Basic and clinical science posters: Beta cells, islets and stem cells

The effect of angiotensin 1–7 on islet function

A Al-Romaiyan

Pharmacology & Therapeutics, Kuwait University, Kuwait

Background Angiotensin 1–7 (ang 1–7) is a product of the renin–angiotensin system generated through the action of angiotensin-converting enzyme-2 (ACE-2). Recently, ACE-2/ang 1–7 and its G-protein coupled receptor (MAS) have been shown to be expressed in beta cell islets of Langerhans. Thus, the ang 1–7/MAS pathway may be involved in modulating islet function.

Aims The aim of this study was to investigate the role of ang 1–7/MAS on islet function.

Methods Using static secretion experiments, isolated mouse islets were used to test the effects of ang 1–7 on insulin secretion and to investigate whether the stimulatory effect of glucose on insulin secretion is dependent on MAS receptor activation.

Results Ang 1–7 [125nM–10μM] caused a concentration dependent increase in insulin secretion from mouse islets at 2mM glucose (2mM glucose: 1.4 ± 0.43ng/5 islets/h, +10μM ang 1–7: 6.04 ± 1.11, p < 0.05, n=4–6). Ang 1–7-induced insulin secretion was not inhibited by the presence of A779, a MAS receptor blocker (2.5μM ang 1–7: 0.86 ± 0.06ng/5 islets/h, +10μM A779: 0.91 ± 0.05, p > 0.05, n=6). Treating mouse islets with 20mM glucose resulted in an augmentation in insulin secretion which was significantly reduced by blocking MAS receptor (20mM glucose: 9.3 ± 0.6ng/5 islets/h, +A779: 5.7 ± 1.3, p < 0.05, n=6).

Conclusions These data indicate that the ang 1–7 and its receptor are important for islet function. Ang 1–7 stimulates insulin secretion from mouse islets in vitro independently of MAS activation. The glucose-induced insulin secretion is partly dependent on MAS activation.

Evidence that signal regulatory protein alpha (SIRPA) and CD47 may contribute to the improvement in beta cell viability mediated by interleukin-13

KA Leslie, MA Russell, SJ Richardson, NG Morgan

Biomedical and Clinical Sciences, University of Exeter, Exeter, UK

Introduction The viability of pancreatic beta cells is improved during exposure to interleukin (IL)-13 by a mechanism involving changes in the expression of target genes controlled by the transcription factor STAT6. One such gene encodes signal regulatory protein alpha (SIRPα), a protein not previously recognised as a regulator of beta cell viability. To understand the role of SIRPα more fully, we studied its effects on beta cell viability.

Materials and Methods Rodent INS-1E and human EndoC-βH1 cells were employed. Cell viability was monitored by a flow cytometer after propidium iodide staining, while qRT-PCR and western blotting were used to study gene and protein expression respectively. Knockdown of target genes was achieved with small interference RNAs. The overexpression involved transfection with a vector encoding SIRPα.

Results IL-13 improved the viability of rodent and human beta cells during exposure to pro-inflammatory cytokines (EndoC-βH1; cytokines 74.2 ± 1.4%, cytokines+IL-13 63.1 ± 1.8; p < 0.001). RNA analysis revealed a strong upregulation of both SIRPα and CD47 under these conditions. Knockdown of SIRPa or CD47 directly reduced beta cell viability (p < 0.001) both in the presence and absence of pro-inflammatory cytokines. Conversely, the overexpression of SIRPα resulted in improved cell viability and this correlated with increased CD47 expression. Both SIRPα and CD47 were seen to be expressed in human islets from non-diabetic subjects but immunofluorescence analysis revealed that CD47 is depleted in type 1 diabetes.

Conclusion We propose that the interaction between SIRPαα and CD47 may play a significant role in the regulation of beta cell viability and that this interaction is impaired in type 1 diabetes.

Acknowledgement IBEX

The chemokine CCL25-CCR9 axis impairs beta cell function: Potential for therapeutic intervention in type 2 diabetes

P Atanes, GC Huang, P Choudhary, SJ Persaud

Department of Diabetes, King's College London, London, UK

Aims To identify alterations in G protein-coupled receptor (GPCR) mRNA expression in islets isolated from donors with normal and elevated BMI and determine the effects of activation of the highly up-regulated chemokine receptor, CCR9, on islet function and viability.

Methods GPCR mRNA expression of human islets retrieved from organ donors with normal (22.7 ± 1.3, n=3) and elevated (34.6 ± 1.3, n=3) BMI was quantified by qPCR. Effects of the CCR9 agonist (CCL25) and antagonist (vercirnon) on glucose-stimulated insulin secretion (GSIS) from human (HI) and mouse islets (MI) were measured by radioimmunoassay. Apoptosis and cAMP accumulation were assessed by caspase 3/7 quantification and fluorescence assay.

Results qPCR analysis showed significantly altered expression of several islet GPCRs between obese and lean donors, with CCR9 showing 92 ± 17-fold upregulation in islets from high BMI donors. CCL25 inhibited GSIS from HI in a concentration-dependent manner (20mM glucose: 1.98 ± 0.08ng/islet/h; +5nM CCL25: 2.02 ± 0.14; +50nM: 1.58 ± 0.08; +500nM: 1.40 ± 0.08; p < 0.0001). Vercirnon blocked inhibition of GSIS by CCL25, without having any effect on its own (20mM glucose: 1.99 ± 0.22; +50nM CCL25: 1.36 ± 0.10; +50nM CCL25/10nM vercirnon: 2.08 ± 0.06; +10nM vercirnon: 1.87 ± 0.16; p < 0.0001). CCL25 enhanced cytokine-induced HI apoptosis (no-cytokines: 104,782 ± 12,991 relative light units (RLU); +cytokines: 207,813 ± 19,627; +50nM CCL25: 275,881 ± 18,438; p < 0.001) and reduced FSK-stimulated cAMP accumulation (basal: 7.5 ± 0.66nM; 1μM FSK: 203.2 ± 6.27; +50nM CCL25: 178.3 ± 5.14; p < 0.0001). Similar effects were observed in MI.

Conclusions CCL25 acts via the Gi-coupled receptor CCR9 to impair beta cell function by inhibiting GSIS, reducing cAMP accumulation and promoting cytokine-induced apoptosis. These observations suggest that CCL25 downregulation or CCR9 inhibition could be explored to treat type 2 diabetes.

Acknowledgement Diabetes Research Group

The novel adipokine WISP1 inhibits insulin release from clonal MIN6 beta cells

OS Idiakheua¹, PK Patel¹, BM Singh^1,2, SJ Dunmore¹

¹Wolverhampton Diabetes Research Group, SMCP, University of Wolverhampton, Wolverhampton, UK, ²Diabetes Centre, New Cross Hospital, Royal Wolverhampton NHS Trust, Wolverhampton, UK

Aims/Background WISP1 is a novel adipokine which has pro-inflammatory effects and is increased in obesity and some forms of diabetes. WISP1 induces insulin resistance in skeletal muscle cells and hepatocytes with effects on the insulin signalling pathway. Since insulin receptor and pathways are expressed in beta cells, affecting their function, we hypothesised that WISP1 may affect insulin release.

Methods Cultured MIN6 cells were stabilised to basal insulin release in 0.2mmol/l glucose and then incubated for 1 h in KRH-BSA in the presence of low (2mmol/l) or high glucose (20mmol/l) with a range of WISP1 concentrations found in lean and obese subjects (0,0.1, 0.5 and 1.0ng/ml). Insulin release was measured using a mouse insulin ELISA-based assay (Mercodia, Uppsala, Sweden).

Results WISP1 dose-dependently inhibited insulin release from MIN6 cells at both high and low glucose concentrations. This was significant at 0.5 and 1.0ng/ml at both high and low glucose concentrations (n=3, p < 0.01).

Conclusion We have adduced evidence that, in addition to its effects of inducing insulin resistance, the novel adipokine WISP1 inhibits insulin release from beta cells at concentrations found in obesity supporting the suggestion that it may play a significant role in the development of type 2 diabetes. We are studying the effects of the adipokine on the phosphorylation status of components of the insulin signalling pathway and on the proliferation of beta cells.

Pleiotrophin is expressed in mouse and human beta cells and can increase DNA synthesis and expression of key genes controlling mature beta cell phenotype

JS Fernández¹, BS Strutt², A Liang³, MP Ramos¹, DJ Hill^2,3

¹Facultad de Farmacia, Universidad San Pablo, Madrid, Spain, ²Lawson Health Research Institute, London, Ontario, Canada, ³Department of Physiology and Pharmacology, Western University, London, Ontario, Canada

Aims Pleiotrophin (PTN), a heparin-binding growth factor, is an epithelial and endothelial cell mitogen whose expression increases in regenerating tissues. 15-month-old Ptn^−/− mice show decreased insulin sensitivity. The study objectives were to determine if PTN is expressed in mouse and human pancreas and its actions on beta cell DNA synthesis, gene expression and glucose-stimulated insulin secretion.

Methods Immunohistochemistry for PTN and insulin was performed on mouse pancreas sections (21–120 days) or human pancreata. Ins1E beta cells were exposed to PTN (24–48 h) before insulin and Ki67 (indicator of DNA synthesis) immunohistochemistry. Messenger RNA expression for insulin, Glut2, Pdx1 and the PTN receptor, RPTPbz was quantified using qPCR. Ins1E cells were exposed to 2.8mM or 28.8mM glucose with or without 0.1mg/ml PTN and insulin release measured using radioimmunoassay (n=4–6).

Results Immunohistochemistry revealed co-localisation of PTN and insulin within beta cells, although in human the number co-expressing PTN decreased in later life. The percent Ins1E cells co-labelled with Ki67 increased significantly following incubation with PTN (24 h, control 74 ± 3%, 1μg/ml PTN 85 ± 2%, p < 0.05). Exposure to 1μg/ml PTN significantly increased the relative mRNA expression for insulin (55% vs control), Pdx1 (80%) and Glut2 (60%). Expression of RPTPbz was unaltered. Insulin release was not significantly altered by PTN in the presence of either low or high glucose.

Summary Our findings reveal that PTN is expressed within mouse and human beta cells, and can increase DNA synthesis and the expression of key genes controlling mature beta cell phenotype, but not GSIS.

Pancreatic islets display sexual dimorphism in insulin secretion in response to liver-expressed antimicrobial peptide-2

NV Hewawasam¹, A King², M Patterson¹, S Reeves¹, AC Hauge-Evans¹

¹Department of Life Sciences, University of Roehampton, London, UK, ²Department of Diabetes, King's College London, London, UK

Aims The prevalence of type 2 diabetes, beta cell function as well as dysfunction are influenced by sex. We have found that treatment with exogenous ghrelin significantly reduced cytokine-induced apoptosis in islets from female mice, but not from male mice. Reportedly, ghrelin acts via the G-protein coupled receptor growth hormone secretagogue receptor type 1A (GHSR1a). Liver-expressed antimicrobial peptide-2, LEAP-2, has recently been characterised as a GHSR1a antagonist. The aim of this study was to investigate whether the action of ghrelin on islet function is modified by LEAP-2 and whether the response to ghrelin and LEAP-2 is subject to sex-related differences.

Methods Apoptosis was detected by measurement of caspase 3/7 activity. Static insulin secretion experiments were conducted and insulin content was measured by radioimmunoassay.

Results Treatment with 100nM LEAP-2 did not significantly affect cytokine-induced apoptosis in islets from both male and female mice and it did not significantly reverse the protective effect of ghrelin observed in islets from females (p=0.27, n=6). In our study, we did not observe an insulinostatic effect of ghrelin in islets from both groups (p=0.99 males and p=0.76 females, n=5). However, in islets from male mice treatment with 100nM LEAP-2 resulted in significantly higher glucose-induced insulin secretion compared to 20mM glucose (p=0.001, n=5–6) but not in female (p=0.49, n=5) and this was significantly reduced in the presence of ghrelin (p=0.0016, n=5–6).

Conclusion Treatment with LEAP-2 did not significantly reverse the protective effect of ghrelin in islets from female mice. However, it significantly increased glucose-induced insulin secretion in islets from male, but not from female mice, suggesting a sexually dimorphic effect of LEAP-2 in islet function.

The impact of TNFAIP3-driven pathways on beta cell survival and function

WW Ratajczak, C Kelly, S Atkinson

Northern Ireland Centre for Stratified Medicine, Ulster University, Londonderry, UK

The key deliverable of this study is to identify TNFAIP3-driven pathways regulating beta cell survival and function.

TNFAIP3 was induced in the BRIN-BD11 beta cell line via the addition of recombinant TNFα (10ng/ml) for 0–24 h. Peak induction of TNFAIP3 at both the mRNA and protein level was detected after 1 h and this timepoint was used for all experiments. To test how the presence/absence of TNFAIP3 affected inflammatory and apoptotic pathways, siRNA targeting TNFAIP3 was used to silence the gene in BRIN-BD11 cells. The impact on inflammatory and apoptotic pathways was assessed using RT² PCR arrays and the outcomes modelled using IPA software. The array results were then validated with independent probes and human islet sequencing data available from HIRN datasets. The effect of TNFAIP3 knockdown was assessed with glucose stimulated insulin secretion and TUNEL assays.

Knockdown of TNFAIP3 in the region of 60% was achieved. PCR arrays identified 15 genes, which displayed significant (>2-fold change p < 0.05) differences in expression between control and TNFAIP3 silenced cells. Validation refined the genes of interest. IPA modelling of the interactive networks regulated by these genes show clear links to the NF-kB pathway, TWEAK and TNFR1 signalling and protein ubiquitination pathways. Silencing of TNFAIP3 reduced glucose-stimulated insulin secretion by approximately 50% and increased apoptosis by over 90% when compared to the positive control (p < 0.0001).

TNFAIP3 plays a major role in insulin secretion and control of apoptosis in BRIN-BD11 beta cell model. Identification of variants in TNFAIP3 in people with type 1 diabetes may prove a useful prognostic tool.

Potential effects of the coffee compounds cafestol and trigonelline on pancreatic beta cell survival and function in type 2 diabetes

K Sarnsamak, A Costabile, AC Hauge-Evans

Health Science Research Centre, Life Science Department, University of Roehampton, London, UK

Aims Cafestol (CFT) and trigonelline (TGL) from coffee have been reported to provide anti-hyperglyceamic effects in in vivo studies. However, it is unclear whether the compounds directly affect pancreatic beta cell function. This study investigated the role of CFT and TGL in the regulation of insulin secretion and in beta cell survival following exposure to glucolipotoxicity, a cellular stress factor typical of type 2 diabetes.

Methods INS1 beta cells were incubated with 1, 3 and 5μmol/l CFT or TGL, respectively, with or without 20mmol/l glucose + 0.25mmol/l palmitate for 20 h. Cell viability was assessed by measurement of adenosine triphosphate (ATP) content. Additionally, insulin release following 1 h static incubations with or without CFT and TGL was determined by a radioimmunoassay.

Results INS1 cell viability significantly decreased after exposure to glucolipotoxicity (p < 0.05) and this detrimental effect was not reversed by co-treatment with CFT or TGL alone or in combination (p > 0.05 vs 20mmol/l glucose + 0.25mmol/l palmitate only, n=3 separate experiments). Furthermore, CFT or TGL alone did not modulate insulin release in beta cells induced by 20mmol/l glucose, 10μmol/l forskolin and 100μmol/l 3-isobutyl-1-methylxanthine (p > 0.05). Nevertheless, a significant increase was observed in response to CFT and TGL combined at 1 (1.95 ± 0.28ng insulin/5000 cells/h vs control only: 1.11 ± 0.11, p < 0.05, n=5), 3 (1.42 ± 0.08, p < 0.01) and 5μmol/l (2.40 ± 0.32, p < 0.001), respectively.

Summary The results suggest that CFT and TGL from coffee do not protect beta cells against glucolipotoxicity. However, they may play a stimulatory role in the regulation of pancreatic insulin release.

Gut microbiota, wholegrain polyphenols and pancreatic beta cell function in type 2 diabetes

AC Hauge-Evans¹, G Corona¹, K Sarnsamak¹, DR Atar-Zwillenberg¹, C Yit¹, D Vauzour², A Costabile¹

¹Health Science Research Centre, University of Roehampton, London, UK, ²Faculty of Medicine and Health Sciences, University of East Anglia, Norwich, UK

Aims The intestinal microbiota plays an important role in host metabolism via production of dietary metabolites. Microbiota imbalances are linked to type 2 diabetes, but dietary modification of the microbiota may promote glycaemic control. This study investigated whether differences in gut microbiota between control and diabetic mice influence the production of polyphenolic metabolites from wheat wholegrain (WW). Second, the study assessed whether changes in metabolite profiles affect pancreatic beta cell function.

Methods Faecal samples (1% (v/v)) from control or high-fat high-fructose fed (HFHF) mice were fermented in vitro with 0.1% (w/v) WW for 0, 6 and 24 h. Polyphenolic profiles were determined by UHPLC-MS/MS and microbiota composition by bacterial 16S rRNA sequencing and qPCR. MIN6 beta cell apoptosis was assessed by measurement of caspase activity and insulin release by radioimmunoassay.

Results Levels of WW polyphenols were decreased in fermentation samples from HFHF with time and HFHF showed an overall dysbiotic microbiome profile (p < 0.05 vs control at 0 h, n=3). WW fermentation led to major changes in microbiota profile in control (6 h) and HFHF (24 h) groups, with increased diversity in HFHF (p < 0.05, n=3). Twenty hours incubation with 1% supernatant from both groups did not alter MIN6 cell apoptosis in response to palmitate and cytokines (p > 0.05, n=3). Insulin secretion was inhibited at 0 h by control, but not HFHF supernatant, and this effect was partly reverted at 24 h (p < 0.05, n=5).

Conclusion Our results suggest that a dysbiotic HFHF microbiota profile affects the production of polyphenolic metabolites with potential implications for beta cell function.

Investigation into the effects of HIV pre-exposure prophylaxis (PrEP) agents Truvada^® and rilpivirine on INS-1E beta cell function and viability

SC Maandi¹, A Patel², O Alsini¹, JG Mabley¹

¹School of Pharmacy and Biomolecular Sciences, University of Brighton, Brighton, UK, ²Brighton and Sussex Medical School, University of Sussex, Brighton, UK

Aims PrEP has gained its status as an important HIV prevention measure in the UK. Truvada^®, a combination of anti-HIV drugs tenofovir disoproxil fumarate (TDF) and emtricitabine, is currently in a successful UK-wide trial, while rilpivirine has recently entered phase II clinical trials. Although older anti-HIV drugs cause beta cell dysfunction, these PrEP agents have not been investigated. The aim of this study was to investigate the effects of TDF, emtricitabine and rilpivirine on beta cell function and viability.

Methods The rat beta cell line INS-1E was exposed to TDF, emtricitabine and rilpivirine (10 or 20μM) for 24 h, before measuring glucose (20mM)-stimulated insulin secretion, cell viability (MTT assay), apoptosis (flow cytometry) and antioxidant capacity (ABTS assay). Data are expressed as mean ± SEM, and statistical analysis was carried out using unpaired Student's t-test.

Results Rilpivirine significantly reduced glucose-stimulated insulin secretion from 100.8 ± 4% to 32.5 ± 4% and 36.2 ± 4% at 10 or 20μM respectively (p < 0.05 vs untreated cells). TDF and emtricitabine had no inhibitory effect on insulin secretion. Additionally, rilpivirine significantly reduced cell viability from 100 ± 1% to 83.8 ± 2% and 55.1 ± 3% at 10 or 20μM respectively (p < 0.05 vs untreated cells). TDF and emtricitabine had no effect on cell viability. Rilpivirine 20μM also significantly increased apoptosis levels from 2.2 ± 0.4% to 4.9 ± 1% and decreased antioxidant capacity from 100 ± 5% to 58.2 ± 5% (p < 0.05 vs untreated cells).

Conclusions Truvada^® appears to be a safe PrEP therapy with no beta cell damaging effects. However, the beta cell dysfunction mediated by rilpivirine may preclude this drug's use in individuals with type 2 diabetes.

SUPPORTING INFORMATION The conference poster for this abstract is available online in the Supporting Information section at the end of this page.

Generation and characterisation of a novel beta cell-specific Gpr56 knockout mouse model

EO Olaniru¹, AJ King¹, PM Jones¹, X Piao², SJ Persaud¹

¹Department of Diabetes, King's College London, London, UK, ²Department of Medicine, Children's Hospital and Harvard Medical School, Boston, Massachussetts, USA

Refer to Oral number A67.

Improving islet transplantation success by increasing expression of the epidermal growth factor receptor (EGFR)

UT Ali^1,2, K Suba^1,2, S Bitsi¹, AM Alonso^1,2, Y Patel^1,2, I Leclerc¹, GA Rutter¹, S Rothery³, A Tomas^1,2, V Salem^1,2

¹Section of Cell Biology and Functional Genomics, Imperial College London, London, UK, ²Investigative Medicine, Division of Diabetes, Endocrinology and Metabolism, Imperial College London, London, UK, ³Facility for Imaging and Light Microscopy, Imperial College London, London, UK

Aims An important factor determining islet transplantation success is the vascularisation of engrafted islets. The epidermal growth factor receptor (EGFR) is a key angiogenic, pro-survival and proliferative factor expressed in many tissues that has recently been implicated in renal transplant biology. In an in vivo mouse model, we aimed to investigate for the first time if EGFR overexpression in pancreatic islets improves engraftment.

Methods Adenoviruses encoding human EGFR alongside a green fluorescent protein reporter were generated. Following successful infection, EGFR-overexpressing or control (empty vector) islets were randomly transplanted into the anterior chambers of the eyes of syngeneic C57BL/6 mice. Using confocal microscopy and IV dextran-Texas red, islet volume and blood vessel density were assessed longitudinally over 30 days. To assess beta cell identity and function, qPCR of angiogenesis, beta cell enriched and beta cell disallowed genes was performed in identically treated islet groups as those transplanted.

Results Blood vessel density increased significantly for both groups of islets over time (p < 0.05), with EGFR-overexpressing islets demonstrating greater density compared to control islets at all time points. Supporting this, qPCR results showed that EGFR-overexpressing islets had greater gene expression of angiogenic factors (e.g. Vegfa). Consistent with preserved beta cell identity, beta cell enriched genes (e.g. Pdx-1) were upregulated in EGFR-overexpressing islets compared to control islets, while beta cell disallowed genes (e.g. Acot7) were further repressed.

Conclusion We are the first to demonstrate that EGFR overexpression can promote vascularisation of transplanted islets in vivo. The transcriptional profiling of these islets also suggested an improved potential for beta cell regenerative capacity and function.

Glucose metabolism is a key regulator of pancreatic beta cell autophagy

AV Williams, FP Zummo, K Lanyi, JAM Shaw, PE Lovat, C Arden

Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK

Refer to Oral number A61.

Olanzapine interacts with the canonical Wnt signalling pathway in MIN6 beta cells to adversely affect function

M Honey, G Scutt, WM Macfarlane, AJ Bone

School of Pharmacy and Biomolecular Sciences, University of Brighton, Brighton, UK

Aims There is evidence that the second generation antipsychotic drug olanzapine alters expression of the Wnt pathway effector TCF7L2 in the liver and skeletal muscle. In beta cells, Wnt signalling is also known to modify glucose-stimulated insulin secretion (GSIS) through fluctuations in free cytoplasmic beta catenin. We have previously demonstrated that olanzapine reduces GSIS in MIN6 cells; this study aimed to explore the interaction between olanzapine and Wnt/beta catenin signalling.

Methods Mouse MIN6 cells were used in all experiments. Following incubation with 10 μM olanzapine and 1μM XAV939, the cell cycle distribution was analysed using flow cytometry. qRT-PCR was used to determine mRNA expression and western blot was used to investigate protein concentration.

Results Exposure to olanzapine for 24h significantly altered the cell cycle distribution (G0/1: control 70.45 ± 1.12%, olanzapine 66.41 ± 1.22%, p < 0.05) (G2/M: control 24.06 ± 1.13%, olanzapine 27.84 ± 1.22%, p < 0.05). There was increased concentration of cytoplasmic beta catenin (p < 0.05) and increased mRNA expression of cell cycle regulators cyclin D1 and c-myc (p < 0.05) which was inhibited by Wnt inhibitor XAV939. GSIS analysis following 72h olanzapine with addition of XAV939 resulted in significantly improved insulin secretion compared to olanzapine alone (p < 0.05).

Conclusion These data suggest that olanzapine accelerates transition between G1 and S/G2 of the cell cycle through Wnt/beta catenin related increases in cyclin D1 and c-myc. Inhibiting the Wnt signalling pathway with XAV939 inhibits increased expression of cell cycle related genes, and also improves GSIS, suggesting that activation of the Wnt pathway by olanzapine is related to loss of beta cell function.

Inhibition of nicotinamide phosphoribosyltransferase (NAMPT) prevents cytokine-mediated beta cell dysfunction

D Egbase¹, S Butterworth², J Greally², S Sayers¹, E Evans¹, G Bewick¹, A Tsakmaki¹, L Smith¹, S Persaud¹, P Caton¹

¹Department of Diabetes, King's College London, London, UK, ²Division of Pharmacy, University of Manchester, Manchester, UK

Aims Intracellular nicotinamide phosphoribosyltransferase (NAMPT) is an NAD biosynthetic enzyme and is the rate-limiting enzyme in the NAD salvage pathway. Abnormally elevated NAMPT has been demonstrated to promote autoimmune and inflammatory processes, many of which are key to type 1 diabetes pathogenesis. Although, the role of NAMPT in diabetes remains unclear. Therefore, we used small-molecule NAMPT modulators to determine the role of abnormally elevated NAMPT levels and activity in cytokine-mediated beta cell death and dysfunction in islets.

Method Mouse islets were incubated for 24 h with either NAMPT inhibitor (10–20nM FK866 or 200–400nM C17) or activator (5–10μM P7C3) ± proinflammatory cytokine cocktail (1ng/ml). Islet function and survival were assessed via static and dynamic glucose-stimulated insulin secretion (GSIS) assays, caspase 3/7-glo bioluminescence assay, annexin V-binding assay.

Results Both NAMPT inhibitors reversed cytokine-mediated increases in islet apoptosis, when measured by caspase 3/7 activity (p < 0.01; C17) (p < 0.001; FK866). Static GSIS assays showed NAMPT inhibitors also protected islets against cytokine-mediated reductions in GSIS (both p < 0.0001), while P7C3 exerted the opposite effects. However, these effects were less pronounced when dynamic GSIS was measured by perifusion. While P7C3 co-incubations did not decrease caspase 3/7 activity, it increased (by 2-fold; p=0.003) islet cell viability and reduced (by 2-fold; p=<0.0001) necrotic single cell islet populations compared to cytokine-only treatment.

Summary NAMPT inhibition reversed cytokine-mediated beta cell apoptosis and dysfunction; however, NAMPT activation reversed cytokine-mediated necrosis. This suggests that NAMPT may play a pathophysiological role in mediating inflammation-induced beta cell damage and may be a therapeutic target for type 1 diabetes.

Mesenchymal stromal cell induced gene expression changes in pancreatic islets

EL Hubber, C Rackham, TJ Pullen, PM Jones

Department of Diabetes, King's College London, London, UK

Aims Mesenchymal stromal cells (MSCs) are a multifunctional adult progenitor cell type that can improve islet secretory function and viability in vitro. Understanding how MSCs do this may allow us to improve clinical islet transplantation protocols in the treatment of type 1 diabetes. The aims of this study are to determine what gene expression changes are occurring in MSC co-cultured islets and if these changes impact islet function and/or viability.

Methods We have employed single cell RNA-Seq to describe the transcriptomes of three biological replicates of mouse islets co-cultured with MSCs for 72 h. Data were analysed in R with the Seurat v3 package. Differentially expressed genes were analysed by qPCR.

Results Eight distinct cell populations that have previously been described in islet single cell studies were identified within sequenced islets: beta, alpha, gamma, PP, endothelial, MSCs, immune and Schwann. Differential gene expression analysis between control beta cells and MSC co-cultured beta cells revealed 26 significantly differentially expressed genes (DEGs) common between all replicates, 20 upregulated and 6 downregulated, with fold change in gene expression ranging from 2.16 (Mt1) to 1.31 (Dio1).These DEGs have a wide range of downstream functional roles including cell death, stress and trafficking. qPCR analysis demonstrated a strong positive correlation (r=0.82) between qPCR and scRNA-Seq fold-change in gene expression, validating RNA-Seq results.

Summary In total, 26 DEGs between control and MSC co-cultured beta cells have been identified and validated. Further work will use gene knockdown and overexpression to determine what effect these genes have on islet cell function and viability.

Acknowledgement Diabetes Research Group

Fibroblast growth factor 7 (FGF7) releasing particles enhance islet engraftment and improve metabolic control following islet transplantation

S Alwahsh¹, O Quatchi², P Starkey Lewis¹, A Bond³, J Mann¹, M Alvarez-Paino², J Noble³, S Forbes¹, K Shakesheff², S Forbes³

¹MRC Centre for Regenerative Medicine, University of Edinburgh, Edinburgh, UK, ²School of Pharmacy, University of Nottingham, Nottingham, UK, ³BHF Centre for Cardiovascular Science, University of Edinburgh, Edinburgh, UK

Background Transplantation of islets in type 1 diabetes is limited by poor islet engraftment into the liver, with 2–3 donor pancreases required per recipient.

Aims To condition the liver with growth factors to enhance islet engraftment to improve long-term graft function.

Methods Diabetic mice received a marginal islet transplant (n=400 islets) via the hepatic portal vein with FGF7 loaded galactosylated poly(dl-lactide-co-glycolic acid) (FGF7-GAL-PLGA) particles, specifically formulated to target the liver. Hepatocyte proliferation in short-term and long-term studies was assessed along with histological analysis of the liver for islets and markers of vascularisation (CD31). Glycaemic control was assessed over a six-week period.

Results In short-term experiments, 26μm diameter FGF7-GAL-PLGA particles specifically targeted the liver, promoting hepatocyte proliferation; 0.1mg FGF7-GAL-PLGA particles (60ng FGF7) induced hepatocyte proliferation with greater efficacy and specificity than subcutaneous FGF7 (1.25mg/kg x2 doses; ˜75μg FGF7) and was well tolerated. Numbers of engrafted islets and vascularisation were greater in liver sections of mice receiving islets and FGF7-GAL-PLGA particles vs mice receiving islets alone, 72 h post-transplant. More mice (six out of eight) that received islets and FGF7-GAL-PLGA particles normalised blood glucose concentrations by 30 days post-transplantation vs nought out of eight mice receiving islets alone with no evidence of increased proliferation of cells within the liver at this stage and normal liver function tests.

Conclusion Liver targeted FGF7-GAL-PLGA particles achieve selective FGF7 delivery to the liver promoting short-term hepatocyte proliferation and islet engraftment to help normalise blood glucose levels with an excellent safety profile.

Two-photon microscopy is an effective imaging modality to detect alpha and beta cell distribution throughout intact living islets in different species

A Garcia-Burgos^1,2, P Burgoyne², G Borthwick², Y Chang³, N Kang³, R Wiegand², A McNeilly⁴, S Forbes², R Duncan¹

¹Institute of Biological Chemistry, Biophysics and Bioengineering, Heriot Watt University, Edinburgh, UK, ²BHF Centre for Cardiovascular Science, University of Edinburgh, Edinburgh, UK, ³Department of Chemistry, POSTECH, South Korea, ⁴Systems Medicine, University of Dundee, UK

Background The normal cellular composition and distribution of cells within an islet is of fundamental importance, aiding our understanding of the defects seen in diabetes. In man, alpha and beta cells appear randomly distributed within an islet, whereas in mouse where there is a central distribution of beta cells. Such studies have relied on immunohistochemistry using fixed tissue, which may alter morphology, to examine islets which are relatively large multicellular structures of 150–200μm in diameter, as it has not previously been possible to image live intact islets with sufficient resolution and depth penetration.

Aims To develop imaging methodologies using super-resolution of live islets with specific labelling of alpha and beta cells throughout the 3D structure of the islet.

Methods Mouse and human islets were obtained and incubated in a 2-glucosamine-based 2-photon fluorescent probe that label the GLUT-2 receptor of the beta cells (TP-β) and a 2-photon fluorescent probe that labels the glucagon moiety (TP-α) for 30min and then transferred to high glucose (20mM) for 4h and imaged using 2-photon microscopy.

Results In living mouse islets, the distribution of alpha and beta cells demonstrated a central beta cell distribution throughout the 3D structure with high definition between cellular structures. In human islets, there was a random distribution of alpha and beta cells throughout the 3D islet structure.

Conclusions Using 2-photon microscopy and fluorescent probes, a 3D model and cellular composition of a live intact islet may be generated negating the need to use fixed tissue.

The impact of a long non-coding RNA at the Pax6 locus on beta cell identity and function

RM Callingham¹, I Leclerc¹, TJ Pullen^1,2, GA Rutter¹

¹Section of Cell Biology and Functional Genomics, Imperial College London, London, UK, ²Department of Diabetes, King's College London, London, UK

Aims It is increasingly clear that loss of beta cell identity contributes to the progression of type 2 diabetes, and is characterised by decreases in beta cell transcription factors. Maintaining expression of the Pax6 transcription factor is essential for maintaining beta cell identity. To understand how it is regulated, we investigated the role of a recently identified antisense-orientated long non-coding RNA (lncRNA), Pax6os1, expressed in beta cells from the Pax6 locus. siRNA knockdown of Pax6os1 in the MIN6 mouse beta cell-line indicated that this molecule may be a negative regulator of Pax6 expression and beta cell identity. We therefore studied the role of Pax6os1 in vivo.

Methods Mice inactivated globally for Pax6os1 were generated via the excision of the promoter and first exon using CRISPR/Cas9 technology. RNA–protein interactions were explored using a proteomic approach (SILAC).

Results Ablation of Pax6os1 expression in male mice led to decreased beta cell mass (p=0.044, n=3–4), consistent with reduced proliferation observed in siRNA knockdown in MIN6 cells. While there was no effect in mice on normal chow, Pax6os1 knockout improved glucose tolerance in females fed high-fat diet (p=0.017, n=6–8). Pax6os1 knockout increased insulin expression (p=0.030, n=4–6) despite having no clear effect on Pax6 expression (p=0.185, n=4–6). In vitro RNA-binding identified histone H1 as a binding partner of Pax6os1.

Conclusions Pax6os1 plays a role in the control of beta cell identity and function, largely independently of changes in Pax6 expression. These actions may be via alterations in chromatin accessibility and the expression of critical beta cell genes.

The long non-coding RNA PAX6-AS1 modulates human beta cell function

L Lopez Noriega¹, A Martinez Sanchez¹, R Callingham¹, E Akalestou¹, P Chabosseau¹, I Leclerc¹, P Marchetti², TJ Pullen³, GA Rutter¹

¹Department of Medicine, Section of Cell Biology and Functional Genomics, Imperial College London, London, London, UK, ²Department of Clinical and Experimental Medicine, University of Pisa, Pisa, Italy, ³Department of Diabetes, King's College London, London, UK

Aims PAX6 is a transcription factor required for normal beta cell function. Long non-coding RNAs are transcripts longer than 200 nucleotides that may regulate the expression of nearby genes. We investigate here roles for an antisense long non-coding RNA expressed from the PAX6 locus (PAX6-AS1). The aim of this study was to determine whether PAX6-AS1 modulates PAX6 expression and impacts human beta cell function through this, or other mechanisms.

Materials and Methods PAX6-AS1 expression was measured in EndoC-βH1 cells and human pancreatic islets incubated at different glucose concentrations and in human islets from type 2 diabetic donors. Additionally, PAX6-AS1 was silenced in EndoC-βH1 cells and in human islets using CRISPR/Cas9 technology and shRNA lentiviral vectors, respectively.

Results PAX6-AS1 expression was induced at high glucose concentrations (3mM vs 17mM) in both EndoC-βH1 cells (n=5, p < 0.01) and human islets (n=7, p < 0.03). In addition, human islets from type 2 diabetic donors showed increased expression of PAX6-AS1 vs controls (n=10-5, p < 0.01). EndoC-βH1 cells, engineered using a tailored CRISPR/Cas9 approach to delete exon 1 of PAX6-AS1, exhibited a 47 ± 6% (n=6, p < 0.0001) decrease in PAX6-AS1, but increased levels of several beta cell signature genes. Interestingly, PAX6-AS1 knockdown in human islets decreased ghrelin mRNA levels, while a strong tendency towards augmented insulin mRNA expression was observed. Furthermore, PAX6-AS1-knockdown islets displayed enhanced glucose-induced insulin secretion and a tendency towards increased calcium dynamics in response to glucose.

Conclusion PAX6-AS1 upregulation may contribute to beta cell failure during the development of some forms of type 2 diabetes.

The type 2 diabetes-associated lipid-binding protein STARD10 could bind phosphatidylinositides and affect islet lipid composition

G Carrat¹, L Haataja², P Arvan², A Tomas¹, K Cheng³, NS Amirruddin⁴, RB Sessions⁵, AK Teo⁴, D Wigley³, GA Rutter¹

¹Section of Cell Biology and Functional Genomics, Imperial College London, London, UK, ²Division of Metabolism, Endocrinology & Diabetes, University of Michigan Medical School, Ann Arbor, Michigan, USA, ³Section of Structural Biology, Imperial College London, London, UK, ⁴Stem Cells and Diabetes Laboratory, Institute of Molecular and Cell Biology, Singapore, Singapore, ⁵School of Biochemistry, University of Bristol, Bristol, UK

Aim Risk alleles for type 2 diabetes in the STARD10 locus on chromosome 11q13 impair glucose-induced insulin secretion and are associated with decreased proinsulin:insulin ratios. We have shown that the type 2 diabetics risk associated with variation at this locus is likely to be mediated through lowered STARD10 expression in the beta cell. Here, we investigate the mechanisms by which STARD10 may regulate insulin secretion.

Methods Islets were isolated from StarD10^fl/fl-Ins1Cre male mice (βStarD10KO). Electron microscopy (EM) images were obtained from islets after chemical fixation and sectioning using an ultramicrotome. Total zinc content was measured by inductively coupled plasma mass spectrometry (MS). STARD10-binding partners were identified after immunoprecipitation and analysis by Tandem Mass Tag/MS at the University of Bristol Proteomic Facility.

Results EM analysis of islets from βStarD10KO mice revealed an altered dense core granule appearance, with a dramatic increase in ‘rod like’ dense cores, and an increase in total islet zinc content. MS analysis of STARD10-binding partners detected several proteins shown to play a role in insulin secretion and the phosphoinositides pathway. Determination of human STARD10 crystal structure and molecular docking revealed that STARD10 could bind phosphatidylinositol in addition to its previously described ligands. In addition, the lipidomic profile was affected in the islets of the βStarD10KO mice, with an increase in cholesteryl ester and phosphatidylinositol.

Conclusion Our data suggest that STARD10 could bind and transport phosphatidylinositols and affect the islet lipid composition. We also identify STARD10 as an important regulator of insulin granule maturation and beta cell zinc homeostasis.