Gut bacterial and fungal dysbiosis in tuberculosis patients

Abstract

Background: Recent studies have more focused on gut microbial alteration in tuberculosis (TB) patients. However, no detailed study on gut fungi modification has been reported till now. So, current research explores the characteristics of gut microbiota (bacteria)- and mycobiota (fungi)-dysbiosis in TB patients and also assesses the correlation between the gut microbiome and serum cytokines. It may help to screen the potential diagnostic biomarker for TB.

Results: The results show that the alpha diversity of the gut microbiome (including bacteria and fungi) decreased and altered the gut microbiome composition of TB patients. The bacterial genera Bacteroides and Prevotella were significantly increased, and Blautia and Bifidobacterium decreased in the TB patients group. The fungi genus Saccharomyces was increased while decreased levels of Aspergillus in TB patients. It indicates that gut microbial equilibrium between bacteria and fungi has been altered in TB patients. The fungal-to-bacterial species ratio was significantly decreased, and the bacterial-fungal trans-kingdom interactions have been reduced in TB patients. A set model including Bacteroides, Blautia, Eubacterium_hallii_group, Apiotrichum, Penicillium, and Saccharomyces may provide a better TB diagnostics option than using single bacterial or fungi sets. Also, gut microbial dysbiosis has a strong correlation with the alteration of IL-17 and IFN-γ.

Conclusions: Our results demonstrate that TB patients exhibit the gut bacterial and fungal dysbiosis. In the clinics, some gut microbes may be considered as potential biomarkers for auxiliary TB diagnosis.

Keywords: Dysbiosis; Gut microbiota; Gut mycobiota; IFN-γ; IL-17; Tuberculosis.

Fig. 1

Comparison of TB patients’ intestinal bacterial diversity and composition versus healthy controls. (a) Simpson index and Chao 1 index represent the alpha diversity (Mann-Whitney U test). (b) The beta diversity was assessed using PCoA, which is based on Unweighted unifrac distance. ANOSIM was used to analyze the differences in the bacterial community between the TB group and the HC group. Comparisons of the relative abundances of intestinal bacteria between the TB and HC groups were conducted at the phylum (c) and genus (d) levels (Wilcoxon Rank Sum test). The LEfSe analysis (e) identified the differentially abundant taxa between the TB patients and HC (LDA > 4, P < 0.05). The taxa enriched in the HC were characterized by a positive LDA score (red), and the TB-enriched taxa were indicated with a negative LDA score (blue). The circular cladogram (f) is a taxonomic diagram showing the taxonomic hierarchy of the signified species in patients with TB and HC by LEfSe. TB: tuberculosis; HC: healthy control. *P < 0.05; **P < 0.01; ***P < 0.001

Fig. 2

Comparison of TB patients’ gut fungal diversity and composition versus healthy controls. (a) Simpson index and Chao 1 index represent the alpha diversity (Unpaired t-tests). (b) The beta diversity was assessed using PCoA analysis based on Unweighted unifrac distance. ANOSIM was used to analyze the differences in the fungal community between the TB group and the HC group. Comparisons of the relative abundances of intestinal fungus between the TB and HC groups were conducted at the phylum (c) and genus (d) levels (Wilcoxon Rank Sum test). (e) The LEfSe analysis identified the differentially abundant taxa between the TB patients and HC (LDA > 4, P < 0.05). (f) The circular cladogram is a taxonomic diagram showing the taxonomic hierarchy of the signified species in patients with TB and HC by LEfSe. PCoA, principal coordinate analysis; LEfSe, linear discriminant analysis effect size; LDA, linear discriminant analysis; TB: tuberculosis; HC: healthy control. *P < 0.05; **P < 0.01; ***P < 0.001

Fig. 3

Fungal-bacterial equilibration analyses in two groups. (a) The ITS2/16S diversity ratio was calculated by the Sobs at the genus level (Mann-Whitney U test). Trans-kingdom abundance correlation networks of HC (b) and TB (c) groups at the genus level using the Spearman coefficient. Each node represents a genus, and the color of the nodes represents the phylum to which it belongs. Node’s shape represents the kingdom to which they belong (square is fungus and circle is bacteria). The size of the node represents the mean abundance of the genus. Edge represents the positive correlations (blue) and negative correlations (red). ****P < 0.0001

Fig. 4

ROC analysis and functional prediction of gut microbiome. ROC analysis of (a) bacteria, (b) fungi, and (c) the combination of fungi and bacteria. Correlation between predicted differential pathways and the top 20 differential bacteria (d) and fungi (e). The depth of color in the heat map indicates the strength of the correlation: red indicates a positive correlation, and blue indicates a negative correlation. ROC: Receiver operating characteristic; AUC: area under the curve (0.5–1.0). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 5

Correlation analysis between intestinal flora and serum cytokine. (a) Compared cytokine levels of IFN-γ and IL-17 in serum between two groups ( Mann-Whitney U test). (b) Correlation analysis was performed between the relative abundance of bacteria and the levels of IFN-γ and IL-17. (c) Spearman was used to evaluate the correlation between fungus and IFN-γ and IL-17. The depth of color in the heat map indicates the strength of the correlation: red indicates a positive correlation, while blue indicates a negative correlation. *P < 0.05; **P < 0.01; ***P < 0.001

Fig. 6

Linear Regression analysis between diagnostics biomarkers and serum cytokine. (a) Linear Regression between the relative abundance of Bacteroides, Blautia, and Eubacterium_hallii_group with IFN-γ and IL-17 expression. (b) Linear Regression between the relative abundance of Apiotrichum, Penicillium, and Saccharomyces with IFN-γ and IL-17 expression