Exploring the intrinsic and extrinsic determinants of heterogeneity in a β-cell network

Pancreatic islets are micro-organs composed of multiple endocrine cell types. β-cells are the most common of these and are highly heterogeneous in both their intrinsic properties, such as ion channel conductances and metabolic activity, and extrinsic properties, including gap junction coupling and paracrine signaling. Capturing these diverse sources of heterogeneity is essential for computational models that aim to reproduce islet function. We evaluate two established multicellular models, the coupled Cha-Noma model of the mouse β-cell network, and the Riz human model. Prior work with these two models suggest that the Cha-Noma model involves minimal intrinsic heterogeneity and therefore bursts with highly synchronized activity, whereas heterogeneity in the Riz model prescribed from in vitro patch-clamp results in highly unsynchronized behavior of the coupled network. We hypothesize that adjusting the number of bursting cells in both model formulations may invoke more physiologically realistic network coordination. We applied a categorical sensitivity analysis to establish which parameters are most important for determining bursting in single cells of the Riz model. We also introduced new heterogeneity (based on single-cell gene expression) in parameters that had previously been treated as invariant among β-cells. We hypothesized that introducing heterogeneity in the small conductance Ca ²⁺ K ⁺ channel in particular could promote a higher proportion of bursting cells. Lastly, both models fail to incorporate the influences of non-β-cells. We introduced paracrine signaling between α and β-cells into the coupled Cha-Noma model and showed it plays a role in accelerating the response time of β-cells to acute glucose stimulation (i.e. promotes a 1 ^st responder phenotype).