Effects of calcium carbonate and CS intervention on physiological parameters in low-calcium diet fed rats

The schematic diagram of the animal experiment is shown in Fig. 1A and the low-calcium base feed formula is shown in Table 1. After 12 weeks of intervention, femoral shaft weight, bone calcium content, and bone mineral density (BMD) were evaluated in the three groups (Fig. 1B–D). There was no significant difference in femoral shaft weight between low calcium group (group C) and normal control group (group N), cartilage sulfate group (group CS) and calcium carbonate group (group Ca) (P > 0.05). On the other hand, The calcium content of femur, bone mineral density at the midpoint of femur and distal femur in group C were significantly lower than those in groups N, Ca and CS (0.05 and 0.01, respectively). (P < 0.05, P < 0.01). The calcium content of femur, bone mineral density of midpoint of femur and bone mineral density of distal femur in CS group were not lower than those in Ca group. HE staining showed that compared with group C, bone mineral density was significantly increased in group CS and Ca (Fig. 1E).

Fig. 1

Animal experiment plan and effects of various interventions on the femur of the rats. A Animal experiment plan. The changes of femoral shaft weight (B), bone calcium content (C), and bone mineral density (D) of the femurs from the rats subjected to the various interventions. HE staining showed bone mineral density of rats (as shown by arrow) in different groups (E). All data is represented as Mean ± Standard Deviation. All data were accessed using One-way ANOVA with Tukey’s multiple comparisons test. *P < 0.05 indicate the significant difference

Structural changes in intestinal microflora induced by calcium carbonate and CS intervention

A total of 3866 operational taxonomic units (OTUs) were found in the N group and the C group, with 838 and 1233 OTUs specific to the two groups, respectively (Fig. 2A). Individual sparse curves approached saturation platforms, indicating high sampling coverage (over 99%) in all samples (Fig. 2B). The Shannon and Simpson index of α diversity was significantly different in the N group and the C groups (Fig. 2C, D). There were 3621 OTUs in the C group and the CS group, with 1051 and 593 OTUs specific to the two groups, respectively (Fig. 2A). The Shannon and Simpson index did not differ between the two groups (Fig. 2C, D). There were 3456 OTUs in the CS and C groups, with 1102 OTUs in the CS group and 886 OTUs in the C group (Fig. 2A). There were significant differences in the Shannon and Simpson indexes in the two groups (Fig. 2C, D). This suggests that calcium has a stronger ability to reduce intestinal flora richness and α diversity under CS intervention.

Fig. 2

Taxonomic Composition of Intestinal Microbiota. A Venn diagram of C and N groups and C and CS groups and CS and Ca groups at OTU level. Individual sparse curves (B), Shannon index (C), Simpson index (D). E PCA Plot. F PCoA Plot. The abscissa represents the first principal component (PC1), the ordinate represents the second principal component (PC2). F Weighted UniFrac PCoA diagram. G Cluster analysis of intestinal microbiota between samples. H Composition of intestinal microbiota at phylum level. (I) Composition of intestinal microbiota at genus level. J at the genus level (P < 0.05, or P < 0.01) different gut microbiota

In terms of β diversity, the similarity of the bacterial communities in the three groups is shown in Fig. 2. Group N was distinguished from group C and group N from group CS according to the weighted UniFrac PCA score (Fig. 2E). Meanwhile, the weighted UniFrac PCoA score distinguished the CS group from the N group (Fig. 2F). An unweighted pair group method with arithmetic mean and clustering tree of weighted UniFrac was used to cluster the differences between the samples (Fig. 2G). The multiple response permutation procedure analysis showed significant differences in microbial community structure between the N group and the C group, and between the CS group and the Ca group (Additional file 1: Table S1). Therefore, calcium has a strong ability to affect the diversity of intestinal microbiota under CS intervention but cannot completely restore it.

Figure 2H, I show the composition of intestinal flora. The top 10 phyla and genera in the most abundant microbiome of all participants are shown in Fig. 2H, I, and included Bacteroidota, Firmicutes, Proteobacteria, Actinobacteriota, Campilobacterota, Spirochaetota, unidentified_Bacteria, Cyanobacteria, Desulfobacterota, Acidobacteriota, which were dominant in all microbial communities. Compared with the N group, the C group showed reduced Firmicutes, Spirochaetota, unidentified_Bacteria, and Desulfobacterota and increased Bacteroidota. CS intervention resulted in significant changes in intestinal microbial community structure. Specifically, Megasphaera was increased with CS intervention compared with the C group. CS intervention elevated Bifidobacterium levels to be higher than the Ca group, and Anaerobiospirillum, Alloprevotella, Quinella, Christensenellaceae_R-7_group, Allobaculum, Bacteroides, Colidextribacter, NK4A214_group,unidentified_[Eubacterium]_coprostanoligenes_group, Marvinbryantia, Oscillibacter, Parasutterella, and Lachnospiraceae_ND3007_group were all lower than the Ca group (Fig. 2J).

Effects of low-calcium diet on intestinal microbiota and metabolomics

A linear discriminant analysis effect size (LEfSe) analysis was used to identify potential biomarkers for intestinal flora in the N group and the C group, and a t-test was used to further examine and visualize the results (P < 0.05; Fig. 3A). A total of 22 significant taxa were identified. There were 16 potential biomarkers of intestinal flora in the N group and eight in the C group. On this basis, we analyzed the correlation between 10 intestinal flora biomarkers and 20 fecal and plasma metabolites involved in the KEGG pathway in these two groups. The results showed that there was a strong correlation between intestinal flora and fecal and plasma metabolic pathways (Fig. 3B, C). To assess the global metabolic changes in the C group, fecal metabolites from N and C rats were measured and analyzed by mass spectrometry combined with liquid chromatography (LC–MS/MS). The fecal metabolites in the N group increased by 106 species and decreased by 117 species (Fig. 3D). Figure 3F shows the KEGG bubble diagram of the metabolic pathways involved in the production of different metabolites. Compared with the C group, metabolites in the N group were significantly different, including nucleosides, nucleotides and analogues, organic oxygen compounds, heterocyclic compounds, and organic oxygen compounds. These differentially expressed metabolites were enriched in the KEGG pathway, and they involve vitamin B6 metabolism, lysosome, pantothenic acid and CoA biosynthesis, and the AMPK signaling pathway (Additional file 1: Table S2). These pathways may be assigned to cofactor and vitamin metabolism, transport and catabolism, and signal transduction. Abundances of the 10 kinds of plasma metabolites were higher in the N group than those in the C group, and 18 kinds of plasma metabolites were lower in the N group than those in the C group (Fig. 3E). Figure 3G shows the KEGG bubble diagram. The differentially expressed metabolites in plasma were enriched in the KEGG pathway and allocated to the lysine degradation of amino acid metabolism (Additional file 1: Table S3).

Fig. 3

Effect of low calcium diet on intestinal microbiota and metabolic profiles. A LEfSe analysis of cladogram in group N and group C. B The correlation graph shows the correlation between intestinal microbiota with fecal metabolites. C The correlation graph shows the correlation between intestinal microbiota with plasma metabolites. D The volcano map of fecal metabolites between the N group and C group. E The volcano map of plasma metabolites between the N group and C group. F The KEGG bubble map of fecal metabonomics between the low calcium group and the control group. G The KEGG bubble map of plasma metabolites between the low calcium group and the control group. The larger the Abscissa in the picture, the higher the enrichment of differential metabolites in the pathway. The color of the point represents the P-value value of the hypergeometric test, and the smaller the value is, the greater the reliability of the test is and the more statistically significant it is. The size of the point represents the number of differential metabolites in the corresponding pathway, and the larger the point, the more differential metabolites in the pathway

Effects of calcium carbonate on intestinal microflora and metabolomics

Through 16S rRNA sequencing and metabonomics analysis, we compared potential biomarkers belonging to intestinal flora with differential metabolites in feces and plasma of groups C and Ca to elucidate the mechanism of calcium carbonate action. A total of 6 taxa were identified as significant (Fig. 4A), with Prevotella copri and Actinobacteria significantly present in group C. Concurrently, we analyzed the correlation between 10 intestinal flora biomarkers and 20 fecal and plasma metabolites involved in the KEGG pathway in groups C and Ca. The results showed that there was a strong correlation between intestinal flora and fecal and plasma metabolic pathways (Fig. 4B, C). To assess the global metabolic changes in the Ca group, fecal metabolites from the C and Ca groups were measured and analyzed by LC–MS/MS. The fecal metabolites in group C increased by 42 and decreased by 22 (Fig. 4D). Figure 4E shows the metabolic pathways involved in the production of different metabolites. Compared with Ca group, the changes in the metabolites in group C included up-regulation of AMP, GMP, PRPP and down-regulation of progesterone and 17 alpha-hydroxyprogesterone. These differentially expressed metabolites are enriched in KEGG pathways, including Antifolate resistance, olfactory transduction, cortisol synthesis and secretion, Cushing's syndrome, and the cGMP-PKG signaling pathway (Additional file 1: Table S2). These pathways may be assigned to antitumor, sensory, endocrine, and signal transduction. Group C had 52 more plasma metabolites than group Ca, and 2 plasma metabolites fewer than group Ca (Fig. 4F). Figure 4G shows the KEGG bubble diagram. Calcium carbonate intervention down-regulated plasma nicotinamide, which is involved in the longevity regulating pathways—worm and platelet-activated ADP (Additional file 1: Table S3).

Fig. 4

Calcium carbonate improved intestinal microbiota and metabolic disorders induced by low calcium diet. A LEfSe analysis of cladogram in C group and Ca group with LDA Score larger than 3. B The correlation graph shows the correlation between intestinal microbiota and fecal metabolites in group C and group Ca. C Correlation graph shows the correlation between intestinal microbiota and plasma metabolites in group C and group Ca. D The volcano map of fecal metabolites between the C group and Ca group. E The volcano map of plasma metabolites between the C group and Ca group. F The KEGG bubble map of fecal metabonomics between the C group and Ca group. G The KEGG bubble map of plasma metabolites between the C group and Ca group. All data is represented as Mean ± Standard Deviation

Effects of chondroitin sulfate on intestinal microflora and metabolomics

LEfSe was used to identify biomarkers characterizing dimensional intestinal bacteria and to identify taxonomic groups with significant differences between groups, and to generate a clade map. A total of six and seven taxa were identified as significant in the comparison of the C group and the CS group (Fig. 5A). Specifically, in the comparison between these two groups where the C group showed Clostridia, the number of Coriobacteriaceae and Collinsella and Collinsella_aerofaciens and Coriobacteriales and Coriobacteriia was higher in the CS group. In the comparison between the C group and CS, numbers of Lachnospiraceae, Lachnospirales, Alloprevotella, Aeromonadales, Succinivibrionaceae, Anaerobiospirillum, and Clostridia were higher in the latter group. Concurrently, we analyzed the correlation between 10 intestinal flora biomarkers and 20 fecal and plasma metabolites involved in the KEGG pathway in the C group and CS group. The results showed that there was a strong correlation between intestinal flora and fecal and plasma metabolic pathways (Fig. 5B, C). To assess the global metabolic changes in the CS group, fecal metabolites and plasma metabolites from the C and CS groups were measured and analyzed by LC–MS/MS. Compared with the C group, fecal metabolites in the CS group increased by 11 and decreased by 8 (Fig. 5D). The differentially expressed metabolites were enriched in linoleic acid metabolism, progemediated oocyte maturation, tryptophan metabolism, and purine metabolism in the KEGG pathway (Additional file 1: Table S2). Figure 5F shows the metabolic pathways involved in the production of different metabolites. These pathways may be assigned to lipid metabolism, the endocrine system, amino acid metabolism and nucleotide metabolism. The plasma metabolites in the CS group were 18 more than those in the C group and 21 less than those in the Ca group (Fig. 5E). Figure 5G shows the KEGG bubble diagram. Plasma differentially expressed metabolites are enriched in the KEGG pathway and assigned to the glutathione metabolism of other amino acids (Additional file 1: Table S3). These fecal metabolites and plasma metabolites have a variety of structures, including lipids, fatty acids, and amino acid derivatives.

Fig. 5

Calcium carbonate improved intestinal microbiota and metabolic disorders induced by low calcium diet. A LEfSe analysis of cladogram in C and CS group and CS and Ca group with LDA Score larger than 3. B The correlation graph shows the correlation between intestinal microbiota with fecal metabolites. C The correlation graph shows the correlation between intestinal microbiota with plasma metabolites. D The volcano map of fecal metabolites between the C and CS group and CS and Ca group. E The volcano map of plasma metabolites between the C and CS group and CS and Ca group. F The KEGG bubble map of fecal metabonomics between the C and CS group and CS and Ca group. G The KEGG bubble map of plasma metabolites between the C and CS group and CS and Ca group. All data is represented as Mean ± Standard Deviation

Compared with the Ca group, the number of fecal metabolites in the CS group increased by 49 and decreased by 14 (Fig. 5D). Differentially expressed metabolites are enriched in olfactory transduction and the CGMP-PKG signaling pathway in the KEGG pathway (Additional file 1: Table S2). Figure 5F shows the metabolic pathways involved in the production of different metabolites. These pathways may be assigned to sensory system functions and signal transduction. The plasma metabolites in the CS group were 33 more than those in the Ca group and 5 less than those in Ca group (Fig. 5E). Figure 5G shows the KEGG bubble diagram. The differentially expressed metabolites in plasma were enriched in the KEGG pathway and allocated to biotin metabolism of cofactors and nutrition (Additional file 1: Table S3). These fecal and plasma metabolites have a variety of structures, including lipids, fatty acids, and amino acid derivatives. The metabolic pathways affected by CS intervention are shown in Fig. 6F.

Fig. 6

Effect of chondroitin sulfate intervention on excreta and plasma metabolites in rats. A Differences in fecal stearic acid between normal and different intervention groups. B Differences in plasma testosterone between normal and different intervention groups. C, D Fecal short-chain fatty acid levels in different intervention groups. E Difference level of glutathione between chondroitin sulfate and low calcium group. F The metabolic pathways affected by chondroitin sulfate intervention

We detected 843 compounds from 2200 + compound controls based on the Novo self-established metabolome database. In the analysis of differential metabolites in feces and plasma, the number of metabolites identified in feces was much greater than that in plasma. Among 290 + lipids (100 + oxidized lipids, 190 + lipids), 107 lipids were detected, among which 93 lipids were different, we obtained lipid classification by matching screened DEL with Lipidmaps database (http://www.lipidmaps.org), delete mismatched entries and count the number of DEL that accompany each category, the most highly ranked category is Fatty Acyls (FA). The levels of plasma testosterone and stearic content in normal feeding rats were significantly higher than those in the other two groups (Fig. 6A, B), indicating that the plasma testosterone and stearic content in the feces of normal feeding rats were decreased due to long-term low calcium levels, and the supplementation of calcium and CS could not be recovered. In the fatty acids metabolites, we confirmed differential changes in short-chain fatty acids in feces, but not in plasma. CS intervention resulted in higher levels of absolute oxoadipic acid and lower levels of absolute 5-aminolevulinate and N-methyl-a-aminoisobutyric acid content in the Ca group compared with the C group (Fig. 6C), and lower levels of absolute 4-methylvaleric acid and hydroxyglutaric acid and 2-hydroxy-2-methylbutanedioic acid content in the C group compared with the Ca group (Fig. 6D). However, in plasma metabolomics, we found that CS induced antioxidant activity and reduced the oxidized glutathione content (Fig. 6E).