Expansion and Impaired Mitochondrial Efficiency of Deep Subcutaneous Adipose Tissue in Recent-Onset Type 2 Diabetes

Abstract

Context/objective: Impaired adipose tissue (AT) function might induce recent-onset type 2 diabetes (T2D). Understanding AT energy metabolism could yield novel targets for the treatment of T2D.

Design/patients: Male patients with recently-diagnosed T2D and healthy male controls (CON) of similar abdominal subcutaneous AT (SAT)-thickness, fat mass, and age (n = 14 each), underwent hyperinsulinemic-euglycemic clamps with [6,6-2H2]glucose and indirect calorimetry. We assessed mitochondrial efficiency (coupling: state 3/4o; proton leak: state 4o/u) via high-resolution respirometry in superficial (SSAT) and deep (DSAT) SAT-biopsies, hepatocellular lipids (HCL) and fat mass by proton-magnetic-resonance-spectroscopy and -imaging.

Results: T2D patients (known diabetes duration: 2.5 [0.1; 5.0] years) had 43%, 44%, and 63% lower muscle insulin sensitivity (IS), metabolic flexibility (P < 0.01) and AT IS (P < 0.05), 73% and 31% higher HCL (P < 0.05), and DSAT-thickness (P < 0.001), but similar hepatic IS compared with CON. Mitochondrial efficiency was ~22% lower in SSAT and DSAT of T2D patients (P < 0.001) and ~8% lower in SSAT vs DSAT (P < 0.05). In both fat depots, mitochondrial coupling correlated positively with muscle IS and metabolic flexibility (r ≥ 0.40; P < 0.05), proton leak correlated positively (r ≥ 0.51; P < 0.01) and oxidative capacity negatively (r ≤ -0.47; P < 0.05) with fasting free fatty acids (FFA). Metabolic flexibility correlated positively with SAT-oxidative capacity (r ≥ 0.48; P < 0.05) and negatively with DSAT-thickness (r = -0.48; P < 0.05). DSAT-thickness correlated negatively with mitochondrial coupling in both depots (r ≤ -0.50; P < 0.01) and muscle IS (r = -0.59; P < 0.01), positively with FFA during clamp (r = 0.63; P < 0.001) and HCL (r = 0.49; P < 0.01).

Conclusions: Impaired mitochondrial function, insulin resistance, and DSAT expansion are AT abnormalities in recent-onset T2D that might promote whole-body insulin resistance and increased substrate flux to the liver.

Trial registration: ClinicalTrials.gov NCT01055093.

Keywords: Adipose tissue; humans; insulin resistance; metabolic flexibility; mitochondrial function; type 2 diabetes.

Figure 1.

Subcutaneous adipose tissue layers of the abdominal wall. (A) Scheme depicting all adipose tissue layers of the abdominal wall from skin to the intestine: visceral adipose tissue (VAT), whole subcutaneous adipose tissue (WSAT) composed of superficial (SSAT) and deep subcutaneous layers (DSAT). (B) Ultrasound image at the level of musculus rectus abdominis showing SSAT, DSAT, and the Scarpa fascia (white line between SSAT and DSAT) dividing both adipose tissue depots in an individual with normal glucose tolerance (CON) and (C) a type 2 diabetes patient (T2D). The red arrow indicates the increase of SSAT thickness in CON and the orange arrow indicates the increase of DSAT thickness in T2D. (D) Ratio of SSAT/WSAT and (E) DSAT/WSAT thickness. Data are shown as mean ± SEM. ***P < 0.001, data were compared by 2-tailed Mann-Whitney U test. All variables were assessed in n = 14 T2D patients and n = 14 CON.

Figure 2.

Muscle (A) and liver insulin sensitivity (B), free fatty acid (FFA) levels during clamp (C) as a basis to assess adipose tissue insulin resistance (D). Individuals with normal glucose tolerance (CON), rate of disappearance (Rd) for muscle insulin sensitivity, type 2 diabetes patient (T2D). The adipose tissue insulin resistance index (Adipo IR) was calculated as the product of the plasma FFA and insulin levels during the clamp and reflects adipose tissue insulin resistance. Hepatic insulin sensitivity was assessed by the difference between basal and insulin-suppressed endogenous glucose production (∆EGP). Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, data were compared by 2-tailed Mann-Whitney U test. All variables were assessed in n = 14 T2D patients and n = 14 CON.

Figure 3.

Metabolic flexibility Data are shown as mean ± SEM. ** P < 0.01, data are compared by 2-tailed Mann-Whitney U test. Difference between basal and insulin-stimulated respiratory quotient (∆RQ) to assess metabolic flexibility, individuals with normal glucose tolerance (CON) and with type 2 diabetes (T2D). In 1 participant of the control group, metabolic flexibility was not assessable, because the insulin-stimulated measurement of respiratory quotient was not performed due to technical problems. Thus, variables were assessed in n = 14 T2D patients and n = 13 CON.

Figure 4.

Mitochondrial oxidative capacity (A), coupling (B) and proton leak (C) in subcutaneous adipose tissue layers. Data are shown as mean ± SEM. * P < 0.05, *** P < 0.001, data were compared by ANCOVA adjusted for age, BMI, and total body fat and by paired Student t-test. All variables were assessed in n = 14 T2D patients and n = 14 CON. Abbreviations: CON, controls (individuals with normal glucose tolerance); DSAT, deep subcutaneous adipose tissue; fccp, carbonylcyanide-4-trifluoromethoxy phenylhydrazone; LCR, leak control ratio (LCR = state 4o/state u) to assess mitochondrial proton leak; RCR, respiratory control ratio (RCR = state 3/state 4o) to assess mitochondrial coupling; SSAT, superficial subcutaneous adipose tissue (SSAT); T2D, patients with type 2 diabetes.

Figure 5.

Association of metabolic flexibility and muscle insulin sensitivity with mitochondrial coupling and proton leak in SSAT/DSAT. Mitochondrial coupling from respiratory control ratio (RCR = state 3/state 4o respiration) and proton leak from leak control ratio (LCR = state 4o/state u) in superficial (SSAT) and deep subcutaneous adipose tissue (DSAT). Circles indicate type 2 diabetes patients (T2D) and triangles indicate individuals with normal glucose tolerance (CON). Of note, after bonferroni correction no correlation remaind significant. § indicates that the results are still significant after adjusting for age, BMI, and body fat still significant. In 1 CON participant, metabolic flexibility was not assessable, because the insulin-stimulated measurement of respiratory quotient was not performed due to technical problems. Thus, variables of metabolic flexibility were assessed in n = 14 T2D patients and n = 13 CON. All other variables were assessed in n = 14 T2D patients and n = 14 CON.