Effects of Soy Isoflavones, Resistant Starch and Antibiotics on Polycystic Ovary Syndrome (PCOS)-Like Features in Letrozole-Treated Rats

Abstract

Polycystic ovary syndrome (PCOS) is the most common endocrine disorder in reproductive-aged women. Recently, various dietary interventions have been used extensively as a novel therapy against PCOS. In the present study, we show that soy isoflavone metabolites and resistant starch, together with gut microbiota modulations, were successful in decreasing the severity of PCOS-like reproductive features while increasing the expression of gut barrier markers and butyric acid in the gut. In the letrozole-induced PCOS model rats, the intake of both 0.05% soy isoflavones and 11% resistant starch, even with letrozole treatment, reduced the severity of menstrual irregularity and polycystic ovaries with a high concentration of soy isoflavones and equol in plasma. Antibiotic cocktail treatment suppressed soy isoflavone metabolism in the gut and showed no considerable effects on reducing the PCOS-like symptoms. The mRNA expression level of occludin significantly increased with soy isoflavone and resistant starch combined treatment. Bacterial genera such as Blautia, Dorea and Clostridium were positively correlated with menstrual irregularity under resistant starch intake. Moreover, the concentration of butyric acid was elevated by resistant starch intake. In conclusion, we propose that both dietary interventions and gut microbiota modulations could be effectively used in reducing the severity of PCOS reproductive features.

Keywords: PCOS; antibiotics; gut microbiota; resistant starch; soy isoflavones.

Figure 1

Schematic diagram of the animal experiment. Five-week-old female Sprague Dawley (SD) rats were used for the experiment. After acclimatization and the prevaginal sampling period, rats were divided into seven groups (n = 8, 7 weeks old). All animals were housed in individual cages. PCOS symptoms were induced using oral gavage of 0.5 mg/kg letrozole for 21 days simultaneously with the diet treatments. The control group was treated with 1% CMC. Control and L groups were given the control diet. The LS group was given a 0.05% soy isoflavone-based diet. The LR group was given an 11% resistant starch (RS)-based diet. The LSR group was given a mixture of both. The PA group was given the control diet along with an antibiotic cocktail (Ab). The PSA group was given the antibiotic cocktail with the control diet. CMC = carboxymethyl cellulose.

Figure 2

Analyses of menstrual irregularity and polycystic ovaries. (A) Number of cycles present during the diet treatment period. This analysis was performed to measure menstrual irregularity. Vaginal cell suspensions were collected daily and stained using Modified Wright–Giemsa stain to identify the cycle stage on each day. The number of cycles was counted during the 21-day diet treatment period. (B) Cystic follicle count and (C) corpora lutea count were measured using ovarian histological analysis. Excised ovaries were embedded in paraffin blocks, and the sections were stained using hematoxylin–eosin stain. All the values are expressed as the mean ± S.E. Statistical analysis was conducted by one-way ANOVA followed by Dunnett’s test for multiple comparison analysis. a, b represent p < 0.05 compared to the C group, * represents p < 0.05 compared to the L group.

Figure 3

Photomicrographs of ovarian sections of a representative rat from each study group. Excised ovaries were embedded in paraffin blocks, and the sections were stained using hematoxylin–eosin stain. (A) C group, (B) L group, (C) LS group, (D) LR group, (E) LSR group, (F) LA group, (G) LSA group. Red arrowheads = corpora lutea; black arrowheads = cystic follicles; scale bars = 200 μm.

Figure 4

Analyses of reproductive hormones. Biochemical analyses of reproductive hormones; (A) testosterone, (B) estradiol and pituitary hormones; (C) LH and (D) FSH. These hormones were measured in the serum samples of the rats using the respective ELISA kits. All the values are expressed as the mean ± S.E. Statistical analysis was conducted by one-way ANOVA followed by Dunnett’s test for multiple comparison analysis. a, b represent p < 0.05 compared to the C group.

Figure 5

Daidzein and equol concentration analyses. Soy isoflavone metabolite concentrations in plasma were analyzed via HPLC. (A,B) show daidzein and equol concentrations, respectively. All the values are expressed as the mean ± S.E. Statistical analysis was conducted by one-way ANOVA followed by Dunnett’s test for multiple comparison analysis. a, b represent p < 0.05 compared to the C group, * represent p < 0.05 compared to the L group.

Figure 6

Analyses of gut microbial profiles. The cecal content of each rat, which was collected at the end of the experiment, was used to analyze the gut microbiota composition using 16S rRNA sequencing. Alpha diversity analysis by (A) Faith’s PD index and (B) Shannon index. (C) Beta diversity analysis by Bray–Curtis method. (D) Relative abundances of significantly different bacterial genera. All the values are expressed as the mean ± S.E. Statistical analysis was conducted by Kruskal–Wallis test followed by Tukey–Kramer test for multiple comparison analysis. a, b represent p < 0.05 compared to the C group, * represent p < 0.05 compared to the L group.

Figure 7

Analyses of gut barrier markers. (A) Relative abundance of microbial functional pathway responsible for bacterial invasion of epithelial cells. This analysis was performed by evaluating the differences in the function of the microbial communities ascertained from 16S rRNA sequencing using PICRUSt software. Relative mRNA expression of gut barrier markers; (B) claudin-2 and (C) occludin. The expression levels of the genes in the colon samples were measured using quantitative RT-PCR. GAPDH was used as a loading control to normalize each sample. All the values are expressed as the mean ± S.E. Statistical analysis of bacterial sequences was conducted by Kruskal–Wallis test followed by Tukey–Kramer test for multiple comparison analysis. Statistical analysis of other parameters was conducted by one-way ANOVA followed by Dunnett’s test for multiple comparison analysis. a, b represent p < 0.05 compared to the C group, * represent p < 0.05 compared to the L group.

Figure 8

Analyses of SCFAs. (A) Relative abundance of microbial functional pathway responsible for butanoate metabolism. This analysis was performed by evaluating the differences in the function of the microbial communities ascertained from 16S rRNA sequencing using PICRUSt software. (B) Butyric acid concentration. Concentrations of the SCFAs were measured using the cecal content samples. Supplementary Figure S5 shows the concentrations of acetic acid and propionic acid. All the values are expressed as the mean ± S.E. Statistical analysis was conducted by one-way ANOVA followed by Dunnett’s test for multiple comparison analysis. Statistical analysis of bacterial sequences was conducted by Kruskal–Wallis test followed by Tukey–Kramer test for multiple comparison analysis. Statistical analysis of other parameters was conducted by one-way ANOVA followed by Dunnett’s test for multiple comparison analysis. a, b represent p < 0.05 compared to the C group, * represent p < 0.05 compared to the L group.