The effect of N-nitro-L-arginine and aminoguanidine treatment on changes in constitutive and inducible nitric oxide synthases in the spinal cord after sciatic nerve transection

Ca2+-dependent and Ca2+-independent nitric oxide synthase (NOS) activity, and neuronal and inducible NOS immunoreactivity (nNOS-IR and iNOS-IR), were investigated in the rabbit lower lumbar spinal cord after i) sciatic nerve transection and survival of experimental animals for 2 weeks, ii) treatment of animals with N-nitro-L-arginine (NNLA), an inhibitor of nNOS dosed at 20 mg/b.w. for 12 days, and iii) after treatment of animals with the inducible NOS (iNOS) inhibitor, aminoguanidine, dosed at 100 mg/b.w. for 4 and 12 days. Our attention was focused on the dorsal part of L4-L6 segments receiving sensory inputs from the sciatic nerve, and on the ventral part consisting of sciatic nerve motor neurons. Sciatic nerve transection increased Ca2+-dependent NOS activity and the density of nNOS in the dorsal part of the spinal cord on the ipsilateral side. NNLA treatment effectively reduced nNOS-IR in both the dorsal horn and the dorsal column, and decreased Ca2+-dependent NOS activity in the lower lumbar segments. Immunocytochemical analysis disclosed the up-regulation of iNOS immunoreactive staining after peripheral axotomy in alpha-motoneurons. The changes in iNOS expression and Ca2+-independent NOS activity were not significantly corrected by aminoguanidine treatment for 4 days. Long-lasting iNOS inhibition decreased Ca2+-independent NOS activity, but caused motor neuron degeneration and mediated small necrotic foci in the ventrolateral portion of the ventral horn. The results of the present study provide evidence that constitutive NOS inhibition by NNLA is more effective than specific long-lasting inhibition of iNOS by aminoguanidine treatment.     

Nitric oxide in the cerebral cortex of amyloid-precursor protein (SW) Tg2576 transgenic mice

Changes in the amyloid-peptide (Abeta), neuronal and inducible nitric oxide (NO)synthase (nNOS, iNOS), nitrotyrosine, glial fibrillary acidic protein, and lectin from Lycopersicon esculentum (tomato) were investigated in the cerebral cortex of transgenic mice (Tg2576) to amyloid precursor protein (APP), by immunohistochemistry (bright light, confocal, and electron microscopy). The expression of nitrergic proteins and synthesis of nitric oxide were analyzed by immunoblotting and NOS activity assays, respectively. The cerebral cortex of these transgenic mice showed an age-dependent progressive increase in intraneuronal aggregates of Abeta-peptide and extracellular formation of senile plaques surrounded by numerous microglial and reactive astrocytes. Basically, no changes to nNOS reactivity or expression were found in the cortical mantle of either wild or transgenic mice. This reactivity in wild mice corresponded to numerous large type I and small type II neurons. The transgenic mice showed swollen, twisted, and hypertrophic preterminal and terminal processes of type I neurons, and an increase of the type II neurons. The calcium-dependent NOS enzymatic activity was higher in wild than in the transgenic mice. The iNOS reactivity, expression and calcium-independent enzymatic activity increased in transgenic mice with respect to wild mice, and were related to cortical neurons and microglial cells. The progressive elevation of NO production resulted in a specific pattern of protein nitration in reactive astrocytes. The ultrastructural study carried out in the cortical mantle showed that the neurons contained intracellular aggregates of Abeta-peptide associated with the endoplasmic reticulum, mitochondria, and Golgi apparatus. The endothelial vascular cells also contained Abeta-peptide deposits. This transgenic model might contribute to understand the role of the nitrergic system in the biological changes related to neuropathological progression of Alzheimer's disease.     

Differential roles of neuronal and endothelial nitric oxide synthases during carrageenan-induced inflammatory hyperalgesia

The present study investigated the role of neuronal nitric oxide synthase (nNOS) in carrageenan-induced inflammatory pain by combining genomic and pharmacological strategies. Intrathecal injection of the nNOS inhibitor 7-nitroindazole dose-dependently inhibited carrageenan-induced thermal hyperalgesia in both early and late phases in wild-type mice. However in nNOS knockout mice, carrageenan-induced thermal hyperalgesia remained intact in the early phase but was reduced in the late phase. Spinal Ca2+ -dependent nitric oxide synthase (NOS) activity in nNOS knockout mice was significantly lower than that in wild-type mice. Following carrageenan injection, although the spinal Ca2+ -dependent NOS activity in both wild-type and knockout mice increased, the enzyme activity in nNOS knockout mice reached a level similar to that in wild-type mice. On the other hand, no significant difference in spinal Ca2+ -independent NOS activity was noted between wild-type and nNOS knockout mice before and after carrageenan injection. Furthermore, intrathecal administration of the endothelial NOS (eNOS) inhibitor L-N5-(1-iminoethyl)-ornithinein nNOS knockout mice inhibited the thermal hyperalgesia in both early and late phases, though this inhibitor had no effect in wild-type mice. Meanwhile, Western blot showed that eNOS expression in the spinal cord of nNOS knockout mice was up-regulated compared with wild-type mice; immunohistochemical staining showed that the spinal eNOS was mainly distributed in superficial laminae of the dorsal horn. Finally, double staining with confocal analysis showed that the enhanced spinal eNOS was expressed in astrocytes, but not in neurons. Our current results indicate that nNOS plays different roles in the two phases of carrageenan-induced inflammatory pain. In this model, enhanced spinal eNOS appears to compensate for the role of nNOS in nNOS knockout mice.     

Pseudorabies virus-induced expression of nitric oxide synthase isoforms

In addition to its role as a neurotransmitter, studies have postulated both neuroprotective and neurotoxic roles for nitric oxide (NO) generated in response to infections with neurotropic viruses. This study examined the expression of neuronal nitric oxide synthase (nNOS) and inducible nitric oxide synthase (iNOS) isoforms of NOS induced by neuronal infection with virulent and attenuated strains of pseudorabies virus (PRV). Caudal brainstem neurons infected by peripheral inoculation of the viscera served as the model system. Neuronal infection induced the expression of nNOS and iNOS, but the timing and the apparent magnitude of NOS expression varied according to the virulence of the infecting strain of virus. Expression of nNOS was observed in infected neurons that did not express this enzyme in control animals, and the onset of expression was earlier in animals infected with virulent PRV. Expression of iNOS was largely restricted to monocytes and macrophages that invaded the brain in response to PRV infection. These iNOS-expressing cells were observed earlier in animals infected with the virulent virus, and were differentially concentrated in areas exhibiting virus-induced neuropathology. Collectively, these data suggest functionally diverse roles for NO in the brain response to PRV neuronal infection.     

The nitric oxide donor LA 419 decreases ischemic brain damage

Stroke represents a major clinical problem with limited available therapeutic treatments. Nitric oxide (NO) and the enzymes that produce it are involved in the pathogenesis of this disease. Here we investigated whether the novel NO donor LA 419 was able to ameliorate the consequences of stroke in an experimental model of global ischemia. We observed a sharp increase in the amounts of inducible NO synthase (iNOS) and nitrotyrosine in the cerebral cortex of experimental rats and a moderate increase of neuronal NO synthase (nNOS), as demonstrated by immunohistochemistry, Western blotting, and enzymatic activity assays. Treatment of these animals with LA 419 completely prevented ischemia-induced upregulation of nitrergic markers. Magnetic resonance imaging of the experimental brains showed a marked decrease in apparent diffusion coefficient (ADC) following ischemia-reperfusion, which was significantly corrected by pre-treatment with LA 419. These results clearly show that LA 419 is an efficient modulator of NO-related pathophysiological events and could eventually be used for the treatment of patients with cerebrovascular pathologies.     

Neuronal nitric oxide synthase immunoreactivity in ependymal cells during early postnatal development

Neuronal nitric oxide synthase (nNOS) immunoreactivity was observed in ependymal cell layer of the central canal of spinal cord of neonatal rats (2-20 days old). Neuronal nitric oxide synthase immunoreactivity was present in postnatal day 2 and this immunoreactivity gradually disappeared by postnatal day 16. The progressive decrease in nNOS staining with the increasing postnatal age may suggest that nNOS staining paralleled the maturation of the central canal and may also suggest that nNOS activity plays a role in the development of the ependymal cells.     

Changes in mRNA of protein inhibitor of neuronal nitric oxide synthase following facial nerve transection

Protein inhibitor of neuronal nitric oxide synthase (PIN) is reported as the protein inhibiting neuronal nitric oxide synthase (nNOS) activity by preventing dimerization of nNOS. It was also reported that PIN inhibits the activity of all nitric oxide synthase (NOS) isozymes. We examined the effects of facial nerve transection on PIN mRNA and NOS expression by in situ hybridization for PIN mRNA and nicotinamide adenine dinucleotide phosphate diaphorase (NADPH-d) staining. PIN mRNA was initially expressed and transiently increased from 3 to 5 days and returned to the basal level at 7 days after axotomy in the motoneurons of the facial nucleus. NADPH-d-positive motoneurons were found from 7 days post-operation in the facial nucleus. These results suggest that PIN may interact with NOS from 7 days post-operation.     

Localization of inducible nitric oxide synthase to mast cells during ischemia/reperfusion injury of skeletal muscle

Nitric oxide contributes to tissue necrosis after ischemia-reperfusion (IR). A biochemical and immunohistochemical study was made of the amounts and localization of both Ca++-independent nitric oxide synthase (NOS) II and Ca++-dependent (NOS I and NOS III) in rat skeletal muscle after ischemia and 0.5, 2, 8, 16, and 24 hours reperfusion. NOS II was not detectable in control muscle or during ischemia, was first detected after 2 hours reperfusion, increased further by 8 hours, and remained elevated at 24 hours. Both NOS II and nitrotyrosine, a marker of peroxynitrite formation, were localized exclusively to mast cells except after 24 hours reperfusion when some macrophages and neutrophils also showed positive immunoreactivity. Mast cells underwent extensive degranulation during reperfusion. NOS I was not detected in injured or control muscle. The level of NOS III, which was localized to the endothelium of blood vessels of all sizes in control muscle, decreased progressively during ischemia and reperfusion to reach undetectable levels after 16 hours reperfusion. These findings indicate that most of the nitric oxide formed during IR injury is generated by NOS II located almost exclusively in mast cells.     

Distribution and expression pattern of the nitrergic system in the cerebellum of the sheep

The nitrergic system produces nitric oxide as an atypical neurotransmitter in the nervous system. Nitric oxide is produced from l-arginine through specific enzymes known as nitric oxide synthases. Of these, the more abundant form in neurons is the constitutive neuronal nitric oxide synthase, although the inducible isoform can be expressed as well, especially following stress or other injuries. The excessive formation of nitric oxide results in protein nitration, particularly at tyrosine residues, thus the presence of nitrotyrosine can be used as a marker of nitric oxide production. In previous studies we have shown the distribution of the components of the nitrergic system in the cerebellum of rodents, where neuronal nitric oxide synthase immunoreactivity was present in stellate and basket cells, and occasionally in granule cells. Here, we present evidence that in the sheep, as a model of larger mammals, most cerebellar neurons display an intense immunostaining for neuronal nitric oxide synthase, including unipolar brush cells, and Lugaro and Golgi neurons, which are not immunoreactive in rodents. In addition, weak immunoreactivity for inducible nitric oxide synthase and nitrotyrosine was found in particular cell types, indicating a basal expression for these markers. Our results suggest a larger dependence on the nitrergic system for the cerebella of larger mammals. Since this increase happens in both activating and inhibitory neurons of the cerebellar circuitry, we propose that in these animals there is a higher steady-state regulation of the cerebellum based on nitric oxide.     

Subcellular localization of neuronal nitric oxide synthase in the superficial gray layer of the rat superior colliculus

The superficial layers of the rat superior colliculus (sSC) receive innervation from retina and include nitric oxide synthase (NOS)-immunoreactive neurons. We used electron microscopic immunocytochemistry to assess the subcellular localization of neuronal NOS (nNOS) in the sSC. nNOS immunoreactivity was detected on the external membrane of mitochondria, endoplasmic reticulum, in pre- and postsynaptic profiles and also diffusely distributed in the cytosol. Postsynaptic labeled regions were often associated with presumptive retinal unlabeled terminals. Microtubules also appeared intensely labeled. These results show that NOS immunoreactive neurons may be innervated by retinal terminals and suggest an association of nNOS with cytoskeletal elements.     

Differential regulation of nitric oxide synthase isoforms in experimental acute chagasic cardiomyopathy

We have previously demonstrated induction and high level expression of IL-1beta, IL-6 and tumour necrosis factor-alpha in the myocardium during the acute stage of experimental Trypanosoma cruzi infection (Chagas' disease). The myocardial depressive effects of these cytokines are mediated in part by the induction of nitric oxide synthase (NOS), production of nitric oxide (NO) and formation of peroxynitrite. In this study we investigated the expression, activity and localization of NOS isoforms, and the levels of NO, malondialdehyde (a measure of oxidative stress), and peroxynitrite in rats at 1.5, 5, 10 and 15 days after infection with T. cruzi trypomastigotes. The myocardial inflammatory infiltrate and number of amastigote nests increased over the course of infection. A significant increase in tissue nitrate + nitrite levels, NOS2 mRNA, and NOS2 enzyme activity was observed at all time points in the infected compared with uninfected animals. The enzyme activity of constitutive NOS, tissue malondialdehyde levels, and NOS3 mRNA levels was only transiently increased after infection. The protein levels of the NOS isoforms paralleled their mRNA expression. While no positive nitrotyrosine immunoreactivity was detected in control myocardium, its levels increased in infected animals over time. Thus, by 1.5 days post-infection, when no parasite or immune cell infiltration could be detected, the myocardium expressed high levels of NOS and NO metabolites. Nevertheless, the early production of NO in the myocardium was not sufficient to clear the parasites.     

Role of spinal nitric oxide synthase-dependent processes in the initiation of the micturition hyperreflexia associated with cyclophosphamide-induced cystitis

The aim of this study was to examine the participation of nitrergic neurotransmission in the initiation of micturition hyperreflexia associated to cyclophosphamide (CP)-induced cystitis in rats. Micturition threshold volume was significantly reduced 4 h after CP administration (100 mg/kg, i.p.); this reduction was attenuated by intra-arterially injected N(G)-nitro-l-arginine-methyl ester (l-NAME), a non selective nitric oxide synthase (NOS) inhibitor, but not by intravesical infusion of S-methyl-l-thiocitrulline (l-SMTC), another structurally different NOS inhibitor. Interestingly, l-NAME failed to affect micturition threshold volume in normal rats. The magnitude of isolated detrusor strips contractions elicited by either carbachol or nerve activation was significantly reduced in CP-treated rats but was unaffected by the addition of N(G)-nitro-l-arginine (l-NOARG), a nonselective NOS inhibitor. In contrast, intrathecal l-NAME and l-SMTC but not N(G)-nitro-d-arginine-methyl ester (d-NAME) administration augmented the micturition threshold volume in CP-treated rats in an l-arginine preventable manner. As with the systemic injection, intrathecal l-NAME also did not affect the micturition threshold volume in normal rats. Four hours after CP injection, the number of neuronal NOS immunoreactive or nicotinamide adenine dinucleotide phosphate diaphorase (NADPH-d) positive neurons in spinal lumbosacral segments (L6-S2) was not altered whereas the number of c-Fos immunoreactive neurons increased significantly in the dorsal gray commissural nucleus (DGC), the parasympathetic sacral nucleus (PSN) and lamina X of these segments. Ca(2+)-dependent, but not Ca(2+)-independent NOS activity increased significantly in spinal L6-S2 segments but not in thoracic segments of CP-treated rats. These data indicate that the micturition hyperreflexia observed in the initial hours of CP-induced cystitis is associated with an increase in Ca(2+)-dependent NOS activity in spinal L6-S2 segments suggesting an increased production of nitric oxide (NO). The increased production of NO in these spinal segments appears to be necessary for the initiation of the micturition hyperreflexia.     

Role of neuronal nitric oxide in the regulation of vasopressin expression and release in response to inhibition of catecholamine synthesis and dehydration

We used neuronal nitric oxide synthase (nNOS) gene knockout mice to study the effects of catecholamines and neuronal nitric oxide on vasopressin expression in the hypothalamic neurosecretory centers. nNOS gene deletion did not change the level of vasopressin mRNA in the supraoptic or paraventricular nuclei. In contrast, vasopressin immunoreactivity was lower in nNOS deficient mice than in wild-type animals. Dehydration increased vasopressin mRNA levels and decreased vasopressin immunoreactivity in both wild-type and nNOS knockout mice, but these responses were more marked in the nNOS knockout mice. Treatment with alpha-mpt, a pharmacologic inhibitor of catecholamine synthesis, resulted in increased vasopressin mRNA levels in wild-type mice and in reduced vasopressin immunoreactivity in both wild-type and nNOS knockout mice. From these results, we conclude: (1) neuronal nitric oxide suppresses vasopressin expression under basal conditions and during activation of the vasopressinergic system by dehydration; (2) catecholamines limit vasopressin expression; (3) nNOS is required for the effects of catecholamines on vasopressin expression.     

Nocturnal motor coordination deficits in neuronal nitric oxide synthase knock-out mice

Nitric oxide is formed in the brain primarily by neurons containing neuronal nitric oxide synthase (nNOS), though some neurons may express endothelial NOS (eNOS), and inducible NOS (iNOS) only occurs in neurons following toxic stimuli. Mice with targeted disruption of nNOS (nNOS-) display distended stomachs with hypertrophied pyloric sphincters reflecting loss of nNOS in myenteric plexus neurons. nNOS- animals resist brain damage following middle cerebral artery occlusions consistent with evidence that excess release of nitric oxide mediates neurotoxicity in ischemic stroke. Neuronal NOS- mice have no grossly evident defects in locomotor activity, breeding long-term depression in the cerebellum, long-term potentiation in the hippocampus, and overall sensorimotor function. However, nNOS- animals display excessive, inappropriate sexual behavior and dramatic increases in aggression. Because the cerebellum possesses the greatest levels of nNOS neurons in the brain, it was surprising that presumed cerebellar functions such as balance and coordination were grossly normal in nNOS- mice. These previous studies were all conducted during the day (between 1400 and 1600, lights on at 0700). We now report striking, discrete abnormalities in balance and motor coordination in nNOS-mice reflected selectively at night.     

Co-induction of growth-associated protein GAP-43 and neuronal nitric oxide synthase in the cochlear nucleus following cochleotomy

Abstract In adult animals, cochlear lesioning leads to a reactive synaptogenesis with a reemergence of growth-associated protein, GAP-43, in the auditory brainstem nuclei. In addition, nitric oxide (NO) is also implicated in synaptogenesis. Three isoforms of nitric oxide synthase (NOS) responsible for generating NO have been identified and, in neurons, the predominant isoform is neuronal NOS (nNOS). Studies in visual or olfactory systems have found that the NOS expression often correlates with periods of axonal outgrowth and synapse formation; whether NO plays a similar role in the auditory brainstem needs to be examined. In the present study, a unilateral cochleotomy was performed in adult mice to examine the relationship between the reemergence of GAP-43 and the expression pattern of nNOS. Following surgery, GAP-43 re-emerged in the ipsilateral anterior ventral cochlear nucleus (AVCN) and the immunoreactivity reached a climax around postoperative day (POD) 8; the same expression pattern as that reported in the previous literature is the indicator of synaptogenesis. As for the nNOS immunoreactivity, a dramatic redistribution from a mostly cytoplasmal to a predominantly membranous localization in the ipsilateral AVCN was found especially at POD 4. A similar redistribution pattern in the ipsilateral AVCN for the N-methyl-D-aspartate (NMDA) receptor was also observed at POD 4, corresponding to the fact that the activation of nNOS is coupled to calcium influx via the NMDA-receptor. Furthermore, the expression of cyclic guanosine monophosphate (cGMP) is an indicator for activity of soluble guanylyl cyclase (sGC), the substrate of NO, which reveals the target area of NO. Therefore, cGMP immunoreactivity was also examined and an obvious increase of cytoplasmal cGMP expression was observed around POD 4. Accordingly, it is suggested that nNOS activity correlates closely with the reactive synaptogenesis following a cochleotomy. Further evidence is shown by the results of fluorescent double staining; nNOS-positive cells were surrounded by GAP-43 labeled regions that appeared to be presynaptic boutons, and the vast majority of nNOS-positive cells also expressed cGMP. The former result indicates that, after surgery, there should be new terminal endings projecting onto the nNOS-positive cells in the AVCN. Furthermore, the latter result suggests a possible role of an autocrine mediator for nNOS in the AVCN.     

NMDA receptor regulation of nNOS phosphorylation and induction of neuron death

Stimulation of NMDA receptors activates neuronal nitric oxide synthase (nNOS) and the production of nitric oxide (NO). Dephosphorylation of nNOS increases nNOS enzymatic activity. We have examined the regulation of nNOS phosphorylation in rat cortical neurons following NMDA receptor activation. We show that nNOS is constitutively phosphorylated and that NMDA receptor activation decreases the level of nNOS phosphorylation by a mechanism that is blocked specifically by NMDA receptor antagonists and inhibitors of the Ca²⁺-regulated phosphatases calcineurin and PP1/PP2A. Using quantitative digital microscopy, we show that NMDA receptor activation induces the accumulation of nitrotyrosine, a measure of nNOS activity, and TdT-mediated fluorescein-dUTP nick end labeling (TUNEL) positivity, a measure of cell death. A calcineurin inhibitor blocked the increase in both TUNEL and nitrotyrosine positivity. Notably, TUNEL was increased in those neurons that were most strongly positive for nitrotyrosine. We conclude that NMDA receptor activation induces death of neurons by a cell autonomous pathway involving nNOS dephosphorylation by a calcineurin-dependent mechanism.     

Neuronal nitric oxide synthase is the dominant nitric oxide supplier for the survival of dorsal root ganglia after peripheral nerve axotomy

This study was designed to determine whether nitric oxide supply may be a major factor in the survival of dorsal root ganglia in a sciatic nerve injury model. Wild-type (WT) mice were compared with knockout (KO) mice lacking neuronal nitric oxide synthase (nNOS) or endothelial (eNOS). The NO-generating capacities were analysed by NOS immunohistochemistry and NADPH-diaphorase staining 1, 2, 6, and 12 weeks after nerve transection. The occurrence and morphological type of neuronal death were determined by TUNEL reaction and ultrastructural examination. Cell loss following nerve section, whist dependent on the availability of NO, as shown by its marked elevation in nNOS KO mice, did not correlate well with nNOS expression in WT animals. Whereas a lack of eNOS was tolerated, deficiency of nNOS led to an enhanced cell loss. The results suggest a crucial role of NO supply after transection of peripheral nerves with a particular significance of the nNOS isoform.     

The immunohistochemical localization of neuronal nitric oxide synthase in the basal forebrain of the dog

The present work describes for the first time the anatomical distribution of neuronal nitric oxide synthase (nNOS) immunoreactivity and NADPH-d activity in the basal forebrain of the dog. As in other species, small, intensely nNOS-immunoreactive cells were seen within the olfactory tubercle, caudate nucleus, putamen, nucleus accumbens and amygdala. In addition, a population of mixed large and small nNOS positive cells was found in the medial septum, diagonal band and nucleus basalis overlapping the distribution of the magnocellular cholinergic system of the basal forebrain. Our results show that the distribution of NOS containing neurons in these nuclei in the dog is more extensive and uniform than that reported in rodents and primates. When double labeling of nNOS and NADPH-d was performed in the same tissue section most neurons were double labeled. However, a considerable number of large perikarya in the diagonal band and nucleus basalis appeared to be single labeled for nNOS. Thought a certain degree of interference between the two procedures could not be completely excluded, these findings suggest that NADPH-d histochemistry, which is frequently used to show the presence of NOS, underestimates the potential of basal forebrains neurons to produce nitric oxide. In addition, a few neurons mainly localized among the fibers of the internal capsule, appeared to be labeled only for NADPH-d. These neurons could be expressing a different isoform of NOS, not recognized by our anti-nNOS antibody, as has been reported in healthy humans and AD patients.     

Nitric oxide synthases modulate progenitor and resident endothelial cell behavior in galactosemia

We used knockout animals of either inducible nitric oxide synthase (iNOS(/)) or endothelial NOS (eNOS(/)) to characterize the role of NOS in galactosemia, a model of diabetic retinopathy. NADH oxidase and nitrotyrosine were used as biomarkers of oxidative stress and vascular dysfunction. These animals were engrafted with hematopoietic stem cells (HSC) expressing green fluorescence protein (gfp(+)) to characterize the contribution of HSC and endothelial progenitor cells to neovascularization. Increased NADH oxidase activity and superoxide generation occurred in all galactose-fed mice. eNOS(/) mice demonstrated increased iNOS immunoreactivity in their retinal vasculature. Nitrotyrosine levels were low at baseline in the wild-type (WT) mice, eNOS(/) and iNOS(/) mice, and the galactose-fed iNOS mice and increased following galactose feeding in eNOS(/) and WT. Galactose-fed WT.gfp and iNOS(/).gfp chimeric animals had areas of perfused new vessels composed of gfp(+) cells. In contrast, galactose-fed eNOS(/).gfp mice produced copious, unbranched, nonperfused tubes. Thus, nitric oxide modulates HSC behavior and vascular phenotype in the retina. Although there is increased NADH oxidase and superoxide in galactosemic mice of all isoforms, iNOS is the source of nitric oxide responsible for peroxynitrite and nitrotyrosine formation that leads to the pathology observed in galactosemic mice.