Estradiol and high fat diet associate with changes in gut microbiota in female ob/ob mice

Abstract

Estrogens protect against diet-induced obesity in women and female rodents. For example, a lack of estrogens in postmenopausal women is associated with an increased risk of weight gain, cardiovascular diseases, low-grade inflammation, and cancer. Estrogens act with leptin to regulate energy homeostasis in females. Leptin-deficient mice (ob/ob) exhibit morbid obesity and insulin resistance. The gut microbiome is also critical in regulating metabolism. The present study investigates whether estrogens and leptin modulate gut microbiota in ovariectomized ob/ob (obese) or heterozygote (lean) mice fed high-fat diet (HFD) that received either 17β-Estradiol (E2) or vehicle implants. E2 attenuated weight gain in both genotypes. Moreover, both obesity (ob/ob mice) and E2 were associated with reduced gut microbial diversity. ob/ob mice exhibited lower species richness than control mice, while E2-treated mice had reduced evenness compared with vehicle mice. Regarding taxa, E2 was associated with an increased abundance of the S24-7 family, while leptin was associated with increases in Coriobacteriaceae, Clostridium and Lactobacillus. Some taxa were affected by both E2 and leptin, suggesting these hormones alter gut microbiota of HFD-fed female mice. Understanding the role of E2 and leptin in regulating gut microbiota will provide important insights into hormone-dependent metabolic disorders in women.

Figure 1

Estradiol and leptin alter weight gain and food intake in adult female ob/ob and lean control mice. (A) Average weight, (B) percent weight gain and (C) food intake (2 mice/cage) of heterozygote controls (Het) E2 (n = 13), Het Veh (n = 12), ob/ob E2 (n = 8) and ob/ob Veh (n = 7) with arrow indicating the start of high fat diet. The effects of genotype on average weight (A) and food intake (C) were present on all days. Days with effects of E2 are denoted by *(p < 0.05; ANOVA). Error bars indicate ± SEM.

Figure 2

Estradiol and obesity reduce gut microbiota alpha diversity. (A) ob/ob genotype is associated with lower species richness as measured by Observed Taxonomic Units. (B) E2 treatment is associated with lower species evenness as measured by the Pielou’s evenness index. “a” and “b” indicate groups with different species richness while “c” and “d” indicate groups with different species evenness (p < 0.05, linear regression). Het E2 (n = 6), Het Veh (n = 6), ob/ob E2 (n = 6) and ob/ob Veh (n = 6).

Figure 3

Gut microbial communities distinctly cluster as an effect of estradiol treatment and leptin. PCo1 and PCo2 clustering of each group. Bray-Curtis distance was used to calculate principal coordinates 1 and 2 and PERMANOVA, to calculate the association with leptin or E2 on aggregate data over all days. Ob E2 = E2-treated ob/ob (n = 6), Ob V = Vehicle ob/ob (n = 6), Het E2 = E2-treated Het (n = 6), and Het V = Vehicle Het mice (n = 6).

Figure 4

Gut microbiota associate with estradiol treatment and obesity at multiple taxa levels. Microbiota community structure at the (A) phylum, (B) family and (C) genus level, separated by estradiol treatment and ob/ob genotype. Data are shown as relative proportion of the taxa identified. Taxa with prevalence of >10% or with a maximum proportion of >0.2% were included. Het E2 (n = 6), Het Veh (n = 6), ob/ob E2 (n = 6), and ob/ob Veh mice (n = 6). An overdispersed Poisson regression was used to calculate associations of taxa with leptin and E2.

Figure 5

Estradiol and obesity associate with changes in the gut microbial composition. (A) Phyla and lower level taxa associated with estradiol (E2) or vehicle (Veh) treatment. (B) Phyla and lower level taxa associated with ob/ob or Het genotype. Het E2 (n = 6), Het Veh (n = 6), ob/ob E2 (n = 6) and ob/ob Veh (n = 6).

Figure 6

Gut microbiota associate with estradiol treatment and obesity at multiple taxa levels. Boxplots showing microbial taxa that are significantly altered by (A) Estradiol (E2) treatment and (B) genotype (overdispersed Poisson regression). Analysis was done on aggregate data from all 6 day points. Data are shown as relative proportion of the taxa identified. Taxa with prevalence of >10% or with a maximum proportion of >0.2% were included. E2-treated ob/ob (n = 6), E2-treated Het (n = 6), Veh ob/ob (n = 6), vehicle Het mice (n = 6).

Figure 7

Estradiol and obesity associate with changes in the relative abundances of Bacteroidetes (phylum) and S24-7 (family) following the start of HFD. Relative abundances of the (A) the phylum Bacteroidetes and (B) its family S24-7 over time. *Indicates effects of E2 and # indicates effects of genotype (q < 0.1, overdispersed Poisson regression). Het E2 (n = 6), Het Veh (n = 6), ob/ob E2 (n = 6) and ob/ob Veh (n = 6). Error bars indicate ± SEM.

Figure 8

Estradiol and obesity associate with changes in the relative abundances of the phylum Firmicutes and its lower taxa. Relative abundances over time of the (A) phylum Firmicutes and genera (B) Clostridium, (C) Lactobacillus and (D) Coprococcus. *Indicates effects of E2 and # indicates effects of genotype (q < 0.1, overdispersed Poisson regression). Het E2 (n = 6), Het Veh (n = 6), ob/ob E2 (n = 6) and ob/ob Veh (n = 6). Error bars indicate ± SEM.

Figure 9

Estradiol and obesity resist increases in the relative abundances of the phylum Actinobacteria and its lower taxa. Relative abundances of the (A) phylum Actinobacteria and (B) family Coriobacteriaceae over time. *Indicates effects of E2 and # indicates effects of genotype (q < 0.1 overdispersed Poisson regression). Het E2 (n = 6), Het Veh (n = 6), ob/ob E2 (n = 6) and ob/ob Veh (n = 6). Error bars indicate ± SEM.