BACKGROUND Mammalian target of rapamycin complex 1 (mTORC1) is a protein kinase that relays nutrient availability signals to control numerous cellular functions including autophagy, a process of cellular self-eating activated by nutrient depletion. Addressing the therapeutic potential of modulating mTORC1 signaling and autophagy in human disease requires active chemicals with pharmacologically desirable properties. METHODOLOGY/PRINCIPAL FINDINGS Using an automated cell-based assay, we screened a collection of {\textgreater}3,500 chemicals and identified three approved drugs (perhexiline, niclosamide, amiodarone) and one pharmacological reagent (rottlerin) capable of rapidly increasing autophagosome content. Biochemical assays showed that the four compounds stimulate autophagy and inhibit mTORC1 signaling in cells maintained in nutrient-rich conditions. The compounds did not inhibit mTORC2, which also contains mTOR as a catalytic subunit, suggesting that they do not inhibit mTOR catalytic activity but rather inhibit signaling to mTORC1. mTORC1 inhibition and autophagosome accumulation induced by perhexiline, niclosamide or rottlerin were rapidly reversed upon drug withdrawal whereas amiodarone inhibited mTORC1 essentially irreversibly. TSC2, a negative regulator of mTORC1, was required for inhibition of mTORC1 signaling by rottlerin but not for mTORC1 inhibition by perhexiline, niclosamide and amiodarone. Transient exposure of immortalized mouse embryo fibroblasts to these drugs was not toxic in nutrient-rich conditions but led to rapid cell death by apoptosis in starvation conditions, by a mechanism determined in large part by the tuberous sclerosis complex protein TSC2, an upstream regulator of mTORC1. By contrast, transient exposure to the mTORC1 inhibitor rapamycin caused essentially irreversible mTORC1 inhibition, sustained inhibition of cell growth and no selective cell killing in starvation. CONCLUSION/SIGNIFICANCE The observation that drugs already approved for human use can reversibly inhibit mTORC1 and stimulate autophagy should greatly facilitate the preclinical and clinical testing of mTORC1 inhibition for indications such as tuberous sclerosis, diabetes, cardiovascular disease and cancer.

Perhexiline, 2-(2,2-dicyclohexylethyl)piperidine, was originally developed as an anti-anginal drug in the 1970s. Despite its success, its use diminished due to the occurrence of poorly understood side effects including neurotoxicity and hepatotoxicity in a small proportion of patients. Recently, perhexiline's mechanism of action and the molecular basis of its toxicity have been elucidated. Perhexiline reduces fatty acid metabolism through the inhibition of carnitine palmitoyltransferase, the enzyme responsible for mitochondrial uptake of long-chain fatty acids. The corresponding shift to greater carbohydrate utilization increases myocardial efficiency (work done per unit oxygen consumption) and this oxygen-sparing effect explains its antianginal efficacy. Perhexiline's side effects are attributable to high plasma concentrations occurring with standard doses in patients with impaired metabolism due to CYP2D6 mutations. Accordingly, dose modification in these poorly metabolizing patients identified through therapeutic plasma monitoring can eliminate any significant side effects. Herein we detail perhexiline's pharmacology with particular emphasis on its mechanism of action and its side effects. We discuss how therapeutic plasma monitoring has led to perhexiline's safe reintroduction into clinical practice and how recent clinical data attesting to its safety and remarkable efficacy led to a renaissance in its use in both refractory angina and chronic heart failure. Finally, we discuss the application of pharmacogenetics in combination with therapeutic plasma monitoring to potentially broaden perhexiline's use in heart failure, aortic stenosis, and other cardiac conditions.

Perhexiline is a potent prophylactic anti-anginal agent that has been shown to inhibit myocardial utilization of long-chain fatty acids and to inhibit the mitochondrial enzyme carnitine palmitoyltransferase (CPT)-1. We compared the hemodynamic and biochemical effects of perhexiline (0.5 and 2.0 microM) and of another CPT-1 inhibitor, oxfenicine (0.5 mM), in Langendorff-perfused rat hearts subjected to 60 min of low-flow ischemia (95{\%} flow reduction) followed by 30 min of reperfusion. Both perhexiline (2 microM only) and oxfenicine attenuated (p {\textless} 0.003, p {\textless} 0.0002, respectively) increases in diastolic tension during ischemia, without significant effects on developed tension, or on cardiac function during reperfusion. Myocardial concentrations of long-chain acylcarnitines (LCAC), products of CPT-1 action, were decreased (p {\textless} 0.05) by oxfenicine, unaffected by 2 microM perhexiline, and increased slightly by 0.5 microM perhexiline. Perhexiline, but not the active metabolite of oxfenicine, also inhibited cardiac CPT-2 with similar IC50 and Emax, although lower Hill slope, compared with CPT-1. Oxfenicine, but not perhexiline, reduced concentrations of the endogenous CPT-1 inhibitor, malonyl-CoA. Perhexiline, but not oxfenicine, inhibited myocardial release of lactate during normal flow. We conclude that (a) perhexiline protects against diastolic dysfunction during ischemia in this model, independent of major changes in LCAC accumulation and (b) this may result from simultaneous effects of perhexiline on myocardial CPT-1 and CPT-2.