A microbiota-informed pipeline for in vitro characterisation of insulinotropic properties of dairy fermentates using static EndoC-βH1 and dynamic microfluidic models

Diabetes mellitus (DM), which encompasses Type 1, Type 2 and gestational diabetes, affects approximately 537 million adults, with prevalence continuing to rise. In addition to pregnancy, factors such as dietary choices, lifestyle and infections can lead to the development of DM. Given the increasing prevalence of DM, it is necessary to continue to identify approaches to control the release of insulin. Emerging evidence indicates that the gut microbiota plays a central role in mediating the metabolic effects of diet on the host, including the regulation of insulin secretion. Dietary interventions have been proven to be successful in this regard, but identifying foods and food components that influence insulin secretion is challenging due to the complexity of diet-microbiota-host interactions and the lack of fast and standardised screening tools. We have developed a novel screening framework based on the β-cell insulin response, using the human cell line EndoC-βH1 to assess candidate foods and ingredients with insulinotropic properties. As an initial proof-of-concept, we applied the framework to food ingredients, i.e., dairy-derived fermentates, which had first undergone simulated digestion followed by addition to an ex vivo colon model, allowing us to mimic the metabolic transformations taking place in the human gut. The processed colonic fermentation supernatants were then evaluated using a static and microfluidic platform, respectively, to capture dynamic insulin release profiles. Distinct fermentate compositions were associated with varying levels of insulin secretion. Our results suggest that this framework could be applied broadly to screen foods and food components for insulinotropic potential, providing an affordable approach for the development for anti-diabetic dietary applications.