Deficiency of histone variant macroH2A1.1 is associated with sexually dimorphic obesity in mice

Abstract

Obesity has a major socio-economic health impact. There are profound sex differences in adipose tissue deposition and obesity-related conditions. The underlying mechanisms driving sexual dimorphism in obesity and its associated metabolic disorders remain unclear. Histone variant macroH2A1.1 is a candidate epigenetic mechanism linking environmental and dietary factors to obesity. Here, we used a mouse model genetically depleted of macroH2A1.1 to investigate its potential epigenetic role in sex dimorphic obesity, metabolic disturbances and gut dysbiosis. Whole body macroH2A1 knockout (KO) mice, generated with the Cre/loxP technology, and their control littermates were fed a high fat diet containing 60% of energy derived from fat. The diet was administered for three months starting from 10 to 12 weeks of age. We evaluated the progression in body weight, the food intake, and the tolerance to glucose by means of a glucose tolerance test. Gut microbiota composition, visceral adipose and liver tissue morphology were assessed. In addition, adipogenic gene expression patterns were evaluated in the visceral adipose tissue. Female KO mice for macroH2A1.1 had a more pronounced weight gain induced by high fat diet compared to their littermates, while the increase in body weight in male mice was similar in the two genotypes. Food intake was generally increased upon KO and decreased by high fat diet in both sexes, with the exception of KO females fed a high fat diet that displayed the same food intake of their littermates. In glucose tolerance tests, glucose levels were significantly elevated upon high fat diet in female KO compared to a standard diet, while this effect was absent in male KO. There were no differences in hepatic histology. Upon a high fat diet, in female adipocyte cross-sectional area was larger in KO compared to littermates: activation of proadipogenic genes (ACACB, AGT, ANGPT2, FASN, RETN, SLC2A4) and downregulation of antiadipogenic genes (AXIN1, E2F1, EGR2, JUN, SIRT1, SIRT2, UCP1, CCND1, CDKN1A, CDKN1B, EGR2) was detected. Gut microbiota profiling showed increase in Firmicutes and a decrease in Bacteroidetes in females, but not males, macroH2A1.1 KO mice. MacroH2A1.1 KO mice display sexual dimorphism in high fat diet-induced obesity and in gut dysbiosis, and may represent a useful model to investigate epigenetic and metabolic differences associated to the development of obesity-associated pathological conditions in males and females.

Figure 1

Effects of high fat diet (HFD) in Fl/Fl and macroH2A1.1 KO mice. (A) Basal body weight (in g) of mice at the beginning of the experiments. (B) Body weight gain (percentage increase with respect to basal body weight) of mice under standard diet (SD) or HFD in the Fl/Fl or KO background, divided by sex. (C) Food intake per day (g) in mice under standard diet (SD) or HFD in the Fl/Fl or KO background, divided by sex. Data are presented as means ± S.E.M (n = 6–12 mice/group). ^*,#p < 0.05, ***^,###P < 0.001, ****p < 0.0001, comparison as indicated, determined by unpaired t-test or two-way ANOVA test.

Figure 2

Responsiveness to glucose in Fl/Fl and macroH2A1.1 KO mice, according to the sex. Glucose tolerance test (GTT) was performed in Fl/Fl and macroH2A1.1 KO mice fed a standard diet (SD) or a high fat diet (HFD) following a 6 h fast. Mice were treated with glucose 2 g/kg, i.p., and blood glucose concentrations were measured at different time points (A–B) and AUC (C–D). Data are expressed as mean ± S.E.M. (n = 6–12 mice/group). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 and #p < 0.05, unpaired t-test or two-way ANOVA test.

Figure 3

Liver histological analysis. Upper panels: representative pictures from hematoxylin and eosin (H&E) staining of liver sections around the lobular areas in Fl/Fl and macroH2A1.1 KO mice fed with standard diet (SD) or with high fat diet (HFD). (A) Males Fl/Fl SD; (B) Males Fl/Fl HFD; (C) Females Fl/Fl SD; (D) Females Fl/Fl HFD; (E) Males macroH2A1.1 KO SD; (F) Males macroH2A1.1 KO HFD; (G) Females macroH2A1.1 KO SD; (H) Females macroH2A1.1 KO HFD. Magnification 200x, scale bar 50 µm. Lower panel: NAFLD and inflammation were scored using a semiquantitative system that grouped histological features into broad categories (steatosis, hepatocellular injury, portal inflammation, fibrosis and miscellaneous features) in males (left panel) and in females (right panel). Data are expressed as mean ± S.E.M (n = 6–8 mice/group). One-way ANOVA followed by Tukey post-hoc test, *p < 0.05, **p < 0.01, ***p < 0.001, HFD condition versus respective SD condition (i.e. same sex or same genotype).

Figure 4

White adipose tissue histological analysis. Upper panels: representative pictures from hematoxylin and eosin (H&E) staining of white adipose tissue sections of Fl/Fl and macroH2A1.1 KO mice fed with standard diet (SD) or with high fat diet (HFD). (A) Males Fl/Fl SD; (B) Males Fl/Fl HFD; (C) Females Fl/Fl SD; (D) Females Fl/Fl HFD; (E) Males macroH2A1.1 KO SD; (F) Males macroH2A1.1 KO HFD; (G) Females macroH2A1.1 KO SD; (H) Females macroH2A1.1 KO HFD. Magnification 400x, scale bar 20 µm. (I) Histograms showed the adipocytes area evaluation obtained by Zeiss Software. The adipocytes area is expressed in µm². Data are expressed as mean ± S.E.M. (n = 6–8 per group). One-way ANOVA followed by Tukey post-hoc test, ***p < 0.001, **p < 0.01, HFD condition versus respective SD condition (i.e., same sex, same genotype); ^$p < 0.05 versus female Fl/Fl HFD.

Figure 5

Taxonomic diversity between the gut microbiota of macroH2A1.1 KO versus control (Fl/Fl, males and females). (A) The panel represents the alpha diversity at family, order and genus level in the different groups. Observed and Simpson indexes calculate the alpha diversity in terms of richness (number of taxa that are present in the sample). (B) Multidimensional scaling (MDS) Ordination on Bray–Curtis Index. (C) Hierarchical Clustering on Bray–Curtis Index. (B) and (C) are two graphical representations of the beta diversity, to visualize the global level of similarity between the sample bacterial profiles based on the Bray–Curtis method.

Figure 6

Taxonomic diversity between the gut microbiota of macroH2A1.1 KO versus control (Fl/Fl, males and females). The figures show the taxa relative abundance per sample at Phylum level in male (A) and female (B) mice, for the two genotypes control (Fl/Fl) and macroH2A1.1 KO. The top 15 bacterial Phyla are visualized. (C) Taxa relative abundance per sample at the Family level in female mice, for the two genotypes control (Fl/Fl) and macroH2A1.1 KO. The top 15 bacterial phyla are visualized. (D) LefSe result cladogram (genus and above levels) indicates the bacterial taxa that are significantly different between control (Fl/Fl) and macroH2A1.1 KO female mice (p value < 0.05, Wilcoxon–Mann–Whitney test).