Gut microbiota-derived metabolite isovalerylcarnitine modulates salt sensitivity of blood pressure and incident hypertension: a multicenter dietary salt intervention trial

Abstract

This study aims to investigate the roles of gut microbiota and plasma metabolites in salt sensitivity (SS) of blood pressure (SSBP) and hypertension. A 23-day, multicenter, dietary salt intervention trial (the MetaSalt study) recruited 528 participants who underwent a baseline observation, low-salt, and high-salt interventions. SSBP was assessed and used as the primary outcome, and fecal shotgun metagenome and plasma targeted metabolome were measured. We found that high salt significantly altered 85 gut-microbial species (p < 9.42 × 10^-5) and 70 metabolites (p < 2.26 × 10^-4). Among them, the changes in 22 species and 8 metabolites were associated with SSBP (p < 0.05), and a gut microbiota-acylcarnitine network implicated in SSBP was identified, with a gut microbiota-derived metabolite, isovalerylcarnitine, as the core metabolite. Isovalerylcarnitine was also inversely associated with SSBP in the GenSalt study (p = 0.0102). Importantly, increased isovalerylcarnitine attenuated SS hypertension and improved endothelial function in rats, and was associated with reduced risk (ranging from 13% to 19%) of BP progression and incident hypertension in a prospective cohort (n = 3907, median follow-up = 5.5 years). This study demonstrated that the gut-acylcarnitine axis may play roles in the development of SS hypertension. Trial number: ChiCTR1900025171.

Fig. 1. Overall flow chart of the study.

BP blood pressure; SSBP salt sensitivity of blood pressure.

Fig. 2. Identification of salt-related biomarkers.

a Study design of the MetaSalt study. b–e Volcano plot for salt-related gut-microbial species, functional pathways (KEGG pathways and CAZy families), and metabolites. Linear mixed models were applied to compare the differences in gut microbiota and metabolites between the low-salt and high-salt interventions among the MetaSalt participants (n = 503), adjusted for the random intercepts of individuals and families. Effect sizes in figures represent the relative changes of biomarkers in the high-salt period, compared with those in the low-salt period. Bonferroni correction was separately applied for the identification of salt-related gut-microbial species, functional pathways, and metabolites. f Overlapped KEGG pathways between those identified based on salt-related metabolites and those identified based on gut microbiota. Over-representation analysis and linear mixed model were used to identify key KEGG pathways for metabolites and gut microbiota, respectively. Data are presented as effect size (95% CI). p-values are two-sided. BP blood pressure; CAZy the Carbohydrate-Active EnZymes database; CI confidence interval; KEGG the Kyoto encyclopedia of genes and genomes database; MAP mean arterial pressure; SR salt resistance; SS salt sensitivity.

Fig. 3. Relationships among gut-microbial species, metabolites, and SSBP.

a Adjusted mean changes of SSBP-related gut-microbial species across SSBP groups. Data are presented as adjusted mean change (95% CI). b Adjusted mean changes of SSBP-related metabolites across SSBP groups. Data are presented as adjusted mean change (95% CI). c Relationships between gut-microbial species and metabolites. Linear mixed models were used to identify SSBP-related biomarkers among the MetaSalt participants (n = 503), estimate their adjusted mean changes (least-squares means), and evaluate the relationships between gut-microbial species and metabolites. Covariates included age, sex, field centers, body mass index, smoking, hypertension, total cholesterol, and random intercepts of families. Co-occurrence networks were visualized using Cytoscape based on relationships with p < 0.05. p-values are two-sided. CI confidence interval; SR salt resistance; SS salt sensitivity; SSBP salt sensitivity of blood pressure.

Fig. 4. Salt-related and SSBP-related acylcarnitines in the MetaSalt and GenSalt studies.

a Study scheme of the GenSalt study (n = 226). b Plasma levels of five acylcarnitines between the low-salt and high-salt interventions. Differences in metabolites between groups were tested using linear mixed models, adjusted for the random intercepts of individuals. In each boxplot, the center line indicates the median, the box represents the interquartile range (IQR), the whiskers extend to the minimum and maximum values within 1.5 IQR, and points beyond the whiskers are outliers. c Adjusted mean change in isovalerylcarnitine in SR and extreme SS individuals. Differences in metabolites between groups were tested using linear mixed models, and the adjusted mean changes (least-squares means) were obtained, adjusted for body mass index, smoking, and total cholesterol, and the random intercepts of matches. Data are presented as adjusted means change (95% CI). p-values are two-sided. CI confidence interval; SR salt resistance; SS salt sensitivity; SSBP salt sensitivity of blood pressure.

Fig. 5. Validation of the effect of isovalerylcarnitine in rats.

a Flow chart of the intervention in Wistar rats. b Non-invasive BP level of Wistar rats in the control and HSD groups (n = 15 animals per group). Data are presented as mean (95% CI). c Plasma level of isovalerylcarnitine of Wistar rats in the control and HSD groups (n = 15 animals per group). d Flow chart of the intervention in Dahl SS rats. e Telemetric monitoring BP of Dahl SS rats in the control, HSD, and HSD + isovalerylcarnitine groups (n = 7, 11, 11 animals, respectively). Data are presented as mean (95% CI). f Concentration-response curves of vasodilatation in mesenteric arteries in response to acetylcholine (n = 11, 13, 13 arteries, respectively) and sodium nitroprusside (n = 10, 10, 12 arteries, respectively) in the control, HSD, and HSD +  isovalerylcarnitine groups of Dahl SS rats. Data are presented as mean (95% CI). Immunofluorescent staining (g) and fluorescence intensity (h) of eNOS in the aortas from the control, HSD, and HSD + isovalerylcarnitine groups of Dahl SS rats (n = 7, 11, 11 animals, respectively). i Representative H&E, Masson’s staining, and Sirius Red staining of the aortas and kidneys of Dahl SS rats. j Quantification of tunica media collagen deposition in thoracic aortas from the control, HSD, and HSD + isovalerylcarnitine groups of Dahl SS rats (n = 7, 11, 11 animals, respectively). k Quantification of fibrotic area of the kidney from the control, HSD, and HSD +  isovalerylcarnitine groups of Dahl SS rats (n = 7, 11, 11 animals, respectively). Differences in BP and vasodilatation between groups throughout the entire experiment were tested using repeated measurement ANOVA. Differences in metabolites and BP between groups at the end of the experiment were tested using t-test or Wilcoxon rank sum test. p-values are two-sided. Scale bar: 50 μm. In each boxplot, the center line indicates the median, the box represents the interquartile range (IQR), the whiskers extend to the minimum and maximum values within 1.5 IQR, and points beyond the whiskers are outliers. ANOVA analysis of variance; BP blood pressure; CI confidence interval; DBP diastolic blood pressure; eNOS endothelial nitric oxide synthase; HSD high-salt diet; NaCl sodium chloride; SBP systolic blood pressure; SS salt sensitivity.

Fig. 6. Impacts of isovalerylcarnitine on BP progression and incident hypertension in a prospective cohort.

a Study scheme of the prospective cohort study (n = 3907). b Association between isovalerylcarnitine and hypertension. Logistic regression and Cox proportional hazards models were used to estimate ORs and HRs, respectively, adjusted for age, sex, body mass index, work-related physical activity, area, region, smoking status, current drinking, education, dietary score, dyslipidemia, diabetes mellitus, and carnitine. Data are presented as OR or HR (95% CI). p-values are two-sided. Before analysis, the concentration of isovalerylcarnitine was normalized, and effect sizes were reported for each standard deviation (0.08 μmol/L) increase in isovalerylcarnitine. The associations of the continuous form and tertiles [T1 (0.03–0.10 μmol/L), T2 (0.11–0.16 μmol/L), and T3 (0.17–0.45 μmol/L)] of isovalerylcarnitine with prevalent hypertension, incident hypertension, and BP status progression were evaluated. For the tertile form, the first tertile (T1) was used as the reference. BP blood pressure; CI confidence interval; DBP diastolic blood pressure; HR hazard ratio; OR odds ratio; SBP systolic blood pressure.