Conditional Vitamin D Receptor Deletion Induces Fungal and Archaeal Dysbiosis and Altered Metabolites

Abstract

A vitamin D receptor (VDR) deficiency leads to the dysbiosis of intestinal bacteria and is associated with various diseases, including cancer, infections, and inflammatory bowel disease. However, the impact of a VDR deficiency on fungi and archaea is unknown. We conditionally deleted the VDR in Paneth cells (VDR^ΔPC), intestinal epithelial cells (VDR^ΔIEC), or myeloid cells (VDR^ΔLyz) in mice and collected feces for shotgun metagenomic sequencing and untargeted metabolomics. We found that fungi were significantly altered in each knockout (KO) group compared to the VDR^Loxp control. The VDR^ΔLyz mice had the most altered fungi species (three depleted and seven enriched), followed by the VDR^ΔPC mice (six depleted and two enriched), and the VDR^ΔIEC mice (one depleted and one enriched). The methanogen Methanofollis liminatans was enriched in the VDR^ΔPC and VDR^ΔLyz mice and two further archaeal species (Thermococcus piezophilus and Sulfolobus acidocaldarius) were enriched in the VDR^ΔLyz mice compared to the Loxp group. Significant correlations existed among altered fungi, archaea, bacteria, and viruses in the KO mice. Functional metagenomics showed changes in several biologic functions, including decreased sulfate reduction and increased biosynthesis of cobalamin (vitamin B12) in VDR^ΔLyz mice relative to VDR^Loxp mice. Fecal metabolites were analyzed to examine the involvement of sulfate reduction and other pathways. In conclusion, a VDR deficiency caused the formation of altered fungi and archaea in a tissue- and sex-dependent manner. These results provide a foundation about the impact of a host factor (e.g., VDR deficiency) on fungi and archaea. It opens the door for further studies to determine how mycobiome and cross-kingdom interactions in the microbiome community and metabolites contribute to the risk of certain diseases.

Keywords: Inflammatory Bowel Disease (IBD); archaeome; cobalamin; innate immunity; metabolite; microbiome; mycobiome; sulfate reduction; vitamin; vitamin B12; vitamin D receptor.

Figure 1

Species-level alpha and beta diversity via different tissue deletions of the VDR. (A) Distribution of Shannon diversity metrics at the species level, divided by sex, for fungi. Color indicates genotype. Subsequent ANOSIM analysis on sex subgroups revealed no statistically significant comparisons (p > 0.05). (B) Within-group versus between-group Bray–Curtis dissimilarity metrics for mice of both sexes. Color indicates genotype. Subsequent ANOSIM analysis found all control-knockout pairs to be significantly different. These results, VDR^ΔIEC vs. VDR^Loxp (p = 0.005), VDR^ΔPC vs. VDR^Loxp (p = 0.001), and VDR^ΔLyz vs. VDR^Loxp (p = 0.001), are indicated on the boxplot.

Figure 2

Differential fungi in the KO mice. Fungal taxa that are significantly (q, # ≤ 0.1, * ≤ 0.05, ** ≤ 0.01, and *** ≤ 0.001) differentially abundant in VDR KO mice relative to the control group, not sub-dividing by sex. “s_” indicates that the taxa is a species and “g_” indicates that it is a genus—the only two taxonomic levels that were studied. LFC indicates log2 fold change, and the color indicates the direction of the shift relative to the control (blue is decreased, and red is increased). (A) VDR^ΔPC, (B) VDR^ΔIEC, and (C) VDR^ΔLyz.

Figure 3

Significant (p ≤ 0.05) correlations between significantly differentially abundant (q ≤ 0.1) species of fungi and significantly differentially abundant (q ≤ 0.1) bacteria, archaea, and viruses. The cell color indicates the magnitude of the correlation (blue = negative; red = positive), while a white space indicates no significant correlation. The colors of the row names indicate the following: peach for archaea, blue for bacteria, green for eukaryotes, and purple for viruses. (A) VDR^ΔPC. (B) VDR^ΔLyz.

Figure 4

Differential fungi in sex subsets of the VDR^ΔPC group. Fungal taxa that are significantly (q # ≤ 0.1, * ≤ 0.05, ** ≤ 0.01, and *** ≤ 0.001) differentially abundant in the VDR knockout mice relative to the control mice via sex subset. Figure (A) shows the genera and figure (B) shows the species—the only two taxonomic levels that were observed. LFC indicates the log2 fold change, and the color indicates the direction of the shift relative to the control (blue is decreased, and red is increased). Purple font indicates that a taxa was found to be differentially abundant in both the male and female groups.

Figure 5

Differential fungi in sex subsets of the VDR^ΔIEC group. Fungal taxa that are significantly (q # ≤ 0.1, * ≤ 0.05, ** ≤ 0.01, and *** ≤ 0.001) differentially abundant in the VDR knockout mice relative to the control mice via sex subset. Figure (A) shows the genera and figure (B) shows the species—the only two taxonomic levels that were observed. LFC indicates the log2 fold change, and the color indicates the direction of the shift relative to the control (blue is decreased, and red is increased). Purple font indicates that a taxa was found to be differentially abundant in both the male and female groups.

Figure 6

Differential fungi in sex subsets of the VDR^ΔLyz group. Fungal taxa that are significantly (q # ≤ 0.1, * ≤ 0.05, ** ≤ 0.01, and *** ≤ 0.001) differentially abundant in the VDR knockout mice relative to the control mice via sex subset. Figure (A) shows the genera and figure (B) shows the species—the only two taxonomic levels that were observed. LFC indicates the log2 fold change, and the color indicates the direction of the shift relative to the control (blue is decreased, and red is increased). Purple font indicates that a taxa was found to be differentially abundant in both the male and female groups.

Figure 7

Species-level alpha and beta diversity in the VDR KO mice compared to the VDR^Loxp control group. (A) Shannon diversity by sample. Presented are the p-values of the subsequent ANOSIM analysis on the sex subgroups that were statistically significant (p ≤ 0.05) or marginally significant (p ≤ 0.1). (B) Within-group versus between-group Bray–Curtis dissimilarity metrics for mice of both sexes. Color indicates genotype. A subsequent ANOSIM analysis found all the control–knockout pairs to be significantly different. These results, VDR^ΔIEC vs. VDR^Loxp (p = 0.012), VDR^ΔPC vs. VDR^Loxp (p = 0.002), and VDR^ΔLyz vs. VDR^Loxp (p = 0.001), are indicated on the boxplot.

Figure 8

Differential archaea and their correlation maps. (A) Archaeal taxa that are significantly (q ≤ 0.1, * ≤ 0.05, and *** ≤ 0.001) differentially abundant in the VDR KO mice relative to the control group, not splitting by sex. “s_” indicates that the taxa is a species and “g_” indicates it is a genus—the only two taxonomic levels that were used as inputs. LFC indicates the log2 fold change, and the color indicates the direction of the shift relative to the control (blue is decreased, and red is increased). (B) Significant (p ≤ 0.05) correlations between significantly (q ≤ 0.1) differentially abundant species of archaea and significantly (q ≤ 0.1) differentially abundant bacteria, fungi, and viruses. The color indicates the magnitude of the correlation (blue = negative; red = positive), while a white space indicates no significant correlation. The colors of the row names indicate the following: peach for archaea, blue for bacteria, green for eukaryotes, and purple for viruses.

Figure 9

Differential archaea in sex subsets. Archaeal taxa that are significantly (q ≤ 0.1, * ≤ 0.05, ** ≤ 0.01, and *** ≤ 0.001) differentially abundant in VDR knockout mice. (A) VDR^ΔPC, (B) VDR^ΔIEC, and (C) VDR^ΔLyz in male and female mice. “s_” indicates that the taxa is a species and “g_” indicates that it is a genus—the only two taxonomic levels that were observed. LFC indicates the log2 fold change, and the color indicates the direction of the shift relative to the control (blue is decreased, and red is increased).

Figure 10

Predictive functional analysis using metagenomics. (A) Genes, (B) pathways, and (C) modules that are significantly (q ≤ 0.1, * ≤ 0.05, ** ≤ 0.01, and *** ≤ 0.001) increased in the KO vs. control mice. The magnitude of this difference is given as the log2 fold change (LFC), and the color indicates the direction of the shift relative to the control.

Figure 11

Metabolites related to sulfate reduction. (A) The primary metabolites involved in dissimilatory and sulfate reduction, along with key downstream molecules [62,63]. (B) The primary metabolites involved in the cobalamin salvage pathway [64]. (C) The difference in the abundances of metabolites from the above pathways detected in the VDR^ΔLyz mice relative to the VDR^Loxp mice. LFC indicates the Log2 fold change. Gray indicates a lack of statistical significance (p ≤ 0.05), while the presence of a significant difference is given by non-gray color and * annotation. A # indicates marginal significance (p ≤ 0.1).

Figure 12

The working model of VDR deletion and altered archaea, fungi, bacteria, viruses, and metabolites.