Inflammation-linked adaptations in dermal microvascular reactivity accompany the development of obesity and type 2 diabetes

Abstract

Background/objectives: The increased prevalence of obesity has prompted great strides in our understanding of specific adipose depots and their involvement in cardio-metabolic health. However, the impact of obesity on dermal white adipose tissue (dWAT) and dermal microvascular functionality remains unclear. This study aimed to investigate the temporal changes that occur in dWAT and dermal microvascular functionality during the development of diet-induced obesity and type 2 diabetes in mice.

Methods: Metabolic phenotyping of a murine model of hypercaloric diet (HCD)-induced obesity and type 2 diabetes was performed at three time points that reflected three distinct stages of disease development; 2 weeks of HCD-overweight-metabolically healthy, 4 weeks of HCD-obese-prediabetic and 12 weeks of HCD-obese-type 2 diabetic mice. Expansion of dWAT was characterized histologically, and changes in dermal microvascular reactivity were assessed in response to pressure and the vasodilators SNP and Ach.

Results: HCD resulted in a progressive expansion of dWAT and increased expression of pro-inflammatory markers (IL1β and COX-2). Impairments in pressure-induced (PIV) and Ach-induced (endothelium-dependent) vasodilation occurred early, in overweight-metabolically healthy mice. Residual vasodilatory responses were NOS-independent but sensitive to COX inhibition. These changes were associated with reductions in NO and adiponectin bioavailability, and rescued by exogenous adiponectin or hyperinsulinemia. Obese-prediabetic mice continued to exhibit impaired Ach-dependent vasodilation but PIV appeared normalized. This normalization coincided with elevated endogenous adiponectin and insulin levels, and was sensitive to NOS, COX and PI3K, inhibition. In obese-type 2 diabetic mice, both Ach-stimulated and pressure-induced vasodilatory responses were increased through enhanced COX-2-dependent prostaglandin response.

Conclusions: We demonstrate that the development of obesity, metabolic dysfunction and type 2 diabetes, in HCD-fed mice, is accompanied by increased dermal adiposity and associated metaflammation in dWAT. Importantly, these temporal changes are also linked to disease stage-specific dermal microvascular reactivity, which may reflect adaptive mechanisms driven by metaflammation.

Fig. 1

Effect of HCD on the development of obesity and metabolic syndrome. Mice were fed either hypercaloric diet (filled circles and bars) or standard chow (open circles and bars) for 2, 4 or 12 weeks. a Body weight of mice during the 12-week feeding programme. b Calorie intake of control and HCD fed mice. c Subcutaneous white adipose tissue (ScWAT) weights. d Epididymal white adipose tissue (EpiWAT) weights. e Circulating leptin levels. f Circulating adiponectin (AdipoQ) levels. g Area under the curve (AUC) of glucose levels during glucose tolerance tests (IP-GTT) and h Insulin tolerance tests (IP-ITT). See also Supplemental Fig. 1. Data represents mean ± SEM (n = 10 in each group or n = 50 in each group for body weight) *p < 0.05, **p < 0.01, ***p < 0.001 vs. age-matched control diet fed mice. ^†p < 0.05, ^††p < 0.01, ^†††p < 0.001 vs. 2-week control diet fed mice

Fig. 2

Effect of HCD on dermal adipose remodeling and inflammation. a Representative photomicrographs of immunohistochemical sections from indicated groups following staining with hematoxylin and eosin, F4/80 or anti-IL1-beta. (Scale bars: 500 µm (low mag) and 25 µm (high mag).) b Thickness of dermal adipose tissue was determined relative to the combined papillary dermis and reticular dermis thickness. For each of five mice, 10 randomly selected measurements were calculated from histological images and measured using image J. Data represents mean ± SEM (n = 5 in each group) *p < 0.05, **p < 0.01, ***p < 0.001 vs. age-matched control diet fed mice. c Effects of age and HCD on COX-2 expression in mouse skin

Fig. 3

Time-dependent effects of HCD on vascular reactivity. Maximal percent increase in skin laser Doppler flowmetry (LDF) was determined in response to a mild pressure, b iontophoretic delivery of SNP and c Ach. Data represents mean ± SEM (n = 10 in each group); *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 vs. age-matched control diet fed mice; ^†p < 0.05 vs. 2-week control diet fed mice

Fig. 4

Mechanisms involved in impaired PIV and Ach responsiveness in overweight metabolically healthy mice. Mice were fed either a standard chow (C; white or light grey bars) or hypercaloric diet (HCD; black or dark grey bars) for 2 weeks. Effects of selected pharmacological inhibitors were determined on microvascular response to local pressure application (a, b) or Ach stimulation (c, d). Inhibitors included PI3-kinase inhibitor, wortmannin (W; n = 10), nitric oxide synthase inhibitor, L-NG-nitro-L-arginine (LNNA; n = 10) and LNNA + anti-inflammatory COX-1/2 inhibitor, indomethacin (LNNA + Indo; n = 5). e Effects in vivo administration of insulin (Ins; n = 10) and adiponectin (AdipoQ; n = 9) on impaired PIV response in 2-week-HCD fed mice. f Effects in vivo administration of insulin (Ins; n = 6) on impaired Ach-induced response in 2-week-HCD fed mice. Data represents mean ± SEM. ^†p < 0.05, ^††p < 0.01 vs. 2-week-C diet fed mice; *p < 0.05, **p < 0.01 vs. age-matched control diet fed mice; and ^§p < 0.05, ^§§p < 0.001, ^§§§§p < 0.00001 vs. 2-week hypercaloric diet fed mice

Fig. 5

Mechanisms involved in normalized PIV and impaired Ach responsiveness in obese pre-diabetic mice. Mice were fed either hypercaloric diet (HCD; black or dark grey bars) or standard chow (C; white or light grey bars) for 4 weeks. Effects of selected pharmacological inhibitors were determined on microvascular response to local pressure application (a, b) or Ach stimulation (c, d). Inhibitors included PI3-kinase inhibitor, wortmannin (W; n = 7), nitric oxide synthase inhibitor, L-NG-nitro-L-arginine (LNNA; n = 10) and LNNA + anti-inflammatory COX-1/2 inhibitor, indomethacin (LNNA + Indo; n = 5). Data represents mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 vs. age-matched mice fed the same diet but not treated with any inhibitor

Fig. 6

Mechanisms involved in enhanced PIV and ACh responsiveness in obese diabetic mice. Mice were fed either hypercaloric diet (HCD; black or dark grey bars) or standard chow (C; white or light grey bars) for 12 weeks. Effects of selected pharmacological inhibitors were determined on microvascular response to local pressure application (a, b) or Ach stimulation (c, d). Inhibitors included PI3-kinase inhibitor, wortmannin (W; n = 10), nitric oxide synthase inhibitor, L-NG-nitro-L-arginine (LNNA) and anti-inflammatory COX-1/2 inhibitor, indomethacin (Indo; n = 7) and SC5812 (SC; n = 7). Data represents mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 vs. age-matched mice fed the same diet but not treated with any inhibitor