Type 2 diabetes mellitus remains an enormous health burden – there are over 300 million sufferers worldwide (1) and 20% of the US healthcare budget is spent on caring for people with diabetes (2). Pancreatic islet dysfunction underlies the development of diabetes. While it is recognised that communication between cells within the islet is essential for proper function, the mechanism responsible is unknown. Connexin 36 is expressed on insulin-producing beta cells, is important for coordinated pulsatile insulin release and may be involved in intercellular communication within pancreatic islets (3). The interaction between pancreatic islets and endothelial cells was investigated in vitro. Pancreatic islets were isolated from male C57B6 mice by infusion of collagenase into the pancreatic duct in situ followed by digestion at 37°C and density purification. Endothelial cells were grown from C57B6 bone marrow by culture on fibronectin in defined media for one week and confirmed to express endothelial markers (CD31 and E-selectin), bind lectin and uptake acetylated low-density lipoprotein by flow cytometry. To investigate soluble mechanisms, islets were cultured (10 islets per well) with endothelial cell-conditioned medium for 3 days and tested for glucose-stimulated insulin release. Secreted insulin was detected by high-sensitivity ELISA and corrected for total protein and stimulation indices were calculated as insulin secretion at high glucose (20 mM) divided by basal insulin secretion (at 3 mM glucose). Values are means±SEM compared by unpaired t-tests. Endothelial cell-conditioned islets had an increased level of basal insulin release (0.7±0.22 ng/min/mg total protein; n=7) compared to controls (0.2±0.04 ng/min/mg total protein; n=6; p=0.05). However, they had a reduced ability to upregulate insulin release in response to high glucose (stimulation index = 1.3±0.5 compared to 4.2±0.9 for controls; p=0.01). Effects on gene expression were analysed by real-time PCR using Taqman primers (normalizing to the house-keeping gene B2m) following 3 day non-contact co-cultures where islets (25 islets per well) were cultured in transwells (0.4 µm pores) above endothelial monolayers. Co-cultured islets significantly down-regulated the expression of connexin 36 (0.4±0.05 fold relative to control islets; n=18; p=0.04). There was no difference in expression of the Insulin gene between groups. These data suggest that connexin 36 expression is modulated during intercellular communication between beta cells and endothelial cells, resulting in dysregulated insulin release. We suggest this represents a ‘survival’ islet phenotype, where baseline insulin secretion is maintained during vascular remodelling. This mechanism may have relevance for islet physiology, where the function and survival of beta cells is critically dependent on the presence of intra-islet vascular endothelium.