Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Removing, replacing, and processing proinsulin in beta-cells Ramzy, Adam Rehim 2019

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2019_september_ramzy_adam.pdf [ 22.09MB ]
Metadata
JSON: 24-1.0379180.json
JSON-LD: 24-1.0379180-ld.json
RDF/XML (Pretty): 24-1.0379180-rdf.xml
RDF/JSON: 24-1.0379180-rdf.json
Turtle: 24-1.0379180-turtle.txt
N-Triples: 24-1.0379180-rdf-ntriples.txt
Original Record: 24-1.0379180-source.json
Full Text
24-1.0379180-fulltext.txt
Citation
24-1.0379180.ris

Full Text

Removing, Replacing, and Processing Proinsulin in Beta-Cells   by  ADAM REHIM RAMZY  B.HSc., The University of Calgary, 2012     A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY   in   THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Cell and Developmental Biology)      THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)      May 2019  © Adam Rehim Ramzy, 2019  ii   The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  Removing, Replacing, and Processing Proinsulin in Beta-Cells  Submitted by Adam Ramzy  in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Cell and Developmental Biology  Examining Committee: Dr. Timothy J Kieffer Supervisor  Dr. Susanne Clee Supervisory Committee Member  Dr. C Bruce Verchere Supervisory Committee Member Dr. William Thomas Gibson University Examiner Dr. Dan Seriano Luciani University Examiner   Additional Supervisory Committee Members: Dr. Garth Warnock Supervisory Committee Member Dr. Scott Covey Supervisory Committee Member    iii  Abstract Diabetes affects over 425 million worldwide, costs billions, and causes morbidity and mortality for patients. Though insulin injections are lifesaving, insufficient β-cell mass and function leaves patients facing risks of chronic hyperglycemia and acute risks of hypoglycemia. Replacement of β-cells via transplantation of cadaveric islets is a functional cure but is limited by a paucity of donor tissue. If β-cell replacement or (re)generation therapies were abundantly available, they could be potential cures for diabetes. To this end, investigating β-cell development and function is worthwhile. In the current thesis, we first characterized the role of insulin on β-cell development and maturation by studying insulin knockout mice (Ins1-/-Ins2-/-). Though insulin was necessary for β-cell maturation, insulin replacement by islet transplantation but not insulin injection, supported maturation of endogenous β-cells. Second, we developed and characterized an adeno associated virus (AAV) carrying Cre recombinase regulated by an insulin promoter (AAV Ins1-Cre) for in vivo genetic manipulations. AAV Ins1-Cre produced efficient recombination in β-cells alongside off-target recombination, making it a useful tool when off-target effects are controlled for or deemed unimportant. Third, we assessed the viability of a gene therapy for the Ins1-/-Ins2-/- mouse model of monogenic diabetes. We delivered an insulin gene to β-cells (an Ins1 promoter driving human insulin (INS) or mouse insulin 1 (Ins1)) using AAV Ins1-INS or AAV Ins1-Ins1. Though the AAV delivered the insulin gene to β-cells, Ins1-/-Ins2-/- β-cells retained a processing defect leading to secretion of insulin’s precursor proinsulin. We created adult insulin knockout mice using AAV Ins1-Cre and failed to prevent onset of diabetes with AAV Ins1-Ins1. Finally, in Chapter 5 we assessed the production of mature insulin in human β-cells. Despite consensus on the role of prohormone convertase 2 (PC2) in proinsulin processing, we provide evidence that unlike mouse β-cells, human β-cells produce mature insulin without PC2. This thesis provides insight into the developmental impact of “removing insulin” from β-cells, assesses the viability of a gene therapy “replacing insulin” in β-cells, and revises a longstanding theory on the “processing of proinsulin” in human β-cells. These findings may guide development of gene- and cell- based therapies for diabetes.  iv  Lay Summary Diabetes affects over 425 million people worldwide and costs healthcare systems billions. All patients with diabetes have insufficient insulin, a hormone produced by β-cells within pancreatic islets. Though insulin injections have been lifesaving for almost 100 years, patients struggle with tremendous morbidity and mortality associated with the challenges and flaws in attempting to regulate insulin through a syringe. Demonstration that transplantation of pancreatic islets from organ donors can be curative, supports research on the development and function of β-cells to guide future gene- and cell-based therapies. The present thesis investigates the development of β-cells, develops a tool to manipulate β-cell genes in mice, assesses the viability of a gene therapy cure for diabetes caused by mutations in the insulin gene, and corrects a longstanding dogma on how human β-cells produce properly processed insulin. This work provides broad insight into the pathogenesis of diabetes and guides the development of potential treatments. v  Preface Components of Chapter 1 were adapted from two published articles: Ramzy A (2017) Moving gene therapies from the lab to the hospital bed: The adeno-associated virus as a promising gene therapy vector to treat disease. UBC Medical Journal. 8(2), 1-2. and Ellis C*, Ramzy A*, Kieffer TJ (2017) Regenerative medicine and cell-based approaches to restore pancreatic function. Nat Rev Gastroenterol & Hepatol. 14(10), 612-628. *These authors contributed equally to this work. The first was written by AR. The second was co-written by AR, CE, and TJK. Figure 1-1 is reprinted from Ellis C, Ramzy A, Kieffer TJ (2017). Figures 1-2 and 1-3 were reprinted from Bruin JE, Kieffer TJ, Stem Cells and Cancer Stem Cells, Volume 8, 2012 with permission from the authors. Studies from Chapter 2 are published in the following article: Ramzy A, Mojibian M, Kieffer TJ (2018) Insulin-deficient mouse β-cells do not fully mature but can be remedied through insulin replacement by islet transplantation. Endocrinology. 159(1), 83-102. AR and TJK designed the experiments. AR performed the experiments. MM performed islet transplantation, optimized insulin therapy, and generated and maintained Ins1−/−Ins2−/−animals. AR analyzed data, and AR and TJK wrote the manuscript with input from MM. Studies from Chapter 3 are in review in the following manuscript: Ramzy A, Tudurí E, Glavas MM, Baker RK, Mojibian M, Fox JK, O'Dwyer SM, Dai D, Hu X, Denroche HC, Edeer N, Gray SL, Verchere CB, Johnson JD, Kieffer TJ (2019) AAV Ins1-Cre can produce efficient β-cell recombination but requires consideration of off-target effects. AR, ET, MMG, RKB, and TJK designed experiments. AR, ET, MMG, RKB, MM, JF, SMO, DD, XH, HCD and NE performed experiments. AR wrote the manuscript with input from ET, MG, CBV, JDJ, and TJK. In vivo experiments presented in Figure 3-1 were done predominantly by ET and in vivo experiments presented in Figure 3-7 were completed predominantly by MMG and DD. Data presented in Figure 3-2, perifusion and calcium imaging in Figure 3-3, and Figure 3-4 were collected vi  predominantly by ET and XH. The remaining work was done predominantly by AR with some tissues used for immunostaining collected by ET, MMG, and SLG. Studies from Chapter 4 were conceived and designed by A Ramzy, RK Baker, and TJ Kieffer. AR performed experiments and analyzed data. RKB designed viral constructs and N Edeer assisted with experiments presented in Figures 4-7, 4-8, and 4-9. M Reid (McMaster University) collected images by transmission electron microscopy. AR wrote the chapter with input from TJK. Studies from Chapter 5 are in peer review the following manuscript: Ramzy A, Asadi A, Kieffer TJ (2019) Human β-cells process proinsulin with prohormone convertase (PC) 1/3 but not PC2. AR and AA performed experiments. AR analyzed data and wrote the manuscript with contributions from AA and TJK. All authors approved the final draft. Data presented in Figures 5-2B and 5-3 were predominantly collected by AA and the remaining data was collected predominantly by AR. A Ramzy wrote this thesis with comments and edits from TJ Kieffer, S Clee, S Covey, CB Verchere, and G Warnock. All animal studies in this thesis were approved by the University of British Columbia Animal Care Committee (Certificates #A14-0162, #A14-0081, #A17-0026, #A18-0113). Work with human islets from the University of Alberta Islet Core are approved with the Research Ethics Board (H14-02949).   vii  Table of Contents Abstract .................................................................................................................................... iii Lay Summary ........................................................................................................................... iv Preface ....................................................................................................................................... v Table of Contents .................................................................................................................... vii List of Tables ............................................................................................................................ xi List of Figures ......................................................................................................................... xii Acknowledgements ................................................................................................................ xx Dedication ............................................................................................................................. xxii Chapter 1: Introduction ............................................................................................................ 1 1.1 Diabetes Mellitus .......................................................................................................1 1.1.1 The burden of diabetes .....................................................................................1 1.1.2 Common types of diabetes ...............................................................................1 1.1.3 Monogenic diabetes ..........................................................................................3 1.1.4 Reduced β-cell mass is a hallmark of diabetes ..............................................5 1.2 Medical management of diabetes ...........................................................................6 1.3 β-cell replacement: the precedent of islet transplantation ....................................9 1.4 β-cell development and maturation ....................................................................... 13 1.5 β-cell replacement ................................................................................................... 16 1.5.1 Generation of β-cells in vitro .......................................................................... 16 1.5.1.1 Insulin secreting cell lines ........................................................................... 16 1.5.1.2 Xenotransplantation .................................................................................... 17 1.5.1.3 Stem-cell-derived β cells ............................................................................. 18 1.5.2 Reprogramming cells in situ ........................................................................... 19 viii  1.5.2.1 Reprogramming pancreas .......................................................................... 20 1.5.2.2 Reprogramming hepatocytes ..................................................................... 24 1.5.2.3 Reprogramming the gastrointestinal tract ................................................. 26 1.6 Gene therapy ........................................................................................................... 29 1.6.1 In vivo genetic manipulations ......................................................................... 29 1.6.2 The adeno-associated virus ........................................................................... 30 1.6.3 Clinical use of AAVs ........................................................................................ 32 1.6.4 Using AAVs to study and treat diabetes ....................................................... 34 1.7 Prohormone processing ......................................................................................... 35 1.7.1 Discovery of prohormones.............................................................................. 35 1.7.2 The prohormone convertases ........................................................................ 35 1.7.2.1 Prohormone convertases 1/3 and 2 .......................................................... 36 1.7.2.2 Predicting target specificity of the prohormone convertases................... 37 1.7.2.3 Carboxypeptidase E .................................................................................... 39 1.7.2.4 The prohormone convertase cofactors proSAAS and 7B2 ..................... 39 1.7.3 Proinsulin processing ...................................................................................... 40 1.7.4 Proinsulin processing in human β-cells ......................................................... 43 1.8 Thesis investigation ................................................................................................ 46 Chapter 2: Insulin-deficient mouse β-cells do not fully mature but can be remedied through insulin replacement by islet transplantation........................................................... 48 2.1 Introduction .............................................................................................................. 48 2.2 Materials and Methods ........................................................................................... 50 2.3 Results ..................................................................................................................... 57 ix  2.4 Discussion ............................................................................................................... 79 Chapter 3: AAV Ins1-Cre can produce efficient β-cell recombination but requires consideration of off-target effects ......................................................................................... 86 3.1 Background.............................................................................................................. 86 3.2 Materials and methods ........................................................................................... 88 3.3 Results ..................................................................................................................... 92 3.4 Discussion ............................................................................................................. 108 Chapter 4: Insulin deficient β-cells retain a prohormone processing defect that impairs an AAV mediated insulin gene therapy .................................................................................... 113 4.1 Background............................................................................................................ 113 4.2 Materials and methods ......................................................................................... 114 4.3 Results ................................................................................................................... 121 4.4 Discussion ............................................................................................................. 143 Chapter 5: Revisiting proinsulin processing: Evidence that human β-cells lack prohormone convertase 2 and can produce mature insulin without its function ............ 152 5.1 Background............................................................................................................ 152 5.2 Materials and methods ......................................................................................... 152 5.3 Results ................................................................................................................... 159 5.4 Discussion ............................................................................................................. 172 Chapter 6: Conclusions, future directions, and major challenges ................................. 177 6.1 Research summary and conclusions .................................................................. 177 6.2 Research limitations ............................................................................................. 180 6.3 Future Directions ................................................................................................... 183 6.3.1 The role of insulin during human β-cell development ................................ 183 x  6.3.2 Investigating the impact of insulin therapy on β-cell maturity ................... 183 6.3.3 Further investigating the observed prohormone processing defect in insulin deficient Ins1-/-Ins2-/- β-cells ............................................................................. 185 6.3.4 Further investigating gene therapy approaches for both insulin knockout PND as well as other models of monogenic diabetes ............................................... 186 6.3.5 Assessing partial proinsulin processing by PC2 in human β-cells ...... 188 6.3.6 Characterizing the role of β-cell prohormone processing defects during diabetes pathogenesis ............................................................................................... 189 6.3.7 Investigating the role of PC2 for processing proinsulin in rat β-cells ....... 190 6.4 Concluding thoughts ............................................................................................. 190 References ............................................................................................................................ 192 Appendices ........................................................................................................................... 217 Appendix A Secondary antibodies used for immunofluorescent staining in all chapters. ............................................................................................................................ 217 Appendix B Maps of plasmid sequences between flanking ITRs that were used for the generation of AAVs. ......................................................................................................... 218 Appendix C Study designs of in vivo experiments in Chapter 4. ................................. 219  xi  List of Tables Table 1-1 List of relevant ongoing or completed trials into cell-based therapies for diabetes .....12 Table 2-1 Primary antibodies used for immunofluorescent staining ...........................................55 Table 2-2 Primer sequences and annealing temperatures. .......................................................57 Table 3-1 Primary antibodies used for immunofluorescent experiments. ...................................91 Table 4-1 Relevant reported cross reactivity of commercial ELISAs for insulin, C-peptide, proinsulin, and proinsulin intermediates (%). ........................................................................... 117 Table 4-2 Primary antibodies used for immunofluorescent staining ......................................... 119 Table 4-3 Summary of studies investigating the relative bioactivities of insulin and proinsulin . 127 Table 5-1 Basic demographics of pancreas and islet donors used in experiments. ................. 153 Table 5-2 Key resources including antibodies used in Chapter 5 ............................................ 155 Table 5-3 Oligonucleotide primers and probes used in qPCR and ISH experiments. .............. 158 Table 5-4 Alignment of PC2 sequences for human, mouse, rat, pig, and dog ......................... 168  xii  List of Figures Figure 1-1 The pancreas .............................................................................................................2 Figure 1-2 Glucose monitoring during diabetes ...........................................................................7 Figure 1-3 Continuous glucose monitoring after islet transplant ................................................11 Figure 1-4 β-cell development ...................................................................................................16 Figure 1-5 Conventional production of recombinant AAV requires a triple transfection of adherent HEK293 cells. ............................................................................................................31 Figure 1-6 Alignment of proinsulin sequences ...........................................................................41 Figure 1-7 Current theory of proinsulin processing ....................................................................42 Figure 1-8 Human β-cells have less PCSK2 than PCSK1 in published RNAseq databases on sorted human and mouse islet cells. .........................................................................................44 Figure 2-1 Ins1-/-Ins2-/- mice have IAPP expressing β-cells........................................................59 Figure 2-2 Variable MafA immunoreactivity in mice with reduced insulin gene copy number. ....60 Figure 2-3 Ins1-/-Ins2-/- mice have enlarged islets with abnormal islet cell number and distribution. ...............................................................................................................................62 Figure 2-4 β-cells in hypoglycemic Ins1-/-Ins2-/- mice lack immunoreactivity for some β-cell transcription factors. ..................................................................................................................64 Figure 2-5 Hyperglycemic Ins1-/-Ins2-/- β-cells lose expression of key β-cell genes. ...................65 Figure 2-6 Hypoglycemic Ins1-/-Ins2-/- β-cells lack expression of key β-cell factors. ...................67 Figure 2-7 Consistent patterns of immunofluorescence with two commercially available NGN3 antibodies. ................................................................................................................................68 Figure 2-8 Ins1-/-Ins2-/- β-cells are not fully differentiated. ..........................................................69 Figure 2-9 Ins1-/-Ins2-/- β-cells lose expression of βGAL after insulin therapy to keep mice alive into adulthood. ..........................................................................................................................71 xiii  Figure 2-10 Adult Ins1-/-Ins2-/- mice treated by insulin injections have islet hyperplasia and fibrosis. .....................................................................................................................................72 Figure 2-11 Insulin therapy by injection is not sufficient for the completed maturation of Ins1-/-Ins2-/- β-cells. ............................................................................................................................74 Figure 2-12 Insulin replacement by islet transplant leads to islet hyperplasia and preservation of β-cell mass, unlike insulin injections. .........................................................................................76 Figure 2-13 β-cells from islet transplant treated Ins1-/-Ins2-/- mice express markers of mature β-cell function. ..............................................................................................................................77 Figure 2-14 β-cells of islet transplant treated Ins1-/-Ins2-/- mice are more mature. ......................78 Figure 3-1 Intraperitoneal administration of AAV Ins1-Cre up to a single dose of 1012 VGP does not significantly alter glucose metabolism. ................................................................................93 Figure 3-2 AAV Ins1-Cre does not elevate liver enzymes..........................................................94 Figure 3-3 Intraperitoneal administration of AAV Ins1-Cre (1012 VGP) does not significantly alter β-cell maturity or function. .........................................................................................................96 Figure 3-4 AAV Ins1-Cre can infect islets in vitro ......................................................................98 Figure 3-5 Intraperitoneal AAV Ins1-Cre produces dose dependent β-cell recombination alongside hypothalamic and acinar tissue recombination. .........................................................99 Figure 3-6 AAV Ins1-Cre intraductal administration does not significantly alter glucose tolerance but induction of foreign fluorescent protein expression causes insulitis. .................................. 101 Figure 3-7 High sequence similarity allows detection of different confetti recombination events using a polyclonal α GFP antibody .......................................................................................... 103 Figure 3-8 Both IP and ID administration of AAV Ins1-Cre can cause recombination in the liver but the extent may be dependent on delivery method and model organism. ........................... 105 Figure 3-9 The AAV Ins1-Cre can be a useful tool for directing β-cell recombination when off-target effects are deemed minimally important. ....................................................................... 107 xiv  Figure 4-1 β-cell infection by AAV Ins1-INS. ........................................................................... 122 Figure 4-2 Transient remission of diabetes in Ins1-/-Ins2-/- mice after treatment with AAV Ins1-INS .......................................................................................................................................... 124  Figure 4-3 Low infection rate and normal immunoreactivity for PC1/3 and PC2 in β-cells of AAV Ins1-INS injected Ins1-/-Ins2-/- mice .......................................................................................... 128 Figure 4-4 Low infectivity and impaired proinsulin processing prevented diabetic remission in AAV Ins1-Ins1 treated Ins1-/-Ins2-/- mice .................................................................................. 131 Figure 4-5 Ins1-/-Ins2-/- β-cells have impaired prohormone processing .................................... 132 Figure 4-6 Ins1-/-Ins2-/- β-cells appear mature and are not immunoreactive for markers of dedifferentiation ...................................................................................................................... 134 Figure 4-7 AAV Ins1-Ins1 has low infectivity for β-cells and may have aberrant hepatic expression .............................................................................................................................. 137 Figure 4-8 AAV Ins1-Ins1 did not prevent diabetes onset when co-delivered with AAV Ins1-Cre ............................................................................................................................................... 138 Figure 4-9 Relatively low rates of C-peptide 1 immunoreactivity after co-delivery of AAV Ins1-Ins1 with AAV Ins1-Cre ........................................................................................................... 140 Figure 4-10 Failed production of AAV Ins1-Ins1 ...................................................................... 142 Figure 4-11 Efficient induction of diabetes in Ins1-/-Ins2f/fmTmGPdxCre-ER mice .................... 143 Figure 5-1 Human β-cells were not immunoreactive for PC2 using four antibodies ................. 162 Figure 5-2 Single channel immunostaining for insulin and PC2 using multiple antibodies and no colocalization of PC1/3 and PC2 in human pancreas. ............................................................. 163 Figure 5-3 In multiple human pancreas donors PC2 did not colocalize with insulin and in multiple mouse pancreata PC2 colocalizes with insulin. ....................................................................... 164 Figure 5-4 Human β-cells have less immunoreactivity for 7B2 than α-cells whereas mouse β-cells have comparable immunoreactivity to α-cells. ................................................................. 165 xv  Figure 5-5 Rat β-cells were not immunoreactive for PC2 and EndoC- βH1 cells have PC2..... 167 Figure 5-6 Human β-cells have less PCSK2 than neighbouring α-cells, unlike mouse β-cells . 169 Figure 5-7 Human β-cells can process proinsulin without PC2 function but require the function of PC1/3, unlike PC2 dependent α-cells ...................................................................................... 171    xvi  List of Abbreviations AAV8 Adeno associated virus, eighth serotype ACTH Adrenocorticotropic hormone ADCY5 Adenylate cyclase type 5 ALDH1A3 Aldehyde dehydrogenase 1 family member A3 ARX Aristaless-related homeobox protein ATP Adenosine triphosphate BMP Bone morphogenetic protein βGAL β-galactosidase Cas9 CRISPR associated protein 9 CDX2 Caudal homeobox protein 2 CRISPR Clustered regularly interspaced short palindromic repeats DAPI 4’,6-diamidino-2-phenylindole dsAAV Double stranded adeno associated virus EGF Epidermal growth factor FGF Fibroblast growth factor FOXO1 Forkhead box protein O1 GATA4 GATA binding protein 4 Gapdh Glyceraldehyde 3-phosphate dehydrogenase GCG Glucagon GCK Glucokinase GIP Glucose-dependent insulinotropic peptide GLP-1 Glucagon-like peptide 1 GLUT1 Glucose transporter 1 GLUT2 (gene: Slc2a2) Glucose transporter 2 xvii  GSIS Glucose stimulated insulin secretion HbA1c Hemoblogin A1C HES1 Hairy and enhancer of split 1 HHEX Haematopoietically expressed Hhex HNF1α Hepatocyte nuclear factor 1 alpha IAPP islet amyloid polypeptide ID Intraductal IGF Insulin-like growth factor IGF1R Insulin-like growth factor 2 receptor INS Insulin  Ins1 Insulin 1 Ins2 Insulin 2 IP Intraperitoneal IPGTT Intraperitoneal glucose tolerance test IR Insulin receptor IRS Insulin receptor substrate ISH In situ hybridization ISL1 Insulin gene enhancer 1 KCNJ11 Potassium voltage-gated channel subfamily J member 11 L-MYC v-myc avian myolocytomatosis viral oncogene lung carcinoma derived MAFA Muscoloaponeurotic fibrosarcoma oncogene homolog A MHC Major histocompatibility complex MODY Monogenic diabetes of the young mRNA Messenger ribonucleic acid xviii  NANOG Homeobox protein NANOG ND Neonatal diabetes NEUROD1 Neurogenic differentiation 1 NGN3 Neurogenin-3 NKX2.2 Homeodomain transcription factor 2.2 NKX6.1 Homeodomain transcription factor 6.1 PAX4 Paired box 4 PAX6 Paired box 6 PBS Phosphate-buffered saline PC1/3 (gene: Pcsk1) Prohormone convertase 1/3 PC2 (gene: Pcsk2) Prohormone convertase 2 PCNA Proliferating cell nuclear antigen PDX1 Pancreatic and duodenal homeobox 1 PND Permanent neonatal diabetes PP (gene: Ppy) Pancreatic polypeptide PTF1a Pancreas associated transcription factor 1a qPCR Quantitative polymerase chain reaction RRID Research resource identifier SOX2 Sry box-2 SST Somatostatin  STZ Streptozotocin SUR1 Sulfonylurea receptor 1 SYN Synaptophysin  TCF7L2 Transcription factor 7-like 2 TND Transient neonatal diabetes xix  UBC University of British Columbia UCN3 Urocortin-3  xx  Acknowledgements A sincere thank you to all of those who contributed to this work both directly and indirectly. This dissertation is the culmination of the work of many. First and foremost, I owe particular thanks to my supervisor Dr. Timothy J Kieffer for his inspirational and supportive guidance. Whether in the form of a hallway chat, a 4-hour lab meeting, or a comment-peppered manuscript, I am grateful for your time to teach, challenge, and push me. Your mentorship has yielded lessons I will use for my entire career ahead. I thank my supervisory committee members, past and present, Drs. Susanne Clee, Scott Covey, Jim Johnson, Bruce Verchere, and Garth Warnock for their support through challenging me, sharing their expertise, and guidance along the winding road to a PhD. A special thank you to Dr. Garth Warnock for sharing his guidance not only in committee meetings, but also for his generosity by taking the time to teach in the operating room. My gratitude to Dr. Jim Johnson for his consistent attention to detail and teaching the highest standards of scientific rigor. Thank you to Dr. Scott Covey for always making the time to help and sharing his substantial experimental knowledge. My appreciation to Dr. Bruce Verchere for his expert guidance and consistent willingness to share resources, time, and support. Thank you to Dr. Susanne Clee for thinking outside the box and helping find unexpected solutions. Thank you to all the members of the Kieffer lab for always asking the tough questions. I have been fortunate to work with many talented and kind graduate students, fellows, and technicians. To my fellow office-mates Chiara Toselli, Nelly Saber, Ursula Neumann, Anna D’Souza, and Blair Gage, thanks for always making time spent in the lab more fun and always taking the time to help. To Travis Webber, Ali Asadi, Robert Baker, Majid Mojibian, Maria Glavas, Cara Ellis, Sandra Pereira, and Shannon O’Dwyer, thanks for your teaching, guidance, assistance, and comradery. Thanks to Nazde Edeer for her hard work and thanks to the many others for their contributions to these studies. xxi  Thank you to the UBC MD/PhD program for their professional and financial support, including the Vancouver Frasier Medical Program MD/PhD Scholarship. My gratitude to the Canadian Institutes of Health Research for funding support (Vanier Canada Graduate Scholarship) and the other research agencies including Diabetes Canada that made doing this work possible. To my parents, thank you for teaching me how to work hard and be happy along the way. I am not sure where I’d be without your example on how to make priorities and live life with moderation, but I am thankful for where it has taken me. I also must give a special thanks to Uncle Rehim for being my friend and teacher. You have always been a source of laughter and been there anywhere, and anytime. Thank you to Matt for being the truest friend one could hope to have and thank you to my sister Miriam for always being someone I can count on. I have been so fortunate to be surrounded by friends and family that have supported me throughout this journey. Finally, a thank you that could never be put into words – thank you to my wife Jessie. No matter the circumstance, you have been the unwavering piece of my life that is kind, positive, and generous. It is because of you that these past five years have been both productive and a sincerely happy phase of life.     xxii  Dedication        This dissertation is dedicated to Jessie, for maximizing all that is good in life.     1  Chapter 1: Introduction 1.1 Diabetes Mellitus 1.1.1 The burden of diabetes Diabetes is a metabolic disease that is defined by elevated blood glucose (hyperglycemia). Diabetes affects over 425 million worldwide1 and bears a significant financial, disability, and mortality cost for health care systems and patients around the world2. In Canada alone, approximately 10 million people have diabetes or prediabetes and this number is estimated to increase by 44% by 20253. With so many affected it is estimated that diabetes will cost the Canadian healthcare system over $15 billion annually by 20224. As a consequence of chronic hyperglycemia patients face a shortened lifespan and deal with complications of diabetes including diabetic retinopathy, nephropathy, neuropathy, limb infections, and cardiovascular disease5. Contrarily, acute low blood sugar (hypoglycemia) is a common side-effect of medications prescribed for the management of diabetes and poses serious risks including loss of consciousness, coma, and even death without prompt management. 1.1.2 Common types of diabetes Almost all diabetes is caused by insulin insufficiency6. Insulin is a hormone produced by β-cells within the pancreatic islets. Islets are clusters of ~1000-2000 cells that form the endocrine portion of the pancreas. Islets predominantly contain insulin-secreting β-cells, glucagon-secreting α-cells, and somatostatin-secreting δ cells, as well as other rare cell types including pancreatic polypeptide (PP) producing PP-cells and ghrelin producing ε-cells (Figure 1-1). Insulin from β-cells is the main hormone that acts to lower blood sugar and functions by binding its receptor on the cell surface and signaling cells to increase glucose uptake, protein synthesis, and energy storage. Glucagon is a counter-regulatory hormone like growth hormone and cortisol that acts to counter the action of insulin in order to raise blood sugar. Most often, diabetes is broadly classified as type 1, type 2, gestational, and other rare  2   Figure 1-1 The pancreas The endocrine pancreas is comprised of four main cell types, including insulin-secreting β cells, glucagon-secreting α cells, somatostatin-secreting δ-cells, and pancreatic polypeptide expressing PP-cells, alongside very rare ghrelin producing ε-cells. The exocrine pancreas makes up >95% of the pancreatic tissue and is comprised of acinar cells that produce digestive enzymes, such as trypsin and amylase, and a network of ductal cells that carry these enzymes to the intestine. Figure adapted from7.  forms. In type 1 diabetes autoreactive T-cells selectively kill β-cells of the pancreatic islets and all patients are treated with insulin administration by injection or pump. It is a polygenic disease triggered by yet unknown environmental factors. Variants in the major histocompatibility complex class II is the most significant genetic predictor for type 1 diabetes with an odds-ratio of almost seven, with all other variants having odds ratios less than two8. Type 2 diabetes affects approximately 90% of patients with diabetes and is characterized by insulin resistance and eventual β-cell loss. Though many patients with type 2 diabetes have hyperinsulinemia, a reduced sensitivity to insulin in peripheral tissues (insulin resistance) and increased circulating glucagon results in a relative insulin insufficiency9. Genetics contribute substantially to risk for type 2 diabetes with a concordance rate of ~70% in monozygotic twins and major environmental 3  risk factors are physical inactivity, unhealthy diet, and obesity10. Many common genetic variants such as single nucleotide polymorphisms in genes like transcription factor 7-like 2 (TCF7L2) or adenylate cyclase type 5 (ADCY5) are associated with type 2 diabetes, and it seems unlikely that rare variants cause this disease11. Gestational diabetes has a similar pathogenesis to type 2 diabetes with hyperglycemia precipitated by the elevated insulin requirements and insulin resistance of pregnancy. Gestational diabetes affects 2-4% of pregnancies and is a risk factor for both future diabetes in pregnancy as well as type 2 diabetes for mom and baby1. 1.1.3 Monogenic diabetes Alongside the more common forms of diabetes, there are also many different monogenic forms that cause approximately 1-6% of all cases of diabetes. Regrettably, monogenic diabetes is often not considered clinically and over 80% of patients get misdiagnosed as having type 1 diabetes or type 2 diabetes12. Monogenic diabetes can be classified as either maturity onset diabetes of the young (MODY) or neonatal diabetes (ND) which is either permanent (PND) or transient (TND). ND presents in the first 6 months of life whereas MODY presents later but usually before 30 years of age13. There have been at least 22 genes identified as causative for monogenic diabetes14, 15 including transcription factors like HNF1α (MODY3), NEUROD1 (MODY6), and PAX4 (MODY9), other β-cell factors like GCK (MODY2) and KCNJ11 (TND and PND), and even insulin (INS) itself (MODY10 and PND)13. These monogenic causes of diabetes provide valuable insight into the likely pathogenesis of more common type 2 diabetes because of the known associations between MODY gene variants and type 2 diabetes13. Additionally, patients with monogenic diabetes may be ideal candidates for gene therapies that could correct or replace the mutated genes. Of particular relevance to the current work, mutations in the insulin gene (INS) can cause both MODY and ND, and this can occur via multiple pathways. Insulin is made from a 4  precursor called proinsulin and the current prevailing theory claims that it requires cleavage at specific internal sites by prohormone convertase 1/3 (PC1/3) and 2 (PC2)16. Correct processing, disulphide bond formation, and folding are essential for production of mature insulin with full bioactivity. Mutation of the insulin gene was first identified in the 1970’s in a patient with a substitution of leucine in place of phenylalanine at residue 25 (F25L) This mutation caused diabetes in the patient and the mutated insulin had an impaired ability to bind its receptor and cause glucose uptake17. Since, more than 20 mutations have been identified and found causative in cases of PND and MODY, alternatively called “mutant INS gene-induced diabetes of youth”18. Most patients with insulin mutations develop severe neonatal diabetes19. Largely based on work in the Akita mouse with a C96Y mutation of the mouse insulin 2 gene (Ins2), the best available evidence suggests that insulin mutations cause diabetes via a roundabout path of endoplasmic reticulum stress18. Specifically, mutations in insulin act in a dominant negative fashion wherein mutated insulin misfolds and forms aggregates that incorporate even the bystander normal insulin. Most causative mutations occur near the cleavage sites to impair processing, and other mutations within the A- and B- chains alter disulfide bond folding and protein maturation20. These defects lead to buildup of misfolded insulin protein aggregates in the endoplasmic reticulum, limit β-cell expansion in the neonatal period21, and β-cells die from endoplasmic reticulum stress22. This bystander effect may also occur in humans because one study showed impaired production of fluorescently labelled human insulin in a heterozygous Akita mouse23. Interestingly, the ER stress hypothesis does not explain the pathogenesis of diabetes for all patients with insulin mutations causative for PND and MODY10. There are patients with mutations altering receptor binding affinity or mutations that alter insulin sorting or processing18. Another example includes one patient that was identified as having homozygous mutations in the intron of INS24. This patient does not produce misfolded proinsulin and thus retains surviving insulin-deficient β-cells as evidenced by normal circulating islet amyloid 5  polypeptide (IAPP); another hormone predominantly made in β-cells. As to whether insulin-deficient, IAPP-producing β-cells would be able to properly produce and secrete insulin after correction or replacement of the mutated insulin genes, is a question as of yet unanswered. 1.1.4 Reduced β-cell mass is a hallmark of diabetes Not only do almost all patients with diabetes face an insulin insufficiency, there is also a loss of β-cell mass. Patients with type 1 diabetes have extremely low β-cell mass by the time of diagnosis25. Though there is growing debate into the pathogenesis of type 1 diabetes including the role of β-cell dysfunction26, few would argue against the dogma that there is an eventual substantial reduction of β-cell mass with contributions from immune cells including cytotoxic T-cells and there is an association of autoimmunity with abnormal autoantibodies targeting β-cell proteins including insulin. This contrasts to type 2 diabetes in which patients may have a more modest reduction in β-cell mass near the time of onset (0-65%)27, 28. Though conventionally thought of as a disease dominated by insulin resistance, there is good evidence that there is an eventual loss of β-cells in patients with type 2 diabetes29. Both cell autonomous30 and systemic factors, including hyperglycemia, hypertriglyceridemia, and inflammatory cytokines31 contribute to β-cell dysfunction, failure, and death. Additionally, it has recently been proposed that though β-cell death does appear to occur eventually in mouse models of progressive type 2 diabetes, β-cell ‘dedifferentiation’ may occur before β-cell death32. Dedifferentiation is a process whereby mature cells revert to a progenitor state and lose their mature function. This may occur before β-cell death in type 2 diabetes because β-cell mass is often normal in postmortem pancreas collected near the time of diagnosis and β-cell mass is inversely correlated with disease duration29. Lineage tracing studies (a technique to label a cell or group of cells as well as all progeny) have shown that mouse β-cells in a hyperglycemic or in vitro setting can lose their hormone-producing status and become insulin negative33, 34. In severe mouse models of diabetes β-cells become immunoreactive for the endocrine progenitor marker neurogenin-3 6  (NGN3) and pluripotency markers octamer-binding transcription factor 4 (OCT4), v-myc avian myolocytomatosis viral oncogene lung carcinoma derived (L-MYC), and homeobox protein NANOG, and may eventually adopt an alternative cell fate, taking on a glucagon producing role35. The process of dedifferentiation appears to have a positive feedback loop of hyperglycemia driving β-cell dedifferentiation, but this process appears reversible with exogenous insulin34 and may be possible with other therapies to decrease insulin demands by reducing food intake36 or perhaps by correcting an underlying β-cell defect. Notably, the process of losing insulin expression as a marker of dedifferentiation has been shown in humans with type 2 diabetes37 but some findings counter that dedifferentiation happens rarely in type 2 diabetes38. Regardless whether patients face β-cell death or dedifferentiation, therapies aimed at regenerating or replacing β-cell mass could correct the underlying pathology of diabetes. Though there is great interest in developing therapies that correct this underlying pathological process driving diabetes progression, the most prevalent current best practice for the medical management of diabetes does not address the loss or dysfunction of β-cells.  1.2 Medical management of diabetes In 1922, researchers at the University of Toronto discovered the hormone insulin. These well-known scientists, John Macleod, James Collip, Frederick Banting, and Charles Best heralded in the next century of diabetes management – insulin replacement by injection. Though insulin injections have substantially decreased mortality associated with diabetes mellitus and its comorbidities, insulin injections come with risks, including dangerous hypoglycemia and inevitable periods of hyperglycemia, as well as causing considerable distress to patients. Patients with low socioeconomic status face disproportionately high rates of complications and comorbidities owing to challenges with self-management, with 2–3 times the risk of cardiovascular disease and death compared with patients with diabetes and 7  higher socioeconomic status39. For some patients that have hypoglycemic unawareness, classically labelled ‘brittle diabetics’, management is particularly challenging, and their blood sugars are extremely variable (Figure 1-2).  Figure 1-2 Glucose monitoring during diabetes Continuous glucose monitoring of a patient with type 1 diabetes using insulin injections to regulate blood glucose. Normal blood glucose for people without diabetes falls within the range in grey (70-140 mg/dL or ~4-8 mM) ~100% of the time. Despite this patient’s best effort, they spend the majority of the time with severe hyperglycemia and enter dangerous hypoglycemia several times during the week. Line colors represent different days. Figure adapted with permission from Cell Metab, 29(3) Latres et al. 545-63 (2019).   Since 1921, there have been tremendous advances in the medications and technologies available to treat diabetes. By 1983, recombinant human insulin was available40 and through research into modifications of structure and formulation, there has been impressive development of rapid and long acting insulins. These variable formulations allow for patients to better manage their diabetes with fewer injections and achieve better glucose control with less morbidity associated with planning injections41. Patients continue to face risks of hypoglycemia with intensive insulin therapies, but access to fast insulin analogues allow better flexibility for meals, and combination therapies have been associated with improved glycated hemoglobin (HbA1c; a measure of average blood glucose levels over ~3 months)42. 8   Though patients with type 1 diabetes are almost fully reliant on insulin and insulin analogues for glucose management, patients with other forms of diabetes like type 2 diabetes wherein many β-cells survive, have other options. Antihyperglycemic medications can be broadly classified as insulin secretagogues or insulin sensitizers. Insulin secretagogues act on β-cells to stimulate insulin secretion whereas insulin sensitizers act on tissues like the liver, adipose, and muscle to improve insulin sensitivity and thus reduce insulin demands. The cornerstone of type 2 diabetes management is the drug metformin, a drug in the biguanide class. Metformin was first approved in 1957 and by 2013 five of six patients with type 2 diabetes took metformin43. Though the specifics of its mechanism of action are unclear, it is generally accepted that a main component of metformin’s action is as an insulin sensitizer on the liver. Metformin suppresses hepatic glucose production to lower blood glucose and is associated with modest weight-loss44. Other insulin sensitizers include thiazolidinediones that act as peroxisome proliferator-activated receptor γ agonists to improve insulin sensitivity in hepatocytes and adipocytes45. Insulin secretagogues include the sulfonylurea, glucagon like peptide-1 (GLP-1) agonist, and dipeptidyl peptidase 4 inhibitor classes of drugs. Sulfonylureas act by binding the sulfonylurea receptor (SUR1; gene: ABCC8) on β-cells to increase the propensity for closure of the linked ATP sensitive potassium channel (Kir6.2; gene: KCNJ11) and thus increase the likelihood of depolarization and subsequent insulin release. Interestingly, sulfonylureas can lead to insulin independence for many patients with monogenic diabetes (both MODY and ND) caused by mutations in the sulfonylurea receptor gene or the linked potassium channel46. GLP-1 agonists are modified versions of the endogenous GLP-1 and bind the GLP-1 receptor to increase β-cell cytoplasmic cAMP to increase insulin secretion. Dipeptidyl peptidase 4 inhibitors suppress the enzyme that breaks down GLP-1, thereby increasing circulating active GLP-1 concentration. Though sulfonylureas pose a risk of hypoglycemia by causing excess insulin 9  secretion during periods of euglycemia47, the action of GLP-1 is glucose dependent, thereby only acting to lower blood sugar during periods of hyperglycemia. It has recently been appreciated that insulin secretagogue-based therapies may in fact worsen the long-term prospects for patients with type 2 diabetes48. As type 2 diabetes is associated with β-cell stress from high insulin demands, drugs that increase insulin secretion like sulfonylureas can lower blood glucose transiently but appear to accelerate diabetes progression over 5+ years compared to sensitizers like TZDs48. The story is much more complex for GLP-1 agonists that offer many benefits independent of stimulating insulin secretion, including delayed gastric emptying, improved satiety, weight-loss, and suppressed glucagon secretion leading to reduced insulin requirements49. Though endogenous GLP-1 with its very short half-life and secretion in meal-responsiveness appears to be protective for β-cells, long acting GLP-1 analogues developed to avoid the need for frequent injections may cause abnormal constant β-cell stimulation, thereby driving hyperinsulinemia and β-cell failure50. Regardless, GLP-1 agonists have gained great interest for their superior weight-loss effects51. Finally, though not falling in either above category, it is worth mentioning the sodium-glucose cotransporter (SGLT1 and SGLT2) inhibitor class of drugs. Both proteins are found in the tubules of the kidney nephrons (alongside the intestine, brain, liver, and other tissues) and normally transport glucose and sodium back into the blood52. Inhibition leads to a retention of glucose in the urine, thereby lowering blood sugar and offering benefits on weight-loss52. Though these therapies are promising and suitable for many patients, they continuously fail to address the fundamental pathological change in diabetes: a loss of functional β-cells.  1.3  β-cell replacement: the precedent of islet transplantation Unlike insulin injections and other currently available antihyperglycemic medications, cell-based therapies offer the potential for an effective cure for diabetes by replacing or 10  repairing the lost insulin-producing β-cells of the pancreas. Critically, it is because of the precedent set by islet transplantation53, 54 that we can be confident that replacement of bona fide pancreatic islets containing functional β-cells is curative. As early as 1894, physicians attempted to replace lost pancreatic tissue via a subcutaneous implantation of sheep pancreas into a 13-year-old boy55. It took another 106 years for a breakthrough in Edmonton, Canada with the development of an islet transplantation protocol that successfully induced insulin independence in seven patients for up to 14.9 months56. Patients with islet transplantation can achieve full insulin independence with normal glucose tolerance (Figure 1-3) and associated reduced risks of both acute hypoglycemic events and progression of chronic complications57. Since 2000, over 1,500 islet transplantation procedures have been performed worldwide including in Canada, the UK, France, the USA, and elsewhere53. Additionally, whole organ vascularized pancreas transplant has been investigated and performed ~30,000 times although is a substantially more invasive and challenging procedure with higher rates of complication58. The latest trials in adults report that ∼50% of recipients remain insulin-independent 5 years after islet transplantation53 and even using previous, less-advanced protocols, 73% of patients had improved glycemic control 10 years after transplantation, albeit not insulin independence53. Regrettably, as diabetes affects hundreds of millions worldwide, even if islet transplantation protocols were improved to achieve long-term insulin independence in all recipients, there will always be a severe shortage of donor islets.  11    Several approaches not reliant on cadaveric donor islets have been investigated to yield an adequate supply of β-cells for patients. These approaches can be broadly categorized as β-cell replacement by exogenous in vitro or non-human derivation of β-like cells from stem cell or other sources for delivery to patients, or in situ (re)generation of insulin-producing cells through reprogramming strategies, often via a gene therapy approach. These approaches have moved closer to being clinically viable therapies for diabetes in the past few years, but have not yet reached their potential to displace exogenous insulin injections as the standard for treating people with diabetes given that extremely few have even started early clinical trials. The standards for a cell or gene therapy that would replace insulin injection must be stringent, including the need for well-regulated insulin secretion in response to stimulation by glucose and other nutrients59 like mature healthy β cells60, avoidance of hyperinsulinemia and hypoglycemia, and improvement of patient quality of Figure 1-3 Continuous glucose monitoring after islet transplant Continuous glucose monitoring of a patient with type 1 diabetes 9 months after receiving an islet transplant. Normal blood glucose for people without diabetes falls within the range in grey (70-140mg/dL, approximately 4-8mM) ~100% of the time. Despite struggling to maintain euglycemia by insulin injections (Figure 1-2), this patient now has near perfect glycemic control and does not require vigilant monitoring of glucose and insulin injections. Line colors represent different days. Figure adapted with permission from Cell Metab, 29(3) Latres et al. 545-63 (2019). 12  life59. The challenges still facing cell-based therapies are, therefore, daunting but not insurmountable61, 62 (see Table 1-1 for relevant recent clinical trials). With this cautious optimism, a review of the present understanding of β-cell development is presented (Section 1.4) to provide context for the subsequent review of the literature attempting to generate β-cells in vitro (Section 1.5.1) or in situ (Section 1.5.2). These fields are summarized to highlight the clear need for more basic research into the development and function of mature β-cells to guide future cell or gene therapies for diabetes.   Table 1-1 List of relevant ongoing or completed trials into cell-based therapies for diabetes Implantation site Pre-clinical model Cells ClinicalTrial.gov Identifiers Phase Studies Subcutaneous Immunocompromised mice hESC-derived pancreatic progenitor cells NCT02239354 I/II Preclinical63, 64 NCT02939118 Follow-up  Subcutaneous Immunocompromised mice hESC-derived pancreatic progenitor cells NCT03163511 I/II Preclinical63, 64 Omentum Diabetic rats, Cynomolgus monkey Human islets NCT02213003 I/II Preclinical65 NCT02821026 I/II  Gastric submucosa Landrace or Yucatan miniature swine Human islets NCT01571817 I  NCT02402439 I Preclinical66, 67 Health CanadaPRO00049850 I  Intraperitoneal Nonhuman primates  Neonatal porcine islets NCT00940173 I/II  Mice NCT01736228 II Preliminary results68   NCT01739829 I/II  Portal vein Rodents, large mammals, and >1500 human islet transplantations Human islets NCT00434811 III  NCT01369082 III Trial results53, 54, 69, 70 NCT01897688 III  NCT00679042 III    13 1.4 β-cell development and maturation  In both humans and mice, the developing embryo forms a trilaminar structure composed of endoderm, mesoderm, and ectoderm. From these three cell types comes every cell of the developing organism. Of relevance to the current work, pancreas and most of the gastrointestinal tract including the liver and intestines arise from an endoderm lineage. After gastrulation, formation of the definitive endoderm, and invagination to form the primitive gut tube, cell types diverge along the anterior-posterior axis. Antagonists of WNT (wingless/integrated)71 and FGF4 (fibroblast growth factor-4) from the primitive streak-mesoderm72 create a gradient along the anterior-posterior axis. Next, the pancreatic organ bud arises from a PDX1 (pancreatic and duodenal homeobox 1) expressing foregut/midgut tissue between the developing SOX2 (Sry box-2) expressing stomach and CDX2 (caudal-type homeobox 2) expressing intestinal progenitors73. Downstream, elevated FGF/BMP (bone morphogenetic protein) from cardiac mesoderm favors liver via HHEX (haematopoietically expressed homeobox protein Hhex) expression in the ventral foregut and lower levels drives Pdx1 and Sox17 expression for pancreatic and biliary system development respectively. Suppression of Shh by activin represses hepatic and biliary fates while retinoic acid drives high PDX174. From PDX1+ cells, specification to the pancreatic lineage involves NKX6.1 (homeobox protein Nkx6.1) induction75 before pancreatic endoderm designates into one of the three key pancreatic lineages: endocrine islets, exocrine acinar tissue, or ductal tissue. Notably, though PDX1 is essential for any pancreatic development76, NKX6.1 is not essential for all pancreatic tissue as there can be development of all pancreatic cell-types except β-cells in a knockout mouse model77. Patterning next occurs along a complicated tubular epithelium duct tube in which antagonistic signals of PTF1a (pancreas associated transcription factor 1a) and NKX6.178 facilitate segregation into GATA4+ (GATA binding protein 4) PTF1a+ “tip” cells that go on to develop into the acinar tissue and GATA4 negative “trunk” cells. Bipotent trunk cells can form  14 either ductal tissue or via suppression of Notch signaling79 there is a transient wave of NGN3 (neurogenin 3) to designate an endocrine fate80. Recent work found that interaction of the extracellular matrix with integrin α5 acts as a cell-autonomous cue favoring Notch signaling and the eventual ductal cell fate81.  Following induction of NGN3, endocrine progenitors must next designate into a specific endocrine cell type. In mice, the first cells formed are α-cells during a primary wave of NGN3 before formation of the other cell types in a secondary wave of NGN382. In humans, there is no evidence for this two-wave nature of endocrine cell development and insulin+ β-cells are the earliest observed endocrine cell type83. Knockout mouse models suggest that some transcription factors are important for the core endocrine program beyond NGN3, including ISL184 (insulin gene enhancer protein 1) and NEUROD185 (neurogenic differentiation 1). Subtype specification includes PAX4 (paired box 4) driving the β-cell and δ-cell pathway86 in balance with the opposing actions of ARX (aristaless related homeobox) driving the α-cell and PP-cell pathway87. There are many other factors that play roles in designating cell development and are important for normal mature function. These include PAX6 (paired box 6) and NKX2.2 (homeobox protein 2.2) that are both required for normal α- and β-cell development and suppress development of ghrelin expressing cells88-90. Interestingly, despite being a pan-pancreas transcription factor early in development, PDX1 later becomes restricted to β-cells and δ-cells and is required to actively maintain the β-cell fate91.  Following development of insulin expressing β-cells by embryonic day 10.5 in mice92 and 7 to 8 weeks post conception in humans75, there is progressive maturation of β-cells until after birth. Immature β-cells contain insulin but lack the necessary machinery for regulated storage and secretion of fully processed insulin. Characterization of β-cell maturity is useful for studies on mouse models of β-cell development and function, the assessment of the β-cell phenotype in health and disease, and the evaluation of β-like-cells developed as potential cell therapies to  15 treat diabetes. Arguably functional assessment is the most direct measure of β-cell maturity but histological examination of markers of maturity is a useful alternative in cases where having a live pure endocrine cell population is not practical. Beyond insulin immunoreactivity, along the developmental trajectory towards mature β-cells the earliest markers include PDX1 and the post-NGN3 factor NKX2.2. After early broad expression, NKX6.1 later becomes restricted to β-cells downstream of NKX2.277. The next markers include pan-islet PAX6 and the hormone islet amyloid polypeptide (IAPP) is found in β-cells and some δ-cells77. Moving closer to mature β-cells, membranous glucose transporter (GLUT1 in humans or GLUT2 in mouse) and the prohormone processing enzyme PC1/3 (prohormone convertase 1/3) become expressed93. Glucose transporters are essential for mature glucose-stimulated insulin secretion and PC1/3 is essential for proper processing of proinsulin into mature insulin. Currently, the most specific markers for mature β-cells are the transcription factor MAFA (V-maf musculoaponeurotic fibrosarcoma oncogene homolog A) and the proposed β-cell maturity marker UCN3 (urocortin-3) that is also found in α-cells in humans94. MAFA binds the insulin promoter and there is good justification for it to be considered the prototypical mature β-cell marker. Loss of MAFA leads to loss of glucose stimulated insulin expression, a reduction in islet insulin content, and mild defects in glycemia95. Additionally, forced overexpression of MAFA in neonatal rat islets led to a lowering of basal insulin secretion and improvements in glucose-stimulated insulin secretion (GSIS)96, both key features of mature β-cells. With that being said, it is possible that despite the functional role of MAFA in the mature β-cell phenotype, it may not be fully specific for functionally mature β-cells. Nonetheless, assessing β-cell markers including MAFA may be a useful way to assess the developmental and maturity status of β-cells (Figure 1-4). Based on the work characterizing normal development briefly summarized in this section, there has been much work into mimicking this process to produce β-cells in vitro (section 1.5.1) or induce formation of β-cells from other cell populations in situ (section 1.5.2).  16  Figure 1-4 β-cell development Summary of key steps along the development trajectory of β-cells in mice. From the endoderm lineage arises the pancreatic buds composed of pancreatic progenitors. Eventual segregation into a tubular epithelium separates GATA4+ tip cells and GATA4- trunk cells that are bipotent endocrine or ductal progenitors. There are two waves of endocrine cell development with transient waves of Ngn3 expression. The secondary transition is responsible for forming most β-cells. Some markers associated with each stage are listed above and suppressed signaling pathways necessary to allow the pancreatic and β-cell program are shown in red.   1.5 β-cell replacement  1.5.1 Generation of β-cells in vitro 1.5.1.1 Insulin secreting cell lines Cultured cell lines have been produced from rodents and used as important tools to probe β-cell function and may be potential cell sources to implant into patients with type 1 diabetes for insulin replacement. Building β cells from non-β cells (such as pituitary cells or human embryonic kidney cells) has been attempted by introducing the insulin gene along with various components of the stimulus–secretion coupling pathways found in native β cells97, 98, but these cells lack the capacity to adequately store fully processed insulin and to release it with the appropriate regulated kinetics that will be necessary to safely treat type 1 diabetes in humans59. Generation of a rat β-cell line (INS-1) with glucose responsiveness has been attempted by either clonal selection99 or forced expression of glucose-sensing molecules GLUT2 and glucokinase100. Moreover, with a view towards potential clinical use for insulin replacement therapy, these INS-1 β-cells have been stably transfected with the  17 human insulin cDNA to create human insulin-secreting cells99, 100. An oncogene approach (large T antigen of simian virus 40) has been used in human fetal tissue to produce human β-cell lines, one of which (EndoCβH1 cells) shares many features with mature human β-cells including glucose-induced release of insulin101. To address the proliferative nature of the cells, conditionally immortalized human β-cell lines were generated using Cre-mediated excision of the immortalizing transgenes101 and arrest of proliferation improves insulin expression, content, and secretion. Although concerns regarding the transformed nature of the cells, along with limited proliferative capacity and stability of the cell phenotype, might preclude clinical use, these cells continue to provide a valuable source of human insulin-secreting cells for research purposes. 1.5.1.2 Xenotransplantation Porcine islets have been suggested as a safe and effective alternative source of human compatible islets102, supported by successes in treating models of type 1 diabetes in heavily immunosuppressed nonhuman primates transplanted with neonatal103 or adult104 porcine islets. Challenges of xenotransplantation include strong xenoimmunity and transmission of porcine endogenous retroviruses. Though it has been suggested that this zoonotic retrovirus might preclude generation of a porcine-derived cell product105-107, genetic editing technologies have successfully led to a >1000fold reduction in transmission of porcine retroviruses to human cells in vitro108. Other have envisioned tackling xenogenicity by using apancreatic pigs to grow human organs such as the pancreas109, 110. Remarkable proof-of-principle interspecies organogenesis has been obtained through a process termed 'conceptus complementation', whereby whole organs were generated from donor pluripotent stem cells (PSCs) using their chimera-forming ability to complement organogenesis-disabled host animals in vivo. Specifically, rat wild-type PSCs were injected into Pdx1−/− (pancreatogenesis-disabled) mouse blastocysts, generating a normally functioning rat  18 pancreas in mice111. Even if it is possible to replicate this process by generating human pancreas with human PSCs in available apancreatic pigs112, considerable technical and ethical issues must be resolved before proceeding with the generation of human cells within pigs. 1.5.1.3 Stem-cell-derived β cells Stem cells have been investigated as a potentially unlimited alternate source of β-cells because of their ability to self-renew indefinitely and their potential to differentiate into different cell types113. On the basis of early animal studies114, and studies showing progressive decreases in insulin requirements following transplantation of human fetal pancreata115, 116 or islets117 into a small number of adults with type 1 diabetes, there is evidence that immature islets can expand and/or mature in vivo to produce sufficient insulin to meet adult requirements. Thus, one approach is to deliver pluripotent stem cells differentiated to pancreatic progenitor cells in vitro64 and allow the maturation process to finish in vivo. Pancreatic progenitor cells can prevent diabetes onset after streptozotocin (STZ)-induced destruction of endogenous mouse β-cells in immunocompromised animals64. Pancreatic progenitor cells have also been shown to mature into functional β-like cells in mice with diabetes in several months118 and with refinements to differentiation protocols the time post-implantation to in vivo maturation is decreasing64, 119, 120. In 2014, ViaCyte launched a phase I/IIa clinical trial (US National Library of Medicine: NCT02239354 (2017)) testing the safety and efficacy of pancreatic progenitor cells differentiated from hESCs and implanted within a macroencapsulation device. Though the survival of implanted cells was evidently inadequate121, a new trial with a modified encapsulation device is currently underway (US National Library of Medicine: NCT03163511 (2018)). To minimize the impact of host microenvironment on maturation post-transplantation, there is also great interest in differentiating pancreatic progenitors fully to mature β-cells in  19 vitro. Efficient differentiation of hESCs into cells that expressed key β-cell transcription factors including MAFA, as well as response to glucose in vitro, has been demonstrated122 but these cells still lack some functional characteristics compared with adult β cells. These more mature hESC-derived cells were able to reverse diabetes in mice approximately four times faster and with 75% less cells than the pancreatic progenitor cells122. Although this progress is encouraging, current differentiation protocols with hESCs still fall short of producing mature β cells61, 62. Despite nuclear immunoreactivity for MAFA, poor phasic insulin secretion and abnormal Ca2+ flux in response to glucose suggest that there are other undefined factors contributing and defining mature β-cells that are not being considered. A better characterization of the signals underlying β-cell development could guide future maturation protocols to yield more mature β-cells and minimize challenges related to cost, efficiency and timelines of production123. 1.5.2 Reprogramming cells in situ An alternative strategy to transplanting insulin-producing cells into patients is to induce formation of insulin-producing cells in situ through ectopic expression of key β cell transcription factors such as PDX1 and MAFA, or insulin itself, usually via gene therapy approaches. These approaches re-direct non-β cells towards the β cell fate through expression of β-cell- specific transcription factors or by recruiting non-β cells for insulin production. Despite considerable optimism that this strategy could lead to a cure for diabetes, there are important technical limitations and clinical considerations of an in situ approach including: 1) There is limited possibility of complex, controlled and multistep strategies. 2) There will be challenges protecting patients from dangerous exposure because it is not readily feasible to selectively remove a delivered gene therapy (e.g. viral vector) nor could it be perceived as a reasonable clinical resolution to surgically remove a target tissue such as the liver or pancreas if an adverse event occurs. 3) Though a cell product could be  20 encapsulated or contained with a predetermined space, a gene therapy product would almost certainly involve systemic spreading with unavoidable off-target infection of host cells. Although some have attempted to use non-endoderm tissues including muscle124 or pituitary97, 125 to generate insulin-producing cells, starting from a more similar endodermal tissue may be a more feasible strategy. 1.5.2.1 Reprogramming pancreas Ductal tissue As ductal tissue develops extremely similarly to endocrine tissue it is a logical cell source to convert (transdifferentiate) to β-cells. Though the postnatal pancreas has well-established cell populations and there is limited expansion and conversion between ductal, acinar, and endocrine cell types during times of normal physiology126, in certain pathological states β-cells may arise from acinar or ductal cells. After pancreatic duct ligation NGN3 is necessary for newly forming β-cells127 and lineage tracing of NGN3+ cells label the majority of new β cells128, 129. Contrarily, by using a HNF1β promoter based lineage tracing model that labelled ductal epithelium130, others demonstrated that it was only during embryogenesis that duct cells formed islets130, 131. But this work was again countered using a duct specific SOX9 promoter-based reporter that suggested that, in certain situations, ducts can be a source of new β cells, such as during a setting of moderate hyperglycemia with low-dose growth factors132. Other work has suggested that inflammatory cytokines are responsible for initiating the epithelial-to-mesenchymal transition to form endocrine cells from ducts133. There has also been evidence that a fasting mimicking diet can induce expression of NGN3 in ductal cells and promote β-cell regeneration in mouse models of diabetes134. In an attempt to avoid the caveats of potential growth factor therapy to induce in vivo transdifferentiation of ducts to endocrine cells, protocols to culture duct cells127 and direct ducts to form insulin-secreting cells in culture135 have been developed. Expressing the endocrine pancreas transcription  21 factors NGN3, PAX6, MAFA, and PDX1 in both human and rodent ductal cells induced formation of islet-like clusters136, 137. However, the clinical utility of human ductal cells might be limited given the need for cadaveric donors unless there is substantial progress towards large-scale duct cell isolation and expansion techniques138. Overall, the literature on the potential of converting duct tissue to β-cells lacks a consensus on mechanism and more research is needed to develop effective and replicable transdifferentiation approaches. Acinar tissue Though not as developmentally similar to the endocrine pancreas as the ducts, acinar tissue is an appealing β-cell source because it makes up >90% of the pancreas. Growth factor therapies like delivery gastrin139, 140 or gastrin with epidermal growth factor (EGF)141 have been proposed to induce expansion of β-cell mass from a ductal source, however this hypothesis was based only on the co-expression of a ductal marker with insulin142 and PDX1143. Lineage tracing experiments suggested that after delivery of EGF with ciliary neurotrophic factor, new islet cells came from NGN3 expressing elastase+ acinar tissue144. This contrasts earlier work that suggested that endocrine cells do not arise from acinar tissue in response to transforming growth factor α overexpression145. The ectopic expression of transcription factors in acinar tissue is another method to form β-like cells. Adenovirus delivery of Ngn3, MafA, and Pdx1 were found to convert infected cells to an insulin-expressing fate in mice146. Lineage tracing suggested an exocrine cell source and the therapy could reverse STZ-induced diabetes within 1 week. Furthermore, it was shown that subsets of these vectors could also be used to selectively induce generation of two other islet cell types, somatostatin-producing δ cells and glucagon-producing α cells147. This work has been replicated with other viral vectors148 and some suggest that acinar reprogramming may be dependent on euglycemia149. Other work highlights inflammation and macrophage infiltration as factors preventing acinar-to-β-cell transdifferentiation in favor of acinar-to-ductal  22 metaplasia150. In related work on isolated human exocrine tissue, combining NGN3, MAFA and PDX1 overexpression with the exogenous transcription factor PAX4 and suppression of homeobox protein ARX151, or suppression of the epithelial–mesenchymal transition using Rho-associated kinase and transforming growth factor-β1 inhibitors152, could induce formation of glucagon negative insulin+ cells. Despite continued investigations in this area, systemic delivery of viral vectors remains problematic for clinical translatability, with poor acinar tissue selectivity and potentially dangerous complications of overexpressing transcription factors153. The literature has highlighted that there is still much to understand relating to cell plasticity and the delicate regulation of cell differentiation in the postnatal pancreas. Non-β islet cells The developmental similarity of all islet cells suggests that non-β cells could also be a source of new β cells. Under conditions of extreme β-cell loss, both α cells and δ cells naturally transdifferentiate to β cells154, 155, albeit they only make a partial contribution to the regeneration of ~1% normal β-cell mass over 1 month. Immature β cells existing at the islet periphery with transcriptomes similar to pancreatic progenitor cells156-158 have been described capable of becoming mature β cells or α cells in mice. Expression of Cfap126 (encoding protein Flattop), a gene that is transcriptionally activated during establishment of planar cell polarity, has been shown to increase as the islet microenvironment and 3D structure develops during β-cell maturation and is associated with a mature and non-proliferative state159. CFAP126 is decreased in human islets from patients with pre-diabetes or type 2 diabetes compared with islets from donors without diabetes, suggesting that the Wnt/planar cell polarity pathway is adversely affected in these metabolic states159. Manipulating or repairing the Wnt/planar cell polarity pathway might enable α-cell transdifferentiation or mature β-cell expansion to meet increasing insulin demands.  23 It has been demonstrated that ectopic expression of Pax4 was sufficient to transdifferentiate α-cells into bona fide β-cells in mice160 and expression of Pax4 under a Gcg promoter could counter STZ induced diabetes in adult mice161. Lineage tracing experiments suggested generation of β-cells from α-cells161 and the authors hypothesized that α-cell depletion activated as of yet unidentified signal(s) to initiate the endocrine program in duct cells to replace lost α cells. These glucagon-positive cells were then forced into a β cell program through ectopic Pax4 expression following activation of the glucagon promoter. This continuous duct–α–β cell cycling was blocked by glucagon supplementation and thereby prevented progressive islet enlargement. The function of PAX4 to induce the β-cell program is mirrored by the function of ARX to maintain the α-cell identity while suppressing PAX487. Mice lacking Arx expression in α cells displayed a similar phenotype to Pax4 overexpressing mice with lineage tracing evidence of an α-cell source of β-cells162. A pair of recent studies purported that GABA has been identified as a suppressor of ARX capable of inducing islet hyperplasia and neogenesis163 and the anti-malarial drug artemisinin acts as a ligand of the GABAA receptor on α cells, capable of increasing β-cell mass in zebrafish embryos and suppressing ARX expression in human α cells164. Disappointingly, others have been unable to replicate these findings165, 166. Potential explanations could include different lineage tracing models or perhaps unidentified and unappreciated confounding variables were not accounted for that contributed to the suppression or activation of the β-cell phenotype.  β-cells Assuming sufficient β cells remain in individuals with diabetes, or pre-diabetes, it is conceivable that these cells might be expandable and, therefore, could be used as a source of new β cells. β-cell mass expands several fold from birth to adulthood, primarily as a result of β-cell replication167. Observations of greater β-cell mass in humans during pregnancy168 and in the setting of obesity169 compared with lean controls provides direct evidence for the  24 expansion capacity of adult human β cells. β-cell replication is likely regulated at multiple checkpoints in the cell cycle170, 171 and signaling through the insulin receptor has been implicated171. Proteins in the insulin signaling pathways are downregulated in islets from donors with type 2 diabetes172, potentially contributing to the inability of β cells to expand sufficiently in these patients. Elegant studies in mice indicate that serotonin acts downstream of lactogen signaling to stimulate β-cell proliferation during pregnancy173 and there is hope that continued investigation of the mechanisms controlling both β-cell proliferation and survival might lead to the development of successful therapeutic strategies to enhance the response of β cells to increased metabolic loads174. Thus, there has been considerable effort to map the cellular pathways involved in β-cell replication170, 175 and screens have identified small molecules that promote proliferation of human β-cells176-179. Collectively, these are promising findings that suggest it might indeed be possible to develop small molecules for restoration of β-cell mass in humans, but several challenges remain in translating these findings into an effective therapy for diabetes180. Rigorous testing is needed to determine how to achieve adequate β-cell proliferation for sufficient expansion in a safe and reversible manner and it will need to be established that approaches to drive entry of β-cells into the cell cycle do not lead to de-differentiation or damage to the β-cells that impair their function. Further, development of drugs that could drive selective expansion of just β-cells but not other cell types is a daunting roadblock yet to be addressed. Additionally, without capabilities to monitor in vivo β-cell mass or suitable biomarkers, dosing and monitoring such therapies would be challenging. 1.5.2.2 Reprogramming hepatocytes Liver and pancreas development diverge based on minimal factors and, therefore, the liver might be a suitable target for reprogramming181. Strategies targeting the liver have utilized two general approaches: transcription-factor-based reprogramming and insulin gene  25 therapy. Adenoviral delivery of Pdx1 to the liver can generate insulin-producing cells capable of reversing streptozotocin-induced diabetes182 for a prolonged period183 and PDX1 overexpression can reprogram human hepatocytes to produce insulin in vitro184. Beyond Pdx1, adenoviral delivery of Ngn3185 or a combination of Ngn3 and Pdx1 can induce insulin expression in rodent liver186, 187. Other cocktails including Pdx1–Ngn3–MafA188, 189, Pdx1–NeuroD1–MafA190, 191, and Ngn3–Btc–Socs1192 have been used to induce insulin secretion from the liver. However, there is presently insufficient evidence that these strategies produce a complete transdifferentiation of hepatocytes into β cells. Nonetheless, even if we posit that these strategies can guide a hepatocyte to gain all the features needed for glucose-regulated insulin production, delivering transcription factors systemically with unpredictable degrees of hepatocyte reprogramming and variable off-target effects probably poses unacceptable risks to patients. These and other gene or cell therapy approaches could benefit from the inclusion of a clinically acceptable safety switch to allow ablation of target or dysfunctional cells193. Insulin gene therapy of the liver Over two decades ago, it was shown that nonspecific hepatic delivery of rat Ins1 by adenoviral infection could protect rats from diabetic ketoacidosis194. This experiment highlighted that, even without full dynamic regulation, a hepatic cell source of insulin could provide life-saving support in rodents. Other hepatic gene delivery methods have been tested in rodents including lentivirus195-197 or adeno-associated viruses198, 199 capable of reducing diabetes complications200-202 in mouse models. Insulin gene therapy generated insulin-secreting human hepatocytes that reversed diabetes when transplanted into pigs, but immortalization of hepatocytes required use of a human telomerase transgene, which could contribute to oncogenesis203. Including glucose-responsive elements in the promoter improved glucose-regulated transcription204 and modifications to the proinsulin processing sites improved processing in hepatocytes205 but glucose-regulated insulin secretion  26 equivalent to that from β cells has not yet been achieved. Much like transcription-factor-guided approaches, risks associated with nonselective delivery of insulin to hepatocytes and other off-target tissues suggests that, without substantial developments in this field, none of these approaches will be able to replace insulin injections as the standard therapy for diabetes. 1.5.2.3 Reprogramming the gastrointestinal tract The gut mucosa is a complex organ with many cell types including a population of enteroendocrine cells secreting serotonin, GLP-1, or glucose-dependent insulinotropic polypeptide (GIP)206. Development of all enteroendocrine cells follows a similar signaling pathway to β cells, dependent on NGN3 and inhibited by the transcription factor HES1 (hairy and enhancer of split 1)207, 208. The observation of co-expression of the transcription factor forkhead box protein O1 (Foxo1) with Ngn3 during enteroendocrine cell development (E14.5 in mice) led to a hypothesis that FOXO1 has a role in enteroendocrine development. Whole-body Foxo1 ablation had no effect on mouse pancreatic endocrine development209 but mice with selective knockout of Foxo1 in Ngn3-expressing cells developed MAFA and PDX1 expressing glucose-responsive, insulin-expressing cells in the gut that despite producing only 1% of the C-peptide of mature β-cells reversed diabetes within a week after STZ injection210. A reduction in serotonin-expressing cells in the gut suggested that this subpopulation of enteroendocrine cells were the source of the β-like-cells210. Drugs that block FOXO1 action in the gut, therefore, might be a target for producing surrogate β cells. In a similar strategy to that used with acinar tissue146, combined expression of Ngn3–MafA–Pdx1 has been utilized to induce formation of so-called neo-β cell islets in intestinal crypts211 and more extensively in the gastric antrum212 which could reverse multiple bouts of STZ induced diabetes and protected mice from diabetes after total pancreatectomy. The cells of the antrum have a long cell lifespan to sustain expression of non-DNA-integrating  27 therapies, express the insulin-processing enzymes PC1/3 and PC2, and have a similar transcriptome to islets, which could be advantageous features over other portions of the gut212. However, genetic modifications and adenoviral vectors carry risks and technical limitations213, and achieving optimal and reproducible dosing might be extremely challenging. Luminal gene214 or small molecule reprogramming factor215 delivery might be an alternative strategy to form pancreatic tissue from the gastric mucosa, though such an approach would also have many technical challenges including controlling the exposure to reprogramming factors and achieving consistent efficiency. Insulin gene therapy of the gastrointestinal tract Despite their main role as exocrine glands, salivary glands have been examined as possible targets for insulin replacement by insulin gene therapy216. The salivary glands are capable of endocrine secretion of growth hormone217, readily accessible, well-encapsulated to limit off target vector spreading, and every salivary gland is not necessary for life so that in the event of an adverse event a gland could be relatively easily removed with limited risk218. Despite poor processing of proinsulin in the salivary glands, delivery of a modified proinsulin with increased insulin bioactivity to the parotid and submandibular glands could ameliorate toxin-induced diabetes in mice, much like gene therapies in the liver219. A phase I clinical trial published in 2012 utilizing adenoviral-mediated transfer of the aquaporin-1 cDNA to the parotid gland for radiation-induced salivary hypofunction yielded positive safety and efficacy results, thereby providing considerable support for the feasibility of this gene therapy approach in humans220. However, whether sufficient quantities of insulin could be released from such genetically modified salivary glands in humans, with the right kinetics to safely and effectively control glucose homeostasis in patients with insulin-dependent diabetes is unclear. Indeed, it has been noted that the relative proportions of exocrine versus endocrine secretion of proteins from salivary glands is unpredictable, and in the absence of an  28 understanding of the mechanisms through which both regulated and constitutive pathway secretory proteins are sorted, further clinical application for this strategy is severely hindered221. In 2000, gut K cells were proposed as surrogate cells for insulin production in view of their natural capacity for meal-regulated secretion of GIP222. Transgenic insertion of a GIP promoter driving the INS gene protected mice from diabetes via STZ-mediated destruction of endogenous β cells222 or autoimmune-mediated β-cell death in NOD mice223. Induction of immune tolerance from gut insulin exposure in both cell-mediated and humoral immune responses may protect insulin-producing cells223. This approach does not induce transdifferentiation of K cells into bona fide β cells but instead takes advantage of the endogenous capacity of K cells for storage and release of GIP in a meal-dependent manner. The pattern of GIP release during feeding is normally like insulin, such that inducing K cells to produce insulin in patients with type 1 diabetes could be a method to re-establish normal patterns of insulin secretion. Notably, using the regulatory sequences from the GIP promoter as a strategy to drive insulin expression conferred cell specificity in the transgenic mouse models, suggesting that a gene therapy vector might not need to be selective to K cells. However, this research has yet to be translated to a viable therapeutic and, much like other gene therapy strategies, it will be a challenge to deliver an insulin gene therapy product in a safe way that can reproducibly produce dose-dependent effects. In consideration of the breadth of β-cell replacement and regeneration strategies, an important common thread is incomplete characterization of the developmental pathways guiding interventions and inconsistencies in adequate formation and identification of mature β-cells. Despite tremendous development progress towards differentiation towards immature β-cells in vitro122, growing mature β-cells remains elusive. Despite tremendous excitement around non-specific gene therapies for the induction of insulin expression in pancreas, liver,  29 and gut, these therapies remain uncontrolled, unregulated, and may lack adequate control of the factors contributing to transdifferentiation. Notwithstanding these hurdles, additional efforts towards characterizing β-cell development and function will be helpful for guiding the production of β-cells in vitro and in situ. Development of genetic tools for in vivo genetic manipulation and gene therapy technologies could be useful for this purpose.  1.6 Gene therapy There are currently over six hundred clinical trials registered in US National Institutes of health studying “Gene Therapy”224. Broadly and conventionally speaking, gene therapy is a treatment method to replace mutated or deleted genes (e.g. mutations in clotting factors) to correct genetic disorders (e.g. hemophilia). Importantly, the possibility for gene therapy can extend beyond such stringencies, to counteract complex pathologies such as autoimmune disorders and provide genes to suppress local autoimmune attack225 or selectively target cancer cells to suppress oncogenesis or hone immune cells226. Additionally, gene therapy can be used to repair mutated or deleted genes. 1.6.1 In vivo genetic manipulations In mouse models, one of the most useful and commonly used genetic tools to study the role of specific genes in specific tissues is the Cre-LoxP system. Cre recombinase is an enzyme that recognizes LoxP sites in the genome and based on orientation, can excise, flip, or translocate targets. By delivering the Cre recombinase in a tissue-specific manner, LoxP flanked sites can be deleted in a tissue specific manner227. Additionally, Cre mediated recombination can be controlled in not only space but also in time, by fusing Cre to a modified estrogen receptor (ER). By fusing Cre to the ER, Cre is retained in the cytoplasm and is thus unable to bind DNA until tamoxifen metabolic products endoxifen or 4-OHT bind the ER and translocate Cre to the nucleus. Though using LoxP site containing mice makes this approach suitable, this  30 is not possible in human patients. As an alternative, CRISPR/Cas9 mediated gene correction or mutation could be used in humans228. Briefly, guiding gRNA interact with the Cas9 protein and identify complementary sequences in the genome for selective cutting followed by either non-homologous end joining, or homology directed repair. With the seemingly endless therapeutic possibilities with this technology, it is important to consider the delivery tools of gene therapy. These vectors must ideally act as an undetected “Trojan Horse”, capable of evading all immune reaction and selectively targeting only cells of interest. Perhaps the vector that has come closest to meeting these criteria, is the adeno-associated virus (AAV). 1.6.2 The adeno-associated virus The AAV was first thought to be an impurity of the adenovirus preparation229 but was in fact a parvovirus thereafter name “adeno-associated”. Since, unique features have been discovered that make the AAV highly suited to use as a clinical gene therapy vector. First, the AAV is considered non-pathogenic as it causes little to no immune response230. Most of the population has been infected by wild-type AAV without any obvious or common symptoms231. Second, there are many AAV serotypes with tropism for a variety of tissues 232 and capsid modifications can improve selectivity and change the immunological profile233. Third, the AAV’s structure suits a gene therapy vector as a non-enveloped single stranded DNA virus. The wild-type genome encodes replication and capsid proteins and is flanked by two inverted terminal repeats (ITRs)234. Though the wildtype single-stranded AAV takes upwards of 4 months to initiate gene expression235, modification to remove a component of the 3’ ITR, can allow packaging as a self-complementary double-stranded virus capable of initiating gene expression less than one week after infection236, 237. Packaging of a double stranded AAV is limited to only ~2.5kb, though as only the ITR is necessary for packaging, the remaining ~2.3kb can be engineered with a suitable promoter and gene of interest. Finally, the AAV genome is highly stable, enabling prolonged transgene expression with reports of expression four years after  31 treatment in a human238. Taken together, non-pathogenicity and fast yet prolonged expression, has resulted in the AAV as being one of the most studied gene therapy vectors. Another important consideration for clinically viability is manufacturing. The AAV is replication deficient and is hence sub-classified as a ‘dependovirus’. The AAV depends on the functions of a helper virus, such as an adenovirus or herpesvirus to replicate in a host mammalian cell239. Clinically, this means that the AAV cannot autonomously replicate in a host but also means that large-scale manufacturing is challenging and costly. Briefly, adenovirus independent production of a pure recombinant AAV is done by a triple plasmid transfection system240 (Figure 1-5). Adherent HEK293 cells are transfected with the construct of interest, a plasmid containing the AAV specific replication and capsid genes, and a third plasmid expressing the essential adenovirus genes232. Use of adherent cell cultures with serum    Figure 1-5 Conventional production of recombinant AAV requires a triple transfection of adherent HEK293 cells. The first plasmid contains the construct of interest flanked by AAV inverted terminal repeats, the second plasmid carries the AAV Rep and Cap genes essential for virus packaging, and the third plasmid carries the five essential adenovirus helper genes. High purity AAV carrying the double–stranded construct of interest can be isolated (e.g., by heparin affinity column chromatography). Virus can be delivered by many methods (intraperitoneally, intramuscularly, intravenously, or by direct delivery to a target organ such as via the pancreatic duct) and infects target cells using an essential receptor. Peak gene expression occurs within two weeks.  32 supplemented media limits production of virus meeting current good manufacturing protocols,241 but there are other technologies,242 and some recent protocols use serum-free suspension cultures thus improving production efficiency243. With these promising advancements, scalability at bearable costs will lead to greater opportunity for an AAV therapy to reach more patients. Despite mostly promising features, as the AAV has gained greater attention, previously unappreciated risks and challenges to using the AAV have been unearthed. To date, there have been limited reports of an association between AAV infection and spontaneous abortion244 and a recent paper suggested that a certain serotype of AAV (AAV2) may be linked to hepatocellular carcinoma245, but these findings have since been heavily challenged246-248. Given the frequent presence of the wild type 3’ ITR in these cancers, the wild-type AAV may pose a small risk247 but causality of AAV integration and carcinogenesis is lacking. Notably, even if carcinogenesis is supported, this may not be a risk for modified AAV vectors since clinical AAVs only use 145bp (the 5’ ITR) from the wildtype genome. The second main challenge is the presence of AAV neutralizing antibodies in humans249. This was unexpected based on preclinical studies and may explain the acute elevation of liver enzymes following liver AAV infection250. Nonetheless, there are many strategies to avoid this roadblock such as improving vector efficiency to reduce doses needed or transient immune suppression before AAV administration251, and the research community has seen many successful clinical trials for many monogenic diseases - more than 120 published clinical trials using AAV vectors have failed to find severe side effects. 1.6.3 Clinical use of AAVs Since the first AAV trial to treat cystic fibrosis in 1996252, AAVs have been studied to treat hemophilia B, rheumatoid arthritis, Duchenne’s muscular dystrophy, Leber’s congenital amaurosis, lipoprotein lipase deficiency, and many other diseases.253 Among the most promising research includes treatment for hemophilia B, a disease characterized by impaired blood clotting due to insufficient factor IX250. In ongoing Phase I/II clinical trials, there have been  33 reports of patients being free of multiple weekly factor IX infusions for over a year after a single AAV injection with maintained factor IX levels sufficient to maintain normal clotting times comparable to healthy counterparts254. There has also been promising clinical research using AAVs to treat Leber’s congenital amaurosis, a cause of childhood blindness. Early clinical trials showed improved visual acuity weeks after replacement of the mutated gene (RPE65) by AAV255. Stage III clinical trials256 by the biotech company Spark Therapeutics have claimed incredible findings: having treated 29 patients, all demonstrate profound improvements in light sensitivity and eye mobility and Spark Therapeutics report that there have been no “product related serious adverse events”257. Through all these promising findings on AAV based therapies, the trailblazer into clinical approval is treatment for congenital metabolic disorder lipoprotein lipase deficiency (LPLD) that reached clinical approval in Europe just six years ago. Alipogene tiparvovec (Glybera®) was first recommended for approval in 2012258. This was the result of a long process requiring four reviews by the Committee on Human Medicinal Products259. Though Glybera® has been removed from market in 2017260, a few key lessons have already been learned from the first AAV approved therapy. First, much like UniQure experienced when studying Glybera®, future phase III clinical trials treating rare monogenic diseases may struggle to find a sufficient population for study and the associated challenges when taking results from small sample sizes to review boards for final clinical approval. Furthermore, regardless whether treating a rare monogenic disease or a more common disease, the cost associated with multiple appeals for drug administrations is often prohibitive and was only successful for Glybera® thanks to private donors. Even after approval, health care systems are still faced with the extreme cost associated with such a therapy – after much speculation and estimates,261 the final cost landed on ~$1.4 million USD for the one time treatment with Glybera®262. For a disease with severe quality and quantity of life cost without effective treatment, an immediately expensive gene therapy product may in fact be cost- 34 effective and bearable to users with extended payment plans263. Notably, UniQure claims that it was not drug failure or adverse events driving their decision to remove Glybera® from market. Instead, they claimed that low patient demand and high costs maintaining Glybera® had been limiting their primary focus on developing gene therapy drugs for hemophilia B, Huntington’s disease, and heart failure260. Perhaps more prevalent diseases like heart failure or even diabetes will be more successful targets for gene therapy products in the future. 1.6.4 Using AAVs to study and treat diabetes Though we are unaware of any AAV based gene therapy clinical trials to treat diabetes, there have been many meaningful steps in this direction. In 2006, it was discovered that the eighth serotype of AAV (AAV8) has very high affinity for the mouse pancreas, including pancreatic islets264. Gene therapy may have merit to treat and investigate polygenic type 2 diabetes, and there have been studies utilizing an AAV to prevent onset of diabetes in models of autoimmune diabetes225. Additionally, patients with monogenic diabetes may in certain cases be ideal candidates for a gene therapy approach to replace the missing or damaged genes and thus cure the disease. For example, though many patients with insulin gene mutations may not be viable candidates for a gene therapy to replace the insulin gene because of a dominant negative function of the mutated gene265, these patients could be viable targets for a gene editing based gene therapy using CRISPR/Cas9 technology. Conversely, patients with mutations that reduce the bioactivity or biosynthesis of insulin266 or have recessive loss of function mutations of the insulin gene24 are ideal targets for an insulin gene therapy. Furthermore, having suitable tools to replace or repair genes in the β-cells of patients with diabetes could be the pathway to a cure for the 1-2% of patients with monogenic diabetes – millions worldwide. One major hurdle for gene therapies for diabetes will be ensuring that target cells; whether dysfunctional β-cells or surrogate cells, are capable of properly storing and processing proinsulin into mature, fully bioactive insulin.  35 1.7 Prohormone processing 1.7.1  Discovery of prohormones In 1967 Chretien and Li discovered that multiple hormones were contained within the precursor β-lipotropin267. From this observation came the prohormone theory that proposed that hormones could be derived through specific endoproteolytic cleavages of a precursor268. The same year, Don Steiner and colleagues independently supported this theory by showing with pulse-chase experiments that insulin is generated from a larger precursor he named ‘proinsulin’269. Since, dozens of prohormones have been discovered and production of hormones and other neuropeptides require the function of prohormone convertases for correct production. Specific processing is paramount for production of the correct hormones and proper processing is dependent on the function of the correct prohormone convertase. 1.7.2 The prohormone convertases To date, nine prohormone convertase (PC) genes have been discovered (PCSK1-PCSK9). With the exception of PCSK8 (Site 1 protease) that cleaves the N-terminal of sterol regulatory element-binding proteins270 and PCSK9 (prohormone convertase 9) that autocleaves at a glutamine and is involved exclusively in a cholesterol recycling pathway271, all other PCs cleave targets at single or dibasic amino acid sites. PCSK5 (PC5/6), PCSK6 (PACE4), and PCSK7 (PC7) are nearly ubiquitous to cell surfaces, PCSK3 (furin) is ubiquitous to the trans-Golgi network, and PCSK4 (PC4) is almost exclusive to the testis272. PCSK1 (PC1/3) and PCSK2 (PC2) are exclusive to neuroendocrine cells and are the PCs responsible for cleaving prohormones273. Despite a significant spectrum of function, the PCs have substantial structural similarities. They are produced with four domains: an N-terminal prodomain, catalytic domain, P-domain, and carboxyl-terminal domain. They all contain a catalytic triad composed of a His, Asp, and Ser, and with the exception of PC2 contain an Asn oxyanion hole to stabilize the transition state during cleavage274. The N-terminal prodomain is removed autocatalytically, the  36 catalytic domain conveys catalytic activity, the P-domain is important for conveying the strict calcium and pH requirements for activity (most notably for PC1/3 and PC2), and the C-terminal domain is responsible for proper intracellular trafficking of the PCs273. 1.7.2.1 Prohormone convertases 1/3 and 2 PC1/3 was first cloned in 1991 and as the third PC, was named PC3. As the same protein had been named the type 1 proinsulin processing enzyme in 1988275, it soon adopted the name PC1/3. PC1/3 is first produced as a 94kDa propeptide. Like all PCs, it contains an N-terminal prodomain, catalytic domain, P domain, and C terminal domain. Autocleavage in the ER first removes the pro-domain yielding an 87kDa intermediate before successive cleavage in the C terminal domain to yield a 74kDa and finally a 66kDa mature PC1/3 in the trans-Golgi network. As clearly demonstrated by its autocleavage, PC1/3 gains catalytic activity within the trans-Golgi network and can continue to be biologically active in the secretory granules. The activity of PC1/3 is quite unique in that it is ~80fold lower than PC2, suggesting that it can have more impactful regulation. PC1/3 has a pH optimum of 5.5-6.5 as the 87kDa intermediate and a lower pH optimum of 5-5.5 with a stringent 2 mM calcium requirement as the maximally bioactive 66kDa form273. The most notable regulator is its binding partner proSAAS (reviewed in 1.7.2.4). Little is known about the transcriptional regulation of PC1/3, but STAT3 (signal transducer and activator of transcription 3) binding sites are contained in the promoter and may be involved in hypothalamic PC1/3 regulation276. There is also evidence of translational regulation of PC1/3 including by glucose in rat islets, unlike PC2277. Finally, PC1/3 is known to oligomerize as it enters acidic compartments278 which reduces its activity. Two different PC1/3 knockout mice have been generated. The most studied PC1/3 knockout mouse was generated in the Steiner lab and has a deletion of a portion of the promoter and the first exon of PC1/3279. Two thirds of these PC1/3 knockout mice die in the first week of life and the survivors have stunted growth279. The other model was generated by deleting exons 2-10 and led to  37 preimplantation lethality, possibly related to this model being bred onto an inbred FVB/N background280, whereas the model generated in the Steiner lab was bred onto an outbred strain (CD1). There are dozens of targets processed by PC1/3 including proinsulin (reviewed in detail in 1.8.3), POMC, and other prohormones and neuropeptides. PC2 was discovered in rat islet cell tumors in 1988275 and was cloned by Smeekens et al in 1990281. PC2 is synthesized as a 75kDa proenzyme that autocleaves to yield a mature 64kDa enzyme. Though sharing most of the same structural features with other PCs, a unique feature of PC2 is the lack of the Asn oxyanion hole. Additionally, unlike PC1/3 it is not until the secretory granules that PC2 gains biological activity to autocleave and activate. This is perhaps partially mediated by a stringent pH requirement of 5.0282 and partially mediated by inhibition by its cofactor 7B2 (reviewed in 1.7.2.4). This delay in activation leads to a proposed sequential processing of prohormones by PC1/3 then PC2 in cells with both prohormone convertases. The PC2 knockout mouse on C57BL/6 background has been generated283, and has a somewhat mild phenotype. The mice have persistent hypoglycemia from a total loss of glucagon production and associated improved glucose tolerance. Interestingly, when crossed onto the 129Sv background PC2 knockout mice develop a lethal phenotype, dying between 5 and 9 weeks of age from hyperadrenocorticotropic hormone (ACTH) secretion284. 1.7.2.2 Predicting target specificity of the prohormone convertases Despite many in vitro studies attempting to clarify the target preferences of PC1/3 and PC2, there is still abundant uncertainty at predicting which targets are processed by which enzyme in vivo. Though furin has the most stringent consensus sequence of R-X-[R/K]-R↓ (X designates any amino acid and K and R are the single letter amino acid codes for lysine and arginine respectively. Single letter amino acid codes are used in the remainder of this thesis.), other enzymes have the vague consensus sequence R/K-2nX-R↓. Notably, there are exceptions to even these rules and predicting targets in vivo is extremely difficult. Arguably the best  38 unbiased data available to date are mass spectrometry studies on the relative processing of neuropeptides in PC1/3 or PC2 knockout mice compared to controls285, 286. By comparing the relative impact of each knockout, authors highlight that single arginine residue cut sites were more likely to be processed by PC1/3 and when bulky amino acids Y, W, F, or P were in the P1’ or P2’ sites (using the system of287 for denoting positions prior to (Px) or after (P’x) the scissile bond) cutting was more likely by PC2. Others have used the crystal structure of furin to approximate the crystal structure of PC1/3, PC2, and other PCs, and compare the amino acids lining the active site cleft288. Though the authors compared mixed species of PCs, we have independently compared the sequences of human, mouse, and rat PC1 and PC2, and found no substitutions at any of the active site cleft residues (data not shown). The authors concluded that enzymes with stringent requirements for basic amino acids N-terminal to the cut site, most notably furin, had more negatively charged amino acids in the cleft. When comparing PC1/3 and PC2 we note a similar number of negatively charged amino acids (10 vs 9 respectively) but more basic amino acids in PC1/3 compared to PC2 (4 vs 2). Perhaps these differences allow for PC2 to be more flexible in its substrate, including providing access to substrates with bulky amino acids near the cut site. We stress again the non-specificity of this proposal with the example of proghrelin – a target processed exclusively by PC1/3 in mouse, despite a bulky proline at the P2 position (KLQPR↓)289. Certainly, more work is required to understand the processing specificity of prohormone convertases. Studying only the sequence of the target is undoubtedly inadequate as highlighted by studies assessing prosomatostatin processing - though studies on mouse brain suggests that PC1/3 is necessary for processing prosomatostatin into somatostatin-14285, studies on mouse islets from PC1/3-/- mice suggests somatostatin-14 production from prosomatostatin does not require PC1/3279.  39 1.7.2.3 Carboxypeptidase E In 1982, researchers purified an enzyme in bovine adrenal chromaffin granules that removed basic amino acids from the C-terminus of Met- and Leu- enkephalin (a neuropeptide)290. The biological importance of this enzyme, now widely known as carboxypeptidase E (CPE; gene: Cpe), is highlighted in the study of a CPE mutant mouse (S202P). In 1995, it was found that these mice had elevated proinsulin including B-chain extended (31,32) diarginyl insulin291 and progressive obesity and hyperglycemia. This finding revealed that CPE was the enzyme responsible for removing the dibasic amino acids of the B chain and C-peptide exposed after PC1/3 and PC2 processing in mouse β-cells and was the first example of a defect in prohormone processing leading to diabetes. The role of CPE extends beyond removal of C-terminal basic residues to the sorting of hormones into the secretory pathway, including proinsulin292. Beyond insulin, it has been shown that CPE is responsible for the C-terminal removal of basic amino acids from dozens of prohormones292 and CPE knockout has been implicated in infertility via defective pro-gonadotropin releasing hormone processing293, low bone mineral density via defective cocaine and amphetamine regulated transcript (CART)294, and impaired processing of progastrin295. 1.7.2.4 The prohormone convertase cofactors proSAAS and 7B2 Both PC2 and PC1/3 have cofactors that regulate their function. proSAAS (gene: PCSK1N) is a 21kDa granin-like neuroendocrine protein that is secreted and is known to be an inhibitor of PC1/3296. proSAAS is co-expressed with PC1/3 in many mouse neuroendocrine tissues297 and proSAAS mutant mice have slightly reduced body weight and no defect in glucose tolerance298. We are unaware if proinsulin processing has ever been studied in proSAAS mutant mice. proSAAS contains an N-terminal domain and a C-terminal domain and is cleaved by furin at many sites, with many known products including PEN and big-LEN299. The C-terminal domain is a potent inhibitor of PC1/3 and the N-terminal domain decreases PC1/3  40 secretion rate suggesting a role as a chaperone296. There is also evidence that proSAAS is further cleaved, including by PC2300, but there is still little clarity on the function of each peptide. 7B2 (gene: SCG5) binds PC2 and is involved in directing PC2 to the secretory granules. 7B2 is produced as a 27kDa precursor and is processed into the 21kDa mature 7B2 and a C-terminal peptide by furin mediated cleavage301. There is evidence that the C-terminal peptide suppresses PC2 activity temporarily until it reaches the secretory granules where proPC2 autocatalyzes to become active301. 7B2 acts as a protein folding chaperone to prevent aggregation of PC2 and protect its ability to proceed into secretory granules and activate properly302. 7B2 deficient mice die from hypersecretion of ACTH and have abnormal secretion of proinsulin and hypoglucagonemia303. This phenotype is more severe than the PC2 knockout mouse on a C57Bl/6 background, though comparable to the PC2 knockout mouse on the 129Sv background284. There is some indication that the role of 7B2 may extend beyond functioning as a PC2 cofactor. Scg5 is expressed more broadly than Pcsk2304 and it has been found to prevent aggregation of proteins including IAPP305, β-amyloid, and α-synuclein306. 1.7.3 Proinsulin processing Like most hormones, insulin is also produced from a larger precursor protein269, 307. The structure of porcine insulin was identified in 1968 as containing an N-terminal B-chain and C-terminal A-chain with a connecting C-peptide308 and the following year there was evidence that rats had two insulins (INS1 and INS2) with near identical sequence309. Though it has since been discovered that the sequence of the C-peptide is relatively divergent among mammals, the mature insulin sequence is extremely well conserved (Figure 1-6). Since the discovery of proinsulin’s structure, many have worked to clarify how proinsulin is processed to remove the signal peptide and C-peptide to liberate mature insulin310 (Figure 1-7). Current theory posits that the B-chain - C-peptide junction is cleaved by prohormone convertase 1/3 (PC1/3; gene PCSK1) before cleavage at the C-peptide – A-chain junction by   41  Figure 1-6 Alignment of proinsulin sequences The amino acid sequences of proinsulin for human, mouse (I and II), rat (I & II), dog, and pig proinsulin. Alignment performed by multiple sequence comparison by log-expectation (MUSCLE). Changes in sequence designated as conservative are highlighted in purple, as semi-conservative are highlighted in blue, and as non-conservative are highlighted in red.  prohormone convertase 2 (PC2; gene PCSK2). This theory first gained traction when early work identified one or more enzymes in rat insulinoma tissue that cleaved proinsulin and were dependent on the high Ca and low pH environment normally found in secretory granules311. In follow-up, Hutton and colleagues showed that there are two endopeptidases in rat insulinoma cells and demonstrated that human proinsulin was only fully processed when it was incubated with both rat endopeptidases in vitro275. In 1990 and 1991, the similarities in function between the unknown endopeptidases and the yeast gene KEX2 were used to identify PC1/3 and PC2281, 312. By 1998, the PC2 knockout mouse was characterized and shown to have impaired processing at the C-A junction, resulting in a buildup of des-31,32 proinsulin313. In 2002, the PC1/3 knockout mouse was characterized and shown to have severely impaired processing at the B-C junction, resulting in a buildup of des-64,65 proinsulin279. Based on relative in vitro processing rates of intact human proinsulin versus des-31,32 proinsulin or des-64,65 proinsulin by rat PC1/3 and PC2 isolated from insulinoma cell granules, it has been proposed that processing at the B-C junction by PC1/3 occurs before processing by PC2 at the C-A junction314, but some data has countered this hypothesis by showing more buildup of des-64,65 proinsulin during the processing of rat INS2 in islets315. Taken together, these studies provide good support for the theory that primary mouse β-cells process insulin sequentially by PC1/3 then PC2, but there have been limited studies in other species, most notably humans.  42  PC1/3 PC2 CPE CPE PC2 PC1/3 CPE C-Peptide Proinsulin Des-31,32 Proinsulin Des-64,65 Proinsulin Diarginyl insulin Insulin Split-32,33 Proinsulin Split-65,66 Proinsulin Figure 1-7 Current theory of proinsulin processing Current understanding of how proinsulin is processed to yield mature insulin (B-chain shown in blue, C-peptide shown in gray, and A-chain shown in orange with connecting dibasic amino acids in red). The illustrated sequence is human. The pathway on the right is thought to be dominant, though this has been challenged for rat Ins 2. First, PC1/3 processes the B-C junction, second CPE removes the dibasic KR amino acids from the B-chain, and third PC2 processes the C-A junction to yield insulin and C-peptide. The pathway on the left occurs via PC2 processing at the C-A junction first before processing by PC1/3 and the B-C junction and a final removal of dibasic RR amino acids by CPE.  43 1.7.4 Proinsulin processing in human β-cells Based on publicly available sequencing datasets, there is some indication that PC2 may not be as important in human β-cells as it is in mouse β-cells. Though a microarray experiment on a human β-cell line (EndoCβH2) found abundant PCSK1 and PCSK2316, RNAseq experiments on sorted primary α-cells and β-cells showed higher expression of PCSK2 than PCSK1 in mouse β-cells317 whereas human β-cells expressed 20x more PCSK1 than PCSK2318 (Figure 1-8). Single cell RNAseq experiments on human pancreas find enrichment for PCSK2 in α-cells and PCSK1 in β-cells319-321, with more abundant expression of PCSK1 than PCSK2 in human β-cells and less PCSK2 in β-cells than α-cells. Additionally, Davalli et al. reported selective insufficiency of PC2 immunoreactivity in the β-cells of human islets transplanted into nude mice322 and others have reported that some human insulinomas are not immunoreactive for PC2323. Given the broad divergence of islet architecture among mammals324, it seems highly possible that other aspects of islet biology, including prohormone processing mechanisms, are not fully conserved between bona fide human and mouse β-cells. Given the obvious limitations in attempting to study human β-cell function in vivo, the best available information on human proinsulin processing comes from patients with gene mutations and in vitro experiments. There have been no reported cases of non-functional PCSK2 mutations and there is an association between PCSK2 polymorphisms and diabetes325 but there are no reported associations between PCSK2 polymorphisms and circulating levels of proinsulin326. Contrarily, there have been at least 21 reported PCSK1 mutant humans and the vast majority have elevated circulating intact proinsulin and/or des-64,65 proinsulin327. Most patients present with severe obesity and hyperphagia328, 329, contrasting PC1/3-/- mice that have stunted growth330. Importantly, patients with PCSK1 mutations also presented with high levels of circulating insulin immunoreactivity of over 1000 pM329. However, as pointed out by authors329, cross-reactivities of the assays for the > 8000pM of circulating proinsulin were likely the cause  44 of these supposed high insulin levels. In fact, highly specific HPLC experiments revealed that there was no detectable mature insulin in the circulation of the patient with the PCSK1 mutations329, suggesting that PC1 is essential for processing proinsulin in humans. Though dissecting the specific impact of impaired processing of proinsulin versus impaired processing of growth hormone releasing hormone on the obesity and impaired linear growth phenotype of patients with PCKS1 mutations is challenging327, it is likely abundant proinsulin with low relative bioactivities (compared to fully processed insulin) is adequate to produce hypertrophic signals to drive growth.  Figure 1-8 Human β-cells have less PCSK2 than PCSK1 in published RNAseq databases on sorted human and mouse islet cells. Reanalysis of publicly available RNAseq databases (Bramswig et al., 2013; Benner et al., 2014). Relative FPKM for PCSK2 and PCSK1 compared in human β-cells and α-cells as well as mouse β-cells and α-cells. There is likely an α-cell impurity contributing to PCSK2 in the sorted β-cells populations given high GCG FPKM in the β-cell samples.   There is evidence that proinsulin processing is impaired during diabetes progression including prior to and after onset of type-1 diabetes. Insulin autoantibody positive patients have elevated circulating proinsulin331 and the proinsulin/C-peptide ratio is positively associated with progression to type 1 diabetes332, 333. The proinsulin/insulin ratio is increased in newly diagnosed  45 type 1 diabetes334, 335 and patients with a lower proinsulin/insulin ratio are more likely to have a honeymoon phase (a period of insulin independence) post diagnosis336. Even among patients that lose detectable circulating C-peptide most retain detectable proinsulin337, 338 and a similar defect of processing may occur for proIAPP in type 1 diabetes339. Notably, reductions in PC1/3 expression have been observed in pancreas from donors with type 1 diabetes, suggesting that an insufficiency of PC1/3 could be a driver for impaired processing during diabetes pathogenesis340. There is also evidence that proinsulin processing is impaired during the pathogenesis of type 2 diabetes as circulating proinsulin is associated with progression to type 2 diabetes341 and patients with type 2 diabetes have an increased proinsulin/insulin ratio342. Additionally, in patients with impaired glucose tolerance, the proinsulin/insulin ratio is elevated343, 344. Though the directions of causation between impaired prohormone processing and diabetes progression cannot be confirmed, it is worth noting that proinsulin and the proinsulin processing intermediates have between 1/5th and 1/100th the bioactivity than fully processed insulin345, 346. Understanding the nature of this β-cell defect could be useful for potential therapeutics, for a better understanding of the β-cell phenotype during diabetes progression, and for the development of diabetes biomarkers. Given the importance of proinsulin processing to β-cell function and diabetes progression, there is a major gap in the literature on studies investigating the processing of human proinsulin by human prohormone convertases in primary human β-cells. Past work has demonstrated that mouse PC1/3 can process human proinsulin alone in a rat pituitary cell line (GH3), albeit at a low efficiency347 and others have shown that mouse PC1/3 alone can fully process human proinsulin in a mouse pituitary cell line348. There have also been studies on the processing of human proinsulin in rat pituitary cells with mixed species of prohormone convertases347 and investigations of human proinsulin processing in transgenic mice349. In human islets, one study investigated the kinetics of proinsulin conversion and found that human  46 β-cells buildup more des-31,32 proinsulin that des-64,65 proinsulin350, but we are unaware of any studies specifically assessing the roles of PC1/3 and PC2 in primary human β-cells.  1.8 Thesis investigation  The past century of diabetes management has centered around insulin injections with growing use of other peptide injections and oral hypoglycemic agents. Though these therapies have been lifesaving, millions continue to deal with the unyielding quality of life reduction and early mortality not fully addressed by these therapies. Work in the field of islet transplantation has provided proof-of-principle data that β-cell replacement therapies could be an effective cure. There has been a tremendous amount of research into viable alternative therapies, including generating insulin-producing cells in vitro from stem cells by mimicking normal development and reprograming or repairing cells in situ to become functional β-like cells. Understanding the factors that contribute to the normal development of β-cells will be paramount to furthering the ability to generate β-cells in vitro. Understanding the factors that contribute to β-cell dysfunction and dedifferentiation will be crucial to therapies aimed at endogenous generation of β-cells in situ and studying the role of diabetes specific genes causative for MODY or ND provides insight into more complex polygenic diabetes. Understanding the basic cell biology of human β-cells will be essential for understanding the pathology of diabetes and viability and success of any β-cell (re)generating therapy. This thesis will provide insight into each of these pursuits in four main data chapters – characterization of insulin-deficient β-cells (chapter 2), development of a genetic tool for β-cell in vivo manipulation (chapter 3), assessing a gene therapy approach to treat an insulin knockout mouse model of PND (chapter 4), and assessing the role of PC2 in proinsulin processing in human β-cells (chapter 5). First, we aimed to determine if insulin is necessary for β-cell development. The insulin knockout mouse is a good model of PND and capable of defining the role of insulin during β-cell  47 development. We assessed the β-cell phenotype at birth and explored the ability of insulin replacement to reverse developmental defects. Second, we aimed to determine if an adeno-associated virus (AAV) can deliver Cre recombinase to pancreatic β-cells for efficient recombination. We designed an AAV with a fragment of the rat insulin 1 promoter driving expression of Cre recombinase for use as a genetic tool for in vivo manipulation (AAV Ins1-Cre). In our third aim, we determine if insulin-deficient β-cells can take on an insulin-producing role when insulin is replaced by an AAV mediated gene therapy. We delivered the insulin gene to insulin-deficient β-cells by AAV to determine if dedifferentiated insulin-deficient β-cells can complete maturation and begin glucose stimulated insulin secretion like mature β-cells. Examining the potential of undifferentiated insulin-deficient β-cells to mature and take on an insulin-producing role when insulin is replaced provides insight into the reversibility of diabetes by facilitating β-cell redifferentiation and could also be a viable therapeutic strategy for patients with MODY or ND. Finally, we sought to investigate proinsulin processing with the specific aim to investigate if PC2 is present in human β-cells and is necessary for processing human proinsulin. Though past work has provided convincing evidence that mouse proinsulin is processed by both PC1/3 and PC2, published findings justified examining if human proinsulin is processed differently than mouse proinsulin. We localized PC2 in human pancreas and using human islets we impaired the function of PC1/3 or PC2 to determine the capabilities of the enzymes to process proinsulin in human β-cells. Collectively, the studies in this thesis assess the role of insulin on β-cell development, develop a useful genetic tool to study β-cells in vivo, provide insight into the viability of a gene therapy approach to reverse monogenic insulin deficient diabetes, and clarify a longstanding theory on proinsulin processing in human β-cells. These findings emphasize the importance of the continued rigorous study of β-cell development and function to better guide gene and cell-based therapies for the treatment of diabetes.    48 Chapter 2: Insulin-deficient mouse β-cells do not fully mature but can be remedied through insulin replacement by islet transplantation 2.1 Introduction  In humans, β-cells containing insulin first appear in the pancreas at week 7 in development and insulin acts as the main anabolic signaling molecule starting in the 26th week of gestation351. In mice, insulin first appears in the pancreas at embryonic day 9.5 (E9.5)352 and begins to regulate cell growth and glycemia at birth; a similar developmental stage to human embryos at 26 weeks gestation. Uptake of glucose into cells is a fundamental process that is almost always dependent on the actions of insulin353. Without insulin, vertebrates face elevated glucose levels, increased stress hormones, ketogenesis, and rapid death from severe ketoacidosis354.  Insulin may also play other non-glucose lowering roles, including serving as an important signaling molecule on β-cells. Mice with targeted disruption of insulin receptors (IRs) on β-cells have impaired glucose-stimulated insulin secretion and glucose intolerance355. IRs on β-cells are necessary for β-cell hyperplasia in response to insulin resistance356. These studies355, 356 utilized a transgenic mouse model with the rat insulin promoter regulating Cre (RIP-Cre), with a floxed insulin receptor gene (Insr). Cre is expressed after E9.5 when insulin first appears in the mouse pancreas352 and recombines to delete the IR from ~80%357 of β-cells and regions of the brain358. Delayed and incomplete recombination allows some insulin signaling in β-cells beyond E9.5 and impaired insulin action in the brain can promote obesity and insulin resistance359. Therefore, these findings do not conclusively demonstrate that a loss of insulin signaling in β-cells is responsible for the observed phenotype. Recently, Trinder et al. addressed the caveat of central recombination by using mice with an inducible Cre transgene driven by a mouse insulin 1 promoter (Ins1-Cre/ERT) to generate IR deficiency selectively in β-cells after E13360. Ins1- 49 Cre/ERT mice had reduced β-cell mass at birth, but like previous studies using the RIP-Cre line, these mice retained alternate insulin signaling pathways and attenuated IGF signaling in β-cells through the IR. These caveats make it challenging to conclude that a loss of insulin action in β-cells was entirely responsible for the phenotype. Alongside incomplete β-cell recombination, delayed loss of IRs until later in development, and the presence of a human growth hormone minigene in the Ins1-Cre/ERT transgene361, 362, there are other important factors to consider when attempting to study a loss of insulin signaling by using a model with deletion of the IR. First, insulin may signal with low affinity through the IGF receptors 1 and 2363, 364 and it is possible that there are other insulin signaling pathways not yet appreciated. Hyperinsulinemia in β-cell insulin receptor deficient mice may increase the likelihood of insulin signaling via the IGF receptors355. Second, IGF1 and IGF2 bind the IR and play essential roles in embryological mammalian development365. In fact, binding of IGF2 to the IR appears to be the most important signal through the IR during mouse development366. Thus, deleting the insulin receptor may not completely block all actions of insulin and impairs the IR-mediated actions of the IGF hormones. We posit that the best available model capable of definitively defining the unique role of insulin in β-cell development and maturation, is a model with a loss of the insulin gene(s). In 1997 Duvillié and colleagues generated an absolute insulin-deficient mouse lacking both copies of the two nonallelic insulin genes, Ins1 and Ins2 (Ins1-/-Ins2-/-) with β-galactosidase (βGAL; gene: lacZ) knocked in to the Ins2 locus367. Ins1-/-Ins2-/- mice are viable at birth, have a slightly (20%) reduced body weight, and though euglycemic at the time of parturition, quickly develop severe hyperglycemia after suckling and die on average within 48 hours367, 368. This aligns with findings in humans with insulin mutations who first present with growth retardation at 27 weeks gestation, a similar developmental stage to mice at birth369. Pancreatic islets of neonatal Ins1-/-Ins2-/- mice are enlarged but have reduced levels of endocrine hormone mRNA  50 (glucagon [Gcg], somatostatin [Sst], and pancreatic polypeptide [Ppy])370. The pancreas of Ins1-/-Ins2-/- mice is full of aggregates of βGAL expressing cells but apart from positive immunoreactivity for PDX1 in βGAL expressing cells370, there are no reports of the functional maturity of these cells to determine the role of insulin signaling during β-cell development. In the present study, we determined the role of insulin during β-cell development and maturation by characterizing pancreatic islets of neonatal Ins1-/-Ins2-/- mice and mice maintained into adulthood by insulin replacement therapy. We found that alongside a complete loss of insulin production, β-cells of Ins1-/-Ins2-/- mice lacked many markers of mature β-cells and were present in islets of abnormal size and endocrine hormone composition. Additionally, insulin-deficient β-cells expressed markers of dedifferentiation and dysfunction suggesting that insulin is a necessary signal for complete β-cell maturation and/or maintaining their maturity. Short-term replacement of insulin in the insulin knockout mice via exogenous injection facilitated a partial normalization of the β-cell phenotype but was associated with islet fibrosis. Long-term replacement of insulin by isogenic islet transplantation was sufficient for the maturation of insulin-deficient β-cells. This model provides direct evidence that insulin itself is required for normal β-cell development and the maintenance of normal β-cell function.  2.2 Materials and Methods Animal breeding and insulin therapy All experiments were approved by the UBC Animal Care Committee and carried out in accordance with the Canadian Council on Animal Care Guidelines. Ins1+/+Ins2+/+ mice on a C57BL/6J background (Jackson Laboratories, Bar Harbour, ME, USA) were used as controls. Ins1-/-Ins2+/- were generated in the lab of Dr. J. Jami (Institut Cochin, Paris, France) and supplied indirectly from the lab of Dr. J. Johnson (University of British Columbia, Vancouver, Canada). Ins1-/-Ins2+/- were on a mixed background (~50:50 C57BL/6 and 129; personal communication  51 from Dr. J. Johnson) then were backcrossed onto a C57BL/6J background (five rounds) before generating animals used in the current study. Animals were housed with a 12-h light/dark cycle with ad libitum access to a standard chow diet (2918, Harlan Laboratories, Madison, WI, USA). In the same rodent facility, Ins1-/-Ins2+/- mice were bred to acquire Ins1-/-Ins2-/- mice and Ins1+/+Ins2+/+ mice were bred to acquire Ins1+/+Ins2+/+ mice. Upon first discovery of neonatal mice (P0 – P1) subsets of mice were euthanized by decapitation and blood sugar was examined using a handheld glucometer (LifeScan; Burnaby, BC, Canada). Insulin therapy was immediately initiated in other cohorts of Ins1-/-Ins2-/- mice, initially identified by reduced body weight and confirmed by genotyping at an older age. We genotyped Ins2 by qPCR assay (wild-type forward primer: 5’-GGT CCT TGG TGG TAG TAA CTT G, reverse primer: 5’-GCC TCT AAA GCC TAC TCA TCT TC, and probe: 5’-GCA GTG CTC TAT GAG GGC CCT AAA; lacZ knock in forward primer: 5’-CTG TAT GAA CGG TCT GGT CTT T, reverse primer: 5’-CGC TAT GAC GGA ACA GGT ATT, and probe: 5’-TTG CCC GGA TAA ACG GAA CTG GAA) and Ins1 by qPCR assay (wild-type forward primer: 5’-CCA TTG TTA GGT TGG ATG ATT, reverse primer: 5’-CGG TTG CCT ACC TTC TT, and probe: 5’-AGT ATC TGG AAT TCT GCT TCC TGC CC; knockout forward primer: 5’-AAA CCA CAC TGC TCG AC, reverse primer: 5’-CAG GAA GCA GAA TTC CAG ATA, and probe: 5’-GGG CTG CAG GAA TTC GAT ATC AAG C). Briefly, animals received twice daily subcutaneous injections of approximately 0.1U Insulin Glargine (Lantus®; diluted to 5 U/mL in F-10 media, Sigma-Aldrich, St. Louis, MO). As regular blood glucose sampling is not possible with young pups, health of the animals was tracked by body weight and glycosuria, and when animals failed to gain weight or lost weight, insulin dosing was halved or skipped. When animals presented with severe glycosuria, we increased the insulin dose. As pups grew, we began regular blood sampling to better refine insulin doses. Therapy continued for up to 291 days. Mortality by this method was initially extremely high (>90%) but decreased with experience to less than 50% for animals analyzed in this study.  52 Some animals received islet transplants into the anterior chamber of the eye at 14 days of age. In all experiments, we present findings from a mix of male and female animals. We did not assess sex in neonates and did not detect significant differences between the sexes in adult mice (data not shown). Islet isolation and transplantation Newborn Ins1-/-Ins2-/- or Ins1+/+Ins2+/+ mice were euthanized by decapitation upon first discovery (P0-P1) and pancreatic islets were isolated by collagenase digestion371. Following euthanization, we performed a rapid dissection and exposed the pancreatic duct. Approximately 1 mL of collagenase solution (1000 U/mL type XI collagenase, Sigma-Aldrich, St. Louis, MO; diluted in Hanks Buffer without CaCl2) was injected into the pancreatic duct. The pancreas was then excised and digested for 6-8 mins at 37oC in 3 mL of collagenase. Islets were then handpicked three times in medium (Hams F10, 7.5% fetal bovine serum, and penicillin/streptomycin, Sigma-Aldrich, St. Louis, MO) to increase purity to over 90%. Islets were washed in PBS and immediately lysed by forceful suction into a 20g needle 20-30 times in 50 µL of QIAzol Lysis Reagent (Qiagen, Toronto, ON, Canada) before storage at -80oC. Islet isolation from donor adult C57Bl/6 mice (used for transplantation) was completed in the same fashion with slight changes (incubation in collagenase for 12-15 mins at 37oC). After handpicking, islets were cultured overnight, visually inspected for purity, and washed in sterile PBS prior to transplantation. Following two weeks of insulin therapy with Insulin Glargine (Lantus®), Ins1-/-Ins2-/- animals received an islet transplant into the anterior chamber of the eye (approximately 100-150 islets) as previously described372. This procedure is technically feasible at two weeks of age and provides reversal of diabetes within 24 hours. Animals were anaesthetized by inhaled isoflurane (5%) and maintained at 1-3%. The cornea was punctured with a 27G needle to gain access to the anterior chamber of the eye. Islets were loaded into a micropipette (MXL3-BP-IND-200; Origio MidAtlantic Devices, Mt Laurel, NJ, USA) and the  53 micropipette was passed through the opening in the cornea and islets were deposited using a micromanipulator. After removal of the micropipette, animals received Isoptears (Alcon Canada, Mississauga, ON, Canada) with 0.3% wt/vol gentamycin to prevent infection. Immunohistochemistry and immunohistofluorescence Pancreata were dissected out of mice, washed in phosphate buffered saline (PBS), and fixed in 4% paraformaldehyde (PFA) overnight before being transferred to 70% ethanol for long term storage prior to paraffin-embedding and sectioning (5 µm thickness; Wax-It Histology Services, Vancouver, Canada). Hematoxylin and Eosin and Trichome staining was performed by standard protocol and slides were scanned using the ScanScope CS system (Aperio; Vista, CA). Immunofluorescent staining was performed as previously described16 and we list antibody source, catalog number, working dilution, and research resource identifier (RRID) in Table 2-1. Secondary antibodies used in Chapters 2 through 5 are described in Appendix A. Briefly, sections were deparaffinized in xylene (15 mins) and rehydrated in graded ethanol (100% 2x5 mins, 95% 5 mins, 70% 5 mins, and PBS 10 mins) before heat induced epitope retrieval in an EZ Retriever microwave oven (BioGenes; Fremont, CA) for 15 mins at 95oC in 10 mM citrate buffer (0.5% Tween 20, pH 6.0; ThermoFisher Scientific®, Waltham, MA). Samples were blocked in DAKO® Protein Block, Serum Free (Dako; Burlington, Canada) and incubated overnight in primary antibody diluted in Dako Antibody Diluent (Dako; Burlington Canada). The following day, slides were washed and incubated in secondary antibody for 1 hour at room temperature before mounting and counterstaining with nuclear stain 4,6-diamidino-2-phenylindole (DAPI) using VECTASHILED® Hard Set Mounting Medium with DAPI (Vector Laboratories; Burlingame, Ca). All images were captured and analyzed with an ImageXpress® Micro XLS System, controlled by MetaXpress® High-Content Image Acquisition & Analysis Software (Molecular Devices Corporation; Sunnyvale, CA) with a scientific CMOS camera, a Nikon 20× Plan Apo objective (NA=0.75, 1-6300-0196; Nikon, Tokyo, Japan) and DAPI (DAPI- 54 5060B), FITC (FITC-3540B), Cy3 (Cy3-4040B), Texas Red (TXRED-4040B), and Cy5 (Cy5-4040A) filter cubes. We performed quantification of histological images using the same software (MetaXpress®). We used IAPP, insulin, or β-galactosidase to label β-cells, glucagon (GCG) to label α-cells, somatostatin (SST) to label δ-cells, and pancreatic polypeptide (PP) to label PP cells. Of note, though δ-cells are known to express some IAPP373, given the relative smaller number of δ-cells, lower expression level, and very limited co-localization of SST and IAPP (data not shown), we use bright IAPP immunoreactivity as a marker of β-cells. We immunostained for synaptophysin (SYN) to identify islet cells and quantify islet size and number. We calculated endocrine cell area of each cell type by quantifying immunoreactive surface area relative to total pancreas surface area and calculated the average of three immunostained sections 50-100 µM apart from each animal. We calculated the proportion of β-cells with positive immunoreactivity for proteins of interest or total immunoreactive area of proteins of interest by thresholding images and using a multi wavelength cell scoring journal in MetaXpress® and used image analysis to find the mean fluorescent intensity of positive signal in relevant cell compartments. Briefly, DAPI was used to localize the nuclear compartment of each cell and defined minimum and maximum cell diameters and total stained area were used to identify cytoplasmic compartments. Settings were based on an assessment of representative cells for each immunostaining experiment with an effort to maximize sensitivity and specificity of signal detection. All quantification analyses for direct comparison were based on immunostaining experiments performed in parallel with identical image acquisition conditions. Imaging conditions aimed to make use of a 16-bit dynamic range of fluorescent intensity while avoiding pixel intensity saturation. After defining parameters, quantification was performed in MetaXpress® and the generated masks for cell compartments and positive/negative signal coding in outputs were reviewed. Settings were adjusted as needed to ensure quantification of colocalization or average pixel fluorescent intensity between sections was representative.  55 Table 2-1 Primary antibodies used for immunofluorescent staining Peptide/protein target Antigen Sequence Name of Antibody Manufacturer, catalog #, and/or name of producer Species raised in; monoclonal or polyclonal Dilution used RRID Aldehyde dehydrogenase family member 1A3 unknown Rabbit anti-ALDH1A3 Novus Biologicals NBP2-15339 Rabbit Polyclonal 1 to 100 AB_2665496 Islet amyloid polypeptide unknown Rabbit anti-IAPP AbCam ab15125 Rabbit Polyclonal 1 to 50 AB_2295631 Beta-galactosidase Whole beta-galactosidase Rabbit anti-βGAL Thermo Scientific A11132 Rabbit Polyclonal 1 to 100 AB_221539 Beta-galactosidase unknown Mouse anti-βGAL DSHB 40-1a-c Mouse Monoclonal 1 to 50 AB_528100 Glucokinase Recombinant glucokinase Rabbit anti-GCK Sigma HPA007034 Rabbit Polyclonal 1 to 50 AB_888431 Glucagon unknown Mouse anti-GCG Sigma G 2654 Mouse Monoclonal 1 to 1000 AB_259852 Glucose transporter 2 First extracellular loop of Glut2 Rabbit anti-GLUT2 Millipore 07-1402 Rabbit Polyclonal 1 to 500 AB_1587076 Insulin Full length human insulin Guinea pig anti-INS Thermo Fisher Scientific PA1-26938 Guinea pig Polyclonal 1 to 100 AB_794668 Insulin unknown Rabbit anti-INS Cell Signaling C27C9 Rabbit Monoclonal 1 to 200 AB_2126503 Insulin Residues near Val36 of human insulin Mouse anti-INS Cell Signaling L6B10 Mouse Monoclonal 1 to 250 AB_10949314 v-myc avian myolocytomatosis viral oncogene lung carcinoma derived AA 105-154 of human l-myc Rabbit anti-L-MYC AbCam Ab28739 Rabbit Polyclonal 1 to 100 AB_2148730 V-maf muscoloapo-neurotic fibrosarcoma oncogene homolog A unknown Rabbit anti-MAFA Betalogics (Johnson & Johnson) LP9872 Rabbit Polyclonal 1 to 1000 AB_2665528 V-maf muscoloaponeurotic fibrosarcoma oncogene homolog B AA 18-140 of MAFB Rabbit anti-MAFB Sigma Life Sciences HPA005653 Rabbit Polyclonal 1 to 100 AB_1079293 homeobox protein NANOG Mouse nanog Rabbit anti-NANOG AbCam Ab80892 Rabbit Polyclonal 1 to 100 AB_2150114 homeodomain transcription factor 6.1 Human Nkx6.1 Goat anti-NKX6.1 R and D Systems AF5857 Goat Polyclonal 1 to 20 AB_1857045 homeodomain transcription factor 2.2 Nkx2.2-GST fusion, E. coli Mouse anti-NKX2.2 DSHB 74.5A5 Mouse Monoclonal 1 to 100 AB_531794 Neurogenin-3 Met1-Leu214 of human Ngn3 Sheep anti-NGN3 R and D Systems AF3444 Sheep Polyclonal 1 to 20 AB_2149527 Neurogenin-3 AA40-69 of human Ngn3 Sheep anti-NGN3 Thermo Fisher Scientific PA5-11893 Rabbit Polyclonal 1 to 100 AB_2149526 Pancreatic polypeptide Ala30-Leu95 of human PP Goat anti-PP R and D Systems AF6297 Goat Polyclonal 1 to 200 AB_10717571 Paired box 6 C-terminus mouse PAX6 Rabbit anti-PAX6 Covance PRB-278P Rabbit Polyclonal 1 to 250 AB_2313780  56 Peptide/protein target Antigen Sequence Name of Antibody Manufacturer, catalog #, and/or name of producer Species raised in; monoclonal or polyclonal Dilution used RRID Prohormone convertase 1/3 unknown Rabbit anti-PC1/3 Lakshmi Devi gift Rabbit Polyclonal 1 to 500 AB_2665530 Prohormone convertase 1/3 unknown Mouse anti-PC1/3 Gunilla Westermark gift Mouse Monoclonal Direct AB_2665529 Prohormone convertase 2 E622-N638 of mouse PC2 Rabbit anti-PC2 Thermo Fisher Scientific PA1-058 Rabbit Polyclonal 1 to 500 AB_2158593 Proliferating cell nuclear antigen Rat PCNA Mouse anti-PCNA Abcam ab29 Mouse Monoclonal 1 to 100 AB_303394 Pancreatic and duodenal homeobox 1 N-terminus of mouse PDX1 Guinea pig anti-PDX1 Abcam ab47308 Guinea pig Polyclonal 1 to 1000 AB_777178 Somatostatin Human somatostatin Mouse anti-SST Β Cell Biology Consortium AB1985 Mouse Polyclonal 1 to 500 AB_10014609 Sulphonylurea receptor 1 Aa1560-1582 Rabbit anti-SUR1 Abcam ab32844 Rabbit Polyclonal 1 to 50 AB_2273320 Synaptophysin C-terminus of human Syn Rabbit anti-Synaptophysin Novus Biologicals NB120-16659 Rabbit Monoclonal 1 to 50 AB_792140  RNA Isolation and RT-qPCR Following islet isolation, lysis, and storage in Qiazol Lysis Reagent (Qiagen, Toronto, ON, Canada) we performed an acid guanidinium thiocyanate-phenol-chloroform extraction by standard methods374. To ensure sufficient quantity of mRNA for downstream applications, we pooled lysed islets from 3 to 5 mice in each group of Ins1-/-Ins2-/-, Ins1-/-Ins2+/-, or Ins1+/+Ins2+/+ animals, for a total of 25-35 islets per sample. cDNA was generated from isolated islet RNA using a RevertAid H Minus First Strand cDNA synthesis kit containing random hexamers and oligo(dT)18s (ThermoFisher Scientific®, Waltham, MA) and was performed according to manufacturer’s protocol. mRNA levels of target genes were quantified by 2-ΔΔCT method (glyceraldehyde 3-phosphate dehydrogenase [Gapdh] used as the reference gene) using the SsoFast EvaGreen Supermix (Bio-Rad®, Mississauga, Canada) kit according to manufacturer’s protocol (for primer details see Table 2-2).   57 Table 2-2 Primer sequences and annealing temperatures. Target Gene(s) Forward primer Reverse Primer Annealing Temperature Efficiency1 Abcc8 GTGGTATCGAGCAGAGATCC ATCAGGATACAAAAGGCGCT 60oC 98.4% Amy1 TAACCACATGTGTGGCGCAG CCAGAAGGCCAGTCAGACGA 59oC 96.9% Gapdh TGGAGAAACCTGCCAAGTA TGTTGCTGTAGCCGTATTCA 60oC 98.5% Gcg CAGGCACGCTGATGGCT GTGAAGATGGTTGTGAATGGTGAAATAC 59oC 105.1% Gck TTTGTGAGCCGATCCTGCTT AGGTGATTTCGCAGTTGGGT 59oC 100.0% Ins1 + Ins2 GCTCTCTACCTGGTGTGTGGGGAGC GGAGCAGATGCTGGTGCAGCACTG 59oC 102.4% Neurog3 TTGAGTCGGGAGAACTAGGA TGGAATTGGAACTGAGCACT 60oC 105.9% Nkx2.2 TTATGTTTGAGTTGCAGCGG AAACCACGTGAAACTGCTGA 59oC 99.1% Nkx6.1 CTTCGCCCTGGAGAAGAC CCGAGTCCTGCTTCTTCTTG 59oC 98.5% MafA AGGCGCACCCGACTTCTTTCTGTGA GGTAGGTACCTGGAGCTGGGGCTT 59oC 95.2% Pax6 CACATTTCCTTGAGAGGGCT AAGACACCACCAAGCTGATT 59oC 101.6% Pcsk1 CCAGTTCTCAGTCCAGCCCTT CCAAACGCAAAAGAAGGCGAA 59oC 97.5% Pcsk2 AGCCATGGAACTAGGTGTGC TTGTATGCTACGCCGACTCC 59oC 104.9% Pdx1 GCTCACGCGTGGAAAGGCCAGT CGCATACTGCTGCCTCTCC 60oC 98.4% Ppy CGCATACTGCTGCCTCTCC CCTGGTCAGTGTGTTGATGTATCTG 59oC 104.9% Slc2a2 ATCAGGACTGTATTGTGGGC GAGGCCAGCAATCTGACTAA 59oC 97.3% Sst CTGAGCAGGACGAGATGAGG TAACAGGATGTGAATGTCTTCCAGAA 59oC 96.2%  1 Efficiency was calculated by CFX Manager software (Version 3.1, Bio-Rad Laboratories, Hercules, Ca) by fitting a standard curve of serial dilutions of islet cDNA.  Statistics Data were subject to the Shapiro-Wilk normality test and when all groups passed the test for normality we analyzed by the Student’s t-test, one-way ANOVA with Tukey test for multiple comparisons, or two-way ANOVA with Bonferonni’s post-hoc testing. When one or more groups failed the test for normality, we used the non-parametric Mann-Whitney U test or Kruskal-Wallis test with Dunn’s multiple comparisons test. Statistical analysis was performed using GraphPad Prism 7.01 (La Jolla, CA) with significance set at p < 0.05.  2.3 Results Ins1-/-Ins2-/- mice are born with β-cells Upon first discovery (P0 – P1), Ins-/-Ins2-/- mice had reduced body weight compared to Ins1+/+Ins2+/+ controls (Figure 2-1A). Ins-/-Ins2-/- mice were euglycemic at the time of delivery but  58 quickly developed hyperglycemia after suckling. Therefore, blood glucose levels were highly variable at the time of blood sampling, unlike Ins1+/+Ins2+/+ controls and Ins1-/-Ins2+/- littermates. To determine if reduced insulin expression had a detrimental impact on the phenotype of β-cells, we immunostained pancreas sections from mice with reduced insulin gene copy number (Ins1-/-Ins2+/+, Ins1+/+Ins2-/-, or Ins1-/-Ins2+/-) for the prototypical mature β-cell marker muscoloaponeurotic fibrosarcoma oncogene homolog A (MAFA, a basic leucine zipper transcription factor; Figure 2-2). There was reduced intensity of MAFA immunoreactivity in mice with reduced insulin gene copy number. Based on previous reports of abnormal β-cell size and islet structure in Ins1-/-Ins2+/+ and Ins1+/+Ins2-/- mice375 paired with reduced nuclear immunoreactivity for MAFA in mice with reduced insulin gene copy number, we compared the Ins1-/-Ins2-/- group with wild-type Ins1+/+Ins2+/+ controls in the majority of our experiments. Histological examination of neonatal pancreata revealed normal aggregates of IAPP+βGAL+INS- clusters resembling neonatal islets of Langerhans in Ins1-/-Ins2-/- mice and IAPP+βGAL-INS+ β-cells in Ins1+/+Ins2+/+ mice (Figure 2-1B). We use IAPP, insulin, and/or β-galactosidase as markers for β-cells.    59  Figure 2-1 Ins1-/-Ins2-/- mice have IAPP expressing β-cells. (A) Upon first discovery (P0 – P1), wild-type Ins1+/+Ins2+/+ and insulin knockout Ins1-/-Ins2-/- mice were weighed and their blood glucose was measured (n = 10-31). Individual animal weight or blood glucose is shown on box and whisker plots. Statistical analysis was performed using a one-way ANOVA with Tukey test for multiple comparisons of body weight between all groups. ***p < 0.001. (B) Immunostaining of P1 pancreata for insulin (INS), islet amyloid polypeptide (IAPP), β-galactosidase (βGAL) and nuclei (DAPI) in wildtype Ins1+/+Ins2+/+ controls (upper panels) and Ins1-/-Ins2-/- mice (lower panels), revealing islet-like clusters of INS-IAPP+βGAL+ cells in Ins1-/-Ins2-/- pancreas (representative micrographs from n = 3). Scale bar is 100 µm.     60 Insulin-deficient islets are enlarged and have an abnormal endocrine cell distribution We characterized the endocrine cell distribution of neonatal islets to determine if a lack of insulin causes a shift in endocrine cell proportions. Ins1+/+Ins2+/+ and Ins1-/-Ins2-/- islets had a similar distribution of GCG+ α-cells and SST+ δ-cells but no obvious pathological immune infiltration nor fibrosis by Hematoxylin and Eosin and Masson’s Trichrome staining (Figure 2-3A). Among all hormone immunoreactive cells, the relative proportion of GCG and PP immunoreactive cells decreased and the proportion of SST immunoreactive cells increased in islets of Ins1-/-Ins2-/- mice compared to Ins1+/+Ins2+/+ controls (Figure 2-3B). The proportion of total pancreas section area immunoreactive for IAPP and SST was increased and PP was decreased in Ins1-/-Ins2-/- pancreas compared to Ins1+/+Ins2+/+ controls (Figure 2-3C; proportion of total pancreas section area immunoreactive for each hormone is used as a surrogate of total endocrine cell mass376). We immunostained for ghrelin and found extremely rare cells as expected in both controls and insulin-deficient islets (data not shown). Insulin-deficient islets were fewer in number (Figure 2-3D) but over triple in size compared to islets in pancreas of Ins1+/+Ins2+/+ control mice (Figure 2-3E). We immunostained for the proliferating cell nuclear antigen (PCNA) as a marker of replicating cells and observed a pronounced increase in proportion of IAPP+PCNA+/IAPP+ cells in Ins1-/-Ins2-/- mice compared to controls (Figure 2-3F). Figure 2-2 Variable MafA immunoreactivity in mice with reduced insulin gene copy number. Mice with reduced insulin gene dose have reduced MafA immunoreactivity. Pancreata from 2-month-old mice were immunostained for insulin and MAFA (representative micrographs from n = 3). Scale bar is 100 μm.    61   62 Figure 2-3 Ins1-/-Ins2-/- mice have enlarged islets with abnormal islet cell number and distribution. (A) P1 pancreata of Ins1+/+Ins2+/+ (top panels) and Ins1-/-Ins2-/- (bottom panels) mice were immunostained for IAPP and GCG or GCG, SST, and PP. Hematoxylin & Eosin, and Masson’s Trichrome staining of P1 pancreas (representative micrographs from n = 3-5). Scale bars are 100 µm. (B) Percentage of total endocrine immunoreactive area, that is immunoreactive for the islet hormones IAPP, GCG, SST, and PP (n = 4-5 animals, 3 sections quantified per animal). (C) β-cell, α-cell, δ-cell, and PP-cell area relative to total pancreas area (n = 4-5 animals, 3 sections quantified per animal). (D) Number of islets identified by synaptophysin immunoreactivity relative to pancreas section area (n = 4-5, 3 sections quantified per animal). (E) Average islet size by synaptophysin immunoreactive area (n=4-5, 3 sections quantified per animal). (F) Quantification of percentage of IAPP+ cells that were PCNA+; representative image shown (n = 5). Scale bar is 100 µm. Individual animals are shown on box and whisker plots. Statistical analysis was performed using a Student’s T-test to assess significance. *p < 0.05, **p < 0.01, ***p < 0.001 vs Ins1+/+Ins2+/+ controls.  Insulin-deficient β-cells lack markers of mature β-cells We examined the maturity of the insulin-deficient β-cells by immunostaining for markers of mature β-cells. For some proteins, the immunoreactivity coincided with the feeding/glycemia status of the animals. IAPP+ cells of hypoglycemic insulin-deficient mice had normal pancreatic and duodenal homeobox 1 (PDX1; Figure 2-4A), but completely lacked immunoreactivity for homeodomain transcription factor 6.1 (NKX6.1), in contrast to Ins1+/+Ins2+/+ controls (Figure 2-4B). We also detected less frequent immunoreactivity of homeodomain transcription factor 2.2 (NKX2.2) in β-cells (IAPP+) of Ins1-/-Ins2-/- mice compared to Ins1+/+Ins2+/+ controls (Figure 2-4C). Ins1-/-Ins2-/- mice had a reduced proportion of IAPP+ cells that expressed paired box 6 (PAX6) and IAPP+PAX6+ cells had reduced intensity of PAX6 immunoreactivity compared to Ins1+/+Ins2+/+ controls (Figure 2-4D), as did the remainder of the islet cells. MAFA immunoreactivity was observed in a lower proportion of βGAL+ β-cells in insulin knockouts compared to INS+ cells in control mice and the intensity of immunoreactivity in cells with positive MAFA immunoreactivity trended to be reduced (Figure 2-4E).    63   64 In hyperglycemic Ins1-/-Ins2-/- neonates, NKX6.1, NKX2.2, PAX6, MAFA, and glucose transporter 2 (GLUT2) immunoreactivity was completely lacking (Figure 2-5A) and there was a reduced percentage of islet area with positive immunoreactivity for NKX6.1, NKX2.2, PAX6, MAFA, GLUT2, SUR1, and PC2 relative to control Ins1+/+Ins2+/+ mice (Figure 2-5B). GLUT2 immunoreactivity was detected in fewer IAPP+ cells in Ins1-/-Ins2-/- animals and IAPP+GLUT2+ cells had reduced intensity of GLUT2 immunoreactivity in Ins1-/-Ins2-/- animals compared to Ins1+/+Ins2+/+ controls (Figure 2-6A). In neonates that had not yet suckled and were hypoglycemic, immunoreactivity of SUR1 (Figure 2-6B), PC1/3 (Figure 2-6C), and PC2 (Figure 2-6D) was comparable to Ins1+/+Ins2+/+ controls. We examined transcript levels by qPCR in isolated islets. There was reduced expression of Ins, Gcg, Sst, and Ppy, as well as Pdx1, Nkx6.1, Nkx2.2, Pax6, Sur1, and Pcsk2 in hypoglycemic Ins1-/-Ins2-/- pups compared to Ins1+/+Ins2+/+ controls and mRNA expression levels of Slc2a2, Gck, and Pcsk1 were similar to levels in islets from control mice (Figure 2-6E). We found comparable expression of Amy1 in samples, indicating that the differences observed in expression levels of other target genes were unlikely a result of variable pancreatic exocrine contamination of samples.   Figure 2-4 β-cells in hypoglycemic Ins1-/-Ins2-/- mice lack immunoreactivity for some β-cell transcription factors. P0 – P0.5 pancreata was immunostained for PDX1 (A), NKX6.1 (B), NKX2.2 (C), PAX6 (D), and MAFA (E). Representative micrographs shown (n = 3). Quantification of percentage of IAPP+ cells (A-D) or INS/βGAL+ cells (E) immunoreactive for the selected transcription factor and fluorescent intensities relative to Ins1+/+Ins2+/+ controls in PDX1+ (A), NKX6.1+ (B), NKX2.2+ (C), PAX6+ (D), or MAFA+ (E) cells shown to the right of representative images. Individual animal values are shown on box and whisker plots. A Student’s T-test (target+ β-cells A-E and fluorescent intensity B, D, E) or Mann-Whitney U test (fluorescent intensity A, C) was used to assess significance. *p < 0.05, **p < 0.01 vs. Ins1+/+Ins2+/+ controls. Scale bar is 100 µm and insets are enlarged 8x.  65  Figure 2-5 Hyperglycemic Ins1-/-Ins2-/- β-cells lose expression of key β-cell genes. (A) P1 pancreata from hyperglycemic Ins1-/-Ins2-/- mice were immunostained for PDX1, NKX2.2, PAX6, MAFA, GLUT2, SUR1, PC1/3, or PC2 (red) and IAPP, βGAL/INS, or GCG (green). Representative micrograph shown (n = 3). Scale bar is 100 μm and insets are enlarged 8x. (B) Quantification of proportion of islet area with positive immunoreactivity to protein of interest. Individual animals are shown on box and whisker plots. Statistical analysis was performed using a Student’s T-test to assess significance. **p < 0.01, ***p < 0.001 vs Ins1+/+Ins2+/+ controls.  66   1  67 Insulin-deficient β-cells express progenitor markers We next determined if insulin-deficient β-cells of Ins1-/-Ins2-/- mice resembled dedifferentiated β-cells. We immunostained pancreas from three neonatal Ins1+/+Ins2+/+ and three neonatal hypoglycemic Ins1-/-Ins2-/- animals for factors associated with a progenitor state that have been detected in dedifferentiated β-cells35. Many βGAL+ β-cells in Ins1-/-Ins2-/- mice were immunoreactive for v-myc avian myolocytomatosis viral oncogene lung carcinoma derived (L-MYC), homeobox protein NANOG (NANOG), and aldehyde dehydrogenase 1 family member A3 (ALDH1A3; Figure 2-8A-C). This contrasts with neurogenin-3 (NGN3), which was present in the perinuclear area of synaptophysin immunoreactive (SYN+) islets cells in Ins1+/+Ins2+/+ controls but absent from Ins1-/-Ins2-/- islet cells (Figure 2-8D). We repeated NGN3 immunostaining using two commercial antibodies (R&D Systems, catalog: AF3444 and Thermo Fisher Scientific, catalog: PA5-11893) and found similar patterns of immunoreactivity (Figure 2-7). The proportion of β-cells that were L-MYC+ or ALDH1A3+ was significantly increased in Ins1-/-Ins2-/- mice relative to controls. Ins1-/-Ins2-/- β-cells had higher intensity for immunoreactive NANOG and L-MYC compared to control samples. We attempted to quantify the corresponding gene levels by qPCR but could only detect sufficient transcript of Aldh1a3 which was increased ~8-fold in islets of in Ins1-/-Ins2-/- mice relative to controls (Figure 2-8C). Figure 2-6 Hypoglycemic Ins1-/-Ins2-/- β-cells lack expression of key β-cell factors. P0 – P0.5 pancreata were immunostained for GLUT2 (A), SUR1 (B), PC1/3 (C), and PC2 (D). Representative micrographs shown (n = 3). Quantification of percentage of INS/βGAL+ cells (A-B) or IAPP+ (C) cells immunoreactive for the selected protein and fluorescent intensities relative to Ins1+/+Ins2+/+ controls in GLUT2+ (A), SUR1+ (B), or PC1/3+ (C) cells shown to the right of representative images. (D) Percentage of GCG- islet cells expressing PC2 and relative fluorescent intensity of PC2 immunoreactivity in GCG-PC2+ cells. Scale bar is 100 µm and insets are enlarged 8x. (E) Relative mRNA quantification from isolated P0 – P05 islets from Ins1+/+Ins2+/+, Ins1-/-Ins2-/-, and Ins1-/-Ins2+/- pups. Individual animals are shown on box and whisker plots. A Student’s T-test (target+ β-cells B, D and fluorescent intensity A-D), Mann-Whitney U (target+ β-cells A, C), one-way ANOVA with Tukey test for multiple comparisons (E [Gcg, Nkx2.2, Slc2a2, Gck, Pcsk1, Amy1]), or Kruskal-Wallis test with Dunn’s multiple comparisons test (E [Sst, Ppy, Pdx1, Nkx6.1, Pax6, Abcc8, Pcsk2]) was used to assess significance. *p < 0.05, **p < 0.01, ***p < 0.001 vs controls.  68   Figure 2-7 Consistent patterns of immunofluorescence with two commercially available NGN3 antibodies. P1 pancreata from Ins1-/-Ins2-/- and Ins1+/+In2+/+ mice were immunostained for NGN3 using two commercially available antibodies. Scale bar is 50 μm.     69  Figure 2-8 Ins1-/-Ins2-/- β-cells are not fully differentiated. P0 – P0.5 pancreata from Ins1+/+Ins2+/+ and Ins1-/-Ins2-/- mice were immunostained for L-MYC (A), NANOG (B), ALDH1A3 (C) and NGN3 (D). Representative micrographs shown (n = 3). Quantification of percentage of INS/βGAL+ cells (A-C) or Synaptophysin+ (SYN) cells (D)  70 immunoreactive for the selected protein and fluorescent intensity relative to Ins1+/+Ins2+/+ controls in L-MYC+ (A), NANOG+ (B), ALDH1A3+ (C), or NGN3+ (D) cells shown to the right of representative images. Relative mRNA expression of Aldh1a3 (D) by 2-ΔΔCt method in isolated P0 – P0.5 islets from Ins1+/+Ins2+/+, Ins1-/-Ins2-/-, and Ins1-/-Ins2+/- pups. Individual animals are shown on box and whisker plots. A Student’s T-test (target+ β-cells A,C, D and fluorescent intensity A-D), Mann- Whitney U (target+ β-cells B), or Kruskal-Wallis test with Dunn’s multiple comparisons test (qPCR relative expression) was used to assess significance. *p < 0.05, ***p < 0.001 vs controls. Scale bar is 50 µm and insets are enlarged 4x.  Insulin replacement by injection leads to islet fibrosis and expanded islet cell area  To determine if restoration of insulin signaling by exogenous insulin therapy was sufficient for completed β-cell maturation, we treated insulin-deficient animals with insulin injections for 2 months. Insulin injected Ins-/-Ins2-/- mice gained weight but at a slower rate than control animals (Figure 2-9A). Following insulin therapy by injection or isogenic islet transplant from Ins1+/+Ins2+/+ donors, animals retained clusters of IAPP+ cells, but the majority of IAPP+ cells were βGAL- (Figure 2-9B-D). Seeking to clarify this observation, we examined INS and βGAL immunoreactivity in Ins1-/-Ins2+/- and Ins1+/+Ins2-/- islets. Evidently there can be unique expression of Ins1 versus Ins2 within β-cells, as many INS+ cells were βGAL- in Ins1+/+Ins2-/- mice. Additionally, it appears there can be unique expression of the two Ins2 alleles within β-cells, as Ins1-/-Ins2+/- islets also had abundant INS+ cells that were βGAL- (Figure 2-9E). Following insulin therapy by injections for two months, there was islet cell hyperplasia and obvious islet fibrosis by Masson’s trichrome staining in insulin-deficient islets (Figure 2-10A). Though a reduced proportion of endocrine cells were IAPP+ β-cells (Figure 2-10B), we observed an expansion of all endocrine cell types relative to total pancreatic area in Ins1-/-Ins2-/- mice compared to Ins1+/+Ins2+/+ controls (Figure 2-10C). Further, Ins1-/-Ins2-/- islets trended to being larger (Figure 2-10D-E) and there was a trend towards an increased proportion of IAPP+PCNA+/IAPP+ islet cells compared to in controls (Figure 2-10F).    71  Figure 2-9 Ins1-/-Ins2-/- β-cells lose expression of βGAL after insulin therapy to keep mice alive into adulthood. (A) Body weight of Ins1-/-Ins2-/- animals treated with insulin injections and Ins1+/+Ins2+/+ controls (n = 7). Data are shown as mean ± SEM. Statistical analysis was performed using a repeated measures ANOVA with Bonferroni post hoc testing. ***p < 0.001 vs Ins1+/+Ins2+/+ controls. (B) After 2 months of insulin therapy by injection alone, the majority of β-cells of Ins1-/-Ins2-/- mice lost expression of βGAL and there were few INS-IAPP+βGAL+ cells (rare examples marked by arrows). After 9 months of insulin therapy by injection alone (C) or 12 months of insulin therapy by islet transplantation (D) the same pattern existed and most β-cells of Ins1-/-Ins2-/- mice lost expression of βGAL and there were few INS-IAPP+βGAL+ cells (rare examples marked by arrows) (E) Immunostaining of pancreata from 2-month-old Ins1+/+Ins2+/+ mice, Ins1+/+Ins2-/- mice, and Ins-/-Ins2+/- mice for INS and βGAL. Scale bars are 100 μm.  72  Figure 2-10 Adult Ins1-/-Ins2-/- mice treated by insulin injections have islet hyperplasia and fibrosis.  73 (A) Pancreata of 2-month-old Ins1+/+Ins2+/+ (top panels) and insulin injection treated Ins1-/-Ins2-/- (bottom panels) mice were immunostained for IAPP and GCG or GCG, SST, and PP. Hematoxylin & Eosin, and Masson’s Trichrome staining of pancreas (representative micrographs from n = 3-5). Scale bars are 100 µm. (B) Percentage of total endocrine (synaptophysin+) immunoreactive area, that expressed the islet hormones IAPP, GCG, SST, or PP. (C) β-cell, α-cell, δ-cell, and PP-cell area relative to total pancreas area (n = 4 animals, 3 sections quantified per animal). (D) Number of islets identified by synaptophysin immunoreactivity relative to pancreas section area. (E) Average islet size by synaptophysin immunoreactive area. (F) Quantification of percentage of IAPP+ cells that were PCNA+; representative image shown (n = 4). Scale bar is 100 µm. (B-F) Individual animal endocrine area is shown on box and whisker plots. Statistical analysis was performed using a Student’s T-test (B [IAPP, SST, PP], C[IAPP, GCG, PP], F) or Mann-Whitney U test (B [GCG], C [SST], D, E) to assess significance. *p < 0.05, **p < 0.01, ***p < 0.001 vs Ins1+/+Ins2+/+ controls.  Insulin replacement by injection facilitates partial maturation of Ins1-/-Ins2-/- β-cells To determine if insulin replacement by injection was sufficient for Ins1-/-Ins2-/- β-cells to complete maturation, gain expression of NKX6.1, NKX2.2, MAFA, PAX6, and GLUT2, and lose expression of L-MYC and NANOG, we immunostained for these factors in Ins1-/-Ins2-/- mice treated with insulin injections. Following 2 months of twice daily insulin injections, Ins1-/-Ins2-/- β-cells gained immunoreactivity for NKX2.2 (Figure 2-11A) but still had reduced intensity of NKX6.1 immunoreactivity compared to INS+ Ins1+/+Ins2+/+ β-cells (Figure 2-11B). Further, Ins1-/-Ins2-/- islet cells were deficient in PAX6, MAFA, and GLUT2 compared to control islets (Figure 2-11C-E). There was cytoplasmic NANOG immunoreactivity in Ins1-/-Ins2-/- IAPP+ cells but no L-MYC+ cells (Figure 2-11F-G). Rare ALDH1A3+ cells were observed (Figure 2-11H) and pericytoplasmic immunoreactivity for NGN3 was frequent in Ins1-/-Ins2-/- IAPP+ β-cells and neighboring cells within the islet (Figure 2-11I), similar to Ins1+/+Ins2+/+ neonatal β-cells (Figure 2-8D).   74  Figure 2-11 Insulin therapy by injection is not sufficient for the completed maturation of Ins1-/-Ins2-/- β-cells.  75 Pancreata of 2-month-old Ins1+/+Ins2+/+ (left panels) and insulin injection treated Ins1-/-Ins2-/- (right panels) mice were immunostained for factors associated with mature β-cells (A-E) and a progenitor state (F-J). Ins1-/-Ins2-/- βGAL+ cells express NKX2.2 (A) but had less intense to absent NKX6.1 (B), PAX6 (C), MAFA (D), and GLUT2 (E) immunoreactivity compared with INS+ Ins1+/+Ins2+/+ β-cells. Ins1-/-Ins2-/- βGAL+ β-cells express NANOG (G), and NGN3 (I), but not L-MYC (F) and only extremely rare ALDH1A3+IAPP+ cells (H). Scale bars are 100 µm and insets are enlarged 8x (A-E) or 4x (F-I).  Insulin replacement by islet transplantation facilitates expansion and completed maturation of Ins1-/-Ins2-/- β-cells We next determined if restoration of insulin signaling by isogenic islet transplant into the anterior chamber of the eye was sufficient for completed β-cell maturation. We attempted to have a side-by-side comparison of the endocrine pancreas of animals treated for ~1 year by insulin injections (n=1) alone or islet transplant into the anterior chamber of the eye at two weeks of age (n=4; Figure 2-12). Due to major technical challenges in keeping Ins1-/-Ins2-/- animals alive for such a prolonged time, we were only able to collect samples from one mouse at 291 days of age. The mouse treated by injections had extreme islet fibrosis (Figure 2-12A) with a near complete loss of IAPP+ cells and islets were >50% GCG+ (Figure 2-13A). Compared to Ins1+/+Ins2+/+ controls, mice treated with an islet transplant had grossly normal appearing islets without signs of fibrosis (Figure 2-12A) and did not have significantly altered islet hormone proportions (Figure 2-12B), but there was a slightly expanded α-cell area (Figure 2-12C) and mice had enlarged islets (Figure 2-12E).  Unlike mice treated with insulin injections for 2 months (Figure 2-11) pancreatic β-cells of Ins1-/-Ins2-/- mice transplanted with healthy islets had normal immunoreactivity for PAX6, MAFA, and GLUT2 (Figure 2-13B) but we detected the abnormal presence of cells with cytoplasmic NANOG immunoreactivity (Figure 2-14). NGN3 immunoreactivity was not detected in pancreas of Ins1-/-Ins2-/- mice transplanted with islets, similar to adult Ins1+/+Ins2+/+ controls (Figure 2-14).  76  Figure 2-12 Insulin replacement by islet transplant leads to islet hyperplasia and preservation of β-cell mass, unlike insulin injections. (A) Pancreata of Ins1+/+Ins2+/+ (top panels; n = 3, 12 months of age), insulin injection treated Ins1-/-Ins2-/- (middle panels; n = 1, 9 months of age), and islet transplant treated Ins1-/-Ins2-/- (bottom panels; n = 4, 12-14 months of age) mice were immunostained for IAPP and GCG or GCG, SST, and PP. Arrows point to examples of polyhormonal cells that are GCG+PP+ or SST+PP+. Hematoxylin & Eosin, and Masson’s Trichrome staining of pancreas. Scale bars are 100 µm. (B) Percentage of total endocrine immunoreactive area, that is immunoreactive for the islet hormones IAPP, GCG, SST, or PP. (C) β-cell, α-cell, δ-cell, and PP-cell area relative to total pancreas area. Number (D) and size (E) of islets identified by synaptophysin immunoreactivity relative to pancreas section area. (F) Quantification of percentage of IAPP+ cells that were PCNA+. Individual animal endocrine area is shown on box and whisker plots. Ins1+/+Ins2+/+ and islet treated Ins1-/-Ins2-/- groups were compared by Student’s T-test (B [IAPP, GCG, SST], C [IAPP, GCG, SST], D, E) or Mann-Whitney U test (B [PP], C [PP], F) to assess significance. *p < 0.05 vs Ins1+/+Ins2+/+ controls.  77  Figure 2-13 β-cells from islet transplant treated Ins1-/-Ins2-/- mice express markers of mature β-cell function. (A) Pancreas from an insulin injection treated Ins1-/-Ins2-/- mouse (9 months of age) was immunostained for IAPP, GCG, SST, PP, and SYN. Proportions and relative area of each  78 endocrine hormone was quantified as well as islet size and number using SYN immunoreactivity. Dashed lines indicate values of Ins1+/+Ins2+/+ mice (12 months of age) shown in Figure 8. (B) Pancreata of adult Ins1+/+Ins2+/+ (left panels; n = 3, 12 months of age), insulin injection treated Ins1-/-Ins2-/- (middle panels; n = 1, 9 months of age), and islet transplant treated Ins1-/-Ins2-/- (right panels; n = 4, 12-14 months of age) mice were immunostained for NKX2.2, PAX6, MAFA, and GLUT2. Scale bar is 100 μm and insets are enlarged 8x.    Figure 2-14 β-cells of islet transplant treated Ins1-/-Ins2-/- mice are more mature. Pancreata of Ins1+/+Ins2+/+ (left panels; n = 3, 12 months of age), insulin injection treated Ins1-/-Ins2-/- (middle panels; n = 1, 9 months of age), and islet transplant treated Ins1-/-Ins2-/- (right panels; n = 4, 12-14 months of age) mice were immunostained for L-MYC, ALDH1A3, NANOG, and NGN3. Scale bar is 100 μm and insets are enlarged 4x.  79 2.4 Discussion  Mice lacking IRs on β-cells have impaired glucose stimulated insulin secretion355 and β-cell mass expands during development360. IR deficient β-cells are unable to sufficiently expand in response to hepatic insulin resistance, resulting in adult onset diabetes356. In the current study, we characterized pancreas in neonatal Ins1-/-Ins2-/- mice and Ins1-/-Ins2-/- mice kept alive into adulthood using insulin therapy. By using the insulin knockout mouse model as an alternate to insulin receptor deficient mice, we provide evidence that insulin itself is necessary for β-cell maturation and insulin deficiency alters islet development. However, we cannot rule out the possibility that IGF signaling though the IR is altered due to changes to expression of the IR or IGF1 receptor (IGF1R) and downstream signaling components. Additionally, it is possible that other secondary defects to the genetic deletion of insulin contribute to the observed β-cell phenotype. Independent of these caveats, we posit that the insulin deficient mouse model provides the best evidence available that signaling by insulin itself is necessary for β-cell maturation. Certainly, the Ins1-/-Ins2-/- mouse model does not have any insulin signaling via an IR independent pathway and unlike the IR deficient model, there is no direct and complete loss of IGF signaling through the IR. At birth, insulin-deficient islets were enlarged with an expanded IAPP+ area. Notwithstanding the caveat that IAPP in δ-cells partially contributed to the assessed IAPP+ area, our findings are consistent with observations of expanded β-cell mass and increased β-cell replication in the Ins1-Cre/ERT; IRflox/flox (tamoxifen at E13) model360. The similar results of these two studies reduce the likelihood that Trinder et al.’s findings of expanded β-cell mass were attributable to the presence of a GH minigene within the Ins1-Cre/ERT transgene377, 378. We also observed a lack of mature β-cell factors PDX1, NKX6.1, MAFA, and membranous GLUT2 in insulin-deficient β-cells. These findings contrast with normal immunoreactivity for PDX1, NKX6.1, MAFA, and GLUT2 in pancreata from Ins1-Cre/ERT; IRflox/flox (tamoxifen at E13)  80 mice. This discrepancy can be potentially explained in three ways: 1) insulin before E13 may be essential for the normal formation of maturing β-cells, 2) though insulin signals through the IGF1R with relatively low affinity379, increased expression of the IGF1R in the islets of Ins1-CreER; IRflox/flox (tamoxifen at E13) mice360 may contribute to meaningful insulin signaling through the IGF1R, and 3) despite normal blood glucose, Ins1-/-Ins2-/- pups have hypertriglyceridemia which could contribute to loss of membranous GLUT2380 as well as β-cell dysfunction and loss of mature β-cell factors381. Dedifferentiated β-cells in patients with type 2 diabetes lose their hormone producing status and lose insulin immunoreactivity37, 382 and in mouse models of diabetes, β-cells express progenitor markers L-MYC and NANOG35. Insulin deficient β-cells in Ins1-/-Ins2-/- mice at birth appeared similar, with expression of both L-MYC and NANOG. Notably, we are unaware of any evidence of L-MYC or NANOG being present in the normal developing pancreas, suggesting that insulin deficient β-cells may not be arrested at a developmental stage. Instead, insulin deficient β-cells appear to dedifferentiate after failed maturation due to a lack of the β-cell defining protein – insulin. Furthermore, like β-cells of patients with type 2 diabetes, an elevated proportion of the insulin deficient β-cells expressed ALDH1A337, a marker of dysfunctional β-cells382. NGN3 was expressed in neonatal Ins1+/+Ins2+/+ islet cells, but not in insulin deficient islet cells. Though Ngn3 was not detectable in islet cells by in situ hybridization at birth383, NGN3 is present in normal post-natal β-cells384 and in non-fully differentiated β-cells that are still dividing385, and β-cells replicate after birth386. A lack of NGN3 in Ins1-/-Ins2-/- β-cells is consistent with reduced expression of the post-NGN3 factor NKX2.288, 387. As NGN3 is necessary for pancreatic endocrine cell formation383, there must be NGN3 early in development to initiate the formation of pancreatic islets but a lack of insulin results in a secondary loss of NGN3 expression later in development. Our results suggest that insulin signaling contributes to the maturation of β-cells, and in the absence of insulin, β-cells fail normal maturation and  81 dedifferentiate to an early replicating embryonic state lacking NGN3 and downstream factors including NKX2.2 and MAFA, and expressing pluripotency markers NANOG and L-MYC. Replacing insulin contributes to the further maturation of insulin deficient β-cells. After insulin injection therapy, insulin deficient β-cells were not immunoreactive for L-MYC or ALDH1A3, but we observed islet fibrosis. Additionally, β-cells of insulin injection treated Ins1-/-Ins2-/- adults resembled neonatal wild-type β-cells, with perinuclear immunoreactivity for NGN3. This cytoplasmic localization of NGN3 has been previously reported as marker of newly forming β-cells: cytoplasmic NGN3 has been observed during β-cell regeneration following immunological destruction of β-cells388 and has been observed in early phases of endocrine cell neogenesis in vitro by stimulation with a growth factor cocktail389. Additionally, unlike the nuclear localization of NANOG observed in neonatal Ins1-/-Ins2-/- islet cells that is conventionally a marker of self-renewal in stem cells390, insulin treated mice (by insulin injection or islet transplantation) had cytoplasmic NANOG which has been used as a marker of an epithelial-to-mesenchymal transition in nasopharyngeal carcinoma391, cervical cancer392, and pancreatic cancer393. This contrasts to endogenous β-cells in Ins1-/-Ins2-/- animals treated with insulin replacement by islet transplantation, that had normal immunoreactivity for all mature β-cell factors examined, including MAFA. Insulin deficient β-cells of islet transplant treated adult mice also had a normal absence of L-MYC, ALDH1A3, and NGN3. Continued presence of cells with cytoplasmic NANOG immunoreactivity provides weak evidence for ongoing epithelial-to-mesenchymal transition of non-endocrine cells as a source of new islet cells contributing to islet hyperplasia. Because the primary deficit of Ins1-/-Ins2-/- mice is a loss of insulin, it is somewhat surprising that replacing insulin by injections alone or islet transplant resulted in dramatically divergent outcomes for the endogenous β-cells. We proposed three variables that could contribute to the differences in β-cell phenotype between islet transplant and insulin injection  82 treated Ins1-/-Ins2-/- mice: 1) superior glycemic control in mice treated with islet transplantation relative to those treated with insulin injections, 2) native mouse insulin produced by transplanted islets may signal in β-cells with higher bioactivity than recombinant insulin394, 395, or 3) insulin deficient islets may fail to produce other essential factor(s) that are replaced by transplanted islets. Though we are unable to conclusively discern which, if any, of these variables contribute to the differences in β-cell phenotype, the striking resemblance in fibrotic islet phenotype between injection treated Ins1-/-Ins2-/- mice and mice with diabetes as a result of inexcitable β-cells (KATP gain of function)34 suggests that differences in glycemia are at least partially responsible. Hyperglycemia leading to glucotoxicity has also been shown to cause reduced expression of Pdx1, Mafa, and Slc2a2381 and reprogramming of exocrine cells to insulin-producing cells was more complete and abundant from a viral therapy (expressing NGN3/PDX1/MAFA) when mice were treated with islet transplant for good control of their toxin induced diabetes compared to crude treatment with insulin pellets149. Similarly, in mice with the KATP gain of function model of neonatal diabetes, those treated with islet transplant did not develop islet fibrosis and isolated endogenous islets treated with sulfonylureas had glucose stimulated insulin secretion, unlike islets from the untreated hyperglycemic group396. These findings further suggest that the crude glycemic regulation by insulin injections (or pellets) compared to the ideally regulated glycemia of islet transplant treated mice is likely an important contributor to the differences in β-cell phenotype. Transplanted islets also secrete additional peptides beyond insulin, including C-peptide, a byproduct of proinsulin processing that may contribute to preservation of islet health397. Finally, in this study the mice treated by islet transplant were kept alive for approximately 1 year whereas injection treated mice were kept alive for 2 months, with the exception of a single mouse that was kept alive by injections for nearly one year. The extreme challenge to keep mice alive by multiple daily injections for a long  83 duration limited our ability to have age matched groups and the different duration of therapy may have contributed to the differences in phenotype.  Ins1-/-Ins2-/- mice develop aggregates of endocrine cells in the pancreas resembling islets but a loss of insulin alters the cellular composition. Ins1-/-Ins2-/- islets have a reduced proportion of GCG+ and PP+ endocrine cells, and increased IAPP+ and SST+ cell mass. Given the known contribution of PAX4 to the β- and δ-cell lineages and ARX to the α- and PP-cell lineages87, 398, reduced GCG+ and PP+ and expanded IAPP+ and SST+ cell populations align with an overall reduction in ARX and increase in PAX4 signaling. Additionally, there was a progressive expansion of GCG+ α-cells in mice treated with insulin injections that was not observed in mice treated with islet transplantation. A loss of insulin signaling in α-cells contributes to α-cell hyperplasia399 and poor glycemic control in insulin injected mice also may contribute to progressive α-cell expansion400. Additionally, we made the surprising observation that the pan-endocrine factors NKX2.2 and PAX6 are absent from not only β-cells, but also the remainder of the islet cells. Reduced PP+ cells in the pancreas aligns with the phenotype of NKX2.2 deficient endocrine pancreas88 but unlike the PAX6 knockout mouse90, we did not observe expanded ghrelin+ cells. Likely there is an undefined paracrine effect inhibiting the normal maturation of non-β islet cells in the pancreas of insulin deficient mice. MAFA immunoreactivity was diminished in β-cells of mice with reduced insulin gene copy number (Ins1-/-, Ins2-/-, and Ins1-/-Ins2+/-). Though we did not follow-up on these findings, they raise the intriguing possibility that despite not having obvious severe abnormalities in glucose homeostasis401, reduced insulin dosed β-cells may not be functionally mature due to MAFA insufficiency402 and perhaps other unidentified factors. Additionally, we made an unexpected observation that there can be unique regulation of the Ins2 and Ins1 loci and the two Ins2 alleles in mouse β-cells. With lacZ knocked into the Ins2 locus, the majority of β-cells in adult Ins1-/-Ins2+/- or Ins1+/+Ins2-/- animals are INS+βGAL-. In Ins1+/+Ins2-/- animals, cells that  84 were INS+βGAL- had an active Ins1 gene but the Ins2 knock-in of βGAL was inactive. In Ins1-/-Ins2+/- animals, cells that were INS+βGAL- had an active wildtype Ins2 allele but the βGAL knock in allele was inactive. Similarly, loss of βGAL in most IAPP expressing cells happens after birth in Ins1-/-Ins2-/- animals. Though chronic hyperglycemia has been shown to suppress insulin transcription403, loss of βGAL cannot be attributed to a diabetic state as postnatal loss of βGAL expression also occurs in euglycemic Ins1-/-Ins2-/- mice treated by islet transplant and Ins1+/+Ins2-/- and Ins1-/-Ins2+/- animals. As expression of insulin is arguably the most important aspect of the pancreatic β-cell identity, loss of β-galactosidase does not align with a mature β-cell identity. It is however important to consider the model organism, because age dependent loss of βGAL in insulin treated Ins1-/-Ins2-/- mice is consistent with reduced lacZ expression when knocked into the globin genes in mice404. The distinct expression pattern of βGAL compared to insulin in the insulin knockout/lacZ knock in model may be a caveat of the mouse model because there can be selective CpG methylation to silence a foreign open reading frame like the lacZ gene405. With that said, there is already some evidence that Ins2 is translated more efficiently than Ins1406 and there is maternal imprinting of Ins2407, suggesting distinct transcriptional and translational regulation of the mouse insulin genes. It may be worthwhile to further explore the possibility of differential regulation of the mouse insulin genes and alleles. There have been many reported human cases of diabetes caused by mutations in the insulin gene20. Most of these patients have a heterozygous dominant negative disease caused by misfolding of the mutated insulin, but there has been a report of patients with homozygous loss of INS408. Though we found no published histology of the pancreas from such patients, a patient with an intronic mutation causing altered splicing of INS, had undetectable C-peptide but readily detectable IAPP in circulation, suggesting that the patient had insulin deficient β-cells in their pancreas24. This hypothesis is supported by the abundant INS-IAPP+ β-cells in Ins1-/-Ins2-/- mouse pancreas. Related, a child with severe sulfonylurea unresponsive permanent neonatal  85 diabetes mellitus from an activating mutation in the ATP-sensitive potassium channel had severely reduced β-cell mass by post mortem histological examination409. Patients with mutations in the ATP-sensitive potassium channel are less likely to become insulin independent with sulfonylurea therapy the longer the patient has been insulin dependent410. Potentially there is a major change in β-cell maturity following transfer of therapy but prolonged insulin replacement by injection causes progressive loss of β-cell mass thus hindering the ability of sulfonylurea therapy to induce remission of diabetes. This hypothesis aligns with our findings that though a lack of insulin during development initially results in an expanded β-cell mass, insulin replacement by injection is not sufficient for the maintenance of β-cell mass. Collectively, we provide evidence that insulin is a necessary signaling molecule for the maturation of β-cells. Replacement of insulin contributes to β-cell maturation, however the ability of insulin therapy to complete β-cell maturation may be dependent on euglycemia, replacement of the native species of insulin, or other factors secreted from islets. Without insulin during development in mice, β-cells appear not fully mature. Extremely rare mutations of the INS gene may also cause altered maturation of β-cells in patients. Understanding how insulin regulates β-cell maturation is relevant to research into the generation of β-cells in vitro and therapies attempting to induce diabetic remission in patients with dedifferentiated β-cells in early type 2 diabetes. Though there has been some debate about the potential that insulin itself signals in an autocrine fashion on β-cells411, our findings provide evidence that insulin itself is in fact an essential signaling molecule on β-cells.    86 Chapter 3: AAV Ins1-Cre can produce efficient β-cell recombination but requires consideration of off-target effects 3.1 Background Specific in vivo genetic modification is a useful way to assess the impact of genes on relevant physiological processes. One of the most useful genetic tools to study the role of specific genes in vivo is the Cre-LoxP system. By using a tissue specific promoter and conjugating Cre to an estrogen receptor (ER), Cre will recognizes LoxP sites in the genome and based on orientation, can excise, flip, or translocate targets in specific tissues at specific times227. Many Cre driver mouse lines have been generated, including dozens that are specific for the pancreas or certain pancreatic cell lineages357. These tools have been used extensively with great success, but this approach faces important caveats.  One of the earliest pancreatic Cre driver mice developed uses 668bp of the rat insulin 2 (Ins2) promoter to drive Cre, often called the “RIP-Cre” mouse412. This mouse model has been used in many studies, but findings have been challenged because of a lack of appropriate controls. There may be developmental defects from lifelong Cre expression and there are potential impairments in insulin secretion and glucose tolerance in Ins2-Cre transgenic mice413. Moreover, β-cells of “RIP-CreER” (Ins2-cre/Esr1) mice have increased rates of apoptosis in response to glucagon-like peptide 1 stimulation414. Additionally, an important caveat faced by almost all Cre driver lines is imperfect specificity of recombination. Cre and CreER driver lines using the Ins2 promoter have abundant recombination throughout the brain358, 415. Alternative lines make use of a pancreatic and duodenal homeobox 1 gene (Pdx1) promoter to control Cre expression, but it has been reported to induce hypothalamic358, α-cell, and δ-cell recombination416. To address this issue, a more specific Ins1 promoter has been used and may limit these off-target confounds417, 418. Further, inducible ER conjugated Cre models like the  87 Ins2-cre/Esr1 mouse can have tamoxifen-independent recombination419 but this can be limited by use of the mutated ER in Cre-ERT2 420. However, it is notable that delivery of tamoxifen itself (to induce recombination) can alter glucose homeostasis and impair β-cell proliferation361. Additionally, inclusion of a growth hormone minigene in many Cre driver lines (including Ins2-Cre, Pdx1-CreLate, Ins1-Cre and others) is also a source of β-cell dysfunction via local activation of prolactin receptors378 that can induce β-cell proliferation377. Finding a way to avoid caveats of developmental deletion of LoxP flanked genes, growth hormone minigene-induced β-cell dysfunction, unintended recombination before tamoxifen administration, and tamoxifen toxicity would be useful. It is also worth noting the time and costs of crossing Cre driver mice with LoxP containing mice to generate mice with suitable genotypes. Finding ways to minimize these costs and delays would be extremely useful and make complex genetic studies more accessible to laboratories that face time, labour, or financial challenges. The AAV is a well-known and simple vector that could be an alternative approach to deliver Cre to pancreatic β-cells in an adequately specific and cost-effective way.  In the current study we characterize an AAV Ins1-Cre (eighth serotype) for delivery of Cre recombinase to pancreatic β-cells. We delivered variable doses of AAV Ins1-Cre and assessed β-cell function and maturity and demonstrate the utility of this AAV by inducing diabetes in Ins1-/-Ins2f/f mice. Though off-target recombination events likely occurred throughout the liver and exocrine pancreas, we posit that this would have little to no impact because these cells do not normally produce insulin. We avoided months of crossing and genotyping, tamoxifen administration, and lifelong Cre expression. In addition, animals served as their own controls by comparing pre and post AAV injection, and there was no recombination prior to AAV Ins1-Cre injection.   88 3.2 Materials and methods Plasmid preparation The AAV Ins1-Cre plasmid was generated from the dsAAV mouse Ins2-EGFP plasmid, kindly provided by Dr. Paul Robbins264. First, the 1.1kb Ins2 promoter (with 5’ UTR) was excised with BamHI and AgeI and replaced with 410-bp of the rat Ins1 promoter with the 5’ UTR (Ins1; primers: Ins1-F: 5’-CACTGGATCCTGAGCTAAGAATCCAGCTATCAATAGAA ACT; Ins1-R: 5’-CACACAACCCCGTGTTGGAACAATGACCTGGAAGATAG). Next, the EGFP open reading frame (ORF) was excised with AgeI and NotI and the 4.9kb vector fragment was gel-purified. The Cre ORF was generated by PCR amplification from an Addgene CMV-Cre vector (pBS185) with a Kozak sequence and XmaI site added upstream and a NotI site downstream. (primers: CreORF-F: 5’-CACACGCCCGGGGCCGCCACCATGTCCAATTTACTG ACCGTACACCAA; CreORF-R: 5’-CACACGCGGCCGCCTAATCGCCATCTTCCAGCAGGC). The PCR product was digested with NotI and XmaI and ligated into the 4.9kb vector to yield the final dsAAV Ins1-Cre plasmid. The plasmid was sent to the Children’s Hospital of Philadelphia (CHOP) for manufacturing of high titre recombinant dsAAV. AAVs were administered IP and via the pancreatic duct (intraductally; ID), as previously described264. Animal models All procedures with animals were approved by the University of British Columbia Animal Care Committee and carried out in accordance with the Canadian Council on Animal Care guidelines. Eight-week-old C57L/6J mice were ordered from Jackson Laboratories (Bar Harbor, Maine). mTmG mice were generated in the lab of Dr. Liqun Luo421 (Stanford University, Stanford, CA) and ordered from Jackson Laboratories. Rosa26-LSL-Luciferase mice were generated in the lab of Dr. William Kaelin422 (Dana Farber Cancer Institute, Boston, MA) and ordered from Jackson Laboratories. Confetti mice were generated in the lab of Dr. Hans Clevers423 (Hubrecht Institute, Utrecht, Netherlands) and ordered from Jackson Laboratories.  89 Ins1-/- mice were generated in the lab of Dr. J Jami367 (Institut Cochin, Paris, France) and Ins2f/f mice were generated in the lab of Dr. Massimo Trucco424 (University of Pittsburgh, Pittsburgh, PA). All mice employed in this study were housed with a 12 h light, 12 h dark cycle and had ad libitum access to chow diet (2918, Harlan Laboratories, Madison, WI). Adult (6-8 weeks old) male mice were used in all studies. Physiological tests and plasma assays To assess body weight and blood glucose mice were fasted for four hours during the morning (9:00-13:00) prior to measurements using a One Touch Ultra glucometer (Life Scan Inc., Burnaby, Canada). For glucose tolerance tests, mice were fasted for 4 (IP) or 6 hours (oral) and then given a glucose load (2 g glucose/kg of body weight for both IP and oral tolerance tests). Blood was sampled from the saphenous vein and measured for glucose and insulin before injection (t = 0 min) and at different time points post-injection. Plasma insulin levels were determined by a Mouse Insulin Ultrasensitive enzyme-linked immunosorbent assay (ELISA; ALPCO Diagnostics, Salem, NH). Serum AST and ALT activities were measured using commercially available kits (Sigma-Aldrich, St. Louis, MO). In vivo luciferase imaging Rosa26-LSL-Luciferase mice were injected with XenoLight RediJect D-Luciferin Ultra Solution (150 mg/kg) and imaged under isoflurane anesthesia using the IVIS® Imaging System Lumina Series. At study termination animals were injected with the luciferin solution then euthanized before rapid dissection and imaging. Islet isolation, dispersion and cell culture Hank’s balanced salt solution (HBSS) containing type XI collagenase (Sigma-Aldrich, St. Louis, MO) was employed to isolate pancreatic islets as previously described425. Briefly, collagenase solution (1000 U/mL) was injected into the common bile duct then the pancreas  90 dissected out and digested at 37ºC for 11-14 min. Islets were then washed with iced-cold HBSS, filtered through a 70 μm cell strainer, and handpicked. Ca2+ imaging Following 48 h culture at 37ºC and 5% CO2 on glass coverslips, pancreatic islets from infected wild-type C57BL/6J mice were loaded with 5 μM Fura 2-AM for 30 min and imaged on a Zeiss Axiovert 200 M inverted microscope equipped with temperature-controlled stage and a FLUAR 20X objective (Carl Zeiss, Thornwood, NY) as previously described426. During the experiments, islets were continuously perifused with Ringer's solution containing 144 mM NaCl, 5.5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 20 mM Hepes (adjusted to pH 7.35 by NaOH).  Tissue harvest and immunohistochemical analysis Tissues were harvested and fixed in paraformaldehyde (PFA) 4% overnight after transcardial perifusion with PBS for 2 min and ice-cold PFA 4% for 10 min. After fixation, tissues were washed in 70% ethanol and embedded in paraffin, for subsequent sectioning (5 µm thickness; Wax-It Histology Services, Vancouver, Canada). Brains were left in PBS containing 20% sucrose for 4 days and then frozen at -40oC in 2-methylbutance prior to sectioning at 30 µm on a sliding microtome, in a one-in-six series through the rostrocaudal extent of the hypothalamus and immunostained by standard method427. For immunofluorescence of paraffin-embedded tissues, sections were deparaffinized in xylene and rehydrated in graded ethanol before heat-induced epitope retrieval in an EZ Retriever microwave oven (BioGenes, Fremont, CA; 95°C for 15 mins in 10 mM citrate buffer with 0.05% Tween-20). We blocked slides in DAKO Protein Block, Serum Free (Dako, Burlington, Canada) and incubated overnight in primary antibody diluted in DAKO Antibody Diluent followed by incubation in secondary antibody for 1 hour at room temperature. Slides were mounted and counterstained with nuclear stain 4′,6-diamidino-2-phenylindole (DAPI) and VECTASHIELD® Hard Set Mounting Medium with DAPI (Vector Laboratories, Burlingame, CA), and images were captured and analyzed with an  91 ImageXpress® Micro XLS System, controlled by MetaXpress® High-Content Image Acquisition & Analysis Software (Molecular Devices Corporation, Sunnyvale, CA) with a scientific CMOS camera, a Nikon 20× Plan Apo objective (NA = 0.75, 1-6300-0196; Nikon, Tokyo, Japan), and DAPI (DAPI-5060B), FITC (FITC-3540B), Cy3 (Cy3-4040B), Texas Red (TXRED-4040B), and Cy5 (Cy5-4040A) filter cubes. We performed quantification of histological images in the same software (MetaXpress®). All antibodies used are listed in Table 3-1.  Table 3-1 Primary antibodies used for immunofluorescent experiments. Peptide/protein target Antigen Sequence Name of Antibody Manufacturer, catalog #, and/or name of individual providing the antibody Species raised in; monoclonal or polyclonal Dilution used RRID CD3 C-terminus of human CD3 Rabbit anti-CD3 Anaspec 29588 Rabbit Polyclonal 1 to 50 2275572 CD45 Unknown Rat anti-CD45 BD Biosciences 550539 Rat Monoclonal 1 to 25 2174426 dsRED Unknown Rabbit anti-dsRED Clontech 632496 Rabbit Polyclonal 1 to 100 10013483 Glucagon unknown Mouse anti-GCG Sigma G 2654 Mouse Monoclonal 1 to 1000 259852 Glucose transporter 2 First extracellular loop of Glut2 Rabbit anti-GLUT2 Millipore 07-1402 Rabbit Polyclonal 1 to 500 1587076 Green fluorescent protein Full length Aequorea GFP GFP (mono) Clontech 632375 Mouse Monoclonal 1 to 200 2756343 Green fluorescent protein GFP GFP (poly) Life Technologies A11122 Rabbit Polyclonal 1 to 500 221569 Islet amyloid polypeptide unknown Rabbit anti-IAPP AbCam ab15125 Rabbit Polyclonal 1 to 50 2295631 Insulin unknown Rabbit anti-INS Cell Signaling C27C9 Rabbit Monoclonal 1 to 200 2126503 Insulin Residues surrounding Val36 of human insulin Mouse anti-INS Cell Signaling L6B10 Mouse Monoclonal 1 to 250 10949314 V-maf muscoloapo-neurotic fibrosarcoma oncogene homolog A unknown Rabbit anti-MAFA Betalogics (Johnson & Johnson) LP9872 Rabbit Polyclonal 1 to 1000 2665528 homeodomain transcription factor 6.1 Human Nkx6.1 Goat anti-NKX6.1 R and D Systems AF5857 Goat Polyclonal 1 to 20 1857045 Somatostatin Human somatostatin Mouse anti-SST Β Cell Biology Consortium AB1985 Mouse Polyclonal 1 to 500 10014609   92 Statistical analysis Data are expressed as mean ± SEM. Statistical analysis was performed using the Mann-Whitney test, one-way ANOVA, or two-way ANOVA as indicated (GraphPad Prism, GraphPad Software Inc., La Jolla, CA, USA). p values < 0.05 were considered significant. Multilevel modeling was performed in hierarchical linear modeling software (HLM 6.0, Scientific Software International Inc., Lincolnwood, IL).  3.3 Results Intraperitoneal administration of AAV Ins1-Cre up to a dose of 1012 viral genome particles (VGP) does not significantly alter glucose metabolism To determine whether AAV Ins1-Cre (410bp of the rat Ins1 promoter and 5’UTR sequence preceding the Cre open reading frame) leads to Cre expression and subsequently Cre recombination in β-cells, we administered AAV Ins1-Cre at doses of 1010, 1011, 1012, and 3x1012 VGP by a single IP injection to adult reporter mTmG mice. This double fluorescent reporter model contains a chicken β-actin core promoter with a CMV enhancer driving a loxP-flanked coding sequence of membrane-targeted tandem dimer Tomato (mT), resulting in tdTomato expression. Upon Cre recombination, the mT sequence is excised allowing expression of membrane-targeted EGFP (mG)421. A group of PBS-injected mTmG mice were used as controls. As several reports have already shown that transgenic Ins2-Cre mice may develop glucose intolerance413, we studied glucose metabolism over 8 weeks (-1 to 7 weeks relative to AAV). There were no significant changes to four hour fasting blood glucose and body weight between the AAV Ins1-Cre injected mice and controls (Figure 3-1A). Seven weeks post virus administration, we performed an oral glucose tolerance test (OGTT, 2 g/kg body weight of    93   Figure 3-1 Intraperitoneal administration of AAV Ins1-Cre up to a single dose of 1012 VGP does not significantly alter glucose metabolism. (a) 4 hour fasting blood glucose and body weight of adult mTmG reporter mice injected with either vehicle (PBS) or different doses of AAV Ins1-Cre. (b) Oral glucose tolerance test (6 hour fast, 2 g glucose/kg body weight) 7 weeks post virus injection. (c) Plasma insulin levels during oral glucose tolerance test and (d) after 6 hours fasting 8 weeks post virus injection. Data are expressed as mean ± SEM and were analyzed using one- or two-way repeated measures ANOVA. n=4-5 mice per group. (* p <0.05)    94  a 30% dextrose solution) and found that delivery of AAV Ins1-Cre up to a dose of 1012 VGP did not alter glucose excursions; however, 3x1012 VGP caused glucose intolerance (Figure 3-1B). Eight weeks after AAV injections, we assessed insulin secretion during a glucose challenge and there were no significant differences among groups (Figure 3-1C). Basal plasma insulin levels were similar between the AAV Ins1-Cre and PBS injected groups (Figure 3-1D). In addition, we measured serum levels of alanine transferase (ALT) and aspartate aminotransferase (AST) pre- and post- virus administration (Figure 3-2A and B) and did not detect elevated serum transaminases in AAV Ins1-Cre mice compared to controls.        Figure 3-2 AAV Ins1-Cre does not elevate liver enzymes AST and ALT levels were measured in serum from PBS and AAV Ins1-Cre injected animals before (day 0) and on day 14 relative to virus delivery.      95 Intraperitoneal administration of AAV Ins1-Cre (1012 VGP) does not significantly alter calcium nor insulin secretion dynamics in isolated islets Cre-expressing β-cells may display altered intracellular Ca2+ responses when exposed to different stimuli including glucose and KCl426. We performed intracellular Ca2+ imaging experiments in islets from AAV Ins1-Cre- and PBS-treated wildtype C57BL/6J mice by means of Fura 2-AM to determine if Cre expression mediated by AAV Ins1-Cre alters intracellular Ca2+ signaling. We used wild-type mice because of abnormal Ca2+ responses in untreated mTmG islets (data not shown). Using a non-toxic dose of 1012 VGP, we carried out Ca2+ experiments three to four weeks post-injection of either PBS or 1012 VGP of AAV Ins1-Cre. Islets from AAV Ins1-Cre injected mice did not display significantly altered intracellular Ca2+ dynamics in response to elevating glucose levels (8 and 15 mM) or in response to depolarizing 30 mM KCl (Figure 3-3A-B). The area under the curve (AUC) during the stimulation conditions (8mM G, 15mM G, and 30 mM KCl) is presented to the right and there were no significant differences between groups (n = 3 per group, 10-20 cells per animal shown and each animal is a different color. Groups were compared by random intercept multilevel modeling with individual cell AUC nested within mouse, b = -50.72, SE = 33.91, t(4) = -1.50, p = 0.209). In addition, we measured  insulin secretion and there we no differences in insulin secretion when comparing islets isolated from PBS and AAV Ins1-Cre injected mice (Figures 3-3C and 3-3D). Similar to controls, β-cells of AAV Ins1-Cre injected mice had bright nuclear MAFA immunoreactivity (Figure 3-3E). Additionally, β-cells in AAV Ins1-Cre injected mice had nuclear NKX6.1, and membranous GLUT2 immunoreactivity (Figure 3-3F), in support of normal glucose responsive insulin secretion dynamics.   96  Figure 3-3 Intraperitoneal administration of AAV Ins1-Cre (1012 VGP) does not significantly alter β-cell maturity or function.  Adult C57BL/6J mice were given either PBS or 1012 VGP of AAV Ins1-Cre by single intraperitoneal injection, and islets were isolated three to four weeks post-injection. Representative [Ca2+]i recordings of an islet from a PBS treated mouse (a) and an AAV Ins1-Cre injected mouse (b), in response to different glucose (G) concentrations and potassium chloride (KCl). Graphs are representative of 31-47 islets from 3 mice per group. Area under the curve for individual cells shown to the right with each mouse a different color. (c) Insulin secretion in response to G and KCl in islets from PBS and AAV injected mice, and (d) area under the curve (AUC) in response to high glucose (20 mM G) from islets perifused in (c). Batches of 80 islets from 3 mice per group were employed. Mann-Whitney test was performed. Pancreas from adult mTmG mice injected with PBS or AAV Ins1-Cre was immunostained for mature β-cell factors MAFA (e), NKX6.1, and GLUT2 (f). Representative images of n=3 shown. Scale bars are 100 μm.  97 IP AAV Ins1-Cre produces robust β-cell recombination alongside off-target hypothalamic and exocrine pancreas recombination We first demonstrated high recombination efficiency with the AAV Ins1-Cre by infecting islets from mTmG mice in vitro. We observed bright GFP signal throughout the islet for 4 days post-infection (Figure 3-4). We next immunostained pancreas from AAV Ins1-Cre injected mice for insulin and GFP (Figure 3-5A). There were high rates of β-cell recombination (81-93%) in 3/3 mice in the 3x1012 VGP and 2/3 mice in the 1012 VGP dosed mice. There were expected lower rates of β-cell recombination in lower dosed mice with essentially zero recombination in the PBS treated control that has never been exposed to any Cre recombinase. As there are hypothalamic cells that express Ins2401 and AAV is known to infect the brain428, we immunostained hypothalamus (Figure 3-5B). There was rare recombination in low dose 1011 VGP dosed mice, but expansive recombination throughout the hypothalamus of 1012 VGP dosed mice, particularly in neurons of the arcuate nucleus and the dorsomedial nucleus of the hypothalamus, but there was a lack of immunoreactivity in the ventromedial nucleus of the hypothalamus. Additionally, we observed abundant acinar cell recombination and occasional recombination in non-β islet cells (Figures 3-5C and 3-5D) that were immunoreactive for glucagon (α-cells) or somatostatin (δ-cells; Figure 3-5F). Rates of recombination were dose dependent with upwards of 10% of insulin negative cells in pancreatic sections being GFP+. As a potential mechanism of activation of the Ins1 promoter fragment in the AAV Ins1-Cre, δ-cells express the insulin promoter binding transcription factor PDX1 (Pdx1)429 and we analyzed a published single-cell RNAseq database320 and found that 12/185 acinar cells had Pdx1 transcript levels higher than the median of the β-cell population (Figure 3-5E). Notably, GFP expression level functions on a binary system: present (Cre has recombined mTmG) or absent (no recombination event and cells continue to make tomato). A tiny amount of off-target Cre is sufficient to produce equivalent GFP signal to a cell with robust Cre expression. This contrasts  98 to other models like delivering an AAV Ins1-GFP in which there can be a wide gradient of GFP expression such that weak off-target expression may not be detectable.    Figure 3-4 AAV Ins1-Cre can infect islets in vitro Islets from adult reporter mTmG mice were isolated and exposed to 106 VGP and 107 VGP of AAV Ins1-Cre per cell, or PBS, overnight. Endogenous EGFP fluorescence was tracked for the following four days.   99  Figure 3-5 Intraperitoneal AAV Ins1-Cre produces dose dependent β-cell recombination alongside hypothalamic and acinar tissue recombination. Pancreas from mTmG mice administered PBS or varying doses of AAV Ins1-Cre was collected 10 weeks post-AAV and immunostained for insulin and GFP (a), and recombination rate in β-cells is quantified (n=3). Groups were compared to PBS control by Kruskal-Wallis test with  100 Dunn’s post hoc test. (b) Brains from these mice were immunostained for GFP with nuclear DAPI counterstain. A labeled coronal section through the mid-hypothalamus is shown (3V: third ventricle, ARC: arcuate nucleus, VMH: ventromedial nucleus of the hypothalamus, and DMG: dorsomedial nucleus of the hypothalamus). Representative image of n=3. (c-d) Significant exocrine and insulin negative islet cell recombination in AAV Ins1-Cre injected mice. Representative images of n=3 shown. Quantification of the proportion of insulin negative cell recombination was compared to PBS control by one-way ANOVA with Tukey’s post-hoc test. (e) Data from Segerstolpe et al. (2016) was analyzed and Pdx1 FPKM is presented for cells coded as β-cells or acinar cells by authors. Twelve of 185 acinar cells (shown in red) had Pdx1 transcript levels above median levels in β-cells. (f) Pancreas was immunostained for GFP and the islet hormones glucagon (GCG) or somatostatin (SST). Representative images of n=3 shown. Scale bars in all panels are 100 μm. Insets are enlarged 4x. (* p < 0.05, ** p < 0.01)  Intraductal (ID) administration of AAV Ins1-Cre up to a dose of 1011 VGP does not significantly affect body weight, blood glucose levels or glucose tolerance To minimize central recombination and other potential off-target sites observed following IP delivery, we delivered the AAV Ins1-Cre via the common bile duct. In this study we used adult confetti reporter mice that express one of four fluorescent proteins upon Cre mediated recombination events423. Body weight and fasting blood glucose were measured over a period of 35 days and there were no significant differences between the sham group and the AAV Ins1-Cre injected animals (0.2x1011 VGP and 1011 VGP groups; Figures 3-6A and 3-6B). Four weeks post-surgery, there were no significant differences in oral glucose tolerance (Figure 3-6C). Pancreas was collected six weeks post-AAV and immunostained for insulin and fluorescent proteins. As the four fluorescent proteins of the confetti mouse have extremely high sequence similarities, we immunostained with a polyclonal GFP antibody or a monoclonal GFP antibody with a polyclonal dsRed antibody to provide good sensitivity to detect all recombination events (Figure 3-7). We detected three distinct immunostaining patterns that likely represent all fluorescent proteins with the exception of nuclear GFP, which is reported to be less   101  Figure 3-6 AAV Ins1-Cre intraductal administration does not significantly alter glucose tolerance but induction of foreign fluorescent protein expression causes insulitis. Four hour fasting body weight (a) and blood glucose (b) of adult confetti reporter mice that received AAV Ins1-Cre by intraductal delivery. (c) Oral glucose tolerance test (2 g glucose/kg  102 body weight) 4 weeks post virus administration. Data are expressed as mean ± SEM and were analyzed using two-way repeated measures ANOVA. n=7-8 mice in each group (a-c). Pancreas was collected seven weeks post-AAV and immunostained for insulin (INS) and GFP, and the proportion of Ins+ or Ins- cells that were GFP+ was quantified (d). Representative images of n =3 shown. Groups were compared to PBS control by Kruskal-Wallis test with Dunn’s post hoc test. (e) Brains were immunostained for GFP and a coronal section through the mid-hypothalamus is shown (3V: third ventricle, ARC: arcuate nucleus, VMH: ventromedial nucleus of the hypothalamus, and DMG: dorsomedial nucleus of the hypothalamus). Representative image of n=1 of control or n=3 of 1011 VGP injected mice shown. There was no detectable GFP in 0.2x1011 VGP dosed mice (data not shown). (f) Adult confetti mice injected with AAV Ins1-Cre had insulitis with CD3+ and CD45+ cells surrounding islets. Representative images of n=3 shown. (g) Adult C57Bl/6 mice were injected with 5x1012 VGP AAV Ins1-GFP IP and developed insulitis by 4 months post-AAV. Scale bars are 100 μm.   common or absent in many tissues including pancreas of confetti mice423. Using the polyclonal GFP antibody we quantified the rates of β-cell and non-β-cell recombination in the pancreas and found significant recombination in 1011 VGP injected mice (Figure 3-6D). We immunostained brain and found no evidence of recombination in the hypothalamus (Figure 3-6E), in contrast to our findings with IP delivery of the AAV Ins1-Cre (Figure 3-5B). Notably, histological experiments identified significant accumulation of DAPI+INS- cells near islets that resembled insulitis. We immunostained for the T-cell co-receptor CD3 and pan-leukocyte protein CD45 and confirmed heterogenous insulitis (Figure 3-6F). This was not observed in previous studies with mTmG mice; animals with fluorescent protein throughout their lifespan. We observed a similar phenomenon of insulitis in a different model that is also naïve to fluorescent protein expression into adulthood until IP delivery of a high dose AAV Ins1-GFP (Figure 3-6G). Development of insulitis did not appear to depend on route of AAV delivery (IP vs ID) but could be explained by sudden induction of fluorescent protein expression in model systems that lacked fluorescent protein expression prior to AAV delivery. Related to off-target recombination in acinar cells shown in Figure 3-5C, in AAV Ins1-GFP injected mice we note bright GFP signal in targeted β-cells, but no detectable acinar GFP. This contrasts to widespread acinar, α-cell, and δ-cell GFP  103 in Figure 3-5C and highlights the binary outcome of recombination or no recombination in the mTmG or any other lineage trace model employing a Cre transgene.   Figure 3-7 High sequence similarity allows detection of different confetti recombination events using a polyclonal α GFP antibody Pancreas collected from adult confetti mice treated with ID AAV Ins1-Cre was immunostained for fluorescent proteins to identify recombination events yielding nuclear GFP (not clearly observed), cytoplasmic RFP, cytoplasmic YFP, and membranous CFP. Representative images of n=3. Scale bar is 100 μm and insets are enlarged 4x.    DAPI INSULIN  GFP(POLY)DAPI DSRED GFP(MONO)2x1011VGP2x1011VGPCFPMembYFP CytoRFPCytoRFPCytoCFPMembYFP Cyto 104 Both IP and ID administration of AAV Ins1-Cre can cause recombination in the liver As the eighth serotype of AAV (AAV8) has a high propensity to infect hepatocytes264, we looked for evidence of recombination in the liver. In the IP AAV Ins1-Cre injected mTmG mice (Figure 3-8A), there was robust recombination in hepatocytes. We also observed some less abundant recombination in transgenic Ins2-Cre;mTmG mice and based on nuclear and cell shape, GFP+ cells are likely not hepatocytes but are perhaps resident Kupffer cells. There were low rates of recombination in the liver of mice given AAV Ins1-Cre via the pancreatic duct (Figure 3-8B). To further explore liver recombination, we delivered 1012 VGP AAV Ins1-Cre IP to Rosa26-LSL-luciferase mice. When Cre is present in Rosa26-LSL-luciferase mice, Cre removes the LSL (loxP-transcriptional stop site-LoxP) sequence to allow transcription of luciferase. We performed in vivo and ex vivo chemiluminescent experiments following luciferin injection (Figure 3-8C). There was significant luminescence for five weeks post-AAV and by dissecting out organs quickly after luciferin injection, it was revealed that the majority of activity was coming from the liver and minority from the pancreas and adipose tissue. Though early reports using AAV8 chicken β-actin promoter-GFP264 would suggest the pancreas has the highest expression levels, we again revisit the binary outcome of a lineage tracing model like the luciferase mouse. Past work reveals that total expression off AAV8 is likely highest in the pancreas (i.e. measuring GFP from an AAV actin-GFP264) but the current work suggests that the infection rate is highest in the liver. When the detection outcome is binary like a recombination event, higher rates of liver infection lead to higher rates of recombination and thus the highest luciferase activity comes from the liver.  105  Figure 3-8 Both IP and ID administration of AAV Ins1-Cre can cause recombination in the liver but the extent may be dependent on delivery method and model organism. Liver from adult mTmG mice given IP AAV Ins1-Cre (a) or confetti mice given ID AAV Ins1-Cre (b) was immunostained for fluorescent proteins to identify hepatic recombination events. Representative images of n=2-3 shown. Scale bars are 100 μm. (c) Adult Rosa26-LSL-luciferase mice were administered 1012 VGP AAV Ins1-Cre IP (n=3) or PBS (n=1) and luciferase activity was assessed following luciferin injection for 5 weeks including after rapid dissection at study termination.  Deletion of a β-cell specific gene with AAV Ins1-Cre To highlight the utility of the AAV Ins1-Cre to be an alternative to Cre driver mouse lines for producing β-cell recombination, we delivered 1012 VGP IP to Ins1-/-Ins2f/f mice at 6-8 weeks of age. Ins1-/-Ins2f/f mice have deletions of the Ins1 open reading frame and have LoxP sites flanking the Ins2 gene thus making them inducible insulin knockouts. This study is an ideal comparison to the transgenic mouse model approach carried out using Ins1-/-Ins2f/fPdx1-CreER mice430 wherein following delivery of tamoxifen there was ~99% recombination efficiency, mice developed hyperglycemia two weeks post-tamoxifen, and had reduced circulating insulin four weeks post-tamoxifen. As expected, delivery of AAV Ins1-Cre did not impact body weight  106 (Figure 3-9A) and there was induction of variable hyperglycemia in 3-6 weeks (Figure 3-9B). We performed an IPGTT at 10 days (Figure 3-9C) and 28 days (Figure 3-9D) post-AAV and found severe glucose intolerance by day 28. Analysis of blood collected during the day 28 IPGTT revealed no significant changes in fasting insulin levels, but a blunted glucose stimulated insulin release (Figure 3-9E). Despite hyperglycemia, we were unable to detect significant changes in fasting insulin levels throughout the study (Figure 3-9F). Six weeks post-AAV, we collected pancreas in PFA and immunostained for Cre and found heterogenous Cre immunoreactivity that was discordant with insulin immunoreactivity (Figure 3-9G). We identified many cells in the core of the islet lacking not only insulin immunoreactivity, but also Cre. In addition, we observed cells with bright nuclear Cre but enduring bright insulin immunoreactivity. We next immunostained for insulin and islet amyloid polypeptide (IAPP; Figure 3-9H) and used IAPP as a marker of β-cells for subsequent analyses. Though a majority of β-cells retained some insulin immunoreactivity, there was an obviously fainter insulin immunoreactivity in many cells and an evident bimodal distribution as shown in a violin plot of average intensity of insulin immunoreactivity in IAPP+ cells. By defining a cutoff (red line) of insulin intensity between the bimodal distribution in AAV injected animals there was a significant difference in percent IAPP+ cells below the cutoff in AAV Ins1-Cre injected animals (35.3%) compared to controls (6.3%).  107  Figure 3-9 The AAV Ins1-Cre can be a useful tool for directing β-cell recombination when off-target effects are deemed minimally important. Four hour fasting body weight (a) and blood glucose (b) of adult Ins1-/-Ins2f/f mice that received AAV Ins1-Cre by IP delivery (indicated by red arrow). IPGTTs were performed on day 10 (c) and  108 day 28 (d) relative to AAV injection. (e) Blood collected during the IPGTT on day 28 was assayed for circulating insulin. (f) Fasting insulin levels throughout the study were assessed by ELISA. Data shown as mean±SEM (a-f) with individual traces (b-f). Pancreata collected six weeks post-AAV were immunostained for insulin and Cre (g). Insets show 4x enlargement of representative cells with bright insulin and Cre immunoreactivity, diminished insulin and Cre immunoreactivity, and cells in the core of the islet with neither insulin nor Cre immunoreactivity. (h) Pancreas was immunostained for insulin and IAPP. The intensity of insulin immunoreactivity in IAPP+ cells was quantified, and individual cells for each animal are presented as violin plots with percent of cells below a designated cut-off (red line) shown as mean ± 95% confidence interval. Data analysed by repeated measures 2-way ANOVA (a-f) or Mann-Whitney test (h). n=4-5, scale bars are 100 μm, and insets are enlarged 4x. (** p<0.01, ***, p<0.001)  3.4 Discussion Cre-LoxP technology has been used to great success to study the role of genes of interest in β-cell function, phenotype, and physiological outcomes357. To date, the field has relied on transgenic mouse lines to deliver Cre and inducible CreERs to cells of interest. Many Cre driver lines face caveats of off-target recombination and constitutive expression of Cre may have detrimental effects on β-cell development and function. Additionally, inducible CreER lines are complicated by GH minigene inclusions within transgenes, leaky recombination before tamoxifen delivery, potential β-cell defects, and tamoxifen toxicity. As an alternative, we propose that the AAV could be used as a vector for delivery of Cre to β-cells. AAV is non-pathogenic and simple IP delivery of the eighth serotype can infect many tissues, including the pancreas. We designed an AAV8 carrying Cre under control of a fragment of the rat Ins1 promoter. Use of AAV Ins1-Cre enabled efficient β-cell recombination via the insulin promoter and zero leak temporal control. Additionally, use of the AAV allowed us to avoid Cre driver mouse lines that include a growth-hormone minigene, we did not need to deliver tamoxifen, and there were no deleterious effects of AAV Ins1-Cre on glucose-stimulated insulin secretion nor Ca2+ dynamics in response to glucose and KCl. Contrarily, islets from Ins2-Cre transgenic mice have severe defects in Ca2+ dynamics in response to glucose stimulation426. Though higher AAV Ins1-Cre doses may cause some defects in β-cell function based on  109 observations of impaired glucose tolerance, excellent recombination efficiency using 1012 VGP was feasible and avoided detectable β-cell defects. As packaging limitations precluded the use of a long insulin promoter like the 8.5kb used in the Ins1-Cre/ERT transgene358, when AAV Ins1-Cre was delivered IP there was off-target recombination in the hypothalamus, specifically the arcuate nucleus and dorsomedial nucleus of the hypothalamus. Though Ins1 is β-cell specific whereas Ins2 is also be expressed in neurons of the hypothalamus431, the short promoter used here may contribute to promiscuity in the hypothalamus. Delivery of AAV via the pancreatic duct resulted in undetectable hypothalamic recombination partially because a lower dose (1011 VGP) could be used to achieve β-cell recombination; however, the same dose produced hypothalamic recombination when delivered IP, suggesting that the magnitude of systemic spreading is lower when delivered via the pancreatic duct. Surprisingly, both IP and ID delivery resulted in abundant recombination in the exocrine pancreas, other islet cells, and liver. Though the infection of AAV8 appears highest in the pancreas264 and certainly use of an insulin promoter would lead to the highest expression in β-cells, recombination is not a linear assessment of expression but rather a binary outcome. As a comparison, when the AAV carried GFP under control of a rat Ins1 promoter (Figure 3-6G), any leaky GFP in hepatocytes or exocrine pancreas remained undetectable and orders of magnitude fainter than in β-cells. As there is notable heterogeneity in expression level of Pdx1 in the mouse pancreas320, it is possible that acinar cells or non-β islet cells with higher Pdx1 were more likely to activate the rat Ins1 promoter fragment and undergo a recombination event, despite such cells not expressing detectable GFP following transduction with AAV Ins1-GFP. Similarly, others have observed activity of the insulin promoter in liver using a fragment of the insulin promoter432. With less sensitive, non-binary, histological assays, there is even detectable proinsulin in many tissues including liver and adipose in models of diabetes433.  110 Generation of a suitable cohort of Cre driver mice with a floxed target gene can take over a year, require dozens of cages, hundreds of genotyping experiments, and hundreds of hours of labor. AAV Ins1-Cre can be readily manufactured for a few thousand dollars, the cost to treat each mouse can be <$50 (1011 VGP ID), and suitable mice can be ordered directly from a rodent breeding facility or can be bred within the lab from a handful of breeders. Additionally, although current AAV manufacturing techniques that require adherent cell cultures and serum supplemental media are costly241, some recent advances permit serum free suspension cultures that could improve production efficiency and perhaps lower cost242, 243. That being said, we acknowledge that ID delivery is technically challenging and an invasive procedure thus presenting unique challenges. Another alternative to using a Cre driver mouse line may be using adenoviral constructs, though we are unaware of any such attempts. Adenoviruses are inexpensive and relatively simple, but there has been limited efficiency at targeting the endocrine pancreas and the adenovirus causes a significant leukocyte response in the pancreas434, 435, a major caveat when attempting to study a disease with an autoimmune component. Other technologies including lipid nanoparticles are being investigated for delivering genes to pancreatic cancer436 but this approach is new and there may be challenges achieving adequate efficiency of delivery. Though faced with the caveat of off-target recombination in the liver, exocrine pancreas, and hypothalamus, the AAV Ins1-Cre approach requires no tamoxifen, avoids use of Cre transgenes with a growth hormone minigene, avoids developmental and functional defects of constitutive Cre expression, and precludes early recombination events prior to delivery. With an appreciation of both the benefits and caveats of using AAV Ins1-Cre, we demonstrate the utility of this tool by inducing hyperglycemia in Ins1-/-Ins2f/f mice. Likely off-target recombination was not concerning because there are no other tissues that could reasonably cause hyperglycemia following loss of the insulin gene. Recombination events to delete Ins2 in tissues like the liver is  111 irrelevant because they do not express insulin. Further, unlike the alternative approach using an inducible Pdx-CreER430 that would require extensive breeding and genotyping, we could generate these mice in a single round of breeding and proceed with our experiment in <3 months. As expected, Ins1-/-Ins2f/f mice injected with AAV Ins1-Cre developed hyperglycemia and had reduced glucose-stimulated insulin secretion. Similar to the previous study on the Ins1-/-Ins2f/fmTmGPdxCreER model430, we saw only a modest reduction in insulin immunoreactivity despite induction of diabetes and reduced glucose-stimulated circulating insulin levels. Given that a >90% pancreatectomy is required to induce diabetes in rodents437, the loss of functional β-cell mass must be greater than the observed ~1/3 reduction in insulin immunoreactive β-cells to cause hyperglycemia. The notion that recombination events and the resulting loss of a functional β-cell does not align with changes in insulin immunoreactivity is highlighted by the surprising presence of cells with bright nuclear Cre and cytoplasmic insulin. Barring unlikely explanations like insulin endocytosis438, presumably such a cell appearance should exist only for a few days as after recombination, stored insulin should be secreted in response to hyperglycemia. Enduring insulin immunoreactivity can be partially attributed to abundant insulin storage and high stability of insulin mRNA430, but assuming recombination occurred near the time of AAV injection, this seems inadequate to explain persistent near normal insulin storage (by comparing brightness of immunoreactivity to control cells) six weeks after recombination during a period of hyperglycemia. For example, mouse islet insulin secretion is ~3% of content per hour at 10 mM glucose439 and though highly stable, insulin mRNA has a half life of only one day at low glucose and three days in settings of hyperglycemia440. Thus, dedifferentiation leading to proliferation430 and associated loss of glucose-stimulated insulin secretion may better explain the presence of cells with likely recombination events but enduring insulin immunoreactivity for 6 weeks. Triple-omics assessment of islets isolated 6 days post-tamoxifen from Ins1-/-Ins2f/fmTmGPdxCreER mice430 found down-regulation of secretory pathways like  112 “Golgi vesicle transport”. At the same time-point, β-cells were GFP+ and had insulin immunoreactivity comparable to controls, suggesting a loss of secretory pathways even prior to loss of all secretory granules. Additionally, others have reported enduring insulin immunoreactivity in models of β-cell dedifferentiation35. The AAV Ins1-Cre is an ideal tool to generate this model of diabetes and repeat investigations using the AAV Ins1-Cre to clarify these observations are worthwhile. We propose that the AAV Ins1-Cre approach described here could be used as a tool for expedited generation of inducible zero leak knockout of β-cell specific genes. This approach avoids inclusion of transgenes and delivery of tamoxifen but faces the caveat of off-target recombination in the hypothalamus, liver, and exocrine and endocrine pancreas. Though these caveats may be problematic for some study designs, when deleting genes whose predominant function is in β-cells, the approach is faster and comparably efficient to Cre driver line models. We highlight this point by inducing diabetes in Ins1-/-Ins2f/f mice. The AAV Ins1-Cre will be useful for studying in vivo β-cell function and the impact of gain or loss of function of β-cell specific genes.   113 Chapter 4: Insulin deficient β-cells retain a prohormone processing defect that impairs an AAV mediated insulin gene therapy 4.1 Background Despite accounting for only 1-6% of all diabetes, monogenic diabetes still affects millions worldwide441. Mutations in the insulin gene (INS) most commonly cause permanent neonatal diabetes (PND) but can also cause monogenic diabetes of the young (MODY)19. In 1997 Duvielle and colleagues developed a mouse model of PND with deletion of both non-allelic insulin genes (Ins1-/-Ins2-/-)367 and we characterized insulin deficient β-cells of these mice in detail (Chapter 2)442. Ins1-/-Ins2-/- mice have dedifferentiated IAPP+ β-cells that lack mature β-cell factors including MAFA and NKX6.1 at birth. Shortly after birth, Ins1-/-Ins2-/- pups become hyperglycemic and β-cells lose immunoreactivity of the prohormone processing enzyme PC2. Replacement of insulin by injection was not sufficient for the maturation of insulin deficient β-cells. Long-term replacement of insulin by islet transplantation better supported endogenous β-cells based on immunoreactivity for mature β-cell factors including MAFA442. Assessing the viability of a gene therapy approach to treat this model of PND can provide insight into the viability of a gene therapy for patients with diabetes caused by mutation in the INS gene.  In the current study we aimed to determine if an eighth serotype AAV (AAV8) carrying an insulin open reading frame regulated by an insulin promoter (Ins1) could restore insulin production to Ins1-/-Ins2-/- β-cells and thereby resolve diabetes. Though AAVs could deliver either human (INS) or mouse (Ins1) insulin genes to β-cells, defects in prohormone processing endured in Ins1-/-Ins2-/- β-cells that prevented their ability to form normal dense core secretory granules. We aimed to determine if developmental defects were the cause of impaired prohormone processing in AAV infected Ins1-/-Ins2-/- β-cells by generating adult inducible insulin deficient mice. Co-delivery of the AAV Ins1-Cre (studied in detail in Chapter 3) with a mouse  114 insulin AAV (AAV Ins1-Ins1) failed to prevent diabetes onset compared to AAV Ins1-Cre alone. Taken together, these findings support the capability of the AAV8 to deliver the insulin gene to pancreatic β-cells, but efficacy at reversing neonatal diabetes is limited by an enduring prohormone processing defect in insulin knockout β-cells.  4.2 Materials and methods Animal models and insulin therapy All experiments were approved by the UBC Animal Care Committee and carried out in accordance with the Canadian Council on Animal Care Guidelines. C57Bl/6J Ins1+/+Ins2+/+ mice were ordered from Jackson Laboratories (Bar Harbour, ME, USA). Ins1-/-Ins2+/- were generated in the lab of Dr. J. Jami (Institut Cochin, Paris, France), and Ins2f/f mice were generated in the lab of Dr. Massimo Trucco424 (University of Pittsburgh, Pittsburgh, PA) and both genotypes were generously supplied indirectly by the lab of Dr. J. Johnson (University of British Columbia, Vancouver, Canada). mTmG mice were generated in the lab of Dr. Liqun Luo421 (Stanford University, Stanford, Ca), Pdx1-CreERT mice were generated in the lab of Dr. Douglas Melton443 (Harvard University, Cambridge, MA), and mTmG; Pdx1-CreERT mice were generously shared by Dr. Francis Lynn (University of British Columbia, Vancouver, Canada). Unless noted, we present findings from a mix of male and female animals and did not detect significant differences between the sexes (data not shown). Notably, studying both sexes was done to maximize sample sizes of Ins1-/-Ins2-/- mice given substantial technical challenges keeping them alive with insulin therapy. Animals were given ad libitum access to a standard chow diet (2918, Harlan Laboratories, Madison, WI, USA) and housed with a 12-h light/dark cycle. Generation of Ins1-/-Ins2-/- mice and insulin therapy was performed as previously described442. Briefly, Ins1-/-Ins2+/- mice were bred to acquire Ins1-/-Ins2-/- and Ins1-/-Ins2+/- mice.  115 We initiated insulin therapy immediately in suspected insulin knockouts (reduced body weight and glycosuria) and confirmed genotype by qPCR at an older age. Animals received twice daily subcutaneous injections of approximately 0.1U Insulin Glargine (Lantus®; diluted to 5 U/mL in F-10 media, Sigma-Aldrich, St. Louis, MO) until fifteen days of age when they were either treated with islet transplantation into the anterior chamber of the eye372 or were treated with insulin AAV and insulin therapy continued as needed. Adeno-associated virus production The AAV Ins1-GFP plasmid was generated from the dsAAV mINS2p-EGFP plasmid, kindly provided by Dr. Paul Robbins264. To generate the Ins1-GFP plasmid, the 1.1kb mINS2p with 5’ UTR was excised with BamHI and AgeI and replaced with 410-bp of the rat Ins1 promoter with 5’ UTR (Ins1; primers: Ins1-F: 5’-CACTGGATCCTGAGCTAAGAATCCAGCTATC AATAG AAACT; Ins1-R: 5’-CACACAACCCCGTGTTGGAACAATGACCTG GAAGATAG). Next, to generate the Ins1-Cre plasmid the EGFP open reading frame (ORF) was excised with AgeI and NotI and the 4.9kb vector fragment was gel-purified. The Cre ORF was generated by PCR amplification from an Addgene CMV-Cre vector (pBS185) with a Kozak sequence and XmaI site added upstream and a NotI site downstream (primers: CreORF-F: 5’-CACACGCCCGGGGCC GCCACCATGTCCAATTTACTGACCGT ACACCAA; CreORF-R: 5’-CACACGCGGCCGCCTAA TCGCCATCTTCCAGCAGGC). The PCR product was digested with NotI and XmaI and ligated into the 4.9kb vector to yield the final dsV Ins1-Cre plasmid. To generate the Ins1-INS plasmid, the human insulin ORF was generated by qPCR from human islet cDNA (primers: INS-ORF-F: 5’- CACACAACCGGTCGCACCATGGCCCTGTGGATGC; INS-ORF-R: 5’- CACACAGCGGCC GCCTAGTTGCAGTAGTTCTCCAGC) and was digested with Age and Not1 before ligation into the digested vector. The Ins1-Cre and Ins1-GFP plasmids were sent to the Children’s Hospital of Philadelphia for manufacturing of high titer dsAAV and the Ins1-INS plasmid was sent to SAB Technology Inc. (Philadelphia, PA). To generate the Ins1-Ins1 plasmid, the mouse Ins1 ORF  116 was generated by qPCR from mouse islet cDNA (primers: mINS-ORF-F: 5’- CACACACCATGGCCC TGTTGGTGCACTTC: mINS-ORF-R: 5’- CACACAGCGGC CGCTTAGTTGCAGTAGTTCTCCA GCTGGTAG) and was digested with Nco1 and Not1 before ligation into a Nco1 and Not1 digested Ins1-INS vector. The Ins1-Ins1 plasmid was sent to Vector BioLabs (Malvern, PA) for manufacturing of high titer dsAAV. We delivered AAV by IP injection. Maps of the sequences between the two ITRs of each plasmid are shown in Appendix B. Islet isolation and transplantation Ins1-/-Ins2+/- mice (genotype confirmed by qPCR) were euthanized and pancreatic islets were isolated by collagenase digestion (1000U/mL type XI collagenase, Sigma-Aldrich, St. Louis, MO)425. After intraductal infusion of collagenase, the pancreas was then excised and digested for 12-15 mins at 37oC in 3 mL of collagenase. Islets were then handpicked three times in medium (Hams F10, 7.5% fetal bovine serum, and penicillin/streptomycin, Sigma-Aldrich, St. Louis, MO) to increase purity to over 90%. After overnight culture, islets were washed in sterile PBS prior to transplantation into the anterior chamber of the eye (approximately 100-150 islets) as previously described372. To access the anterior chamber of the eye, we punctured the cornea with a 27G needle and a micropipette loaded with islets was used to deposit islets (MXL3-BP-IND-200; Origio MidAtlantic Devices, Mt Laurel, NJ, USA). Animals were treated with perioperative Isoptears (Alcon Canada, Mississauga, ON, Canada) with 0.3% wt/vol gentamycin to prevent infection. To remove the graft, we enucleated the eye with the graft as previously described368. Physiological tests Blood glucose monitoring was done by random-fed measurements or after a four hour fast in the morning (0900-1300) using a One Touch Ultra glucometer (Life Scan Inc., Burnaby, Canada). For intraperitoneal glucose tolerance test (IPGTT) and fast/refeed (F/RF), mice were  117 fasted four hours (0900-1300) before glucose injection (1-2 g glucose/kg body weight IP) or return of food. Blood was sampled by the saphenous vein. For insulin tolerance tests mice were administered Novolin® at a dose of 0.8 U/kg body weight IP. Assays We assessed circulating plasma levels of human C-peptide (Mercodia Cat# 10-1141-01, Uppsala, Sweden), mouse C-peptide (Alpco Cat# 80-CPTMS-E01, Salem, NH), and proinsulin (human: Mercodia Cat# 10-1118-01, Uppsala, Sweden, mouse: Mercodia Cat# 10-1232-01, Uppsala, Sweden) by commercially available enzyme linked immunosorbent assay (ELISA). A list of relevant cross-reactivities of each assay is detailed in Table 4-1. Assessment of  Table 4-1 Relevant reported cross reactivity of commercial ELISAs for insulin, C-peptide, proinsulin, and proinsulin intermediates (%).  Mouse C-peptide (Alpco, 80-CPTMS-E01) Rat/mouse proinsulin (Mercodia, 10-1232-01) Human C-peptide (Mercodia 10-1141-01) Human proinsulin (Mercodia, 10-1118-01)  Target  Reactivity  Target  Reactivity  Target  Reactivity  Target  Reactivity C-peptide 1 100 Mouse insulin ND Insulin <0.0006 Insulin <0.03 C-peptide 2 151 Mouse C-peptide ND Proinsulin 5 C-peptide <0.006 Proinsulin 1 2 Human C-peptide ND Des-31,32 3 Des-31,32 95 Proinsulin 2 ND Human proinsulin 5 Split-32,33 2 Split-32,33 95 Human proinsulin ND   Des-64,65 74 Des-64,65 84     Split-65,66 10 Split-65,66 90        Mouse proinsulin 1  3       Mouse proinsulin 2 16  hemoglobin A1c (HbA1c) in whole blood was performed using a Siemens DCA 200 Vantage Analyzer (Siemens Healthcare Diagnostics, Tarrytown, NY). For detection of insulin autoantibodies, plasma was sent to the Barbara Davis Center (University of Colorado, Aurora, CO) for a micro insulin autoantibody radiobinding assay. Human insulin was used as the antigen  118 thus allowing detection of mouse IgG specific for epitopes of human insulin. The sequences of human and mouse insulin are almost identical (Figure 1-6) thus allowing detection of insulin autoantibodies in a mouse exposed to mouse insulin, like the non-obese diabetes (NOD) model. Immunohistofluorescence Pancreata were fixed in 4% paraformaldehyde (PFA) overnight before storage in 70% ethanol until paraffin-embedding and sectioning (5 µm thickness; Wax-It Histology Services, Vancouver, Canada). Immunofluorescent staining was performed as previously described442. All antibodies, including RRID, are listed in Table 4-2.  Transmission electron microscopy Samples that had been fixed in PFA and embedded in paraffin were sent to the electron microscopy facility at McMaster University (Hamilton, Canada). Samples were deparaffinized and re-fixed in glutaraldehyde and osmium tetroxide and embedded in Spurr’s epoxy resin. Grids were visualized using a Tecnai G2 Spirit electron microscope (FEI Co., Eindhoven, The Netherlands) and images were captured using a 4Kx4K FEI Eagle HS CCD camera and we present representative images. Titering AAV by quantitative PCR  We performed quantitative PCR on manufactured AAV and plasmids of known concentration to determine AAV titer. We used stocks of the Ins1-Ins1 plasmid sent to VectorBiolabs for manufacturing AAV Ins1-Ins1 and a stock of Ins1-GFP plasmid. To liberate AAV DNA we heat treated at 95oC for 10 minutes. Next, to generate comparable template structure, we digested all plasmids and AAV genomes with Not1 and BamH1 to remove complex inverted terminal repeats from AAVs and generate linear fragments from circular plasmids. We then diluted all DNA and prepared amplification reactions with SsoFast SYBR Green Supermix (BioRad, Mississauga, Canada) starting from known numbers of plasmid templates (106, 107, 108, and 109 templates) and reported copies AAV templates. We performed  119 qPCR for Ins1 (Ins1-F: 5’- AAGCGTGGCATTGTGGAT; Ins1-R: 5’- GCTTAGTTGCAGTAGTTC TCCAGCTGGTAG) and GFP (GFP-F: 5’- CACCATCTTCTTCAAGGACGA; GFP-R: 5’- GTTGT AGTTGTACTCCAGCTTGT). We interpolated AAV titer by fitting Ct values for the AAV onto the standard curve generated using the four starting copies of plasmid. Statistical analysis Given use of variable and sometimes small sample sizes, data were subject to the Shapiro-Wilk normality test and when all groups passed the test for normality we analyzed by parametric test. We used one-way ANOVA with Tukey test for multiple comparisons or two-way ANOVA with Bonferonni’s post-hoc testing. When any of the groups failed the test for normality, we used the non-parametric Mann-Whitney U test or Kruskal-Wallis test with Dunn’s multiple comparisons test. Statistical analysis was performed using GraphPad Prism 8.00 (La Jolla, CA) with significance set at p < 0.05. Summary of in vivo study design  Outlines of all in vivo studies performed in Chapter 4 are presented in Appendix C.  Table 4-2 Primary antibodies used for immunofluorescent staining  Peptide/protein target Antigen Sequence Name of Antibody Manufacturer, catalog #, and/or name of individual providing the antibody Species raised in; monoclonal or polyclonal Dilution used RRID Aldehyde dehydrogenase family member 1A3 unknown Rabbit anti-ALDH1A3 Novus Biologicals NBP2-15339 Rabbit Polyclonal 1 to 100 AB_2665496 Islet amyloid polypeptide unknown Rabbit anti-IAPP AbCam ab15125 Rabbit Polyclonal 1 to 50 AB_2295631 Beta-galactosidase Whole beta-galactosidase Rabbit anti-βGAL Thermo Scientific A11132 Rabbit Polyclonal 1 to 100 AB_221539 Carboxypeptidase-E Human CPE (AA49-200) Mouse anti-CPE BD Transduction 610758 Mouse Monoclonal 1 to 100 AB_398081 Glucagon unknown Mouse anti-GCG Sigma G 2654 Mouse Monoclonal 1 to 1000 AB_259852 Glucose transporter 2 First extracellular loop of Glut2 Rabbit anti-GLUT2 Millipore 07-1402 Rabbit Polyclonal 1 to 500 AB_1587076 Insulin unknown Rabbit anti-INS Cell Signaling C27C9 Rabbit Monoclonal 1 to 200 AB_2126503  120 Peptide/protein target Antigen Sequence Name of Antibody Manufacturer, catalog #, and/or name of individual providing the antibody Species raised in; monoclonal or polyclonal Dilution used RRID Insulin Residues surrounding Val36 of human insulin Mouse anti-INS Cell Signaling L6B10 Mouse Monoclonal 1 to 250 AB_10949314 v-myc avian myolocytomatosis viral oncogene lung carcinoma derived AA 105-154 of human l-myc Rabbit anti-L-MYC AbCam Ab28739 Rabbit Polyclonal 1 to 100 AB_2148730 V-maf muscoloapo-neurotic fibrosarcoma oncogene homolog A unknown Rabbit anti-MAFA Betalogics (Johnson & Johnson) LP9872 Rabbit Polyclonal 1 to 1000 AB_2665528 homeobox protein NANOG Mouse nanog Rabbit anti-NANOG AbCam Ab80892 Rabbit Polyclonal 1 to 100 AB_2150114 homeodomain transcription factor 6.1 Human Nkx6.1 Goat anti-NKX6.1 R and D Systems AF5857 Goat Polyclonal 1 to 20 AB_1857045 homeodomain transcription factor 2.2 Nkx2.2-GST fusion protein from E. coli Mouse anti-NKX2.2 DSHB 74.5A5 Mouse Monoclonal 1 to 100 AB_531794 Human C-peptide Full length human C-peptide GP anti- C Peptide AbCam Ab30477 Guinea pig Polyclonal 1 to 100 Ab_726924 Paired box 6 C-terminus of mouse PAX6 Rabbit anti-PAX6 Covance PRB-278P Rabbit Polyclonal 1 to 250 AB_2313780 Prohormone convertase 1/3 unknown Rabbit anti-PC1/3 Lakshmi Devi gift Rabbit Polyclonal 1 to 500 AB_2665530 Prohormone convertase 2 E622-N638 of mouse PC2 Rabbit anti-PC2 Thermo Fisher Scientific PA1-058 Rabbit Polyclonal 1 to 500 AB_2158593  proIAPP (C-terminus) C-terminal fragment of proIAPP Rabbit anti- N-terminus proIAPP Dr. C Bruce Verchere N/A Rabbit Polyclonal 1 to 200 Pending proIAPP (N-terminus) N-terminal fragment of proIAPP Rabbit anti C-terminus proIAPP Dr. C Bruce Verchere N/A Rabbit Polyclonal 1 to 200 Pending Proinsulin (BC junction)  Recombinant human proinsulin Rabbit anti-proinsulin(BC) DSHB GA-9A8 Mouse Monoclonal 1 to 250 AB_532383 Proinsulin (CA junction) Unkncown Rabbit anti-proinsulin(CA) Alpco 82-PIN-AC-mAb Mouse Monoclonal 1 to 1000 AB_2783531     121 4.3 Results AAV Ins1-INS can produce dose-dependent infection of mouse β-cells To validate the infectivity of AAV Ins1-INS, we injected wild-type Ins1+/+Ins2+/+ mice with variable doses of AAV or saline as a control. Pancreas was collected 42 days post-AAV and we immunostained for islet amyloid polypeptide (IAPP) for use as a marker of β-cells and assessed infection rate by immunostaining for human C-peptide. We used human pancreas as a positive control (Figure 4-1A). There was a dose dependent significant increase in the proportion of infected β-cells (Figure 4-1B). Incomplete colocalization of IAPP and C-peptide in human pancreas is partially attributable to IAPP immunoreactivity in δ-cells373, 444. We collected plasma during a glucose challenge on day 20 relative to AAV delivery (2 g glucose/kg body weight) and assayed for mouse C-peptide and human C-peptide. For many experiments in this thesis we used commercially available ELISAs to assay for C-peptide and proinsulin. These kits have published cross-reactivities for insulin, C-peptide, proinsulin, and/or proinsulin intermediates (Table 4-1). There were no changes to circulating mouse C-peptide (Figure 4-1C) and human C-peptide immunoreactivity was increased in a dose dependent manner (Figure 4-1D). IP glucose tolerance tests (2 g glucose/kg body weight) were performed on day -1 (Figure 4-1E), day 13 (Figure 4-1F), and day 40 (Figure 4-1G) relative to the day of AAV delivery. There were no differences between groups at baseline but there was a transient improvement in glucose tolerance in the two higher dosed groups (Figure 4-1F). The highest infection rate was with a dose of 1012 viral genome particles (VGP) or approximately 5x1010 VGP/g body weight (animals were approximately 20 g at the time of AAV injection). This dose produced an average β-cell infection rate of 68%, transiently improved glucose tolerance, and led to significant increases in circulating human C-peptide to approximately 2-3 ng/mL. Based on past work in the Kieffer lab, circulating human C-peptide levels of 1-2 ng/mL in mice treated with sources of human insulin (stem cell derived insulin producing cells) is sufficient to lower blood glucose in diabetic mice122.  122  Figure 4-1 β-cell infection by AAV Ins1-INS. Immunofluorescence for human C-peptide in pancreas collected six weeks after 8-week-old wild-type mice were treated with AAV Ins1-INS. Human pancreas is used as positive control (A; representative images of n = 2-3, scale bar is 100 μm). Quantification of the proportion of IAPP+ cells immunoreactive for human C-peptide (B). Plasma collected during an intraperitoneal glucose tolerance test (IPGTT; 2 g glucose/kg body weight) on day 20 relative to AAV Ins1-INS injection was assayed for mouse C-peptide (C) and human C-peptide (D). IPGTTs on days -1 (E), 13 (F), and 40 (G) relative to AAV Ins1-INS treatment. Data presented on box and whisker plots (B), individual traces (C-D), or mean ± SEM (E-G). All groups compared by one-way ANOVA (B) or repeated measures two-way ANOVA (C-G). * p < 0.05, ** p < 0.01, *** p < 0.001. 4 x 1011VGPDAPI HUMAN C-PEPTIDE  IAPPSaline 1 x 1011VGP 1 x 1012VGP Human pancreas020406080100Human C-peptide+IAPP+ cellsSaline 1x1011VGP 4x1011VGP 1x1012VGP Human Pancreas0 6001234TimeHumanC-Peptide(ng/ml)7 15********* ******* *0 60051015TimeMouse C-Peptide(ng/ml)7 150 30 60 900102030TimeBloodGlucose(mM)15 0 30 60 900TimeBloodGlucose(mM)15 0 30 60 900102030TimeBloodGlucose(mM)15Day -1 Day 13 Day 40AB C DE F G(%)102030************ 123 AAV Ins1-INS transiently reverses diabetes in Ins1-/-Ins2-/- mice We generated a cohort of Ins1-/-Ins2-/- pups and determined if treatment with AAV Ins1-INS could reverse diabetes. We monitored Ins1-/-Ins2+/- littermates and wild-type Ins1+/+Ins2+/+ mice as controls. Body weight (Figure 4-2A) and blood glucose (Figure 4-2B) were monitored from birth for seven weeks. Ins1-/-Ins2-/- pups were given twice or three times daily insulin injections (Lantus®) until 15 days of age when groups were injected with either 5x1010 VGP AAV Ins1-INS per gram body weight or saline (shown by arrows on Figures 4-2A and B). Ins1-/-Ins2-/- mice became insulin independent one week after AAV treatment (Figure 4-2C) but were glucose intolerant at 28 days of age during an IP glucose tolerance test (1g glucose/kg body weight; Figure 4-2D). Based on highly variable glucose tolerance (Figure 4-2D) the 21 Ins1-/-Ins2-/- mice were split into 3 groups – the group with the worst glucose tolerance (AAV/- n = 11) and the remaining mice with better glucose tolerance were split into two groups (AAV/AAV n = 5 and AAV/Saline n = 5). The AAV/AAV group was reinjected with 5x1010 VGP AAV Ins1-INS per gram body weight at 35 days of age and the AAV/Saline group was injected with saline (shown by arrows on Figures 4-2A and B). We assayed plasma collected during the IPGTT at 28 days of age for mouse and human C-peptide. The area under the curve of glucose during the IP glucose tolerance test at 28 days of age was inversely correlated with levels of fasting plasma human C-peptide, b = -353.8, SE = 142.3, p < 0.05, CI95 [-56.03, -651.6] (Figure 4-2E). Ins1-/-Ins2-/- mice had no circulating mouse C-peptide and AAV treated Ins1-/-Ins2+/- littermate controls and AAV treated Ins1+/+Ins2+/+ wild-type controls did not have significantly different levels of mouse C-peptide than PBS treated Ins1-/-Ins2+/- or Ins1+/+Ins2+/+ controls respectively (Figure 4-2F). AAV treated littermates and controls had low levels of circulating human C-peptide and AAV Ins1-INS treated animals had abundant human C-peptide (Figure 4-2G). Despite retreatment at 5 weeks of age (AAV/AAV group) and severe relapse to diabetes at ~4-5 weeks of age, all groups of Ins1-/-Ins2-/- animals had comparable levels of human C-peptide at 4, 6, and   124  Figure 4-2 Transient remission of diabetes in Ins1-/-Ins2-/- mice after treatment with AAV Ins1-INS Random fed body weight (A) and blood glucose (B) of Ins1-/-Ins2-/-, Ins1-/-Ins2+/- littermates, and Ins1+/+Ins2+/+ control mice monitored from birth. At 15 days of age, all groups were treated with  125 AAV or PBS control (shown by arrow). Insulin requirements (Lantus®) of Ins1-/-Ins2-/- animals from birth to 4 weeks of age (C). Highly variable glucose tolerance (IPGTT, 1 g glucose/kg body weight) of Ins1-/-Ins2-/- mice, unlike Ins1-/-Ins2+/- littermates and Ins1+/+Ins2+/+ controls (D; day 14-16 relative to AAV treatment). Levels of human C-peptide correlated with area under the curve of the glucose tolerance test (E) and Ins1-/-Ins2-/- animals were grouped as non-responders with the worst glucose tolerance (AAV/-) and responders with better glucose tolerance (responders were retreated at five weeks of age with AAV or Saline – AAV/AAV or AAV/saline groups respectively). Plasma collected during the IPGTT at 28 days of age was assayed for mouse C-peptide (F) and human C-peptide (G; 0 and 30min relative glucose injection; letters a, b, and c represent statistical differences between groups, p < 0.05). Human C-peptide levels in fasted plasma collected at 4 weeks (replotted from panel G), 6 weeks, and 8 weeks of age (H). Plasma collected at 7 weeks of age alongside plasma from known autoimmune diabetic non-obese diabetic (NOD) mice (NODHIGH had severe diabetes and NODLOW had mild diabetes) was assayed for insulin autoantibodies (I). Plasma collected after a 4 hour fast was used in an assay specific for proinsulin and proinsulin intermediates (J; 18/21 Ins1-/-Ins2-/- mice have detectable proinsulin). HbA1c of blood collected at 8 weeks of age (K). Pancreas was immunostained for human C-peptide (green), mouse C-peptide (blue), and either an unprocessed C-A junction proinsulin or and unprocessed B-C junction proinsulin16. Cytoplasmic immunoreactivity suggests there is impaired processing at the B-C junction in AAV treated Ins1-/-Ins2-/- mice (L). Representative images of n = 4-5, the scale bar is 100 μm, and insets are enlarged 4x. Data presented as mean ± SEM (A-D) or box and whisker plots (F-K). Individual traces or points shown on D-E. All groups were compared by repeated measures 2-way ANOVA (F-H) or one-way ANOVA (I-K) * p < 0.05, ** p < 0.01.  8 weeks of age (Figure 4-2H). As a potential explanation for discordant observations of abundant human C-peptide and severe hyperglycemia, we assayed plasma for neutralizing insulin autoantibodies (Figure 4-2I). There was no elevation of insulin autoantibodies in AAV Ins1-INS treated animals. We assayed plasma from two non-obese diabetic (NOD) mice with autoantibody levels at the high end of expected (NODHIGH) and low end (NODLOW) of expected titer for the NOD model of autoimmune diabetes as positive controls. As an alternative explanation, we considered the published cross-reactivity of the human C-peptide assay used in Figure 4-2H (see Table 4-1; proinsulin: 2%, des-31,32 proinsulin: 3%, split-32,33 proinsulin: 2% des-64,65 proinsulin: 74%, and split-65,66 proinsulin: 10%). We used an assay specific for proinsulin and the proinsulin intermediates (Mercodia, 10-1118-01) and found that there was circulating proinsulin in most AAV treated Ins1-/-Ins2-/- mice (18/21 mice; 4-2J). Proinsulin has some bioactivity (Table 4-3) that may lower blood glucose in highly insulin sensitive pups but  126 may not be adequate in larger mice reaching adulthood. HbA1c was elevated in Ins1-/-Ins2-/- animals at study termination (Figure 4-2K). To support evidence of impaired proinsulin processing, we performed immunohistofluorescence with antibodies specific for unprocessed proinsulin. The “C-A junction” antibody binds an intact C-A junction found in intact proinsulin, des-31,32 proinsulin, and split-32,33 proinsulin. The “B-C junction” antibody binds an intact B-C junction found in intact proinsulin, des-64,65 proinsulin, and split-65,66 proinsulin16. In normal β-cells there was an expected perinuclear pattern of immunoreactivity using both antibodies (Ins1+/+Ins2+/+ and Ins1-/-Ins2+/- pancreas; Figure 4-2L). We found evidence of impaired B-C site processing of AAV Ins1-INS infected Ins1-/-Ins2-/- β-cells based on pancytoplasmic immunoreactivity using the B-C junction antibody. Successful β-cell infection and normal immunoreactivity for proinsulin processing enzymes in AAV Ins1-INS injected Ins1-/-Ins2-/- mice At 8-11 weeks of age AAV Ins1-INS injected mice and saline injected controls were sacrificed and pancreas was collected in 4% PFA. We immunostained for human C-peptide and IAPP (Figure 4-3A) and used IAPP as a marker for β-cells in subsequent analyses. In all Ins1-/-Ins2-/- groups (AAV/-, AAV/AAV, and AAV/Saline) and the Ins1-/-Ins2+/- AAV group, approximately 5-15% of β-cells were immunoreactive for human C-peptide. To investigate the elevated proinsulin in circulation (Figure 4-2J) and evidence for impaired proinsulin processing in AAV Ins1-INS infected Ins1-/-Ins2-/- β-cells (Figure 4-2L), we immunostained for the essential prohormone processing enzymes PC1/3 and PC2. We also immunostained for the PC2 cofactor neuroendocrine protein 7B2 and the enzyme carboxypeptidase E (CPE) that removes exposed C-terminal dibasic amino acids after internal cleavage by PC1/3 or PC2 (Figure 4-3B). There was normal immunoreactivity for PC1/3 and PC2 in Ins1-/-Ins2-/- islets and present but reduced immunoreactivity for 7B2 and CPE. We quantified the proportion of PC1/3 or PC2 positive INS+ cells and the fluorescent intensity of PC1/3 and PC2 signal (Figure 4-3C). There were normal  127 rates and intensities of immunoreactivity for PC1/3 and PC2 in Ins1-/-Ins2-/- infected β-cells. Based on previous observations that most Ins1-/-Ins2-/- β-cells stop expressing β-galactosidase off of the Ins2 knock in after birth367, we immunostained pancreas for human C-peptide and β-galactosidase (Figured 4-3D). If the observed low rate of human C-peptide immunoreactive cells (Figure 4-3A) was because insulin promoters in many AAV Ins1-INS infected β-cells were silenced (like the β-galactosidase knock in in the Ins2 locus) rather than because of low rates of infection, then most human C-peptide positive cells, would also be positive for βGAL. This was not the case with only occasional colocalization of human C-peptide and βGAL. We also immmunostained for the mature β-cell factors GLUT2 and NKX6.1, and there was normal nuclear NKX6.1 but a lack of membranous GLUT2 in hyperglycemic AAV treated Ins1-/-Ins2-/- mice (Figure 4-3D).   Table 4-3 Summary of studies investigating the relative bioactivities of insulin and proinsulin  Relative receptor binding affinity of human insulin, proinsulin, and proinsulin intermediates based on the ED50 of displacing insulin in three different cell types. Based on data from 445.  Insulin molecule Human lymphocytes  Isolated rat adipocytes Purified rat liver membranes Insulin 100 100 100 Proinsulin 1.4 0.56 1.1 Split-32,33 proinsulin 7 3 3 Des-31,32 proinsulin 8 5 2 Split-65,66 proinsulin 22 14 12 Des-64,65 proinsulin 20 12 10   Relative in vitro bioactivity of bovine insulin and proinsulin for the conversion of glucose to CO2 in isolated rat adipocyte cells. Based on data from 446. Insulin 100   Proinsulin 11   Split-32,33 proinsulin 15   split-65,66 proinsulin 30        Relative in vitro bioactivity of diarginyl insulin for the conversion of glucose to CO2 in isolated rat adipocyte cells. Based on data from 446. Porcine insulin 100 Porcine Diarginyl insulin 38  128  Figure 4-3 Low infection rate and normal immunoreactivity for PC1/3 and PC2 in β-cells of AAV Ins1-INS injected Ins1-/-Ins2-/- mice Pancreas was immunostained for human C-peptide and IAPP (A). Representative image of n = 4-5. Quantification of the proportion of human C-peptide immunoreactive IAPP+ β-cells shown to the right. Immunostaining of pancreas from Ins1-/-Ins2-/- AAV, Ins1+/+Ins2+/+ AAV, and Ins1-/-Ins2+/- mice for PC1/3, PC2, 7B2, and carboxypeptidase-E (B; representative images of n = 4-5). The proportion and intensity of PC1/3 or PC2 immunoreactivity in INS+ cells were quantified (C). Pancreas was immunostained for human C-peptide and β-galactosidase and there is rare colocalization (D; representative images of n=3). Immunostaining for NKX6.1 and GLUT2 in pancreas (D). Representative images of n = 3. All groups were compared by one way ANOVA with Tukey’s post-hoc test (A and C). Scale bars are 100 μm and insets are enlarged 4x. Ins1-/-Ins2-/-AAVDAPI HUMAN C-PEP IAPPDAPI HUMAN C-PEPIns1-/-Ins2+/-AAV Ins1-/-Ins2+/-SalineIns1 -/-Ins2 +/- AAVIns1 -/-Ins2 -/- Non-respondersIns1 -/-Ins2 -/- AAV/SalineIns1 -/-Ins2 -/- AAV/AAV5101520Infected β-cells (%)Ins1-/-Ins2-/- AAV Ins1+/+Ins2+/+ AAV Ins1-/-Ins2+/- AAVDAPI HUMAN C-PEPTIDE β-GALACTOSIDASEDAPI CARBOXYPEPTIDASE-E IAPPDAPI PC2 HUMAN C-PEPTIDE INSULINDAPI PC1/3 HUMAN C-PEPTIDE INSULINDAPI 7B2 INSULINIns1-/-Ins2-/- AAV Ins1+/+Ins2+/+ AAV Ins1-/-Ins2+/- AAVDAPI NKX6.1 GLUT-2 INSULINAB CDIns+PC+Ins+0255075100012RelativeFluorescentIntensityPC1/3 PC1/3PC2 PC2 129 AAV Ins1-Ins1 infects fewer β-cells than the AAV Ins1-INS and Ins1-/-Ins2-/- β-cells are unable to process mouse insulin 1 Given that Ins1-/-Ins2-/- β-cells were unable to process human insulin delivered via AAV Ins1-INS we aimed to determine if they could process a native mouse insulin. We designed and delivered an AAV carrying mouse insulin 1 under control of a 401bp fragment of the rat insulin 1 promoter (AAV Ins1-Ins1). We tracked body weight (Figure 4-4a) and blood glucose (Figure 4-4B) of Ins1-/-Ins2-/- mice, Ins1+/+Ins2+/+ wild-type controls and Ins1-/-Ins2+/- littermate controls given either 5x1010 VGP AAV Ins1-Ins1 per gram body weight or saline at 15 days of age. Ins1-/-Ins2-/- pups were treated with exogenous insulin therapy (Lantus®) as needed for the duration of the study. Seven Ins1-/-Ins2-/- mice were treated with AAV Ins1-Ins1 at 15 days of age (red) and treatment did not induce diabetic remission (Figure 4-4B). Four Ins1-/-Ins2-/- mice were given isogenic islet transplants into the anterior chamber of the eye at 15 days of age (purple) but two mice died during or shortly after surgery, so we present data from the remaining n = 2 (group denoted Ins1-/-Ins2-/- islets + AAV). The Ins1-/-Ins2-/- islets + AAV group was then treated with high dose 5x1011 VGP AAV Ins1-Ins1 per gram body weight at 30 days of age and the islet grafts were removed by enucleation at 49 days of age. Plasma collected at 4 weeks of age was assayed for mouse C-peptide and there were low but detectable levels in the Ins1-/-Ins2-/- AAV group and similar levels in the Ins1-/-Ins2-/- islets + AAV, Ins1-/-Ins2+/- saline, and Ins1-/-Ins2+/- AAV groups (Figure 4-4C). At 7 weeks of age, Ins1-/-Ins2-/- AAV mice did not have detectable mouse C-peptide (Figure 4-4D) but had detectable proinsulin (Figure 4-4E) in plasma collected after a 4 hour fast. After Ins1-/-Ins2-/- islets + AAV mice were enucleated at 7 weeks of age, plasma collected during an IPGTT (0 min and 30 min relative to 1 g glucose/kg body weight at 8 weeks of age) and fast/refeed (4 hours fast and 1 hour refeed at 10 weeks of age) was assayed for C-peptide. Ins1-/-Ins2-/- islets + AAV mice had normal C-peptide immunoreactivity (Figure 4-4F). C-peptide immunoreactivity is likely at least largely attributable to cross-reactivity to  130 extreme hyperproinsulinemia in Ins1-/-Ins2-/- islets + AAV mice (Figure 4-4G). One of the Ins1-/-Ins2-/- islets + AAV mice had normal glucose tolerance and the other had glucose intolerance (Figure 4-4H; 1 g glucose/kg body weight). We collected pancreas at 10 weeks of age and immunostained for glucagon and C-peptide 1 (Figure 4-4I). There were low rates of C-peptide 1 immunoreactivity (4.4±1.2%) in IAPP+ β-cells of Ins1-/-Ins2-/- AAV mice and an expected higher rate of infection (63.7±7.7%) in high-dose treated Ins1-/-Ins2-/- islets + AAV mice (Figure 4-4J).  Ins1-/-Ins2-/- β-cells have impaired proinsulin and proIAPP processing We assessed the processing of the IAPP precursor proIAPP and the insulin precursor proinsulin in β-cells of wild-type and AAV or islets + AAV treated Ins1-/-Ins2-/- mice. In mice, proIAPP is processed by PC1/3 then PC2 sequentially at the C-terminus then N-terminus respectively447, 448. Using antibodies specific for the C-terminal fragment of proIAPP and N-terminal fragment of proIAPP there was cytoplasmic immunoreactivity for the latter in Ins1-/-Ins2-/- β-cells, suggesting poor N-terminal proIAPP processing (Figure 4-5A). Using proinsulin specific B-C junction and C-A junction antibodies there was immunoreactivity throughout the cytoplasm using the C-A junction antibody. Both proIAPP and proinsulin antibodies suggest that there was poor processing at PC2 dependent sites – the C-A junction of proinsulin and N-terminus of proIAPP. Similar to past findings (Figure 4-3B), there was immunoreactivity for both PC1/3 and PC2 in Ins1-/-Ins2-/- β-cells (Figure 4-5B) and immunoreactivity for both 7B2 and CPE in islets + AAV treated mice. PC2, 7B2, and CPE immunoreactivity was present but reduced in islets of hyperglycemic Ins1-/-Ins2-/- AAV mice. We performed transmission electron microscopy on pancreas sections (Figure 4-5C). There were normal dense core secretory granules in wild-type mice but only immature low-density granules lacking halos in infected β-cells of Ins1-/-Ins2-/- islets + AAV mice. This was the third finding, along with cytoplasmic immunoreactivity for prohormones shown (Figure 4-5A) and elevated circulating proinsulin (Figures 4-4 E and 4-4G), suggesting impaired proinsulin processing in AAV Ins1-Ins1 infected Ins1-/-Ins2-/- β-cells.  131  Figure 4-4 Low infectivity and impaired proinsulin processing prevented diabetic remission in AAV Ins1-Ins1 treated Ins1-/-Ins2-/- mice Random fed body weight (A) and blood glucose (B) of Ins1-/-Ins2-/-, Ins1-/-Ins2+/-, and Ins1+/+Ins2+/+ mice. At 15 days of age all animals were injected with either AAV (light blue and red), PBS (dark blue), or given an islet transplant into the anterior chamber of the eye (purple). Four hour fasted plasma collected 2 weeks (C) and 5 weeks (D) after AAV injection was assayed for mouse C-peptide (RF = random fed, F = fasted, G = 15 min after 1 g glucose IP/kg body weight). Circulating proinsulin in 4 hour fasted plasma collected 5 weeks post AAV (E). Circulating C-peptide (F) and proinsulin (G) in control mice and mice treated with islet transplant at two weeks of age (shown by purple triangle), high dose AAV (5x1011VGP/g) at five weeks of 0 7 14 21 28 35 42 49 56 63 70010203040Days of AgeBloodGlucose(mM)0 7 14 21 28 35 42 49 56 63 70010203040Days of AgeBody weight(g)Ins1-/-Ins2+/-AAV AAVIslet transplant EnucleationIns1-/-Ins2+/-SalineIns1-/-Ins2-/-AAV Ins1-/-Ins2-/-Islets+AAVIns1+/+Ins2+/+0.00.51.01.5MouseC-peptide(ng/mL)4 weeks old0.00.10.2Proinsulin(ng/mL)0246MouseC-Peptide(ng/mL) 7 weeks old 7 weeks oldRF F G RF F GRF F G RF F G RF F G RF F GRF F G RF F G012345MouseC-Peptide(ng/mL)0.51.051525Proinsulin(ng/mL)8 weeks old 10 weeks old 8 weeks old 9 weeks old10 weeks oldDAPI GLUCAGON C-PEPTIDE IIns1-/-Ins2-/-Islets+AAV Ins1-/-Ins2-/-AAVIns1-/-Ins2+/-AAV Ins1+/+Ins2+/+0510152025Time (mins)BloodGlucose(mM)15 30 60 90BC DAEF G HI DAPI C-PEPTIDE I IAPP020406080C-PEP I+IAPP+(%)IAPP+J 132 age (shown by purple arrow), and enucleation at seven weeks of age (shown by purple ┴). Plasma was collected at eight (4 hour fasted and 1 hour post-refeed) and ten (4 hour fasted) weeks of age. IP glucose tolerance test (1 g glucose/kg body weight) at 9 weeks of age (H). Immunostaining for glucagon and C-peptide 1 (I) and β-cell infection rate using IAPP as a marker for β-cells (J). Data shown as mean ± SEM in bold colors with individuals shown in faint colors (A, B, and H) or on box and whisker plots (C-G). Representative images of n = 3, scale bars are 100 μm.    Figure 4-5 Ins1-/-Ins2-/- β-cells have impaired prohormone processing DAPI INSULIN N-TERMINAL PROIAPPIns1-/-Ins2-/-Islets/AAVIns1-/-Ins2-/-AAVIns1+/+Ins2+/+DAPI INSULIN C-TERMINAL PROIAPPDAPI CARBOXYPEPTIDASE-E INSULINDAPI INSULIN PROHORMONE CONVERTASE 2DAPI INSULIN PROHORMONE CONVERTASE 1/3DAPI INSULIN NEUROENDOCRINE PROTEIN 7B2DAPI INSULIN B-C JUNCTIONDAPI INSULIN C-A JUNCTIONA Ins1-/-Ins2-/-Islets/AAVIns1-/-Ins2-/-AAVIns1+/+Ins2+/+BIns1-/-Ins2-/-Islets/AAVIns1+/+Ins2+/+C 133 Pancreas collected from mice studied in Figure 4-4 was immunostained for insulin (green) and the C-terminal fragment of proIAPP or intact proIAPP (top), the N-terminal fragment of proIAPP or N-terminal extended proIAPP (second row), the intact unprocessed B-C junction of proinsulin (third row), or the intact unprocessed C-A junction of proinsulin (bottom row). Cytoplasmic immunoreactivity for N-terminal proIAPP and the intact C-A junction of proinsulin suggests that there is impaired processing of the N-terminal junction in proIAPP and the C-A junction in proinsulin in β-cells of Ins1-/-Ins2-/- mice (A). Immunostaining of pancreas for PC1/3 (top row), PC2 (second row), 7B2 (third row), and CPE (bottom row; B). Representative images of n = 3. Scale bars are 100 μm. Transmission electron microscopy of β-cells in wild-type Ins1+/+Ins2+/+ and AAV Ins1-Ins1 infected endogenous β-cells of Ins1-/-Ins2-/- islets + AAV mice (C). Normal dense core insulin secretory granules are abundant in Ins1+/+Ins2+/+ control β-cells, but Ins1-/-Ins2-/- β-cells contain mostly immature granules. The scale bar is 3 μm. Representative images of n = 2.   After treatment with AAV Ins1-Ins1, β-cells of Ins1-/-Ins2-/- mice appear mature and are not immunoreactive for markers of dedifferentiation We immunostained pancreas from AAV Ins1-Ins1 treated mice and controls for markers of β-cell maturity and dedifferentiation. Both Ins1-/-Ins2-/- AAV mice and high dose AAV treated Ins1-/-Ins2-/- islets + AAV mice had normal nuclear immunoreactivity in the cells in the core of the islet for the β-cell transcription factors PDX1, NKX2.2, NKX6.1, and MAFA (Figure 4-6A). Ins1-/-Ins2-/- islets + AAV mice had normal membranous GLUT2 immunoreactivity but hyperglycemic low dose AAV treated Ins1-/-In2-/- AAV mice lacked GLUT2 immunoreactivity. Immunostaining for progenitor markers that have been used as β-cell dedifferentiation markers35, 442, revealed no evidence for β-cell dedifferentiation (Figure 4-6B). There was no immunoreactivity for NANOG or L-MYC and rare immunoreactivity for ALDH1A3. Thus, despite an apparent mature phenotype, no evidence for dedifferentiation, and immunoreactivity for prohormone processing enzymes, AAV Ins1-Ins1 infected Ins1-/-Ins2-/- β-cells have a severe prohormone processing defect.   134   Figure 4-6 Ins1-/-Ins2-/- β-cells appear mature and are not immunoreactive for markers of dedifferentiation Pancreas collected from mice studied in Figure 4-4 was immunostained for Insulin alongside PDX1 and GLUT2 (top row), NKX6.1 and NKX 2.2 (middle row), or MAFA (bottom row; A). Pancreas was immunostained for IAPP (red) as a marker for β-cells and progenitor markers used to probe β-cell dedifferentiation – NANOG (top row), L-MYC (middle row), and ALDH1A3 (bottom row; B). Representative images of n = 2-4, scale bars are 100 μm.   AAV Ins1-Ins1 has low infectivity for β-cells and may have aberrant hepatic expression Based on persistent prohormone processing defects in AAV infected Ins1-/-Ins2-/- β-cells, we hypothesized that developmental effects led to irreversible defects. Thus, we aimed to avoid developmental defects by treating adult insulin knockout mice. Before treating adult insulin knockout mice, we aimed to determine the optimal AAV dose for β-cell infection by injecting Ins1-/-Ins2+/+ adult mice with variable doses of AAV Ins1-Ins1. We treated 8-week-old male mice with doses ranging from 5 x 1011 VGP up to 5 x 1012 VGP of both AAV Ins1-Ins1 and AAV Ins1-GFP. We delivered AAV Ins1-GFP as a positive control because of known high efficiency of β-cell infection in past experiments in the Kieffer lab449. We collected plasma after a 4 hour fast and 1 hour refeed (Figure 4-7A; 14 days post-AAV) and after a 4 hour fast and 30 minutes post DAPI NANOG IAPPIns1-/-Ins2-/-Islets/AAVIns1-/-Ins2-/-AAVIns1+/+Ins2+/+DAPI L-MYC IAPPDAPI ALDH1A3 IAPPDAPI MAFA INSULINIns1-/-Ins2-/-Islets/AAVIns1-/-Ins2-/-AAVIns1+/+Ins2+/+DAPI INSULIN NKX6.1 NKX2.2DAPI INSULIN PDX1 GLUT-2A B 135 2 g glucose/kg body weight (Figure 4-7B; 16 days post-AAV) and assayed for proinsulin. There was an AAV dose dependent increase in circulating proinsulin using an assay with no reported cross reactivity to mature insulin or C-peptide (Mercodia, 10-1232-01; Table 4-1). We collected pancreas and liver 17 days post-AAV and immunostained using an insulin antibody as a marker for β-cells and a C-peptide 1 or GFP antibody to assess infection rates of the AAV Ins1-Ins1 or AAV Ins1-GFP respectively (Figure 4-7C). In this experiment we used an uninfected Ins1-/-Ins2+/- mouse pancreas as a negative control and note issues with non-specific immunoreactivity of the C-peptide 1 antibody in the perinuclear region of islet cells. When accounting for this background signal, there is a dose-dependent infection rate for both AAVs. Though this background signal was unexpected based on previous optimizations16, this experiment was not repeated because of the highly limited supply of this polyclonal antibody. AAV Ins1-Ins1 was less efficient at infecting β-cells than AAV Ins1-GFP at all doses. A dose of 2.5 x 1012 VGP produced the highest infection rate using the AAV Ins1-Ins1. To determine if the abundant circulating proinsulin was from a β-cell source, we immunostained pancreas for insulin (INS) and proinsulin specific uncut B-C junction (B/C) and uncut C-A junction (C/A) antibodies (Figure 4-7D). There was consistent perinuclear immunoreactivity suggesting normal processing. We immunostained liver as a potential source of off-target expression of the AAVs and thus a source of circulating proinsulin. Though there was no detected immunoreactivity using an antibody that has reactivity for both insulin and proinsulin, there was abundant bright GFP immunoreactivity suggesting activation of the Ins1 promoter and production of GFP (Figure 4-7E). Given that insulin is rapidly secreted whereas GFP would be expected to be stored, and both the AAV Ins1-GFP and AAV Ins1-Ins1 have identical promoter sequences, it is possible that infected hepatocytes not only make GFP, but also produce and rapidly secrete (pro)insulin thus not having detectable (pro)insulin immunoreactivity.  136  DAPI C-PEPTIDE I INSULIN5 x 1011VGP 1 x 1012VGP 2.5 x 1012VGP 5 x 1012VGPDAPI INS B/C5 x 1011VGP 1 x 1012VGP 2.5 x 1012VGP 5 x 1012VGPInfected INS+ cells (%)BCDAIns1-/-Ins2+/-DAPI C-PEP1DAPI INS C/ADAPI INS GFP5 x 1011VGP 1 x 1012VGP 2.5 x 1012VGP 5 x 1012VGPEFastFed246Proinsulin(ng/ml)5x1011 1012 2.5x1012 5x10120 min30 min246Proinsulin(ng/ml)5x1011 1012 2.5x1012 5x101220406080100mINS1GFPDAPI INSULIN GFP5 x 1011VGP 1 x 1012VGP 2.5 x 1012VGP 5 x 1012VGPAAV dose2.5x1012 5x10121012 137 Figure 4-7 AAV Ins1-Ins1 has low infectivity for β-cells and may have aberrant hepatic expression Adult Ins1-/-Ins2+/+ mice were treated with varying doses of AAV Ins1-Ins1 and AAV Ins1-eGFP. Blood was collected during a fast-refeed (4 hour fast and 1 hour refeed) 14 days post-AAV (A) and an IPGTT (2 g glucose/kg body weight) 16 days post-AAV (B) and assayed for proinsulin and proinsulin intermediates. Pancreas collected 17 days post-AAV was immunostained for insulin as a marker of β-cells alongside AAV products C-peptide 1 (top row) or GFP (bottom row; C). Uninfected Ins1-/-Ins2+/- mouse pancreas was used as a negative control and is shown on the left. Quantification of percent insulin immunoreactive cells that were immunoreactive for C-peptide 1 or GFP shown to the right. There is perinuclear immunoreactivity in β-cells for the intact unprocessed B-C junction and C-A junction of proinsulin (D). Liver was immunostained for insulin and GFP (E). Scale bars in all panels are 100 μm and representative images of n = 1-2 shown.   AAV Ins1-Ins1 did not prevent diabetes onset when co-delivered with AAV Ins1-Cre Using the optimized dose of AAV Ins1-Ins1 (Figure 4-7; 2.5 x1012 VGP) we attempted to prevent diabetes onset in Ins1-/-Ins2f/f mice co-treated with AAV Ins1-Cre. Upon Cre mediated recombination, Ins1-/-Ins2f/f mice lose functional Ins2 genes and become insulin deficient. Adult male mice (6-8 weeks of age) were treated with 1.5 x 1012 VGP AAV Ins1-Cre, 1.5 x 1012 VGP AAV Ins1-Cre plus 2.5 x 1012 VGP AAV Ins1-Ins1, or PBS. There were no differences in body weight over the duration of the study (Figure 4-8A; -1 to 9 weeks relative to AAV) and both AAV treated groups developed fasting hyperglycemia (4 hour fast) 4 weeks after infection (Figure 4-8B). We collected plasma after 4 hour fasts throughout the study and assayed for C-peptide (Figure 4-8C). There were no significant differences between groups. We also assayed plasma for proinsulin (Figure 4-8D) and detected a transient increase in proinsulin in the AAV Ins1-Cre + AAV Ins1-Ins1 treated group. Both AAV treated groups had similar glucose intolerance during IPGTTs (2 g glucose/kg body weight) 22 days (Figure 4-8E) and 46 days (Figure 4-8F) post-infection compared to controls. We assayed plasma collected at 0 minutes and 15 minutes relative to glucose injection during the day 46 IPGTT and there were no significant differences in stimulation index (C-peptide at 15 minutes/C-peptide at 0 minutes) between groups (Figure 4-8G). This fails to replicate the reduced glucose stimulated insulin secretion presented in Figure  138 3-9 though there was a non-significant trend to reduced C-peptide stimulation index in the AAV Ins1-Cre treated group (p = 0.2). As a potential explanation for observations of glucose intolerance but non-significant changes to circulating C-peptide levels, we performed an insulin tolerance test on day 51 relative to AAV to assess insulin sensitivity (Figure 4-8H). There were no significant differences in relative changes to blood glucose when normalized to baseline nor when presented as variable non-normalized blood glucose (data not shown).   Figure 4-8 AAV Ins1-Ins1 did not prevent diabetes onset when co-delivered with AAV Ins1-Cre 0 14 28 42 56 70 84253035Day relative to AAVBodyweight(g)PBS Ins1-Cre (1.5x1012VGP) Ins1-Cre (1.5x1012VGP) + Ins1-Ins1 (2.5x1012VGP)0 14 28 42 56 70 84510152025Day relative to AAVBloodGlucose(mM)**** *** ***A B0 30 60 90 1205101520253035Time (min)BloodGlucose(mM)15****** **** ****0 30 60 90 1205101520253035Time (min)BloodGlucose(mM)15*** ******-7 0 7 14 21 28 35 42 49012345DayC-peptide(ng/mL)-7 0 7 14 21 28 35 42 490.00.51.0DayProinsulin(ng/ml)*****012345Stimulation index0 30 60 900.00.51.01.5Time (min)RelativeBloodGlucose1020 45CFD EG H 139 Fasting body weight (A) and blood glucose (B) of inducible insulin knockout Ins1-/-Ins2f/f mice that received AAV Ins1-Cre, AAV Ins1-Cre plus AAV Ins1-Ins1, or PBS by IP injection on day 0. C-peptide (C) and proinsulin (D) in plasma collected after 4 hour fasts. Glucose tolerance (IP 2 g glucose/kg body weight) on days 22 (E) and 46 (F) relative to AAV injection. Stimulation index of C-peptide in plasma collected during the day 46 glucose tolerance test (C-peptide at 15 minutes/C-peptide at 0 minutes relative to glucose injection; G). Insulin tolerance (0.8 U Novolin®/kg body weight) on day 51 relative to AAV injection (H). Data presented as mean ± SEM (A-F, H) or mean ± 95% confidence interval (G), and individual animals are shown in all panels.   AAV Ins1-Ins1 produced unexpectedly low rates of β-cell C-peptide 1 immunoreactivity Thirteen weeks after AAV delivery, pancreas was collected in PFA. We immunostained for IAPP as a marker of β-cells. We assessed AAV Ins1-Cre mediated recombination by determining the proportion of IAPP+ cells immunoreactive for C-peptide 2. We assessed AAV Ins1-Ins1 infection rate by determining the proportion of IAPP+ cells immunoreactive for C-peptide 1. There was a 10-16% infection rate for the AAV Ins1-Ins1 and ~half of β-cells in AAV Ins1-Cre treated mice were not immunoreactive for C-peptide 2 (Figure 4-9A). There was normal PC1/3 immunoreactivity in all groups but reduced PC2 immunoreactivity in insulin immunoreactive cells in AAV Ins1-Ins1 and AAV Ins1-Ins1 + AAV Ins1-Cre treated animals (Figure 4-9B). There were no substantial changes to proinsulin immunoreactivity using unprocessed B-C junction and C-A junction antibodies (Figure 4-9B). Notably, this study required production of a new batch of AAV as the batch used in the two previous studies (Figures 4-4 to 4-7) was depleted. Given the unexpectedly low rate of infection by AAV Ins1-Ins1 (10-16% presented in Figure 4-9A compared to 51% using the same dose of 2.5x1012 VGP presented in Figure 4-7C) and previous experience with AAV being produced at a lower titer than reported by the manufacturer, we assessed the new batch of AAV. We performed a quantitative PCR experiment to titer four different AAVs; the highly infectious AAV Ins1-GFP used in Figure 4-7C and three batches of AAV Ins1-Ins1. The first was previously studied and found to be at a titer 1/10th expected (AAV Ins1-Ins1a), the second was used in   140  Figure 4-9 Relatively low rates of C-peptide 1 immunoreactivity after co-delivery of AAV Ins1-Ins1 with AAV Ins1-Cre Thirteen weeks post-AAV pancreas was collected and immunostained for IAPP (used as a marker of β-cells) and C-peptide 1 (CP1) or C-peptide 2 (CP2). Representative images shown and quantification of the proportion of IAPP+ cells immunoreactive for C-peptide 1 or C-peptide 2 is shown to the right (A). Pancreas was immunostained for insulin and either PC1, PC2, an unprocessed B-C junction of proinsulin (BC), or an unprocessed C-A junction of proinsulin (CA; panel B). Titer of AAVs (AAV Ins1-GFP produced by the Children’s Hospital of Philadelphia and three batches of AAV Ins1-Ins1 produced by Vector Biolabs) was calculated by quantitative PCR using plasmids of known concentrations (C). Individual animals (A) or technical replicates at three concentrations along a standard curve (C) are shown on box and whisker plots. Groups were compared by one-way ANOVA with Tukey’s post hoc test (A). * p < 0.05, ** p < 0.01.  PBSDAPIIAPPCP2AAV Cre AAV Cre + Ins1DAPI INS PC2DAPIIAPPCP1DAPI INS PC1 DAPI INS BC DAPI INS CAPBSAAV CreAAV Cre + Ins1051015200255075100C-peptide1+/IAPP+(%) C-peptide2+/IAPP+(%)ABGFP Ins1a Ins1b Ins1c0.10.20.31.52.5Kieffer/ManufacturertiterC** *** *PBS Ins1-Cre Ins1-Cre + Ins1-Ins1 141 Figures 4-4 to 4-7 (AAV Ins1-Ins1b), and the third was used for studies presented in Figures 4-8 to 4-9. Using plasmids of known concentration as the control, we found that there was an expected discrepancy with the highly infectious AAV Ins1-GFP virus ~50% more concentrated than reported (ratio of titer calculated in the Kieffer lab/titer reported by the manufacturer) and the failed batch ~1/10th as concentrated as reported (AAV Ins1-Ins1a). Interestingly, batches AAV Ins1-Ins1b and AAV Ins1-Ins1c both were similar in concentration of ~1/5th the concentration reported by the manufacturer. Though these results are disappointing, discrepancies in titer do not seem to explain the differences in infection rate reported using batch AAV Ins1-Ins1b (Figure 4-7C) versus batch AAV Ins1-Ins1c (Figure 4-9A). We delivered 1012 VGP of AAV Ins1-Ins1c to adult Ins1-/-Ins2f/f mice and assessed infection rate in β-cells. In a first experiment, we did not include a positive control (Figure 4-10A) thus making it difficult to interpret weak perinuclear immunoreactivity. We repeated the experiment with wild-type Ins1+/+Ins2+/+ controls and found that there were no C-peptide 1 immunoreactive cells in most AAV Ins1-Ins1 treated islets (Figure 4-10B). Rare immunoreactive cells were observed (see cells in white boxes in Figure 4-10A). Infection rate was far less than expected given infection rates of 11.7±3.5% using 1012 VGP of batch Ins1b (Figure 4-7C). These findings justify using a new batch of AAV Ins1-Ins1 and it may be worthwhile to utilize an alternative adult inducible insulin knockout mouse model relying on less total AAV and producing higher β-cell recombination rates.   142  Figure 4-10 Failed production of AAV Ins1-Ins1 Adult 10-week-old Ins1-/-Ins2f/f mice were treated with 1012 VGP AAV Ins1-Ins1 (Batch C). Two weeks post-AAV pancreas was collected and immunostained for insulin (green) and C-peptide 1 (red) in two distinct experiments (A-B). Pancreas from adult Ins1-/-Ins2-/-, Ins1-/-Ins2+/-, and Ins1+/+Ins2+/+ mice were used as controls. Representative images of n = 2-3 shown. Rare C-peptide 1 immunoreactive cells shown in white boxes in (A). Scale bars are 100 μm.   Optimizing tamoxifen dose to induce diabetes in Ins1-/-Ins2f/fmTmGPdx-CreER mice In preparation for using Ins1-/-Ins2f/fmTmGPdx-CreER mice as an alternative model for adult inducible insulin knockout mice, we investigated doses of tamoxifen for induction of diabetes. We treated adult male mice (12-14 weeks of age) with tamoxifen by IP injections on four consecutive days. Upon delivery of tamoxifen to these animals, Cre translocates to the nucleus and the Ins2 genes undergo recombination. A past report using Ins1-/-Ins2f/fmTmGPdx-CreER mice reported a >95% recombination rate in β-cells after delivery of 3 mg tamoxifen/40 g body weight430. We compared this previously reported dose430 with an alternative dose of 6 mg tamoxifen/40 g body weight. In agreement with Szabat et al., 3 mg tamoxifen/40 mg body weight was an ideal dose as it induced robust diabetes in approximately three weeks (Figure 4-11A) with no significant effects on body weight (Figure 4-11B). Induction of diabetes suggests  143 efficient recombination of the floxed Ins2 gene in β-cells and minimal weight loss suggests tolerable tamoxifen toxicity for animals.   Figure 4-11 Efficient induction of diabetes in Ins1-/-Ins2f/fmTmGPdxCre-ER mice Adult male Ins1-/-Ins2f/fmTmGPdxCre-ER mice were given tamoxifen on four sequential days by IP injection (shown by shaded region and four arrows). Random fed blood glucose (A) and body weight (B) was monitored over 20 days. Data presented as mean (solid lines) and individual data points.   4.4 Discussion Monogenic diabetes is a candidate disease for a gene therapy cure. With the availability of the AAV as a clinically relevant viral vector, we investigated the viability of an AAV mediated therapy for the Ins1-/-Ins2-/- mouse model of PND. Despite efficacy at delivering the insulin gene to endogenous Ins1-/-Ins2-/- β-cells, AAVs carrying either a mouse or human insulin gene could at best transiently reverse diabetes because of modest infectivity when delivered at two weeks of age and a severe enduring prohormone processing defect in Ins1-/-Ins2-/- β-cells. Multiple experimental techniques suggest that Ins1-/-Ins2-/- β-cells have a prohormone processing defect that endures after replacement of an insulin gene by AAV. First, by commercial ELISA we found elevated circulating proinsulin. Second, by immunohistofluorescence we detected cytoplasmic proinsulin immunoreactivity and cytoplasmic proIAPP immunoreactivity. Finally, by electron  144 microscopy there were abundant immature insulin granules in infected β-cells. In follow-up to these observations, we avoided developmental effects of insulin gene deletion and replaced the insulin gene to adult inducible insulin knockout animals. Unfortunately, this approach failed to prevent onset of diabetes but warrants follow-up using a better prepared and validated batch of AAV Ins1-Ins1 and a different inducible knockout model. These findings suggest that insulin deficient β-cells develop a prohormone processing defect that is not reversed by replacing an insulin gene by AAV. In our first attempt to reverse PND, we treated Ins1-/-Ins2-/- pups at two weeks of age with AAV Ins1-INS and observed a transient remission of diabetes. We propose that circulating proinsulin was responsible for lowering blood glucose during this period. By immunostaining pancreatic islets, we observed impaired processing at the B-C junction in AAV Ins1-INS infected Ins1-/-Ins2-/- β-cells (Figure 4-2L) suggesting a buildup of des-64,65 proinsulin and/or split-65,66 proinsulin. Notably, these proinsulin intermediates have bioactivities of ~20% that of mature insulin (Table 4-3). It is plausible that immediately following AAV therapy animals were small and insulin sensitive and insulin intermediates with bioactivities 1/5th that of mature insulin were enough to lower blood glucose. In the following weeks, their body weight quickly doubled and as animals reached 5-6 weeks of age there was an expected peak in the counterregulatory hormone growth hormone during the pubertal period450. Additionally, as has been observed in patients following islet transplantation, there is a progressive need for higher levels of circulating C-peptide to maintain blood glucose levels451, suggesting reduced insulin sensitivity after being exposed to a cell source of insulin. As C-peptide levels were consistent throughout our study, evidently any cells producing (pro)insulin were not capable of compensating and thus secreted insufficient (pro)insulin molecules to prevent relapse to hyperglycemia. Additionally, given that an identical promoter led to aberrant GFP production in the liver (AAV Ins1-GFP; Figure 4-7E), it is possible that non-β-cells were an extra source of proinsulin in the first week post-AAV and  145 contributed to transient diabetic remission. Furthermore, retreatment of a subset of animals with extra AAV (AAV/AAV group) was not successful at improving infection rate nor levels of circulating human C-peptide. Given that following exposure to high dose AAV mice mount an immune response and form memory immune cells452, it is somewhat unsurprising that retreatment failed to effectively deliver genes to additional β-cells. After making observations of impaired proinsulin processing in insulin knockout pups treated with the human insulin virus, we designed and produced an AAV carrying the native mouse Ins1 open reading frame (AAV Ins1-Ins1). Regrettably, using an identical study design to the AAV Ins1-INS study, we failed to produce even a transient remission of diabetes. Based on histological evaluation of infection rate we found that AAV Ins1-Ins1 only infected 4.4±1.2% of β-cells compared to an infection rate of 8.5±1.0% after treatment with AAV Ins1-INS. We note that these percentages may not be representative of the actual infection rate at the time of treatment because of the expansion of β-cell mass that occurred between AAV delivery at two weeks of age and tissue collection several weeks later. There is a rapid doubling of β-cell mass between 17 and 31 days of age in Sprague Dawley rats453 and a similar doubling in β-cell mass in mice between 14 and 25 days of age454. Regardless of the actual infection rate at the time of AAV delivery, given the associated marginally detectable circulating C-peptide immunoreactivity of < 0.5 ng/mL two weeks after AAV Ins1-Ins1 compared to ~1-2 ng/mL after AAV Ins1-INS, the lower infection rate of AAV Ins1-Ins1 likely contributed to its inability to cause a reversal of diabetes. Furthermore, we direct the reader to compare the impaired proinsulin processing phenotype in AAV Ins1-INS versus AAV Ins1-Ins1 infected Ins1-/-Ins2-/- β-cells. Though AAV Ins1-INS infected β-cells had impaired processing at the B-C junction of proinsulin, AAV Ins1-Ins1 infected β-cells had impaired processing at the C-A junction. This observation is important because proinsulin unprocessed at the C-A junction has bioactivities of only 2-11% that of mature insulin whereas proinsulin unprocessed at the B-C junction has bioactivities of 10-30%  146 that of mature insulin (Table 4-3). This difference in relative bioactivities may have contributed to the failure of the AAV Ins1-Ins1 to produce even a transient remission of diabetes in Ins1-/-Ins2-/- treated pups. Past investigations on human insulin transgenic mice349, 455, suggest that mouse β-cells are capable of processing human proinsulin. In consideration of this work, it is unsurprising that providing a mouse insulin AAV was not successful at resolving the proinsulin processing defect observed after delivery of the human insulin gene. Interestingly, the proinsulin processing defect in Ins1-/-Ins2-/- β-cells was different after delivery of human insulin via AAV Ins1-INS versus delivery of mouse insulin via AAV Ins1-Ins1. Though the current accepted theory is that proinsulin; regardless the species, is processed by PC1/3 at the B-C junction then by PC2 at the C-A junction, this theory is mostly based on studies on knockout mouse models313, 330. In unpublished work, we provide evidence that PC2 is not involved in the processing of human proinsulin in human β-cells (Chapter 5). In these human islet studies, we did not investigate the sequential processing of human proinsulin, but it is possible that it is first processed at the C-A junction much like rat insulin 2315. This could contribute to the surprising finding that AAV Ins1-INS infected Ins1-/-Ins2-/- β-cells appear to process human proinsulin at the C-A junction but not the B-C junction. Further, we make the discordant observations of impaired proinsulin processing paired with the presence of PC1/3, PC2, and 7B2 immunoreactivity. These findings can be explained in many ways, but generally speaking, highlight the complex process of normal prohormone processing. The presence of PC1/3 and PC2 is not adequate for prohormone processing – they must be properly sorted in secretory granules. Furthermore, secretory granules must be adequately acidified and contain sufficient calcium for proper proenzyme maturation to gain function. Additionally, prohormones like insulin itself must be properly sorted and folded to be processible by the prohormone convertases. Overall, the mechanism underlying the observed prohormone processing defect in Ins1-/-Ins2-/- β-cells is unclear. There have been no reports of impaired proinsulin processing in  147 β-cell selective insulin receptor knockout mice365 nor in β-cell selective IGF1r knockout mice456. Examining proinsulin processing in these mouse models by proinsulin immunostaining and commercial ELISAs is worthwhile and could provide insight into the mechanism by which a loss of insulin results in a β-cell prohormone processing defect. This wide breadth of potential explanation underlying impaired prohormone processing in Ins1-/-Ins2-/- β-cells justifies examination of the transcriptomic and proteomic landscape of β-cells after insulin gene delivery and potentially IR and/or IGF1r deficient β-cells. Exposure to hyperglycemic conditions can have dramatic effects on islets, including dedifferentiation35, dysfunction382, and reduced expression of mature β-cell factors Pdx1, Mafa, and Slc2a2381. Furthermore, β-cells may transdifferentiate and begin expressing glucagon457. In parallel to β-cell dysfunction and dedifferentiation during hyperglycemia, others have reported a compensatory upregulation of PC1/3 in α-cells458, 459. Increased PC1/3 increases production of GLP-1 in α-cells, which can have a protective effect on β-cells and reduce rates of apoptosis460, 461. Additionally, reduced production of glucagon may lower blood glucose. In the current thesis, we made several observations of reduced PC2 immunoreactivity in β-cells exposed to hyperglycemia. First, in Chapter 2 we found that hyperglycemic Ins1-/-Ins2-/- pups had reduced PC2 immunoreactivity and Pcsk2 expression (Figure 2-5A-B). Second, in Figure 4-5B hyperglycemic Ins1-/-Ins2-/- mice treated with AAV Ins1-Ins1 had reduced PC2 immunoreactivity, but euglycemic high-dose treated Ins1-/-Ins2-/- islets + AAV mice had normal PC2 immunoreactivity. Finally, in Figure 4-9B, hyperglycemic Ins1-/-Ins2f/f mice treated with AAV Ins1-Cre had reduced PC2 immunoreactivity in the core of the islet, but not in the islet mantle where α-cells are localized. Potentially during times of hyperglycemia, β-cells have reduced PC2 but preserve levels of PC1/3 because PC1/3 is substantially more capable of fully processing proinsulin, even in the absence of PC2313.   148 We make the interesting observation that after treatment with AAV Ins1-Ins1, both infected β-cells as well as neighboring insulin negative cells appear mature with nuclear immunoreactivity for MAFA (Figure 4-6). Previous studies on Ins1-/-Ins2-/- mice kept alive by insulin therapy revealed that long-term replacement of insulin by islet transplantation but not short-term replacement by injections was sufficient for more complete maturation of Ins1-/-Ins2-/- β-cells (Chapter 2). Given that AAV Ins1-Ins1 treated mice were hyperglycemic and sacrificed at a young age (7.5 weeks old), these findings suggest that factors beyond hyperglycemia and duration of insulin therapy likely contributed to the different β-cell phenotype in injection vs islet transplant treated Ins1-/-Ins2-/- mice. It is possible that AAV Ins1-Ins1 infected β-cells produced high local concentrations of native mouse insulin and perhaps began producing other factors (including free C-peptide, proinsulin, and proinsulin intermediates including des-31,32 proinsulin) sufficient for maturation of not only infected, but also neighboring uninfected β-cells.  Given severe prohormone processing defects observed in mice with lifelong insulin gene deletion, we attempted to prevent diabetes onset in adult inducible insulin knockout mice. Unfortunately, co-delivery of AAV Ins1-Ins1 and AAV Ins1-Cre failed to prevent onset of diabetes in Ins1-/-Ins2f/f mice. We propose that this is due an inadequate population of β-cells with restoration of Ins1 following loss of Ins2. Limited by lack of a lineage trace in this model, we are unable to precisely determine which proportion of the 10-16% of C-peptide 1 immunoreactive cells have undergone recombination and have lost Ins2. With that being said, we note that though only ~50% of IAPP+ cells are C-peptide 2 negative, recombination was likely higher because of onset of hyperglycemia. As discussed at length in Section 3-4, recombination is likely higher than loss of immunoreactivity because of observations of Cre+Ins+ cells. By considering this caveat, assuming random infection by AAV Ins1-Ins1, half or more of the 10-16% of C-peptide 1 immunoreactive cells have lost their endogenous Ins2 genes. Thus, delivery of AAV Ins1-Ins1 restored an insulin gene to 5-16% of β-cell mass. Given  149 that normal variation is more substantial462, this rate is unlikely to produce detectable changes in circulating insulin or glycemia. In follow-up, we verified the infectivity of AAV Ins1-Ins1c and confirmed extremely poor infection rate even when not co-delivering with AAV Ins1-Cre. Low AAV titer does not explain low rates of C-peptide 1 immunoreactivity because the AAV Ins1-Ins1c used in the AAV Ins1-Cre + AAV Ins1-Ins1 study was found to have a titer highly comparable to the AAV Ins1-Ins1b that produced a 50% infection rate using the same dose (Figure 4-7C). It is apparent that the AAV Ins1-Ins1 batch C is somehow failing and we are in communication with the manufacturer. Related, we note that the use of extremely high doses of AAV (1.5 x1012 AAV Ins1-Cre plus 2.5 x1012 AAV Ins1-Ins1) could have caused β-cell toxicity. Indeed, doses of only 3x1012 VGP of AAV Ins1-Cre led to impaired glucose tolerance (Figure 3-1B). Though not possible to confirm, loss of infected β-cells due to β-cell toxicity could have led to detection of only 10-16% infection rate by AAV Ins1-Ins1 at the time of pancreas collection 13 weeks post-AAV. Perhaps actual infection rate at the time of AAV delivery was closer to the 50% expected, but toxicity led to a loss of infected β-cells or infected β-cells were defective and failed to activate expression from AAV genomes. It is notable that the induction of diabetes by delivery of AAV Ins1-Cre to Ins1-/-Ins2f/f mice was less robust in the studies presented in Chapter 4 than those in Chapter 3, despite use of a higher dose of AAV (1 x 1012 VGP in Chapter 3 vs 1.5 x 1012 VGP in Chapter 4). This suggests that the infectivity or function of AAV Ins1-Cre may also be somehow impaired like the AAV Ins1-Ins1. Follow-up using an AAV independent adult inducible insulin knockout mouse model would be worthwhile to rule-in or out this possibility. A cohort of Ins1-/-Ins2f/fmTmGPdx-CreER mice have been generated in the Kieffer lab and we aim to determine if a carefully validated batch of AAV Ins1-Ins1 can reverse diabetes when delivered two-weeks after tamoxifen administration. Investigation into the efficacy of gene therapy to treat other models of monogenic diabetes is warranted. Notably, most INS mutations cause diabetes with a dominant negative  150 mechanism, by causing severe ER stress and thus likely such patients have few surviving endogenous β-cells. Without surviving β-cells an insulin gene replacement approach is not a viable approach to cure their diabetes. Instead, assessing the viability of a gene editing approach would be clinically relevant and using the readily available Akita mouse model would be informative. Using AAV8, delivery of guide RNAs, Cas9463, and a repair template to Akita mice could theoretically correct their C96Y mutation and restore normal insulin production and prevent onset of diabetes. Alternatively, early intervention with other therapies aimed at degrading misfolded proinsulin aggregates could preserve endogenous β-cells thereby making an insulin gene replacement therapy a viable option464, potentially even for homozygous mutants. Furthermore, patients with recessive loss of function mutations in insulin would be ideal candidates for a gene replacement. Patients with other mutations that suppress insulin secretion like those with KCNJ11 gain of function mutations, would also be ideal candidates for a gene therapy to replace or correct relevant mutations. Studies using other mouse models of monogenic diabetes like those with mutant Kcnj11 causing PND465 will be informative. Based on the current work it is possible that β-cell prohormone processing defects endure beyond gene replacement or correction. The findings in Chapter 4 provide insight into the viability of gene therapies for patients with monogenic diabetes. Despite use of multiple AAVs and multiple study designs, we have been largely unable to reverse PND caused by deletion of the insulin genes in mice. Notably, besides transient diabetic remission after treating Ins1-/-Ins2-/- pups with AAV Ins1-INS our only successful attempt to induce diabetic remission was via treatment of two mice with extremely high dose virus after islet transplantation. Nonetheless, despite β-cells becoming immunoreactive for markers of mature β-cells and not immunoreactive for markers of dedifferentiation, this approach did not resolve β-cell prohormone processing defects as animals had extreme hyperproinsulinemia and β-cells contained abundant unprocessed proinsulin and  151 proIAPP. We followed up on these findings by attempting to prevent onset of diabetes using adult inducible insulin knockout mice. Though unsuccessful, we propose that further investigations into reversing diabetes in adult inducible insulin knockout Ins1-/-Ins2f/fmTmGPdx-CreER mice using a new batch of AAV are worthwhile. Much work is still required to better understand the complex β-cell phenotype in the Ins1-/-Ins2-/- mouse model of PND. In summary, our findings urge conservatism in the pursuit of a gene therapy approach to resolve lifelong genetic defects in patients with INS mutations given potential secondary defects to genetic mutations.  152 Chapter 5: Revisiting proinsulin processing: Evidence that human β-cells lack prohormone convertase 2 and can produce mature insulin without its function  5.1 Background Since the discovery of proinsulin in 1967 much work has investigated the processing pathway to produce mature insulin. Mostly based on studying the PC2 knockout313 and PC1/3 knockout279 mouse models and in vitro experiments314, current theory posits that proinsulin is processed by PC1/3 and PC2 sequentially at the B-C and C-A junctions to excise C-peptide and yield mature insulin (Figure 1-7). For decades it has been assumed that proinsulin processing in mice recapitulates the process in humans. Careful review of the literature suggests that PC2 may not be involved in human proinsulin processing. There have been publications revealing low levels of PCSK2 mRNA in sorted human β-cells (Figure 1-8) and in single cell RNAseq databases319, 320 and one paper reported “β-cell selective PC2 deficiency”322. We interrogate the assumption that human β-cells process proinsulin like mice and reveal that human β-cells do not rely on PC2. Using rigorously validated antibodies and in situ hybridization there was no PC2 immunoreactivity and little PCSK2 in human β-cells. Using a pulse-chase study design, we revealed that human β-cells can produce mature insulin without the function of PC2. Through careful and rigorous use of multiple experimental techniques and experimental conditions we have provided evidence that there is a need to revise the longstanding theory of proinsulin processing.  5.2 Materials and methods Experimental models and subject details Human pancreas tissue biopsies were collected by the Irving K. Barber Human Islet Isolation Laboratory (Vancouver, Canada). Cadaveric human islets were provided by the  153 University of Alberta Hospital IsletCore (Edmonton, Canada) after isolation by standardized protocol 466 and all work with human tissues were approved with the Research Ethics Board (H14-02949). Basic donor demographics are detailed in Table 5-1 (IsletCore, University of Alberta, Edmonton, Canada; Ike Barber Human Islet Transplant Laboratory, University of British Columbia, Vancouver, Canada). All experiments with animals were approved by the UBC Animal Care Committee and carried out in accordance with the Canadian Council on Animal Care Guidelines. We express our sincere gratitude to organ donors and their families for making our studies on human tissues possible.   Table 5-1 Basic demographics of pancreas and islet donors used in experiments. Source Code Age Sex Body weight BMI Ike Barber 588 41 Male 68.2kg Unknown Ike Barber 594 40 Male 94.9kg Unknown Ike Barber 634 23 Female 105kg Unknown IsletCore R265 64 Female Unknown 23.7 IsletCore R267 64 Male Unknown 37.7 IsletCore R271 60 Female Unknown 26.0 IsletCore R278 57 Male Unknown 27.6 IsletCore R282 57 Male Unknown 26.4 IsletCore R283 22 Male Unknown 22.5 IsletCore R286 69 Female Unknown 27.7 IsletCore R145 55 Female Unknown 24.1 IsletCore R146 52 Male Unknown 25.3 IsletCore R147 62 Male Unknown 26.0 IsletCore R148 40 Male Unknown 29.1 IsletCore R151 46 Female Unknown 26.7   Paraffin embedded samples C57Bl/6J mice (Jackson Laboratories, Bar Harbour, ME, USA) were sacrificed at 12 weeks of age. Following euthanasia, the pancreas was quickly dissected out of mice, washed in phosphate buffered saline (PBS), and fixed in 4% paraformaldehyde (PFA) overnight before being transferred to 70% ethanol for storage prior to paraffin-embedding and sectioning (5 µm  154 thickness; Wax-It Histology Services, Vancouver, Canada). A similar protocol was used to collect pancreas from rats, pigs, and dogs. Human pancreas tissue biopsies were collected using the same protocol (Irving K. Barber Human Islet Isolation Laboratory; Vancouver, BC). Immunohistofluorescence Immunofluorescent staining was performed as previously described 16. Briefly, sections were deparaffinized in xylene (3x5 mins) and rehydrated in graded ethanol (100% 2x5 mins, 95% 5 mins, 70% 5 mins, and PBS 10 mins) before heat induced epitope retrieval in an EZ Retriever microwave oven (BioGenes; Fremont, CA) for 15 mins at 95oC in 10 mM citrate buffer (0.5% Tween 20, pH 6.0; ThermoFisher Scientific®, Waltham, MA). Samples were blocked in DAKO® Protein Block, Serum Free (Dako; Burlington, Canada) and incubated overnight in primary antibody diluted in Dako Antibody Diluent (Dako; Burlington Canada). The following day, slides were washed and incubated in secondary antibody (Alexa Fluor conjugated secondary antibodies, Life Technologies) for 1 hour at room temperature before mounting and counterstaining with VECTASHILED® Hard Set Mounting Medium with nuclear stain 4,6-diamidino-2-phenylindole (DAPI; Vector Laboratories; Burlingame, Ca). All images were captured and analyzed with an ImageXpress® Micro XLS System (Molecular Devices Corporation; Sunnyvale, CA) with a scientific CMOS camera, a Nikon 20× Plan Apo objective (NA=0.75, 1-6300-0196; Nikon, Tokyo, Japan) and DAPI (DAPI-5060B), FITC (FITC-3540B), Cy3 (Cy3-4040B), Texas Red (TXRED-4040B), and Cy5 (Cy5-4040A) filter cubes. All primary antibodies and other key resources using in Chapter 5 are listed in Table 5-2.       155 Table 5-2 Key resources including antibodies used in Chapter 5 REAGENT or RESOURCE SOURCE IDENTIFIER Primary antibodies Rabbit α-Insulin (C27C9) Cell Signaling Cat#: 3014; AB_2126503 Mouse α-Insulin (L6B10) Cell Signaling Cat#: 8138; AB_10949314  Guinea pig α-Insulin Linco Cat#: 4011-01F; AB_433703 Rabbit α-PC1/3 Lakshmi Devi AB_2665530  Mouse α-PC1/3 Gunilla Westermark 467, AB_2665529  Rabbit α-PC2C-term Epitope at C-terminus of mouse PC2 (residues 622-638) Thermo Fisher Scientific Cat#: PA1-058; AB_2158593 Rabbit α-PC2N-term Epitope near N-terminus of human PC2 (residues 87-116) Thermo Fisher Scientific Cat#: PA5-14595; AB_2299126 Mouse α-PC2R&D Unknown epitope within human PC2 (residues 110-638) R&D Systems Cat#: MAB6018; AB_10718108 Rabbit α-PC2IgM Polyclonal IgM antibody shared by Dr. Westermark Gunilla Westermark 467, AB_2737594 Mouse α-α-tubulin Sigma-Aldrich T9026, AB_477593 Mouse α-glucagon antibody Sigma-Aldrich Cat#: G2654; AB_259852 Rabbit α-glucagon antibody AbCam Cat#: EP3070; Ab10561971 Rabbit α-7B2 antibody Nabil Seidah AB_2737595 Mouse α-digoxigenin antibody Novus Biologicals Cat#: NB100-1879; AB_530922 Goat α-mouse IgG, alkaline phosphatase conjugated antibody Sigma-Aldrich Cat#: A3562; AB_258091 IRdye 680RD goat α-mouse IgG LI-COR Biosciences Cat#: 925-68070; AB_2651128    Biological Samples   Isolated human islets University of Alberta IsletCore https://www.ualberta.ca/alberta-diabetes/facilities/core-services/isletcore Human pancreas biopsies Ike Barber Human Islet Laboratory http://surgery.med.ubc.ca/research/labs/ike-barber-human-islet-transplant-laboratory/    Chemicals, Peptides, and Recombinant Proteins Recombinant human PC2 R&D Systems 6018-SE-010 3H-Leucine Perkin Elmer NET460005MC  156 REAGENT or RESOURCE SOURCE IDENTIFIER Chemicals, Peptides, and Recombinant Proteins Brefeldin-A Sigma-Aldrich B5936 1,3-Cyclohexanediamine, N1,N3-bis(4,5-dihydro-1H-imidazol-2-yl)-4,6-bis[4-[(4,5-dihydro-1H-imidazol-2-yl)amino]phenoxy]-, (1R,3S,4S,6R)- MedChem Shortcut LLC 166830, 468 Guanidine, N,N'''-[(1R,3S,4S,6R)-4-[[4-(aminoiminomethyl)amino]-1-naphthalenyl] oxy]-6-[4[(aminoiminomethyl)amino]phenoxy]-1,3-cyclohexanediyl]bis-, rel- MedChem Shortcut LLC 166811, 468 1-(1,3-thiazolan-2-yl)pentane-1,2,3,4,5-pentaol ChemDiv 5408-0471, 469 Proprotein convertase inhibitor - Calbiochem Millipore 537076, 470 Critical Commercial Assays In Situ Hybridization Kit BioChain K2191050    Experimental Models: Cell Lines Human: EndoC-βH1 (passage 70) Dr. Raphael Scharfmann 101 Mouse: αTC-1 (passage 28) American Type Culture Collection 471 Mouse: αTCΔPC2 (passage 12) Dr. Don Steiner 472 Mouse: βTC-3 (passage 12) Dr. Shimon Efrat 473    Experimental Models: Organisms/Strains PC2-/- mouse pancreas (18-week-old) The Jackson Laboratory B6;129-Pcsk2tm1Dfs/J Dog pancreas (adult male) Dr. Alan Charrington Vanderbilt University Pig pancreas (4-year-old) Sinclair Bio Resources YucatanTM miniature swine Mouse pancreas (12-week-old) The Jackson Laboratory C57Bl/6J Rat Pancreas (8-week-old) UBC Center for Disease Modeling Wistar    Software and Algorithms MetaXpress® High-Content Image Acquisition & Analysis Software Molecular Devices Corporation https://www.moleculardevices.com/products/cellular-imaging-systems/acquisition-and-analysis-software/metaxpress Odyssey Infrared Imaging System Application Software Version 3.0.16 LI-COR Biosciences https://www.licor.com/ GraphPad Prism V7 GraphPad Software https://www.graphpad.com/scientific-software/prism/   157 Western blotting Groups of 250 mouse islets, 1000 human islets, or ~1 million EndoC-βH1 cells were lysed in 200 μL of lysis buffer (50 mM Tris pH8.0, 150 mM NaCl, 0.02% Na Azide, 0.1% SDS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 1 mM phenylmethyl sulfonyl fluoride, and protease inhibitor cocktail (Sigma-Aldrich, St Louis, MO)) as per a published protocol474. Levels of proteins of interest were assessed by fluorescent western blotting methods using two PC2 antibodies (1:1000, R&D Systems, MAB6018; 1:1000, Thermo Fischer, PA5-14594), anti-PC1/3 antibody (1:2500, gift from Lakshmi Devi), and an anti-α-tubulin antibody (1:1000, Sigma Aldrich) and detected on an Li-Cor Odyssey 9120 Imaging system (Li-Cor, Lincoln, NE). In situ hybridization For all in situ hybridization experiments, we used tissue samples collected and sectioned in all RNAse free solutions (including PFA, 70% ethanol, and water for sectioning) and cleaned all equipment with RNAse Away (ThermoFisher Scientific®, Waltham, MA). We performed a standard protocol using a modified version of a commercially available in situ hybridization kit (BioChain; Newark, CA). We used three probes specific for PCSK2 (Table 5-3) in regions highly conserved between human and mouse and utilized a primary mouse α-digoxigenin antibody and secondary goat α-mouse secondary antibody conjugated to alkaline phosphatase (Table 5-2). Quantitative PCR qPCR was performed using a standard protocol. mRNA expression level is represented as 2-ΔCt using the average of two internal control genes (GAPDH and ACTB). Primer sequences are detailed in Table 5-3.    158 Table 5-3 Oligonucleotide primers and probes used in qPCR and ISH experiments. Oligonucleotides 5'/DIG/GTG GTC ATG AAA GGC CAC TTG TCA AAG CCC ACC TTG GAG TCG TC -3' Integrated DNA Technologies PCSK2 probe A 5’/DIG/CTT TAT GCA ACT CCA CAA GAA AAT GAT TCG TGA AGA CSG GTC -3’ Integrated DNA Technologies PCSK2 probe B 5'/DIG/CAC CAT GGC ACC TGC ATC AAG GAC CCC GTA GCC AAA GAG GTG ATT AAA TTC CAG GCC -3’ Integrated DNA Technologies PCSK2 probe C 5’/AAG ATG GCT TTG CAG CAG GAA GGA Integrated DNA Technologies PCSK2 forward primer 5’/AGC CAC ATT CAA ATC AAG GCC AGG Integrated DNA Technologies PCSK2 reverse primer 5’/AAG CAA ACC CAA ATC TCA CCT GGC Integrated DNA Technologies PCSK1 forward primer 5’/TCA CCA TCA AGC CTG CTC CAT TCT Integrated DNA Technologies PCSK1 reverse primer 5’/GCT CTC TAC CTG GTG TGT GGG GAG C Integrated DNA Technologies Ins forward primer 5’/GGA GCA GAT GCT GGT GCA GCA CTG Integrated DNA Technologies Ins reverse primer 5’/G(W)G T(S)A ACG GAT TTG G(Y)C GT Integrated DNA Technologies GAPDH forward primer 5’/GAC AAG CTT CCC (R)TT CTC (R)G Integrated DNA Technologies GAPDH reverse primer 5’/GGA CTT CGA GCA AGA GAT GG Integrated DNA Technologies ACTB forward primer 5’/AGC ACT GTG TTG GCG TAC AG Integrated DNA Technologies ACTB reverse primer  Pulse-chase experiments After receiving islets, they were incubated overnight in complete media (CMRL, insulin-transferrin-sodium selenite (Corning, Corning, NY), GlutaMAX (ThermoFisher Scientific®, Waltham, MA), bovine serum albumin (Roche Diagnostics, Laval, Canada), and penicillin/streptomycin). The next day, islets were pre-incubated for 60 minutes in Krebs Ringer buffer with bicarbonate, HEPES, and 16.7 mM glucose (KRBH-16.7) at 37oC, pulsed for 40 minutes in 1mL of KRBH-16.7 with 800 μCi/mL of 3H-Leu (PerkinElmer, Waltham, MA), and chased for 120 minutes. Islets were split into groups of 1000 and chased in KRBH-2.8 in eight conditions: 37oC, 20oC, 37oC with 10 μg/mL Brefeldin A (BFA; Sigma-Aldrich, St Louis, MO),  159 37oC with 30 μM chloroquine, 37oC with one of two PC1/3 inhibitors (EMD Millipore, 537076; MedChem Shortcut LLC, 166811)468, 470, or 37oC with one of two PC2 inhibitors (MedChem Shortcut LLC, 166830; ChemDiv, 5408-0471)468, 469. Islets were lysed in 150 μL lysis buffer474. Immunoprecipitation, SDS-PAGE, and Scintillation counting Islet lysates were incubated overnight at 4oC with 5 μg of carrier-free anti-insulin antibody (C27C9, Cell Signalling, Danvers, MA) followed by 6-hour incubation at 4oC with 50 μL of protein-A/G magnetic beads (50% slurry in lysis buffer; ThermoFisher Scientific®, Waltham, MA). After washing three times, immunoprecipitated insulin was eluted by heating to 95oC in tris-tricine sample buffer (Bio-Rad Laboratories). We immunoprecipitated glucagon with the same protocol using an anti-glucagon antibody (EP3070, AbCam). Samples were then separated using tris-tricine PAGE approach to separate small proteins475. After separation, gels were stained by colloidal Coomassie476 and visualized mature and prohormone bands were excised manually. Gel slices were then dissolved in 0.3 mL of 30% hydrogen peroxide for three days at 50oC. Scintillation counting was performed on a LS 6500 scintillation system (Beckman, Brea, CA) with quench curve correction for final disintegrations per minute counts in Ultima Gold uLLT scintillation fluid (PerkinElmer, Waltham, MA). Counts were adjusted for background and number of leucines in prohormones compared to processed mature hormones477. Quantification and statistical analysis All statistical analyses were done on GraphPad Prism V7 and details of statistical tests performed are in Figure captions. We used p < 0.05 as the cut-off for significance.  5.3 Results Human β-cells were not immunoreactive for PC2 using four antibodies To minimize risk of non-specific findings, we performed antibody validation experiments. There were identical immunostaining patterns for PC2 using multiple antibodies in human and  160 mouse pancreas (Figure 5-1A). To negatively validate with a knockout model, we used an antibody that binds to the deleted N-terminal prodomain of PC2283. This antibody produced no immunoreactivity in pancreas from PC2 knockout mice (Figure 5-1B). By western blot two antibodies (R&D, MAB6018, red; Thermo-Fisher, PA5-14594, green) immunoreacted with recombinant human PC2 and produced a blot with the expected bands for proPC2 (75kDa) and mature PC2 (64kDa) from both human and mouse islet extracts (Figure 5-1C). Using four antibodies, we detected robust immunoreactivity for PC2 in mouse β-cells and neighbouring islet cells, but no immunoreactivity for PC2 in human β-cells (Figure 5-1D; see Figure 5-2A for single channels). Notably, though not detectable in human β-cells, we observed the expected robust PC2 immunoreactivity in the α-cells. These findings were replicated in four additional organ donors (Figures 5-3). We determined if there is a distinct distribution of PC2 in the head, body, and tail regions of human pancreas and whole mouse pancreas by immunostaining for insulin, glucagon, and PC2 (Figure 5-1E). Unlike mouse, in human pancreas the intensity of immunoreactivity for PC2 in β-cells is dramatically less than in α-cells and there is almost no colocalization of insulin and PC2 in all regions of human pancreas. The rate of colocalization of PC2 and insulin was not significantly different then the rate of colocalization of insulin and glucagon in the head, body, or tail of human pancreas (p > 0.05; data not shown). In all regions of human and mouse pancreas, there is high colocalization of PC2 with glucagon. Unlike in the mouse, PC1/3 and PC2 do not colocalize in human pancreas (Figure 5-2B). We immunostained for the essential PC2 cofactor neuroendocrine protein 7B2 (7B2; gene SCG5) and though there is immunoreactivity for 7B2 in human α-cells and β-cells, signal intensity is lower in β-cells, unlike in mice where there is similar immunoreactivity in both cell types (Figure 5-4).     161   162 Figure 5-1 Human β-cells were not immunoreactive for PC2 using four antibodies (A) Immunostaining for PC2 using three antibodies (see Table 5-2 for antibody details) in adult human and mouse pancreas. Representative images shown (n=3). Scale bar is 100 μm. (B) Immunostaining for insulin (green) and PC2 (red) in wild-type and PC2-/- mouse pancreas. Representative images shown (n=2). Scale bar is 100 μm. (C) Western blot for PC2 using two antibodies (R&D, MAB6018, red; N-terminal, Thermo PA5-14595, green) in human islet and mouse islet lysates with a recombinant human PC2 (rhPC2) control. (D) Immunostaining for insulin (green) and PC2 (red) using four PC2 antibodies. Representative images shown (n=5). Scale bar is 100 μm.  (E) Immunostaining for insulin, glucagon, and PC2C-term (see key resources table) in head, body, and tail of human pancreas and mouse pancreas (representative images of n = 3 shown on the left, scale bar is 100 μm, and insets are enlarged 4x). Quantification of the relative PC2 immunoreactivity intensity in α-cells versus β-cells (middle) and the rate of colocalization of PC2 with insulin or glucagon (right). Groups were compared by one-way ANOVA with Tukey post hoc test (** p < 0.01, *** p < 0.001)   163  Figure 5-2 Single channel immunostaining for insulin and PC2 using multiple antibodies and no colocalization of PC1/3 and PC2 in human pancreas. (A) Individual channels from Figure 1D. Immunostaining for insulin (green) and PC2 (red) using four antibodies (see Table 5-2 for antibody details) in adult human pancreas with nuclei labelled with DAPI (grey). Scale bar is 100 μm. (B) Immunostaining for PC2 (green) and PC1/3 (red) in adult human and mouse pancreas with nuclei labelled with DAPI (grey). Scale bar is 100 μm. Human pancreasDAPI PC2 PC1/3Mouse pancreasDAPI PC2 DAPI PC1/3PC2 R&D PC2 IgMPC2 C-termPC2 N-termAB 164  Figure 5-3 In multiple human pancreas donors PC2 did not colocalize with insulin and in multiple mouse pancreata PC2 colocalizes with insulin. Immunostaining for insulin (red) and PC2 (green) in adult human (A) and mouse (B) pancreas with nuclei labelled with DAPI (grey). Scale bars are 100 μm.  Human PancreasMouse PancreasAB 165  Figure 5-4 Human β-cells have less immunoreactivity for 7B2 than α-cells whereas mouse β-cells have comparable immunoreactivity to α-cells. (A) Immunostaining for 7B2 (red) and insulin (green) in human and mouse pancreas. Overlays shown with nuclei labelled with DAPI (grey). Representative image shown (n=4-5). Scale bar is 100 μm. (B) Immunostaining for 7B2 (red) and glucagon (green) in human and mouse pancreas. Overlays shown with nuclei labelled with DAPI (grey). Representative image shown (n=4-5). Scale bar is 100 μm. (C) Quantification of relative mean fluorescent intensity of 7B2 in Ins+ or Gcg+ cells. Data shown on box and whisker plots. *p < 0.05 (two-tailed t-test).  166 PC2 immunoreactivity was not detected in rat β-cells using three validated antibodies and EndoC-βH1 cells have PC2 We immunostained pancreas from pig, rat, and dog (Figure 5-5A). Unexpectedly, β-cells were not immunoreactive for PC2 in rat pancreas. β-cells in pig and dog pancreas were immunoreactive for PC2 using two antibodies, but one antibody failed to produce immunoreactivity in any islet cells of these species, likely due to incompatible epitopes (Table 5-4). This observation further supports the epitope specificity of these antibodies. We investigated the presence of PC2 in the human β-cell line EndoC-βH1101. We use the PC2 knockout αTC1ΔPC2 mouse α-cell line, the α-TC1 mouse α-cell line that has PC2 nearly exclusively, and β-TC3 mouse β-cell line with both PC2 and PC1/3 as controls. EndoC-βH1 cells are immunoreactive for both PC1/3 and PC2 (Figure 5-5B). Transcript levels of PCSK1 and PCSK2 are comparable in EndoC-βH1 cells, similar to human islets (Figure 5-5C). We investigated protein levels of PC1/3 and PC2 by western blots on EndoC-βH1 cells and human islets and confirm histological findings of comparable levels of PC1/3 and PC2 in EndoC-βH1 cells (Figure 5-5D).  Human β-cells have less PCSK2 than neighbouring α-cells, unlike mouse β-cells  To localize PCKS2 mRNA within pancreatic islets, we performed in situ hybridization on sections of human and mouse pancreas (Figure 5-6). With overlaying immunofluorescence for insulin and glucagon, intense PCSK2 signal within human islets is localized to α-cells with marginally detectable signal in β-cells (Figure 5-6A). This contrasts to the expected pan-islet signal in mouse islets (Figure 5-6B).    167  Figure 5-5 Rat β-cells were not immunoreactive for PC2 and EndoC- βH1 cells have PC2 (A) Representative immunofluorescence for insulin and PC2 using three antibodies in rat, pig, and dog pancreas (n = 1-3). One antibody produces no immunofluorescence in pig pancreas (PC2 C-terminal, PA1-058) or dog pancreas (PC2 N-terminal, PA5-14595) likely due to incompatible epitopes (see Table 5-2 for antibody details and Table 5-4 for PC2 sequence alignment). Scale bar is 100 μm.  168 (B) Immunostaining for PC1/3 and PC2 in human EndoC-βH1, PC2 knockout mouse αTCΔPC2, PC2 expressing mouse αTC-1, and mouse βTC-3 cell lines. Scale bar is 50 μm. (C) Relative expression of PCSK1 and PCSK2 in EndoC-βH1 cells and whole human islets by qPCR. 2-ΔCt presented in box and whisker plots with GAPDH and ACTB as reference genes. (D) Western blot for PC2 (R&D, MAB6018, red; N-terminal, PA5-14595, green) and PC1/3 (Gift from Dr. Lakshmi Devi, AB_2665530) in human islet and EndoC-βH1 cell lysates with a rhPC2 control and with α-tubulin (α-tub; Sigma-Aldrich, T9026) used as a loading control. Quantification presented on the right.    Table 5-4 Alignment of PC2 sequences for human, mouse, rat, pig, and dog Residues 94-116 and 622-638 from PC2 corresponding to the published epitopes of two commercially available PC2 antibodies. Sequences highlighted in yellow produced no immunoreactivity in pancreas. Residues in purple are conservative and residues in red are non-conservative. Proposed epitopes are shown.   N-terminal PC2 antibody (Thermo Fisher Scientific, PA5-14595, RRID#: AB_2299126): Human R V K M A L Q Q E G F D R K K R G Y R D I N E Mouse R I K M A L Q Q E G F D R K K R G Y R D I N E Rat R V K M A L Q Q E G F D R K K R G Y R D I N E Pig R V K R A L Q Q E G F D R K K R G Y R D I N E Dog R V K M A L Q Q E G F N R K K R G Y R D I N E     C-terminal PC2 antibody (Thermo Fisher Scientific, PA1-058, RRID#: AB_2158593): Human L D E A  V E R S L K S I L N K N Mouse L D E A  V E R S L Q S I L R K N Rat L D E A  V E R S L Q S I L R K N Pig L D E A  V E R S L K S I L G K H Dog L D E A  V E R S L K S I L R K N     169  Figure 5-6 Human β-cells have less PCSK2 than neighbouring α-cells, unlike mouse β-cells (A-B) PCSK2 detected in human and mouse pancreas by in situ hybridization using digoxigenin labelled probes and detected using alkaline phosphatase conjugated antibody with NBT/BCIP substrate. After ISH, sections were immunostained for insulin (red) and glucagon (green) and overlaid with ISH image. Representative images shown (n=3). Scale bar is 100 μm. Insets are enlarged 4x.    170 PC2 does not play a significant role in the processing of proinsulin in human β-cells Given that PC2 is ~80-100x more catalytically active than PC1/3478, we sought to validate histological findings with sensitive biochemical approaches to assess whether human β-cells have low yet catalytically relevant levels of PC2. Given that neighbouring α-cells within human islets require PC2 to produce glucagon from proglucagon479, we assessed both proinsulin processing as well as proglucagon processing in pulse-chase experiments. Importantly, we choose to study primary human islets in order to specifically answer the question as to how human β-cells process proinsulin. It seems that PC2 can access and cleave human proinsulin347, but studying the ability of PC2 to process human proinsulin in models like EndoC-βH1 cells or other cell lines would not clarify the important question of how proinsulin is processed in bona fide human β-cells. In human islets, there was a near total blockade of glucagon production by all PC2 inhibiting chase conditions (Figure 5-7A) but no significant difference in proinsulin processing (Figure 5-7B). To determine if PC1/3 is predominantly responsible for proinsulin processing in human β-cells, we employed two PC1/3 inhibitors. Neither inhibitor impaired the processing of proglucagon, but both led to a significant inhibition of proinsulin processing.     171 Figure 5-7 Human β-cells can process proinsulin without PC2 function but require the function of PC1/3, unlike PC2 dependent α-cells (A) Proportion of 3H labelled processed proglucagon into glucagon in human islet lysates after 40-minute pulse in 800 μCi/mL 3H-Leu and 120-minute chase in various conditions. Specific organ donors are designated by unique symbols. Groups were compared by one-way ANOVA with Tukey post hoc test (*** p < 0.001). (B) Proportion of 3H labelled processed proinsulin into insulin in human islet lysates after 40-minute pulse in 800 μCi/mL 3H-Leu and 120-minute chase in various conditions. Specific organ donors are designated by unique symbols. Groups were compared by one-way ANOVA with Tukey post hoc test (* p < 0.05). We made the surprising observation of absent PC2 immunoreactivity in rat β-cells  172 5.4 Discussion The prevailing theory that proinsulin processing requires both PC1/3 and PC2 is best supported by studies of islets from PC2 and PC1/3 knockout mice279, 313. To our knowledge, the roles of PC1/3 and PC2 for processing human proinsulin has never been rigorously examined. We performed immunohistofluorescence with well validated antibodies and made the surprising observation of virtually absent PC2 immunoreactivity in both human and rat β-cells. Interestingly, others have observed minimal PC2 immunoreactivity in human β-cells480 and in one case, interpreted the finding as a sign of β-cell dysfunction after transplantation into immunodeficient rodents322. Yet others have reported abundant PC2 immunoreactivity in human β-cells481, but that work used a polyclonal antibody482 with an immunizing peptide containing significant homology between human PC1/3 (RRDELEE) and PC2 (KKEELEE), raising the possibility of significant cross-reactivity with PC1/3. Likewise, prior detection of PC2 immunoreactivity in rat β-cells by immunohistochemistry483 and immunogold electron microscopy484 relied on polyclonal antibodies generated with large PC2 immunogens containing regions of high homology to PC1/3 (PC2: TNACEGKEN vs PC1/3: TRACEGQEN483 or identical sequences FALALEAN and LTWRDMQHL in both PC2 and PC1/3484). By western blot using an antisera to a large fragment of PC2 (D174-S384) containing substantial regions of homology to PC1/3, rat insulinoma granules have immunoreactivity for PC2485. However, using a specific antibody with a similar immunogen to one of the antibodies we use in the current study (PC2C-term) there seems to be little PC2 in rat β-cells at ~95% purity486. Given that rat insulin 2 is processed first at the C-A junction487, a site favorable for PC1/3 (RQKR)488, it supports the theory that PC1/3 could be exclusively responsible for proinsulin processing in rat, though does not rule out the possibility that M at the P4 position (using the system of 287 for denoting positions prior to (Px) or after (P’x) the scissile bond) is the cause for slower B-C junction processing. Additionally, only PC1/3 is regulated by high glucose in rat islets277, suggesting that  173 rat β-cells may not depend on the function of PC2. The processing of proinsulin in rat β-cells is worth studying in greater detail in the future and rat islets may be an excellent model for studying human proinsulin processing. Unlike in human and rat pancreas, abundant PC2 immunoreactivity was evident in β-cells from mouse, pig, and dog pancreas. In these three species, the CA junction of proinsulin has PC1/3 unfavourable amino acids at the P4 position (mINS2: Q, pigINS: P, dogINS: L)489, while both rat insulins have a PC1/3 favourable R488. Mouse pro-islet amyloid polypeptide (proIAPP) requires PC2 to process its N-terminal site448. Mouse, dog, and pig N-terminal proIAPP all have the amino acid M in the P4 position. Contrarily, both rat and human have a V at P4. We propose that the P4 M at the N-terminal processing site of proIAPP is unfavorable for processing by PC1/3 and a contributor to the species differences in β-cell PC2 expression. This is supported by observations that rat INS2 with a P4 M at the BC junction of proinsulin is unique among rodent insulins in that it is processed at the CA junction before the BC junction315 perhaps because rat β-cells lack PC2 and PC1/3 is slow to process the BC junction. Additionally, in PC2-/- mice that possess a human IAPP transgene, there is no increase in the ratio of circulating NH2-proIAPP1-48 to IAPP and no increased amyloid formation490 suggesting no defect in human proIAPP processing in the absence of PC2. Taken together, our findings of the species differences of PC2 immunoreactivity in β-cells suggest that PC2 is essential for the processing of N-terminal proIAPP when there is a P4 M. In addition to examining protein levels of PC2, we assessed mRNA levels of PCSK2 in human pancreas. Our detection of low level PCSK2 in human β-cells by in situ hybridization aligns with single cell RNAseq experiments319, 320 reporting detectable PCSK2 in human β-cells, albeit at much lower levels than in α-cells, unlike in mice where PCSK2 is abundant in both α-cells and β-cells491. Occasional intense PCSK2 signals in non-α-cells within human islets are likely δ-cells, given the known role for PC2 in excising somatostatin-14283. Detectable PCSK2  174 mRNA in conjunction with a lack of PC2 immunoreactivity suggest that there is post-transcriptional regulation of PCSK2 in human β-cells, a possibility warranting follow-up. We also detected immunoreactivity for the essential PC2 cofactor 7B2 (gene: SCG5) in human β-cells, albeit at lower intensity than in human α-cells. These results are not surprising given high SCG5 mRNA levels in sorted human β-cells318. Though all brain PCSK2+ cells are also SCG5+, there are many SCG5+ cells that lack PCSK2304. 7B2 is known to suppress aggregation of β-amyloid and synuclein306 and could have other roles in human β-cells.  It is notable that convertase expression can be regulated. In rodent models of hyperglycemia, there can be a compensatory upregulation of PC1/3 in α-cells458, 459 which can increase intra-islet production of GLP-1 and reduce rates of apoptosis in β-cells460, 492. PC2 deficient mouse α-cells used to generate the αTC1ΔPC2 cells spontaneously produce PC1/3 and begin excising GLP-1 from proglucagon461. Intestinal L-cells of PC1/3-/- mice abnormally produce mature glucagon, likely attributable to PC2 action493. Potentially, a similar initiation of PC2 expression occurs in β-cells of humans with PCSK1 mutations who present with hyperglycemia and elevated circulating des-64,65 proinsulin328. Additionally, expression of PC2 in the human EndoC-βH1 cell line could reflect immaturity of the cells as PC2 is expressed during fetal development of early endocrine cells494. Whether there may be dynamic regulation of PC1/3 and PC2 in human β-cells under normal or pathological conditions warrants further investigation. By pulse-chase, human β-cells did not have significant impairment of proinsulin processing in conditions capable of blocking the PC2 dependent processing of proglucagon to glucagon. Based on our immunostaining experiments, if β-cells do contain any PC2, it is far less than the PC2 content of α-cells, thus making blockade of proglucagon processing a stringent positive control of PC2 inhibition. It is unsurprising that temperature blockade at 20oC and brefeldin-A had variable blockade in the processing of proinsulin because both block shuttling of  175 proinsulin into secretory granules and temperature blockade has been demonstrated to partially supress proneuropeptide Y processing by PC1/3, albeit to a lesser degree than by PC2 495. Additionally, given that we could only confirm robust PC2 inhibition based on blockade of proglucagon processing but cannot confirm a lack of PC1/3 suppression, non-significant yet numerical decreases in proinsulin processing in PC2 inhibited conditions may be best attributed to modest reductions in PC1/3 activity. Both PC2 inhibitors used in pulse-chase experiments function as allosteric or mixed inhibitors and have been shown to inhibit PC1/3 by 10-20% when employed in cell-free experiments at lower concentrations (10 μM 468 and 25 μM 469). Notably, β-cells are sensitive to PC1/3 inhibition as PC1/3 heterozygous mouse β-cells with likely a ≤50% loss of PC1/3 (and abundant PC2) have modestly impaired proinsulin processing279. In contrast to PC2 inhibited conditions, PC1/3 inhibited conditions that did not impair proglucagon processing, significantly impaired proinsulin processing. We cannot confirm the extent of PC1/3 inhibition and incomplete PC1/3 inhibition is a potential explanation for the incomplete blockade of proinsulin processing. Regardless, these findings provide validation for our experimental technique and clearly show that PC1/3 is important to the processing of proinsulin.  The pulse-chase study design was limited by a lack of a molecular knockdown approach for PCSK2 (e.g. lentiviral delivery of shRNA)496. Though the specificity of a molecular approach is appealing and may be a worthwhile follow-up study, the confounding effects of islet dispersion, cell reaggregation, and prolonged culture to allow time for mRNA knockdown and PC2 protein turnover, are considerable. Another important limitation is that as we were unable to differentiate intact proinsulin from proinsulin processing intermediates using size-based proinsulin/insulin separation by SDS-PAGE, we were unable to rule out the potential that some single-site processing of proinsulin is occurring in PC1/3 inhibited conditions. Given relatively normal production of fully mature insulin (necessarily processed at both the B-C and C-A junctions to liberate C-peptide and reduce the molecular mass to ~6kDa) in PC2 inhibited  176 conditions that were sufficient for a near 100% blockade proglucagon processing, this appears unlikely. Impaired proinsulin processing has prognostic value for progression from autoantibody positivity to type 1 diabetes332 and from impaired glucose tolerance to type 2 diabetes341. Some patients with type 1 diabetes lose detectable circulating C-peptide337 but almost all retain detectable proinsulin338 and some have detectable proIAPP339. Type 1 diabetes is associated with a reduction of islet PCSK1340 and understanding normal processing of prohormones in healthy β-cells is important to understand the contribution of PCSK1 to defective prohormone processing during diabetes. The current work challenges the prevailing assumption that human β-cells process proinsulin by PC1/3 then PC2 sequentially. Though evidence supports this dogma in mouse β-cells, our findings suggest PC2 has little to no role in human β-cells.   177 Chapter 6: Conclusions, future directions, and major challenges 6.1 Research summary and conclusions  Though insulin injections have proven lifesaving for an uncountable number of patients, insulin and almost all currently available therapies for diabetes continue to fail to address the key underlying pathology: a loss of mature, functional β-cells. Whether faced with a near complete loss of β-cells in type 1 diabetes, a relative insulin insufficiency and eventual β-cell dedifferentiation and death in type 2 diabetes, or a genetic defect in monogenic diabetes, all patients with diabetes lack adequate β-cell number or function. Replacement, genetic repair, or (re)generation of β-cells could offer the near flawless glycemic regulation that those without diabetes take for granted. With the precedent set by clinical islet transplantation, the need for abundantly available therapeutic strategies for β-cell replacement or regeneration is paramount. Continued basic research is important to advance our understanding of potential therapies currently failing to meet minimum standards to enter the clinic. In the current thesis we undertook a broad spectrum of research with a central focus on insulin as a key factor involved in β-cell function and maturation.  First, we performed a rigorous characterization of β-cells in insulin knockout mice. Research has identified many key factors involved in pancreas and β-cell development and maturation, but there has been limited investigation into the role of insulin itself. Past research used a β-cell selective IR knockout model355, 360, an approach faced with many limitations including incomplete β-cell recombination, delayed loss of IR until further into development, recombination in the brain, GH minigene inclusion in transgenes360, impairment of IGF2 signaling via the IR, and no blockade of insulin action via IGF receptors. By studying insulin knockout mice, we avoided these limitations and provide the best to date available insight into the role of insulin on mouse β-cell development. At birth, insulin knockout mice have expanded β-cell mass, but β-cells have a dedifferentiated phenotype. Insulin  178 replacement by injection facilitated a partial maturation of endogenous insulin deficient β-cells whereas long-term insulin replacement by islet transplantation better supported maturation. These findings provide evidence that insulin is a necessary signaling molecule for the maturation of β-cells and suggest that a lack of insulin signaling contributes to β-cell dedifferentiation. In follow-up to the work of Chapter 2, in Chapter 4 we present our investigations on the viability of a gene therapy approach to be curative for this mouse model of patients with monogenic diabetes.  Before assessing a gene therapy approach to cure the insulin knockout mouse model of PND, we first undertook the development of a novel genetic tool for in vivo genetic editing. In vivo genetic editing has been most commonly studied using Cre driver mice and enables the study of the role of specific genes on both cellular and physiological outcomes. To avoid caveats and costs associated with use of Cre driver transgenic mice, we designed and characterized AAV Ins1-Cre. As expected, without delivery of AAV there was zero recombination and induction of recombination did not require delivery of tamoxifen; both of which are challenges faced by the inducible CreER approach. Additionally, when considering cage costs, genotyping costs, and labor costs, AAV Ins1-Cre was produced more quickly and for a lower cost than breeding multiple transgenic Cre driver mice with floxed target genes. Though AAV Ins1-Cre produced efficient recombination in β-cells, there was off-target recombination in the liver, exocrine pancreas, non-β islet cells, and in regions of the hypothalamus. Given the use of a relatively short regulatory sequence, a tiny amount of leaky expression in off-target tissues is conceivable. Given the binary outcome of recombination or no recombination, apparent comparable signal intensity in cells with leaky off-target expression is expected. Nonetheless, when targeting floxed β-cell specific genes, these caveats will not be impactful because liver and exocrine cells would not be affected by the loss of a gene that they do not express. To demonstrate this, we induced hyperglycemia in Ins1-/-Ins2f/f mice by  179 delivering the AAV Ins1-Cre IP. We propose that with prudent consideration of potential implications of off-target recombination, this AAV will be useful for other studies, including those presented in Chapter 4.  With successful characterization of insulin deficient β-cells and characterization of the AAV Ins1-Cre for in vivo genetic editing, we sought to investigate the viability of a gene therapy cure for the Ins1-/-Ins2-/- mouse model of PND. Though AAV Ins1-INS delivered human insulin to β-cells, when delivered to Ins1-/-Ins2-/- pups it only produced a transient reversal of diabetes and led to circulating hyperproinsulinemia. We theorized that though pups were initially highly insulin sensitive allowing low-bioactivity proinsulin to lower blood glucose, reduced insulin sensitivity after exposure to a new cell source of insulin, increased insulin demands from animal growth, and likely increased growth hormone levels as animals enter the pubertal period, drove relapse to diabetes. To attempt to address elevated circulating proinsulin, we delivered an AAV carrying native a mouse insulin (AAV Ins1-Ins1) to Ins1-/-Ins2-/- pups. AAV Ins1-Ins1 failed to produce any diabetic remission due to low β-cell infectivity and unresolved poor proinsulin processing. We present data of elevated circulating proinsulin, cytoplasmic prohormone immunoreactivity in β-cells, and a lack of mature insulin granules in infected β-cells to repeatedly confirm impaired proinsulin processing in infected β-cells. We conclude that Ins1-/-Ins2-/- β-cells have a severe prohormone defect leading to a buildup of incompletely processed proIAPP and proinsulin after infection by AAV. Operating under the theory that the AAV was effectively delivering insulin to β-cells, but insulin deficient β-cells developed an irreversible prohormone processing defect, we generated an adult inducible insulin knockout model. This approach avoided any developmental effects of insulin deficiency. Co-delivery of AAV Ins1-Ins1 with the AAV Ins1-Cre was unable to prevent onset of adult insulin gene loss induced diabetes and warrants follow-up that is discussed in detail in section 6.3.4. Taken together, these findings highlight the necessary role for insulin on β- 180 cell maturation. Our results provide disappointing insight into the pursuit of a gene therapy-based treatment for PND caused by insulin mutations. Such approaches may face previously unforeseen challenges of enduring β-cell defects even after replacement or correction of mutated genes.  In the final experiments of the thesis, we explored proinsulin processing in human β-cells. Based on decades of research using knockout mouse models and in vitro experiments, current theory posits that proinsulin is processed by PC1/3 and PC2 sequentially to excise C-peptide and yield mature insulin. It has been assumed that proinsulin processing in mice recapitulates the process in humans. In the current work, we interrogated this assumption and revealed that human β-cells do not rely on PC2. We showed a lack of PC2 immunoreactivity in human β-cells using well validated antibodies and little PCSK2 by in situ hybridization. We confirmed these findings by sensitive biochemical technique. Using the ideal cell population of genuine primary human islets, none of five PC2 suppressing interventions impaired production of mature insulin despite fully blocking processing of proglucagon to glucagon; a PC2 dependent process. Contrarily, suppression of PC1/3 using two inhibitors impaired proinsulin processing with no significant impact on proglucagon processing. Based on work using careful and rigorous experimental techniques and multiple experimental conditions, we suggest that there is a need to revise the longstanding theory of proinsulin processing.  6.2 Research limitations The work presented in this thesis has various caveats discussed within each chapter. These include limitations of methodology, model systems, and experimental approaches. Here, we discuss three general limitations. The first is the use of immunohistofluorescent techniques in every chapter. Unlike other common techniques like qPCR, western blotting, or  181 RNA sequencing, immunohistofluorescent investigation provides insight on immunoreactivity for specific proteins on a cellular and subcellular basis. Additionally, unlike other single cell approaches like single cell qPCR or RNA sequencing, histological approaches assess protein levels rather than mRNA levels. The correlation between mRNA and protein is often extremely low497. However, it is important to acknowledge that immunostaining is limited by offering only semi-quantitative results as well as usually relying on commercial antibodies. Though rigorous antibody validation in Chapter 5 essentially rules-out the possibility of non-specific immunoreactivity driving non-specific findings, this is not always possible or practical. For example, this was not possible for the research presented in Chapter 2 because twenty-nine different primary antibodies were used, and other antibodies or knockout mouse models are not always available. Though some were better validated (e.g. NGN3 antibodies highlighted in Figure 2-7), the majority relied on the presence of expected positive signal in the correct cells and cellular compartment in control animals as validation. Also, histological examination of tissue provides only a snapshot in time. In other words, histological approaches collect images that are static representations of a dynamic organism. This limits the ability to make mechanistic observations. Without effective lineage tracing models, it is not possible to conclusively determine cell origin and as is almost always the case with histological approaches, adaptive cell identity or migration is not clearly examinable. Notwithstanding these limitations, adequate sample size, highly optimized repeatable experimental technique, and redundancy of investigation (e.g. performing qPCR or ISH to complement histological experiments) limits the potential of non-specific errors. Additionally, we have avoided unjustified mechanistic conclusions in our interpretation of our findings. The second limitation of the experiments presented in this dissertation is our reliance on mouse model organisms in many experiments. Investigations in Chapters 2 to 4 use transgenic mouse models. Though mouse models are a standard in the field, there are  182 several limitations in extrapolating findings to humans. For example, the role of insulin during β-cell development examined in Chapter 2 may not recapitulate human pathology. Additionally, though the Ins1-/-Ins2-/- mouse is a useful model of PND caused by insulin mutations, an important limitation is that these animals fail to produce any insulin. In contrast, most patients with monogenic diabetes caused by mutations in INS produce misfolded, aggregative, or reduced bioactivity insulin498. Use of multiple model organisms including the insulin deficient pig model499 could complement findings and provide greater insight into the translatability of findings to humans. The third limitation we acknowledge is the use of small sample sizes that limit the conclusions that can be drawn in some experiments. In Chapter 2, Figures 2-4 to 2-8, 2-11, and 2-14 present representative images of n = 3 and associated quantification of histology. In an effort to be transparent and rigorous, we performed multiple analyses of each immunostain and examined multiple markers of β-cell maturity or dedifferentiation. Given the highly consistent nature of the findings and our use of non-parametric statistical methods when data failed to meet assumptions of normality, risk of type 1 error is reduced. More critically, data in Figure 2-13 has only an n = 1 for the long-term insulin injection treated Ins1-/-Ins2-/- group. We urge conservatism in interpreting these findings but felt that given the extreme challenge to ever replicate this experiment, sharing these data was worthwhile. In Chapters 3-4 we aimed to use larger sample sizes for histological experiments but in some cases sample sizes were 3. In Chapter 5 sample sizes were also modest (usually n = 4-5) but we perform multiple experiments that yield consistent results that repeatedly support the central finding. We analyzed data with non-parametric statistics when appropriate. We hope that full transparency of presenting individual data points in almost every figure will aid readers in being able to assess findings and reach justifiable conclusions.   183 6.3 Future Directions 6.3.1 The role of insulin during human β-cell development In an effort to improve the human translatability of this research, using human model systems is an important area of future direction. Extending on our work published on insulin knockout mice, in December 2018 a paper was published showing results of experiments where authors attempted in vitro differentiation of induced pluripotent stem cells from a human donor with PND resulting from homozygous mutations at the INS gene transcriptional start site500. Interestingly, both before and after CRISPR/Cas9 technology mediated repair they report successful in vitro differentiation to “β-like cells” with nuclear immunoreactivity for NKX6.1 and PDX1. These findings suggest that defects in the neonatal mouse Ins1-/-Ins2-/- β-cell phenotype are either unique to mice but not humans or are a result of non-cell-autonomous factors. An ethically controversial experiment to clarify could involve taking advantage of the chimera-forming ability of complement organogenesis in organ disabled host animals in vivo501. Specifically, incorporation of INS-/- human PSCs into Pdx1-/- mouse blastocysts would allow assessment of the development of insulin mutant human β-cells in the context of a human pancreas within a developing embryo, rather than in a highly controlled in vitro system. Less ethically fraught models like in utero injection of partially differentiated INS-/- human embryonic stem cell derived definitive endoderm or pancreatic progenitors into developing mouse embryos could be complementary502. These experiments would take advantage of assessing human β-cell development in the context of normal mammalian fetal and post-natal growth periods. 6.3.2 Investigating the impact of insulin therapy on β-cell maturity The results presented in Chapter 2 of this thesis provide insight on the reversibility of a dedifferentiated β-cell phenotype after initiation of intensive insulin therapy. In follow-up, we propose assessing the impact of insulin replacement therapy on the endogenous β-cells during  184 the pathogenesis of type 2 diabetes. Like mouse β-cells, adult human β-cells also contain much of the necessary machinery for insulin signaling including insulin receptors and insulin receptor substrate-1 and -2 (IRS-1 and -2)503, suggesting that insulin signaling may be important to human β-cells. Functionally, suppression of IR levels in human islets by shRNA impaired glucose stimulated insulin secretion503 and human islets with an IRS-1 polymorphism had impaired glucose stimulated insulin secretion, elevated proinsulin secretion, and reduced insulin content504. Further, there is evidence of reduced insulin signaling in β-cells of patients with type 2 diabetes: IRS1, IRS2, Tyr612 IRS1, and Tyr612 IRS2 immunoreactivity is reduced in islets from patient with type 2 diabetes compared to islets from non-diabetic donors172. Taken together, these findings lead us to propose that though often attributed to reduced glucolipotoxicity and decreased endogenous insulin demands505, diabetic remission following intensive insulin therapy in patients with type 2 diabetes506 may be partially attributed to augmented β-cell insulin signaling from increased circulating insulin. Increased insulin signaling in β-cells could lead to redifferentiation of dedifferentiated β-cells and a virtuous cycle of increased insulin secretion by β-cells. Though increased insulin signaling was inadequate to resolve severe prohormone processing defects in Ins1-/-Ins2-/- β-cells, β-cells did appear more mature based on loss of immunoreactivity for progenitor markers and perhaps increased insulin signaling could resolve mildly impaired processing in early type 2 diabetes. Future investigation could involve histological comparisons of pancreas from donors with recent onset type 2 diabetes, to those that have begun insulin therapy with and without remission. Examination of proinsulin processing and markers of β-cell maturity or dedifferentiation would be informative and if feasible, pairing histological experiments with functional assessment of β-cells by perifusion experimental design would be extremely useful. As revealed in the current thesis (e.g. experiments presented in Figures 4-4 to 4-6), so-called markers of β-cell maturity like MAFA may not be specific for functionally mature β-cells.  185 We make the observation that replacing insulin by injections alone or islet transplant resulted in dramatically divergent outcomes for the endogenous β-cells. We propose three variables that could contribute to the differences in β-cell phenotype: 1) superior glycemic control in mice treated with islet transplantation relative to those treated with insulin injections, 2) native mouse insulin produced by transplanted islets may signal in β-cells with higher bioactivity than recombinant insulin, or 3) insulin deficient islets may fail to produce other essential factors that are replaced by transplanted islets but not exogenous insulin therapy. Follow-up investigations to clarify would be insightful when considering early management approaches to maximize the likelihood of diabetes remission for patients with prediabetes and early type 2 diabetes. To address variable 1, we propose attempting to establish more rigorous insulin therapy protocols for insulin deficient mice and assess the impact of better glucose control on β-cell maturity. To address variable 2, we propose comparing replacement of recombinant human insulin with mouse insulin. To begin to address variable 3, a proteomics assay of islet protein content comparing Ins1-/-Ins2-/- and control islets could direct exploration on which specific molecules are deficient in Ins1-/-Ins2-/- islets, including known lost β-cell factors like C-peptide. Additionally, in vitro experiments using isolated Ins1-/-Ins2-/- islets from neonates could address these questions. Islets could be cultured in low or high glucose media, cultured with or without human or mouse insulin in the media, and with or without wildtype islets in the well. Together, these experiments would provide insight into the mechanism underlying the divergent β-cell phenotype in insulin treated Ins1-/-Ins2-/- mice. 6.3.3 Further investigating the observed prohormone processing defect in insulin deficient Ins1-/-Ins2-/- β-cells In Chapter 4 we investigated the feasibility of reversing the Ins1-/-Ins2-/- mouse model of PND. Disappointingly, despite feasibility of returning an insulin gene to endogenous insulin deficient β-cells, therapeutic benefit was hindered by impaired proinsulin processing. Though  186 the precise pathological mechanism underlying this defect was not clarified by avoiding developmental effects by studying a model of adult insulin knockout (delivering a failed batch of AAV Ins1-Ins1 plus AAV Ins1-Cre to Ins1-/-Ins2f/f mice), multiple experimental techniques (immunostaining for both proinsulin and proIAPP, transmission electron microscopy, and assessing circulating proinsulin by commercial ELISA) repeatedly suggest a prohormone processing defect in infected insulin knockout β-cells. Given the implications of these findings for the viability of a gene therapy approach for patients with PND caused by INS mutations, attempting to clarify the mechanism underlying the β-cell prohormone processing defect is worthwhile. Given that normal prohormone processing is a complex pathway that could require (a non-exhaustive list) signal peptide cleavage, disulfide bond formation, glycosylation, proteolysis, amidation, granule acidification, calcium pumps, and chaperone proteins507, a large-scale screen may be worthwhile. For example, a focused single-cell mass spectrometry experiment508 or a transcriptomics or proteomics screen could be informative. A single-cell approach would allow comparison of AAV Ins1-Ins1 infected and uninfected Ins1-/-Ins2-/- β-cells with littermate and wild-type controls. 6.3.4 Further investigating gene therapy approaches for both insulin knockout PND as well as other models of monogenic diabetes In addition to exploring the proinsulin processing defect of Ins1-/-Ins2-/- mice, we also propose reattempting an insulin gene therapy approach for an adult inducible insulin knockout mouse model. Though the AAV Ins1-Cre was rapid to implement and when delivered to Ins1-/-Ins2f/f mice was an affordable approach to generate adult inducible insulin knockout mice, we suggest using an AAV independent approach. Given that high dose AAV Ins1-Cre impaired β-cell function (3x1012 VGP; Chapter 3), the high combined dose of both AAV Ins1-Ins1 plus AAV Ins1-Cre (total 4x1012 VGP) could have impaired the function of infected β-cells. Using the well-studied Ins1-/-Ins2f/fmTmGPdx-CreER mouse model430 we could deliver a lower total  187 dose of AAV and assess a reversal of diabetes rather than solely attempting to prevent diabetes onset. This approach would allow us to determine if developmental insulin insufficiency was the distal cause of adult β-cell dysfunction. Furthermore, this model would allow for careful examination of the efficacy of therapy after different durations of diabetes. Given the observed dedifferentiation phenotype that arises in insulin deficient β-cells442, assessing if the gene therapy approach is effective if delivered before, during, or after genomic insulin gene deletion or diabetes onset would provide insight into the capability of dedifferentiated β-cells to redifferentiate after replacement of an essential gene. Should this approach again fail to reverse or even prevent diabetes despite use of a validated batch of AAV (we are investigating a new manufacturer), we would need to evoke explanations not involving AAV toxicity to explain failed prevention of diabetes onset after co-delivery of AAV Ins1-Ins1 with AAV Ins1-Cre. Inadequate infection rate is a viable explanation but given that the AAV appeared to be an adequate titer for ~50% β-cell infection rate, this may not fully explain our findings of failed protection from diabetes. For example, it is possible that there could have been a rapid cell-autonomous β-cell defect upon loss of all endogenous insulin genes that is not corrected by delivering an extragenomic insulin gene by AAV. Another important follow-up question is assessing the viability of a gene therapy approach to correct or replace other mutations causative for monogenic diabetes. For example, though in vivo genetic editing is in its infancy, with future progress improving efficiency of genetic repair509, patients with activating mutations in the ATP-sensitive potassium channel may be candidates for corrective AAV-CRISPR/Cas9 mediated gene repair463. These patients are less likely to become insulin independent with sulphonylurea therapy the longer the patient has been insulin dependent410. Impaired insulin signaling before sulphonylurea therapy may contribute to impaired diabetic remission and examining the maturation status of β-cells from donors may be worthwhile. When investigating the viability of an AAV mediated correction of a  188 model of KATP gain-of-function mutations34, assessing the impact of disease duration and management strategies on the success of the AAV therapy would be important. Repairing pathogenic mutations in HNF1A would also be worth studying because it is the most common cause of MODY, and recent evidence suggests that abundant dysfunctional β-cells survive in these patients510. 6.3.5 Assessing partial proinsulin processing by PC2 in human β-cells In Chapter 5, we assessed proinsulin processing in human islets in pulse-chase experiments. Our method of detection relied on the size differential of proinsulin and all proinsulin intermediates at ~9kDa versus mature insulin at ~6kDa, thus limiting our ability to determine if any single site processing occurred. In follow-up, pulsing islets with non-radioactive 2H-Leu would allow for mass spectrometry to be utilized as the outcome measure. Mass spectrometry would provide superior resolution to allow us to determine what proportion of proinsulin molecules are processed at the B-C or C-A junction (with or without CPE mediated cleavage of C-terminal basic amino acids). It is possible that though in the presence of PC1/3 inhibitors human islets are unable to normally produce mature insulin, there is still some processing of proinsulin at either the B-C or C-A junction. Determining if proinsulin intermediates are produced in PC1/3 inhibited conditions would allow us to determine if PC2 is playing a minor role in the processing of proinsulin but is unable to meaningfully produce mature insulin without the function of PC1/3. Notably, highly efficient PC2 inhibition by five techniques that could fully block proglucagon processing did not significantly inhibit the production of mature insulin. We suggest that this observation paired with the likelihood that PC1/3 inhibitors were likely <100% efficient, meaningfully diminishes the possibility that PC2 is playing a role in proinsulin processing in human β-cells.  189 6.3.6 Characterizing the role of β-cell prohormone processing defects during diabetes pathogenesis Defective proinsulin processing has been implicated in the pathogenesis of both type 1 and type 2 diabetes and is both associated with and predictive of the disease. Patients with type 1 diabetes who lose detectable circulating C-peptide can retain detectable proinsulin337 and some research suggests a similar defect of processing may occur for proIAPP339. Notably, reduced expression of PC1/3 in islets of patients with type 1 diabetes could be a driver for impaired processing340. We propose investigating whether PC2 plays a role in processing proinsulin during diabetes pathogenesis or during times of β-cell stress. This hypothesis is rooted in our findings from Chapter 5 that human β-cells appear to be fully reliant on PC1/3 for proinsulin processing in times of normal physiology paired with past work suggesting that endocrine cells are capable of “switching” between PC1/3 and PC2. Studies on mouse models of PC1/3 or PC2 deletion reveal that after PC2 knockout αTC1-ΔPC2 cells produce PC1/3461 instead of PC2 and in PC1/3-/- mice their intestinal L cells abnormally produce mature glucagon, likely attributable to PC2 action493. Additionally, such an α-cell “switch” to PC1/3 has been reported in models of β-cell stress including pregnancy and leptin deficient ob/ob mice458. Thus, it is possible that a mechanism of coping with β-cell stress during diabetes pathogenesis may be activation of PC2 in human β-cells. Repeating past investigations of immunostaining for PC2, performing ISH for PCSK2, and performing pulse-chase experiments with manipulated function of PC1/3 or PC2 in pancreas and islets of patients with diabetes would probe this question. Additionally, given the potentially transient nature of such a compensatory approach, investigating many pancreata from donors with prediabetes or in the early phases of diabetes would be important, perhaps using tissues from the Network for Pancreatic Organ Donors with Diabetes database340.  190 6.3.7 Investigating the role of PC2 for processing proinsulin in rat β-cells We made the surprising observation that unlike mice, dogs, and pigs, rat β-cells are not immunoreactive for PC2, similar to humans. Given the high reliance on rats as a model for the study of diabetes and the potential that they best mimic human proinsulin processing, we propose following up on these findings. First, it would be valuable to perform pulse-chase experiments on rat islets. These findings could be directly compared to pulse-chase results presented in Chapter 5. Second, generation of PC2 deficient rats would be useful. Given the systemic impact of global PC2 knockout on mice283 and the unknown but potential embryonic lethality of PC2 mutations in humans, generation of a β-cell specific PC2 deficient rats would be ideal. This approach would require rats with a β-cells specific Cre like the well studied Ins1-Cre driver mice417 and would require generation of PC2 floxed rats that could be made by CRISPR/Cas9 approach. Examination of circulating proinsulin immunoreactivity by commercially available ELISA, histological examination of proinsulin intermediates in pancreas sections, and western blot assessment of the ratio of proinsulin to insulin in pancreas could all determine if a loss of PC2 has a detrimental impact on rat proinsulin processing. Given the lack of immunoreactivity for PC2 in rat β-cells, we hypothesize that β-cell selective PC2-deficient rats would have normal proinsulin and proIAPP processing.  6.4 Concluding thoughts Despite major advances in diabetes medications511, including improvements in insulin therapy512, these options continue to provide suboptimal glucose control and daily insulin injections and glucose monitoring cause substantial distress to patients. Islet transplantation has clearly demonstrated the potential for cell sources of insulin to be effective cures57 but are limited by supply. Without substantial improvements to alternative and abundant cell-based approaches, they will not meet safety requirements and will fail to be more effective  191 than insulin injections. Basic research into the development and function of β-cells will provide insight into improving available in vitro and in vivo β-cell replacement, regeneration, or genetic correction strategies.  The current thesis has clarified how β-cells normally develop and function with a focus on the specific role of insulin and its precursor proinsulin. Our findings are broadly relevant, including suggesting factors contributing to discrepancies of the ability of therapies to induce the generation of β-cells in situ (e.g. growth factor therapies142, 144, 145, 150) and factors necessary for differentiating β-cells in vitro. Additionally, revealing the essential function of insulin for β-cell maturation has direct implications for the understanding and treatment of monogenic diabetes and our findings provide disappointing insight into potential challenges of gene therapies for monogenic diabetes. Our results are also relevant to research into the potential to induce diabetic remission in patients with type 2 diabetes by inducing the redifferentiation of dedifferentiated β-cells. Furthermore, we develop and characterize a novel in vivo genetic editing tool (AAV Ins1-Cre) that will be useful for further exploration of β-cell biology in mouse models. Finally, we clarified the fundamental process of excising mature insulin from proinsulin in humans and revise a twenty-year-old theory. This research clarifies the potential β-cell defects contributing to the pathological prohormone processing defect observed in both type 1 and type 2 diabetes. With a better understanding of the impact of removing insulin on β-cell development and maturation, the development of an in vivo genetic editing tool, new understanding into the challenges of replacing insulin to β-cells by an AAV mediated gene therapy, and a revised theory on the production of mature insulin in human β-cells, the work in this thesis meaningfully contributes to a better understanding of β-cell biology. It is our hope that the findings presented here will contribute to the eventual development of successful gene- and cell-based cures for diabetes.  192 References 1. International Diabetes Foundation, IDF Diabetes. 7th ed. 2015, Brussels, Belgium: International Diabetes Foundation. 2. Anderson, R.J., et al., The prevalence of comorbid depression in adults with diabetes - A meta-analysis. Diabetes Care, 2001. 24(6): p. 1069-1078. 3. Association, C.D. Diabetes Statistics in Canada. 2015  [cited 2018 July 7]; Available from: http://www.diabetes.ca/how-you-can-help/advocate/why-federal-leadership-is-essential/diabetes-statistics-in-canada. 4. Bilandzic, A. and L. Rosella, The cost of diabetes in Canada over 10 years: applying attributable health care costs to a diabetes incidence prediction model. Health Promot Chronic Dis Prev Can, 2017. 37(2): p. 49-53. 5. Nathan, D.M., Long-term complications of diabetes mellitus. N Engl J Med, 1993. 328(23): p. 1676-85. 6. Flannick, J., S. Johansson, and P.R. Njolstad, Common and rare forms of diabetes mellitus: towards a continuum of diabetes subtypes. Nat Rev Endocrinol, 2016. 7. Ellis, C., A. Ramzy, and T.J. Kieffer, Regenerative medicine and cell-based approaches to restore pancreatic function. Nat Rev Gastroenterol Hepatol, 2017. 14(10): p. 612-628. 8. Todd, J.A., et al., Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet, 2007. 39(7): p. 857-64. 9. Faerch, K., et al., Insulin Resistance Is Accompanied by Increased Fasting Glucagon and Delayed Glucagon Suppression in Individuals With Normal and Impaired Glucose Regulation. Diabetes, 2016. 65(11): p. 3473-3481. 10. Kaprio, J., et al., Concordance for type 1 (insulin-dependent) and type 2 (non-insulin-dependent) diabetes mellitus in a population-based cohort of twins in Finland. Diabetologia, 1992. 35(11): p. 1060-7. 11. Fuchsberger, C., et al., The genetic architecture of type 2 diabetes. Nature, 2016. 536(7614): p. 41-47. 12. Shields, B.M., et al., Maturity-onset diabetes of the young (MODY): how many cases are we missing? Diabetologia, 2010. 53(12): p. 2504-8. 13. Vaxillaire, M. and P. Froguel, Monogenic diabetes in the young, pharmacogenetics and relevance to multifactorial forms of type 2 diabetes. Endocr Rev, 2008. 29(3): p. 254-64. 14. Anik, A., et al., Maturity-onset diabetes of the young (MODY): an update. J Pediatr Endocrinol Metab, 2015. 28(3-4): p. 251-63. 15. Bansal, V., et al., Spectrum of mutations in monogenic diabetes genes identified from high-throughput DNA sequencing of 6888 individuals. BMC Med, 2017. 15(1): p. 213. 16. Asadi, A., J.E. Bruin, and T.J. Kieffer, Characterization of Antibodies to Products of Proinsulin Processing Using Immunofluorescence Staining of Pancreas in Multiple Species. J Histochem Cytochem, 2015. 63(8): p. 646-62. 17. Tager, H., et al., A structurally abnormal insulin causing human diabetes. Nature, 1979. 281(5727): p. 122-5. 18. Liu, M., et al., Proinsulin misfolding and diabetes: mutant INS gene-induced diabetes of youth. Trends Endocrinol Metab, 2010. 21(11): p. 652-9. 19. Edghill, E.L., et al., Insulin mutation screening in 1,044 patients with diabetes: mutations in the INS gene are a common cause of neonatal diabetes but a rare cause of diabetes diagnosed in childhood or adulthood. Diabetes, 2008. 57(4): p. 1034-42. 20. Stoy, J., et al., Insulin gene mutations as a cause of permanent neonatal diabetes. Proc Natl Acad Sci U S A, 2007. 104(38): p. 15040-4.  193 21. Riahi, Y., et al., Inhibition of mTORC1 by ER stress impairs neonatal beta-cell expansion and predisposes to diabetes in the Akita mouse. Elife, 2018. 7. 22. Ron, D., Proteotoxicity in the endoplasmic reticulum: lessons from the Akita diabetic mouse. J Clin Invest, 2002. 109(4): p. 443-5. 23. Hodish, I., et al., Misfolded proinsulin affects bystander proinsulin in neonatal diabetes. J Biol Chem, 2010. 285(1): p. 685-94. 24. Carmody, D., et al., Continued lessons from the INS gene: an intronic mutation causing diabetes through a novel mechanism. J Med Genet, 2015. 52(9): p. 612-6. 25. Oram, R.A., et al., The majority of patients with long-duration type 1 diabetes are insulin microsecretors and have functioning beta cells. Diabetologia, 2014. 57(1): p. 187-91. 26. Soleimanpour, S.A. and D.A. Stoffers, The pancreatic beta cell and type 1 diabetes: innocent bystander or active participant? Trends Endocrinol Metab, 2013. 24(7): p. 324-31. 27. Matveyenko, A.V. and P.C. Butler, Relationship between beta-cell mass and diabetes onset. Diabetes Obes Metab, 2008. 10 Suppl 4: p. 23-31. 28. Butler, A.E., et al., Beta-cell deficit and increased beta-cell apoptosis in humans with type 2 diabetes. Diabetes, 2003. 52(1): p. 102-10. 29. Rahier, J., et al., Pancreatic beta-cell mass in European subjects with type 2 diabetes. Diabetes Obes Metab, 2008. 10 Suppl 4: p. 32-42. 30. Weyer, C., et al., The natural history of insulin secretory dysfunction and insulin resistance in the pathogenesis of type 2 diabetes mellitus. J Clin Invest, 1999. 104(6): p. 787-94. 31. Ferrannini, E., The stunned beta cell: a brief history. Cell Metab, 2010. 11(5): p. 349-52. 32. Dor, Y. and B. Glaser, beta-cell dedifferentiation and type 2 diabetes. N Engl J Med, 2013. 368(6): p. 572-3. 33. Weinberg, N., et al., Lineage tracing evidence for in vitro dedifferentiation but rare proliferation of mouse pancreatic beta-cells. Diabetes, 2007. 56(5): p. 1299-304. 34. Wang, Z., et al., Pancreatic beta cell dedifferentiation in diabetes and redifferentiation following insulin therapy. Cell Metab, 2014. 19(5): p. 872-82. 35. Talchai, C., et al., Pancreatic beta cell dedifferentiation as a mechanism of diabetic beta cell failure. Cell, 2012. 150(6): p. 1223-34. 36. Sheng, C., et al., Reversibility of beta-Cell-Specific Transcript Factors Expression by Long-Term Caloric Restriction in db/db Mouse. J Diabetes Res, 2016. 2016: p. 6035046. 37. Cinti, F., et al., Evidence of beta-Cell Dedifferentiation in Human Type 2 Diabetes. J Clin Endocrinol Metab, 2016. 101(3): p. 1044-54. 38. Butler, A.E., et al., beta-Cell Deficit in Obese Type 2 Diabetes, a Minor Role of beta-Cell Dedifferentiation and Degranulation. J Clin Endocrinol Metab, 2016. 101(2): p. 523-32. 39. Rawshani, A., et al., Impact of Socioeconomic Status on Cardiovascular Disease and Mortality in 24,947 Individuals With Type 1 Diabetes. Diabetes Care, 2015. 38(8): p. 1518-27. 40. Johnson, I.S., Human insulin from recombinant DNA technology. Science, 1983. 219(4585): p. 632-7. 41. Taton, J., et al., How recombinant insulin analogs improve insulin therapy of diabetes mellitus: pathophysiology, clinical practice and recommendations. Med Sci Monit, 2001. 7(4): p. 848-59. 42. Hartman, I., Insulin analogs: impact on treatment success, satisfaction, quality of life, and adherence. Clin Med Res, 2008. 6(2): p. 54-67.  194 43. Sharma, M., I. Nazareth, and I. Petersen, Trends in incidence, prevalence and prescribing in type 2 diabetes mellitus between 2000 and 2013 in primary care: a retrospective cohort study. BMJ Open, 2016. 6(1): p. e010210. 44. Domecq, J.P., et al., Clinical review: Drugs commonly associated with weight change: a systematic review and meta-analysis. J Clin Endocrinol Metab, 2015. 100(2): p. 363-70. 45. Diabetes Canada Clinical Practice Guidelines Expert, C., et al., Pharmacologic Glycemic Management of Type 2 Diabetes in Adults. Can J Diabetes, 2018. 42 Suppl 1: p. S88-S103. 46. Song, J., et al., KCNJ11, ABCC8 and TCF7L2 polymorphisms and the response to sulfonylurea treatment in patients with type 2 diabetes: a bioinformatics assessment. BMC Med Genet, 2017. 18(1): p. 64. 47. van Dalem, J., et al., Risk of hypoglycaemia in users of sulphonylureas compared with metformin in relation to renal function and sulphonylurea metabolite group: population based cohort study. BMJ, 2016. 354: p. i3625. 48. van Raalte, D.H. and C.B. Verchere, Improving glycemic control in type 2 diabetes: stimulate insulin secretion or provide beta-cell rest? Diabetes Obes Metab, 2017. 49. Baggio, L.L. and D.J. Drucker, Biology of incretins: GLP-1 and GIP. Gastroenterology, 2007. 132(6): p. 2131-57. 50. Vilsboll, T., SUSTAIN – 6 clinical and metabolic outcomes. EASD Oral Presentation # S353, in EASD. 2016: Munich, Germany. 51. Shyangdan, D.S., et al., Glucagon-like peptide analogues for type 2 diabetes mellitus. Cochrane Database Syst Rev, 2011(10): p. CD006423. 52. Tahrani, A.A., A.H. Barnett, and C.J. Bailey, SGLT inhibitors in management of diabetes. Lancet Diabetes Endocrinol, 2013. 1(2): p. 140-51. 53. Shapiro, A.M., M. Pokrywczynska, and C. Ricordi, Clinical pancreatic islet transplantation. Nat Rev Endocrinol, 2016. 54. Bruni, A., et al., Islet cell transplantation for the treatment of type 1 diabetes: recent advances and future challenges. Diabetes, Metabolic Syndrome and Obesity: Targets and Therapy, 2014. 7: p. 211-223. 55. Williams, P., Notes on diabetes treated with extract and by grafts of sheep's pancreas. Br Med J, 1894. 2: p. 1303-1304. 56. Shapiro, A.M., et al., Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. The New England journal of medicine, 2000. 343(4): p. 230-238. 57. Warnock, G.L., et al., A multi-year analysis of islet transplantation compared with intensive medical therapy on progression of complications in type 1 diabetes. Transplantation, 2008. 86(12): p. 1762-6. 58. Moassesfar, S., et al., A Comparative Analysis of the Safety, Efficacy, and Cost of Islet Versus Pancreas Transplantation in Nonuremic Patients With Type 1 Diabetes. Am J Transplant, 2016. 16(2): p. 518-26. 59. Halban, P.A., et al., Gene and cell-replacement therapy in the treatment of type 1 diabetes: how high must the standards be set? Diabetes, 2001. 50(10): p. 2181-91. 60. MacDonald, P.E., J.W. Joseph, and P. Rorsman, Glucose-sensing mechanisms in pancreatic beta-cells. Philos Trans R Soc Lond B Biol Sci, 2005. 360(1464): p. 2211-25. 61. Johnson, J.D., The quest to make fully functional human pancreatic beta cells from embryonic stem cells: climbing a mountain in the clouds. Diabetologia, 2016. 59(10): p. 2047-57. 62. Kieffer, T.J., Closing in on Mass Production of Mature Human Beta Cells. Cell Stem Cell, 2016. 18(6): p. 699-702.  195 63. D'Amour, K.A., et al., Production of pancreatic hormone-expressing endocrine cells from human embryonic stem cells. Nat Biotechnol, 2006. 24(11): p. 1392-401. 64. Kroon, E., et al., Pancreatic endoderm derived from human embryonic stem cells generates glucose-responsive insulin-secreting cells in vivo. Nat Biotechnol, 2008. 26(4): p. 443-52. 65. Berman, D.M., et al., Bioengineering the endocrine pancreas: intraomental islet transplantation within a biologic resorbable scaffold. Diabetes, 2016. 66. Caiazzo, R., et al., Evaluation of Alternative Sites for Islet Transplantation in the Minipig: Interest and Limits of the Gastric Submucosa. Transplantation Proceedings, 2007. 39(8): p. 2620-2623. 67. Echeverri, G.J., et al., Endoscopic Gastric Submucosal Transplantation of Islets (ENDO-STI): Technique and Initial Results in Diabetic Pigs. American Journal of Transplantation, 2009. 9(11): p. 2485-2496. 68. Garkavenko, O., et al., Islets transplantation: New Zealand experience. Xenotransplantation, 2011. 18(1): p. 60-60. 69. Ricordi, C., et al., National Institutes of Health-Sponsored Clinical Islet Transplantation Consortium Phase 3 Trial: Manufacture of a Complex Cellular Product at Eight Processing Facilities. Diabetes, 2016. 65(11): p. 3418-3428. 70. Hering, B.J., et al., Phase 3 Trial of Transplantation of Human Islets in Type 1 Diabetes Complicated by Severe Hypoglycemia. Diabetes Care, 2016. 39(7): p. 1230-40. 71. Kimura-Yoshida, C., et al., Canonical Wnt signaling and its antagonist regulate anterior-posterior axis polarization by guiding cell migration in mouse visceral endoderm. Dev Cell, 2005. 9(5): p. 639-50. 72. Wells, J.M. and D.A. Melton, Early mouse endoderm is patterned by soluble factors from adjacent germ layers. Development, 2000. 127(8): p. 1563-72. 73. Jorgensen, M.C., et al., An illustrated review of early pancreas development in the mouse. Endocrine reviews, 2007. 28(6): p. 685-705. 74. McCracken, K.W. and J.M. Wells, Molecular pathways controlling pancreas induction. Semin Cell Dev Biol, 2012. 23(6): p. 656-62. 75. Jennings, R.E., et al., Development of the human pancreas from foregut to endocrine commitment. Diabetes, 2013. 62(10): p. 3514-22. 76. Ahlgren, U., J. Jonsson, and H. Edlund, The morphogenesis of the pancreatic mesenchyme is uncoupled from that of the pancreatic epithelium in IPF1/PDX1-deficient mice. Development, 1996. 122(5): p. 1409-16. 77. Sander, M., et al., Homeobox gene Nkx6.1 lies downstream of Nkx2.2 in the major pathway of beta-cell formation in the pancreas. Development, 2000. 127(24): p. 5533-40. 78. Schaffer, A.E., et al., Nkx6 transcription factors and Ptf1a function as antagonistic lineage determinants in multipotent pancreatic progenitors. Developmental cell, 2010. 18(6): p. 1022-1029. 79. Apelqvist, A., et al., Notch signalling controls pancreatic cell differentiation. Nature, 1999. 400(6747): p. 877-81. 80. Gradwohl, G., et al., Neurogenin3 is Required for the Development of the Four Endocrine Cell Lineages of the Pancreas. Proceedings of the National Academy of Sciences of the United States of America, 2000. 97(4): p. 1607-1611. 81. Mamidi, A., et al., Mechanosignalling via integrins directs fate decisions of pancreatic progenitors. Nature, 2018. 564(7734): p. 114-118. 82. Villasenor, A., D.C. Chong, and O. Cleaver, Biphasic Ngn3 expression in the developing pancreas. Dev Dyn, 2008. 237(11): p. 3270-9.  196 83. Nair, G. and M. Hebrok, Islet formation in mice and men: lessons for the generation of functional insulin-producing beta-cells from human pluripotent stem cells. Curr Opin Genet Dev, 2015. 32: p. 171-80. 84. Ahlgren, U., et al., Independent requirement for ISL1 in formation of pancreatic mesenchyme and islet cells. Nature, 1997. 385(6613): p. 257-60. 85. Naya, F.J., et al., Diabetes, defective pancreatic morphogenesis, and abnormal enteroendocrine differentiation in BETA2/neuroD-deficient mice. Genes Dev, 1997. 11(18): p. 2323-34. 86. Collombat, P., et al., The simultaneous loss of Arx and Pax4 genes promotes a somatostatin-producing cell fate specification at the expense of the alpha- and beta-cell lineages in the mouse endocrine pancreas. Development, 2005. 132(13): p. 2969-80. 87. Collombat, P., et al., Opposing actions of Arx and Pax4 in endocrine pancreas development. Genes Dev, 2003. 17(20): p. 2591-603. 88. Sussel, L., et al., Mice lacking the homeodomain transcription factor Nkx2.2 have diabetes due to arrested differentiation of pancreatic beta cells. Development, 1998. 125(12): p. 2213-21. 89. Sander, M., et al., Genetic analysis reveals that PAX6 is required for normal transcription of pancreatic hormone genes and islet development. Genes Dev, 1997. 11(13): p. 1662-73. 90. Ahmad, Z., et al., Pax6 Inactivation in the Adult Pancreas Reveals Ghrelin as Endocrine Cell Maturation Marker. PLoS One, 2015. 10(12): p. e0144597. 91. Gao, T., et al., Pdx1 maintains beta cell identity and function by repressing an alpha cell program. Cell Metab, 2014. 19(2): p. 259-71. 92. Herrera, P.L., et al., Embryogenesis of the murine endocrine pancreas; early expression of pancreatic polypeptide gene. Development, 1991. 113(4): p. 1257-65. 93. Vieira, A., et al., beta-Cell Replacement Strategies: The Increasing Need for a "beta-Cell Dogma". Front Genet, 2017. 8: p. 75. 94. van der Meulen, T. and M.O. Huising, Maturation of stem cell-derived beta-cells guided by the expression of urocortin 3. Rev Diabet Stud, 2014. 11(1): p. 115-32. 95. Nishimura, W., S. Takahashi, and K. Yasuda, MafA is critical for maintenance of the mature beta cell phenotype in mice. Diabetologia, 2015. 58(3): p. 566-74. 96. Aguayo-Mazzucato, C., et al., Mafa expression enhances glucose-responsive insulin secretion in neonatal rat beta cells. Diabetologia, 2011. 54(3): p. 583-93. 97. Faradji, R.N., et al., Glucose-induced toxicity in insulin-producing pituitary cells that coexpress GLUT2 and glucokinase. Implications for metabolic engineering. J Biol Chem, 2001. 276(39): p. 36695-702. 98. Xie, M., et al., beta-cell-mimetic designer cells provide closed-loop glycemic control. Science, 2016. 354(6317): p. 1296-1301. 99. Hohmeier, H.E., et al., Isolation of INS-1-derived cell lines with robust ATP-sensitive K+ channel-dependent and -independent glucose-stimulated insulin secretion. Diabetes, 2000. 49(3): p. 424-30. 100. Clark, S.A., et al., Novel insulinoma cell lines produced by iterative engineering of GLUT2, glucokinase, and human insulin expression. Diabetes, 1997. 46(6): p. 958-67. 101. Ravassard, P., et al., A genetically engineered human pancreatic beta cell line exhibiting glucose-inducible insulin secretion. J Clin Invest, 2011. 121(9): p. 3589-97. 102. Ellis, C.E. and G.S. Korbutt, Justifying clinical trials for porcine islet xenotransplantation. Xenotransplantation, 2015. 22(5): p. 336-44. 103. Cardona, K., et al., Long-term survival of neonatal porcine islets in nonhuman primates by targeting costimulation pathways. Nat Med, 2006. 12(3): p. 304-6.  197 104. Hering, B.J., et al., Prolonged diabetes reversal after intraportal xenotransplantation of wild-type porcine islets in immunosuppressed nonhuman primates. Nat Med, 2006. 12(3): p. 301-3. 105. van der Windt, D.J., et al., Clinical islet xenotransplantation: how close are we? Diabetes, 2012. 61(12): p. 3046-55. 106. Samy, K.P., et al., Islet cell xenotransplantation: a serious look toward the clinic. Xenotransplantation, 2014. 21(3): p. 221-9. 107. van der Laan, L.J., et al., Infection by porcine endogenous retrovirus after islet xenotransplantation in SCID mice. Nature, 2000. 407(6800): p. 90-4. 108. Yang, L., et al., Genome-wide inactivation of porcine endogenous retroviruses (PERVs). Science, 2015. 350(6264): p. 1101-4. 109. Nagashima, H. and H. Matsunari, Growing human organs in pigs-A dream or reality? Theriogenology, 2016. 86(1): p. 422-6. 110. Matsunari, H., et al., Blastocyst complementation generates exogenic pancreas in vivo in apancreatic cloned pigs. Proc Natl Acad Sci U S A, 2013. 110(12): p. 4557-62. 111. Kobayashi, T., et al., Generation of rat pancreas in mouse by interspecific blastocyst injection of pluripotent stem cells. Cell, 2010. 142(5): p. 787-99. 112. Kang, J.D., et al., Apancreatic pigs cloned using Pdx1-disrupted fibroblasts created via TALEN-mediated mutagenesis. Oncotarget, 2017. 8(70): p. 115480-115489. 113. Murry, C.E. and G. Keller, Differentiation of embryonic stem cells to clinically relevant populations: lessons from embryonic development. Cell, 2008. 132(4): p. 661-80. 114. Castaing, M., et al., Blood glucose normalization upon transplantation of human embryonic pancreas into beta-cell-deficient SCID mice. Diabetologia, 2001. 44(11): p. 2066-76. 115. Tuch, B.E., et al., Recovery of human fetal pancreas after one year of implantation in the diabetic patient. Transplantation, 1988. 46(6): p. 865-70. 116. Lafferty, K.J. and L. Hao, Fetal pancreas transplantation for treatment of IDDM patients. Diabetes Care, 1993. 16(1): p. 383-6. 117. Farkas, G. and S. Karacsonyi, Clinical transplantation of fetal human pancreatic islets. Biomed Biochim Acta, 1985. 44(1): p. 155-9. 118. Rezania, A., et al., Maturation of human embryonic stem cell-derived pancreatic progenitors into functional islets capable of treating pre-existing diabetes in mice. Diabetes, 2012. 61(8): p. 2016-29. 119. Rezania, A., et al., Enrichment of human embryonic stem cell-derived NKX6.1-expressing pancreatic progenitor cells accelerates the maturation of insulin-secreting cells in vivo. Stem Cells, 2013. 31(11): p. 2432-42. 120. Bruin, J.E., et al., Maturation and function of human embryonic stem cell-derived pancreatic progenitors in macroencapsulation devices following transplant into mice. Diabetologia, 2013. 56(9): p. 1987-98. 121. Inc, V., Two-year Data from ViaCyte’s STEP ONE Clinical Trial Presented at ADA 2018. 2018: https://viacyte.com/archives/press-releases/two-year-data-from-viacytes-step-one-clinical-trial-presented-at-ada-2018. 122. Rezania, A., et al., Reversal of diabetes with insulin-producing cells derived in vitro from human pluripotent stem cells. Nat Biotechnol, 2014. 32(11): p. 1121-33. 123. Quiskamp, N., J.E. Bruin, and T.J. Kieffer, Differentiation of human pluripotent stem cells into beta-cells: Potential and challenges. Best Pract Res Clin Endocrinol Metab, 2015. 29(6): p. 833-47. 124. Mas, A., et al., Reversal of type 1 diabetes by engineering a glucose sensor in skeletal muscle. Diabetes, 2006. 55(6): p. 1546-53.  198 125. Lipes, M.A., et al., Insulin-secreting non-islet cells are resistant to autoimmune destruction. Proc Natl Acad Sci U S A, 1996. 93(16): p. 8595-600. 126. Inada, A., et al., Carbonic anhydrase II-positive pancreatic cells are progenitors for both endocrine and exocrine pancreas after birth. Proc Natl Acad Sci U S A, 2008. 105(50): p. 19915-9. 127. Xu, X., et al., Beta cells can be generated from endogenous progenitors in injured adult mouse pancreas. Cell, 2008. 132(2): p. 197-207. 128. Swales, N., et al., Plasticity of adult human pancreatic duct cells by neurogenin3-mediated reprogramming. PLoS One, 2012. 7(5): p. e37055. 129. Van de Casteele, M., et al., Neurogenin 3+ cells contribute to beta-cell neogenesis and proliferation in injured adult mouse pancreas. Cell Death Dis, 2013. 4: p. e523. 130. Solar, M., et al., Pancreatic exocrine duct cells give rise to insulin-producing beta cells during embryogenesis but not after birth. Dev Cell, 2009. 17(6): p. 849-60. 131. Kopp, J.L., et al., Sox9+ ductal cells are multipotent progenitors throughout development but do not produce new endocrine cells in the normal or injured adult pancreas. Development, 2011. 138(4): p. 653-65. 132. Zhang, M., et al., Growth factors and medium hyperglycemia induce Sox9+ ductal cell differentiation into beta cells in mice with reversal of diabetes. Proc Natl Acad Sci U S A, 2016. 113(3): p. 650-5. 133. Valdez, I.A., et al., Proinflammatory Cytokines Induce Endocrine Differentiation in Pancreatic Ductal Cells via STAT3-Dependent NGN3 Activation. Cell Rep, 2016. 15(3): p. 460-70. 134. Cheng, C.W., et al., Fasting-Mimicking Diet Promotes Ngn3-Driven beta-Cell Regeneration to Reverse Diabetes. Cell, 2017. 168(5): p. 775-788 e12. 135. Bonner-Weir, S., et al., In vitro cultivation of human islets from expanded ductal tissue. Proc Natl Acad Sci U S A, 2000. 97(14): p. 7999-8004. 136. Lee, J., et al., Expansion and conversion of human pancreatic ductal cells into insulin-secreting endocrine cells. Elife, 2013. 2: p. e00940. 137. Yamada, T., et al., Reprogramming Mouse Cells With a Pancreatic Duct Phenotype to Insulin-Producing beta-Like Cells. Endocrinology, 2015. 156(6): p. 2029-38. 138. Gmyr, V., et al., Human pancreatic ductal cells: large-scale isolation and expansion. Cell Transplant, 2001. 10(1): p. 109-21. 139. Rooman, I., et al., Mitogenic effect of gastrin and expression of gastrin receptors in duct-like cells of rat pancreas. Gastroenterology, 2001. 121(4): p. 940-9. 140. Rooman, I., J. Lardon, and L. Bouwens, Gastrin stimulates beta-cell neogenesis and increases islet mass from transdifferentiated but not from normal exocrine pancreas tissue. Diabetes, 2002. 51(3): p. 686-90. 141. Wang, T.C., et al., Pancreatic gastrin stimulates islet differentiation of transforming growth factor alpha-induced ductular precursor cells. J Clin Invest, 1993. 92(3): p. 1349-56. 142. Rooman, I. and L. Bouwens, Combined gastrin and epidermal growth factor treatment induces islet regeneration and restores normoglycaemia in C57Bl6/J mice treated with alloxan. Diabetologia, 2004. 47(2): p. 259-65. 143. Suarez-Pinzon, W.L., et al., Combination therapy with epidermal growth factor and gastrin induces neogenesis of human islet {beta}-cells from pancreatic duct cells and an increase in functional {beta}-cell mass. J Clin Endocrinol Metab, 2005. 90(6): p. 3401-9. 144. Baeyens, L., et al., Transient cytokine treatment induces acinar cell reprogramming and regenerates functional beta cell mass in diabetic mice. Nat Biotechnol, 2014. 32(1): p. 76-83.  199 145. Blaine, S.A., et al., Adult pancreatic acinar cells give rise to ducts but not endocrine cells in response to growth factor signaling. Development, 2010. 137(14): p. 2289-96. 146. Zhou, Q., et al., In vivo reprogramming of adult pancreatic exocrine cells to beta-cells. Nature, 2008. 455(7213): p. 627-32. 147. Li, W., et al., In vivo reprogramming of pancreatic acinar cells to three islet endocrine subtypes. Elife, 2014. 3: p. e01846. 148. Li, W., et al., Long-term persistence and development of induced pancreatic beta cells generated by lineage conversion of acinar cells. Nat Biotechnol, 2014. 32(12): p. 1223-30. 149. Cavelti-Weder, C., et al., Hyperglycaemia attenuates in vivo reprogramming of pancreatic exocrine cells to beta cells in mice. Diabetologia, 2016. 59(3): p. 522-32. 150. Clayton, H.W., et al., Pancreatic Inflammation Redirects Acinar to beta Cell Reprogramming. Cell Rep, 2016. 17(8): p. 2028-2041. 151. Lima, M.J., et al., Generation of Functional Beta-Like Cells from Human Exocrine Pancreas. PLoS One, 2016. 11(5): p. e0156204. 152. Lima, M.J., et al., Suppression of epithelial-to-mesenchymal transitioning enhances ex vivo reprogramming of human exocrine pancreatic tissue toward functional insulin-producing beta-like cells. Diabetes, 2013. 62(8): p. 2821-33. 153. Park, J.Y., et al., Pdx1 expression in pancreatic precursor lesions and neoplasms. Appl Immunohistochem Mol Morphol, 2011. 19(5): p. 444-9. 154. Thorel, F., et al., Conversion of adult pancreatic alpha-cells to beta-cells after extreme beta-cell loss. Nature, 2010. 464(7292): p. 1149-1154. 155. Chera, S., et al., Diabetes recovery by age-dependent conversion of pancreatic delta-cells into insulin producers. Nature, 2014. 514(7523): p. 503-7. 156. Beamish, C.A., et al., Insulin-positive, Glut2-low cells present within mouse pancreas exhibit lineage plasticity and are enriched within extra-islet endocrine cell clusters. Islets, 2016. 8(3): p. 65-82. 157. Razavi, R., et al., Diabetes enhances the proliferation of adult pancreatic multipotent progenitor cells and biases their differentiation to more beta-cell production. Diabetes, 2015. 64(4): p. 1311-23. 158. van der Meulen, T., et al., Virgin Beta Cells Persist throughout Life at a Neogenic Niche within Pancreatic Islets. Cell Metab, 2017. 25(4): p. 911-926 e6. 159. Bader, E., et al., Identification of proliferative and mature beta-cells in the islets of Langerhans. Nature, 2016. 535(7612): p. 430-4. 160. Collombat, P., et al., The ectopic expression of Pax4 in the mouse pancreas converts progenitor cells into alpha and subsequently beta cells. Cell, 2009. 138(3): p. 449-62. 161. Al-Hasani, K., et al., Adult duct-lining cells can reprogram into beta-like cells able to counter repeated cycles of toxin-induced diabetes. Dev Cell, 2013. 26(1): p. 86-100. 162. Courtney, M., et al., The inactivation of Arx in pancreatic alpha-cells triggers their neogenesis and conversion into functional beta-like cells. PLoS Genet, 2013. 9(10): p. e1003934. 163. Ben-Othman, N., et al., Long-Term GABA Administration Induces Alpha Cell-Mediated Beta-like Cell Neogenesis. Cell, 2016. 164. Li, J., et al., Artemisinins Target GABAA Receptor Signaling and Impair alpha Cell Identity. Cell, 2016. 165. van der Meulen, T., et al., Artemether Does Not Turn alpha Cells into beta Cells. Cell Metab, 2018. 27(1): p. 218-225 e4. 166. Ackermann, A.M., N.G. Moss, and K.H. Kaestner, GABA and Artesunate Do Not Induce Pancreatic alpha-to-beta Cell Transdifferentiation In Vivo. Cell Metab, 2018.  200 167. Meier, J.J., et al., Beta-cell replication is the primary mechanism subserving the postnatal expansion of beta-cell mass in humans. Diabetes, 2008. 57(6): p. 1584-94. 168. Van Assche, F.A., L. Aerts, and F. De Prins, A morphological study of the endocrine pancreas in human pregnancy. Br J Obstet Gynaecol, 1978. 85(11): p. 818-20. 169. Saisho, Y., et al., beta-cell mass and turnover in humans: effects of obesity and aging. Diabetes Care, 2013. 36(1): p. 111-7. 170. Fiaschi-Taesch, N., et al., Survey of the human pancreatic beta-cell G1/S proteome reveals a potential therapeutic role for cdk-6 and cyclin D1 in enhancing human beta-cell replication and function in vivo. Diabetes, 2009. 58(4): p. 882-93. 171. Shirakawa, J., et al., Insulin Signaling Regulates the FoxM1/PLK1/CENP-A Pathway to Promote Adaptive Pancreatic beta Cell Proliferation. Cell Metab, 2017. 172. Folli, F., et al., Altered insulin receptor signalling and beta-cell cycle dynamics in type 2 diabetes mellitus. PLoS One, 2011. 6(11): p. e28050. 173. Kim, H., et al., Serotonin regulates pancreatic beta cell mass during pregnancy. Nat Med, 2010. 16(7): p. 804-8. 174. Rieck, S. and K.H. Kaestner, Expansion of beta-cell mass in response to pregnancy. Trends Endocrinol Metab, 2010. 21(3): p. 151-8. 175. Lakshmipathi, J., et al., PKCzeta Is Essential for Pancreatic beta-Cell Replication During Insulin Resistance by Regulating mTOR and Cyclin-D2. Diabetes, 2016. 65(5): p. 1283-96. 176. Dirice, E., et al., Inhibition of DYRK1A Stimulates Human beta-Cell Proliferation. Diabetes, 2016. 65(6): p. 1660-71. 177. Shen, W., et al., Inhibition of DYRK1A and GSK3B induces human beta-cell proliferation. Nat Commun, 2015. 6: p. 8372. 178. Wang, P., et al., A high-throughput chemical screen reveals that harmine-mediated inhibition of DYRK1A increases human pancreatic beta cell replication. Nat Med, 2015. 21(4): p. 383-8. 179. El Ouaamari, A., et al., SerpinB1 Promotes Pancreatic beta Cell Proliferation. Cell Metab, 2016. 23(1): p. 194-205. 180. Wang, P., et al., Diabetes mellitus--advances and challenges in human beta-cell proliferation. Nat Rev Endocrinol, 2015. 11(4): p. 201-12. 181. Slack, J.M., Developmental biology of the pancreas. Development, 1995. 121(6): p. 1569-80. 182. Ferber, S., et al., Pancreatic and duodenal homeobox gene 1 induces expression of insulin genes in liver and ameliorates streptozotocin-induced hyperglycemia. Nat Med, 2000. 6(5): p. 568-72. 183. Ber, I., et al., Functional, persistent, and extended liver to pancreas transdifferentiation. J Biol Chem, 2003. 278(34): p. 31950-7. 184. Sapir, T., et al., Cell-replacement therapy for diabetes: Generating functional insulin-producing tissue from adult human liver cells. Proc Natl Acad Sci U S A, 2005. 102(22): p. 7964-9. 185. Yechoor, V., et al., Neurogenin3 is sufficient for transdetermination of hepatic progenitor cells into neo-islets in vivo but not transdifferentiation of hepatocytes. Dev Cell, 2009. 16(3): p. 358-73. 186. Donelan, W., et al., Pancreatic and duodenal homeobox gene 1 (Pdx1) down-regulates hepatic transcription factor 1 alpha (HNF1alpha) expression during reprogramming of human hepatic cells into insulin-producing cells. Am J Transl Res, 2015. 7(6): p. 995-1008.  201 187. Tang, D.Q., et al., Genetically reprogrammed, liver-derived insulin-producing cells are glucose-responsive, but susceptible to autoimmune destruction in settings of murine model of type 1 diabetes. Am J Transl Res, 2013. 5(2): p. 184-99. 188. Banga, A., et al., In vivo reprogramming of Sox9+ cells in the liver to insulin-secreting ducts. Proc Natl Acad Sci U S A, 2012. 109(38): p. 15336-41. 189. Banga, A., et al., Stable insulin-secreting ducts formed by reprogramming of cells in the liver using a three-gene cocktail and a PPAR agonist. Gene Ther, 2014. 21(1): p. 19-27. 190. Nagasaki, H., et al., Generation of insulin-producing cells from the mouse liver using beta cell-related gene transfer including Mafa and Mafb. PLoS One, 2014. 9(11): p. e113022. 191. Ham, D.S., et al., Generation of functional insulin-producing cells from neonatal porcine liver-derived cells by PDX1/VP16, BETA2/NeuroD and MafA. PLoS One, 2013. 8(11): p. e79076. 192. Li, R., et al., Gene therapy with neurogenin3, betacellulin and SOCS1 reverses diabetes in NOD mice. Gene Ther, 2015. 22(11): p. 876-82. 193. Di Stasi, A., et al., Inducible apoptosis as a safety switch for adoptive cell therapy. N Engl J Med, 2011. 365(18): p. 1673-83. 194. Kolodka, T.M., et al., Gene therapy for diabetes mellitus in rats by hepatic expression of insulin. Proc Natl Acad Sci U S A, 1995. 92(8): p. 3293-7. 195. Ren, B., et al., Long-term correction of diabetes in rats after lentiviral hepatic insulin gene therapy. Diabetologia, 2007. 50(9): p. 1910-20. 196. Elsner, M., et al., Reversal of diabetes through gene therapy of diabetic rats by hepatic insulin expression via lentiviral transduction. Mol Ther, 2012. 20(5): p. 918-26. 197. Elsner, M., A. Jorns, and S. Lenzen, Diabetes therapy by lentiviral hepatic insulin gene expression without transformation of liver. Diabetologia, 2008. 51(4): p. 694-5; author reply 696. 198. Hsu, P.Y., R.M. Kotin, and Y.W. Yang, Glucose- and metabolically regulated hepatic insulin gene therapy for diabetes. Pharm Res, 2008. 25(6): p. 1460-8. 199. Thule, P.M., et al., Long-term glycemic control with hepatic insulin gene therapy in streptozotocin-diabetic mice. J Gene Med, 2015. 17(8-9): p. 141-52. 200. Thule, P.M., et al., Hepatic insulin gene therapy prevents deterioration of vascular function and improves adipocytokine profile in STZ-diabetic rats. Am J Physiol Endocrinol Metab, 2006. 290(1): p. E114-E122. 201. You, S., et al., Hepatic insulin gene therapy prevents diabetic enteropathy in STZ-treated CD-1 mice. Mol Ther Methods Clin Dev, 2015. 2: p. 15028. 202. You, S., et al., [Effects of hepatic insulin gene therapy on enteric neuropathy in STZ-diabetic mice]. Zhong Nan Da Xue Xue Bao Yi Xue Ban, 2011. 36(6): p. 546-53. 203. Okitsu, T., et al., Transplantation of reversibly immortalized insulin-secreting human hepatocytes controls diabetes in pancreatectomized pigs. Diabetes, 2004. 53(1): p. 105-12. 204. Olson, D.E., et al., Glucose-responsive hepatic insulin gene therapy of spontaneously diabetic BB/Wor rats. Hum Gene Ther, 2003. 14(15): p. 1401-13. 205. Groskreutz, D.J., M.X. Sliwkowski, and C.M. Gorman, Genetically engineered proinsulin constitutively processed and secreted as mature, active insulin. J Biol Chem, 1994. 269(8): p. 6241-5. 206. Jensen, J., et al., Control of endodermal endocrine development by Hes-1. Nat Genet, 2000. 24(1): p. 36-44. 207. Schonhoff, S.E., M. Giel-Moloney, and A.B. Leiter, Minireview: Development and differentiation of gut endocrine cells. Endocrinology, 2004. 145(6): p. 2639-44.  202 208. Fujita, Y., A.T. Cheung, and T.J. Kieffer, Harnessing the gut to treat diabetes. Pediatr Diabetes, 2004. 5 Suppl 2: p. 57-69. 209. Kitamura, T., et al., Regulation of pancreatic juxtaductal endocrine cell formation by FoxO1. Mol Cell Biol, 2009. 29(16): p. 4417-30. 210. Talchai, C., et al., Generation of functional insulin-producing cells in the gut by Foxo1 ablation. Nat Genet, 2012. 44(4): p. 406-12, S1. 211. Chen, Y.J., et al., De novo formation of insulin-producing "neo-beta cell islets" from intestinal crypts. Cell Rep, 2014. 6(6): p. 1046-58. 212. Ariyachet, C.e.a., Reprogrammed Stomach Tissue as a Renewable Source of Functional b Cells for Blood Glucose Regulation. Cell Stem Cell, 2016(18): p. 1-12. 213. Young, L.S. and V. Mautner, The promise and potential hazards of adenovirus gene therapy. Gut, 2001. 48(5): p. 733-6. 214. Kriegel, C., H. Attarwala, and M. Amiji, Multi-compartmental oral delivery systems for nucleic acid therapy in the gastrointestinal tract. Adv Drug Deliv Rev, 2013. 65(6): p. 891-901. 215. Wang, Y., et al., Conversion of Human Gastric Epithelial Cells to Multipotent Endodermal Progenitors using Defined Small Molecules. Cell Stem Cell, 2016. 216. Isenman, L., C. Liebow, and S. Rothman, The endocrine secretion of mammalian digestive enzymes by exocrine glands. Am J Physiol, 1999. 276(2 Pt 1): p. E223-32. 217. Goldfine, I.D., et al., The endocrine secretion of human insulin and growth hormone by exocrine glands of the gastrointestinal tract. Nat Biotechnol, 1997. 15(13): p. 1378-82. 218. Samuni, Y. and B.J. Baum, Gene delivery in salivary glands: from the bench to the clinic. Biochim Biophys Acta, 2011. 1812(11): p. 1515-21. 219. Rowzee, A.M., et al., Expression and secretion of human proinsulin-B10 from mouse salivary glands: implications for the treatment of type I diabetes mellitus. PLoS One, 2013. 8(3): p. e59222. 220. Baum, B.J., et al., Early responses to adenoviral-mediated transfer of the aquaporin-1 cDNA for radiation-induced salivary hypofunction. Proc Natl Acad Sci U S A, 2012. 109(47): p. 19403-7. 221. Baum, B.J., et al., Advances in salivary gland gene therapy - oral and systemic implications. Expert Opin Biol Ther, 2015. 15(10): p. 1443-54. 222. Cheung, A.T., et al., Glucose-dependent insulin release from genetically engineered K cells. Science, 2000. 290(5498): p. 1959-62. 223. Mojibian, M., et al., Insulin-producing intestinal K cells protect nonobese diabetic mice from autoimmune diabetes. Gastroenterology, 2014. 147(1): p. 162-171 e6. 224. Clinicaltrials.gov. Search Results: "Gene Therapy". 2016  [cited 2016 Oct 12]; Available from: https://clinicaltrials.gov/ct2/results?term=%22Gene+Therapy%22&Search=Search. 225. Montane, J., et al., Prevention of murine autoimmune diabetes by CCL22-mediated Treg recruitment to the pancreatic islets. J Clin Invest, 2011. 121(8): p. 3024-8. 226. Meijer, D.H., et al., Controlling brain tumor growth by intraventricular administration of an AAV vector encoding IFN-beta. Cancer Gene Ther, 2009. 16(8): p. 664-71. 227. Gu, H., et al., Deletion of a DNA polymerase beta gene segment in T cells using cell type-specific gene targeting. Science, 1994. 265(5168): p. 103-6. 228. Chew, W.L., et al., A multifunctional AAV-CRISPR-Cas9 and its host response. Nat Methods, 2016. 13(10): p. 868-74. 229. Atchison, R.W., B.C. Casto, and W.M. Hammon, Adenovirus-Associated Defective Virus Particles. Science, 1965. 149(3685): p. 754-6. 230. Kotin, R.M., et al., Site-specific integration by adeno-associated virus. Proc Natl Acad Sci U S A, 1990. 87(6): p. 2211-5.  203 231. Mingozzi, F. and K.A. High, Immune responses to AAV in clinical trials. Curr Gene Ther, 2007. 7(5): p. 316-24. 232. Grieger, J.C. and R.J. Samulski, Adeno-associated virus vectorology, manufacturing, and clinical applications. Methods Enzymol, 2012. 507: p. 229-54. 233. Bowles, D.E., et al., Phase 1 gene therapy for Duchenne muscular dystrophy using a translational optimized AAV vector. Mol Ther, 2012. 20(2): p. 443-55. 234. Lusby, E., K.H. Fife, and K.I. Berns, Nucleotide sequence of the inverted terminal repetition in adeno-associated virus DNA. J Virol, 1980. 34(2): p. 402-9. 235. McCarty, D.M., Self-complementary AAV vectors; advances and applications. Mol Ther, 2008. 16(10): p. 1648-56. 236. Hauck, B., et al., Intracellular viral processing, not single-stranded DNA accumulation, is crucial for recombinant adeno-associated virus transduction. J Virol, 2004. 78(24): p. 13678-86. 237. Wang, Z., et al., Rapid and highly efficient transduction by double-stranded adeno-associated virus vectors in vitro and in vivo. Gene therapy, 2003. 10(26): p. 2105-2111. 238. Jiang, H., et al., Evidence of multiyear factor IX expression by AAV-mediated gene transfer to skeletal muscle in an individual with severe hemophilia B. Mol Ther, 2006. 14(3): p. 452-5. 239. Wright, J.F., Manufacturing and characterizing AAV-based vectors for use in clinical studies. Gene Ther, 2008. 15(11): p. 840-8. 240. Xiao, X., J. Li, and R.J. Samulski, Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J Virol, 1998. 72(3): p. 2224-32. 241. Ayuso, E., Manufacturing of recombinant adeno-associated viral vectors: new technologies are welcome. Mol Ther Methods Clin Dev, 2016. 3: p. 15049. 242. Kotin, R.M., Large-scale recombinant adeno-associated virus production. Hum Mol Genet, 2011. 20(R1): p. R2-6. 243. Grieger, J.C., S.M. Soltys, and R.J. Samulski, Production of Recombinant Adeno-associated Virus Vectors Using Suspension HEK293 Cells and Continuous Harvest of Vector From the Culture Media for GMP FIX and FLT1 Clinical Vector. Mol Ther, 2016. 24(2): p. 287-97. 244. Pereira, C.C., et al., Molecular detection of adeno-associated virus in cases of spontaneous and intentional human abortion. J Med Virol, 2010. 82(10): p. 1689-93. 245. Nault, J.C., et al., Recurrent AAV2-related insertional mutagenesis in human hepatocellular carcinomas. Nat Genet, 2015. 47(10): p. 1187-93. 246. Berns, K.I., et al., Adeno-Associated Virus Type 2 and Hepatocellular Carcinoma? Hum Gene Ther, 2015. 26(12): p. 779-81. 247. Buning, H. and M. Schmidt, Adeno-associated Vector Toxicity-To Be or Not to Be? Mol Ther, 2015. 23(11): p. 1673-5. 248. Nault, J.C., et al., AAV2 and Hepatocellular Carcinoma. Hum Gene Ther, 2016. 27(3): p. 211-3. 249. Liu, Q., et al., Neutralizing antibodies against AAV2, AAV5 and AAV8 in healthy and HIV-1-infected subjects in China: implications for gene therapy using AAV vectors. Gene Ther, 2014. 21(8): p. 732-8. 250. Nathwani, A.C., et al., Adenovirus-associated virus vector-mediated gene transfer in hemophilia B. N Engl J Med, 2011. 365(25): p. 2357-65. 251. Mingozzi, F. and K.A. High, Immune responses to AAV in clinical trials. Curr Gene Ther, 2011. 11(4): p. 321-30. 252. Flotte, T., et al., A phase I study of an adeno-associated virus-CFTR gene vector in adult CF patients with mild lung disease. Hum Gene Ther, 1996. 7(9): p. 1145-59.  204 253. Mingozzi, F. and K.A. High, Therapeutic in vivo gene transfer for genetic disease using AAV: progress and challenges. Nat Rev Genet, 2011. 12(5): p. 341-55. 254. Therapeutics, S. Spark Therapeutics Announces Updated Data from First Cohort in Hemophilia B Phase 1/2 Trial Demonstrating Consistent, Sustained Therapeutic Levels of Factor IX Activity. 2016  [cited 2016 October 10]; Available from: http://ir.sparktx.com/phoenix.zhtml?c=253900&p=irol-newsArticle&ID=2177020. 255. Maguire, A.M., et al., Safety and efficacy of gene transfer for Leber's congenital amaurosis. N Engl J Med, 2008. 358(21): p. 2240-8. 256. Clinicaltrials.gov. A Safety and Efficacy Study in Subjects With Leber Congenital Amaurosis (LCA) Using Adeno-Associated Viral Vector to Deliver the Gene for Human RPE65 to the Retinal Pigment Epithelium (RPE) [AAV2-hRPE65v2-301]. 2016; Available from: https://clinicaltrials.gov/show/NCT00999609. 257. Therapeutics, S. Spark Therapeutics Announces New Positive Data from Continuation of Phase 3 Trial of Voretigene Neparvovec. 2016  [cited Oct 12, 2016; Available from: http://ir.sparktx.com/phoenix.zhtml?c=253900&p=irol-newsArticle&ID=2194535. 258. Press release: European Medicines Agency recommends first gene therapy for approval E.M. Agency, Editor. 20 July 2012. 259. Yla-Herttuala, S., Endgame: glybera finally recommended for approval as the first gene therapy drug in the European union. Mol Ther, 2012. 20(10): p. 1831-2. 260. uniQure. uniQure Announces It Will Not Seek Marketing Authorization Renewal for Glybera in Europe. 2017  [cited 2018 July 20]; Available from: http://uniqure.com/GL_PR_Glybera%20withdrawal_FINAL_PDF.pdf. 261. Morrison, C., $1-million price tag set for Glybera gene therapy. Nat Biotechnol, 2015. 33(3): p. 217-8. 262. Yla-Herttuala, S., Glybera's second act: the curtain rises on the high cost of therapy. Mol Ther, 2015. 23(2): p. 217-8. 263. Han, X. and W. Ni, Cost-Effectiveness Analysis of Glybera for The Treatment of Lipoprotein Lipase Deficiency. Value Health, 2015. 18(7): p. A756. 264. Wang, Z., et al., Widespread and stable pancreatic gene transfer by adeno-associated virus vectors via different routes. Diabetes, 2006. 55(4): p. 875-884. 265. Colombo, C., et al., Seven mutations in the human insulin gene linked to permanent neonatal/infancy-onset diabetes mellitus. J Clin Invest, 2008. 118(6): p. 2148-56. 266. Garin, I., et al., Recessive mutations in the INS gene result in neonatal diabetes through reduced insulin biosynthesis. Proc Natl Acad Sci U S A, 2010. 107(7): p. 3105-10. 267. Chretien, M. and C.H. Li, Isolation, purification, and characterization of gamma-lipotropic hormone from sheep pituitary glands. Can J Biochem, 1967. 45(7): p. 1163-74. 268. Chretien, M., How the prohormone theory solved two important controversies in hormonal and neural Peptide biosynthesis. Front Endocrinol (Lausanne), 2013. 4: p. 148. 269. Steiner, D.F., et al., Insulin biosynthesis: evidence for a precursor. Science, 1967. 157(3789): p. 697-700. 270. Brown, M.S. and J.L. Goldstein, A proteolytic pathway that controls the cholesterol content of membranes, cells, and blood. Proc Natl Acad Sci U S A, 1999. 96(20): p. 11041-8. 271. Benjannet, S., et al., NARC-1/PCSK9 and its natural mutants: zymogen cleavage and effects on the low density lipoprotein (LDL) receptor and LDL cholesterol. J Biol Chem, 2004. 279(47): p. 48865-75.  205 272. Seidah, N.G., et al., Testicular expression of PC4 in the rat: molecular diversity of a novel germ cell-specific Kex2/subtilisin-like proprotein convertase. Mol Endocrinol, 1992. 6(10): p. 1559-70. 273. Muller, L. and I. Lindberg, The cell biology of the prohormone convertases PC1 and PC2. Prog Nucleic Acid Res Mol Biol, 1999. 63: p. 69-108. 274. Bryan, P., et al., Site-directed mutagenesis and the role of the oxyanion hole in subtilisin. Proc Natl Acad Sci U S A, 1986. 83(11): p. 3743-5. 275. Davidson, H.W., C.J. Rhodes, and J.C. Hutton, Intraorganellar calcium and pH control proinsulin cleavage in the pancreatic beta cell via two distinct site-specific endopeptidases. Nature, 1988. 333(6168): p. 93-6. 276. Fox, D.L. and D.J. Good, Nescient helix-loop-helix 2 interacts with signal transducer and activator of transcription 3 to regulate transcription of prohormone convertase 1/3. Mol Endocrinol, 2008. 22(6): p. 1438-48. 277. Alarcon, C., B. Lincoln, and C.J. Rhodes, The biosynthesis of the subtilisin-related proprotein convertase PC3, but no that of the PC2 convertase, is regulated by glucose in parallel to proinsulin biosynthesis in rat pancreatic islets. J Biol Chem, 1993. 268(6): p. 4276-80. 278. Hoshino, A., et al., Modulation of PC1/3 activity by self-interaction and substrate binding. Endocrinology, 2011. 152(4): p. 1402-11. 279. Zhu, X., et al., Severe block in processing of proinsulin to insulin accompanied by elevation of des-64,65 proinsulin intermediates in islets of mice lacking prohormone convertase 1/3. Proc Natl Acad Sci U S A, 2002. 99(16): p. 10299-304. 280. Mbikay, M., et al., A targeted deletion/insertion in the mouse Pcsk1 locus is associated with homozygous embryo preimplantation lethality, mutant allele preferential transmission and heterozygous female susceptibility to dietary fat. Dev Biol, 2007. 306(2): p. 584-98. 281. Smeekens, S.P. and D.F. Steiner, Identification of a human insulinoma cDNA encoding a novel mammalian protein structurally related to the yeast dibasic processing protease Kex2. J Biol Chem, 1990. 265(6): p. 2997-3000. 282. Vivoli, M., Lindberg, I, Prohormone Convertase 2. Handbook of Biologically Active Peptides, 2013. 283. Furuta, M., et al., Defective prohormone processing and altered pancreatic islet morphology in mice lacking active SPC2. Proc Natl Acad Sci U S A, 1997. 94(13): p. 6646-51. 284. Peinado, J.R., et al., Strain-dependent influences on the hypothalamo-pituitary-adrenal axis profoundly affect the 7B2 and PC2 null phenotypes. Endocrinology, 2005. 146(8): p. 3438-44. 285. Wardman, J.H., et al., Analysis of peptides in prohormone convertase 1/3 null mouse brain using quantitative peptidomics. J Neurochem, 2010. 114(1): p. 215-25. 286. Zhang, X., et al., Neuropeptidomic analysis establishes a major role for prohormone convertase-2 in neuropeptide biosynthesis. J Neurochem, 2010. 112(5): p. 1168-79. 287. Schechter, I. and A. Berger, On the size of the active site in proteases. I. Papain. Biochem Biophys Res Commun, 1967. 27(2): p. 157-62. 288. Henrich, S., et al., Proprotein convertase models based on the crystal structures of furin and kexin: explanation of their specificity. J Mol Biol, 2005. 345(2): p. 211-27. 289. Zhu, X., et al., On the processing of proghrelin to ghrelin. J Biol Chem, 2006. 281(50): p. 38867-70.  206 290. Fricker, L.D. and S.H. Snyder, Enkephalin convertase: purification and characterization of a specific enkephalin-synthesizing carboxypeptidase localized to adrenal chromaffin granules. Proc Natl Acad Sci U S A, 1982. 79(12): p. 3886-90. 291. Naggert, J.K., et al., Hyperproinsulinaemia in obese fat/fat mice associated with a carboxypeptidase E mutation which reduces enzyme activity. Nat Genet, 1995. 10(2): p. 135-42. 292. Ji, L., et al., Dissecting carboxypeptidase E: properties, functions and pathophysiological roles in disease. Endocr Connect, 2017. 6(4): p. R18-R38. 293. Srinivasan, S., et al., Deficits in reproduction and pro-gonadotropin-releasing hormone processing in male Cpefat mice. Endocrinology, 2004. 145(4): p. 2023-34. 294. Cawley, N.X., et al., Obese carboxypeptidase E knockout mice exhibit multiple defects in peptide hormone processing contributing to low bone mineral density. Am J Physiol Endocrinol Metab, 2010. 299(2): p. E189-97. 295. Lacourse, K.A., et al., Disturbed progastrin processing in carboxypeptidase E-deficient fat mice. FEBS Lett, 1997. 416(1): p. 45-50. 296. Fortenberry, Y., et al., Functional characterization of ProSAAS: similarities and differences with 7B2. J Biol Chem, 2002. 277(7): p. 5175-86. 297. Feng, Y., S.E. Reznik, and L.D. Fricker, Distribution of proSAAS-derived peptides in rat neuroendocrine tissues. Neuroscience, 2001. 105(2): p. 469-78. 298. Morgan, D.J., et al., The propeptide precursor proSAAS is involved in fetal neuropeptide processing and body weight regulation. J Neurochem, 2010. 113(5): p. 1275-84. 299. Wardman, J.H. and L.D. Fricker, ProSAAS-derived peptides are differentially processed and sorted in mouse brain and AtT-20 cells. PLoS One, 2014. 9(8): p. e104232. 300. Sayah, M., et al., Tissue distribution and processing of proSAAS by proprotein convertases. J Neurochem, 2001. 76(6): p. 1833-41. 301. Mbikay, M., N.G. Seidah, and M. Chretien, Neuroendocrine secretory protein 7B2: structure, expression and functions. Biochem J, 2001. 357(Pt 2): p. 329-42. 302. Lee, S.N. and I. Lindberg, 7B2 prevents unfolding and aggregation of prohormone convertase 2. Endocrinology, 2008. 149(8): p. 4116-27. 303. Westphal, C.H., et al., The neuroendocrine protein 7B2 is required for peptide hormone processing in vivo and provides a novel mechanism for pituitary Cushing's disease. Cell, 1999. 96(5): p. 689-700. 304. Seidel, B., et al., Neuroendocrine protein 7B2 is essential for proteolytic conversion and activation of proprotein convertase 2 in vivo. DNA Cell Biol, 1998. 17(12): p. 1017-29. 305. Peinado, J.R., et al., Blockade of islet amyloid polypeptide fibrillation and cytotoxicity by the secretory chaperones 7B2 and proSAAS. FEBS Lett, 2013. 587(21): p. 3406-11. 306. Helwig, M., et al., The neuroendocrine protein 7B2 suppresses the aggregation of neurodegenerative disease-related proteins. J Biol Chem, 2013. 288(2): p. 1114-24. 307. Steiner, D.F. and P.E. Oyer, The biosynthesis of insulin and a probable precursor of insulin by a human islet cell adenoma. Proc Natl Acad Sci U S A, 1967. 57(2): p. 473-80. 308. Chance, R.E., R.M. Ellis, and W.W. Bromer, Porcine proinsulin: characterization and amino acid sequence. Science, 1968. 161(3837): p. 165-7. 309. Clark, J.L. and D.F. Steiner, Insulin biosynthesis in the rat: demonstration of two proinsulins. Proc Natl Acad Sci U S A, 1969. 62(1): p. 278-85. 310. Davidson, H.W., (Pro)Insulin processing: a historical perspective. Cell Biochem Biophys, 2004. 40(3 Suppl): p. 143-58. 311. Davidson, H.W., M. Peshavaria, and J.C. Hutton, Proteolytic conversion of proinsulin into insulin. Identification of a Ca2+-dependent acidic endopeptidase in isolated insulin-secretory granules. Biochem J, 1987. 246(2): p. 279-86.  207 312. Smeekens, S.P., et al., Identification of a cDNA encoding a second putative prohormone convertase related to PC2 in AtT20 cells and islets of Langerhans. Proc Natl Acad Sci U S A, 1991. 88(2): p. 340-4. 313. 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, 1998. 273(6): p. 3431-7. 314. Rhodes, C.J., B. Lincoln, and S.E. Shoelson, 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, 1992. 267(32): p. 22719-27. 315. Halban, P.A., Proinsulin processing in the regulated and the constitutive secretory pathway. Diabetologia, 1994. 37 Suppl 2: p. S65-72. 316. Scharfmann, R., et al., Development of a conditionally immortalized human pancreatic beta cell line. J Clin Invest, 2014. 124(5): p. 2087-98. 317. Benner, C., et al., The transcriptional landscape of mouse beta cells compared to human beta cells reveals notable species differences in long non-coding RNA and protein-coding gene expression. BMC Genomics, 2014. 15: p. 620. 318. Bramswig, N.C., et al., Epigenomic plasticity enables human pancreatic alpha to beta cell reprogramming. J Clin Invest, 2013. 123(3): p. 1275-84. 319. Muraro, M.J., et al., A Single-Cell Transcriptome Atlas of the Human Pancreas. Cell Syst, 2016. 3(4): p. 385-394 e3. 320. Segerstolpe, A., et al., Single-Cell Transcriptome Profiling of Human Pancreatic Islets in Health and Type 2 Diabetes. Cell Metab, 2016. 24(4): p. 593-607. 321. Baron, M., et al., A Single-Cell Transcriptomic Map of the Human and Mouse Pancreas Reveals Inter- and Intra-cell Population Structure. Cell Syst, 2016. 322. Davalli, A.M., et al., Disproportionate hyperproinsulinemia, beta-cell restricted prohormone convertase 2 deficiency, and cell cycle inhibitors expression by human islets transplanted into athymic nude mice: insights into nonimmune-mediated mechanisms of delayed islet graft failure. Cell Transplant, 2008. 17(12): p. 1323-36. 323. Itoh, Y., et al., Prohormone convertases (PC1/3 and PC2) in rat and human pancreas and islet cell tumors: subcellular immunohistochemical analysis. Pathol Int, 1996. 46(10): p. 726-37. 324. Steiner, D.J., et al., Pancreatic islet plasticity: interspecies comparison of islet architecture and composition. Islets, 2010. 2(3): p. 135-45. 325. Chang, T.J., et al., Genetic polymorphisms of PCSK2 are associated with glucose homeostasis and progression to type 2 diabetes in a Chinese population. Sci Rep, 2015. 5: p. 14380. 326. Zheng, X., et al., Association of type 2 diabetes susceptibility genes (TCF7L2, SLC30A8, PCSK1 and PCSK2) and proinsulin conversion in a Chinese population. Mol Biol Rep, 2012. 39(1): p. 17-23. 327. Stijnen, P., et al., PCSK1 mutations and human endocrinopathies: from obesity to gastrointestinal disorders. Endocr Rev, 2016: p. er20151117. 328. Jackson, R.S., et al., Obesity and impaired prohormone processing associated with mutations in the human prohormone convertase 1 gene. Nat Genet, 1997. 16(3): p. 303-6. 329. O'Rahilly, S., et al., Brief report: impaired processing of prohormones associated with abnormalities of glucose homeostasis and adrenal function. N Engl J Med, 1995. 333(21): p. 1386-90.  208 330. Zhu, X., et al., Disruption of PC1/3 expression in mice causes dwarfism and multiple neuroendocrine peptide processing defects. Proc Natl Acad Sci U S A, 2002. 99(16): p. 10293-8. 331. Rodriguez-Calvo, T., et al., Increase in Pancreatic Proinsulin and Preservation of beta-Cell Mass in Autoantibody-Positive Donors Prior to Type 1 Diabetes Onset. Diabetes, 2017. 66(5): p. 1334-1345. 332. Truyen, I., et al., Proinsulin levels and the proinsulin:c-peptide ratio complement autoantibody measurement for predicting type 1 diabetes. Diabetologia, 2005. 48(11): p. 2322-9. 333. Sims, E.K., et al., Elevations in the Fasting Serum Proinsulin-to-C-Peptide Ratio Precede the Onset of Type 1 Diabetes. Diabetes Care, 2016. 39(9): p. 1519-26. 334. Ludvigsson, J. and L. Heding, Abnormal proinsulin/C-peptide ratio in juvenile diabetes. Acta Diabetol Lat, 1982. 19(4): p. 351-8. 335. Watkins, R.A., et al., Proinsulin and heat shock protein 90 as biomarkers of beta-cell stress in the early period after onset of type 1 diabetes. Transl Res, 2016. 168: p. 96-106 e1. 336. Scholin, A., et al., Proinsulin/C-peptide ratio, glucagon and remission in new-onset Type 1 diabetes mellitus in young adults. Diabet Med, 2011. 28(2): p. 156-61. 337. Steenkamp, D.W., et al., Preserved Proinsulin Secretion in Long-Standing Type 1 Diabetes. Endocr Pract, 2017. 23(12): p. 1387-1393. 338. Sims, E.K., et al., Proinsulin Secretion Is a Persistent Feature of Type 1 Diabetes. Diabetes Care, 2019. 42(2): p. 258-264. 339. Courtade, J.A., et al., Measurement of Pro-Islet Amyloid Polypeptide (1-48) in Diabetes and Islet Transplants. J Clin Endocrinol Metab, 2017. 102(7): p. 2595-2603. 340. Wasserfall, C., et al., Persistence of Pancreatic Insulin mRNA Expression and Proinsulin Protein in Type 1 Diabetes Pancreata. Cell Metab, 2017. 26(3): p. 568-575 e3. 341. Kahn, S.E., et al., Proinsulin as a marker for the development of NIDDM in Japanese-American men. Diabetes, 1995. 44(2): p. 173-9. 342. Kahn, S.E. and P.A. Halban, Release of incompletely processed proinsulin is the cause of the disproportionate proinsulinemia of NIDDM. Diabetes, 1997. 46(11): p. 1725-32. 343. Larsson, H. and B. Ahren, Relative hyperproinsulinemia as a sign of islet dysfunction in women with impaired glucose tolerance. J Clin Endocrinol Metab, 1999. 84(6): p. 2068-74. 344. Mykkanen, L., et al., Serum proinsulin levels are disproportionately increased in elderly prediabetic subjects. Diabetologia, 1995. 38(10): p. 1176-82. 345. Revers, R.R., et al., The effects of biosynthetic human proinsulin on carbohydrate metabolism. Diabetes, 1984. 33(8): p. 762-70. 346. Galloway, J.A., et al., Biosynthetic human proinsulin. Review of chemistry, in vitro and in vivo receptor binding, animal and human pharmacology studies, and clinical trial experience. Diabetes Care, 1992. 15(5): p. 666-92. 347. Kaufmann, J.E., et al., Proinsulin conversion in GH3 cells after coexpression of human proinsulin with the endoproteases PC2 and/or PC3. Diabetes, 1997. 46(6): p. 978-82. 348. Irminger, J.C., et al., Human proinsulin conversion in the regulated and the constitutive pathways of transfected AtT20 cells. J Biol Chem, 1994. 269(3): p. 1756-62. 349. Marban, S.L., Roth, J., Transgenic hyperinsulinemia: A mouse model of insulin resistance and glucose intolerance without obesity. 1996: p. 201-224. 350. Sizonenko, S., et al., Kinetics of proinsulin conversion in human islets. Diabetes, 1993. 42(6): p. 933-6.  209 351. Pan, F.C. and M. Brissova, Pancreas development in humans. Curr Opin Endocrinol Diabetes Obes, 2014. 21(2): p. 77-82. 352. Teitelman, G., et al., Precursor cells of mouse endocrine pancreas coexpress insulin, glucagon and the neuronal proteins tyrosine hydroxylase and neuropeptide Y, but not pancreatic polypeptide. Development, 1993. 118(4): p. 1031-9. 353. Richter, E.A. and M. Hargreaves, Exercise, GLUT4, and skeletal muscle glucose uptake. Physiol Rev, 2013. 93(3): p. 993-1017. 354. Schade, D.S. and R.P. Eaton, Pathogenesis of diabetic ketoacidosis: a reappraisal. Diabetes Care, 1979. 2(3): p. 296-306. 355. Kulkarni, R.N., et al., Tissue-specific knockout of the insulin receptor in pancreatic beta cells creates an insulin secretory defect similar to that in type 2 diabetes. Cell, 1999. 96(3): p. 329-39. 356. Okada, T., et al., Insulin receptors in beta-cells are critical for islet compensatory growth response to insulin resistance. Proc Natl Acad Sci U S A, 2007. 104(21): p. 8977-82. 357. Magnuson, M.A. and A.B. Osipovich, Pancreas-specific Cre driver lines and considerations for their prudent use. Cell Metab, 2013. 18(1): p. 9-20. 358. Wicksteed, B., et al., Conditional gene targeting in mouse pancreatic ss-Cells: analysis of ectopic Cre transgene expression in the brain. Diabetes, 2010. 59(12): p. 3090-8. 359. Bruning, J.C., et al., Role of brain insulin receptor in control of body weight and reproduction. Science, 2000. 289(5487): p. 2122-5. 360. Trinder, M., et al., beta-cell insulin receptor deficiency during in utero development induces an islet compensatory overgrowth response. Oncotarget, 2016. 361. Carboneau, B.A., et al., Unexpected effects of the MIP-CreER transgene and tamoxifen on beta-cell growth in C57Bl6/J male mice. Physiol Rep, 2016. 4(18). 362. Oropeza, D., et al., Phenotypic Characterization of MIP-CreERT1Lphi Mice With Transgene-Driven Islet Expression of Human Growth Hormone. Diabetes, 2015. 64(11): p. 3798-807. 363. Ullrich, A., et al., Insulin-like growth factor I receptor primary structure: comparison with insulin receptor suggests structural determinants that define functional specificity. EMBO J, 1986. 5(10): p. 2503-12. 364. Morgan, D.O., et al., Insulin-like growth factor II receptor as a multifunctional binding protein. Nature, 1987. 329(6137): p. 301-7. 365. Kitamura, T., C.R. Kahn, and D. Accili, Insulin receptor knockout mice. Annu Rev Physiol, 2003. 65: p. 313-32. 366. Louvi, A., D. Accili, and A. Efstratiadis, Growth-promoting interaction of IGF-II with the insulin receptor during mouse embryonic development. Dev Biol, 1997. 189(1): p. 33-48. 367. Duvillie, B., et al., Phenotypic alterations in insulin-deficient mutant mice. Proceedings of the National Academy of Sciences of the United States of America, 1997. 94(10): p. 5137-5140. 368. Neumann, U.H., et al., Glucagon receptor gene deletion in insulin knockout mice modestly reduces blood glucose and ketones but does not promote survival. Mol Metab, 2016. 5(8): p. 731-6. 369. Danan, C., et al., [Physiopathological approach and antenatal diagnosis of diabetes mellitus insulin resistant: apropos of a case with leprechaunism]. Arch Pediatr, 1994. 1(3): p. 268-72. 370. Duvillie, B., et al., Increased islet cell proliferation, decreased apoptosis, and greater vascularization leading to beta-cell hyperplasia in mutant mice lacking insulin. Endocrinology, 2002. 143(4): p. 1530-1537.  210 371. Lacy, P.E. and M. Kostianovsky, Method for the isolation of intact islets of Langerhans from the rat pancreas. Diabetes, 1967. 16(1): p. 35-39. 372. Mojibian, M., et al., Implanted islets in the anterior chamber of the eye are prone to autoimmune attack in a mouse model of diabetes. Diabetologia, 2013. 56(10): p. 2213-21. 373. Lukinius, A., et al., Expression of islet amyloid polypeptide in fetal and adult porcine and human pancreatic islet cells. Endocrinology, 1997. 138(4): p. 5319-25. 374. Chomczynski, P. and N. Sacchi, The single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction: twenty-something years on. Nat Protoc, 2006. 1(2): p. 581-5. 375. Leroux, L., et al., Compensatory responses in mice carrying a null mutation for Ins1 or Ins2. Diabetes, 2001. 50 Suppl 1: p. S150-3. 376. Bock, T., B. Pakkenberg, and K. Buschard, Genetic background determines the size and structure of the endocrine pancreas. Diabetes, 2005. 54(1): p. 133-7. 377. Baan, M., et al., Transgenic expression of the human growth hormone minigene promotes pancreatic beta-cell proliferation. Am J Physiol Regul Integr Comp Physiol, 2015. 309(7): p. R788-94. 378. Brouwers, B., et al., Impaired islet function in commonly used transgenic mouse lines due to human growth hormone minigene expression. Cell Metab, 2014. 20(6): p. 979-90. 379. Kjeldsen, T., et al., The ligand specificities of the insulin receptor and the insulin-like growth factor I receptor reside in different regions of a common binding site. Proc Natl Acad Sci U S A, 1991. 88(10): p. 4404-8. 380. Lee, Y., et al., Beta-cell lipotoxicity in the pathogenesis of non-insulin-dependent diabetes mellitus of obese rats: impairment in adipocyte-beta-cell relationships. Proc Natl Acad Sci U S A, 1994. 91(23): p. 10878-82. 381. Shimo, N., et al., Short-term selective alleviation of glucotoxicity and lipotoxicity ameliorates the suppressed expression of key beta-cell factors under diabetic conditions. Biochem Biophys Res Commun, 2015. 467(4): p. 948-54. 382. Kim-Muller, J.Y., et al., Aldehyde dehydrogenase 1a3 defines a subset of failing pancreatic beta cells in diabetic mice. Nat Commun, 2016. 7: p. 12631. 383. Gradwohl, G., et al., neurogenin3 is required for the development of the four endocrine cell lineages of the pancreas. Proc Natl Acad Sci U S A, 2000. 97(4): p. 1607-11. 384. Wang, S., et al., Sustained Neurog3 expression in hormone-expressing islet cells is required for endocrine maturation and function. Proc Natl Acad Sci U S A, 2009. 106(24): p. 9715-20. 385. Jensen, J., et al., Independent development of pancreatic alpha- and beta-cells from neurogenin3-expressing precursors: a role for the notch pathway in repression of premature differentiation. Diabetes, 2000. 49(2): p. 163-76. 386. Georgia, S. and A. Bhushan, Beta cell replication is the primary mechanism for maintaining postnatal beta cell mass. J Clin Invest, 2004. 114(7): p. 963-8. 387. Anderson, K.R., et al., Identification of known and novel pancreas genes expressed downstream of Nkx2.2 during development. BMC Dev Biol, 2009. 9: p. 65. 388. Tonne, J.M., et al., Beta cell regeneration after single-round immunological destruction in a mouse model. Diabetologia, 2015. 58(2): p. 313-23. 389. Baeyens, L., et al., Ngn3 expression during postnatal in vitro beta cell neogenesis induced by the JAK/STAT pathway. Cell Death Differ, 2006. 13(11): p. 1892-9. 390. Cavaleri, F. and H.R. Scholer, Nanog: a new recruit to the embryonic stem cell orchestra. Cell, 2003. 113(5): p. 551-2.  211 391. Luo, W., et al., Embryonic stem cells markers SOX2, OCT4 and Nanog expression and their correlations with epithelial-mesenchymal transition in nasopharyngeal carcinoma. PLoS One, 2013. 8(2): p. e56324. 392. Gu, T.T., S.Y. Liu, and P.S. Zheng, Cytoplasmic NANOG-positive stromal cells promote human cervical cancer progression. Am J Pathol, 2012. 181(2): p. 652-61. 393. Amsterdam, A., et al., LGR5 and Nanog identify stem cell signature of pancreas beta cells which initiate pancreatic cancer. Biochem Biophys Res Commun, 2013. 433(2): p. 157-62. 394. Werner, H. and E.A. Chantelau, Differences in bioactivity between human insulin and insulin analogues approved for therapeutic use- compilation of reports from the past 20 years. Diabetol Metab Syndr, 2011. 3(1): p. 13. 395. Pepper, A.R., et al., Diabetic rats and mice are resistant to porcine and human insulin: flawed experimental models for testing islet xenografts. Xenotransplantation, 2009. 16(6): p. 502-10. 396. Remedi, M.S., et al., Secondary consequences of beta cell inexcitability: identification and prevention in a murine model of K(ATP)-induced neonatal diabetes mellitus. Cell Metab, 2009. 9(2): p. 140-51. 397. Bugliani, M., et al., Effects of C-peptide on isolated human pancreatic islet cells. Diabetes Metab Res Rev, 2007. 23(3): p. 215-9. 398. Collombat, P., et al., Embryonic endocrine pancreas and mature beta cells acquire alpha and PP cell phenotypes upon Arx misexpression. J Clin Invest, 2007. 117(4): p. 961-70. 399. Kawamori, D., et al., Insulin signaling in alpha cells modulates glucagon secretion in vivo. Cell metabolism, 2009. 9(4): p. 350-361. 400. Li, Z., F.A. Karlsson, and S. Sandler, Islet loss and alpha cell expansion in type 1 diabetes induced by multiple low-dose streptozotocin administration in mice. J Endocrinol, 2000. 165(1): p. 93-9. 401. Mehran, A.E., et al., Hyperinsulinemia drives diet-induced obesity independently of brain insulin production. Cell Metab, 2012. 16(6): p. 723-37. 402. Zhang, C., et al., MafA is a key regulator of glucose-stimulated insulin secretion. Mol Cell Biol, 2005. 25(12): p. 4969-76. 403. Poitout, V., et al., Regulation of the insulin gene by glucose and fatty acids. J Nutr, 2006. 136(4): p. 873-6. 404. Robertson, G., et al., Age-dependent silencing of globin transgenes in the mouse. Nucleic Acids Res, 1996. 24(8): p. 1465-71. 405. Kirov, J.V., et al., Reporter Gene Silencing in Targeted Mouse Mutants Is Associated with Promoter CpG Island Methylation. PLoS One, 2015. 10(8): p. e0134155. 406. Wentworth, B.M., et al., The ratio of mouse insulin I:insulin II does not reflect that of the corresponding preproinsulin mRNAs. Mol Cell Endocrinol, 1992. 86(3): p. 177-86. 407. Deltour, L., et al., Differential expression and imprinting status of Ins1 and Ins2 genes in extraembryonic tissues of laboratory mice. Gene Expr Patterns, 2004. 5(2): p. 297-300. 408. Raile, K., et al., Diabetes caused by insulin gene (INS) deletion: clinical characteristics of homozygous and heterozygous individuals. Eur J Endocrinol, 2011. 165(2): p. 255-60. 409. Greeley, S.A., et al., Case Report: Preservation of Reduced Numbers of Insulin-Positive Cells in Sulfonylurea-Unresponsive KCNJ11-related Diabetes. J Clin Endocrinol Metab, 2016: p. jc20162826. 410. Babiker, T., et al., Successful transfer to sulfonylureas in KCNJ11 neonatal diabetes is determined by the mutation and duration of diabetes. Diabetologia, 2016. 59(6): p. 1162-6.  212 411. Rhodes, C.J., et al., Direct autocrine action of insulin on beta-cells: does it make physiological sense? Diabetes, 2013. 62(7): p. 2157-63. 412. Postic, C., et al., Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-specific gene knock-outs using Cre recombinase. J Biol Chem, 1999. 274(1): p. 305-15. 413. Lee, J.Y., et al., RIP-Cre revisited, evidence for impairments of pancreatic beta-cell function. J Biol Chem, 2006. 281(5): p. 2649-53. 414. Teitelman, G. and M. Kedees, Mouse insulin cells expressing an inducible RIPCre transgene are functionally impaired. J Biol Chem, 2015. 290(6): p. 3647-53. 415. Gannon, M., et al., Analysis of the Cre-mediated recombination driven by rat insulin promoter in embryonic and adult mouse pancreas. Genesis, 2000. 26(2): p. 139-42. 416. Sabatini, P.V., et al., Neuronal PAS Domain Protein 4 Suppression of Oxygen Sensing Optimizes Metabolism during Excitation of Neuroendocrine Cells. Cell Rep, 2018. 22(1): p. 163-174. 417. Thorens, B., et al., Ins1(Cre) knock-in mice for beta cell-specific gene recombination. Diabetologia, 2015. 58(3): p. 558-65. 418. Cheng, Y., et al., Generation and Characterization of Transgenic Mice Expressing Mouse Ins1 Promoter for Pancreatic beta-Cell-Specific Gene Overexpression and Knockout. Endocrinology, 2015. 156(7): p. 2724-31. 419. Liu, Y., et al., Tamoxifen-independent recombination in the RIP-CreER mouse. PLoS One, 2010. 5(10): p. e13533. 420. Feil, R., et al., Regulation of Cre recombinase activity by mutated estrogen receptor ligand-binding domains. Biochem Biophys Res Commun, 1997. 237(3): p. 752-7. 421. Muzumdar, M.D., et al., A global double-fluorescent Cre reporter mouse. Genesis, 2007. 45(9): p. 593-605. 422. Safran, M., et al., Mouse reporter strain for noninvasive bioluminescent imaging of cells that have undergone Cre-mediated recombination. Mol Imaging, 2003. 2(4): p. 297-302. 423. Snippert, H.J., et al., Intestinal crypt homeostasis results from neutral competition between symmetrically dividing Lgr5 stem cells. Cell, 2010. 143(1): p. 134-44. 424. Fan, Y., et al., Thymus-specific deletion of insulin induces autoimmune diabetes. EMBO J, 2009. 28(18): p. 2812-24. 425. Salvalaggio, P.R., et al., Islet filtration: a simple and rapid new purification procedure that avoids ficoll and improves islet mass and function. Transplantation, 2002. 74(6): p. 877-9. 426. Tuduri, E., et al., Impaired Ca(2+) signaling in beta-cells lacking leptin receptors by Cre-loxP recombination. PLoS One, 2013. 8(8): p. e71075. 427. Denroche, H.C., et al., Disrupted Leptin Signaling in the Lateral Hypothalamus and Ventral Premammillary Nucleus Alters Insulin and Glucagon Secretion and Protects Against Diet-Induced Obesity. Endocrinology, 2016. 157(7): p. 2671-85. 428. Zincarelli, C., et al., Analysis of AAV serotypes 1-9 mediated gene expression and tropism in mice after systemic injection. Mol Ther, 2008. 16(6): p. 1073-80. 429. Leonard, J., et al., Characterization of somatostatin transactivating factor-1, a novel homeobox factor that stimulates somatostatin expression in pancreatic islet cells. Mol Endocrinol, 1993. 7(10): p. 1275-83. 430. Szabat, M., et al., Reduced Insulin Production Relieves Endoplasmic Reticulum Stress and Induces beta Cell Proliferation. Cell Metab, 2016. 23(1): p. 179-93. 431. Johnson, J.D., A practical guide to genetic engineering of pancreatic beta-cells in vivo: getting a grip on RIP and MIP. Islets, 2014. 6(3): p. e944439.  213 432. Chen, X., et al., In vivo detection of extrapancreatic insulin gene expression in diabetic mice by bioluminescence imaging. PLoS One, 2010. 5(2): p. e9397. 433. Kojima, H., et al., Extrapancreatic insulin-producing cells in multiple organs in diabetes. Proc Natl Acad Sci U S A, 2004. 101(8): p. 2458-63. 434. Wang, A.Y., et al., Comparison of adenoviral and adeno-associated viral vectors for pancreatic gene delivery in vivo. Human Gene Therapy, 2004. 15(4): p. 405-413. 435. Morro, M., et al., Pancreatic transduction by helper-dependent adenoviral vectors via intraductal delivery. Hum Gene Ther, 2014. 25(9): p. 824-36. 436. Chen, J., et al., Production and clinical development of nanoparticles for gene delivery. Mol Ther Methods Clin Dev, 2016. 3: p. 16023. 437. Bouwens, L., Beta cell regeneration. Curr Diabetes Rev, 2006. 2(1): p. 3-9. 438. Rorsman, P. and E. Renstrom, Insulin granule dynamics in pancreatic beta cells. Diabetologia, 2003. 46(8): p. 1029-45. 439. Anello, M., P. Gilon, and J.C. Henquin, Alterations of insulin secretion from mouse islets treated with sulphonylureas: perturbations of Ca2+ regulation prevail over changes in insulin content. Br J Pharmacol, 1999. 127(8): p. 1883-91. 440. Tillmar, L., C. Carlsson, and N. Welsh, Control of insulin mRNA stability in rat pancreatic islets. Regulatory role of a 3'-untranslated region pyrimidine-rich sequence. J Biol Chem, 2002. 277(2): p. 1099-106. 441. Hattersley, A.T., et al., ISPAD Clinical Practice Consensus Guidelines 2018: The diagnosis and management of monogenic diabetes in children and adolescents. Pediatr Diabetes, 2018. 19 Suppl 27: p. 47-63. 442. Ramzy, A., M. Mojibian, and T.J. Kieffer, Insulin-Deficient Mouse beta-Cells Do Not Fully Maturebut Can Be Remedied Through Insulin Replacementby Islet Transplantation. Endocrinology, 2018. 159(1): p. 83-102. 443. Gu, G., J. Dubauskaite, and D.A. Melton, Direct evidence for the pancreatic lineage: NGN3+ cells are islet progenitors and are distinct from duct progenitors. Development, 2002. 129(10): p. 2447-57. 444. Mulder, H., A.C. Lindh, and F. Sundler, Islet amyloid polypeptide gene expression in the endocrine pancreas of the rat: a combined in situ hybridization and immunocytochemical study. Cell Tissue Res, 1993. 274(3): p. 467-74. 445. Peavy, D.E., et al., Receptor binding and biological potency of several split forms (conversion intermediates) of human proinsulin. Studies in cultured IM-9 lymphocytes and in vivo and in vitro in rats. J Biol Chem, 1985. 260(26): p. 13989-94. 446. Yu, S.S. and A.E. Kitbachi, Biological activity of proinsulin and related polypeptides in the fat tissue. J Biol Chem, 1973. 248(11): p. 3753-61. 447. Marzban, L., et al., Role of beta-cell prohormone convertase (PC)1/3 in processing of pro-islet amyloid polypeptide. Diabetes, 2004. 53(1): p. 141-8. 448. Wang, J., et al., The prohormone convertase enzyme 2 (PC2) is essential for processing pro-islet amyloid polypeptide at the NH2-terminal cleavage site. Diabetes, 2001. 50(3): p. 534-9. 449. Riedel, M.J., et al., DsAAV8-mediated expression of glucagon-like peptide-1 in pancreatic beta-cells ameliorates streptozotocin-induced diabetes. Gene Ther, 2010. 17(2): p. 171-80. 450. Courtland, H.W., et al., Growth hormone mediates pubertal skeletal development independent of hepatic IGF-1 production. J Bone Miner Res, 2011. 26(4): p. 761-8. 451. Ryan, E.A., et al., Five-year follow-up after clinical islet transplantation. Diabetes, 2005. 54(7): p. 2060-9.  214 452. Nayak, S. and R.W. Herzog, Progress and prospects: immune responses to viral vectors. Gene Ther, 2010. 17(3): p. 295-304. 453. Scaglia, L., et al., Apoptosis participates in the remodeling of the endocrine pancreas in the neonatal rat. Endocrinology, 1997. 138(4): p. 1736-41. 454. Stolovich-Rain, M., et al., Weaning triggers a maturation step of pancreatic beta cells. Dev Cell, 2015. 32(5): p. 535-45. 455. Marban, S.L., J.A. DeLoia, and J.D. Gearhart, Hyperinsulinemia in transgenic mice carrying multiple copies of the human insulin gene. Dev Genet, 1989. 10(5): p. 356-64. 456. Kulkarni, R.N., et al., beta-cell-specific deletion of the Igf1 receptor leads to hyperinsulinemia and glucose intolerance but does not alter beta-cell mass. Nat Genet, 2002. 31(1): p. 111-5. 457. Brereton, M.F., et al., Reversible changes in pancreatic islet structure and function produced by elevated blood glucose. Nat Commun, 2014. 5: p. 4639. 458. Kilimnik, G., et al., Intraislet production of GLP-1 by activation of prohormone convertase 1/3 in pancreatic alpha-cells in mouse models of ss-cell regeneration. Islets, 2010. 2(3): p. 149-55. 459. O'Malley, T.J., et al., Progressive change of intra-islet GLP-1 production during diabetes development. Diabetes Metab Res Rev, 2014. 30(8): p. 661-8. 460. Zummo, F.P., et al., Glucagon-Like Peptide 1 Protects Pancreatic beta-Cells From Death by Increasing Autophagic Flux and Restoring Lysosomal Function. Diabetes, 2017. 66(5): p. 1272-1285. 461. Wideman, R.D., et al., A switch from prohormone convertase (PC)-2 to PC1/3 expression in transplanted alpha-cells is accompanied by differential processing of proglucagon and improved glucose homeostasis in mice. Diabetes, 2007. 56(11): p. 2744-52. 462. Golson, M.L., W.S. Bush, and M. Brissova, Automated quantification of pancreatic beta-cell mass. Am J Physiol Endocrinol Metab, 2014. 306(12): p. E1460-7. 463. Ran, F.A., et al., In vivo genome editing using Staphylococcus aureus Cas9. Nature, 2015. 520(7546): p. 186-91. 464. Cunningham, C.N., et al., Chaperone-Driven Degradation of a Misfolded Proinsulin Mutant in Parallel With Restoration of Wild-Type Insulin Secretion. Diabetes, 2017. 66(3): p. 741-753. 465. Koster, J.C., et al., Targeted overactivity of beta cell K(ATP) channels induces profound neonatal diabetes. Cell, 2000. 100(6): p. 645-54. 466. Shapiro, A.M., et al., International trial of the Edmonton protocol for islet transplantation. N Engl J Med, 2006. 355(13): p. 1318-30. 467. Paulsson, J.F. and G.T. Westermark, Aberrant processing of human proislet amyloid polypeptide results in increased amyloid formation. Diabetes, 2005. 54(7): p. 2117-25. 468. Vivoli, M., et al., Inhibition of prohormone convertases PC1/3 and PC2 by 2,5-dideoxystreptamine derivatives. Mol Pharmacol, 2012. 81(3): p. 440-54. 469. Yongye, A.B., et al., Identification of a small molecule that selectively inhibits mouse PC2 over mouse PC1/3: a computational and experimental study. PLoS One, 2013. 8(2): p. e56957. 470. Becker, G.L., et al., Highly potent inhibitors of proprotein convertase furin as potential drugs for treatment of infectious diseases. J Biol Chem, 2012. 287(26): p. 21992-2003. 471. Powers, A.C., et al., Proglucagon processing similar to normal islets in pancreatic alpha-like cell line derived from transgenic mouse tumor. Diabetes, 1990. 39(4): p. 406-14. 472. Webb, G.C., et al., Altered proglucagon processing in an alpha-cell line derived from prohormone convertase 2 null mouse islets. J Biol Chem, 2004. 279(30): p. 31068-75.  215 473. Efrat, S., et al., Beta-cell lines derived from transgenic mice expressing a hybrid insulin gene-oncogene. Proc Natl Acad Sci U S A, 1988. 85(23): p. 9037-41. 474. Marzban, L., G. Trigo-Gonzalez, and C.B. Verchere, Processing of pro-islet amyloid polypeptide in the constitutive and regulated secretory pathways of beta cells. Mol Endocrinol, 2005. 19(8): p. 2154-63. 475. Schagger, H., Tricine-SDS-PAGE. Nat Protoc, 2006. 1(1): p. 16-22. 476. Dyballa, N. and S. Metzger, Fast and sensitive colloidal coomassie G-250 staining for proteins in polyacrylamide gels. J Vis Exp, 2009(30). 477. Bukowski, T.R., et al., Triple-label beta liquid scintillation counting. Anal Biochem, 1992. 204(1): p. 171-80. 478. Lamango, N.S., X. Zhu, and I. Lindberg, Purification and enzymatic characterization of recombinant prohormone convertase 2: stabilization of activity by 21 kDa 7B2. Arch Biochem Biophys, 1996. 330(2): p. 238-50. 479. Rouille, Y., S. Martin, and D.F. Steiner, Differential processing of proglucagon by the subtilisin-like prohormone convertases PC2 and PC3 to generate either glucagon or glucagon-like peptide. J Biol Chem, 1995. 270(44): p. 26488-96. 480. Lam, C.J., et al., Low-Level Insulin Content Within Abundant Non-beta Islet Endocrine Cells in Long-standing Type 1 Diabetes. Diabetes, 2019. 68(3): p. 598-608. 481. Teitelman, G., Heterogeneous Expression of Proinsulin Processing Enzymes in Beta Cells of Non-diabetic and Type 2 Diabetic Humans. J Histochem Cytochem, 2019: p. 22155419831641. 482. Scopsi, L., et al., Proprotein convertases (PC1/PC3 and PC2) in normal and neoplastic human tissues: their use as markers of neuroendocrine differentiation. J Clin Endocrinol Metab, 1995. 80(1): p. 294-301. 483. Nie, Y., et al., Regulation of pancreatic PC1 and PC2 associated with increased glucagon-like peptide 1 in diabetic rats. J Clin Invest, 2000. 105(7): p. 955-65. 484. Tanaka, S., et al., Immunocytochemical localization of prohormone convertases PC1/PC3 and PC2 in rat pancreatic islets. Arch Histol Cytol, 1996. 59(3): p. 261-71. 485. Bennett, D.L., et al., Identification of the type 2 proinsulin processing endopeptidase as PC2, a member of the eukaryote subtilisin family. J Biol Chem, 1992. 267(21): p. 15229-36. 486. Neerman-Arbez, M., V. Cirulli, and P.A. Halban, Levels of the conversion endoproteases PC1 (PC3) and PC2 distinguish between insulin-producing pancreatic islet beta cells and non-beta cells. Biochem J, 1994. 300 ( Pt 1): p. 57-61. 487. Sizonenko, S.V. and P.A. Halban, Differential rates of conversion of rat proinsulins I and II. Evidence for slow cleavage at the B-chain/C-peptide junction of proinsulin II. Biochem J, 1991. 278 ( Pt 3): p. 621-5. 488. Jean, F., et al., Fluorescent peptidyl substrates as an aid in studying the substrate specificity of human prohormone convertase PC1 and human furin and designing a potent irreversible inhibitor. J Biol Chem, 1995. 270(33): p. 19225-31. 489. Ledgerwood, E.C., et al., The specificity of the neuroendocrine convertase PC3 is determined by residues NH2- and COOH-terminal to the cleavage site. Biochem Mol Biol Int, 1996. 39(6): p. 1167-76. 490. Courtade, J.A., et al., Loss of prohormone convertase 2 promotes beta cell dysfunction in a rodent transplant model expressing human pro-islet amyloid polypeptide. Diabetologia, 2017. 60(3): p. 453-463. 491. Xin, Y., et al., Use of the Fluidigm C1 platform for RNA sequencing of single mouse pancreatic islet cells. Proc Natl Acad Sci U S A, 2016. 113(12): p. 3293-8.  216 492. Wideman, R.D., et al., Improving function and survival of pancreatic islets by endogenous production of glucagon-like peptide 1 (GLP-1). Proc Natl Acad Sci U S A, 2006. 103(36): p. 13468-73. 493. Ugleholdt, R., et al., Impaired intestinal proglucagon processing in mice lacking prohormone convertase 1. Endocrinology, 2004. 145(3): p. 1349-55. 494. Ramond, C., et al., Understanding human fetal pancreas development using subpopulation sorting, RNA sequencing and single-cell profiling. Development, 2018. 145(16). 495. Paquet, L., et al., Peptide biosynthetic processing: distinguishing prohormone convertases PC1 and PC2. Mol Cell Endocrinol, 1996. 120(2): p. 161-8. 496. Harata, M., et al., Delivery of shRNA via lentivirus in human pseudoislets provides a model to test dynamic regulation of insulin secretion and gene function in human islets. Physiol Rep, 2018. 6(20): p. e13907. 497. Maier, T., M. Guell, and L. Serrano, Correlation of mRNA and protein in complex biological samples. FEBS Lett, 2009. 583(24): p. 3966-73. 498. Liu, M., et al., INS-gene mutations: from genetics and beta cell biology to clinical disease. Mol Aspects Med, 2015. 42: p. 3-18. 499. Cho, B., et al., Generation of insulin-deficient piglets by disrupting INS gene using CRISPR/Cas9 system. Transgenic Res, 2018. 27(3): p. 289-300. 500. Ma, S., et al., beta Cell Replacement after Gene Editing of a Neonatal Diabetes-Causing Mutation at the Insulin Locus. Stem Cell Reports, 2018. 11(6): p. 1407-1415. 501. Yamaguchi, T., et al., Interspecies organogenesis generates autologous functional islets. Nature, 2017. 542(7640): p. 191-196. 502. Ma, H., et al., Establishment of human pluripotent stem cell-derived pancreatic beta-like cells in the mouse pancreas. Proc Natl Acad Sci U S A, 2018. 115(15): p. 3924-3929. 503. Muller, D., et al., Identification of insulin signaling elements in human beta-cells: autocrine regulation of insulin gene expression. Diabetes, 2006. 55(10): p. 2835-42. 504. Marchetti, P., et al., Insulin secretory function is impaired in isolated human islets carrying the Gly(972)-->Arg IRS-1 polymorphism. Diabetes, 2002. 51(5): p. 1419-24. 505. Robertson, R.P., et al., Beta-cell glucose toxicity, lipotoxicity, and chronic oxidative stress in type 2 diabetes. Diabetes, 2004. 53 Suppl 1: p. S119-24. 506. Kramer, C.K., B. Zinman, and R. Retnakaran, Short-term intensive insulin therapy in type 2 diabetes mellitus: a systematic review and meta-analysis. Lancet Diabetes Endocrinol, 2013. 1(1): p. 28-34. 507. Bonnemaison, M.L., B.A. Eipper, and R.E. Mains, Role of adaptor proteins in secretory granule biogenesis and maturation. Front Endocrinol (Lausanne), 2013. 4: p. 101. 508. Wang, Y.J., et al., Single-Cell Mass Cytometry Analysis of the Human Endocrine Pancreas. Cell Metab, 2016. 24(4): p. 616-626. 509. Bengtsson, N.E., et al., Muscle-specific CRISPR/Cas9 dystrophin gene editing ameliorates pathophysiology in a mouse model for Duchenne muscular dystrophy. Nat Commun, 2017. 8: p. 14454. 510. Haliyur, R., et al., Human islets expressing HNF1A variant have defective beta cell transcriptional regulatory networks. J Clin Invest, 2019. 129(1): p. 246-251. 511. Mazzola, N., Review of current and emerging therapies in type 2 diabetes mellitus. Am J Manag Care, 2012. 18(1 Suppl): p. S17-26. 512. Russell, S.J., et al., Outpatient glycemic control with a bionic pancreas in type 1 diabetes. N Engl J Med, 2014. 371(4): p. 313-25.   217 Appendices Appendix A  Secondary antibodies used for immunofluorescent staining in all chapters. Peptide/protein target Antigen Sequence Name of Antibody Manufacturer, catalog #, and/or name of producer Species raised in; monoclonal or polyclonal Dilution used RRID Rabbit IgG  Whole IgG  Goat anti-rabbit AF488  Life Technologies  A11034  Goat  Polyclonal  1:1000  AB_2576217  Rabbit IgG  Whole IgG  Goat anti-rabbit AF555  Life Technologies  A21429  Goat  Polyclonal  1:1000  AB_151761  Rabbit IgG  Whole IgG  Goat anti-rabbit AF594  Life Technologies  A11037  Goat  Polyclonal  1:1000  AB_2534095  Rabbit IgG  Whole IgG  Goat anti-rabbit AF647  Life Technologies  A21245  Goat  Polyclonal  1:1000  AB_2535813  Rabbit IgG  Whole IgG  Donkey anti-rabbit AF488  Life Technologies  A21206  Donkey  Polyclonal  1:1000  AB_141708  Rabbit IgG  Whole IgG  Donkey anti-rabbit AF555  Life Technologies  A31572  Donkey  Polyclonal  1:1000  AB_162543  Rabbit IgG  Whole IgG  Donkey anti-rabbit AF594  Life Technologies  A21207  Donkey  Polyclonal  1:1000  AB_141637  Rabbit IgG  Whole IgG  Donkey anti-rabbit AF647  Life Technologies  A31573  Donkey  Polyclonal  1:1000  AB_2536183  Mouse IgG  Whole IgG  Goat anti-mouse AF488  Life Technologies  A11029  Goat  Polyclonal  1:1000  AB_2534088  Mouse IgG  Whole IgG  Goat anti-mouse AF555  Life Technologies  A21424  Goat  Polyclonal  1:1000  AB_141780  Mouse IgG  Whole IgG  Goat anti-mouse AF594  Life Technologies  A11032  Goat  Polyclonal  1:1000  AB_141672  Mouse IgG  Whole IgG  Goat anti-mouse AF647  Life Technologies  A21236  Goat  Polyclonal  1:1000  AB_2535805  Mouse IgG  Whole IgG  Donkey anti-mouse AF488  Life Technologies  A21202  Donkey  Polyclonal  1:1000  AB_2535788  Mouse IgG  Whole IgG  Donkey anti-mouse AF555  Life Technologies  A31570  Donkey  Polyclonal  1:1000  AB_2536180  Mouse IgG  Whole IgG  Donkey anti-mouse AF594  Life Technologies  A21203  Donkey  Polyclonal  1:1000  AB_141633  Mouse IgG  Whole IgG  Donkey anti-mouse AF647  Life Technologies  A31571  Donkey  Polyclonal  1:1000  AB_162542  Goat IgG  Whole IgG  Donkey anti-goat AF594  Life Technologies  A11058  Donkey  Polyclonal  1:1000  AB_142540  Sheep IgG  Whole IgG  Donkey anti-sheep AF488  Life Technologies  A11015  Donkey  Polyclonal  1:1000  AB_141362  Guinea pig IgG  Whole IgG  Goat anti–guinea pig AF488  Life Technologies  A11073  Goat  Polyclonal  1:1000  AB_2534117  Guinea pig IgG  Whole IgG  Goat anti–guinea pig AF647  Life Technologies  A21450  Goat  Polyclonal  1:1000  AB_151882     218 Appendix B  Maps of plasmid sequences between flanking ITRs that were used for the generation of AAVs.     mINS2p-eGFPIns1-GFPIns1-CreIns1-INSIns1-Ins1 219 Appendix C  Study designs of in vivo experiments in Chapter 4. Animals used to generate data for Figure 4-1     Animals used to generate data for Figures 4-2 and 4-3    220 Animals used to generate data for Figures 4-4 to 4-6     Animals used to generate data for Figure 4-7     221 Animals used to generate data for Figures 4-8 and 4-9     Animals used to generate data for Figure 4-10     Animals used to generate data for Figure 4-11  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0379180/manifest

Comment

Related Items