The gut microbiome is closely interconnected with the host physiology. Evidence continues to mount that the cardiovascular (CV) system, although seemingly distant from the gut, is also influenced by the gut microbiota. Sitting at the interface between the host and the environment, the gut microbiome mediates and modulates ways in which the environment affects CV risk and progression of cardiovascular disease (CVD), including heart failure (HF). One of the chief mechanisms involves microbially produced small molecules that act at or beyond the host gut barrier. Biological processes dysregulated in HF, such as myocardial energetics, cardiac remodeling, and repair capacity, systemic vascular and inflammatory tone, are some of the processes affected by metabolites produced by our gut microbiome, including short-chain fatty acids, trimethylamine-N-oxide, indole-3-propionate, and phenacetylglutamine. Several of these are direct metabolites of nutrients we consume daily, and this represents one of the mechanisms by which our diet may modify CV risk. In this issue of JACC: Heart Failure, Molinaro et al[1] offer the first evidence for another microbially produced metabolite with potential relevance for CV and HF pathophysiology.
The study by Molinaro et al[1] focused on imidazole propionate (ImP), a gut microbial metabolite of the amino acid histidine. Having previously found circulating ImP to be elevated in patients with type 2 diabetes (T2D)[2] and to cause insulin resistance (IR) in an animal model,[3] the authors sought to explore the relationship between circulating ImP and atherosclerotic CVD and HF using 2 large, independent cohorts, a European population-based cohort (MetaCardis) (n = 1,985) and a North American hospital-based cohort (GeneBank) (n = 2,155). The European cohort had a smaller proportion of patients with HF (n = 133) and CVD (n = 282), with a much larger group of individuals without known CVD, made up of metabolically diseased and healthy individuals (combined n = 1,569). Individuals with established CVD or HF had significantly higher levels of circulating ImP, compared with those without, and the highest levels were in individuals with HF. Compared with the lowest quartile of ImP levels, individuals in the highest quartile were 3 times more likely to have HF even after adjustment for multiple traditional CV risk factors, several of which were different between the groups, as would be expected. Additional analyses showed that ImP levels were inversely associated with left ventricular ejection fraction (LVEF) and directly associated with levels of pro-atrial natriuretic peptide and N-terminal pro–B-type natriuretic peptide. These findings were replicated in the North American cohort, which was largely a disease cohort (HF, n = 407; CVD, n = 1,331; no-CVD/HF, n = 417). Finally, the authors took advantage of the 5-year longitudinal follow-up data available for the North American cohort to demonstrate an increased risk of overall mortality across the ImP quartiles.
Altogether, this study supports the association link between gut microbially produced ImP and both atherosclerotic CVD and HF, as well as important HF phenotypes characterized by reduced LVEF and elevated cardiac filling pressures. Furthermore, it demonstrates that circulating ImP is independently predictive of medium-term all-cause mortality risk. The study has some important strengths, in particular the large sample size, presence of a replication cohort, and the geographical diversity, which together make it an important new observation in the microbiome story. Nonetheless, like many other gut microbiome–focused studies, it leaves many unknowns and poses more questions than it answers. Although ImP appears clearly associated with the presence of CVD and HF, the survival association is less convincing because that analysis did not stratify or categorize by disease groups, and the magnitude of effect was substantially mitigated after adjustment (raw HR: 3.60 and adjusted HR: 1.18). Although the large sample size is clearly a strength, it could be showing a real signal that still might reflect only a small effect with substantial between-person variability in the measure. This seems to be the case in both analyzed cohorts, as the boxplots clearly show mean differences but also very wide and overlapping ranges between the subject categories. Some additional detail in phenotype would help as well; while a good starting point, HF phenotypes are described by more than just LVEF and N-terminal pro–B-type natriuretic peptide. Specific CVD and HF phenotypes, as well as comorbidities common to these conditions and prescribed medications may be associated with unique patterns of ImP. As the authors point out, these questions have yet to be answered. Finally, it is worth reminding that all the associations are descriptive and provide little insight regarding causality or mechanism.
Considering the mechanism of how ImP might relate to the pathogenesis of atherothrombotic CVD and HF and disease progression, if it is indeed an active contributor, there are several plausible hypotheses to be evaluated. Promisingly, animal, and in vitro data have shown that ImP causes sustained activation of the mechanistic target of rapamycin complex 1 (mTORC1) leading to dysglycemia and IR.[3] In addition, ImP inhibits adenosine monophosphate–activated protein kinase (AMPK) to counteract the action of metformin,[4] the first-line treatment for T2D. Both mTORC1 and AMPK signaling pathways are critical for myocardial physiology and have been implicated in the development of cardiac dysfunction. For example, mTOR is a master regulator of several crucial cellular processes, including cell growth, proliferation, autophagy, and metabolism, and although some level of activity is necessary for cardiomyocyte survival, its overactivation results in maladaptive cardiac remodeling and hypertrophy.[5] AMPK is a major regulator of myocardial metabolism and energetics and is especially critical at the times of myocardial stress, such as HF.[6] It remains to be seen whether ImP, at circulating levels seen in this study, affects cardiac function directly, and if so, by what mechanism. Certainly, the pathways referenced previously seem like a very reasonable place to start to answer this question. If ImP is indeed found to be cardiotoxic, this would potentially represent another discrete physiologic link between the gut microbiome, IR/T2D-related dysmetabolism, and HF, with clear clinical relevance. Diabetes can cause cardiomyopathy and chronic HF, and IR can develop secondary to progressive HF.[7] HF and IR/T2D share many gut microbiome features, including low community diversity, overgrowth of the pro-inflammatory microbes, and depletion of the beneficial, short chain fatty acid–producing bacteria.[8] It is possible that the gut microbiome compositional and functional changes, including potentially increased capacity for ImP biosynthesis, contribute to the development of cardiac dysfunction in the context of preexisting T2D, and to the development of IR in individuals with progressive HF.
Given the high prevalence and morbidity of coincident HF and T2D, identifying a metabolite or pathway that could potentially be targeted to disrupt the pathophysiologic synergy between the 2 would be very impactful. Whether ImP fills this void will depend on clarifying whether it is a marker or mediator of CVD and HF risk. If proven to be the latter, detailed investigations clarifying its mechanisms (eg, interrogating proposed downstream pathways) will be necessary, as will the efforts to identify interventions that affect ImP levels or target organ activity. The ultimate step will be to ascertain if altering ImP levels or activity changes downstream CVD or HF risk, or disease progression. Interestingly, this research group already found that pirfenidone, a drug used to treat idiopathic pulmonary fibrosis, blocks ImP-induced activation of the protein kinase p38γ. This results in downstream AMPK activation to overcome ImP-related metformin resistance.[4] Whether this or other drugs may effectively block ImP activity in cardiac or vascular tissues, and with the same effectiveness, remains to be studied. On the other hand, what is already clear is that dietary habits are related to the circulating ImP levels. In their previous work on this same MetaCardis cohort, this research group showed that the ImP levels were not related to the amount of dietary histidine consumed, but rather to unhealthy eating patterns.[2] For example, ImP was directly correlated with intake of saturated fats (eg, cheese), and was inversely correlated with intake of fiber and unsaturated fat (eg, vegetables and nuts). Such dietary patterns have been previously associated with gut dysbiosis; this study similarly found that ImP levels directly correlated with low gut microbial richness, pro-inflammatory microbiota composition with augmented capacity for ImP production, and high circulating inflammatory markers. As such, ImP might be simply a biomarker of dysregulated gut microbiome—from unhealthy diet or disease. If it is more than that in CVD and HF, dietary modifications may be a plausible intervention to mitigate ImP. In fact, it is likely that following healthier dietary habits—those rich in unprocessed, whole foods such as vegetables, fruits, whole grains, legumes, nuts, and seeds, and low in processed foods and animal products—would alter much more than just circulating ImP to provide a more holistic cardiometabolic benefit.
In their paper, Molinaro et al[1] have elevated the priority of further research on ImP and the gut microbiome vis-à-vis CVD and HF. However, as the authors[1] themselves point out, this really is just the beginning, and much work remains to be done before we can truly understand the importance of ImP in pathogenesis and progression of CVD and HF. Although clearly enjoying a recent boom, gut microbiome research in general is still in its infancy. What we know so far likely represents just the tip of the iceberg. The current study has revealed to us more of this iceberg, but until we dive much deeper under its surface, we will not fully understand the complexities of our gut microbiome–host meta-organism.
JACC Heart Fail. 2023;11(7):822-824. © 2023 American College of Cardiology Foundation
