Associations Among Adipose Tissue Immunology, Inflammation, Exosomes and Insulin Sensitivity in People With Obesity and Nonalcoholic Fatty Liver Disease

Abstract

Background and aims: Insulin resistance is a key factor in the pathogenesis of nonalcoholic fatty liver disease (NAFLD). We evaluated the importance of subcutaneous abdominal adipose tissue (SAAT) inflammation and both plasma and SAAT-derived exosomes in regulating insulin sensitivity in people with obesity and NAFLD.

Methods: Adipose tissue inflammation (macrophage and T-cell content and expression of proinflammatory cytokines), liver and whole-body insulin sensitivity (assessed using a hyperinsulinemic-euglycemic clamp and glucose tracer infusion), and 24-hour serial plasma cytokine concentrations were evaluated in 3 groups stratified by adiposity and intrahepatic triglyceride (IHTG) content: (1) lean with normal IHTG content (LEAN; N = 14); (2) obese with normal IHTG content (OB-NL; N = 28); and (3) obese with NAFLD (OB-NAFLD; N = 28). The effect of plasma and SAAT-derived exosomes on insulin-stimulated Akt phosphorylation in human skeletal muscle myotubes and mouse primary hepatocytes was assessed in a subset of participants.

Results: Proinflammatory macrophages, proinflammatory CD4 and CD8 T-cell populations, and gene expression of several cytokines in SAAT were greater in the OB-NAFLD than the OB-NL and LEAN groups. However, with the exception of PAI-1, which was greater in the OB-NAFLD than the LEAN and OB-NL groups, 24-hour plasma cytokine concentration areas-under-the-curve were not different between groups. The percentage of proinflammatory macrophages and plasma PAI-1 concentration areas-under-the-curve were inversely correlated with both hepatic and whole-body insulin sensitivity. Compared with exosomes from OB-NL participants, plasma and SAAT-derived exosomes from the OB-NAFLD group decreased insulin signaling in myotubes and hepatocytes.

Conclusions: Systemic insulin resistance in people with obesity and NAFLD is associated with increased plasma PAI-1 concentrations and both plasma and SAAT-derived exosomes. ClinicalTrials.gov number: NCT02706262 (https://clinicaltrials.gov/ct2/show/NCT02706262).

Keywords: Cytokines; Insulin Resistance; Macrophages; PAI-1; T Cells.

Figure 1.

Macrophages in subcutaneous abdominal adipose tissue. (A and B) Number of total, pro-inflammatory (M1-like), and anti-inflammatory (M2-like) macrophages per gram of tissue (A) and M1-like and M2-like macrophages as a percentage of total macrophages and the M1-like:M2-like ratio (B) in SAAT in LEAN, OB-NL and OB-NAFLD groups. (C and D) Relationship between the hepatic (HISI; C) and skeletal muscle (Glucose Rd/I; D) insulin sensitivity and M1-like and M2-like macrophages as a proportion of total macrophages and the M1-like:M2-like ratio in LEAN (black circles), OB-NL (grey circles) and OB-NAFLD (white circles) participants. The values in (A) and (B) are means ± SEM. *Value significantly different from the LEAN value, P < .05. ^†Value significantly different from the OB-NL value, P < .05. ^#linear trend, P < .05.

Figure 2.

T cells in subcutaneous abdominal adipose tissue. (A) Absolute numbers (top row) and proportions (bottom row) of naïve and memory CD4 and CD8 T cells. (B) Absolute Th1 and Th17 cell numbers and proportions within total CD4 T cells. (C) Frequencies of IFN-γ and IL-17 positive cells within CD4 (left two graphs) and CD8 (right two graphs) T cells, following in vitro stimulation. (D) Absolute numbers of CD69⁺ CD4, CD8 and NK cells. Values are means ± SEM. *Value significantly different from the LEAN value, P < .05. ^†Value significantly different from the OB-NL value, P < .05. ^#linear trend, P < .05.

Figure 3.

Gene expression of markers of inflammation in subcutaneous abdominal adipose tissue. Expression of genes encoding for cytokines [PAI-1 (encoded by SERPINE1), TNF-α, IL-1β, IL-6, CCL2, CCL3, CCL5, CXCL16, IL-10] and for proteins highly expressed in macrophages (CD68 and CD206, encoded by MRC1) and lipid-associated macrophages (TREM2) in SAAT in LEAN, OB-NL and OB-NAFLD groups. Values are means ± SEM *Value significantly different from the LEAN value, P < .05. ^†Value significantly different from the OB-NL value, P < .05. ^#linear trend, P < .05.

Figure 4.

24-h plasma cytokine profiles. Plasma cytokine concentrations determined hourly between 7 am and 11 pm on day 1 and at 5 am and 7 am on day 2 (left panels) and 24-h AUCs (right panels) in LEAN, OB-NL and OB-NAFLD groups. Values are means ± SEM. *Value significantly different from the LEAN value, P < .05. ^†Value significantly different from the OB-NL value, P < .05.

Figure 5.

Relationships among plasma PAI-1 24-h area-under-the-curve (AUC) and both hepatic (HISI) and skeletal muscle (Glucose Rd/I) insulin sensitivity (A), adipose tissue SERPINE1 expression and both adipose tissue CD68 expression and M1-like macrophages (B), and plasma PAI-1 24-h AUC and both adipose tissue CD68 expression and M1-like macrophages (C) in LEAN (black circles), OB-NL (grey circles) and OB-NAFLD (white circles) participants.

Figure 6.

Exosomes and insulin action. (A and B) Insulin-stimulated phosphorylation of serine 473 of Akt in human skeletal muscle myotubes (A) and mouse hepatocytes (B) treated with exosomes isolated from plasma from LEAN, OB-NL and OB-NAFLD participants and SAAT explants from OB-NL and OB-NAFLD participants. (C) Relationship between skeletal muscle insulin sensitivity (glucose Rd/I) and phosphorylation of serine 473 of Akt in myotubes treated with exosomes isolated from plasma and SAAT in LEAN (black circles), OB-NL (grey circles) and OB-NAFLD (white circles) participants. (D) Relationship between the hepatic insulin sensitivity index and phosphorylation of serine 473 of Akt in hepatocytes treated with exosomes isolated from plasma and SAAT in LEAN (black circles), OB-NL (grey circles) and OB-NAFLD (white circles) participants. Values are means ± SEM. *Value significantly different from the LEAN value, P < .05. ^†Value significantly different from the OB-NL value, P < .05.