Diabetes impairs heart mitochondrial function without changes in resting cardiac performance

Diabetes is a chronic disease associated to a cardiac contractile dysfunction that is not attributable to underlying coronary artery disease or hypertension, and could be consequence of a progressive deterioration of mitochondrial function. We hypothesized that impaired mitochondrial function precedes Diabetic Cardiomyopathy. Thus, the aim of this work was to study the cardiac performance and heart mitochondrial function of diabetic rats, using an experimental model of type I Diabetes. Rats were sacrificed after 28 days of Streptozotocin injection (STZ, 60 mg kg⁻¹, ip.). Heart O₂ consumption was declined, mainly due to the impairment of mitochondrial O₂ uptake. The mitochondrial dysfunction observed in diabetic animals included the reduction of state 3 respiration (22%), the decline of ADP/O ratio (∼15%) and the decrease of the respiratory complexes activities (22–26%). An enhancement in mitochondrial H₂O₂ (127%) and NO (23%) production rates and in tyrosine nitration (58%) were observed in heart of diabetic rats, with a decrease in Mn-SOD activity (∼50%). Moreover, a decrease in contractile response (38%), inotropic (37%) and lusitropic (58%) reserves were observed in diabetic rats only after a β‐adrenergic stimulus. Therefore, in conditions of sustained hyperglycemia, heart mitochondrial O₂ consumption and oxidative phosphorylation efficiency are decreased, and H₂O₂ and NO productions are increased, leading to a cardiac compromise against a work overload. This mitochondrial impairment was detected in the absence of heart hypertrophy and of resting cardiac performance changes, suggesting that mitochondrial dysfunction could precede the onset of diabetic cardiac failure, being H₂O₂, NO and ATP the molecules probably involved in mitochondrion-cytosol signalling.     

Mechanism of ATP synthesis by mitochondrial ATP synthase from beef heart

Previous studies of the rate constants for the elementary steps of ATP hydrolysis by the soluble and membrane-bound forms of beef heart mitochondrial F₁ supported the proposal that ATP is formed in high-affinity catalytic sites of the enzyme with little or no change in free energy and that the major requirement for energy in oxidative phosphorylation is for the release of product ATP.The affinity of the membrane-bound enzyme for ATP during NADH oxidation was calculated from the ratio of the rate constants for the forward binding step (k ₊₁) and the reverse dissociation step (k _–1).k _–1 was accelerated several orders of magnitude by NADH oxidation. In the presence of NADH and ADP an additional enhancement ofk _–1 was observed. These energy-dependent dissociations of ATP were sensitive to the uncoupler FCCP.k ₊₁ was affected little by NADH oxidation. The dissociation constant (K _d ATP) increased many orders of magnitude during the transition from nonenergized to energized states.     

Voltage-driven ATP synthesis by beef heart mitochondrial F0F1-ATPase

The F0F1-ATPase of the inner mitochondrial membrane catalyzes the conversion of a proton electrochemical energy into the chemical bond energy of ATP (Boyer, P.D., Chance, B., Ernster, L., Mitchell, P., Racker, E., and Slater, E.C. (1977) Annu. Rev. Biochem. 46, 955-1026). To assess the role of the membrane potential (delta psi) in this process and to study the effect of very short pulses on ATP synthesis, we employed a high voltage pulsation method (Kinosita, K., and Tsong, T.Y. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 1923-1927) to induce a delta psi of controlled magnitude and duration in a suspension of submitochondrial particles and F0F1-ATPase vesicles. Cyanide-treated submitochondrial particles were exposed to electric pulses of 10-30 kV/cm of magnitude (generating a peak delta psi of 150-450 mV) and 1-100 microseconds duration. Net [32P]ATP synthesis from [32P]Pi and ADP was observed with maximal values of 410 pmol/mg X pulse for a 30 kV/cm-100-microseconds pulse. This corresponds to a yield of 10-12 mol of ATP per mol of F0F1 complex per pulse. As many as 4 nmol/mg were produced after pulsing the same sample 8 times. By varying the ionic strength of the suspending medium, and consequently the pulse width, it is clearly shown that the synthesis was electrically driven and did not correlate with Joule heating of the sample. Titrations using specific inhibitors and ionophores were performed. The voltage-induced ATP synthesis was 50% inhibited by 0.11 microgram/mg of oligomycin and 2.4 nmol/mg of N,N'-dicyclohexylcarbodiimide. Ionophores and uncouplers had varying degrees of inhibition. The dependence of ATP synthesis on pulse width was nonlinear, exhibiting a threshold at 10 microseconds and a biphasic behavior above this value. Isolated F0F1-ATPase reconstituted into asolectin vesicles also synthesized ATP when pulsed with electric fields. A 35 kV/cm pulse induced the synthesis of 115 pmol of ATP per mg of protein, which corresponds to approximately 0.34 mol of ATP per mol of F0F1-ATPase. This synthesis was also sensitive to oligomycin and dicyclohexylcarbodiimide. The possibility of turnover of the ATPase in microseconds is considered.     

Catalysis of partial reactions of ATP synthesis by beef heart mitochondrial adenosine triphosphatase

We have found that when the ATP hydrolysis activity of beef heart mitochondrial adenosine triphosphatase (F1) is eliminated by either cold treatment or chemical modification, the enzyme attains the ability to catalyze the Pi in equilibrium ATP exchange reaction. The ATP hydrolysis activity of isolated F1 was lost upon chemical modification by phenyglyoxal, butanedione, or 7-chloro-4-nitrobenzene-2-oxa-1,3-diazole. The F1 thus chemically modified was able to catalyze an ADP-dependent Pi in equilibrium ATP exchange reaction. In addition F1 that had been cold-treated to eliminate ATP hydrolysis activity, also catalyzed the Pi in equilibrium ATP exchange reaction. The Pi in equilibrium ATP exchange catalyzed by modified F1 was shown to be totally inhibited by the F1-specific antibiotic efrapeptin. We have previously shown that isolated beef heart mitochondrial ATPase will catalyze the formation of a transition state analog of the ATP synthesis reaction (Bossard, M. J., Vik, T. A., and Schuster, S. M. (1980) J. Biol. Chem. 255, 5342-5346). While the F1-catalyzed ATP hydrolysis activity was lost rapidly upon chemical modification or cold treatment, the ability of the enzyme to produce Pi . adenosine 5'-diphosphate (chromium(III) salt) from phosphate and monodentate adenosine 5'-diphosphate (chromium(III) salt) was unimpaired. The implications of these data with regard to the mechanism of ATP synthesis are discussed.     

tRNA-triggered ATP hydrolysis and generation of membrane potential by the leishmania mitochondrial tRNA import complex

Translocation of tRNAs across mitochondrial membranes is a receptor-mediated active transport process requiring ATP. A large tRNA import complex from the inner membrane of Leishmania mitochondria catalyzes translocation into phospholipid vesicles. In this reconstituted system, the import substrate tRNA(Tyr)(GUA) specifically stimulated hydrolysis of ATP within the vesicles, with the subsequent generation of a membrane potential by pumping out of protons, as shown by the protonophore-sensitive uptake of the potential-sensitive dye rhodamine 123. Generation of membrane potential was dependent on ATP hydrolysis, and inhibited by oligomycin, recalling the proton-translocation mechanism of the respiratory F(1)-F(0)-ATPase. For translocation of tRNA, ATP could be replaced by low pH of the medium, but proton-dependent import was resistant to oligomycin. Moreover, ATP hydrolysis, generation of membrane potential and tRNA uptake were inhibited by carboxyatractyloside, a specific inhibitor of mitochondrial ATP-ADP translocase, implying an ATP requirement within the vesicles. These observations imply a gating mechanism in which tRNA, on binding to its receptor, triggers the energetic activation of the complex, leading to the opening of import channels.     

ATP synthesis, mitochondrial function, and steroid biosynthesis in rodent primary and tumor Leydig cells

Previous studies in MA-10 tumor Leydig cells demonstrated that disruption of the mitochondrial electron-transport chain (ETC), membrane potential (ΔΨ(m)), or ATP synthesis independently inhibited steroidogenesis. In contrast, studies of primary Leydig cells indicated that the ETC, ΔΨ(m), and ATP synthesis cooperatively affected steroidogenesis. These results suggest significant differences between the two systems and call into question the extent to which results from tumor Leydig cells relate to primary cells. Thus, to further understand the similarities and differences between the two systems as well as the impact of ATP disruption on steroidogenesis, we performed comparative studies of MA-10 and primary Leydig cells under similar conditions of mitochondrial disruption. We show that mitochondrial ATP synthesis is critical for steroidogenesis in both primary and tumor Leydig cells. However, in striking contrast to primary cells, perturbation of ΔΨ(m) in MA-10 cells did not substantially decrease cellular ATP content, a perplexing finding because ΔΨ(m) powers the mitochondrial ATP synthase. Further studies revealed that a significant proportion of cellular ATP in MA-10 cells derives from glycolysis. In contrast, primary cells appear to be almost completely dependent on mitochondrial respiration for their energy provision. Inhibitor studies also suggested that the MA-10 ETC is impaired. This work underscores the importance of mitochondrial ATP for hormone-stimulated steroid production in both MA-10 and primary Leydig cells while indicating that caution must be exercised in extrapolating data from tumor cells to primary tissue.     

Mitochondrial K(ATP) channel activation reduces anoxic injury by restoring mitochondrial membrane potential

Mitochondrial membrane potential (DeltaPsi(m)) is severely compromised in the myocardium after ischemia-reperfusion and triggers apoptotic events leading to cell demise. This study tests the hypothesis that mitochondrial ATP-sensitive K(+) (mitoK(ATP)) channel activation prevents the collapse of DeltaPsi(m) in myocytes during anoxia-reoxygenation (A-R) and is responsible for cell protection via inhibition of apoptosis. After 3-h anoxia and 2-h reoxygenation, the cultured myocytes underwent extensive damage, as evidenced by decreased cell viability, compromised membrane permeability, increased apoptosis, and decreased ATP concentration. Mitochondria in A-R myocytes were swollen and fuzzy as shown after staining with Mito Tracker Orange CMTMRos and in an electron microscope and exhibited a collapsed DeltaPsi(m), as monitored by 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolcarbocyanine iodide (JC-1). Cytochrome c was released from mitochondria into the cytosol as demonstrated by cytochrome c immunostaining. Activation of mitoK(ATP) channel with diazoxide (100 micromol/l) resulted in a significant protection against mitochondrial damage, ATP depletion, cytochrome c loss, and stabilized DeltaPsi(m). This protection was blocked by 5-hydroxydecanoate (500 micromol/l), a mitoK(ATP) channel-selective inhibitor, but not by HMR-1098 (30 micromol/l), a putative sarcolemmal K(ATP) channel-selective inhibitor. Dissipation of DeltaPsi(m) also leads to opening of mitochondrial permeability transition pore, which was prevented by cyclosporin A. The data support the hypothesis that A-R disrupts DeltaPsi(m) and induces apoptosis, which are prevented by the activation of the mitoK(ATP) channel. This further emphasizes the therapeutic significance of mitoK(ATP) channel agonists in the prevention of ischemia-reperfusion cell injury.     

Effects of dibromothymoquinone on mung bean mitochondrial electron transfer and membrane fluidity

The effects of the quinone analog dibromothymoquinone on electron transfer in isolated mung bean mitochondria are described. Both the main, cyanidesensitive and the alternate, cyanide-insensitive pathways are inhibited by dibromothymoquinone but in markedly different fashions. Half-maximal inhibition appeared at 40 μM and 20 μM dibromothymoquinone for the cyanide-sensitive and alternate pathways, respectively. With succinate as the electron donor, dibromothymoquinone inhibited the alternate pathway at a single site; showing a mixed, non-competitive type inhibition. On the succinate, cyanide-sensitive pathway dibromothymoquinone showed two sites of inhibition and neither coincides with the site of inhibition associated with the alternate pathway. With malate as the electron donor, two sites of inhibition by dibromothymoquinone were observed regardless of the pathway measured.Dibromothymoquinone also inhibited the rate of valinomycin-induced swelling of isolated mung bean mitochondria. Steady-state kinetics showed the inhibition to be non-competitive with respect to valinomycin. Additionally dibromothymoquinone was observed to increase the fluorescence polarization associated with the hydrophobic probe 1,6-diphenylhexatriene. The results indicated that dibromothymoquinone decreased the fluidity of the inner mitochondrial membrane and suggested that the inhibition of mitochondrial electron transfer by dibromothymoquinone may be associated with this decrease in membrane fluidity.The relationship of the multisite nature of the inhibition of electron transfer by dibromothymoquinone and the possible role of mobile electron carriers such as ubiquinone on the main and alternate respiratory pathways of higher plants is discussed.     

Stretch-induced injury alters mitochondrial membrane potential and cellular ATP in cultured astrocytes and neurons

Energy deficit after traumatic brain injury (TBI) may alter ionic homeostasis, neurotransmission, biosynthesis, and cellular transport. Using an in vitro model for TBI, we tested the hypothesis that stretch-induced injury alters mitochondrial membrane potential (delta(psi)m) and ATP in astrocytes and neurons. Astrocytes, pure neuronal cultures, and mixed neuronal plus glial cultures grown on Silastic membranes were subjected to mild, moderate, and severe stretch. After injury, delta(psi)m was measured using rhodamine-123, and ATP was quantified with a luciferin-luciferase assay. In astrocytes, delta(psi)m dropped significantly, and ATP content declined 43-52% 15 min after mild or moderate stretch but recovered by 24 h. In pure neurons, delta(psi)m declined at 15 min only in the severely stretched group. At 48 h postinjury, delta(psi)m remained decreased in severely stretched neurons and dropped in moderately stretched neurons. Intracellular ATP content did not change in any group of injured pure neurons. We also found that astrocytes and neurons release ATP extracellularly following injury. In contrast to pure neurons, delta(psi)m in neurons of mixed neuronal plus glial cultures declined 15 min after mild, moderate, or severe stretch and recovered by 24-48 h. ATP content in mixed cultures declined 22-28% after mild to severe stretch with recovery by 24 h. Our findings demonstrate that injury causes mitochondrial dysfunction in astrocytes and suggest that astrocyte injury alters mitochondrial function in local neurons.     

ATP synthesis kinetics and mitochondrial function in the postischemic myocardium as studied by 31P NMR

The effects of ischemia on mitochondrial function and the unidirectional rate of ATP synthesis (Pi----ATP rate) were studied using a Langendorff-perfused heart preparation and 31P NMR spectroscopy. There was significant postischemic depression of mechanical function assessed as the heart rate pressure product, and the myocardial oxygen consumption rate at a given rate pressure product was elevated. Experiments performed on glucose- and pyruvate-perfused hearts demonstrated the presence of a large contribution to the unidirectional Pi----ATP rate catalyzed by glyceraldehyde-3-phosphate dehydrogenase and phosphoglycerate kinase. This rate was much greater than the maximal glucose utilization rate in the myocardium, demonstrating that the glyceraldehyde-3-phosphate dehydrogenase/phosphoglycerate kinase reactions are near equilibrium both before and after ischemia. In the pyruvate-perfused postischemic hearts, the glycolytic contribution was eliminated and the net rate of ATP synthesis by oxidative phosphorylation was measurable. Despite the reduced mechanical function and increased myocardial oxygen consumption rate, the ratio of the net rate of ATP synthesis by oxidative phosphorylation to oxygen consumption rate (the P:O ratio) was not altered subsequent to ischemia (2.34 +/- 0.12 and 2.36 +/- 0.09 in normal and postischemic hearts, respectively). Therefore, mitochondrial uncoupling cannot be the cause of postischemic depression in mechanical function; instead, the data suggest the existence of ischemia-induced inefficiency in ATP utilization.     

A deafness-associated tRNAHis mutation alters the mitochondrial function,ROS production and membrane potential

In this report, we investigated the molecular genetic mechanism underlying the deafness-associated mitochondrial tRNA^His 12201T>C mutation. The destabilization of a highly conserved base-pairing (5A-68U) by the m.12201T>C mutation alters structure and function of tRNA^His. Using cybrids constructed by transferring mitochondria from lymphoblastoid cell lines derived from a Chinese family into mtDNA-less (ρ^o) cells, we showed ∼70% decrease in the steady-state level of tRNA^His in mutant cybrids, compared with control cybrids. The mutation changed the conformation of tRNA^His, as suggested by slower electrophoretic mobility of mutated tRNA with respect to the wild-type molecule. However, ∼60% increase in aminoacylated level of tRNA^His was observed in mutant cells. The failure in tRNA^His metabolism was responsible for the variable reductions in seven mtDNA-encoded polypeptides in mutant cells, ranging from 37 to 81%, with the average of ∼46% reduction, as compared with those of control cells. The impaired mitochondrial translation caused defects in respiratory capacity in mutant cells. Furthermore, marked decreases in the levels of mitochondrial ATP and membrane potential were observed in mutant cells. These mitochondrial dysfunctions caused an increase in the production of reactive oxygen species in the mutant cells. The data provide the evidence for a mitochondrial tRNA^His mutation leading to deafness.     

Inhibition of steady-state mitochondrial ATP synthesis by bicarbonate, an activating anion of ATP hydrolysis.

Bicarbonate, an activating anion of ATP hydrolysis, inhibited ATP synthesis coupled to succinate oxidation in beef heart submitochondrial particles but diminished the lag time and increased the steady-state velocity of the (32)Pi-ATP exchange reaction. The latter effects exclude the possibility that bicarbonate is inducing an intrinsic uncoupling between ATP hydrolysis and proton translocation at the level of F(1)F(o) ATPase. The inhibition of ATP synthesis was competitive with respect to ADP at low fixed [Pi], mixed at high [Pi] and non-competitive towards Pi at any fixed [ADP]. From these results we can conclude that (i) bicarbonate does not bind to a Pi site in the mitochondrial F(1); (ii) it competes with the binding of ADP to a low-affinity site, likely the low-affinity non-catalytic nucleotide binding site. It is postulated that bicarbonate stimulates ATP hydrolysis and inhibits ATP synthesis by modulating the relative affinities of the catalytic site for ATP and ADP.     

Effects of dibromothymoquinone on mung bean mitochondrial electron transfer and membrane fluidity.

The effects of the quinone analog dibromothymoquinone on electron transfer in isolated mung bean mitochondria are described. Both the main, cyanide-sensitive and the alternate, cyanide-insensitive pathways are inhibited by dibromothymoquinone but in markedly different fashions. Half-maximal inhibition appeared at 40 microM and 20 microM dibromothymoquinone for the cyanide-sensitive and alternate pathways, respectively. With succinate as the electron donor, dibromothymoquinone inhibited the alternate pathway at a single site; showing a mixed, non-competitive type inhibition. On the succinate, cyanide-sensitive pathway dibromothymoquinone showed two sites of inhibition and neither coincides with the site of inhibition associated with the alternate pathway. With malate as the electron donor, two sites of inhibition by dibromothymoquinone were observed regardless of the pathway measured. Dibromothymoquinone also inhibited the rate of valinomycin-induced swelling of isolated mung bean mitochondria. Steady-state kinetics showed the inhibition to be non-competitive with respect to valinomycin. Additionally dibromothymoquinone was observed to increase the fluorescence polarization associated with the hydrophobic probe 1,6-diphenylhexatriene. The results indicated that dibromothymoquinone decreased the fluidity of the inner mitochondrial membrane and suggested that the inhibition of mitochondrial electron transfer by dibromothymoquinone may be associated with this decrease in membrane fluidity. The relationship of the multisite nature of the inhibition of electron transfer by dibromothymoquinone and the possible role of mobile electron carriers such as ubiquinone on the main and alternate respiratory pathways of higher plants is discussed.     

Cytochrome c maintains mitochondrial transmembrane potential and ATP generation after outer mitochondrial membrane permeabilization during the apoptotic process

During apoptosis, cytochrome c is released into the cytosol as the outer membrane of mitochondria becomes permeable, and this acts to trigger caspase activation. The consequences of this release for mitochondrial metabolism are unclear. Using single-cell analysis, we found that when caspase activity is inhibited, mitochondrial outer membrane permeabilization causes a rapid depolarization of mitochondrial transmembrane potential, which recovers to original levels over the next 30-60 min and is then maintained. After outer membrane permeabilization, mitochondria can use cytoplasmic cytochrome c to maintain mitochondrial transmembrane potential and ATP production. Furthermore, both cytochrome c release and apoptosis proceed normally in cells in which mitochondria have been uncoupled. These studies demonstrate that cytochrome c release does not affect the integrity of the mitochondrial inner membrane and that, in the absence of caspase activation, mitochondrial functions can be maintained after the release of cytochrome c.     

The effect of venturicidin on light and oxygen-dependent electron transport,proton translocation,membrane potential development and ATP synthesis in intact cells of Rhodopseudomonas capsulata

Venturicidin behaves as an orthodox energy transfer inhibitor in intact cells of Rhodopseudomonas capsulata as judged by the following criteria. 1. It led to inhibition of respiration. Inhibition was relieved by low concentrations of uncoupling agent. 2. It enhanced light-induced and oxygen dependent H⁺ efflux. 3. It stimulated light-induced and oxygen dependent carotenoid band shifts. The rate of decay of the band shifts after short flash excitation was decreased in the presence of venturicidin. 4. It stimulated light-induced and oxygen dependent butyltriphenylphosphonium uptake. 5. It inhibited the rise in cellular ATP concentration accompanying either photosynthesis or respiration.     

Influence of cytosolic and mitochondrial Ca2+, ATP,mitochondrial membrane potential,and calpain activity on the mechanism of neuron death induced by 3-nitropropionic acid

3-Nitropropionic acid (3NP), an irreversible inhibitor of succinate dehydrogenase, induces both rapid necrotic and slow apoptotic death in rat hippocampal neurons. Low levels of extracellular glutamate (10 microM) shift the 3NP-induced cell death mechanism to necrosis, while NMDA receptor blockade results in predominantly apoptotic death. In this study, we examined the 3NP-induced alterations in free cytosolic and mitochondrial calcium levels, ATP levels, mitochondrial membrane potential, and calpain and caspase activity, under conditions resulting in the activation of apoptotic and necrotic pathways. In the presence of 10 microM glutamate, 3NP administration resulted in a massive elevation in [Ca(2+)](c) and [Ca(2+)](m), decreased ATP, rapid mitochondrial membrane depolarization, and a rapid activation of calpain but not caspase activity. In the presence of the NMDA receptor antagonist MK-801, 3NP did not induce a significant elevation of [Ca(2+)](c) within the 24h time period examined, nor increase [Ca(2+)](m) within 1h. ATP was maintained at control levels during the first hour of treatment, but declined 64% by 16h. Calpain and caspase activity were first evident at 24h following 3NP administration. 3NP treatment alone resulted in a more rapid decline in ATP, more rapid calpain activation (within 8h), and elevated [Ca(2+)](m) as compared to the results obtained with added MK-801. Together, the results demonstrate that 3NP-induced necrotic neuron death is associated with a massive calcium influx through NMDA receptors, resulting in mitochondrial depolarization and calpain activation; while 3NP-induced apoptotic neuron death is not associated with significant elevations in [Ca(2+)](c), nor with early changes in [Ca(2+)](m), mitochondrial membrane potential, ATP levels, or calpain activity.     

The low affinity neurotensin receptor antagonist levocabastine impairs brain nitric oxide synthesis and mitochondrial function by independent mechanisms

Neurotensin is known to inhibit neuronal Na⁺, K⁺‐ATPase, an effect that is rescued by nitric oxide (NO) synthase inhibition. However, whether the neurotensinergic and the nitrergic systems are independent pathways, or are mechanistically linked, remains unknown. Here, we addressed this issue and found that the administration of low affinity neurotensin receptor (NTS2) antagonist, levocabastine (50 μg/kg, i.p.) inhibited NO synthase (NOS) activity by 74 and 42% after 18 h in synaptosomal and mitochondrial fractions isolated from the Wistar rat cerebral cortex, respectively; these effects disappeared 36 h after levocabastine treatment. Intriguingly, whereas neuronal NOS protein abundance decreased (by 56%) in synaptosomes membranes, it was enhanced (by 86%) in mitochondria 18 h after levocabastine administration. Levocabastine enhanced the respiratory rate of synaptosomes in the presence of oligomycin, but it failed to alter the spare respiratory capacity; furthermore, the mitochondrial respiratory chain (MRC) complexes I–IV activities were severely diminished by levocabastine administration. The inhibition of NOS and MRC complexes activities were also observed after incubation of synaptosomes and mitochondria with levocabastine (1 μM) in vitro. These data indicate that the NTS2 antagonist levocabastine regulates NOS expression and activity at the synapse, suggesting an interrelationship between the neurotensinergic and the nitrergic systems. However, the bioenergetics effects of NTS2 activity inhibition are likely to be independent from the regulation of NO synthesis.