The nitric oxide synthases (NOS-I, neuronal, NOS-II, inducible, and NOS-III, endothelial) are the most recent additions to the large number of heme proteins that contain cysteine thiolate-liganded protoporphyrin IX heme prosthetic groups. This group of oxygenating enzymes also includes one of the largest gene families, that of the cytochromes P450, which have been demonstrated to be involved in the hydroxylation of a variety of substrates, including endogenous compounds (steroids, fatty acids, and prostaglandins) and exogenous compounds (therapeutic drugs, environmental toxicants, and carcinogens). The substrates for cytochromes P450 are universally hydrophobic while the physiological substrate for the nitric oxide synthases is the amino acid L-arginine, a hydrophilic compound. This review will discuss the approaches being used to study the structure and mechanism of neuronal nitric oxide synthase in the context of its known prosthetic groups and regulation by Ca(2+)-calmodulin and/or tetrahydrobiopterin (BH4).

The crystal structure of the endothelial nitric oxide synthase (NOS) heme domain complexed with NO reveals close hydrogen bonding interactions between NO and the terminal guanidino nitrogen of the substrate, L-arginine. Dioxygen is expected to bind in a similar mode which will facilitate proton abstraction from L-Arg to dioxygen, a required step for O-O bond cleavage. Structures of mechanism-based NOS inhibitors, N(5)-(1-iminoethyl)-L-ornithine and N-(3-(aminomethyl)benzyl)acetamidine, provide clues on how this class of compounds operate as suicide substrate inhibitors leading to heme oxidation.

Type 2 diabetes is an insulin-resistant state characterized by hyperinsulinemia and accelerated atherosclerosis. In vitro and in vivo studies in rodents have suggested that nitric oxide generation plays an important role in glucose transport and insulin action. We determined nitric oxide synthase (NOS) activity in skeletal muscle of 10 type 2 diabetic (hemoglobin A(1C) = 6.8 +/- 0.1%) and 11 control subjects under basal conditions and during an 80 mU/m(2).min euglycemic insulin clamp performed with vastus lateralis muscle biopsies before and after 4 h of insulin. In diabetics, insulin-stimulated glucose disposal (Rd) was reduced by 50%, compared with controls (5.4 +/- 0.3 vs. 10.4 +/- 0.5 mg/kg.min, P < 0.01). Basal NOS activity was markedly reduced in the diabetic group (101 +/- 33 vs. 457 +/- 164 pmol/min.mg protein, P < 0.05). In response to insulin, NOS activity increased 2.5-fold in controls after 4 h (934 +/- 282 pmol/min.mg protein, P < 0.05 vs. basal), whereas insulin failed to stimulate NOS activity in diabetics (86 +/- 28 pmol/min.mg protein, P = NS from basal). Basal NOS protein content in muscle was similar in controls and diabetics and did not change following insulin. In controls, insulin-stimulated NOS activity correlated inversely with fasting plasma insulin concentration (r = -0.58, P = 0.05) and positively with Rd (r = 0.71, P = 0.03). In control and diabetic groups collectively, Rd correlated with insulin-stimulated NOS activity (r = 0.52, P = 0.02). We conclude that basal and insulin-stimulated muscle NOS activity is impaired in well-controlled type 2 diabetic subjects, and the defect in insulin-stimulated NOS activity correlates closely with the severity of insulin resistance. These results suggest that impaired NOS activity may play an important role in the insulin resistance in type 2 diabetic individuals.

Nitric oxide is generated under normal and pathophysiological conditions by three distinct isoforms of nitric oxide synthase (NOS). A small-molecule inhibitor of NOS (3-Br-7-nitroindazole, 7-NIBr) is profoundly neuroprotective in mouse models of stroke and Parkinson's disease. We report the crystal structure of the catalytic heme domain of endothelial NOS complexed with 7-NIBr at 1.65 A resolution. Critical to the binding of 7-NIBr at the substrate site is the adoption by eNOS of an altered conformation, in which a key glutamate residue swings out toward one of the heme propionate groups. Perturbation of the heme propionate ensues and eliminates the cofactor tetrahydrobiopterin-heme interaction. We also present three crystal structures that reveal how alterations at the substrate site facilitate 7-NIBr and structurally dissimilar ligands to occupy the cofactor site.

A versatile diatomic physiological messenger, nitric oxide (NO), is biosynthesized by a group of flavo-heme enzymes, the nitric oxide synthases. We have examined the active site of the neuronal isoform by EPR spectroscopy of the ferrous nitric oxide complex. The nitric oxide complex of the substrate-free enzyme exhibits a cytochrome P450-type EPR spectrum typical of a hexacoordinate NO-heme complex with a non-nitrogenous proximal axial heme ligand. The NO complex of the substrate-free enzyme is rather unstable and spontaneously converts to a cytochrome P420 type pentacoordinate denatured form. Binding of L-arginine (l-Arg) enhances the stability of the hexacoordinate NO form. The EPR spectrum of the NO adduct of the enzyme-substrate complex has an increased g-anisotropy and well-resolved hyperfine couplings due to the 14N of nitric oxide. Significant perturbations in the NO EPR spectrum were observed upon Nomega-monomethyl-L-Arg and Nomega-hydroxy-L-Arg binding. The perturbations in the EPR spectrum indicate that L-Arg and its derivatives bind on the distal site of the heme in very close proximity to the bound NO to cause alterations in the heme-NO coordination structure. Interactions between the bound NO and the substrate or its analogues appear to affect the Fe-NO geometry, resulting in the observed spectral changes. We infer that analogous interactions with oxygen might be involved in the hydroxylation events during enzyme catalysis of nitric oxide synthase.

Comparative CO photolysis kinetics studies on wildtype and autoregulatory (AR) insert-deletion mutant of rat nNOS holoenzyme were conducted to directly investigate the role of the unique AR insert in the catalytically significant FMN-heme intraprotein electron transfer (IET). Although the amplitude of the IET kinetic traces was decreased two-to three-fold, the AR deletion did not change the rate constant for the calmodulin-controlled IET. This suggests that the ratelimiting conversion of the electron-accepting state to a new electron-donating (output) state does not involve interactions with the AR insert, but that AR may stabilize the output state once it is formed.

No abstract

The presence of NADPH diaphorase staining was compared with the immunohistochemical localization of four NADPH-dependent enzymes-neuronal (type I), inducible (type II), and endothelial (type III) nitric oxide synthase (NOS) and cytochrome P450 reductase. Cell types that were immunoreactive for the NADPH-dependent enzymes were also stained for NADPH diaphorase, suggesting that endothelial and neuronal NOS and cytochrome P450 reductase all show NADPH diaphorase activity in formaldehyde-fixed tissue. However, in some tissues, the presence of NADPH diaphorase staining did not coincide with the presence of any of the NADPH-dependent enzymes we examined. In vascular endothelial cells, the punctate pattern of staining observed with NADPH diaphorase histochemistry was identical to that seen following immunohistochemistry using antibodies to endothelial NOS. In enteric and pancreatic neurons and in skeletal muscle, the presence of NADPH diaphorase staining correlated with the presence of neuronal NOS. In the liver, sebaceous glands of the skin, ciliated epithelium, and a subpopulation of the cells in the subserosal glands of the trachea, zona glomerulosa of the adrenal cortex, and epithelial cells of the lacrimal and salivary glands, the presence of NADPH diaphorase staining coincided with the presence of cytochrome P450 reductase immunoreactivity. In epithelial cells of the renal tubules and zona fasciculata and zona reticularis of the adrenal cortex, NADPH diaphorase staining was observed that did not coincide with the presence of any of the enzymes. Inducible NOS was not observed in any tissue. Thus, while tissues that demonstrate immunoreactivity for neuronal and endothelial NOS also stain positively for NADPH diaphorase activity, the presence of NADPH diaphorase staining does not reliably or specifically indicate the presence of one or more NOS isoforms.