Roles of oral microbiota and oral-gut microbial transmission in hypertension

Abstract

Introduction: Considerable evidence has linked periodontitis (PD) to hypertension (HTN), but the nature behind this connection is unclear. Dysbiosis of oral microbiota leading to PD is known to aggravate different systematic diseases, but the alteration of oral microbiota in HTN and their impacts on blood pressure (BP) remains to be discovered.

Objectives: To characterize the alterations of oral and gut microbiota and their roles in HTN.

Methods: We performed a cross-sectional (95 HTN participants and 39 controls) and a 6-month follow-up study (52 HTN participants and 26 controls) to analyze the roles of oral and gut microbiota in HTN. Saliva, subgingival plaques, and feces were collected for 16S rRNA gene sequencing or metagenomic analysis. C57BL/6J mice were pretreated with antibiotics to deplete gut microbiota, and then transplanted with human saliva by gavage to test the impacts of abnormal oral-gut microbial transmission on HTN.

Results: BP in participants with PD was higher than no PD in both cross-sectional and follow-up cohort. Relative abundances of 14 salivary genera, 15 subgingival genera and 10 gut genera significantly altered in HTN and those of 7 salivary genera, 12 subgingival genera and 6 gut genera significantly correlated with BP. Sixteen species under 5 genera were identified as oral-gut transmitters, illustrating the presence of oral-gut microbial transmission in HTN. Veillonella was a frequent oral-gut transmitter stably enriched in HTN participants of both cross-sectional and follow-up cohorts. Saliva from HTN participants increased BP in hypertensive mice. Human saliva-derived Veillonella successfully colonized in mouse gut, more abundantly under HTN condition.

Conclusions: PD and oral microbiota are strongly associated with HTN, likely through oral-gut transmission of microbes. Ectopic colonization of saliva-derived Veillonella in the gut may aggravate HTN. Therefore, precise manipulations of oral microbiota and/or oral-gut microbial transmission may be useful strategies for better prevention and treatment of HTN.

Keywords: Blood pressure; Gut microbiota; Hypertension; Oral microbiota; Oral-gut axis; Veillonella.

Graphical abstract

Fig. 1

Participants with PD have higher blood pressure. A-C, SBP of all participants (A), non-hypertensive participants (B), and hypertensive participants (C) with or without PD. D-F, DBP of all participants (D), non-hypertensive participants (E), and hypertensive participants (F) with or without PD. G-H, SBP and DBP of individuals with or without PD in the follow-up study. n = 59:75 for A and D, 23:16 for B and E, 36:59 for C and F, and 34:44 for G and H. Student’s t-test was used for statistical analysis in A-F and Two-way ANOVA followed by Sidak’s multiple comparison test in G and H.

Fig. 2

Shifts of microbiota diversity in oral cavity and intestine of participants with HTN. A, Chao1, Faith’s phylogenetic diversity (Faith’s), Shannon index, and Pielou’s evenness (Pielou’s) of oral (saliva and subgingival plaques) microbiota and gut (feces) microbiota in participants with or without HTN. B, Rarefaction curves of oral and gut microbiota in participants with or without HTN. C, PCoA of oral and gut microbiota in participants with or without HTN. D, Bray-Curtis distance of oral and gut microbiota in participants with or without HTN. All microbiota was analyzed using 16S rRNA gene sequencing. n = 39:94 for saliva, 39:93 for subgingival plaques, and 24:52 for feces. Kruskal-Wallis rank sum test was used for statistical analysis in A and permutational multivariate analysis of variance in D.

Fig. 3

Compositional alterations of oral and gut microbiota in participants with HTN. A. Stacked bar plots showing relative abundances of oral (saliva and subgingival plaques) microbiota and gut (feces) microbiota at phylum level in participants with or without HTN. B. Box plots showing differential enrichment of the top 5 phyla in participants with or without HTN. C. Clustering heatmaps showing relative abundances of the top 50 genera in participants with or without HTN. All microbiota was analyzed using 16S rRNA gene sequencing. n = 39:94 for saliva, 39:93 for subgingival plaques, and 24:52 for feces. Mann-Whitney U rank sum (MW) test was used for statistical analysis in B and Kruskal-Wallis rank sum test in C.

Fig. 4

Associations between oral/gut microbiota and blood pressure and other clinical parameters. Heatmaps of Spearman’s correlation coefficients between clinical parameters (blood pressure, immunity, inflammatory markers) and relative abundances of the top 50 genera in microbiota of saliva (A), subgingival plaques (B), and feces (C). IL-6: Interleukin-6, PCT: Procalcitonin, CRP: C-reactive protein, WBC: White blood cells, Eo_c: Eosinophil count, Baso_c: Basophil count, Neut_c: Neutrophil count, Mono_c: Monocyte count, Ly_c: Lymphocyte count, Eo_r: Eosinophil ratio, Baso_ r: Basophil ratio, Neut_ r: Neutrophil ratio, Ly_ r: Lymphocyte ratio, Mono_r: Monocyte ratio, DBP: Diastolic blood pressure, SBP: Systolic blood pressure. n = 77 for saliva and subgingival plaques, and 76 for feces. #p(FDR) < 0.1, *p(FDR) < 0.05, **p(FDR) < 0.01, ***p(FDR) < 0.001.

Fig. 5

Communications between oral and gut microbiota in participants with or without HTN. A. Venn diagrams showing the unique and shared OTUs among saliva, subgingival plaques, and feces. The 30 most abundant genera out of the 121 shared genera among salivary, subgingival, and gut microbiota were used for analyses in B-G. B. Prevalence and relative abundances of the top 30 shared genera in no HTN and HTN. The insets of the ‘Feces’ column show the relative abundance of the 15 low-abundance gut genera on enlarged y axes. C-D. Heatmaps of Spearman’s correlation coefficients between relative abundances of shared genera in salivary (C) or subgingival (D) microbiota and those in gut microbiota. E-G. Heatmaps of Spearman’s correlation coefficients among relative abundances of the top 30 shared genera within salivary (E), subgingival (F), or gut (G) microbiota. n = 133 for saliva, 132 for subgingival plaques, and 76 for feces in A and E-G. n = 39:94 for saliva, 39:93 for subgingival plaques, and 24:52 for feces in B. n = 75 for saliva, subgingival plaques, and feces in C and D. #p(FDR) < 0.1, *p(FDR) < 0.05, **p(FDR) < 0.01, ***p(FDR) < 0.001.

Fig. 6

Oral-gut microbiota transmission in participants with or without HTN at species level. All microbiota was analyzed using metagenomic sequencing. A. Overview of microbial species prevalent in oral samples (saliva & subgingival plaques, abbreviated as S&P), gut samples (feces, abbreviated as F), or both oral and gut samples (S&P&F). B. Oral-gut transmission scores of shared species in oral and gut microbiota based on metaSNV analysis. Filled diamonds indicate p < 0.05 by Wilcoxon rank-sum test and the corresponding species are classified as frequent transmitters. Open diamonds indicate p > 0.05 by Wilcoxon rank-sum test but p < 0.05 by Mann-Whitney U rank sum test, and the corresponding species are classified as occasional transmitters. C. Heatmap of Spearman’s correlation coefficients between relative abundances of the 16 oral-gut transmitters in oral (salivary and subgingival) microbiota and those in gut microbiota. D-F. Relative abundances of frequent transmission and occasional transmission species in saliva (D), subgingival plaques (E), and feces (F). n = 60 in A-C, and 24:36 in D-F. #p(FDR) < 0.1, *p(FDR) < 0.05, **p(FDR) < 0.01, ***p(FDR) < 0.001.

Fig. 7

Sustained oral-gut transmission in patients with HTN after 6 months. A. Chao1, Faith’s phylogenetic diversity (Faith’s), Shannon index, and Pielou’s evenness (Pielou’s) of oral (saliva and subgingival plaque) microbiota and gut microbiota in participants with or without HTN. B. Bray-Curtis distance of oral and fecal microbiota in participants with or without HTN. C. SourceTracker analysis to estimate microbial communications from oral cavity to gut and within oral cavity. D. Clustering heatmaps showing relative abundances of the top 50 genera in participants with or without HTN. E. Relative abundances of the 5 oral-gut-transmitting genera in feces. All microbiota was analyzed using 16S rRNA gene sequencing. n = 24:43 for saliva, 25:43 for subgingival plaques, and 24:38 for feces. Kruskal-Wallis rank sum test was used for statistical analysis in A and B, Student’s t test in C and E, and permutational multivariate analysis of variance in D. *p < 0.05.

Fig. 8

Ectopic colonization of salivary microbiota from HTN participants in mouse intestine aggravates HTN. A. Schematic illustration of experimental design. ABX, antibiotic cocktail; Ang II, Angiotensin II. B. Noninvasive tail-cuff monitoring of SBP and DBP in mice. C. Representative H&E staining of thoracic aortas (left) and the quantification of wall thickness (right). Magnification = 200 ×. Three fields were randomly chosen in each sample for quantification. D. Representative Picrosirius red staining of thoracic aortas (left) and the quantification of fibrotic areas (right). Magnification = 200 ×. Three fields were randomly chosen in each sample for quantification. E-F. QRT-PCR analysis of collagen-I in thoracic aortas (E) as well as atrial natriuretic peptide (ANP) and B-type natriuretic peptide (BNP) in left ventricles (F). G. PCoA of mouse feces. H. Stacked bar plots showing relative abundances of the top 20 microbial genera in human saliva and mouse feces. I. LEfSe of mouse gut microbiota. The threshold of LDA score was 2. J. The relative abundance of Veillonella in mouse gut microbiota. K. Random forest regression analysis of mouse gut microbiota at genus level. The heatmap represents relative abundances of microbes. n = 3:3:3 for H2O + Vehicle vs no HTN-saliva + Vehicle vs HTN-saliva + Vehicle; n = 5:5:4 for H2O + Ang II vs no HTN-saliva + Ang II vs HTN-saliva + Ang II. ## p < 0.01 for no HTN-saliva + Ang II vs HTN-saliva + Ang II at day 35 in (B); *p < 0.05, **p < 0.01, *** p < 0.001.