Transient Dietary Intervention Induces Healthy Adipose Tissue Expansion and Metabolically Healthy Obesity in Mice

Abstract

As obesity progresses, dynamic tissue remodeling of adipose tissue occurs over time, that is, adipocyte hypertrophy, chronic inflammation, and interstitial fibrosis. Some obese individuals exhibit healthy adipose tissue expansion, characterized by modest inflammation and fibrosis despite adipocyte hypertrophy, resulting in "Metabolically Healthy Obesity (MHO)". In this study, we investigated the effects of transient weight loss on adipose tissue remodeling during the development of obesity. Male C57BL6/J mice received various types of transient weight loss treatments during diet-induced obesity. A 2-week weight loss intervention during the inflammatory phase promoted healthy adipose tissue expansion, reduced ectopic lipid accumulation, and improved glucose metabolism. In contrast, protocols with shorter duration and delayed intervention, failed to induce MHO. Since serum concentrations of ketone bodies were elevated during weight loss, we examined the effects of hyperketonemia on obesity-induced adipose tissue remodeling. Transient treatment with 1,3-butanediol (BD), which increased serum ketone body concentrations to levels similar to those observed during weight loss, induced healthy adipose tissue expansion and reduced hepatic steatosis even during continuous high-fat diet (HFD) feeding. Ketone bodies effectively suppressed activation of adipose tissue fibroblasts in vivo and in vitro. This study provides evidence that an appropriate dietary intervention can promote healthy adipose tissue expansion in mice, even after the regaining of weight, thereby leading to MHO. As the underlying mechanism, our data revealed a key role for ketone bodies in suppressing activation of adipose tissue fibroblasts. This study paves the way for nutritional approaches to induce MHO.

Keywords: chronic inflammation; fibroblast; fibrosis; ketone body; obesity; weight‐cycling.

FIGURE 1

Diet‐induced weight‐cycling induces metabolically healthy obesity in mice. (A) Schematic diagram of weight‐cycling model 1 (WC1) and its control (Ctrl). The WC1 mice were fed high‐fat diet (HFD) for 8 weeks, followed by standard diet (SD) for 2 weeks for weight loss, then HFD again for additional 6 weeks, whereas the Ctrl mice were fed HFD continuously for 14 weeks. n = 6 per group. (B) Body weight curves. (C) Tissue weights and the percentages of tissue weight relative to body weight. (D) Blood glucose levels and serum parameters. (E, F) RNA‐sequencing analysis of the epididymal adipose tissue. n = 2 per group. (E) Clustering analysis (k) and (F) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis (k = 2). (G) Serum concentrations of branched‐chain amino acids (BCAA) and branched‐chain α‐keto acids (BCKA). (H) mRNA expression levels in epididymal adipose tissue. Data are expressed as means ± SEM. *p < 0.05 and **p < 0.01 by Student's t test.

FIGURE 2

WC leads to healthy adipose tissue expansion and reduced hepatic steatosis. (A) Histogram of adipocyte diameters in epididymal adipose tissue from WC1 and Ctrl mice. n = 10–11. (B) Representative images of F4/80 immunostaining and the number of crown‐like structures (CLSs) in epididymal adipose tissue. (C) Representative images of Sirius red staining and quantification of the Sirius red–positive area in epididymal adipose tissue. (D) Hydroxyproline content in epididymal adipose tissue. n = 4 per group. (E) mRNA expression levels in liver. (F) Representative images of hematoxylin and eosin staining in liver. (G) Hepatic triglyceride content. (H) Insulin tolerance test (lt.) and glucose tolerance test (rt.). Serum insulin concentrations were measured before and 15 min after glucose administration during the glucose tolerance test. n = 6 per group. Data are expressed as means ± SEM. *p < 0.05 and **p < 0.01 by Student's t test. Scale bars, 100 μm.

FIGURE 3

Other WC protocols show differential effects on adipose tissue. (A–I) Weight‐cycling model 2 (WC2). (A) Schematic diagram of WC2 and its control (Ctrl). The WC2 mice underwent two 1‐week periods of SD feeding with a 2‐week HFD period in between, whereas the Ctrl mice were fed HFD continuously for 14 weeks. n = 6 per group. (B) Body weight curves. (C) Tissue weights. (D) Serum insulin concentrations. (E) Histogram of adipocyte diameters in epididymal adipose tissue. (F) The number of crown‐like structures (CLSs) in epididymal adipose tissue. (G) Quantification of the Sirius red–positive area in epididymal adipose tissue (H) mRNA expression levels in liver. (I) Hepatic triglyceride content. (J–R) Weight‐cycling model 3 (WC3). (J) Schematic diagram of WC3 and Ctrl. After 11 weeks of HFD feeding, WC3 mice underwent a 2‐week weight loss intervention. n = 5–6. (K) Body weight curves. (L) Tissue weights. (M) Serum insulin concentrations. (N) Histogram of adipocyte diameters in epididymal adipose tissue. (O) The number of CLSs in epididymal adipose tissue. (P) Quantification of the Sirius red–positive area in epididymal adipose tissue (Q) mRNA expression levels in liver. (R) Hepatic triglyceride content. Data are expressed as means ± SEM. *p < 0.05 and **p < 0.01 by Student's t test.

FIGURE 4

Gene expression profiles in adipose tissue during weight loss. (A) Schematic diagram of the experimental protocol. n = 5–6. (B) Body and tissue weights. “Pre” represents the mice fed HFD for 8 weeks. Weight loss 1 (“WL1”) and “WL2” represent the mice fed SD for 1 and 2 weeks after 8 weeks of HFD feeding, respectively. (C) Blood glucose levels and serum parameters, including β‐hydroxybutyrate (βHB) and hydroxyproline. (D–G) RNA‐sequencing analysis of epididymal adipose tissue during weight loss. n = 3 per group. (D) Principal component (PC) analysis and (E) K‐means clustering analyses (k = 5). (F) Gene enrichment analysis of cluster E. (G) Gene set enrichment analysis results of collagen biosynthesis and modifying enzymes. Data are expressed as means ± SEM. *p < 0.05 and **p < 0.01 by Tukey–Kramer test.

FIGURE 5

Adipose tissue fibrosis is partially reversed during weight loss. Mice were fed HFD for 8 weeks (HFD 8 w), then fed HFD or SD for an additional 2 weeks (HFD 10 w and weight loss [WL], respectively). (A) mRNA expression levels in epididymal adipose tissue. (B) mRNA expression levels in fibroblasts isolated from epididymal adipose tissue. (C) Histogram of adipocyte diameters in epididymal adipose tissue. (D, E) Representative images of Sirius red staining (D) and α‐smooth muscle actin (αSMA) immunostaining (E) in epididymal adipose tissue and their quantification. (F) Western blot analysis of type I and VI collagen in epididymal adipose tissue. Data are expressed as means ± SEM. *p < 0.05 and **p < 0.01 versus HFD 8 w; ^† p < 0.05 and ^†† p < 0.01 versus HFD 10 w by Student's t test. n = 4–6. Scale bars, 100 μm.

FIGURE 6

Treatment with β‐hydroxybutyrate suppresses obesity‐induced adipose tissue fibrosis. (A) Blood βHB concentrations in standard diet‐fed mice treated with 20% 1,3‐butanediol (BD) or vehicle (Ctrl). n = 3 per group. (B) Schematic diagram of the experimental protocol. During the last 2 weeks of a 10‐week HFD, mice were treated with BD or vehicle (Ctrl). n = 9 per group. (C) Body and tissue weights. (D) Serum parameters. (E) mRNA expression levels in epididymal adipose tissue. (F) mRNA expression levels in liver. (G, H) Representative images of Sirius red staining (G) and αSMA immunostaining (H) in epididymal adipose tissue and their quantification. (I) FACS gating for identification of CD45(+), F4/80 (+) (macrophages) cells in the SVF of epididymal adipose tissue (lt.). Proportion of macrophages in the SVF (rt.). n = 5 per group. (J) mRNA expression levels in fibroblasts isolated from epididymal adipose tissue. n = 9 per group. (K) Effect of βHB (5 mM) on mRNA expression levels in adipose tissue fibroblasts stimulated with TGFβ. n = 4. Data are expressed as means ± SEM. *p < 0.05 and **p < 0.01 by Student's t test (A, C–J) or analysis of variance with Tukey–Kramer test (K). Scale bars, 100 μm.

FIGURE 7

Transient BD administration during HFD feeding is sufficient to induce healthy adipose tissue expansion. (A) Schematic diagram of the experimental protocol. Mice received a transient treatment of 1,3‐butanediol (BD) for 2 weeks during a continuous 14‐week HFD feeding. n = 12–15. (B) Body and tissue weights, and the percentages of tissue weights relative to body weight. Ctrl, control (C) Serum parameters. (D) Histogram of adipocyte diameters in the epididymal adipose tissue. (E) Representative images of F4/80 immunostaining and the number of crown‐like structures (CLSs) in epididymal adipose tissue. (F) Representative images of Sirius red staining and quantification of the Sirius red–positive area in epididymal adipose tissue. (G) Representative images of hematoxylin and eosin staining in the liver. (H) Hepatic triglyceride content. Data are expressed as means ± SEM. *p < 0.05 and **p < 0.01 by Student's t test. Scale bars, 100 μm.