Protein phosphorylation is crucial for regulating synaptic transmission. We describe a novel mechanism for the phosphorylation of the GABA A receptor, which mediates fast inhibition in the brain. A protein copurified and coimmunoprecipitated with the phosphorylated receptor ␣1 subunit; this receptor-associated protein was identified by purification and microsequencing as the key glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Molecular constructs demonstrated that GAPDH directly phosphorylates the long intracellular loop of GABA A receptor ␣1 subunit at identified serine and threonine residues. GAPDH and the ␣1 subunit were found to be colocalized at the neuronal plasma membrane. In keeping with the GAPDH/GABA A receptor molecular association, glycolytic ATP produced locally at plasma membranes was consumed for this ␣1 subunit phosphorylation, possibly within a single macrocomplex. The membrane-attached GAPDH is thus a dual-purpose enzyme, a glycolytic dehydrogenase, and a receptor-associated kinase. In acutely dissociated cortical neurons, the rundown of the GABA A responses was essentially attributable to a Mg 2ϩ -dependent phosphatase activity, which was sensitive to vanadate but insensitive to okadaic acid or fluoride. Rundown was significantly reduced by the addition of GAPDH or its reduced cofactor NADH and nearly abolished by the addition of its substrate glyceraldehyde-3-phosphate (G3P). The prevention of rundown by G3P was abolished by iodoacetamide, an inhibitor of the dehydrogenase activity of GAPDH, indicating that the GABA A responses are maintained by a glycolysis-dependent phosphorylation. Our results provide a molecular mechanism for the direct involvement of glycolysis in neurotransmission.

When neuroblastoma cells were transferred to a medium of low (6 nM) thiamine concentration, a 16‐fold decrease in total intracellular thiamine content occurred within 8 days. Respiration and ATP levels were only slightly affected, but addition of a thiamine transport inhibitor (amprolium) decreased ATP content and increased lactate production. Oxygen consumption became low and insensitive to oligomycin and uncouplers. At least 25% of mitochondria were swollen and electron translucent. Cell mortality increased to 75% within 5 days. [3H]PK 11195, a specific ligand of peripheral benzodiazepine receptors (located in the outer mitochondrial membrane) binds to the cells with high affinity (KD = 1.4 ± 0.2 nM). Thiamine deficiency leads to an increase in both Bmax and KD. Changes in binding parameters for peripheral benzodiazepine receptors may be related to structural or permeability changes in mitochondrial outer membranes. In addition to the high‐affinity (nanomolar range) binding site for peripheral benzodiazepine ligands, there is a low‐affinity (micromolar range) saturable binding for PK 11195. At micromolar concentrations, peripheral benzodiazepines inhibit thiamine uptake by the cells. Altogether, our results suggest that impairment of oxidative metabolism, followed by mitochondrial swelling and disorganization of cristae, is the main cause of cell mortality in severely thiamine‐deficient neuroblastoma cells.

A number of recent studies have suggested that phosphorylation of the y-aminobutyric acid A (GABAA) receptor exam-ined the ability of specific kinases to catalyze significant phosphorylation of the GABAA receptor that has been purified to near homogeneity. The GABAA receptor was purified as previously described using benzodiazepine affinity chromatography. The purified receptor possessed no detectable kinase activity. Protein kinase C and cAMP-dependent protein kinase catalyzed the phosphorylation of the (i and a subunits of the receptor. However, most of the phosphate incorporation was associated with the .8 subunit. Two muscimol binding polypeptides designated s58 (Mr 58,000) and 365 (Mr 56,000) were present in the preparation. The higher molecular weight polypeptide, 858, was phosphorylated specifically by cAMPdependent protein kinase. / 56was phosphorylated specifically by protein kinase C. 1s58 and .856 gave distinct patterns in a one-dimensional phosphopeptide analysis. The stoichiometry of phosphorylation (mol of phosphate/mol of muscimol binding) catalyzed by cAMP-dependent protein kinase was 0.52 and that catalyzed by protein kinase C was 0.38. Taken together these data confirm that there are two forms of the (8 subunit of the GABAA receptor and suggest that these two forms of the /3 subunit are phosphorylated by distinct kinases.Various ligand-gated ion channels [nicotinic acetylcholine receptor, glycine receptor, and y-aminobutyric acid A (GABAA) receptor] have been cloned and their primary sequences have been deduced. Significant homologies exist among all of these proteins (1, 2). A number of recent reports have shown that the nicotinic acetylcholine receptor (AcChoR) is phosphorylated and that this phosphorylation regulates the rate of desensitization of the receptor (3-5). The AcChoR has been shown to be phosphorylated by the cAMPdependent protein kinase (PKA) and by the Ca2+/phosphatidylserine-dependent protein kinase (PKC) (6-8).The GABAA receptor mediates the majority of inhibitory synaptic transmission in the central nervous system. Given the significant import of this receptor to central nervous system function, we have been particularly interested in the possibility that the GABAA receptor might exhibit posttranslational modification that resembles that seen in the AcChoR.A number of recent reports have suggested that phosphorylation of the GABAA receptor could modulate receptor function. Sigel and Baur (9) reported that activators of PKC led to decreased activity of the GABAA receptor that had been expressed in oocytes. In addition, activators of PKA decreased GABAA receptor-mediated Cl-flux in rat brain synaptoneurosomes (10) and induced a decline in GABAgated chloride currents in cultured chick cortical neurons (11). Conversely, phosphorylation has also been reported to maintain GABA receptor function in dissociated hippocampal neurons and in cultured chick spinal neurons (12,13). A consensus sequence preferred by PKA has also been identified in the sequence of the 8 subunit o...

Heterogeneity of binding affinities for a variety of ligands was observed for gamma-aminobutyric acid type A (GABAA) receptors in the rat CNS, at both GABA and benzodiazepine recognition sites. Photoaffinity labeling by [3H]flunitrazepam and [3H]muscimol to affinity column-purified receptor proteins was examined by gel electrophoresis in sodium dodecyl sulfate. Anesthetic barbiturates (pentobarbital) and steroids (alphaxalone) both differentially stimulated the incorporation of [3H]flunitrazepam more so into the 51-kDa alpha 1 subunit than into the 53-kDa alpha 2 polypeptide, and incorporation of [3H]muscimol into the 55-kDa beta 2 subunit more so than the 58-kDa beta 3 polypeptide. Binding to these polypeptides was also affected differentially by other allosteric modulators and competitive inhibitors, including the benzodiazepine "type 1" selective ligand CL218,872. Heterogeneity in affinity of this drug for the single 51-kDa alpha 1 polypeptide strongly suggests that type I receptors, like type II, are heterogeneous. In brain sections, the extent of enhancement of [3H]muscimol binding showed significant regional variation, similar for both steroids and barbiturates, and the GABA analogues THIP and taurine inhibited muscimol binding with regional variations in affinity that were almost opposites of each other. Modulation of [3H]flunitrazepam binding by steroids, barbiturates, and THIP significantly varied with regions. Taken together, ligand binding heterogeneity exhibited by photoaffinity labeling and autoradiography demonstrate the existence of multiple pharmacological-binding subtypes resulting from the combination of multiple polypeptide gene products into several oligomeric isoreceptors. Comparison of the regional distribution of binding subtypes with that of different subunit gene products allows the following conclusions about possible subunit compositions of native pharmacological receptor subtypes present in the brain: Benzodiazepine pharmacology of the oligomeric receptor isoforms is dependent on the nature of alpha and subunits other than alpha, GABA-benzodiazepine coupling is dependent on the nature of the alpha subunits, GABA site pharmacology is dependent on the nature of the beta subunits, and several subunits including alpha and beta contribute to the degree of sensitivity to steroids and barbiturates. Finally, the presence of discrete subunits may be necessary but is not sufficient to postulate a defined pharmacological property.

No abstract