Alterations in Intestinal Permeability and Bone Density Following Ovariectomy in Rats
To investigate the pivotal role of alterations in intestinal permeability in the bone loss resulting from ovariectomy, we conducted an examination of the pathological changes in both the gut and bone of ovariectomized rats. As illustrated in Fig.1B, compared to the control group, ovariectomy resulted in a slender femur, lighter staining, a reduced number of trabeculae, and significantly increased spacing. The bone trabeculae in the tibial tissue exhibited significant reduction and slimness, accompanied by an increase in fat vacuoles and inflammatory cell infiltration. These findings indicate that ovariectomy induces bone loss and osteoporosis in rats. Furthermore, osteoporosis induced by bilateral ovariectomy did not cause notable changes in the thickness of the ileum's intestinal wall. However, the intestinal villus height (VH) and the ratio of villus height to crypt depth (VH/CD) in the OVX group significantly decreased compared to the control group. The depth of crypt (CD) in the OVX group was significantly increased compared to the control group (Fig.1C-D). In addition, we also measured the concentrations of lactulose and mannitol in rat urine to evaluate the intestinal permeability of ovariectomized rats. The content of lactulose in the urine of ovariectomized rats was significantly higher than that of the control group (P<0.05) (Fig.1E). This data suggests that osteoporosis resulting from ovarian removal significantly impacts the integrity and permeability of the intestinal mucosal barrier. Ovariectomy-induced osteoporosis in rats is accompanied by a significant increase in peripheral blood TNF-α, IL-1, and IL-6 (Fig.1F), possibly linked to alterations in the integrity and permeability of the intestinal mucosal barrier.
Fig.1: Ovariectomy-Induced Osteoporosis in Rats and Its Association with Altered Intestinal Permeability. (A) Experimental design of ovarian removal surgery in female SD Rats (Experiment 1) and Ruminococcus intervention experimental design (Experiment 2). (B) Pathological changes in the femur and tibia observed at 100X (upper) and 200X (lower) magnifications in control and OVX groups. (C) Pathological changes in the ileum at 40X (left), 100X (middle), and 200X (right) magnifications in control and OVX groups. (D) Measurements of the thickness of the ileal wall, villus height, crypt depth, and the ratio of villus height to crypt depth. Data are expressed as the means ± SD. n=6 per group. (E) Lactulose concentration in urine for control and OVX groups. (F) Levels of peripheral blood IL-1, IL-6, and TNF-α in control and OVX groups. The difference between control and OVX groups was analyzed by the student’s t-test. Data are expressed as the means ± SD. n =10 per group (E, F). Exp, experiment; OVX, ovariectomized group; w, weeks.
Ovariectomy-Induced Alterations in Gut Microbiota Composition in Rats
After filtering, quality control, and chimeric removal, a total of 1,054,679 sequences were obtained from 16S rDNA gene sequencing data across 20 fecal samples, averaging 52,733 ± 4,640 sequences per sample (ranging from 45,375 to 62,614). In Fig.2A, the control group exhibits 4926 unique ASVs, the OVX group has 4294 unique ASVs, and there are 1491 shared ASVs between the two groups. To assess sequencing data adequacy and bacterial abundance characteristics, we performed sparse analysis and computed the observed OTU, Chao1, and Shannon indices. The saturation curves for each group approach saturation (Fig.2B), affirming the robustness of the sequencing data with no detection of new species. This implies that ovarian resection did not significantly impact the gut microbiota's α-diversity, species richness, and evenness within the two groups, as they remain comparable. At the genus level, β-diversity was calculated. The principal coordinates analysis (PCoA) results indicated distinct clusters between the control group and OVX rats, with PCoA1 scores at 17.07% and PCoA2 scores at 14.55%, respectively. PCoA has a dominant impact on separation and clustering, whether based on Bray_ Curtis distance or Weighted_ Unifrac distance (Fig.2C). Furthermore, cluster analysis unveiled distinct patterns between the control group and the OVX group, exhibiting robust clustering between the two groups (Fig.2C). At the phylum level, illustrated in Fig.2D, Firmicutes and Bacteroidetes constitute the major components of the gut microbiota, collectively constituting over 90% of the total gut microbiota in both the control and OVX groups. Comparatively, Firmicutes exhibited a declining trend in the OVX group, while Bacteroidetes demonstrated an increasing trend compared to the control group (Fig.2F). Upon conducting 16S rDNA gene amplicon sequencing analysis at the genus level, it was observed that, relative to the control group, the abundance of g_Ruminococcus and g_Lachnospiraceae_NK4A136_group significantly decreased in the OVX group, whereas the abundance of g_UCG-005 increased (Fig.2E and G). Among these three altered genera, g_Ruminococcus exhibited the most significant changes. Therefore, we speculate that the decreased abundance of g_Ruminococcus may contribute to osteoporosis in estrogen-deficient rats. Further exploration of the differential bacteria in g_Ruminococcus is warranted.
Fig.2 Ovariectomy can lead to dysbiosis of the gut microbiota in rats. (A) Venn diagram analysis comparing the control and OVX groups. (B) Rarefaction curves illustrating alpha richness in the control and OVX groups. (C) Scatter plots depicting the Principal Coordinates Analysis (PCoA) of fecal microbial communities at the genus level, utilizing Bray-Curtis distance and Weighted UniFrac distance. (D) Composition of the gut microbiome at the phylum level. (E) Composition of the gut microbiome at the genus level. (F) Phylum-level composition in female rats, presented in bar graph format. (G) Genus-level composition in female rats, displayed in bar graph format. Data are expressed as means ± SD, n = 10 per group (A-G). Treat, OVX group.
After Ovariectomy: Impact on BMSCs Osteogenic and Adipogenic Differentiation Abilities
BMSCs were isolated from OVX rats and observed under light microscopy. Primary BMSCs cultures from the control and OVX groups showed densely clustered cells, with spindle-shaped cells extending from the central region. Continuous media exchange revealed comparable cell morphologies between the OVX and control groups, both displaying whirlpool-like growth patterns (Fig. 3A). Although the control group exhibited a slightly higher growth rate and density under the same culture conditions, flow cytometry analysis of BMSCs surface markers (CD34, CD29, CD45, CD90) indicated positive expression for CD29 and CD90, and negative expression for CD23 and CD45 (Fig. S1), confirming the identity of the isolated cells as BMSCs. Upon osteogenic induction in both OVX and control group BMSCs, Alizarin Red and Oil Red O staining were performed at different time points during osteogenic induction. Additionally, osteogenic- and adipogenic-related genes, including Runx2, Bmp2, Osterix, Tgf-β, Wnt3A, and Ppar-γ, were assessed for expression. During osteo-induction, spindle-shaped cells transitioned into irregular shapes, forming clustered structures in both OVX and control group BMSCs. Alizarin red staining revealed the development of calcium nodules, with the control group exhibiting higher numbers and density of calcium nodules compared to the OVX group within the same induction cycle (14 or 21 days) (Fig.3B). Oil Red staining demonstrated a significant increase in intracellular lipid droplet formation on the 14th day of adipogenic induction, with lipid droplet numbers gradually increasing over time. By the 21st day, lipid droplets fused with each other. Notably, BMSCs in the OVX group displayed higher numbers and density of lipid droplets compared to the control group under the same induction period (Fig.3C). Gene expression analysis indicated that, compared to the control group, ppar γ expression significantly increased, while Tgf-β, Wnt, RunX2, and Osterix expression levels significantly decreased in the OVX group after 14 and 21 days of osteogenic induction (Fig.4C and E). Similarly, during adipogenic induction, Ppar γ expression significantly increased, while Bmp2, Tgf-β, Wnt, Runx2, and Osterix expression levels significantly decreased in the OVX group compared to the control group after 14 and 21 days (P<0.05).
Fig.3 Influence of ovarian removal on osteogenic and adipogenic differentiation capacities in rat BMSCs. (A) Representative images of primary BMSCs from P0-P3. (B) Representative images of alizarin red S staining of BMSCs at 100X (upper) and 200X (lower) after 14-day and 21-day osteogenic induction. (C) Representative images of oil red staining of BMSCs after 14 day and 21 day adipogenic induction. (D-E) Wnt3A, Bmp2, Tgf-β, Runx2 Osterix and Pparγ mRNA expressions of rat BMSCs after 14 day and 21day osteogenic or adipogenic induction, from control and OVX group, respectively. n = 6.
Characterization and miRNA Profiling of Exosomes from BMSCs
The systems by which an altered microbiota can participate in the progression of osteoporosis diseases are manifold, and, among these, a fundamental moment could be constituted by the ability of the microbiome to intervene in the expression and functioning of miRNAs. Following the isolation and culture of exosomes derived from BMSCs in ovariectomized rats, the morphology of the exosomes was examined using transmission electron microscopy (TEM), revealing a distinctive cup-shaped structure (Fig. 4A). The size analysis indicated that these spherical nanoparticles had an approximate size of 75.89 ± 15.39 nm, consistent with TEM findings (Fig. 4B). Flow cytometry analysis demonstrated a significant increase in surface markers, including CD9 and CD81, on the isolated nanoparticles compared to BMSCs (Fig. 4C), confirming the successful isolation of exosomes from BMSCs. Subsequent miRNA sequencing and bioinformatics analysis were conducted to unveil the comprehensive miRNA expression profile of BMSCs-derived exosomes. Table S1 shows the differential expression of miRNAs in the exosomes of osteoporotic rat BSMCs. A total of 28 miRNAs were differentially expressed, with 25 up-regulated and 3 down-regulated miRNAs (P < 0.05), as depicted in Fig.4D and E. Significant miRNA differences (|log2FC| > 1, P < 0.05) were summarized. Two miRNAs with the significant differential expression were selected for further analysis, miR-23b-3p exhibiting the significant upregulation, and miR-151-3p showing the significant downregulation. Moreover, KEGG pathway annotation classification and enrichment analysis were performed on the upregulated genes. Enrichment analysis was conducted using the phyper function in R software to calculate the P-value, followed by FDR correction to obtain the Q-value. Pathways significantly enriched in differentially expressed genes encompassed the oxytocin signaling pathway, MAPK signaling pathway, Phospholipase D signaling pathway, and PI3K-AKT signaling pathway. Examination of the enriched diseases within these pathways revealed associations with bone metabolism balance, osteoporosis, osteogenesis imperfecta, and osteosclerosis. These findings suggest that the exosomal miR-23b-3p and miR-151-3p may exert their influence on the osteogenic differentiation of BMSCs through modulation of the oxytocin, MAPK signal pathway, Phospholipase D, and PI3K-Akt signaling pathways.
Fig.4 Sequencing and bioinformatics analysis of Exosomal miRNA in BMSCs. (A) Morphology of BMSCs exosomes. (B) Diameter distribution of BMSCs exosomes. (C) exosomal marker analysis. (D) Volcano plot of differential miRNA analysis. (E-F) The miRNA expressions of BMSCs exosomes after 14 day and 21day osteogenic or adipogenic induction, from control and OVX group, respectively. n = 6. (G) KEGG enrichment analysis of genes in the BMSCs exosomal miRNA.
The Influence of Ruminococcus on Osteogenic Differentiation in Ovariectomized Rats
Numerous experiments have also demonstrated the presence and actions of a microbiota–bone axis capable of inducing the onset of osteoporotic disease. The systems by which an altered microbiota can participate in the progression of osteoporosis diseases are manifold, and, among these, a fundamental moment could be constituted by the ability of the microbiome to intervene in the expression and functioning of miRNAs. To further identify variations in the gut microbiome triggered by estrogen deficiency, we conducted LDA effect size (LEfSe) analysis to determine differences between the control and OVX groups. The results revealed a significant decrease in the relative abundance of g_Lachnospiraceae_NK4A136 and g_Ruminococcus, while g_UCG_005 showed a significant increase in the OVX group compared to the control group (Fig.5A and B). The histological examination of mouse colon tissue sections through HE staining revealed notable changes in the ileum. In comparison to the control, the OVX group exhibited a significant reduction in intestinal villus height (VH) and VH/CD, accompanied by inflammatory cell infiltration, although there were no significant alterations in intestinal wall thickness and crypt depth. Contrastingly, the OVX+Ruminococcus group displayed marked increases in intestinal villus height (VH), crypt depth (CD), intestinal wall thickness, and VH/CD, coupled with a reduction in inflammatory cells (Fig. 5C and D). The ELISA results of peripheral blood in each group demonstrated elevated levels of IL-1, IL-6, and TNF-α in the OVX group compared to the control group. Although the levels of these inflammatory markers in the OVX+Ruminococcus group were significantly higher than those in the control group after ingestion of Ruminococcus, they were significantly reduced compared to the OVX group (P<0.05) (Fig.5E). The observed pathological changes in the intestine might be associated with heightened peripheral blood inflammation. Bone density testing uncovered a significant decrease in bone mineral density (BMD) values in the OVX group compared to the control group. Notably, after Ruminococcus ingestion, the BMD value of the OVX group mice showed a slight increase (P<0.05), and the bone density value became comparable to that of the control group (Fig.5F). Further examination of the femur and tibia in rats after ovariectomy revealed a decrease in bone trabeculae, increased fat vacuoles, and inflammatory cells. However, following the administration of Ruminococcus, there was an improvement in bone trabeculae, a reduction in fat vacuoles, and inflammatory cells in the femur and tibia tissues (Fig.5G and H). Moreover, induced osteogenic and adipogenic differentiation of BMSCs for 14 and 21 days revealed distinct outcomes. Alizarin red staining illustrated that the osteogenic differentiation degree of the OVX group was significantly hindered compared to the control group. In contrast, Ruminococcus ingestion in OVX rats significantly enhanced the degree of osteogenic differentiation, indicating a notable alleviating effect on the inhibition caused by ovariectomy (Fig.5I). Furthermore, oil red staining demonstrated that the degree of adipogenic differentiation in the OVX group was notably elevated compared to the control group. Treatment with Ruminococcus substantially mitigated the degree of adipogenic differentiation in ovariectomized rats, underscoring its significant alleviating effect on the increase induced by OVX (J). Finally, the examination of genes related to osteogenesis and adipogenic differentiation unveiled significant increases in miR-23b-3p and PPARγ expression in the OVX group compared to the control group. However, following Ruminococcus intervention, the expression of miR-23b-3p and Pparγ was reduced. Ovariectomy in rats significantly diminished miR-151-3p, Bmp2, and Tgf-β expression, as well as Wnt3A, Runx2, and Osterix. Ruminococcus induced an elevation in miR-151-3p, Bmp2, Tgf-β, Wnt3A, Runx2, and Osterix expression (Fig.S2). The intervention of Ruminococcus alleviated the changes in gene expression related to osteogenesis and adipogenesis induced by osteoporosis.
Fig.5 Ruminococcus administration improves osteoporosis caused by ovarian removal in rats. (A) The taxonomic representation of statistically and biologically differences between OVX group and control. (B) Pathological changes in the ileum at 100X (upper), and 200X (lower) magnifications after Ruminococcus administration. (D) Measurements of the thickness of the ileal wall, villus height, crypt depth, and the ratio of villus height to crypt depth. Data are expressed as the means ± SD. n=6 per group. (E) Levels of peripheral blood IL-1, IL-6, and TNF-α after Ruminococcus administration. Data are expressed as the means ± SD. n=10 per group. (F) Detection of bone density in rats after administration of Ruminococcus. (G-H) Pathological changes in the femur and tibia observed at 100X (upper) and 200X (lower) magnifications after Ruminococcus administration. (I) Representative images of alizarin red S staining of BMSCs at 100X (upper) and 200X (lower) after 14-day and 21-day osteogenic induction. (J) Representative images of oil red staining of BMSCs at 100X (upper) and 200X (lower) after 14-day and 21-day osteogenic induction.