Correlation between gut microbiota and metabolomics under intermittent hypoxic conditions

Abstract

Objective: This study investigates the role of gut microbiota and metabolites as biomarkers in the diagnosis and assessment of Obstructive Sleep Apnea (OSA) by analyzing the correlation between microbiota and metabolomics in the host's gut under conditions of chronic intermittent hypoxia (CIH).

Methods: We analyzed the composition and function of gut microorganisms and metabolites in OSA and CIH mouse models using 16S rRNA sequencing combined with targeted metabolomics to evaluate the association between clinical indicators and severity of OSA and gut flora and metabolites.

Results: Compared with normal controls, we found: (1) The apnea hypoventilation index (AHI) and fasting blood glucose were higher in the moderate and severe OSA groups, the eosinophilia score (EP) and neutrophil percentage were significantly higher in the severe OSA group, and the percentage of lymphocytes in the blood count and platelet pressures were lower in the moderate group (P < 0.05).(2) The OSA and CIH groups showed an increase in the abundance of Firmicutes and Tenotrophomonas, and a decrease in the abundance of Bacteroidota and Ligilactobacillus (P < 0.05).(3) The OSA and CIH groups had elevated Nervonic acid and Erucic acid and decreased Arachidonic acid (P < 0.05). In addition, correlation analysis showed that: There was a negative correlation between Coprobacillus and percentage of lymphocytes, and Bacteroidota and EP and AHI;and a positive correlation between Lactococcus and Neutrophils, Arachidonic acid and width of erythrocyte distribution, and Erucic and Neuronic acids and AHI (P < 0.05).

Conclusion: The CIH condition can lead to microbial and metabolite alterations, which correlate with disease severity and clinical indicators, and is expected to be a new biomarker for the diagnosis and assessment of severity of OSA.

Keywords: Biomarkers; Chronic Intermittent Hypoxia; Disease Evaluation; Gut microbiota; Targeted metabolomics.

Fig. 1

Flowchart of the study

Fig. 2

A, B Microbial genus and phylum levels in the gut of the population. E, F Relative abundance differences of the top 20 species at the level of microbial genera and phylum in the gut of the mouse. C, D The NC group vs. patients with OSA—Phylogenetic evolutionary tree and between-groups abundance distribution heat and Venn diagrams. G, H The CIH mice vs.NC group—Phylogenetic evolutionary tree and between-groups abundance distribution heat and Venn diagram. I The NC group vs. patients with OSA -DESeq2 volcano diagram. K, J The NC group vs. patients with OSA—LDA histogram and cladogram diagram. l The CIH mice vs.NC group-DESeq2 volcano diagram. N.M. The CIH mice vs.NC group—LDA value distribution histogram and cladogram graph. High-resolution versions of all subfigures are provided in Supplementary Files (Fig. 2)

Fig. 3

A and C Alpha diversity analysis and Principal Coordinate Analysis (PCoA) comparing the NC group vs. patients with OSA. B and D Alpha diversity analysis and PCoA comparing the CIH mice vs.NC group. E KEGG Level 3 pathway abundance profiles: the NC group vs. patients with OSA. F KEGG Level 3 pathway abundance profiles: the CIH mice vs. NC group. High-resolution versions are available in Supplementary Files (Fig. 3)

Fig. 4

A Stacked bar chart showing the top 20 metabolites in the NC group vs. patients with OSA. B Heatmap of metabolite profiles. C and D Top 15 metabolites ranked by importance in Random Forest analysis and results from Support Vector Machine (SVM) analysis. E OPLS-DA score plot. G OPLS-DA permutation test results showing the distribution of the test statistic and p-value. F PLS-DA score plot. H PLS-DA permutation test results showing the distribution of the test statistic and p-value. I Volcano plot based on T-test results. J ANOVA boxplot. High-resolution versions are available in Supplementary Files (Fig. 4). SVM: Support Vector Machine. PLS-DA: Partial Least Squares Discriminant Analysis. OPLS-DA: Orthogonal Partial Least Squares Discriminant Analysis. ANOVA: Analysis of Variance

Fig. 5

A Stacked bar chart of the top 20 metabolites in the CIH group vs. NC group. B Heatmap of metabolite profiles. C Top 15 metabolites ranked by importance in Random Forest analysis. D SVM classification results. E OPLS-DA score plot. F PLS-DA score plot. G Volcano plot of statistical significance. H ANOVA boxplot comparing metabolite abundance across groups. High-resolution versions are available in Supplementary Files (Fig. 5). SVM: Support Vector Machine. PLS-DA: Partial Least Squares Discriminant Analysis. OPLS-DA: Orthogonal Partial Least Squares Discriminant Analysis. ANOVA: Analysis of Variance

Fig. 6

A Heatmap of correlations between targeted metabolites identified in the human gut and clinical parameters. B Microbial-metabolite correlation network diagram and associated heatmap. High-resolution versions are available in Supplementary Files (Fig. 6)