Faculty Bibliography

Elevated vascular endothelial growth factor (VEGF) levels are required for ocular and tumor angiogenesis in animal models. Ischemic hypoxia is strongly correlated with increased VEGF expression in these systems and is considered a physiologically relevant stimulus. Because ischemic hypoxia is often followed by reperfusion and reactive oxygen intermediate (ROI) generation, we examined the potential role of ROI in the control of VEGF gene expression. Human retinal pigment epithelial cells exposed to superoxide or hydrogen peroxide rapidly increased VEGF mRNA levels. Superoxide-associated mRNA increases were dose dependent, blocked by antioxidants, and associated with elevated VEGF protein levels in conditioned media. Increases in VEGF mRNA levels were also observed in cultured human melanoma and rat glioblastoma cells with superoxide or hydrogen peroxide. Cycloheximide prevented the ROI-associated increases in VEGF mRNA. Transcriptional inhibition with actinomycin D revealed an inducible increase in VEGF mRNA half-life, but nuclear run-on experiments showed no increase in VEGF transcriptional rate. Reoxygenation of human retinal pigment epithelial cells in vitro and ocular reperfusion in vivo increased retinal VEGF mRNA levels. Antioxidants prevented the reperfusion-associated VEGF mRNA increases in retina. We conclude that ROIs increase VEGF gene expression in vitro and during the reperfusion of ischemic retina in vivo. The ROI-associated increases are mediated largely through increases in VEGF mRNA stability.

These studies were carried out to characterize the activation of rat striatal tyroxine hydroxylase produced by depolarization of the medial forebrain bundle and to evaluate the possible role of cyclic AMP as a mediator of this activation. The enzymatic properties of tyrosine hydroxylase following in vivo depolarization were compared to those produced by treatment of striatal synaptosomes with dibutyryl cyclic AMP (dbcAMP). Similar effects were observed with regard to enzyme distribution, altered sensitivity to dopamine-induced inhibition, and activity as a function of tyrosine concentration. However, differences between the two treatments were also apparent. First, treatment with dbcAMP shifted the pH optimum from 6.2 to 7.0. In contrast, electrical stimulation decreased the rate of decline in activity as the pH was increased above the optimum, but did not shift the pH optimum. Second, plots of tyrosine hydroxylase activity versus cofactor concentration revealed two enzyme forms for both control and electrically stimulated preparations. However, dbcAMP treatment converted the enzyme to a single high affinity form. These results can be explained by one of the following: (1) cyclic AMP is the sole mediator of enzyme activation, but does not produce a maximally activated enzyme following in vivo depolarization, (2) cyclic AMP is only one of several mediators involved or (3) cyclic AMP is not involved in depolarization-induced activation, with activation occurring via the mediation of other intracellular messengers, such as calcium.

The present studies were carried out to determine if tyrosine hydroxylase phosphorylation in rat brain striatal synaptosomes is activated by dibutyryl cyclic AMP treatment. Incubation of synaptosomes with [32P]orthophosphate, followed by immunoprecipitation and sodium dodecyl sulfate polyacrylamide gel electrophoresis, produced a band of radioactivity associated with a 62 kDa polypeptide. Treatment with the catecholamine neurotoxin, 6-hydroxydopamine, produced parallel losses of: (1) tyrosine hydroxylase enzyme activity, (2) dopamine content, and (3) the 62 kDa band of radioactivity. These data support the identification of this band as a tyrosine hydroxylase-derived polypeptide. Incubation with dibutyryl cyclic AMP produced an increase in soluble tyrosine hydroxylase activity and phosphorylation. These results suggest that the increase in synaptosomal catecholamine synthesis produced by dibutyryl cyclic AMP is mediated by an increase in tyrosine hydroxylase phosphorylation.

Tracer efflux studies were used to determine the effect of activation of protein kinase C on K channel function in rat brain synaptosomes. Hippocampal synaptosomes were treated with sn-1,2-dioctanoylglycerol (diC8), a synthetic diacylglycerol (DG) analog that activates protein kinase C. DiC8 inhibited depolarization-induced 86Rb efflux through voltage-gated K channels but did not affect the component of efflux corresponding to Ca-activated K channels. In time-course experiments, diC8 inhibited two components of 86Rb efflux: efflux through a rapidly inactivating, voltage-gated K channel (responsible for the "A" current) and that through a slowly inactivating, voltage-gated K channel (believed to be the "delayed rectifier"). Experiments with specific blockers of these voltage-gated K channels supported this observation. Inhibition of K-stimulated 86Rb efflux by diC8 was time dependent: at least 15 sec of preincubation was required before the effect could be observed. The effect of diC8 was concentration dependent: 50 microM diC8 produced a half-maximal inhibition of K-stimulated 86Rb efflux. The idea that the inhibition of synaptosome K channels by diC8 resulted from activation of C kinase was supported by pharmacological evidence. The action of diC8 was mimicked by 1-oleoyl-2-acetylglycerol, another DG analog that activates protein kinase C, but not by deoxy-diC8, a DG analog that does not activate C kinase. Inhibition of C kinase by sphingosine or H-7 prevented the diC8 effect. These studies demonstrate that synaptosomes are a good model in which to study modulation of mammalian CNS K channels.(ABSTRACT TRUNCATED AT 250 WORDS)