Leucine accelerates atherosclerosis through dose-dependent MTOR activation in macrophages

ABSTRACT

The role of diet in driving cardiovascular disease (CVD) is well-recognized, particularly in the case of lipids. Dietary protein on the other hand has been heralded as an overall metabolically beneficial nutrient with popularity in the fitness community and in weight-loss regimens. Pursuant to epidemiological studies raising a CVD risk signal for excessive protein intake, we initially conducted murine studies establishing an atherogenic role for dietary protein, the critical involvement of macrophage MTORC1 signaling, and downstream inhibition of protective macroautophagy/autophagy pathways. In recent work, we confirm these findings in monocytes from humans consuming protein and dissect the MTORC1-autophagy cascade in human macrophages. We also identify leucine as the single most important amino acid, observing dose-dependent activation of MTOR whereby only leucine concentrations above a threshold trigger pathogenic signaling and monocyte/macrophage dysfunction. Using mouse models fed diets with modulated protein and leucine content, we confirm this threshold effect in driving atherosclerosis. Our findings establish a pathogenic role for dietary leucine in CVD and raise the promise of therapeutic strategies aimed at selective inhibition of macrophage leucine-MTOR signaling.

High-protein diets (HPDs) have been relied on as a primary strategy for weight loss and ensuing metabolic benefit. However, accumulating epidemiological and animal studies suggest an increased CVD risk albeit without a defined mechanism. In prior work, we discovered that HPD induces atherosclerosis in mice by activating amino acid-MTORC1 signaling in macrophages leading to inhibition of atheroprotective autophagy pathways. The key amino acid inducers and the clinical relevance of these findings remained unclear.

We thus conducted two clinical studies modulating protein intake in humans to assess MTORC1 signaling in circulating monocytes. Our first study examined the protein extremes, comparing intake of low (10% energy) and very high (50% energy) liquid protein meals. Participants who consumed very high-protein showed increased MTORC1 activation and autophagy inhibition in circulating monocytes, similar to our earlier work in animal models. The second study examined protein intake in a more typical dietary context, comparing a standard (15% of energy) with a moderately high (22% of energy) protein meal. Even in this “real-world” scenario, modestly elevated protein resulted in MTORC1 activation and autophagy inhibition in monocytes, highlighting the atherogenic role of dietary protein.

To define the culprit amino acid driving pathogenesis, we profiled the kinetics of circulating amino acids in our clinical studies and used human monocyte-derived macrophages to relate amino acid concentrations with MTORC1 activation kinetics and downstream signaling [1]. Only leucine is the predominant MTORC1-activator at concentrations observed in human plasma upon protein intake while classic inducers including arginine and the other branched-chain amino acids (BCAAs), isoleucine and valine, do not induce MTORC1. Leucine-mediated MTORC1 signaling is sufficient to impair autophagy/mitophagy, cause accumulation of dysfunctional mitochondria, elevate reactive oxygen species, and trigger apoptosis, all of which underlie CVD risk. Interestingly, leucine levels required to initiate deleterious MTOR signaling are observed only above modestly elevated protein intake (>22% energy), demarcating a threshold at which risk rises. Murine studies with modulation of protein intake supplemented with/without leucine recapitulate these findings, implicating leucine as both necessary and sufficient for diet-driven MTOR signaling.

To further investigate the threshold effect of protein intake on atherosclerosis in vivo, we designed three atherogenic diets with varied protein content (7, 21, and 45% energy), correlating with the degree of protein intake in our human studies. Whereas atherosclerosis-prone apoe-null mice have a similar plaque burden and macrophage MTORC1 activation with low- and moderate-protein intake, a threshold is reached in the HPD resulting in elevations of MTORC1 and lesion size/complexity. Finally, we further modified moderate-protein atherogenic diets with either supplemental leucine, all amino acids, or all amino acids without leucine. Only the addition of leucine reproduces plaque progression observed in the high-protein setting, establishing leucine as necessary and sufficient in protein-induced atherosclerosis.

Our data raise an essential question, namely why leucine is the predominant pathogenic amino acid in atherosclerosis as compared to all other amino acids? After all, BCAAs including leucine, isoleucine, and valine, are associated with CVD. However, studies have traditionally considered all BCAAs as equivalent risk factors without distinguishing their roles. Recent work has also shown that limiting another BCAA, isoleucine, can enhance metabolic health and increase lifespan. While atherosclerosis was not specifically evaluated, such data provide a compelling reason to explore the selective impact of BCAAs and other amino acids on cardiovascular and cardiometabolic diseases. Examination of how protein over-consumption or restriction might affect atherosclerosis in the context of each amino acid would also be prudent.

Our findings suggest that nutritional interventions targeting leucine could be a new therapeutic strategy for CVD. The variable abundance of leucine in different protein sources indicates that a “designer” diet with reduced leucine content could conceivably reduce CVD risk. The degree to which MTOR signaling can be modulated by consumption of different protein sources (e.g., animal versus plant) needs to be evaluated. We refer to dietary manipulation for promoting desired signaling in immune cells as a “precision nutrition” approach to CVD therapeutics.

Inhibition of leucine-induced MTORC1 signaling offers another therapeutic avenue. Broad MTOR inhibitors such as rapamycin and related rapalogs suppress atherosclerosis in animal models and provide benefit in human coronary artery disease (e.g., drug-eluting stents). Yet, nonselective MTOR inhibition promotes hyperglycemia and insulin resistance as metabolic side-effects. The pathogenic role of leucine-MTOR signaling raises the potential for modulating its uptake, sensing, or catabolism as a selective approach. Transporters of leucine, such as the SLC/LAT family, are overexpressed in several tumors which leads to pharmacological targeting in ongoing cancer clinical trials to block pro-proliferative MTOR signaling. Similar approaches to inhibit MTORC1 in macrophages and CVD can be equally efficacious.

Modulation of cellular leucine sensing and metabolism are other promising therapeutic approach. Leucine sensors such as SESN2 (sestrin 2), LARS, and SAR1B utilize distinct mechanisms to mediate MTORC1 activation. Targeting each of these sensors can abrogate MTOR signaling as a therapeutic strategy in CVD. As amino acids cannot be stored outside of cellular protein synthetic needs, unused amounts undergo catabolism in mitochondria; therefore, promoting leucine metabolism may help deplete its levels below deleterious thresholds.

One should also consider the impact of other atherogenic nutrients such as lipids. Data in cell lines implicate cholesterol as an activator of MTOR. Lipoprotein-derived cholesteryl esters are hydrolyzed in lysosomes freeing cholesterol for sensors such as SLC38A9 and GPR155 to recruit MTORC1 to the lysosomal surface. Such lysosome-dependent signaling in macrophages might be an overlooked mechanism among more established pathogenic roles of cholesterol in atherosclerosis. Overall, a plethora of novel concepts arise from identification of leucine as the culprit amino acid in CVD risk (Figure 1 summarizes the leucine-MTOR signaling paradigm).

Figure 1.

Summary of the relevant cellular machinery involved in the leucine-mtor signaling paradigm, particularly focusing on leucine uptake, lysosome-related nutrient sensing and MTOR activation (leucine vs cholesterol), and the inhibition of autophagy/mitophagy and tfeb-mediated lysosome-autophagy biogenesis.

Funding Statement

The work was supported by the National Heart, Lung, and Blood Institute [R01 HL159461]; National Institute of Diabetes and Digestive and Kidney Diseases [R01 DK131188]; National Heart Lung and Blood Institute [R01 HL125838]; United States Department of Veterans Affairs [VA MERIT I01 BX003415].

Disclosure statement

No potential conflict of interest was reported by the author(s).