Copyright and use of this thesis This thesis must be used in accordance with the provisions of the Copyright Act 1968. Reproduction of material protected by copyright may be an infringement of copyright and copyright owners may be entitled to take legal action against persons who infringe their copyright. Section 51 (2) of the Copyright Act permits an authorized officer of a university library or archives to provide a copy (by communication or otherwise) of an unpublished thesis kept in the library or archives, to a person who satisfies the authorized officer that he or she requires the reproduction for the purposes of research or study. The Copyright Act grants the creator of a work a number of moral rights, specifically the right of attribution, the right against false attribution and the right of integrity. You may infringe the author’s moral rights if you: - fail to acknowledge the author of this thesis if you quote sections from the work - attribute this thesis to another author - subject this thesis to derogatory treatment which may prejudice the author’s reputation For further information contact the University’s Director of Copyright Services sydney.edu.au/copyright

Faculty of Medicine

Strategies to combat inflammation in pancreas and islet transplantation

Tina Ghoraishi A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

3rd of September 2014

Declaration

This thesis is the result of full time and part time study during the period from February

2007 to August 2014. The work, except where duly acknowledged, is my own, and has not been submitted for any other degree at this, or any other University. In vivo experiments in the porcine brain death model involved the participation of a large number of collaborators at the Victor Chang Cardiac Research Institute. The work in the murine brain death model was performed jointly with Zane Wang, packaging and purification of the rAAV vectors was carried out by Moumita Paul.

Tina Ghoraishi August 2014

ii Acknowledgements

I would like to express sincere appreciation to the many individuals who have offered support and encouragement throughout this project. First and foremost, my deepest gratitude goes to my supervisor A/Prof Alexandra Sharland, for her constant support and inspiration, for patiently guiding me through this research effort with her generosity and wealth of knowledge. Appreciation is extended to collaborators from the Victor

Chang Cardiac Research Institute for their great contribution to development of the porcine brain death model. Special thanks go to Zane Wang and Chuanmin Wang for their endeavours in establishment of murine brain death model. I also would like to sincerely thank Moumita Paul for her efforts in virus preparation, and all her assistance.

Additionally, appreciation is expressed to The Transplant Society of Australia and New

Zealand, and Professor Christopher Jackson for providing resources. Last but not least, I am very grateful to assistance and support from all the group members; Alexander

Bishop, Eithne Cunningham, Miriam Habib, Mamdouh Moawadh and Anar Ganbold.

Thank you all, without your valuable contribution this degree would have never been completed!

iii For my husband, his love and encouragement keeps me pursuing my dreams.

For my mother, her unconditional love and kindness brighten difficult days.

For my sisters, for having faith in me.

And, to the loving memory of my father, for all-time.

iv Contents Abstract ...... viii

List of tables ...... xii

List of figures ...... xiii

List of abreviations ...... xviii

CHAPTER 1 ...... 1

Introduction ...... 1 1.1 Deceased donor organ transplantation ...... 2 1.2 Indications for pancreas and islet transplantation ...... 2 1.3 Organ quality and the outcomes of pancreas and islet transplantation ...... 10 1.4 Clinical outcomes ...... 16 1.5 The innate immune system...... 21 1.6 Brain death ...... 41 1.7 Donor management/preconditioning ...... 51 1.8 Experimental strategies to reduce graft inflammation ...... 53 1.9 Modulation of immune responses using soluble RAGE ...... 59

CHAPTER 2 ...... 62

Materials and methods ...... 62 2.1 Porcine neonatal islet cell cluster (pNICC) preparation and culture ...... 63 2.2 Porcine Aortic Endothelial Cell isolation from piglet aortas ...... 79 2.3 Preparing tissue segments for immunohistostaining purposes ...... 81 2.4 Immunohistostaining methods ...... 82 2.5 Brain death induction in pigs and different donor management protocols . 85 2.6 Apoptosis detection Assays ...... 91 2.7 Splenocyte isolation from C57BL/6 mouse spleen ...... 96 2.8 Brain death induction in C57BL/6 mice ...... 97

CHAPTER 3 ...... 113

Inflammation in porcine neonatal islet cell clusters (PNICCs) ...... 113 3.1 Introduction ...... 114 3.2 Porcine neonatal islet cell clusters in culture after isolation ...... 115

v 3.3 Gene expression profiling of porcine neonatal islet cell clusters in culture 117 3.4 Activated protein C and the endothelial protein c receptor ...... 129 3.5 Obtaining full length EPCR from porcine neonatal islet cell clusters and other cell types ...... 135 3.6 Porcine aortic endothelial Cell (PAEC) treatment with rHuAPC ...... 145 3.7 Antibody against human EPCR is active on porcine tissue but EPCR protein was not detected in islets ...... 156 3.8 Treatment of pNICCs with recombinant human activated protein C ...... 160 3.9 Discussion ...... 167

CHAPTER 4 ...... 170

Hormone resuscitation of the brain-dead donor and its effect on pancreatic inflammation and function ...... 170 4.1 Introduction ...... 171 4.2 Physiological changes in BD pig ...... 172 4.3 Gene expression in pancreas from brain-dead pigs ...... 177 4.4 Measurement of insulin production in pancreatic islets ...... 186 4.5 Detection of apoptosis in pancreas tissue from brain-dead pigs ...... 190 4.6 Discussion ...... 203

CHAPTER 5 ...... 209

Systematic overexpression of esrage and its effects on the pancreas and islets in brain-dead mice ...... 209 5.1 Introduction ...... 210 5.2 Expression of esRAGE in vivo after inoculation with rAAV esRAGE ...... 211 5.3 Physiological changes and systematic inflammatory response in brain dead mice 215 5.4 Investigation of inflammation in the pancreas post brain death ...... 217 5.5 Insulin protein and gene expression in pancreas post brain death ...... 242 5.6 Detection of apoptosis in brain-dead mouse pancreatic tissue ...... 245 5.7 Discussion ...... 251

CHAPTER 6 ...... 255

General discussion ...... 255

vi Appendix : Composition of Buffers, Solutions and media ...... 269

References ...... 274

vii ABSTRACT

Pancreas and islet transplantation are life-saving treatments for patients with type 1 diabetes mellitus and intractable hypoglycaemia unawareness. Donated pancreata are predominantly derived from deceased donors, most of whom have succumbed to brain death. The pancreas and islets are particularly susceptible to damage during donor brain death, and early inflammatory changes are compounded by further damage during islet isolation and intraportal infusion. The growing prevalence of T1DM and scarcity of the organ pool necessitate the development of strategies to minimise inflammatory events and improve organ quality and retrieval rates. This thesis spans three separate but related studies: evaluating the effects of isolation and anti-inflammatory treatment on isolated pancreatic islets, and modulation of inflammation in pancreas from brain-dead donors using two different anti-inflammatory approaches, one already established in clinical practice and one experimental. For this purpose, we examined isolated porcine neonatal cell clusters islets in vitro, and established porcine and murine brain death models recapitulating the human condition.

Porcine neonatal islet cell clusters are considered as a potential alternative to human islets, and they are extensively utilised in experimental models. Activated protein C is a serine protease with cytoprotective, anti-apoptotic and anti-coagulant properties, which can attenuate inflammation and potentially protect islets. We hypothesized that rHuAPC could exert anti-inflammatory effects on porcine islets in culture, based on the activity of rHuAPC on tissues from several other species, and the high level of homology between porcine and human protein C and its receptor, EPCR. We studied islets from

1-3 day old piglets after isolation during 14 days of culture, measuring inflammatory

viii gene expression. Progressively increased insulin gene expression was evident over this period as the NICCs matured. Expression of most proinflammatory mediators increased after isolation, peaking around day 4-6 and subsequently declining. The principal exception to this pattern occurred for the NKG2D ligands, pMIC2 and pULBP1, whose expression remained elevated throughout the duration of culture. Whereas we were able to discern an effect of rHuAPC on primary porcine aortic endothelial cells in culture, we did not demonstrate an effect of rHuAPC treatment in altering gene expression in

NICCs.

Cardiovascular, metabolic and inflammatory consequences of brain death impact negatively upon the quality of donor organs and upon the outcomes of subsequent transplant procedures. Novel donor management strategies are being employed to support donor haemodynamic status and cellular metabolism and to reduce inflammation in donor organs. Replacement of essential hormones including triiodothyronine, insulin, vasopressin and corticosteroids can maintain the physiological status. We tested the effects of this hormone cocktail on pancreatic function and inflammation in an established porcine model of brain death. Monitoring of the animals revealed that physiological indices such as haemodynamic status, organ perfusion and function were substantially improved by hormone resuscitation compared with those in the control group receiving conventional pressor support with noradrenaline.

The pattern of gene expression for a number of inflammatory mediators was evaluated in whole pancreas specimens – these genes included danger signals, pattern recognition receptors, cytokines, chemokines and stress ligands downstream of TLR signalling, as well as the protective molecule A20. Expression of pro-inflammatory markers was, in

ix general, upregulated in pigs receiving noradrenaline, whereas expression of these genes either showed a more limited increase or was downregulated in the HR group. Insulin gene and protein expression levels were evaluated in the pancreas. Noradrenaline treatment upregulated insulin mRNA transcription, yet the islet insulin protein content decreased from baseline levels, HR maintained insulin content close to normal, despite downregulation of mRNA transcription. We used TUNEL and the activity of cleaved caspase-3 to evaluate the level of apoptosis in porcine pancreas after BD. The TUNEL technique resulted in overestimation of apoptotic cells when the outcomes were compared to c-casp3, indicating that TUNEL might not be an ideal technique for apoptosis detection in pancreas.

Engagement of pattern-recognition receptors by Damage Associated Molecular Patterns

(DAMPs or danger signals) released during brain death results in the expression of various proinflammatory mediators and the propagation of organ damage. esRAGE is a naturally-occuring soluble isoform of RAGE, which is able to attenuate inflammation by competitively binding to RAGE ligands such as HMGB1 and S100B, thus preventing ligation of their receptors TLR2, TLR4, TLR9 and RAGE and also by preventing the formation of RAGE homodimers at the cell surface – a prerequisite for

RAGE signal transduction. To evaluate the effects of systemic esRAGE over-expression upon pancreatic function and inflammation in brain-dead mice, we transduced mice with an rAAV2/8 vector encoding esRAGE. esRAGE gene delivery to the mouse liver yielded high levels of esRAGE in the serum and transplantable organs. BD induction resulted in a systemic inflammatory state, characterised by increased serum levels of the

DAMPs HMGB1 and S100B and of a number of pro-inflammatory cytokines and chemokines including TNFα, IL-6, MCP1 and KC. Gene expression in the pancreas

x was also increased for these proinflammatory cytokines and chemokines, and was modestly reduced by esRAGE overexpression.

Only the early stages of neutrophil migration into the pancreas were apparent during the three-hour observation period following the induction of brain death. Neutrophil margination was elevated in the pancreas of BD mice, and was attenuated in BD mice transduced with esRAGE. Expression of the stress ligand RAE-1 was increased in the pancreas of mice transduced with esRAGE. Whilst this may increase susceptibility of the pancreatic parenchyma to attack by NK cells and CD8+ T cells following transplantation, this tendency may be counterbalanced by the inhibitory effects of

RAGE blockade on activated and memory T cells.

Overall, the most promising of these interventions appears to be hormone resuscitation of the brain-dead donor, a management protocol which is already in clinical use.

However, further evaluation of the effects of HR and of esRAGE expression after islet isolation and allotransplantation will be necessary to fully understand the potential of these treatments to improve transplant outcomes.

xi LIST OF TABLES

Table 1.1 Indication for pancreatic islet transplantation 9 Table 2.1 pNICC medium change schedule 64 Table 2.2 Taqman primer/probe sequences for porcine genes 73 Forward and reverse primers to amplify EPCR sequence in porcine Table 2.3 75 tissue Table 2.4 List of primary antibodies against islet endocrine hormones 84 Table 2.5 Staining scoring system for anti-insulin immunostaining intensity 85 Table 2.6 Hormone-resuscitation protocol 87 Table 2.7 Taqman primer/probe sequences for murine genes 111 Accession numbers for human and porcine mRNA and proteins Table 3.1 131 (protein C and EPCR) Table 3.2 EPCR binding residue in human and pig 134 Table 4.1 H&E staining of BD pig pancreas 175

xii LIST OF FIGURES

Figure 1.1 Factors contributing to loss of islet grafts 15 Figure 1.2 Immune activation by TLR signalling 25 Figure 1.3 Downstream effects of TLR signalling on parenchymal cells 26 Figure 1.4 Activation of the endothelium by TLRs and cytokines 27 Figure 1.5 Monocyte and DC maturation and migration in response to TLR 28 signalling Figure 1.6 Activation of protein C and activity pathways 40 Figure 2.1 pNICCs post isolation in culture stained with Acridine orange + 65 Ethidium Bromide Figure 2.2 Absorbance spectrum of total RNA generated from 69 electrophoresis in the Bioanalyzer Figure 2.3 Sequences of brain death induction, treatment commencement 89 and specimen collection Figure 2.4 Typical races for the optimized model of BD with Gelofusine as 99 fluid support Figure 2.5 Creation of burr hole 101 Figure 2.6 Depiction of tracheostomy 101 Figure 2.7 Exposure of the femoral artery 103 Figure 2.8 Depiction of femoral artery cannulation 103 Figure 2.9 Codon-optimised nucleotide sequence of esRAGE designed by 106 GeneART Figure 2.10 Triple transfection with pAM2AA, pXX6 and p5E18VD2/8, 107 yields rAAV2/8 virions after purification Figure 3.1 A cluster of porcine neonatal islets 116 Figure 3.2 pNICCs at day 3 and 6 post isolation and bright field microscopy 117 x20 Figure 3.3 Comparison of two housekeeping genes in pNICCs 119 Figure 3.4 Insulin gene expression in pNICCs during culture 121 Figure 3.5 Toll-like Receptors 2 and 4 gene expression in pNICCs during 122 culture

xiii Figure 3.6 HSP72 and HMGB1 gene expression in pNICCs during culture 123 Figure 3.7 Chemokines gene expression in pNICCs during culture 125 Figure 3.8 EPCR gene expression in pNICCs during culture 126 Figure 3.9 pULBP1 and MIC2 gene expression in pNICCs during culture 127 Figure 3.10 A20 gene expression in pNICCs during culture 128 Figure 3.11 Alignment between human and porcine protein C mRNA 132 Figure 3.12 Human and porcine protein C preproprotein sequences alignment 133 Figure 3.13 Human and porcine EPCR proteins alignment 134 Figure 3.14 Agarose gel under UV illumination showing PCR product 136 Figure 3.15 PCR product quantification by nanodrop after isolation from gel 136 Figure 3.16 Alignment between PCR product sequencing result from PAECs 137 and pig EPCR CDS Figure 3.17 Nanodrop estimation of DNA amplicons to be used for TOPO- 138 cloning Figure 3.18 Gel electrophoresis of the PCR product 139 Figure 3.19 Nanodrop quantification of the plasmid purified from the 139 bacterial colonies Figure 3.20 Chromatogram for cloned EPCR from PAECs 140 Figure 3.21 Chromatogram for cloned EPCR from pNICCs day 4 post- 141 isolation Figure 3.22 Chromatogram for cloned EPCR from pNICCs day 10 post- 142 isolation Figure 3.23 Alignment between the cloned sequence from pNICC at day 4 143 post-isolation and porcine EPCR mRNA Figure 3.24 Alignment between the cloned sequence from pNICCs at day 10 144 post-isolation and porcine EPCR mRNA Figure 3.25 EPCR gene expression in PAECs 146 Figure 3.26 MMP-2 gene expression in PAECs 147 Figure 3.27 APC is expected to inhibit the expression of these inflammatory- 148 response mediators Figure 3.28 HSP72 gene expression by PAECs 149 Figure 3.29 TNFα gene expression by PAECs 150

xiv Figure 3.30 Chemokines gene expression by PAECs 152 Figure 3.31 MIC2 and pULBP1 gene expression by PAECs 154 Figure 3.32 A20 gene expression by PAECs 155 Figure 3.34 Anti-EPCR staining on human carotid 157 Figure 3.35 Anti-EPCR staining on piglet thoracic aorta 158 Figure 3.36 Anti-EPCR staining on pNICCs staining 159 Figure 3.37 Insulin gene expression in pNICCs 161 Figure 3.38 EPCR and MMP-2 gene expression in pNICCs 162 Figure 3.39 TLR2 and TLR4 gene expression in pNICCs 162 Figure 3.40 HSP72 gene expression in pNICCs 163 Figure 3.41 Chemokines gene expression in pNICCs 165 Figure 3.42 MIC2 and oULBP1 gene expression in pNICCs 165 Figure 3.43 A20 gene expression in pNICCs 166 Figure 4.1 Monitoring of haemodynamic status in brain dead pigs 173 Figure 4.2 Renal artery flow and creatinine clearance in brain dead pigs 174 Figure 4.3 Pancreatic enzyme release in brain dead pigs 174 Figure 4.4 H&E staining on brain dead pig pancreas 176 Figure 4.5 Depiction of the histological damage caused by brain death in pig 176 pancreas Figure 4.6 Absorbance spectrum of total RNA generated from 177 electrophoresis in the Bioanalyzer Figure 4.7 Absorption data for RNA from brain-dead pig pancreas 178 Figure 4.8 HMGB1 and HSP72 gene expression in pig pancreas 179 Figure 4.9 TLR4 gene expression in pig pancreas 180 Figure 4.10 Chemokines gene expression in pig pancreas 182 Figure 4.11 IL-6 gene expression in pig pancreas 183 Figure 4.12 MIC2 and pULBP1 gene expression in pig pancreas 184 Figure 4.13 A20 gene expression in pig pancreas 185 Figure 4.14 C3 gene expression in pig pancreas 186 Figure 4.15 Insulin gene expression in pig pancreas 187 Figure 4.16 Insulin immunostaining scoring based on intensity. 188 Figure 4.17 Insulin content measurement by immunohistostaining 189

xv Figure 4.18 TUNEL assay and localisation of islets in a normal pig pancreas 192 Figure 4.19 TUNEL assay on control pig pancreas at 0hr and 6hr post BD 192 Figure 4.20 TUNEL assay analysis in brain-dead pig pancreas 193 Figure 4.21 TUNEL assay versus cleaved caspase-3 activity in normal mouse 195 thymus Figure 4.22 TUNEL versus anti C. Casp-3 staining in pig and mouse 197 pancreatic islets Figure 4.23 Anti-cleaved caspase-3 staining on brain-dead pig pancreas 198 Figure 4.24 Cleaved Caspase-3 staining analysis in brain-dead pig pancreas 198 Figure 4.25 Genomic DNA from mouse splenocyte measured by Nanodrop 200 Figure 4.26 Genomic DNA from BD pig pancreas measured by Nanodrop 200 Figure 4.27 Genomic DNA electrophoresis visualised on agarose gel 201 Figure 4.28 Example of fragmented genomic DNA due to apoptosis 202 Figure 5.1 Dose-response curve for rAAV-esRAGE 211 Figure 5.2 Administration of rAAV-esRAGE does not result in liver 212 inflammation Figure 5.3 Kinetics of expression for rAAV-esRAGE injected 213 intraperitoneally (IP) Figure 5.4 Levels of esRAGE in various organs 213 Figure 5.5 Serum protein levels of HMGB1 and S100 in mouse 214 Figure 5.6 Serum protein levels of various cytokines and chemokines were 215 measured Figure 5.7 HMGB1 gene expression in mouse pancreas 218 Figure 5.8 Hyaluronan Synthase 1 gene expression in mouse pancreas 219 Figure 5.9 HSPa2 gene expression in mouse pancreas 220 Figure 5.10 TLR4 gene expression in mouse pancreas 222 Figure 5.11 RAGE gene expression in mouse pancreas 223 Figure 5.12 MCP1 gene expression in mouse pancreas 225 Figure 5.13 MIP2 gene expression in mouse pancreas 226 Figure 5.14 IP10 gene expression in mouse pancreas 227 Figure 5.15 KC gene expression in mouse pancreas 228 Figure 5.16 IL-1β gene expression in mouse pancreas 230

xvi Figure 5.17 TNFα gene expression in mouse pancreas 231 Figure 5.18 IL-10 gene expression in mouse pancreas 232 Figure 5.19 IL-33 gene expression in mouse pancreas 233 Figure 5.20 IL-6 gene expression in mouse pancreas 234 Figure 5.21 RAE-1 gene expression in mouse pancreas 236 Figure 5.22 A20 gene expression in mouse pancreas 238 Figure 5.23 C3 gene expression in mouse pancreas 270 Figure 5.24 Insulin gene expression in mouse pancreas 272 Figure 5.25 Anti-insulin immunostaining on mouse pancreas 243 Figure 5.26 Example of the insulin staining intensity values by Image Pro- 244 Premier 9.1 Figure 5.27 Analysis of anti-insulin immunostaining in mouse pancreas 244 Figure 5.28 Normal C57BL/6 mouse pancreas immunostained for C-Casp3 246 and TUNEL Figure 5.29 Cleaved Caspase 3 activity in murine pancreas 248 Figure 5.30 Anti-Ly-6B.2 immunostaining for neutrophils on mouse pancreas 249 Figure 5.31 Analysis of neutrophil presence evaluated by immunostaining 250

xvii LIST OF ABREVIATIONS

AAV Adeno-associated virus

ACIS Automated Cellular Imaging System

ACTH Adrenocorticotropic hormone

ADAM10 A disintegrin and metalloproteinase domain-containing protein

10

ADH Antidiuretic hormone

ATG Antithymocyte immunoglobulin

ATN Acute tubular necrosis

ATP Adenosine triphosphate

Bcl-2 B-cell lymphoma 2

BD Brain death/brain-dead

BMI Body mass index

C-casp3 Cleaved caspase 3

C3 Complement component 3

CBP Cyclic AMP response element binding (CREB) binding protein

CDS Complete coding sequence

CIT Cold ischemia time

xviii CML N(6)-Carboxymethyllysine

CNS Central nervous system

CO Carbone monoxide

CoPP Cobalt protoporphyrin

CpG DNA Oligodeoxynucleotides DNA

CTL Cytotoxic T cells

CVP Central venous pressure

CYLD Cylindromatosis

DAMPs Damage-associated molecular pattern molecules

DBD Donation after brain death

DC Dendritic cells

DCD Donation after cardiac death

DEPC Diethylpyrocarbonate

DMEM Dulbecco’s modified eagle’s medium dNTP Deoxyribonucleotide triphosphate dUTP Deoxyuridine triphosphate

ELISA Enzyme-linked immunosorbent assay

EPCR Endothelial protein C receptor esRAGE Endogenous secretory RAGE

EtOH Ethanol

xix FasL Fas ligand

GAL Galactose-α-1, 3-galactose

GAPDH Glyceraldehyde 3-phosphate dehydrogenase

GFP Green fluorescent protein

GILZ Glucocorticoid-induced leucine zipper

GR Glucocorticoid receptor

HAS2 Hyaloronan synthase 2

HAT Histone acetyltransferase

HbA1c Glycated hemoglobin

HBSS Hank’s balanced salt solution

HEK cells Human embryonic kidney cells

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

HIF-1α Hypoxia-inducible factor 1-alpha

HMGB1 High-mobility group protein B1

HO-1 Heme oxygenase 1

HR Hormone resuscitation

HRP Horseradish peroxidase

HSA Human serum albumin

HSP Heat shock protein

HTK Histidine-tryptophan-ketoglutarate

IBMIR Instant blood-mediated inflammatory reaction

xx ICAM-1 Intercellular Adhesion Molecule 1

ICCs Islet cell clusters

ICU Intensive care unit

IEQ Islet equivalent

IGF-1 Insulin-like growth factor 1

IKK Inhibitor of kappa B kinase

IL-10 Interleukin 10

IL-12 Interleukin 12

IL-1R Interleukin-1 receptor

IL-2 Interleukin 2

IL-33 Interleukin 33

IL-6 Interleukin 6

INF-gamma Interferon gamma iNOS Nitric oxide synthases

IP Intraperitoneal

IP10 Interferon gamma-induced protein 10, C-X-C motif chemokine

10 (CXCL10)

IPTR International pancreas transplant registry

IRAK Interleukin-1 receptor-associated kinase

IRI Ischemia-reperfusion injury

xxi IV Intravenous

JAK/STAT Janus kinase/signal transducer and activator of transcription

JNK c-Jun N-terminal kinases

LDL Low-density lipoprotein

LFA-1 Lymphocyte function-associated antigen 1

LPS Lipopolysaccharide

MAP Mean arterial pressure

MAP kinases Mitogen-activated protein kinases

MCP1 Monocyte chemotactic protein 1

MGB Minor-groove binding

MHCII Major histocompatibility complex II

MIC2 MHC class I-related protein 2

MICA MHC class I polypeptide-related sequence A

MIF Macrophage migration inhibitory factor

MIP1 Macrophage inflammatory protein 1

MMP-2 Matrix metalloproteinase 2 (Gelatinase A)

MMP-9 Matrix metalloproteinase 9 (Gelatinase B) mPVECs Murine portal vein endothelial cells mRNA Messenger ribonucleic acid

MULT1 Murine ULBP-Like transcript 1

xxii MyD88 Myeloid differentiation factor 88

NA Noradrenaline pressor support

NF-κB Nuclear factor kappa-light-chain-enhancer of activated B cells

NICC Neonatal islet cell clusters

NK cells Natural killer cells

NKG2D Natural killer group 2, member D

NO Nitric oxide

NOD mouse Non-obese diabetic mouse

NOD-like receptor The nucleotide-binding oligomerization domain receptors

OD Optical density

P-DRI Pancreas donor risk index

P-PASS Pre-Procurement pancreas suitability score p5E18VD2/8 AAV plasmid containing a AAV serotype 8 cap sequence

PAECs Porcine aortic endothelial cells

PAK Pancreas after kidney transplant

PAMPS Pathogen-associated molecular patterns

PAR Protease-activated receptor

PAWP Pulmonary wedge pressure

PBS Phosphate-buffered saline

PEEP Positive end-expiratory pressure

PET Paraffin-embedded tissue

xxiii PFCs Perfluorohydrocarbons

PRRs Pattern recognition receptors

PSGL-1 High-affinity counter-receptor for P- and E-selectin

PTA Pancreas transplant alone rAAV2/8 AAV2 vector genome pseudo-serotyped with the type 8 capsid

RAE-1 Retinoic acid early transcript 1

RAGE The receptor for advanced glycation endproducts rHuAPC Recombinant human activated protein C

RIG-I-like receptors (Retinoic acid-inducible gene 1)-like receptors

RIP 1 and 2 Receptor interacting protein kinase 1 and 2

ROS Reactive oxygen species

RPMI Roswell Park Memorial Institute medium

S100B S100 calcium binding protein B

S1P Sphigosine-1-phosphate

SOCS1 Suppressor of cytokine signalling 1

SPK Simultaneous pancreas/kidney transplant ssDNA Single-stranded DNA

T1DM Type-1 diabetes mellitus

T2DM Type-2 diabetes mellitus

T3 Triiodothyronine

T4 Thyroxine

xxiv TAB TGF-β-activated protein kinase 1-binding protein 1beta

TACE Tumor necrosis factor (TNF)-α converting enzyme

TAE buffer Tris-acetate EDTA buffer

TAK TGF-β-activated kinase

TBS Tris Buffered Saline

TBST Tris Buffered Saline+Tween-20

TGF-β Transforming growth factor beta

TIR Toll/interleukin-1 receptor

TLR Toll-like receptor

TNFα Tumor necrosis factor alpha

TOLLIP Toll-interacting protein

TRAF TNF- receptor- associated factor

TRIF TIR-domain-containing adaptor inducing interferon-β

TUNEL Terminal deoxynucleotidyl transferase dUTP nick end labelling

ULBP1 UL16-binding proteins 1

UV Ultra-violet

UW solution University of Wisconsin solution

VCAM-1 Vascular cell adhesion molecule 1

xxv

CHAPTER 1 INTRODUCTION

1 1.1 Deceased donor organ transplantation

For many medical conditions, organ transplantation is the ideal, and sometimes the only, available treatment. In Australia, in 2013, the total number of organ donors (DBD and DCD) was 354. A total of 1110 organs were transplanted into 1053 recipients [1], comprising 607 kidneys, 230 livers, 144 lungs, 72 hearts, 38 pancreata and 4 pancreatic islet grafts. Nonetheless, by the end of 2012, 1605 patients remained on the waiting list.

Some of the procured organs are deemed unsuitable for transplantation, including up to

50% of the retrieved pancreata [2]. Strategies to improve the quality of donated organs are urgently needed so that this scarce resource can be fully utilised.

1.2 Indications for pancreas and islet transplantation

Diabetes Mellitus

Diabetes Mellitus is a disturbance of carbohydrate metabolism induced by relative or absolute deficiency of insulin, or resistance to its actions, with many short-term and long-term complications. Diabetes mellitus and its vast medical, social and financial implications represent a huge and growing burden, affecting 347 million people worldwide [3]. High blood glucose is directly or indirectly responsible for 6% of deaths globally. In 2004, an estimated 3.4 million people died from consequences of diabetes mellitus [4, 5]. Increasing global overweight and obesity presages a global diabetes epidemic, with countless complications [6].

Type 1 diabetes mellitus (T1DM)

The prevalence of type I diabetes mellitus is 0.2-0.5 % worldwide [7]. The onset is usually before age 30. T1DM is considered an autoimmune disease due to increased

2 frequency of auto-antibodies to islet cells. Patients usually have a lean body structure and are prone to ketosis caused by severe lack of insulin production. However, in children and young adults, obesity increases the rate of progression to T1DM in susceptible individuals [6, 8][9, 10]

Type 2 diabetes mellitus (T2DM)

T2DM has a prevalence of 2-4% worldwide and comprises 85-95% of all diabetes [11], the rest being T1DM and gestational diabetes. It is characterized by a combination of resistance to insulin action and inadequate compensatory insulin secretory response.

The body build is usually overweight or obese, and patients are resistant to ketosis with high, normal or low levels of insulin.

Complications of diabetes

Short-term complications of diabetes and its treatment include hypoglycaemia, ketoacidosis, and infections. Hypoglycaemia is one of the most dangerous of these conditions and can result in permanent brain damage and death. Long-term complications (end organ damage) are divided into microvascular complications

(retinopathy, nephropathy, peripheral and autonomic neuropathy) and macrovascular complications (affecting the peripheral, cerebral and cardiac vessels).

Shortcomings of exogenous insulin therapy

T1DM, a fatal disease, was transformed to a treatable condition with the introduction of commercially-produced insulin. Apart from a strict diet and healthy lifestyle, the conventional treatment for T1DM is exogenous insulin. As a result, the acute form of the disease is rarely lethal in those countries where exogenous insulin is accessible [4,

3 12, 13]. However, the main shortcoming of insulin treatment is that it cannot restore the complex balance of normal glucose homeostasis; this balance involves strict minute-to- minute feedback between blood glucose levels and insulin secretion from the islet beta cells, along with other paracrine activities of islets, secreting glucagon and somatostatin. Therefore, despite maintaining the blood glucose level within normal limits, insulin therapy doesn’t react to the constant blood glucose fluctuations [12].

Intensive insulin therapy has been shown to improve long-term complications; however, a 3-fold increase in the risk of serious hypoglycaemic periods causing seizures and coma has been the cost of improved control [13, 14]. Additionally, the current insulin treatment schedules are demanding and difficult to maintain for this life- long condition. Restoration of endogenous insulin production that responds to the physiologic changes in glucose levels would fundamentally change the treatment of diabetes. Currently, whole pancreas transplantation and isolated islet transplantation are the dominant sources of clinical β-cell replacement [14, 15].

Pancreas and islet transplant: milestones

The first clinical transplantation of pancreatic tissue was performed in 1893 by Watson-

Williams and Harsant. They transplanted 3 pieces of a freshly slaughtered sheep’s pancreas into the thigh of a young patient dying from diabetic ketoacidosis. The glucose level improved temporarily, but the xenograft failed within 3 days [14]. In

1921, Best and Banting discovered insulin. Despite records of complications such as hypoglycaemic insulin reactions since 1923, insulin remained the only treatment for

T1DM until the 1970s. Kelly and Lillehei at the University of Minnesota transplanted the first vascularized whole human pancreas in 1966 [16]. The preliminary outcomes and survival rates were extremely poor, but from the 1980s onwards, improvements in

4 immunosuppression, surgical techniques, infection prophylaxis and other areas of management brought pancreas transplantation from an experimental treatment into clinical practice for patients with T1DM [17-19]. Since 1970, >27,000 pancreas transplants in the USA and a further >15,000 pancreas transplants worldwide have been reported to the International Pancreas Transplant Registry (IPTR) [19, 20].

Paul Lacy in 1967 first showed that the pancreas could be digested with collagenase, and that islets could be isolated form the pancreatic digest [21]. Then Ballinger and

Lacy in 1972 corrected diabetes in an experimental model using islet isografts from normal rats implanted under the kidney capsule of streptozotocin-induced diabetic rats

[22]. Preliminary attempts at islet autografting in humans were undertaken in 1980 when a series of 10 patients with painful chronic pancreatitis were subjected to islet autotransplant [23]. Three patients reached insulin independence for 1, 9 and 38 months. The retrieved pancreas was digested with collagenase and the digest was infused into the patient’s portal vein within two-and-a-half hours of recovery.

Largiader et al in 1980 reported a human case of insulin independence with normal glucose levels in an allogenic islet cell transplant recipient 9 months post-transplant

[24]. The subsequent outcomes of intra-portal islet infusion remained poor, contradictory to what was predicted, and despite using conventional immunosuppression, insulin independence at 1 year was less than 10% [14]. In 1992

Pyzdrowski reported that 265,000 islets were sufficient to achieve insulin independence

[25]. A major advance in islet cell transplantation occurred in 2000 when the Edmonton group successfully treated seven T1DM patients following a glucocorticoid-free immunosuppression protocol [26]. The recipients had hypoglycaemia unawareness and

5 received fresh islets from multiple donors infused into their portal vein. The immunosuppression regimen consisted of dacluzimab, sirolimus and low dose tacrolimus. Other specialised centers have been able to reproduce the Edmonton group’s results and achieve high success rates, according to reports from multicentre trials sponsored by the Immune Tolerance Network [20]. Approximately 1,085 patients have received islet transplantation at 40 international sites since the introduction of the

Edmonton Protocol in 2000 (752 allografts, 333 autografts), as documented by the

Collaborative Islet Transplant Registry [27].

Who should undergo pancreas/ islet transplantation?

Selecting patients for pancreas or islet transplantation is a complex matter, and a variety of factors should be acknowledged in decision making. The severity of glycemic complications (hypoglycemia, hyperglycemia, and ketoacidosis) should be carefully evaluated. Pre-existence of long-term complications such as retinopathy, cardiovascular or renal disease can adversely affect transplant outcome [15, 19]. The psychosocial situation of the patient and family needs to be addressed. Severe or longstanding illness imposes a great burden, exhausting the patient and family both physically, mentally and financially [27]. Patients must fully understand and be aware of all the potential risks of the transplant procedure, including those related to the long-term immunosuppressive regimen [17, 19]. Islet or pancreas transplantation is proposed only when all available approaches to satisfactory diabetes management have failed [15]. Finally, it is absolutely crucial for the patient undergoing pancreas or islet transplantation to consider blood glucose control and the absence of hypoglycemia as their primary therapeutic goal rather than insulin independence [15].

6 Pancreas transplants are performed in patients who need insulin because of T1DM, type 2 diabetes mellitus or a total pancreatectomy. Successful pancreas transplant is currently the only definitive long-term treatment for T1DM patients that restores glycaemic control and prevents the incidence of hypoglycaemic attacks [19, 28]. Three categories for pancreas allotransplantation include: simultaneous pancreas- kidney transplantation (SPK) (the most common type), pancreas after kidney transplantation [29] and pancreas transplant alone [30] (mostly in brittle diabetes cases) [15]. Insulin-independence rates of up to 85% at one year after transplantation have been observed [31]. Pancreas transplantation is generally well-tolerated nowadays, with patient survival rates 95% to 98% at one year in all categories (SPK, PTA, and PAK) and graft survival rates of at one year over 80% [15, 19].

SPK is defined as the “gold standard” treatment in T1DM patients with end-stage renal disease, in the absence of contra-indications. Malignancies, chronic infections, insufficient compliance, and severe cardiovascular conditions are contraindications for SPK transplantation [32].

Indications for pancreas transplant alone [30] have been developed by the American Diabetes Association: “In the absence of indications for kidney transplantation, pancreas transplantation should only be considered a therapy in patients who exhibit these three criteria: 1) a history of frequent, acute, and severe metabolic complications (hypoglycemia, hyperglycemia, ketoacidosis) requiring medical attention; 2) clinical and emotional problems with exogenous insulin therapy that are so severe as to be incapacitating; and 3) consistent failure of insulin-based management to prevent acute complications [15].” However, the risks of invasive surgery, along with the lifelong need for immunosuppressive therapy and its side effects [17] have to be considered alongside the medical risks of living with DM.

7 The indications for islet transplantation were established based on international agreement, and are very similar to those for pancreas transplant [15, 20]. In the table below [15] inclusion and exclusion criteria for islet transplant are listed (Table 1.1).

8 Relative Inclusion criteria Exclusion criteria contraindications Diabetes mellitus type 1 Age < 18 year Age > 70 year (or complete insulin deficiency after Diabetes duration < 5 year pancreatectomy) Repeated severe Remaining C-peptide secretion Body mass index > hypoglycemic episodes, (stimulated C-peptide > 0.5 ng/mL) 28 kg/m² metabolic lability (“brittle diabetes”) despite optimal Malignancy (except for Relevant insulin therapy nonmelanomatous skin cancers) resistance (insulin requirement > 0.8 IE/kg body weight)

Highly impaired quality of Untreated proliferative retinopathy life

Rapid progress of Portal hypertension complications

Creatinine clearance > Prohibitive cardiovascular risk 40 mL/min (myocardial infarction within last 6 month, ejection fraction < 20%, Compliance with medical uncontrolled hypertension, management significant noncorrectable coronary artery disease)

Active infections including hepatitis B/C, HIV, tuberculosis Alcohol/drug abuse Pregnancy, nursing, intention for pregnancy, insufficient contraception

Table 1.1 Indication (inclusion and exclusion criteria) for pancreatic islet transplantation. Ludwig, B., et al., Current Diabetes Reports, 2010 [15].

Uncontrolled hypoglycemic episodes are the most common indication for islet transplantation alone. Islet after kidney transplantation is limited to the T1DM patients

9 with end-stage renal disease who have already undergone kidney transplantation, or who have rejected the pancreas after SPK transplantation. The success rate of islet transplantation alone depends on rigorous patient selection [15, 19, 33].

1.3 Organ quality and the outcomes of pancreas and islet transplantation

1.3.1 Donor characteristics

The outcome of pancreas or islet transplant is highly dependent on donor selection [34].

Currently, no “donor test” can reliably foretell the outcome of transplantation; however scoring systems have been introduced to categorize potential donors [35]. An ideal donor in terms of demographic factors is described as a donation after brain death

(DBD) donor, between 10 and 40 years of age, with a BMI less than 27.5 kg/m2 and a cause of brain death which is traumatic or other than cerebrovascular [36]. Reports from International Pancreas Transplant Registry also confirm the importance of donor characteristics. Due to the increasing number of extended criteria donors and nonrecovery (organs declined after procurement) rates, the Eurotransplant Pancreas

Advisory Committee introduced the Pre-Procurement-Pancreas-Suitability-Score (P-

PASS), are defined according to donor parameters at the time of reporting. These parameters are listed as age, BMI, ICU stay, cardiac arrest, sodium, amylase, lipase, and catecholamine dose. Total points range from 9 to 28 [37]. Retrospective studies show a 3-fold increase in the chance of nonrecovery, with a score greater than 17 [2].

The Pancreas-Donor-Risk-Index (P-DRI) was formulated by reviewing the Scientific

Registry of Transplant Recipients to determine which factors influencing organ quality are predictive for 1-year graft survival [38]. Variables such as sex, age, race, BMI, height, cause of death, serum creatinine and donation after circulatory death (DCD) are

10 included, as well as preservation time. The median donor has a P-DRI of 1 and is defined as male, 28 years of age, nonblack, BMI of 24kg/m2, with a height of 173 cm,

DBD with a cause of death other than cerebrovascular, serum creatinine less than 2.5 mg/dl and 12hr cold ischemia time (CIT). Age above 45 years, BMI higher than 30 kg/m2 and DCD increase the P-DRI and contribute to higher graft failure risk [39], although successful organ procurement from these groups has been reported. Pancreatic grafts from DCD donors – currently limited to controlled DCD – have also dramatically increased [40]. Analysis of 1000 pancreatic grafts from the UK demonstrates no difference between DCD and DBD in the 1-year patient and graft survival [41]. This result was confirmed by other centres. Madison, Wisconsin reported 1, 3 and 10-year graft survival rates from DCD donors similar to those from brain-dead donors [42]. It is of particular importance to avoid accumulation of risk factors in one donor, for example a longer organ removal time should be compensated by keeping the cold ischemia time as short as possible [36, 43].

These factors can predict the outcomes of islet procurement as well as pancreas transplant. The results of islet isolation can be altered by additional variables. Islets from paediatric donors are reportedly more difficult to separate from the exocrine components of the pancreas without excessive fragmentation [20]. Higher body mass index (BMI) of the donor results in greater islet recovery [44] and pancreatic insulin content is increased in donors with an increased body surface area. A damaged-looking pancreas has also been correlated with reduced islet isolation rate and graft survival

[27].

11 1.3.2 Pancreas retrieval and preservation

Pancreas excision requires advanced surgical techniques to keep the organ intact and preserve the capsular integrity, which is important for subsequent enzymatic digestion [27]. Pancreas procurement in DBD donors is mainly performed using either of two methods: rapid dissection in the cold after flush-out [45], or warm dissection before cold perfusion [46]. During retrieval and before perfusion the organ must be evaluated to decide whether it is transplantable or not [37]. Minimal handling is advised to prevent pancreatitis and local injury prior to aortic cross- clamping [47]. After dissection the pancreas is cooled rapidly with intracellular preservation solutions and topical ice, and then transported for whole organ transplant or islet isolation. Both standard solutions, University of Wisconsin (UW) and histidine-tryptophan-ketoglutarate (HTK) are equally suitable to protect the pancreas during transport for islet isolation. However UW has been shown to be superior to HTK for whole pancreas preservation [48, 49]. Ideally, cold ischemia time should be kept below 6-8 hours [27, 35, 45].

1.3.3 Ischemia-reperfusion injury

The pancreas is highly sensitive to ischemia [50]. So far, no reliable early markers for pancreatic ischemia-reperfusion injury (IRI) have been identified. For instance, serum amylase, lipase and C-reactive protein levels do not reflect the severity of pancreatic inflammation until several days after transplantation [37]. The major feature of pancreatic IRI is microcirculatory impairment and venous thrombosis in the graft accounts for the majority of early graft failures [51-53]. Therefore, improving the posttransplant blood circulation in the graft has been attempted through administration

12 of nitric oxide [54] or anticoagulants [55] or by preventing neutrophil adhesion [56], although none of these approaches has been validated by prospective randomized trials.

Remote ischaemic preconditioning, utilising short occlusion of the lower extremity blood flow prior to organ retrieval, has demonstrated promising results in decreasing

IRI in pancreas grafts [57, 58]. A current randomized controlled trial is now evaluating this approach [59].

1.3.4 Islet isolation

The total number of islets in a human pancreas has been estimated to vary between 3.6–

14.8 million, 1.3-2% of the whole pancreas. Islets vary in size; 75% of the total islet volume in humans comprises islets ≥ 100 µm, which constitute less than 20% of the total islet number. Most centres with an active clinical islet cell transplant programme obtain 300–600,000 islet equivalents (IEQ)/pancreas after isolation, which is about

50% of the islet tissue content [20, 27].

After retrieval, the organ is transported to the islet isolation facility [27]. Islets are particularly susceptible to irreversible ischemic damage. To minimise hypoxia, continuously oxygenated perfluorohydrocarbons (PFCs), known for their high oxygen- solubility coefficients, are being combined with standard University of Wisconsin preservation solution for pancreas which are procured for islet isolation and, this seemingly improves islet cell function [20].

Briefly, after retrieval, the pancreas is catheterized and collagenase enzyme is infused into the duct to start the digestion process [27]. For human islet isolation, the pancreas

13 is cut into pieces and transferred to a Ricordi chamber along with the enzymatic solution, and incubated at 37 °C until the islets are freed from the surrounding exocrine tissue, usually for 15-30 min [60-62]. For large or small animals, the organ pieces are incubated with enzyme at 37 °C in a water bath [63]. After digestion and washing, the digest is purified using either discontinous Ficoll density-gradient centrifugation [20] or continuous iodixanol density-gradient centrifugation [64, 65]. The latter has been proven to decrease inflammation and also improve efficiency of islet isolation [66]. The duration of centrifugation is typically 10 min. Larger islets may be damaged during this process, but the medium-sized majority are recovered. Islets are then transferred to culture, and maintained for 24 hr before clinical transplantation; or used for further research purposes. Culture before use increases islet purification from exocrine tissue, may reduce inflammation, boost islet recovery and allow islet or donor manipulation

[67, 68].

14

Figure 1.1 Factors contributing to loss of islet grafts. To achieve long-term graft function, many aspects of the islet transplant procedure must be optimised, from donor management to islet intraportal infusion. Islets in the photo are stained with dithizone. Shapiro, J., et al., The review of diabetic studies, 2012 [27].

Instant Blood-Mediated Inflammatory Reaction

Upon infusion into the portal vein, the islets provoke a non-specific inflammatory response called instant blood-mediated inflammatory reaction (IBMIR). Direct contact of islets with circulating blood initiates the complement and coagulation cascades [69].

Tissue factor, along with an array of proinflammatory mediators, including monocyte chemotactic protein (MCP)-1, interleukin-8, and macrophage migration inhibitory factor (MIF) expressed by isolated islets, activates platelets and complement. Islets become trapped within clot and are infiltrated with neutrophils and macrophages within

15 min of infusion [70].

15 The clinical success of allotransplantation is inversely correlated with the level of proinflammatory mediators expression by islets [71, 72]. Suppression of tissue factor or thrombin formation prevents islet damage in vitro [73]. However, in vivo IBMIR is not completely abrogated by anticoagulation and can cause portal thrombosis [70]. IBMIR destroys most of the infused graft and less than 50% of the islets seed in the liver after

2 hr [72]. Various clinical reports show that only a small fraction of total infused islets engraft and function, causing inferior islet transplant outcomes. Reduction of the inflammatory status of islets at the time of intarportal infusion is thus an important therapeutic goal.

1.4 Clinical outcomes

1.4.1 What constitutes a successful transplant for the treatment of type 1 diabetes mellitus?

A successful pancreas transplant is the only available long-term treatment for T1DM that can restore normoglycaemia, prevent hypoglycaemia, arrest progression of, or in some instances cure, the secondary complications of diabetes [19]. Developments in whole pancreas transplantation techniques, organ retrieval and anti-rejection strategies over the past 4 decades have continually improved the outcome of pancreas transplantation. Unadjusted patient survival rates have increased significantly to >96% at 1 year post-transplant and >80% at 5 years in the three categories (SPK, PTA, and

PAK) over the last decade [12, 17, 19, 74]. Using insulin independence as the marker for graft survival, unadjusted graft survival rates at 1 year were 89% (SPK), 86% [29] and 82% [30]. 5 year survival rates are 71% (SPK), 65% [29] and 58% [12, 19, 30, 74].

The estimated half-life of a pancreas graft is now 7–14 years.

16 Analysis of transplant outcomes in a large cohort of patients indicates that secondary diabetic complications affecting many organs and systems have been improved significantly after transplantation [19]. Pancreas transplantation can prolong life [28,

75], reverse established nephropathy [76] and improve quality of life [77].

Improvements in diabetic nephropathy, neuropathy, gastroparesis, retinopathy, microvascular and macrovascular disease, cardiac function and sexual function have generally been observed [15, 19]. Better control of the levels of HbA1c compared to intensive insulin therapy was also reported [78]. However, pancreas transplantation requires invasive surgery with considerable attendant morbidity and life-long immunosuppression [14, 19].

Many groups believe that to gauge the success of an islet transplant, resolution of life- threatening hypoglycaemia and HbA1c control are more appropriate outcome measures than simply measuring insulin-independence [12].

Islet transplantation is yet to be perfected; and until achievement of treatment from a single donor becomes the norm, it cannot replace pancreas transplantation [27].

However, fundamental developments such as refinements to donor selection criteria, isolation method and immunosuppression protocols have substanitially improved the clinical outcomes of islet transplantation.

Recipients in 10 programs demonstrated graft survival and insulin independence comparable to multi-donor rates with single-donor islet infusion [65, 79-87]. The current “Edmonton protocol” results in 1-year insulin independence rates of 80%-85%

[27]. Insulin independence rates of up to 50-70% at 5 years are now being reported

17 from 7 centres following different protocols [65, 81, 88, 89]. O’Connell et al. reported the following results on clinical islet transplantation in Australia from 2012 to 2013

[90]: “islet transplantation can improve outcomes in patients with severe hypoglycemia and metabolic instability. Eighty two percent of recipients achieved the primary endpoint of absence of hypoglycemia and HbA1c <7%. Interestingly, patients who were C-peptide positive but required supplemental insulin had excellent control of their diabetes, with equivalent HbA1c to that achieved in patients who were insulin independent with similar improvement in hypoglycemia. These outcomes were a marked improvement from our originally reported single centre study where 50% of recipients were insulin independent and all had returned to insulin within 24 months”.

The beneficial impact of euglycaemia is in accordance with the results from pancreas transplantation [27]. Islet transplantation has been shown to halt micro- and macroangiopathic changes, with or without insulin independence [91-93]. Islet transplantation slows progression of retinopathy, and improved nerve conduction velocity is observed [94, 95]. Further, islet grafting has been shown to arrest progression of diabetic nephropathy [95] and improve vascular function [96].

Nonetheless, without controlled studies to prove the effects of the islet transplantation on secondary complications, no definitive conclusions can be established [14, 15].

1.4.2 Current challenges

Pancreas and islet transplant programmes have expanded vastly over the past two decades as a successful treatment for T1DM. Nevertheless, there are a number of challenges that should be addressed to render these treatments more accessible and

18 efficient. Overall shortage of donor pancreata and the immunologic factors relating to both allo- and autoimmunity are two main issues.

Organ shortage

Scarcity of donated human pancreata is limiting potential transplantation. Donor shortage is exacerbated by underutilization [27, 36, 97]. Identifying potential donors

[36], improving organ retrieval rates [98] , pancreas procurement from donation-after- cardiac-death (DCD) donors [37, 51, 98], xenotransplantation [37, 99] and derivation of insulin-secreting tissue form stem cells [100, 101] are all strategies which could improve the availability of pancreas and islet grafts.

Alternative source: xenotransplantation

Immense organ shortage necessitates the development of replaceable sources of tissue for transplantation. With significant ethical, logistic and safety barriers surrounding organ procurement from nonhuman primates, the prospect of utilizing islets from pigs represents one possible alternative. Porcine insulin has long been used in T1DM before production of recombinant human insulin; structurally they differ by only one amino acid. Although the first scientific attempt at pig islet transplant to a human recipient in

1994 did not restore normoglycaemia, the porcine c-peptide was detectable for more than 300 days [102].

Wild-type pig solid organ transplantation (rather than islets or cells) into nonhuman primates leads to hyperacute rejection. The naturally-occuring anti-Gal antibodies of human and old-world primate recipients detect the galactose-α1,3-galactose (Gal) oligosaccharide which is expressed by the porcine vascular endothelium, and

19 immediate activation of complement destroys the graft [103]. Foetal and neonatal porcine islets also express Gal, which decreases during maturation. Xenotransplant of wild-type porcine islets of any age into the circulation triggers a severe instant blood mediated immune reaction (IBMIR), destroying the graft [104]. The IBMIR response to islets from GAL knock-out pig appears to be milder [105]. Murine transgenic islets over-expressing human CD39, an ectonucleotidase that degrades the platelet agonists

ATP; also provoked reduced IBMIR and coagulation [106]. Following IBMIR, the adaptive immune response to islet xenografts is T cell–mediated. Strategies for overcoming the primary obstacles towards islet xenotransplantation include genetic engineering of the donor pigs, development of immunosuppressive regimens effective against xenoimmunity, immunomodulation/immunoprotection (such as co- transplantation with other donor/recipient cells) and the use of alternative anatomical sites for islet transplant [99, 107].

Alloimmunity

The current immunosuppressive regimens are successfully controlling allorejection of pancreas or islet grafts, and are also at least partly effective in preventing recurrence of autoimmunity. The most encouraging graft survival rates are from regimens with tacrolimus and mycophenolate maintenance therapy [27, 97].

This acceptance comes with the great cost of life-long immunosuppression and its many side effects. If pancreas or islet transplantation is to be the standard treatment for children, or early after diabetes is recognised, and/or xenotransplants are going to become reality, immune rejection most probably needs to be addressed by means other than pharmacologic immunosuppression.

20 Reducing immunogenicity of the organ is an essential step towards optimized engraftment. Donor treatment strategies [51, 98, 108] immunoalteration through the expression of protective genes [109, 110] [111], and inflammation reduction [112,

113], combination of these strategies with therapies aimed at tolerance induction in recipients [114] [115, 116] are areas where the graft can be modulated at multiple levels to minimise the immune reaction. Understanding the physiology of the instant blood-mediated immune reaction (IBMIR) [117] after infusion of islets into the portal vein, and the mechanisms underlying recurrent autoimmunity against β-cells will also facilitate the treatment of T1DM and secondary complications by islet or pancreas transplantation [110] [12].

1.5 The innate immune system

Also called natural or native immunity, innate immunity is present in healthy individuals and the first system responding to infection or injury. The major components of innate immunity are epithelial barriers, phagocytes (macrophages and neutrophils), dendritic cells, natural killer cells, the complement system and various cytokines and chemokines. Pattern recognition receptors permit the recognition of broad families of microbial ligands or endogenous “danger signals” released from damaged tissues. In addition to restriction and eradication of infection, innate immunity stimulates adaptive responses that will follow, and plays an important role in the initiation of tissue repair [118].

Immunosuppressive regimens have made organ transplantation a reliable effective treatment for many end-stage diseases. Allotransplant rejection was initially thought to

21 be entirely mediated by adaptive immunity, yet this does not explain observations such as the improved outcome of renal transplantation from living unrelated donors compared to brain dead donors, even in the absence of HLA-matching. Based on

Matzinger’s “danger model” and reports describing immune activation of organs by brain-death and ischemia-reperfusion injury, the role of innate immunity in pre and post-transplant graft failure was recognised [119, 120]. Early events in brain death activate innate immunity and increase the immunogenicity of organs through pattern recognition receptor (PRR) signalling, upregulation of proinflammatory and adhesion molecules and endothelial/complement/coagulation activation.

1.5.1 Components of innate immunity contributing to the outcomes of transplantation

Pattern recognition receptors

Pattern recognition receptors (PRRs) are essential to the function of the innate immune system and include C-type lectin receptors, NOD-like receptors, RIG-like receptors,

Toll-like receptors and RAGE [121]. Among these, Toll-like receptors and RAGE have been implicated in transplant-associated immunity.

A correlation between factors known to be related to organ quality, including brain death, ischemia and donor characteristics, and activation of innate immunity and subsequently adaptive immunity is established [122]. According to the “danger model” [119], various non-nspecific threats arising from pre- and peritransplant events

(ischemia, brain death, or surgical manipulations) can trigger the innate immune system

[123]. The molecular mechanisms linking these events with immune reaction are yet to

22 be fully understood, but numerous experimental studies provide solid evidence for the involvement of pattern recognition receptors [124].

Toll-like Receptors

Toll-like receptors (TLRs) are present on the cell surface and in the intracellular compartment where microbial components are encountered. TLRs are highly conserved across species from Drosophila to humans. Most species posess 10-15 TLRs which are expressed by a variety of cell types including epithelial cells, endothelial cells, leukocytes, dendritic cells and the parenchymal cells of many organs including pancreatic islet cells. They recognise and are activated by families of molecules called pathogen-associated molecular patterns (PAMPs) and danger associated molecular patterns (DAMPs) [125, 126] (Figure 1.2). DAMPs, alarmins or danger signals are non- infectious endogenous molecules released by injured tissue [127], namely heat shock protein (HSPs), high-mobility group box 1 (HMGB1), hyaloronan, fibrinogen, fibronectin, heparan sulphate, surfactant protein A and minimally modified low-density lipoprotein (LDL) [127-129]. When engaged by their ligands, TLRs recruit adaptor molecules for signal propagation through their Toll/interleukin-1 receptor (TIR) domain. Four adaptor molecules are known, amongst these myeloid differentiation factor 88 (MyD88) and TIR-domain-containing adaptor inducing interferon-β (TRIF) predominate. Accordingly, TLRs signal through two MyD88-dependant and – independent pathways [130]. In humans, stimulation of all TLRs except TLR3 results in signal transduction through the MyD88 pathway. Recruitment of MyD88 activates

IL-1 receptor-associated kinases (IRAKs) 1, 2 and 4 as well as tumor necrosis factor

(TNF)-receptor-associated factor 6 (TRAF6). Then, the IRAK-1/TRAF6 complex separates from the receptor and binds to TGF-β-activated kinase 1 (TAK1) and TAK1-

23 binding proteins, TAB1 and TAB2, at the cell or compartment membrane. While

IRAK-1 is degraded at the membrane, the complex of TRAF6, TAK1, TAB1, and

TAB2 is released and forms a large complex with other cytosolic proteins. This complex catalyses the synthesis of a Lys 63-linked polyubiquitin chain of TRAF6 and subsequently induces TRAF6-mediated activation of TAK1 finally causing NF-κB translocation to the nucleus, where it induces the downstream expression of genes encoding cytokines, chemokines, co-stimulatory and adhesion molecules, cellular membrane proteins and others [130, 131]. TLRs can activate both innate and adaptive immunity based on the pathways they signal through, and thus are among the key molecular linking the two systems [132]. Depending on the cell type, downstream events can differ (Figures 1.3-1.5).

24

Figure 1.2 Immune activation by TLR signalling. Donor brain death and ischemia as acute injuries release DAMPs that initiate TLR signalling on antigen- presenting and parenchymal cells. TLR activation results in production of cytokines, chemokines and growth factors, dendritic cell maturation and the migration of activated dendritic cells to the T cell zone of draining lymph nodes. These events stimulate different cell types, results in leukocyte recruitment and activation of the alloimmune response. Graft immune activation induces further alloimmunity and promotion of chronic damage and dysfunction. Leventhal, J.S., et al., Kidney Int, 2012 [133].

25

Figure 1.3 Downstream effects of TLR signalling on parenchymal cells. Engagement of TLR4 on parenchymal cells activates the MyD88-dependant pathway, leading to production of chemokines and recruitment of leukocytes. Expression of stress ligands which bind to the activating receptor NKG2D increase the susceptibility of the parenchymal cells to killing by CD8+ T cells and NK cells.

26

Figure 1.4 Activation of the endothelium by TLRs and cytokines. Signalling through TLRs on the endothelium and on adjacent leucocytes leads to the production of chemokines which bind to the endothelial surface. Local expression of chemokines attracts leukocytes, and induces the production of cytokines such as TNFα. Adhesion molecules including P/E- selectins and intercellular adhesion molecule-1 (ICAM-1) are expressed in response to cytokine stimulation of the endothelium, and facilitate leukocyte arrest, transmigration and tissue infiltration.

27

Figure 1.5 Monocyte and DC maturation and migration in response to TLR signalling. TLR engagement on dendritic cells initiates adaptive immunity by upregulating MHC class II and co-stimulatory molecules which are important for antigen presentation to naïve T cells , and by modulating the expression of chemokine receptors (decreases in CCR6 and increase in CCR7), thereby directing the DC to the T cell regions of peripheral lymph nodes where antigen presentation can occur.

RAGE

The Receptor for Advanced Glycation Endproducts (RAGE) is a member of the immunoglobulin superfamily, encoded in the Class III region of the major histocompatability complex. This receptor consists of one V type domain, two C type domains, a transmembrane domain, and a cytoplasmic tail. The V domain has two N- glycosylation sites where most of the extracellular ligands bind. The cytoplasmic tail is critical for intracellular signalling [134].

28 This multilagand receptor was initially found to be activated by advanced glycation endproducts (AGEs), but other ligands including damage-associated molecular patterns

(DAMPs) have been discovered subsequently [135]. Based on its multiligand recognition, RAGE is now categorised as a pattern-recognition receptor (PRR) [134].

Many tissue and cell types constitutively express RAGE at low levels; it is found in all transplantable organs, but mature lung type-I pneumocytes express it at significantly higher levels than other resting cell types. RAGE is related to the pathogenesis of many conditions such as cardiovascular disease, diabetic vascular disease, neuronal degeneration and cancer [136, 137] and is upregulated at the site of ligand accumulation.

Ligands for RAGE are drawn from several large families including the High Mobility

Group family proteins such as the prototypic HMGB1/amphoterin, members of the

S100/calgranulin protein family, matrix proteins such as Collagen I and IV, amyloid- beta peptide, and advanced glycation end products such as carboxymethyllysine (CML-

AGE) [138] [139].

RAGE shares binding affinity to HMGB1with TLRs 2, and 9. The signal transduced following ligation of RAGE by S100/calgranulins or HMGB1 varies depending on the cell type. The key mediators propagating the signal are erk1/2 (p44/p42) MAP kinases. p38 and SAPK/JNK MAP kinases, rho GTPases, phosphoinositol-3 kinase and the

JAK/STAT pathway, resulting in nuclear translocation of NF-κB and downstream gene upregulation of pro-inflammatory and adhesion molecules [134, 140, 141] [142]. The unique effects of RAGE-mediated NF-κB activation can be attributed to its prolonged

29 time course [138], therefore RAGE can act as a master switch converting a transient proinflammatory response into sustained cellular dysfunction, whereby this sustained

NF-κB activation mediates the crosstalk between innate and adaptive immunity. In addition to the NF-κB pathway, RAGE acts as an adhesion receptor for leukocyte recruitment, mainly through the β2-integrin Mac-1 [143].

Although the role of RAGE in adaptive immunity is less well elucidated compared to that in innate immunity, significant findings have been reported. RAGE-deficient T cells in mouse showed impaired proliferative responses in vitro to nominal and alloantigens, in parallel with decreased production of IFN-gamma and IL-2. Dendritic cells from RAGE-deficient mice did not show impairment in Ag presentation, maturation, or migratory capacities [144]; although other studies have suggested that

RAGE and HMGB1 are necessary for animal and human DC maturation and homing to lymph nodes [145].

RAGE is expressed as a variety of isoforms. Major forms are full-length, membrane- bound forms (fl-RAGE or mRAGE) and a soluble splice variant including the extracellular domain alone named endogenous secretory RAGE (esRAGE). Soluble

RAGE (sRAGE) may also be produced by enzymatic cleavage of the transmembrane receptor by ADAM-10 or MMP-9 [146]. Further isoforms with no binding activity lack the N-terminus and the dominant negative form does not have the cytoplasmic tail

(DN-RAGE) [137]. The various RAGE isoforms have different regulatory effects. esRAGE, for instance, is a naturally-occurring decoy receptor that binds to HMGB1,

S100B as well as AGE and other RAGE ligands. esRAGE is known to mediate

HMGB1 clearance from the circulation in healthy humans [147]. DN-RAGE

30 suppresses RAGE-mediated signalling, even in cells expressing the full-length form

[148]. Notably, sRAGE and esRAGE both can intercept the RAGE-ligand interaction and thus suppress RAGE response and expression [134]. esRAGE and RAGE can also form heterodimers with flRAGE at the cell surface, preventing signal transduction which requires homodimerisation and association of the two cytoplasmic tails [149].

Danger-associated molecular patterns, endogenous danger signals or alarmins

Danger-associated molecular patterns (DAMPs) are endogenous mediators released from injured cells under non-infectious conditions; they provoke leukocyte chemotaxis and infiltration and antigen-presenting cell activation as part of innate immune signalling. Some such as HMGB1 or HSP72 are stored in distinct anatomical compartments within cells, whilst others, including low molecular weight hyaluronic acid, fibronectin and byglycan, result from breakdown of extracellular matrix components. They can be found in leukocytes, epithelial and endothelial cells, and many other cell types. These immunostimulants serve as early warning signals to activate the innate and adaptive immune systems [150].

HMGB1

One of the best-characterized endogenous alarmins is high mobility group box 1

(HMGB1). It is a DNA-binding protein which modulates transcription and chromatin modelling and is ubiquitously found in the nuclei of all vertebrate cells. It is released upon cell injury or necrosis or from activated inflammatory cells, and acts as an extracellular immune-signalling molecule [151]. Proinflammatory activity of HMGB1 through TLR4/TNFα signalling depends on the presence of cysteine at position 106; during apoptosis, this residue is oxidised, causing reduced HMGB1 binding affinity

31 with TLR4; hence only HMGB1 released during necrotic death and not apoptosis has proinflammatory capabilities [151].

Other molecules such as ssDNA, LPS, IL-1β and nucleosomes join HMGB1 to shape specific complexes which are recognised by TLR9, TLR4, IL-1R and TLR2 respectively [152, 153]. The inflammatory response provoked by complex structures is significantly stronger compared to that of uncomplexed alarmins [154]. HMGB1 is known to bind to RAGE and induce inflammatory events. For instance, binding of

HMGB1 to CpG DNA stimulates augmented cytokine production through joint RAGE and TLR9 signalling [155]. The serum level of HMGB1 increases in any severe inflammation such as tissue damage or sepsis and can be used as a predictor of morbidity in many conditions. Production of IL-1, IL-6, TNFα and macrophage inflammatory proteins (MIP1α and MIP1β) is directly mediated by HMGB1 [156].

S100B

S100 calcium binding protein B or S100B belongs to the family of S100/calgranulins and acts as a ligand for RAGE. It is found at various levels in the cytoplasm and nucleus of many cells including dendritic cells, langerhans cells, and mononuclear phagocytes, myocardium, renal epithelium, adipocytes, melanocytes and chondrocytes; however astroglia of the CNS express it most abundantly [157]. Different organs can release S100B into circulation following damage and it is cleared by the kidney [158].

Acute brain injury and brain death in particular have shown to increase S100B levels in blood, thus it is being employed as a reliable biochemical marker to measure the severity and outcome of CNS damage [159, 160].

32 Intracellular S100B is involved in regulation of a number of cellular processes such as cell cycle progression, differentiation and cytoskeletal structure; nonetheless it can impose detrimental effects at high level [160, 161]. High extracellular levels of S100B can stimulate endothelial, vascular smooth muscle cells, monocytes, macrophages and

T cells through RAGE and induce the production of downstream proinflammatory mediators [143, 162, 163], which subsequently causes migration of inflammatory cells to the stressed tissue as well as apoptosis [164]. Downstream events of RAGE signalling by S100B also include apoptosis induction by elevation of reactive oxygen species, cytochrome C release and activation of the caspase cascade [161]. Extensive studies have demonstrated the release of S100B from the injured CNS; however the source of the circulating S100B in brain dead individuals may include other organs undergoing hypoperfusion and damage due to the haemodynamic sequalae of brain death.

Cells involved in the innate immunity

The innate response to injury is executed by a variety of leukocyte subsets. Neutrophils, macrophages, natural killer cells (NK cells) and some T cells are the principal leukocytes migrating rapidly to the site of injury.

Neutrophils are the cellular defence frontline of the innate immune response to foreign bodies or injury. They are recruited from the circulation or bone marrow reserve in response to chemoattractants. Interleukin-8, IFN-γ, C5a and leukotriene are the best known chemoattractants; E/P-selectin, intercellular adhesion molecule 1 (ICAM1) and vascular cell adhesion molecule 1 (VCAM1) expressed on injured vascular endothelium are the adhesion molecules that neutrophils bind to, primarily in the

33 postcapillary venules. At the site of injury, neutrophil properties include phagocytosis, production of reactive oxygen species (ROS), cytokine secretion, leukotriene synthesis, degranulation and bactericidal activity [165-167]. Neutrophils can directly phagocytose intruding cells, or detect “danger signals” including PAMPs or DAMPs via their pattern recognition receptors (PRRs) such as toll-like receptors (TLRs). Stimulation of TLRs initiates downstream immune reactions [168].

When circulating monocytes migrate to the site of inflammation, they differentiate to macrophages. These phagocytic cells play an important role in removal of pathogens and cellular debris and display a number of functions including antigen internalisation and presentation, cytokine production and stimulation of other immune cells [169].

Phagocytosis is triggered when macrophages encounter opsonised microbes or their

PRR recognise microbe–associated molecular patterns (MAMPs). The remnants of ingested pathogens are presented as peptide antigens in the context of major histocompatibility complex II (MHC II), inducing T cell activation, and thus contributing to the development of an adaptive immune response. Macrophages can also secrete proinflammatory cytokines and other soluble molecules when directly stimulated by pathogens- these include TNF, IL-1, IL-6, IL-12, nitric oxide and indoleamine dioxygenase [170, 171].

Natural killer (NK) cells are important effector lymphocytes which recognise

“foreignness” by detecting the absence of self-MHC I, in association with activating ligands induced by infection or malignant transformation.

34 These activating ligands include UL16-binding proteins 1 (ULBP1), MHC class I polypeptide-related sequence A (MICA) and MHC class I polypeptide-related sequence

B (MICB) in humans, MHC class I-related protein 2 (MIC2) and porcine ULBP1 in pigs and retinoic acid early transcript 1 (RAE-1), murine ULBP-Like transcript 1

(MULT1) and histocompatibility 60 (H60) in mice. In common with macrophages, NK cells express Fc receptors including CD16 and CD 32 which trigger their activation and causing killing of target cells that have been coated with antibody (antibody-dependant cellular cytotoxicity). Target cells are destroyed by NK cells in two ways. The first is via the release of proteins called perforins, which form pores, so proteolytic enzymes called granzymes, pass into the target and induce apoptosis [172]. Alternatively the NK cell’s Fas ligand (FasL) binds to Fas on the target cell ultimately causing the activation of caspases leading to apoptosis [173].

Additionally, NK cells secrete a variety of mediators, including proinflammatory cytokines, chemokines, colony-stimulating factors, and other molecules functioning as regulators of body defenses. NK cells regulate the activity of other immune cells through cytokine production and cell-cell contact [174].

T cells are mainly considered as components of adaptive immunity; however both

CD4+ and CD8+ T cells can amplify the activity of the innate immune system. CD4+ T cells mainly release proinflammatory cytokines, whereas CD8+ T cells, also known as cytotoxic T cells (CTL), can destroy targets by mechanisms similar to those of NK cells. The cytokines released by CD4+ T cells such as IFN-γ, IL-2 and IL-15 induce the recruitment, activation and proliferation of cytotoxic T cells, NK cells and macrophages [175].

35

Dendritic cells (DCs) are critical for the presentation of cognate antigens to naïve CD4

T cells and initiation of adaptive immune response. However, dendritic cell activation is dependent upon the presence of non-antigen-specific danger signals which are required to facilitate and amplify the subsequent adaptive response. Hypoxia and oxidative stress activate dendritic cell differentiation and maturation, DCs express hypoxia inducible factor-1 (HIF-1 ) and other downstream target genes, such as vascular endothelial growth factor or glucose transporter-1, along with CD40, CD80,

CD86 and increased levels of MHC class II which initiate antigen presentation and T cell activation [176].

1.5.2 Inhibitors of innate immune signalling

A20

A20 is one the endogenous negative regulators of pattern recognition receptor (PRR) signalling. Along with MyD88s (the short form of MyD88), suppressor of cytokine signalling 1 (SOCS1), toll-interacting protein (TOLLIP) and cylindromatosis (CYLD), these suppress over-or unnecessary stimulation of PRRs including TLRs [177]. A20 was first discovered in 1990 as a zinc finger protein that is strongly induced by TNFα, so called TNFAIP3 (tumor necrosis factor, alpha-induced protein 3), and is capable of inhibition of TNF-mediated NF-κB activation [178]. A20 is also crucial for regulation of TNF-independent inflammatory signalling [179]. A20 is expressed by many cell types, including β cells, emphasizing its potential role in regulation of TLR signalling within pancreatic islets. The nuclear NF-κB directly induces A20 when it binds to two

κB binding sites in the A20 promoter, increasing transcription re-initiation rates and subsequent total levels of A20 mRNA [180]. Therefore, in general, by inhibiting NF-

36 κB activation, A20 in turn self-regulates its expression. After production, this cysteine de-ubiquitinating protease exerts anti-inflammatory effects in several pathways, mainly by de-ubiquitination of central molecules such as TRAF6 in TLR4/IL-1 and IL-

17 signalling, RIP1 in TNFα and RIP2 in NOD2 pathways [177]. Both TLR4 ligands,

LPS and TNF, can induce A20 expression and both MyD88-dependent and MyD88- independent TLR-signalling pathways are inhibited by A20 activity [181].

Macrophages from A20 deficient mice produce considerably higher levels of proinflammatory cytokines when stimulated with TLR2, TLR3 and TLR9 ligands [182]

[179].

A20-deficient mice die prematurely due to spontaneous multi-organ inflammation and cachexia, they exhibit defects in resolving acute inflammatory responses and are unable to terminate TNF-induced IKK and NF-κB activation; indicating the importance of the negative regulatory effect of A20 to whole organism homeostasis.

Studying mouse pancreatic islets revealed the crucial role of A20 activity in the islet response to inflammation and injury. Loss of A20 expression renders β-cells susceptible to apoptotic death; whereas enhancing A20 expression could improve the survival of islets exposed to traumatic conditions [183, 184].

Activated Protein C (APC)

Protein C which was first isolated from bovine plasma by Stenflo [185], is a serine proteinase synthesized by liver, endothelial cells, leukocytes, and keratinocytes. The most overt function of protein C is as an endogenous anti-coagulant. Severe protein C deficiency is characterised by massive, usually lethal, thrombotic complications

37 occurring in infancy. Adults heterozygous for protein C deficiency manifest a significantly increased risk for venous thrombosis [186]. When activated, APC exhibits a repertoire of additional physiological effects including cytoprotective (mainly through its receptor EPCR), anti-inflammatory, anti-apoptotic (via EPCR and protease-activated receptor-1, PAR-1) and pro-angiogenic functions (via matrix metalloproteinase activation), endothelial barrier protection (via induction of sphingosine kinase-1

(SpK1), and upregulation of sphingosine-1-phosphate (S1P) formation via SpK1)[187-

192].

Protein C is activated by the thrombin-thrombomodulin complex when it is bound to its cell surface ligand, endothelial protein C receptor (EPCR). EPCR augments thrombin- thrombomodulin- dependent PC activation by about almost 10-fold, with the same affinity for both PC and APC [189, 193]. This receptor mediates the pleiotropic cytoprotective effects of APC on cells through PAR1 signalling (Figure 1.6). EPCR has been detected on endothelial cells, keratinocytes, monocytes CD56+ NK cells, neutrophils, eosinophils, tenocytes and other cell types [192], EPCR gene transcription can be upregulated by APC and inflammatory cytokines such as TNFα and IL-1β

[194]. APC can modulate pathways of inflammation and apoptosis, with the general effect of down-regulation of proinflammatory and proapoptotic pathways and up- regulation of anti-inflammatory and anti-apoptotic pathways. APC blocks expression of

TNFα, suppresses NF-κB nuclear translocation, and subsequently the transcription of downstream proinflammatory genes [188, 191]. However, APC has also been observed to upregulate expression of the chemokines MCP1 and IL-8, and cytokine IL-6 [195,

196].

38 Matrix metalloproteinase-2 (MMP-2) is resistant to cleavage by most enzymes, but

APC has been shown to selectively upregulate both gene and protein expression of

MMP-2 and directly cleave and activate the proenzyme to its mature form. Protein C and matrix-metalloproteinase 2 activity are thus closely linked [197].

According to in vivo studies, leukocyte recruitment can be blocked by exogenous APC in various inflammatory conditions [198, 199]. Recombinant human APC inhibited

TNF-induction of cell surface adhesion molecules such as VCAM-1, ICAM-1, selectins, through regulatory effect on the NF-κB pathway [188]. Xue et al. demonstrated that APC can promote angiogenesis in human skin and wound healing in mice as well as the development of healing phenotype in cultured tenocytes [200, 201].

Stimulation of endothelial cell proliferation and angiogenesis by APC through activating MAP kinase pathway mediated by EPCR and PAR-1 signalling is also reported in vitro and in vivo [9].

39

Figure 1.6 Activation of protein C and activity pathways. (A) In the presence of thrombomodulin (TM) and endothelial cell protein C receptor (EPCR), protein C is activated by thrombin (IIa). The anticoagulant activity is exerted when APC is detached from its receptor (B). If APC remains bound to EPCR, multiple effects on cellular functions follow (C). (B) Anticoagulant activity is induced at cell surfaces by cleavage and deactivation of the activated cofactors Va (FVa) and VIIIa (FVIIIa) to, FVi and FVIIIi. (C) Pleiotropic cytoprotective effects of APC are transduced through the cellular receptors EPCR and PAR-1. These activities include modulation of gene expression, anti-inflammatory and anti-apoptotic activities and protection of endothelial barrier functions. Griffin, J. H., et al., Journal of Thrombosis and Haemostasis, 2007 [187].

1.5.3 Inflammatory pathways in warm ischemia-reperfusion injury

Extensive animal model studies in heart [202], lung [203], liver [204], kidney [205] and pancreatic islets [206] have demonstrated a significant role for TLR4 signalling in the pathogenesis of warm ischemia-reperfusion injury (IRI). Although TLR4 absence has a general protective effect in ischemia and IRI in all organs, the extent of engagement and effect of TLR4 and adaptor molecules vary among tissue and organs.

HMGB1 is released from hepatocytes after ischemic injury and induces inflammation through TLR4 signalling, primarily on non-parenchymal cells [207]. Both mRNA and

40 protein levels of HMGB1 are upregulated in ischemic mouse kidneys [205], but in this instance, signalling through TLR4 on renal parenchymal cells is the principal contributor to further inflammation and damage.

In several experimental studies, IRI was attenuated by neutralising HMGB1 with antibodies. Mouse kidneys were protected from IRI by blocking HMGB1, manifested as less tubular damage and apoptosis, less infiltration and lower creatinine level [208].

In liver, administration of HMGB1 neutralising antibody led to significantly reduced damage after IRI, whereas recombinant HMGB1 treatment exacerbated IRI [209, 210].

1.6 Brain death

Most organs for transplantation are retrieved from brain dead donors. The scarcity of the graft source along with the global detrimental effects of brain death on transplantable organs, necessitate expansion of our knowledge of the events surrounding brain death at a cellular and molecular level and improved management of potential donors.

According to The American Academy of Neurology [211], brain death is the irreversible loss of brain and brain stem function, usually due to major haemorrhage, hypoxia, or metabolic dysregulation. The diagnosis is confirmed by comprehensive neurological evaluation. To establish absence of brain stem reflexes, and apnea under standardized conditions. Additional tests like electroencephalography, transcranial

Duplex-ultrasound, or cerebral angiography, may be performed. Brain death is accompanied by complicated sequelae including hemodynamic, hormonal, metabolic,

41 and inflammatory consequences, which all profoundly affect the state of the patient’s organs.

1.6.1 Physiological and hormonal changes

Brain stem herniation introduces massive haemodynamic instability, starting with a short period of hypertension and bradycardia. These were observed in BD-induced

Chacma baboon [212], then in small and other large animal models [213, 214]. The succeeding event is an explosive release of catecholamines into the circulation, called the “catecholamine storm”, which causes a hyperdynamic cardiovascular response

[215]. Higher heart rate despite increased systemic and pulmonary vascular resistance is observed. Brain dead patients manifest increased cardiac indices and initially enhanced tissue oxygenation [216]. When serum catecholamines are depleted, peripheral vascular tone falls, cardiac output falls rapidly and this may be compounded by hypovolaemia. Neurogenic pulmonary oedema may supervene, and all lead to cardiovascular collapse [212]. Due to the erratic hemodynamic state, organ perfusion, particularly for abdominal organs, is disrupted [217]. The trio of hemodynamic changes, compromised organ perfusion and reduced oxygen levels characterize established brain death.

BD disturbs the hypothalamic-pituitary axis. Loss of vassopresin, [anti-diuretic hormone (ADH)] from the posterior pituitary gland causes diabetes insipidus which worsens hypovolemia. Decreased Thyroid-stimulating hormone (TSH) and adrenocorticotropic hormone, ACTH from the anterior pituitary gland dampens circulating levels of thyroid hormones and corticosteroids, which contribute to

42 increased hypovolemia and acidosis, as well as triggering immune responses and general inflammation [215].

Ischemia affects cell metabolism at different levels including disturbed cellular ionic gradients, increased calcium concentration and activation of numerous phospholipases, leading to degradation of cell membranes [219]. Depending on the intensity of damage, oxidative stress is induced by uncoupling of the respiratory chain in the inner mitochondria membrane after accumulation of free fatty acids [50]. Inhibition of oxidative phosphorylation can result in cell death due to ATP depletion and lack of energy. A shift from aerobic to anaerobic metabolism can be clinically manifested as high levels of circulating free fatty acids and lactate, lower insulin secretion and hyperglycaemia [220, 221].

1.6.2 Proinflammatory states accompanying brain death

Brain death constitutes a systemic inflammatory state. This state is the result of a complex interaction between the immune system, vascular endothelium, complement and coagulation systems. Preliminary studies, particularly those in small animal models, revealed a general increase in cytokines such as IL-6, IL-1β, TNFα and IFN-γ in the circulation early after BD induction [213] [In some human studies, only systemic

IL-6 and TNFα increased in circulation [222]]. Gene expression for these molecules is also upregulated in the transplantable organs; with each ogan manifesting a slightly different pattern of gene expression [223].

Activation of the vascular endothelium occurs after BD [224], with the expression of adhesion molecules, such as P-/E-selectins, intercellular adhesion molecule (ICAM)-1

43 and vascular cell adhesion molecule (VCAM)-1, and of chemokines such as monocyte chemoattractant protein (MCP)-1, and this combination results in cellular infiltration in all transplantable organs [225, 226]. The infiltrating leukocytes are activated, and secrete proinflammatory chemokines such as TNFα and IFN-γ, which induce the upregulation of major histocompatibility complex (MHC) class I and II on parenchymal cells. This event is associated with increased organ immunogenicity and affects transplant outcomes through the T-cell recognition process [227].

The complement system is also activated as a consequence of brain death. Increase in complement C3 synthesis, C3a serum level, and also complement deposition were observed, adversely affecting the engraftment of kidney and heart transplants [228,

229].

Coagulation pathways are triggered after BD. The extrinsic pathway is initiated by release of thromboplastin from tissues with promotion of fibrin clotting. Platelets are activated through contact with damaged endothelium and release coagulation factors and inflammatory molecules [230].

In a porcine model, HMGB1 migration from the nucleus to cytoplasm, a precursor to release from the cells, expression was noted in liver early after BD induction, TLR4 but not TLR2 expression was upregulated in both liver and kidney [231].

Epithelial and gut barrier dysfunction is induced by elevated levels of HMGB1, leading to increased bacterial translocation to mesenteric lymph nodes [232]. Furthermore,

44 following endotoxaemia induced by brain death the HMGB1-LPS complex can form an extremely potent ligand for TLR4 signalling [233].

In summary, brain death and its severe physiological and metabolic consequences can induce the release of proinflammatory mediators. Inflammation increases the immunologic activity of transplantable organs and renders tissues more detectable by the host’s immune system. Histological damage, inferior function, and lower engraftment in comparison with organs from living donors are transplant-related outcomes of such a catastrophe [227, 234, 235]. Variability in the extent of inflammatory changes has been reported in different studies- in general, greater degrees of haemodynamic instability are correlated with more significant increases in proinflammatory mediators within the transplantable organs suggesting that donor management protocols which optimise haemodynamic parameters could attenuate graft inflammation [236-239].

1.6.3 Endotoxaemia in BD donors

Endotoxin is a lipopolysaccharide produced from the cell wall of Gram-negative bacteria. This macromolecule is normally absorbed into the portal circulation and is cleared by hepatocytes and/or Kupffer cells. The process of endotoxin clearance can be disturbed due to humoral, bacterial or host alterations. When the endotoxin is not cleared by the liver, it reaches the systemic circulation and causes endotoxaemia [210].

It is widely known that integrity of the intestinal barrier may be compromised in critically ill patients, such as those exposed to trauma, burns, shock, etc. [240]. The gut response to a major trauma like brain death induces an intestinal cytokine response

45 [241], increased intestinal permeability [242], and translocation of intestinal bacteria and endotoxin [243], causing systemic inflammatory response syndrome (SIRS) and sepsis [244].

BD contributes to high systemic endotoxin levels through several different pathways.

Liver damage and Kupffer cell exhaustion following BD can interrupt the clearance and promote endotoxaemia [245]. Moreover, the damage to gut’s structure and barrier function can happen early after BD induction [246] [244] and is manifested by upregulation of the adhesion molecules ICAM-1, VCAM-1, and E-selectin, infiltration of leukocytes, and apoptosis in duodenum and jejunum. As a result, inflammation promotes intestinal permeability, which causes bacterial translocation and accelerates more systemic cytokine release and cell infiltration, leading to further inflammation and immune activation of potential donor organs. Endotoxin is a ligand for TLR4 and can contribute to the activation of antigen presenting and parenchymal cells expressing this pattern-recognition receptor [246]. As described above, endotoxin signalling is amplified when endotoxin binds to the HMGB1 released from damaged tissues following brain death. In liver transplantation the level of endotoxaemia is inversely correlated with graft outcome [210].

1.6.4 Role of TLRs in experimental islet transplantation

β cell death is induced by TLR4 signalling. Islet isolation and transplantation have been shown to increase TLR4 expression which adversely affects transplant outcomes.

Allogeneic islets from TLR4−/− mice injected into the renal subcapsular space

46 exhibited reduced inflammatory signals, increased presence of regulatory T cells and indefinite graft survival. TLR4 suppression with blocking treatments led to similar results [247, 248].

In other studies utilizing a mouse model of islet transplant under the kidney capsule, inflammation in the graft was mediated by both TLR2 and TLR4 and related to early graft dysfunction, the absence of TLR4, NF-kB activation and graft dysfunction were prevented [206].

These findings were confirmed in mice receiving intraportal islet infusion. Following transplant of TLR4−/− islets, the immune response in liver with the engrafted islets in the measured by proinflammatory cytokines, HMGB1 expression, NF-κB activation, caspase-3 and TUNEL positive cells, was dramatically suppressed compared to the grafts from wild type animals, concomitant with a prolonged graft survival. HMGB1- mediated stimulation of TLR4, TLR2 and RAGE was inhibited by administration of anti-HMGB1 antibody in mouse recipients. By blocking the circulating and tissue

HMGB1, production of TNFα, L-1ß and IFN-γ (by NK cells and Gr-1+CD11b+ cells) were suppressed and massive early islet graft loss was halted, which led to improved islet graft viability [249] [250]. Reports suggest that serum HMGB1 levels can be a useful clinical biomarker to evaluate islet damage [251]. These studies suggest that inhibition of TLR4 signalling represents a novel strategy to attenuate early graft failure and preserve graft mass and function.

Nonetheless, another model employing TLR4-KO mice exhibited conflicting outcomes.

In vitro and in vivo stimulation of TLR4 signalling by LPS increased inflammation and

47 infiltration within the islet graft under the kidney capsule; however no evidence of graft survival prolongation was observed when either TLR4 deficient donor or recipient.

Survival of the MyD88-KO islet allograft was significantly prolonged when recipients were treated with anti-CD40L, suggesting that a combination of treatments targeting innate and adaptive alloresponses can synergize to prevent graft rejection [252].

1.6.5 Brain death in clinical solid organ transplantation

Extensive clinical data have demonstrated that donor BD has detrimental effects on organ transplant outcomes, in accordance with the findings of experimental procedures. In 1995, Terasaki et al. reported that kidney transplants from deceased donors have significantly lower survival rates than those from spousal (living unrelated) donors, despite the fact that the deceased donor organs were better matched than those from spousal donors [253]. The principal factor contributing to the reduced survival of kidneys from deceased donors was postulated to be damage due to peri- mortem events. BD can increase the rate of acute rejection and delayed graft function in kidney transplants [254].

Several studies of postreperfusion clinical kidney specimens confirmed triggering of inflammation and endothelial activation by BD [255] [256] [257]. Endothelial adhesion molecules, (E-selectin, ICAM-1, and VCAM-1) as well as proinflammatory cytokines and chemokines such as MCP1, TNFα, INFγ, IL-10 were significantly upregulated in these biopsies, comparable to those from kidneys undergoing acute rejection. In heart transplant recipients, expression of IL-6 and TNFα in the graft predicted the development of right ventricular failure and heart malfunction [258, 259]. Release of sympathomimetic mediators results in increased rates of apoptosis and necrosis in

48 cardiomyocytes [260]. Furthermore, BD can promote chronic allograft vasculopathy, in particular explosive/traumatic BD in donors caused inferior graft survival rates and increased intimal coronary artery thickening [220, 261, 262].

It was shown by Van Der Hoeven in 1999 that adhesion molecules ICAM-1 and

VCAM-1 are overexpressed in liver after BD, immunologically activating the organ and causing increased leukocyte infiltration [224]. Upregulation of many cytokines and chemokines (IL-6, IL-10, macrophage inflammatory protein (MIP)-1, TNFα and IFN-γ as well as MHC-II) is observed in transplanted livers from BD donors [234]. Recipients of BD donor grafts had more acute rejection episodes compared with recipients of grafts from live donors [263]. Increased serum transaminases are detected after liver transplantation from BD donors compared with living donors [234], and these grafts are subject to a higher rate of primary non-function [264].

1.6.6 Impact of brain death on pancreas and islet transplant

Contreras et al. investigated the effect of brain death on rat pancreas [265]. These authors reported increased serum levels of TNFα, IL-1β and IL-6 post BD induction.

Transcription of these molecules was upregulated in pancreas tissue. Islet recovery, viability and functionality (in vitro and after intraportal transplant) were significantly reduced in islets isolated from BD pancreatic tissue. Overall, brain death decreased islet yields and functionality in rats.

Tissue factor (TF) and monocyte chemoattractant protein-1 (MCP1) expressed by isolated islets trigger the instant blood-mediated inflammatory reaction (IBMIR) following intraportal infusion [70, 71]. The regulators leading to expression of these

49 molecules by islets are yet to be elucidated. Interestingly, in a rodent model of BD, TF and MCP1 gene expression in the pancreatic tissues did not change after BD compared to the controls, but freshly isolated islets from the BD group displayed significantly higher mRNA expression for these genes than islets isolated from the control group

[266]. BD may thus be of paramount importance as an initiator of TF and MCP1 induction in the isolated islets, and as a contributor to islet graft loss through IBMIR.

In one clinical brain death study, TNFα protein levels were higher in BD than in control pancreas, but despite systemic inflammation and increases in serum IL-6 and TNFα, pancreatic levels of IFN-γ, IL-1β and IL-6 were unchanged by BD [222]. This was the first study to detect BD-induced damage to the pancreas before commencement of islet isolation and transplantation. TNFα induces apoptosis in β cells through NF-κb activation [183], and resultant loss of islet β cells may contribute to reduced yield from

BD donors.

Similar to findings in rodent models, whole pancreas tissue from BD donors does not express tissue factor [222] whereas human isolated islets do [267]. Such differences can be explained by the detrimental effects of cold ischemia and the isolation procedure on islets compounding the effects of brain death [70, 266].

When islet isolation outcomes were compared in cadaveric versus living donors who had pancreatecomy due to trauma or lesion [268], superior results were observed from living donors. Islet purity, islet isolation success rate, and equivalent islet numbers adjusted by pancreas weight were all significantly higher in the living donors; the islet function in vitro was also superior in living compared to BD islets, though this

50 difference did not reach statistical significance. Analysis of human islets post isolation showed that they express significantly high mRNA levels of A20 compared to islets within the native pancreas. High expression of NF-κB-regulated A20 is consistent with the observation that organ preservation and or islet isolation results in activation of NF-

κB [269].

1.7 Donor management/preconditioning

The great disparity between the pool of available donor organs and the demand for these organshas prompted endeavours to increase the organ retrieval rate through improved management of potential donors and broadening of the donor eligibility criteria [270].

Goals are defined for donor management in order to reach standards. These intend to preserve physiology close to normal by maintaining body temperature, adequate oxygenation, haemodynamic stability and sufficient urinary output [271]. These management protocols may confer additional benefits if they can attenuate inflammation in transplantable organs.

Based on a broad understanding of the pathophysiology of brain death, treatment with three hormones; T3, cortisol and insulin, was proposed in the 1980s, primarily for administration to BD prospective heart donors with suboptimal haemodynamic parameters [272]. This application led to a series of experimental and clinical studies.

In pigs, Novitzky and Cooper showed that treatment with hormones significantly restores cardiac output and stroke volume compared to values in control animals [214]; in baboons conversion of anaerobic to aerobic metabolism was observed [221]. 21

51 consecutive BD donors were treated with hormones and compared with donors receiving traditional therapy including i.v. fluids, inotropes and bicarbonate. The treated group showed significantly improved cardiac indices and organ function after transplant [273]. A retrospective study of BD donors during 1 year in the United States had established the superiority of “hormone resuscitation”, comprising a different three agents, methylprednisolone, arginine vasopressin and T3 or T4. Stability of a group of

“marginal” BD donors was significantly improved and the number of organs retrieved and transplanted per donor increased significantly [274]. Some protocols include a forth drug, adding insulin to the cocktail. In a BD pig model case control study where the four-drug hormone resuscitation was compared to norepinephrine and regular fluid replacement [231], results showed that inotrope requirement was decreased and hemodynamic status was significantly improved in the treated animals, supporting HR application in the management of the brain-dead donor.

High-dose steroid treatment in BD is now known to suppress the inflammatory response [275], prevent hypoxic damage [276] and improve transplant function and engraftment [277, 278].

Overall, hormone resuscitation has been resulted in significant augmentation of retrieval rates, reduction of inflammation and improved 1-year graft survival for pancreas, kidney, liver, heart and lung [270, 279].

52 1.8 Experimental strategies to reduce graft inflammation

1.8.1 Treatments applied to the donor

Blockade of IL-1 receptor

IL-1β is highly upregulated early after injury [280] and plays a major role in mediation of acute inflammatory responses. Blockade of IL-1 receptor in rhesus macaque donors led to suppression of proinflammatory molecule production, improved islet function in culture, and most importantly increased engraftment and function after transplantation

[281]. These finding are in accordance with the effects of IL-1β and TNFα combined blockade in in recipients in enhancing the efficacy of clinical islet transplantation and achieving single-donor islet transplant [82].

Anti-thymoglobulin treatment

Antithymocyte immunoglobulin (ATG), a polyclonal rabbit IgG against human T cells, is generally used in recipients. ATG antibodies interrupt leukocyte-endothelium interaction and cell migration by modulating expression or binding of mediators such as ICAM1 and lymphocyte function-associated antigen 1 (LFA-1), and also by binding to the chemokine receptors and thereby blocking their function [282]. ATG can induce changes in different lymphocyte cell subsets in blood and promote regulatory T cell expansion [283]. Donor treatment with rATG may improve the organ quality. In a rat model, rabbit anti-rat thymoglobulin was administered to the brain-dead donors, and hearts, lungs and small bowels were harvested. Histological changes including neutrophil infiltration and structural damage were studied. The histological injuries were significantly lower in the lungs of the rATG-treated group, with a trend towards reduction in hearts and intestines as well [283].

53 The immunosuppressive effect of ATG has been attributed to T cell depletion, mediated by lymphocyte surface antigen modulation and alterations in the patterns of cytokine production, chemotaxis, endocytosis, stimulation, and proliferation [282]. In another study of BD in rates by the same group [284], acute tubular necrosis (ATN) was significantly attenuated in the rATG-treated group. This effect was overexpression of IL-10, inhibiting pro-inflammatory cytokines such as TNFα. These results pave the way for further application of ATG in donor cytoprotective strategies to improve organ viability.

Antibodies against selectin

Selectin molecules are among the inflammatory mediators induced early after brain injury and subsequent hypoperfusion, initiating immune activation and leukocyte migration. High-affinity counter-receptor for P- and E-selectin, PSGL-1, is expressed constitutively on cell surfaces and essential as a trigger for the diapedesis of leukocytes

[285]. rPSGL-Ig, a recombinant soluble form of PSGL-1, abrogates the interaction of circulating leukocytes and platelets with activated endothelium. Administration of this ligand in animal models of total ischemia/reperfusion has been shown to inhibit the initial selectin activity and subsequent cellular and humoral events [286, 287]. Selectin blockade with rPSGL-Ig in rat models of kidney allograft reduced early inflammatory events, chronic rejection rate, improved engraftment and organ survival, although in conjuction with minimal dose of sirolimus and/or cyclosporin A. In these models the protective effects are particularly boosted when both donor and recipient are treated with the antibody [288, 289].

54 1.8.2 Treatment/manipulation of isolated islets

Extended duration of culture alone

After isolation, islets express several mediators with powerful biological activity, including IL-1β, IL-8, MIP2, MCP1, and MIF. After long-term culture (7–11 days), expression of these pro-inflammatory molecules considerably decreases compared to the islets at early stages after isolation. According to the literature, it’s certainly advantageous to culture islets before transplantation to reduce expression of these mediators. Cultured islets are more likely for successful engraftment [290]. Culturing would also permit enhanced expression of specific genes that may improve islet engraftment, e.g., protective genes (heme oxygenase-1, HO-1; A20), anti-apoptotic genes (Bcl-2), and immunomodulatory genes (IL-10, TGFβ).

Islets in vitro:

Gene transfer: A20, haemoxygenase-1, Fas Ligand

TNFα-induced protein 3 (TNFAIP3), or A20, is a regulator of inflammation and a potent anti-apoptotic gene in many cells types including β cells. A20 is considered as a potential therapeutic target for intervention in conditions characteristically apoptosis and/or the inflammatory response, due to its activity on NF-κB and cell death [109,

179].

In mouse models of islet transplant, adenovirus-mediated gene transfer of A20 to donors protects the isolated islets from IL-1β/IFN-γ and Fas-induced apoptosis. A20 expression protects functional β cell mass and promotes survival. The dual

55 antiapoptotic and antiinflammatory effect of A20 on β cells is related to abrogation of cytokine-induced NO production due to transcriptional blockade of iNOS induction

[291] [183] [111]. Normoglycaemia was restored and maintained in mice with less number of A20-expressing islets compared to normal islets [111].

Haemoxygenase-1 is an enzyme which degrades Heme to (HO-1) metabolites including bilirubin and biliverdin producing carbon monoxide during the process. HO-1 has anti- inflammatory; proliferative anti apoptotic properties, there is increasing evidence of its efficacy in improving transplant outcomes [292]. HO-1 expression is observed in islets under stress conditions such as after isolation [293]. HO-1 application both pharmacologically or by gene transfer in donor islets protects them from stress-induced apoptosis in both in vitro and in vivo settings, graft function and survival and tolerance is also promoted [294, 295].

Exposure of rat isolated islets to Cobalt protoporphyrin and zinc protoporphyrin in culture induced HO-1 over-expression, when these islets were transplanted into mice, improved graft function and survival was observed. The protective effects of HO-1 in this xenotransplantation model are thought to be mediated by IL-10 as HO-1 generates

CO, which downregulates iNOS and upregulates IL-10, protecting the β cells [296].

Administration of bilirubin, a product of HO-1 activity, either pharmacologically or by gene delivery in donors and in isolated islets, induces upregulation of antiapoptotic molecules like Bcl-2 and hepatopoietin (HPO-1) and suppression of the proapoptotic and proinflammatory genes caspase-3, caspase-8, and MCP1 both in vitro and post- transplant and promotes tolerance [297, 298].

56

Allotransplant of FasL–engineered islet grafts in a mouse model established euglycaemia and limited the need for immunosuppression, inducing robust localized tolerance in all recipients. Tolerance was initiated and maintained by

CD4+CD25+Foxp3+ regulatory T (Treg) cells [110]. CTLA4-Ig and TGF-β transduction of mouse islets prior to transplantation also prolongs graft survival [299]

1.8.3 Treatment of the recipient

Anti-inflammatory agents

Anti-inflammatory agents have been incorporated within immunosuppressive regimens in recent clinical allogeneic islet transplant protocols [27], although no large clinical study has yet evaluated the impact of these interventions. TNFα inhibitors along with a

T-cell depletion agent are being largely administered in peri-transplant immunosuppressive regimens, resulting in significantly higher rates of insulin independence and long-term graft survival than conventional immunosuppression [88,

300].

In a human to rodent islet xenotransplant model, etanercept (an TNFα inhibitor), could reverse hyperglycaemia only when administered in combination with anakinra (an IL-

1β receptor antagonist) [300], suggesting that TNF inhibition alone cannot fully halt the inflammation. More importantly, clinical application of etanercept and anakinra in islet allotransplantation rendered all patients insulin-independent after a single infusion [82].

57 In a rodent model, withaferin A, an anti-inflammatory compound and a potent NF-κB inhibitor abolished proinflammatory cytokines and apoptosis and significantly enhanced islet graft survival [301].

Several anti-inflammatory agents are under investigation and more emerging to optimise islet transplantation outcomes including diannexin [302], and the combination of JNK inhibitor SP600125 with nicotinamide and simvastatin [303].

Haemoxygenase-1

In rodent models, administration of HO-1 pharmacologically along with cobalt protoporphyrin (CoPP) to recipients enhances islet function and survival, along with promoting tolerance induction [295]. Moreover, preconditioning of only the donor, islets, or the recipient or combinations of such treatments with carbon monoxide (CO), supresses TLR4 signalling and production of proinflammatory mediators such as TNFα and MCP1, and decreases graft infiltration, resulting in prolonged graft survival [248,

304].

Activated protein C

In a “loop system” where human islets were exposed to ABO-mismatched blood, rHuAPC prevented IBMIR in a dose-dependent manner, significantly preserving islet integrity while maintaining platelet and leukocyte counts [305]. It has been demonstrated that APC limits IBMIR in vivo and in vitro through all the three anti- coagulation, fibrinolysis and NF-κB inhibition pathways [117]. Administration of murine APC to culture media decreased TNFα production and translocation of NF-κB in the murine portal vein endothelial cells (PVECs) co-cultured with syngeneic islets

58 [306]. Though the treatment in this study involved the portal vein endothelial cells,

APC actions may also have been mediated through an effect upon the islets in culture.

Contreras et al. showed that following intraportal islet allotransplant into recipient mice treated with APC, the fibrin formation and deposition due to islet infusion was reduced in liver; endothelial activation, inherent to islet infusion, and measured by plasma levels of sICAM-1 and soluble P/E-selectin was almost abolished in the treated group.

Cytokine production and inflammatory response of the liver (the engraftment site) was significantly suppressed. Islets in treated recipients showed superior engraftment and functionality and reduced apoptosis and immunogenicity in biopsies [306].

Induction of APC generation by liposomal formulations of thrombomodulin, increased the engraftment of transfused islets through reduction of coagulation and inflammation

[307]. In NOD mice, exogenous APC totally prevented development of type 1 diabetes by preserving β cells from apoptosis, stimulation of growth, reduction of inflammation and suppressing immune cell infiltration into islets [308]. It is noteworthy that plasma levels of PC/APC are significantly lower in T1DM patients compared to the controls

[309].

Taken together these findings suggest that APC might also be able to inhibit recurrent anti-islet autoimmunity in islet graft recipients with type I DM.

1.9 Modulation of immune responses using soluble RAGE

Soluble RAGE results in an effective blockade of RAGE signalling, both in vitro and in vivo a range of mostly rodent animal models. In mice, sRAGE administration has been shown to suppress production of pro-inflammatory markers such as adhesion

59 molecules, cytokines, chemokines, matrix metalloproteinase 9 activity, tissue factor, and phosphorylation of p38 MAP kinase. Endothelial activation and leukocyte migration and infiltration are also inhibited, all leading to significantly suppressed innate immune activation in response to inflammatory stimuli [143, 310].

In a study showing the association of RAGE activity with the development of diabetic angiopathy, the activity of human esRAGE in mice was documented [311].

Blockade of RAGE improves outcomes after hepatic IRI in mouse, modifies the activity of NF-κB and AP-1 and the expression of TNFα and IL-6, and modulates the activation of inflammatory and proapoptotic pathways. Hepatocyte cell death is suppressed and regeneration is promoted [312]. It should be noted that because of the ability of some RAGE ligands (such as HMGB1) to change multiple pattern recognition receptors, soluble RAGE also has inhibitory effect on signalling through

TLRs 2, 4 and 9.

In accordance with its role in both innate and adaptive immunity, interruption of RAGE signalling using either soluble RAGE isoforms or small molecule inhibitors of RAGE promotes favourable outcomes of allotransplantation. In a lung allograft model, IRI was attenuated, and oxygenation, capillary leakage and tissue injury scenes all improved in animals pretreated with soluble RAGE [313]. RAGE blockaded by esRAGE delays allograft rejection in murine heart transplantation whilst small-molecule RAGE inhibition [314], dampened islet allograft rejection [315], highlighting the role of

RAGE-induced donor-reactive T cell priming in acute rejection. The ability of soluble

60 RAGE to attenuate inflammation in pancreas and islets after donor brain death has not previously been evaluated.

Here we will evaluate the ability of systemic esRAGE overexpression to reduce pancreatic inflammation in a murine model of brain death, alongside two other strategies to reduce pancreas or islet inflammation - treatment of brain-dead donor pigs with a hormone resuscitation cocktail and treatment of porcine neonatal islet cell clusters in vitro with recombinant HuAPC.

61

CHAPTER 2 MATERIALS AND METHODS

62 2.1 Porcine neonatal islet cell cluster (pNICC) preparation and culture

2.1.1 PNICC isolation

Donor pancreata were obtained from 1-3 day old inbred Westran or outbred Large

White neonatal piglets of either sex (Body weight: 1.2-2.5 kg, Pancreas weight: 2-4 gr).

The pancreata were removed and digested to isolate islet clusters following a method modified from that described by Korbutt et al [316]. Briefly, piglets were anesthetised with Halothane. The chest and abdomen were opened using a parasternal incision and the piglets were allowed to exsaguinate. The pancreas was then removed and placed in cooled washing medium (4°C) containing HBSS with 1% pig serum (isolated from fresh adult large white pig blood). In a class II biosafety cabinet, each of the pancreata was chopped into ~2 mm3 fragments with long curved scissors in a 50 ml Falcon conical tube (BD Biosciences, San Jose, California).The tissue was left to settle under gravity and the supernatant was removed and replaced by fresh medium. After a further

3 rounds of chopping and washing the tissue (by brief centrifugation at 1000rpm),

HBSS plus Collagenase Type V (1 mg per ml of the chopped tissue), (Sigma, Castle

Hill, NSW, Australia) was added to the finely cut pancreata. Digests were gently agitated in a shaking water bath at 37°C for 14-18 min, then filtered through a metal filter mesh and washed 4 times with washing medium to remove all the collagenase.

The yield of pNICCs from each pancreas depends on the starting weight of the neonatal pig pancreas (weight in grams x 20000 = Number of pNICCs).

Enough pNICC culture medium (consisting of Hams F-10 medium (Sigma) + 10% pig serum) was added to 150 mm x 20 mm Petri dishes to give a final volume of 35 ml in

63 each plate. The islet aggregates were then transferred into the dishes at a concentration of 1000 ICC/ 3 ml, which was maintained during the culture period.

2.1.2 pNICC culture

Dishes containing the islet aggregates were maintained in culture medium at a37°C in a humidified incubator with 5% CO2, 95% air. The medium and dishes were completely changed on the first day post isolation and then every 2nd day for 5 days by briefly spinning the islets in 50 ml tubes and pouring off the supernatants. During the second week of culture, half and full medium changes were done every 2nd day alternately

(Table 2.1).

Full medium and plate Half medium change change

1 -

3 - Days post- 6 -

isolation - 8

10 -

- 12

14 -

Table 2.1 pNICC medium change schedule.

64 2.1.3 Preparation of pNICCs for counting and viability assessment

Using a sterile tip and pipette, 3 x 50μl aliquots of resuspended pNICCs were taken from each islet preparation. 2 x 50μl were placed in two lines on a large Petri dish for counting purposes. The number of pNICCs in each of the lines was counted, or islet size according to isolated pancreatic islet equivalents (IEQ) was measured, depending on what was required. Results were recorded in the NICC Isolation Data Sheet.

Prior to staining the islet clusters for viability assessment, a labelled frosted glass slide was marked with Dako pen (Dako, Denmark) to create a surrounding thick barrier and air dried. 1 x 50μl of the islet suspension was transferred into a 0.2ml tube, an equal amount (50μl) of Acidrine orange + Ethidium Bromide was added to this, then the suspension was transferred onto the marked glass slide using a transfer pipette, without introducing bubbles. This was then cover-slipped without pressing and left for 15 min to develop. The slides were examined under a mercury lamp-equipped, fluorescence

BX60 Olympus microscope (Olympus, Tokyo, Japan). Images were taken using a

DP72 Olympus camera mounted on the microscope (Figure 2.1). In live cells Acridine orange fluoresces green, and in dead cells Ethidium Bromide fluoresces orange.

Day 3 Day 6 Day 10

Figure 2.1 pNICCs post isolation in culture stained with Acridine orange + Ethidium Bromide. Live cells fluoresce green, dead cells fluoresce orange (x200 magnification).

65 2.1.4 Recombinant Human Activated Protein C preparation

Xiris [Drotrecogin alfa (activated) recombinant human Activated Protein C; Eli Lilly

Pty Ltd, West Ryde, NSW, Australia], 5mg as powder in a vial was provided by Dr

Chris J. Jackson (Sutton Arthritis Research Laboratories at Royal North Shore Hospital,

Sydney).

The powder was reconstituted with double-distilled autoclaved water to make a 2μg/μl solution prior to use. The solution was then stored in sufficient working aliquots at -

80ºC.

2.1.5 Treatment of Porcine Neonatal Islet Cell Clusters (pNICCs) with rHu APC

Recombinant human APC [187] was added to culture medium at a concentration of

10μg/ml from day 3 post-isolation. The dose 10μg/ml was decided based on the literature, majorly from the studies performed by Jackson et al [317, 318]. rHuAPC was applied to the treatment-receiving dishes along with every full or half change of the medium.

2.1.6 Harvesting pNICCs

ICCs were harvested days 0, 3, 4, 6, 10 and 14 for gene expression and immunohistochemical staining. 7,000 to 10,000 ICCs were harvested from culture by centrifugation at 1000 rpm for 2 minutes. After removing the supernatant, ICCs were directly disrupted in lysis buffer for nucleic acid extraction. Some lysates were frozen at -80 °C at this stage and batched for further RNA extraction.

66 For immunostaining around 1,000 ICCs were frozen in Tissue Tek freezing medium

(Sakura Finetek, AJ Alphen aan den Rijn , The Netherlands). A thin layer of medium was applied to the bottom of a pre-cooled cryomold and frozen down in liquid nitrogen, the ICCs were placed on the frozen medium and allowed to freeze immediately, the cryomold was filled with Tissue Tek, which was then frozen in the vapour phase of liquid nitrogen and stored at -80°C. The frozen blocks were cut in a Cryotome, in 5μm sections. The slides were stored at -80°C for further use and were fixed in acetone at room temperature for 10 min immediately prior to immunostaining.

2.1.7 Total RNA preparation from porcine islet cell clusters

RNA extraction

Total RNA was extracted using the Ambion RNAqueous Mini Kit (Ambion, TX) following a protocol modified for the manufacturer’s instructions. Approximately 7000

ICCs were pelleted by briefly spinning them and isolating the supernatant; then 700μl of Lysing/Binding buffer (Ambion) was added to the tissue pellet. After homogenising the tissue with a T10 Basic Ultra Turrax homogeniser (IKA, Staufen, Germany), the solution was syringed through a 25 gauge needle to reduce the viscosity and shear the

DNA. The following solutions were added subsequently with thorough mixing between additions: 70μl of 2M sodium acetate, pH 4.0 (Sigma) followed by 700 µl of RNA grade Phenol solution, pH 4.2 (Sigma). The mixture was vortexed and left on ice for 5 minutes before 140μl of a 49:1chloroform : isoamyl alcohol (both BDH AnalaR,

Kilsyth, Australia) solution was added. After vortexing for 30 seconds, the contents were left to cool on ice for 15 minutes. The organic and aqueous phases were separated at 11,000rpm for 45 minutes at 4°C in a refrigerated Eppendorf microcentrifuge. The

67 aqueous phase was then removed to a fresh 2ml Eppendorf tube; an equal volume

(800µl) of 64% ethanol solution (Ambion) was added to this. The solution was then purified and washed on an RNAqueous filter cartridge (Ambion) following the manufacturer’s protocol. After eluting the RNA in 180µl of elution solution, 2µl of glycogen (Ambion), 12µl of 7.5M ammonium acetate and 390µl of 100% ethanol was added to the solution and left overnight to precipitate at -20°C before being centrifuged at 11,000rpm at 4°C for 45 min as above. The RNA pellet was then washed with 1.5ml of 70% ethanol, and allowed to air dry at room temperature. Dried pellets were then reconstituted with 20µl of DEPC-treated water before the quantity and quality of RNA were assessed.

Assessment of RNA concentration, integrity and contamination

The RNA concentration of each sample was determined by optical density analysis.

The RNA content and protein contamination in 1µl of solution was estimated using a

NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific, MA, USA). A tester sample was then diluted if necessary and 1µl of this tester sample was loaded on an

Agilent 2100 Bioanalyzer RNA chip (Agilent Technologies, CA, USA) following the manufacturer’s instruction and run in an Agilent 2100 Bioanalyzer (Agilent

Technologies) to assess the precise concentration and obtain the RIN (RNA Integrity

Number), which is calculated based on the absorbance spectrum. All the RNA samples from porcine NICCs show a RIN of more than 7, indicative of highly preserved total

RNA (Figure 2.2). RNA samples to be used as templates for reverse transcription were then diluted with DEPC-treated water to a final concentration of 0.2µg/µl and stored in

NUNC cryogenic vials in liquid nitrogen as three 1µg aliquots and 1 aliquot of the remaining RNA.

68

Figure 2.2 absorbance spectrum of total RNA generated from electrophoresis in the Bioanalyzer.First peak in the graphs shows distinct 18S ribosomal subunit, the second peak shows distinct 28S ribosomal subunit.

2.1.8 Complementary DNA synthesis from extracted RNA

cDNA was synthesised using the SuperScript III kit (Invitrogen, CA, USA). All samples to be compared were included in the same cDNA synthesis reaction. Minus RT controls (reaction prepared without reverse transcriptase enzyme), and cDNA for preparing standard dilution series as well as no-template-controls were included in all

69 reactions. Two cDNA samples were synthesized from each RNA sample in two separate procedures.

1µg of total RNA was made up to 11µl with DEPC-treated water and primed with 2µl

Oligo-dT primer (Roche diagnostics, IN, USA). The contents were mixed and heated at

65ºC for 5 mins, then cooled on ice before adding the master mix. The master mix (7µl) contained 4µl of SuperScript III reaction buffer (Invitrogen), 1µl of 0.1M DTT

(Invitrogen), 1µl of 10mM dNTP (Ultrapure dNTP set, Amersham Pharmacia, Sweden) and 1µl if SuperScript III enzyme (Invitrogen). In the Minus RT reaction 1µl of water replaced the SuperScript enzyme and for the no template control no RNA was added to the DEPC-treated water. The 20µl mixture was then heated at 50ºC for 60 min followed by 15 min at 70ºC. After the completion of the reaction, the cDNA was diluted 1:5 with

80µl of DEPC treated water and aliquoted in 5µl portions in 0.2µl tubes or directly into

PCR plates which were then stored at -80 ºC prior to analysis.

2.1.9 Taqman primer and probe design for real-time RT-PCR of porcine genes

Taqman primers and probes for quantitative real-time RT-PCR analysis were designed using sequence data obtained from NCBI, ANGIS (Sydney Bioinformatics, NSW,

Australia) and EnsEMBL databases using Primer Express (PE Biosystems, Foster City,

CA) software. In order to avoid potential amplification of small amounts of contaminating genomic DNA in PCR, all the primer/probe sets were designed to either span intron/exon boundaries or to span long intronic sequences. The Taqman primer and probes sets were synthesized by Sigma-Aldrich (Sydney, Australia) or Applied

Biosystems (PE Biosystems) and supplied in lyophilised form. All Taqman probes

70 were labelled with a 5’ 6-FAM reporter and a 3’ TAMRA quencher, apart from the probe for MMP-2, designed as an MGB (minor-groove binding) probe. The oligonucleotide pellets were reconstituted with DEPC-treated H2O, and the stock solution was stored at -70ºC. The working solution of primers was made up from a mixture of reverse and forward primers at a 30μM concentration of each, and the working solution of probe was diluted to a 20μM concentration, both working aliquots were stored at -20ºC. Primer/probe Sequences appear in table below (Table 2.2).

Gene Name Sequence (5’ → 3’)

Forward CAGCAATGCCTCCTGTACCA

GAPDH Reverse GATGCCGAAGTTGTCATGGA

Probe AACTGCTTGGCACCCCTGGCC

Forward CCGTGGCTCGTACAAAGCA

H3 Histone 3A Reverse TTGCGAGCGGCTTTTGTAG

Probe CCCGCAAATCGACTGGTGGTAAAGC

Forward CCCCGAAAACAGAGGTGTCA

A20 Reverse TTTGAGAAGTACTGGGCTGAGAAA

Probe AGCCAGTTGTCACTCAACCCAGTCCATC

Forward GCGGCTGATGAGCTACAGAAG

MCP1 Reverse CCGCGATGGTCTTGAAGATC

Probe CACCAGCAGCAAGTGTCCTAAAGAAGCAG

Forward AGCGCTCTCAGCACCAATG

MIP1 Reverse TTCCGCACGGTGTATGTGA

Probe CAGACCCTCCCACCTCCTGCTGC

Forward TCAAGAACATCCAGGACCTGAA

MIP2 Reverse GCTATGACTTCCGTTTGGTCACA

71 Probe TGACGTCCCCGGGACCCCA

Forward CCAAGAGGGAGCCTCAAGCT

IL-1β Reverse CTCTGCTGACTTTTCTTTCCATAGC

Probe TGCCCCCCAGACATAGCCCAAA

Forward GACAAAGCCACCACCCCTAA

IL-6 Reverse CTCGTTCTGTGACTGCAGCTTATC

Probe CACCACAAATGCCGGCCTGCT

Forward GGGTGGAAAGGTGTGGAATG

Reverse GGCTGCAGTTCTGGCAAGAG IL-8 Probe TGCACTGGCATCGAAGTTCTGCACTT

Forward GCGCTGCAGGAGGCTAAAG

Reverse AGTCTGCGAGGAAGTCTCTTGC Complement C3 Probe CATCTGTGAACCACAGGTCAATAGCCTGTTG

Forward GCATGGATGTGATCAAGCAAGA

INF-γ Reverse CTGGCCTTGGAACATAGTCTGA

Probe TGCAGATCCAGCGCAAAGCCAT

Forward CCAGATTGAGGTGACCTTCGA

HSP72 Reverse TGCCTGTGCTCCTGTCAGT

Probe TGTCACACTCAGGATGCCATTGGCA

Forward CCAGGCAAGTGGATTATTGACA

TLR2 Reverse CGGAAAGCACGAAGATGGTT

Probe ATCATCGACTCCATCGAAAAGAGCCAGA

Forward GAGCTGAGCAAATGGTGTGTACA

TLR4 Reverse CAGGTGGCGTTCCTGAAACT

Probe ACCTTTAGAAATGCAGGACCTGCCTGTGC

Forward AAGGAGAGCATCCTGGCCTATC

HMGB1 Reverse TCTGCAGCGGTGTTATTCCA

Probe ATCTCTCCCAGTTTCTTTGCAACATCACCA

Forward TCTGGCAAGCGTTGTTTTTGTA

72 MIC2 Reverse ACCGTGAAGTTGTACCAAAGATTG

Probe CCCGGGTACCGTCGTTGCACA

Forward AGACAGCGGCATCACCAACT

ULBP1 Reverse GATGGCTGCACCCTTCGA

Probe CAGCGCCCCCCATAGCCCA

Forward CCTGGAGGCTTACCTGAAGGA

EPCR Reverse AGCGAACGATGAGAGGAAAGG

Probe TTCCAGGGTCTGGTGCAGGTGGTG

Forward CCCGGAAAAGATCGATGCT

MMP-2* Reverse AGTACTCATTCCCTGCAAAGAACA

Probe CCCCACAGGAGGAGA*

Forward TTGAAATGATTCCTGCAAGTCAA

IP10 Reverse TTTCTCCCCATTCTTTTTCATTG

Probe CTTGCCCACATGTTGAGATCATTGCC

Forward AGAAGCGTGGCATCGTGG

Insulin Reverse GGCCTAGTTGCAGTAGTTCTCCAG

Probe TGCCCTCTGGGAGCCCAAAC

Table 2.2 Taqman primer/probe sequences for porcine genes. *MGB probe: Minor Groove Binding Probe.

2.1.10 Quantitative real-time RT-PCR of porcine genes

The Real-Time RT-PCR was set up as follows. Each reaction contained 5μl of template cDNA (in duplicate), 12.5μl of 2x SensiMix buffer (Quantace Ltd, Finchley, UK),

0.25μl of 30μM Forward+Reverse Taqman primers, 0.25μl of 20μM Taqman probe and

7μl H2O to make up a 25μl reaction volume. To minimise pipetting errors and variabilities between samples, 20μl of pre-assembled Master Mix was added to 5μl of

73 cDNA template. A gene specific standard curve (comprising a 4-fold dilution series of a high expressing cDNA sample) was included in each assay.

Thermal cycling parameters were as following:

2 holds at 50ºC for 2 min and 95ºC for 10 min, followed by 40 cycles of 95ºC for 15 sec (melting) and 60ºC for 1 min (annealing). The reactions were run in a Model 7700 or 7000 Sequence DetectorTM (Applied Biosystems) or a Corbett Rotor-Gene 6000

(Qiagen) real time RT-PCR machines. Results were analysed using ABI Prism

Sequence Detector SoftwareTM v1.7 and Corbett Rotor-Gene 6000 Application

Software, v1.7. The expression level of housekeeping genes, Glyceraldehyde-3- phosphate dehydrogenase (GAPDH) and H.3 Histone 3A, was measured with gene specific Taqman primers and probes and the formula Gene of Interest

ExpressionsampleX/ Housekeeping GenesampleX was used to normalise the expression of all cDNA samples.

Analysis of Real-Time RT-PCR data

The results were saved in Excel format (Microsoft Office); the concentration of desired gene amplicons were calculated by incorporating the CtsampleX of gene of interest in the linear regression formula obtained from the gene specific Standard Curve; the average concentration of the two DNA syntheses was used as final data to be analysed. To analyse the differential expression between individual specimens, Graphpad Prism version 5.0 (GraphPad Software, CA) was used. One-way or two-way ANOVA and

Bonferroni post-hoc analyses were performed and a P value smaller than 0.05 was considered statistically significant.

74 2.1.11 Amplification of full-length porcine endothelial protein C receptor (EPCR) by RT-PCR

Primers for amplification of EPCR cDNA by PCR were designed on a porcine mRNA sequence (NCBI ref.: NM_001163406) using Sigma and ANGIS primer-design software (Table 2.3). The primer sets were obtained from Sigma in lyophilised form and reconstituted in DEPC-treated water; the working aliquots were stored at -20 at

10mM concentration.

Gene Name Sequence (5’ → 3’)

Forward CACCATGTTGACAACATTGCT EPCR

Primers Reverse CAGAGGGAGCCAGTGAGTAATTA

Table 2.3 Forward and reverse primers to amplify EPCR sequence in porcine tissue.

Reaction set up was performed using the KOD hot start DNA polymerase kit

(Novagen, EMD Biosciences, Inc., Darmstadt, Germany) and following a protocol modified from the manufacturer’s instructions. Each 50 μl reaction contained 5 μl of

10x buffer, 3 μl of dNTPs, 28 μl of PCR grade water, 1.5 μl of 10 μM of each of sense and anti-sense primers, 5 μl of template DNA and 1 μl of KOD hot start DNA polymerase. Thermal cycler conditions were as following: Polymerase activation at

95ºC for 2 min, denaturation at 95ºC for 20 sec, annealing at 63ºC for 10 sec, extension at 70ºC for 15 sec; steps 2-4 were repeated for 40 cycles in a Palm Cycler Gradient

PCR Machine (Corbett, Qiagen).

75 Visualisation of PCR product

To check the specificity of the rt-PCR reaction, PCR product size was verified by agarose gel electrophoresis and ethidium bromide staining, and observed under UV transillumination. 1% agarose gel was made as follows: 1.6 g of agarose powder

(Promega, Madison, WI) was dissolved in 160 ml of 1x TAE buffer by heating in microwave oven, then 9.06µl of ethidium bromide was added and poured into a an Owl

Gel caster (Thermo Fisher Scientifics). The PCR products were mixed with 1x TAE buffer to make up to 10 μl and loaded with 2 μl of 6 x DNA Loading Dye (Cat# R0611,

Thermo Fisher Scientific) in the gel wells. 12μl of GeneRuler 1kb Plus DNA Ladder

(Cat#SM1331/2/3, Thermo Scientific) was loaded in the ladder lane. The gel was electrophoresed for 1 to 2 hours at 100 V.; then visualized in a Syngene G:Box Chemi

HR16 apparatus (Syngene, Synoptics Ltd, Cambridge, United Kingdom) under UV light.

DNA purification from agarose gel

Bands of the predicted size (~729bp) were cut out from the agarose gel under UV light.

DNA was purified from the gel using the Gel-Advance Gel Extraction System

(Viogene). Briefly, the gel slices containing PCR product were dissolved and DNA was purified and washed from the solution onto a filter cartridge by centrifugation. The

DNA was then eluted from the cartridge and stored at -20ºC. The DNA quality and quantity was measured using a NanoDropTM 1000 spectrophotometer.

76 2.1.12 TOPO-cloning of porcine EPCR sequence and transfection into bacteria

Using the blunt-ended pcDNA Gateway Directional TOPO expression kit (Invitrogen), the TOPO cloning reaction was set up following the manufacturer’s instructions.

Briefly, PCR product, around 400ng/μl, was put into the reaction at a 1:1 volume ratio to the TOPO vector (i.e. 1μl PCR product plus 1μl Vector). Then the reaction was made up to 6μl by adding 1μl salt solution and 3μl H2O; gently mixed and incubated for 15 min at room temaprature. 2μl of this solution was added to a freshly thawed vial of One

Shot TOP10 chemically competent E. coli (Invitrogen, Life Technologies). Empty vector, as negative control, and pUC19 DNA (provided with the competent cells), as positive control DNA were also each added to a vial of cells. The competent cell vials were then incubated on ice for 15min. The vials were transferred to a 42ºC water bath for 30sec to transform cells with the vector by heat-shock, and returned to ice. 250μl of

S.O.C medium (provided with the competent cells) was then added to the cell vials and left to shake at 200rpm for 1hr at 37ºC.

Autoclaved LB-Agar was heated up to liquefy. Ampicillin was added to the liquid LB-

Agar to yield a final concentration of 100μg/ml and mixed gently. The liquid was poured into sterile 10cm Petri dishes in a fume hood and let set.

50-200µl of the E. coli transformation suspension was added to the LB-Agar dishes and evenly spread by vigorous shaking with sterile glass beads. The plates were incubated upside-down at 37ºC overnight. Colonies appeared by the following morning, 3 colonies from each dish were picked and each was transferred into a 15ml tissue culture tube containing 5ml of liquid LB-Agar+100μg/ml Ampicillin. The tubes were shaken at

77 200rpm at 37ºC for 6 hrs. The E. coli cultures were then passaged into 250ml of LB-

Agar+100μg/ml Ampicillin and shaken overnight at 200rpm at 37ºC to produce enough bacteria for MaxiPrep.

Plasmid DNA extraction by Maxiprep

JETSTAR 2.0 Plasmid Maxiprep kit (Astral scientific, Caringbah, NSW, Australia) was used to isolated the plasmid from transformed cells. E .coli cells were pelleted by centrifugation, the medium was removed and cells resuspended in solution E1

(provided in the kit). Cells were then lysed by adding solution E2 and mixing gently, to get a homogenous lysate, incubated at RT for 5min. After adding solution E3 and mixing again, the lysate was centrifuged at RT and 12.000 xg for 10min. The supernatant was then applied to the JETSTAR column, which had been previously equilibrated with E4 solution, and allowed to run through the cartridge under gravity.

The column was washed with solution E5 twice. After adding solution E6, the DNA pellet was precipitated pellet using isopropanol followed by centrifugation at 4 ºC and

12.000 xg for 30 min. The plasmid pellet was washed with 70% ethanol and allowed to air-dry, the dried pellet was dissolved in H2O, and the quantity of the plasmid DNA was then measured using a NanoDrop 1000 spectrophotometer (Thermo Fisher

Scientific). Plasmid DNA was stored at -20 ºC prior to sequencing.

Plasmid DNA sequencing

An adequate amount of extracted plasmid DNA from E.coli (approx. 2μg) was sequenced using a forward primer binding the T7 promoter and reverse sequencing primers from Invitrogen.

78 2.2 Porcine Aortic Endothelial Cell isolation from piglet aortas

Isolation of piglet thoracic aorta

2-3-day old large white piglets, weighing 1.5-2.5 kg were sacrificed for multiple organ harvest. The animals were exsanguinated and 2-4 cm of the thoracic aortae were isolated from the surrounding organs and tissues, then transferred to cold sterile PBS.

Using a 1ml syringe, the inside and outside of the aorta was flushed thoroughly with cold PBS, to remove blood clots.

PAEC isolation by scraping method for RNA extraction

Aortae were then put into 15cm Petri dishes and were cut longitudinally using a surgical blade. Any adhering lymph node or excess tissue was dissected away. The vessel was fixed using sterile plastic forceps to avoid tissue slippage. With a sterile cell scraper, the lumen of the aortae was gently scraped to detach the endothelium thoroughly. The scraper carrying the cells was then immersed into 700μl of cell lysing solution (Ambion) and cells were disrupted immediately by syringing up and down.

Tissue lysates were used for RNA extraction using the RNAqueous kit (Ambion).

PAEC isolation with enzymatic digestion method for culture

After brief soaking in cold PBS, aortae were transferred to cold M199 medium, containing 10U/ml penicillin and 10mg/ml streptomycin and left on ice. For dissection, the aortae were transferred into a 10cm petri dish, with a drop of medium. The vessel was fixed to the plate using sterile forceps, and the adhering tissues were dissected using a surgical blade. The vessel was cut longitudinally, and sliced into 2-3 equal pieces, then transferred to a 6-well plate with the luminal side facing up. Collagenase I

79 solution (0.26%) was prepared just before use. 3ml of enzyme solution was added to each well containing tissue, and incubated for 20 min at 37°C.

The supernatant containing detached cells was transferred into a 15ml tube, the luminal surface was scraped gently, the well was washed with 10ml of medium, and this was added to the tube. The cells were washed by centrifugation at 220 xg for 7 min, and the supernatant was discarded. The cell pellet was resuspended in 6 ml of complete M199 medium, transferred into the pre-warmed gelatine-coated T25 flask, and incubated at

37°C in an atmosphere of 95% O2 and 5% CO2.

Subculturing primary PAECs

When the cells were reaching 80% confluence, supernatant was removed from culture flasks by aspiration. The pure colonies of macrovascular endothelial cells exhibit typical “cobblestone” morphology, the cultures with dominant fibroblast growth were discarded. The adherent cell monolayer was washed with PBS and dislodged from the flask by trypsinization. The cells were then washed with medium by spinning at 220 xg for 5 min. The cell pellet was resuspended in complete M199 medium; after counting, cells were seeded in gelatine-coated flasks and incubated at 37°C in 95% O2 and 5%

CO2.

PAECs treatment with recombinant human Activated Protein C

Confluent cells isolated from fresh piglet aortae between passages 3 and 4 were used for the experiment. After counting the cells, 2 x 106 cells were transferred immediately into Ambion lysing/binding solution for RNA extraction as a baseline control. The same number of cells were allocated to each treatment group of the experiment. To the gelatine-coated T75 flasks, 15 ml of complete M199 medium was added, followed by

80 addition of reconstituted recombinant human activated protein C (Xiris; Eli Lilly Pty

Ltd, West Ryde, NSW, Australia) to treatment flasks at 10 µg/ml, immediately prior to inoculation with 2 X 106 of cells. Half medium change and APC treatment were carried out after 24 hrs for 48 hr reactions. Control and treated reactions were harvested after

24 hr and 48 hr in Ambion lysing/binding solution for further RNA extraction.

2.3 Preparing tissue segments for immunohistostaining purposes

Using a surgical blade, 5mm of the washed piglet aorta was cut horizontally, without pressing the lumen.

Snap-freezing

A thin layer of TissueTek freezing medium (Ted Pella, Inc. and PELCO International,

Redding, CA) was put on the bottom of a pre-cooled cryomold and frozen down in liquid nitrogen. The 3mm specimen was then placed vertically on top of the frozen

OCT, surrounded by extra freezing medium, frozen down immediately in the vapour phase of liquid nitrogen and stored at -80°C. The frozen blocks were cut in a cryotome, in 8μm sections. The slides were stored at -80°C until further use and were fixed in acetone at room temperature for 10 min just prior to immunostaining.

Formalin fixation and paraffin embedding

The tissue was fixed overnight in 10% paraformaldehyde in phosphate buffer. The tissue segment was then subjected to serial dehydration and clearing and finally infiltrated with molten paraffin wax. Paraffin embedded tissues (PETs) were sectioned on a microtome at 5μm thickness and collected onto SuperFrost slides (Menzel-Gläser).

81 The slides were dried overnight at 42°C. Slides were deparaffinised in two changes of

Xylene (15 min each) and hydrated through graded ethanol prior to immunostaining.

2.4 Immunohistostaining methods

2.4.1 Indirect immunoperoxidase staining for Endothelial Protein C Receptor

Slides bearing fixed tissue sections were rinsed in PBS with 0.1% Triton X-100

(Sigma). To uncover the antigens, the paraffin embedded sections were boiled in

10mM Citric acid in water in a microwave oven for 15 mins, cooled down to room temperature and rinsed in PBS. All the slides were incubated in 0.5% H2O2 in methanol for 15 min followed by multiple washes with PBS. Sections were blocked with 10% normal goat serum in PBS for 20 min. After blotting dry, the specimens were incubated overnight at 4°C with polyclonal rabbit anti-human EPCR primary antibody

(Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:100 in diluent (PBS with 0.1%

Triton and 1.5% normal goat serum). After rinsing with PBS, incubation at room temperature with biotinylated goat anti-rabbit secondary antibody (Dako, Denmark), diluted 1:380 in diluent, was performed for 30 min before further washing. Using the

Vectastain Avidin-Biotin ABC kit (Vector laboratories, Burlingame, CA), sections were incubated with the mixed buffer for 30 min at room temperature according to the manufacturer’s instructions. Specimens were washed again before incubation with chromogen DAB solution (Dako Denmark A/S) for 10 min. Slides were then rinsed thoroughly in water and counter-stained in Harris’ Haematoxylin, dehydrated through graded ethanol; mounted and coverslipped. (Note: positive, negative and isotype controls were included in all procedures)

82 2.4.2 Islet marker immunostaining on pancreas tissue

The tissue was fixed and paraffinised as described above, serial sections were cut and and collected on SuperFrost slides, then subsequently dewaxed in Xylene and graded ethanol as described above, prior to use.

The intensity of islet marker staining was to be subjected to quantitative analysis, thus, in order to minimise variability, Autostainer Link 48 [319] -an automated slide processing instrument- was utilised for the following series of staining experiments.

Deparaffinised pancreas tissue sections were boiled in 10mM Citric Acid in water (pH

6) for 10 min in a microwave oven, cooled down in a water bath and washed in water.

The slides then were transferred into the Dako Autostainer; all the steps up to counterstaining were carried out by the Autostainer at room temperature. Sections were then incubated with 0.5% H2O2 in methanol for 10 min to block the endogenous peroxidase activity. Samples were blocked with IDetect Super Block solution (ID

Laboratories, Ontario, Canada) for 5 min. The liquid was blown off and the following diluted primary antibodies were added to the sections and incubated for 30 min (Table

2.4).

83 Primary Antibody Dilution Diluent Company

Guinea pig anti human Invitrogen, Life 1:100 TBS+1.5% NGS Insulin antibody Technologies

Rabbit anti Glucagon Invitrogen, Life 1:100 TBS+1.5% NGS antibody Technologies

Rabbit anti Invitrogen, Life 1:100 TBS+1.5% NGS Somatostatin antibody Technologies

Table 2.4 List of primary antibodies against islet endocrine hormones.

The sections were then washed with washing buffer, TBS with 0.1%Tween20 before adding IDetect SuperStain Anti-Polyvalent (ID Laboratories, Ontario, Canada) and incubating for 10 min. Sections were then rinsed with washing buffer, and subsequently incubated with IDetect SuperStain HRP (ID Laboratories, Ontario, Canada) for another

10 min. Following a further washing step, slides were developed with DAB chromogen [319] for 10 min. Slides were rinsed thoroughly in buffer, then removed from the instrument and counterstained with Harris’ Hematoxylin. Sections were dehydrated in graded ethanol, mounted with DPX and coverslipped. Isotype and negative controls for each antibody were included in the staining procedure.

2.4.3 Dako ACIS (Automated Cellular Imaging System)

Stained slides were labelled and scanned using the ACIS slide-scanner at 4 different magnifications in order to create digital slides. The digital slides were then saved in the

ACIS workstation and the brown positive staining intensity (DAB-bound epitopes) was analysed and numerated by the ACIS software.

84 2.4.4 Insulin immunostaining intensity analysis

Subsequently, a scoring system was designed to grade the numerical staining intensity for statistical comparison of groups of study (Table 2.5). The mean scores of each treatment group were compared using One-way ANOVA and Bonferroni post-hoc analyses and a P value of smaller than 0.05 was considered statistically significant.

Staining intensity value Assigned Score acquired from ACIS

90-129 +1

130-149 +2

>150 +3

Table 2.5 Staining scoring system designed according to the anti-insulin immunostaining intensity.

2.5 Brain death induction in pigs and different donor management protocols

A porcine model of brain-death donor was set up to closely imitate the human brain- dead donor. 16 large white pigs were used during the experiment.

2.5.1 Methods of anaesthesia and ventilation

Animals were sedated by an intramuscular injection of ketamine 10 mg/kg, midazolam

1 mg/kg and atropine 0.05 mg/kg. General anaesthesia was induced using intravenous

(IV) thiopentone 2–10 mg/kg. Pigs were then intubated and ventilated with 100% oxygen. Isoflurane inhalation and injection of fentanyl were administered when

85 required. Intravenous saline (0.9%) was infused from one hour post-induction of BD to maintain the central venous pressure (CVP) at 0–5 mmHg. Lignocaine 1 mg/kg IV was given for arrhythmia prophylaxis.

Monitoring of the animals included: electrocardiogram, invasive arterial blood pressure (via a femoral arterial catheter), CVP (via an internal mammary central venous catheter), pulse oximetry, end tidal expired CO2 and core temperature. Arterial blood gas and glucose analyses were performed at hourly intervals, and when clinically indicated, using an i-STAT Portable Clinical Analyzer (i-STAT Corporation, USA).

Urine and bile production was monitored and recorded.

2.5.2 Instrumentation

To monitor the haemodynamic status, ultrasonic dimension transducers (Sonometrics

Corp., Canada) were placed to measure pulmonary wedge pressure (PAWP), cardiac output and mean arterial pressure (MAP).

Flow probes (Transonic Systems Inc., USA) were placed around the left renal artery, hepatic artery and portal vein to measure the perfusion of the organs.

2.5.3 Induction of brain death

Brain death was induced in animals following the protocol described by Hing et al.

[231]. A burrhole was drilled in the right frontoparietal skull, a Foley catheter was inserted in the subdural space, and brain death was induced by inflating the Foley catheter balloon with water in 3 mL increments every 30 s to a total of 24 mL. By increasing the intracranial pressure an autonomic storm was induced. Brain death was

86 confirmed by the typical hemodynamic changes, as well as the absence of response to painful stimuli, and the absence of pupillary and corneal reflexes. Anaesthesia was ceased when BD was confirmed. The animals were maintained for 6 hours before termination of the experiment.

2.5.4 Donor management

Brain-dead pigs were randomly assigned to two different donor management protocols, noradrenaline pressor support (NA) versus hormone resuscitation (HR). One hour after induction of brain death, either of the treatments was commenced. To the NA-receiving group noradrenaline (3.3µg/min) was administered and pigs in the HR group were given a cocktail of methylprednisolone, triiodothyronine, arginine-vasopressin and insulin (Table 2.6). In both groups infusions were titrated to maintain mean arterial pressure at 60-70 mmHg (Figure 2.3).

Hormone Dose

Methylprednisolone 15 mg/kg IV bolus 4 μg IV bolus, followed by 4 μg/h IV Triiodothyronine infusion Arginine-vasopressin 0.5–4.0 units/h titrated to mean arterial pressure 60–70 mmHg Titrated to blood sugar level 6–10 Insulin mmol/L at minimum 1 unit/h IV

Table 2.6 Hormone-resuscitation protocol.

87 2.5.5 Organ procurement

Core-needle or wedge biopsies were harvested from visceral organs including pancreas, kidneys, liver and intestine at baseline (0 hr) and 6 hrs post brain death. In a sterile Petri dish the biopsies were cut if necessary into 2x2mm and snap-frozen separately in the vapour phase of liquid nitrogen, then stored in NUNC cryovials in liquid nitrogen for molecular biology studies. Also 5x5mm tissue sections were fixed in formalin for further paraffinisation as previously outlined as well as snap-frozen by immersion in

TissueTek freezing medium (Ted Pella, Inc. and PELCO International, Redding, CA) in cryomolds following the method described above, and stored at -80 ºC subsequently.

The experiment was terminated 6 hours post BD, the abdominal aorta was cross- clamped and the visceral organs were removed en-bloc, and then transferred to the backtable where the final specimens collected. Serum and urine samples were also harvested at 0 hr and 6 hr time points and frozen in liquid nitrogen.

88

Figure 2.3 Sequences of brain death induction, treatment commencement and specimen collection.

Collection of the pancreas from brain-dead donor pigs

Pancreas biopsies were collected at both 0 hr and 6 hrs after brain death as described above. After procurement, the pancreas was freed from the surrounding tissues and dissected. The pieces were transferred into a sterile Petri dish and rapidly cut into

2x2mm tissue samples that were snap-frozen separately in the vapour phase of liquid nitrogen and stored in liquid nitrogen for molecular biology studies. 5x5mm tissue sections were stored in formalin for fixation and further paraffinisation as previously outlined and also snap-frozen in TissueTek freezing medium.

89 RNA extraction from pig pancreas tissue

Total RNA was isolated from pig pancreas tissue as described above using Ambion

RNAqueous mini kit. RNA concentration and integrity were also determined as described above.

Complementary DNA synthesis from total RNA with Random Primers cDNA was synthesised using the SuperScript III kit (Invitrogen, CA, USA) following a protocol adapted from the manufacturer’s instructions. All samples to be compared were included in the same cDNA synthesis reaction. The RNA integrity number from all pancreatic specimens was reported suboptimal by Nanorop and Bioanalyzer. It is a challenging task to obtain intact RNA from a tissue that contains high amounts of

RNase and is prone to autolysis, such as pancreas [320]. For this reason, random hexamers (Roche diagnostics, IN, USA) were used instead of Oligo-dT to prime the mRNA for cDNA synthesis. Oligo dT only binds to mRNA poly-A tail whereas random hexamers prime mRNA, with or without a poly-A tail, and this can be used as universal, non-specific primers to any site on RNA). Minus RT controls (reaction prepared without reverse transcriptase enzyme) and cDNA for preparing standard dilution series were included in all reactions. Two cDNA samples were synthesized from each RNA sample in two separate procedures.

1µg of total RNA was made up to 11µl with DEPC-treated water and primed with 2µl

Random Primers (Invitrogen). The contents were mixed and heated at 65 ºC for 5 mins, then cooled on ice before adding the Master Mix. The Master Mix (7µl) consisted of

4µl of SuperScript III reaction buffer (Invitrogen), 1µl of 0.1M DTT (Invitrogen), 1µl of 10mM dNTP (Ultrapure dNTP set, Amersham Pharmacia, Sweden) and 1µl if

SuperScript III enzyme (Invitrogen). In the Minus RT reaction 1µl of water replaced

90 the SuperScript enzyme. The 20µl mixture was then incubated at 25 ºC for 5 min, followed by heating at 50ºC for 60 min and 15 min at 70 ºC. After the completion of the reaction, the cDNA was diluted 1:5 by adding 80µl of DEPC treated water and aliquoted in 5µl portions in sterile 0.2µl tubes.

Quantitative real-time PCR of porcine genes

Taqman primer /probe sets for porcine genes are described earlier. The real-time rt-

PCR reaction set up and thermal cycler conditions were as indicated above. Reactions were run in a Corbett Rotor-Gene 6000 (Qiagen). The expression level of housekeeping genes Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and H.3 Histone 3A, was also measured and the formula Gene ExpressionsampleX/ Maintenance GenesampleX was applied to normalise the expression level of all cDNA samples.

Analysis of Real-Time RT-PCR data

Analysis of the results was performed as described in section 2.1.10.

2.6 Apoptosis detection Assays

2.6.1 TUNEL assay

Terminal deoxynucleotidyl transferase dUTP Nick End Labelling, (TUNEL), was used as an established method for the detection of DNA strand breaks in the pancreas samples. TUNEL staining was performed following the instructions provided along with the In Situ cell death detection POD kit (Roche Applied Sciences,

Penzberg, Upper Bavaria, Germany). The TUNEL staining was accompanied by anti- insulin co-staining, to identify the islets throughout the specimen.

91 In brief, slides were deparaffinised and endogenous peroxidase was blocked using 3%

H2O2 in methanol for 10 min. After washing in PBS, the specimens were treated with

20ug/ml proteinase K (Roche) in PBS for 15 min at room temperature to uncover the antigens masked by fixation. Slides were washed in PBS again and blocked with 20% fetal bovine serum (Gibco) in PBS for 20 min at room temperature.

Terminal deoxynucleotidyl transferase enzyme (Roche In Situ cell death detection POD kit) was diluted 1:3 in dilution buffer (Roche). The diluted enzyme was then mixed

1:10 with labelled-nucleotide containing solution (provided with the kit), and sections were incubated with this solution for 30 min at 37 ºC in a humidified atmosphere in darkness. For the positive control, one section was treated with 2000U/ml DNase I recombinant (Invitrogen) in 50mM Tris-HCL, pH7.5, 10mM MgCl2, 1 mg/ml BSA for

10 min at room temperature to induce DNA strand breaks prior to the labelling procedure. The negative control specimen was incubated with label solution only.

After through washing, sections were incubated with peroxidase converter agent for 30 min at 37ºC in a dark humidified chamber, washed, and visualised with 3,3'- diaminobenzidine, DAB (Dako). Slides were then rinsed in water before counterstaining in Harris’ Hematoxylin, dehydration and coverslipping.

TUNEL-Insulin Double staining

Consecutive sections of each pancreas tissue sample were double-stained for TUNEL and insulin. A double-staining protocol was developed with AlkalinePhosphatase for anti-insulin staining, to complement the peroxidase colour development used with

TUNEL. After Primary anti-insulin incubation, sections were incubated with

92 biotinylated anti Guinea pig IgG secondary antibody (Vectastain ABC-AP) + 10% normal goat serum for 30 min at RT. After washing in PBS, Vectastain ABC-AP

Reagent was added to the specimens for 30 min at RT. Washing was followed by visualisation with liquid Permanent Red chromogen (Dako). After a quick wash, slides were briefly counterstained with Harris’ Hematoxylin, then left to dry overnight, mounted in DPX and coverslipped.

Imaging and analysis of TUNEL-stained pancreas sections

A minimum of 6 fields at 20x magnification, each showing at least one pancreatic islet, were photographed from each of the TUNEL-stained pancreas sections, using an

Olympus BX60 microscope and Olympus DP72 mounted camera. The images were then transferred into the Image J image processing program (NIH, Bethesda, MD).

Using the particle analysis plugin, TUNEL positive (dark brown) nuclei versus negative (hematoxylin-stained) nuclei throughout islet and acinar tissue were manually counted and recorded. The average counts from each specimen, including total positive and negative and the ratio of these two in both exocrine and endocrine tissue, was then subjected to statistical analysis by Mann-Whitney and One-way ANOVA tests. P values smaller than 0.05 were considered statistically significant.

2.6.2 Anti Cleaved Caspase-3 immunostaining on pancreas tissue

An alternative method of apoptosis detection is based on identification of the activated

(cleaved) form of caspase-3. Cleaved caspase-3 was detected by immunohistochemical staining of formalin-fixed paraffin-embedded pancreas tissue sections.

93 Using a microwave oven, dewaxed and hydrated slides were brought to boil in 10mM sodium citrate buffer pH 6.0 and maintained at a sub-boiling temperature (i.e. 98ºC) for

15 minutes. Slides were left to cool on the bench top for 30 minutes in the buffer.

Slides were washed in distilled water and incubated in 3% hydrogen peroxide in methanol for 10 minutes. They were washed in water and 1x TBS/0.1% Tween-20 wash buffer subsequently. Sections were blocked with 5% NGS in TBS for 30 min at

RT, followed by incubation overnight at 4ºC with Rabbit anti cleaved caspase-3

(Asp157) primary antibody (Cell Signaling technology, Danvers, MA, USA) diluted

1:10000 in SignalStain Antibody Diluent.

Isotype control (Rabbit Ig from Dako), blocking peptide (Cell Signalling technology) and negative control (diluent only) were also included in the staining process. After washing, detection was performed with SignalStain® Boost IHC Detection Reagent

(Cell Signalling technology) for 30min at RT. 3,3'-diaminobenzidine DAB (Dako) was added to the section for 5min for colour development before the slides were rinsed in water. Sections were then counterstained in Harris’ hematoxylin, dehydrated, mounted and coverslipped.

Analysis of the staining results was performed as described for the TUNEL staining section, above.

2.6.3 Extraction of fragmented genomic DNA

TACS Apoptotic DNA Laddering Kit (Cat# 4850-20-ET, Cat# 4859-20-K,

Gaithersburg, MD) was used to isolate fragmented genomic DNA from both tissue specimens and cells. We used porcine and murine pancreas biopsies and murine

94 splenocytes to extract genomic DNA. For tissue samples: Around 50 mg of the tissue was ground using a sterile frozen mortar and pestle. The powdered tissue was then incubated in 200μl of Sample Buffer and 20μl of 10x Tissue Buffer from the kit for 18 hours, before proceeding to the rest of protocol. For cells: 105 to 107 cells were incubated in 100μl sample buffer for 15 minutes, then 100μl of lysis solution was added, and mixed thoroughly by inversion.

The same protocol was followed for both tissue and cells subsequently. After shaking,

700μl of Extraction Solution 2 was added to the sample in a 2ml microcentrifuge tube.

400μl of Extraction Buffer 3 to was added to the sample and vortexed for 10 seconds.

The solution was centrifuged at 12,000 x g for 5 minutes. The upper layer (aqueous) was transferred to a new microcentrifuge tube. The darker, lower organic phase was discarded.

The following reagents were added sequentially to the aqueous sample: 40μl (i.e. 0.1 volume) sodium acetate and 440μl (i.e. equal volume) 2-propanol, mixed by inversion between additions, then centrifuged at 12,000 x g for 10 minutes at room temperature.

The supernatant was carefully removed without disrupting the pellet. 1 ml of 70% ethanol was added to the pellet to wash, and centrifuged at 12,000 x g for 5 minutes at room temperature. The supernatant was carefully discarded; the pellet was let dry by inverting the tube on a laboratory tissue. The pellet was resuspended in 20-100μl

DNase-free Water. DNA concentration and quality was determined using a Nanodrop

1000 spectrophotometer (Thermo Fisher Scientific, MA), and DNA was stored at -20

ºC for further use.

95 Visualisation of DNA laddering

The extent of DNA fragmentation caused by apoptosis can be observed by horizontal electrophoresis of DNA in agarose gel. For this purpose, 1% agarose gel was made by dissolving Trevigel 500 from TACS kit in 1x TAE buffer, using microwave heat. The liquid was poured into a in large gel casting tray and left to set. 2 μg of the genomic

DNA was made up to 10 μl and loaded with 2 μl of 5X Gel Loading Buffer in the gel wells. 12 μl of GeneRuler 1kb Plus DNA Ladder (Cat#SM1331/2/3, Thermo Scientific) was loaded in the ladder lane. The gel was electrophoresed for 2 hours at 100V. Before imaging, the gel was soaked in 0.5 μg/ml Ethidium Bromide in 1x TAE buffer for 30 minutes. The gel was then visualized in a Syngene G:Box Chemi HR16 apparatus

(Syngene, Synoptics Ltd, Cambridge, United Kingdom) under UV light.

2.7 Splenocyte isolation from C57BL/6 mouse spleen

Scissors and forceps were rinsed in 70% EtOH before use. A normal C57BL/6 mouse was sprayed with 70% ethanol. A vertical incision was made in the abdomen, and the spleen collected into 10ml of 4ºC RPMI medium (Sigma), then mashed, with a sterile syringe plunger, through a 100μm metal sieve into a cold 100mm Petri dish (Sigma).

The cell suspension was collected in a 50ml conical Falcon tube (BD Biosciences). The sieve was washed with a further 40ml of medium (50 ml final medium). The suspension was centrifuged at 600 x g, 4ºC for 10 mins. Supernatant was removed afterwards.

The cell pellet was thoroughly resuspended and 2ml of NH4Cl (red blood cell) lysis buffer was added with a pasteur pipette. The mixture was then incubated at room temperature for 2 minutes, during which the suspension changed from slightly opaque

96 to clear. The tube was filled with RPMI medium and centrifuged at 600 x g, 4ºC for 10 mins. Supernatant was then discarded. The pellet was resuspended in ~25ml RPMI medium with a pipette and filtered through a 40μm cell strainer (BD Biosciences) into new a Falcon tube to remove residual tissue pieces and clumps. The strainer was rinsed through with RPMI until the tube contained 50ml of cell suspension. The tube was centrifuged at 600 x g, 4ºC for 10 mins. The supernatant was removed. Cells were made up to 20ml in RPMI medium, stained with trypan blue to determine the viability and counted by haemocytometer, then seeded as required. The yield is about 2-3 x 107 cells.

Induction of apoptosis in mouse splenocytes using Etoposide

The treatment groups received Etoposide at a final concentration of 8.5μM, and were incubated for 2, 6 and 20 hours before harvesting. Genomic DNA was extracted from the cells as it was explained above using the TACS DNA laddering kit.

2.8 Brain death induction in C57BL/6 mice

2.8.1 Development of a brain death model in mice

Induction of brain-death and maintenance of brain-dead mice prior to tissue harvesting was performed by Mr. Zane Wang in our laboratory. This model is adapted from the protocol published by Atkinson et al [229].

Anaesthesia and Ventilation

Mice were anaesthetised using IP injection of a mixture of Ketamine (Hospira Australia,

VIC, Australia) (81mg/kg) and Xylazine (Troy Laboratories, NSW, Australia)

97 (9mg/kg). The depth of the sedation was monitored every few minutes and before any invasive procedure through the withdrawal reflex and observation of the whiskers of the mouse.

Top-up doses of ketamine and Xylazine were administered when required.

The Inspira advanced safety rodent ventilator (Harvard Apparatus, MA, USA) was used for ventilation with an FIO2 of 60%. The ventilator tubing was attached to a 22G blunt tipped cannula (BD Medical, MC, USA) used for insertion into the trachea. The ventilator auto-adjusts its ventilation settings according to the input weights of the mice.

Pressure-controlled mechanical ventilation was substantially better than volume- controlled ventilation at maintaining good oxygen saturation in the mice.

Fluid Support

The colloid solution we chose to use was Gelofusine (B.Braun, AG, Germany), a colloidal plasma volume replacement fluid based on modified fluid gelatine which has iso-oncotic pressure, physiological pH and a low chloride content.

Immediately prior to induction of brain-death, mice received one bolus of 0.5ml IV

Gelofusine and 2.0ml SC of normal saline at four sites. Following the initial bolus,

1.5ml of normal saline was administered hourly SC divided amongst three sites. Further

IV boluses of 0.25ml Gelofusine were given following femoral artery cannulation if the

MAP was below 55mmHg, and then as required when the MAP dropped to 40mmHg or below (Figure 2.4).

98

Figure 2.4 Typical races for the optimized model of BD with Gelofusine as fluid support. MAP peaked rapidly at the time of BD induction. This was accompanied by a permanent increase in HR, followed by a slow decline in BP over the subsequent hours. Gelofusine fluid support maintained blood pressure without compromising oxygen saturation. Temperature was kept between 37oC to 38.5oC.

99 Organ Harvesting

Pancreas (along with other organs including kidney, heart, liver and spleen) was collected into formalin for paraffin embedding and sectioning, liquid nitrogen as snap frozen pieces for RNA extraction and analysis and also into OCT for frozen sectioning.

Invasive and non-invasive monitoring

All procedures performed below were done on C57BL/6 male mice. Following anaesthesia with Ketamine and Xylazine, the left thighs of the mice were shaved in preparation for femoral artery cannulation.

The scalp was incised vertically, and a 14G cannula stylet was used to drill a burr hole in the bottom left corner of the right parasagittal bone (Figure 2.5). A 2F Fogarty balloon catheter was then inserted into the burr hole and secured using masking tape.

Brain death will be induced after mechanical ventilation and invasive BP monitoring have been established. Upon completion of the burr hole, mice were placed into the dorsal position. A rectal thermometer was then inserted. A vertical midline incision was made in the anterior neck. The two halves of the sternohyoid muscle were then separated and trachea exposed. A loose 5-0 silk tie was then placed around the trachea.

The ventilator was set to pressure controlled ventilation with mouse weight entered,

PEEP of 5cmH2O and FIO2 60%, and gas flow was started. A small transverse incision was then made in the sub-laryngeal area between two rings of cartilage (Figure 2.6 A).

The 22G blunt tipped cannula attached to the ventilator tubing was then inserted into the trachea to a pre-determined mark, avoiding cannulation of either of main bronchus and ensuring ventilation of both lungs. The cannula was then secured with the loose tie. An

100 additional 5-0 silk tie was then placed around the trachea and cannula to maintain a snug fit preventing gas leakage (Figure 2.6 B).

A B

Figure 2.5 Creation of burr hole. A) Theoretical location of burr hole marked by “X”; B) Actual burr hole made in mice.

A B

Figure 2.6 Depiction of tracheostomy. A) The exposure and opening of trachea; B) The securing of the ventilator tube to the trachea.

101 The last surgical procedure was the cannulation of the left femoral artery for MAP monitoring. An incision was made in the left inguinal area. The femoral neurovascular bundle was exposed using blunt dissection with cotton buds (Figure 2.7 A). The femoral artery was then separated from the surrounding tissues and vessels in the area between the abdominal wall and the superficial epigastric artery, (Figure 2.7 B). Lignocaine was used to help prevent artery vasospasm. Once separated, the distal end of the femoral artery was ligated with a 5-0 silk tie, slightly retracted with a clamp. A 5-0 silk tie was also placed around the proximal end of the artery, with a loose knot. This tie was also retracted and elevated using a clamp until the point where blood stopped running into the area in between the two ties (Figure 2.8 B and step 1 of Figure 2.8 A). A small transverse incision was made 2mm proximal to the distal tie (step 2 of figure 2.8 A).

One tip of a micro-forceps was then inserted into the artery for 10 seconds to dilate it, preparing it for cannulation. A 1.2G cannula (Solomon Scientific, TX, USA) was inserted into the femoral artery beyond the proximal tie (step 3 of Figure 2.8 A). The cannula was secured by tightening the proximal and distal ties (step 4 of Figure 2.8 A).

A pulsation within the cannula should be observed at this point as a sign of success. The

MAP was read and recorded via the Physiomex® data recorder system. Finally, a

MouseOX® pulse oximeter (Starr Lifesciences Corporation, Oakmont, PA) with mouse tail clip was used to detect, monitor and record oxygen saturation and heart rate.

102 B A

Figure 2.7 Exposure of the femoral artery. A) Exposure of the femoral neurovascular bundle; B) Separating the artery from surrounding tissues and vessels.

Figure 2.8 Depiction of femoral artery cannulation. A) Steps to performing a successful femoral artery cannulation. Step 1 shows the ligation of the distal end of the femoral artery and retraction with suture of the proximal end to prevent bleeding when the artery is transversely slit open in step 2. Step 3 shows the cannulation of the femoral artery and in step 4 the cannula is secured within the artery; B) Overview of the setting when doing the procedure.

Body Temperature

Core temperature was measured using a rectal probe (Columbus Instruments, Ohio,

USA), and was maintained within the range of 37oC to 38.5oC (Figure 2.4) using a

103 heating pad (Redzone Heating, Australia). Temperature was displayed on a Physiomex small animal physiological monitor (Columbus Instruments, Ohio, USA).

Brain Death Induction

Induction of brain death was performed in a similar manner to that described in other protocols for induction of brain death in rodents. The balloon of the Fogarty catheter is slowly inflated to progressively increase intracranial pressure leading to brain stem herniation and irreversible brain death. The burrhole was sited in the right parasagittal bone (Figure 2.5).

The syringe connected to the Fogarty catheter was placed into an automated syringe driver (Harvard Apparatus, MA, USA) and saline was infused into the balloon at a constant rate of 5µl/min over 20 minutes up to a maximum of 100µl. Inflation was ceased at the point where mean arterial pressure (MAP) started rising rapidly (figure

2.4). This was accompanied by a permanent increase in heart rate (HR), followed by a slow decline in MAP over the subsequent hours (Figure 2.4).

2.8.2 Cloning and subcloning of esRAGE cDNA

A cDNA encoding human esRAGE was obtained in two different ways. Firstly, esRAGE was amplified from normal human lung explanted in the course of resecting a pulmonary tumour. Amplification primers were forward

CACCATGGCAGCCGGAACAGCAGTTGGAGC and reverse

GCGGATATCCTTGTTGACCATCCCCCCAG. PCR products were run on 1% agarose gel to confirm that a band of expected size was amplified, then sequenced using the forward primer above as a sequencing primer (Figure 2.9).

104 In parallel, a codon-optimised cDNA sequence for esRAGE had been ordered from

GeneArt (Life Technologies, NY, USA). 5µg of the pMA plasmid containing this cDNA was resuspended in 50µL of ultrapure H2O and then transformed into SURE-2 competent E. coli (Stratagene, CA, USA) and allowed to grow on agar plates. A few colonies from the plate were selected for Mini-preparation and colony PCR using a

Quick Mini-prep Kit (Invitrogen, CA, USA) to verify that plasmids contained the esRAGE cDNA. Once the inserts were verified, a colony was selected for Maxi- preparation. Plasmid DNA concentration was estimated by measuring absorbance at

260nm on a nanodrop spectrophotometer (Thermo Scientific, MA, USA). Restriction enzyme digestion with NotI (New England Biolabs, MA, USA) and EcoRV (New

England Biolabs, MA, USA) confirmed the presence of a band running at the expected size of 1113 base pairs (bp) on agarose gel electophresis. The esRAGE insert was amplified from the maxi-prepped plasmid DNA using esRAGE-specific primers as above, then both the GeneArt sequence and the native sequence were TOPO cloned into pcDNA3.2 (Invitrogen, CA, USA), a mammalian expression vector. The pcDNA3.2 plasmids containing each of the esRAGE cDNAs were then transformed into one shot

Top-10 E. coli. Correct sequencing was confirmed once again with Mini-preparation and PCR. Following this, expression of the esRAGE protein from these constructs was tested in vitro by transfecting the mammalian cell line HEK293D with the pcDNA 3.2- esRAGE plasmid, collecting the supernatant at 48 and 72 hours after transfection and quantitating esRAGE using a commercially-available human esRAGE sandwich ELISA

(B-Bridge International, CA, USA). Protein expression in vitro was slightly higher from the GeneArt than from the native construct, and so the codon-optimised GeneArt construct was used in all further experiments.

105

Figure 2.9 Codon-optimised nucleotide sequence of esRAGE designed by GeneART.

106 Production and Packaging of an rAAV 2/8 vector encoding human esRAGE

Packaging of recombinant rAAV 2/8 vectors requires that the packaging cell line

HEK293D is triple-transfected with the following plasmids: pAM2AA (encoding the gene of interest and the inverted terminal repeats of the AAV serotype 2 genome) p5E18VD2/8 encoding the rep gene from AAV2 but the viral capsid protein from AAV serotype 8, and pXX6 (encoding adenoviral helper factors) (Figure 2.10).

I ApoE/h esRAGE WPRE, I

p5 Rep2 Cap8

p5 Adenovirus TF

Figure 2.10 Triple transfection with pAM2AA, pXX6 and p5E18VD2/8, yields rAAV2/8 virions after purification.

To produce complementary sticky ends and ensure rapid directional ligation, the template expression vector pAM2AA-eGFP and PCR fragments of esRAGE were simultaneously digested with EcoRV and NotI to release esRAGE fragments as well as to linearize the pAM2AA-eGFP vector.

The digested pAM2AA and esRAGE fragments were run on a 1% agarose gel and extracted using a Wizard Gel Extraction Kit (Promega, WI, USA). These were the ligation products. The amount of PCR product used in the ligation was set at a 3:1

107 vector: insert ratio, and the ligation was performed using Quick T4 DNA Ligase (New

England Biolabs, MA, USA) in a 5µl volume for 5 minutes at room temperature.

The products of the ligation were transformed into SURE-2 Supercompetent cells

(Stratagene, CA, USA). The transformed cells were plated onto agar plates with ampicillin (100µg/mL) at 2 different concentrations of 20µL and 200µL. A resistant colony was randomly selected for Mini-preparation and colony PCR to verify the sequence. Once verified, correct clones underwent Maxi-preparation and then sequence confirmation as described above in 13, ‘Cloning and subcloning of esRAGE cDNA’.

Once the sequence was verified, p5E18 and pXX6 were used along with pAM2AA-

o esRAGE for packaging. HEK293D cells were cultured at 37 C in 5% CO2 with standard medium: DMEM supplemented with 10% fetal calf serum (FCS). These cells were seeded into 40x10cm2 tissue dishes. The following day, the medium was refreshed in the morning and a triple transfection using calcium phosphate was performed 3 hours later. At 16 hours post-transfection, standard medium was replaced with DMEM + 2%

FCS. Cells were then harvested 48 hours post-transfection and immediately stored at -

80oC.

Virions were separated from the cell lysate by cesium chloride gradient purification, and then concentrated. Vector quantification was performed using quantitative real time

PCR and vector was aliquotted at 5x1011 viral genome copies for use.

108 Control vectors

A vector encoding Human serum Albumin was produced by PhD student Miriam Habib using an albumin cDNA cloned from explanted human liver, and was used as a control vector in subsequent experiments. Vector production and packaging were as outlined here for esRAGE, and a dose of 5x1011 VG yielded a concentration of human albumin of 90µg/mL in mouse serum.

2.8.3 mRNA extraction, cDNA synthesis and real time RT- PCR from mouse pancreatic tissue

Total RNA extraction from snap-frozen mouse pancreas and cDNA synthesis were performed following the protocols outlined under section 2.1.7. Real time RT-PCR was set up and run using Taqman primers and probes as described above in section 2.1.10.

Taqman Primer and probe sets for murine genes

The Taqman primers and probes were either ordered as “primers on demand” from

Applied Biosystems or designed using Primer Express and purchased from Sigma

(Table 2.7). The sets purchased as all-in-one products from Applied Biosystems were

(Name of the gene and accession number on Applied Biosystems website): Histone3f3a

(Mm01612808), CXCL10 (Mm00445235), IL-10 (Mm00439616), IL-6

(Mm00446190), IL-33 (Mm00505403), HSP2 (Mm00434069), TNFα (Mm00443258),

TLR4 (Mm00445274), Complement Component3 (Mm00437858), A20

(Mm00437121). The primers and probes we designed are shown in the table below

(Table 2.7).

109

Gene Name Sequence (5’ → 3’)

Forward TGCACCACCAACTGCTTAGC GAPDH Reverse GGCATGGACTGTGGTCATGAG Probe CCTGGCCAAGGTCATCCATGACAACTT Forward GAGCATCCACGTGTTGGCT MCP1 Reverse TGGTGAATGAGTAGCAGCAGGT

Probe AGCCAGATGCAGTTAACGCCCCACT

Forward GCCCCCAGGACCCCA MIP2 Reverse CTTTTTGACCGCCCTTGAGA Probe TGCGCCCAGACAGAAGTCATAGCCA Forward GCACACCCACCCTGCAG IL-1β Reverse AACCGCTTTTCCATCTTCTTCTT

Probe TGGAGAGTGTGGATCCCAAGCAATACCC

Hyaloronan Forward CAGCCTTCAGAGCACTGGG

synthase 2 Reverse TGAGGCAGGGTCAAGCATAGT

(HAS2) Probe CGAAGCGTGGATTATGTACAGGTGTGTGACTC

Forward GGACCCTTAGCTGGCACTTAGA RAGE Reverse GAGTCCCGTCTCAGGGTGTCT

Probe ATTCCCGATGGCAAAGAAACACTCGTG

Forward TGGGCGACTCTGTGCCTC

HMGB1 Reverse GCCTCTCGGCTTTTTAGGATC

Probe CGGAGGAAAATCAACTAAACATGGGCAAA

Forward TTGGTTCAGAAAATTGTCCAAAAG

110 KC Reverse CAGGTGCCATCAGAGCAGTCT

Probe TGCTAAAAGGTGTCCCCAAGTAACGGAGAAAG

Forward ATGAATTCTTAAAGCAGTCCAAGGAA

RAE-1 Reverse GGAAGTTGATGTAAGCTGGGTGATA

Probe AGCCAAGATCAACCTC

Forward CAGCAAGCACGTATTGTTT

Insulin Reverse GGACCACAAAGATGCTGTT Probe TGCCCTCGGGAGCCC

Table 2.7 Taqman primer/probe sequences for murine genes.

2.8.4 Immunohistostaining on mouse pancreas

Anti-insulin immunostaining was performed on paraffin-embedded mouse tissues following the protocols described earlier in section 2.4.

2.8.5 Apoptosis detection in mouse pancreas

Programmed cell death was investigated in mouse pancreas utilising TUNEL and anti

Caspase3 immunostaining as well as genomic DNA studies as per the methods outlined in earlier section 2.6.

2.8.6 Indirect immunoperoxidase staining for neutrophils

Deparaffinised PET sections were heated in 10mM Sodium Citrate (pH 6) in the

Decloaking Chamber (Biocare Medical, Concord, CA) for 15min for antigen retrieval

111 and then washed in TBST buffer (50mM Tris pH 7.4, 300mM NaCl, 0.05% Tween-20).

Sections were blocked for 20 minutes in 20% normal goat serum in TBST buffer. The primary rat anti-mouse Ly-6B.2 (neutrophil) antibody (AbD Serotec, Raleigh, NC) was diluted 1:100 in 1% normal goat serum in TBST buffer; specimens were incubated with the primary antibody for 1hr at room temperature. After washing in TBST buffer, slides were incubated in 3% H2O2 in methanol for 5 minutes to block endogenous peroxidase activity. Slides were washed again and then incubated with the secondary biotinylated goat anti-rat antibody (BD Pharmingen, San Diego, CA) diluted 1:150 in 1% normal goat serum in TBST buffer. Following another wash, the detection solution was made by mixing 1:100 reagent A and 1:100 reagent B (Vector Laboratories, Peroxidase standard,

Cat # PK 4000) in 1% normal horse serum in TBST buffer, and incubated slides with this solution for 30min at room temperature. Slides were washed in TBST and water, and incubated with DAB chromogen (DAKO) for 1min, washed in water, slide were then counterstained in Harris’ Hematoxylin, followed by dehydration and coverslipping.

112

CHAPTER 3 INFLAMMATION IN PORCINE NEONATAL ISLET CELL CLUSTERS (PNICCS)

113

3.1 Introduction

Pancreatic islet allotransplantation is proven to successfully treat uncontrollable T1DM and in some instances to replace exogenous insulin therapy. However, this therapeutic option is limited by the low availability of islet donors compounded by the poor quality of many donor pancreata rendering them unsuitable for transplantation. Optimisation of the quality of donor organs and development of a sustainable supply of tissue for islet replacement are two approaches which might allow an expansion of clinical islet transplantation and both are being actively pursued.

Procured pancreatic islets undergo sequential injuries beginning with the ischaemic and inflammatory consequences of donor brain death, the subsequent exposure to digestive enzymes and loss of interactions with extracellular matrix components which characterise the isolation procedure and further changes during culture, and the instant blood-mediated inflammatory reaction, or IBMIR, all of which accelerate inflammation and graft loss [14]. Islets are routinely cultured prior to infusion, and the culture period represents an interval during which islets can be studied to determine their response to various stimuli. Additionally they can be subjected to anti-inflammatory and immunomodulatory strategies, such as preconditioning with anti-inflammatory substances, gene therapy etc. Activated protein C is known for its unique cytoprotective properties, and has been demonstrated to protect islets, decrease IBMIR and increase engraftment in human and murine islet transplant studies [117, 191, 305, 306].

Human islets for research are very limited. Apart from their suitable physiology, porcine neonatal islet cell clusters are particularly favourable for both experimental and

114 clinical applications; they are more robust and less immunogenic than mature islets and can be manipulated more easily.

The endothelial protein C receptor (EPCR) was initially identified on endothelium but has subsequently been demonstrated on a range of cell types. To our knowledge, the direct effect of rHuAPC on porcine neonatal islet cell clusters (pNICCs) has not been investigated. We aimed to evaluate the inflammatory status of pNICCs in culture and hypothesized that rHuAPC could exert anti-inflammatory activity on pNICCs.

3.2 Porcine neonatal islet cell clusters in culture after isolation

Porcine neonatal islet cell clusters (pNICCs) were isolated by physical and enzymatic procedures from the whole piglet pancreata according to the methods described in chapter 2 (section 2.1 and 2.2). The pNICCs were seeded in 150 mm x 20 mm Petri dishes containing culture medium, [Hams F-10 medium (Sigma) + 10% pig serum] at a concentration of 1000 ICC/ 3 ml. Medium change and culture procedures were performed as outlined above.

Observation of the islets during the culture period showed that they were shedding most of the excess acinar tissue until day 6 in culture (Figure 3.1).

115 A large cluster of islets at day 0

Figure 3.1 A cluster of porcine neonatal islets surrounded by exocrine tissue freshly after isolation form pancreas. Staining is double fluorescent acridine orange/ethidium bromide. Dead cells stained red, live cells stained green (200x magnification).

The integrity of the ICCs, as identified by the surrounding capsule condition, is well maintained until day 6 and then gradually declines until day 14. Viability of the islets is assessed by staining with acridine orange/ethidium bromide. Green cells are viable whereas red cells are not. According to the literature and our experience, smaller-sized islets maintain viability in vitro better than larger islets (Figure 3.2).

116 Day 3 Day 6

A C

B D

Figure 3.2 pNICCs at day 3 and 6 post isolation and bright field microscopy x20. Images A and C show pNICCs stained with acridine orange/ethidium bromide for viability assessment under a fluorescence microscope. Images B and D show islets from the same day of culture, under a bright field microscope. In A, the surrounding acinar cells which are being shed at day 3, fluoresce in red; in C, increased dead cells, also in red, at the core of islets, are observed on day 6 (200x magnification).

3.3 Gene expression profiling of porcine neonatal islet cell clusters in culture

3.3.1 RNA extraction and CDNA synthesis from PNICCS

pNICCs were harvested after various durations of culture for nucleic acid extraction as described in chapter 2. Total RNA was isolated and reverse-transcribed to cDNA as outlined above. In total, 8 islet preparations from 8 piglets in 6 independent experiments were collected for this study.

117 3.3.2 Real Time RT-PCR and gene expression evaluation

The real time RT-PCR reaction was set up as described above. The list of Taqman primer and probe sets is given in chapter 2 (section 2.1). GAPDH and Histone H3 genes were used to normalise the gene expression across all cDNA samples. The results were analysed using One-way ANOVA and Mann Whitney U tests.

3.3.3 Histone H3 versus GAPDH as a housekeeping gene for pancreatic islets in culture

Quantitative gene expression results are often normalized to the expression levels of

“housekeeping” or “maintenance” genes as internal standards.

GAPDH (Glyceraldehyde-3-phosphate dehydrogenase) is a high-expressing gene which is frequently used for normalisation of RT-PCR results. However, according to the literature [321, 322], GAPDH, as a key enzyme in metabolism, is affected by a number of factors including the type of tissue and nutritional variations. Particularly in the upper digestive tract, GAPDH expression is largely modulated by nutritional factors. Our results from testing GAPDH expression in pNICCs and whole pancreas at different time points and conditions confirmed the fluctuation of GAPDH (Figure 3.3).

Therefore it was decided to select a more constant and preferably structural molecule as the maintenance gene. Studies show that Histone H3, a basic protein in chromatin structure, is a reliable choice [323, 324].

Real time RT-PCR was performed for both GAPDH and Histone H3 in pNICCS from 5 piglet pancreata to compare the expression level at different times, and the results were analysed by One-way ANOVA.

118 Comparing the expression level of the two housekeeping genes, GAPDH and Histone

H.3, in pNICCs by RT-PCR demonstrates significantly different patterns of expression on different days during culture. Histone H.3, a structural gene, maintains a steady expression after islet isolation throughout the culture period, whereas GPADH, a metabolic gene, is at its lowest level at day 0 after isolation. GAPDH expression is significantly upregulated between d0 and d4, followed by a modest increase until day 6.

This difference in gene expression suggests that Histone H.3 is more suitable as the internal control gene for RT-PCR in pNICCs (Figure 3.3).

GAPDH in pNICCs Histone H3 in pNICCs

 500  400  P<0.001 400  P<0.01 300 300 200 200

100 100 mRNA ExpressionmRNA level mRNA ExpressionmRNA level 0 0

Day 0 Day 4 Day 6 Day 0 Day 4 Day 6

Figure 3.3 Comparison of two housekeeping genes in pNICCs is indicative of a significant increase in GAPDH mRNA level in pNICCs post isolation until day 4 and 6, whereas Histone H3 expression remains constant during the culture period. Therefore Histone H3 is a more reliable control for normalising RT-PCR results. Error bars refer to standard deviation.

3.3.4 Gene expression in porcine neonatal islet cell clusters during the culture period

We studied a panel of genes with different functions to assess the behaviour of pNICCs post isolation and in culture. These “biomarkers” are major indicators of islet maturity and function, and can predict the engraftment prognosis [112]. β cells, the major

119 endocrine population within pancreatic islets, secrete insulin; this mechanism is regulated in a complex and multifactorial manner [325-327].

Tissue injury caused by pancreas procurement and islet isolation leads to release of chemokines and cytokines such as IL-8, MCP1 and TNFα, as well as alarmins or

DAMPs (damage-associated molecular patterns) like HMGB1 and HSP72 [113, 250,

328] [329]. These act as endogenous ligands for PRRs (pattern recognition receptors) including TLRs [330]. TLR signalling contributes to production of downstream proinflammatory molecules particularly those related to inflammation, apoptosis, cell growth, and angiogenesis in islets [206, 269].

Matrix metalloproteinases are typically found in any remodelling, diseased or inflamed tissue [331]. MMPs play a central role in cell proliferation, migration, differentiation, angiogenesis, apoptosis and host defences. Matrix metalloproteinase-2 (from the gelatinase subgroup) degrades type IV collagen, the major structural component of basement membranes.

MMP-2 is often constitutively expressed and expression is regulated at both transcriptional and post-transcriptional levels by transforming growth factor (TGF) β.

[332, 333]. Activated protein C acts via its receptor EPCR to increase gene and protein expression of MMP-2, and is able to directly cleave and activate the MMP-2 proenzyme [334]. EPCR expression can be upregulated by the presence of activated protein C (APC) and EPCR has been found to be present on endothelial cells, keratinocytes, monocytes, CD56+ NK cells, neutrophils, eosinophils, tenocytes, etc

[192].

120 The following results are RT qPCR data normalised to Histone H.3 values and analysed with One-way ANOVA and Mann Whitney U tests. Error bars are representative of standard error of the mean.

Insulin gene expression:

As the β cells within the neonatal islet cell clusters mature, they acquire the ability to secrete insulin and this is reflected in a progressive increase of insulin mRNA expression from d4 to 14 post-isolation (Figure 3.4).

Figure 3.4 Insulin gene expression, as a key marker of islet function, demonstrates a significant gradual rise from d0 to d14. This is expected as neonatal islet cell clusters gradually mature during this period.

121 Toll-like receptors:

The temporal pattern of gene expression differs between Toll-like receptor (TLR) 2 and

Toll-like receptor (TLR) 4. TLR2 expression is highest immediately post isolation and falls steadily during the culture period to reach a minimum by d10. Conversely, mRNA expression for TLR4 is initially low, reaches a peak between day 3 and 6 post-isolation with a subsequent decline, though at day 14 levels remain above baseline (Figure 3.5).

Figure 3.5 Toll-like Receptors 2 and 4 follow different patterns of expression. TLR4 is at its minimum just after isolation, increases significantly until day 6 and declines again by day 14. TLR2 on the other hand is most highly expressed at day 0, and after some fluctuations decreases significantly towards end of culture until d14.

122 Danger-associated molecular pattern molecules:

Danger signals or damage-associated molecular pattern molecules (DAMPs) are expressed by pNICCs immediately post-isolation. Both HMGB1 and HSP72 exhibit some fluctuations from day to day during culture (Figure 3.6).

Figure 3.6 Alarmins HSP72 and HMGB1, are expressed immediately after isolation and levels fluctuate until day 14 of culture.

123 Chemokines:

Different chemokines manifested very different patterns of expression - this may in part reflect the evolution of cell types in the neonatal islet cell clusters during the culture period, with loss of vascular remnants and exocrine tissue. Expression of both MIP1 and IP10 (CXCL10) fell precipitously from the day of isolation. MIP2 levels increased only modestly to d4, declining thereafter, whilst a more robust rise in the expression of

IL-8 and MCP1 was noted with peak levels attained between d4 and d6 pot-isolation, followed by a return to levels approaching baseline by d14 (Figure 3.7).

124

Figure 3.7 The chemokines examined in this study demonstrate different patterns of gene expression. MIP1 and IP10 are at their highest at day 0, and fall dramatically from day 3 onwards. MIP2 increases modestly to day 4 followed by a decline until day 14. IL-8 expression rises significantly from baseline to day 6, declining thereafter. MCP1 is significantly upregulated by day 4 falling subsequently until day 14.

125 Activated protein C-induced molecules:

EPCR gene expression in pNICCs decreases significantly from d0 to d3. This event is most likely due to loss of vascular remnants during the first days of culture. A modest increase in gene expression is observed on d4 and is maintained until d6.

mRNA transcription of MMP-2 is at its lowest levels just after isolation, but increases significantly by day 4, remaining at this level until d6 then declining towards the end of culture (Figure 3.8). MMP expressed in islets at this time may be involved in remodelling of the basement membrane and restoration of the islet structure.

Figure 3.8 EPCR is expressed by pNICCs. Expression is at its highest immediately after isolation, decreasing significantly thereafter. This probably reflects the loss of vascular remnants from the pNICCs during culture. Gene expression of MMP-2 increases significantly until d6, before declining again towards the end of culture.

126 NKG2D Ligands:

Ligands for the activating receptor NKG2D can be induced on parenchymal cells under conditions of cellular stress and the expression of these ligands renders tissues susceptible to killing by both NK cells and CD8+ cytotoxic T cells. Human NKG2D can recognise porcine ULBP1, and potentially other porcine ligands as well [335]. Thus expression of NKG2D ligands on pNICCs may have deleterious consequences following islet grafting into human recipients. Of the two known porcine ligands, pMIC2 was already maximally upregulated by d3 after isolation, with levels remaining elevated thereafter. pULBP1 steadily increases to reach a peak by d6 post-isolation.

Again, levels remain elevated through d14 (Figure 3.9).

Figure 3.9 The pattern of expression of stress-induced ligands pULBP1 and MIC2, which bind to the receptor NKG2D, differs. pULBP1 increases gradually to reach a peak on d6 of culture. MIC2 is elevated more rapidly, and is already maximally upregulated by d3. Thereafter levels remain constant.

127 Protective molecule A20:

A20 is a protective molecule which acts to prevent NF-κB translocation into the nucleus and thus prevents transcription of pro-inflammatory cytokines and chemokines.

Since NF-κB is an important mediator of signalling downstream of Toll-like receptor engagement, A20 can be considered as an inhibitor of TLR-mediated effects.

A20 mRNA levels peaked between d4-6 post-isolation declining thereafter to levels approximating baseline values (Figure 3.10).

Figure 3.10 Expression of A20, known to moderate inflammation by abolishing TNF- induced NF-kappa B signalling, is greatest on d 4-6 post isolation and returns to baseline levels by the end of culture.

128 3.4 Activated protein C and the endothelial protein c receptor

Protein C is a serine proteinase made by the liver. When activated, it has multiple properties including: anti-coagulant, cytoprotective (mainly through its receptor

EPCR), anti-inflammatory, anti-apoptotic (via EPCR and protease-activated receptor-1,

PAR-1), pro-angiogenic (via matrix metalloproteinase activation), and endothelial barrier protection (via induction of sphingosine kinase-1, SpK1, and up-regulation of sphingosine-1-phosphate [187, 188, 194, 201, 318].

Protein C is activated by the thrombin-thrombomodulin complex when it is bound to its cell surface receptor, endothelial protein C receptor (EPCR). It mediates the pleiotropic cytoprotective effects of activated protein C (APC) on cells through PAR1 signalling,

(PARs are members of the G-protein-coupled receptor superfamily whose cellular activation is propagated through various G-protein and non-G-protein pathways, resulting in a diverse array of physiologic outputs) [336]. Activated protein C can modulate the expression of some critical proinflammatory and proapoptotic pathway genes resulting in downregulation of these pathways and reciprocal up-regulation of anti-inflammatory and anti-apoptotic pathways. Activated protein C suppresses NFκB nuclear translocation and blocks expression of TNFα and other downstream genes

[188]; APC also upregulates expression of chemokines MCP1 and IL-8, and cytokines such as IL-6 [195, 196]. Matrix metalloproteinase-2 (MMP-2) unlike other MMPs, is resistant to cleavage by most enzymes, but APC has been shown to selectively upregulate gene and protein expression of MMP-2 and directly cleave and activate the proenzyme [197]. Modulation of MMP-2 expression is mediated by EPCR, blocking

EPCR prevents APC-induced MMP-2 synthesis and proliferation [201].

129 Endothelial protein C receptor is engaged by Factor VII/VIIa, as well as binding activated protein C [193]. It is a type I transmembrane protein which is similar to the major histocompatibility class 1/CD1 family of proteins. EPCR is essential for the limitation of endotoxin-induced innate immune responses [337, 338] and is expressed on endothelial cells as well as a number of other cell types including islet cells [308].

EPCR is expressed at significantly higher levels on the endothelium of large vessels than on the microvascular endothelium of most organs. It mediates activation of protein C by thrombin, as well as the cytoprotective and anti-inflammatory effects of protein C. Activated protein C upregulates expression of EPCR, while thrombin and

IL-1β are also known to increase EPCR gene expression [339].

EPCR is shed from the cell surface into the circulation upon proteolytic cleavage by

ADAM (A disintegrin and metalloproteinase) 17, also known as tumor necrosis factor

(TNF)-α converting enzyme (TACE), resulting in truncation of the cytoprotective effects of protein C. Increased shedding is observed during severe inflammation, due to high local levels of IL-1β, TNF α, and thrombin [339].

3.4.1 Application of rhuAPC in islet transplantation

According to published reports, application of rhuAPC can reduce inflammation, apoptosis and immunogenicity of islets prior to transplant [192, 306]. It also decreases

IBMIR through NFkB inhibition [117] and promotes angiogenesis and engraftment [9,

200]. Therefore we hypothesized that rHuAPC could ameliorate inflammation in pNICCs in culture.

130 3.4.2 Preliminary data suggesting possible rHuAPC effectiveness on pig tissue

Human Protein C mRNA is 83% homologous to pig protein C mRNA. Human and porcine APC proteins are 74% identical and the thrombin cleavage site on protein C is

100% homologous in pig and human (Figures 3.11 and 3.12).

The sequence of porcine EPCR had been established by the group of Associate

Professor Peter Cowan. These investigators kindly gave us access to the data prior to deposition. The cDNA sequence was subsequently assigned the accession number

NM_001163406 (Table 3.1).

Human Pig

Protein C mRNA NM_000312 NM_213918.1

Protein C preprotein NP_000303 NP_999083.1

EPCR mRNA NM_006404 NM_001163406

EPCR protein NP_006395 NP_001156878

Table 3.1 List of accession numbers for human and porcine protein C and endothelial protein C receptor mRNA and proteins.

Alignment of the human and porcine EPCR (protein sequences translated in silico from the cDNA) demonstrate that 7/10 of the residues important for APC binding were identical between pig and human while the remaining 3 residues were from similar families (Figure 3.13 and Table 3.2). These findings suggest that rHuAPC may be effective in engaging EPCR on porcine tissue.

131

Figure 3.11 Alignment between human (Query) and porcine (Subject) protein C mRNA. The two sequences are 83% homologous in human and pig.

132

Figure 3.12 Human (Query) and porcine (Subject) protein C preproprotein sequences alignment. The proteins are 74% identical, the cleavage site by thrombin (highlighted in blue) is 100% homologous in human and pig.

133

Figure 3.13 Human (Query) and porcine (Subject) EPCR proteins alignment. In the binding sites (highlighted in blue) in the picture above, 7 residues out of 10 are identical in human and pig proteins, 3 other residues are from the same amino acid families (i.e. conservative substitutions).

Human EPCR Pig EPCR Amino Acid family

R (Arginine) Q (Glutamine) Amid

L (Leucine) V (Valine) Hydrophobic

F (Phenylalanine) Y (Tyrosine) Aromatic

Table 3.2 EPCR binding site residues in human and pig. There is 100% protein similarity and 72% identity between porcine and human activated protein C binding sites of EPCR proteins.

134 3.5 Obtaining full length EPCR from porcine neonatal islet cell clusters and other cell types

3.5.1 Amplification of EPCR from pNICCs, PAECs and PIECs by rt-PCR

To investigate the expression of EPCR by pNICCs, we attempted to amplify the full length EPCR coding sequence (CDS, 729 bp) from pNICCs along with porcine aortic endothelial cells (PAECs) and porcine intestinal endothelial cells (PIECs) as positive controls. As described earlier in the methods chapter, extracted RNA from these cells was used as a template to set up conventional PCR using a KOD hot start DNA polymerase kit (Novagen, EMD Biosciences, Inc., Merck KGaA, Darmstadt, Germany) in a Palm Cycler Gradient PCR Machine (Corbett, Qiagen). The PCR products were loaded onto a 1% agarose gel at the maximum possible volume (e.g. 50 μl). After electrophoresis, the bands were visualized using a Syngene G:Box Chemi HR16 apparatus under UV light. In the picture below, the presence of a PCR product of the expected size, 729 bp can be seen in the gel (Figure 3.14).

135 Ladder PAEC 1 PAEC 2 PAEC 3 PIEC 1 PIEC 2 NICC 1 NICC 2

Figure 3.14 Agarose gel under UV illumination showing that a PCR product of the expected size (729 bp) can be amplified from PAEC, PIEC and NICCs.

The bands were cut out of the gel and DNA was extracted following the method explained in chapter two (section 2.1.11). DNA was quantified by absorption using a

Nanodrop spectrophotometer. The yield from PCR was relatively low in repeated experiments, suggesting that the template was not present at high concentration (Figure

3.15).

Figure 3.15 PCR product quantification by nanodrop after isolation from gel.

136 The DNAs were sent for sequencing along with EPCR forward and reverse sequencing primers (chapter 2, section 2.1.12). The full length EPCR CDS sequence was only obtained from PAECs. The PIECs and pNICCs failed to yield a result in repeated experiments. The Table below shows the alignment between the PCR product sequencing report and pig EPCR CDS. This sequencing report was blasted against the

GenBank database and aligned with the pig EPCR CDS with 99% identity (Figure

3.16).

Figure 3.16 Alignment between PCR product sequencing result from PAECs (Query) and pig EPCR CDS (Subject).

137 Since we were unable to obtain reliable sequence data from directly PCR products, the products were cloned into a plasmid vector for large scale DNA preparation prior to repeated attempts at sequencing.

3.5.2 Cloning the porcine EPCR sequence

The porcine EPCR sequence from pNICCs and the positive control cell line PAEC were amplified again by rt-PCR (Figure 3.17). The blunt-ended DNA amplicons were then cloned into the pcDNA/3.2/V5/GW/TOPO vector. Next, chemically competent E.Coli were transformed by the plasmid containing the EPCR insert. Bacterial colonies were cultured; subsequently the plasmid of interest was purified from the competent cells as above. The extracted DNA was quantitated and sequenced using EPCR forward and reverse primers (Figures 3.18 and 3.19). The procedure is described in chapter 2

(section 2.1.12)

Figure 3.17 Nanodrop estimation of the quantity of DNA amplicons to be used for TOPO-cloning into vector and transformation of chemically competent bacteria.

138 Ladder PAEC NICC d4 NICC d10

Figure 3.18 Gel electrophoresis of the PCR product shows the band of interest at the expected size (729 bp). After that the amplicons are cloned into vectors, chemically competent E. Coli are transformed with the vector and cultured, subsequently plasmid containing the EPCR sequence are purified from the bacteria.

Figure 3.19 Nanodrop quantification of the plasmid purified from the bacterial colonies. The plasmid containing the EPCR sequence is tested for quantity and quality, and sent for sequencing.

139

Figure 3.20 Chromatogram for cloned EPCR from PAECs.

140

Figure 3.21 Chromatogram for cloned EPCR from pNICCs day 4 post-isolation.

141

Figure 3.22 Chromatogram for cloned EPCR from pNICCs day 10 post-isolation.

142

Figure 3.23 Alignment between the cloned sequence from pNICC at day 4 post- isolation and porcine EPCR mRNA. The sequences are 99% homologous.

143

Figure 3.24 Alignment between the cloned sequence from pNICCs at day 10 post- isolation and porcine EPCR mRNA. The sequences are 98% homologous.

Sequencing of cloned PCR products demonstrates that full-length EPCR is transcribed in pNICCs day 4 and day 10 post-isolation, as well as in the positive control cell line

PAEC (Figures 3.20-3.24). Hence rHuAPC could potentially render its effects through its receptor EPCR in pNICCs, assuming that that protein expression by the cells reflects gene transcription.

144 3.6 Porcine aortic endothelial Cell (PAEC) treatment with rHuAPC

To examine the efficacy of our protein of interest on porcine tissue, it was decided to test it on a primary porcine cell line which strongly expresses EPCR.

PAECs were isolated from piglet thoracic aortas as described in the methods chapter.

The primary cells were isolated by enzymatic digestion then cultured and treated with rHuAPC at a 10μg/ml concentration for 24 or 48 hours. The treatment and control wells, as well as freshly isolated cells (baseline) were harvested for mRNA extraction and cDNA synthesis. This experiment was repeated 3 times. A panel of genes were studied by real time q-PCR. RT-PCR results are normalised to Histone H.3 values.

Data are analysed using One-way ANOVA and t tests.

During culture of PAECs, gene expression for EPCR is increased significantly within

24 hours and returns to baseline levels after 48 hrs. Although the mRNA level is lower in the non-treated group compared to the treated one at 24hr, no significant effect of

APC on EPCR gene expression is observed in this experiment (Figure 3.25).

145 Endothelial Protein C receptor:

Figure 3.25 EPCR is known to be upregulated by the presence of activated protein C and downregulated by cytokines and inflammatory mediators. A) EPCR expression level increases significantly within 24 hours then declines towards baseline by 48 hours. B) Despite a modest increase in mRNA level in the APC-treated group after 24 hour, no significant difference of EPCR gene expression following APC treatment observed.

146 Matrix Metalloproteinase 2:

Endothelial cells express matrix metalloproteinase (MMP)-2 in response to stimulation.

Activated protein C induces gene expression of this multi-effect enzyme and directly activates the protein. Expression of MMP-2 decreases significantly compared to the baseline level after 24 hrs in culture, and remains low until 48 hours. APC did not change the expression of MMP-2 in this experiment (Figure 3.26).

Figure 3.26 MMP-2 is known to be upregulated by APC. A) A dramatic decline in MMP-2 mRNA expression is observed within 24 hours of culture. B) MMP-2 gene expression is not affected by the APC treatment.

147 Toll-Like Receptors:

Toll-like receptor expression can be upregulated by inflammatory mediators, further amplifying downstream proinflammatory signalling. Endothelial cells express TLRs.

During 48 hours of culture, expression of both TLR2 and TLR4 are significantly decreased in PAECs. TLR2 is downregulated by APC at 48 hours, but the extent of change in TLR4 expression in the presence of APC was much less (Figure 3.27).

Figure 3.27 APC is expected to inhibit the expression of these inflammatory-response mediators. A) The expression of both TLR2 and 4 significantly declines during culture. B) Activated protein C-treated cells express decreased levels TLR2 after 48 hours, but little difference in the expression of TLR4 is detected.

148

Danger-associated molecular pattern molecules:

Heat shock protein 72, a DAMP, can be released during tissue injury and activate TLR signalling. PAECs can release HSP72 under stress. APC treatment does not alter the expression profile of this molecule during culture (Figure 3.28).

Figure 3.28 Danger-signal HSP72 stimulates Toll-like receptor signalling during inflammation. A) PAECs show fluctuations in gene expression for HSP72 during the culture period. B) APC does not change the HSP72 expression pattern in PAECs in culture.

149 Cytokines and chemokines:

The anti-inflammatory effects of activated protein C are partly rendered thorough blocking of TNFα expression and suppression of NF-κB nuclear translocation with subsequent downregulation of downstream proinflammatory gene transcription. APC is known to suppress the expression of various chemokines and adhesion molecules, but has been shown to upregulate MCP1 and IL-8. We did not observe a specific effect of

APC treatment on the expression of TNFα by PAECs, although the expression level declined significantly within 48 hours of culture. MCP1 and IL-8 gene expression, as expected, were significantly elevated within 24 hours of culture in the APC treated cells, when compared to the baseline values. The levels returned to baseline after 48 hours in both treated and non-treated cells. MIP1, MIP2 and IP10 levels were decreased dramatically during the 48 hours of culture, APC treatment did not alter MIP1 and

MIP2 expression, although a modest increase in the level of IP10 was detected after treatment for 48 hrs (Figures 3.29 and 3.30).

Tumour necrosis factor:

Figure 3.29 A) TNFα expression by PAEC decreases significantly within 48 hours in culture. B) Treatment with APC does not affect the TNFα mRNA levels.

150 Chemokines:

151

Figure 3.30 Gene expression of chemokines MCP1, IL-8, MIP1, MIP2 and IP10 was evaluated. MCP1 and IL-8 exhibit different patterns of expression compared to the other chemokines. A) mRNA levels for MCP1 and IL-8 are significantly increased after 24 hrs, but return to baseline by the end of culture. MIP1, MIP2 and IP10 mRNA expression falls dramatically during the experiment period. B) Treatment with APC, as expected, significantly elevated MCP1 and IL-8 expression at 24hr, but not at the later timepoints. MIP1 and MIP2 did not change noticeably with APC, although IP10 was modestly increased in the treated cells after 48 hours.

152 NKG2D ligands:

Ligands for NKG2D are stress-induced molecules causing immune activation of cells.

TLR signalling and cytokine stimulation increase the expression of these ligands. They are also expressed by primary cell lines in culture during proliferation. Activated protein C might downregulate the expression of these molecules by suppressing inflammatory pathways. pULBP1 is decreased in PAECs after 24 hours and remains low, without responding to APC treatment. However, MIC2 rises significantly within

24 hours of culture. Similar to the pULBP1, MIC2 levels are not altered by the presence of APC (Figure 3.31).

153

Figure 3.31 A) Expression of pULBP1 and MIC2 follows different patterns in PAECs during culture. pULBP1 gene expression decreases significantly during 48 hrs, whereas MIC2 expression is elevated within 24 hrs and remains elevated. B) Neither of these genes is affected by the presence of activated protein C.

154 Protective molecule A20:

An endogenous regulator of pattern recognition receptor signalling, A20 is strongly induced by TLR signalling and TNFα. Activated protein C can upregulate A20 expression mainly through PAR1 signalling. PAECs expressed A20 in culture, although the levels decreased significantly after 24 hrs and remained low. Treatment of these cells with activated protein C did not change A20 mRNA levels during the culture period (Figure 3.32).

Figure 3.32 A) A20 mRNA level in PAECs is reduced significantly after 24 hrs in culture. B) A20 expression is higher in the APC-treated group after 24 hrs, but falls to a lower level compared to the non-treated group within 48 hrs.

Of the genes investigated, some, but not all, changed expression as expected in response to APC treatment. Expression of the chemokines MCP1 and IL-8 was significantly increased in response to APC treatment within 24 hours of culture, in accordance with the known effects of APC. Expression of other genes such as EPCR,

A20 and TLR4, was modestly induced within 24hrs, whilst IP10 expression was augmented in treated cells after 48hrs. Expression of MMP-2 was unaltered, contrary to

155 expectations. In general the effects of culture overshadowed the effects of APC treatment.

3.7 Antibody against human EPCR is active on porcine tissue but EPCR protein was not detected in islets

To confirm the functionality of the polyclonal rabbit anti-human EPCR primary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) on a porcine tissue, HRP immunostaining was performed on human and porcine carotid arteries as a positive control. If this antibody is able to stain porcine endothelial tissue, it can be assumed that it is active on other porcine tissues and therefore useful as a reagent to determine the presence of EPCR protein in pNICCs

3.7.1 Anti-EPCR staining of human carotid artery

Immunostaining of paraffin-embedded human carotid artery sections for EPCR was performed following the protocol described in chapter two (section 2.4).

156

Figure 3.34 Anti-EPCR staining on human carotid. To verify the activity of the primary antibody on a positive control tissue, human carotid section was stained with anti-EPCR antibody. The endothelium of the vessel is disrupted; nevertheless, the vasa vasorum endothelial layer shows positive staining.

The endothelium of the lumen of the artery appears to be compromised in most areas throughout these specimens. However, focusing on the vasa vasorum in the tunica adventitia shows some intact endothelium staining positive for EPCR. The primary antibody is active on the human tissue (Figure 3.34).

157

3.7.2 Anti-EPCR staining of piglet thoracic artery

Next, we proceeded to staining piglet aorta for EPCR to detect the activity of the antibody on porcine tissue (Figure 3.35).

Figure 3.35 Anti-EPCR staining on piglet thoracic aorta. The antibody is active on porcine tissue, staining the EPCR protein in different layers of this tissue where endothelial elements are found.

Similar to the staining pattern from human carotid, a loose layer of endothelium is present lining the lumen. The endothelium of vasa vasorum is strongly positive for

EPCR. The anti-EPCR primary antibody is active on porcine tissue.

158 3.7.3 Anti-EPCR staining of porcine neonatal islet cell clusters

According to the previous tests on human and porcine tissues, expression of EPCR protein in pNICCs should be able to be determined using this method. Immunostaining was performed as described in chapter two, on both paraffin-embedded formalin-fixed and frozen porcine neonatal islet cell clusters from day 6 and day 10 post-isolation in culture.

Figure 3.36 Anti-EPCR immunostaining on pNICCs from day 6 and day 10 post- isolation. The positive staining surrounding the islets is related to chormogen entrapment or tissue necrosis rather than positive antigen detection (200x magnification).

In the previous experiments it was shown that anti-EPCR antibody can detect EPCR protein in both human and porcine tissues. However when investigating the presence of this protein in pNICCs – from both paraffin-embedded and frozen sections - there appears to be false positive staining surrounding the islets but no true positivity,

159 according to examination by islet experts. This can be explained by various factors such necrosis of the islet rim or trapping of the chromogen under the specimen edges.

3.8 Treatment of pNICCs with recombinant human activated protein C

3.8.1 Effectiveness of APC on porcine tissue

Earlier in this chapter, and in chapter one, we proposed that rHuAPC could be active on porcine tissue, in the presence of endothelial protein C receptor. A dose of rHuAPC which had been previously shown to be effective in vitro on a variety of cell types from different species was chosen [308, 340] [200] and treatment of positive control primary porcine endothelial cells (PAEC) yielded some changes in gene expression consistent with APC activity on these cells. Whereas we had been able to isolate full-length cDNA transcripts for EPCR from pNICC at both d4 and d10 following isolation, we were unable to detect the presence of EPCR protein on the surface of pNICCs. We could not exclude the possibility that EPCR protein had been present in NICCs in culture but had been removed by enzymatic cleavage during sample collection or preparation; it was decided to investigate the effect of rHuAPC on pNICCs in culture.

This series of experiments was repeated 3 times with islets from 8 animals in total.

160 3.8.2 Treatment, harvest and analysis pNICCs isolation, culture and treatment with rHuAPC was performed as described in chapter two. The amount of serum was maintained at 10% in the media, and rHuAPC was applied at 10μg/ml concentration (since in vivo investigation of rHuAPC action on islets was not feasible, this concentration was determined based on the literature and previous works with rHuAPC on various cell and tissue types). pNICCs were harvested for different analyses including gene expression and immunohistochemistry. The analysis below shows gene expression results from quantitative rt-PCR, the values are normalised to histone H.3 and analysed by One-way ANOVA and t test (Figures 3.37-

3.43).

The baseline values for gene expression from day 0 and day 3 are shown earlier in this chapter (section 3.3.4). The figures below represent only the groups included in the rHuAPC-treatment experiment.

Islet marker:

Figure 3.37 mRNA transcription of insulin increases during culture, as previously observed. APC can potentially exert protective effects, maximising the β cell mass of islets, however the insulin gene expression is not altered by treatment with this agent.

161 APC-induced molecules:

Figure 3.38 Both EPCR and MMP-2 are known to be upregulated at the gene level in the presence of APC. Porcine neonatal islets do express these molecules; nonetheless we did not observe an effect of APC on the mRNA levels for these genes.

Toll-Like Receptors:

Figure 3.39 Toll-like receptor 2 and 4 expression decreases from day 3 in culture. No significant result of APC-treatment is detected here.

162 Danger-associated molecular pattern molecules:

Figure 3.40 HSP72 transcription appeared reduced on d14 of culture in pNICCs treated with APC, but in other respects, gene expression of DAMP molecules HSP72 and HMGB1 did not alter in response to APC treatment.

163 Chemokines:

Figure 3.41 (To be continued in next page)

164

Figure 3.41 Expression of all five chemokines, MIP1, MIP2, MCP1, IL-8 and IP10 gradually decreased in culture from day 6. Gene expression of MCP1 and IL-8 are known to be upregulated by APC, and whilst their expression was increased in PAEC treated with rHuAPC, we did not observe an effect of this anti-inflammatory agent in pNICCs.

NKG2D ligands:

Figure 3.42 Gene expression of stress-induced ligands pULBP1 and MIC2 remains elevated change between day 6 and day 14 of culture. Expression of pULBP1 appeared reduced in APC-treated islets on d14, in other regards the expression is not influenced by addition of APC to the culture medium.

165 Protective molecule A20:

Figure 3.43 A20 expression in cultured pNICCs is not stimulated by APC treatment.

Gene expression of 15 molecules in porcine neonatal islet cell clusters in response to addition of recombinant human activated protein C to the culture medium was evaluated by RT-PCR. The experiment commenced at day 6 post-isolation, when most of the proinflammatory genes had reached their maximum level of transcription. We did not detect any particular pattern of rHuAPC effect on gene expression within pNICCs. Although we had demonstrated some effect of rHuAPC treatment on freshly- isolated porcine aortic endothelial cells, relative lack of effect on pNICCs may relate to inability of the rHuAPC to penetrate the islet cell mass, or lack of expression of EPCR on the cells of the islet cluster.

166 3.9 Discussion

Human islets are only available for research purposes if they have been deemed unsuitable for transplantation, and as such, these islets may not be typical or representative of isolated islets as a whole. Porcine neonatal islet clusters are isolated from pigs without prior brain death, but in other regards are subjected to similar stressful events as human islets procured for transplantation.

Porcine neonatal islet cell clusters are isolated from the pancreas using enzymatic digestion, and maintained in culture for up to several weeks. During this time the islets mature, shedding exocrine tissue and acquiring the ability to secrete insulin and other hormones involved in glucose homeostasis. We studied the pattern of expression of a number of genes involved in inflammation within isolated islets – these include danger signals, the pattern recognition receptors thorough which they act, cytokines, chemokines and stress ligands downstream of TLR signalling, and finally the protective molecule A20.

Expression of these contributors to inflammatory processes peaked at different times post-isolation, such that it was difficult to determine unequivocally a time when inflammation was at its minimum (and thus the most favourable time for intraportal infusion).

Levels of most proinflammatory cytokines and chemokines had declined by d14 and this was coincident with robust expression of insulin by ICCs. Although from this standpoint d14 would appear to be a favourable time for transfer, expression of the

167 NKG2D ligands pMIC2 and pULBP1 remained high at this time, possibly exposing

ICCs to heightened cell-mediated cytotoxicity by NK cells and CD8+ T cells.

Treating murine islets in culture with recombinant APC had been reported to reduce inflammation, prevent islet apoptosis and prevent IBMIR [306]. We hypothesized that rHuAPC could have beneficial activity on porcine NICCs. Similarities between the sequences of human and porcine EPCR suggested likely cross-species reactivity and we were able to isolate full-length transcripts for EPCR from NICCs at both d4 and day 10 of culture. However, immunostaining failed to provide evidence of the presence of

EPCR at a protein level in porcine ICCs. In accordance with this, we were unable to demonstrate changes in gene expression for molecules such as IL-8 and MCP1which are known to be affected by APC signalling through EPCR. Conversely, treatment of the positive control cell type porcine aortic endothelial cells did result in a modest increase in the expression of these genes and in a reduction in TLR4 expression. In these studies, whilst we did not perform a dose response experiment, we used a dose which had been effective in vitro with a variety of cell types from different species

[200, 308, 318] [201]. The inconsistent effects of APC on the expression of different genes could partly be accounted for by loss of activity of APC during the culture period. APC has a half-life of 20-30 minutes in circulation due to inactivation by serum components. Nonetheless, its biological effect has a much longer duration: twice- weekly dosing with rHuAPC protected NOD mice from the development of autoimmune diabetes over a 10 week period [308]. Other possible arguments for our observations include lack of penetration of rHuAPC into ICCs as well as overpowering of EPCR-PAR1 signalling by other pathways.

168 Based on these findings we decided not to pursue the strategy of treating NICCs in culture with rHuAPC further, and instead to address the ability of a clinically relevant protocol for hormone resuscitation of the brain-dead donor to reduce inflammation and improve function of the pancreas and islets in the setting of donor brain death. The results of these experiments will be described in chapter 4.

169

CHAPTER 4 HORMONE RESUSCITATION OF THE BRAIN-DEAD DONOR AND ITS EFFECT ON PANCREATIC INFLAMMATION AND FUNCTION

170

4.1 Introduction

Brain death results in a series of haemodynamic, neurohormonal and pro-inflammatory perturbations, all of which contribute to donor organ dysfunction, and pancreatic islets are particularly susceptible to damage during brain death and its sequelae.

Observations of pituitary hormone depletion after brain death (BD) prompted the use of hormone replacement (HR) treatment protocols, which may ameliorate some of the metabolic consequences of brain death, and improve the quality of donor organs.

Hormone resuscitation allows the salvage of some hearts that would otherwise be deemed unsuitable for transplantation, but it has not been determined whether the benefits of hormone resuscitation extend to other organs such as pancreas [231, 270]. A large animal model study is desirable to establish the effect of HR versus standard noradrenaline-based resuscitation protocols on pancreas quality.

The aim of this study is to utilise a well-characterised pig model of brain death to compare the effects of hormone resuscitation and standard noradrenaline treatment on hemodynamic parameters, and markers of pancreatic function and inflammation.

171 4.2 Physiological changes in BD pig

4.2.1 Haemodynamic status and pancreatic enzyme release

Inflation of an intracranial Foley catheter was accompanied by a typical Cushing response, with hypertension and tachycardia. One hour after the induction of brain death, animals were commenced on either noradrenaline infusion titrated to maintain a mean arterial pressure 60-70 mmHg or a hormone-resuscitation cocktail comprising

Methylprednisolone, triiodothyronine, insulin and arginine-vasopressin titrated to maintain MAP as above. Heart rate was significantly greater in noradrenaline-treated pigs than in those receiving the hormone cocktail at all times between 2 and 6 hours post brain-death (Figure 4.1a).

For evaluation of haemodynamic status, data from 9 pigs in each group was analysed.

Mean arterial pressure was better maintained in HR pigs (64±5 vs. 36±14 mmHg at 6 hrs post BD induction, p<0.05) (Figure 4.1b). Cardiac output and stroke work were also better preserved in pigs receiving hormone resuscitation (Figure 4.1c and d).

172 a b

c d

Figure 4.1 Monitoring of haemodynamic status in brain dead pigs. a) tachycardia is less pronounced in HR. b) Mean arterial pressure is better-maintained with HR. c) Cardiac output increases with hormone treatment. d) Improved stoke work is observed in the HR group.

PaO2/FiO2 ratio was higher in HR than in NA pigs (470±95 cf. 345±106, p<0.05), and alveolar-arterial gradient was elevated in NA pigs (311±101 vs. 192±91 mmHg, p<0.05) implying worse lung function in the latter group.

Renal arterial flow was greater in HR than in noradrenaline-treated animals, and accordingly the HR pigs had higher creatinine clearance during the 6 hours following brain-death (Figure 4.2). There were no differences in hepatic arterial or portal venous flow, and no differences in AST, ALT, bilirubin or bile production between groups.

173

Figure 4.2 Renal artery flow and creatinine clearance are improved in pigs receiving hormone resuscitation compared to those receiving noradrenaline treatment.

Perfusion of the pancreas was not directly measured but levels of the pancreatic enzymes amylase and lipase did not change significantly during the observation period nor differ between treatment groups (Figure 4.3).

Figure 4.3 Pancreatic enzyme release was measured during the maintenance period following induction of brain death. The level of amylase and lipase in serum did not change after 6 hours. There was no difference between two treatment methods.

174 4.2.2 Histological appearance of pancreas from brain-dead pigs

H&E stained sections of paraffin embedded formalin-fixed pancreas were examined by an experienced transplant pathologist, blinded to the treatment group and scored according to the matrix below (Table 4.1).

Grade Necrosis Perivascular Interstitial

neutrophils neutrophils

Mild, focal vessels 1 Single cell necrosis Mild, focal only Clusters of necrotic Mild, general OR 2 Mild, most vessels cells moderate, focal Confluent necrosis Moderate (> 1 cell Moderate or marked, 3 >10% thick) most vessels general

Table 4.1 Sections were stained with H&E and scored based on this matrix by an experienced transplant pathologist blinded to the treatment groups.

Whereas the appearance of pancreas samples from HR pigs at 6 hours post BD induction remained normal, pancreas samples from NA pigs displayed focal necrotic areas with nuclear debris and neutrophil infiltration (Figures 4.4 and 4.5).

175 A B

Figure 4.4 Haematoxylin and Eosin-stained sections from the pancreas of pigs in the hormone (HR) and noradrenaline (NA) treatment groups at 6 hrs post brain death. The appearance of the pancreas in the HR group remained normal (A), whereas the pancreas from NA animals showed areas of focal necrosis (B). The arrows in image B are pointing at focal necrotic areas.

A B

Cellular necrosis in pancreas Total damage in pancreas ** 3 4

3 2

2 Score Score 1 1

0 0 Noradrenalin Hormone Noradrenalin Hormone

** p<0.01

Total damage= Sum score of necrosis+ perivascular neutophils+ interstitial neutrophil

Figure 4.5 The histological damage caused by brain death within 6hrs in pig pancreas was examined by an experienced transplant pathologist and scored according to the matrix shown in table 4.1. Pancreata from noradrenaline-treated (NA) animals exhibited significantly more cellular necrosis than those from hormone resuscitated (HR) pigs. The total damage to the tissue was also higher in NA group; this difference did not reach statistical significance.

176 4.3 Gene expression in pancreas from brain-dead pigs

4.3.1 Pancreas yields RNA of relatively low integrity

1μg of total extracted RNA from frozen pancreas specimens was transcribed to complementary DNA using reverse transcriptase. However, as determined by

Bioanalyzer measurements, the integrity of RNA from all the pancreatic samples was suboptimal after repeated attempts at RNA extraction (Figures 4.6 and 4.7). Thus, random hexamers were used instead of Oligo-dT primer to prime the mRNA for subsequent cDNA synthesis. Oligo-dT only binds to the mRNA poly A tail, whereas random hexamers bind to a variety of different sites on the mRNA and initiate DNA strand elongation. Random hexamers can thus be used as universal, non-specific primers for cDNA synthesis, even from fragmented RNA.

A B

Figure 4.6 Absorbance spectra of total RNA generated from electrophoresis in the Bioanalyzer. (A) A poor quality RNA spectrum versus a good quality RNA spectrum (B), both from BD pig pancreas.

177

Figure 4.7 Absorbance data for RNA from brain-dead pig pancreas. The 260/280 ratio indicating the purity of RNA should be 1.8 or more.

The real time RT-PCR reaction was set up as described in chapter two. Results were normalised to GAPDH and Histone H3 values. Statistical analysis was performed using

Mann Whitney U and One-way ANOVA tests.

4.3.2 Evaluating inflammation by determining gene expression for inflammatory mediators in BD pig pancreas

Signalling through pattern-recognition receptors such as TLR4 has been implicated in the inflammatory consequences of brain-death.

Here, we used RT-PCR to analyse gene expression for two prominent danger signals

HSP72 and HMGB1 which are released from cells during injury, as well as pattern- recognition receptors (TLR 2 and 4) and downstream genes in the TLR signalling pathway, including cytokines, chemokines and NKG2D ligands. Additionally, we examined the expression of the protective molecule A20, which acts to terminate NF-

κB- mediated gene transcription, and of insulin, as a measure of islet activity and

178 function. The following figures display mRNA expression as fold-change between baseline and termination of the experiment at 6 hr post BD, the values are normalised to the Histone H3 housekeeping gene. Each individual data point represents an animal, with the median indicated.

Damage-associated molecular pattern molecules:

Expression of two DAMPs (damage-associated molecular pattern molecules) or endogenous danger signals HMGB1 (high mobility group box 1) and HSP72 (heat- shock protein 72) were evaluated by RT-PCR. They act as extracellular immune- signalling molecules. Hormone- resuscitation downregulates mRNA levels of HMGB1 and HSP72 compared to noradrenaline treatment; however, the difference between groups is not statistically significant (Figure 4.8).

Figure 4.8 Expression of both danger signals HMGB1 and HSP-72 is elevated in the noradrenaline-treated group compared to the hormone-treated animals after 6 hours of brain death, although no significant difference is observed.

179 Toll-Like Receptor 4:

Gene expression of TLR4, a pattern recognition receptor (PRR) which is activated during innate immune responses was evaluated by RT-PCR. Expression of this receptor was upregulated in the pancreas of animals receiving Noradrenaline treatment within 6 hrs of brain death, but not in that of pigs treated with hormone resuscitation, (Figure

4.9).

Figure 4.9 Expression of TLR4, a pattern recognition receptor (PRR) that can be activated by danger signals, is upregulated in the pancreas of pigs treated with noradrenaline, but not in the pancreas of those receiving hormone resuscitation during 6 hours post BD.

180 Chemokines:

Expression of four chemokines was evaluated; MCP1 (recruiting macrophages), MIP1

(activating granulocytes/neutrophils), IP10 or CXCL10 (recruiting NK cells, DCs, macrophages and T cells) and IL-8 (recruiting neutrophils and other target cells).

Pancreatic expression of all these chemokines was downregulated in hormone resuscitation pigs, whilst in pigs receiving noradrenaline treatment, levels increased or remained stable during the six hours following BD induction. The increase in MCP1 gene expression over baseline was significantly greater in noradrenaline-treated pigs than in animals receiving HR (Figure 4.10).

181

Figure 4.10 Gene expression of MCP1, MIP1, IP10 and IL-8 were measured by RT- PCR. In the hormone-receiving groups, pancreatic chemokine expression was downregulated, whereas chemokine expression increased or remained stable in the noradrenaline-treated group during the 6 hours after BD induction. The increase in MCP1 gene expression over baseline was significantly greater in the noradrenaline- treated pigs than in the animals receiving hormone resuscitation.

182 Cytokines:

The expression level of two pro-inflammatory cytokines was measured by RT-PCR.

Both IL-1β and IL-6 play an important role in immune response mediation, cell activation and proliferation. The expression of IL-6 in the pancreas was increased in noradrenaline-treated pigs (median 1.74 at baseline, 12.97 at 6 hours post BD induction), whilst in the hormone-resuscitation animals, IL-6 expression declined over the same period (median 2.22 at baseline and 0.61 at 6 hours). IL-1β gene expression remained stable in noradrenaline-treated pigs but was significantly downregulated by hormone-resuscitation (median 4.75 at baseline decreasing to 2.32 after 6 hours)(Figure

4.11).

Figure 4.11 Gene expression of IL-6 was upregulated within 6 hours of brain death in the pancreas of pigs receiving noradrenaline pressor support. In contrast, IL-6 expression declined in HR pigs over the same interval. IL-1β expression level is downregulated in the hormone-treated group during this period.

183 NKG2D Ligands:

Expression levels of two NKG2D ligands, stress-induced molecules MIC2 and pULBP1 were evaluated by RT-PCR. pULBP1 did not amplify despite multiple attempts. Brain death induction increases the mRNA levels of MIC2 in pancreas compared to the baseline levels, in the noradrenaline-treated group (mean 0.64±0.14 at baseline, 1.97±0.4 after 6 hours, P=0.046) but not in pigs receiving hormone resuscitation (0.69±0.13 at baseline and 0.76±0.17 at 6 hours, P = NS) (Figure 4.12).

Figure 4.12 Gene expression of the stress-induced protein MIC2, a ligand for NKG2D, can be induced by brain death in porcine pancreas. Noradrenaline treated pigs manifest a significant increase in MIC2 expression in the pancreas at 6 hours after BD induction, which was not observed in HR pigs. pULBP1 did not amplify from pig pancreas despite multiple attempts.

184 Protective molecule A20:

Gene expression of A20, a protective immune regulator which is induced by NF-κB activators was increased in HR pancreas in comparison to the pancreas of noradrenaline-treated pigs (Figure 4.13).

Figure 4.13 mRNA expression of A20 is maintained in HR animals at 6hr post BD when compared to baseline, whereas A20 expression declines in noradrenaline-treated pigs.

Complement Component-3 gene expression:

The central component of the complement system is the C3 molecule, Cleavage of C3 into C3b and C3a is a pivotal step in the complement activation cascade. C3 functions include opsonization and phagocytosis of pathogens/apoptotic cells, clearance of immune complexes and chemotaxis. Although 90% of C3 synthesis is in the liver, its synthesis has been shown in many other tissues and cell types including pancreas, brain, kidney, monocytes, B lymphocytes and several tumor cell lines. C3 synthesis is elevated in conditions such as pancreatic cancer, acute pancreatitis, pancreas rejection and IBMIR (instant blood mediated inflammatory reaction) following islet transfusion

185 [27, 229]. Here we measured the gene expression level of C3 in the snap frozen pancreas samples from the different groups of brain-dead pigs. The results were analyzed using t test (Figure 4.14).

According to the analysis, the expression of complement component-3 in pancreas is upregulated in the animals receiving noradrenaline by 6hr post brain-death (mean

0.81±0.22 at baseline and 2.26±0.28 at 6 hours, P=0.004), whereas hormone resuscitation-treated pigs manifest a lesser degree of C3 mRNA upregulation

(0.82±0.39 at baseline and 1.65±0.67 at 6 hours, P=0.04).

Figure 4.14 Complement Component-3 gene expression in pancreas from noradrenaline-treated or HR pigs. mRNA is increased more in the group receiving noradrenaline compared to the hormone-resuscitated animals; the difference is not statistically significant.

4.4 Measurement of insulin production in pancreatic islets

Insulin production, as a significant biomarker for pancreas and islet function, was evaluated in brain-dead pigs treated with either hormone resuscitation or noradrenaline.

Both gene expression and islet insulin content were determined.

186 4.4.1 Insulin gene expression Real-time RT-PCR was used to measure insulin gene expression. The values were normalized to Histone H3 expression. Insulin gene expression is tightly regulated.

Nonetheless, insulin mRNA levels were downregulated at 6 hours after the induction of brain death compared to baseline values in BD pigs receiving hormone resuscitation

(which includes exogenous insulin)(mean 2.6±0.6 at baseline and 1.3±0.2 at 6 hours,

P=0.04). Conversely, insulin mRNA levels were increased at 6 hours compared to baseline in BD pigs maintained with noradrenaline (mean 1.3± 0.3 at baseline and 1.7±

0.3 at 6 hours) (Figure 4.15).

Figure 4.15 Evaluation of insulin gene expression in the pancreas of brain dead pigs after 6 hours indicates a significant reduction in insulin gene transcription in pigs treated with hormone resuscitation, whereas insulin mRNA levels increased over the same interval in the noradrenaline-treated group.

187 4.4.2 Insulin protein measurement

Immunohistostaining of paraffin-embedded formalin-fixed sections from different groups of BD pigs was performed following the protocol previously described (Figure

4.16).

Score +1 Score +2 Score +3

Figure 4.16 Insulin immunostaining scoring based on intensity (200x magnification).

After staining, slides were scanned in a Dako ACIS (Automated Cellular Imaging

System), and staining intensity (DAB-bound epitopes) of the tissue sections was analysed and numerated using the ACIS software. A scoring system was developed as described in chapter two (section 2.4.4) to allow us to compare the mean score of each treatment group using One-way ANOVA and Bonferroni post-hoc analyses (Figure

4.17).

188 Scored Insulin staining intensity

3  p<0.05 p = 0.23

2

1

0

6hr NorAd 0hr baseline 6hr Horm one

Figure 4.17 Comparing insulin content between the two treatment groups using immunohistostaining. There is a reduction in the protein content during 6hrs after brain death in the noradrenaline- treated group, compared to physiological levels (p<0.05). In contrast, the hormone-resuscitation management protocol maintains the insulin content close to the physiological level. Error bars represent SD.

Overall, insulin protein content is better preserved in the hormone resuscitation group; conversely, noradrenaline increases the mRNA level, a seeming paradox. A possible mechanism to explain these changes is that noradrenaline induces glycogenolysis in the liver and subsequently raises blood glucose levels [341]. Insulin secretion increases in response to glucose level; which causes enhanced insulin transcription and protein secretion in the noradrenaline receiving group. On the other hand, the exogenous insulin component in the hormone cocktail may have controlled the blood glucose level, rectified the potential brain death-induced hyperglycemia, and caused insulin retention and mRNA downregulation in the group receiving hormone resuscitation.

Prevention of islet overload and exhaustion by maintenance of glucose homeostasis is crucial in islet graft recovery post-transplantation outcomes [270, 341, 342], and hormone resuscitation would seem to offer an advantage over noradrenaline treatment in this regard.

189 4.5 Detection of apoptosis in pancreas tissue from brain-dead pigs

Tissue injury and trauma can trigger two main types of cell death: necrosis and apoptosis. Necrosis is a form of cell death which starts with the loss of cell membrane integrity and damage to the ion pump. This results in cell swelling, lysis, and release of intracellular enzymes, followed by neutrophil migration, inflammation, and oedema.

Another form of cell death, apoptosis, is in fact a programmed cell death which is morphologically distinguished by overall cellular condensation, shrinkage, and plasma membrane blebbing. Activation of a cascade of reactions, such as activation of caspase enzymes leads to digestion of cell structures as well as damage to nuclei characterized by chromatin degradation and DNA fragmentation. Eventually, the apoptotic cell is broken up into smaller membrane-wrapped apoptotic bodies that are usually taken up by phagocytic cells. An important morphologic difference from necrosis is that programmed cell death lacks inflammation [343].

Different methods detect events which occur at different stages during the complex process of apoptosis. These include TUNEL assay, DNA fragmentation detection and staining for the active (cleaved) form of caspase-3.

4.5.1 TUNEL ASSAY (Terminal deoxynucleotidyl transferase dUTP Nick End Labelling)

This method is used to detect apoptosis (programmed cell death) in situ at the single cell level. Cleavage of genomic DNA resulting in single strand breaks (nicks) occurs during apoptosis or severe damage and the TUNEL method is based on labelling the free 3’-OH ends with biotin-conjugated dUTPs using the enzyme terminal

190 deoxynucleotidyl transferase. The procedure was performed as outlined in chapter 2

(section 2.6.1).

4.5.2 TUNEL-Insulin double staining assay

To differentiate the islets from the surrounding acinar tissue, the original TUNEL and anti-insulin staining protocols were modified as described in chapter 2 (section 2.6.1) into a two-step conjoint horseradish peroxidase and alkaline phosphatase method.

Subsequently two serial slide sections from each specimen were subjected to immunostaining, the first with TUNEL and anti-Insulin and the second with TUNEL assay only. Slides were then examined and photographed using an Olympus BX60 microscope and Olympus DP72 mounted camera and the double-stained sections were used to localise the islets on the second slide. A minimum of 6 islets per pancreas with the surrounding exocrine tissue were photographed and subjected to further analysis

(Figures 4.18 and 4.19).

191 TUNEL+insulin TUNEL

100 µm A B

Figure 4.18 TUNEL assay and localisation of islets in a normal pig pancreas. Image A shows an islet stained red for insulin, and TUNEL-positive nuclei in brown. Image B is the same islet in the consecutive section stained only for TUNEL. The brown-positive nuclei are surprisingly abundant (200x magnification).

TUNEL at 0hr TUNEL at 6hr post BD

Islets Islet

Exocrine Exocrine µ pancreas 100 m pancreas

Figure 4.19 TUNEL assay on control pig pancreas at 0hr and 6hr post BD; there are a significant number of positively-stained nuclei in exocrine and endocrine tissue in both specimens (200x magnification).

192 Quantification of apoptosis by TUNEL assay

Following imaging of the islets, several tools were evaluated for the quantitation and interpretation of TUNEL staining results in the pancreas, including the KS400 software package (Kontron Elektronik GmbH, Munich, Germany), Metamorph (Molecular

Devices, CA) and ACIS (Dako, Glostrup, Denmark). Eventually ImageJ (National

Institute of Health) was chosen to analyse the TUNEL assay results. t-test was used to analyse the results, the bars are showing the mean (Figure 4.20).

Figure 4.20 TUNEL assay analysis in brain-dead pig pancreas. 6 hrs of brain-death increases the TUNEL positivity in all animals, regardless of the treatment type.

Induction of brain death in pigs can increase the rate of TUNEL positive nuclei to up to fourfold in the islets, although not statistically significant (data not shown, p=0.2 for

193 baseline to 6hr in NA group; p=0.3 for baseline to 6hr in HR group). However, the extent of this increase is not affected by the type of treatment received by the animals

(NA compared to HR, p=0.8).

More importantly, the overall frequency of TUNEL positive nuclei in all tissue sections from normal and brain-dead pig pancreas ranged between 25 and 50%, an improbably high number. This result raised the question of whether TUNEL assay is a suitable method to detect apoptosis in pancreas. According to the literature, high levels of false positive TUNEL staining can occur in tissues with high rates of protein synthesis and are ascribed to DNA strand breaks happening during RNA transcription. Moreover, the exocrine pancreas contains deoxyribonucleases which could induce strand breaks in tissue samples between collection and analysis. TUNEL assay should be accompanied and verified by other parallel detection methods including Annexin V, Caspase activity, single strand DNA, mitochondrial membrane potential and Cytochrome C release [344,

345].

Among these, detection of caspase activity is a well-established and reliable technique which can detect apoptosis from its early to late stages in situ in formalin-fixed tissues.

4.5.3 Comparison of TUNEL and Cleaved Caspase-3 activity in mouse thymus

To verify the specifity and sensitivity of TUNEL and determine the quality of our

TUNEL technique, TUNEL assay and anti-Cleaved Caspase-3 immunostaining were performed on C57BL/6 mouse thymus tissue as a positive control tissue. Thymus is one

194 of the sites in the body with a very high rate of lymphocyte apoptosis during normal T cell development. Assays were carried out following the protocols described in chapter

2 (section 2.6.2). Anti Cleaved Caspase-3 staining was performed using rabbit anti human cleaved caspase-3 (ASP175) (Cell Signaling), counterstaining was done with

Harris Haematoxylin. This antibody detects endogenous levels of the large fragment

(17/19 kDa) of activated caspase-3 resulting from cleavage adjacent to Asp175 and does not recognize full-length caspase-3 (Figure 4.21).

195 A

TUNEL Anti Cleaved Caspase-3

Normal

C57BL/6

thymus

100 µm 100 µm

B

Normal C57BL/6 mouse Thymus

40

30

20

10 Positive cells / HPF / cells Positive 0

TUNEL C. Casp-3

Figure 4.21 TUNEL assay versus cleaved caspase-3 activity in normal mouse thymus. Figure A shows the immmunostaining result (200x magnification), figure B illustrates the analysis with no difference between assays (+/- SEM).

Comparable numbers of apoptotic cells were detected in normal mouse thymus using either method (29 cell/HPF for TUNEL vs. 28 cells/HPF for CC3).

It can be concluded from this experiment that the very high rate of TUNEL-positive nuclei in pig pancreas is likely to be due to a factor intrinsic to this tissue, and not to more widespread technical problems with the way in which the assay was performed.

196 4.5.4 Comparison of TUNEL and Cleaved Caspase-3 activity in pig and mouse pancreatic tissue

According to the previous results, possible factors accounting for the high rate of

TUNEL positivity in pig pancreas could include degradation of genomic DNA by pancreatic deoxyribonucleases active after sample collection causing false positive identification of cells by the TUNEL method. To obtain information about apoptosis which was independent of the DNA integrity, we stained pig and mouse pancreas tissue for Cleaved caspase-3 in addition to TUNEL.

Baseline pancreatic biopsies were obtained from pigs which were instrumented and anaesthetised for brain-death studies. Normal pancreas from C57BL/6 mice was also harvested. Paraffin-embedded biopsies were subjected to TUNEL and anti Cleaved-

Caspase-3 immunostaining. An even a higher percentage of TUNEL positive nuclei was observed in mouse islets compared with porcine islets (Figure 4.22).

The proportion of apoptotic cells detected by immunostaining for cleaved caspase-3 is approximately fivefold lower, in both pig and mouse than that detected by TUNEL.

197 Normal pig pancreatic islets Normal C57BL/6 mouse pancreatic islet

 50  P<0.05 60  P<0.01 40 40 30

20 20 10 Percentage positive of cells Percentage positive of cells 0 0

TUNEL TUNEL C. Casp-3 C. Casp-3

Figure 4.22 TUNEL versus anti C. Casp-3 staining in pig and mouse pancreatic islets. Significantly more TUNEL positive cells than cleaved-Caspase-3-positive cells are detected in pancreatic islets from either normal pigs or normal mice (Error bars are +/- SEM).

These outcomes confirm our hypothesis that TUNEL assay on pancreas samples is subject to high levels of false positive staining. Thus another reliable method, such as anti-cleaved caspase-3 immunostaining, must be implemented for apoptosis detection.

198 4.5.5 Cleaved Caspase activity in BD pigs treated according to different resuscitation protocols

Islet in BD pig pancreas stained for cleaved caspase-3

A B counted

Figure 4.23 Anti cleaved caspase-3 staining on paraffin-embedded tissue sections from brain-dead pig pancreas. A) An islet containing positively stained cells. B) The positive (stained brown) and negative cells within islets were counted (negative cells marked green and positive marked red here). Both images are shown with 400x magnification.

C. Casp-3 in noradrenalin-treated pig islets C. Casp-3 in hormone-treated pig islets

0.20 0.15

0.15 0.10

0.10

0.05 0.05

0.00 Ratio of positive/negative cells positive/negative of Ratio 0.00 Ratio of positive/negative cells positive/negative of Ratio

0hr 6hr 0hr 6hr

Figure 4.24 Cleaved Caspase-3 staining analysis in brain-dead pig pancreas. Caspase-3 activity does not change during the 6 hours following brain death in the pancreas of animals receiving noradrenaline or hormone resuscitation treatments.

199

According to the analyses, induction of brain death does not increase the level of

Cleaved Caspase-3 activity within 6 hours in the pancreas of the tested animals. No significant difference between cleaved caspase-3 level in pancreatic islets from pigs receiving either treatment was detected (Figures 4.23 and 4.24).

4.5.6 Genomic DNA studies

Irreversible fragmentation of genomic DNA is one of the hallmarks of later stages of apoptosis. This event is initiated by the activation of endogenous endonucleases called

Caspase activated DNase or DNA Fragmentation Factor (DFF40), which subsequently cleave chromatin DNA into a ladder of monomers (180-200 bp) and multimers. This effect can be used to detect apoptosis via the DNA laddering assay. DNA fragments are visualised by electrophoresis on agarose gel followed by ethidium bromide staining.

Induction of apoptosis in mouse splenocytes using etoposide

To achieve fragmented genomic DNA as a positive control for apoptosis detection,

DNA strand breaks and programmed cell death were induced in mouse splenocytes by addition of the cytoctoxic drug Etoposide to the culture medium. Splenocytes were obtained from normal C57BL/6 mouse spleen following the protocol outlined in chapter 2 (section 2.7). Cells were cultured in RPMI (Sigma). The treatment groups received Etoposide at a final concentration of 8.5 μM, and were incubated for 2, 6 and

20 hours before harvesting. Total genomic DNA was extracted from the cells using the

TACS DNA laddering kit, as explained in chapter two (Figure 4.25).

200

Figure 4.25 Genomic DNA from mouse splenocytes measured by Nanodrop.

Genomic DNA extraction and analysis from brain-dead pig pancreas

The extracted genomic DNA from pig tissues was used to study fragmentation by gel electrophoresis (Figure 4.26).

Figure 4.26 Genomic DNA from BD pig pancreas measured by Nanodrop.

Visualisation of DNA laddering

We used horizontal electrophoresis of DNA in agarose gel to visualise apoptosis- induced DNA fragmentation in pancreatic specimens from brain dead pig and mouse, as well as mouse splenocytes as controls (Figure 4.27 and 4.28).

201

A

1 2 3 4 5 6 7 8 9 10 14 15 11 12 13

B

1 DNA Ladder

2 BD pig pan. No.14/0hr 9 BD p. pan. 24/6hr

3 BD pig pan. No.14/6hr 10 30min BD mouse pan.

4 BD pig pan. No.16/0hr 11 3hr BD mouse pan.

5 BD pig pan. No.16/6hr 12 M. Splenocyte/ Etoposide 2hr/I

6 BD pig pan. No.22/0hr 13 M. Splenocyte/ Etoposide 2hr/II

7 BD pig pan. No.22/6hr 14 M. Splenocyte/ Etoposide 6hr

8 BD pig pan. No.24/0hr 15 M. Splenocyte/ Etoposide 20hr

Figure 4.27 A) Genomic DNA electrophoresis of pig and mouse pancreas and mouse splenocytes visualised on agarose gel. B) The table indicates the corresponding specimens. Lanes 2-9 and 11 (from pancreas) look like a smear of completely degraded DNA. This can be correlated to the pancreas high content of ribonucleases and

202 deoxyribonucleases as well as proteases - thus autodigestion of the biopsies cannot be ruled out. A ladder pattern of DNA fragmentation is evident in splenocytes.

Figure 4.28 Example of fragmented genomic DNA due to apoptosis. M is ladder, 1 is control, 2-4 are genomic DNA from an apoptotic mammalian cell line.

Genomic DNA laddering, a hallmark of the late stages of programmed cell death, was observed in the mouse splenocytes treated with etoposide as positive controls. All of the pancreas biopsies from BD pigs and one from the BD mouse exhibit a smear of

DNA that is most likely fully degraded DNA. This observation can be related to the presence of ribonucleases, deoxyribonucleases and proteases within the pancreas and autodigestion of the tissue upon injury.

4.6 Discussion

Detrimental effects of donor brain death on transplantable organs are inevitable; however the inflammatory responses in the organs can be ameliorated to a great extent by optimised management of the donor. Prevention of hypoperfusion, maintenance of cellular metabolism and inhibition of immune activity are the essentials of these optimised management protocols. Pancreas, highly in demand for T1DM treatment, is particularly vulnerable to ischemia and manipulation, and thus the outcomes of

203 pancreas and islet transplantation stand to gain much from the optimisation of donor pancreas quality.

In this study we developed a brain death model in pigs mimicking the human condition, and examined the effects of two different management protocols, traditional pressor support with noradrenaline and hormone replacement therapy, on pancreatic function and inflammation.

Management of the donor with hormone resuscitation was able to maintain hemodynamic status and organ perfusion significantly better than treatment with noradrenaline during the 6 hours following induction of BD. In assessing the status of the transplantable abdominal organs, we measured a number of different parameters.

HR pigs had improved renal function and creatinine clearance was improved by HR.

Serum levels of pancreatic enzymes amylase and lipase, and liver function tests were similar were similar between pigs treated with either protocol. Pancreatic tissue integrity (evaluated by the presence of cellular necrosis) was better preserved with HR.

The pattern of gene expression for a number of inflammatory mediators was evaluated in whole pancreas specimens – these genes included danger signals, pattern recognition receptors, cytokines, chemokines and stress ligands downstream of TLR signalling, as well as the protective molecule A20. Expression of pro-inflammatory markers was, in general, upregulated in pigs receiving noradrenaline, whereas expression of these genes either showed a more limited increase or was downregulated in the HR group.

204 Activation of the complement cascade contributes to inflammatory events following brain death and helps to trigger adaptive immune responses against the graft. The C3 mRNA level in pancreas was measured by RT-PCR, showing reduced gene expression in pigs treated with hormone resuscitation compared with noradrenaline-treated animals. This result is in accordance with the overall reduction in proinflammatory gene expression in the pancreas of HR pigs.

Based on this analysis, we came to the conclusion that administration of the hormone resuscitation cocktail can modulate the general inflammatory status within the pancreatic tissue. The possible explanations for these observations are discussed below.

− The hormone resuscitation treatment cocktail, comprising methylprednisolone,

triiodothyronine, insulin and arginine-vasopressin maintains hemodynamic

instability and perfusion of the transplantable organs better than noradrenaline,

which results in less tissue damage, less release of DAMPs/danger signals and

less inflammation. There are emerging data on the protective effects of

combination of hormones on donor management, organ quality and transplant

outcomes. Protocols using hormone resuscitation have been shown to improve

systemic perfusion, preserve homeostasis and cellular function; and thus

diminish inflammation and immune activation in all the organs studied [279].

− Inflammation in the organ is induced by both endogenous and exogenous

catecholamine sources in this model. Increase in intracranial pressure releases

catecholamines into circulation; in addition to this catecholamine storm,

noradrenaline is infused from one hour post-induction of BD. Adrenaline and

noradrenaline have been revealed to directly upregulate the synthesis and release

205 of TNFα, principally through engagement of alpha(2)-adrenergic receptors.

Translocation of NF-κB, production of downstream proinflammatory molecules

including IL-8, IL-1β, IL-6, and augmentation of inflammatory responses also

occur. Pre-transplant exposure to noradrenaline (for hemodynamic maintenance)

is related to cytokine release after reperfusion in human liver graft recipients

[217, 346, 347].

− Methylprednisolone directly inhibits the transcription of many proinflammatory

molecules. Administration of methylprednisolone to BD donors has recently

been shown to improve hemodynamic stability, perfusion and increase organ

procurement rate [348]. Antiinflammatory effects of glucocorticoids are

mediated through binding to receptors (GR), translocation into the nucleus and

binding to steroid-sensitive genes. This complex inhibits the downstream

proinflammatory activity of NF-κB signalling by trans-repression (suppression

of co-activator molecules), and trans-activation (induction of anti-inflammatory

genes such as NF-κB inhibitor, MAP kinase-1)[349, 350]. The downregulation

of proinflammatory mediators such as MCP1, IL-8, IL-1β, MIC2 could thus be

related to suppression of NF-κB signalling by methylprednisolone in this model.

− Methylprenisolone upregulates the expression of A20, a key inhibitor for

termination of canonical NF-κB activation. This immune regulator is induced by

a number of NF-κB activators, such as TNFα, TLR4/IL-1, IL-17 etc.; and

suppresses NF-κB upstream signalling through deubiquination of several

mediators including RIP1, RIP2, and TRAF6 [177]. Our observations show that

HR downregulates multiple proinflammatory regulators (all involved NF-κB

signalling) but increases A20 at the same time. This selective effect of

methylprednisolone can be related to the recent findings that corticosteroids

206 selectively spare the expression of anti-inflammatory targets of TNF through

context-dependent cooperation between GR and NF-κB [350].

Insulin gene and protein expression levels were evaluated in the pancreas.

Noradrenaline treatment upregulated insulin mRNA transcription, yet the islet insulin protein content decreased from baseline levels, HR maintained insulin content close to normal, despite downregulation of mRNA transcription. The pattern of insulin expression in the two treatment groups could be related to multiple factors. One possibility is that noradrenaline increases blood glucose levels by inducing glycogenolysis in the liver. Insulin secretion increases in response to the glucose level, and subsequently insulin transcription and protein secretion are upregulated in the noradrenaline receiving group [341].

Another potential mechanism may be due to correction of brain death-induced low circulating insulin and hyperglycaemia by exogenous insulin in the hormone cocktail

[270]. Controlled glucose levels prevent islet overload and exhaustion, resulting in preserved insulin content and lower gene expression in the pancreas from HR animals compared to NA-treated group. Islet exhaustion has been implicated in inferior islet graft outcome and survival, while improved clinical transplant success rates are reported following peritransplant exogenous insulin administration and prevention of islet exhaustion [342]. These findings can be correlated to the outcomes from clinical islet transplant, where dead cells within the islets release excessive insulin/C-peptide in the first 24 hours after stressful events of transfusion [89].

207 We used TUNEL and the activity of cleaved caspase-3 to evaluate the level of apoptosis in porcine pancreas after BD. The TUNEL technique resulted in overestimation of apoptotic cells when the outcomes were compared to c-casp3, indicating that TUNEL might not be an ideal technique for apoptosis detection in pancreas. TUNEL may yield false positive results due to the high prevalence of single-strand breaks in DNA in metabolically active tissues [344], and/or degradation of genomic DNA by tissue deoxyribonuclease activity. The extent of apoptosis was evaluated by a third method, detection of the typical cleavage pattern of genomic DNA by endonucleases happening during programmed cell death upon agarose gel electrophoresis. Our attempts however yielded only smears of degraded DNA from pancreas samples, where the typical ladder pattern was observed in mouse splenocytes treated with etoposide. The exocrine pancreas contains high amounts of nucleases and proteases and degradation of genomic

DNA by these enzymes during the extraction process is likely, making this technique unreliable for the detection of apoptosis in pancreas.

The hormone resuscitation protocol demonstrated several advantages over traditional pressor support with noradrenaline; improved hemodynamic maintenance of the donor, restriction of immune activation and inflammation in pancreatic tissue, and reduction in islet stress. Donor modulation can be expanded to interventions such as transfer of protective genes. For this purpose we decided to evaluate a strategy to attenuate TLR and RAGE signalling in the BD donor using gene-delivery of a naturally-occurring anti- inflammatory molecule - esRAGE. Evaluation of the effects of systematic esRAGE overexpression on the pancreas in a mouse model of brain death is discussed in chapter

5.

208

CHAPTER 5 SYSTEMATIC OVEREXPRESSION OF esRAGE AND ITS EFFECTS ON THE PANCREAS AND ISLETS IN BRAIN-DEAD MICE

209 5.1 Introduction

Perimortem events in the donor can substantially alter the graft microenvironment, and can influence the success or failure of subsequent therapeutic interventions. To optimise transplant outcomes, immunological interventions must address the effects of these donor and procurement factors in addition to the recipient immune response.

One of the key initiators of innate immunity following brain death is signalling through pattern recognition receptors (PRRs) including toll-like receptors (TLRs) and the receptor for advanced glycation end-products (RAGE), resulting in a systemic inflammatory response in the donor, and increased expression of inflammatory mediators within transplantable organs. TLR engagement is linked with immune activation, cellular infiltration, ischemia-reperfusion injury and poor transplant outcomes [204, 351]. Stimulation of RAGE by its various ligands such as alarmins

HMGB1 and S100B also contributes to release of inflammatory mediators, promoting

IRI and inferior engraftment [134, 313]. Blockade of TLR4 and/or RAGE has been shown to protect against IRI and improve transplant results [143, 205, 313].

Endogenous secretory (es) RAGE is a naturally-occurring soluble isoform of RAGE. It functions as a decoy receptor that binds to HMGB1, S100B as well as other RAGE ligands, preventing them engaging PRRs including RAGE, TLR2, TLR4 and TLR9.

Moreover, esRAGE can form heterodimers at the cell surface with full-length transmembrane RAGE, thus blocking the homodimerisation of the RAGE receptor which is necessary for signal transduction [149]. Administration of esRAGE in a mouse model of hepatic IRI blocked RAGE and TLR4 from binding to their ligands, inhibited

NF-κB signalling and protected against lethal hepatic ischemia-reperfusion injury

210 [312]. Knowing that human esRAGE is active in mice [311], we developed a liver- specific rAAV 2/8 vector to allow us to express high levels of esRAGE systemically in

C57BL/6 mice. Using this technology, esRAGE is produced by the liver and secreted into body fluids – concentrations of esRAGE many thousand-fold above physiological levels can be detected in the serum and within organs such as kidney, heart or spleen.

Mice transduced with this vector were used to investigate the immunoprotective effects of esRAGE over-expression on early events following brain death induction in pancreas. The ultimate aim of this work is to further evaluate the effects of esRAGE expression in donor and/or recipient on transplantation and tolerance-induction outcomes.

5.2 Expression of esRAGE in vivo after inoculation with rAAV esRAGE

5.2.1 In-vivo expression of esRAGE

Following the packaging of rAAV-esRAGE, we performed a dose-response experiment to determine the most appropriate vector dose. Male C57BL/6 mice 8-10 weeks-old received doses of rAAV-esRAGE of 1x1010, 5x1010, 1x1011, and 5x1011 vector genomes via intraperitoneal injection (IP). The concentration of esRAGE was then measured in the serum of the injected mice at 10 days post-injection using a sandwich

ELISA (B-Bridge). Systemic expression of esRAGE was dose-dependent, and the dose of 5x1011 VG resulted in the highest esRAGE serum level (Figure 5.1). This vector dose did not result in liver inflammation as measured by serum ALT concentration or the presence of an inflammatory infiltrate upon H and E staining of the liver (Figure 5.2

A and B), and was therefore chosen for use in later experiments.

211 We next conducted a time-course experiment using 5x1011 VG intraperitoneally (IP) to determine the kinetics of rAAV2/8-mediated esRAGE expression. Murine esRAGE is undetectable in serum, and physiological concentrations of esRAGE in humans are in the range 0.1-1 ng/mL. By d2 following IP injection, levels around 5000-fold greater than physiological concentrations in human’s serum had already been achieved in injected mice. esRAGE levels continued to increase until d42 post-injection, then started to decline. Nonetheless, durable high-level expression was demonstrable throughout 100 days after injection. For practical reasons, it was determined that sufficient levels of esRAGE for our experimental purposes were reached by 7 days after IP injection. Brain-death induction was therefore performed 7 days post- inoculation.

Figure 5.1 Dose-response curve for rAAV-esRAGE. Mice were injected IP with esRAGE at doses ranging from 1x1010 to 5x1011 VG. Levels of esRAGE measured in serum by ELISA were dose-dependent and many thousand-fold greater than physiological concentration in humans (0.1-1 ng/mL).

212 A B

Figure 5.2 Administration of rAAV-esRAGE does not result in liver inflammation. Neither serum ALT levels (A) nor the appearances of liver on H and E staining (B) suggested liver inflammation with the highest dose of 5 x 1011 VG rAAV-esRAGE (200x magnification). The increase in ALT level at day 10 was observed in only one mouse; however considering the fact that this rise was still well below pathological levels correlated with inflammation, along with absence of any accompanied histological changes, this observation was not concerned abnormal.

Finally, to verify our assumption that esRAGE secreted by the liver would be distributed throughout the body, and would appear in high concentration within transplantable organs, we collected liver, spleen, kidney and heart from two C57BL/6 mice at 5 days following IP injection with 5x1011 VG rAAV-esRAGE, and measured the concentration of esRAGE per gram of total protein in tissue lysates (Figure 5.3). Not surprisingly, the highest levels were detected in liver, where esRAGE is produced, followed by kidney where esRAGE is reabsorbed from the urine. Spleen and heart contained lower, but still substantial levels of esRAGE (Figure 5.4). No human esRAGE was detected in organs from control C57BL/6 mice. Regarding the pancreas, according to a study by our group using an analogous vector, this viral construct does not infect the pancreas [218]. Although the concentration of esRAGE was not directly measured in these mice pancreatic tissue, as it was mentioned above the secreted esRAGE circulating in serum is supposedly delivered to this organ as well.

213

Figure 5.3 Kinetics of expression for rAAV-esRAGE injected intraperitoneally (IP) at 5x1011 VG. Serum esRAGE expression was already substantial by d2, and reached a peak at day 42 followed by a slow decline.

Figure 5.4 Levels of esRAGE in various organs. Concentrations of esRAGE in relation to total protein in tissue lysates were greatest for liver and kidney, and less, though still substantial for spleen and heart. No human esRAGE was detected in organs from control C57BL/6 mice.

214 5.3 Physiological changes and systematic inflammatory response in brain dead mice

Animals were monitored for physiological parameters before, during and after the brain death procedure. Typical recordings of blood pressure, heart rate, and oxygen saturation were presented in chapter 2 where the model was described. The serum protein levels of danger signals, chemokines and cytokines were measured by ELISA and Luminex techniques.

In response to brain death and consequent tissue ischemia, both danger-associated molecular pattern molecules HMGB1 and S100B (Figure 5.5), and an array of chemokines and cytokines are released into the circulation. We observed a dramatic increase in HMGB1 and S100B within 3 hours of brain death induction. Serum levels of MCP1, KC, MIP2, IP10, TNFα, IFN-γ, IL-10 and IL-1β reached very high levels during first hour after BD (Figure 5.6). These values are indicative of the systemic proinflammatory state resulting from brain death.

30 15000

20 10000

10 5000 serum S100B (pg/ml) Serum HMGB1Serum (ng/ml) 0 0

3 hrs 3 hrs

baseline 30 mins 60 m ins baseline 30 mins 60 mins

vent ctrl 3 hrs vent ctrl 3 hrs

Figure 5.5 Serum protein levels of danger-associated molecular patterns HMGB1 and S100. These danger –associated molecules are increased dramatically following brain death, with a peak occurring at 3 hours after brain death induction.

215

Figure 5.6 Serum protein levels of various cytokines and chemokines were measured. Within 60 min of brain death very high levels of these proinflammatory mediators were attained.

216 5.4 Investigation of inflammation in the pancreas post brain death

Pancreatic specimens were collected from normal animals, mice anaesthetised and ventilated for 3 hours, brain-dead animals at 1 and 3 hrs following induction of brain death, mice previously transduced with rAAV 2/8 vectors expressing either esRAGE or the control protein human serum albumin (HSA) at 3 hours following induction of brain death, and esRAGE or HSA-expressing animals 1 week after inoculation with the vector (but not ventilated or subjected to brain-death). There were 4 animals included in each study group.

5.4.1 RNA extraction and cDNA synthesis from brain-dead mouse pancreas

Total RNA was extracted from the snap-frozen pancreas harvested from brain-dead and control animals following the instructions described before. One μg of RNA was used to synthesize complementary DNA as per protocol in chapter 2. This cDNA was used as template in RT-PCR studies.

5.4.2 Investigation of gene expression by Real Time RT-PCR

Real time RT-PCR was set up for various genes as described in chapter 2, using

Taqman primers and probes. Values were normalized to Histone H3 values. Results were analysed with One-way ANOVA and t-test.

217 Danger Signals:

Gene Expression of three danger molecules HMGB1, Hyaluronan synthase 1 (HAS1) and HSPa2 was evaluated by RT-PCR. These are known to initiate pattern recognition receptor (PRR) signalling. Although HMGB1 protein was released rapidly into the circulation following the induction of brain-death, no increase in mRNA transcription in the pancreas was detected within three hours, except in the mice which were maintained ventilated without the induction of brain death. esRAGE or HSA expression alone in animals without any surgical manipulations did not alter HMGB1 mRNA levels, and the mRNA levels were comparable between brain dead mice expressing either esRAGE or HSA (Figure 5.7 ).

Hyaluronan synthase 1 (HAS1) is the enzyme responsible for synthesis of the low molecular weight form of hyaluronan, a ligand for TLR4. HAS1 expression is elevated in mouse pancreas at 3 hours after BD induction. esRAGE expression did not alter the gene expression of hyaluronan synthase 1 compared to the baseline values in unmanipulated animals, nor did levels of HAS1 differ between BD mice transduced with either esRAGE or HSA (Figure 5.8).

Heat shock-related 70 kDa protein 2 (HSPa2), the murine counterpart of HSP70 in humans, is a molecular chaperone responsible for re-folding damaged proteins. At the intracellular level, HSPa2 has a protective effect. When released into the extracellular milieu, HSPa2 can function as a ligand for TLR4. Pancreatic HSPa2 gene expression was increased after three hours of anaesthesia and mechanical ventilation, regardless of whether brain death had been induced or not. esRAGE expression did not influence levels of HSPa2 in either BD or unmanipulated animals (Figure 5.9).

218

Figure 5.7 A) Gene expression of HMGB1 in pancreas remains close to baseline after 1 and 3 hours post BD, whilst a modest increase is observed in the ventilated animals. B) esRAGE or HSA-transduction alone does not influence HMGB1 gene expression in mouse pancreas at d7 post-inoculation. C) HMGB1 gene expression is comparable in mice transduced with either esRAGE or HSA at 3 hours post-induction of BD.

219

Figure 5.8 A) Hyaluronan Synthase 1 expression was increased in mouse pancreas at 3 hours after BD induction. B) Transduction of unmanipulated mice with either esRAGE or HSA does not alter levels of hyaluronan synthase 1 at d7 post-inoculation. C) Similarly esRAGE does not influence HAS1 gene expression after brain death.

220

Figure 5.9 A) After 3 hours of anaesthesia and instrumentation, levels of the chaperone and DAMP HSPa2 were increased in the pancreas of both ventilated and brain-dead animals. Gene expression for HSPa2 was not affected by esRAGE transduction, either at baseline (B) or following brain death (C).

221 Receptors:

Expression of the pattern recognition receptors toll-like receptor 4 and receptor for advanced glycation endproducts (RAGE) was examined in the pancreas of brain-dead mice and ventilated controls. Similar to findings for HSPa2, TLR4 expression increased in both BD and ventilated mice after three hours of anaesthesia and instrumentation.

Transduction with esRAGE or HSA did not influence the level of TLR4 expression, either in steady state at d7 post-inoculation or at 3 hours after BD induction. Although it has been reported that the presence of soluble RAGE and resultant sequestration of

RAGE ligands causes a downregulation in transcription of the full-length transmembrane isoform of RAGE, we did not observe any changes in RAGE mRNA expression during any of the conditions during the period of study (Figures 5.10 and

5.11).

222

Figure 5.10 A) Some elevation in the pancreatic gene expression of TLR4 is observed after three hours of maintenance in both the brain-dead and ventilated control mice. B) Treatment with rAAV-HSA or rAAV-esRAGE does not change the mRNA levels compared to the normal. C) Although the levels increase modestly in the pancreas of BD mice, there are no differences between the treated and control groups.

223

Figure 5.11 A-C) Gene expression of RAGE in mouse pancreas did not change in brain-dead or ventilated control mice during the period of observation, nor were changes detected at d7 after transduction in unmanipulated mice.

224 Chemokines:

Through its ability to block signalling block through TLRs and RAGE, esRAGE can suppress the downstream production of chemokines such MCP1, MIP2, KC and IP10.

We studied the gene expression of these four chemokines in pancreas by RT-PCR.

Normal mouse pancreas expresses very low levels of these chemokines, and no changes in expression resulted when mice were transduced with esRAGE or HSA vectors but not otherwise exposed to any surgical procedures. Expression of all four chemokine genes was substantially upregulated in mice subjected to brain death, or to ventilation alone (MCP1 normal 0.3±0.08 v/s peak 9.4 ±3.1, P=0.02; MIP2 normal 0.17±0.03 v/s peak 7.6 ±2.8, P=0.02; IP10 normal 0.3±0.05 v/s peak 78 ±77, P=0.2; KC normal

0.05±0.01 v/s peak 8.6±7.2, P=0.02). At 3 hours after the induction of brain-death, modest suppression of the pancreatic expression of all four chemokines is evident in the esRAGE-expressing group compared to HSA and other controls; however, this difference did not reach statistical significance. IP10 is significantly higher in the HSA- treated group compared to normal values after 3 hours of BD (Figures 5.12-15).

225

Figure 5.12 A) MCP1 gene expression in mouse pancreas reaches a peak at 3 hours after the induction of BD. B) MCP1 gene expression changed little in mice transduced with vectors encoding esRAGE or HSA, but at three hours after the induction of BD, MCP1 mRNA levels were reduced in the esRAGE treated mice, compared with the HSA controls, though this difference did not reach statistical significance (C).

226

Figure 5.13 A) Prolonged mechanical ventilation or brain death upregulates gene expression of MIP2 in this model. B) Transcription of this chemokine is not changed following esRAGE overexpression. C) Expression of MIP2 is modestly decreased in the pancreas of BD mice transduced with esRAGE, compared with that of HSA controls.

227

Figure 5.14 A) Brain death induction or anaesthesia/ventilation each increase IP10 gene expression in pancreas. B) IP10 gene expression is not altered when animals are transduced with HSA or esRAGE vectors. C) Following BD, IP10 expression in the HSA control group is significantly greater than that in the normal controls, whereas esRAGE-expressing animals exhibit a lower increase in IP10 mRNA, closer to that of ventilated-only controls.

228

Figure 5.15 A) KC gene expression in mouse pancreas is substantially upregulated by either BD induction or instrumentation and sustained mechanical ventilation. B) HSA or esRAGE overexpression per se does not affect the expression of KC. C) Following brain death, pancreatic KC expression level is reduced in the esRAGE group in comparison to the HSA-treated mice.

229 Cytokines:

Gene expression of 5 cytokines (IL-1β, TNFα, IL-6, IL-10 and IL-33) was evaluated by

RT-PCR. Transcription of these cytokines is known to be activated by TLR and/or

RAGE signalling. These cytokines were expressed at very low levels in the pancreas of normal mice. mRNA transcription was increased by brain death and generally also increased by instrumentation and prolonged mechanical ventilation. (IL-1β normal

0.21±0.03 v/s peak 8.1±3.8, P=0.02; TNFα normal 19±8.0 v/s peak 164±83, P=0.2; IL-

6 normal 0.01±0.007 v/s peak 93±78, P=0.02, IL-10 normal 1.2±0.4 v/s peak 5±2.7,

P=0.05, IL-33 normal 0.46±0.27 v/s peak 1.5±0.7, P=0.2). For IL-1β, IL-6 and IL-10, gene expression in mouse pancreas at 3 hours after induction of brain death was considerably greater than that in ventilated controls, whereas for TNFα and IL-33, either brain death or ventilation alone resulted in upregulation. Transduction alone, in the absence of other manipulations, did not cause any particular change in the expression of these cytokines. There was a modest trend towards reduction of TNFα,

IL-6 and IL-10 mRNA in BD mice treated with esRAGE compared with HSA controls.

IL-1β mRNA expression was unaffected by esRAGE overexpression, whilst gene expression of IL-33 was, if anything, increased in mice receiving esRAGE (Figures

5.16-5.20).

230

Figure 5.16 A) There is a noticeable increase in the expression of cytokine IL-1β after 3 hrs of BD compared to the baseline values. B) When mice are transduced with esRAGE or HSA vectors, pancreatic expression of IL-1β does not change. C) The pancreas shows similar levels of IL-1β mRNA in both esRAGE and HSA-treated animals at 3 hours after the induction of brain death.

231

Figure 5.17 A) Both ventilation/anaesthesia and BD induction can induce gene expression of TNFα in pancreas. B) Expression of this cytokine is similar in the pancreas of otherwise unmanipulated mice transduced with esRAGE or control vector. C) At 3 hours following the induction of brain death, there is a modest reduction in TNFα expression in the esRAGE-treated animals compared to the HSA controls.

232

Figure 5.18 A) In contrast to the other cytokines studied, transcription of IL-10 in pancreas was not increased in ventilated animals, and increases only after 3 hours of brain death. B) no change in IL-10 expression is observed in the pancreas of esRAGE or HSA-transduced control mice. C) There is a reduction in the upregulation of IL-10 in esRAGE-treated mice compared to that in HSA group at 3 hours post brain death induction.

233

Figure 5.19 A) Gene expression of IL-33 increases in response to instrumentation and prolonged mechanical ventilation, with or without the induction of brain death. B) Overexpression of esRAGE or HSA has no effect on IL-33 expression in the pancreas at baseline. C) IL-33 mRNA level is increased in esRAGE-transduced mice compred with that in the HSA-expressing group at three hours following induction of BD.

234

Figure 5.20 A) Both ventilation and brain death induction can increase IL-6 expression in pancreas; the most dramatic change was detected at three hours after brain death induction, where IL-6 mRNA had increased almost 10,000-fold from baseline. B) Transduction of the animals with esRAGE or HSA per se does not affect IL-6 expression in mouse pancreas. C) A minor decrease in IL-6 expression in pancreas was noted in the esRAGE-treated mice compared with HSA controls following brain death.

235 Stress ligands:

Expression of the NKG2D ligand RAE-1 (retinoic acid early inducible gene 1) expression is strictly regulated in normal cells; therefore, little is expressed on healthy adult tissues, including normal pancreas [352]. It has been reported to be induced by retinoic acid, transcription factor AP-1 and MyD-88-dependant-TLR4-signalling [353].

Ischemia/reperfusion injury induces expression of RAE-1 and MULT-1 (another mouse

NKG2D ligand) in renal tubular cells, and this effect was abolished in the absence of

TLR4 signalling [208] In this model, the mRNA level for RAE-1 in mouse pancreas increases about threefold within one hour post BD, but this increase is not sustained. A comparable increase was observed in the pancreas of mice transduced with HAS, but not otherwise subjected to any experimental manipulations. Unexpectedly, we observed a significant increase in RAE-1 transduction in the pancreas of brain-dead mice treated with esRAGE, when compared with control animals transduced with the HSA vector.

Possible reasons for this are discussed later in the chapter. Increased stress ligand expression can render the tissue prone to immune attack by NK cells and activated

CD8+ cells (Figure 5.21).

236

Figure 5.21 A) There is a threefold elevation in RAE-1 gene expression in mouse pancreas at 1 hour post induction of brain death, which is not maintained thereafter. B) Of interest, baseline pancreatic RAE-1 mRNA level was elevated to a comparable degree to that of the BD mice (1 hr group) in HSA-expressing mice (d7 post- inoculation) but not in mice transduced with esRAGE. C) Unexpectedly, RAE-1 gene expression was significantly increased in the esRAGE-transduced animals after brain death compared to expression in all of the controls.

237 Protective molecule A20:

This protective molecule is induced as part of the endogenous regulation of inflammatory responses mediated by TLR or TNFα signalling. Feedback mechanisms controlling the expression of A20 are complex; A20 transcription is promoted by NF-

κB, yet A20 prevents NF-κB nuclear translocation, thus becoming subject to self- regulation. Similarly, upstream blockade of TNFα signalling has been shown to reduce the gene expression of this anti-inflammatory product [350]. In parallel with the changes in TNFα expression, A20 gene expression in this model increased in the pancreas of mice subjected to instrumentation and mechanical ventilation, whether or not BD was also induced. Transduction with esRAGE or HSA per se do not result in changes to the baseline expression of A20 in the pancreas. However, levels of A20 mRNA in the pancreas of esRAGE-transduced animals following brain death were reduced compared to those in HSA-transduced mice (Figure 5.22). This trend was also in accordance with the pattern of TNFα expression in this study, and may also reflect a reduction in NF-κB signalling as a result inhibition of RAGE and/or TLR downstream pathways. Whether the net effect of these changes in A20 and other gene expression is harmful or protective is not possible to determine without further information.

238

Figure 5.22 A) Ventilation and brain death induction can both increase the level of A20 gene in mouse pancreas as early as 1 hour after commencement. B) Expression of A20 is not altered by esRAGE-overexpression in the pancreatic tissue before manipulation. C) Animals expressing esRAGE exhibit a lower level of A20 after BD when compared with mice transduced with HSA; this is concordant with TNFα gene expression.

239 Complement Component 3 measurement:

Total RNA extraction, cDNA synthesis and real time RT-PCR for C3 were performed as described in chapter 2. The results were normalized to histone H3 values and analyzed using One-way ANOVA and t test.

Gene expression of C3 in pancreas rises slightly due to anaesthesia and mechanical ventilation. However, contrary to our expectations, C3 mRNA levels were, if anything, suppressed following brain death induction in this model. Further slight downregulation of complement component-3 mRNA transcription in pancreas was observed in brain- dead animals expressing esRAGE compared with the HSA controls, although this difference did not reach statistical significance (Figure 5.23).

240

Figure 5.23 A) C3 mRNA levels in mouse pancreas fall following brain death induction, contrary to our expectations. B) Baseline expression of complement component-3 in pancreas is not affected by esRAGE transduction. C) Further slight downregulation of pancreatic C3 mRNA level was noted in brain-dead animals expressing esRAGE compared with HSA controls, although this difference did not reach statistical significance.

241 5.5 Insulin protein and gene expression in pancreas post brain death

Pancreatic insulin content is frequently employed as a marker of pancreatic integrity and endocrine function. We studied both insulin gene expression and the insulin protein content of pancreatic islets to determine how these indices were affected by brain death and by systemic esRAGE over-expression.

5.5.1 Insulin gene expression

Insulin gene expression is normally tightly regulated in order to maintain glucose homeostasis. After three hours of anaesthesia and mechanical ventilation, increased insulin gene transcription was evident in mice whether or not they had also been subjected to the induction of brain death. Anaesthesia and surgical procedures are commonly accompanied by a decline in IGF-1 production with concomitant hyperglycaemia [354]. The liver plays a role in glucose homeostasis, and disturbances of hepatic glucose metabolism occur in viral hepatitis, possibly secondary to endoplasmic reticulum stress [355]. At d7 post-inoculation, transduction of mouse liver with either esRAGE or HSA vectors was associated with a small increase in Insulin mRNA, suggesting that overexpression of the transgene in liver may cause a modest metabolic perturbation. Conversely, at 3 hours after the induction of BD, the animals expressing esRAGE manifested a reduced level of insulin gene expression compared to that of the controls (Figure 5.24).

242

Figure 5.24 A) Gene expression of insulin increases in mice exposed to prolonged anaesthesia and mechanical ventilation whether or not they are subjected to brain death. B) A slight increase in insulin gene transcription was noted in mice transduced with either esRAGE or HSA vectors suggesting the possibility of mild perturbation of hepatic glucose homeostasis when large amounts of transgene product are synthesised. C) Conversely, at three hours after the induction of brain death, insulin mRNA levels were lower in esRAGE-transduced mice in comparison to those in the ventilated or HSA-transduced brain dead control groups. There were 4 animals in each group of the study, and results were analysed using One-way ANOVA.

5.5.2 Anti-Insulin immunohistostaining

5μm sections from the paraffin-embedded pancreatic tissue were subjected to anti-

Insulin immunohistostaining as described in chapter 2 (Figure 5.25). The slides were then examined by light microscopy, and all the islets throughout the sections were photographed to be analysed. The images were subsequently assessed using Image Pro-

Premier 9.1 image analysis software (Media Cybernetics, Rockville, MD). The

243 numerical value of staining intensity from each islet was generated by dividing the

Integrated Optical Density (OD) by an Area factor. The mean intensity value for the islets in each mouse pancreas was calculated and used in further analysis (Figure 5.26).

Insulin mRNA levels are increased in response to ventilation and brain-death manipulations, and in general, the insulin protein content increased in parallel with the changes in gene transcription. However, brain dead mice transduced with either HSA or esRAGE exhibited the highest islet insulin content of all groups (Figure 5.27). This may represent a synergistic effect of high-level transgene expression in the liver with surgical stress.

Score +1 Score +2 Score +3

(Normal) (BD 3hr) (BD esRAGE)

50 µm 50 µm 50 µm

Figure 5.25 Anti-insulin immunostaining on mouse pancreas. The scoring system is based on the intensity of anti-insulin immunostaining which correlates with protein content (400x magnification).

244

Figure 5.26 Example of the insulin staining intensity values from one islet generated by the Image Pro-Premier 9.1 image analysis software. The OD/Area is inversely correlated with the immunostaining intensity.

Figure 5.27 Analysis of anti-insulin immunostaining in mouse pancreas. Both prolonged anaesthesia/ventilation and brain-death induction can increase the insulin protein content; the increase is greatest in the brain-dead HSA and RAGE-transduced animals, suggesting possible synergism between the effects of surgery/anaesthesia and those of high-level hepatic transgene expression.

5.6 Detection of apoptosis in brain-dead mouse pancreatic tissue

Interruption of RAGE signalling through the administration of soluble RAGE has been reported to prevent apoptosis and promote regeneration of liver and kidney

245 parenchymal cells following ischemia/reperfusion injury. We examined apoptosis within the pancreas of mice subjected to brain death and mechanical ventilation with and without systemic overexpression of esRAGE or the control transgene HSA.

5.6.1 TUNEL assay

TUNEL assay (Terminal deoxynucleotidyl transferase dUTP Nick End Labelling) is used to detect apoptosis (programmed cell death) by tagging the cleavage of genomic

DNA to single strand breaks (nicks) with labelled dUTPs. These breaks in DNA occur during apoptosis or severe damage. The procedure was performed as described in chapter 2 on all 7 groups of study, each including specimens from 4 mice. However, as outlined in chapter 4, we concluded that TUNEL assay on either murine or porcine pancreas might be prone to high rates of false positivity (Figure 5.28, showing 3 normal pancreas sections) and we employed the alternative method of specific immunostaining for cleaved caspase-3 activity to detect apoptosis in situ.

246

Mouse pancreatic islet Mouse pancreatic islet stained with TUNEL stained for Cleaved- Caspase 3

A B

100 µm 100 µm

Figure 5.28 Normal C57BL/6 paraffin-embedded pancreatic tissue sections immunostained for Cleaved-Caspase3 (column A) and TUNEL (column B) assays to detect apoptosis. A significant number of cells are positive for TUNEL in the normal tissue (200x magnification).

247 5.6.2 Cleaved Caspase-3 activity

A reliable technique for investigating apoptosis is anti cleaved caspase-3 immunostaining. This assay was performed on paraffin-embedded pancreas sections from brain-dead mice from different experimental groups, using rabbit anti human cleaved caspase-3 (ASP175) (Cell Signalling) and following the protocol described in chapter 2. After the completion of staining, the slides were photographed using an

Olympus microscope and camera. Pictures were taken with x10 magnification, and the number of positive cells per 16mm2 of the tissue were counted with ImageJ software.

Results were analysed using One-way ANOVA and t test. Data are shown as scatter plots with the mean value indicated (Figure 5.29).

248

Figure 5.29 Evaluation of Cleaved Caspase 3 activity in murine pancreas by immunostaining. A) Overexpression of esRAGE in otherwise unmanipulated mice does not alter the C-Casp3 activity, when compared to the activity level in the normal pancreas. B) ventilation or brain death induction resulted in a modest increase of the C- Casp3 activity. C) There is a trend towards reduced C-Casp3 in esRAGE-treated mice after brain death, compared with that in HSA-transduced controls.

Cleaved Caspase3 activity can be used as a consistent method to evaluate apoptosis in mouse pancreas as well it was for pig pancreas.

The analysis shows that C-Casp3 activity rises due to mechanical ventilation and brain death induction, reaching the peak at 3 hours. The BD-esRAGE-treated pancreas expresses lesser C-Casp3 level compared to others, although the difference doesn’t reach significance level.

249 5.6.3 Neutrophil infiltration assessment

Neutrophils are amongst the vanguard of innate immune cells, migrating to sites of inflammation within 6 hours after the initiation of tissue damage. Lymphocyte Antigen

6 Complex family (Ly6) including Ly-6B.2 are expressed prominently by murine neutrophils and are frequently used as surface markers for identification of this leukocyte subset. We detected the presence of neutrophils in mouse pancreas by immunostaining for Ly-6B.2 antigen (using a rat anti-mouse primary antibody (AbD

Serotec, Raleigh, NC). Immunostaining was performed as described in chapter 2

(Figure 5.30). The slides were then examined under an Olympus DP72 microscope. 8 random fields at 20x magnification were captured for each specimen, containing vessels, islets and areas of exocrine pancreas. The number of positively-stained cells was counted in the tissue interstitium as well as in the lumen of the microvasculature using ImageJ. Values were analysed with One-way ANOVA and t test, data sets are shown as scatterplots with the mean indicated (Figure 5.31).

Normal BD 3hr BD+esRAGE 3hr 100x 200x 200x

200 µm 100 µm 100 µm

Figure 5.30 Anti-Ly-6B.2 immunostaining for neutrophils on mouse pancreas, conterstained with Harris Haematoxylin. The presence of neutrophils was evaluated by staining to determine the effect of esRAGE expression on neutrophil infiltration following brain death. A significant increase of marginating neutrophils within the pancreas is noticeable at 3 hours after induction of brain death, compared with the normal mice. This increase was inhibited in the esRAGE-expressing BD animals.

250 A B

Neutrophil migration to mouse pancreas Neutrophil migration to mouse pancreas

  p<0.05 250 250 p =0.48 2 2

200 200

150 150

100 100

50 50 Neutrophil per 6.5mm per Neutrophil Neutrophil per 6.5mm per Neutrophil 0 0

Norm al BD 1hr B D 3 h r Norm al Vent 3hr Vent 3hr BD HSA

BD esRAGE

Norm al Vent 3hr BD 1hr BD 3hr BD HSA BD esRAGE

Figure 5.31 Analysis of neutrophil presence evaluated by immunostaining. A) Brain death can isignificantly increase the presence of marginating neutrophils by 3 hrs after induction. B) A modest decline in the number of marginating neutrophils was noted in the BD animals expressing esRAGE when compared with the HSA-expressing controls, although this difference did not reach statistical significance.

5.7 Discussion

For these studies, we were able to develop a mouse model of brain death which recapitulates many of the features of the human condition, including haemodynamic instability and systemic inflammation. Using an AAV2/8 gene delivery system, we were able to achieve high-level expression of esRAGE in circulation and throughout the transplantable organs. Because mice can only be maintained reliably for three hours after brain-death induction, our examination of changes in gene expression or cellular infiltration was necessarily confined to this early period. Similar to published reports, it is likely that damage to the pancreas and islets which is initiated during brain death will

251 be amplified by islet isolation and/or transplantation. In subsequent studies, this model will be extended to incorporate those later processes.

Only the earliest stages of neutrophil infiltration are apparent within three hours of BD induction. The effect of esRAGE in reducing the number of marginating neutrophils detected within the pancreas at this time may be mediated through a reduction in expression of neutrophil chemokines such as KC (the murine equivalent of IL-8) and

MCP1, but also through a direct effect on adhesion, as RAGE is a known binding partner for the leucocyte integrin Mac-1 [143] and esRAGE could disrupt this interaction, preventing neutrophils from binding to the endothelium.

Release of DAMPs such as HMGB1 follows a biphasic pattern. Whilst pre-formed

HMGB1 is released into the circulation within three hours of the induction of brain death, upregulation of HMGB1 gene synthesis has typically been reported to occur over several days after an acute ischemic insult, and TLR4 upregulation is seen over a similar timecourse [205]. In contrast, upregulation of cytokine and chemokine gene expression in pancreas was noted by three hours post-BD. esRAGE overexpression was generally associated with modest reductions in cytokine and chemokine expression, though these differences usually did not reach statistical significance. esRAGE can prevent RAGE signalling almost completely via sequestration of RAGE ligands or prevention of RAGE homodimerisation at the cell surface. However, the extent to which esRAGE inhibits signalling through other pattern-recognition receptors depends upon the extent to which these receptors are engaged by ligands which are shared with

RAGE. Whilst HMGB1 is paramount amongst these, other ligands for TLRs will be unaffected by esRAGE overexpression, as will ligands for other PRRs such as RIG and

252 NOD-like receptors. Regulation of expression by alternative PRRs may explain why

IL-33 expression in the pancreas appears to be increased in mice transduced with esRAGE.

Complement and TLR signalling are activated by common ligands and there is synergistic yet complex crosstalk between the two systems in shaping the innate immune response. For instance, C3 expression and deposition and serum C5a are increased in wide type mouse intestine after IRI, but not in TLR4 knockout animals, whilst treatment with a local inhibitor of complement can reduce TLR4 expression following IRI, indicating the close interaction of these systems [356].

Following brain death, increased C3a serum levels, local and systematic complement activity and deposition have been reported. Moreover, clinical transplant biopsies show an even greater extent of complement deposition compared to those from animals [229,

357]. Higher levels of C3 activity in the brain-dead donor are correlated to increased expression of the proinflammatory cytokines TNFα and IL-1β, as well as to endothelial activation and expression of the adhesion molecules VCAM-1 and E-selectin. Given the intimate link between complement expression and TLR signalling, it was surprising that complement levels in BD mice should be reduced in comparison to mice subjected to anaesthsesia and prolonged mechanical ventilation alone.

Both esRAGE and HSA vectors increased pancreatic insulin mRNA expression in the steady state and significantly increased pancreatic islet insulin content was evident in the BD mice transduced with either of these vectors. Although the mechanisms contributing to this increase are unclear, it is possible to postulate that complex effects

253 of high levels of transgene expression in the liver upon glucose homeostasis may be involved. In further studies, it would be possible to substitute administration of recombinant soluble esRAGE for use of an esRAGE-encoding vector to avoid any such effects.

One striking and unexpected effect of esRAGE overexpression was increase in transcription of the NKG2D ligand RAE-1. RAE-1 expression often accompanies cellular proliferation [358]. The pancreas recovers from ischaemia-reperfusion injury through regeneration and proliferation of its parenchymal cells, and such cellular proliferation is promoted in situations where RAGE signalling is interrupted, such as in the presence of soluble RAGE [359]. Another possible factor contributing to the increased levels of RAE-1 in esRAGE-transduced mice, is the reported reciprocal relationship between RAE-1 and levels of various pro-inflammatory cytokines.

Downregulation of RAE-1 protein expression by factors such as MCP1, MIP1, IL-1α,

IL-1β, IL-33 and TNFα, has been described in salivary gland [360], another exocrine gland which may behave similarly to the pancreas in this regard. Expression of RAE-1 or other “stress ligands” could potentially increase the susceptibility of the pancreas or islets to attack by CD8+ T cells and/or NK cells following transplantation. However, given that RAGE itself is also expressed by activated and memory T cells, suppressive effects of esRAGE on T cell activation and function may outweigh RAE-1 expression in determining the outcome of islet or pancreas grafting in the presence of esRAGE.

254

CHAPTER 6 GENERAL DISCUSSION

255 Type 1 diabetes mellitus is an autoimmune disease, with a growing global prevalence estimated at 0.2-0.5%. Short-term complications of T1DM include hypoglycaemia, ketoacidosis, and infections. Long-term complications comprise microvascular and macrovascular complications, leading to end-stage organ damage. Treatment of type I diabetes mellitus with exogenous insulin prevents ketoacidosis, but can contribute to the development of hypoglycaemic episodes, sometimes accompanied by hypoglycaemia unawareness, as insulin therapy cannot respond to the constant changes in blood glucose. Pancreas, or alternatively, islet transplantation have the potential to restore normoglycaemia and reverse the long-term complications of diabetes mellitus, while preventing the life-threatening hypoglacaemic periods, even when unable to procure insulin-independence [14, 15].

Currently, donor pancreata comes predominantly from deceased donors and their availability is extremely limited. Donor brain death has detrimental effects on pancreatic tissue, which is very sensitive to ischemia, release of proteolytic enzymes and nucleases from the exocrine pancreas amplifies the inflammation. Moreover, when pancreatic islet procurement is undertaken, the damaging effects of isolation compound the islet’s susceptibility to the deleterious consequences of brain death. Brain death in combination with exposure to enzymatic digestion and cold ischemia during purification, results in the expression of proinflammatory markers such as tissue factor

(TF), MCP1 and IL-6 [213, 222, 267, 361, 362].

Expression of tissue factor and monocyte chemoattractant protein (MCP)-1 by islets triggers the instant blood mediated immune reaction (IBMIR) following intraportal infusion. IBMIR consequent upon the exposure of islets to whole blood induces

256 coagulation, neutrophil infiltration, and extensive loss of islet mass with poor engraftment [363]. For the islets surviving the initial stages of IBMIR, the proinflammatory microenvironment which has been established within the graft after the successive insults of brain death, isolation and IBMIR contributes to the severity of subsequent rejection and opposes the development of immunological tolerance [72,

222].

The present lack of a sustainable source for β cell replacement compels us to explore the possibilities of islet xenotransplantation. Porcine insulin is active in humans, and the relative compatibility between porcine and human glucose homeostasis suggests that porcine islets, and particularly porcine neonatal islet cell clusters (pNICCs) are a feasible alternative source for insulin replacement. IBMIR triggered by porcine islets is particularly intense due to pre-existing anti-Gal antibodies, although the severity of this reaction can be attenuated when islets are derived from genetically modified pigs lacking the Gal epitope. Various strategies to overcome the xenogeneic barrier are being proposed, such as immunomodulation, genetic engineering and immunosuppression.

The ability of such strategies to promote the acceptance and function of pNICCs is being actively studied in a number of experimental models [99, 105].

Interventions to reduce inflammation in donor pancreata and isolated islets are urgently required to increase organ availability and improve the outcomes of pancreas and islet transplantation. Anti-inflammatory treatments, optimised donor management protocols and immunoalteration techniques are amongst the strategies being vigorously pursued

[89, 114].

257

In this project, we evaluated the effect of three potential strategies for the attenuation of inflammation within donor pancreas or islets. These were:

1. Treatment of porcine neonatal islet cell clusters in culture with recombinant human activated protein C.

2. Use of a hormone resuscitation protocol instead of conventional pressor support with noradrenaline or other catecholamines in a porcine model of brain death.

3. Overexpression of the natural soluble anti-inflammatory molecule esRAGE (a decoy receptor binding HMGB1 and other danger signals) in a mouse model of brain death.

Treatment of porcine neonatal islet cell clusters with rHuAPC

Recombinant human activated protein C has been reported to reduce IBMIR in vivo and in vitro through anti-coagulation, fibrinolysis and NF-B inhibition pathways [117]

When human islets were exposed to ABO-incompatible blood, APC protected islet integrity in a dose-dependent manner [305]. Treatment of murine recipients with APC during intraportal transplant reduced endothelial activation in the recipient, increased engraftment and functionality, and reduced immunogenicity and apoptosis in biopsies

[306].

We studied pNICCs during a 14-day-culture period. They carry a considerable amount of exocrine tissue for the first few days, shedding this mostly by day 3 post–isolation.

When the expression of a panel of genes was investigated using quantitative rt-PCR, maturation of the pNICCs during the 14 days in culture was accompanied by a gradual

258 increase in insulin mRNA levels. The expression of most proinflammatory mediators typically reached the peak by day 6 of culture followed by a reduction. However, stress- induced ligands pULBP1 and MIC2 increased and remained elevated.

Protein and mRNA sequence comparison of porcine and human protein C reveals a high level of similarity – the mRNA sequences are 83% identical, with 74% identity at the protein level. The thrombin cleavage site on protein C is completely homologous in pig and human. Endothelial protein C receptor demonstrated a high level of conservation.

In particular, 7/10 of the residues recognised to be involved in APC binding to EPCR are identical between pig and human while the remaining 3 residues were from similar families.

According to these findings and the literature, we proposed that rHuAPC should be active on porcine tissues. When porcine aortic endothelial cells were treated with rHuAPC in culture, they produced a small increase in IL-8 and MCP1 levels suggestive of possible APC activity [195, 196], although alterations in gene expression were modest when compared to those resulting from culture itself.

Whilst EPCR is characteristically found on vascular endothelial cells, expression of this receptor in cells of non-endothelial lineage (e.g. keratinocytes, tenocytes, lymphocytes, monocytes) has been described [192] EPCR gene expression was detected in pNICCs.

However, we could not identify expression of the receptor at a protein level, and it is possible that mRNA expression in pNICCs originated from vascular remnants, that mRNA transcription is not reflected by the presence of EPCR protein at the cell surface,

259 or that EPCR previously present at the cell surface has been cleaved by enzymes such as

ADAM 17/TACE.

Although APC has a half-life of 20-30 minutes in circulation due to inactivation by serum components, its biological effect has a much longer duration: twice-weekly dosing with rHuAPC protected NOD mice from the development of autoimmune diabetes over a 10 week period [308]. Distinct from the results obtained when PAECs were treated with rHuAPC, treatment of isolated pNICCs with rHuAPC in culture did not result in any change in the gene expression profile. The absence of detectable EPCR protein on the surface of islet cells along with these results suggests that the addition of rHuAPC to culture media is not likely to be a useful intervention for the reduction of inflammation in porcine islets. In contrast rHuAPC treatment of the recipients of pNICCs could potentially attenuate IBMIR and islet graft destruction and should be further investigated. Lack of penetration of rHuAPC into the islet clusters in addition to overshadowing of other pathways are some other propositions for our observations.

Effects of hormone resuscitation on brain-dead pig pancreas

Adult pigs were subjected to brain death and two different donor management protocols, hormone resuscitation (HR) and standard pressor support with noradrenaline

(NA).

Animals receiving HR maintained higher mean arterial pressure, lower heart rate and increased stroke work. HR was associated with better renal and pulmonary function and preservation of morphology in pancreas biopsies; although serum levels of the pancreatic enzymes amylase and lipase and various parameters of liver function were comparable between the two groups. Of note, amylase and lipase are late markers of

260 damage after pancreatic IRI. These features are indicative of superior haemodynamic status compared to the NA-receiving group.

The pattern of gene expression for a number of inflammatory mediators was evaluated in whole pancreas specimens – these genes included danger signals, pattern recognition receptors, cytokines, chemokines and stress ligands downstream of TLR signalling, as well as the protective molecule A20. Expression of pro-inflammatory markers was, in general, upregulated in pigs receiving noradrenaline, whereas expression of these genes either showed a more limited increase or was downregulated in the HR group.

Activation of the complement cascade contributes to inflammatory events following brain death and helps to trigger adaptive immune responses against the graft. The C3 mRNA level in pancreas was measured by RT-PCR, showing reduced gene expression in pigs treated with hormone resuscitation compared with noradrenaline-treated animals. This result is in accordance with the overall reduction in proinflammatory gene expression in the pancreas of HR pigs.

Whereas the islet content of insulin was less in the pigs receiving noradrenaline 6 hours following BD than at baseline, insulin mRNA expression was increased in this group, implying that insulin synthesis was increased in an attempt to maintain glucose homeostasis, potentially contributing to islet stress. In contrast, HR pigs receiving exogenous insulin as part of the hormonal cocktail maintained islet insulin content with reduced insulin synthesis.

Given these results, it can be concluded that hormone resuscitation has beneficial effects on pancreatic inflammation and islet integrity. Possible mechanisms include:

261

1. Improved haemodynamic status results in better perfusion of pancreata, less ischaemic damage and therefore less release of DAMPs. Therefore, the innate immune activation of the organ is diminished.

2. The methylprednisolone within the HR cocktail exerts direct anti-inflammatory effects. Application of glucocorticoids, and commonly methylprednisolone as part of the invasive donor management protocol has been shown to support organ perfusion, recovery and possibly survival [348]. However, corticosteroids can suppress the inflammatory response in tissues as well; anti-inflammatory properties of these agents are propagated by the binding of the corticosteroid to glucocorticosteroid receptors

(GRs) and translocation of this complex to the nucleus. This results in two subsequent events: downregulation of downstream proinflammatory molecules by suppression of co-activator molecules with intrinsic histone acetyltransferase (HAT) activity such as cyclic AMP response element binding (CREB) binding protein (CBP) which are induced by NF-κB (trans-repression); and upregulation of anti-inflammatory genes

[such as NF-κB inhibitor IκB-α, Annexin-1, Antileukoproteinase Mitogen-activated protein kinase (MPK)-1, glucocorticoid-induced leucine zipper (GILZ),] through direct binding to glucocorticoid response elements [364] in regulatory regions, promoting transcription of these steroid-sensitive genes (trans-activation)[349].

3. The methylprednisolone component of the hormone cocktail induces A20 expression, which then mediates anti-inflammatory effects. As a negative feedback regulator of innate immunity, A20 is directly induced when signalling by any of the activators such as TNFα, TLR4/IL-1, NOD2 or IL-17 leads to NF-κB translocation [177, 180]. This

262 serine ubiquitinase mainly acts through deubiquitination to suppress inflammatory signalling, deactivating RIP1 in the TNFα, RIP2 in the NOD2, and TRAF6 in the

TLR4/IL-1 and IL-17 pathways. It has been shown that corticosteroids selectively spare the expression of anti-inflammatory targets of TNF through context-dependent cooperation between GR and NF-κB [350]. This finding can explain our observations showing downregulation of multiple NF-κB-related proinflammatory regulators, and simultaneous increase in A20 gene expression in hormone-resuscitated animals.

4. Low circulating insulin and hyperglycaemia caused by brain death is rectified by exogenous insulin in the hormone cocktail [270], this adjustment protects the endocrine pancreas from overload, preserves the insulin content while reducing insulin gene transcription in the HR group. Glucotoxicity and the related islet stress/exhaustion has deleterious effects on islets manifest as reduced glucose sensitivity, downregulation of metabolic genes, diminished proliferation and differentiation, increased inflammation and apoptosis [365] Hyperglycaemia also increases oxygen consumption, exacerbating the ischemic condition of islets [366]. While diminished islet graft outcome and survival is known to be accelerated by islet exhaustion - manifested by excessive release of insulin and C-peptide by dead β cells within the garft within 24 hrs - peritransplant exogenous insulin administration was shown to reduce islet exhaustion and substantially improve clinical transplant success[79] [342].

5. The increased proinflammatory status observed in the noradrenaline-receiving group could be related to the exogenous catecholamine. Release of adrenaline and noradrenaline into the circulation following increased intracranial pressure or sudden hypovolemia induces a strong inflammatory response. Engagement of α2-adrenregic

263 receptors upregulates synthesis and release of TNFα and leads to translocation of NF-

κB, initiating production of downstream proinflammatory molecules including IL-8, IL-

1β, and IL-6. Release of TNFα, IL-1β, IL-2 and IL-8 into the circulation after reperfusion in human liver transplant recipients is correlated with increased levels of noradrenaline administered to the donor for blood pressure maintenance [346, 347,

367].

Neither TUNEL nor cleaved caspase-3 staining was significantly increased at 6 hours after induction of BD in either treatment group compared with baseline. However, we did find that TUNEL staining of both porcine and murine pancreas substantially overestimated the numbers of apoptotic cells when the results were compared with detection of cleaved caspase-3. According to the literature, TUNEL assay may detect single-stranded DNA breaks with free 3’-OH ends that appear during DNA repair. High levels of false positive staining can also occur in tissues with high rates of protein synthesis and are ascribed to DNA strand breaks happening during RNA transcription.

Moreover, tissue deoxyribonucleases can degrade genomic DNA within samples. Thus the TUNEL method should be accompanied and verified by other parallel detection methods including Annexin V, Caspase activity, single strand DNA, etc [344, 345]. We attempted to identify the extent of apoptosis in pancreas using a third method, DNA laddering. Detection of the typical cleavage pattern of genomic DNA by endonucleases upon agarose gel electrophoresis, yielded only smears of degraded DNA from pancreas samples, where the typical ladder pattern was observed in mouse splenocytes treated with etoposide. The exocrine pancreas contains high concentration of nucleases as well as proteases and degradation of genomic DNA by these enzymes during the extraction process make this technique unreliable for the detection of apoptosis in pancreas. Thus

264 we propose that apoptosis detection results from pancreas or islets should be interpreted with caution.

Systemic overexpression of esRAGE

Signalling through pattern recognition receptors is central to the inflammation which accompanies brain death and ischemia-repercussion injury. Endogenous soluble RAGE

(esRAGE) acts in two ways to block this signalling. Firstly, esRAGE is a soluble decoy receptor which sequesters HMGB1, S100B and other ligands for the pattern recognition receptors TLR2, TLR4, TLR9 and RAGE – in addition, esRAGE forms heterodimers with transmembrane RAGE, thus preventing signal transduction [149].

We have used an AAV system to achieve systemic overexpression of esRAGE in mice undergoing brain death in order to determine whether this could attenuate inflammation in pancreas and other transplantable organs.

Overexpression of esRAGE was well-tolerated, without liver inflammation and could be manifested for at least 100 days after transduction of mice in-vivo with AAV-esRAGE.

The mouse brain death model recapitulated features of brain death in a clinical setting with an initial spike in mean arterial pressure followed by a slow decline, and tachycardia which was sustained. Serum levels of proinflammatory cytokines and chemokines were upregulated up to thousands of folds within one hour after brain death, and both HMGB1 and S100B were released into the systemic circulation with the greatest levels being attained at 3 hours after the induction of brain death. These soluble mediators may originate from multiple sources. S100B is particularly abundant in the

265 CNS but can also be released by other damaged tissues. Cytokines and chemokines such as IL-6 and IL-8 can be synthesized by a variety of parenchymal cell types, while

HMGB1 is ubiquitous, and is released either passively following necrosis or actively via a vesicle-mediated mechanism from stressed cells. Gene expression of proinflammatory mediators in the pancreas was examined in mice at 1 and 3 hours after the induction of brain death, in normal controls and in mice undergoing mechanical ventilation for 3 hours but without the induction of brain death.

Gene expression of inflammatory mediators was not influenced by esRAGE in animals without surgical manipulation. Levels of proinflammatory molecules, including MCP1,

MIP2, KC, IL-6, IL-1β, TNFα, and IL-33 were upregulated by brain death and generally also increased by instrumentation and prolonged mechanical ventilation over 3 hours.

Expression of these proinflammatory genes in pancreas was modestly reduced in esRAGE-expressing animals compared to the control groups. esRAGE is thought to suppress both TLR and RAGE signalling through sequestration of RAGE ligands or prevention of RAGE homodimerisation at the cell surface, which downregulates proinflammatory cytokine and chemokine production. However, the extent to which esRAGE inhibits signalling through other pattern-recognition receptors depends upon the extent to which these receptors are engaged by ligands which are shared with

RAGE. Whilst HMGB1 is paramount amongst these, other ligands for TLRs will be unaffected by esRAGE overexpression, as will ligands for other PRRs such as RIG and

NOD-like receptors. Regulation of expression by alternative PRRs [368] may explain why IL-33 expression in the pancreas appears to be increased in mice transduced with esRAGE.

266 The protective molecule A20 was upregulated in response to brain death and also ventilation. Levels were lower in esRAGE-expressing BD animals compared to the controls; the mechanism might involve general suppression of NF-κB signalling by esRAGE overexpression.

Interestingly, gene expression of retinoic acid early inducible gene 1 (RAE-1) was substantially greater in the esRAGE-expressing BD animals compared to the other groups. RAE-1 expression has been observed to accompany cellular proliferation [358].

The pancreas recovers from ischaemia-reperfusion injury through regeneration and proliferation of its parenchymal cells, and such cellular proliferation is promoted in situations where RAGE signalling is interrupted, such as in the presence of soluble

RAGE [359]. Another explanation for RAE-1 upregulation might be provided by the reciprocal relationship between RAE-1 and levels of various pro-inflammatory cytokines. Upregulation of this ligand is shown to be independent of TNFα [353], distinguishing its pathway from TNFα-induced cytokines and chemokines. In fact,

RAE-1 can be downregulated at a protein level by factors such as the proinflammatory chemokines and cytokines MCP1, MIP1, IL-1α, IL-1β, IL-33 and TNFα, in tissues such as salivary gland, another exocrine gland which might behave similarly to the pancreas

[360]. NKG2D ligand gene expression in porcine neonatal islet clusters in vitro also persisted whilst islet cells were proliferating, despite the noticeable reduction in levels of other pro-inflammatory genes. Expression of RAE-1 or other “stress ligands” could potentially increase the susceptibility of the pancreas or islets to attack by CD8+ T cells and/or NK cells following transplantation. However, given that RAGE itself is also expressed by activated and memory T cells, suppressive effects of esRAGE on T cell

267 activation and function may outweigh RAE-1 expression in determining the outcome of islet or pancreas grafting in the presence of esRAGE.

Overall, the most promising therapy for attenuation of inflammation in pancreas and islets would appear to be hormone resuscitation of the BD donor. This therapy is already in clinical usage, and while the rationale for its introduction was to improve the quality of hearts from potential donors with unacceptable haemodynamic parameters, beneficial effects in pancreas and other transplantable organs suggest that HR may find broader application in donor management. Over expression of esRAGE showed some promise in reducing inflammation in BD donor pancreas, with the caveat that a general reduction in pro-inflammatory cytokines was accompanied by an increase in expression of the NKG2D ligand RAE-1.

More information about the ability of these two treatments to protect pancreas and islets would be obtained after isolation of islets from BD donors, subsequent functional testing and transplantation. It is also possible that HR and administration of soluble recombinant esRAGE would have a synergistic effect if used in combination.

Treatments aimed at reducing inflammation in isolated islets could also be combined with the donor treatments above. Although rHuAPC may not be useful in this context, other manipulations such as transfer of genes encoding protective molecules to islets

[109, 111] may confer additional benefit, as may treatment of the islet recipient with rHuAPC.

268 APPENDIX : COMPOSITION OF BUFFERS, SOLUTIONS AND MEDIA

1. 0.1% DEPC H20

1 mM of DEPC (Sigma) was dissolved in 1L of double-distilled autoclaved water by shaking the solution and incubated for 3 hours at 37ºC. The solution was then autoclaved at 70ºC for 1 hour to deactivate the DEPC.

2. 2M Sodium Acetate

8.12 g of Sodium Acetate (Sigma) was dissolved in 44.2 ml of DEPC-treated/ autoclaved water. The pH was adjusted to 4.0 with 5.8 ml concentrated HCl (BDH

AnalaR). After adding 0.1% DEPC (Sigma), the solution was incubated for 1 hour at room temperature, heated at 60ºC overnight, then filtered through an 0.22 μm filter and aliquoted into 5ml lots.

3. 49:1 Chloroform: Isoamyl Alcohol

2 ml of isoamyl alcohol (Sigma) was added to 98 ml of chloroform (BDH AnalaR) in a

DEPC-treated glass container.

4. 75% Ethanol

37.5 ml of Ethanol (BDH AnalaR) was diluted with 12.5 ml of DEPC-treated/ autoclaved water and stored at -20ºC.

269 5. 50x Tris-acetate EDTA buffer (TAE)

242 g of Tris.HCL (BDH AnalaR), 57.1 ml glacial acetic acid (Histolabs, Sydney,

Australia) and 100 ml of 0.5M EDTA, pH 8.0 (Ajax Univar) was mixed and made up to

1 L with triple-distilled water (50x stock solution). 1x TAE was made up by diluting 40 mL of the 50x stock solution in 1.96 L of triple-distilled water.

6. 1M Tris.HCl, pH 7.5

121.1 g of Tris.HCl (BDH AnalaR) was dissolved in 800 mL DEPC- treated water and the pH was adjusted to 7.5 with concentrated HCl (BDH AnalaR). The solution was made up to 1 L with DEPC-treated water and autoclaved for 20 min at 121ºC.

7. 0.5M EDTA, pH 8.0

186.1 g of EDTA (Ajax Univar) was dissolved in 800 mL DEPC-treated water and pH adjusted to 8.0 with 15M NaOH. This solution was made up to 1L with DEPC-treated water and autoclaved for 20 min at 121ºC.

8. 1x Tris EDTA buffer

1 mL of 1 M Tris.HCl (pH 7.5) was mixed with 0.2 mL of 0.5 EDTA solution (pH 8.0).

The solution was then made up to 1 L with 988 mL autoclaved double-distilled H2O.

9. 7.5 M Ammonium Acetate (mw=77.08)

57.81 g ammonium acetate (Ajax Univar) was dissolved in 45ml DEPC-treated water.

The solution was made up to 100 ml with DEPC-treated water and filter sterilized through an 0.22 μm filter.

270 10. Hanks’ Buffered Salt Solution (HBBS) 1X

To 100 mL of HBBS 10X , 20 mL of Pen/Strep solution (Invitrogen), 12.5 mL HEPES

1M solution and 5 mL of NaHCO3 were added and was made up to 1L with double- distilled water. 10% pig serum (10mL) was added prior to use.

11. HAMS-F10 medium for PNICC culture

After dissolving 9.8 g of HAMS-F10 powder in 750 mL of double-distilled Mili Q water and filtering through 0.22 μm plastic filter, following sterile material were added to the solution:

20 mL Pen/Strep solution (Invitrogen), 20 mL L-Glutamin 200mM, 1 mL CaCl2

236g/L, 14 mL D-glucose 50g/L, 10 mL Nicotinamide 1M, 10 mL IBMX 5mM, 12.5 mL HEPES 1M, 16 mL 7.5% NaHCO3. The solution was then made up to 1 L with

Mili Q water and 110 mL of heat inactivated porcine serum was added afterwards.

12. M199 complete medium

M199 media (Gibco), 10% heat-deactivated FBS, 100 μg/ml L-glutamine, 100 U/ml

Penicillin, 100 mg/ml Streptomycin, 10 mm HEPES (Sigma), 50 μg/ml ECGS (Sigma).

The solution was filter-sterilised with an 0.22 μm filter, and stored at 4 °C.

13. Collagenase type I solution (0.26%)

Just before use, 780 μg of Collagenase type I (Sigma) powder was dissolved in 3 ml of enzyme buffer and sterilised by passing through and 0.22 μm filter.

271 14. 1% w/v Gelatine (for coating culture flasks)

1g of bovine gelatine (Sigma) was dissolved in 1L of hot autoclaved distilled water. The solution was poured at 1ml/cm2 into culture flasks, covering the surface thoroughly, allowed to set for 5 min at room temperature, then the excess liquid was discarded. The flasks were stored at 4°C for up to a week.

15. LB-Agar

In a 2 liter flask 10g bacto-tryptone (BD, Franklin Lakes, NJ), 5g bacto-yeast (BD,

Franklin Lakes, NJ), 5g sodium chloride and 15g of bacto-agar (BD, Franklin Lakes,

NJ), was placed. Distilled water was added to the mixture until the volume was about

500 ml. The flask was covered with foil and autoclaved for 20 minutes at 121°C, then allowed to cool. The LB-agar was again heated and dissolved before use for bacterial culture.

16. Phosphate Buffered Saline (PBS)

1 sachet of PBS powder (Invitrogen) was dissolved in 1L of d H20.

17. Tris Buffered Saline (TBS x10)

To 1L dH2O, 24.2 g Trizma base (Sigma) and 80g sodium chloride (Sigma) were added. pH wasadjusted to 7.6 with concentrated HCl.

18. Tris Buffered Saline (TBS x1)

100ml of 10x TBS was added to 900 ml dH2O.

272 19. Tris Buffered Saline+Tween-20 TBST:

100 ml of 10x TBS was added to 900 ml dH2O. Add 1 ml Tween-20(Amresco, Solon,

OH) and mix.

273 REFERENCES

1. TSANZ, Australia and New Zealand Organ Donation Registry: 2012 Report, 2012. 2. Vinkers, M.T., et al., How to recognize a suitable pancreas donor: a eurotransplant study of preprocurement factors. Transplantation Proceedings, 2008. 40(5): p. 1275-1278. 3. Global health risks. Mortality and burden of disease attributable to selected major risks. Geneva, World Health Organization, 2009. 4. Danaei, G., et al., 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, 2011. 378(9785): p. 31-40. 5. Del Guerra, S., et al., Functional and molecular defects of pancreatic islets in human type 2 diabetes, 2005. p. 727. 6. Wentworth, J.M., et al., Reappraising the stereotypes of diabetes in the modern diabetogenic environment. Nat Rev Endocrinol, 2009. 5(9): p. 483-489. 7. Maahs, D.M., et al., Epidemiology of type 1 diabetes. Endocrinology and Metabolism Clinics of North America, 2010. 39(3): p. 481-497. 8. Jiang, X., et al., Early life factors and type 2 diabetes mellitus. Journal of Diabetes Research, 2013. 2013. 9. Uchiba, M., et al., Activated protein c induces endothelial cell proliferation by mitogen-activated protein kinase activation in vitro and angiogenesis in vivo. Circulation research. 2004. 95: p. 34-41. 10. Liaw, P.C.Y., et al., Identification of the protein c/activated protein c binding sites on the endothelial cell protein c receptor. Implications for a novel mode of ligand recognition by a major histocompatibility complex class 1-type receptor. The Journal of Biological Chemistry, 2001. 276(11): p. 8364-8370. 11. Dabelea, D., et al., Prevalence of type 1 and type 2 diabetes among children and adolescents from 2001 to 2009. JAMA. 311(17): p. 1778-1786. 12. Johnson, P.R.V., et al., Pancreatic islet transplantation. Seminars in pediatric Sudrgery, 2012. 21: p. 272-280 13. Awoniyi, O., et al., Hypoglycemia in patients with type 1 diabetes: epidemiology, pathogenesis, and prevention. Current Diabetes Reports, 2013. 13(5): p. 669-678. 14. Shapiro, A.M.J., et al., Clinical islet transplant: current and future directions towards tolerance. Immunological Reviews, 2003. 196(1): p. 219-236. 15. Ludwig, B., et al., Islet versus pancreas transplantation in type 1 diabetes: competitive or complementary? Current Diabetes Reports, 2010. 10(6): p. 506- 511. 16. Kelly, W.D., et al., Allotransplantation of Pancreas and Duodenum Along with Kidney in Diabetic Nephropathy. Surgery, 1967. 61(6): p. 827. 17. Gruessner, R.W.G., et al., Over 500 solitary pancreas transplants in nonuremic patients with brittle diabetes mellitus. Transplantation. 2008. 85(1): p. 42-47. 18. Odorico, J.S., et al., Improved solitary pancreas transplant graft survival in the modern immunosuppressive era. Cell Transplantation, 2000. 9(6): p. 919-927.

274 19. Gruessner, R.W.G., et al., The current state of pancreas transplantation. Nature Reviews Endocrinology, 2013. 20. Srinivasan, P., et al., Islet cell transplantation. Postgraduate Medical Journal, 2007. 83(978): p. 224-229. 21. Lacy, P.E. and Kostiano.M, Method for Isolation of Intact Islets of Langerhans from Rat Pancreas. Diabetes, 1967. 16(1): p. 35. 22. Ballinger and L. PE, Transplantation of intact pancreatic-islets in rats. Surgery, 1972. 72(2): p. 175. 23. Najarian, J.S., et al., Total or near total pancreatectomy and islet auto- transplantation for treatment of chronic-pancreatitis. Annals of Surgery, 1980. 192(4): p. 526-542. 24. Largiader, F., et al., Long-term functioning human pancreatic-islet allotransplant. Transplantation, 1980. 29(1): p. 76-77. 25. Pyzdrowski, K.L., et al., Preserved insulin-secretion and insulin independence in recipients of islet autografts. New England Journal of Medicine, 1992. 327(4): p. 220-226. 26. Shapiro, A.M., et al., Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. New England Journal of Medicine, 2000. 343: p. 230-238. 27. Shapiro, A.M.J., Islet transplantation in type 1 diabetes: ongoing challenges, refined procedures, and long-term outcome. The review of diabetic studies, 2012. 9(Feb). 28. Tydén, G., et al., Combined pancreas and kidney transplantation improves survival in patients with end-stage diabetic nephropathy. Clinical Transplantation, 2000. 14(5): p. 505-508. 29. Rouhiainen, A., et al., Regulation of monocyte migration by amphoterin (HMGB1). Blood, 2004. 104(4): p. 1174-82. 30. O'Brien, L.A., et al., Activated protein c decreases tumor necrosis factor related apoptosis-inducing ligand by an EPCR- independent mechanism involving EGR-1/ERK-1/2 activation. Arteriosclerosis, Thrombosis, and Vascular Biology, 2007. 27: p. 2634-2641. 31. Kissler, H.J., et al., Validation of methodologies for quantifying isolated human islets: an islet cell resources study. Clinical Transplantation, 2009. 24(2): p. 236-242. 32. Ojo, A.O., et al., The impact of simultaneous pancreas-kidney transplantation on long-term patient survival1. Transplantation, 2001. 71(1): p. 82-89. 33. Alejandro, R., et al., Update from the collaborative islet transplant registry. Transplantation, 2008. 86(12): p. 1783-1788. 34. Sutherland, D.E.R., et al., Lessons learned from more than 1,000 pancreas transplants at a single institution. Annals of Surgery, 2001. 233(4): p. 463-501. 35. Dan, R., et al., Preservation of the donor pancreas for whole pancreas and islet transplantation. Clinical Transplantation, 2009. 24: p. 1-19. 36. Fridell, J.A., et al., The pancreas allograft donor: current status, controversies, and challenges for the future. Clinical Transplantation, 2010. 24(4): p. 433-449. 37. Maglione, M., et al., Donor risk factors, retrieval technique, preservation and ischemia/reperfusion injury in pancreas transplantation. Current Opinion in Organ Transplantation, 2013. 18: p. 83-88. 38. Axelrod, D.A., et al., Systematic evaluation of pancreas allograft quality, outcomes and geographic variation in utilization. American Journal of Transplantation, 2009. 10(4): p. 837-845.

275 39. Boggi, U., et al., Pancreas transplants from donors aged 45 years or older. Transplantation Proceedings, 2005. 37(2): p. 1265-1267. 40. Domínguez-Gil, B., et al., Current situation of donation after circulatory death in European countries. Transplant International, 2011. 24(7): p. 676-686. 41. Muthusamy, A.S.R., et al., Pancreas Transplantation From Donors After Circulatory Death From the United Kingdom. American Journal of Transplantation, 2012. 12(8): p. 2150-2156. 42. Qureshi, M.S., et al., Outcomes of simultaneous pancreas–kidney transplantation from brain-dead and controlled circulatory death donors. British Journal of Surgery, 2012. 99(6): p. 831-838. 43. Bellingham, J.M., et al., Donation after cardiac death: A 29-year experience. Surgery, 2011. 150(4): p. 692-702. 44. Brandhorst, H., et al., Body mass index of pancreatic donors: a decisive factor for human islet isolation. Exp Clin Endocrinol Diabetes, 1995. 103(S 02): p. 23- 26. 45. Brockmann, J.G., et al., Retrieval of abdominal organs for transplantation. British Journal of Surgery, 2006. 93(2): p. 133-146. 46. Wunderlich, H., et al., DTG procurement guidelines in heart beating donors. Transplant International, 2011. 24(7): p. 733-757. 47. Kin, T., et al., Surgical aspects of human islet isolation. Islets, 2010. 2(5): p. 265-273. 48. Stewart, Z.A., et al., Histidine-Tryptophan-Ketoglutarate (HTK) is associated with reduced graft survival in pancreas transplantation. American Journal of Transplantation, 2009. 9(1): p. 217-221. 49. Alonso, D., et al., Increased pancreatitis in allografts flushed with histidine- tryptophan-ketoglutarate solution: a cautionary tale. American Journal of Transplantation, 2008. 8(9): p. 1942-1945. 50. Castillo, C.F.n.-d., et al., Risk factors for pancreatic cellular injury after cardiopulmonary bypass. N Engl J Med, 1991. 325: p. 382-387. 51. Maglione, M., et al., Donor pretreatment with tetrahydrobiopterin saves pancreatic isografts from ischemia reperfusion injury in a mouse model. American Journal of Transplantation, 2010. 10(10): p. 2231-2240. 52. Drognitz, O., et al., Effects of organ preservation, ischemia time and caspase inhibition on apoptosis and microcirculation in rat pancreas transplantation. American Journal of Transplantation, 2004. 4(7): p. 1042-1050. 53. Schaser, K.-D., et al., In vivo imaging of human pancreatic microcirculation and pancreatic tissue injury in clinical pancreas transplantation. American Journal of Transplantation, 2005. 5(2): p. 341-350. 54. Hegyi, P., et al., The role of nitric oxide in the physiology and pathophysiology of the exocrine pancreas. Antioxidants & Redox Signaling, 2011. 15: p.2723- 2741 55. Hackert, T., et al., Reduction of ischemia/reperfusion injury by antithrombin III after experimental pancreas transplantation. The American Journal of Surgery, 2005. 189(1): p. 92-97. 56. Preissler, G., et al., Intercellular adhesion molecule-1 blockade attenuates inflammatory response and improves microvascular perfusion in rat pancreas grafts. Pancreas, 2012. 41: p. 1112-1118. 57. Oehmann, C., et al., Remote preconditioning reduces microcirculatory disorders in pancreatic ischemia/reperfusion injury. Pancreas, 2007. 35: p. e45- e50.

276 58. Nikeghbalian, S., et al., The effect of ischemic preconditioning of the pancreas on severity of ischemia/reperfusion-induced pancreatitis after a long period of ischemia in the rat. Transplantation Proceedings, 2009. 41(7): p. 2743-2746. 59. NCT00975702, n., Remote Ischemic Preconditioning In Abdominal Organ Transplantation (RIPCOT). ClinicalTrials.gov, 2012. 60. Ricordi C, L.P., et al., Automated method for isolation of human pancreatic islets. Diabetes, 1988. 37(4): 423-420 61. Lakey, J.R.T., et al., Intraductal collagenase delivery into the human pancreas using syringe loading or controlled perfusion. Cell Transplantation, 1999. 8(3): p. 285-292. 62. Kin, T., et al., Factors influencing the collagenase digestion phase of human islet isolation. Transplantation, 2007. 83: p. 7-12. 63. Balamurugan, A.N., B. Ramakrishna, and S. Gunasekaran, Effect of cyclophosphamide pretreatment and culture on insulin secretion by isolated monkey pancreatic islets (Macaca radiata radiata). Transplantation Proceedings, 1999. 31(5): p. 2154-2156. 64. Mita, A., et al., Antiproinflammatory Effects of Iodixanol (OptiPrep)-Based Density Gradient Purification on Human Islet Preparations. Cell Transplantation, 2010. 19(12): p. 1537-1546. 65. Hering, B.J., et al., Single-donor, marginal-dose islet transplantation in patients with type 1 diabetes. JAMA, 2005. 293(7): p. 830-835. 66. Barbaro, B., et al., Improved human pancreatic islet purification with the refined UIC-UB density gradient, 2007. p. 1200-1203. 67. Froud, T., et al., Islet transplantation in type 1 diabetes mellitus using cultured islets and steroid-free immunosuppression: Miami experience. American Journal of Transplantation, 2005. 5(8): p. 2037-2046. 68. Rabinovitch, A., et al., Human pancreatic islet beta-cell destruction by cytokines is independent of nitric oxide production. The Journal of Clinical Endocrinology and Metabolism, 1994. 79(4): p. 1058-1062. 69. Bennet, W., et al., Isolated human islets trigger an instant blood mediated inflammatory reaction: Implications for intraportal islet transplantation as a treatment for patients with type 1 diabetes. Upsala Journal of Medical Sciences, 2000. 105(2): p. 125-133. 70. Moberg, L., et al., Production of tissue factor by pancreatic islet cells as a trigger of detrimental thrombotic reactions in clinical islet transplantation. The Lancet, 2002. 360(9350): p. 2039-2045. 71. Johansson, H., et al., Tissue factor produced by the endocrine cells of the islets of langerhans is associated with a negative outcome of clinical islet transplantation. Diabetes, 2005. 54: p. 1755-1762. 72. Tjernberg, J., et al., Acute antibody-mediated complement activation mediates lysis of pancreatic islets cells and may cause tissue loss in clinical islet transplantation. Transplantation, 2008. 85(8): p. 1193-1199. 73. Özmen, L., et al., Inhibition of Thrombin Abrogates the Instant Blood-Mediated Inflammatory Reaction Triggered by Isolated Human Islets: Possible Application of the Thrombin Inhibitor Melagatran in Clinical Islet Transplantation. Diabetes, 2002. 51(6): p. 1779-1784. 74. Gruessner, A.C., et al., Long-term outcome after pancreas transplantation. Transplantation, 2012. 17(1): p. 100-105. 75. Reddy, K.S., et al., Long-term survival following simultaneous kidney-pancreas transplantation versus kidney transplantation alone in patients with type 1

277 diabetes mellitus and renal failure. American Journal of Kidney Diseases, 2003. 41(2): p. 464-470. 76. Fioretto, P., et al., Reversal of lesions of diabetic nephropathy after pancreas transplantation. N Engl J Med, 1998. 339(2): p. 69-75. 77. Secchi, A., et al., Effect of pancreas transplantation on life expectancy, kidney function and quality of life in uraemic Type 1 (insulin-dependent) diabetic patients. Diabetologia, 1991. 34(1): p. S141-S144. 78. Robertson, M.D.R.P., Pancreas and islet transplants for patients with diabetes: taking positions and making decisions. Endocrine Practice, 1999. 5(1): p. 24-28. 79. Koh, A., et al., Insulin-heparin infusions peritransplant substantially improve single-donor clinical islet transplant success. Transplantation, 2010. 89(4): p. 465-471. 80. Markmann, J.F., et al., Insulin independence following isolated islet transplantation and single islet infusions. Annals of Surgery, 2003. 237(6): p. 741-750. 81. Posselt, A.M., et al., Islet transplantation in type 1 diabetics using an immunosuppressive protocol based on the anti-lfa-1 antibody efalizumab. American Journal of Transplantation, 2010. 10(8): p. 1870-1880. 82. Matsumoto, S., et al., Improving efficacy of clinical islet transplantation with iodixanol-based islet purification, thymoglobulin induction, and blockage of IL- 1β; and TNF-α. Cell Transplantation, 2011. 20(10): p. 1641-1647. 83. Matsumoto, S., et al., Insulin independence after living-donor distal pancreatectomy and islet allotransplantation. The Lancet, 2005. 365(9471): p. 1642-1644. 84. Turgeon, N.A., et al., Experience with a novel efalizumab-based immunosuppressive regimen to facilitate single donor islet cell transplantation. American Journal of Transplantation, 2010. 10(9): p. 2082-2091. 85. Gangemi, A., et al., Islet transplantation for brittle type 1 diabetes: the uic protocol. American Journal of Transplantation, 2008. 8(6): p. 1250-1261. 86. Ghofaili, K.A., et al., Effect of exenatide on β cell function after islet transplantation in type 1 diabetes. Transplantation, 2007. 83(1): p. 24-28. 87. Faradji, R.N., et al., Long-term metabolic and hormonal effects of exenatide on islet transplant recipients with allograft dysfunction. Cell Transplantation, 2009. 18(10-1): p. 1247-1259. 88. Bellin, M.D., et al., Potent induction immunotherapy promotes long-term insulin independence after islet transplantation in type 1 diabetes. American Journal of Transplantation, 2012. 12(6): p. 1576-1583. 89. Shapiro, A.M.J., State of the art of clinical islet transplantation and novel protocols of immunosuppression. Current Diabetes Reports, 2011. 11(5): p. 345-354. 90. O'Connell, P.J., et al., Multicenter australian trial of islet transplantation: improving accessibility and outcomes. American Journal of Transplantation, 2013. 13(7): p. 1850-1858. 91. Fiorina, P., et al., The clinical impact of islet transplantation. American Journal of Transplantation, 2008. 8(10): p. 1990-1997. 92. Venturini, M., et al., Early increase of retinal arterial and venous blood flow velocities at color doppler imaging in brittle type 1 diabetes after islet transplant alone. Transplantation, 2006. 81(9): p. 1274-1277.

278 93. Del Carro, U., et al., Evaluation of polyneuropathy markers in type 1 diabetic kidney transplant patients and effects of islet transplantation: neurophysiological and skin biopsy longitudinal analysis. Diabetes Care, 2007. 30(12): p. 3063-3069. 94. Warnock, G.L., et al., A multi-year analysis of islet transplantation compared with intensive medical therapy on progression of complications in type 1 diabetes. Transplantation, 2008. 86(12): p. 1762-1766. 95. Thompson, D.M., et al., Reduced Progression of Diabetic Microvascular Complications With Islet Cell Transplantation Compared With Intensive Medical Therapy. Transplantation, 2011. 91(3): p. 373-378. 96. Danielson, K.K., et al., Reduction in carotid intima-media thickness after pancreatic islet transplantation in patients with type 1 diabetes. Diabetes Care, 2012. 36(2): p. 450-456. 97. Gruessner, R.W.G. and A.C. Gruessner, Pancreas Transplant Alone: A procedure coming of age. Diabetes Care, 2013. 36(8): p. 2440-2447. 98. Barlow, A.D., et al., Current state of pancreas preservation and implications for dcd pancreas transplantation. Transplantation, 2013. 86(12): p. 1419-1424. 99. O'Connell, P., Pancreatic islet xenotransplantation. Xenotransplantation, 2002. 9(6): p. 367-371. 100. McCall, M.D., et al., Are stem cells a cure for diabetes? Clinical Science, 2009. 118(2): p. 87-97. 101. Jiang, J., et al., generation of insulin-producing islet-like clusters from human embryonic stem cells. Stem Cells, 2007. 25(8): p. 1940-1953. 102. Groth, C.G., et al., Transplantation of porcine fetal pancreas to diabetic patients. The Lancet, 1994. 344(8934): p. 1402-1404. 103. Good AH, C.D., Malcolm AJ, Ippolito RM, Koren E, Neethling FA, Ye Y, Zuhdi N, Lamontagne LR, Identification of carbohydrate structures that bind human antiporcine antibodies: implications for discordant xenografting in humans. Transplantation Proceedings, 1992. 24(559-562). 104. Ji, M., et al., The Importance of Tissue Factor Expression by Porcine NICC in Triggering IBMIR in the Xenograft Setting. Transplantation, 2011. 91(8): p. 841-846. 105. van der Windt, D.J., et al., Early Islet Damage After Direct Exposure of Pig Islets to Blood:Has Humoral Immunity Been Underestimated? Cell Transplantation, 2012. 21(8): p. 1791-1802. 106. Dwyer, K.M., et al., The transgenic expression of human CD39 on murine islets inhibits clotting of human blood. Transplantation, 2006. 82(3): p. 428-432. 107. van der Windt, D.J., et al., Clinical islet xenotransplantation: how close are we? Diabetes, 2012. 61(12): p. 3046-3055. 108. Hing, A.J., et al., The effects of hormone resuscitation on cardiac function and hemodynamics in a porcine brain-dead organ donor model. American Journal of Transplantation, 2007. 7(4): p. 809-817. 109. Hongjun Wang, C.F., et al., Induction of protective genes leads to islet survival and function. Journal of Transplantation, 2011. 2011. 110. Yolcu, E.S., et al., Pancreatic islets engineered with sa-fasl protein establish robust localized tolerance by inducing regulatory T cells in mice. The Journal of Immunology, 2011. 187(11): p. 5901-5909. 111. Grey, S.T., et al., Genetic engineering of a suboptimal islet graft with A20 preserves β-cell mass and function. Journal of Immunology, 2003. 170(12): p. 6250-6256.

279 112. Kanak, M.A., et al., Inflammatory response in islet transplantation. international journal of endocrinology, 2014. 2014: p. 13. 113. Citro, A., E. Cantarelli, and L. Piemonti, anti-inflammatory strategies to enhance islet engraftment and survival. Current Diabetes Reports, 2013. 13(5): p. 733-744. 114. Chhabra, P., et al., Overcoming barriers in clinical islet transplantation: Current limitations and future prospects. Current Problems in Surgery, 2014. 51(2): p. 49-86. 115. Ashkenazi, E., et al., Pancreatic islet xenograft survival in mice is extended by a combination of alpha-1-antitrypsin and single-dose Anti-CD4/CD8 therapy. PloS ONE, 2013. 8. 116. Mirenda, V., et al., Achieving permanent survival of islet xenografts by independent manipulation of direct and indirect T-cell responses. Diabetes, 2005. 54(4): p. 1048-1055. 117. Van Der Windt, D.J., et al., Rapid loss of intraportally transplanted islets: an overview of pathophysiology and preventive strategies. Xenotransplantation, 2007. 14(4): p. 288-297. 118. Abbas A. K., et al., Basic immunology: functions and disorders of the immune system 2014. 119. Matzinger, P., The danger model: A renewed sense of self. Science, 2002. 296(5566): p. 301-305. 120. Pratschke, J., et al., Brain death and its influence on donor organ quality and outcome after transplantation1. Transplantation, 1999. 67(3): p. 343-348. 121. Kumagai, Y., et al., Identification and functions of pattern-recognition receptors. J Allergy Clin Immunol, 2010. 125(5): p. 985-92. 122. Timsit, M.O., et al., Consequences of transplant quality on chronic allograft nephropathy. Kidney Int, 2010. 78(S119): p. S54-S58. 123. Paterson, H.M., et al., Injury primes the innate immune system for enhanced toll-like receptor reactivity. The Journal of Immunology, 2003. 171(3): p. 1473- 1483. 124. Tsan, M.F. and B. Gao, Endogenous ligands of Toll-like receptors. J Leukoc Biol, 2004. 76(3): p. 514-9. 125. Iwasaki, A., et al., Toll-like receptor control of the adaptive immune responses. Nat Immunol, 2004. 5(10): p. 987-995. 126. Andrade, C.F., et al., Innate immunity and organ transplantation: the potential role of toll-like receptors. American Journal of Transplantation, 2005. 5(5): p. 969-975. 127. Beg, A.A., Endogenous ligands of Toll-like receptors: implications for regulating inflammatory and immune responses. Trends Immunol, 2002. 23(11): p. 509-12. 128. Tesar, B.M., et al., The role of hyaluronan degradation products as innate alloimmune agonists. Am J Transplant, 2006. 6(11): p. 2622-35. 129. Yanai, H., et al., HMGB proteins function as universal sentinels for nucleic- acid-mediated innate immune responses. Nature, 2009. 462(7269): p. 99-103. 130. O'Neill, L.A., How Toll-like receptors signal: what we know and what we don't know. Curr Opin Immunol, 2006. 18(1): p. 3-9. 131. Damman, J., et al., Crosstalk between Complement and Toll-like Receptor Activation in Relation to Donor Brain Death and Renal Ischemia-Reperfusion Injury. American Journal of Transplantation, 2011. 11(4): p. 660-669.

280 132. Kim, I.K., et al., Impact of Innate and Adaptive Immunity on Rejection and Tolerance. Transplantation, 2008. 86(7): p. 889-894. 133. Leventhal, J.S., et al., Toll-like receptors in transplantation: sensing and reacting to injury. Kidney Int, 2012. 81(9): p. 826-832. 134. Schmidt, A.M., et al., The multiligand receptor RAGE as a progression factor amplifying immune and inflammatory responses. The Journal of Clinical Investigation, 2001. 108(7): p. 949-955. 135. Yan, S.F., et al., Mechanisms of disease: advanced glycation end-products and their receptor in inflammation and diabetes complications. Nat Rev Endocrinol, 2008. 4(5): p. 285-293. 136. Yan, S., et al., Receptor for AGE (RAGE) and its ligands—cast into leading roles in diabetes and the inflammatory response. Journal of Molecular Medicine, 2009. 87(3): p. 235-247. 137. Hudson, B.I., et al., Identification, classification, and expression of RAGE gene splice variants. The FASEB Journal, 2008. 22(5): p. 1572-1580. 138. Bierhaus, A., et al., Understanding RAGE, the receptor for advanced glycation end products. Journal of Molecular Medicine, 2005. 83(11): p. 876-886. 139. Ramasamy, R., et al., Arguing for the motion: Yes, RAGE is a receptor for advanced glycation endproducts. Molecular Nutrition & Food Research, 2007. 51(9): p. 1111-1115. 140. Yeh, C.H., et al., Requirement for p38 and p44/p42 mitogen-activated protein kinases in RAGE-mediated nuclear factor-kappaB transcriptional activation and cytokine secretion. Diabetes, 2001. 50(6): p. 1495-504. 141. Yan, S.D., et al., Enhanced cellular oxidant stress by the interaction of advanced glycation end products with their receptors/binding proteins. J Biol Chem, 1994. 269(13): p. 9889-97. 142. Mosquera, J.A., Role of the receptor for advanced glycation end products (RAGE) in inflammation. Invest Clin, 2010. 51(2): p. 257-68. 143. Chavakis, T., et al., The pattern recognition receptor (RAGE) is a counterreceptor for leukocyte integrins: a novel pathway for inflammatory cell recruitment. J Exp Med, 2003. 198(10): p. 1507-15. 144. Moser, B., et al., Receptor for advanced glycation end products expression on T cells contributes to antigen-specific cellular expansion in vivo. Journal of Immunology, 2007. 179(12): p. 8051-8058. 145. Dumitriu, I.E., et al., Requirement of HMGB1 and RAGE for the maturation of human plasmacytoid dendritic cells. European Journal of Immunology, 2005. 35(7): p. 2184-2190. 146. Lue, L.F., et al., Receptor for advanced glycation end products. Future neurology, 2009. 4(2): p.167-177. 147. Wang, L.J., et al., Increased serum high-mobility group box-1 and cleaved receptor for advanced glycation endproducts levels and decreased endogenous secretory receptor for advanced glycation endproducts levels in diabetic and non-diabetic patients with heart failure. Eur J Heart Fail, 2011. 13(4): p. 440-9. 148. Taguchi, A., et al., Blockade of RAGE-amphoterin signalling suppresses tumour growth and metastases. Nature, 2000. 405(6784): p. 354-60. 149. Zong, H., et al., Homodimerization is essential for the receptor for advanced glycation end products (RAGE)-mediated signal transduction. J Biol Chem, 2010. 285(30): p. 23137-46. 150. Oppenheim, J.J., et al., Alarmins: chemotactic activators of immune responses. Curr Opin Immunol, 2005. 17(4): p. 359-65.

281 151. Yang, H., et al., A critical cysteine is required for HMGB1 binding to Toll-like receptor 4 and activation of macrophage cytokine release. Proc Natl Acad Sci USA, 2010. 107: p. 11942-11947. 152. Andrade, C.F., et al., Innate immunity and organ transplantation: the potential role of toll-like receptors. Am J Transplant, 2005. 5(5): p. 969-75. 153. Bianchi, M.E., HMGB1 loves company. J Leukoc Biol, 2009. 86(3): p. 573-6. 154. Hreggvidsdottir, H.S., et al., The alarmin HMGB1 acts in synergy with endogenous and exogenous danger signals to promote inflammation. J Leukoc Biol, 2009. 86(3): p. 655-62. 155. Tian, J., et al., Toll-like receptor 9-dependent activation by DNA-containing immune complexes is mediated by HMGB1 and RAGE. Nat Immunol, 2007. 8(5): p. 487-96. 156. Goldstein, R.S., et al., Elevated high-mobility group box 1 levels in patients with cerebral and myocardial ischemia. Shock, 2006. 25(6): p. 571-574 157. Donato, R., Intracellular and extracellular roles of S100 proteins. Microsc Res Tech, 2003. 60(6): p. 540-51. 158. Sedaghat, F. and A. Notopoulos, S100 protein family and its application in clinical practice. Hippokratia, 2008. 12(4): p. 198-204. 159. Rainey, T., et al., Predicting outcome after severe traumatic brain injury using the serum S100B biomarker: results using a single (24h) time-point. Resuscitation, 2009. 80(3): p. 341-5. 160. Gonçalves, C.-A., et al., Biological and methodological features of the measurement of S100B, a putative marker of brain injury. Clinical Biochemistry, 2008. 41(10–11): p. 755-763. 161. Sen, J. and A. Belli, S100B in neuropathologic states: the CRP of the brain? J Neurosci Res, 2007. 85(7): p. 1373-80. 162. Yan, S.S., et al., Suppression of experimental autoimmune encephalomyelitis by selective blockade of encephalitogenic T-cell infiltration of the central nervous system. Nat Med, 2003. 9(3): p. 287-93. 163. Kislinger, T., et al., Receptor for advanced glycation end products mediates inflammation and enhanced expression of tissue factor in vasculature of diabetic apolipoprotein E-null mice. Arterioscler Thromb Vasc Biol, 2001. 21(6): p. 905-10. 164. Rothermundt, M., et al., S100B in brain damage and neurodegeneration. Microsc Res Tech, 2003. 60(6): p. 614-32. 165. Suzuki, K., et al., Priming effect of recombinant human interleukin-2 and recombinant human interferon-gamma on human neutrophil superoxide production. Arzneimittelforschung, 1990. 40(10): p. 1176-9. 166. Daniels, R.H., et al., Recombinant human monocyte IL-8 primes NADPH- oxidase and phospholipase A2 activation in human neutrophils. Immunology, 1992. 75(1): p. 157-63. 167. Jaeschke, H., et al., Superoxide generation by Kupffer cells and priming of neutrophils during reperfusion after hepatic ischemia. Free Radic Res Commun, 1991. 15(5): p. 277-84. 168. Hayashi, F., T.K. Means, and A.D. Luster, Toll-like receptors stimulate human neutrophil function. Blood, 2003. 102(7): p. 2660-9. 169. Brown, S.B. and J. Savill, Phagocytosis triggers macrophage release of Fas ligand and induces apoptosis of bystander leukocytes. J Immunol, 1999. 162(1): p. 480-5.

282 170. Hsieh, C.S., et al., Development of TH1 CD4+ T cells through IL-12 produced by Listeria-induced macrophages. Science, 1993. 260(5107): p. 547-9. 171. Novak, T.E., et al., NF-kappa B inhibition by omega -3 fatty acids modulates LPS-stimulated macrophage TNF-alpha transcription. Am J Physiol Lung Cell Mol Physiol, 2003. 284(1): p. L84-9. 172. Trapani, J.A., et al., Functional significance of the perforin/granzyme cell death pathway. Nat Rev Immunol, 2002. 2(10): p. 735-47. 173. Zamai, L., et al., Natural killer (NK) cell-mediated cytotoxicity: differential use of TRAIL and Fas ligand by immature and mature primary human NK cells. J Exp Med, 1998. 188(12): p. 2375-80. 174. Cooper, M.A., et al., NK cell and DC interactions. Trends Immunol, 2004. 25(1): p. 47-52. 175. Kared, H., et al., T cells and their cytokines in persistent stimulation of the immune system. Curr Opin Immunol, 2014. 29(0): p. 79-85. 176. Rama, I., et al., Hypoxia stimulus: An adaptive immune response during dendritic cell maturation. Kidney Int, 2008. 73(7): p. 816-825. 177. Majumdar, I., et al., The deubiquitinase A20 in immunopathology of autoimmune diseases. Autoimmunity, 2014. 47(5): p. 307-319. 178. Opipari, A.W., et al., The A20 cDNA induced by tumor necrosis factor alpha encodes a novel type of zinc finger protein. Journal of Biological Chemistry, 1990. 265(25): p. 14705-8. 179. Boone, D.L., et al., The ubiquitin-modifying enzyme A20 is required for termination of Toll-like receptor responses. Nat Immunol, 2004. 5(10): p. 1052- 1060. 180. Shembade, N., et al., Regulation of NF-κB signaling by the A20 deubiquitinase. Cell Mol Immunol, 2012. 9(2): p. 123-130. 181. Jeong, E., et al., Intrinsic and Extrinsic Regulation of Innate Immune Receptors. Yonsei Med J, 2011. 52(3): p. 379-392. 182. Lee, E.G., et al., Failure to regulate TNF-induced NF-κB and cell death responses in A20-deficient mice. Science, 2000. 289(5488): p. 2350-2354. 183. Liuwantara, D., et al., Nuclear Factor-κB regulates β-cell death: a critical role for A20 in β-cell protection. Diabetes, 2006. 55(9): p. 2491-2501. 184. Emamaullee, J.A., et al., Interventional strategies to prevent β-cell apoptosis in islet transplantation. Diabetes, 2006. 55(7): p. 1907-1914. 185. Stenflo, J., New Vitamin-K-dependent protein - purification from bovine plasma and preliminary characterization. Journal of Biological Chemistry, 1976. 251(2): p. 355-363. 186. Griffin, J.H., et al., Deficiency Of Protein-C In Congenital Thrombotic Disease. Journal of Clinical Investigation, 1981. 68(5): p. 1370-1373. 187. Griffin, J.H., et al., Activated protein C. Journal of Thrombosis and Haemostasis, 2007. 5(s1): p. 73-80. 188. Joyce, D.E., et al., Gene Expression Profile of Antithrombotic Protein C Defines New Mechanisms Modulating Inflammation and Apoptosis. The Journal of Biological Chemistry, 2001. 276(14): p. 11199-11203. 189. Esmon, C.T., Structure and functions of the endothelial cell protein C receptor. Critical Care Medicine, 2004. 32(5): p. S298-S301. 190. Jackson, C.J., et al., Activated protein C--An anticoagulant that does more than stop clots. The International Journal of Biochemistry & Cell Biology. In Press, Corrected Proof.

283 191. Xue, M., et al., Chapter Thirteen - Activated protein C and its potential applications in prevention of islet β-cell damage and diabetes. Vitamins & Hormones, Academic Press, 2014. 95: p. 323-363. 192. Mosnier, L.O., et al., The cytoprotective protein C pathway. Blood, 2007. 109(8): p. 3161-3172. 193. Montes, R., et al., Is EPCR a multi-ligand receptor? Pros and cons. Thrombosis and Haemostasis, 2012. 107(5): p. 815-826. 194. Esmon, C.T., Crosstalk between inflammation and thrombosis. Maturitas, 2004. 47(4): p. 305-314. 195. Hooper, W.C., et al., and M.A. Renshaw, Activated Protein C Induction of MCP1 in Human Endothelial Cells: A Possible Role for Endothelial Cell Nitric Oxide Synthase. Thrombosis Research, 2001. 103(3): p. 209-219. 196. Riewald, M., et al., Activation of Endothelial Cell Protease Activated Receptor 1 by the Protein C Pathway. Science, 2002. 296(5574): p. 1880-1882. 197. Xue, M., et al., Targeting matrix metalloproteases to improve cutaneous wound healing. Expert Opinion on Therapeutic Targets, 2006. 10(1): p. 143-155. 198. Isermann, B., et al., Activated protein C protects against diabetic nephropathy by inhibiting endothelial and podocyte apoptosis. Nat Med, 2007. 13(11): p. 1349-1358. 199. Elphick, G.F., et al., Recombinant human activated protein C inhibits integrin- mediated neutrophil migration. Blood, 2009. 113(17): 4078-4085. 200. Xue, M., et al., Activated protein C stimulates expression of angiogenic factors in human skin cells, angiogenesis in the chick embryo and cutaneous wound healing in rodents. Clinical Hemorheology and Microcirculation, 2006. 34(1): p. 153-161. 201. Xue, M., et al., Activated protein C mediates a healing phenotype in cultured tenocytes. Journal of Cellular and Molecular Medicine, 2009. 13(4): p. 749-757. 202. Chong, A.J., et al., Toll-like receptor 4 mediates ischemia/reperfusion injury of the heart. J Thorac Cardiovasc Surg, 2004. 128(2): p. 170-9. 203. Shimamoto, A., et al., Toll-like receptor 4 mediates lung ischemia-reperfusion injury. Ann Thorac Surg, 2006. 82(6): p. 2017-23. 204. Tsung, A., et al., Hepatic ischemia/reperfusion injury involves functional TLR4 signaling in nonparenchymal cells. J Immunol, 2005. 175(11): p. 7661-8. 205. Wu, H., et al., TLR4 activation mediates kidney ischemia/reperfusion injury. J Clin Invest, 2007. 117: p. 2847-2859. 206. Krüger, B., et al., Islet-expressed TLR2 and TLR4 sense injury and mediate early graft failure after transplantation. European Journal of Immunology, 2010. 40(10): p. 2914-2924. 207. Tsung, A., et al., The nuclear factor HMGB1 mediates hepatic injury after murine liver ischemia-reperfusion. J Exp Med, 2005. 201(7): p. 1135-43. 208. Chen, G.E., et al., Toll-like receptor 4 engagement contributes to expression of NKG2D ligands by renal tubular epithelial cells. Nephrol Dial Transplant, 2011. 26(12): p. 3873-81. 209. Izuishi, K., et al., Cutting edge: high-mobility group box 1 preconditioning protects against liver ischemia-reperfusion injury. J Immunol, 2006. 176(12): p. 7154-8. 210. Zipfel, A., et al., Endotoxemia in organ donors: graft function following liver transplantation. Transplant International, 2000. 13(1): p. S286-S287.

284 211. Wijdicks, E.F.M., et al., Evidence-based guideline update: Determining brain death in adults: Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology, 2010. 74(23): p. 1911-1918. 212. Novitzky, D., ey al., Electrocardiographic, hemodynamic & endocrine changes occurring during experimental brain death in the Chacma baboon. Journal of Heart Transplantation, 1984. 4(1): p. 63-69. 213. Takada, M., et al., Effects of explosive brain death on cytokine activation of peripheral organs in the rat1. Transplantation, 1998. 65(12): p. 1533-1542. 214. Novitzky, D., et al., Improved cardiac function following hormonal therapy in brain dead pigs: Relevance to organ donation. Cryobiology, 1987. 24(1): p. 1- 10. 215. Ranasinghe, A.M., et al., Endocrine changes in brain death and transplantation. Best Practice & Research Clinical Endocrinology & Metabolism, 2011. 25(5): p. 799-812. 216. Belzberg, H., et al., Hemodynamic and Oxygen Transport Patterns After Head Trauma and Brain Death: Implications for Management of the Organ Donor. Journal of Trauma and Acute Care Surgery, 2007. 63(5): p. 1032-1042 10.1097/01.ta.0000235995.86162.d2. 217. Herijgers, P., et al., Changes In Organ Perfusion After Brain Death In The Rat And Its Relation To Circulating Catecholamines1. Transplantation, 1996. 62(3): p. 330-335. 218. Cunningham, E.C., et al., Gene therapy for tolerance: high-level expression of donor major histocompatibility complex in the liver overcomes naive and memory alloresponses to skin grafts. Transplantation, 2013. 95(1): p. 70-7. 219. Borutaite, V., et al., In the eye of the storm: mitochondrial damage during heart and brain ischaemia. FEBS Journal, 2013. 280(20): p. 4999-5014. 220. Floerchinger, B., et al., Effects of brain death on organ quality and transplant outcome. Transplantation Reviews, 2012. 26(2): p. 54-59. 221. Novitzky, D., et al., Change from aerobic to anaerobic metabolism after brain death, and reversal following triiodothyronine therapy. Transplantation, 1988. 45(1): p. 32-36. 222. Rech, T.H., et al., Brain death-induced inflammatory activity in human pancreatic tissue: a case-control study. Transplantation. 2014. 97(2): p.212- 219. 223. Feuerstein, G.Z., et al., Barone, Inflammatory Gene Expression in Cerebral Ischemia and Trauma. Annals of the New York Academy of Sciences, 1997. 825(1): p. 179-193. 224. Van Der Hoeven, J.A.B., et al., Induction of organ dysfunction and up- regulation of inflammatory markers in the liver and kidneys of hypotensive brain dead rats: A model to study marginal organ donors. Transplantation, 1999. 68(12): p. 1884-1890. 225. Morariu, A.M., et al., Early Events in Kidney Donation: Progression of Endothelial Activation, Oxidative Stress and Tubular Injury After Brain Death. American Journal of Transplantation, 2008. 8(5): p. 933-941. 226. Kusaka, M., et al., Early and late inflammatory changes occurring in rat renal isografts from brain dead donors. Transplantation Proceedings, 2001. 33(1–2): p. 867-868. 227. Pratschke, J., et al., What can be learned from brain-death models? Transplant International, 2005. 18(1): p. 15-21.

285 228. Kusaka, M., et al., Activation of inflammatory mediators in rat renal isografts by donor brain death1. Transplantation, 2000. 69(3): p. 405-410. 229. Atkinson, C., et al., Complement-dependent inflammation and injury in a murine model of brain dead donor hearts. Circulation Research, 2009. 105(11): p. 1094-1101. 230. Stein, S., et al., Coagulopathy in traumatic brain injury. Neurocritical Care, 2004. 1(4): p. 479-488. 231. Hing, A.J., et al., The effects of hormone resuscitation on cardiac function and hemodynamics in a porcine brain-dead organ donor model. American Journal of Transplantation, 2007. 7(4): p. 809-817. 232. Sappington, P.L., et al., HMGB1 B box increases the permeability of Caco-2 enterocytic monolayers and impairs intestinal barrier function in mice. Gastroenterology, 2002. 123(3): p. 790-802. 233. Mantovani, A., et al., Extracellular and intracellular decoys in the tuning of inflammatory cytokines and Toll-like receptors: the new entry TIR8/SIGIRR. J Leukoc Biol, 2004. 75(5): p. 738-42. 234. Weiss, S., et al., Brain death activates donor organs and is associated with a worse I/R injury after liver transplantation. American Journal of Transplantation, 2007. 7(6): p. 1584-1593. 235. Bugge, J.F., Brain death and its implications for management of the potential organ donor. Acta Anaesthesiologica Scandinavica, 2009. 53(10): p. 1239- 1250. 236. Mangino, M.J., et al., The effects of donor brain death on renal function and arachidonic acid metabolism in a large animal model of hypothermic preservation injury. Transplantation, 2003. 75(10): p. 1640-1647. 237. McLean, K.M., et al., Glucocorticoids Alter the Balance Between Pro- and Anti-inflammatory Mediators in the Myocardium in a Porcine Model of Brain Death. The Journal of Heart and Lung Transplantation, 2007. 26(1): p. 78-84. 238. Barklin, A., et al., Does brain death induce a pro-inflammatory response at the organ level in a porcine model? Acta Anaesthesiologica Scandinavica, 2008. 52(5): p. 621-627. 239. Barklin, A., Systemic inflammation in the brain-dead organ donor. Acta Anaesthesiologica Scandinavica, 2009. 53(4): p. 425-435. 240. Doig, C.J., et al., Increased Intestinal Permeability Is Associated with the Development of Multiple Organ Dysfunction Syndrome in Critically Ill ICU Patients. American Journal of Respiratory and Critical Care Medicine, 1998. 158(2): p. 444-451. 241. Grotz, M.R.W., et al., Intestinal Cytokine Response After Gut Ischemia: Role of Gut Barrier Failure. Annals of Surgery, 1999. 229(4): p. 478-486. 242. Faries, P.L., et al., Intestinal Permeability Correlates with Severity of Injury in Trauma Patients. Journal of Trauma and Acute Care Surgery, 1998. 44(6): p. 1031-1036. 243. Gregory M., et al., Role of the gut in multiple organ failure: bacterial translocation and permeability changes. World Journal of Surgery, 1996. 20: p. 411-417. 244. Chun-Hua Hang, J.-X.S., Wei Wu, Hong-Xia Yin, Alterations of intestinal mucosa structure and barrier function following traumatic brain injury in rats. World Journal of Gasteroenterology, 2003.

286 245. Leber, B., et al., Effect of oxidative stress and endotoxin on human serum albumin in brain-dead organ donors. Translational Research, 2012. 159(6): p. 487-496. 246. Koudstaal, L.G., et al., Increased Intestinal Permeability in Deceased Brain Dead Rats. Transplantation, 2009. 88(3): p. 444-446. 247. Wang, C., Z. Ling, and D. Pipeleers, Comparison of cellular and medium insulin and GABA content as markers for living β-cells, 2005. p. E307. 248. Goldberg, A., et al., Toll-like receptor 4 suppression leads to islet allograft survival. The FASEB Journal, 2007. 21(11): p. 2840-2848. 249. Gao, Q., et al., TLR4 Mediates Early Graft Failure After Intraportal Islet Transplantation. American Journal of Transplantation, 2010. 10(7): p. 1588- 1596. 250. Matsuoka, N., et al., High-mobility group box 1 is involved in the initial events of early loss of transplanted islets in mice. The Journal of Clinical Investigation, 2010. 120(3): p. 735-743. 251. Itoh, T., et al., Correlation of Released HMGB1 Levels With the Degree of Islet Damage in Mice and Humans and With the Outcomes of Islet Transplantation in Mice. Cell Transplantation, 2012. 21(7): p. 1371-1381. 252. Ro, H., et al., Roles of Toll-Like Receptors in Allogeneic Islet Transplantation. Transplantation, 2012. 94(10): p. 1005-1012. 253. Terasaki, P.I., et al., High survival rates of kidney transplants from spousal and living unrelated donors. New England Journal of Medicine, 1995. 333(6): p. 333-336. 254. Sánchez-Fructuoso AI, P.D., et al., Does donor brain death influence acute vascular rejection in the kidney transplant? Transplantation, 2004. 78(2): p. 142-146. 255. Nijboer, W.N., et al., Effect of Brain Death on Gene Expression and Tissue Activation in Human Donor Kidneys. Transplantation, 2004. 78(7): p. 978-986. 256. Kim, Y.S., et al., Cadaveric renal allograft at the time of implantation has the similar immunological features with the rejecting allograft1. Transplantation, 2000. 70(7): p. 1080-1085. 257. Koo, D.D.H., et al., Cadaver versus living donor kidneys: Impact of donor factors on antigen induction before transplantation. Kidney International, 1999. 56(4): p. 1551-1559. 258. Birks, E.J., et al., Tumor Necrosis Factor-α Is Expressed in Donor Heart and Predicts Right Ventricular Failure After Human Heart Transplantation. Circulation, 2000. 102(3): p. 326-331. 259. Birks, E.J., et al., Elevated tumor necrosis factor-alpha and interleukin-6 in myocardium and serum of malfunctioning donor hearts. Circulation, 2000. 102(19): p. 352-358. 260. Pérez López, S., et al., Brain Death Effects on Catecholamine Levels and Subsequent Cardiac Damage Assessed in Organ Donors. The Journal of Heart and Lung Transplantation, 2009. 28(8): p. 815-820. 261. Cohen, O., et al., Donor Brain Death Mechanisms and Outcomes After Heart Transplantation. Transplantation Proceedings, 2007. 39(10): p. 2964-2969. 262. Mehra, M.R., et al., The impact of mode of donor brain death on cardiac allograft vasculopathy: An intravascular ultrasound study. Journal of the American College of Cardiology, 2004. 43(5): p. 806-810.

287 263. Jassem, W., et al., Leukocyte infiltration and inflammatory antigen expression in cadaveric and living-donor livers before transplant1. Transplantation, 2003. 75(12): p. 2001-2007. 264. Olinga, P., et al., The influence of brain death on liver function. Liver International, 2005. 25(1): p. 109-116. 265. Contreras, J.L., et al., Brain Death Significantly Reduces Isolated Pancreatic Islet Yields and Functionality In Vitro and In Vivo After Transplantation in Rats. Diabetes, 2003. 52(12): p. 2935-2942. 266. Saito, Y., et al., The Influence of Brain Death on Tissue Factor Expression in the Pancreatic Tissues and Isolated Islets in Rats. Transplantation Proceedings, 2009. 41(1): p. 307-310. 267. Moberg, L., et al., Nicotinamide inhibits tissue factor expression in isolated human pancreatic islets: implications for clinical islet transplantation1. Transplantation, 2003. 76(9): p. 1285-1288. 268. Jung, H.S., et al., A better yield of islet cell mass from living pancreatic donors compared with cadaveric donors. Clinical Transplantation, 2007. 21(6): p. 738- 743. 269. Cowley, M.J., et al., Human Islets Express a Marked Proinflammatory Molecular Signature Prior to Transplantation. Cell Transplantation, 2012. 21(9): p. 2063-2078. 270. Novitzky, D., et al., Hormonal Therapy of the Brain-Dead Organ Donor: Experimental and Clinical Studies. Transplantation, 2006. 82(11): p. 1396- 1401. 271. Darby, J.M., et al., Approach to management of the heartbeating brain dead organ donor. JAMA, 1989. 261(15): p. 2222-2228. 272. Cooper, D.K., Orthotopic and heterotopic transplantation of the heart: the Cape Town experience. Annals of The Royal College of Surgeon of England, 1984. 66(July): p. 228-234. 273. Novitzky, D., et al., Hemodynamic and metabolic responses to hormonal therapy in brain-dead potential organ donors. Transplantation, 1987. 43(6): p. 852-854. 274. Rosendale, J.D., et al., Aggressive pharmacologic donor management results in more transplanted organs1. Transplantation, 2003. 75(4): p. 482-487. 275. Wan, S., et al., Steroid administration in heart and heart—lung transplantation: Is the timing adequate? The Annals of Thoracic Surgery, 1996. 61(2): p. 674- 678. 276. Busuttil, R.W., et al.,Protective effect of methylprednisolone on heart during ischemic arrest. Journal of Thoracic and Cardiovascular Surgery, 1975. 70(6): p. 955-965. 277. Pratschke, J., et al., Improvements in Early Behavior of Rat Kidney Allografts After Treatment of the Brain-Dead Donor. Annals of Surgery, 2001. 234(6): p. 732-740. 278. Rosendale, J.D., et al., Hormonal resuscitation yields more transplanted hearts, with improved early function1. Transplantation, 2003. 75(8): p. 1336-1341. 279. McKeown, D.W., R.S. Bonser, and J.A. Kellum, Management of the heartbeating brain-dead organ donor. British Journal of Anaesthesia, 2012. 108(suppl 1): p. i96-i107. 280. Danobeitia, J.S., et al., Early Activation of the Inflammatory Response in the Liver of Brain-Dead Non-Human Primates. Journal of Surgical Research, 2012. 176(2): p. 639-648.

288 281. Danobeitia, J.S., et al., Donor Pre-treatment with IL-1 Receptor Antagonist Attenuates Inflammation and Improves Functional Potency in Islets from Brain- Dead Non-human Primates. Cell Transplantataion, 2014. 282. Aiello, S., et al., Rabbit anti-rat thymocyte immunoglobulin preserves renal function during ischemia/reperfusion injury in rat kidney transplantation. Transplant International, 2011. 24(8): p. 829-838. 283. Cicora, F., et al., Evaluation of histological damage of solid organs after donor preconditioning with thymoglobulin in an experimental rat model. Transplant Immunology, 2013. 28(4): p. 203-205. 284. Cicora, F., et al., Amelioration of renal damage by administration of anti- thymocyte globulin to potential donors in a brain death rat model. Clinical and Experimental Immunology, 2012. 169(3): p. 330-337. 285. Sako, D., et al., A sulfated peptide segment at the amino terminus of PSGL-1 is critical for P-selectin binding. Cell, 1995. 83(2): p. 323-331. 286. Hayward, R., et al., Recombinant soluble P-selectin glycoprotein ligand-1 protects against myocardial ischemic reperfusion injury in cats. Cardiovascular Research, 1999. 41(1): p. 65-76. 287. Takada, M., et al., Prevention of late renal changes after initial ischemia/reperfusion injury by blocking early selectin binding1. Transplantation, 1997. 64(11): p. 1520-1525. 288. Gasser, M., et al., Normalization of brain death—induced injury to rat renal allografts by recombinant soluble p-selectin glycoprotein ligand. Journal of the American Society of Nephrology, 2002. 13(7): p. 1937-1945. 289. Kusaka, M., et al., Exploitation of the continuum between early ischemia/reperfusion injury and host alloresponsiveness: indefinite kidney allograft survival by treatment with a soluble p-selectin ligand and low-dose cyclosporine in combination1. Transplantation, 1999. 67(9): p. 1255-1261. 290. Johansson, U., et al., Inflammatory mediators expressed in human islets of Langerhans: implications for islet transplantation. Biochemical and Biophysical Research Communications, 2003. 308(3): p. 474-479. 291. Chng, H.W., et al., A new role for the anti-apoptotic gene A20 in angiogenesis. Experimental Cell Research, 2006. 312(15): p. 2897-2907. 292. Öllinger, R., et al., Role of heme oxygenase-1 in transplantation. Transplant International, 2010. 23(11): p. 1071-1081. 293. Ahn, Y., et al., Changes in gene expression in beta cells after islet isolation and transplantation using laser-capture microdissection. Diabetologia, 2007. 50(2): p. 334-342. 294. Li, Y.X., et al., Protection of human islets from induction of apoptosis and improved islet function with HO-1 gene transduction. Chinese Medical Journal, 2006. 119(19): p. 1639-1645. 295. Pileggi, A., et al., Prolonged allogeneic islet graft survival by protoporphyrins. Cell Transplantation, 2005. 14(2-3): p. 85-96. 296. Chen, X., et al., Protective Effect of Heme Oxygenase-1 to Pancreas Islet Xenograft. Journal of Surgical Research, 2010. 164(2): p. 336-343. 297. Lee, S.S., et al., Heme oxygenase-1, carbon monoxide, and bilirubin induce tolerance in recipients toward islet allografts by modulating T regulatory cells. FASEB Journal, 2007. 21(13): p. 3450-3457. 298. Fu, Y.Y., et al., Hyperbilirubinemia reduces the streptozotocin-induced pancreatic damage through attenuating the oxidative stress in the gunn Rat. Tohoku Journal of Experimental Medicine, 2010. 222(4): p. 265-273.

289 299. Londrigan, S.L., et al., Optimizing transduction of pig islet cell clusters for xenotransplantation. Xenotransplantation, 2009. 16(1): p. 45-46. 300. McCall, M., et al., Anakinra Potentiates the Protective Effects of Etanercept in Transplantation of Marginal Mass Human Islets in Immunodeficient Mice. American Journal of Transplantation, 2012. 12(2): p. 322-329. 301. SoRelle, J.A., et al., Withaferin A inhibits pro-inflammatory cytokine-induced damage to islets in culture and following transplantation. Diabetologia, 2013. 56(4): p. 814-824. 302. Cheng, E.Y., et al., Diannexin decreases inflammatory cell infiltration into the islet graft, reduces β-cell apoptosis, and improves early graft function. Transplantation, 2010. 90(7): p. 709-716. 303. Cure, P., et al., Improved metabolic control and quality of life in seven patients with type 1 diabetes following islet after kidney transplantation. Transplantation, 2008. 85(6): p. 801-812. 304. Wang, H., et al., Donor treatment with carbon monoxide can yield islet allograft survival and tolerance. Diabetes, 2005. 54(5): p. 1400-1406. 305. Akima, S., et al., Tirofiban and Activated Protein C Synergistically Inhibit the Instant Blood Mediated Inflammatory Reaction (IBMIR) from Allogeneic Islet Cells Exposure to Human Blood. American Journal of Transplantation, 2009. 9(7): p. 1533-1540. 306. Contreras, J.L., et al., Activated Protein C Preserves Functional Islet Mass After Intraportal Transplantation: A Novel Link Between Endothelial Cell Activation, Thrombosis, Inflammation, and Islet Cell Death. Diabetes, 2004. 53(11): p. 2804-2814. 307. Cui, W., et al., Thrombomodulin Improves Early Outcomes After Intraportal Islet Transplantation. American Journal of Transplantation, 2009. 9(6): p. 1308- 1316. 308. Xue, M., et al., Activated Protein C Inhibits Pancreatic Islet Inflammation, Stimulates T Regulatory Cells, and Prevents Diabetes in Non-obese Diabetic (NOD) Mice. Journal of Biological Chemistry, 2012. 287(20): p. 16356-16364. 309. Gruden, G., et al., Activated protein C resistance in type I diabetes. Diabetes Care, 1997. 20(3): p. 424-425. 310. Bucciarelli, L.G., et al., RAGE blockade stabilizes established atherosclerosis in diabetic apolipoprotein E-null mice. Circulation, 2002. 106(22): p. 2827-35. 311. Shoji, T., et al., Receptor for Advanced Glycation End Products Is Involved in Impaired Angiogenic Response in Diabetes. Diabetes, 2006. 55(8): p. 2245- 2255. 312. Zeng, S., et al., Blockade of receptor for advanced glycation end product (RAGE) attenuates ischemia and reperfusion injury to the liver in mice. Hepatology, 2004. 39(2): p. 422-32. 313. Sternberg, D.I., et al., Blockade of receptor for advanced glycation end product attenuates pulmonary reperfusion injury in mice. J Thorac Cardiovasc Surg, 2008. 136(6): p. 1576-1585. 314. Moser, B., et al., Blockade of RAGE suppresses alloimmune reactions in vitro and delays allograft rejection in murine heart transplantation. American Journal of Transplantation, 2007. 7(2): p. 293-302. 315. Chen, Y., et al., RAGE ligation affects T cell activation and controls T cell differentiation. Journal of Immunology, 2008. 181(6): p. 4272-4278.

290 316. Korbutt, G.S., et al., Large Scale Isolation, Growth, and Function of Porcine Neonatal Islet Cells. Journal of Clinical Investigation, 1996. 97(9): p. 2119- 2129. 317. Xue, M., et al., Endothelial protein C receptor and protease-activated receptor- 1 mediate induction of a wound-healing phenotype in human keratinocytes by activated protein C. J Investig Dermatol, 2005. 125(6): p. 1279-1285. 318. Jackson, C.J., et al., Activated protein C prevents inflammation yet stimulates angiogenesis to promote cutaneous wound healing. Wound Repair and Regeneration, 2005. 13(3): p. 284-294. 319. Borissova, A.M., et al., The effect of vitamin D3 on insulin secretion and peripheral insulin sensitivity in type 2 diabetic patients. Int. J. Clin. Pract., 2003. 57: p. 258-261. 320. Dastgheib, S., et al., Optimization of RNA Extraction from Rat Pancreatic Tissue. Iranian Journal of Medical Sciences, 2014. 39: p. 282-288. 321. Hsiao, L.L., et al., A compendium of gene expression in normal human tissues. Physiological Genomics, 2001. 7(2): p. 97-104. 322. Yamada, H., et al., Effects of Fasting on the Expression of Gastrin, Cholecystokinin, and Somatostatin Genes and of Various Housekeeping Genes in the Pancreas and Upper Digestive Tract of Rats. Biochemical and Biophysical Research Communications, 1997. 231(3): p. 835-838. 323. Warrington, J.A., et al., Comparison of human adult and fetal expression and identification of 535 housekeeping/maintenance genes. Physiological Genomics, 2000. 2(3): p. 143-147. 324. Barber, R.D., et al., GAPDH as a housekeeping gene: analysis of GAPDH mRNA expression in a panel of 72 human tissues, 2005. p. 389-395. 325. Dufrane, D. and P. Gianello, Pig Islet Xenotransplantation Into Non-human Primate Model. Transplantation, 2008. 86(6): p. 753-760. 326. Rorsman, P., et al., Regulation of Insulin Secretion in Human Pancreatic Islets. Annual Reviews of Physiology, 2013. 75(1): p. 155-179. 327. Komatsu, M., et al., Glucose-stimulated insulin secretion: A newer perspective. Journal of Diabetes Investigation, 2013. 4(6): p. 511-516. 328. Itoh, T., et al., HMGB1-Mediated Early Loss of Transplanted Islets Is Prevented by Anti-IL-6R Antibody in Mice. Pancreas, 2014. Publish Ahead of Print. 329. Hårdstedt, M., et al., Post-transplant upregulation of chemokine messenger RNA in non-human primate recipients of intraportal pig islet xenografts. Xenotransplantation, 2005. 12(4): p. 293-302. 330. Beg, A.A., Endogenous ligands of Toll-like receptors: implications for regulating inflammatory and immune responses. Trends in Immunology, 2002. 23(11): p. 509-512. 331. Parks, W.C., et al., Matrix metalloproteinases as modulators of inflammation and innate immunity. Nat Rev Immunol, 2004. 4(8): p. 617-629. 332. Strongin, A.Y., et al., Mechanism of cell surface activation of 72-kda type iv collagenase: isolation of the activated form of the membrane metalloprotease. Journal of Biological Chemistry, 1995. 270(10): p. 5331-5338. 333. Overall, C.M., et al., Transcriptional and post-transcriptional regulation of 72- kDa gelatinase/type IV collagenase by transforming growth factor-beta 1 in human fibroblasts. Comparisons with collagenase and tissue inhibitor of matrix metalloproteinase gene expression. Journal of Biological Chemistry, 1991. 266(21): p. 14064-71.

291 334. Nguyen, M., et al., Human endothelial gelatinases and angiogenesis. The International Journal of Biochemistry & Cell Biology, 2001. 33(10): p. 960- 970. 335. Tran, P.D., et al., Porcine cells express more than one functional ligand for the human lymphocyte activating receptor NKG2D. Xenotransplantation, 2008. 15(5): p. 321-332. 336. Coughlin, S.R., Thrombin signalling and protease-activated receptors. Nature, 2000. 407(6801): p. 258-264. 337. Li, W., et al., Extraembryonic expression of EPCR is essential for embryonic viability. Blood, 2005. 106(8): p. 2716-2722. 338. Li, W., et al., Overexpressing endothelial cell protein C receptor alters the hemostatic balance and protects mice from endotoxin. Journal of Thrombosis and Haemostasis, 2005. 3(7): p. 1351-1359. 339. Gu, J.-M., et al., Endotoxin and thrombin elevate rodent endothelial cell protein C receptor mRNA levels and increase receptor shedding in vivo. Blood, 2000. 95(5): p. 1687-1693. 340. Xue, M., et al., Differential regulation of matrix metalloproteinase 2 and matrix metalloproteinase 9 by activated protein C: Relevance to inflammation in rheumatoid arthritis. Arthritis and Rheumatology, 2007. 56(9): p. 2864-2874. 341. Barth, E., et al., Glucose metabolism and catecholamines. Critical Care Medicine, 2007. 35(9): p. S508-S518. 342. Kikawa, K., et al., Beneficial effect of insulin treatment on islet transplantation outcomes in akita mice. PLoS ONE, 2014. 9(4): p. e95451. 343. Garrity, M.M., et al., Identifying and Quantifying Apoptosis: Navigating Technical Pitfalls. Mod Pathol, 2003. 16(4): p. 389-394. 344. Singh, S.S., et al., Comparison of ACINUS, caspase-3, and TUNEL as apoptotic markers in determination of tumor growth rates of clinically localized prostate cancer using image analysis. The Prostate, 2009. 69(15): p. 1603-1610. 345. Pulkkanen, K.J., et al., False-positive apoptosis signal in mouse kidney and liver detected with TUNEL assay. Apoptosis, 2000. 5(4): p. 329-333. 346. Le Tulzo, Y., et al., Hemorrhage increases cytokine expression in lung mononuclear cells in mice: involvement of catecholamines in nuclear factor- kappaB regulation and cytokine expression. The Journal of Clinical Investigation, 1997. 99(7): p. 1516-1524. 347. Sancho-Bru, P., et al., Norepinephrine induces calcium spikes and proinflammatory actions in human hepatic stellate cells. Gastrointestinal and Liver Physiology, 2006. 291(5): p.G877-G884. 348. Dupuis, S., et al., Corticosteroids in the management of brain-dead potential organ donors: a systematic review. British Journal of Anaesthesia, 2014. 349. Barnes, P.J., How corticosteroids control inflammation: Quintiles Prize Lecture 2005. British Journal of Pharmacology, 2006. 148(3): p. 245-254. 350. Altonsy, M.O., et al., Context-dependent cooperation between nuclear factor-κb and the glucocorticoid receptor at a tnfaip3 intronic enhancer: a mechanism to maintain negative feedback control of inflammation. Journal of Biological Chemistry, 2014. 289(12): p. 8231-8239. 351. Wu, H., et al., Roles of Toll-like receptors in transplantation. Current Opinion in Organ Transplantation, 2014. 19(1): p. 1-7. 352. Ogasawara, K., et al., NKG2D blockade prevents autoimmune diabetes in NOD mice. Immunity, 2004. 20(6): p. 757-767.

292 353. Hamerman, J.A., et al., Cutting Edge: Toll-Like Receptor Signaling in Macrophages Induces Ligands for the NKG2D Receptor. The Journal of Immunology, 2004. 172(4): p. 2001-2005. 354. Mats KEB Wallin, E.S., et al., Amino acid infusion during anesthesia attenuates the surgery induced decline in IGF-1 and diminishes the "diabetes of injury". Nutrition and metabolism, 2007. 4. 355. Bernsmeier, C., et al., Hepatitis C virus dysregulates glucose homeostasis by a dual mechanism involving induction of PGC1α and dephosphorylation of FoxO1. Journal of Viral Hepatitis, 2014. 21(1): p. 9-18. 356. Pope, M.R., et al., Complement regulates TLR4-mediated inflammatory responses during intestinal ischemia reperfusion. Molecular Immunology, 2010. 48(1–3): p. 356-364. 357. Damman, J., et al., Local renal complement C3 induction by donor brain death is associated with reduced renal allograft function after transplantation. Nephrology Dialysis Transplantation, 2011. 26(7): p. 2345-2354. 358. Jung, H., et al., RAE-1 ligands for the NKG2D receptor are regulated by E2F transcription factors, which control cell cycle entry. The Journal of Experimental Medicine, 2012. 209(13): p. 2409-2422. 359. Riuzzi, F., et al., The amphoterin (HMGB1)/receptor for advanced glycation end products (RAGE) pair modulates myoblast proliferation, apoptosis, adhesiveness, migration, and invasiveness: functional inactivation of rage in l6 myoblasts results in tumor formation in vivo. Journal of Biological Chemistry, 2006. 281(12): p. 8242-8253. 360. Mishra, R., et al., Inflammatory cytokine-mediated evasion of virus-induced tumors from nk cell control, 2013, Journal of Immunology. p. 961-970. 361. Pratschke, J., et al., Acute rejection of rat renal allografts is accelerated by donor brain death. Transplantation Proceedings, 1999. 31(1–2): p. 874-875. 362. Pratschke, J., et al., Influence of donor brain death on chronic rejection of renal transplants in rats. J Am Soc Nephrol, 2001. 12(11): p. 2474-81. 363. Moberg, L., et al., Neutrophilic granulocytes are the predominant cell type infiltrating pancreatic islets in contact with ABO-compatible blood. Clinical & Experimental Immunology, 2005. 142(1): p. 125-131. 364. Muller, S., et al., The double life of HMGB1 chromatin protein: architectural factor and extracellular signal. EMBO J, 2001. 20(16): p. 4337-4340. 365. Bensellam, M., et al., The molecular mechanisms of pancreatic β-cell glucotoxicity: Recent findings and future research directions. Molecular and Cellular Endocrinology, 2012. 364(1–2): p. 1-27. 366. Sweet, I.R., et al., Glucose-stimulated increment in oxygen consumption rate as a standardized test of human islet quality. American Journal of Transplantation, 2008. 8(1): p. 183-192. 367. Bezinover, D., et al., Release of cytokines and hemodynamic instability during the reperfusion of a liver graft. Liver Transplantation, 2011. 17(3): p. 324-330. 368. Arshad, M.I., et al., IL-33 and HMGB1 alarmins: sensors of cellular death and their involvement in liver pathology. Liver International, 2012. 32(8): p. 1200- 1210.

293