Characterization of adipose tissue and liver innate immune cells in obesity-related insulin resistance

Abstract

Obesity is a highly prevalent disease associated with a number of chronic and noncommunicable diseases including type 2 diabetes (T2D), nonalcoholic fatty liver disease (NAFLD), and cardiovascular disease (CVD). Obesity is associated with a chronic low-grade systemic inflammation, which is thought to cause insulin resistance as well as other co-morbidities. The systemic inflammation may originate in the adipose tissue (AT), and is characterized by an infiltration of immune cells into the tissue and paracrine and endocrine secretion of pro-inflammatory cytokines. Adipose tissue macrophages (ATMs) and NK cells has previously been shown to be important for obesity-associated AT inflammation, and NK cells have also been linked to liver inflammation in mice models of non-alcoholic steatohepatitis (NASH). In paper I, we investigated circulating NK cells in patients with NAFLD and NASH and found that the proportions of CD56dim to CD56bright NK cells were similar to healthy controls. Further, NK cells from patients and healthy controls showed similar expression levels of activating and inhibitory receptors, with the exception of NK cells from patients with NASH, which showed increased expression of the activating receptor NKG2D. NK cells from NAFLD and NASH patients also had retained functionality compared to healthy controls, and the frequency of liver and AT NK cells were unaltered in the patients. In paper II, we characterized ATMs in individuals with obesity, and found that the visceral adipose tissue (VAT) harbored relatively more pro-inflammatory M1-like macrophages than the subcutaneous adipose tissue (SAT). The pro-inflammatory ratio between M1- and M2-like ATMs correlated positively with the degree of insulin resistance and dyslipidemia. For other markers of inflammation, such as expression of inflammatory genes or the numbers of crown-like structures in the AT, we did not find any significant associations with the metabolic disease parameters. Interestingly, improved insulin sensitivity after weight loss following bariatric surgery was not correlated to a simultaneous decrease in circulating inflammatory factors, suggesting that these two events were not linked. Lastly, in paper III we characterized the surface proteome of ATMs and adipocyte progenitor cells (APCs) and found subpopulations of the macrophages and progenitors with distinct expression of surface proteins. We compared our surface protein data with an available scRNA-seq dataset, and found enrichment of both ATM and APC-specific markers in macrophage and progenitor clusters. We also used the scRNA-seq data to explore the local interactome of the macrophage and progenitor cells in adipose tissue, using a computational method guided by our surface protein expression data.