Engineering a More Physiologically Relevant, Vascularized Model of Hepatic Insulin Resistance

Introduction:: Hepatic insulin resistance, characterized by impaired signaling and dysregulated metabolic regulation in the liver, is a major feature of the progression of metabolic disorders including nonalcoholic fatty liver disease and type 2 diabetes. However, much remains to be understood about the interplay between molecular and cellular cues driving insulin resistance establishment. Current in vitro models are limited in their ability to accurately recapitulate the complexity of this disease state, often using supraphysiological levels of nutrients and hormones or neglecting to incorporate relevant cell types. Here, we utilize co-culture of hepatic spheroids in a microvascular network to better model cell-cell interactions within the liver, while altering nutrient and insulin levels within a physiological range to better recapitulate a clinically relevant insulin resistant disease state.


Materials and Methods:: In this study, we begin using alginate microwells to generate multicellular spheroids composed of primary human hepatocytes, Kupffer cells, human umbilical vein endothelial cells (HUVECs) and normal human lung fibroblasts (NHLFs). Next, spheroids are seeded in microfluidic devices with additional HUVECs and NHLFs in a fibrin hydrogel to create vascularized hepatic spheroids (Fig 1A). To validate that hepatocytes remain functional in this co-culture platform, we performed ELISAs to measure albumin secretion over time, as well as luciferase-based assays to measure Cytochrome P450 (CYP3A4) activity. Over the course of each experiment, these cells are cultured in media containing either physiological or pathological levels of glucose, insulin, and free fatty acids, in to simulate either a healthy or Type 2 Diabetic (T2D) state. To assess the development of hepatic insulin resistant resulting from chronic exposure to this T2D media, we measured insulin clearance via ELISA, gluconeogenesis in response to insulin stimulation, and qPCR to measure expression of genes in affected by insulin signaling. Furthermore, we used time-lapse imaging to study changes vascular morphology over time in T2D media compared to the physiological condition. Perfusion of fluorescent dye in these microvascular networks allowed us to quantify changes in vascular permeability in the T2D devices.


Results, Conclusions, and Discussions:: Results:

Our results demonstrate that our co-culture platform is supportive of vascularized hepatic spheroids, with images showing extensive interactions between hepatic spheroids and GFP-endothelial cells (Fig 1C). Perfusion of fluorescent dextran showed that not only are the spheroids and networks in contact but moreover the self-organized networks have open lumens which allow for transport of dextran directly into the hepatic spheroids (Fig 1D). Measurements of hepatocyte function showed stable secretion of albumin over time, as well as CYP3A4 activity in both media conditions. When comparing the effect of a physiological vs high nutrient “T2D” media, we saw the development of an insulin resistant phenotype over time indicating by impaired insulin clearance over the course of culture (Fig 1E). Furthermore, T2D cultures showed increased pro-inflammatory cytokine secretion (Fig 1F) and altered expression of gluconeogenesis pathway genes. Finally, microvascular networks themselves showed a disease phenotype with narrowing of the networks and increased permeability over time in the T2D media compared to the physiological media condition (Fig 1G).

Conclusions:

These results demonstrate that we have established an in vitro culture platform supportive of multiple cell types relevant for understanding cell-cell interactions in the liver. By appropriately tuning levels of nutrients and insulin in our system, we can apply this platform to model the development of hepatic insulin resistance, with results consistent with known clinical features of Type 2 Diabetes, such as chronically elevated levels of IL-6 and increased vascular permeability.

Discussion:

This platform represents a valuable improvement in modeling liver disease, allowing for independent tuning of molecular or cellular parameters to better understand mechanisms of insulin resistance establishment. Future work will more deeply characterize the model utilizing transcriptomics to deconvolute the contribution of each cell type in the system, alongside proteomic analysis of the chemokine and cytokine profiles to compare our physiological and T2D cultures. Furthermore, this model is adaptable to a range of applications, serving as a valuable tool for therapeutic testing and validation to study prevention or reversal of hepatic insulin resistance.




Acknowledgements (Optional): :

References (Optional): :