Hepatic insulin resistance is a fundamental phenomenon observed in both Type 2 diabetes (T2D) and metabolic (dysfunction) associated fatty liver disease (MAFLD). The relative contributions of nutrients, hyperinsulinemia, hormones, inflammation, and other cues are difficult to parse in vivo as they are convoluted by interplay between the local and systemic events. Here, we used a well-established human liver microphysiological system (MPS) to establish a physiologically-relevant insulin-responsive metabolic baseline and probe how primary human hepatocytes respond to controlled perturbations in insulin, glucose, and free fatty acids (FFAs). Replicate liver MPS were maintained in media with either 200 pM (normal) or 800 pM (T2D) insulin for up to 3 weeks. Conditions of standard glucose (5.5 mM), hyperglycemia (11 mM glucose), normal (20µM) and elevated FFA (100 µM), alone and in combination were used at each insulin concentration, either continuously or reversing back to standard media after 2 weeks of simulated T2D conditions. Hepatic glucose production, activation of signaling pathways, insulin clearance, transcriptome analysis, and intracellular lipid and bile acid accumulation were assessed. Hyperinsulinemia alone induces insulin resistance after one week of exposure, while hyperglycemia and increased FFAs significantly exacerbate this phenotype. Hyperinsulinemia, along with elevated glucose and FFAs, transcriptionally predisposes hepatocytes to insulin resistance through altered metabolic and immune signaling pathways. The phenotypes observed in hyperinsulinemia and nutrient overload are partially reversible upon return to normophysiologic conditions. Our enhanced in vitro model, replicating multiple aspects of the insulin-resistant condition, offers improved insights into disease mechanisms with relevance to human physiology. The many nutritional, hormonal, and systemic inflammatory factors that contribute to the loss of sensitivity to insulin in Type 2 Diabetes and other metabolic disorders are difficult to parse in human patients, and animal models fail to capture all the human dimensions of the relevant biology. We, therefore, developed a microphysiological systems model, involving a microfluidic platform that cultures a simulated human liver for weeks at a time, under controlled nutrient and hormone concentrations, to analyze how the effects of nutrition (glucose and free fatty acids) together with increased insulin cause pathological features seen in human liver in vivo . We found that continually high levels of insulin predispose the liver to some of these features, along with reversal when the nutritional and insulin levels were restored to healthy values.
Read full abstract