Stem Cells from Patients with Congenital Hyperinsulinism

Home , Hyperinsulinism, Insulitis

Stem cells from patients with Congenital

Hyperinsulinism

A thesis submitted to the University of Manchester

for the degree of doctor of philosophy in the Faculty of Biology, Medicine and Health

School of Medical Sciences

Sophie Grace Kellaway

2016 Contents

0.1 Abbreviations ...... 9

0.2 Abstract ...... 11

0.3 Declaration ...... 12

0.4 Copyright statement ...... 12

0.5 Acknowledgements ...... 13

0.6 Thesis organisation ...... 14

0.7 Author note ...... 15

1 Literature Review 16

1.1 Introduction ...... 16

1.2 The Pancreas ...... 17

1.2.1 Anatomy ...... 17

1.2.2 Development ...... 20

1.2.3 Post developmental endocrine proliferation ...... 23

1.2.4 Cell cycle in the pancreas ...... 23

1.2.5 Signalling and markers ...... 25

1.3 Congenital hyperinsulinism (CHI) ...... 27

1.3.1 Symptoms and causes ...... 27

1.3.2 Hyperproliferation ...... 30

1.3.3 Treatment ...... 30

1.3.4 Experimental models ...... 30

1.4 Diabetes ...... 31

1.4.1 Epidemiology ...... 31

1.4.2 Pathogenesis ...... 32

1.4.3 Treatment ...... 33

1.5 Stem Cells ...... 34

1.5.1 Pancreatic stem cells ...... 34

1.5.2 Mesenchymal stem cells ...... 35

2 1.5.3 Pluripotent stem cells ...... 38

1.5.4 Generation of iPSCs ...... 41

1.6 Aims ...... 43

1.7 References ...... 43

2 A Novel Source of Mesenchymal Stem Cells derived from the Pancreas of Patients

with Congenital Hyperinsulinism. 64

2.1 Abstract ...... 65

2.2 Introduction ...... 66

2.3 Methods ...... 68

2.3.1 Derivation of CHI pancreatic mesenchymal stem cell (CHI pMSC) lines . . . . 68

2.3.2 Cell culture ...... 68

2.3.3 Karyotyping ...... 69

2.3.4 Short tandem repeat (STR) proﬁling ...... 69

2.3.5 RT-PCR and qRT-PCR ...... 69

2.3.6 Genotyping ...... 70

2.3.7 Adipogenesis, osteogenesis and chondrogenesis ...... 72

2.3.8 Flow Cytometry ...... 73

2.3.9 Immunocytochemistry (ICC) ...... 73

2.3.10 Bisulﬁte sequencing ...... 75

2.4 Results ...... 76

2.4.1 MSCs were derived from three CHI patients ...... 76

2.4.2 Derived cells were karyotypically stable, unique and contained the CHI causing

mutations ...... 83

2.4.3 CHI derived MSCs express pancreatic markers ...... 84

2.4.4 CHI pMSCs have altered gene expression and methylation patterns compared

to bmMSCs ...... 88

2.5 Discussion ...... 94

2.6 References ...... 97

3 Increased Proliferation and Altered Cell Cycle Regulation in Pancreatic Stem

Cells derived from Patients with Congenital Hyperinsulinism 104

3.1 Abstract ...... 105

3.2 Introduction ...... 106

3.3 Methods ...... 110

3.3.1 Cell culture ...... 110

3 3.3.2 Derivation of adult pancreatic mesenchymal stem cell (A pMSC) lines . . . . . 111

3.3.3 Flow Cytometry (propidium iodide) ...... 111

3.3.4 Immunocytochemistry (ICC) ...... 111

3.3.5 Western Blotting ...... 112

3.3.6 Ki67 stain and count ...... 112

3.4 Results ...... 114

3.4.1 Increased proliferation in CHI tissue and CHI-derived pMSCs ...... 114

3.4.2 Cell cycle regulation in CHI MSCs ...... 116

Kip1 3.4.3 Distribution of G1/S molecules - decreased nuclear p27 in CHI-derived pMSCs ...... 118

3.4.4 Distribution of G2/M molecules - possible increase of CDK1 in CHI-derived pMSCs ...... 120

3.4.5 Glucose and insulin aﬀect p27Kip1 protein levels and localisation ...... 122

3.5 Discussion ...... 125

3.6 References ...... 128

4 Stem Cell Derived β-cells: a new resource for Congenital Hyperinsulinism? 134

4.1 Abstract ...... 135

4.2 Introduction ...... 136

4.3 Methods ...... 140

4.3.1 MSC culture ...... 140

4.3.2 Feeder dependent iPSC culture ...... 140

4.3.3 Feeder free iPSC culture ...... 141

4.3.4 Derivation of iPSCs ...... 141

4.3.5 Three germ layer iPSC diﬀerentiation ...... 141

4.3.6 MSC diﬀerentiation ...... 142

4.3.7 Diﬀerentiation: Melton/Kieﬀer ...... 142

4.3.8 RT-PCR and qRT-PCR ...... 144

4.3.9 Bisulﬁte sequencing ...... 146

4.3.10 Immunocytochemistry (ICC) ...... 147

4.3.11 Western blot ...... 148

4.3.12 Glucose-stimulated insulin secretion (GSIS) assay ...... 148

4.4 Results ...... 151

4.4.1 Serum withdrawal was insuﬃcient to induce diﬀerentiation of CHI MSCs . . . 151

4.4.2 A step-wise diﬀerentiation protocol shows stage speciﬁc gene induction, en-

hanced by the addition of DKK1 ...... 152

4 4.4.3 Phenotypic analysis of diﬀerentiated CHI pMSCs ...... 156

4.4.4 Spheroid formation was essential for diﬀerentiation ...... 160

4.4.5 Derivation of iPSCs ...... 161

4.4.6 Diﬀerentiation of CHI-iPSCs to deﬁnitive endoderm ...... 167

4.4.7 Diﬀerentiation of CHI-iPSCs to pancreatic endoderm ...... 168

4.4.8 Diﬀerentiation of CHI-iPSCs to hormone positive islet cells ...... 169

4.5 Discussion ...... 174

4.6 References ...... 180

5 Conclusions 189

5.1 CHI pMSCs were derived and have a pancreatic phenotype ...... 189

5.2 CHI pMSCs were diﬀerentiated to immature β-like cells ...... 190

5.3 A CHI iPSC line could be diﬀerentiated to pancreatic cells ...... 192

5.4 Increased proliferation in CHI ...... 193

5.5 Conclusion ...... 194

5.6 Future work ...... 194

5.7 References ...... 197

5 List of Figures

1.1 Islet structure ...... 17

1.2 Regulation of insulin secretion ...... 18

1.3 Development of the pancreas ...... 20

1.4 Cell cycle in the pancreas ...... 24

1.5 Histological diﬀerences between focal and diﬀuse CHI ...... 29

1.6 Insulin secretion in CHI ...... 29

1.7 Sources of MSCs ...... 36

1.8 EMT in pMSCs ...... 37

2.1 Derived cell lines showed characteristic features of MSCs ...... 77

2.2 Derived cell lines showed multipotency, characteristic of MSCs ...... 79

2.3 Mesenchymal proteins were expressed in CHI pMSCs ...... 81

2.4 CHI pMSCs are genetically stable and unique ...... 82

2.5 Pancreatic genes are expressed in CHI pMSCs ...... 84

2.6 Growing cells on vitronectin and ﬁbronectin increased cell number and insulin gene

expression ...... 86

2.7 Pancreatic proteins are expressed in CHI pMSCs ...... 87

2.8 Comparison of gene expression between CHI pMSCs and bmMSCs ...... 88

2.9 Comparison of gene methylation between CHI pMSCs and bmMSCs ...... 91

2.10 Bisulﬁte sequencing and gene sequencing of SPP1 ...... 93

3.1 Increased proliferation in CHI tissue compared to age-matched controls ...... 114

3.2 Increased proliferation in CHI-derived MSCs compared to adult controls ...... 115

3.3 Alterations to cell cycle control in CHI-derived MSCs compared to adult controls . . . 117

3.4 Altered levels of nuclear p27Kip1 in CHI MSCs ...... 119

3.5 Levels of G2/M regulators in CHI and adult MSCs ...... 121

3.6 Glucose and insulin concentrations determine p27Kip1 localisation ...... 124

4.1 Melton/Kieﬀer diﬀerentiation protocols ...... 143

6 4.2 Serum withdrawal enhanced spheroid formation but was not suﬃcient alone ...... 151

4.3 Step-wise diﬀerentiation induces and reduces gene expression ...... 153

4.4 Spheroids treated with diﬀerentiation inducers were of a similar size to human pan-

creatic islets ...... 155

4.5 Some untreated spheroids after 21 days have a capsule ...... 156

4.6 Diﬀerentiated CHI pMSCs express INS and PDX1 ...... 158

4.7 Diﬀerentiated spheroids contained and secreted more insulin than control spheroids . . 159

4.8 Cells diﬀerentiated as monolayer do not have the same expression proﬁle as spheroids 160

4.9 MSCs were transduced and cleared the SeV ...... 162

4.10 CHI-iPSCs were phenotypically and functionally pluripotent ...... 164

4.11 Transition of iPSCs to feeder-free culture ...... 165

4.12 CHI-iPSCs did not maintain hypomethylation seen in CHI-pMSCs ...... 170

4.13 CHI iPSCs efficiently differentiate to definitive endoderm ...... 171

4.14 CHI iPSCs diﬀerentiate to deﬁnitive endoderm by D’amour protocol only when cul-

tured on feeders ...... 172

4.15 CHI iPSCs eﬃciently diﬀerentiate to pancreatic endoderm ...... 172

4.16 CHI iPSCs diﬀerentiated to cells with characteristics of immature α and β-cells . . . . 173

7 List of Tables

1.1 Cross-section of MSC to β-cell diﬀerentiation protocols ...... 39

1.2 Cross-section of MSC to β-cell diﬀerentiation protocols ...... 40

2.1 Primers used for RT-PCR, details on design and use are in Section 2.3.5. Primers

indicated with * from Ellard et al. [2007] ...... 72

2.2 Antibodies used in the study ...... 74

2.3 Primers used for bisulﬁte sequencing ...... 75

2.4 Details of cell lines and patients ...... 76

2.5 CD marker expression ...... 78

3.1 Details of cell lines and patients ...... 110

3.2 Antibodies used in the study ...... 113

4.1 Details of cell lines and patients ...... 140

4.2 Primers used for RT-PCR ...... 146

4.3 Primers used for bisulﬁte sequencing ...... 147

4.4 Antibodies used in the study ...... 150

8 0.1 Abbreviations

AU - arbitrary units bFGF - basic ﬁbroblast growth factor bmMSCs - bone marrow mesenchymal stem cells

BSA - bovine serum albumin

Btc - betacellulin

CDK - cyclin dependent kinase

CHI - congenital hyperinsulinism

CKI - cyclin dependent kinase inhibitor

CT - cycle threshold DAPI - 4’,6-diamidino-2-phenylindole

ECL - enhanced chemiluminescent

EGF - epidermal growth factor

ELISA - enzyme-linked immunosorbent assay

EMT - epithelial-mesenchymal transition

Ex-4 - exendin-4

FBS - foetal bovine serum

G6P - glucose-6-phosphate

GLP - glucagon-like peptide

GSIS - glucose stimulated insulin secretion hESCs - human embryonic stem cells

HGF - hepatocyte growth factor

HRP - horse radish peroxidase

IBMX - 3-isobutyl-1-methylxanthine

ICC - immunocytochemistry

IGFR - insulin-like growth factor receptor

INS - insulin iPSCs - induced pluripotent stem cells

IR - insulin receptor

ITS - insulin/transferrin/selenium

KGF - keratinocyte growth factor

KRH - Krebs-Ringer-HEPES

LoH - loss of heterozygosity

MET - mesenchymal-epithelial transition

Mol - moles

9 MPCs - multipotent progenitor cells

MSCs - mesenchymal stem cells

NaB - sodium butyrate

NEAA - non-essential amino acids

NGS - normal goat serum

Nic - nicotinamide

P - passage

PBS - phosphate buﬀered saline

PCR - polymerase chain reaction

PdBU - phorbol 12,13-dibutyrate

PDL - pancreatic duct ligation PI - propidium iodide pMSCs - pancreatic mesenchymal stem cells

PVDF - polyvinylidene ﬂuoride

SDS - sodium dodecyl sulphate

STR - short tandem repeat

TPB - ((2S,5S)-(E,E)-8-(5-(4-(triﬂuoromethyl)phenyl)-2,4-pentadienoylamino)benzolactam

TSA - trichostatin-A

10 0.2 Abstract

The University of Manchester Sophie Grace Kellaway Doctor of Philosophy 2016

Stem cells from patients with Congenital Hyperinsulinism

Diabetes and congenital hyperinsulinism (CHI) are severe diseases aﬀecting the pancreas. Cur- rent models for testing drugs to treat these diseases are in vivo in rodents or isolated rodent islets.

Diﬀerences between the human and rodent pancreas, and ethical issues, mean that in vitro human models are needed. To develop a novel in vitro model for pancreatic diseases, mesenchymal stem cells

(MSCs) and induced pluripotent stem cells (iPSCs) were derived from the pancreas of patients with

CHI. MSCs from three forms of CHI were phenotypically normal for MSCs, and maintained the CHI- causing mutation. When compared to MSCs from bone marrow, the CHI pancreatic MSCs expressed pancreas-speciﬁc gene ISL1 and showed promoter hypomethylation of other pancreatic genes, including PDX1. The CHI pMSCs could be diﬀerentiated to cells resembling immature β-cells, with some

β-cell gene expression (INS, PDX1 ), but no glucose responsive insulin secretion. CHI associated hypersecretion of insulin was not seen as the ATP-sensitive potassium KATP channels were not being expressed. Addition of the Wnt inhibitor DKK1 markedly enhanced diﬀerentiation via induction of neuronal genes. Alongside high insulin secretion, CHI also features increased proliferation. CHI

MSCs were also hyperproliferative, and showed alterations to the cell cycle. These changes were related to p27Kip1 localisation, a known affected protein in CHI tissue, and CDK1, a novel regulator for CHI. iPSCs were also derived from focal CHI MSCs and were also phenotypically normal, but did not maintain the pancreatic hypomethylation present in MSCs. The CHI iPSCs were efficiently differentiated to definitive endoderm and PDX1 positive cells. Terminally differentiated iPSCs were endocrine, but were not mature β-cells. In conclusion, authentic MSCs and iPSCs were derived for the first time from patients with CHI. These stem cells could be differentiated towards β-cells, but mature glucose responsive β-cells were not produced. MSC derived β-like cells secreted insulin but did not have KATP channels, whereas iPSC derived β-like cells had KATP channel gene expression but not INS. With further optimisation to resolve these, CHI stem cell derived β-cells may be used for in vitro modelling. Further, the undifferentiated MSCs only show hyperproliferation associated with p27Kip1 and CDK1 and so can be a useful resource for modelling hyperproliferation seen in CHI.

11 0.3 Declaration

I declare that no portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualiﬁcation of this or any other university or other institute of learning.

0.4 Copyright statement i. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the Copyright) and s/he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes.

ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the

University has from time to time. This page must form part of any such copies made.

iii. The ownership of certain Copyright, patents, designs, trade marks and other intellectual property (the Intellectual Property) and any reproductions of copyright works in the thesis, for example graphs and tables (Reproductions), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions.

iv. Further information on the conditions under which disclosure, publication and commer- cialisation of this thesis, the Copyright and any Intellectual Property University IP Policy (see http://documents.manchester.ac.uk/display.aspx?DocID=24420), in any relevant Thesis restriction declarations deposited in the University Library, The University Librarys regulations (see http://www.library.manchester.ac.uk/about/regulations/) and in The Universitys policy on Presen- tation of Theses

12 0.5 Acknowledgements

First and foremost, I would like to thank Mark and Karen for having me in their lab for the last

3 years, and for setting up this project. Thanks also, to my colleagues for their helpful discussions and assistance: Karolina, Bing, Amna, Helga, Saba, Aleks and Alex. Particularly to Karolina, who has worked closely with me on getting the cells to where they needed to be, and for the many cell culture swaps when I needed a day oﬀ, and to Alex for some superb comments on the manuscripts!

For their technical assistance essential to this project, my thanks to: Roger Meadows (slides- canner), Steve Marsden (confocal microscopy), Fraser Coombe and Paul Fuller (DNA sequencing),

Graeme Fox (STR proﬁling), Mike Jackson (ﬂow cytometry). Similarly, thank you to the surgical team for helping us with obtaining the tissue in a viable state, and to the parents of the patients who gave their consent for the tissue to be used for research.

Thank you for the NC3Rs for funding this project.

Finally, thank you to my mum and Ste, who have kept me going throughout, especially when I found out that the ﬁrst cells were not what I thought they were!

13 0.6 Thesis organisation

My thesis is based on derivation, characterisation and diﬀerentiation of both mesenchymal stem cells

(MSCs) and induced pluripotent stem cells (iPSCs) derived from patients with congenital hyperinsulinism (CHI).

I have written the thesis in the ‘alternative format’, comprising three paper-style chapters, a general literature review and overall discussion, which connects and compares the results chapters.

None of these chapters are published yet, although a manuscript based on the ﬁrst chapter is being submitted for publication. I have chosen this format as:

My research easily separates into these main sections

Some work included only makes sense in the context of my colleagues’ unpublished work

I hope that having the work written in a format suitable for publication will allow dissemination

of the results

This format has allowed practice of writing manuscripts, a skill I will need going forward

In the ﬁrst chapter I describe the derivation and characteristics of three MSC lines from patients with CHI. All work in this chapter, including deriving cell lines from primary tissue was done by me, except for some replicates of ﬂow cytometry for CD markers, which were done by Karolina Mosinska.

The second chapter is based on exploration of the hypothesis that hyperproliferation seen in CHI will also be seen in CHI stem cells. In this chapter, the data for the first figure was generated and analysed by Bing Han, who has characterised proliferation in multiple CHI patients including those which I derived cells from. All other data, experimental design and analysis was my own. The final results chapter is on differentiation of the CHI MSCs to β-cells. The protocol used was based on a published protocol, and was chosen based on screening of multiple published protocols by Karolina and myself. The negative results of this screening are not included but were essential for the work that followed, hence the attribution to Karolina. The assessment of the final protocol and discovery of enhancement of β-cell differentiation by DKK1 was my own work. This chapter also explains the derivation and differentiation of iPSCs, derived from the previously discussed MSCs. This was all my own work, except for one experimental replicate of figure 5.5C, SOX17, contributed by Karolina where I found mine had failed after ceasing lab work. Other authorships are attributed to the medical team who care for the CHI patients, Alexander Ryan for assistance with preparation and proofing the manuscripts, and to Karen Cosgrove and Mark Dunne who conceived the project and provided guidance throughout.

14 The thesis therefore forms a body of work based on my own hypotheses and work, including the following contributions to knowledge:

Proof of principle of deriving MSCs from CHI tissue, and a replicable method to do so

Demonstration that CHI pancreatic MSCs (pMSCs) have gene expression and methylation not

seen in non-pancreatic (bone marrow) MSCs

The observation that CHI pMSCs show innately higher rates of proliferation than healthy adult

pMSCs, correlated with the CHI tissue

The ﬁnding that increased CHI pMSC proliferation may be linked to p27Kip1 localisation

(known from CHI tissue) and CDK1 (not previously studied)

That the use of DKK1 in diﬀerentiation of pMSCs enhances generation of β-like cells, via

induction of neuronal genes which can be switched to pancreatic fate with further factors

Published diﬀerentiation protocols used on CHI pMSCs and iPSCs can be used to produce

endocrine cells, but further work is needed to develp them for in vitro modelling of CHI or

diabetes

iPSCs can be generated from pMSCs, but this does not allow maintenance of hypomethylation

Word count: 55,093

0.7 Author note

This thesis involves characterisation of novel cell lines, and includes a strong focus on confirming the identity and origins of the cells. The level to which this is done, using multiple methods for each cell line may seem above what is normally expected. This is because for the first year of the project, I was working on cell lines identified as CHI pancreatic progenitors, which had been cross-contaminated or misidentified by a former student without our knowledge. A recent article in Nature News (Butler,

2015 1) suggests that despite the known phenomenon of misidentification of large numbers of cell lines, scientists are still lax in confirming the identity of what they are working with, and so many results may be invalidated. As such, I wanted to be sure that the cell lines I derived were the correct cell type, from the patients that they were supposed to be and had a genetic fingerprint for future use.

Some journals require this information now with manuscript submission, although not necessarily in the main text of the paper. With this caveat, I have chosen to include this information in the main text of these papers as I feel it is important for reporting novel cell lines after our experiences.

1D. Butler. Biomedical researchers lax in checking for imposter cell lines. Nature News, 2015

15 Chapter 1

Literature Review

1.1 Introduction

Diabetes is the most common, and fastest growing endocrine disease; it is treatable with insulin injections, medications and lifestyle changes, but not curable. Islet transplantation can oﬀer a cure in some cases, but relies on more donors than are available (Shapiro et al., 2000). Therefore human embryonic stem cells (hESCs), induced pluripotent stem cells (iPSCs) and pancreatic endocrine progenitors have been considered to provide a means of producing a large volume of β-cells for transplantation. Research into other cures for diabetes is ongoing, options being explored include drugs to cause β-cell proliferation or neogenesis, using stem cells (Godfrey et al., 2012). This research is largely dependent on rodent β-cell lines, many of which are poorly responsive to glucose and so not ideal models (Skelin et al., 2010). In order to maintain proper glucose-stimulated insulin secretion

(GSIS), β-cells should be derived from animals immediately prior to experimentation (Rees and Al- colado, 2005). Unfortunately, this leads to large numbers of animals being sacriﬁced, and still yields an imperfect model as there are diﬀerences between human and rodent islets such as structure and metabolism (MacDonald et al., 2011, Pan and Wright, 2011, Cabrera et al., 2006). In line with the

‘3Rs’ - reduction, reﬁnement and replacement of animals in scientiﬁc research (Russell and Burch,

1959), and to reduce the effect of differences between human and rodent cells, a human stem cell model which can be induced to differentiate to β-cells as required could be used. HESCs and iP-

SCs have been researched for this purpose too, but diﬀerentiation protocols are long and ineﬃcient

(D’Amour et al., 2006, Maehr, 2011, Lumelsky et al., 2001), and final differentiated cells can have genetic or proteomic profiles not consistent with true β-cells (Liew et al., 2008, Hansson et al., 2004,

Hori et al., 2002).

We hypothesise that by using human pancreatic stem cells, a number of these problems will be

16 negated. The pancreatic origin means diﬀerentiated cells are likely to be as authentic as possible

(Bar-Nur et al., 2011), there will be none of the differences associated with rodent islets, differentiation will not require numerous intermediate stages and a reduction in the number of animals sacrificed should occur. Differentiation of stem cells typically can rely on matching the developmental processes (Best et al., 2008, Hanley, 2014) as such this review will cover both the development and architecture of the pancreas, as well as current stem cell based research.

1.2 The Pancreas

1.2.1 Anatomy

Gross Anatomy

Figure 1.1: Islet structure Schematic showing how human and rodent islet structure diﬀers; rodents have a core of β-cells and a mantle of other cell types whilst humans have all the cells mixed throughout (Cabrera et al., 2006).

The pancreas is a glandular organ comprised of exocrine and endocrine tissue. The exocrine tissue is made up of acinar cells clustered at the end of ducts, the purpose of which is to secrete and channel digestive enzymes, mucins and bicarbonate to the duodenum. The endocrine tissue is arranged in islets of Langerhans and made up of 5 major cell types, along with blood vessels, neurons and mesenchymal support tissue. Islets in rodents are made up of a core and mantle whilst in humans are more mixed (Cabrera et al., 2006); this is shown in Figure 1.1. The cells are α-cells which secrete glucagon, β-cells which secrete insulin, δ-cells which secrete somatostatin, -cells which secrete ghrelin and PP-cells which secrete pancreatic polypeptide (Elayat et al., 1995, Brissova et al., 2005).

All of these secretions are into the surrounding microvasculature and in response to signals from the blood such as glucose levels. The endocrine and exocrine are compartmentally, morphologically and functionally discrete, and do not interact except in cases of extreme stresses, as reviewed by (Pan and Wright, 2011).

17 β-cells

The role of β-cells is to synthesise, package and release insulin in response to varying blood glucose levels, in a process known as GSIS. Insulin is a protein synthesised in the cells as preproinsulin, this is cleaved to active insulin and C-peptide. Transcription of the INS gene is caused by several transcription factors including NEUROD1, PDX1, PAX6 and ISL1 (Melloul et al., 2014, Petersen et al., 1994, Naya et al., 1995, Sander et al., 1997, Zhang et al., 2009). Glucose sensing occurs due to a signalling cascade triggered when the glucose transporters (GLUT1, 2, 4) move glucose from neighbouring blood vessels into the cells. This stimulates ATP generation which closes ATP- sensitive potassium (KATP) channels made of Kir6.2/Sur1 subunits (Aguilar-Bryan et al., 1998), causing depolarisation of the cell (Petit and Loubatieres-Mariani, 1992, Ashcroft and Rorsman, 1990,

Koster et al., 2005). Intracellular calcium rises, which causes exocytosis of stored insulin granules into the bloodstream (Petit and Loubatieres-Mariani, 1992, Braun et al., 2008, Fridlyand et al., 2013).

This is shown schematically in Figure 1.2. This signal is ampliﬁed by GCK which phosphorylates intracellular glucose to glucose-6-phosphate (G6P); this process is not inhibited by G6P and so metabolism is driven constantly by any glucose available, amplifying insulin release (Matschinsky,

2002). As stored insulin is quickly secreted in response to glucose sensing, glucose also causes further transcription of the insulin genes (Leibiger et al., 2002, Evans-Molina et al., 2007). Glucose modulates

PDX1 which binds and activates the promoters of INS, GLUT2 and glucokinase (GCK ), whilst also aﬀecting chromatin accessiblity (Evans-Molina et al., 2007). This glucose dependent transcription is independent of secretion (Macfarlane et al., 2000).

Figure 1.2: Regulation of insulin secretion Insulin is secreted in response to glucose due to calcium inﬂux caused by depolarisation when potassium is pumped out of ATP-sensitive channels, this occurs because of increased levels of ATP generated by glucose metabolism after transport by GLUT receptors (Koster et al., 2005, Braun et al., 2008).

18 β-cells are essential for the regulation of glucose levels in the body by their secretion of insulin, and dysfunction and/or destruction of β-cells causes diabetes, reviewed in Atkinson et al. (2013),

Kahn et al. (2014). One method that has been considered to cure diabetes therefore, is replacement of islets, β-cells, diﬀerentiation of progenitor cells within the pancreas to β-cells or drugs to cause proliferation of β-cells (Shapiro et al., 2000, Godfrey et al., 2012, Yi et al., 2013). This will be discussed further throughout the literature review.

Structure and Extracellular Matrix

Extracellular matrix (ECM) is a network of proteins and sugars which surrounds all cells in the body.

Cells interact with the ECM primarily via surface receptors called integrins. These interactions are important for pancreatic islets.

Adult islets typically exist in an ECM capsule composed of ﬁbroblasts and secreted collagen, which is in turn surrounded by the periinsular basement membrane, composed of laminin and collagen

(Stendahl et al., 2009). During development, however, the ECM is mostly composed of vitronectin,

ﬁbronectin and collagen IV, particularly around progenitor cells. Vitronectin and ﬁbronectin levels decrease as progenitors develop into islets (Stendahl et al., 2009).

There are many diﬀerent integrins which recognise unique motifs within ECM components. For example, the α5 integrin recognises ﬁbronectin and the αvβ3 heterodimeric integrin acts as a receptor for vitronectin. Most integrins are present in the adult and/or foetal pancreas, allowing binding to varied ECM. The αvβ3 integrin is highly expressed in the developing foetal pancreas but is down regulated as islets begin to form, suggesting an involvement with progenitor cell maintenance (Cirulli et al., 2000).

Culturing cells on extracellular matrix enhances survival. Whilst cells secrete their own ECM in order to prevent anoikis, it is often not the native matrix that is best for them. Isolated islets or puriﬁed β-cells show improved survival and maintenance of islet structure when cultured with ECM, whilst incomplete islets with native ECM show better survival than complete islets with no ECM

(Stendahl et al., 2009). However, β-cells cultured on collagen-IV, fibronectin or vitronectin show diminished insulin secretion (Kaido et al., 2006), perhaps due to the fact these matrix components are mainly seen in early differentiation, so may be involved in anti-differentiation signalling.

19 1.2.2 Development

Organogenesis

Figure 1.3: Development of the pancreas The graphical representation of the development of the pancreas shows each stage as follows. The primitive gut tube forms dorsal and ventral buds, these become the early pancreas. This is followed by cluster budding in the epithelium in the primary transition, then the secondary transition, where microlumina form containing the progenitor cells. The committed endocrine progenitors delaminate from the microlumina and assemble into the early islets. The blue, green and yellow circles represent progenitor cells which are committed to becoming β, α and δ cells, the branched structure represents the microlumen.

The pancreas is derived from the endoderm (Wells and Melton, 1999), and as with all organs goes through a number of stages during formation before all cells are mature and functional. The development has been well characterised in rodents but due to technical and ethical challenges less is known in humans. From what has been observed though, there seem to be strong similarities in developmental processes and signalling, just with diﬀerent time scales. Several signalling pathways

20 are involved in the development process, with most of them being involved in multiple stages in context or level dependent functions (Wilson et al., 2003). The final organ size, potential for cell division and specific differentiation seems to be pre-determined early, potentially through epigenetic mechanisms and may affect how signalling pathways function on individual cells (Pan and Wright,

2011).

Initially the pancreas is derived from the embryonic foregut (primitive gut tube) and separates from cells destined to be intestine or liver (Wells and Melton, 1999). Speciﬁcation of the pancreas is by expression of PDX1, which commits the cells early on to a pancreatic fate (Bouwens and Rooman,

2005, Puri and Hebrok, 2010). Dorsal and ventral buds are formed, which elongate, rotate and fuse to become the definitive pancreas (Piper et al., 2004, Gittes, 2009). The first major stage of pancreatic development is known as the primary transition. In this stage there is a large amount of proliferation in pancreatic progenitors to generate stratified epithelium, and endocrine cells begin to develop by cluster budding (Pan and Wright, 2011). Most of the endocrine cells at this stage are short lived, but Ngn3 lineage tracing shows at least some of the cells become incorporated into the newborn islet mantle (Gu et al., 2002). These Ngn3+ cells may last to adulthood, based on the assumption (and evidence seen in this experiment and others by the group) that Ngn3 cells do not renew themselves like stem cells; this has been disputed by other groups (Pan and Wright, 2011).

In the secondary transition, microlumina are formed and segregate into distinct areas. The trunk of the microlumen contains endocrine biased progenitors and the tip contains multipotent progenitor cells (MPCs) which become exocrine cells (Kopp et al., 2011). At this stage the endocrine biased progenitors in the trunk become committed and delaminate from the epithelium whilst exocrine destined cells remain (Puri and Hebrok, 2010). This delamination process resembles epithelial-to- mesenchymal transition (EMT) and allows the cells to assemble into islets as they divide. The endocrine progenitors divide to become endocrine precursors, then the distinct endocrine cell types.

The progenitors form the separate cell types whilst also forming into the eventual islet structures, which become histologically evident and express hormones approximately 12-13 weeks post concep- tion. Most of the mature endocrine cells are derived from the early endocrine progenitors, but around

20% of β-cells are derived from β-cell replication which becomes a predominate, but not exclusive mechanism for increasing mass after birth (Bonner-Weir et al., 2010).

The stages of development are shown graphically in Figure 1.3.

21 β-cells

β-cells develop from early progenitors, through to endocrine precursors, immature β-cells, then ﬁ- nally mature β-cells. Foetal progenitors are stimulated to divide by FGF, and to diﬀerentiate by

Notch signalling in a process known as neogenesis. Foetal neogensis is the main method used to increase β-cell mass at these early stages and is stimulated by glucose, Glp-1, gastrin, exendin-4 and insulin; glucose also acts to suppress apoptosis during neogensis. In contrast to the adult pancreas, neogenesis contributes around 90% of the β-cell mass, with only 10% being produced by β-cell replication (Bouwens and Rooman, 2005). It is unclear whether the adult pancreas loses the ability to respond to or to produce these signals, or whether the eﬀect is dependent on levels of expression of the signals or other factors.

The best marker of a mature β-cell is functional i.e. GSIS. Immature cells are not able to ade- quately respond to glucose challenges, so maturity is important for generating β-cells in vitro. Insulin is an end stage β-cell marker not generally expressed along the diﬀerentiation pathway; but insulin is not a suitable marker of β-cell maturity for two reasons. INS can be expressed in endocrine precursors and immature β-cells (Blum et al., 2012, Pan and Wright, 2011), and if in vitro cell cultures are examined many cells absorb insulin from medium (Hansson et al., 2004). β-cell populations are heterogeneous, with various levels of maturity seen (Pipeleers, 1992, Szabat et al., 2009, 2012).

When working with β-cells a marker of maturity is essential, as it is not always practical to measure

GSIS. The transcription factors MafA/B are one option. In rodents, immature β-cells are MafB positive, mature β-cells are MafA positive, mature alpha cells remain MafB positive, and during the maturation process the β-cells also go through a MafA/MafB double positive stage related to Pdx1 upregulation (Nishimura et al., 2006, Artner et al., 2007). The case in humans is less clear though, although MAFA is still used as a marker of β-cells in many cases. Urocortin 3 (UCN3) is also a good marker of mature β-cells, positive cells are seen in adult islets and authentically diﬀerentiated in vitro β-cells, whilst those in vitro cells which demonstrate poor GSIS are UCN3 negative (Blum et al., 2014). It is important to note though, that UCN3 is a marker only and not an instigator of maturation - it may be that it is upregulation by transcription factors involved in maturation, or acts in combination with other maturation factors (Blum et al., 2012). UCN3 may also be expressed in

α-cells, particularly when derived from ESCs although this may be suggestive of poor diﬀerentiation

(van der Meulen et al., 2012).

22 1.2.3 Post developmental endocrine proliferation

In order to increase β-cell mass in diabetic patients, a source for the new β-cells must be found.

The neonatal developmental pathways may not be applicable to adults, if the stem/progenitor cells present during embryogenesis have been terminally diﬀerentiated. To this end, experiments have been done to ascertain if the adult pancreas is capable of regeneration.

After a major insult it is possible for the rodent pancreas to regenerate by neogenesis from duct-like progenitors (Li et al., 2010). Depending on the type of injury, the pancreas regenerates diﬀerently - pancreatic duct ligation only allows neogenesis of exocrine cells, not β-cells (Rankin et al., 2013, Cavelti-Weder et al., 2013). This regeneration requires signalling from EGF, Glp-1 and

INGAP, reviewed by Bonner-Weir and Weir (2005), but without insult these factors are not enough to trigger neogenesis, although INGAP can initiate duct proliferation (Pittenger et al., 2009). Ad- ditionally after injury, Ngn3+ cells can convert to islet cells, seemingly via a duct-like progenitor role due to gastrin signalling (Rooman et al., 2002, Bouwens and Rooman, 2005). It is possible that the progenitors involved in these processes are not distinct progenitors, but just plastic and less diﬀerentiated β-cells. Alongside neogenesis from progenitor cells, experiments have shown β-cells duplicate (Dor et al., 2004, Yi et al., 2013). All of these experiments were performed in rodents, and there is little evidence of whether the same phenomena occur in humans and so ex vivo human models are used to investigate if endocrine cells can be produced from the adult pancreas (see section 1.5).

1.2.4 Cell cycle in the pancreas

Control of the cell cycle is important for regulating the number and size of cells. Quiescent cells are said to be in G0, but all cells which are mitotically active can be allocated to a stage of the cell cycle. These cell cycle stages are Gap 1 or G1-phase where cells are resting or growing, Synthesis or

S-phase where DNA is duplicated, Gap 2 or G2-phase where the cell prepares to divide and Mitosis or M-phase where cell division occurs and two daughter cells are produced, reviewed by Harashima et al. (2013). The progression through the G1/S checkpoint commits a cell to division (Bertoli et al.,

2013), the G2/M checkpoint controls both DNA integrity and cell size at the point of division (Stark and Taylor, 2004). Cell cycle progression, particularly through G1/S, is controlled by both positive and negative regulator proteins, with members of the cyclin and cyclin-dependent kinase (CDK) families of proteins being involved in positive regulation, and members of the Cip/Kip and Ink families being negative cell cycle regulators, via repression of the cyclins and CDKs as cyclin-dependent kinase inhibitors (CKIs).

23 Figure 1.4: Cell cycle in the pancreas Cell cycle control diﬀers between cell types: In A-D green shapes indicate cell cycle activators and red shapes indicate cell cycle inhibitors. (A) In proliferative cells, cell cycle activators reside in the nucleus and promote cell cycle entry and progression. (B) In non-proliferative cells such as adult β-cells, cell cycle activators reside in the cytoplasm and inhibitors can be found in the nucleus. (C) In proliferative β-cells, the cell cycle activators shuttle to the nucleus and (D) In non-proliferative β-cells subjected to the same stimuli as the cells in panel C, both activators and inhibitors shuttle to the nucleus, maintaining the quiescent state. Figure taken from Wang et al. (2015).

24 One problem with studying pancreatic tissue, or attempting to regrow damaged pancreatic tissue is that it is a largely non-proliferative organ. This quiescence is due to the cell cycle regulators, which diﬀer between pancreatic cells and more proliferative cell types. In actively proliferating cells, generally, CDK4/6 binds to cyclinD causing phosphorylation of retinoblastoma (Rb) which leads to transcription and promotion of entry to the G1 stage of the cell cycle. CDK4 and CDK6 both act in a similar capacity, showing redundancy in most cells, similarly cyclinD1, 2 and 3 can partner

CDK4/6. However, in β-cells speciﬁcally, only CDK6 is active, partnered to cyclinD3 (Fiaschi-Taesch et al., 2013). On the whole, β-cells are non-proliferative, with cell cycle entry prevented by p16INK4a, p57KIP2 and p27KIP2. On the occasions when β-cells are induced to duplicate these CKIs move to the cytoplasm allowing the nuclear cell cycle promoters to function. This nuclear/cytoplasmic shuttling is a major method for cell cycle regulation. Indeed, in the majority of β-cells, most of the time, the only cell cycle molecules in the nucleus are Rb and p57KIP2 which is why they are considered a non-proliferative cell type (Wang et al., 2015). The diﬀerent molecules involved in regulation, both in proliferative and non-proliferative pancreatic tissues are shown in Figure 1.4.

1.2.5 Signalling and markers

PDX1

PDX1 is a transcription factor which is one of the key drivers of the pancreatic lineage. Whilst all pancreatic developmental factors are also used for specifying other lineages, PDX1 is the most speciﬁc for the pancreatic lineage. The importance of PDX1 in islet development is evidenced by the fact that

Pdx1 null mice can only produce immature endocrine cells, with a mature pancreas not fully developing (Ahlgren et al., 1996). This is likely because PDX1 is a master transcription factor, with the gene having cis-regulatory binding sites for HNF1α, FOXA2, PAX6, HNF6, MAFA and PTF1α (Gerrish et al., 2001, Lee et al., 2002, Samaras et al., 2002, 2003, Jacquemin et al., 2003, Wiebe et al., 2007).

In turn, PDX1 drives transcription of other developmental factors, and of insulin itself (Iype et al.,

2005). Forced PDX1 expression is sufficient for α- to β-cell transdifferentiation and in anti-apoptosis signalling in β-cells (Johnson et al., 2006). Together these experiments strongly implicate PDX1 in development of the endocrine pancreas, and specifically β-cells and also in the maintenance of β-cells.

SOX9

SOX9 is a transcription factor used in many cells throughout development, in the pancreas it is used in multipotent stem cells during pancreas formation as well as in ductal cells in the mature pancreas

(Furuyama et al., 2011, Seymour, 2013, Kawaguchi, 2013). During development, or pancreas diﬀer- entiation from stem cells, SOX9 is essential for stem/progenitor cell maintenance, where inactivation

25 causes progenitor apoptosis and decreased proliferation (Seymour et al., 2007).

GATA4/6

The GATA factors are also used in numerous developmental processes. In the pancreas they act as a switch for specification of endocrine cells, rather than other lineages. GATA6 marks endocrine cells and GATA4 exocrine (Ketola et al., 2004). Null mice for either Gata4 or Gata6 show early lethality due to hyperglycaemia, but a single allele is sufficient for a normal pancreas (Carrasco et al., 2012) so further specifics of their functions are still relatively unknown.

NEUROD1, NGN3 and the NOTCH pathway

The NOTCH pathway is heavily involved in the development of endocrine progenitors, during the secondary transition, and functions downstream through main NOTCH target HES1, followed by

NGN3 and NEUROD1. The NOTCH ligands are involved with lateral inhibition, preventing neighbouring NGN3 cells both becoming endocrine cells. The NOTCH pathway activates HES1, which represses NGN3 expression. Knockout mice of Hes1, downstream of the NOTCH ligands, show early diﬀerentiation and loss of the progenitor pool (Jensen et al., 2000), and overexpression inhibits diﬀerentiation of the progenitors and maintenance of the progenitor pool (Murtaugh et al., 2003).

Similarly premature Ngn3 expression causes over diﬀerentiation to endocrine cells, particularly α- cells (Schwitzgebel et al., 2000) and Ngn3 knock out mice contain no endocrine cells (Gradwohl et al.,

2000). These data shown Ngn3 to be essential during rodent pancreatic development, and it is similar in humans, but it is key that NGN3 is only expressed for a short window during development (Salis- bury et al., 2014). After this, NGN3 cells produce SNAI2 in order to undergo epithelial-mesenchymal transition, to escape the trunk and make islets (Rukstalis and Habener, 2007). Together these show that expression and correct timing of NOTCH and its downstream targets are essential for correct endocrine development, or diﬀerentiation.

ISL1

Islet1 (ISL1), as the name suggests, is essential for islet cell formation. Isl1 knock out mice do not contain islet cells (Ahlgren et al., 1997). It is present in Ngn3+ precursors, and is thought to trigger their proliferation and diﬀerentiation (Du et al., 2009). Isl1 is also present in the mesenchyme of mice (Ahlgren et al., 1997) and in MSCs derived from the human pancreas (Eberhardt et al., 2006).

One function of ISL1 is to enhance insulin gene expression (Zhang et al., 2009); other functions,

26 including in MSCs are still being investigated.

PAX4 and ARX

The divide between β-cells and α-cells is mainly decided by a switch of PAX4 or ARX expression.

Endocrine precursors are remarkably plastic, if 99% of β-cells are destroyed α-cells are able to convert to β-cells via an insulin/glucagon double positive intermediate (Thorel et al., 2010). In isolated endocrine precursors it is possible to determine the fate to either α- or β-cells using only the transcription factors Arx or Pax4 respectively. Despite this plasticity, there is evidence that differentiation is affected or predetermined by methylation patterns. In a β-cell, if Dnmt1 is ablated preventing methylation patterns being maintained during division, during the next round of mitosis the cell will become an α-cell. This appears to happen via a mechanism in which Arx is derepressed due to the loss of CpG methylation (Collombat et al., 2003). Conversely forced expression of PAX4 markedly enhances differentiation to β-cells rather than α-cells (Liew et al., 2008). Finally, ex vivo expanded

β-cells can be induced to re-diﬀerentiate to β-cells by inhibition of ARX (Friedman-Mazursky et al.,

2016).

NES

The role of Nestin (NES) in the pancreas has been the subject of much debate. NES is a cytoskeletal protein, which has been suggested as putative marker of pancreatic stem cells (Hunziker and Stein,

2000). The evidence on this is mixed though, as NES is seen in multiple cells (Street et al., 2004) and is not required for β-cell generation during normal development (Piper et al., 2002). NES is also seen in MSCs derived from the pancreas (Eberhardt et al., 2006), it is not clear if these MSCs are derived from NES postive cells within the pancreas though.

1.3 Congenital hyperinsulinism (CHI)

1.3.1 Symptoms and causes

CHI is an inherited disorder, characterised by high levels of disregulated insulin secretion (Dunne et al., 2004). It affects neonatal patients, being a genetic disorder. There are three forms of CHI - focal, diffuse and atypical. All three forms show the main phenotype of excessive insulin secretion, but have histological and genetic differences, the histological differences are shown in Figure 1.5.

The most common cause of CHI is defects in the KATP channels, which cause constitutive eﬄux of

27 potassium ions and inﬂux of calcium ions causing a release of insulin granules, shown schematically in Figure 1.6.

The inappropriately high levels of insulin cause numerous problems, resulting in most patients being hospitalised within a few days of being born (Hussain and Aynsley-Green, 2004). The insulin results in severely low blood glucose, glucose is essential for metabolism in every cell. This results in lethargy, seizures, cognitive problems, feeding problems or coma (De Leon and Stanley, 2007,

Hussain, 2008). If blood glucose is not swiftly normalised it will lead to degradation of body tissues, including the brain and eventually death (Hussain, 2008). This means that lifelong management of the condition is essential.

Focal CHI is named for a focal lesion, comprising almost exclusively β-cells. It is due to a recessive heterozygous mutation, combined with loss of heterozygosity in the lesion alone. The mutation is paternally inherited, and the maternal allele lost in a sporadic fashion, resulting in expression changes in several maternally imprinted genes (Giurgea et al., 2006). ABCC8, encoding the SUR1 subunit of the KATP channel, is one of the genes deleted which causes the loss of regulation of insulin secretion. Other genes are aﬀected, including p57Kip2 (James et al., 2009), which may lead to the observed increase in islet-cell proliferation in the lesion (Kassem et al., 2000).

Diﬀuse CHI, unlike focal CHI, aﬀects every β-cell in the pancreas. It is caused by a variety of gene mutations, most commonly ABCC8, KCNJ11, GCK and HNF4α (Dunne et al., 1997, Glaser et al.,

2000). An homozygous mutation, or two diﬀerent inactivation mutations on each allele inactivate the insulin secretion pathway in every β-cell. With diﬀuse CHI being predominately caused by homozygous mutations, it is relatively rare in some populations such as the Netherlands (1 in 50,000), but common in regions with high rates of consanguinity such as Saudi Arabia (1 in 2500) (Bruining, 1990).

Atypical CHI is markedly different from focal and diffuse. Where focal and diffuse CHI present shortly after birth, atypical has a later onset (Arnoux et al., 2011). The genetic background of focal and diffuse CHI is well characterised, whereas for atypical CHI it is not (Arnoux et al., 2011). It may be caused by a wide range of mutations in different patients which would make linkage studies impossible, as pancreatic development is controlled by a myriad of genes which could be causing the disease (Pan and Wright, 2011). One factor that is known to be associated with atypical CHI is altered hexokinase expression, but the role of this has not been fully worked out yet (Henquin et al.,

2013).

28 Figure 1.5: Histological differences between focal and diffuse CHI In diffuse CHI every β-cell is affected, in focal CHI only one area of β-cells is affected. Adapted from (Dunne et al., 2004).

Figure 1.6: Insulin secretion in CHI Despite low glucose levels and therefore low intracellular ATP levels, KATP channels cannot open, β-cells remain depolarised with constant calcium inﬂux and therefore insulin is constitutively secreted (Koster et al., 2005, Braun et al., 2008, Dunne et al., 2004).

29 1.3.2 Hyperproliferation

Aside from altered insulin secretion, CHI also features increased proliferation in the pancreas, compared to healthy age-matched pancreas. In the lesion of focal CHI this is caused by loss of p57Kip2

(Fournet et al., 2001, Kassem et al., 2001, James et al., 2009). In diﬀuse CHI, increased nuclear

CDK6 and p27Kip1 is also associated with increased proliferation although the ultimate cause is unknown (Salisbury et al., 2015). Unpublished results from our lab also show increased proliferation in atypical CHI and outside of the lesion in focal CHI. As this is universal to all forms, it may be that the increased proliferation is due to the high levels of insulin. In this context insulin could be acting as a growth factor (Denley et al., 2004, Belﬁore and Malaguarnera, 2011, Ish-Shalom et al., 1997), causing a cycle of proliferation leading to more insulin-secreting cells, leading to more proliferation.

1.3.3 Treatment

Due to the variability in causes and presentation, treatment of CHI can be very diﬀerent. First line treatments for CHI are pharmaceutical, with immediate glucose and glucagon infusions (Moens et al.,

2002) to stabilise blood glucose and prevent permanent damage. More permanent treatments for ongoing management of the condition include octreotide (Glaser et al., 1989) and diazoxide (Touati et al., 1998), which in some cases prevent insulin secretion. Where pharmaceutical treatments fail, for example, when the precise mutation renders the drugs non-functional, the only remaining option is surgery (Arnoux et al., 2010). For atypical and diﬀuse CHI this involves a subtotal pancreatectomy

(Dunne et al., 2004, Hussain, 2008), but this can lead to diabetes later in life (Leibowitz et al., 1995) and so is a last resort. In focal CHI, (18)F-L-dopa positron emission tomography allows identiﬁcation of the lesion, allowing precise surgical excision (Otonkoski et al., 2006). This makes surgery a good option for focal CHI, with it being a ﬁrst line treatment at some centres (Hussain, 2008), although surgery on neonates is still not an ideal situation. The excessive proliferation in CHI can lead to regrowth of the pancreas following surgery in some cases. When this occurs repeated surgery is required to further resect the pancreas (Jack et al., 2003, Meissner et al., 2003). Due to these problems with medical management, novel treatments are desired (Meissner et al., 2003).

1.3.4 Experimental models

CHI research primarily relies on looking at post-operative pancreatic tissue. This has the advantage of allowing study of authentic diseased tissue. On the other hand, because CHI is a rare disease and surgery is last resort, fresh tissue is limited. This restricts the number of experiments which can be performed. Studies on ﬁxed tissue are more readily achievable but only provide a snapshot, not a dynamic system which can be probed. To combat these diﬃculties, other models have been

30 generated. One option is transfection of constitutively active genes containing the CHI causing mutations, into HEK293 cells which do not normally express KATP channels (Arya et al., 2014). This allows study of the effect of the mutations on ion flux, using radioactive rubidium, but no way to understand the effects on insulin secretion. One other option for studying CHI is using rodent models. Generation of an ABCC8 mutant knock in mouse has allowed study of the effect on channels and insulin secretion (Shimomura et al., 2013). However, the differences between human and rodent islet structure (Cabrera et al., 2006, Balboa and Otonkoski, 2015), discussed in section 1.2.1, may mean that rodent experiments cannot be extrapolated to humans. Further, there are differences between cell cycle in humans and rodents (Kulkarni et al., 2012, Fiaschi-Taesch et al., 2009), so it is likely that hyperproliferation in CHI cannot be accurately studied in these rodent models. As such, novel, human models of CHI to study channel ion flux and insulin secretion are needed.

1.4 Diabetes

Diabetes is a chronic disease of the pancreas which leads to life-long therapeutic treatment and lifestyle management. There are 2 major forms of the disease, known as Type 1 and Type 2 diabetes; Type 1 diabetes is further subdivided into Type 1A and Type 1B. Type 1A diabetes is the most common form and is an autoimmune disease which results in an absolute insulin deﬁciency due to β-cell destruction (Atkinson et al., 2013). This β-cell destruction means that as yet Type

1A diabetes is not curable and so is the focus of islet replacement research. Type 1B diabetes is non-immune mediated but also results in a severe insulin deﬁciency, the causes are largely unknown

(Atkinson and Eisenbarth, 2001). Type 2 diabetes is usually a result of lifestyle and diet and is characterised by insulin resistance (insulin produced in the body does not function as well as it should) and a gradual progression to insulin deﬁciency (Kahn et al., 2014). Whilst Type 1 diabetes requires daily insulin supplementation for survival, in early and well managed cases Type 2 diabetes can be treated by a change in diet and a reduction in a weight - over time this can progress to needing insulin supplementation though and the disease symptoms converge.

1.4.1 Epidemiology

Diabetes is one of the most common chronic diseases; in the UK the rate is estimated to be around 1 in 16 people having the disease (Diabetes UK, 2015). There are approximately 2.9 million diagnosed cases of diabetes in England alone (Diabetes UK, 2015), with the rate increasing to 1 in 5 in adults over 85 years old, due to the disease being permanent.

31 Type 2 diabetes is primarily a disease of adults due to the nature of its development from lifestyle and diet - obesity is the major risk factor (Hu et al., 2001, Mokdad et al., 2003, CDC, 2004). Type

2 accounts for around 85% of diabetes cases in the UK and 90% of adult cases of diabetes. Type

2 diabetes is most common worldwide in more developed countries where obesity is prevalent but within this there are separate ethnic predispositions - non-whites have a lower threshold for obesity related risk of diabetes (Ntuk et al., 2014).

Type 1 diabetes, being an autoimmune disease, is usually diagnosed in children. In the UK 18-20 children per 100,000 are diagnosed with type 1 diabetes each year (Daneman, 2006). 50-60% of Type

1 diabetes patients are under 18 years old at diagnosis (Daneman, 2006). Geographical variance in this disease is also high for incidence rates. The highest incidences worldwide are in Finland (where

48-49 members of the population are diagnosed per 100,00 each year), Sardinia, Kuwait and Puerto

Rico (Atkinson and Eisenbarth, 2001). The disease also shows strong familial links alongside age and ethnicity - the risk of developing the disease is 15 times higher for siblings of patients than in the general population (Redondo et al., 2001). Despite the disease usually being diagnosed in children, incidence of patients is highest in adults due to the disease being highly treatable, but not curable.

1.4.2 Pathogenesis

Type 1 diabetes as stated above is a disease caused by β-cell destruction which causes an absolute insulin deﬁciency. The β-cell destruction is as a result of auto immune attacks, the body incorrectly produces antibodies against its own islets. Initially this usually includes anti-insulin or glutamic acid decarbyoxylase antibody (GADA). GADA is seen in 5-30% of type 2 diabetes cases which are therefore thought to have been incorrectly diagnosed (Atkinson and Eisenbarth, 2001). Other antibodies seen include protein tyrosine phosphatase IA2 auto-antibody (IA-2AA) and ICA512A (Atkinson and

Eisenbarth, 2001), where the presence of 2 or more of these antibodies give a 90% predictive risk of developing the disease (Devendra et al., 2004). The antibodies alone cannot cause the disease pathology, but are seen in a latent period before the disease becomes symptomatic known as the prodrome. This period can last years before the disease graduates to the next stage, wherein more antibodies appear and β-cell destruction accelerates and becomes measurable. At this stage the ﬁrst phase insulin response in which insulin is rapidly released in response to increased blood glucose levels is decreased and eventually lost as a result of β-cell mass falling below a critical threshold.

β-cell destruction continues and insulin secretion continues to fall until there is a full deﬁciency and the patient will have overt clinical diabetes (Daneman, 2006). Insulitis triggers the β-cell destruction which causes clinical presentation due to cytotoxic T cells expressing the Fas ligand. Upon binding

32 Fas forms the death-inducing signalling complex and causes apoptosis. This Fas-mediated apoptosis is thought to account for the vast majority of β-cell destruction in type 1 diabetes (Devendra et al.,

2004).

Type 2 diabetes is largely due to lifestyle factors such as obesity, wherein adiposity of the body cavities leads to insulin resistance (Cnop et al., 2002). Insulin resistance - where the cells of the body are less sensitive to insulin, combines with β-cell dysfunction to cause the overt clinical symptoms.

Initially β-cells may still secrete insulin normally but insuﬃciency develops (Stancakova et al., 2009) in a feedback loop with resistance (Kahn et al., 2014), this can eventually lead to the β-cells ceasing to function (Kahn et al., 2014) and being destroyed (Butler et al., 2003).

1.4.3 Treatment

For type 2 diabetes, the ﬁrst line treatment is lifestyle changes, to reduce the causes - normally reducing weight, increasing exercise and decreasing sugar intake (Kahn et al., 2014). This is usually combined with the oral drug Metformin (National Institute for Health and Care Excellence, 2015,

Maruthur et al., 2016). Metformin acts to reduce glucose released by the liver and increasing the sensitivity of cells to what insulin is being produced (Viollet et al., 2012). Another common drug treatment is the sulphonylureas which increase insulin production by remaining β-cells by acting on

KATP channels (Proks et al., 2002). All treatments for type 2 diabetes involve daily management and in many cases do not prevent progression of the disease. The long term complications due to the drugs themselves or the diabetes (which is still biologically present, but masked by the drugs) are not clear (Maruthur et al., 2016).

The treatment for type 1 diabetes is injection of insulin, which must be performed daily for life and requires careful monitoring. A combination of short and long acting insulin/insulin analogues are used to replace the insulin that the destroyed β-cells cannot produce. Typically a long acting insulin is used across the day, with rapid acting insulin taken with each meal (Atkinson and Eisen- barth, 2001). This is adjusted to allow for meal size and activity levels, with genetically engineered analogues with varying durations used to attempt to replicate the steady levels of insulin normally found in the body. Blood glucose levels are measured regularly to ensure correct dosage is achieved, but euglycaemia can be diﬃcult to achieve and glucose/insulin spikes are common place (Daneman,

2006).

Due to the problems associated with insulin treatment, permanent cures and prevention of disease development have been investigated. Over 100 treatments prevented type 1 diabetes developing in

33 the non-obese diabetic (NOD) mouse model or in rats treated with streptozotocin but none of these worked in humans suggesting that this is a poor model (Devendra et al., 2004, Gale, 2004). The insufficiency of prevention or treatment belies the need for a true cure for diabetes. The closest to a cure that has been achieved is islet transplants, where cadaveric islets are transplanted into the hepatic portal vein (Shapiro et al., 2000). Originally this was attempted using immunosuppresion with glucocorticoids but this had a very low success rate, only 12.4% of transplants functioned for more than 1 week as glucocorticoids are diabetogenic. The Edmonton Protocol revolutionised this using immune suppression by antibodies - daclizumab, sirolimus and tacrolimus which showed no immune rejection and long term insulin independence. Multiple transplants were required to achieve insulin independence (2 on average, but 3 was needed in some cases), but all patients trialled did not meet clinical diagnostic criteria for diabetes 6 months on, due to adequate production of insulin and measurable C-peptide in response to oral glucose challenges. The long-term follow up suggested this was not a permanent cure but would need repeating; with refinement islet transplants could offer a permanent cure (Ryan et al., 2005). There are not enough donor sources of islets for the operation to become commonplace when combined with the difficulties, and so stem cell therapies may be put to use in this way once developed.

1.5 Stem Cells

1.5.1 Pancreatic stem cells

Stem cells are present in most organs throughout the body and are responsible for generation of the organ, as well as post-developmental growth and repair. It has therefore been considered that there are likely to be pancreatic stem cells. The evidence regarding pancreatic stem cells is, however, mixed as I will discuss. There are also diﬀerences between humans and rodents which complicates matters.

Early evidence supporting the existence of pancreatic stem cells was that performing at 95% pancreatectomy, pancreatic duct ligation (PDL) or selective β-cell ablation in rodents would lead to regeneration of the pancreas (Li et al., 2010). Attempts to discern the source of the regeneration have been mixed. Some regeneration is due to the β-cell duplication, and some due to ductal cells, as discussed in section 1.2.3. PDL has also been shown to cause an increase in nestin within the pancreas (Peters et al., 2004), which as discussed earlier (1.2.4.6), may mark stem cells including

MSCs, but is not speciﬁc (Street et al., 2004). These reports support the existence of pancreatic stem cells in rodents, but cannot of course be replicated in humans.

34 A separate phenomenon by which pancreatic mass can increase, is during pregnancy (Sorenson and Brelje, 1997). This occurs in both humans and rodents. Experimental lineage tracing, and histological studies have suggested this is due to β-cell duplication in rodents, meaning stem cells, should they exist may only respond to certain extreme stimuli.

As it has proved challenging to identify, or even demonstrate the existence of pancreatic stem cells, attempts have been made to culture and purify them. The post-developmental pancreas is not proliferative, so any proliferative cells in culture may represent progenitor cells, with the exception of fibroblasts. Fibroblasts can be cultured from most tissues and proliferate in culture for a few passages. In direct culture with serum-containing media, endocrine, exocrine and ductal tissue all produce MSCs. These will be discussed further in section 1.5.2 but express nestin, linking to the observation of increased nestin seen following PDL. Variations on this culture have produced different cells. The addition of a ROCK inhibitor and TGFβ inhibitor generates epithelial rather than mesenchymal cells, but with a much lower capacity for proliferation (Lima et al., 2013). Complex ductal structures can also be induced to proliferate and differentiate (Huch et al., 2013, Corritore et al., 2014). These are of interest as it has been hypothesised that stem cells may reside in the ducts, based on regeneration of the pancreas from ductal cells (Li et al., 2010). However, these ductal cultures have not been well studied.

It is unclear from the evidence generated to date if pancreatic stem cells exist, or are capable of pancreas regeneration. The two potential sources are ducts and MSCs, with the most readily proliferative cells being the MSCs.

1.5.2 Mesenchymal stem cells

Sources and characteristics

MSCs represent a diverse set of cells from throughout the body, with largely overlapping characteristics. Initially, and most commonly, MSCs are derived from the bone marrow, and bone marrow

MSCs (bmMSCs) have been investigated as a source for diabetes treatment (Li and Ikehara, 2013).

They have since been derived from adipose tissue, the pancreas, Wharton’s jelly, umbilical cord, periosteum and dental pulp. MSCs from each of these sources have the same deﬁning characteristics, with some variation based on tissue of origin.

The characteristics to deﬁne MSCs, as set down by The International Society for Cellular Ther-

35 Figure 1.7: Sources of MSCs pMSCs have been derived from exocrine cells and β-cells, confirmed by lineage tracing (Lima et al., 2013, Russ et al., 2008) but may also be derived from duct cells (Corritore et al., 2014) or CD90/CD105 tissue resident cells (Carlotti et al., 2010). apy, are plastic adherence, a panel of cell surface markers and the ability to differentiate to three mesodermal lineages (Dominici et al., 2006). Differentiation and the ability to serially passage the cells, set MSCs aside from fibroblasts. Fibroblasts also show plastic adherence and may contaminate cell cultures for the first few passages. Differentiation of fibroblasts has been reported (Brohem et al., 2013) but the evidence to suggest that they were not in fact MSCs is unclear. The cell surface markers required are CD73, CD90 and CD195, with at least 95% of cells being positive for these markers. Other markers may also be used to support the identification. Cells should also be lacking any of CD14, CD19, CD34 and CD45, with less than 2% of cells expressing these, to rule out the presence of contaminating haematopoeitic stem cells. This is more of a problem when deriving MSCs from the bone marrow, which is what these characteristics were developed for - there is no reason to expect haematopoeitic stem cells to be present outside of the bone marrow. The functional assay, of differentiation to fat, bone and cartilage shows MSCs are multipotent. As with the cell surface markers, this characteristic was determined on bmMSCs. It has been shown that pancreatic MSCs

(pMSCs) demonstrate inefficient differentiation to adipocytes, whilst adipose-derived MSCs show much higher efficiency. Other differences between MSCs from various sources include the proliferation rates, duration of proliferation in culture, and efficiency of differentiation to other lineages, including β-cells.

36 Figure 1.8: EMT in pMSCs Reversible EMT in pMSCs may be responsible for phenotypes in culture; serum addition and/or monolayer culture yields mesenchymal cells, whilst serum withdrawal or non-adherent culture yields epithelial-like cells (Dalvi et al., 2009, Davani et al., 2009).

PMSCs have been derived from endocrine, exocrine and ductal cells. It is nigh impossible to separate these diﬀerent tissues, so pMSCs are grown from enriched, but ultimately mixed cell populations. It has been suggested that pMSCs are tissue-resident such as the observed CD90/CD105 positive cells inside islets (Carlotti et al., 2010), or that they are the product of de-diﬀerentiation via epithelial-mesenchymal transition (EMT). Lineage tracing has shown MSCs can be derived from

β-cells, but are not solely (Russ et al., 2008, 2009, Joglekar et al., 2009, Sordi et al., 2010, Lima et al., 2013). From a mixed pancreas digest, it is therefore likely that MSCs from multiple sources will be obtained, as shown in Figure 1.5. It has been suggested that MSCs, generated by EMT are an artifact of cell culture conditions (Seeberger et al., 2008). What is unclear is if that is a problem.

Particularly for in vitro modelling, so long as the terminal cells are functional, the manner in which they were derived may not matter. Further, the process seems to be at least partially reversible, with the formation of islet-like clusters (Dalvi et al., 2009). These cell clusters are unable to graft in mice though (Davani et al., 2009). The authors of the study in which this was assessed, suggested that the cues in the mouse blood may again be switching the cell phenotype back (Davani et al.,

2009), shown in Figure 1.7.

Whilst the evidence is mixed, these experiments most strongly suggest that the MSC, or MSC-like cells produced from the pancreas are a result of de-differentiation of mature cells and are an effect of the in vitro culture. Cells containing a selection of the same markers as these MSCs are present in the pancreas in vivo as discussed further in Section 1.5.2, but this is not definitive proof of the existence of MSCs in vivo as further concurrent markers would be needed, as well as the ability to isolate these cells. The production of ductal structures mentioned above shows better evidence for a stem cell, however, in vitro ductal structures are produced from mixed progenitor populations rather than isolated single cells. A lack of unique cell surface markers for pancreatic stem cells present during development has hindered identification of any remaining post-natal stem cells, but with the

37 current evidence it seems unlikely that there are stem cells present in the adult pancreas. Rather the balance of evidence suggests that cells with the potential to re-diﬀerentiate can be produced in vitro by de-diﬀerentiation of mature cells.

Diﬀerentiation

Diﬀerentiation of MSCs to β-cells or β-like cells has been documented in MSCs from multiple sources, using varied protocols, and with varied outcomes. A cross-section of these are shown in Table 1.1-2.

Whilst growth factors have been used to reprogram bone marrow, adipose, umbilical cord and Whar- ton’s jelly MSCs to β-cells, groups have reported that serum withdrawal alone induces formation of hormone-positive clusters of cells - ‘islet-like clusters’. The use of additional growth factors seems to enhance whether GSIS is seen following diﬀerentiation, and many protocols use the 3D culture.

These cell clusters are not unique to pMSCs, with most, if not all MSCs forming dense spheroids in low adherence culture, reviewed in Cesarz and Tamama (2015). As the authors discuss, spheroids may be autoﬂuorescent and are used to aid diﬀerentiation of bmMSCs to bone and cartilage, so there are some question marks over the authenticity and accuracy of pancreatic islet-like cluster results.

1.5.3 Pluripotent stem cells

Characteristics

Whereas MSCs are multipotent, able to diﬀerentiate to several lineages, ESCs and iPSCs are pluripotent. Pluripotent cells are more ‘basal’ and able to generate all body tissues.

ESCs are derived from the inner cell mass of blastocysts (Thomson et al., 1998), an early stage of embryonic development. They proliferate in culture for a prolonged period (>100 passages) and generate cells of every germ layer. This means that when these cells are transplanted into immuno- compromised mice, they form teratomas. Teratomas are tumours containing multiple cell types, randomly diﬀerentiated, and may include hair and full teeth. ESCs are the gold standard pluripotent cells, as they are derived from the earliest stages of human development.

Pluripotent stem cells can also be derived from any somatic cell, these are iPSCs. In the land- mark study, by simply transducing cells with the pluripotency factors OCT4, NANOG, KLF4 and c-Myc, they were reprogrammed to a state similar to ESCs (Takahashi et al., 2007) There are some differences between ESCs and iPSCs but the defining characteristic of pluripotency is the same. The differences are primarily epigenetic, with incomplete reprogramming occurring in iPSCs, potentially

38 Cell type Number Number With Factors Insulin Reference of stages of days non- monolayer stage Human pMSCs (duct) 4 12 Yes 1% FBS, 10 mM nic, 10 nM Ex-4, 10 PCR only, immunoﬂuo- Seeberger et al. P2-6 ng/ml TGFβ1 rescence negative (2006) Human bmMSCs & pla- 2 12 No 1% N2, 1% B27, 25 ng/ml bFGF, 10 mM PCR only Battula et al. cental MSCs nic, gelatin (2007) Human pMSCs (islet) 1 7-21 Yes 1% ITS, 1% BSA 8.8 µU/ml in 16.7 mM Gallo et al. glucose, 13.2 µU/ml in (2007) 3.3 mM glucose Human umbilical cord, 3 <7 No 1 mM NaB, 50 ng/ml activin a, 0.2- 0.25-0.37 µU/ml in 15 Bhandari et al. Wharton’s jelly & amni- 1% BSA, 50 ng/ml KGF, custom culture mM glucose (0.0-0.25 (2011) otic MSCs media µU/ml for controls) Human bmMSCs & pM- 3 21 Yes 5% platelet lysate, 10 µM retinoic 30 pmol for pMSCs, 11 Zanini et al. SCs (islet) acid,10 µg/ml activin a, 200 µg/ml Glp- pmol bmMSCs in 25 (2011) 39 1, 20 ng/ml EGF, 10 ng/ml FGF, 10 mM glucose µg/ml Btc, 10 mM nic Rat bmMSCs 6 12 No 2-15% FBS, 100 ng/ml activin a, 10 µM >1.5 ng/ml in 25 mM Wang et al. retinoic acid,10 ng/ml bFGF, 10 mM glucose, <0.5 ng/ml in 3 (2012) nic, 10 nM Ex-4, matrigel mM glucose Human bmMSCs, adM- 1 14-21 No 10 mM nic, 2 mM activin a, 10 nM Ex-4, 4-12 ng/mg in 25 mM Kim et al. SCs, Wharton’s jelly & 100 pM HGF, 2% B27, 2% N2 glucose, 3-4 in 3 mM (2012) periosteum glucose, insulin/total protein Mouse pMSCs (islet) 3 10 Yes 1% BSA, 0.3-3 mM taurine, 10 mM nic, 498 µIU/ml in 16.5 mM Gopurappilly 10 mM Glp-1 glucose et al. (2013)

Table 1.1: Cross-section of MSC to β-cell diﬀerentiation protocols Abbreviations used in this table: P - passage, FBS - foetal bovine serum, ITS - insulin/transferrin/selenium, BSA - bovine serum albumin, bFGF - basic ﬁbroblast growth factor, HGF - hepatocyte growth factor, GLP - glucagon like peptide, Nic - nicotinamide, NaB - sodium butyrate, Ex-4 - Exendin-4, NEAA - non-essential amino acids, IBMX - 3-isobutyl-1-methylxanthine, Btc - betacellulin, EGF - epidermal growth factor, TSA - trichostatin-A Cell type Number Number With Factors Insulin Reference of stages of days non- monolayer stage Human adMSCs 4 17 No 5-10 ng/ml activin a, 50 ng/ml Wnt3a, Immunocytochemistry Li et al. (2012) 10 µM retinoic acid, 10 ng/ml DKK1, 20 only (Insulin, C- ng/ml EGF, 60 ng/ml bFGF, 1x NEAA, peptide, PDX1) 2% B27, 10 ng/ml Ex-4, 10 mM Nic Human bmMSCs 3 21 Yes 1% BSA, 1% ITS, 2 nM Activin A, 0.1-5 71.6 µU/ml in 25 mM Jafarian et al. mM taurine, 1 mM NaB, 100 pM HGF, glucose, 19.4 µU/ml in (2014) 10-100 nM Ex-4, 1x NEAA, 2% B27, 2% 5.5 mM glucose (both N2 with IBMX) Human bmMSCs 1 15 No 10% FBS, 10 ng/ml conophylline, 1 nM 0.005/0.017 ng/µg pro- Gabr et al.

40 Btc tein/h in 5.5/25 mM (2014) glucose Human bmMSCs 2 10 No 55 nM TSA, 10 nM GLP-1 0.006/0.019 ng/µg pro- Gabr et al. tein/h in 5.5/25 mM (2014) glucose Human bmMSCs 3 18 No 0.5 mM β-mercaptoethanol, 1% NEAA, 0.006/0.019 ng/µg pro- Gabr et al. 20 ng/ml bFGF, 20 ng/ml EGF, 2% B27, tein/h in 5.5/25 mM (2014) 2 mM L-glutamine, 10 ng/ml btc, 10 glucose ng/ml activin a, 10 mM nic

Table 1.2: Cross-section of MSC to β-cell differentiation protocols Abbreviations used in this table: P - passage, FBS - foetal bovine serum, ITS - insulin/transferrin/selenium, BSA - bovine serum albumin, bFGF - basic fibroblast growth factor, HGF - hepatocyte growth factor, GLP - glucagon like peptide, Nic - nicotinamide, NaB - sodium butyrate, Ex-4 - Exendin-4, NEAA - non-essential amino acids, IBMX - 3-isobutyl-1-methylxanthine, Btc - betacellulin, EGF - epidermal growth factor, TSA - trichostatin-A in a lineage specific fashion (Kim et al., 2011, Bar-Nur et al., 2011, Beagan et al., 2016). This may lead to a bias in differentiation of iPSCs to certain cells which is not exhibited by ESCs. Recently, it has been found that simply erasing epigenetic marks in mouse epiblast cells will also induce reprogramming to a pluripotent state (Zhang et al., 2016). The main advantage of iPSCs over ESCs is that they can be created in a patient-specific fashion, for specific diseases (Yamanaka, 2012). There is considerable variation between iPSC lines (Koyanagi-Aoi et al., 2013, Rouhani et al., 2014), even from the same healthy or diseased background, representing the natural genetic variability in the population. To combat this, large banks are being generated of both healthy iPSCs and those from patients with monogenic disorders (Moran, 2013). It is hoped that banks, such as HipSci (Leha et al.,

2016), will enable development of more robust and repeatable diﬀerentiation protocols for cells of diﬀerent lineages (Moran, 2013).

1.5.4 Generation of iPSCs

Initial generation of iPSCs was by retroviral transduction with lentiviruses, using the ‘Yamanaka factors’ (Takahashi et al., 2007). These factors as mentioned previously are OCT4, SOX2, KLF4 and c-Myc. Whilst OCT4 and SOX2 are also markers of pluripotency, KLF4 aids with MET and formation of epithelial colonies, and acts in complex with OCT4, SOX2 or NANOG (Kim et al.,

2008). C-MYC is used to aid long term growth of the cells. Alternatively LIN28 has been used instead of c-Myc (Yu et al., 2007), which is not functionally similar but rather is involved in pro- moting translation of OCT4 mRNA. OCT4 and SOX2 are the key regulators of pluripotency, acting in feed-forward loops wherein OCT4 and SOX2 together bind to an OCT4 enhancer (Chew et al.,

2005) as well as transcriptionally regulating NANOG, the third primary pluripotency marker (Rodda et al., 2005). OCT4, in concert with the other factors further causes speciﬁc gene activation (Sharov et al., 2008) and prevents diﬀerentiation by gene silencing via polycomb complexes and miRNAs

(Kashyap et al., 2009).

The retroviruses used in the initial iPSC induction are integrative viruses, and in the original study the authors noted at least three retroviral integration sites for each of the four factors. De- pending on which loci the viruses integrate at (Stocking et al., 1993), there can be impacts upon the cells normal function or induction of a cancerous phenotype via insertional mutagenesis (Li and

Dahiya, 2002, Hacein-Bey-Abina et al., 2003). Further, the viruses can reactivate causing problems such as cancers (Okita et al., 2007). Due to these issues, non-integrative transduction methods have been developed. One such methods is transduction via Sendai virus (Fusaki et al., 2009, Seki et al., 2010) which is passively cleared from the cells over time, but remains present suﬃciently

41 long enough to induce reprogramming. A downside to this method is that any imbalance of factors between the diﬀerent vectors can aﬀect reprogramming and so clones must be fully assessed (Carey et al., 2011). An alternative method which is fully virus free is to reprogram by mRNA transduction

(Warren et al., 2010). This method guarantees no virus will be found in cells at any stage which is optimum when cells may be used in vivo but is labour intensive, requiring multiple rounds of transfection. Other methods have been investigated such as directly introducing the proteins, which are reviewed by Robinton and Daley (2012) but these methods have very low efficiency, with normal reprogramming already being at most around 1% efficient. Efficiency of reprogramming is generally low due to barriers, such as the difficulty in full erasure of histone modifications associated with the differentiated cell phenotype, reorganisation of chromatin and activation of the pluripotency associated genes, with numerous examples reviewed by Ebrahimi (2015). Whilst only one high quality iPSC line is required per patient from the selected clones, efficiency of reprogramming to ensure fully reprogrammed iPSCs can be boosted with modulators of epigenetics. These modulators enhance the reorganisation of the genome, an example of which is valproic acid, a histone deacetylase inhibitor which eliminates the need for c-Myc (Huangfu et al., 2008a,b). This is also desirable for in vivo applications, or applications wherein a normal cell cycle is required. Other factors including GLIS1

(Maekawa et al., 2011), SALL4 (Tsubooka et al., 2009) and NANOG (Silva et al., 2009) have also been shown to enhance reprogramming eﬃciency and in some cases the quality of reprogramming, but these require the use of a further vector when reprogramming.

Diﬀerentiation

Differentiation of pluripotent cells to β-cells is a long, step-wise process, which has proven very difficult to date, discussed by (Quiskamp et al., 2015). The pancreas is endodermal, and generation of definitive endoderm is relatively straight-forward. Following this, cells must be differentiated to gut tube, pancreatic endoderm, endocrine progenitors and finally immature to mature β-cells. The

first major report of generating β-cells was D’Amour et al. (2006), but these cells were not mature and appeared to be polyhormonal. The length of the protocols, use of dozens of factors and number of weakly-defined stages make improving protocols challenging. This is further complicated by the fact that seemingly small factors, such as seeding density (Gage et al., 2013), can have a big effect.

Work on deﬁning developmental stages in human foetal tissue has helped to improve on protocols.

For example, it was shown that NGN3 is expressed only for a window during human development

(Salisbury et al., 2014). Subsequently, it was found that prolonged or premature expression of NGN3 during iPSC diﬀerentiation primes cells to become polyhormonal or glucagon expressing (Johansson et al., 2007) and addition of ascorbic acid (vitamin C) can downregulate NGN3 to prevent this and

42 enrich for β-cells (Rezania et al., 2014). This addition, and numerous other reﬁnements led to two similar protocols being published within weeks of each other, demonstrating a major step forward in diﬀerentiation of pluripotent cells to β-cells (Rezania et al., 2014, Pagliuca et al., 2014). These papers demonstrated functional β-cells, which were glucose responsive across multiple repeated stimuli, and able to reverse diabetes in mice. These have heralded a new wave of β-cell research based on pluripotent stem cells, with the two papers amassing a cumulative 241 citations in under two years.

This means that in vitro modelling of pancreatic disease such as diabetes and CHI, using pluripotent stem cells, may now be possible (Balboa and Otonkoski, 2015). Indeed iPSCs have now been derived from patients with type 1 diabetes, wherein the production of β-cells was not adversely aﬀected by the genetic make-up of the patients (Millman et al., 2016).

1.6 Aims

The aims of this project were to derive stem cells (MSCs and/or iPSCs) from patients with CHI.

Following successful derivation, confirmed by phenotyping, these stem cells would be assessed for their usefulness for in vitro modelling of diabetes and CHI. To do this, the aim was to differentiate the stem cells to functional β-cells, by surveying and modifying protocols from the published literature as necessary. The success of these differentiation protocols would be assessed by gene and protein expression, and insulin secretion. Further, an investigation into whether the cells could be used for modelling hyperproliferation in CHI would be carried out, based on the recent data demonstrating this in CHI tissue. Together these separate threads would give an overall picture of whether CHI stem cells could be used for modelling CHI or diabetes in vitro.

1.7 References

L. Aguilar-Bryan, J. P. Clement, G. Gonzalez, K. Kunjilwar, A. Babenko, and J. Bryan. Toward

Understanding the Assembly and Structure of KATP Channels. Physiological Reviews, 78(1):

227–245, 1998.

U. Ahlgren, J. Jonsson, and H. Edlund. The morphogenesis of the pancreatic mesenchyme is uncou-

pled from that of the pancreatic epithelium in IPF1/PDX1-deﬁcient mice. Development, 122(5):

1409–1416, 1996.

U. Ahlgren, S. L. Pfaﬀ, T. M. Jessell, T. Edlund, and H. Edlund. Independent requirement for ISL1

in formation of pancreatic mesenchyme and islet cells. Nature, 385(6613):257–260, 1997.

J.-B. Arnoux, P. de Lonlay, M.-J. Ribeiro, K. Hussain, O. Blankenstein, K. Mohnike,

V. Valayannopoulos, J.-J. Robert, J. Rahier, C. Sempoux, C. Bellanne, V. Verkarre, Y. Aigrain,

43 F. Jaubert, F. Brunelle, and C. Nihoul-Fekete. Congenital hyperinsulinism. Early Human Devel-

opment, 86(5):287–294, 2010.

J.-B. Arnoux, V. Verkarre, C. Saint-Martin, F. Montravers, A. Brassier, V. Valayannopoulos,

F. Brunelle, J.-C. Fournet, J.-J. Robert, Y. Aigrain, C. Bellanne-Chantelot, and P. de Lonlay.

Congenital hyperinsulinism: current trends in diagnosis and therapy. Orphanet Journal of Rare

Diseases, 6(1):1750–1172, 2011.

I. Artner, B. Blanchi, J. C. Raum, M. Guo, T. Kaneko, S. Cordes, M. Sieweke, and R. Stein. MafB

is required for islet β cell maturation. Proceedings of the National Academy of Sciences, 104(10):

3853–3858, 2007.

V. B. Arya, Q. Aziz, A. Nessa, A. Tinker, and K. Hussain. Congenital hyperinsulinism: clinical and

molecular characterisation of compound heterozygous ABCC8 mutation responsive to Diazoxide

therapy. International Journal of Pediatric Endocrinology, 2014(1):1–5, 2014.

F. M. Ashcroft and P. Rorsman. ATP-sensitive K+ channels: a link between B-cell metabolism and

insulin secretion. Biochemical Society Transactions, 18(1):109–111, 1990.

M. A. Atkinson and G. S. Eisenbarth. Type 1 diabetes: new perspectives on disease pathogenesis

and treatment. The Lancet, 358(9277):221–229, 2001.

M. A. Atkinson, G. S. Eisenbarth, and A. W. Michels. Type 1 diabetes. Lancet, 383(9911):69–82,

2013.

D. Balboa and T. Otonkoski. Human pluripotent stem cell based islet models for diabetes research.

Best Practice & Research Clinical Endocrinology & Metabolism, 29(6):899–909, 2015.

O. Bar-Nur, H. Russ, S. Efrat, and N. Benvenisty. Epigenetic Memory and Preferential Lineage-

Speciﬁc Diﬀerentiation in Induced Pluripotent Stem Cells Derived from Human Pancreatic Islet

Beta Cells. Cell Stem Cell, 9(1):17–23, 2011.

V. L. Battula, P. M. Bareiss, S. Treml, S. Conrad, I. Albert, S. Hojak, H. Abele, B. Schewe, L. Just,

T. Skutella, and H.-J. Bahring. Human placenta and bone marrow derived MSC cultured in

serum-free, b-FGF-containing medium express cell surface frizzled-9 and SSEA-4 and give rise to

multilineage diﬀerentiation. Diﬀerentiation, 75(4):279 – 291, 2007.

J. Beagan, T. Gilgenast, J. Kim, Z. Plona, H. Norton, G. Hu, S. Hsu, E. Shield, X. Lyu, E. Apostolou,

K. Hochedlinger, V. Corces, J. Dekker, and J. Phillips-Cremins. Local genome topology can exhibit

an incompletely rewired 3d folding state during somatic cell reprogramming. Cell Stem Cell, 18

(5):611–624, 2016.

44 A. Belﬁore and R. Malaguarnera. Insulin receptor and cancer. Endocrine-Related Cancer, 18(4):

R125–R147, 2011.

C. Bertoli, J. M. Skotheim, and R. A. M. de Bruin. Control of cell cycle transcription during G1

and S phases. Nat Rev Mol Cell Biol, 14(8):518 – 528, 2013.

M. Best, M. Carroll, N. A. Hanley, and K. Piper Hanley. Embryonic stem cells to beta-cells by

understanding pancreas development. Molecular and Cellular Endocrinology, 288(1–2):86–94, 2008.

D. R. Bhandari, K.-W. Seo, B. Sun, M.-S. Seo, H.-S. Kim, Y.-J. Seo, J. Marcin, N. Forraz, H. L.

Roy, D. Larry, M. Colin, and K.-S. Kang. The simplest method for in vitro β-cell production from

human adult stem cells. Diﬀerentiation, 82(3):144 – 152, 2011.

B. Blum, S. Hrvatin, C. Schuetz, C. Bonal, A. Rezania, and D. A. Melton. Functional beta-cell

maturation is marked by an increased glucose threshold and by expression of urocortin 3. Nat

Biotech, 30(3):261–264, 2012.

B. Blum, A. N. Roose, O. Barrandon, R. Maehr, A. C. Arvanites, L. S. Davidow, J. C. Davis, Q. P.

Peterson, L. L. Rubin, and D. A. Melton. Reversal of β cell de-diﬀerentiation by a small molecule

inhibitor of the TGFβ pathway. eLife, 3:e02809–, 2014.

S. Bonner-Weir and G. C. Weir. New sources of pancreatic β-cells. Nat Biotech, 23(7):857–861, 2005.

S. Bonner-Weir, W.-C. Li, L. Ouziel-Yahalom, L. Guo, G. C. Weir, and A. Sharma. β-Cell Growth

and Regeneration: Replication Is Only Part of the Story. Diabetes, 59(10):2340–2348, 2010.

L. Bouwens and I. Rooman. Regulation of Pancreatic Beta-Cell Mass. Physiological Reviews, 85(4):

1255–1270, 2005.

M. Braun, R. Ramracheya, M. Bengtsson, Q. Zhang, J. Karanauskaite, C. Partridge, P. R. Johnson,

and P. Rorsman. Voltage-Gated Ion Channels in Human Pancreatic β-Cells: Electrophysiological

Characterization and Role in Insulin Secretion. Diabetes, 57(6):1618–1628, 2008.

M. Brissova, M. J. Fowler, W. E. Nicholson, A. Chu, B. Hirshberg, D. M. Harlan, and A. C. Powers.

Assessment of Human Pancreatic Islet Architecture and Composition by Laser Scanning Confocal

Microscopy. Journal of Histochemistry & Cytochemistry, 53(9):1087–1097, 2005.

C. Brohem, C. de Carvalho, C. Radoski, F. Santi, M. Baptista, B. Swinka, C. de A Urban,

L. de Araujo, R. Graf, I. Feferman, and M. Lorencini. Comparison between ﬁbroblasts and mes-

enchymal stem cells derived from dermal and adipose tissue. Int J Cosmet Sci., 35(5):448–57,

2013.

45 G. J. Bruining. Recent advances in hyperinsulinism and the pathogenesis of diabetes mellitus. Current

Opinion in Pediatrics, 2(4):758–765, 1990.

A. E. Butler, J. Janson, S. Bonner-Weir, R. Ritzel, R. A. Rizza, and P. C. Butler. β-Cell Deﬁcit and

Increased β-Cell Apoptosis in Humans With Type 2 Diabetes. Diabetes, 52(1):102–110, 2003.

O. Cabrera, D. M. Berman, N. S. Kenyon, C. Ricordi, P.-O. Berggren, and A. Caicedo. The unique

cytoarchitecture of human pancreatic islets has implications for islet cell function. Proceedings of

the National Academy of Sciences of the United States of America, 103(7):2334–2339, 2006.

B. Carey, S. Markoulaki, J. Hanna, D. Faddah, Y. Buganim, J. Kim, K. Ganz, E. Steine, J. Cas-

sady, M. Creyghton, G. Welstead, Q. Gao, and R. Jaenisch. Reprogramming factor stoichiometry

inﬂuences the epigenetic state and biological properties of induced pluripotent stem cells. 9(6):

588–598, 2011.

F. Carlotti, A. Zaldumbide, C. J. Loomans, E. van Rossenberg, M. Engelse, E. J. de Koning, and

R. C. Hoeben. Isolated human islets contain a distinct population of mesenchymal stem cells.

Islets, 2(3):164–173, 2010.

M. Carrasco, I. Delgado, B. Soria, F. Martin, and A. Rojas. GATA4 and GATA6 control mouse

pancreas organogenesis. J Clin Invest, 122(10):3504–3515, 2012.

C. Cavelti-Weder, M. Shtessel, J. E. Reuss, A. Jermendy, T. Yamada, F. Caballero, S. Bonner-

Weir, and G. C. Weir. Pancreatic Duct Ligation After Almost Complete β-Cell Loss: Exocrine

Regeneration but No Evidence of β-Cell Regeneration. Endocrinology, 154(12):4493–4502, 2013.

CDC. Prevalence of overweight and obesity among adults with diagnosed Diabetes United States,

1988-1994 and 1999-2000. Centers for Disease Control and Prevention (CDC) (November 2004)

MMWR. Morbidity and Mortality Weekly Report, 53(45):1066–68, 2004.

Z. Cesarz and K. Tamama. Spheroid Culture of Mesenchymal Stem Cells. Stem Cells International,

2016:9176357–, 2015.

J.-L. Chew, Y.-H. Loh, W. Zhang, X. Chen, W.-L. Tam, L.-S. Yeap, P. Li, Y.-S. Ang, B. Lim, P. Rob-

son, and H.-H. Ng. Reciprocal transcriptional regulation of pou5f1 and sox2 via the oct4/sox2

complex in embryonic stem cells. Molecular and Cellular Biology, 25(14):6031–6046, 2005.

V. Cirulli, G. M. Beattie, G. Klier, M. Ellisman, C. Ricordi, V. Quaranta, F. Frasier, J. K. Ishii,

A. Hayek, and D. R. Salomon. Expression and Function of α3 and β5 Integrins in the Developing

Pancreas: Roles in the Adhesion and Migration of Putative Endocrine Progenitor Cells. The

Journal of Cell Biology, 150(6):1445–1460, 2000.

46 M. Cnop, M. J. Landchild, J. Vidal, P. J. Havel, N. G. Knowles, D. R. Carr, F. Wang, R. L. Hull,

E. J. Boyko, B. M. Retzlaﬀ, C. E. Walden, R. H. Knopp, and S. E. Kahn. The Concurrent Ac-

cumulation of Intra-Abdominal and Subcutaneous Fat Explains the Association Between Insulin

Resistance and Plasma Leptin Concentrations: Distinct Metabolic Eﬀects of Two Fat Compart-

ments. Diabetes, 51(4):1005–1015, 2002.

P. Collombat, A. Mansouri, J. Hecksher-Sorensen, P. Serup, J. Krull, G. Gradwohl, and P. Gruss.

Opposing actions of Arx and Pax4 in endocrine pancreas development. Genes & Development, 17

(20):2591–2603, 2003.

E. Corritore, E. Dugnani, V. Pasquale, R. Misawa, P. Witkowski, J. Lei, J. Markmann, L. Piemonti,

E. M. Sokal, S. Bonner-Weir, and P. A. Lysy. β-Cell Diﬀerentiation of Human Pancreatic Duct-

Derived Cells After In Vitro Expansion. Cellular Reprogramming, 16(6):456–466, 2014.

M. P. Dalvi, M. R. Umrani, M. V. Joglekar, and A. A. Hardikar. Human pancreatic islet progenitor

cells demonstrate phenotypic plasticity in vitro. Journal of Biosciences, 34(4):523–528, 2009.

K. A. D’Amour, A. G. Bang, S. Eliazer, O. G. Kelly, A. D. Agulnick, N. G. Smart, M. A. Moorman,

E. Kroon, M. K. Carpenter, and E. E. Baetge. Production of pancreatic hormone-expressing

endocrine cells from human embryonic stem cells. Nat Biotech, 24(11):1392–1401, 2006.

D. Daneman. Type 1 diabetes. The Lancet, 367(9513):847–858, 2006.

B. Davani, S. Ariely, L. Ikonomou, Y. Oron, and M. C. Gershengorn. Human islet-derived precursor

cells can cycle between epithelial clusters and mesenchymal phenotypes. Journal of Cellular and

Molecular Medicine, 13:2570–2581, 2009.

D. D. De Leon and C. A. Stanley. Mechanisms of Disease: advances in diagnosis and treatment of

hyperinsulinism in neonates. Nat Clin Pract End Met, 3(1):57–68, 2007.

A. Denley, E. R. Bonython, G. W. Booker, L. J. Cosgrove, B. E. Forbes, C. W. Ward, and J. C.

Wallace. Structural Determinants for High-Aﬃnity Binding of Insulin-Like Growth Factor II to

Insulin Receptor (IR)-A, the Exon 11 Minus Isoform of the IR. Molecular Endocrinology, 18(10):

2502–2512, 2004.

D. Devendra, E. Liu, and G. S. Eisenbarth. Type 1 diabetes: recent developments. BMJ, 328(7442):

750–754, 2004.

Diabetes UK. Diabetes UK Facts and Stats, Nov 2015. URL

https://www.diabetes.org.uk/About us/What-we-say/Statistics/.

47 M. Dominici, K. Le Blanc, I. Mueller, I. Slaper-Cortenbach, F. Marini, D. Krause, R. Deans, A. Keat-

ing, D. Prockop, and E. Horwitz. Minimal criteria for deﬁning multipotent mesenchymal stromal

cells. The International Society for Cellular Therapy position statement. Cytotherapy, 8(4):315–7,

2006.

Y. Dor, J. Brown, O. I. Martinez, and D. A. Melton. Adult pancreatic β-cells are formed by self-

duplication rather than stem-cell diﬀerentiation. Nature, 429(6987):41–46, 2004.

A. Du, C. S. Hunter, J. Murray, D. Noble, C.-L. Cai, S. M. Evans, R. Stein, and C. L. May. Islet-1

is Required for the Maturation, Proliferation, and Survival of the Endocrine Pancreas. Diabetes,

58(9):2059–2069, 2009.

M. J. Dunne, C. Kane, R. M. Shepherd, J. A. Sanchez, R. F. James, P. R. Johnson, A. Aynsley-

Green, S. Lu, J. P. Clement, K. J. Lindley, S. Seino, L. Aguilar-Bryan, G. Gonzalez, and P. J. Milla.

Familial Persistent Hyperinsulinemic Hypoglycemia of Infancy and Mutations in the Sulfonylurea

Receptor. New England Journal of Medicine, 336(10):703–706, 1997.

M. J. Dunne, K. E. Cosgrove, R. M. Shepherd, A. Aynsley-Green, and K. J. Lindley. Hyperinsulinism

in Infancy: From Basic Science to Clinical Disease. Physiological Reviews, 84(1):239–275, 2004.

M. Eberhardt, P. Salmon, M.-A. von Mach, J. G. Hengstler, M. Brulport, P. Linscheid, D. Seboek,

J. Oberholzer, A. Barbero, I. Martin, B. Muller, D. Trono, and H. Zulewski. Multipotential nestin

and Isl-1 positive mesenchymal stem cells isolated from human pancreatic islets. Biochemical and

Biophysical Research Communications, 345(3):1167–1176, 2006.

B. Ebrahimi. Reprogramming barriers and enhancers: strategies to enhance the eﬃciency and kinetics

of induced pluripotency. 4(1):1–12, 2015.

A. A. Elayat, M. M. el Naggar, and M. Tahir. An immunocytochemical and morphometric study of

the rat pancreatic islets. Journal of Anatomy, 186(Pt 3):629–637, 1995.

C. Evans-Molina, J. C. Garmey, R. Ketchum, K. L. Brayman, S. Deng, and R. G. Mirmira. Glucose

Regulation of Insulin Gene Transcription and Pre-mRNA Processing in Human Islets. Diabetes,

56(3):827–835, 2007.

N. Fiaschi-Taesch, T. A. Bigatel, B. Sicari, K. K. Takane, F. Salim, S. Velazquez-Garcia, G. Harb,

K. Selk, I. Cozar-Castellano, and A. F. Stewart. Survey of the Human Pancreatic β-Cell G1/S

Proteome Reveals a Potential Therapeutic Role for Cdk-6 and Cyclin D1 in Enhancing Human

β-Cell Replication and Function In Vivo. Diabetes, 58(4):882–893, 2009.

48 N. M. Fiaschi-Taesch, J. W. Kleinberger, F. G. Salim, R. Troxell, R. Wills, M. Tanwir, G. Casinelli,

A. E. Cox, K. K. Takane, D. K. Scott, and A. F. Stewart. Human pancreatic β-cell g1/s molecule

cell cycle atlas. Diabetes, 62(7):2450, 2013.

J.-C. Fournet, C. Mayaud, P. de Lonlay, M.-S. Gross-Morand, V. Verkarre, M. Castanet, M. Devillers,

J. Rahier, F. Brunelle, J.-J. Robert, C. Nihoul-Fekete, J.-M. Saudubray, and C. Junien. Unbalanced

Expression of 11p15 Imprinted Genes in Focal Forms of Congenital Hyperinsulinism. The American

Journal of Pathology, 158(6):2177–2184, 2001.

L. E. Fridlyand, D. A. Jacobson, and L. Philipson. Ion channels and regulation of insulin secretion

in human β-cells: A computational systems analysis. Islets, 5(1):1–15, 2013.

O. Friedman-Mazursky, R. Elkon, and S. Efrat. Rediﬀerentiation of expanded human islet β cells by

inhibition of ARX. Scientiﬁc Reports, 6:20698–, 2016.

K. Furuyama, Y. Kawaguchi, H. Akiyama, M. Horiguchi, S. Kodama, T. Kuhara, S. Hosokawa,

A. Elbahrawy, T. Soeda, M. Koizumi, T. Masui, M. Kawaguchi, K. Takaori, R. Doi, E. Nishi,

R. Kakinoki, J. M. Deng, R. R. Behringer, T. Nakamura, and S. Uemoto. Continuous cell supply

from a Sox9-expressing progenitor zone in adult liver, exocrine pancreas and intestine. Nat Genet,

43(1):34–41, 2011.

N. Fusaki, B. A. N. Hiroshi, A. Nishiyama, K. Saeki, and M. Hasegawa. Eﬃcient induction of

transgene-free human pluripotent stem cells using a vector based on sendai virus, an rna virus

that does not integrate into the host genome:. 85(8):348–362, 2009.

M. M. Gabr, M. M. Zakaria, A. F. Refaie, S. M. Khater, S. A. Ashamallah, A. M. Ismail, N. El-Badri,

and M. A. Ghoneim. Generation of Insulin-Producing Cells from Human Bone Marrow-Derived

Mesenchymal Stem Cells: Comparison of Three Diﬀerentiation Protocols. BioMed Research In-

ternational, 832736:9, 2014.

B. K. Gage, T. D. Webber, and T. J. Kieﬀer. Initial Cell Seeding Density Inﬂuences Pancreatic

Endocrine Development During in vitro Diﬀerentiation of Human Embryonic Stem Cells. PLoS

ONE, 8(12):e82076–, 2013.

E. A. Gale. European Nicotinamide Diabetes Intervention Trial (ENDIT): a randomised controlled

trial of intervention before the onset of type 1 diabetes. The Lancet, 363(9413):925–931, 2004.

R. Gallo, F. Gambelli, B. Gava, F. Sasdelli, V. Tellone, M. Masini, P. Marchetti, F. Dotta, and

V. Sorrentino. Generation and expansion of multipotent mesenchymal progenitor cells from cul-

tured human pancreatic islets. Cell Death Diﬀer, 14(11):1860–1871, 2007.

49 K. E. Gerrish, M. A. Cissell, and R. Stein. The role of hepatic nuclear factor 1a and PDX-1 in

transcriptional regulation of the pdx-1 gene. Journal of Biological Chemistry, 276(51):47775–84,

2001.

G. K. Gittes. Developmental biology of the pancreas: A comprehensive review. Developmental

Biology, 326(1):4–35, 2009.

I. Giurgea, C. Bellanna-Chantelot, M. Ribeiro, L. Hubert, C. Sempoux, J.-J. Robert, O. Blankenstein,

K. Hussain, F. Brunelle, C. Nihoul-Fekete, J. Rahier, F. Jaubert, and P. de Lonlay. Molecular

Mechanisms of Neonatal Hyperinsulinism. Horm Res Paediatr, 66(6):289–296, 2006.

B. Glaser, H. Landau, A. Smilovici, and R. Nesher. Persistent hyperinsulinaemic hypoglycaemia of

infancy: long-term treatment with the somatostatin anologue sandostatin. Clinical Endocrinology,

31(1):71–80, 1989.

B. Glaser, P. Thornton, T. Otonkoski, and C. Junien. Genetics of neonatal hyperinsulinism. Archives

of Disease in Childhood - Fetal and Neonatal Edition, 82(2):F79–F86, 2000.

K. J. Godfrey, B. Mathew, J. C. Bulman, O. Shah, S. Clement, and G. I. Gallicano. Stem cell-based

treatments for Type1 diabetes mellitus: bone marrow, embryonic, hepatic, pancreatic and induced

pluripotent stem cells. Diabetic Medicine, 29(1):14–23, 2012.

R. Gopurappilly, V. Bhat, and R. Bhonde. Pancreatic tissue resident mesenchymal stromal cell

(msc)-like cells as a source of in vitro islet neogenesis. Journal of Cellular Biochemistry, 114(10):

2240–2247, 2013.

G. Gradwohl, A. Dierich, M. LeMeur, and F. Guillemot. neurogenin3 is required for the development

of the four endocrine cell lineages of the pancreas. Proceedings of the National Academy of Sciences,

97(4):1607–1611, 2000.

G. Gu, J. Dubauskaite, and D. A. Melton. Direct evidence for the pancreatic lineage: NGN3+ cells

are islet progenitors and are distinct from duct progenitors. Development, 129(10):2447–2457,

2002.

S. Hacein-Bey-Abina, C. von Kalle, M. Schmidt, F. Le Deist, N. Wulﬀraat, E. McIntyre, I. Radford,

J.-L. Villeval, C. C. Fraser, M. Cavazzana-Calvo, and A. Fischer. A serious adverse event after

successful gene therapy for x-linked severe combined immunodeﬁciency. New England Journal of

Medicine, 348(3):255–256, 2003.

N. Hanley. Closing in on pancreatic beta cells. Nat Biotech, 32(11):1100–1102, Nov. 2014.

50 M. Hansson, A. Tonning, U. Frandsen, A. Petri, J. Rajagopal, M. C. Englund, R. S. Heller, J. Hakans-

son, J. Fleckner, H. N. Skold, D. Melton, H. Semb, and P. Serup. Artifactual Insulin Release From

Diﬀerentiated Embryonic Stem Cells. Diabetes, 53(10):2603–2609, 2004.

H. Harashima, N. Dissmeyer, and A. Schnittger. Cell cycle control across the eukaryotic kingdom.

Trends in Cell Biology, 23(7):345 – 356, 2013.

J.-C. Henquin, C. Sempoux, J. Marchandise, S. Godecharles, Y. Guiot, M. Nenquin, and J. Rahier.

Congenital Hyperinsulinism Caused by Hexokinase I Expression or Glucokinase-Activating Muta-

tion in a Subset of β-Cells. Diabetes, 62(5):1689–1696, 2013.

Y. Hori, I. C. Rulifson, B. C. Tsai, J. J. Heit, J. D. Cahoy, and S. K. Kim. Growth inhibitors promote

diﬀerentiation of insulin-producing tissue from embryonic stem cells. Proceedings of the National

Academy of Sciences, 99(25):16105–16110, 2002.

F. B. Hu, J. E. Manson, M. J. Stampfer, G. Colditz, S. Liu, C. G. Solomon, and W. C. Willett. Diet,

Lifestyle, and the Risk of Type 2 Diabetes Mellitus in Women. N Engl J Med, 345(11):790–797,

2001.

D. Huangfu, R. Maehr, W. Guo, A. Eijkelenboom, M. Snitow, A. E. Chen, and D. A. Melton. Induc-

tion of pluripotent stem cells by deﬁned factors is greatly improved by small-molecule compounds.

26(7):795–797, 2008a.

D. Huangfu, K. Osafune, R. Maehr, W. Guo, A. Eijkelenboom, S. Chen, W. Muhlestein, and D. A.

Melton. Induction of pluripotent stem cells from primary human ﬁbroblasts with only oct4 and

sox2. 26(11):1269–1275, 2008b.

M. Huch, P. Bonfanti, S. F. Boj, T. Sato, C. J. M. Loomans, M. van de Wetering, M. Sojoodi,

V. S. W. Li, J. Schuijers, A. Gracanin, F. Ringnalda, H. Begthel, K. Hamer, J. Mulder, J. H. van

Es, E. de Koning, R. G. J. Vries, H. Heimberg, and H. Clevers. Unlimited in vitro expansion of

adult bi-potent pancreas progenitors through the Lgr5/R-spondin axis. The EMBO Journal, 32

(20):2708–2721, 2013.

E. Hunziker and M. Stein. Nestin-Expressing Cells in the Pancreatic Islets of Langerhans. Biochemical

and Biophysical Research Communications, 271(1):116–119, 2000.

K. Hussain. Diagnosis and Management of Hyperinsulinaemic Hypoglycaemia of Infancy. Horm Res

Paediatr, 69(1):2–13, 2008.

K. Hussain and A. Aynsley-Green. Hyperinsulinaemic hypoglycaemia in preterm neonates. Archives

of Disease in Childhood - Fetal and Neonatal Edition, 89(1):F65–F67, 2004.

51 D. Ish-Shalom, C. Christoﬀersen, P. Vorwerk, N. Sacerdoti-Sierra, R. Shymko, D. Naor, and

P. De Meyts. Mitogenic properties of insulin and insulin analogues mediated by the insulin recep-

tor. Diabetologia, 40(S2):S25–31, 1997.

T. Iype, J. Francis, J. C. Garmey, J. C. Schisler, R. Nesher, G. C. Weir, T. C. Becker, C. B. Newgard,

S. C. Griﬀen, and R. G. Mirmira. Mechanism of insulin Gene Regulation by the Pancreatic Tran-

scription Factor Pdx-1: Application of pre-mRNA analysis and chromatin immunoprecipitation

to assess formation of functional transcriptional complexes. Journal of Biological Chemistry, 280

(17):16798–16807, 2005.

M. M. Jack, R. M. Greer, M. J. Thomsett, R. M. Walker, J. R. Bell, C. Choong, D. M. Cowley,

A. C. Herington, and A. M. Cotterill. The outcome in Australian children with hyperinsulinism

of infancy: early extensive surgery in severe cases lowers risk of diabetes. Clinical Endocrinology,

58(3):355–364, 2003.

P. Jacquemin, F. P. Lemaigre, and G. G. Rousseau. The Onecut transcription factor HNF-6 (OC-1)

is required for timely speciﬁcation of the pancreas and acts upstream of Pdx-1 in the speciﬁcation

cascade. Developmental Biology, 258(1):105–116, 2003.

A. Jafarian, M. Taghikhani, S. Abroun, Z. Pourpak, A. Allahverdi, and M. Soleimani. Generation

of high-yield insulin producing cells from human bone marrow mesenchymal stem cells. Molecular

Biology Reports, 41(7):4783–4794, 2014.

C. James, R. R. Kapoor, D. Ismail, and K. Hussain. The genetic basis of congenital hyperinsulinism.

Journal of Medical Genetics, 46(5):289–299, 2009.

J. Jensen, R. S. Heller, T. Funder-Nielsen, E. E. Pedersen, C. Lindsell, G. Weinmaster, O. D. Madsen,

and P. Serup. Independent development of pancreatic alpha- and beta-cells from neurogenin3-

expressing precursors: a role for the notch pathway in repression of premature diﬀerentiation.

Diabetes, 49(2):163–176, 2000.

M. V. Joglekar, V. M. Joglekar, S. V. Joglekar, and A. A. Hardikar. Human fetal pancreatic insulin-

producing cells proliferate in vitro. Journal of Endocrinology, 201(1):27–36, 2009.

K. A. Johansson, U. Dursun, N. Jordan, G. Gu, F. Beermann, G. Gradwohl, and A. Grapin-Botton.

Temporal Control of Neurogenin3 Activity in Pancreas Progenitors Reveals Competence Windows

for the Generation of Diﬀerent Endocrine Cell Types. Developmental Cell, 12(3):457–465, 2007.

J. D. Johnson, E. Bernal-Mizrachi, E. U. Alejandro, Z. Han, T. B. Kalynyak, H. Li, J. L. Beith,

J. Gross, G. L. Warnock, R. R. Townsend, M. A. Permutt, and K. S. Polonsky. Insulin protects

52 islets from apoptosis via Pdx1 and speciﬁc changes in the human islet proteome. Proceedings of

the National Academy of Sciences, 103(51):19575–19580, 2006.

S. E. Kahn, M. E. Cooper, and S. Del Prato. Pathophysiology and treatment of type 2 diabetes:

perspectives on the past, present, and future. The Lancet, 383(9922):1068–1083, 2014.

T. Kaido, M. Yebra, V. Cirulli, C. Rhodes, G. Diaferia, and A. M. Montgomery. Impact of Deﬁned

Matrix Interactions on Insulin Production by Cultured Human β-Cells: Eﬀect on Insulin Content,

Secretion, and Gene Transcription. Diabetes, 55(10):2723–2729, 2006.

V. Kashyap, N. C. Rezende, K. B. Scotland, S. M. Shaﬀer, J. L. Persson, L. J. Gudas, and N. P.

Mongan. Regulation of stem cell pluripotency and diﬀerentiation involves a mutual regulatory

circuit of the nanog, oct4, and sox2 pluripotency transcription factors with polycomb repressive

complexes and stem cell micrornas. Stem Cells and Development, 18(7):1093–1108, 2009.

S. A. Kassem, I. Ariel, P. S. Thornton, I. Scheimberg, and B. Glaser. Beta-cell proliferation and

apoptosis in the developing normal human pancreas and in hyperinsulinism of infancy. Diabetes,

49(8):1325–1333, 2000.

S. A. Kassem, I. Ariel, P. S. Thornton, K. Hussain, V. Smith, K. J. Lindley, A. Aynsley-Green, and

B. Glaser. p57KIP2 Expression in Normal Islet Cells and in Hyperinsulinism of Infancy. Diabetes,

50(12):2763–2769, 2001.

Y. Kawaguchi. Sox9 and programming of liver and pancreatic progenitors. J Clin Invest, 123(5):

1881–1886, 2013.

I. Ketola, T. Otonkoski, M.-A. Pulkkinen, H. Niemi, J. Palgi, C. M. Jacobsen, D. B. Wilson, and

M. Heikinheimo. Transcription factor GATA-6 is expressed in the endocrine and GATA-4 in the

exocrine pancreas. Molecular and Cellular Endocrinology, 226(1–2):51–57, 2004.

J. Kim, J. Chu, X. Shen, J. Wang, and S. H. Orkin. An extended transcriptional network for

pluripotency of embryonic stem cells. Cell, 132(6):1049–1061, 2008.

K. Kim, R. Zhao, A. Doi, K. Ng, J. Unternaehrer, P. Cahan, H. Hongguang, Y.-H. Loh, M. J. Aryee,

M. W. Lensch, H. Li, J. J. Collins, A. P. Feinberg, and G. Q. Daley. Donor cell type can inﬂuence

the epigenome and diﬀerentiation potential of human induced pluripotent stem cells. Nat Biotech,

29(12):1117–1119, 2011.

S.-J. Kim, Y.-S. Choi, E.-S. Ko, S.-M. Lim, C.-W. Lee, and D.-I. Kim. Glucose-stimulated insulin

secretion of various mesenchymal stem cells after insulin-producing cell diﬀerentiation. Journal of

Bioscience and Bioengineering, 113(6):771 – 777, 2012.

53 J. L. Kopp, C. L. Dubois, E. Hao, F. Thorel, P. L. Herrera, and M. Sander. Progenitor cell domains

in the developing and adult pancreas. Cell Cycle, 10(12):1921–1927, 2011.

J. C. Koster, M. A. Permutt, and C. G. Nichols. Diabetes and Insulin Secretion: The ATP-Sensitive

K+ Channel (KATP) Connection. Diabetes, 54(11):3065–3072, 2005.

M. Koyanagi-Aoi, M. Ohnuki, K. Takahashi, K. Okita, H. Noma, Y. Sawamura, I. Teramoto,

M. Narita, Y. Sato, T. Ichisaka, N. Amano, A. Watanabe, A. Morizane, Y. Yamada, T. Sato,

J. Takahashi, and S. Yamanaka. Diﬀerentiation-defective phenotypes revealed by large-scale anal-

yses of human pluripotent stem cells. Proceedings of the National Academy of Sciences, 110(51):

20569–20574, 2013.

R. N. Kulkarni, E.-B. Mizrachi, A. G. Ocana, and A. F. Stewart. Human β-Cell Proliferation and

Intracellular Signaling: Driving in the Dark Without a Road Map. Diabetes, 61(9):2205–2213,

2012.

C. S. Lee, N. J. Sund, M. Z. Vatamaniuk, F. M. Matschinsky, D. A. Stoﬀers, and K. H. Kaestner.

Foxa2 Controls Pdx1 Gene Expression in Pancreatic β-Cells In Vivo. Diabetes, 51(8):2546–2551,

2002.

A. Leha, N. Moens, R. Meleckyte, O. J. Culley, M. K. Gervasio, M. Kerz, A. Reimer, S. A. Cain,

I. Streeter, A. Folarin, O. Stegle, C. M. Kielty, R. Durbin, F. M. Watt, and D. Danovi. A high-

content platform to characterise human induced pluripotent stem cell lines. Methods, 96:85–96,

2016.

B. Leibiger, T. Moede, S. Uhles, P.-O. Berggren, and I. B. Leibiger. Short-term regulation of insulin

gene transcription. Biochemical Society Transactions, 30(2):312–317, 2002.

G. Leibowitz, B. Glaser, A. A. Higazi, M. Salameh, E. Cerasi, and H. Landau. Hyperinsulinemic

hypoglycemia of infancy (nesidioblastosis) in clinical remission: high incidence of diabetes mellitus

and persistent beta-cell dysfunction at long-term follow-up. The Journal of Clinical Endocrinology

& Metabolism, 80(2):386–392, 1995.

J. Li, L. Zhu, X. Qu, J. Li, R. Lin, L. Liao, J. Wang, S. Wang, Q. Xu, and R. C. Zhao. Stepwise

Diﬀerentiation of Human Adipose-Derived Mesenchymal Stem Cells Toward Deﬁnitive Endoderm

and Pancreatic Progenitor Cells by Mimicking Pancreatic Development In Vivo. Stem Cells and

Development, 22(10):1576–1587, 2012.

L.-C. Li and R. Dahiya. MethPrimer: designing primers for methylation PCRs. Bioinformatics, 18

(11):1427–1431, 2002.

54 M. Li and S. Ikehara. Bone Marrow Stem Cell as a Potential Treatment for Diabetes. Journal of

Diabetes Research, 2013:5, 2013.

W.-C. Li, J. M. Rukstalis, W. Nishimura, V. Tchipashvili, J. F. Habener, A. Sharma, and S. Bonner-

Weir. Activation of pancreatic-duct-derived progenitor cells during pancreas regeneration in adult

rats. Journal of Cell Science, 123(16):2792–2802, 2010.

C. G. Liew, N. N. Shah, S. J. Briston, R. M. Shepherd, C. P. Khoo, M. J. Dunne, H. D. Moore, K. E.

Cosgrove, and P. W. Andrews. PAX4 Enhances Beta-Cell Diﬀerentiation of Human Embryonic

Stem Cells. PLoS ONE, 3(3):e1783–, 2008.

M. J. Lima, K. R. Muir, H. M. Docherty, R. Drummond, N. W. McGowan, S. Forbes, Y. Heremans,

I. Houbracken, J. A. Ross, S. J. Forbes, P. Ravassard, H. Heimberg, J. Casey, and K. Docherty.

Suppression of Epithelial-to-Mesenchymal Transitioning Enhances Ex Vivo Reprogramming of

Human Exocrine Pancreatic Tissue Toward Functional Insulin-Producing β-Like Cells. Diabetes,

62(8):2821–2833, 2013.

N. Lumelsky, O. Blondel, P. Laeng, I. Velasco, R. Ravin, and R. McKay. Diﬀerentiation of Embryonic

Stem Cells to Insulin-Secreting Structures Similar to Pancreatic Islets. Science, 292(5520):1389–

1394, 2001.

M. J. MacDonald, M. J. Longacre, S. W. Stoker, M. Kendrick, A. Thonpho, L. J. Brown, N. M. Hasan,

S. Jitrapakdee, T. Fukao, M. S. Hanson, L. A. Fernandez, and J. Odorico. Diﬀerences between

Human and Rodent Pancreatic Islets: Low pyruvate carboxylase, atp citrate lyase, and pyruvate

carboxylation and high glucose-stimulated acetoacetate in human pancreatic islets. Journal of

Biological Chemistry, 286(21):18383–18396, 2011.

W. M. Macfarlane, R. M. Shepherd, K. E. Cosgrove, R. F. James, M. J. Dunne, and K. Docherty.

Glucose modulation of insulin mRNA levels is dependent on transcription factor PDX-1 and occurs

independently of changes in intracellular Ca2+. Diabetes, 49(3):418–423, 2000.

R. Maehr. iPS Cells in Type 1 Diabetes Research and Treatment. Clinical Pharmacology & Thera-

peutics, 89(5):750–753, 2011.

M. Maekawa, K. Yamaguchi, T. Nakamura, R. Shibukawa, I. Kodanaka, T. Ichisaka, Y. Kawamura,

H. Mochizuki, N. Goshima, and S. Yamanaka. Direct reprogramming of somatic cells is promoted

by maternal transcription factor glis1. 474(7350):225–229, 2011.

N. M. Maruthur, E. Tseng, S. Hutﬂess, L. M. Wilson, C. Suarez-Cuervo, Z. Berger, Y. Chu, E. Iyoha,

J. B. Segal, and S. Bolen. Diabetes Medications as Monotherapy or Metformin-Based Combination

Therapy for Type 2 Diabetes: A Systematic Review and Meta-analysisDiabetes Medications as

55 Monotherapy or Metformin-Based Combination Therapy. Annals of Internal Medicine, Epub

ahead of print, 2016.

F. M. Matschinsky. Regulation of Pancreatic β-Cell Glucokinase: From Basics to Therapeutics.

Diabetes, 51(suppl 3):S394–S404, 2002.

T. Meissner, U. Wendel, P. Burgard, S. Schaetzle, and E. Mayatepek. Long-term follow-up of 114

patients with congenital hyperinsulinism. European Journal of Endocrinology, 149(1):43–51, 2003.

D. Melloul, S. Marshak, and E. Cerasi. Regulation of insulin gene transcription. Diabetologia, 45(3):

309–326, 2014.

J. Millman, X. Chunhui, A. Van Devort, M. Gurtler, F. Pagliuca, and D. Melton. Generation of stem

cell-derived β-cells from patients with type 1 diabetes. Nature Communications, 7:11463, 2016.

K. Moens, V. Berger, J.-M. Ahn, C. Van Schravendijk, V. J. Hruby, D. Pipeleers, and F. Schuit.

Assessment of the Role of Interstitial Glucagon in the Acute Glucose Secretory Responsiveness of

In Situ Pancreatic β-Cells. Diabetes, 51(3):669–675, 2002.

A. H. Mokdad, E. S. Ford, B. Bowman, W. H. Dietz, F. Vinicor, V. S. Bales, and J. S. Marks.

Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. JAMA, 289(1):

76–79, 2003.

N. Moran. Banking iPS cells. Nat Biotech, 31(1):11–11, 2013.

L. C. Murtaugh, B. Z. Stanger, K. M. Kwan, and D. A. Melton. Notch signaling controls multiple

steps of pancreatic diﬀerentiation. Proceedings of the National Academy of Sciences, 100(25):

14920–14925, 2003.

National Institute for Health and Care Excellence. Type 2 diabetes in adults: management (NICE

guidelines NG28), December 2015. URL https://www.nice.org.uk/guidance/ng28.

F. J. Naya, C. M. Stellrecht, and M. J. Tsai. Tissue-speciﬁc regulation of the insulin gene by a novel

basic helix-loop-helix transcription factor. Genes & Development, 9(8):1009–1019, 1995.

W. Nishimura, T. Kondo, T. Salameh, I. El Khattabi, R. Dodge, S. Bonner-Weir, and A. Sharma.

A switch from MafB to MafA expression accompanies diﬀerentiation to pancreatic β-cells. Devel-

opmental Biology, 293(2):526–539, 2006.

U. E. Ntuk, J. M. Gill, D. F. Mackay, N. Sattar, and J. P. Pell. Ethnic-Speciﬁc Obesity Cutoﬀs for

Diabetes Risk: Cross-sectional Study of 490,288 UK Biobank Participants. Diabetes Care, 37(9):

2500–2507, 2014.

56 K. Okita, T. Ichisaka, and S. Yamanaka. Generation of germline-competent induced pluripotent

stem cells. Nature, 448(7151):313–317, 2007.

T. Otonkoski, K. Nanto-Salonen, M. Seppnen, R. Veijola, H. Huopio, K. Hussain, P. Tapanainen,

O. Eskola, R. Parkkola, K. Ekstrom, Y. Guiot, J. Rahier, M. Laakso, R. Rintala, P. Nuutila, and

H. Minn. Noninvasive Diagnosis of Focal Hyperinsulinism of Infancy With [18F]-DOPA Positron

Emission Tomography. Diabetes, 55(1):13–18, 2006.

F. Pagliuca, J. Millman, M. Gurtler, M. Segel, A. VanDervort, J. Ryu, Q. Peterson, D. Greiner, and

D. Melton. Generation of Functional Human Pancreatic β Cells In Vitro. Cell, 159(2):428–439,

2014.

F. C. Pan and C. Wright. Pancreas organogenesis: From bud to plexus to gland. Dev. Dyn., 240(3):

530–565, 2011.

K. Peters, R. Panienka, J. Li, G. Kloppel, and R. Wang. Expression of stem cell markers and

transcription factors during the remodeling of the rat pancreas after duct ligation. Virchows

Archiv, 446(1):56–63, 2004.

H. V. Petersen, P. Serup, J. Leonard, B. K. Michelsen, and O. D. Madsen. Transcriptional Regula-

tion of the Human Insulin Gene is Dependent on the Homeodomain Protein STF1/IPF1 Acting

Through the CT Boxes. Proceedings of the National Academy of Sciences of the United States of

America, 91(22):10465–10469, 1994.

P. Petit and M. Loubatieres-Mariani. Potassium channels of the insulin-secreting B cell. Fundamental

& Clinical Pharmacology, 6(3):123–134, 1992.

D. G. Pipeleers. Heterogeneity in Pancreatic β-cell Population. Diabetes, 41(7):777–781, 1992.

K. Piper, S. Ball, L. Turnpenny, S. Brickwood, D. Wilson, and N. Hanley. Beta-cell diﬀerentiation

during human development does not rely on nestin-positive precursors: implications for stem cell-

derived replacement therapy. Diabetologia, 45(7):1045–7, 2002.

K. Piper, S. Brickwood, L. Turnpenny, I. Cameron, S. Ball, D. Wilson, and N. Hanley. Beta cell

diﬀerentiation during early human pancreas development. Journal of Endocrinology, 181(1):11–23,

2004.

G. L. Pittenger, D. Taylor-Fishwick, and A. I. Vinik. The Role of Islet Neogeneis-Associated Protein

(INGAP) in Pancreatic Islet Neogenesis. Current Protein & Peptide Science, 10(1):37–45, 2009.

P. Proks, F. Reimann, N. Green, F. Gribble, and F. Ashcroft. Sulfonylurea Stimulation of Insulin

Secretion. Diabetes, 51(suppl 3):S368–S376, 2002.

57 S. Puri and M. Hebrok. Cellular Plasticity within the Pancreas - Lessons Learned from Development.

Developmental Cell, 18(3):342–356, 2010.

N. Quiskamp, J. E. Bruin, and T. J. Kieﬀer. Diﬀerentiation of human pluripotent stem cells into

β-cells: Potential and challenges. Best Practice & Research Clinical Endocrinology & Metabolism,

29(6):833–847, 2015.

M. M. Rankin, C. J. Wilbur, K. Rak, E. J. Shields, A. Granger, and J. A. Kushner. β-Cells Are Not

Generated in Pancreatic Duct Ligation-Induced Injury in Adult Mice. Diabetes, 62(5):1634–1645,

2013.

J. M. Redondo, L. Yu, M. Hawa, T. Mackenzie, A. D. Pyke, S. G. Eisenbarth, and G. R. D. Leslie.

Heterogeneity of Type I diabetes: analysis of monozygotic twins in Great Britain and the United

States. Diabetologia, 44(3):354–362, 2001.

D. A. Rees and J. C. Alcolado. Animal models of diabetes mellitus. Diabetic Medicine, 22(4):359–370,

2005.

A. Rezania, J. E. Bruin, P. Arora, A. Rubin, I. Batushansky, A. Asadi, S. O’Dwyer, N. Quiskamp,

M. Mojibian, T. Albrecht, Y. H. C. Yang, J. D. Johnson, and T. J. Kieﬀer. Reversal of diabetes

with insulin-producing cells derived in vitro from human pluripotent stem cells. Nat Biotech, 32

(11):1121–1133, 2014.

D. A. Robinton and G. Q. Daley. The promise of induced pluripotent stem cells in research and

therapy. 481(7381):295–305, 2012.

D. J. Rodda, J.-L. Chew, L.-H. Lim, Y.-H. Loh, B. Wang, H.-H. Ng, and P. Robson. Transcriptional

regulation of nanog by oct4 and sox2. Journal of Biological Chemistry, 280(26):24731–24737, 2005.

I. Rooman, J. Lardon, and L. Bouwens. Gastrin Stimulates β-Cell Neogenesis and Increases Islet

Mass From Transdiﬀerentiated but Not From Normal Exocrine Pancreas Tissue. Diabetes, 51(3):

686–690, 2002.

F. Rouhani, N. Kumasaka, M. C. de Brito, A. Bradley, L. Vallier, and D. Gaﬀney. Genetic Back-

ground Drives Transcriptional Variation in Human Induced Pluripotent Stem Cells. PLoS Genet,

10(6):e1004432–, 2014.

J. M. Rukstalis and J. F. Habener. Snail2, a mediator of epithelial-mesenchymal transitions, ex-

pressed in progenitor cells of the developing endocrine pancreas. Gene Expression Patterns, 7(4):

471–479, 2007.

58 H. A. Russ, Y. Bar, P. Ravassard, and S. Efrat. In Vitro Proliferation of Cells Derived From Adult

Human β-Cells Revealed By Cell-Lineage Tracing. Diabetes, 57(6):1575–1583, 2008.

H. A. Russ, P. Ravassard, J. Kerr-Conte, F. Pattou, and S. Efrat. Epithelial-Mesenchymal Transition

in Cells Expanded In Vitro from Lineage-Traced Adult Human Pancreatic Beta Cells. PLoS ONE,

4(7):e6417, 2009.

W. Russell and R. Burch. The Principles of Humane Experimental Technique. Methuen, 1959.

E. A. Ryan, B. W. Paty, P. A. Senior, D. Bigam, E. Alfadhli, N. M. Kneteman, J. R. Lakey, and

A. J. Shapiro. Five-Year Follow-Up After Clinical Islet Transplantation. Diabetes, 54(7):2060–2069,

2005.

R. J. Salisbury, J. Blaylock, A. A. Berry, R. E. Jennings, R. De Krijger, K. Piper Hanley, and N. A.

Hanley. The window period of NEUROGENIN3 during human gestation. Islets, 6(3):e954436–,

2014.

R. J. Salisbury, B. Han, R. E. Jennings, A. A. Berry, A. Stevens, Z. Mohamed, S. A. Sugden,

R. De Krijger, S. E. Cross, P. P. Johnson, M. Newbould, K. E. Cosgrove, K. P. Hanley, I. Banerjee,

M. J. Dunne, and N. A. Hanley. Altered Phenotype of β-Cells and Other Pancreatic Cell Lineages

in Patients With Diﬀuse Congenital Hyperinsulinism in Infancy Caused by Mutations in the ATP-

Sensitive K-Channel. Diabetes, 64(9):3182–3188, 2015.

S. E. Samaras, M. A. Cissell, K. Gerrish, C. V. E. Wright, M. Gannon, and R. Stein. Conserved

Sequences in a Tissue-Speciﬁc Regulatory Region of the pdx-1 Gene Mediate Transcription in

Pancreatic β Cells: Role for Hepatocyte Nuclear Factor 3β and Pax6. Molecular and Cellular

Biology, 22(13):4702–4713, 2002.

S. E. Samaras, L. Zhao, A. Means, E. Henderson, T.-a. Matsuoka, and R. Stein. The Islet β Cell-

enriched RIPE3b1/Maf Transcription Factor Regulates pdx-1 Expression. Journal of Biological

Chemistry, 278(14):12263–12270, 2003.

M. Sander, A. Neubuser, J. Kalamaras, H. C. Ee, G. R. Martin, and M. S. German. Genetic analysis

reveals that PAX6 is required for normal transcription of pancreatic hormone genes and islet

development. Genes & Development, 11(13):1662–1673, 1997.

V. Schwitzgebel, D. Scheel, J. Conners, J. Kalamaras, J. Lee, D. Anderson, L. Sussel, J. Johnson, and

M. German. Expression of neurogenin3 reveals an islet cell precursor population in the pancreas.

Development, 127(16):3533–3542, 2000.

59 K. L. Seeberger, J. M. Dufour, A. M. J. Shapiro, J. R. T. Lakey, R. V. Rajotte, and G. S. Korbutt.

Expansion of mesenchymal stem cells from human pancreatic ductal epithelium. Lab Invest, 86

(2):141–153, 2006.

K. L. Seeberger, A. Eshpeter, R. V. Rajotte, and G. S. Korbutt. Epithelial cells within the human

pancreas do not coexpress mesenchymal antigens: epithelial-mesenchymal transition is an artifact

of cell culture. Lab Invest, 89(2):110–121, 2008.

T. Seki, S. Yuasa, M. Oda, T. Egashira, K. Yae, D. Kusumoto, H. Nakata, S. Tohyama, H. Hashimoto,

M. Kodaira, Y. Okada, H. Seimiya, N. Fusaki, M. Hasegawa, and K. Fukuda. Generation of induced

pluripotent stem cells from human terminally diﬀerentiated circulating t cells. 7(1):11–14, 2010.

P. A. Seymour. Sox9: A Master Regulator of the Pancreatic Program. The Review of Diabetic

Studies : RDS, 11(1):51–83, 2013.

P. A. Seymour, K. K. Freude, M. N. Tran, E. E. Mayes, J. Jensen, R. Kist, G. Scherer, and M. Sander.

SOX9 is required for maintenance of the pancreatic progenitor cell pool. Proceedings of the National

Academy of Sciences, 104(6):1865–1870, 2007.

A. J. Shapiro, J. R. Lakey, E. A. Ryan, G. S. Korbutt, E. Toth, G. L. Warnock, N. M. Kneteman,

and R. V. Rajotte. Islet Transplantation in Seven Patients with Type 1 Diabetes Mellitus Using

a Glucocorticoid-Free Immunosuppressive Regimen. New England Journal of Medicine, 343(4):

230–238, 2000.

A. A. Sharov, S. Masui, L. V. Sharova, Y. Piao, K. Aiba, R. Matoba, L. Xin, H. Niwa, and M. S. H.

Ko. Identiﬁcation of pou5f1, sox2, and nanog downstream target genes with statistical conﬁdence

by applying a novel algorithm to time course microarray and genome-wide chromatin immunopre-

cipitation data. BMC Genomics, 9(1):1–19, 2008.

K. Shimomura, M. Tusa, M. Iberl, M. F. Brereton, S. Kaizik, P. Proks, C. Lahmann, N. Yaluri,

S. Modi, H. Huopio, J. Ustinov, T. Otonkoski, M. Laakso, and F. M. Ashcroft. A Mouse Model of

Human Hyperinsulinism Produced by the E1506K Mutation in the Sulphonylurea Receptor SUR1.

Diabetes, 62(11):3797–3806, 2013.

J. Silva, J. Nichols, T. W. Theunissen, G. Guo, A. L. van Oosten, O. Barrandon, J. Wray, S. Ya-

manaka, I. Chambers, and A. Smith. Nanog is the gateway to the pluripotent ground state. 138

(4):722–737, 2009.

M. Skelin, M. Rupnik, and A. Cencic. Pancreatic beta cell lines and their applications in diabetes

mellitus research. ALTEX, 27(2):105–13, 2010.

60 V. Sordi, R. Melzi, A. Mercalli, R. Formicola, C. Doglioni, F. Tiboni, G. Ferrari, R. Nano,

K. Chwalek, E. Lammert, E. Bonifacio, and L. Piemonti. Mesenchymal Cells Appearing in Pan-

creatic Tissue Culture Are Bone Marrow-Derived Stem Cells With the Capacity to Improve Trans-

planted Islet Function. Stem Cells, 28(1):140–151, 2010.

R. L. Sorenson and T. C. Brelje. Adaptation of Islets of Langerhans to Pregnancy: β-Cell Growth,

Enhanced Insulin Secretion and the Role of Lactogenic Hormones. Horm Metab Res, 29(06):

301–307, 1997.

A. Stancakova, M. Javorsky, T. Kuulasmaa, S. M. Haﬀner, J. Kuusisto, and M. Laakso. Changes

in Insulin Sensitivity and Insulin Release in Relation to Glycemia and Glucose Tolerance in 6,414

Finnish Men. Diabetes, 58(5):1212–1221, 2009.

G. R. Stark and W. R. Taylor. Analyzing the G2/M Checkpoint. In Checkpoint Controls and Cancer,

volume 280 of Methods in Molecular Biology, pages 51–82. 2004.

J. C. Stendahl, D. B. Kaufman, and S. I. Stupp. Extracellular Matrix in Pancreatic Islets: Relevance

to Scaﬀold Design and Transplantation. Cell Transplantation, 18(1):1–12, 2009.

C. Stocking, U. Bergholz, J. Friel, K. Klingler, T. Wagener, C. Starke, T. Kitamura, A. Miyajima,

and W. Ostertag. Distinct classes of factor-independent mutants can be isolated after retroviral

mutagenesis of a human myeloid stem cell line. 8(3):197–209, 1993.

C. Street, Lakey, K. Seeberger, L. Helms, R. Rajotte, A. Shapiro, and G. Korbutt. Heterogenous

expression of nestin in human pancreatic tissue precludes its use as an islet precursor marker.

Journal of Endocrinology, 180(2):213–225, 2004.

M. Szabat, D. S. Luciani, J. M. Piret, and J. D. Johnson. Maturation of Adult β-Cells Revealed

Using a Pdx1/Insulin Dual-Reporter Lentivirus. Endocrinology, 150(4):1627–1635, 2009.

M. Szabat, F. C. Lynn, B. G. Hoﬀman, T. J. Kieﬀer, D. W. Allan, and J. D. Johnson. Maintenance of

β-Cell Maturity and Plasticity in the Adult Pancreas: Developmental Biology Concepts in Adult

Physiology. Diabetes, 61(6):1365–1371, 2012.

K. Takahashi, K. Tanabe, M. Ohnuki, M. Narita, T. Ichisaka, K. Tomoda, and S. Yamanaka. Induc-

tion of Pluripotent Stem Cells from Adult Human Fibroblasts by Deﬁned Factors. Cell, 131(5):

861–872, 2007.

J. A. Thomson, J. Itskovitz-Eldor, S. S. Shapiro, M. A. Waknitz, J. J. Swiergiel, V. S. Marshall, and

J. M. Jones. Embryonic Stem Cell Lines Derived from Human Blastocysts. Science, 282(5391):

1145–1147, 1998.

61 F. Thorel, V. Nepote, I. Avril, K. Kohno, R. Desgraz, S. Chera, and P. L. Herrera. Conversion of

adult pancreatic α-cells to β-cells after extreme β-cell loss. Nature, 464(7292):1149–1154, 2010.

G. Touati, F. Poggi-Travert, H. Ogier de Baulny, J. Rahier, F. Brunelle, C. Nihoul-Fekete, P. Czerni-

chow, and J. M. Saudubray. Long-term treatment of persistent hyperinsulinaemic hypoglycaemia

of infancy with diazoxide: a retrospective review of 77 cases and analysis of eﬃcacy-predicting

criteria. Eur J Pediatr., 157(8):628–33, 1998.

N. Tsubooka, T. Ichisaka, K. Okita, K. Takahashi, M. Nakagawa, and S. Yamanaka. Roles of sall4

in the generation of pluripotent stem cells from blastocysts and ﬁbroblasts. Genes to Cells, 14(6):

683–694, 2009.

T. van der Meulen, R. Xie, O. G. Kelly, W. W. Vale, M. Sander, and M. O. Huising. Urocortin

3 Marks Mature Human Primary and Embryonic Stem Cell-Derived Pancreatic Alpha and Beta

Cells. PLoS ONE, 7(12):e52181–, 2012.

B. Viollet, B. Guigas, N. Sanz Garcia, J. Leclerc, M. Foretz, and F. Andreelli. Cellular and molecular

mechanisms of metformin: an overview. Clinical Science (London, England : 1979), 122(6):253–

270, 2012.

P. Wang, N. M. Fiaschi-Taesch, R. C. Vasavada, D. K. Scott, A. Garcia-Ocana, and A. F. Stewart.

Diabetes mellitus - advances and challenges in human β-cell proliferation. Nat Rev Endocrinol, 11

(4):201–212, 2015.

Q. Wang, L. Ye, H. Liu, X. Liu, S. Li, and Z. Chen. Reprogramming of bone marrow-derived

mesenchymal stem cells into functional insulin-producing cells by chemical regimen. American

Journal of Stem Cells, 1(2):128–137, 2012.

L. Warren, P. D. Manos, T. Ahfeldt, Y.-H. Loh, H. Li, F. Lau, W. Ebina, P. K. Mandal, Z. D. Smith,

A. Meissner, G. Q. Daley, A. S. Brack, J. J. Collins, C. Cowan, T. M. Schlaeger, and D. J. Rossi.

Highly eﬃcient reprogramming to pluripotency and directed diﬀerentiation of human cells with

synthetic modiﬁed mrna. 7(5):618–630, 2010.

J. M. Wells and D. A. Melton. Vertebrate Endoderm Development. Annu. Rev. Cell Dev. Biol., 15

(1):393–410, 1999.

P. O. Wiebe, J. D. Kormish, V. T. Roper, Y. Fujitani, N. I. Alston, K. S. Zaret, C. V. E. Wright,

R. W. Stein, and M. Gannon. Ptf1a Binds to and Activates Area III, a Highly Conserved Region of

the Pdx1 Promoter That Mediates Early Pancreas-Wide Pdx1 Expression. Molecular and Cellular

Biology, 27(11):4093–4104, 2007.

62 M. E. Wilson, D. Scheel, and M. S. German. Gene expression cascades in pancreatic development.

Mechanisms of Development, 120(1):65–80, 2003.

S. Yamanaka. Induced Pluripotent Stem Cells: Past, Present, and Future. Cell Stem Cell, 10(6):

678–684, 2012.

P. Yi, J.-S. Park, and D. Melton. Betatrophin: A Hormone that Controls Pancreatic β Cell Prolif-

eration. Cell, 153(4):747–758, 2013.

J. Yu, M. A. Vodyanik, K. Smuga-Otto, J. Antosiewicz-Bourget, J. L. Frane, S. Tian, J. Nie, G. A.

Jonsdottir, V. Ruotti, R. Stewart, I. I. Slukvin, and J. A. Thomson. Induced pluripotent stem cell

lines derived from human somatic cells. Science, 318(5858):1917–1920, 2007.

C. Zanini, S. Bruno, G. Mandili, D. Baci, F. Cerutti, G. Cenacchi, L. Izzi, G. Camussi, and M. Forni.

Diﬀerentiation of Mesenchymal Stem Cells Derived from Pancreatic Islets and Bone Marrow into

Islet-Like Cell Phenotype. PLoS ONE, 6(12):e28175, 2011.

H. Zhang, W.-P. Wang, T. Guo, J.-C. Yang, P. Chen, K.-T. Ma, Y.-F. Guan, and C.-Y. Zhou. The

LIM-Homeodomain Protein ISL1 Activates Insulin Gene Promoter Directly through Synergy with

BETA2. Journal of Molecular Biology, 392(3):566–577, 2009.

H. Zhang, S. Gayen, J. Xiong, B. Zhou, A. Shanmugam, Y. Sun, H. Karatas, L. Liu, R. Rao, S. Wang,

A. Nesvizhskii, S. Kalantry, and Y. Dou. MLL1 Inhibition Reprograms Epiblast Stem Cells to

Naive Pluripotency. Cell Stem Cell, page ePub, 2016.

63 Chapter 2

A Novel Source of Mesenchymal

Stem Cells derived from the

Pancreas of Patients with

Congenital Hyperinsulinism.

Sophie G Kellaway, Karolina Mosinska, Zainaba Mohamed,

Indi Banerjee, Alexander Ryan, Karen E Cosgrove and Mark J Dunne

64 2.1 Abstract

Diabetes and congenital hyperinsulinism (CHI) are severe diseases affecting the pancreas. Current models for testing drugs to treat these diseases are in vivo in rodents or isolated rodent islets. Differ- ences between the human and rodent pancreas mean in vitro human models are needed. Developing islets from mesenchymal stem cells (MSCs) has shown promise but the final cells produced are not mature and MSCs show limited replicative capacity. In this study we have derived MSCs from the pancreas of patients with CHI. The pancreas of neonatal patients with CHI have greater proliferation, plasticity and insulin secretion than from healthy counterparts. MSCs from these patients have the typical cell surface markers CD29, CD44, CD73, CD90 and CD105, mesenchymal cytoskeletal proteins Vimentin, Nestin and α-smooth muscle actin, and differentiation capacity to osteoblasts, chondrocytes and adipocytes. The MSCs also expressed a number of transcription factors associated with pancreatic development, including ISLET1, a unique marker of pancreas-derived MSCs.

Pancreatic β-cell related genes that were not expressed, including PDX1 showed decreased promoter methylation compared to MSCs derived from bone marrow suggesting a possible ‘epigenetic memory’. We have demonstrated here the feasibility of deriving pancreatic MSCs from patients with CHI which maintain some features of the pancreas. Re-diﬀerentiation of these cells to β-cells may provide a valuable new resource to study CHI in vitro.

65 2.2 Introduction

Diabetes is a chronic disease affecting nearly 10% of adults worldwide (Danaei et al., 2011) and cannot currently be cured. Cadaveric islet transplantation offers a long term treatment or potential cure by increasing β-cell mass (Ryan et al., 2005, Shapiro et al., 2000) but as a treatment is limited by lack of donors. Consequently, alternative sources of islets for cell therapies have been investigated, including derivation from human embryonic stem cells (hESCs) and induced pluripotent stem cells (iPSCs). HESCs and iPSCs do not typically have the epigenetic profile associated with mature

β-cells when diﬀerentiated, unless the iPSCs are derived from pancreatic tissue originally (Bar-Nur et al., 2011) and so we have considered stem cells from the pancreas.

Multiple putative stem cells have been isolated from the adult pancreas and diﬀerentiated to

β-like or islet-like cells; initially from rodents (Zulewski et al., 2001) and later humans. This includes proliferative ductal structures (Corritore et al., 2014), de-diﬀerentiated endocrine (Russ et al., 2009,

Joglekar et al., 2009) and exocrine tissue (Lima et al., 2013) which resemble mesenchymal stem cells (MSCs) and de-diﬀerentiated endocrine tissue which remains epithelial (Lima et al., 2013).

Additionally, MSCs from bone marrow and adipose tissue have also been diﬀerentiated to insulin secreting cells discussed by Bhonde et al. (2014). It can be hypothesised that if these cells can be re-diﬀerentiated in vitro then it can also take place in vivo - a process known as β-cell neogenesis.

Whilst hESCs and iPSCs may oﬀer an exogenous source of islet cells for transplantation, somatic tissue stem cells can potentially provide a model to investigate neogenesis for allogeneic islet cell generation. However, pancreatic MSCs have proven challenging as a model due to slow growth, early senescence at around 30 population doublings (Hu et al., 2003, Carlotti et al., 2010) and limited insulin release by re-diﬀerentiated cells (Bhonde et al., 2014).

The exact origins of MSCs from the pancreas is unclear, whether they are tissue resident such as CD90/CD105 double positive cells (Carlotti et al., 2010) or de-diﬀerentiated cells. Numerous experiments have shown that they can de-diﬀerentiate from various cell types. When placed into culture, endocrine and non-endocrine cells produce similar MSCs, with lineage tracing being required to reveal the origins of MSCs from mixed pancreas digests (Russ et al., 2009). Pancreatic

MSCs can be readily derived from exocrine pancreas, demonstrated with lineage tracing (Lima et al.,

2013). Exocrine cells are a useful source when only a small portion of tissue is available, such as from neonates, as the exocrine pancreas comprises over 95% of the pancreas, or when the islets are required for other in vitro studies. It was shown by Lima et al. (2016) that exocrine-derived MSCs can be readily diﬀerentiated to functional insulin secreting cells with a transcription factor regimen, suggesting plasticity or transdiﬀerentiation from these cells is possible. Even after extended in vitro

66 culture these exocrine-derived MSCs were able to undergo mesenchymal-epithelial transition, high- lighting the ﬂexibility of this in vitro cell phenomenon (Muir et al., 2015).

We hypothesised that MSCs could be derived from the exocrine pancreas of patients with congenital hyperinsulinism (CHI). If these MSCs maintain the phenotype of CHI, this would be useful for modelling both CHI and other pancreatic disorders. There are three forms of CHI - focal, diﬀuse and atypical (also known as localised islet cell nuclear enlargement, LINE). Focal CHI is caused by a recessive mutation in the ABCC8 gene (which encodes the SUR1 subunit of the pancreatic

ATP-sensitive potassium (KATP) channel), where loss of heterozygosity leads to no functional allele and non-functional channels (Verkarre et al., 1998, Glaser et al., 1999, Fournet et al., 2001). Diﬀuse

CHI can be caused by a number of mutations (Glaser et al., 2000), including ABCC8 (Dunne et al.,

1997), but is homozygous and so affects the whole pancreas. Atypical CHI typically has a later onset than focal or diffuse CHI, is not caused by any known mutation (screening of the genes associated with focal and diffuse CHI excludes these, but some post-zygotic mutations have now been reported at meetings) and affects the whole pancreas (Arnoux et al., 2011). All three forms of CHI exhibit high levels of insulin secretion (Dunne et al., 2004) and hyperproliferation (Salisbury et al., 2015,

James et al., 2009). Increased proliferation would mean the cells are easier to work with than other

MSC types, high levels of insulin secretion would be beneﬁcial for high-throughput screening of novel putative drug treatments for diabetes or CHI and as the cells are pancreatic they may maintain a pancreatic epigenetic proﬁle. To explore these hypotheses, in this study we aimed to derive and characterise MSCs from the pancreas of patients with CHI.

67 2.3 Methods

All reagants were from Sigma-Aldrich or Fisher-Scientiﬁc.

2.3.1 Derivation of CHI pancreatic mesenchymal stem cell (CHI pMSC) lines

Patient tissue was received from surgeons with ethical permission and full consent from the patients’ parents or legal guardians. Tissue was rinsed in Krebs-Ringer-Hepes buﬀer (KRH buﬀer comprised:

129 mM sodium chloride, 5 mM sodium bicarbonate, 4.8 mM potassium chloride, 1.2 mM potassium orthophosphate, 1.2 mM magnesium sulphate, 10 mM HEPES, 2.5 mM calcium chloride, 5.6 mM

D-glucose, 0.1% bovine serum albumin (BSA) and the pH was adjusted to 7.4) to remove excess blood. The tissue was ﬁrst digested with 0.75 mg/ml Liberase (Roche, Mannheim, Germany) in

KRH buﬀer with 1.5 mg/ml egg white trypsin inhibitor (Sigma-Aldrich, Irvine, UK) and 1.5 mg/ml soybean trypsin inhibitor (Gibco, Paisley, UK), for 3 minutes at 37°C with agitation, followed by

1 minute of vigorous shaking, and this was done three times. Ice cold KRH buffer was added and the digestion mixture centrifuged for 1 minute at 150 x g. The secondary digestion was with 0.5 mg/ml Liberase in KRH buffer with trypsin inhibitors, for 3 minutes at 37°C with agitation, followed by 1 minute of vigorous shaking and this was done twice. Ice cold KRH buffer was added and the digestion mixture centrifuged for 1 minute at 150 x g, this was done 4 times to ensure the preparation was as clean as possible. Where possible islets were removed from the mixture under a dissection microscope but due to the age and pathology of patients, islets were not clearly delineated. The resulting digestion mixture was transferred to 6-well tissue culture treated plate in RPMI 1640 5.5 mM glucose (0 mM glucose:RPMI 1640 11 mM glucose; 50:50; Gibco/Sigma-Aldrich) with 10% foetal bovine serum (FBS, Gibco), around 1 well per 20 mg wet weight tissue received. After 48 hours, cells and tissue which had not adhered were removed and media was changed thereon every 48 hours until 70% confluency was reached at which point the cells underwent the first passage.

2.3.2 Cell culture

Cells were routinely cultured on tissue culture plastic (plates or vented cap ﬂasks, Corning, Schiphol-

Rijk, Netherlands) coated with human ﬁbronectin (Chemicon, Merck Millipore, Nottingham, UK) to 0.5 µg/cm2 and vitronectin (Life Technologies, Paisley, UK) to 0.1 µg/cm2, the reasoning behind these growth conditions is shown and explained in Supplementary Information, Figure 2.8. Cells were cultured in RPMI-1640 5.5 mM glucose (mixed from 11 mM glucose RPMI, Sigma-Aldrich and

0 mM glucose RPMI, Gibco, Life Technologies to 5 mM) with 10% FBS. Cultured cells were kept in

68 a humidified incubator at 37°C, 5% CO2. Subculturing/passage was performed when cells reached approximately 70% confluency - cells were washed with Dulbecco’s phosphate buffered saline (dPBS), and detached from tissue culture plastic with 0.05% trypsin-EDTA (Sigma-Aldrich) for 3 minutes.

Trypsinisation was halted using 1 mg/ml soybean trypsin inhibitor (Gibco). Cells were pelleted at

50 x g and transferred to new vessels at a cell line and condition speciﬁc density, between 1:2 and

1:5. Medium was changed every 48 hours. Brightﬁeld images were taken on an Olympus CKX41 microscope using 4x and 10x objectives and captured with QCapture Pro software. All experiments were performed on cells between passages 3 and 16. Total number of population doublings was estimated from number of passages and split ratios.

2.3.3 Karyotyping

Three days after seeding the cells, culture medium was changed and 20 ng/ml epidermal growth factor (Calbiochem, Merck Millipore), 20 ng/ml ﬁbroblast growth factor (Peprotech, Rocky Hill,

USA) and 100 ng/ml Colcemid (Karyomax, Gibco) was added. At least 12 hours later cells were trypsinised as above and put into hypotonic solution, comprising 0.0375M potassium chloride (Fisher

Scientific, Loughborough, UK), for 5 minutes at 37°C. The cells were pelleted at 180 x g, then fixed using ice cold 3:1 methanol/acetic acid solution, added in a drop-wise fashion. The cells were pelleted again and more fixative added, this process was repeated one more time and the resulting solution frozen. G-banding and analysis was then performed by TDL genetics (The Doctor’s Laboratory,

London).

2.3.4 Short tandem repeat (STR) proﬁling

Cells were lysed and DNA ampliﬁed using a PowerPlex 16 HS system (Promega, Madison, USA), the samples were analysed on a 3730 DNA analyser (Applied Biosystems, California, USA) using

POP-7 polymer (Life Technologies). The number of repeats at each locus was then called using the

GeneMapper 4.0 software (Life Technologies). The PowerPlex 16 HS system is a multiplex PCR based assay which ampliﬁes the 13 combined DNA index system loci (a set of nucleotides which have high copy-number variation, used in forensic applications for identifying individuals), two control low-stutter pentanucleotide markers and amelogenin for sex determination.

2.3.5 RT-PCR and qRT-PCR

Total RNA was extracted using an RNeasy Mini Kit (Qiagen, Crawley, UK) with on column DNase

1 (Qiagen) genomic DNA digestion according to the manufacturer’s instructions and quantiﬁed using a NanoDrop 1000 spectrophotometer. cDNA was synthesised from 500 ng RNA using a nanoScript

2 reverse transcription kit (Primer Design, Chandler’s Ford, UK) with oligo-dT primers and random

69 nonamer primers. RT-PCR was carried out using a Taq DNA Polymerase kit (Invitrogen, Life

Technologies), 1.5mM magnesium chloride and 2.5pmol of each primer with conditions as detailed in Table 2.1. PCR cycling conditions were 94°C for 5 minutes followed by 33 cycles (22 cycles for

18S rRNA, 38 cycles for INS) of 94°C for 30 seconds, annealing temperature as detailed in Table 2.1 for 30 seconds and 72°C for 90 seconds, with a ﬁnal 72°C extension of 2 minutes. RT-PCR products were run on a 2% agarose gel stained with GelRed (Biotium, California, USA) and visualised on a Gel Doc XR system (Bio-Rad, Hemel Hempstead, UK). PCR images had the colour inverted in

IrfanView, and where necessary brightness and contrast were adjusted, to ensure that fainter bands visible on the transilluminator would be visible when printed. qRT-PCR was carried out using power

SYBR green PCR master mix (Life Technologies) and 2.5 pmol of each primer as detailed in Table

2.1 and run on a StepOnePlus Real-Time PCR System (Life Technologies). For qRT-PCR, the gene of interest CT values were normalised to the geometric mean of 2 reference genes, 18S rRNA and

GAPDH (∆CT), following which the sample comparisons were done (∆∆CT). Statistical significance was calculated by one way ANOVA with Bonferroni post-hoc test if appropriate. Bone marrow mesenchymal stem cell (bmMSC) mRNA was gratefully received from Dr Stephen Richardson, all procedures were the same. Primers were designed using Primer-BLAST (NCBI) and synthesised by Eurofins (Wolverhampton, UK). Specificity of the primers was assessed by BLAST (NCBI) and fragment size, optimal annealing temperatures were determined by temperature gradient PCR to ensure only one fragment was present and the most intense band present. QRT-PCR primers had an efficiency between 85 and 115% confirmed by serial dilution of standards.

2.3.6 Genotyping

Genomic DNA was extracted from cells using a PureLink genomic DNA kit (Invitrogen). PCR was carried out as Section 2.3.5, for the aﬀected exon as determined in the patient by clinicians.

PCR products were purified using a QIAquick PCR purification kit (Qiagen). Products were then sequenced, using 15 ng of purified PCR product, 4 pmoles of each primer used previously (forward and reverse sequencing carried out separately) with the BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems) on a 3730 DNA analyser (Applied Biosystems) using POP-7 polymer (Life

Technologies).

Gene Sequence (5’ - 3’) Annealing tem- Product

perature (°C) size (base

pairs)

18S rRNA F - GTAACCCGTTGAACCCCATT 60 150

R - CCATCCAATCGGTAGTAGCG

70 Gene Sequence (5’ - 3’) Annealing tem- Product

perature (°C) size (base

pairs)

ABCC8 exon F - CCTATTTGCAGAGGATGTGAC 60 454

6* R - GATACTGCTATGGGCTTTGC

ABCC8 exon F - CCATGCACACATTTTCCAAC 60 325

37* R - ATCCCACTAAACCCTTTCCAA

ARX F - CTCCAACATCCACTCTCTCT 59 257

R - CTATCTTACAGGCTCGCAT

CDH1 F - CACCAGCCTCCTTTTTATTT 58 479

R - CTTTTCATAGTTCCGCTCTG

CK19 F - CGATGTGCGAGCTGATAGTG 58 258

R - TTGGCAGGTCAGGAGAAGA

COL2A1 F - CTGCATGAGGGCGCGGTA 62 384

R - CAGTGGCGAGGTCAGTTGG

FOXA2 F - CCGTTCTCCATCAACAACCT 59 325

R - GACCCCCACTTGCTCTCTC

GAPDH F - ATTGCCCTCAACGACCACTT 60 95

R - GGTCCACCACCCTGTTGC

GCG F - TTCTGAGGCCACATTGCTTT 59 289

R - CTGTGGCTACCAGTTCTTC-

TATTCT

INS F - ACCAGCATCTGCTCCCTCTA 60 114

R - GGTTCAAGGGCTTTATTCCA

ISL1 F - TATTTTGCCACAAGCGTCTC 62 282

R - TTCAAAGACCACCGTACAACC

ISL1 (qRT- F - GCAGCCCAATGACAAAACTAA 60 78

PCR) R - CCGTCGTGTCTCTCTGGACT

MAFB F - CGAGAGAGACGCCTACAAGG 59 237

R - CCCTCTCGCTCAAGTCAAAC

MYT1 F - AAGTGCGTTTTCTGTTTTA 58 419

R - TTTACCCTCTGACCTTGTG

NES F - AAGGTGAAAAGGGGTGTGG 62 95/103

R - ATGAGATGGAGCAGGCAAG

71 Gene Sequence (5’ - 3’) Annealing tem- Product

perature (°C) size (base

pairs)

NEUROD1 F - ACCAGCCCTTCCTTTGATGG 60 131

R - AGTGTCGCTGCAGGATAGTG

NGN3 F - CGAATGCACGACCTCAAC 59 125

R - AGTCAGCGCCCAGATGTAGT

OCLN F - CTCTGAAGTGAAACTGCTTG 55 136

R - AAAAATAGGGAGGCTGGTAG

PAX6 F - AAAAATCCCCAAAAGCAAAT 59 245

R - AGGCTGACACTGAACAAAAC

PDX1 F - AGTGATACTGGATTGGCGTTG 59 138

R - TAGGGAGCCTTCCAATGTGT

PPARG2 F - GCAAACCCCTATTCCATGCTG 59 355

R - GCAGGCTCCACTTTGATTGC

SST F - AGGATGAAATGAGGCTTGA 55 191

R - GGAGGATTAGGGAAGAGAGA

SOX9 F - AGCGAAATCAACGAGAAACT 60 222

R - ATCCCCTCAAAATGGTAATG

SOX17 F - CGCACGGAATTTGAACAGTAT 59 182

R - GGATCAGGGACCTGTCACAC

SPP1 F - TCCTAGCCCCACAGACCCTT 59 391

R - AACTCCTCGCTTTCCATGTGT

VIM F - GTGAAATGGAAGA- 58 274

GAACTTTGC R - GTATCAACCA-

GAGGGAGTGAAT

Table 2.1: Primers used for RT-PCR, details on design and use are in Section 2.3.5. Primers indicated with * from Ellard et al. (2007)

2.3.7 Adipogenesis, osteogenesis and chondrogenesis

For adipogenesis and osteogenesis, cells were seeded in duplicate at 1 x 104 cells/cm2 on uncoated plastic (six-well plate) and incubated for 4 days in standard growth medium (RPMI with 10% FBS).

On day 4, cells were switched to StemPro adipogenesis or osteogenesis complete media (Gibco) with one well being kept in standard growth media as a negative control. For chondrogenesis cells were seeded as a micromass culture - 8 x 104 cells in a 5 µl droplet in a six-well plate was allowed to

72 adhere for 2 hours, after which StemPro chondrogenesis complete medium (Gibco) or RPMI 1640 with 10% FBS was added. Medium was changed every three days. After 14 days for adipogenesis and chondrogenesis or 21 days for osteogenesis in these conditions, cells were harvested for RT-PCR as above or were were fixed for 30 minutes in 4% paraformaldehyde and washed in dPBS. To confirm adipogenesis, cells were stained for lipid droplets - a 5 minute incubation with 60% isopropanol was performed to reduce non-specific stains, followed by a 5 minute incubation in oil red O (0.3% w/v in

60% isopropanol; Sigma-Aldrich). This was washed oﬀ in distilled water and counterstained for 60 seconds with Mayer’s haematoxylin (Sigma-Aldrich), which was in turn washed oﬀ in distilled water.

To conﬁrm osteogenesis, cells were stained for calcium deposits with a 3 minute incubation in Alizarin

Red S solution (2% w/v in distilled water, pH 4.2; Alfa Aesar, Heysham, UK) which was washed oﬀ in distilled water. Chondrogenesis was conﬁrmed by staining for sulphated proteoglycans with

Alcian Blue solution (1% w/v in 0.1 N HCl; Alfa Aesar) for 30 minutes and rinsing in distilled water.

Lipid droplets or calcium deposits stained red and proteoglycans stained blue were observed on an

Olympus CKX41 microscope using 10x and 40x objectives and images captured using QCapture Pro.

2.3.8 Flow Cytometry

Cells were harvested by trypsinisation, spun and resuspended in dPBS. A cell count was performed, the cells were spun again and then resuspended to 1 x 106 cells/ml in PBS, 10% normal goat serum

(NGS, Sigma-Aldrich) and 0.1% sodium azide for a 15 minute incubation on ice to prevent non- specific binding. The cells were divided such that there were 1x105 cells per condition required and neat or diluted primary antibodies added to the final concentration listed in Table 2.2 for 30 minutes at room temperature. Cells were spun at 150 x g and antibody solution removed, 3 x 500 µl dPBS washes were performed, with spins to remove the solution from the cells. Cells were resuspended in 100 µl dPBS with 3% NGS and diluted secondary antibodies added to the final concentration in

Table 2.2 for 30 minutes at room temperature. Cells were spun at 150 x g and antibody solution removed, 3 x 500 µl dPBS washes were performed, with spins to remove the solution from the cells.

Cells were resuspended in 500 µl dPBS with 3% NGS and analysed on a Beckman Coulter Cyan

ADP with 635nm excitation. Gating was performed using Summit V4.3, based on forward scatter and side scatter of unlabelled control cells to exclude debris (low forward scatter), dead cells (low forward scatter, high side scatter) and doublets/cell clumps (high forward scatter). Markers were compared to the relevant isotype control.

2.3.9 Immunocytochemistry (ICC)

Cells were cultured on glass coverslips until of a suitable density and were ﬁxed in 4% paraformaldehyde or 4% paraformaldehyde/0.025% glutaraldehyde (for cytoskeleton proteins) for 20 minutes.

73 Cells were permeabilised in 0.3% Triton x-100 (Sigma-Aldrich) or in ice cold methanol (for cytoskeleton proteins) and non-speciﬁc staining was prevented by incubation in 10% NGS in PBS

(with 0.3M glycine for cytoskeleton proteins). Primary antibodies were applied as Table 2.2 for 1 hour in 0.1% Triton x-100, 3% NGS in PBS and washed oﬀ in PBS/0.1% Tween-20 (Sigma-Aldrich).

Fluorophore-conjugated secondary antibody, speciﬁc to the species the primary antibody was raised in, was applied as Table 2.2 in 0.1% triton x-100, 3% NGS in PBS for 1 hour. Cells were washed in

PBS/0.1% Tween-20, then distilled water and allowed to dry. Coverslips were mounted using Pro

Long Gold antifade reagent (Life Technologies) with DAPI. Slides were viewed using an Olympus

BX51 upright microscope using 20x 0.50 UPlanFLN, 40x 0.75 UPlanFLN or 60x 1.25 UPlanFLN

(Ph 3) objectives and captured using a Coolsnap EZ camera (Photometrics, USA) through MetaVue

Software (Molecular Devices, USA). Filter sets for DAPI (31000v2), FITC (41001) and Cy3 (41007a) were used to prevent bleed through of colour channels whilst maximising signal. Images were pro- cessed and analysed using ImageJ (http://rsb.info.nih.gov/ij).

Antigen Species Concentration Manufacturer (µg/ml) α-smooth mus- Rabbit 0.5 Abcam (Cambridge, UK) cle actin CD29 Mouse 10 Abcam CD44 Mouse 1 Abcam CD45 Rabbit 2 Abcam CD73 Mouse 10 Miltenyi (Surrey, UK) CD90 Mouse 1 Abcam CD105 Mouse 10 Miltenyi E-Cadherin Mouse 1.66 BD Biosciences (Oxford, UK) IgG Rabbit Variable AbD Serotech (Oxford, UK) IgG1 Mouse Variable Cell Signalling Technology (Leiden, The Netherlands) IgG2a Mouse Variable Fisher Scientiﬁc Islet1 Rabbit 5 Abcam MafB Mouse 10 R&D (Abingdon, UK) Nestin Mouse 5 Millipore Occludin Mouse 10 R&D Pax6 Rabbit 5 Abcam Sox9 Rabbit 5 Millipore Vimentin Mouse 5 Abcam Cy2 anti- Goat 5 Jackson/Abcam rabbit Cy3 anti- Goat 5 Abcam mouse Cy5 anti- Goat 2.5 Eurogentec mouse Cy5 anti- Goat 2.5 Abcam rabbit

Table 2.2: Antibodies used in the study

74 2.3.10 Bisulﬁte sequencing

Genomic DNA was extracted from cells using a PureLink genomic DNA kit (Invitrogen). 1 µg of genomic DNA was bisulﬁte converted using the EpiTect fast DNA bisulﬁte kit (Qiagen) according to the manufacturer’s instructions. PCR was carried out as Section 2.3.5 with primers and temperatures detailed in Table 2.3. PCR primers were designed with MethPrimer (Li and Dahiya, 2002).

BmMSC DNA was gratefully received from Dr Stephen Richardson, it had been previously extracted with the QIAamp DNA Mini Kit but all subsequent procedures were identical and carried out at the same time. PCR products purified with a Qiaquick PCR purification kit (Qiagen) according to the manufacturer’s instructions. Products were then sequenced, using 15 ng of purified PCR product, 4 pmoles of each primer used previously (forward and reverse sequencing carried out separately) with the BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems) on a 3730 DNA analyser

(Applied Biosystems) using POP-7 polymer (Life Technologies). Percentage methylation was determined by measuring the peak heights of the cytosine or thymine for the forward trace, or adenine or guanine for the reverse trace on ChromasLite (Technelysium, Australia), with the proportion of cytosine or guanine (methylated) to total cytosine/guanine plus thymine/adenine being calculated. Data was visualised with Methylation plotter (Mallona et al., 2014) with statistical signiﬁcance calculated within this program by the Kruskal-Wallis test.

Gene Sequence (5’ - 3’) Annealing Genomic loca- Product tempera- tion size ture (°C) (base pairs) PDX1 F - TTTGGTTTTTTGTGGTT- 64 27918743 to 226 TAAGTTTT R - AACCACTTTTTC- 27918969 TACACCCACTAAC AMY2A F - GTTTTTGGAAA- 61 103616149 to 292 GAAGTTGTTTTTGA R - CCC- 103616441 TACCCTATTCTATTTCTCCCTA CK19 F - TTTAATATTTTTGTG- 61 41529473 to 442 GTTGAATGAA R - TCCCTAT- 41529031 TAAAAACCCCTTACTTAC SPP1 F - GGTGGGTTGGGTAGTGGTA- 62 87975604 to 268 GAAAAT R - TTTTCCTTAAAATT- 87975872 TAAACTTCCCCATAAA INS F - TGTGAGTAGGGATAGGTTTG- 64 2161202 to 266 GTTAT R - AAAAACTAAAAACTAC- 2161468 TAAACCCCC

Table 2.3: Primers used for bisulﬁte sequencing

75 2.4 Results

2.4.1 MSCs were derived from three CHI patients

Cell lines were derived from three patients with CHI, the details of which are summarised in Table

2.4. The ages of the patients varied and they were from all three major disease backgrounds.

Cell line Age at Disease form Mutation Number of Number of surgery population passages doublings F-CHI 4 months Focal CHI ABCC8 exon 6 (het- 46 16 erozygous) A-CHI 17 months Atypical CHI Unknown 106 24 D-CHI 11 weeks Diﬀuse CHI ABCC8 exon 37 60 16

Table 2.4: Details of cell lines and patients

The cells discussed herein were initially identiﬁed by their morphology in culture. Between two and four days after the operation tissue had adhered to the plastic, following which cells were seen to be growing out around the tissue. Within two days of adherence, the remaining tissue was lost, either through the cells having moved into a monolayer form, or through the cells in the tissue having died and been removed with media changes.

The cells were maintained until they reached approximately 70% confluency. Some flexibility was required for the primary culture as there was not a uniform monolayer as seen at later stages, but rather cells were growing in islands based on where chunks of tissue had adhered. The different cultures also took different amounts of time to reach confluency - F-CHI took 12 days, A-CHI took

7 days and D-CHI took 14 days, this may reﬂect the starting number of cells which may therefore impact the number of population doublings we observed, which varied from approximately 50 to 100 across the three lines. At this stage cells were passaged ‘P1’ and the cell lines began to proliferate more uniformly until senescence which began after 16 passages in F-CHI and D-CHI and 24 passages in A-CHI. The cell lines had a similar and stable mesenchymal appearance throughout culture as shown in Figure 2.1, panel A.

The cells were initially identified as MSCs as detailed above by the characteristic morphology and plastic adherence but it was important to confirm that they were authentic MSCs. Authentic MSCs must express (or not express) a panel of specific cell surface markers and be able to differentiate to multiple mesodermal lineages (Dominici et al., 2006).

76 Figure 2.1: Derived cell lines showed characteristic features of MSCs (A) Brightfield images show a similar plastic adherent, mesenchymal morphology in each cell line, images shown are from 2 days after passaging (B) Flow cytometry showed the cells to be positive for CD29, CD44, CD73, CD90 and CD105 vs isotype control and negative for CD45. 1 x 104 cells were counted in each condition. Graphs are representative for between 1 and 4 experiments each. i is A-CHI, ii is D-CHI and iii is F-CHI. 77 We tested these characteristics in all three cell lines and found similar results. All three cell lines were negative for CD45 when tested by flow cytometry, a marker of haematopoeitic stem cells confirming no cross contamination with blood. This was important to exclude as the tissue post- surgery contains blood. Further, as shown in Figure 2.1, all three were positive for CD29, CD44,

CD73, CD90 and CD105. Each of these markers alone is insuﬃcient to identify the cell type, but the combination is highly speciﬁc for MSCs giving further evidence that we had successfully isolated

MSCs from the patients’ pancreata. Authentic MSCs should be >95% positive for CD73, CD90 and

CD105 and <2% positive for CD45 (Dominici et al., 2006), which our cells were - the percentage of cells positive for each marker is shown in Figure 2.5.

Marker F-CHI A-CHI D-CHI CD29 84.3±4.4 93.7 77.2 CD44 95.1 88.06±10.6 95.2±4.0 CD45 2.21 1.9±0.7 1.9±1.0 CD73 97.6 96.1 99.7±0.04 CD90 81.0 96.7 94.6±3.1 CD105 98.7 99.9 99.8

Table 2.5: CD marker expression: Average percentage of cells expressing CD markers relative to isotype control is given, with standard deviation. Markers were assessed by ﬂow cytometry at passages 9, 11 and 12 in F-CHI, 5, 11, 14 and 16 in D-CHI and 9, 12 and 14 in A-CHI

The average total cells expressing each CD marker is shown in Table 2.5. These are the results from between one and four experiments, in the case of CD29, CD73 and CD105 there are only one to two experiments for each cell line as the ﬁrst antibodies used (all from Abcam) did not work well. For CD90, results were varied, with results showing anything from fully negative to >95% of cells expressing the antigen, despite cells being identical - this may have been due to the aliquots of antibody, further repeats using a new antibody may be helpful in this case.

One of the defining characteristics of MSCs is multipotency, which defines their ‘stemness’. This is assessed by the ability of the MSCs to differentiate to three mesoderm derived mature cell types

- adipocytes, osteoblasts and chondrocytes. The diﬀerentiated cells are normally assessed by a phenotypic assay that can be seen using histological stains (Dominici et al., 2006, Marion and Mao,

2006). Mature osteoblasts deposit calcium as in bone, mature adipocytes swell and contain lipid droplets as their function is to store fat, mature chondrocytes secrete proteoglycans in order to form connective tissues (Marion and Mao, 2006). Again, similar results were seen for all three cell lines. There was widespread positive staining seen for sulfated proteoglycans in all micromass chondrogenesis cultures and for calcium in the osteogenesis cultures. Only rare positive stains were

78 Figure 2.2: Derived cell lines showed multipotency, characteristic of MSCs (A) Histo- logical stains were done for differentiated osteoblasts (alizarin red), chondrocytes (alcian blue) and adipocytes (oil red) in all three cell lines after differentiation at passage 14 in D-CHI, 16 in A-CHI and 7 in F-CHI, and in control cells cultured in standard media, for the undifferentiated A-CHI only is shown, representative for all three. (B) shows further validation of the differentiation, where RT-PCR for a key marker gene was performed for each lineage in F-CHI P7 only.

79 seen in the adipogenesis cultures, <1% of cells contained lipid droplets, but those which did were packed full of lipid droplets. No positive stains were seen in the controls for calcium or lipids, slight staining was seen throughout the controls for the proteoglycan staining, but this was less intense than in the differentiated cells. Figure 2.2 shows images of each of the stains, with the positive cells being shown for adipogenesis rather than a representative field of the mostly negative stains. To confirm that what we were seeing was differentiation to the lineage, RT-PCR was performed for a marker of each lineage. We felt this was necessary, as proprietary kits were used so there was a chance that they were inducing formation of calcium deposits, lipid droplets or proteoglycan deposition without the cells being actually differentiated. The markers used were seen in F-CHI and so only histological stains were used for the other two cell lines. The three CHI derived cell lines showed plastic adherence, specific cell surface markers and the ability to differentiate to two of three cell types, and potentially the third (adipocytes). Together with the flow cytometry, these data confirm that the cells we had derived were MSCs.

One of the deﬁning characteristics of MSCs, as suggested by the name is that they are mesenchymal, not epithelial. Most mature cells in the body, including islet cells, are epithelial. Mesenchymal cells are seen primarily in connective tissue and during development, and are characterised by an ab- sence of junction proteins and the presence of certain cytoskeletal components. We therefore wanted to investigate the mesenchymal or epithelial phenotype of the cells. The cell lines were assessed for gene and protein expression of a panel of markers of epithelial and mesenchymal cells as shown in

Figure 2.3.

Strong expression was seen for all three mesenchymal genes, and also for OCLN which is part of tight junctions, and at a much lower level, CDH1 which encodes E-cadherin, part of adherens junctions. The immunostaining showed similar results, clear expression was seen of the mesenchymal cytoskeleton proteins, with the characteristic localisation of the proteins. Similarly tight junction proteins were not clearly seen in the cells showing that they are not epithelial, although there was some signal for Occludin. The fact that expression of the genes encoding junction proteins E-Cadherin and Occludin is seen was unexpected but may suggest that the cells are able to switch to an epithelial phenotype given the necessary signals.

80 Figure 2.3: Mesenchymal proteins were expressed in CHI pMSCs Markers were assessed between passages 9 and 14 in all three lines to conﬁrm if cells are epithelial or mesenchymal by (A) RT-PCR for 3 mesenchymal markers and 2 epithelial markers, mesenchymal genes NES, VIM and ACTA2 were present in all lines, epithelial markers OCLN and CDH1 were also present. (B) The mesenchymal markers were also present as protein in the characteristic cytoskeleton form Nestin and Vimentin shown in D-CHI and α-smooth muscle actin shown in A-CHI; epithelial marker E-Cadherin was not present at all as protein shown in F-CHI. Immunostaining images are representative for all three lines, all scale bars are 50 µm. Rabbit and mouse panels show secondary antibody only stains. (C) shows positive controls for the antibodies, wherein western blot was used to conﬁrm that the antibodies bind at the correct size to cells known to express the target (HeLa, A549 and MCF7s were used).

81 Figure 2.4: CHI pMSCs are genetically stable and unique(A) G-band karyotypes of each cell line, chromosomes from 1 cell are shown, representative of 19-20 metaphase spreads from passage 10 in D-CHI, 12 in A-CHI and 9 in F-CHI. (B) DNA sequencing of the ABCC8 gene showed F-CHI to have both adenine and guanine at the indicated base, showing heterozygosity, and D-CHI to have just an adenine indicating homozygosity (C) STR proﬁles were unique for each cell line, the number of repeats at each locus in each cell line is given, assessed at passage 3 and 16 in D-CHI, 4 in A-CHI and 4 in F-CHI.

82 2.4.2 Derived cells were karyotypically stable, unique and contained the CHI causing mutations

To confirm that the cells were unique and specific to the patients, STR profiling was carried out.

This gave the unique fingerprint at 16 loci for each patient and therefore each cell line. Each STR profile was compared to the other cell lines and the ATCC catalogue, which includes STR profiles for all banked cell lines. No matches over 80% were seen to lines from the ATCC catalogue across the 8 loci which are free to compare, where 80% signifies a match (Capes-Davis et al., 2013). The number of repeats at each locus is shown in Figure 2.4 panel C.

Additional verification was through sequencing of the ABCC8 gene. As part of the clinical diagnosis of CHI the full ABCC8 gene (among other genes known to cause CHI) is sequenced to identify the mutation responsible for the disease. We amplified and sequenced the exon reported by the hos- pital and compared this to the information given to verify the cells’ origins. The affected nucleotide and surrounding sequence are shown in Figure 2.4B, these were compared to the wild type sequence from NCBI to identify any mutations.

In F-CHI mutation was therefore identiﬁed as ABCC8 exon6 heterozygous d. 16894G>A; c.1016G>A ; p. Gly316Arg, matching the details provided by the clinicians. The sequencing con-

ﬁrmed the origins of the tissue from this patient, and also highlighted the fact that the cells grew out of non-lesion tissue as the DNA present is heterozygous. Focal CHI is caused by the secondary loss of the chromosome 11 p-arm so aﬀected lesion cells would not have the wild type allele present.

In D-CHI the mutation was identified as ABCC8 exon37 homozygous c.4481G>A ; p.Arg1494Gln, again matching the details provided by the clinicians and confirming the origins of the tissue from this patient. The affected nucleotide is shown in Figure 2.4, panel B, the only adenine shown would be a guanine in healthy individuals.

A-CHI was not genotyped by us, as the clinicians did not detect a mutation in any known CHI- associated gene, which as discussed above includes ABCC8, KCNJ11, GCK and HNF4A.

The three cell lines were also conﬁrmed as karyotypically normal, as shown in Figure 2.4A. This allowed conﬁrmation that the cell lines were stable in culture and could therefore be used for further studies. For each cell line, 19-20 cells were analysed by G-banding and all cells were identical, except for 1 cell in the A-CHI line which had a reciprocal translocation. Low levels of random mutations are known to occur in proliferative cell lines in vitro, the A-CHI line was most proliferative meaning there was more chance of random mutations occurring. As this was only seen in one cell we determined

83 the cell line was still suﬃciently normal to be used.

2.4.3 CHI derived MSCs express pancreatic markers

Figure 2.5: Pancreatic genes are expressed in CHI pMSCs RT-PCR for genetic markers of pancreatic development in A-CHI at passage 2, 16 days after the start of culture (labelled early) and for all three CHI lines at passage 9 (labelled late). 18S rRNA is shown as a housekeeping reference gene for all other genes. The developmental stages at which the genes are expressed is shown in the center.

Having conﬁrmed that the CHI derived cells were MSCs we examined numerous pancreatic lineage markers in comparison to bmMSCs, which are the usual MSCs used in research. Having pancreatic

84 genes already expressed would be a beneﬁt for diﬀerentiation in pancreatic research.

A panel of transcription factors from various stages throughout the classical pancreatic developmental pathways was assessed (Pan and Wright, 2011). Gene expression was extremely stable from around passage 5 until senescence occurred when cells were grown on tissue culture plastic in standard culture media. Prior to passage 5, additional genes were expressed which were lost by passage 5, this is shown in Figure 2.5. This included PDX1, NEUROD, INS, GCG, SST and the transcription factors expressed after passage 5. The transcription factors expressed were the same across the three cell lines later, this is shown in Figure 2.5. Several of these transcription factors are expressed at multiple stages during development, for clarity these are shown at the earliest stage at which they are expressed only.

CHI derived MSCS expressed three of five early stage markers, covering early endoderm and multipotent progenitors, as shown in Figure 2.5. SOX9 and SOX17 were clearly expressed, FOXA2 was not expressed at all, PDX1 was possibly expressed, but the band was not clearly visible and further quantitation would be needed to confirm this. The SOX genes, or SRY-related HMG box genes, are involved in multiple developmental pathways and alone are not specific for pancreatic development. HES1 is a downstream target of the Notch pathway and is co-expressed with SOX9 in pancreatic progenitors during development - they were also co-expressed in our CHI-derived MSCs.

The cells also expressed two transcription factors associated with endocrine progenitors and endocrine precursors. NEUROG3 is expressed transiently during development but was not seen at any point in the cell lines as shown in Figure 2.5. PAX6 is induced as the cells mature towards diﬀerentiated endocrine cells and was seen throughout culture in our cell lines. However as with the

SOX genes, PAX6 is a member of the paired box family of transcription factors which are involved in development in many lineages. ISL1 was expressed in our three cell lines, this gene as the name indicates is highly speciﬁc for pancreatic endocrine lineages.

There were few markers of immature or mature final endocrine cells present. MAFB is involved with differentiation of β-cells and also involved in developmental pathways throughout the body and is expressed in the cell lines. ARX is a transcription factor involved with α-cell differentiation but was not expressed in our cells. GCG, INS and SST encode the hormones expressed by α-, β- and

δ-cells respectively and all seemed to be expressed, although this was not as clear after passage 5 as before so quantiﬁcation would be required. CK19 is a marker of duct cells within the pancreas and was clearly expressed.

85 Culturing cells on uncoated plasticware led to the loss of some gene expression after several pas-

Figure 2.6: Growing cells on vitronectin and fibronectin increased cell number and insulin gene expression (A) Increased cells were counted after 5 days growth on vitronectin and fibronectin, compared to no coating. Bars are mean of 2 experiments in A-CHI at passages 10 and 11, where 2 x 104 cells were seeded, error bars are standard deviation. (B) INS gene expression was maintained for additional passages in A-CHI when cultured on vitronectin and fibronectin but not when they were cultured on uncoated plasticware. Image shown is from passage 6. sages, which was shown in Figure 2.5. However, when cells were cultured routinely on vitronectin and fibronectin, early results suggested that this allowed maintenance of INS expression, as well as increased cell numbers, shown in Figure 2.6. Further comparisons were carried out of other ECMs, which was inconclusive . This was primarily because differences in gene expression from one passage to another were small, so longer experiments on each separate ECM would need to be carried out.

Attempts to ascertain for how long the gene expression was maintained were also inconclusive, as the qRT-PCR primers for INS did not have good sensitivity at the low expression end, with no true negatives being seen. Vitronectin and ﬁbronectin were initially selected based on their presence in developing islets (Stendahl et al., 2009) and due to the initial results were used as standard culture conditions.

To summarise, the three CHI-derived pMSC lines strongly expressed a number of transcription factors associated with pancreatic development - SOX17, SOX9, ISL1, PAX6 and MAFB. They also seemed to express PDX1 and the mature hormones GCG, INS and SST. The genes expressed are not speciﬁc to any stage but range from early endoderm through to mature hormone expressing cells.

The presence of these genes seems indicative that the cells are of pancreatic lineage, have maintained

86 Figure 2.7: Pancreatic proteins are expressed in CHI pMSCs Immunocytochemistry showed SOX9, ISL1, PAX6 and MAFB mRNA was translated and expressed as proteins, but was not necessarily correctly located. Images shown are from D-CHI, representative for all three cell lines, from passages between 7 and 12. All scale bars are 50 µm. White arrow on Islet1 indicates the only nucleus stained positive with this antibody. Rabbit and mouse panels show secondary antibody only stains.

87 some aspects of the lineage and may therefore be able to be re-diﬀerentiated to mature pancreatic cells. The presence of factors from multiple stages of development, endocrine and ductal cells may be due to a mixed population of cells being present.

We assessed whether the markers, SOX9, MAFB, PAX6 and ISL1 were translated to proteins.

As these are transcription factors expression should be primarily nuclear. This was the case for

MAFB, but ISL1 was only seen to be nuclear in a minority of cells, shown in Figure 2.7. PAX6 and

SOX9 appeared to be solely located to the cytoplasm. This indicated that the proteins were likely not functional and causing downstream transcription of pancreatic genes, conditions to cause nuclear translocation of these proteins may help re-diﬀerentiation of these progenitor cells.

2.4.4 CHI pMSCs have altered gene expression and methylation patterns compared to bmMSCs

Figure 2.8: Comparison of gene expression between CHI pMSCs and bmMSCs (A) RT- PCR images for transcription factors involved in pancreatic diﬀerentiation that are expressed in our cells show they are also expressed in bmMSCs except for ISL1, A-CHI is shown in the ﬁrst column as a reference (B) qRT-PCR fold changes for ISL1 in CHI lines vs bmMSCs show increased ISL1 expression in CHI lines, n=3 individual lines for the bmMSCs, passages 3 and 4, and n=3 passages for each of our CHI lines, passages 3 and 4, error bars show standard deviation.* p<0.05, *** p<0.005.

SOX9, ISL1, PAX6, MAFB and CK19 are implicated in diﬀerentiation in the pancreas, but are also seen in other roles and tissues, and so their gene expression was assessed in three independent bmMSC lines shown in Figure 2.8.

88 All of these genes were also expressed in bmMSCs except for ISL1 which did not appear to be expressed. This difference was quantified in our three cell lines across three passages and the three independent bmMSCs which confirmed that the bmMSCs did not express ISL1 but our cells did. In the bmMSCs there was no expression detected at all i.e. the CT value was called at 40, the maximum in the assay used, but this could in fact be much higher (infinite) giving an artificial fold change.

ISL1 is a gene speciﬁcally involved in the development of the endocrine pancreas, in particular the islets of langerhans for which it is named. The fact it is expressed in the pancreas derived lines but not the bmMSCs, highlights the fact that the cells have a remnant of their tissue of origin and may therefore be more suited to rediﬀerentiating to this lineage than bmMSCs. Based on this we hypothesised that there may be epigenetic memory of the pancreatic lineage in these cells.

Of the genes which were not clearly expressed, we tested the methylation in the 5’ CpG islands in CHI derived pMSCs against the bmMSCs, to see if there was epigenetic memory in the cells.

Both PDX1 and CK19 showed overall less methylation in the CHI derived cells than the bmMSCs, with significantly less methylation at two loci in each, shown in Figure 2.9. In the PDX1 promoter, the differentially methylated loci align to the region closest to the coding sequence, just after the enhancer sequences (area II/III). The loci both show differences of 20% - locus 7: 36% and 56% in pMSCs and bmMSCs respectively, locus 9: 22% and 42%. CK19 had been used for comparison, as it appeared to be expressed in both CHI and bmMSCs, but showed differences in methylation at the 2 loci of over 10% (locus 2: 69% and 87%, locus 6: 61% and 73% in pMSCs and bmMSCs respectively). The differential degree of methylation may indicate alternate regulation in the different tissue types, such as for enhancer binding. INS is the gene which encodes insulin, the protein most essential when producing β-cells. There was no difference in methylation levels between the CHI derived and bmMSCs for INS. INS does not have a CpG island upstream and primers were therefore designed to amplify the most dense region of CpGs proximal to the gene – they may not be involved in regulation of the gene as this has not been studied. AMY2A, the gene encoding pancreatic amylase in exocrine cells showed different methylation in the CHI derived cells compared to the bmMSCs, with significantly less methylation at one locus (24% methylation in pMSCs and 49% methylation in bmMSCs, p=0.0122) and somewhat increased at another locus with the rest being identical. It is unclear whether this would mean the cells were able to preferentially differentiate to an exocrine lineage but exocrine cell markers may need to be excluded when attempting to differentiate the cells down the endocrine lineage.

To conﬁrm that the decreased methylation seen was not global, we checked the 5’ region of SPP1

89 90 Figure 2.9: Comparison of gene methylation between CHI pMSCs and bmMSCs The percentage methylation at each CpG in the 5’ regions of PDX1, INS, AMY2A and CK19 was assessed by bisulﬁte sequencing. The top panel for each gene shows the percentage methylation (shade of grey) at each locus (individual circle), with the averages for the three CHI and three bmMSCs shown below to aid clarity. * p<0.05 assessed by Kruskal-Wallis test within MethPlotter. The boxed regions on the top panels indicate the locus at which signiﬁcance was seen. Methylation was assessed in all samples at passage 3 or 4.

91 (osteopontin) - a marker of a non-pancreatic lineage that MSCs diﬀerentiate to i.e. osteoblasts.

Analysis was not fully possible as the three CHI pMSC lines did not have a wild type sequence, with an extra CpG being present at the -65 position of the promoter (Hijiya et al., 1994) as shown in Fig- ure 2.10, the blue and black peaks representing cytosine and guanine respectively within the boxed regions.It is unclear why this mutation would be present in only the CHI derived cells, although

SPP1 is a downstream target of SOX9 (Pritchett et al., 2012) and a putative pancreatic ductal marker (Roberts, 2011). A mutation affecting regulation of SPP1 expression in the pancreas might therefore have implications in the aetiology of CHI. We checked this mutation with normal DNA sequencing on the non-bisulfite converted gDNA. The sequencing revealed that the gDNA did not have the extra CpG, shown in Figure 2.10B - within the box where the sequence reads ”CTCCG”, the T would become G. This suggests that an artifact of the bisulfite conversion is responsible for the altered bisulfite sequencing result, potentially due to the degree of methylation at the true CpG locus.

92 Figure 2.10: Bisulfite sequencing and gene sequencing of SPP1 (A) Bisulfite sequencing traces of each of the three bmMSC samples and the three CHI pMSC samples used previously for the promoter region of SPP1. Boxed regions highlight the region of the trace where differences between CHI and bmMSCs were seen. (B) Sequencing of the same region on non bisulfite converted gDNA of the F-CHI line, boxed region corresponds to the boxed regions in panel A.

93 2.5 Discussion

In this study we report the reproducible derivation of cell lines from the pancreatic tissue of three patients with CHI. Each patient had a diﬀerent form of the disease which had no impact on the ability to derive a cell line. The derived cells were not clonal, as the origins of pMSCs or the best cell type to use are not fully understood (Bhonde et al., 2014).

The cell lines have the defining characteristics of MSCs as set down by The International Society for Cellular Therapy (Dominici et al., 2006), shown in Figures 2.1 and 2.2. The derived cells show plastic adherence (which is how they are initially selected for, from the tissue), they are able to differentiate to osteoblasts, chondrocytes and partially to adipocytes and express cell surface markers CD29, CD44, CD73, CD90 and CD105, whilst being negative for CD45. The low percentage of adipocytes arising from the differentiation is in concordance with previous data from pancreas derived MSCs (Russ et al., 2009). The authors of Russ et al. (2009) suggested that this may mean pMSCs are not authentic MSCs, however we believe this may be due to tissue specific biases as discussed below. An opposing study showed robust differentiation to adipocytes in pMSCs which had been lineage traced and shown to not be derived from INS/PDX1 positive cells (Chase et al.,

2007). This could suggest that poor adipocyte diﬀerentiation is a feature of MSCs derived from the pancreatic lineages and that in Chase et al. (2007), the derived cells were in fact contaminating bmMSCs as has happened in other studies (Sordi et al., 2010).

CHI pMSCs also express mesenchymal cytoskeletal proteins vimentin, nestin and α-smooth muscle actin and are negative for epithelial proteins E-cadherin and occludin as shown in Figure 2.3. This was in agreement with other studies of pMSCs (Davani et al., 2007, Gallo et al., 2007). E-cadherin is quickly lost in MSCs (Russ et al., 2008) with networks of epithelial-mesenchymal transition markers being seen (Kutlu et al., 2009) suggesting pMSCs are derived from epithelial cells within the pancreas. Nestin has been discussed as a putative stem cell marker, but is co-expressed with vimentin,

α-smooth muscle actin, CK19 and amylase (Street et al., 2004) or with MSC markers CD90 and

CD105 within the pancreas (Carlotti et al., 2010). All of these (bar amylase) were expressed in the

CHI pMSCs, providing evidence of their similarity to other pMSCs. However this provides little information on whether they are derived from mesenchymal or ductal cells within the pancreas, or de-diﬀerentiated epithelial endocrine cells.

The source of MSCs within the body has been shown to be important, with differences seen between proliferation and differentiation ability. BmMSCs undergo senescence before other types and only periosteum derived MSCs could be differentiated to β-cells sufficiently for glucose stimulated

94 insulin secretion, of four cell types tested (Kim et al., 2012). BmMSCs and pMSCs have been shown to be diﬀerent at a basal level, and after diﬀerentiation to insulin secreting cells (Kutlu et al., 2009,

Zanini et al., 2011). We therefore investigated pancreatic markers within the CHI derived pMSCs compared to bmMSCs, the most common MSC source shown in Figures 2.4-8. We observed that the derived cells express a number of genes associated with pancreatic development – SOX17, SOX9,

PAX6, MAFB, HES1 and ISL1. Some of these markers were similar to those reported before in pMSCs (Eberhardt et al., 2006) although PAX6 and ISL1 have been shown to be downregulated in pMSCs compared to islets in one study (Kutlu et al., 2009), although another showed pMSCs to be

ISL1 negative (Gallo et al., 2007). HES1 has been reported before to be involved in adaptation of

β-cells to in vitro proliferation (Bar et al., 2008), which is likely the role of its expression in our cells.

Similarly, SOX9 is known to be associated with maintenance of progenitors in an undifferentiated state (Seymour et al., 2007). The genes we saw expressed are not solely associated with the pancreas, so we compared the expression of them in our cell lines to non-pancreatic MSCs, derived from bone marrow. All the genes were expressed in the bmMSCs except for ISL1. This suggests that there is some tissue specific genetic memory in the CHI derived MSCs, which may be an advantage for redifferentiating the cells to pancreatic lineages.

The cells also expressed CK19, a ductal marker (Sahin et al., 2005, Bouwens et al., 1994). How- ever, CK19 was also expressed in bmMSCs, it is not clear the role that CK19 may play in MSCs, and it has not been reported in the bone marrow in vivo. CK19 has been reported previously in pMSCs which were also derived from exocrine enriched pancreatic digest, however expression of CK19 was lost by P3 in this report (Lima et al., 2013). The bmMSCs used in this study which expressed CK19 were only available at P3, as in Lima et al. (2013), and so later gene expression was not studied; the

CHI derived cells expressed CK19 after P3 though. The prolonged CK19 expression in the CHI- derived pMSCs may be indicative of a ductal origin, which has been hypothesised as the source of

MSCs within the exocrine compartment (Fanjul et al., 2010). The ducts are frequently hypothesised to be the primary source of pancreatic stem cells lending weight to this theory.

Some genes essential for endocrine diﬀerentiation to β-cells, such as PDX1 and INS, were not expressed in the CHI-derived pMSCs. It has been shown that pMSCs have more active histone acetylation for INS and GCG than bmMSCs, although INS promoter accessibility was inhibited in both MSC types (Wilson et al., 2009). In line with this, we saw no diﬀerence in 5’ methylation of

CpGs upstream of INS, but there was signiﬁcantly less methylation in the CpG islands upstream of

PDX1, CK19 and AMY2A. The location and magnitude of the diﬀerential methylation of the PDX1 promoter makes it most likely that this gene, of the genes tested, would have transcription aﬀected

95 by methylation. This suggests that despite CHI pMSCs not expressing PDX1, re-induction may be easier than in bmMSCs. However, as several of the markers the MSCs expressed were from before endocrine commitment, there may also be the possibility of re-induction of amylase and diﬀerentia- tion to an exocrine lineage which would need to be excluded. De-methylation of INS may occur with expression of PDX1, but the use of DNA methyltransferase inhibitors may be required for successful diﬀerentiation.

In conclusion, pMSCs were reproducibly derived from three patients with CHI, with each form of the disease. All three lines showed a similar phenotype, characteristic of MSCs with a bias against adipose diﬀerentiation which may be characteristic of MSCs from the pancreas. Further contributing to the idea that the CHI pMSCs were unique for the pancreas, they expressed ISL1 which was not seen in bmMSCs and show diﬀerential promoter methylation to bmMSCs as we showed in Figures

2.8 and 2.9. This may mean CHI pMSCs are able to preferentially redifferentiate to β-cells over other MSC types. We therefore present the first derivation of MSCs from CHI patients and propose their use for pancreatic research by differentiation to β-cells.

96 2.6 References

J.-B. Arnoux, V. Verkarre, C. Saint-Martin, F. Montravers, A. Brassier, V. Valayannopoulos,

F. Brunelle, J.-C. Fournet, J.-J. Robert, Y. Aigrain, C. Bellanne-Chantelot, and P. de Lonlay.

Congenital hyperinsulinism: current trends in diagnosis and therapy. Orphanet Journal of Rare

Diseases, 6(1):1750–1172, 2011.

Y. Bar, H. A. Russ, S. Knoller, L. Ouziel-Yahalom, and S. Efrat. HES-1 Is Involved in Adaptation

of Adult Human β-Cells to Proliferation In Vitro. Diabetes, 57(9):2413–2420, 2008.

O. Bar-Nur, H. Russ, S. Efrat, and N. Benvenisty. Epigenetic Memory and Preferential Lineage-

Speciﬁc Diﬀerentiation in Induced Pluripotent Stem Cells Derived from Human Pancreatic Islet

Beta Cells. Cell Stem Cell, 9(1):17–23, 2011.

R. R. Bhonde, P. Sheshadri, S. Sharma, and A. Kumar. Making surrogate β-cells from mesenchymal

stromal cells: Perspectives and future endeavors. The International Journal of Biochemistry &

Cell Biology, 46:90 – 102, 2014.

L. Bouwens, R.-N. Wang, E. D. Blay, D. G. Pipeleers, and G. Kloppel. Cytokeratins as Markers of

Ductal Cell Diﬀerentiation and Islet Neogenesis in the Neonatal Rat Pancreas. Diabetes, 43(11):

1279–1283, 1994.

A. Capes-Davis, Y. A. Reid, M. C. Kline, D. R. Storts, E. Strauss, W. G. Dirks, H. G. Drexler,

R. A. MacLeod, G. Sykes, A. Kohara, Y. Nakamura, E. Elmore, R. W. Nims, C. Alston-Roberts,

R. Barallon, G. V. Los, R. M. Nardone, P. J. Price, A. Steuer, J. Thomson, J. R. Masters, and

L. Kerrigan. Match criteria for human cell line authentication: Where do we draw the line? Int.

J. Cancer, 132(11):2510–2519, 2013.

F. Carlotti, A. Zaldumbide, C. J. Loomans, E. van Rossenberg, M. Engelse, E. J. de Koning, and

R. C. Hoeben. Isolated human islets contain a distinct population of mesenchymal stem cells.

Islets, 2(3):164–173, 2010.

L. G. Chase, F. Ulloa-Montoya, B. L. Kidder, and C. M. Verfaillie. Islet-Derived Fibroblast-Like

Cells Are Not Derived via Epithelial-Mesenchymal Transition From Pdx-1 or Insulin-Positive Cells.

Diabetes, 56(1):3–7, 2007.

E. Corritore, E. Dugnani, V. Pasquale, R. Misawa, P. Witkowski, J. Lei, J. Markmann, L. Piemonti,

E. M. Sokal, S. Bonner-Weir, and P. A. Lysy. β-Cell Diﬀerentiation of Human Pancreatic Duct-

Derived Cells After In Vitro Expansion. Cellular Reprogramming, 16(6):456–466, 2014.

G. Danaei, M. M. Finucane, Y. Lu, G. M. Singh, M. J. Cowan, C. J. Paciorek, J. K. Lin, F. Farzadfar,

Y.-H. Khang, G. A. Stevens, M. Rao, M. K. Ali, L. M. Riley, C. A. Robinson, and M. Ezzati.

97 National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980:

systematic analysis of health examination surveys and epidemiological studies with 370 country-

years and 2.7 million participants. The Lancet, 378(9785):31–40, 2011.

B. Davani, L. Ikonomou, B. M. Raaka, E. Geras-Raaka, R. A. Morton, B. Marcus-Samuels, and

M. C. Gershengorn. Human Islet-Derived Precursor Cells Are Mesenchymal Stromal Cells That

Diﬀerentiate and Mature to Hormone-Expressing Cells In Vivo. Stem Cells, 25(12):3215–3222,

2007.

M. Dominici, K. Le Blanc, I. Mueller, I. Slaper-Cortenbach, F. Marini, D. Krause, R. Deans, A. Keat-