Search

Depending on its concentration, nitric oxide (NO) has beneficial or toxic effects. In pathological conditions, NO reacts with superoxide to form peroxynitrite, which nitrates proteins forming nitrotyrosine residues (3NY), leading to loss of protein function, perturbation of signal transduction, and cell death. 3NY immunoreactivity is present in many CNS diseases, particularly multiple sclerosis. Here, using the high flux NO donor, spermine-NONOate, we report that oligodendrocytes are resistant to NO, while motor neurons are NO sensitive. Motor neuron sensitivity correlates with the NO-dependent formation of 3NY, which is significantly more pronounced in motor neurons when compared with oligodendrocytes, suggesting peroxynitrite as the toxic molecule. The heme-metabolizing enzyme, heme-oxygenase-1 (HO1), is necessary for oligodendrocyte NO resistance, as demonstrated by loss of resistance after HO1 inhibition. Resistance is reinstated by peroxynitrite scavenging with uric acid further implicating peroxynitrite as responsible for NO sensitivity. Most importantly, differential sensitivity to NO is also present in cultures of primary oligodendrocytes and motor neurons. Finally, motor neurons cocultured with oligodendrocytes, or oligodendrocyte-conditioned media, become resistant to NO toxicity. Preliminary studies suggest oligodendrocytes release a soluble factor that protects motor neurons. Our findings challenge the current paradigm that oligodendrocytes are the exclusive target of multiple sclerosis pathology.

Neuregulins play a major role in the formation and stabilization of neuromuscular junctions, and are produced by both motor neurons and muscle. Although the effects and mechanism of neuregulins on skeletal muscle (e.g. regulation of acetylcholine receptor expression) have been studied extensively, the effects of neuregulins on motor neurons remain unknown. We report that neuregulin-1beta (NRGbeta1) inhibited apoptosis of rat motor neurons for up to 7 days in culture by a phosphatidylinositol 3 kinase-dependent pathway and synergistically enhanced motor neuron survival promoted by glial-derived neurotrophic factor (GDNF). However, binding of neurotrophins, including brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF), to the p75 neurotrophin receptor (p75NTR) abolished the neuregulin anti-apoptotic effect on motor neurons. Inhibitors of the c-Jun N-terminal kinase (JNK) mitogen-activated protein kinase prevented motor neuron death caused by co-incubation of NRGbeta1 and BDNF or NGF, as well as by trophic factor deprivation. Motor neuron apoptosis resulting from both trophic factor deprivation and exposure to NRGbeta1 plus neurotrophins required the induction of neuronal nitric oxide synthase and peroxynitrite formation. Because motor neurons express both p75NTR and neuregulin erbB receptors during the period of embryonic programmed cell death, motor neuron survival may be the result of complex interactions between trophic and death factors, which may be the same molecules acting in different combinations.

Fibroblast growth factor-1 (FGF1 or acidic FGF) is highly expressed in motor neurons. FGF-1 is released from cells by oxidative stress, which might occur from SOD-1 aberrant function in amyotrophic lateral sclerosis (ALS). Although FGF-1 is known to be neuroprotective after spinal cord injury or axotomy, we found that FGF-1 could activate spinal cord astrocytes in a manner that decreased motor neuron survival in co-cultures. FGF-1 induced accumulation of the FGF receptor 1 (FGFR1) in astrocyte nuclei and potently stimulated nerve growth factor (NGF) expression and secretion. The FGFR1 tyrosine kinase inhibitor PD166866 prevented these effects. Previously, we have shown that NGF secretion by reactive astrocytes induces motor neuron apoptosis through a p75(NTR)-dependent mechanism. Embryonic motor neurons co-cultured on the top of astrocytes exhibiting activated FGFR1 underwent apoptosis, which was prevented by PD166866 or by adding either anti-NGF or anti-p75(NTR) neutralizing antibodies. In the degenerating spinal cord of mice carrying the ALS mutation G93A of Cu, Zn superoxide dismutase, FGF-1 was no longer localized only in the cytosol of motor neurons, while FGFR1 accumulated in the nuclei of reactive astrocytes. These results suggest that FGF-1 released by oxidative stress from motor neurons might have a role in activating astrocytes, which could in turn initiate motor neuron apoptosis in ALS through a p75(NTR)-dependent mechanism.

Reactive astrocytes frequently surround degenerating motor neurons in patients and transgenic animal models of amyotrophic lateral sclerosis (ALS). We report here that reactive astrocytes in the ventral spinal cord of transgenic ALS-mutant G93A superoxide dismutase (SOD) mice expressed nerve growth factor (NGF) in regions where degenerating motor neurons expressed p75 neurotrophin receptor (p75(NTR)) and were immunoreactive for nitrotyrosine. Cultured spinal cord astrocytes incubated with lipopolysaccharide (LPS) or peroxynitrite became reactive and accumulated NGF in the culture medium. Reactive astrocytes caused apoptosis of embryonic rat motor neurons plated on the top of the monolayer. Such motor neuron apoptosis could be prevented when either NGF or p75(NTR) was inhibited with blocking antibodies. In addition, nitric oxide synthase inhibitors were also protective. Exogenous NGF stimulated motor neuron apoptosis only in the presence of a low steady state concentration of nitric oxide. NGF induced apoptosis in motor neurons from p75(NTR +/+) mouse embryos but had no effect in p75(NTR -/-) knockout embryos. Culture media from reactive astrocytes as well as spinal cord lysates from symptomatic G93A SOD mice-stimulated motor neuron apoptosis, but only when incubated with exogenous nitric oxide. This effect was prevented by either NGF or p75(NTR) blocking-antibodies suggesting that it might be mediated by NGF and/or its precursor forms. Our findings show that NGF secreted by reactive astrocytes induce the death of p75-expressing motor neurons by a mechanism involving nitric oxide and peroxynitrite formation. Thus, reactive astrocytes might contribute to the progressive motor neuron degeneration characterizing ALS.

Nerve growth factor (NGF) prevents apoptosis induced by the oxidant peroxynitrite in undifferentiated PC12 rat pheochromocytoma cells. Previous studies have shown that activation of phosphatidylinositol 3-kinase (PI 3-kinase) by NGF via the TrkA receptor tyrosine kinase protects PC12 cells from serum deprivation-induced apoptosis. We found that two PI 3-kinase inhibitors, wortmannin and LY294002, eliminated the protection NGF provided against peroxynitrite-induced apoptosis at concentrations consistent with their effectiveness as PI 3-kinase inhibitors. When the activity of PI 3-kinase was assayed in phosphotyrosine immunoprecipitates after treatment of PC12 cells with peroxynitrite, PI 3-kinase activity was reduced by 50% of that detected in control cells, whereas PI 3-kinase activity in NGF-treated cells was unaffected by peroxynitrite. If an antibody against PI 3-kinase was used to immunoprecipitate the enzyme, treatment with peroxynitrite had no effect on activity. Therefore, peroxynitrite appeared to disrupt interactions between PI 3-kinase and phosphotyrosine proteins, rather than directly inhibiting the enzyme. NGF also activates p21Ras-dependent pathways, but this did not appear to be required for NGF to exert its protective effect against peroxynitrite. PC12 cells expressing a dominant inhibitory mutant of p21Ras were equally susceptible to peroxynitrite-induced apoptosis, which was prevented by NGF. Wortmannin was also able to block the protective effect of NGF in the p21Ras mutant cell line. Although many signaling pathways are activated by NGF, these results suggest that a PI 3-kinase-dependent pathway is important for inhibiting peroxynitrite-induced apoptosis.

Peroxynitrite is a powerful oxidant formed by the near-diffusion-limited reaction of nitric oxide with superoxide. Large doses of peroxynitrite (> 2 mM) resulted in rapid cell swelling and necrosis of undifferentiated PC12 cells. However, brief exposure to lower concentrations of peroxynitrite (EC50 = 850 microM) intially (3-4 h) caused minimal damage to low-density cultures. By 8 h, cytoplasmic shrinkage with nuclear condensation and fragmentation became increasingly evident. After 24 h, 36% of peroxynitrite-treated cells demonstrated these features associated with apoptosis. In addition, 46% of peroxynitrite-treated cells demonstrated DNA fragmentation (by terminal-deoxynucleotide transferase-mediated dUTP-digoxigenin nick end-labeling) after 7 h, which was inhibited by posttreatment with the endonuclease inhibitor aurintricarboxylic acid. Serum starvation also resulted in apoptosis in control cells (23%), the percentage of which was not altered significantly by peroxynitrite treatment. Although peroxynitrite is known to be toxic to cells, the present study provides a first indication that peroxynitrite induces apoptosis. Furthermore, pretreatment of cells with nerve growth factor or insulin, but not epidermal growth factor, was protective against peroxynitrite-induced apoptosis. However, both acidic and basic fibroblast growth factors greatly increased peroxynitrite-initiated apoptosis, to 63 and 70%, respectively. Thus, specific trophic factors demonstrate differential regulation of peroxynitrite-induced apoptosis in vitro.