Rationale: According to general view, aldehyde dehydrogenase-2 (ALDH2) catalyzes the high-affinity pathway of vascular nitroglycerin (GTN) bioactivation in smooth muscle mitochondria. Despite having wide implications to GTN pharmacology and raising many questions that are still unresolved, mitochondrial bioactivation of GTN in blood vessels is still lacking experimental support.Objective: In the present study, we investigated whether bioactivation of GTN is affected by the subcellular localization of ALDH2 using immortalized ALDH2-deficient aortic smooth muscle cells and mouse aortas with selective overexpression of the enzyme in either cytosol or mitochondria. Methods and Results:Quantitative Western blotting revealed that ALDH2 is mainly cytosolic in mouse aorta and human coronary arteries, with only approximately 15% (mouse) and approximately 5% (human) of the enzyme being localized in mitochondria. Infection of ALDH2-deficient aortic smooth muscle cells or isolated aortas with adenovirus containing ALDH2 cDNA with or without the mitochondrial signal peptide sequence led to selective expression of the protein in mitochondria and cytosol, respectively. Cytosolic overexpression of ALDH2 restored GTN-induced relaxation and GTN denitration to wild-type levels, whereas overexpression in mitochondria (6-fold vs wild-type) had no effect on relaxation. Overexpression of ALDH2 in the cytosol of ALDH2-deficient aortic smooth muscle cells led to a significant increase in GTN denitration and cyclic GMP accumulation, whereas mitochondrial overexpression had no effect. Key Words: adenovirus Ⅲ aldehyde dehydrogenase-2 Ⅲ mitochondria Ⅲ nitroglycerin Ⅲ vasodilation A ldehyde dehydrogenase-2 (ALDH2) has a wellestablished function in the detoxification of reactive aldehydes, in particular ethanol-derived acetaldehyde, in the liver. Because the liver enzyme is almost exclusively located in the mitochondrial matrix space, it is commonly designated as mitochondrial aldehyde dehydrogenase to differentiate it from the cytosolic isoform (ALDH1). 1 In 2002, Stamler et al 2 discovered that vascular ALDH2 catalyzes bioconversion of nitroglycerin to yield 1,2-glycerol dinitrate (GTN) and inorganic nitrite. This reaction appears to be associated with formation of a disulfide in the catalytic site, leading to mechanism-based enzyme inactivation in the absence of an appropriate reductant, like dithiothreitol, 2 dihydrolipoic acid, 3 or the thioredoxin/thioredoxin reductase system. 4 Based on the mitochondrial localization of liver ALDH2, it has been proposed that GTN bioactivation takes place in mitochondria of vascular smooth muscle cells. Reduction of GTN-derived nitrite to nitric oxide (NO) by components of the respiratory chain would then couple ALDH2-catalyzed GTN metabolism to activation of soluble guanylate cyclase (sGC) and vascular relaxation. 2 Mitochondrial GTN bioactivation, supported by the observation that isolated mitochondria generate NO bioactivity from added GTN, 5,6 has been generally accepted in the field and has fo...

Aldehyde dehydrogenase-2 (ALDH2) catalyzes vascular bioactivation of the antianginal drug nitroglycerin (GTN) to yield nitric oxide (NO) or a related species that activates soluble guanylate cyclase (sGC), resulting in cGMP-mediated vasodilation. Accordingly, established ALDH2 inhibitors attenuate GTNinduced vasorelaxation in vitro and in vivo. However, the ALDH2 hypothesis has not been reconciled with early studies demonstrating potent inhibition of the GTN response by diphenyleneiodonium (DPI), a widely used inhibitor of flavoproteins, in particular NADPH oxidases. We addressed this issue and investigated the effects of DPI on GTN-induced relaxation of rat aortic rings and the function of purified ALDH2. DPI (0.3 mM) inhibited the high affinity component of aortic relaxation to GTN without affecting the response to NO, indicating that the drug interfered with GTN bioactivation. Denitration and bioactivation of 1-2 mM GTN, assayed as 1,2-glycerol dinitrate formation and activation of purified sGC, respectively, were inhibited by DPI with a half-maximally active concentration of about 0.2 mM in a GTN-competitive manner. Molecular modeling indicated that DPI binds to the catalytic site of ALDH2, and this was confirmed by experiments showing substrate-competitive inhibition of the dehydrogenase and esterase activities of the enzyme. Our data identify ALDH2 as highly sensitive target of DPI and explain inhibition of GTN-induced relaxation by this drug observed previously. In addition, the data provide new evidence for the essential role of ALDH2 in GTN bioactivation and may have implications to other fields of ALDH2 research, such as hepatic ethanol metabolism and cardiac ischemia/reperfusion injury.

Graphical abstract

No abstract

No abstract