BACKGROUND.l-Carnitine, an abundant nutrient in red meat, accelerates atherosclerosis in mice via gut microbiota–dependent formation of trimethylamine (TMA) and trimethylamine N-oxide (TMAO) via a multistep pathway involving an atherogenic intermediate, γ-butyrobetaine (γBB). The contribution of γBB in gut microbiota–dependent l-carnitine metabolism in humans is unknown. METHODS. Omnivores and vegans/vegetarians ingested deuterium-labeled l-carnitine (d3-l-carnitine) or γBB (d9-γBB), and both plasma metabolites and fecal polymicrobial transformations were examined at baseline, following oral antibiotics, or following chronic (≥2 months) l-carnitine supplementation. Human fecal commensals capable of performing each step of the l-carnitine→γBB→TMA transformation were identified. RESULTS. Studies with oral d3-l-carnitine or d9-γBB before versus after antibiotic exposure revealed gut microbiota contribution to the initial 2 steps in a metaorganismal l-carnitine→γBB→TMA→TMAO pathway in subjects. Moreover, a striking increase in d3-TMAO generation was observed in omnivores over vegans/vegetarians (>20-fold; P = 0.001) following oral d3-l-carnitine ingestion, whereas fasting endogenous plasma l-carnitine and γBB levels were similar in vegans/vegetarians (n = 32) versus omnivores (n = 40). Fecal metabolic transformation studies, and oral isotope tracer studies before versus after chronic l-carnitine supplementation, revealed that omnivores and vegans/vegetarians alike rapidly converted carnitine to γBB, whereas the second gut microbial transformation, γBB→TMA, was diet inducible (l-carnitine, omnivorous). Extensive anaerobic subculturing of human feces identified no single commensal capable of l-carnitine→TMA transformation, multiple community members that converted l-carnitine to γBB, and only 1 Clostridiales bacterium, Emergencia timonensis, that converted γBB to TMA. In coculture, E. timonensis promoted the complete l-carnitine→TMA transformation. CONCLUSION. In humans, dietary l-carnitine is converted into the atherosclerosis- and thrombosis-promoting metabolite TMAO via 2 sequential gut microbiota–dependent transformations: (a) initial rapid generation of the atherogenic intermediate γBB, followed by (b) transformation into TMA via low-abundance microbiota in omnivores, and to a markedly lower extent, in vegans/vegetarians. Gut microbiota γBB→TMA/TMAO transformation is induced by omnivorous dietary patterns and chronic l-carnitine exposure. TRIAL REGISTRATION. ClinicalTrials.gov NCT01731236. FUNDING. NIH and Office of Dietary Supplements grants HL103866, HL126827, and DK106000, and the Leducq Foundation.
Robert A. Koeth, Betzabe Rachel Lam-Galvez, Jennifer Kirsop, Zeneng Wang, Bruce S. Levison, Xiaodong Gu, Matthew F. Copeland, David Bartlett, David B. Cody, Hong J. Dai, Miranda K. Culley, Xinmin S. Li, Xiaoming Fu, Yuping Wu, Lin Li, Joseph A. DiDonato, W.H. Wilson Tang, Jose Carlos Garcia-Garcia, Stanley L. Hazen
Submitter: Mario Bonomini | email@example.com
Authors: Mario Bonomini, Arduino Arduini, Victor Zammit
University of Chieti-Pescara, Italy
Published March 13, 2019
We read with interest the research paper of Koeth et al. on the gut microbiota-dependent transformations of L-carnitine to trimethylamine N-oxide (TMAO) in humans (1). Beside the novel discovery of the multistep transformation involving different elements of the gut microbiome, the authors make very definitive statements regarding the increased cardiovascular (CVD) risk posed by TMAO- generating nutrients like L-carnitine. Such statements are not justified in view of the current debate within the biomedical community of the validity of the claims for a causative association between TMAO and CVD risk. As outlined by Fennema et al (2), most of the human studies showing a direct correlation between higher plasma TMAO levels and increased CVD risk may simply represent an epiphenomenon. There are also studies in humans showing an inverse correlation (3-6) or, as in a cohort of patients with high cardiometabolic risk, plasma TMAO concentrations are confounded by impaired kidney function and poor metabolic control (7). Other factors acting as potential confounders are high salt diet and loop diuretics (8,9). Moreover, a significant dysbiosis of the gut microbiota along with reduced TMAO blood levels were present in patients with large-artery atherosclerotic ischemic stroke and transient ischemic attack, though those patients with asymptomatic atherosclerosis did not show any change either in gut microbiota or in blood TMAO levels (6). There are also preclinical studies conducted in different genetic animal models showing a lack of association between higher TMAO levels and atherosclerotic lesions (10-11). Intriguingly, TMAO has been shown to possess pharmacological properties even with regard to CVD (10, 11) as well as L-carnitine (12). It is outside the scope of this letter to enter into a detailed discussion of the propensity of different foods to contain or generate TMAO, but it is noteworthy that fish-rich diets have a content of TMAO which is an order of magnitude higher than the L-carnitine content of red meat-based ones (13), and no association has ever been found between fish consumption and cardiac disease. Therefore, it is unfortunate that Koeth et al (1) did not consider all the literature informing the current debate on whether plasma TMAO levels are a genuine risk factor for CVD.