox factors can bind to the same DNA sequence, 5’-A/T

A/T CAA A/T G-3’, contributing to their functional redundancy between group members62,68-70. It is known that Sox proteins have low DNA-binding affinity on their own. However, by interacting with other transcription factors, they acquire high affinity and specificity61,68,71-73. Depending on the cellular context and the availability of interacting partners, Sox proteins can act as either transcriptional

18 activators or repressors. Studies have also shown that Sox proteins act in part to remodel local chromatin architecture61,71.

Sox Proteins in Endocrine Pancreas Development and Function.

Transcripts for several Sox subgroups have been identified in the murine and human pancreas (reviewed in 74). In the developing and adult mouse pancreas, mRNA has been detected for Sox subgroup B1 (Sox2), C (Sox4, 11,

12), D (Sox5, 6, 13), E (Sox8, 9, 10), F (Sox7, 17, 18), G (Sox15), and H (Sox30).

In adult mouse islets, several Sox transcripts have been identified including Sox4,

5, 6, 9, 10, 11, 12, 13, and Sox15. However, the Sox protein expression in is less well characterized. Microarray study comparing 8-21 weeks of fetal age versus adult human pancreas showed the expression of Sox4, 9, 10, 11,

Sox12 and Sox17 to be higher in fetal stage than the adult. Sox13, however, showed a lower expression in the fetal human pancreas. In addition, Sox17 was found to be highly expressed in fetal and at lower levels in adult pancreas by

QPCR, and Sox9 was expressed both in human fetal and adult75,76.

Functionally, several Sox factors have been implicated in the maintenance of islet homeostasis (all reviewed in 74). Within Sox subgroup C, Sox4-/- mice die of circulatory failure at day embryonic days 14 (e14) and Sox11-/- mice die at birth with congenital cyanosis. Sox4-deficient mice had abnormal development of the endocrine pancreas and Sox11-deficient mice developed pancreatic hypoplasia77-79. Sox4-deficient explant studies showed reduced endocrine cell

!"

19 differentiation80. This explant study also showed that endocrine cells failed to migrate and organize into islets. In addition, Sox4 haploinsufficient mice and

Sox4-""#$&"$"" and decreased glucose-stimulated insulin secretion (GSIS), respectively81. These studies suggest that Sox4 plays an important role in regulating insulin secretion in response to glucose in mouse. On the contrary, Sox12-deficient mice are normal, suggesting that not all Sox proteins in the pancreas are required for normal development and function77.

For Sox subgroup D, Sox13 was identified to encode islet cell antigen 12

(ICA12), an autoantigen associated with type 1 diabetes82. Another member of

Sox D subgroup, Sox6, was reduced in ob/ob mice and mice high fat diet-

83 induced diabetes . Its HMG domain binds to the NH2 domain of Pdx1, repressing the ability of Pdx1 to activate Insulin1 and 2 genes. In addition,

' ! & #$& " "" *>?@*V? ratio (due to reduced expression of mitochondrial metabolism genes: NADH dehydrogenase complex, ATP synthase, and subunit of Cytochrome c oxidase) and calcium mobilization, leading to reduced insulin secretion. Thus, the attenuation of Sox6 in high fat diet induced diabetic mice may be one of the key events that occurs to govern a compensatory mechanism in increasing the

"!!83.

For Sox subgroup E, Sox9 haploinsufficiency in humans leads to campomelic dysplasia, a skeletal malformation and sex reversal condition.

Interestingly, these patients also had abnormal pancreatic morphology, smaller

20 islets, and variable expression of [ $ \ deficiency (Pdx1-Cre; Sox9flox/flox) caused severe pancreatic hypolasia with reduced pancreatic progenitors. This suggested that Sox9 is involved in maintaining pancreatic progenitors and preventing premature cell differentiation75.

Pancreatic Sox9-haploinsufficient (Pdx1-Cre; Sox9+/flox) mice had 50% reduction in Ngn3-positive cells by birth, and decreased in endocrine cell numbers. While there was no difference in body weight, resting and fasting blood glucose level, fasting blood glucose progressed to diabetic levels as the mice were aged up to

60 weeks84. By 6 weeks of age, total pancreatic insulin content was 70% of littermate controls and the mice started to become glucose intolerant. In addition,

Sox9 haploinsufficient mice were largely unaffected in response to high fat diet; fasting glucose levels and response to glucose challenge were similar to the mice without high fat diet. This milder phenotype in the Sox9-haploinsufficient mice compared to the human Sox9-haploinsufficient patients suggested that mice are better able to compensate for the reduction in gene dosage than those of human.

Lastly, in the Sox subgroup F, Sox17 is the earliest members of the Sox family to play a role in endoderm and pancreas development. Sox17 is required for endoderm formation and maintenance in frog, , and mouse embryos85-88.

Sox17 null mouse embryos have defects in definitive endoderm development; endodermal cells were reduced in size and number in the posterior88. These mice also had elevated apoptosis levels in the prospective foregut endoderm. Sox17 mutants died before e10.5; however, Sox17+/- mice are viable and fertile. In

21 Xenopus ]]^"]- catenin to activate Sox17 endodermal target genes including Gata6, Hnf1 and

Foxa2, which are also important regulators of pancreas development and

!89-95. In addition, Sox17 was shown to be expressed in pool of foregut progenitor cells that gives rise to the ventral pancreatic and biliary primordium95. Gain- and loss-of-function experiments showed that Sox17 acts to promote biliary fate and repress pancreatic . The involvement of

Sox17 in later stages of pancreas development or in adult pancreatic function has not been explored and is the subject of this thesis.

The Roles of Sox17 in Different Biological Contexts.

In addition to its involvement in endoderm organ development88-91,95,

Sox17 is known to be a key regulator in other developmental and disease contexts including the primitive hematopoietic stem cell development, vascular development, lung development, oligodendrocyte development, and colon cancer cell proliferation96-102. Part of the ability of Sox17 to function in such diverse contexts is through its interactions with a diverse array of transcriptional co-

!"-catenin, TCF/LEF and Smad transcription factors.

Studies in the hematopoietic system showed that Sox17 is expressed in both fetal and neonatal hematopoietic stem cells (HSC), and its deficiency in

Sox17GFP/GFP mice led to severe defects in fetal hematopoiesis. This study suggested that Sox17 autonomously acts to maintain fetal but not adult HSCs. It was also suggested that Sox17 may act by promoting Wnt signaling in these

22 fetal/neonatal hematopoietic cells. Furthermore, another study showed that ectopic Sox17 is sufficient to confer fetal HSC properties in adult mouse hematopoietic cells, increasing their self-renewing potential, and their long-term multilineage reconstitution properties103.

During early cardiovascular endothelial development in mice, Sox17 and

Sox18 function redundantly to form the anterior dorsal aorta and head/cervical microvasculature100. Sox17/Sox18-double mutants also had defects in the differentiation of endocardial cells in the developing heart tube. Later in development, it was found that Sox17 is required for proper postnatal angiogenesis in the liver, kidney, and reproductive tract101.

In the lungs, Sox17 was observed to induce plasticity and alter cell fate of the respiratory epithelial cells97. Ectopic expression of Sox17 in embryonic mouse lung epithelial cells disrupted branching morphogenesis and differentiation of proximal/distal epithelial cell types. In the adult lung, conditional expression of Sox17 in alveolar type II cells resulted in ectopic clusters of cells expressing markers of proximal airway epithelial lineages. This proximal reprogramming of alveolar cells was reversible. Sox17 further promoted expression of progenitor cell markers such as Sca1, as well as proliferation of epithelial cells via stimulated expression of cell cycle regulatory genes, cyclin-D1.

In this context, Sox17 was shown to regulate target genes by interacting with

Smad3 to inhibit TGF-_"-3-mediated transcriptional responses that normally act to inhibit proliferation of epithelial cells98.

23 In contrast to the lungs, study showed that Sox17 is important in controlling both oligodendrocyte progenitor cell cycle exit and myelin gene expression and differentiation in the brain96. Specifically, recent study suggested that Sox17 suppressed cyclin D1 expression and cell proliferation by directly

-catenin activity104. These studies of Sox17 showed that functional role of Sox17 is very context-dependent. Therefore, it is important to perform functional studies of Sox17 in many different biological contexts to dissect the many possible molecular roles of Sox17.

In disease contexts, in particular intestinal neoplasia, canonical Wnt signaling that governs cell proliferation is known to be aberrant105-108.One study found that Sox17 is expressed in normal gut epithelium and its expression is reduced in intestinal neoplasia102. Further analysis showed that Sox17 protein

] >` " -catenin and forms a complex to promote the

""!>`"-'{|-independent mechanism.

This suggested that Sox17 can repress canonical Wnt signaling; thus, affecting proliferation of colon carcinoma cells. This is in contrast with the role of Sox4 in colon carcinoma proliferation, where Sox4 functioned to stabilize -catenin protein.

In another disease context, such as the congenital anomalies of the kidney and the urinary tract (CAKUT), mutation in Sox17 had been identified in these human patients109. The mutation is a point mutation at nucleotide 755 to the translation start site in the Sox17 gene. This maternal origin mutation resulted in a protein change from a tyrosine to an asparagine residue located between the

24 HMG-box and the glycine-proline rich segment at the C-terminal of the protein

(referred to as p.Y259N mutant). As measured by the Sox17 and Wnt reporter plasmids, the p.Y259N mutant protein levels were shown to be significantly higher, had increased transcriptional activity, and was able to suppress the

}_-catenin signaling better than wildtype Sox17. This study suggested that elevated p.Y259N mutant protein may be responsible in causing the congenital defect that occurs in CAKUT patients.

Taken together, all these studies combined further highlighted the different biological activities of Sox proteins. They also demonstrated the importance of

Sox interacting proteins in regulating the transcriptional output of Wnt signaling in different developmental and pathological processes (reviewed in 110).

25 References.

1 Di Guglielmo, G. M. et al. Insulin receptor internalization and signalling. Mol Cell Biochem 182, 59-63 (1998). 2 Ogawa, W., Matozaki, T. & Kasuga, M. Role of binding proteins to IRS-1 in insulin signalling. Mol Cell Biochem 182, 13-22 (1998). 3Izumi, T.et al. Dominant negative pathogenesis by mutant proinsulin in the Akita diabetic mouse. Diabetes 52, 409-416 (2003). 4 Hartling, S. G. et al. Longitudinal study of fasting proinsulin in 148 siblings of patients with insulin-dependent diabetes mellitus. Study Group on Childhood Diabetes in Finland. Eur J Endocrinol 137, 490-494 (1997). 5 Roder, M. E., Porte, D., Jr., Schwartz, R. S. & Kahn, S. E. Disproportionately elevated proinsulin levels reflect the degree of impaired B cell secretory capacity in patients with noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 83, 604-608 (1998). 6 Roder, M. E. et al. Disproportionately elevated proinsulin levels precede the onset of insulin-dependent diabetes mellitus in siblings with low first phase insulin responses. The Childhood Diabetes in Finland Study Group. J Clin Endocrinol Metab 79, 1570-1575 (1994). 7 Saad, M. F. et al. Disproportionately elevated proinsulin in Pima Indians with noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 70, 1247-1253 (1990). 8 Kahn, S. E. & Halban, P. A. Release of incompletely processed proinsulin is the cause of the disproportionate proinsulinemia of NIDDM. Diabetes 46, 1725-1732 (1997). 9 Tersey, S. A. et al. Islet -cell endoplasmic reticulum stress precedes the onset of type 1 diabetes in the nonobese diabetic mouse model. Diabetes 61, 818-827, doi:db11-1293 [pii] 10.2337/db11-1293 (2012). 10 Basu, A., Pedersen, M. G. & Cobelli, C. Prediabetes: evaluation of -cell function. Diabetes 61, 270-271, doi:61/2/270 [pii] 10.2337/db11-1677 (2012). 11 Kanat, M. et al. Distinct -cell defects in impaired fasting glucose and impaired glucose tolerance. Diabetes 61, 447-453, doi:db11-0995 [pii] 10.2337/db11-0995 (2012). 12 Stoy, J. et al. Insulin gene mutations as a cause of permanent neonatal diabetes. Proc Natl Acad Sci U S A 104, 15040-15044, doi:0707291104 [pii] 10.1073/pnas.0707291104 (2007). 13 Zuber, C., Fan, J. Y., Guhl, B. & Roth, J. Misfolded proinsulin accumulates in expanded pre-Golgi intermediates and endoplasmic reticulum subdomains in pancreatic cells of Akita mice. FASEB J 18, 917-919, doi:10.1096/fj.03-1210fje 03-1210fje [pii] (2004).

26 14 Anderson, M. S. & Bluestone, J. A. The NOD mouse: a model of immune dysregulation. Annu Rev Immunol 23, 447-485, doi:10.1146/annurev.immunol.23.021704.115643 (2005). 15 Mako, M. E., Starr, J. I. & Rubenstein, A. H. Circulating proinsulin in patients with maturity onset diabetes. Am J Med 63, 865-869 (1977). 16 Hartling, S. G., Lindgren, F., Dahlqvist, G., Persson, B. & Binder, C. Elevated proinsulin in healthy siblings of IDDM patients independent of HLA identity. Diabetes 38, 1271-1274 (1989). 17 Kahn, C. R. Diabetes. Causes of insulin resistance. Nature 373, 384-385, doi:10.1038/373384a0 (1995). 18 Gordon, E. S. Diabetes mellitus. New developments. Postgrad Med 55, 145-149 (1974). 19 Temple, R. C. et al. Insulin deficiency in non-insulin-dependent diabetes. Lancet 1, 293-295 (1989). 20 Marchetti, P. et al. The endoplasmic reticulum in pancreatic cells of type 2 diabetes patients. Diabetologia 50, 2486-2494, doi:10.1007/s00125-007- 0816-8 (2007). 21 Fajans, S. S., Bell, G. I. & Polonsky, K. S. Molecular mechanisms and clinical pathophysiology of maturity-onset diabetes of the young. N Engl J Med 345, 971-980, doi:10.1056/NEJMra002168 (2001). 22 Giuffrida, F. M. & Reis, A. F. Genetic and clinical characteristics of maturity-onset diabetes of the young. Diabetes Obes Metab 7, 318-326, doi:DOM399 [pii] 10.1111/j.1463-1326.2004.00399.x (2005). 23 Servitja, J. M. & Ferrer, J. Transcriptional networks controlling pancreatic development and cell function. Diabetologia 47, 597-613, doi:10.1007/s00125-004-1368-9 (2004). 24 Murphy, R., Ellard, S. & Hattersley, A. T. Clinical implications of a molecular genetic classification of monogenic -cell diabetes. Nat Clin Pract Endocrinol Metab 4, 200-213, doi:ncpendmet0778 [pii] 10.1038/ncpendmet0778 (2008). 25 Lee, C. S. et al. Foxa2 controls Pdx1 gene expression in pancreatic - cells in vivo. Diabetes 51, 2546-2551 (2002). 26 Servitja, J. M. et al. Hnf1alpha (MODY3) controls tissue-specific transcriptional programs and exerts opposed effects on cell growth in pancreatic islets and liver. Mol Cell Biol 29, 2945-2959, doi:MCB.01389-08 [pii] 10.1128/MCB.01389-08 (2009). 27 Welters, H. J. et al. Conditional expression of hepatocyte nuclear factor- 1, the maturity-onset diabetes of the young-5 gene product, influences the viability and functional competence of pancreatic -cells. J Endocrinol 190, 171-181, doi:190/1/171 [pii] 10.1677/joe.1.06768 (2006). 28 Ahlgren, U., Jonsson, J., Jonsson, L., Simu, K. & Edlund, H. -cell- specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the -

27 cell phenotype and maturity onset diabetes. Genes Dev 12, 1763-1768 (1998). 29 Johnson, J. D. et al. Increased islet apoptosis in Pdx1+/- mice. J Clin Invest 111, 1147-1160, doi:10.1172/JCI16537 (2003). 30 Brissova, M. et al. Reduction in pancreatic transcription factor PDX-1 impairs glucose-stimulated insulin secretion. J Biol Chem 277, 11225- 11232, doi:10.1074/jbc.M111272200 M111272200 [pii] (2002). 31 Suckale, J. & Solimena, M. Pancreas islets in metabolic signaling--focus on the -cell. Front Biosci 13, 7156-7171, doi:3218 [pii] (2008). 32 Schuit, F. C., In't Veld, P. A. & Pipeleers, D. G. Glucose stimulates proinsulin biosynthesis by a dose-dependent recruitment of pancreatic cells. Proc Natl Acad Sci U S A 85, 3865-3869 (1988). 33 Dean, P. M. Ultrastructural morphometry of the pancreatic -cell. Diabetologia 9, 115-119 (1973). 34 Straub, S. G., Shanmugam, G. & Sharp, G. W. Stimulation of insulin release by glucose is associated with an increase in the number of docked granules in the -cells of rat pancreatic islets. Diabetes 53, 3179-3183, doi:53/12/3179 [pii] (2004). 35 Howell, S. L. The mechanism of insulin secretion. Diabetologia 26, 319- 327 (1984). 36 Dodson, G. & Steiner, D. The role of assembly in insulin's biosynthesis. Curr Opin Struct Biol 8, 189-194, doi:S0959-440X(98)80037-7 [pii] (1998). 37 Karala, A. R., Lappi, A. K. & Ruddock, L. W. Modulation of an active-site cysteine pKa allows PDI to act as a catalyst of both disulfide bond formation and isomerization. J Mol Biol 396, 883-892, doi:S0022- 2836(09)01519-8 [pii] 10.1016/j.jmb.2009.12.014 (2010). 38 Rajpal, G., Schuiki, I., Liu, M., Volchuk, A. & Arvan, P. Action of protein disulfide isomerase on proinsulin exit from endoplasmic reticulum of pancreatic -cells. J Biol Chem 287, 43-47, doi:C111.279927 [pii] 10.1074/jbc.C111.279927 (2012). 39 Weiss, M. A. The structure and function of insulin: decoding the TR transition. Vitam Horm 80, 33-49, doi:S0083-6729(08)00602-X [pii] 10.1016/S0083-6729(08)00602-X (2009). 40 Fan, J. Y., Roth, J. & Zuber, C. Ultrastructural analysis of transitional endoplasmic reticulum and pre-Golgi intermediates: a highway for cars and trucks. Histochem Cell Biol 120, 455-463, doi:10.1007/s00418-003- 0597-1 (2003). 41 Huang, X. F. & Arvan, P. Intracellular transport of proinsulin in pancreatic -cells. Structural maturation probed by disulfide accessibility. J Biol Chem 270, 20417-20423 (1995). 42 Davidson, H. W., Rhodes, C. J. & Hutton, J. C. Intraorganellar calcium and pH control proinsulin cleavage in the pancreatic cell via two distinct site- specific endopeptidases. Nature 333, 93-96, doi:10.1038/333093a0 (1988).

28 43 Howell, S. L., Tyhurst, M., Duvefelt, H., Andersson, A. & Hellerstrom, C. Role of zinc and calcium in the formation and storage of insulin in the pancreatic -cell. Cell Tissue Res 188, 107-118 (1978). 44 Steiner, D. F., Park, S. Y., Stoy, J., Philipson, L. H. & Bell, G. I. A brief perspective on insulin production. Diabetes Obes Metab 11 Suppl 4, 189- 196, doi:DOM1106 [pii] 10.1111/j.1463-1326.2009.01106.x (2009). 45 Furuta, M. et al. Incomplete processing of proinsulin to insulin accompanied by elevation of Des-31,32 proinsulin intermediates in islets of mice lacking active PC2. J Biol Chem 273, 3431-3437 (1998). 46 Lipkind, G. & Steiner, D. F. Predicted structural alterations in proinsulin during its interactions with prohormone convertases. Biochemistry 38, 890-896, doi:10.1021/bi981556q bi981556q [pii] (1999). 47 Rhodes, C. J., Lincoln, B. & Shoelson, S. E. Preferential cleavage of des- 31,32-proinsulin over intact proinsulin by the insulin secretory granule type II endopeptidase. Implication of a favored route for prohormone processing. J Biol Chem 267, 22719-22727 (1992). 48 Baker, E. N. et al. The structure of 2Zn pig insulin crystals at 1.5 A resolution. Philos Trans R Soc Lond B Biol Sci 319, 369-456 (1988). 49 Marsh, B. J. et al. Regulated autophagy controls hormone content in secretory-deficient pancreatic endocrine -cells. Mol Endocrinol 21, 2255- 2269, doi:me.2007-0077 [pii] 10.1210/me.2007-0077 (2007). 50 Michael, J., Carroll, R., Swift, H. H. & Steiner, D. F. Studies on the molecular organization of rat insulin secretory granules. J Biol Chem 262, 16531-16535 (1987). 51 Kennedy, H. J. et al. Glucose generates sub-plasma membrane ATP microdomains in single islet -cells. Potential role for strategically located mitochondria. J Biol Chem 274, 13281-13291 (1999). 52 Sener, A. & Malaisse, W. J. Stimulation by D-glucose of mitochondrial oxidative events in islet cells. Biochem J 246, 89-95 (1987). 53 Cook, D. L. & Hales, C. N. Intracellular ATP directly blocks K+ channels in pancreatic B-cells. Nature 311, 271-273 (1984). 54 Meglasson, M. D. & Matschinsky, F. M. Pancreatic islet glucose metabolism and regulation of insulin secretion. Diabetes Metab Rev 2, 163-214 (1986). 55 Rorsman, P. et al. The Cell Physiology of Biphasic Insulin Secretion. News Physiol Sci 15, 72-77 (2000). 56 Safayhi, H. et al. L-type calcium channels in insulin-secreting cells: biochemical characterization and phosphorylation in RINm5F cells. Mol Endocrinol 11, 619-629 (1997). 57 Satin, L. S. & Cook, D. L. Voltage-gated Ca2+ current in pancreatic B-cells. Pflugers Arch 404, 385-387 (1985). 58 Curry, D. L., Bennett, L. L. & Grodsky, G. M. Dynamics of insulin secretion by the perfused rat pancreas. Endocrinology 83, 572-584 (1968).

29 59 Henquin, J. C., Nenquin, M., Stiernet, P. & Ahren, B. In vivo and in vitro glucose-induced biphasic insulin secretion in the mouse: pattern and role of cytoplasmic Ca2+ and amplification signals in -cells. Diabetes 55, 441-451, doi:55/2/441 [pii] (2006). 60 Kahn, S. E. et al. Importance of early phase insulin secretion to intravenous glucose tolerance in subjects with type 2 diabetes mellitus. J Clin Endocrinol Metab 86, 5824-5829 (2001). 61 Wilson, M. & Koopman, P. Matching SOX: partner proteins and co-factors of the SOX family of transcriptional regulators. Curr Opin Genet Dev 12, 441-446, doi:S0959437X02003234 [pii] (2002). 62 Pevny, L. H. & Lovell-Badge, R. Sox genes find their feet. Curr Opin Genet Dev 7, 338-344, doi:S0959-437X(97)80147-5 [pii] (1997). 63 Schepers, G. E., Teasdale, R. D. & Koopman, P. Twenty pairs of sox: extent, , and nomenclature of the mouse and human sox transcription factor gene families. Dev Cell 3, 167-170, doi:S153458070200223X [pii] (2002). 64 Bowles, J., Schepers, G. & Koopman, P. Phylogeny of the SOX family of developmental transcription factors based on sequence and structural indicators. Dev Biol 227, 239-255, doi:10.1006/dbio.2000.9883 S0012-1606(00)99883-X [pii] (2000). 65 Lefebvre, V., Dumitriu, B., Penzo-Mendez, A., Han, Y. & Pallavi, B. Control of cell fate and differentiation by Sry-related high-mobility-group box (Sox) transcription factors. Int J Biochem Cell Biol 39, 2195-2214, doi:S1357-2725(07)00175-6 [pii] 10.1016/j.biocel.2007.05.019 (2007). 66 Dong, C., Wilhelm, D. & Koopman, P. Sox genes and cancer. Cytogenet Genome Res 105, 442-447, doi:10.1159/000078217 78217 [pii] (2004). 67 Kiefer, J. C. Back to basics: Sox genes. Dev Dyn 236, 2356-2366, doi:10.1002/dvdy.21218 (2007). 68 Wegner, M. From head to toes: the multiple facets of Sox proteins. Nucleic Acids Res 27, 1409-1420, doi:gkc266 [pii] (1999). 69 Mertin, S., McDowall, S. G. & Harley, V. R. The DNA-binding specificity of SOX9 and other SOX proteins. Nucleic Acids Res 27, 1359-1364, doi:gkc251 [pii] (1999). 70 van Beest, M. et al. Sequence-specific high mobility group box factors recognize 10-12- minor groove motifs. J Biol Chem 275, 27266- 27273, doi:10.1074/jbc.M004102200 M004102200 [pii] (2000). 71 Kamachi, Y., Uchikawa, M. & Kondoh, H. Pairing SOX off: with partners in the regulation of embryonic development. Trends Genet 16, 182-187, doi:S0168-9525(99)01955-1 [pii] (2000). 72 Wiebe, M. S., Nowling, T. K. & Rizzino, A. Identification of novel domains within Sox-2 and Sox-11 involved in autoinhibition of DNA binding and partnership specificity. J Biol Chem 278, 17901-17911, doi:10.1074/jbc.M212211200

30 M212211200 [pii] (2003). 73 Wissmuller, S., Kosian, T., Wolf, M., Finzsch, M. & Wegner, M. The high- mobility-group domain of Sox proteins interacts with DNA-binding domains of many transcription factors. Nucleic Acids Res 34, 1735-1744, doi:34/6/1735 [pii] 10.1093/nar/gkl105 (2006). 74 McDonald, E., Krishnamurthy, M., Goodyer, C. G. & Wang, R. The emerging role of SOX transcription factors in pancreatic endocrine cell development and function. Stem Cells Dev 18, 1379-1388, doi:10.1089/scd.2009.0240 (2009). 75 Seymour, P. A. et al. SOX9 is required for maintenance of the pancreatic progenitor cell pool. Proc Natl Acad Sci U S A 104, 1865-1870, doi:0609217104 [pii] 10.1073/pnas.0609217104 (2007). 76 Piper, K. et al. Novel SOX9 expression during human pancreas development correlates to abnormalities in Campomelic dysplasia. Mech Dev 116, 223-226, doi:S0925477302001454 [pii] (2002). 77 Hoser, M. et al. Sox12 deletion in the mouse reveals nonreciprocal redundancy with the related Sox4 and Sox11 transcription factors. Mol Cell Biol 28, 4675-4687, doi:MCB.00338-08 [pii] 10.1128/MCB.00338-08 (2008). 78 Schilham, M. W. et al. Defects in cardiac outflow tract formation and pro- B-lymphocyte expansion in mice lacking Sox-4. Nature 380, 711-714, doi:10.1038/380711a0 (1996). 79 Sock, E. et al. Gene targeting reveals a widespread role for the high- mobility-group transcription factor Sox11 in tissue remodeling. Mol Cell Biol 24, 6635-6644, doi:10.1128/MCB.24.15.6635-6644.2004 24/15/6635 [pii] (2004). 80 Wilson, M. E. et al. The HMG box transcription factor Sox4 contributes to the development of the endocrine pancreas. Diabetes 54, 3402-3409, doi:54/12/3402 [pii] (2005). 81 Goldsworthy, M. et al. Role of the transcription factor sox4 in insulin secretion and impaired glucose tolerance. Diabetes 57, 2234-2244, doi:db07-0337 [pii] 10.2337/db07-0337 (2008). 82 Kasimiotis, H. et al. Sex-determining region Y-related protein SOX13 is a diabetes autoantigen expressed in pancreatic islets. Diabetes 49, 555-561 (2000). 83 Iguchi, H. et al. SOX6 attenuates glucose-stimulated insulin secretion by repressing PDX1 transcriptional activity and is down-regulated in hyperinsulinemic obese mice. J Biol Chem 280, 37669-37680, doi:M505392200 [pii] 10.1074/jbc.M505392200 (2005). 84 Dubois, C. L. et al. Sox9-haploinsufficiency causes glucose intolerance in mice. PLoS One 6, e23131, doi:10.1371/journal.pone.0023131 PONE-D-11-05019 [pii] (2011).

31 85 Hudson, C., Clements, D., Friday, R. V., Stott, D. & Woodland, H. R. Xsox17alpha and - mediate endoderm formation in Xenopus. Cell 91, 397-405, doi:S0092-8674(00)80423-7 [pii] (1997). 86 Clements, D. & Woodland, H. R. Changes in embryonic cell fate produced by expression of an endodermal transcription factor, Xsox17. Mech Dev 99, 65-70, doi:S0925477300004767 [pii] (2000). 87 Clements, D., Cameleyre, I. & Woodland, H. R. Redundant early and overlapping larval roles of Xsox17 subgroup genes in Xenopus endoderm development. Mech Dev 120, 337-348, doi:S0925477302004501 [pii] (2003). 88 Kanai-Azuma, M. et al. Depletion of definitive gut endoderm in Sox17-null mutant mice. Development 129, 2367-2379 (2002). 89 Sinner, D. et al. Global analysis of the transcriptional network controlling Xenopus endoderm formation. Development 133, 1955-1966, doi:133/10/1955 [pii] 10.1242/dev.02358 (2006). 90 Sinner, D., Rankin, S., Lee, M. & Zorn, A. M. Sox17 and -catenin cooperate to regulate the transcription of endodermal genes. Development 131, 3069-3080, doi:10.1242/dev.01176 dev.01176 [pii] (2004). 91 Zorn, A. M. et al. Regulation of Wnt signaling by Sox proteins: XSox17 alpha/ and XSox3 physically interact with -catenin. Mol Cell 4, 487-498, doi:S1097-2765(00)80200-2 [pii] (1999). 92 Haumaitre, C. et al. Lack of TCF2/vHNF1 in mice leads to pancreas agenesis. Proc Natl Acad Sci U S A 102, 1490-1495, doi:0405776102 [pii] 10.1073/pnas.0405776102 (2005). 93 Edghill, E. L., Bingham, C., Ellard, S. & Hattersley, A. T. Mutations in hepatocyte nuclear factor-1 and their related phenotypes. J Med Genet 43, 84-90, doi:jmg.2005.032854 [pii] 10.1136/jmg.2005.032854 (2006). 94 Gao, N. et al. Foxa2 controls vesicle docking and insulin secretion in mature  cells. Cell Metab 6, 267-279, doi:S1550-4131(07)00260-4 [pii] 10.1016/j.cmet.2007.08.015 (2007). 95 Spence, J. R. et al. Sox17 regulates organ lineage segregation of ventral foregut progenitor cells. Dev Cell 17, 62-74, doi:S1534-5807(09)00214-7 [pii] 10.1016/j.devcel.2009.05.012 (2009). 96 Sohn, J. et al. Identification of Sox17 as a transcription factor that regulates oligodendrocyte development. J Neurosci 26, 9722-9735, doi:26/38/9722 [pii] 10.1523/JNEUROSCI.1716-06.2006 (2006). 97 Park, K. S., Wells, J. M., Zorn, A. M., Wert, S. E. & Whitsett, J. A. Sox17 influences the differentiation of respiratory epithelial cells. Dev Biol 294, 192-202, doi:S0012-1606(06)00141-2 [pii] 10.1016/j.ydbio.2006.02.038 (2006).

32 98 Lange, A. W., Keiser, A. R., Wells, J. M., Zorn, A. M. & Whitsett, J. A. Sox17 promotes cell cycle progression and inhibits TGF- /Smad3 signaling to initiate progenitor cell behavior in the respiratory epithelium. PLoS One 4, e5711, doi:10.1371/journal.pone.0005711 (2009). 99 Kim, I., Saunders, T. L. & Morrison, S. J. Sox17 dependence distinguishes the transcriptional regulation of fetal from adult hematopoietic stem cells. Cell 130, 470-483, doi:S0092-8674(07)00777-5 [pii] 10.1016/j.cell.2007.06.011 (2007). 100 Sakamoto, Y. et al. Redundant roles of Sox17 and Sox18 in early cardiovascular development of mouse embryos. Biochem Biophys Res Commun 360, 539-544, doi:S0006-291X(07)01244-2 [pii] 10.1016/j.bbrc.2007.06.093 (2007). 101 Matsui, T. et al. Redundant roles of Sox17 and Sox18 in postnatal angiogenesis in mice. J Cell Sci 119, 3513-3526, doi:jcs.03081 [pii] 10.1242/jcs.03081 (2006). 102 Sinner, D. et al. Sox17 and Sox4 differentially regulate -catenin/T-cell factor activity and proliferation of colon carcinoma cells. Mol Cell Biol 27, 7802-7815, doi:MCB.02179-06 [pii] 10.1128/MCB.02179-06 (2007). 103 He, S., Kim, I., Lim, M. S. & Morrison, S. J. Sox17 expression confers self- renewal potential and fetal stem cell characteristics upon adult hematopoietic progenitors. Genes Dev 25, 1613-1627, doi:25/15/1613 [pii] 10.1101/gad.2052911 (2011). 104 Chew, L. J. et al. SRY-box containing gene 17 regulates the Wnt/ - catenin signaling pathway in oligodendrocyte progenitor cells. J Neurosci 31, 13921-13935, doi:31/39/13921 [pii] 10.1523/JNEUROSCI.3343-11.2011 (2011). 105 Groden, J. et al. Identification and characterization of the familial adenomatous polyposis coli gene. Cell 66, 589-600, doi:0092- 8674(81)90021-0 [pii] (1991). 106 Morin, P. J. et al. Activation of -catenin-Tcf signaling in colon cancer by mutations in -catenin or APC. Science 275, 1787-1790 (1997). 107 Miyaki, M. et al. Frequent mutation of -catenin and APC genes in primary colorectal tumors from patients with hereditary nonpolyposis colorectal cancer. Cancer Res 59, 4506-4509 (1999). 108 Oving, I. M. & Clevers, H. C. Molecular causes of colon cancer. Eur J Clin Invest 32, 448-457, doi:1004 [pii] (2002). 109 Gimelli, S. et al. Mutations in SOX17 are associated with congenital anomalies of the kidney and the urinary tract. Hum Mutat 31, 1352-1359, doi:10.1002/humu.21378 (2010). 110 Kormish, J. D., Sinner, D. & Zorn, A. M. Interactions between SOX factors and Wnt/-catenin signaling in development and disease. Dev Dyn 239, 56-68, doi:10.1002/dvdy.22046 (2010).

33 Figure Legends.

Figure 1. MODY genes islet transcriptional network.

A network of transcription factors, including several MODY genes, that works together to maintain islet homeostasis. Each transcription factor functionally relates to each other: a solid line showed direct target correlation, a dotted line showed indirect target correlation. The present study in Chapter 3 is focused on the MODY4 gene, Pdx1. All the genes that are associated with Pdx1 in this islet transcriptional network is linked with an orange line.

Figure 2. Insulin biosynthesis in pancreatic cells.

Insulin biosythesis begins with its transcription in the nucleus, followed by its translation in the ribosome. Proinsulin continues to be processed in the secretory organelles, which then trafficked to the plasma membrane where it is ready to be secreted to the plasma in response to glucose.

Figure 3. Glucose sensing and glucose-stimulated insulin release pathway.

A schematic process of glucose sensing and its corresponding insulin release from pancreatic cell.

34 Figure 1.

35 Figure 2.

36 Figure 3.

37 CHAPTER 2

The Role of Sox17 in Insulin Processing and  Cell Secretory Pathway.

Diva Jonatan1, Jason Spence2, Leena Haataja2, Peter Arvan2, Gail Deutsch3, and

James Wells1

1. Division of Developmental Biology, Cincinnati Children’s Hospital Medical

Center, 2. Division of Metabolism, Endocrinology, and Diabetes, University of

Michigan, 3. Seattle Children’s Hospital, Seattle, WA, 98105

38 ABSTRACT

Defect in any stages of insulin regulation, or loss of cell mass, can lead to elevated blood glucose levels, resulting in diabetes. Thus, regulators of these biological processes might be manipulated therapeutically to improve cell function. Sox17 is a key transcriptional regulator and acts, in part, by regulating

! " " † ] ]

! ![ ' , we hypothesized that Sox17

[‡!^in the pancreas and mature cells (Sox17-paLOF) had no observable impact on pancreas development, islet architecture, and glucose regulation. However, Sox17-paLOF mice had higher islet proinsulin protein content, abnormal ER subcellular distribution of proinsulin, and dilated elles. In addition, these mice were more susceptible to high fat diet-induced glucose dysregulation. These phenotypes suggested that Sox17-paLOF mice were prediabetic.

These data suggested that Sox17 regulates insulin processing and/or transit through the secretory system. To investigate these possibilities, we

'"^ resulting in decreased accumulation of proinsulin protein in cells. Over time this resulted in a 4-fold increase in serum proinsulin levels and diabetes. At the transcriptional level, Sox17 regulated expression of genes involved in hormone transport and secretion. Together, these data suggest that Sox17 regulates pathways involved in both the insulin processing and secretion and that perturbation of these pathways results in loss of normal glucose homeostasis.

39 INTRODUCTION.

Secretion of insulin by the pancreatic cells in response to glucose is central for glucose homeostasis, and dysregulation of this process is a hallmark of early stages of diabetes. In healthy cells, insulin is translated as preproinsulin in the ribosome where the amino terminal signal sequence is removed, converting it to proinsulin. Proinsulin is then transported to the ER where three disulphide bridges will form between the cysteine residues.

Proinsulin will then traffic to the pre-Golgi, a dynamic intermediate structure, and finally to the , where it is packaged into mature secretory granules. In the secretory granules, the endoproteases PC1/3, PC2, and

Carboxypeptidase E cleave proinsulin to become mature insulin and C- peptide1,2,3-5. While much is known about the dynamic nature of insulin processing and trafficking through the secretory pathway, less is known about the regulatory molecules that control these processes.

Unlike Type 1 diabetes, which results from autoimmune destruction of cells, Type 2 diabetes is caused by chronic disruption of glucose homeostasis at several levels. Early stages of T2D can be observed in human patients as a prediabetes stage that is defined by having blood glucose levels that are higher than normal but not yet high enough to be diagnosed as diabetic. Prediabetes is associated with a number of other characteristics including impaired fasting blood glucose, impaired glucose tolerance, disproportionately elevated proinsulin:insulin levels, and dilated endoplasmic reticulum (ER)6-11. These data

40 suggest that defective cell function, particularly in the insulin trafficking process,

contributes to the development of prediabetes.

Our understanding of the biology of prediabetes is limited, in large part

due to a lack of animal models that reproducibly develop symptoms of the

disease. One current model for defective insulin trafficking is the Akita mouse,

which harbors a mutation in the Insulin-2 gene that results in improper folding of

the mutant protein. These animals have a progressive deterioration of secretory

organelle structure and function that accompanied the development of

prediabetes into full diabetes. And this organelle dysfunction in individual cell is

thought to be the primary cause of diabetes in the Akita mouse6. In addition to

the prediabetes phase, an elevated ratio of proinsulin:insulin in the plasma is also

a well-recognized feature of type 2 diabetes9-11. It has also been shown that there

is an increased overall size of the ER in type 2 diabetic patients12. Taken

together, these studies show that defects in insulin processing and folding can

disrupt the function of the secretory machinery, resulting in impaired cell

function and failure to maintain glucose homeostasis. Thus, the regulatory

molecules involved in insulin trafficking and secretion may be exploited

therapeutically to prevent or delay the progression of diabetes.

Sox17 is an HMG box transcription factor and a key regulator in various

developmental and disease contexts, including endoderm organ development13-

19, primitive hematopoietic stem cell development, vascular development and

colon cancer cell proliferation20-23. Part of the ability of Sox17 to function in such

diverse contexts is through its interactions with a diverse array of transcriptional

41 co-factors including -catenin, TCF/LEF and Smad transcription factors. During endoderm development Sox17 acts, in part, as a transcriptional regulator of other

"" ! " " † ]

]!![} have been implicated in islet cell development and homeostasis24-26, a role for

Sox17 in the adult cell has not been described.

Here we have used a combination of mouse genetics, metabolic functional assays, high-resolution quantitation of subcellular localization of proinsulin and

""!^ development and physiological function. We identified a novel role for Sox17 in

" in the regulation of insulin trafficking and secretion. While not

ˆ" ! development or maintenance of ", pancreatic loss- and gain-of-function experiments demonstrate that Sox17 plays a role in proinsulin trafficking and secretion, as manipulations of Sox17 levels perturb the ratio of proinsulin to mature insulin in the islet cells and in the plasma.

Changes in Sox17 levels cause altered morphological integrity of the ER, pre-

Golgi, and Golgi secretory organelles in a manner that is strikingly similar to animals with prediabetes. Moreover, when placed on a high fat diet, animals lacking Sox17 in the pancreas develop pronounced glucose dysregulation. At the molecular level, we demonstrate that Sox17 does not regulate any previously identified transcriptional targets, but rather regulates expression of genes involved in hormone transport, secretion, and subcellular localization. In summary, we identify a novel role for Sox17 in the regulation of insulin trafficking,

42 processing and secretion and that perturbations in Sox17 levels result in a prediabetic phenotype.

MATERIALS AND METHODS.

Mice.

All mice used in these studies; Pdx1-Cre, Sox17Fl, Sox17GFP, Ins2-rtTA, tetO-

Sox17; have been previously described and were housed at the Cincinnati

Children’s Hospital Research Foundation mouse facility17,18,20,27,28. All animal procedures were approved under institutional protocols. For loss-of-function experiments, twelve- to sixteen-week-old adult male and female mice were used.

For gain-of-function experiments, sixteen-week-old adult male mice were used.

In order to upregulate Sox17 expression in the gain-of-function system, doxycycline was given in the food and water for 24 hours before analysis, unless otherwise stated. All mice were originally maintained on an outbred background and were then backcrossed to C57BL/6 background.

Immunofluorescence and confocal microscopy analysis.

Tissues were prepared and stained as previously described18. For Z-stack analysis, images were acquired using confocal microscopy (Zeiss LSM 5.10 with

40x dry objective and 63x PlanApo oil NA 1.4 objective at Nyquist limit and Nikon

A1R si with 100x PlanApo oil NA 1.49 objective at Nyquist limit) and compressed

3D Z-stack images were created and analyzed using Bitplane Imaris 7.2 software.

43 See Supplementary Table 1 for a list of primary and secondary antibodies used in these studies.

Islet isolation and total pancreatic insulin content.

Islets were isolated using standard collagenase digestion followed by purification through a Ficoll gradient and islet gravity sedimentation. For total pancreatic insulin content, ten islets were handpicked, washed with Hank’s buffer (Gibco

#14175), and lysed in 10mM Tris-EDTA, 1% Triton-X 100, pH 8.0. Insulin content was measured using Mouse Insulin ELISA kit (Crystal Chem #90080).

Non-fasting and fasting glucose, proinsulin, and insulin assays.

For glucose measurement, glucose was measured using Freestyle Freedom

Blood Glucose Monitoring System. For fasting glucose, mice were fasted overnight or for 4 hours, as indicated. For plasma proinsulin and insulin levels, blood samples were taken either using tail-vein or submandibular bleed. Blood samples were incubated at room temperature for 20 to 30 minutes, followed by centrifugation for 5 minutes at 1,000 rpm. Plasma samples were then taken and centrifuged for 10 minutes at 2,000 rpm to eliminate the platelets. Plasma samples were stored at -800C until ready to be analyzed using ELISA assays

(Mouse Insulin ELISA kit and Mouse Proinsulin ELISA – Alpco #80-PINMS-E01).

44 Western blot analysis of proinsulin and insulin.

Proteins (3 mg/lane by BCA protein assay) were separated on 4-12% NuPAGE

Novex Bis-Tris gels (Invitrogen), electrotransferred to nitrocellulose (Bio-Rad), and immunoblotted with guinea pig anti-insulin (Linco/Millipore) and mouse anti- (Sigma). Horseradish peroxidase-conjugated secondary antibodies were from Jackson ImmunoResearch with proteins visualized by enhanced chemiluminescence (ECL, Millipore).

Glucose tolerance test and insulin tolerance test.

Glucose tolerance test was performed as previously described29. For insulin tolerance test, mice were fasted for 8-12 hours and intraperitoneally injected with recombinant human insulin (1U/kg, Novo Nordisk, Novolin?R NDC 0169-1833-

11). Blood glucose levels were measured at indicated time points.

Real-time PCR and Microarray analysis.

Mouse islets were isolated as described above, and total islet RNA was extracted using either RNeasy Micro Kit (Qiagen, cat. no. 74004) or PureLink RNA Mini Kit

(Invitrogen, cat. no. 12183-018A). RNA samples were then reverse transcribed into cDNA using the SuperScript III First-Strand Synthesis System (Invitrogen).

For Real-time PCR, QuantiTect SYBR Green (Qiagen) was used on BioRad

CFX96. For microarray analysis, RNA was isolated from islets and used to create target DNA for hybridization to Affymetrix Human 1.0 Gene ST Arrays using standard procedures (Affymetrix, Santa Clara, CA). Independent biological

45 triplicates were performed for each condition. Affymetrix microarray Cel files were subjected to RMA normalization in GeneSpring 10.1. Probe sets were first filtered for those that are over expressed or underexpressed and then subjected to statistical analysis for differential expression by 1.3 fold or more between controls and either Sox17-paLOF or Sox17-GOF islets with p<0.05 using the

Students T-test. Log2 gene expression ratios were then subjected to hierarchical clustering using the standard correlation distance metric as implemented in

GeneSpring. Heat map was created using GeneSpring. The differentially expressed genes were subjected to functional enrichment analyses using the ToppCluster server30.

Electron microscopy

Mouse pancreas were dissected and fixed in 3% glutaraldehyde and 0.175 M sodium cacodylate buffer, pH 7.4 at 4 C for one hour. The samples were then post fixed in 1% osmium tetroxide in 0.2 M sodium cacodylate buffer for 1 hour at

4 C, processed thorugh a graded series of , infiltrated, and embedded in

LX-112 resin. After polymerization at 60 C for three days, ultrathin sections (100 nm) were cut using a Leica EM UC7 microtome and counterstained in 2% aqueous uranyl acetate and Reynolds lead citrate. Images were taken with a transmission electron microscope (Hitachi H-6750) equipped with a digital camera (AMT 2k x 2K tem CCD).

46 Analysis of insulin subcellular localization

Surfaces were created using pixel intensity thresholds to identify positively stained proinsulin in the respective organelles and structures from background fluorescence. Imaris software calculated volumes for these structures and organelles using standard algorithm to measure total volume and relative percent colocalization. Images acquired with Zeiss LSM 5.10 were always acquired at

512x512 by 13 optical sections on average. This standard image volume size was used to compare the Sox17 animals to controls. Nikon images were tile- scanned to image the entire islet, and were always acquired at 512x512 by 56 optical sections on average. Due to variable size of the islets, total volume analysis was not included using the Nikon images.

Statistical analysis.

All the data are expressed as mean ± SEM, and Student t-tests were used for statistical analysis.

RESULTS.

Sox17 is not required for cell development.

Previous studies have demonstrated that Sox17 is required for early endoderm formation and in regulating pancreatic/biliary lineage segregation from a common pool of foregut progenitor cells13,18. Moreover, during early endoderm development, Sox17 regulates several key transcription factors and signaling pathways that are also known to play a central role in adult cell homeostasis.

47 We therefore investigated whether Sox17 might have a functional role in the adult pancreas by deleting Sox17 throughout the pancreas (Sox17-paLOF) using either a Pdx1-Cre line and both Sox17fl and Sox17-GFP or using Pdx1-Cre and

Sox17fl only as indicated (Figure 1A)18,20,31. Surprisingly, Sox17 deletion had no effect on overall pancreas development (data not shown) or on gross islet architecture at 12 weeks of age as indicated by normal distribution of insulin- expressing cells at the core and alpha and delta cells at the mantle of the islet

(Figure 1C, 1D). Pdx1 protein expression was also indistinguishable between control and Sox17-paLOF mice (Figure 1E, 1F). All together, these results showed that Sox17-paLOF mice had morphologically normal islets, suggesting that Sox17 is not required for islet development. We confirmed Sox17 deletion in islets isolated from 12-16 week old of Sox17-paLOF animals using quantitative

PCR. We observed a 70% reduction in Sox17 mRNA (Figure 1B, p-'0‹[‹Œ).

The remaining expression is likely due to expression of Sox17 in endothelial cells, which are not targeted by Pdx1-Cre. Quantitative analysis of islet mRNAs demonstrated that loss of Sox17 had no effect on Insulin, Somatostatin,

Glucagon, and Glut2, nor were there any changes in islet transcription factors such as Foxa2, Gata6 and HNF1b, which are known Sox17 targets in Xenopus endoderm14,15,32 (data not shown). As a separate confirmation of the Pdx1-Cre activity in the islets, we crossed Pdx1-Cre mice with a -gal reporter line and observed robust activation of the reporter allele in cells expressing Cre protein

(data not shown). We also observed reduced levels of Sox17 protein in the

Sox17-paLOF animals with 2 Sox17 antibodies (Supplementary Figure 1),

48 however due to cross reactivity of Sox17 antibodies with other Sox proteins in the pancreas, we found that quantitative RT-PCR of isolated islets was the most specific method for showing quantitative loss of Sox17 expression. Sox17-paLOF animals had no changes in ductal or exocrine compartments (data not shown).

Sox17-paLOF results in elevated proinsulin protein in the islets.

We next investigated if Sox17 plays a role in the physiological function of mature cells. At 12-16 weeks of age, Sox17-paLOF mice had normal resting blood glucose and total plasma insulin levels by ELISA that detects all forms of insulin including proinsulin, mature insulin, and C-peptide (Figure 2A, 2B, control mice consisted of Sox17fl/+, Sox17GFP/fl; Pdx1Cre;Sox17fl/+, n=19 for glucose measurement, n=18 for plasma insulin measurement; Sox17-paLOF mice consisted of Pdx1Cre;Sox17GFP/fl and Pdx1Cre;Sox17fl/fl, n=22 for glucose measurement, n=20 for plasma insulin measurement). However, analysis of the ratio of proinsulin to total insulin levels showed a trend of accumulation of proinsulin protein in the plasma of Sox17-paLOF animals compared to controls

(Figure 2C). This could be due to elevated secretion of proinsulin, or due to an accumulation of proinsulin in the islet, or a combination of both. To determine if there was accumulation of proinsulin in islets we analyzed protein extracts from isolated islets by western blot (Figure 2D). While there was no difference in the amount of mature insulin islet protein between genotypes (Figure 2D, quantitation in Figure 2F), Sox17-paLOF islets (Pdx1Cre;Sox17fl/fl, n=5) had significantly elevated proinsulin levels relative to control islets (Pdx1Cre;Sox17fl/+,

49 n=5) (Figure 2D, quantitation in Figure 2E, p-'0‹[‹). In addition to islets containing more proinsulin, we found that the total insulin levels in the plasma tend to be higher in Sox17-paLOF mice in response to glucose challenge

(Figure 2H, control mice: Sox17fl/+, n=3; Sox17-paLOF mice: Pdx1Cre;Sox17GFP/fl, n=4). Despite higher levels of total insulin, Sox17-paLOF mutant mice exhibited normal glucose tolerance profile (control mice: Sox17fl/+, n=4; Sox17-paLOF mice: Pdx1Cre;Sox17GFP/fl, n=4) suggesting that the increase in total insulin was due to secretion of inactive proinsulin (Figure 2G). The increase in total plasma insulin level were not due to an increase in total insulin production, shown in normal insulin islet content, or a decrease in insulin utilization by other cells, shown in normal peripheral insulin sensitivity (Figure 2J, 2I, control mice:

Sox17fl/+, n=2; Sox17-paLOF mice: Pdx1Cre;Sox17GFP/fl, n=4). Taken together, these data suggest that loss of Sox17 results in altered insulin prohormone processing and secretion, which could be due to dysregulation of insulin trafficking through the secretory pathway or defects in insulin processing machinery.

Loss of Sox17 in the pancreas causes accumulation of proinsulin in the ER and structural changes in secretory organelles.

We investigated if loss of Sox17 results in transcriptional changes in factors that regulates insulin prohormone processing and found no difference in the mRNA levels of PC1, PC2, and CPE in Sox17-paLOF islets (data not shown).

Since Sox17 did not appear to regulate the insulin processing machinery, we

50 investigated whether insulin trafficking through the ER, ER-Golgi intermediate

(pre-Golgi), and Golgi was impaired by quantifying proinsulin subcellular localization within these secretory pathway using the Imaris Software (control:

Sox17fl/+ and Sox17GFP/fl, n=7-8 mice, Sox17-paLOF: Pdx1Cre;Sox17GFP/fl,, n=7, up to 6-10 islets analyzed per mouse). Imaris software is an image analysis program that allows colocalization quantitation between two fluorescence channels using compressed confocal Z-stack images. We found that there was no difference in the accumulation of proinsulin in both the pre-Golgi (ERGIC) and

Golgi (GM130) compartments across genotypes (Figure 3H-3S, quantitation in

Supplementary Figure 2B-E). However, there was a higher percentage of proinsulin in the ER (KDELR) of Sox17-paLOF animals (Control - 7.26%, Sox17- paLOF – 13.795%, p-'0‹[‹); there was more proinsulin resided in the ER.

Correlated to this, there was also more ER region that contained proinsulin

(Control - 13.749%, Sox17-paLOF – 24.44%, p-' 0 ‹[‹) (Figure 3A-3G).

There was no change in the total ER area volume between these two genotypes, quantitated by Imaris software (Supplemental Figure 2A). These findings suggested that the proinsulin trafficking is abnormal in Sox17-paLOF mice, with proinsulin accumulating in the ER. Previous reports have demonstrated that defects in proinsulin trafficking through the secretory organelles resulted in altered insulin processing and secretion6,33,34.

In the Akita mouse model, proinsulin trafficking defects correlated with secretory organelles becoming distended and dilated during the prediabetes phase6,33. Given that Sox17-paLOF mice had similar proinsulin trafficking defects,

51 we examined the ultrastructural appearance of the secretory organelles in the cells of Sox17-paLOF using electron microscopy. We observed severely distended ER, pre-Golgi, and Golgi in Sox17-paLOF (Pdx1Cre;Sox17fl/fl , n=3) cells compared to Pdx1Cre;Sox17fl/+ (n=3) littermates (Figure 3T-3W). This further suggested that these mice were at their prediabetic stage. These marked organelle distentions were specific to the cells, whereas the alpha and delta cells had normal, well-organized secretory organelles (data not shown).

Sox17-paLOF mice are prediabetic and prone to high fat diet-induced stress of cells.

Previous studies showed that an elevated of proinsulin to insulin ratio in patients is an indicator of prediabetes for both type 2 diabetes and maturity onset diabetes7-9,11,35,36. Our data suggested that Sox17-paLOF mice are prediabetic, with a characteristically elevated proinsulin:insulin ratio. To determine if loss of

Sox17 has an impact in diabetes progression, we subjected the mice to high-fat diet-induced diabetes, which is a model of type 2 diabetes in humans. We fed

Sox17-paLOF male (Pdx1Cre;Sox17fl/fl, n=4-6) mice and their Pdx1Cre;Sox17fl/+

(n=3) littermates a high-fat diet starting at 6 months of age and followed them for

6 months. We observed that mice of all genotypes had a similar weight gain in response to a high-fat diet (Figure 4A) and there was a similar amount of food intake and insulin peripheral sensitivity between genotypes (Figure 4D, 4E).

However, Sox17-paLOF mice had higher blood glucose levels after a 4-hour fast

(Figure 4B), and when these animals were glucose challenged, Sox17-paLOF

52 mice were unable to restore normoglycemia as quickly as control animals (Figure

4C). Consistent with these data, we also observed a trend of reduction in their 4- hour fasting plasma proinsulin and total insulin levels; however, the ratio of proinsulin:total insulin remained close to control levels (Supplementary Figure 3).

The development of diabetes in Sox17-paLOF mice in response to a high fat diet is consistent with the fact that these animals were phenotypically prediabetic and that loss of Sox17 made these mice more prone to diet-induced cells stress.

This glucose homeostasis defect is not due to the cell peripheral sensitivity response to insulin, owning that the defect is specific to cell function impairment.

Sox17 overexpression for 24 hours is sufficient to alter proinsulin:total insulin protein ratio and proinsulin secretion in vivo.

Since Sox17-paLOF animals lacked Sox17 during pancreas development it is possible that Sox17 was not having a direct regulatory role in proinsulin trafficking and secretion. To determine the immediate impact of altered Sox17 levels on these cellular processes, we utilized a tetracycline-inducible approach to transiently overexpress Sox17 in mature cells in vivo. We overexpressed

Sox17 for 24 hours using an Ins2-driven reverse tetracycline transactivator transgenic line (Ins2-rtTA) to drive expression of a tetracycline regulated Sox17 transgene (Ins2-rtTA = control, Ins2-rtTA;tetO-Sox17 = Sox17-GOF)28,17,18

(Figure 5A). Sixteen-week old mice were administered doxycycline, resulting in robust Sox17 transgene expression in the islets within 24 hours, as measured by

53 mRNA and protein levels (Figure 5B-D). At the mRNA level there were no significant changes in insulin or insulin processing enzymes (Supplemental

Figures 4, 7B). In addition, gross islet architecture is unchanged in these animals, shown in the pattern of insulin, glucagon, Pdx1, and Nkx2.2 immunostainings

(Figure 5E-J). However, 24 hours of Sox17 overexpression caused a significant reduction in proinsulin protein level in the islets with no significant difference in mature insulin level (Figure 5K, quantitation in Figure 5L and 5M). This is opposite to the accumulation of proinsulin that we observed in the Sox17-paLOF mice. These data suggest that a 24-hour pulse of Sox17 overexpression in adult

cells was sufficient to directly impact insulin trafficking through the secretory pathway in vivo, resulting in reduced levels of islet proinsulin levels.

Overexpression of Sox17 in adult cells causes accumulation of proinsulin in the serum and results in diabetes

To determine the long-term impact of Sox17 on the secretory pathway, we analyzed the effects of prolonged Sox17 overexpression on insulin secretion and glucose homeostasis. Over the course of 7 days, total insulin level tended to be lower in the serum and similar range of plasma proinsulin was found in Sox17-

GOF mice compared to controls (Supplementary Figure 5). By 2-3 weeks, animals were diabetic with a 4-fold increase in the proinsulin to total insulin ratio

(Figure 5N, 5O). The effects of Sox17 expression are reversible such that within

25 days of inactivating the Sox17 transgene by removing doxycycline, animals returned their blood glucose levels to control level (Figure 5P). After 41 days of

54 Sox17 overexpression, the islets had reduced insulin and Pdx1 protein expressions (Figure 5Q,R,U,V and 5S,W). Islet architecture remained intact in these mice as the glucagon and somatostatin expressing cells remain localized at the mantle of the islets (Figure 5Q,R,U,V). However, we observed altered cell shape morphology in the islet seen in the E-cadherin immunostaining (Figure

5S,T,W,X). Some cells appeared to adopt a more elongated shape compare to the control spherical cell shape. Similarly to the blood glucose level, islet protein expressions and cell shape returned back to normal when dox was removed (Figure 5AA-AB), suggesting that the effect is specific to Sox17 misexpression level in these mice. Together, these data suggested that Sox17 is regulating the secretory machinery causing an increase in secretion of improperly processed insulin. Overtime, this leads to significant increased of proinsulin plasma level and reduction of insulin, Pdx1, and Glut2 (Figure 5O, 5Q-T vs 5U-X, and Supplementary Figure 5, data not shown) protein levels in the islet, which together caused diabetes in these mice. This suggested that prolonged Sox17 overexpression caused insulin trafficking to continue worsen and impacted more

cell markers that are critical to maintain proper insulin secretion.

24 hours Sox17 overexpression alters proinsulin trafficking through the secretory organelle machinery.

Since the reduction in islet proinsulin is not due to a change in processing enzymes, we investigated if proinsulin trafficking through the secretory system was affected. Proper transit of proinsulin through the ER-pregolgi-golgi network

55 and assembly into secretory vesicles is important for regulated secretion of mature insulin. It is therefore possible that overexpression of Sox17 might alter proinsulin trafficking along the secretory pathway, resulting in perturbation of proinsulin conversion to insulin. We quantified proinsulin levels in each organelle using different organelle markers (ER (KDELR), pre-Golgi (ERGIC), and Golgi

(GM130) markers) with Imaris software. There was no difference in proinsulin subcellular localization in the ER and Golgi regions (quantitation shown in

Supplementary Figure 6). However, Sox17-GOF cell showed decreased proinsulin percentage colocalization in the pre-Golgi region (Control – 17.408,

Sox17-GOF – 8.993 (Figure 6A-F, quantitation in 6G, p-value 0‹[‹), suggesting that the proinsulin trafficking through the pregolgi was abnormal. Consistent to this, total pre-Golgi area was also found to be decreased in these mice

(Supplementary Figure 6C, asterisk shows p-value 00.01). As described above, there was no change in insulin transcription and overall insulin protein levels remained unchanged between Sox17 overexpressing animals at 24 hours and control groups (Supplementary Figure 7B, Figure 5E,H). EM image analysis was performed to assess the organelle morphology in high resolution. Data showed dilated ER and occasionally dilated mitochondria 24 hours after Sox17 overexpression (Figure 6H-6K), indicating that the cell secretory machinery is impaired immediately after misexpression of Sox17. These data, in combination with elevated proinsulin levels in the serum suggest that Sox17 is stimulating transit of insulin through the secretory pathway, resulting in inappropriate secretion of unprocessed proinsulin.

56 Data showed that prolonged Sox17 overexpression resulted in diabetes with a 4-fold increase in plasma proinsulin:total insulin ratio (Figure 5O). At this later time point, we observed significant quantitative reductions in proinsulin colocalization in the ER and Golgi regions (Figure 6L-O, 6T-W, quantitation not shown) and an overall total decreased in the ER, preGolgi, and Golgi total volume (Figure 6L-W, measured qualitatively). We did not observe any significant difference in proinsulin colocalization in the pre-Golgi region at later stages of

Sox17 overexpression (Figure 6P-S). This suggests that the initial pulse of Sox17 for 24 hours induced a burst of proinsulin trafficking and that continued Sox17- mediated stimulation of insulin trafficking resulted in secretion of improperly processed proinsulin and led to a reduction in processed insulin levels in the plasma (Figure 5O, Supplementary Figure 5). As described above, our data also showed that persistent Sox17 overexpression caused a significant reduction of islet insulin, Pdx1, Glut2, and also a change in the cell membrane morphology

(Figure 5Q-AB and data not shown). This suggests that long-term Sox17 expression is not only affecting insulin secretion, but other cellular pathways as well. Taken together, these gain- and loss-of-function data suggest a new biological role for Sox17 in regulating the secretory pathway in cells.

Sox17 transcriptome: Sox17 regulates pathways involved in insulin transport, secretion.

In order to identify the molecular basis by which Sox17 regulates the secretory pathway in cells, we performed a transcriptional microarray analysis

57 on isolated islets following a 24-hour pulse of Sox17 expression (Figure 7A). We found 1844 transcripts in islets that were altered by 1.3 fold or more in response to Sox17 expression, 972 genes were reduced and 872 genes were elevated (a complete list is shown in Supplementary Table 2). Transcripts were classified into gene ontology categories that represent significantly altered cell-biological processes using ToppCluster software (p<0.05) with no correction (Figure 7A).

Some of the enriched biological processes in the Sox17-regulated genes included in peptide/insulin transport, localization and secretion, consistent with a role of Sox17 in regulating insulin trafficking and secretion (Figure 7A). Other enriched processes include cell development and cell morphogenesis (see

Figure 7A for a complete list of enriched GO terms). Data showed validation of altered genes in Supplementary Figure 7A–U. These data suggest that Sox17 plays a significant role in modulating cell susceptibility to changes in insulin subcelullar trafficking.

Among the genes that were regulated by Sox17, there were several with known functions in insulin trafficking and secretion, summarized in a model depicted in Figure 7B. One category of genes was those involved in ER-Golgi transport, such as Reticulon (increased), Use1 (increased), Serca2/ATP2A2,

COG1, and GolginB1 (decreased). These altered genes may affect insulin trafficking through these secretory organelles. In addition, recent studies have shown that the insulin granule is not only acting as a container, it also acts as a signaling node where many regulatory pathways intersect37. Sox17 expression may impact several factors that control insulin secretion during fasting and

58 feeding periods. Genes that were upregulated in this class included the transmembrane receptors Ephrin (Eph)A5 and A7, which repress insulin secretion in low glucose38. Genes that were decreased in response to

Sox17 expression included GLP1R and Glut2, both involved in nutrient- stimulated insulin secretion, as well as the small G-protein Rab27a. These changes in molecules that regulate basal and nutrient stimulated insulin secretion could explain the elevated glucose levels in Sox17-GOF mice overtime. Lastly, we observed changes in cell transcription factors including Pdx1, NeuroD1,

Foxo1, all reduced between 1.3 to 1.4 fold and Sox6 was increased by ~ 1.6 fold.

While this is a relatively modest change in expression, these factors have central roles in cell homeostasis and changes in the levels of these factors could have a significant combined impact.

Taken together, these results suggested that Sox17 expression had the largest impact on biological processes involved in insulin trafficking and secretion, as well as genes involved in nutrient-regulated insulin secretion. These findings are entirely consistent with our loss- and gain-of-function data showing that

Sox17 plays a central role in insulin trafficking and secretion.

DISCUSSION.

Collectively, our results suggest a new role for the transcription factor

Sox17 in regulating insulin trafficking through the secretory pathway, a dynamic and highly regulated process. Pancreatic loss of Sox17 resulted in accumulation of proinsulin in the ER, dilated and distended secretory organelles, and a trend of

59 increased in secretion of proinsulin all of which are hallmarks of prediabetes6-11.

Consistent with this conclusion, Sox17-paLOF mice went on to develop additional symptoms of type 2 diabetes when placed on a high fat diet, including an impaired ability to maintain normal blood glucose levels in response to fasting and a glucose challenge. Our gain-of-function data suggests that Sox17 directly regulates insulin trafficking since Sox17 expression in the mature cells caused reduced proinsulin levels in the pregolgi region within 24 hours and over time caused a 4-fold increased in unprocessed proinsulin found in the serum. Our microarray data suggests a direct role for Sox17 in the regulation of proinsulin trafficking and secretion in vivo by regulating key genes involved in hormone transport, secretion, and cellular localization.

Disruption in insulin trafficking process is associated with prediabetes, which in part is characterized by elevated proinsulin:insulin ratio levels and dilated ER morphology6-9. Thus, the identification of signals and pathways that govern this early phase of diabetes is of great clinical interest. Such knowledge could be used to target therapeutic intervention to potentially prevent or delay the diabetes progression in patients. Our study suggests that Sox17 regulates a number of biological processes in the cell including protein transport and hormone secretion and that dysregulation of Sox17 can lead to a prediabetic phenotype. Therefore, the mechanisms by which Sox17 regulates these processes might shed light on the pathology of prediabetes. For example, Sox17 expression caused a 4-fold increase in Reticulon, which may explain the faster trafficking of proinsulin from the ER to Golgi. Reticulon is known as curvature-

60 stabilizing protein that is localized in the ER exit sites (ERES), important to form and maintain tubular ER39. This ER morphology and ERES localization are known to affect the organization of the ER to Golgi transport system40.

Overexpression of Reticulon in PC12 cells was also found to significantly enhance human growth hormone secretion41. Additionally we also found an increased in the Use1 SNARE gene (increased by 1. 37 fold), a part of the quarternary SNARE complex that interacts with Golgi-to-ER retrograde COPI vesicle to function in retrograde transport from the Golgi to the preGolgi or the

ER42. It is known that Reticulon interacts with several SNARE proteins involved in vesicle exocytosis, including syntaxin13, 7, 1, and Vamp239-41,43. Therefore, the

Sox17-mediated increase in both Rtn1 and SNARE proteins may lead to precocious anterograde trafficking and an increase in overall transit through the

ER and Golgi, possibly causing an increase in secretion of improperly processed proinsulin.

Sox17 might also directly impact secretion through regulation of secretory granule components. For example Rab27a was decreased by 1.43 fold change in the Sox17-GOF islets. Rab27a is a member of the GTPase protein family that has been associated with insulin-containing granules and is known to regulate secretion at several levels depending on the downstream effectors being regulated44,45. Mice with Rab27a mutation (ashen mice) had an impaired glucose-stimulated insulin secretion46. However, the different Rab27a effectors,

MyRIP, Granuphillin, and Coronin 3, all appear to have unique functions in the secretory pathway. Reduced expression MyRIP in INS-1 cells resulted in

61 decreased insulin release in response to secretagogues47; however, granuphilin null mice showed enhanced insulin release, improved glucose tolerance, and yet had reduced docking granules48,49. To add to the complexity, it was shown that

GDP-Rab27a binds to coronin 3 to regulate the endocytosis retrograde transport of the secretory granule membrane from the plasma membrane50,51. Because

Rab27a regulates several different levels of granule transport process, its deficiency affected long lasting glucose induced insulin secretion46. The reduced

Rab27a in the Sox17-GOF mice might disrupt the balance of granule, movement, localization and secretion.

There were also subtle changes in known cell transcriptional regulators including Pdx1, NeuroD1, Sox6. Pdx1 and NeuroD1 have been previously linked to congenital forms of diabetes in human maturity-onset diabetes of the young,

MODY4 and MODY6, respectively52-56. In mice, complete loss of Pdx1 resulted in pancreas agenesis, whereas Pdx1 heterozygotes and cell specific deletion of

Pdx1 are glucose intolerant, showing impaired glucose stimulated insulin secretion52,57,58. NeuroD1 null mice developed severe hyperglycemia and resulted in early neonatal death59,60. In mice with specific cell deletion of

NeuroD1, glucose stimulated insulin secretion was reduced, resulting in glucose intolerant mice60. Pdx1, NeuroD1, Glut2, GIPR and GLP1R genes were reduced in the Sox17-GOF islet, suggesting a role for Sox17 in regulating nutrient stimulated insulin secretion. The hyperglycemia observed in the obese Sox17- paLOF mice in response to fasting and glucose challenge is consistent with this conclusion. While there were a lot of genes that were downregulated by Sox17

62 overexpression and knowing that there is no evidence of Sox17 acting as a repressor, we did observe an increase in the expression of Pdx1 co-repressor,

Sox6. Previous study in MIN6 cells showed that Sox6 decreased Pdx1 stimulation through changes in chromatin structure on the insulin promoter26.

Sox6 is also known to be downregulated in hyperinsulinemic obese mice and by acting as corepressor of Pdx1, it is known to reduce glucose-stimulated insulin secretion. We also observed a 1.4-fold reduction in Foxo1, which is required for the maintenance of insulin secretion61. These studies and our study further suggested that there is a complex cell network of transcription factors that work together to govern insulin secretion process, and a change in one of the important nodes in the network can dramatically change the effects of the other transcription factor.

We also observed changes in Glucose Transporter 2 (Glut2) and in components of the UPR pathway in the ER, PERK and WFS1, which were identified to be the cause of monogenic permanent neonatal diabetes62. Glut2 is involved in glucose sensing pathway, working side by side with glucokinase. The

2-fold reduction of Glut2 found 24 hours after Sox17 overexpression may not affect the ability of the cell to sense glucose yet because Glut2 is not the rate limiting step of this pathway, glucokinase is, and there is no difference in the level of glucokinase seen in these Sox17-GOF islets (data not shown). In terms of the

ER UPR pathway, PERK, WFS1, IRE1, and Atf4 are known to be involved and previous studies showed that misregulation of these genes may result in profound impact in how the cell responds to misfolded protein64-67,34,63. This is

63 particularly of interest because recent evidence shows that ER stress plays a significant role in the pathogenesis of type 1 and 2 diabetes64-67. Although mRNA data cannot be equated to protein levels, these results nevertheless argue that the reduced gene expressions involved in the UPR pathway overall may play a role in the proinsulin processing defects. However, we still need to clarify in future studies how Sox17 affect the protein levels of these genes due to many known translational modifications that occur in this pathway.

We conclude that Sox17 affects several regulatory pathways that converge to coordinate several cellular processes to achieve a balance between the proinsulin processing trafficking through the secretory machinery and the exocytosis of the insulin granules to the plasma in response to metabolic demands.

Since the discovery of the initial Sox transcription factor, SRY, in 1990, 20

Sox genes have been identified and classified into 8 subgroups based on sequence similarity and genomic organization. Studies have shown that there are several Sox proteins expressed in the murine pancreas, including Sox4, 5, 6, 9,

10, 11, 12, 13, and Sox1524-26. Moreover, Sox17 is expressed at high levels in the human fetal pancreas development and at lower levels in adult islets68. While a role for Sox17 in the developing embryo is largely known, we identified a new role for Sox17 in the mouse postnatal pancreas. During endoderm development,

Sox17 acts, in part, by transcriptionally regulating other important endodermal transcription factors including " † ] ]

! ![ ]' data indicated that these

64 factors are not Sox17 targets in mature cells (data not shown), suggesting that

Sox17 has a unique set of targets in this context. This could be due to the presence and/or abundance of Sox17 co-factors in adult islets versus endoderm.

Sox17 h]]-catenin, TCF/LEF factors, and Smad factors, as in the case of other Sox proteins that also have been shown to have a wide range of interacting co-factors which impact their functions14,15,19,23,69.

Given the requirement of Sox17 in endoderm and extrahepatobiliary cell fate specification, we expected that Sox17 may play a role in pancreatic development. However, taking into account the many Sox family members that are expressed in the developing and adult murine pancreas it is possible that other Sox proteins in the adult islets compensate for the loss of Sox17 in our

Sox17-paLOF mice21,22,25,68. For example, Sox4 is the most highly expressed in the islets; it is known to be important for endocrine differentiation, islet organization24, and for mediating insulin secretion in response to glucose70. In addition, Sox9 is also known to regulate Pdx1 expression and glucose-stimulated insulin release71. Taken together, previous studies showed how other Sox proteins in the adult mouse pancreas may compensate for the Sox17 loss in the islets. Further analysis of the phenotypes of pancreas-specific Sox17/Sox4/Sox9 null mutants, for example, will be required to directly determine the exact contribution of each Sox gene to proinsulin trafficking process.

In summary, our findings support a model in which Sox17 controls multiple aspects of insulin trafficking in mature cells through transcriptional regulation of genes involved in the maintenance of secretory machinery and insulin secretion.

65 Our results suggested that in addition to controlling endoderm formation and pancreatic/biliary lineage segregation, Sox17 plays a key role in regulating homeostasis maintenance of adult pancreatic cells in prediabetes context. This model could provide a better insight into the causes of prediabetic phase, and will inform effort to help design a preventive therapeutical strategy that can be useful for human patients.

Acknowledgements

We thank James Wells, Jeff Whitsett, Aaron Zorn, Jonathan Katz, Gail Deutsch,

Jason Spence, Peter Arvan, Leena Haataja, Matt Kofron, Anil Jegga,

Georgianne Ciraolo, Debora Sinner, Alex Lange, Randy Seeley, Israel

Hodish, Microarray Core, Suh-Chin Lin, Kyle McCracken, and other Wells

and Zorn lab members for reagents and input on the work. We also thank

Sean Morrison (University of Michigan) for the Sox17GFP mice and Andrew

Lowy for the Pdx1-Cre mice.

Sources of Funding

This work was supported by JDRF 2-2003-530 and Beta Cell Consortium-BCBC

(Grant #U01DK072473).

66 References.

1 Hou, J. C., Min, L. & Pessin, J. E. Insulin granule biogenesis, trafficking and exocytosis. Vitam Horm 80, 473-506, doi:S0083-6729(08)00616-X [pii] 10.1016/S0083-6729(08)00616-X (2009). 2 MacDonald, P. E., Joseph, J. W. & Rorsman, P. Glucose-sensing mechanisms in pancreatic -cells. Philos Trans R Soc Lond B Biol Sci 360, 2211-2225, doi:LXG4LT730R156775 [pii] 10.1098/rstb.2005.1762 (2005). 3 Arvan, P. Secretory protein trafficking: genetic and biochemical analysis. Cell Biochem Biophys 40, 169-178, doi:CBB:40:3S:169 [pii] (2004). 4 Arvan, P. & Halban, P. A. Sorting ourselves out: seeking consensus on trafficking in the -cell. Traffic 5, 53-61, doi:152 [pii] (2004). 5Orci, L.et al. Nonconverted, amino acid analog-modified proinsulin stays in a Golgi-derived clathrin-coated membrane compartment. J Cell Biol 99, 2187-2192 (1984). 6Izumi, T.et al. Dominant negative pathogenesis by mutant proinsulin in the Akita diabetic mouse. Diabetes 52, 409-416 (2003). 7 Hartling, S. G. et al. Longitudinal study of fasting proinsulin in 148 siblings of patients with insulin-dependent diabetes mellitus. Study Group on Childhood Diabetes in Finland. Eur J Endocrinol 137, 490-494 (1997). 8 Roder, M. E. et al. Disproportionately elevated proinsulin levels precede the onset of insulin-dependent diabetes mellitus in siblings with low first phase insulin responses. The Childhood Diabetes in Finland Study Group. J Clin Endocrinol Metab 79, 1570-1575 (1994). 9 Roder, M. E., Porte, D., Jr., Schwartz, R. S. & Kahn, S. E. Disproportionately elevated proinsulin levels reflect the degree of impaired B cell secretory capacity in patients with noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 83, 604-608 (1998). 10 Saad, M. F. et al. Disproportionately elevated proinsulin in Pima Indians with noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 70, 1247-1253 (1990). 11 Kahn, S. E. & Halban, P. A. Release of incompletely processed proinsulin is the cause of the disproportionate proinsulinemia of NIDDM. Diabetes 46, 1725-1732 (1997). 12 Marchetti, P. et al. The endoplasmic reticulum in pancreatic cells of type 2 diabetes patients. Diabetologia 50, 2486-2494, doi:10.1007/s00125-007- 0816-8 (2007). 13 Kanai-Azuma, M. et al. Depletion of definitive gut endoderm in Sox17-null mutant mice. Development 129, 2367-2379 (2002). 14 Sinner, D. et al. Global analysis of the transcriptional network controlling Xenopus endoderm formation. Development 133, 1955-1966, doi:133/10/1955 [pii] 10.1242/dev.02358 (2006).

67 15 Sinner, D., Rankin, S., Lee, M. & Zorn, A. M. Sox17 and -catenin cooperate to regulate the transcription of endodermal genes. Development 131, 3069-3080, doi:10.1242/dev.01176 dev.01176 [pii] (2004). 16 Zorn, A. M. et al. Regulation of Wnt signaling by Sox proteins: XSox17 alpha/ and XSox3 physically interact with -catenin. Mol Cell 4, 487-498, doi:S1097-2765(00)80200-2 [pii] (1999). 17 Park, K. S., Wells, J. M., Zorn, A. M., Wert, S. E. & Whitsett, J. A. Sox17 influences the differentiation of respiratory epithelial cells. Dev Biol 294, 192-202, doi:S0012-1606(06)00141-2 [pii] 10.1016/j.ydbio.2006.02.038 (2006). 18 Spence, J. R. et al. Sox17 regulates organ lineage segregation of ventral foregut progenitor cells. Dev Cell 17, 62-74, doi:S1534-5807(09)00214-7 [pii] 10.1016/j.devcel.2009.05.012 (2009). 19 Lange, A. W., Keiser, A. R., Wells, J. M., Zorn, A. M. & Whitsett, J. A. Sox17 promotes cell cycle progression and inhibits TGF- /Smad3 signaling to initiate progenitor cell behavior in the respiratory epithelium. PLoS One 4, e5711, doi:10.1371/journal.pone.0005711 (2009). 20 Kim, I., Saunders, T. L. & Morrison, S. J. Sox17 dependence distinguishes the transcriptional regulation of fetal from adult hematopoietic stem cells. Cell 130, 470-483, doi:S0092-8674(07)00777-5 [pii] 10.1016/j.cell.2007.06.011 (2007). 21 Sakamoto, Y. et al. Redundant roles of Sox17 and Sox18 in early cardiovascular development of mouse embryos. Biochem Biophys Res Commun 360, 539-544, doi:S0006-291X(07)01244-2 [pii] 10.1016/j.bbrc.2007.06.093 (2007). 22 Matsui, T. et al. Redundant roles of Sox17 and Sox18 in postnatal angiogenesis in mice. J Cell Sci 119, 3513-3526, doi:jcs.03081 [pii] 10.1242/jcs.03081 (2006). 23 Sinner, D. et al. Sox17 and Sox4 differentially regulate -catenin/T-cell factor activity and proliferation of colon carcinoma cells. Mol Cell Biol 27, 7802-7815, doi:MCB.02179-06 [pii] 10.1128/MCB.02179-06 (2007). 24 Wilson, M. E. et al. The HMG box transcription factor Sox4 contributes to the development of the endocrine pancreas. Diabetes 54, 3402-3409, doi:54/12/3402 [pii] (2005). 25 Lioubinski, O., Muller, M., Wegner, M. & Sander, M. Expression of Sox transcription factors in the developing mouse pancreas. Dev Dyn 227, 402-408, doi:10.1002/dvdy.10311 (2003). 26 Iguchi, H. et al. SOX6 attenuates glucose-stimulated insulin secretion by repressing PDX1 transcriptional activity and is down-regulated in hyperinsulinemic obese mice. J Biol Chem 280, 37669-37680, doi:M505392200 [pii] 10.1074/jbc.M505392200 (2005).

68 27 Milo-Landesman, D. et al. Correction of hyperglycemia in diabetic mice transplanted with reversibly immortalized pancreatic cells controlled by the tet-on regulatory system. Cell Transplant 10, 645-650 (2001). 28 Nir, T., Melton, D. A. & Dor, Y. Recovery from diabetes in mice by cell regeneration. J Clin Invest 117, 2553-2561, doi:10.1172/JCI32959 (2007). 29 Stiles, B. L. et al. Selective deletion of Pten in pancreatic cells leads to increased islet mass and resistance to STZ-induced diabetes. Mol Cell Biol 26, 2772-2781, doi:26/7/2772 [pii] 10.1128/MCB.26.7.2772-2781.2006 (2006). 30 Kaimal, V. et al. Integrative systems biology approaches to identify and prioritize disease and drug candidate genes. Methods Mol Biol 700, 241- 259, doi:10.1007/978-1-61737-954-3_16 (2011). 31 Spence, J. R. & Wells, J. M. Translational embryology: using embryonic principles to generate pancreatic endocrine cells from embryonic stem cells. Dev Dyn 236, 3218-3227, doi:10.1002/dvdy.21366 (2007). 32 Ketola, I. et al. Transcription factor GATA-6 is expressed in the endocrine and GATA-4 in the exocrine pancreas. Mol Cell Endocrinol 226, 51-57, doi:S0303-7207(04)00236-9 [pii] 10.1016/j.mce.2004.06.007 (2004). 33 Zuber, C., Fan, J. Y., Guhl, B. & Roth, J. Misfolded proinsulin accumulates in expanded pre-Golgi intermediates and endoplasmic reticulum subdomains in pancreatic cells of Akita mice. FASEB J 18, 917-919, doi:10.1096/fj.03-1210fje 03-1210fje [pii] (2004). 34 Gupta, S., McGrath, B. & Cavener, D. R. PERK (EIF2AK3) regulates proinsulin trafficking and quality control in the secretory pathway. Diabetes 59, 1937-1947, doi:db09-1064 [pii] 10.2337/db09-1064 (2010). 35 Mako, M. E., Starr, J. I. & Rubenstein, A. H. Circulating proinsulin in patients with maturity onset diabetes. Am J Med 63, 865-869 (1977). 36 Kahn, C. R. Diabetes. Causes of insulin resistance. Nature 373, 384-385, doi:10.1038/373384a0 (1995). 37 Suckale, J. & Solimena, M. Pancreas islets in metabolic signaling--focus on the -cell. Front Biosci 13, 7156-7171, doi:3218 [pii] (2008). 38 Konstantinova, I. et al. EphA-Ephrin-A-mediated cell communication regulates insulin secretion from pancreatic islets. Cell 129, 359-370, doi:S0092-8674(07)00368-6 [pii] 10.1016/j.cell.2007.02.044 (2007). 39 Voeltz, G. K., Prinz, W. A., Shibata, Y., Rist, J. M. & Rapoport, T. A. A class of membrane proteins shaping the tubular endoplasmic reticulum. Cell 124, 573-586, doi:S0092-8674(06)00067-5 [pii] 10.1016/j.cell.2005.11.047 (2006). 40 Okamoto, M. et al. High-curvature domains of the ER are important for the organization of ER exit sites in Saccharomyces cerevisiae. J Cell Sci, doi:jcs.100065 [pii] 10.1242/jcs.100065 (2012).

69 41 Steiner, P. et al. Reticulon 1-C/neuroendocrine-specific protein-C interacts with SNARE proteins. J Neurochem 89, 569-580, doi:10.1111/j.1471- 4159.2004.02345.x JNC2345 [pii] (2004). 42 Verrier, S. E. et al. Members of a mammalian SNARE complex interact in the endoplasmic reticulum in vivo and are found in COPI vesicles. Eur J Cell Biol 87, 863-878, doi:S0171-9335(08)00115-5 [pii] 10.1016/j.ejcb.2008.07.003 (2008). 43 Shibata, Y. et al. Mechanisms determining the morphology of the peripheral ER. Cell 143, 774-788, doi:S0092-8674(10)01251-1 [pii] 10.1016/j.cell.2010.11.007 (2010). 44 Regazzi, R. et al. Expression, localization and functional role of small of the Rab3 family in insulin-secreting cells. J Cell Sci 109 ( Pt 9), 2265-2273 (1996). 45 Yi, Z. et al. The Rab27a/granuphilin complex regulates the exocytosis of insulin-containing dense-core granules. Mol Cell Biol 22, 1858-1867 (2002). 46 Kasai, K. et al. Rab27a mediates the tight docking of insulin granules onto the plasma membrane during glucose stimulation. J Clin Invest 115, 388- 396, doi:10.1172/JCI22955 (2005). 47 Waselle, L. et al. Involvement of the Rab27 binding protein Slac2c/MyRIP in insulin exocytosis. Mol Biol Cell 14, 4103-4113, doi:10.1091/mbc.E03- 01-0022 E03-01-0022 [pii] (2003). 48 Izumi, T., Gomi, H. & Torii, S. Functional analysis of Rab27a effector granuphilin in insulin exocytosis. Methods Enzymol 403, 216-229, doi:S0076-6879(05)03018-1 [pii] 10.1016/S0076-6879(05)03018-1 (2005). 49 Gomi, H., Mizutani, S., Kasai, K., Itohara, S. & Izumi, T. Granuphilin molecularly docks insulin granules to the fusion machinery. J Cell Biol 171, 99-109, doi:jcb.200505179 [pii] 10.1083/jcb.200505179 (2005). 50 Kimura, T. et al. The GDP-dependent Rab27a effector coronin 3 controls endocytosis of secretory membrane in insulin-secreting cell lines. J Cell Sci 121, 3092-3098, doi:121/18/3092 [pii] 10.1242/jcs.030544 (2008). 51 Kimura, T. & Niki, I. Rab27a, actin and -cell endocytosis. Endocr J 58, 1- 6, doi:JST.JSTAGE/endocrj/K10E-391 [pii] (2011). 52 Ahlgren, U., Jonsson, J., Jonsson, L., Simu, K. & Edlund, H. -cell- specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the - cell phenotype and maturity onset diabetes. Genes Dev 12, 1763-1768 (1998). 53 Stoffers, D. A., Ferrer, J., Clarke, W. L. & Habener, J. F. Early-onset type- II diabetes mellitus (MODY4) linked to IPF1. Nat Genet 17, 138-139, doi:10.1038/ng1097-138 (1997).

70 54 Malecki, M. T. et al. Mutations in NEUROD1 are associated with the development of type 2 diabetes mellitus. Nat Genet 23, 323-328, doi:10.1038/15500 (1999). 55 Kristinsson, S. Y. et al. MODY in Iceland is associated with mutations in HNF-1alpha and a novel mutation in NeuroD1. Diabetologia 44, 2098- 2103, doi:10.1007/s001250100016 (2001). 56 Gonsorcikova, L. et al. Autosomal inheritance of diabetes in two families characterized by obesity and a novel H241Q mutation in NEUROD1. Pediatr Diabetes 9, 367-372, doi:PDI379 [pii] 10.1111/j.1399-5448.2008.00379.x (2008). 57 Brissova, M. et al. Reduction in pancreatic transcription factor PDX-1 impairs glucose-stimulated insulin secretion. J Biol Chem 277, 11225- 11232, doi:10.1074/jbc.M111272200 M111272200 [pii] (2002). 58 Johnson, J. D. et al. Increased islet apoptosis in Pdx1+/- mice. J Clin Invest 111, 1147-1160, doi:10.1172/JCI16537 (2003). 59 Naya, F. J. et al. Diabetes, defective pancreatic morphogenesis, and abnormal enteroendocrine differentiation in  2/neuroD-deficient mice. Genes Dev 11, 2323-2334 (1997). 60 Gu, C. et al. Pancreatic cells require NeuroD to achieve and maintain functional maturity. Cell Metab 11, 298-310, doi:S1550-4131(10)00076-8 [pii] 10.1016/j.cmet.2010.03.006 (2010). 61 Kobayashi, M. et al. FoxO1 as a double-edged sword in the pancreas: analysis of pancreas- and -cell-specific FoxO1 knockout mice. Am J Physiol Endocrinol Metab 302, E603-613, doi:10.‚Äã1152/‚Äãajpendo.‚Äã00469.‚Äã2011 ajpendo.00469.2011 [pii] 10.1152/ajpendo.00469.2011 (2012). 62 Rubio-Cabezas, O., Klupa, T. & Malecki, M. T. Permanent neonatal diabetes mellitus--the importance of diabetes differential diagnosis in neonates and infants. Eur J Clin Invest 41, 323-333, doi:10.1111/j.1365- 2362.2010.02409.x (2011). 63 Sachdeva, M. M. & Stoffers, D. A. Minireview: Meeting the demand for insulin: molecular mechanisms of adaptive postnatal -cell mass expansion. Mol Endocrinol 23, 747-758, doi:me.2008-0400 [pii] 10.1210/me.2008-0400 (2009). 64 Fonseca, S. G., Urano, F., Burcin, M. & Gromada, J. Stress hypERactivation in the -cell. Islets 2, 1-9, doi:10456 [pii] 10.4161/isl.2.1.10456 (2010). 65 Cnop, M., Foufelle, F. & Velloso, L. A. Endoplasmic reticulum stress, obesity and diabetes. Trends Mol Med 18, 59-68, doi:S1471- 4914(11)00145-6 [pii] 10.1016/j.molmed.2011.07.010 (2012).

71 66 Volchuk, A. & Ron, D. The endoplasmic reticulum stress response in the pancreatic -cell. Diabetes Obes Metab 12 Suppl 2, 48-57, doi:10.1111/j.1463-1326.2010.01271.x (2010). 67 Oslowski, C. M. & Urano, F. The binary switch that controls the life and death decisions of ER stressed cells. Curr Opin Cell Biol 23, 207-215, doi:S0955-0674(10)00211-5 [pii] 10.1016/j.ceb.2010.11.005 (2011). 68 McDonald, E., Krishnamurthy, M., Goodyer, C. G. & Wang, R. The emerging role of SOX transcription factors in pancreatic endocrine cell development and function. Stem Cells Dev 18, 1379-1388, doi:10.1089/scd.2009.0240 (2009). 69 Kormish, J. D., Sinner, D. & Zorn, A. M. Interactions between SOX factors and Wnt/-catenin signaling in development and disease. Dev Dyn 239, 56-68, doi:10.1002/dvdy.22046 (2010). 70 Goldsworthy, M. et al. Role of the transcription factor sox4 in insulin secretion and impaired glucose tolerance. Diabetes 57, 2234-2244, doi:db07-0337 [pii] 10.2337/db07-0337 (2008). 71 Dubois, C. L. et al. Sox9-haploinsufficiency causes glucose intolerance in mice. PLoS One 6, e23131, doi:10.1371/journal.pone.0023131 PONE-D-11-05019 [pii] (2011).

72 Figure Legends.

Figure 1. Sox17 is not required for cell development.

A) Schematic representation of the Sox17-paLOF mice. B) Quantitation of fold difference in Sox17 transcript showed significant reduction of Sox17 mRNA levels in Sox17-paLOF islets (asterisk (*) shows p-value  0.05). Real Time RT-

PCR samples were normalized to GAPDH mRNA. C-F) Immunofluorescence using anti-insulin, -somatostatin, -glucagon, and –Pdx1 antibodies in control and

Sox17-paLOF mice showed no difference in islet architecture between genotypes.

Figure 2. Sox17-paLOF results in elevated proinsulin protein in the islets.

A, B) Non-fasting blood glucose levels and total plasma insulin levels were measured for control and Sox17-paLOF mice, showing no difference. C) Non- fasting ratio of plasma proinsulin to total insulin showed a trend of increase in

Sox17-paLOF. D) Western blot of insulin and proinsulin from pancreatic islets.

Quantitation of western blot of insulin and proinsulin (n=5 for Pdx1Cre;Sox17fl/+ islets, n=5 for Pdx1Cre;Sox17fl/fl) in 2E and 2F showed significant increased in proinsulin protein in islets (asterisk shows p-value  0.01). G) Glucose tolerance test showed similar glucose clearance. H) Glucose-stimulated insulin secretion showed higher trend of total plasma insulin level in Sox17-paLOF mice. I, J)

Insulin tolerance test and total pancreatic islet insulin content showed no difference in insulin peripheral sensitivity and insulin production between genotypes.

73 Figure 3. Loss of Sox17 in the pancreas causes accumulation of proinsulin in the ER and structural changes in secretory organelles.

A-F) Immunofluorescence confocal staining of adult pancreas section in 63x oil magnification at Nyquist limits showed colocalization between ER (KDELR) and proinsulin. -catenin stained the cell membrane. Mouse-on-mouse staining kit was used to help reduce the background staining. These Z-stacks compressed images (13 optical sections on average) were acquired using Zeiss LSM 510. G)

Quantitation of percent ER and proinsulin colocalization using Bitplane Imaris software (asterisks show p-value  0.01). H-M) Immunofluorescence staining showed no difference in colocalization between pre-Golgi (ERGIC) and proinsulin.

N-S) Immunofluorescence staining showed no difference in colocalization between Golgi (GM130) and proinsulin. T-V) EM image analysis showed dilated morphology of ER, Golgi, and mitochondria in the Sox17-paLOF islet sections.

N=nucleus, ER=endoplasmic reticulum, G=Golgi, M=mitochondria.

Figure 4. Sox17-paLOF mice are prediabetic and prone to high fat diet- induced stress of cells.

A) Body weight measurement over 23 weeks of high fat diet showed no difference between genotypes. B) 4 hours fasting blood glucose level over 23 weeks of high fat diet showed a tendency of Sox17-paLOF (Pdx1Cre;Sox17fl/fl) mice to have higher blood glucose levels than controls (Pdx1Cre;Sox17fl/+).

Asterisk show p-value  0.01. C) Glucose tolerance test showed impaired glucose clearance in the Sox17-paLOF mice. D) Food intake measurement

74 starting at 21 weeks after the mice have been on high fat diet showed no difference between genotypes. E) Insulin tolerance test showed no difference in peripheral insulin sensitivity in the Sox17-paLOF. The asterisks indicate p-value

 0.05.

Figure 5. Sox17 overexpression for 24 hours is sufficient to alter proinsulin:total insulin protein ratio and proinsulin secretion in vivo, followed by accumulation of proinsulin in the plasma, leading to diabetes after prolonged exposure of Sox17.

A) Schematic representation of the Sox17-GOF mice. B) Quantitation of fold difference in Sox17 transcript showed significant increased of Sox17 mRNA levels in Sox17-GOF islets (The asterisk indicates p-value  0.01). Real Time

RT-PCR samples were normalized to GAPDH mRNA. C,D).

Immunofluorescence of adult pancreas sections was performed using anti-Sox17 antibody showing elevated Sox17 protein in the islets. Images taken with Zeiss

LSM 510 with 40x dry objective. E-J) Immunofluorescence stainings for islet markers, such as insulin, Pdx1, glucagon, and Glut2, showed no islet architecture defect found in 24 hours after Sox17 was overexpressed. K)

Western blot of insulin and proinsulin from pancreatic islets (n=3 for controls and n=5 for Sox17-GOF islets). L,M) Quantitation of western blot of insulin and proinsulin in (K) showed significant decreased in proinsulin protein in islets (p- value  0.01). N) Non-fasting blood glucose level measurement over 20 days of

Sox17 overexpression showed the mice developed diabetes overtime (n=6 for

75 controls and n=5 for Sox17-GOF islets, asterisks show p-value  0.05). O) Non- fasting ratio of plasma proinsulin:total insulin level measurement showed increased presence of proinsulin in the blood overtime (n=3 for controls and n=3 for Sox17-GOF islets, asterisks show p-value  0.05). P) Non-fasting blood glucose level measurement over 41 days of Sox17 overexpression. Sox17-GOF mice became diabetics, and when dox was removed from the diet of some of these diabetic mice, their blood glucose returned to control levels (1 asterisk (*) shows statistical significant comparison between controls and Sox17-GOF mice

ON DOX, 2 asterisks (**) shows statistical significant comparison between controls and Sox17-GOF mice OFF DOX). Q-X) Immunofluorescence staining for islet markers after prolonged Sox17 overexpression. Sox17 was still highly overexpressed, glucagon and somatostatin were normal, but insulin and Pdx1 were highly reduced in patches of islets 41 days after DOX administration. E- cadherin stained the cell membrane, showing a change in cell shape of the

Sox17-GOF islets. Y-AB) Sox17-GOF mice were taken OFF DOX and showed restored islet marker expressions and cell shape.

Figure 6. 24 hours of Sox17 overexpression alters proinsulin trafficking through the secretory organelle machinery.

A-F) Immunofluorescence staining of adult pancreas section showed colocalization between pre-Golgi (ERGIC) and proinsulin. DRAQ5 stained nucleus. Mouse-on-mouse staining kit was used to help reduce the background staining. These Z-stacks compressed images were acquired using Zeiss LSM

76 510 with 63x oil objective. G) Quantitation of percent pre-Golgi and proinsulin colocalization using Bitplane Imaris software (asterisk shows p-value  0.01). H-

K) EM image analysis showed dilated morphology of ER and mitochondria in the

Sox17-paLOF islet sections. ER=endoplasmic reticulum, G=Golgi,

M=mitochondria. L-W) Immunofluorescence staining showed colocalizations between ER (KDELR), Pre-Golgi (ERGIC), Golgi (GM130) and proinsulin 20 days after DOX administration. Z-stacks compressed images were taken using

Nikon A1R si with 100x PlanApo oil objective and tile-scanned to image the entire islet.

Figure 7. Sox17 regulates pathways involved in insulin transport and secretion.

A) Microarray heat map and the gene ontology of the biological process that are involved in the Sox17-misregulated genes cluster (upregulated genes in red, downregulated genes in blue). B) Schematic model representation of how Sox17 may function in regulating the transcriptome of the genes involved in insulin processing and secretion pathway. Genes that are upregulated by Sox17 overexpression are shown in red, and those that are downregulated are shown in blue. Fold change difference from the microarray result is shown in parantheses.

77 78 79 80 81 82 83 84 85 Supplementary Figure Legends.

Supplementary Figure 1. Sox17 immunostaining in Sox17-paLOF, wildtype, and Sox17-GOF islets.

A) Sox17 immunostaining using DAB in 20x magnification showed low reduction of Sox17 protein in Sox17-paLOF islets (A) compared to the wildtype (B) and its protein overexpression in Sox17-GOF islets (C). Immunofluorescence staining of adult pancreas section in 40x magnification at Nyquist limits showed low reduction of Sox17 staining in Sox17-paLOF islets (D) compared to the basal level of Sox17 in wildtype islet (E) and overexpression level in Sox17-GOF islets

(F). Insulin is stained to mark cells and E-cadherin is marking cell membrane.

These images were acquired using Zeiss LSM 510.

Supplementary Figure 2. Percent colocalizations between proinsulin and organelle markers, and their total regional areas.

A) No difference found in the total proinsulin and ER volume area between genotypes, quantitated by the Imaris software. B, D) Quantitation of percent pre-

Golgi, Golgi and proinsulin colocalization using Bitplane Imaris software. C) No change in the total proinsulin and pre-Golgi volume area. E) Total proinsulin area was slightly increased in the Sox17-paLOF islets (asterisk shows p-value  0.05), but no difference found in the Golgi total volume area.

86 Supplementary Figure 3. Plasma proinsulin, total insulin, and their ratio in

Sox17-paLOF mice on high fat diet.

A-C) After high fat diet intake, mice were fasted for 4 hours before plasma samples were taken to measure proinsulin, total insulin, and the ratio between proinsulin and total insulin levels. There was a trend of reduction in both proinsulin and total insulin levels (A, C), with no difference in the overall plasma ratio of proinsulin:total insulin (B).

Supplementary Figure 4. Transcriptional insulin processing enzymes level.

Sox17 does not regulate the PC1/3 and PC2 insulin processing enzymes transcriptionally. Quantitation of fold difference in PC1/3 (A) and PC2 (B) transcripts showed no significant difference between genotypes. Real Time RT-

PCR samples were normalized to GAPDH mRNA.

Supplementary Figure 5. Plasma proinsulin and total insulin levels in

Sox17-GOF mice.

A-B) Proinsulin and total insulin plasma levels showed continued trend of increase in the proinsulin level with no difference in total insulin level by 2-3 weeks after Dox administration.

Supplementary Figure 6. Percent colocalization and total regional areas of proinsulin and various organelle markers.

87 A-B) Percent colocalization and total volume area of proinsulin and ER analysis, showing no difference between genotypes. C) Total area of Pre-Golgi showed lower volume in the Sox17-GOF islets, with no difference in the total proinsulin region. D-E) No change found in the percent colocalization and total regions of proinsulin and Golgi.

Supplementary Figure 7. Microarray validation of altered genes in Sox17-

GOF islets.

Data showed mRNA increased level in Sox17 (A), Rtn1 (H), Use1 (J), Gsta4 (M),

Lipf (N), Mobp (O), Defb1 (P), decreased level in Glut2 (C), Pdx1 (D), Glp1R (E),

Foxo1 (F), Atf4 (G), Hdac6 (I), Pkd1 (K), Prkca (L), Lpl (Q), Ppp1r1a (R), Vilip-1

(S), Insrr (T), Cpb2 (U), no change in Insulin (B). Asterisk indicates p-value 

0.05.

Supplementary Table 1. Primary and Secondary Antibodies.

Supplementary Table 2. Sox17-GOF islet microarray gene list.

List of misregulated genes in Sox17-GOF islets by 1.55 fold change and above.

Red highlighted genes were upregulated, blue highlighted genes were downregulated.

88 89 90 91 92 93 94 Supplementary Table 1:

Primary and Secondary Antibodies

Primary Antibody Source Dilution

Rat anti-BrdU Abcam 1:100

Guinea Pig anti-Sox17 Whitsett Lab 1:1000

Mouse anti-FoxA2 Novus Biologicals 1:500

Mouse anti-proinsulin R&D 1:500

95 Rab anti-Cre Novagen 1:5000

Chicken anti-BGal Abcam 1:1000

Rat anti-Insulin R&D 1:200

Goat anti-Somatostatin Santa Cruz 1:200

Guinea Pig anti-Glucagon Linco 1:1000

Rabbit anti-active Caspase 3 Cell Signaling 1:100

Mouse anti-KDELR Enzo 1:100

Rabbit anti-ERGIC Sigma 1:200

Mouse anti-E-Cadherin BD Biosciences 1:500

Mouse anti-GM130 BD Transduction 1:100

Goat anti-Glut2 Santa Cruz 1:1000

Mouse anti-Nkx2.2 74.5A5-C 1:100

Goat anti-Pdx1 Abcam 1:100

Goat anti-Hnf1B Santa Cruz 1:100

Rabbit anti-B-catenin Santa Cruz 1:100

Guinea Pig anti-Insulin DAKO 1:1000

Secondary Antibody Source Dilution

Goat anti-mouse IgG2a 549 Jackson Immuno 1:400

Goat anti- Biotin Jackson Immuno 1:200

Goat anti-mouse IgG1 488 Jackson Immuno 1:400

Goat anti-rabbit 488 Invitrogen 1:400

Goat anti-mouse Biotin Jackson Immuno 1:200

Goat anti-guinea pig Cy5 Jackson Immuno 1:200

96 Goat anti-rabbit Cy5 Jackson Immuno 1:200

Donkey anti-rabbit Cy3 Jackson Immuno 1:200

Donkey anti-rat 546 Invitrogen 1:200

Donkey anti-goat 549 Jackson Immuno 1:200

Donkey anti-rabbit 488 Invitrogen 1:200

Donkey anti-guinea pig Cy5 Jackson Immuno 1:200

Donkey anti-guine pig 549 Jackson Immuno 1:200

Donkey anti-guine pig biotin Jackson Immuno 1:200

Donkey anti-goat 488 Jackson Immuno 1:200

Donkey anti-goat biotin Invitrogen 1:200

Donkey anti-mouse 647 Invitrogen 1:200

Donkey anti-mouse Cy3 Jackson Immuno 1:200

Conjugates/Nuclear Dyes

Strepavidin cy5 Jackson Immuno 1/500

TOPRO3 Jackson Immuno 1/20,000

DRAQ5 Cell Signaling 1/10,000

TSA KIT, tyr 488 Invitrogen 1:200

TSA KIT, tyr 549 Invitrogen 1:200

Supplementary Table 2: misregulated genes in Sox17-GOF islets

Fold Genbank Change P-value Common ENSMUST0000002568 0 29.45 0.0356 Lipf XM_917532 27.57 1.66E-05 EG640530

97 NM_008614 21.66 0.00165 Mobp ENSMUST0000003490 3 17.76 0.00606 Gsta4 NM_019790 13.95 0.00112 Tmeff2 BC055924 13.93 0.00831 EG545886 AF498319 11.11 3.95E-05 Rgs13 BC023359 10.09 0.00228 3632451O06Rik NM_013927 9.731 0.00064 Cngb3 NM_175647 7.219 0.00273 Dmrta1 NM_145629 6.465 0.000132 Pls3 BC048664 6.401 0.0203 1700016D06Rik ENSMUST0000001424 8 6.234 0.0462 Sval2 M65237 6.227 0.0246 Mug-ps1 NM_026713 6.055 0.0265 Mogat1 NM_177747 5.989 0.000235 Zfp711 NM_133348 5.557 0.00198 Acot7 ENSMUST0000009909 1 5.332 6.18E-05 Gm410 NM_009037 5.249 9.30E-06 Rcn1 NM_134022 5.221 0.0024 6330403K07Rik NM_172781 5.206 0.0462 Klhl4 NM_019754 5.117 0.00182 Tagln3 NM_009801 4.937 0.00438 Car2 NM_080448 4.575 0.00224 Srgap3 NM_008278 4.511 0.0137 Hpgd NM_175465 4.459 0.00123 Sestd1 NM_008885 4.426 0.000901 Pmp22 NM_175771 4.37 0.00805 Tmem47 NM_011807 4.339 0.00296 Dlg2 NM_177373 4.209 0.00233 Ppfia2 NM_010408 4.145 0.00216 Hcn1 ENSMUST0000008546 9 4.066 0.00618 Pik3cg NM_153457 4.037 0.0176 Rtn1 NM_008594 4.014 0.00237 Mfge8 NM_138595 3.958 0.0192 Gldc NM_011348 3.824 0.000452 Sema3e ENSMUST0000006405 4 3.821 0.0149 Syt1 XM_146632 3.723 0.00103 9030420J04Rik NM_027997 3.721 0.000584 Serpina9

98 NM_011254 3.718 0.000254 Rbp1 NM_013930 3.683 0.00191 Aass NM_019703 3.605 0.00272 Pfkp NM_019656 3.586 0.00143 Tspan6 NM_019641 3.54 2.97E-05 Stmn1 NM_010251 3.515 0.0172 Gabra4 NM_027168 3.498 0.00134 Hddc2 ENSMUST0000003030 3 3.493 0.00774 Cyp2j6 NM_019641 3.488 9.45E-05 Stmn1 NM_008481 3.432 0.000217 Lama2 NM_001011875 3.419 0.024 Rnase12 NM_173379 3.4 0.000368 Leprel1 NM_013657 3.395 0.00456 Sema3c NM_009807 3.341 0.00549 Casp1 NM_025760 3.303 0.0254 Ptplad2 NM_001082976 3.288 0.00142 Tc2n NM_080448 3.273 0.0118 Srgap3 NM_029942 3.273 0.00056 Prelid2 NM_029447 3.19 0.00042 Nln NM_181414 3.187 0.0058 Pik3c3 ENSMUST0000004042 3 3.152 0.0204 Cd59a AK046516 3.132 0.00687 ENSMUSG00000071543 ENSMUST0000010780 2 3.11 0.00124 Trim59 NM_080462 3.108 0.00883 Hnmt AF184981 3.073 0.0402 Fmo2 ENSMUST0000002987 6 3.068 0.0018 Calb1 NM_198702 3.058 0.00054 Lphn3 NM_010807 3.034 0.0218 Marcksl1 NM_010634 3.033 0.00105 Fabp5 NM_018863 3.002 0.0238 Pdyn BC051224 3 0.0172 A530053G22Rik NM_001042592 2.992 0.0478 Arrdc4 ENSMUST0000000449 7 2.948 0.00089 Large NM_010597 2.922 0.00749 Kcnab1 NM_146100 2.918 0.000279 Ina ENSMUST0000004651 3 2.902 0.00286 Phyhipl

99 NM_007697 2.889 0.0137 Chl1 BC002008 2.871 2.00E-05 Fabp5 NM_001033773 2.856 0.0351 Ube2u NM_001081128 2.833 0.00035 Mtr NM_001033331 2.807 0.0296 Gas2l3 NM_178779 2.803 0.00601 Rnf152 NM_001009935 2.786 0.0377 Txnip BC055351 2.782 0.00489 Hn1l NM_144855 2.759 0.0175 Cbs ENSMUST0000003936 6 2.745 0.00467 Kcnh8 NM_019691 2.745 0.000159 Gria4 NM_024279 2.738 0.0112 1700094C09Rik NM_029619 2.724 0.00179 Msrb2 ENSMUST0000003355 4 2.72 0.0178 Gpr165 NM_026167 2.706 0.0107 Klhl13 ENSMUST0000000734 0 2.704 0.0014 Atp12a NM_207685 2.701 0.00626 Elavl2 NM_013709 2.674 0.000149 Sh3yl1 ENSMUST0000008172 1 2.659 0.000244 Ezh2 NM_011035 2.658 0.013 Pak1 NM_172805 2.655 0.00654 Kcnh5 NM_018779 2.647 0.0405 Pde3a NM_027971 2.619 0.0132 Serpinb12 NM_001033346 2.612 0.0106 Lrrc55 NM_172824 2.598 0.0016 Ccdc14 NM_029466 2.574 0.00589 Arl5b ENSMUST0000001929 0 2.561 0.00513 Cacng2 BC100412 2.553 0.00586 1700001E04Rik NM_019631 2.526 0.0471 Tmem45a NM_172671 2.526 0.0002 Lgr4 NM_001001804 2.519 0.00706 Abhd7 NM_001029930 2.517 0.00418 ENSMUSG00000068790 BC117712 2.514 0.000213 Clec12b NM_021099 2.482 0.00177 Kit ENSMUST0000007802 1 2.468 0.0145 Glmn NM_172310 2.461 0.0409 Tarsl2

100 NM_178005 2.442 0.00535 Lrrtm2 NM_008604 2.442 0.00172 Mme BC049669 2.428 5.64E-05 1700047I17Rik1 BC049669 2.428 5.64E-05 1700047I17Rik1 NM_201371 2.423 0.0186 Prmt8 BC037216 2.397 0.00309 Lrrtm4 NM_153124 2.393 0.0103 St8sia5 NM_001083897 2.391 0.00386 Mpzl1 NM_023245 2.375 0.0079 Palmd NM_198302 2.374 0.0123 Rbm11 NM_028719 2.37 0.00957 Cpne4 BC050850 2.359 0.00977 AK129302 NM_007675 2.357 0.0206 Ceacam10 NM_028651 2.357 0.0115 Tmtc4 NM_153587 2.355 0.0248 Rps6ka5 ENSMUST0000007052 2 2.355 0.00154 Plod2 NM_011943 2.354 0.00135 Map2k6 NM_153163 2.35 0.0056 Cadps2 NM_013540 2.34 0.00106 Gria2 NM_026139 2.338 0.0061 Armcx2 XM_001479202 2.328 0.0112 LOC100047943 NM_198422 2.324 0.00589 Paqr3 NM_029001 2.312 0.00623 Elovl7 BC060180 2.309 0.00646 Ccng2 NM_133239 2.309 0.000543 Crb1 EF651808 2.291 0.00278 Hn1l NM_009121 2.288 0.000817 Sat1 NM_178673 2.284 0.014 Fstl5 NM_177173 2.277 0.0142 A830018L16Rik NM_028979 2.273 0.000678 Cyp2j9 NM_010315 2.271 0.00905 Gng2 NM_146140 2.265 0.00681 Tram1l1 NM_001081121 2.243 0.027 4931429I11Rik EF651808 2.241 0.00108 Hn1l NM_001083628 2.238 0.000428 AK220484 NM_010164 2.234 0.00572 Eya1 NM_007913 2.228 0.0139 Egr1 ENSMUST0000007585 3 2.217 0.0298 Cks2 NM_001031664 2.209 0.0172 Nudt10 AK129372 2.209 0.00784 9430031J16Rik

101 NM_172256 2.208 0.00251 Dync2li1 ENSMUST0000003503 8 2.204 0.00423 Faim BC016222 2.201 0.00273 LOC544988 NM_001099298 2.2 0.0211 Scn2a1 NM_145575 2.198 0.00136 Cald1 NM_028876 2.192 0.0163 Tmed5 NM_134072 2.189 0.00578 Akr1c14 NM_028810 2.184 0.0282 NM_008633 2.17 0.00062 Mtap4 NM_010237 2.162 0.0357 Frk NM_013754 2.162 0.0192 Insl6 NM_028665 2.158 0.00374 Ankrd42 NM_177290 2.15 0.0394 Itgb8 ENSMUST0000003068 4 2.14 0.00432 Gnl2 NM_026470 2.135 0.00448 Spata6 NM_001037725 2.135 4.07E-05 Als2cr13 ENSMUST0000000292 6 2.134 0.00151 Pla1a NM_133227 2.133 0.0154 Nup155 ENSMUST0000002836 9 2.127 0.0305 Dapl1 ENSMUST0000005395 0 2.125 0.0244 Lrrc28 EF651808 2.12 0.00233 Hn1l NM_017377 2.119 0.00507 B4galt2 ENSMUST0000009934 3 2.118 0.0403 Nr2c1 DQ112091 2.117 0.00164 B230118H07Rik NM_023249 2.103 0.0172 Ypel1 BC100412 // 1700001E04Rik // BC100412 2.102 0.000781 1700001E04Rik ENSMUST0000005786 6 2.101 0.0291 Nrsn1 ENSMUST0000002942 3 2.098 0.0155 Serpini1 NM_032465 2.094 0.00344 Cd96 AB073967 2.09 0.00404 1100001E04Rik NM_001024706 2.08 0.00767 EG432825 BC129903 2.072 0.0127 Fgd4 EF651808 2.053 0.00224 Hn1l NM_008083 2.048 0.0377 Gap43

102 NM_009923 2.048 0.0257 Cnp NM_009199 2.042 0.00277 Slc1a1 NM_026825 2.037 0.012 Lrrc16 BC035305 2.034 0.00588 Aig1 NM_181579 2.028 0.0136 Pof1b ENSMUST0000011429 2 2.027 0.0474 Cadm2 BC032970 2.026 0.00122 2810026P18Rik NM_001043335 2.022 0.0199 Eml1 NM_010882 2.016 0.000483 Ndn NM_177591 2.014 0.00468 Igsf1 NM_001098230 2.005 0.0129 Ppm2c NM_019958 2.002 0.00457 Rgs17 NM_001081664 1.998 0.0102 4833423E24Rik ENSMUST0000005243 1 1.996 0.0161 Armcx6 NM_027409 1.994 0.0266 Mospd1 ENSMUST0000007585 3 1.992 0.0287 Cks2 NM_145928 1.99 0.0211 Tspan14 NM_008576 1.984 0.0433 Abcc1 XR_032130 1.981 0.0146 LOC667519 NM_145743 1.978 0.0181 Lace1 NM_023598 1.97 0.0278 Arid5b NM_008485 1.97 0.0101 Lamc2 ENSMUST0000003480 1 1.964 0.0364 Bckdhb NM_175393 1.964 0.0122 4930555G01Rik NM_175393 1.964 0.0122 4930555G01Rik NM_027934 1.961 0.000703 Rnf180 NM_001083628 1.96 0.00932 AK220484 NM_001081257 1.96 0.00124 LOC545291 NM_009255 1.948 0.0428 Serpine2 NM_008585 1.943 0.000297 Mep1a NM_027569 1.938 0.00124 Spag9 NM_020574 1.936 0.0376 Kcne3 NM_008017 1.934 0.0141 Smc2 ENSMUST0000004005 6 1.934 0.00436 Ppfibp2 NM_030165 1.931 0.00906 Galnact2 NM_145940 1.927 0.00287 Wipi1 ENSMUST0000005701 1.917 0.00807 Agtrl1

103 9 NM_023732 1.917 0.00754 Abcb6 NM_146201 1.913 0.0233 Zfp553 NM_153195 1.91 0.0385 Fbxo7 NM_177742 1.9 0.0292 Triml1 NM_011883 1.887 0.0291 Rnf13 NM_013751 1.887 0.00922 Hrasls NM_173760 1.881 0.0251 Hisppd1 ENSMUST0000004968 1 1.88 0.0466 Itgbl1 NM_054088 1.88 0.00532 Pnpla3 NM_025816 1.879 0.000175 Tax1bp1 ENSMUST0000000817 9 1.878 0.0061 Mid1ip1 NM_027571 1.877 0.0477 P2ry12 NM_026115 1.877 0.00982 Hat1 NM_172770 1.875 0.00825 Ttc12 AK043745 1.871 0.0485 Tmem71 NM_027927 1.871 0.0367 Ints12 ENSMUST0000002717 2 1.869 0.00788 Ica1l NM_013935 1.868 0.0402 Ptpla NM_001099633 1.867 0.0129 Dnahc9 NM_023119 1.866 0.0156 Eno1 NM_025588 1.86 0.00208 Exoc2 NM_025968 1.85 0.0479 Ltb4dh ENSMUST0000009042 9 1.85 0.0464 Cd59b NM_011920 1.848 0.0222 Abcg2 NM_172496 1.845 0.0189 Cobl NM_175563 1.835 0.0203 Prr11 NM_025949 1.832 0.0462 Rps6ka6 XM_143339 1.83 0.00888 Wdr49 NM_026013 1.83 0.000681 Tmem77 NM_008597 1.825 0.0482 Mgp NM_001110222 1.825 0.0294 Dcx NM_175116 1.823 0.0114 P2ry5 NM_028696 1.823 0.00171 Obfc2a NM_010360 1.821 0.00192 Gstm5 NM_033561 1.82 0.00615 Eif4h NM_177167 1.817 0.00828 Ppm1e NM_001081477 1.808 0.0164 Brwd3

104 NM_001029912 1.807 0.0359 Zswim5 NM_183389 1.805 0.0168 Duxbl NM_183389 1.805 0.0168 Duxbl NM_019724 1.804 0.00317 Mmp16 NM_027401 1.801 0.00284 1700010C24Rik NM_009846 1.795 0.000545 Cd24a NM_026268 1.794 0.0408 Dusp6 NM_009133 1.791 0.00372 Stmn3 NM_009384 1.788 0.00374 Tiam1 NM_181577 1.783 0.0181 Ccdc85a ENSMUST0000002925 9 1.782 0.0127 Mccc1 NM_023119 1.779 0.0094 Eno1 NM_172685 1.777 0.00177 Slc25a24 NM_178674 1.774 0.0337 Fbxl21 NM_023233 1.773 0.0325 Trim13 ENSMUST0000003278 1 1.773 0.00278 Nox4 ENSMUST0000002945 3 1.769 0.00302 Vangl1 NM_177052 1.766 0.00394 Kif6 NM_022032 1.764 0.0172 Perp NM_025866 1.761 0.0344 Cdca7 NM_023119 1.76 0.00799 Eno1 NM_001085378 1.758 0.0148 Myh7b NM_028031 1.754 0.0147 Zdhhc13 NM_001077202 1.753 0.00507 Hs6st2 ENSMUST0000006268 4 1.752 0.00714 Tmem64 ENSMUST0000002760 6 1.75 0.00637 Rgs2 BC100412 1.749 0.00156 1700001E04Rik ENSMUST0000002163 4 1.744 0.0273 Akr1c13 NM_011627 1.744 0.00322 Tpbg NM_197990 1.743 0.0362 1700025G04Rik NM_019413 1.741 0.0289 Robo1 ENSMUST0000002156 4 1.735 0.0239 Smoc1 NM_001081316 1.731 0.0195 Dsel NM_008591 1.731 0.0176 Met NM_172597 1.728 0.0426 Txndc16 NM_144860 1.717 0.00151 Mib1

105 NM_001080995 1.716 0.00122 4632434I11Rik BC069874 1.715 0.025 2810408A11Rik AK143414 1.712 0.00495 Zfp7 NM_030677 1.711 0.0308 Gpx2 NM_030709 1.709 0.0253 Tmprss5 BC085192 1.709 0.0214 L3mbtl3 NM_007483 1.706 0.00011 Rhob BC047214 1.705 0.00949 Fbxl5 NM_009776 1.703 0.0147 Serping1 ENSMUST0000010277 7 1.697 0.0207 Lepr ENSMUST0000002472 4 1.696 0.028 Crisp2 BC100412 1.696 0.00573 1700001E04Rik NM_178705 1.694 0.0395 Luzp2 NM_178628 1.685 0.0444 Spg3a NM_172742 1.685 0.0322 Mtmr10 NM_201600 1.684 0.0405 Myo5b NM_028724 1.683 0.00308 Rin2 NM_207229 1.679 0.0364 Plac9 NM_207229 1.679 0.0364 Plac9 NM_176968 1.676 0.0252 Nt5dc1 NM_007630 1.673 0.0236 Ccnb2 AJ306625 1.673 0.00719 Synpo2 BC026813 1.672 0.0329 6330416L07Rik NM_176860 1.672 0.0255 Ubash3b NM_013755 1.67 0.0381 Gyg NM_008829 1.669 0.0441 Pgr NM_178797 1.668 0.015 Mlstd1 XM_884776 1.668 0.00802 EG241989 BC027537 1.666 0.00233 Gins1 NM_178920 1.665 0.00906 Mal2 NM_178712 1.664 0.0209 Gpr64 NM_134130 1.662 0.0449 Abhd3 NM_001033322 1.661 0.0454 Gucy1a2 NM_008252 1.659 0.0297 Hmgb2 NM_182999 1.658 0.0398 Rnf20 NM_009704 1.658 0.0227 Areg NM_016710 1.658 0.00314 Nsbp1 NM_021470 1.657 0.0255 Rnf32 NM_027759 1.656 0.0142 Fsip1 NM_010305 1.654 0.0321 Gnai1

106 NM_001033214 1.653 0.0146 E330034G19Rik ENSMUST0000004284 2 1.652 0.0172 Cdon NM_009819 1.652 0.000555 Ctnna2 XM_983501 1.65 0.0132 EG666383 NM_183187 1.647 0.0159 BC055107 AF401531 1.646 0.0114 Cklf NM_001007460 1.645 0.00141 Zdhhc23 NM_178202 1.642 0.0205 Hist1h2bp NM_019878 1.642 0.00306 Sult1b1 NM_025284 1.639 0.00224 Tmsb10 NM_207229 1.636 0.041 Plac9 NM_008288 1.635 0.0246 Hsd11b1 ENSMUST0000003270 1 1.634 0.0303 Tdrd12 NM_008413 1.634 0.00444 Jak2 ENSMUST0000005815 9 1.633 0.00999 Cnrip1 NM_009466 1.632 0.00812 Ugdh NM_175937 1.631 0.00554 Cpeb2 NM_153162 1.631 0.00281 Txnrd3 NM_028841 1.629 0.000456 Tspan17 NM_001081208 1.628 0.0228 Hs3st5 NM_001081189 1.628 0.00101 Uprt NM_022332 1.626 0.0291 St7 NM_007709 1.626 0.00164 Cited1 NM_008252 1.625 0.0352 Hmgb2 NM_172689 1.624 0.0123 Ddx58 NM_016669 1.615 0.0137 Crym NM_008646 1.613 0.0126 Mug2 NM_001008499 1.609 0.0149 Taar4 ENSMUST0000004178 0 1.609 0.00143 Endod1 NM_010930 1.607 0.0351 Nov NM_001080818 1.607 0.0321 Cdc14a NM_019635 1.605 0.00531 Stk3 NM_007874 1.604 0.000857 Reep5 NM_011326 1.603 0.0153 Scnn1g NM_019985 1.603 0.015 Clec1b NM_028804 1.603 0.003 Ccdc3 NM_178202 1.601 0.0238 Hist1h2bp NM_008252 1.599 0.0432 Hmgb2

107 ENSMUST0000005109 1 1.599 0.0119 Hist1h2be NM_152915 1.598 0.0378 Dner NM_007399 1.597 0.0166 Adam10 NM_025284 1.597 0.00608 Tmsb10 NM_009817 1.596 0.0219 Cast NM_028615 1.596 0.0075 Dppa2 AF461091 1.595 0.000172 Cspg5 NM_024255 1.594 0.0294 Hsdl2 ENSMUST0000002996 4 1.593 0.0348 Epha7 NM_130447 1.593 0.0273 Dusp16 BC100412 1.591 0.00196 1700001E04Rik NM_009848 1.59 0.0246 Entpd1 NM_153594 1.589 0.00421 Pcmtd2 NM_007639 1.587 0.029 Cd1d1 NM_011441 1.585 0.00441 Sox17 NM_052976 1.583 0.0362 Ophn1 NM_178202 1.581 0.0184 Hist1h2bp NM_153803 0.645 0.0151 BC038479 NM_053247 0.645 0.00795 Lyve1 XR_034456 0.645 0.00447 LOC668888 NM_022331 0.645 0.0024 Herpud1 NM_025999 0.645 0.000198 Rnf141 NM_013559 0.643 0.02 Hsp110 NM_010800 0.642 0.0197 Bhlhb8 NM_025790 0.642 0.000105 Them2 BC022225 0.641 0.039 Kif12 NM_145981 0.641 0.0197 Phyhip NM_145131 0.641 0.0182 Pitrm1 NM_031999 0.641 0.000439 Gpr137b NM_172927 0.64 0.0205 E330026B02Rik NM_023913 0.639 0.0142 Ern1 NM_180678 0.639 0.00675 Gars NM_029851 0.638 0.00786 Dync2h1 NM_007499 0.637 0.0227 Atm NM_010276 0.637 0.0198 Gem NM_008102 0.637 0.00209 Gch1 NM_019733 0.636 0.000391 Rbpms NM_030098 0.635 0.01 Rnase6 NM_031169 0.633 0.0215 Kcnmb1 NM_001011794 0.633 0.00924 Olfr1322

108 NM_177708 0.632 0.0102 Rtn4rl1 NM_207228 0.632 0.00658 Tsga10 NM_199449 0.631 0.00118 Zhx2 U01841 0.631 0.000811 Pparg NM_001007583 0.63 0.0175 Best3 NM_030565 0.63 0.00391 BC004044 NM_009767 0.629 0.00581 Chic1 NM_133769 0.628 5.97E-05 Cyfip2 NM_009903 0.627 0.0373 Cldn4 NM_001004173 0.627 0.00587 Sgpp2 NM_001081347 0.626 0.041 Rhobtb1 NM_010612 0.626 0.00504 Kdr NM_008018 0.625 0.0101 Sh3pxd2a NM_018852 0.624 0.0253 Scn9a ENSMUST0000002217 6 0.623 0.0271 Hmgcr NM_145491 0.623 0.0175 Rhoq NM_173745 0.623 0.000983 Dusp18 NM_001039934 0.622 0.000897 Mtap2 ENSMUST0000010269 8 0.62 0.000145 Rapgef4 NM_177769 0.619 0.0276 Elmod1 NM_009665 0.619 0.0076 Amd1 NM_019802 0.619 0.00132 Ggcx NM_009665 0.618 0.0162 Amd1 NM_011957 0.618 0.0139 Creb3l1 NM_018882 0.617 0.00504 Gpr56 NM_011406 0.615 0.0377 Slc8a1 NM_172537 0.615 0.00204 Sema6d NM_026053 0.614 0.0154 Gemin6 NM_007444 0.614 0.0149 Amd2 AK122385 0.614 0.0117 Rab3gap2 NM_029787 0.613 0.00196 Cyb5r3 NM_001099323 0.611 0.00184 RP23-211P15.2 NM_013682 0.61 0.0164 T2 NM_011311 0.608 0.0494 S100a4 NM_173028 0.607 0.000867 Vps13a NM_007674 0.606 0.0172 Cdx4 NM_148922 0.606 0.00211 Mdm1 NM_015737 0.605 0.0142 Galnt4 NM_026004 0.605 0.0011 Nt5c3 NM_023503 0.604 0.012 Ing2

109 NM_010290 0.604 0.00494 Gjd2 NM_198419 0.602 0.0214 Phactr1 NM_001033316 0.602 0.0203 Ffar3 NM_013885 0.602 0.00618 Clic4 ENSMUST0000002647 5 0.602 0.00217 Ddit3 ENSMUST0000002878 0 0.6 0.0122 Chac1 NM_021331 0.599 0.0213 G6pc2 NM_007444 0.599 0.0154 Amd2 NM_009218 0.599 0.00573 Sstr3 NM_177606 0.598 0.0349 Plekhh2 ENSMUST0000005959 5 0.597 0.0176 Prkca NM_130878 0.597 0.00252 Pcdh21 NM_008212 0.597 0.00111 Hadh ENSMUST0000003106 9 0.596 0.00495 Sepsecs NM_181751 0.595 0.026 Gpr119 NM_008855 0.595 0.0098 Prkcb1 ENSMUST0000004618 8 0.595 0.000207 EG328644 AK047968 0.594 0.0117 Zdhhc20 D82072 0.592 0.00528 Ptgds2 ENSMUST0000002049 7 0.59 0.00386 Aldh1l2 NM_175272 0.587 0.00744 Nav2 AK043588 0.586 0.0122 BC026590 NM_011864 0.586 0.0029 Papss2 ENSMUST0000003316 1 0.586 0.00257 Scnn1b NM_033525 0.584 0.0151 Npnt ENSMUST0000010292 5 0.584 0.00537 Uap1l1 ENSMUST0000004555 7 0.584 4.64E-05 Slc7a5 ENSMUST0000005270 0 0.581 0.039 Ffar1 ENSMUST0000000171 3 0.581 0.012 Gstt1 NM_026053 0.576 0.00658 Gemin6 NM_175007 0.575 0.0303 Amph ENSMUST0000010256 8 0.575 0.0156 Phactr4

110 NM_028247 0.574 0.00187 Slc16a10 NM_133995 0.573 0.0401 Upb1 NM_022030 0.573 0.0315 Sv2a NM_019516 0.573 0.0026 Lgals12 NM_008638 0.572 0.0005 Mthfd2 NM_027398 0.57 0.0101 Kcnip1 NM_018878 0.567 0.000501 Paxip1 ENSMUST0000002144 3 0.566 0.00814 Mthfd1 NM_015781 0.564 0.00738 Nap1l1 NM_010700 0.563 0.00442 Ldlr NM_178114 0.562 0.0094 Amigo2 NM_001080780 0.559 0.0115 Ret NM_025436 0.558 0.0112 Sc4mol XM_147850 0.557 0.0224 BC030046 AK009333 0.555 0.00714 2310014D11Rik NM_026950 0.553 0.0137 Ociad2 ENSMUST0000011270 1 0.552 0.028 Cdh7 AY566864 0.552 0.0158 A1cf XM_001475233 0.55 0.00495 4932438A13Rik ENSMUST0000006901 1 0.55 0.000204 Ang NM_153598 0.549 0.00256 Ugt2b34 NM_001033175 0.545 0.000454 Cln6 ENSMUST0000003320 1 0.542 0.00621 Anks4b NM_001081388 0.535 0.00255 Rimbp2 NM_011995 0.535 0.00079 Pclo XM_147850 0.531 0.00446 BC030046 NM_001081243 0.53 0.000405 Filip1 NM_133882 0.524 0.00763 C8b ENSMUST0000003219 8 0.523 0.0197 Usp18 NM_001081205 0.523 0.0047 Npal1 NM_009377 0.522 0.00237 Th NM_007514 0.521 0.0416 Slc7a2 NM_011299 0.521 0.00139 Rps6ka2 NM_001081262 0.517 0.0227 4932431H17Rik NR_002900 0.507 0.0153 Snora69 NM_145399 0.506 0.00214 Scgn NM_153420 0.505 0.00391 Acpl2

111 ENSMUST0000002586 6 0.505 0.00356 Vldlr NM_026185 0.501 0.00619 1300007F04Rik NM_001012723 0.5 0.0237 Wfdc16 NM_001005423 0.498 0.0074 Mreg NM_031197 0.498 0.0038 Slc2a2 NM_001003913 0.497 0.00231 Mars NM_021332 0.497 0.000388 Glp1r XM_911780 0.496 0.0323 EG626952 NM_001081014 0.491 0.000929 Dennd4c NM_031250 0.49 0.00179 Ucn3 NM_016717 0.489 0.0236 Scly NM_011374 0.482 0.0158 St8sia1 XM_147850 0.476 0.00707 BC030046 NM_021883 0.475 0.0394 Tmod1 NM_001080815 0.475 0.000876 Gipr ENSMUST0000005812 6 0.473 0.0174 Nr1h4 NM_022312 0.472 0.0046 Tnr NM_011843 0.472 0.00255 Mbc2 NM_018869 0.468 0.00426 Grk5 ENSMUST0000004423 1 0.467 0.00773 Serpina10 NM_183168 0.456 0.00771 P2ry6 NM_016863 0.451 0.000185 Fkbp1b NM_020279 0.445 0.00234 Ccl28 ENSMUST0000006030 4 0.443 0.00332 Tox3 NM_001081346 0.438 0.0305 Rtkn2 NM_021455 0.438 0.00587 Mlxipl NM_146214 0.432 0.00642 Tat NM_015744 0.426 0.00227 Enpp2 ENSMUST0000004938 9 0.418 0.0035 Zdhhc2 NM_018857 0.414 0.039 Msln ENSMUST0000007229 9 0.399 0.00873 Vsnl1 NM_019775 0.392 0.0264 Cpb2 NM_011832 0.363 0.00231 Insrr NM_008509 0.322 0.00291 Lpl NM_021391 0.322 0.000233 Ppp1r1a NM_026853 0.319 0.0122 Asb11

112 NM_026935 0.285 0.017 Sult1c2 NM_007843 0.226 0.0141 Defb1

113 CHAPTER 3

The Partial Rescue of MODY4 Phenotypes by Sox17

Diva Jonatan1, Jason Spence2, and James Wells1

1. Division of Developmental Biology, Cincinnati Children’s Hospital Medical

Center, 2. Division of Metabolism, Endocrinology, and Diabetes, University of

Michigan.

114 Summary.

Mature Onset Diabetes of the Young (MODY) is a monogenic type of diabetes that has been linked to single gene mutations in several transcription factors and glucose signaling molecules expressed in cells. The resulting cell dysfunction in MODY is thought to be due to disruption of cell regulatory network. In this chapter, we will focus our study on how the transcription factor

Sox17 impacts cell function in the context of the MODY4/Pdx1 diabetic background.

In Chapter 2, we demonstrated that in non-diabetic mice, Sox17 can modulate insulin processing and secretion and regulate several genes involved in these pathways, such as Pdx1, GLP1R, ATF4, and WFS1; some of these genes have been implicated in the cell pathology observed in in MODY4 mice. These studies suggest that Pdx1 and Sox17 may regulate the same biological pathway in the cell and that Sox17 might be able to positively impact the MODY4 pathology.

To test this we utilized mice with a ±2-fold increased of Sox17 islet expression in a MODY4 background. Transgenic expression of Sox17 in adult  cells was sufficient to temporarily restored euglycemia in normally hyperglycemic

MODY4 animals. Moreover, Sox17 was also able to reverse the defective islet architecture found in MODY4 mice, but did not affect MODY4-associated glucose intolerance. Taken together, this study suggested that physiological dose of

Sox17 plays a key role in stabilizing basal glucose homeostasis and maintenance of islet morphology. Future studies will focus on identifying these

115 Sox17-regulated pathways and if they can be therapeutically manipulated to improve cell function in the context of diabetes.

Introduction.

Mature Onset Diabetes of the Young (MODY) is a monogenic type of diabetes that is characterized by early age onset and autosomal dominant transmission. It has been linked to single gene mutations in several transcription factors and glucose signaling molecules expressed in cells. Despite of the heterogeneity of the clinical phenotypes between different subtypes of MODY patients, the primary cause of diabetes in all patients appears to be cell dysfunction. This common phenotype is thought to be due to the fact that MODY genes have a common function in a cell regulatory network and that one mutation can impact the entire network. In this chapter, we will focus our study on how the transcription factor Sox17 impacts the MODY network in direct relation to the MODY4 gene, which encodes the transcription factor Pdx1.

Pdx1 is a homeobox gene that is expressed developmentally throughout pancreatic development and in adult endocrine cells. Prior to the onset of pancreatic bud formation, Pdx1 is expressed in foregut epithelium. At later stages, it is abundantly expressed in proliferating pancreatic progenitor stage. Postnatally,

Pdx1 is expressed at high levels in and  cells, with low levels of expression in subpopulations of acinar and ductal cells1-3. Mice and humans lacking functional

Pdx1 protein have pancreatic agenesis resulting from a block in pancreatic branching morphogenesis and differentiation4-6. Pdx1 haploinsufficiency in

116 patients develop early onset diabetes (MODY4) with severe impairment of insulin secretion and enhanced insulin sensitivity7. Mutations in Pdx1 are also associated with predisposition to late-onset type 2 diabetes8.

There are several studies of Pdx1/MODY4 in the mouse that describe the impact of Pdx1 loss-of-function on adult cell function. One study used a tetracycline-inducible Pdx1 transgene to rescue the Pdx1 loss-of-function lethal phenotye to generate adult mice (Pdx1-tTA;tetO-Pdx1 mice). Subsequent loss of

Pdx1 in adults resulted in animals with impaired glucose tolerance, decreased insulin expression, absence of glucose transporter protein (Glut2) and an impaired islet architecture as evidenced by the presence of alpha cells (glucagon positive cells) in the core of the islets rather than at their normal location on the islet periphery9. This led to diabetes within 14 days of Pdx1 inactivation in the adult mice. Another experimental approach generated conditional Pdx1 null cells in adult mice using a RIP (Rat Insulin )-Cre driver. In these animals, where Pdx1 deletion became prominent starting in between 3-5 weeks after birth,

Glut2 and insulin levels were similarly reduced and glucagon expression was increased10. In particular, the ratio between insulin- to glucagon-expressing cells in these RIP-Cre;Pdx1fl/fl mice was 1:1, whereas in wild type mice it is usually 5:1.

RIP-Cre;Pdx1fl/fl mice develop diabetes at 17-19 weeks of age.

While these studies showed a profound and rapid phenotype in animals lacking Pdx1, studies of Pdx1 haploinsufficiency in mouse suggested a much milder phenotype that develops over time. In one study, in which exon 2 was replaced with the neomycin-resistance gene, data showed that Pdx1+/- mice (8-12

117 weeks of age) had worsening glucose tolerance, defective islet architecture, increased islet microvasculature, reduced insulin pancreatic content and reduced

cell mass as animals aged11. Perfused pancreas experiment showed reduced insulin secretion in response to glucose, KCl, and arginine. Surprisingly, insulin secretion in perifusion or static incubation of glucose and other secretagogues in isolated islets showed no difference between genotypes and so does the calcium signaling-dependent exocytosis analysis, suggesting that the insulin secretory physiology is normal in these Pdx1+/- isolated islets. This discrepancy in the results from perfused pancreas and perifusion islets experiments showed that the heterogeneous responses of single cells to glucose may not always be the same as predicted from the activity of the whole islets. This further suggests that studies from different techniques of experiments need to be taken into account to explain the complexity of the Pdx1+/- phenotype. With that in mind, this particular study also showed Pdx1+/- islets to be fragmented, smaller, and had increased apoptosis susceptibility when cultured in basal glucose concentrations.

Consistent with this, BCl-XL and Bcl2 were reduced in Pdx1-haploinsufficient islets and there was an increased in active-caspase 3 levels in vivo. Taken together, Johnson, et al concluded that islet apoptosis preceded a defect in glucose-stimulated insulin secretion that in turn led to diabetes11.

Another study of Pdx1 haploinsufficient mouse, using an allele of Pdx1 in which LacZ was knocked into the coding region, showed a similar impaired glucose intolerance that was worsened with aged as well as reduced Glut2 levels12. However, these Pdx1+/LacZ had normal pancreatic insulin content and

118 normal islet architecture. Perfusion experiments with isolated islets suggested a defect in mitochondria metabolism and in the K+ATP-channel dependent stimulated-secretion coupling. This particular study showed that the GLP-1- stimulated insulin secretory pathway was not dependent on Pdx1. This is in contrast to the cell specific deletion of Pdx1 study by Li, et al where GLP-1 signaling was attenuated13. Consistent with this idea, expression of dominant negative forms of Pdx1 in the INS1 cell line also supported a role for Pdx1 in

GLP1R and cAMP signaling14. In these experiments, inhibition of Pdx1 caused a

90% reduction of GLP1R expression and reduced intracellular cAMP levels, with normal glucose utilization, mitochondria oxidation, and calcium-induced exocytosis. As mentioned earlier, it is not clear which model most closely resembles the MODY4 patients; therefore, data from these various studies should be taken into account when explaining the MODY4 phenotype.

Our studies, detailed in the previous chapter, support a model in which

Sox17 controls multiple aspects of proinsulin processing, trafficking, and secretion in mature cells through transcriptional regulation of genes involved in these processes. In particular, pancreatic loss of Sox17 showed severe distention of ER, pre-Golgi, Golgi, and mitochondria organelles by EM, insulin trafficking defects, and increased proinsulin protein levels in the islets. Animals lacking pancreatic Sox17 became more glucose intolerant after being placed on high fat diet as compared to Sox17 sufficient animals. We have transcriptional data that Sox17 regulates expression of Pdx1, GLP1R, GIPR, and key ER genes expressions involved in the unfolded protein response (UPR) pathway, such as

119 ATF4 and WFS1. Study by Sachdeva, et al suggested that Pdx1 regulates cell susceptibility to ER stress15. This study showed that after five months of high fat diet, glucose intolerance seen in Pdx1+/- mice were worsened. Compensatory hyperinsulinemia and of cell mass were decreased, with no difference in body weight gain between genotypes. The ER of the cell was severely distended and the ER-related UPR genes, including Atf4 and WFS1, were downregulated in these mice. It was characterized in this study that Atf4 and WFS1 were direct targets of Pdx1.

Taken all these studies together, these data suggest that Pdx1 and Sox17 may regulate some of the same biological processes in cells. We therefore wanted to explore the possibility that Sox17 might be used to positively impact pathologic phenotypes observed in mature cells in the context of diabetic mouse model, in particular MODY4. We were fortunate in that our Sox17 overexpression model using the MODY4 driver mouse model (Pdx1-tTA hemizygote mice) resulted in roughly 2 fold of Sox17 expression, a more physiological overexpression range. Transgenic overexpression of Sox17 (Pdx1- tTA;tetO-Sox17) in adult  cells is sufficient to temporarily restore euglycemia in normally hyperglycemic MODY4 animals. Preliminary data showed that the reduced plasma insulin level in MODY4 mice was also transiently rescued by transgenic Sox17 overexpression. Lastly, Sox17 overexpression also corrects the aberrant islet architecture that is a hallmark of islet dysfunction in MODY4 mice. However, Sox17 overexpression failed to rescue glucose intolerance in these MODY4 background and there was no difference in islet cell proliferation or

120 apoptosis seen across genotypes. This study shows that Sox17-regulated pathways can ameliorate some of the diabetic symptoms in MODY4 animals, suggesting that these pathways might be viable therapeutic targets to improve cell function. Future molecular studies need to be carried out to identify these pathways.

Materials and Methods.

Mice.

All mice used in these studies; Pdx1-tTA, tetO-Sox17; have been previously described and were housed at the Cincinnati Children’s Hospital Research

Foundation mouse facility9,16. All animal procedures were approved under institutional protocols. For all experiments, sixteen- to twenty-week-old adult mice were used, unless otherwise noted. We regulated Sox17 transgene expression with doxycycline as previously described17. Briefly, to keep the Sox17 transgene off, Pdx1-tTA;tetO-Sox17animals were maintained on doxycycline (Dox) chow, which maintains repression of the transgene. We removed doxycycline (Dox) chow to induce expression of the Sox17 transgene as indicated in Figure 1. All mice were originally maintained on an outbred background.

Immunofluorescence and confocal microscopy analysis.

Tissues were prepared and stained as previously described17. Images were acquired using confocal microscopy (Zeiss LSM 5.10 with 40x PlanApo NA 1.4

121 objective at Nyquist limit). See Chapter 2, Supplementary Table 1, for a list of primary and secondary antibodies used in these studies.

Islet counting analysis.

Four mice per genotype were used to analyze the islet architecture, proliferation, and apoptosis. From each mouse, three different sections of the pancreas were analyzed, and three sizes of islets were included, large (±110 < μm), medium

(±60-100 μm), and small (< ±50 μm) islets. For each size, three islets were counted. Each section was then stained with the appropriate antibodies, glucagon and insulin, BrdU, or active caspase-3 with nuclear staining. For the islet architecture analysis, the number of glucagon positive at the core of the islet was counted and divided by the total number of glucagon positive cells in the islet to get a percent core glucagon cells. For the proliferation analysis, 50 mg

BrdU per kg of mouse was given before mice were sacrificed. The number of

BrdU positive cells were counted in each islet and divided by the total number of cells in that islet section. For the apoptosis analysis, the number of cells expressing active Caspase-3 were counted and divided by the total number of islet cells in that particular section. Additionally, in order to measure the number of islets found in the whole pancreas, pancreas was sectioned all the way through and every fifth sections, two serial sections were stained with insulin and hematoxylin for contrast. Different sizes of islets were counted from these two serial sections and were averaged.

122 Blood glucose and plasma insulin analysis.

Blood glucose and insulin levels at resting and glucose challenge conditions were analyzes as described in Chapter 2.

Statistical analysis.

All the data are expressed as mean ± SEM, and Student t-tests were used for statistical analysis.

Results.

Sox17 overepression in the context of MODY4 background temporarily rescued resting hyperglycemia.

MODY4 symptoms include hyperglycemia, reduced serum insulin, impaired glucose tolerance, and disrupted islet architecture10-12,14,18. Various molecular factors were found to contribute to the MODY4 phenotypes, including a reduction in GLP1R signaling followed by reduced cAMP level14. Given that we have demonstrated that transgenic expression of Sox17 in wild type animals stimulates insulin secretion and can impact GLP1R expression, we hypothesized that misexpression of Sox17 in the MODY4 disease background mice would help alleviate some of the disease phenotypes. To do this, we generated a tetracycline-regulated model whereby in our hands, we found modest upregulation of Sox17 levels in the context of a MODY4 background. We used the Pdx1-tTA knock in mice as the MODY4 model and we crossed these with a tet-inducible Sox17 mice (tetO-Sox17) to allow for Sox17 overexpression in

123 MODY4 background mice (Figure 1A). Both control and experimental mice were maintained on Dox from the time of breeding until weaning in order to keep the transgene repressed. To induce expression of Sox17, the Sox17 transgene animals were taken off of the Dox-containing diet until time of analysis, unless otherwise stated. (Figure 1A). The tetracycline-regulated approach allowed us to exquisitely control the expression of the Sox17 transgene, shown in Figure 1B

(mice were off Dox diet at embryonic days 12.5 (e12.5), pancreas was collected at 8 weeks of age). Adult mice were analyzed for several cell physiological assays. Detailed of the experimental timeline is depicted in Figure 1C.

Since the Pdx1-tTA mice had not yet been studied as a MODY model, we first confirmed that these mice had higher blood glucose level compared to wildtype littermates (Figure 1D, E). We then transgenically expressed Sox17 in the MODY4 background (Pdx1-tTA;tetO-Sox17 mice) for 13-17 weeks and found that this restored the resting blood glucose to wildtype levels (Figure 1D, wt, n=4;

Pdx1-tTA, n=6; Pdx1-tTA;tetO-Sox17, n=5, mice were ±16-20 weeks of age).

This was correlated with the rescued plasma insulin levels (Figure 1G, n=3 for

Pdx1-tTA;tetO-Sox17 mice – off Dox; n=3 for Pdx1-tTA – n=2 on Dox and n=1 off

Dox; n=3 for tetO-Sox17 – n=2 on Dox and n=1 off Dox, Sox17 was overexpressed for ±12 weeks, starting at 8 weeks of age). It was not clear if the restoration of normal blood glucose and serum insulin levels was an acute effect of Sox17 in stimulating insulin secretion, or a chronic effect of long-term Sox17 expression for 13-17 weeks. We investigated the acute effects of Sox17 on hyperglycemia and found that within 8 days of Sox17 overexpression, normal

124 blood glucose levels were achieved in MODY4 animals (Figure 1E, wt, n=4;

Pdx1-tTA, n=3; Pdx1-tTA;tetO-Sox17, n=4). However, it is important to note that chronic long term expression of Sox17 was detrimental to the mice, they became diabetics (Figure 1F, wt, n=4; Pdx1-tTA, n=5; Pdx1-tTA;tetO-Sox17, n=3). Sox17 overexpression alleviated MODY4 phenotype temporarily, but as Sox17 overexpression was accumulated overtime, it increased resting blood glucose level similar to what happened in Ins-rtTA;tetO-Sox17 mice. Therefore, precise level of Sox17 dose needs to be considered carefully to ensure proper cell homeostasis. In addition, it is also important to note that in our hands, the blood glucose level of Pdx1-tTA mice returned to normal level as the mice aged to about 32-36 weeks (Figure 1F). These mice appeared to lose their MODY4 phenotype at this late time point.

We then focused our analysis to the 13-17 weeks of Sox17 overexpression. Using these mice, we challenged the mice using glucose tolerance test to test how efficiently animals could clear a bolus of injected glucose. Sox17 overexpression was not able to compensate for the glucose intolerance defect of the Pdx1-tTA mice (Figure 1H, wt, n=4; Pdx1-tTA, n=6;

Pdx1-tTA;tetO-Sox17, n=5). These data suggested that Sox17 expression can transiently restore glucose homeostasis in MODY4 mice during resting conditions; however, cells are still unable to handle extreme spike in blood glucose levels brought about by glucose challenge.

125 Sox17 overexpression rescued MODY4 mice disrupted islet architecture.

One of the hallmark pathologies in MODY4 mice is the disruption of normal islet architecture as measured by an increase in glucagon positive alpha cells that are no longer predominantly localized to the periphery of the islet10-12,18.

Analysis of Pdx1-tTA animals indicated that this MODY4 model also has a disrupted islet architecture, with an increase in glucagon-expressing alpha cells in the core of the islet (Figure 2 compare G to H). While preliminary data showed that a transient burst of transgenic Sox17 expression for 5 days to have no effect on islet architecture, prolonged expression of Sox17 for 13-17 weeks in Pdx1-tTA islet background resulted in a significant restoration of islet architecture as compared to wildtype controls (Figure 2G, H, I, quantitated in 2J, n=4 per genotype). Taken together these data suggested that prolonged Sox17 overexpression was sufficient to partially restore normal islet architecture in

MODY4 mice. Previous studies by Johnson, et al. showed that the MODY4 phenotype coincided with increased in cell apoptosis11. However, analysis of islet apoptosis and proliferation showed no difference in apoptosis or proliferation between any of the genotypes at this end timepoint (Figure 2K-R, n=4 per genotype, using 13-17 weeks Sox17 overexpressing islets), suggesting that this

MODY4 model does not recapitulate the apoptotic phenotype of other MODY4 models. Further apoptosis and proliferation analysis studies using different time points during Sox17 overexpression still need to be performed in order to show definitively that cell death and cell proliferation were not affected in this MODY4

126 mouse model or in the restoration process of MODY4 phenotypes in the Sox17 overexpressing mice.

Sox17 altered the distribution of islet sizes and cell-cell adhesion contacts of the MODY4 mice.

We investigated mechanisms by which Sox17 expression might rescued the islet architecture of the MODY4 mice. To do this, we investigated changes in cell adhesion, shape, size and/or changes in the overall sizes of islet morphology throughout the whole pancreas. -catenin and E-cadherin immunostaining of the islets showed changes in the cell shape of Sox17 overexpressing cells, suggesting that the cell adhesion contacts between the cells were altered (Figure

3A-I). In general, cells were more spindle shaped and less spherical in Pdx1- tTA;tetO-Sox17 animals, as compared to control or Pdx1-tTA animals. This suggested that the ability of Sox17 to restore MODY4 islet architecture may involve induction of a cell shape change. This may not be a general mechanism in how proper islet architecture is regulated as we did not observe any changes in the cell shape of the MODY4 abnormal islet architecture. However, further studies need to be performed to investigate whether Sox17-induced cell shape changes play a role in maintaining islet architecture to ensure localization of non-

cells to the periphery and cells localization toward the center of the islets.

We also explored the impact of Sox17 on islet size distribution throughout pancreas. Preliminary data showed that wild type and MODY4/Pdx1-tTA mice had relatively comparable islet size distribution (Figure 3J, n=1). However, the

127 average size of islets was increased in the Sox17-overexpressing mouse (Sox17 was overexpressed for ±22 weeks), relative to wildtype and Pdx1-tTA mice

(Figure 3J, n=1). There were more large islets found in Sox17-overexpressing mouse. Taken together these data suggest that Sox17 expression induced a cell shape change and also altered islet size distribution. This is an interesting observation given the published reports that islets are capable of replicating by islet fission19,20, suggesting that Sox17 expression was regulating this process.

Discussion.

Previous studies showed that expression of a dominant negative of Pdx1 in INS1 cell line resulted in 80% decrease of PC1/3 enzyme, a proinsulin processing enzyme14. This correlated with a delay in proinsulin conversion by

60%; however, the cell line also showed a decreased in growth hormone secretion, which is not known to be processed by the PC1/3 enzymes. These data suggested that the proinsulin processing defect could be due to an insulin secretion defect, given that these cells also had reduced cAMP levels and 90% reduction of GLP1R expression. Given that our data in chapter 2 indicates that

Sox17 promotes insulin trafficking throughout the secretory pathway and affects insulin processing in a manner that is independent of PC1/3, we hypothesized that Sox17-regulated pathways may alleviate the proinsulin processing and secretion defects that are Pdx1-dependent and help rescue the glucose homeostasis impairment in MODY4 mice.

128 Taken together, our data suggested that a burst of Sox17 was found to have therapeutic benefit and restored normoglycemia in a MODY4 background.

13-17 weeks of transgenic Sox17 overexpression transiently reversed the hyperglycemia, hypoinsulinemia, and restored normal islet architecture in

MODY4 animals. However, Sox17 overexpression did not reverse the glucose intolerance caused by Pdx1 haploinsufficiency. There were no differences in islet cell proliferation or apoptosis between wild type, MODY4 or Sox17-expressing animals (phenotype summary table is shown in Figure 3K). However, extensive prolonged Sox17 overexpression was detrimental to cell, similar to the overexpression effect of Sox17 in the InsrtTA;tetO-Sox17 mice discussed in

Chapter 2.

Previous studies showed that compensatory changes of cell mass exist to maintain glucose homeostasis. For example, cell mass is known to be linearly related to body weight21. During pregnancy, pancreatic cell mass had a

50% increase due to larger islets, increased cell replication, and hyperthrophy of cells22-24. Furthermore, compensatory increase in cell mass also occurred in obese patients and mice with insulin resistance (insulin receptor, IRS-1+/- mice)25,26. In our study, preliminary data suggested that Sox17 transiently rescued the MODY4 phenotype in part by playing a role in changing the islet cell- to-cell contacts, affecting the overall islet cell shape, and in changing the islet size distribution throughout the whole pancreas. The cell shape of Sox17 overexpressing islets appeared elongated, and less spherical in shape.

Preliminary data showed that there were a higher percentage of large islets in the

129 Sox17-overexpressing mouse compared to wildtype and Pdx1-tTA mice. This could be due to a role in Sox17 in regulating islet mass or in regulating islet fission, which is a proposed mechanism for islet replication in postnatal pancreas19,20. Previous study showed that islet fission is suggested to play an important role in expanding islet number during neonatal period. It was thought that expanding islet distribution via islet fission optimized the association of islet cells with blood vessels19,20. Furthermore, it was shown that islet cell size differences is known to reflect the changes in cell mass; however, it is rarely analyzed due to the difficulties in measuring it accurately because of the nonrandom location of different islet size21. Further studies need to be performed to determine the role of Sox17 in regulating islet fission as a compensatory mechanism in these MODY4 background mice.

In summary, our studies suggested that early burst of Sox17 play a key role in stabilizing basal glucose homeostasis and islet morphology. Further studies need to be performed to analyze the molecular mechanism in which

Sox17 regulate these compensatory pathways in the context of a MODY4 background. This study, in combination with the loss- and gain-of-function studies of Sox17 in a wildtype background, will also narrow down which direct factors are regulated by Sox17 and/or Pdx1.

Acknowledgements

We thank James Wells, Jeff Whitsett, Aaron Zorn, Jonathan Katz, Gail Deutsch,

Jason Spence, Matt Kofron, Debora Sinner, Alex Lange, and the Wells

130 and Zorn labs for input on the work. We also thank Raymond MacDonald

(University of Texas Southwestern) for the Pdx1tTA mice.

Sources of funding

This work was supported by JDRF 2-2003-530 and  Cell Consortium-BCBC

(Grant #U01DK072473).

References.

1Guz, Y.et al. Expression of murine STF-1, a putative insulin gene transcription factor, in cells of pancreas, duodenal epithelium and pancreatic exocrine and endocrine progenitors during ontogeny. Development 121, 11-18 (1995). 2 Li, H., Arber, S., Jessell, T. M. & Edlund, H. Selective agenesis of the dorsal pancreas in mice lacking homeobox gene Hlxb9. Nat Genet 23, 67- 70, doi:10.1038/12669 (1999). 3Wu, K. L.et al. Hepatocyte nuclear factor 3 is involved in pancreatic - cell-specific transcription of the pdx-1 gene. Mol Cell Biol 17, 6002-6013 (1997). 4 Jonsson, J., Carlsson, L., Edlund, T. & Edlund, H. Insulin-promoter-factor 1 is required for pancreas development in mice. Nature 371, 606-609, doi:10.1038/371606a0 (1994). 5 Offield, M. F. et al. PDX-1 is required for pancreatic outgrowth and differentiation of the rostral duodenum. Development 122, 983-995 (1996). 6 Stoffers, D. A., Ferrer, J., Clarke, W. L. & Habener, J. F. Early-onset type- II diabetes mellitus (MODY4) linked to IPF1. Nat Genet 17, 138-139, doi:10.1038/ng1097-138 (1997). 7 Clocquet, A. R. et al. Impaired insulin secretion and increased insulin sensitivity in familial maturity-onset diabetes of the young 4 (insulin promoter factor 1 gene). Diabetes 49, 1856-1864 (2000). 8 Hani, E. H. et al. Defective mutations in the insulin promoter factor-1 (IPF- 1) gene in late-onset type 2 diabetes mellitus. J Clin Invest 104, R41-48, doi:10.1172/JCI7469 (1999). 9 Holland, A. M., Gonez, L. J., Naselli, G., Macdonald, R. J. & Harrison, L. C. Conditional expression demonstrates the role of the homeodomain transcription factor Pdx1 in maintenance and regeneration of -cells in the adult pancreas. Diabetes 54, 2586-2595, doi:54/9/2586 [pii] (2005). 10 Ahlgren, U., Jonsson, J., Jonsson, L., Simu, K. & Edlund, H. -cell- specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the -

131 cell phenotype and maturity onset diabetes. Genes Dev 12, 1763-1768 (1998). d11 Johnson, J. D. et al. Increased islet apoptosis in Pdx1+/- mice. J Clin Invest 111, 1147-1160, doi:10.1172/JCI16537 (2003). 12 Brissova, M. et al. Reduction in pancreatic transcription factor PDX-1 impairs glucose-stimulated insulin secretion. J Biol Chem 277, 11225- 11232, doi:10.1074/jbc.M111272200 M111272200 [pii] (2002). 13 Li, Y. et al. -Cell Pdx1 expression is essential for the glucoregulatory, proliferative, and cytoprotective actions of glucagon-like peptide-1. Diabetes 54, 482-491, doi:54/2/482 [pii] (2005). 14 Wang, H. et al. Suppression of Pdx-1 perturbs proinsulin processing, insulin secretion and GLP-1 signalling in INS-1 cells. Diabetologia 48, 720- 731, doi:10.1007/s00125-005-1692-8 (2005). 15 Sachdeva, M. M. et al. Pdx1 (MODY4) regulates pancreatic cell susceptibility to ER stress. Proc Natl Acad Sci U S A 106, 19090-19095, doi:0904849106 [pii] 10.1073/pnas.0904849106 (2009). 16 Park, K. S., Wells, J. M., Zorn, A. M., Wert, S. E. & Whitsett, J. A. Sox17 influences the differentiation of respiratory epithelial cells. Dev Biol 294, 192-202, doi:S0012-1606(06)00141-2 [pii] 10.1016/j.ydbio.2006.02.038 (2006). 17 Spence, J. R. et al. Sox17 regulates organ lineage segregation of ventral foregut progenitor cells. Dev Cell 17, 62-74, doi:S1534-5807(09)00214-7 [pii] 10.1016/j.devcel.2009.05.012 (2009). 18 Kulkarni, R. N. et al. PDX-1 haploinsufficiency limits the compensatory islet hyperplasia that occurs in response to insulin resistance. J Clin Invest 114, 828-836, doi:10.1172/JCI21845 (2004). 19 Jo, J. et al. Formation of pancreatic islets involves coordinated expansion of small islets and fission of large interconnected islet-like structures. Biophys J 101, 565-574, doi:S0006-3495(11)00772-7 [pii] 10.1016/j.bpj.2011.06.042 (2011). 20 Seymour, P. A., Bennett, W. R. & Slack, J. M. Fission of pancreatic islets during postnatal growth of the mouse. J Anat 204, 103-116, doi:10.1111/j.1469-7580.2004.00265.x JOA265 [pii] (2004). 21 Bonner-Weir, S. -cell turnover: its assessment and implications. Diabetes 50 Suppl 1, S20-24 (2001). 22 Parsons, J. A., Bartke, A. & Sorenson, R. L. Number and size of islets of Langerhans in pregnant, human growth hormone-expressing transgenic, and pituitary dwarf mice: effect of lactogenic hormones. Endocrinology 136, 2013-2021 (1995). 23 Parsons, J. A., Brelje, T. C. & Sorenson, R. L. Adaptation of islets of Langerhans to pregnancy: increased islet cell proliferation and insulin

132 secretion correlates with the onset of placental lactogen secretion. Endocrinology 130, 1459-1466 (1992). 24 Scaglia, L., Smith, F. E. & Bonner-Weir, S. Apoptosis contributes to the involution of cell mass in the post partum rat pancreas. Endocrinology 136, 5461-5468 (1995). 25 Kloppel, G., Lohr, M., Habich, K., Oberholzer, M. & Heitz, P. U. Islet pathology and the pathogenesis of type 1 and type 2 diabetes mellitus revisited. Surv Synth Pathol Res 4, 110-125 (1985). 26 Bruning, J. C. et al. Development of a novel polygenic model of NIDDM in mice heterozygous for IR and IRS-1 null alleles. Cell 88, 561-572, doi:S0092-8674(00)81896-6 [pii] (1997).

133 Figure Legends.

Figure 1. Sox17 overepression in the context of MODY4 background rescued resting hyperglycemia.

A) Schematic representation of the Pdx1-tTA;tetO-Sox17 mice. B)

Immunofluorescence stainings of Sox17 and Pdx1, showing elevated Sox17 expression in the Pdx1-tTA;tetO-Sox17 mice. C) Experimental timeline used in this study showing the age of mice and its corresponding period of Sox17 overexpression days at times of analysis. D) Resting blood glucose levels of 16-

20 weeks old mice, 4 months after Sox17 was overexpressed, showed Sox17 overexpression rescued MODY4 (Pdx1-tTA) hyperglycemia (1 asterisk (*) indicates comparison between wildtype (n=4) vs Pdx1-tTA (n=6) mice with p- value  0.01; 2 asterisk (**) shows comparison between Pdx1-tTA vs Pdx1- tTA;tetO-Sox17 (n=5) mice with p-value  0.01). E) Resting blood glucose of 16-

20 weeks old mice, 8 days after Sox17 was overexpressed, showed similar rescue. (1 asterisk (*) indicates comparison between wildtype (n=4) vs Pdx1-tTA

(n=3) mice with p-value  0.01; 2 asterisk (**) shows comparison between Pdx1- tTA vs Pdx1-tTA;tetO-Sox17 (n=4) mice with p-value  0.01). F) Resting blood glucose of 32-36 weeks old mice, 29-33 weeks of Sox17 overexpression, showed diabetes level in these Sox17 overexpressing mice. (1 asterisk (*) indicates comparison between wildtype (n=4) vs Pdx1-tTA;tetO-Sox17 (n=3) mice with p-value  0.01; 2 asterisk (**) shows comparison between Pdx1-tTA

(n=5) vs Pdx1-tTA;tetO-Sox17 mice with p-value  0.01). G) Resting plasma insulin of 5 months old mice, after 3 months of Sox17 overexpression, showed

134 rescued of MODY4 hypoinsulinemia, albeit not statistically significant. H) Glucose tolerance test showed that both MODY4 and Sox17 overpressing mice were glucose intolerant (1 asterisk (*) indicates comparison between wildtype (n=4) vs

Pdx1-tTA (n=6) mice with p-value  0.05; 2 asterisk (**) shows comparison between wildtype vs Pdx1-tTA;tetO-Sox17 (n=5) mice with p-value  0.05).

Figure 2. Sox17 overexpression rescued MODY4 disrupted islet architecture.

A-C) Immunofluorescence stainings for insulin, glucagon, and somatostatin showed Pdx1-tTA;tetO-Sox17 (on Dox, Sox17 off) islets to have similar disrupted islet architecture as MODY4 islets. D-F) Preliminary data showed that islet architecture was not rescued after early Sox17 overexpression of 5 days

(wildtype-n=3, Pdx1-tTA-n=1, Pdx1-tTA-n=1). G-I) Immunofluorescence stainings for insulin, glucagon, and somatostatin showed Sox17 partially rescued disrupted islet architecture of MODY mice after 13-17 weeks of overexpression, quantitated in 2J (1 asterisk (*) indicates comparison between wildtype (n=4) vs Pdx1-tTA

(n=4) mice with p-value  0.01; 2 asterisk (**) shows comparison between Pdx1- tTA vs Pdx1-tTA;tetO-Sox17 (n=4) mice with p-value  0.01). K-N) This rescued was not due to change in cell proliferation, measured by BrdU incorporation (n=4 per genotype). Islets stained with BrdU antibody. O-R) Apoptosis level measured by active caspase-3 immunostaining also showed no difference between genotypes (n=4 per genotype). Circles in images pointed out the positive-stained cells.

135 Figure 3. Sox17 altered the distribution of islet sizes and cell-cell adhesion contacts of the MODY4 mice.

A-I) Immunofluorescence stainings for glucagon, Pdx1, E-cadherin, and - catenin showed changes in cell shape of the islets of Pdx1-tTA;tetO-Sox17 after

13-17 weeks of overexpression. J) Quantitation of percent size distribution of the islets in the pancreas. K) Summary table of different phenotypes of MODY4 and its rescued phenotypes in response to Sox17 overxexpression.

136 137 138 139 140 141 142 143 Chapter 4.

Summary and Discussion

143 Summary and Discussion

Major Findings.

The study detailed in Chapter 2 was the first to describe a model whereby Sox17 regulates insulin trafficking and secretion in the pancreatic cells. Perturbed

Sox17 level in the islet leads to a prediabetic phenotype shown in the disrupted islet proinsulin:insulin ratio, altered organelle morphology, and further glucose misregulation in response to high fat diet. These data also suggested that Sox17 overexpression alters genes involved in insulin secretion pathway and that it plays a role in modulating cell function to regulate changes in insulin subcellular trafficking. Furthermore, the study in Chapter 3 demonstrated that a more physiological dose of transgenic overexpression of Sox17 in adult  cells is sufficient to temporarily alleviate MODY4 diabetes phenotypes. Data showed that

Sox17 overexpression transiently rescued MODY4 hyperglycemia and disrupted islet architecture. Together, these data are the first to show that Sox17 regulates insulin trafficking process and islet morphology maintenance in the adult pancreas that are important to ensure proper glucose homeostasis.

Sox17 in normal cells.

The studies detailed here demonstrate that Sox17 is a key factor in the insulin trafficking process, and perturbation of Sox17 level resulted in mice with prediabetes. Unfortunately, 79 million people are considered to be prediabetics in the United States alone (2011 National Diabetes Fact Sheet, released by the

American Diabetes Association). Prediabetes phase is in part characterized by

144 higher blood glucose level that is not high enough to be diagnosed as diabetes yet, increased plasma ratio of proinsulin:insulin, and dilated ER morphology, which together suggest that defects in proinsulin processing contribute to the progression of diabetes1-6. The loss of function study detailed in Chapter 2 demonstrates that losing Sox17 altered proinsulin prohormone processing in particular by disrupting the proinsulin trafficking through the secretory organelle machinery (Figure 1). This was evident in the accumulation of proinsulin in the dilated ER. These mice were susceptible to high fat diet-induced diabetes shown by glucose misregulation in fasting and challenge states, confirming that these

Sox17-paLOF mice were prediabetics. Moreover, the gain of function study suggests that Sox17 is sufficient to directly impact insulin trafficking through the secretory pathway, resulting in reduced proinsulin level in the islet cell and continued precocious secretion of improperly processed proinsulin to the plasma

(Figure 1). This eventually leads to diabetes in the mice. This study also demonstrates that Sox17 modulates several genes that are intricately involved in orchestrating the insulin secretory pathway (Chapter 2, Figure 7B). Further studies are required to determine the mechanism in which Sox17 directly or indirectly regulate any of these genes transcriptionally and post-translationally. It is hoped that a thorough understanding of the molecular mechanism of how proinsulin is processed, trafficked, and secreted in the context of prediabetes will allow this knowledge to be exploited therapeutically to either delay or prevent diabetes progression in patients. For example, we may target potential downstream effectors of Sox17 that were altered in our islet microarray analysis,

145 such as EphA5 receptor tyrosine kinase. EphA5 was elevated by 1.44 fold in

Sox17 overexpressing islets. EphA5 is highly expressed in the endocrine cells compare to the surrounding exocrine tissue in mouse and human pancreas7. The

EphA forward signaling was suggested to be involved in insulin secretion suppression under basal condition, which under glucose stimulatory condition, was attenuated by the phosphorylation of EphA and thus, allows ephrin-A reverse signaling to be activated and enhanced insulin secretion7. Konstantinova, et al. suggested that potential drugs that target the Eph-Ephrin signaling pathway in cells might circumvent the hypoglycemic side effects that most current diabetic drugs have. Thus, we may make use of these downstream effectors to impact insulin secretion in diabetes context.

Sox17 in pathological cells.

The studies of Sox17 in a MODY4 diabetic mouse background suggested that

Sox17 is able to ameliorate part of the disease symptoms. In the context of wildtype background, Sox17 study in Chapter 2 demonstrates that Sox17 can modulate insulin secretion and can regulate several genes involved in insulin trafficking and secretion pathway, such as Pdx1, GLP1R, ATF4, and WFS1. In the context of MODY4 disease, it was known that haploinsufficiency of Pdx1 led to hyperglycemia, impaired glucose tolerance, decreased insulin and Glut2 levels, and disrupted islet architecture8-12. Several studies showed that the defects in

Pdx1 haploinsufficient mice are due to GLP1 and its consequent cAMP signaling in the cell12. In addition, recent study highlighted the role of Pdx1 in modifying

146 cell susceptibility to ER stress, in particular by targeting Atf4 and WFS113. These studies suggest that Pdx1 and Sox17 may regulate the same biological pathway in the cell. Taken all the studies together, we analyzed in Chapter 3 whether

Sox17 can positively impact the MODY4 pathology. This study demonstrates that

Sox17 can rescue MODY4 hyperglycemia and disrupted islet architecture, but failed to improve glucose tolerance in this MODY4 background. Preliminary data showed Sox17 was able to reverse the MODY islet architecture in part by inducing cell shape change and increasing the islet size in the pancreas. Further studies need to be performed to dissect the molecular mechanism that governs this rescue by Sox17. This study along with the data presented earlier confirms that the signaling of Sox17 under a physiological dose can be used therapeutically to improve cell function in the context of diabetes.

Experimental limitations and alternative approaches.

The experimental approaches used in this dissertation were designed to generate specific and quantitative data in order to rigorously assess the role of

Sox17 in normal and pathological pancreatic cells. These data provide evidence in support of the main hypothesis; however, there exist some experimental limitations that can be addressed by using alternative approaches of experiments.

In the study of Sox17 loss of function, using Pdx1-Cre driver resulted in deletion of Sox17 throughout the pancreas since the beginning of pancreas development. We tested the developmental defect of Sox17 in these mice, and

147 found that the lineage segregation defect as a result of losing Sox17 using this

Pdx1-Cre driver was minimum. However, by using an alternate driver, such as

Pdx1-CreER, it would allow a more direct analysis to assess the role of Sox17 in the mature cell only14. There is an inherent experimental concerns in using the

Cre-loxP system; therefore, careful analysis of the Cre toxicity and the efficiency of the recombination will need to be performed when introducing a new driver in the breeding system. The proper control mice of Pdx1-CreER need to be used in order to make sure that the driver mice themselves do not develop any adverse phenotype as in the case of the Rip-Cre transgene driver mice15.

Another experimental limitation, which must be addressed, is the use of specific insulin processing and secretion assays. A pulse-chase experiment to determine the proinsulin processing will further determined where the processing is delayed (reviewed in16). In addition, insulin secretion assays using perfused pancreas and isolated islet perifusion will also further analyze whether the insulin secretory physiology is affected in these islets9. Using these techniques, a cytosolic calcium imaging and intracellular cAMP levels can also be measured17.

The experiments detailed in Chapter 3 suggested that Sox17 is able to rescue some of the MODY4 disease phenotypes. However, further studies need to be performed to analyze the molecular mechanism of how Sox17 fits into the

MODY transcriptional network. In addition, it is also important to determine the interaction between Sox17 and Pdx1 in regulating the insulin secretory pathway both transcriptionally and post-translationally. Sox17 may act as a cofactor of

Pdx1, as in the case of Sox619. In addition, it was shown that phosphorylation of

148 Pdx1 leads to its subnuclearlocalization in the cell in response to glucose18.

Hence, it would also be critical to study the degree of phosphorylation of Sox17 and Pdx1 proteins and how it affects each other’s function and their effects on cell function.

Experiments detailed in this study were designed to demonstrate that

Sox17 plays a critical role in mature cells to maintain glucose homeostasis.

This knowledge may make it possible for future studies to analyze specific therapeutical targets that can be modulated by Sox17 to improve cell function and delay or prevent diabetes disease progression.

Acknowledgements

We thank James Wells, Jeff Whitsett, Aaron Zorn, Jonathan Katz, Gail Deutsch,

Jason Spence, Peter Arvan, Leena Haataja, Matt Kofron, Anil Jegga,

Debora Sinner, Alex Lange, Georgianne Ciraolo, Microarray Core, and we

thank Suh-Chi Lin, Kyle McCracken, Anna Method, Christopher Mayhew,

Jonathan Howell, and the many wonderful colleagues in the Wells and

Zorn lab for communication of results prior to publication. We also thank

Sean Morrison (University of Michigan) and Raymond MacDonald

(University of Texas Southwestern) for the Sox17GFP and Pdx1tTA mice.

Andy Lowy for the Pdx1-Cre mice, also Ben Stanger for the Pdx1-CreER

mice.

149 Sources of funding

This work was supported by JDRF 2-2003-530 and  Cell Consortium-BCBC

(Grant #U01DK072473).

References.

1Izumi, T.et al. Dominant negative pathogenesis by mutant proinsulin in the Akita diabetic mouse. Diabetes 52, 409-416 (2003). 2 Hartling, S. G. et al. Longitudinal study of fasting proinsulin in 148 siblings of patients with insulin-dependent diabetes mellitus. Study Group on Childhood Diabetes in Finland. Eur J Endocrinol 137, 490-494 (1997). 3 Roder, M. E. et al. Disproportionately elevated proinsulin levels precede the onset of insulin-dependent diabetes mellitus in siblings with low first phase insulin responses. The Childhood Diabetes in Finland Study Group. J Clin Endocrinol Metab 79, 1570-1575 (1994). 4 Roder, M. E., Porte, D., Jr., Schwartz, R. S. & Kahn, S. E. Disproportionately elevated proinsulin levels reflect the degree of impaired B cell secretory capacity in patients with noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 83, 604-608 (1998). 5 Saad, M. F. et al. Disproportionately elevated proinsulin in Pima Indians with noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 70, 1247-1253 (1990). 6 Kahn, C. R. Diabetes. Causes of insulin resistance. Nature 373, 384-385, doi:10.1038/373384a0 (1995). 7 Konstantinova, I. et al. EphA-Ephrin-A-mediated cell communication regulates insulin secretion from pancreatic islets. Cell 129, 359-370, doi:S0092-8674(07)00368-6 [pii] 10.1016/j.cell.2007.02.044 (2007). 8 Ahlgren, U., Jonsson, J., Jonsson, L., Simu, K. & Edlund, H. -cell- specific inactivation of the mouse Ipf1/Pdx1 gene results in loss of the - cell phenotype and maturity onset diabetes. Genes Dev 12, 1763-1768 (1998). 9 Johnson, J. D. et al. Increased islet apoptosis in Pdx1+/- mice. J Clin Invest 111, 1147-1160, doi:10.1172/JCI16537 (2003). 10 Brissova, M. et al. Reduction in pancreatic transcription factor PDX-1 impairs glucose-stimulated insulin secretion. J Biol Chem 277, 11225- 11232, doi:10.1074/jbc.M111272200 M111272200 [pii] (2002). 11 Kulkarni, R. N. et al. PDX-1 haploinsufficiency limits the compensatory islet hyperplasia that occurs in response to insulin resistance. J Clin Invest 114, 828-836, doi:10.1172/JCI21845 (2004).

150 12 Wang, H. et al. Suppression of Pdx-1 perturbs proinsulin processing, insulin secretion and GLP-1 signalling in INS-1 cells. Diabetologia 48, 720- 731, doi:10.1007/s00125-005-1692-8 (2005). 13 Sachdeva, M. M. et al. Pdx1 (MODY4) regulates pancreatic cell susceptibility to ER stress. Proc Natl Acad Sci U S A 106, 19090-19095, doi:0904849106 [pii] 10.1073/pnas.0904849106 (2009). 14 Gu, G., Dubauskaite, J. & Melton, D. A. Direct evidence for the pancreatic lineage: NGN3+ cells are islet progenitors and are distinct from duct progenitors. Development 129, 2447-2457 (2002). 15 Lee, J. Y. et al. RIP-Cre revisited, evidence for impairments of pancreatic -cell function. J Biol Chem 281, 2649-2653, doi:M512373200 [pii] 10.1074/jbc.M512373200 (2006). 16 Steiner, D. F., Park, S. Y., Stoy, J., Philipson, L. H. & Bell, G. I. A brief perspective on insulin production. Diabetes Obes Metab 11 Suppl 4, 189- 196, doi:DOM1106 [pii] 10.1111/j.1463-1326.2009.01106.x (2009). 17 Gao, N. et al. Foxa2 controls vesicle docking and insulin secretion in mature  cells. Cell Metab 6, 267-279, doi:S1550-4131(07)00260-4 [pii] 10.1016/j.cmet.2007.08.015 (2007). 18 An, R. et al. Pancreatic and duodenal homeobox 1 (PDX1) phosphorylation at serine-269 is HIPK2-dependent and affects PDX1 subnuclear localization. Biochem Biophys Res Commun 399, 155-161, doi:S0006-291X(10)01337-9 [pii] 10.1016/j.bbrc.2010.07.035 (2010). 19 Iguchi, H. et al. SOX6 attenuates glucose-stimulated insulin secretion by repressing PDX1 transcriptional activity and is down-regulated in hyperinsulinemic obese mice. J Biol Chem 280, 37669-37680, doi:M505392200 [pii] 10.1074/jbc.M505392200 (2005).

151 Figure Legends.

Figure 1. The role of Sox17 in regulating insulin trafficking.

Sox17 loss of function study suggested that Sox17 functions to fine-tune proinsulin trafficking between the secretory machinery organelles. Additionally,

Sox17 gain of function study further showed that precise level of Sox17 in the cell is important in order to maintain effective proinsulin transport and secretion.

Figure 1.

152 “I only hope that we don't lose sight of one thing - that it was all started by a mouse.” – Walt Disney

153