The synthesis of ATP by glycolytic enzymes in the postsynaptic density and the effect of endogenously generated nitric oxide.

The major contribution of this paper is the finding of a glycolytic source of ATP in the isolated postsynaptic density (PSD). The enzymes involved in the generation of ATP are glyceraldehyde-3-phosphate dehydrogenase (G3PD) and phosphoglycerate kinase (PGK). Lactate dehydrogenase (LDH) is available for the regeneration of NAD+, as well as aldolase for the regeneration of glyceraldehyde-3-phosphate (G3P). The ATP was shown to be used by the PSD Ca2+/calmodulin-dependent protein kinase and can probably be used by two other PSD kinases, protein kinase A and protein kinase C. We confirmed by immunocytochemistry the presence of G3PD in the PSD and its binding to actin. Also present in the PSD is NO synthase, the source of NO. NO increases the binding of NAD, a G3PD cofactor, to G3PD and inhibits its activity as also found by others. The increased NAD binding resulted in an increase in G3PD binding to actin. We confirmed the autophosphorylation of G3PD by ATP, and further found that this procedure also increased the binding of G3PD to actin. ATP and NO are connected in that the formation of NO from NOS at the PSD resulted, in the presence of NAD, in a decrease of ATP formation in the PSD. In the discussion, we raise the possible roles of G3PD and of ATP in protein synthesis at the PSD, the regulation by NO, as well as the overall regulatory role of the PSD complex in synaptic transmission.

Light and electron microscopic localization of alpha subunits of GTP-binding proteins, G(o) and Gi, in the cerebral cortex and hippocampus of rat brain.

Antibodies that recognize alpha subunits of G(o), Gi2 and Gi3 were used to evaluate their association with synaptic junctions. G(o), but not Gi, was concentrated within perikaryal and dendritic cytoplasm of a small population of bipolar neurons. All three G-proteins were associated with the intracellular surface of dendritic, axonal and astrocytic plasma membranes and postsynaptic densities (PSDs). However, association with PSDs was more prevalent for the two Gi's than for G(o) while the association with terminals forming putatively excitatory synapses was more prevalent for G(o) and Gi3 than for Gi2. Thus, neuromodulators may modulate the release of excitatory transmitters via activation of presynaptic Gi3 and G(o) and also regulate the opening of Ca2+ and/or K+ channels via activation of Gi's and G(o) at PSDs.

Occurrence of the alpha subunits of G proteins in cerebral cortex synaptic membrane and postsynaptic density fractions: modulation of ADP-ribosylation by Ca2+/calmodulin.

We have examined the isolated postsynaptic density (PSD) fraction for the presence of a G protein. First, we found specific binding of guanosine 5'-[gamma-[35S]thio]triphosphate to the PSD. Second, pertussis toxin-activated ADP-ribosylation of the isolated PSD fraction resulted in the appearance of a G protein with an apparent molecular mass of 41 kDa, and two G proteins with apparent molecular masses of 41 kDa and 39 kDa in synaptic membrane (SM) fraction and total homogenate (H). The amount of the 41-kDa G protein per unit protein was in the order of SM greater than H greater than PSD. Anti-G(i0 antibodies recognized the 41-kDa G protein in both PSD and SM, whereas anti-G(o) antibodies reacted with the 39-kDa G protein in the SM. The absence of G(o) protein in the PSD suggested that there was no contamination with SM. Moreover, unlabeled PSD incubated with an extract of SM that contained the labeled G proteins resulted in no label in the subsequently reisolated PSD, suggesting that the G protein found in the PSD was not due to adsorption of the G protein onto the PSD during its isolation from the SM. PSD pretreated with EGTA gave an 11-fold increase in the ADP-ribosylation reaction of the G(i) protein; similar effects on the G(i) and G(o) proteins of SM were obtained. Restoration of Ca2+/calmodulin to the PSD, but not of either Ca2+ or calmodulin alone, removed the effect of EGTA, indicating a strong complex formation between G(i) and Ca2+/calmodulin that decreased the ADP-ribosylation reaction. Preincubation with the Ca(2+)-channel blocker nifedipine decreased the ADP-ribosylation reaction in the PSD. We conclude that G(i) is present in the PSD, that it may interact with calmodulin and that it is involved in the regulation of voltage-dependent Ca2+ channel. We present a theory of the involvement of the G protein and calmodulin in postsynaptic neurophysiological events.

Possible role for calmodulin and the Ca2+/calmodulin-dependent protein kinase II in postsynaptic neurotransmission.

The theory presented here is based on results from in vitro experiments and deals with three proteins in the postsynaptic density/membrane-namely, calmodulin, the Ca2+/calmodulin-dependent protein kinase, and the voltage-dependent Ca2+ channel. It is visualized that, in vivo in the polarized state of the membrane, calmodulin is bound to the kinase; upon depolarization of the membrane and the intrusion of Ca2+, Ca2(+)-bound calmodulin activates the autophosphorylation of the kinase. Calmodulin is visualized as having less affinity for the phosphorylated form of the kinase and is translocated to the voltage-dependent Ca2+ channel. There, with its bound Ca2+, it acts as a Ca2+ sensor, to close off the Ca2+ channel of the depolarized membrane. At the same time, it is thought that the configuration of the kinase is altered by its phosphorylated states; by interacting with Na+ and K+ channels, it alters the electrical properties of the membrane to regain the polarized state. Calmodulin is moved to the unphosphorylated kinase to complete the cycle, allowing the voltage-dependent Ca2+ channel to be receptive to Ca2+ flux upon the next cycle of depolarization. Thus, the theory tries to explain (i) why calmodulin and the kinase reside at the postsynaptic density/membrane site, and (ii) what function autophosphorylation of the kinase may play.

Effect of septal kindling on glutamate binding and calcium/calmodulin-dependent phosphorylation in a postsynaptic density fraction isolated from rat cerebral cortex.

Postsynaptic density (PSD) fractions were isolated from the cerebral cortices of control and kindled rats and assayed for glutamate and gamma-aminobutyric acid-binding capacities and for the Ca2+/calmodulin-dependent protein kinase. Glutamate binding was found to be increased by approximately 50% in the PSDs isolated from kindled rats as compared to controls; this increase was almost completely from an increase in Bmax; Kd decreased only slightly. Studies with inhibitors indicate that the receptors involved were of the N-methyl-D-aspartate and quisqualate types. PSDs isolated from control and kindled rats did not differ in gamma-aminobutyric acid or flunitrazepam binding. The in vitro autophosphorylation of the Ca2+/calmodulin-dependent protein kinase was depressed by 45-76% in PSDs isolated from kindled rats as compared to controls, with little change in amount of the kinase. Therefore, we infer that (i) the kindled state is associated with an increase in glutamate activation of postsynaptic sites, allowing Ca2+ to enter dendritic spines, (ii) a change has occurred in activity of the protein kinase, which is the major cerebral cortex PSD protein, and (iii) perhaps major alterations in the PSD are a concomitant to the long-lasting nature of the kindled state.

NIH conflict-of-interest guidelines.

In the News & Comment article "Super Collider advocates tangle with cost cutters" by Mark Crawford (12 Jan., p. 152), the collision energy of the Superconducting Super Collider (SSC) was incorrectly described as being two orders of magnitude greater than that of the Tevatron. The SSC's collision energy is supposed to be 40 trillion election volts-20 times that of the Tevatron. In addition, the $7.2-billion cost estimate for the project is based on inflated dollars, not constant dollars, as was stated.

Properties of a protein kinase C activity in synaptic plasma membrane and postsynaptic density fractions isolated from canine cerebral cortex.

Protein kinase C (PKC) activity (phosphorylation increased by addition of Ca2+/phosphatidylserine or Ca2+/phosphatidylserine/phorbol ester) was found in both a synaptic plasma membrane (SPM) and a postsynaptic density (PSD) fraction. The SPM fraction had as endogenous substrates 87K-, 60K-, 50K-, and 20K-Mr proteins, whereas the PSD fraction had only the 20K-Mr protein. The PKC activity was also detected using histone III-S as a substrate, in SPM but much less in PSD. Phosphorylations of histone and the endogenous substrates of PKC, assayed in the absence of Ca2+, were enhanced in the SPM prepared after treatment of brain homogenate with phorbol 12-myristate 13-acetate (TPA), but very little enhancement was found in PSD after such treatment. The SPM PKC activity (both for endogenous substrate proteins and for histone), which was enhanced by TPA treatment of brain homogenate, was inhibited by calcium (IC50, 3 x 10(-7) M). The phosphorylations of the 20K-Mr protein in PSD, and in SPM prepared with and without TPA treatment, were all inhibited by H-7. The 20K-Mr protein in the PSD fraction is also phosphorylated by a PSD Ca2+/calmodulin-dependent protein kinase II. The evidence indicates that both SPM and PSD fractions contain a PKC activity. Detergent treatment of SPM, to produce a purified PSD fraction, results in a PSD fraction that has lost most of the endogenous substrates, lost the TPA-induced enhanced activity assayed in the absence of Ca2+, and lost the inhibitory effect of low Ca2+ concentration.

Mammalian cerebral cortical tissue responds to low-intensity visible light.

Low levels of visible light directed onto slices of rat cerebral cortical tissue enhanced net potassium-induced release of the neurotransmitter gamma-aminobutyric acid (GABA) from these brain slices. At higher light intensity, net potassium-induced release was suppressed. These effects were apparently not from increased temperature. The amount of light enhancing this neurotransmitter release is approximately equal to the amount of light that can penetrate the head and reach the brain at the intensities of sunlight; this was determined by measuring the light entering the rat head through fur, scalp, skull, and dura mater and considering several natural lighting conditions. These results suggest that ambient light may be sufficient to alter the release of transmitters from mammalian cerebral cortex in vivo.

Neurochemical characteristics of a postsynaptic density fraction isolated from adult canine hippocampus.

Postsynaptic density and synaptic membrane fractions isolated from hippocampal tissue have been compared to those previously isolated from cerebellum and cerebral cortex. In all respects examined, the isolated hippocampal preparations are similar to the cerebral cortex fractions. The morphology of the postsynaptic density (PSD) preparation is the same and the protein composition is similar, but with higher concentrations of the 51-kDa major protein and of calmodulin, and lower concentrations of actin, in the hippocampal PSD fraction. The binding characteristics for glutamate and GABA are also similar between the two fractions, but with higher Bmax and KD glutamate values and lower Bmax and higher KD GABA values for the hippocampal PSD preparation. Both preparations contain GABAA and GABAB receptors. The PSD fraction contains, as does the cerebral cortex fraction, a calmodulin-dependent binding of the Ca2+ channel antagonist, nitrendipine, as well as a cAMP-dependent and a Ca2+/calmodulin-dependent protein kinase, with the same respective substrates. The value of the hippocampal fractions for studies on long-term potentiation and on kindling in the hippocampus is discussed.

Characteristics of a Ca2+/calmodulin-dependent binding of the Ca2+ channel antagonist, nitrendipine, to a postsynaptic density fraction isolated from canine cerebral cortex.

Synaptic membrane (SM) and postsynaptic density (PSD) fractions isolated from the cerebral cortex (CTX) and cerebellum (CL) of the canine brain were found to contain one class of specific nitrendipine binding sites. The specific binding constants were: CTX-SM, Kd = 110 pM (Bmax = 126 fmol/mg protein); CTX-PSD, Kd = 207 pM (Bmax = 196 fmol/mg); CL-SM, Kd = 100 pM (Bmax = 65 fmol/mg); CL-PSD, Kd = 189 pM (Bmax = 80 fmol/mg). Treatment of the CTX-SM and CTX-PSD fractions with 0.5% deoxycholate and 1.0% N-lauroyl sarcosinate removed 88-91% and 42-51% of the nitrendipine binding, respectively, indicating that the major nitrendipine binding present in the SM fractions are of non-synaptic origin. Moreover, the percentages of total protein and specific nitrendipine binding removed from PSDs by these detergents were similar, indicating no preferential dissociation of the latter, and suggesting that the receptor protein is firmly bound and is probably an intrinsic component of the PSD fraction. Both Ca2+ and calmodulin were found to be important for the binding of nitrendipine to the CTX-SM and CTX-PSD fractions since: R24571, a calmodulin antagonist, was found to inhibit nitrendipine binding to the CTX-SM and CTX-PSD fractions with IC50 values of 1.1 microM and 0.9 microM, respectively; removal of Ca2+ from the CTX-SM and CTX-PSD fractions with 0.2 mM EGTA resulted in losses of specific nitrendipine binding of 80 and 90%, respectively; Ca2+ alone restored nitrendipine binding to EGTA-pretreated CTX-SM fractions and not to CTX-PSD fractions, with the latter needing both Ca2+ and calmodulin to restore nitrendipine binding; EGTA treatment removed 14-16% and 89-91% of nitrendipine bound to the CTX-SM and CTX-PSD fractions, respectively, suggesting that calmodulin (but not Ca2+) is needed to maintain the nitrendipine-nitrendipine receptor-calmodulin complex; Ca2+-reconstituted EGTA-pretreated CTX-SM fractions and the Ca2+ plus calmodulin-reconstituted EGTA-pretreated CTX-SM and CTX-PSD fractions were found to have similar binding constants to those for the corresponding native, untreated fractions; and the Ca2+/calmodulin dependency on nitrendipine binding was similar to the well-known Ca2+/calmodulin dependency on phosphorylation in EGTA-pretreated PSD fractions. It needed much less Ca2+ to saturate Ca2+/calmodulin-dependent phosphorylation of the pretreated CTX-PSD fractions than the nitrendipine binding. Yet, less calmodulin was needed to saturate nitrendipine binding than the phosphorylation.(ABSTRACT TRUNCATED AT 400 WORDS)

Binding of L-[3H]glutamate to fresh or frozen synaptic membrane and postsynaptic density fractions isolated from cerebral cortex and cerebellum of fresh or frozen canine brain.

Synaptic membrane (SPM) and postsynaptic density (PSD) fractions isolated from cerebral cortex (CTX) and cerebellum (CL) of canine brain, either fresh or frozen and isolated from either fresh or frozen tissue, were found to contain L-[3H]glutamate binding sites. It was found that there was a concentration of L-glutamate binding sites in CTX-PSD and CL-PSD over the respective membrane fractions, and the Bmax value of CL-PSD (92.0 pmol/mg protein) was about three times that of CTX-PSD (28.9 pmol/mg). The results, together with those of others, suggest that the thin CL-PSD are probably derived from the excitatory synapses in the molecular layer. The ion dependency of L-glutamate binding to canine CTX-SPM fraction was found to be similar to that reported for a rat brain SPM fraction: (a) Cl- increased the number of L-glutamate binding sites and the effect was enhanced by Ca2+; Ca2+ alone had no significant effect; (b) the Cl-/Ca2+-sensitive binding sites were abolished by 2-amino-4-phosphonobutyrate (APB) or freezing and thawing; (c) the effect of Na+ ion was biphasic; low concentration of Na+ (less than 5 mM) decreased Cl-/Ca2+-dependent L-glutamate binding sites, whereas at higher concentrations of Na+ the binding of glutamate was found to increase either in the presence or absence of Ca2+ and Cl-. In addition, the K+ ion (50 mM) was found to decrease the Na+-independent and Cl-/Ca2+-independent binding of L-glutamate to fresh CTX-SPM by 18%, but it decreased the Na+-dependent and Cl-/Ca2+-independent L-glutamate binding by 93%; in the presence of Cl-/Ca2+, the K+ ion decreased the Na+-dependent binding by 78%. Freezing and thawing of CTX-SPM resulted in a 50% loss of the Na+-dependent L-glutamate binding sites assayed in the absence of Ca2+ and Cl-. The CL-SPM fraction showed similar ion dependency of L-glutamate binding except for the absence of Na+-dependent glutamate binding sites. The CTX-PSD fraction contained neither Na+-dependent nor APB (or Cl-/Ca2+)-sensitive L-glutamate binding sites and its L-glutamate binding was unaffected by freezing and thawing, in agreement with the reported findings using rat brain PSD preparation. L-Glutamate binding to CTX-SPM or CTX-PSD fraction was not affected by pretreatment with 10 mM L-glutamate, nor by simultaneous incubations with calmodulin.(ABSTRACT TRUNCATED AT 400 WORDS)

Existence of a Ca2+-dependent K+ channel in synaptic membrane and postsynaptic density fractions isolated from canine cerebral cortex and cerebellum, as determined by apamin binding.

Apamin, a 18-amino acid neurotoxin isolated from bee venom, is a specific blocker of one class of the Ca2+-dependent K+ channels. The monoiodo derivative of the toxin with high specific radioactivity (1600 Ci/mmol) has been used to study its binding to synaptic membrane (SM) and postsynaptic density (PSD) fractions isolated from cerebral cortex (CTX) and cerebellum (CL) of canine brains. The Bmax (30.2 fmol/mg protein) for CTX-PSD is about twice that for CTX-SM (17.3 fmol/mg protein), suggesting a concentration of the apamin receptor protein in CTX-PSD over CTX-SM fractions. The lower value of Bmax for CL-PSD (12.3 fmol/mg protein), and the higher Kd value (51 pM) than for CTX-SM (33 pM), CTX-PSD (24 pM), and CL-SM (39 pM), may reflect the disruptive effect of Triton X-100 on these thin structures. The values of Bmax and Kd for CTX-SM are similar to those (22.0 fmol/mg protein and 33 pM) for rat CTX-SM. Both Ca2+ and Na+ inhibit apamin binding to CTX-PSD with K0.5 values of 14 and 31 mM, respectively, while the optimum concentration of KCl for activation is 5 mM. All these values are similar to those found for rat synaptosomes. Covalent labeling of the apamin binding protein, using the non-cleavable cross-linker, disuccinimidyl suberate, reveals an apamin binding polypeptide of 27 kdaltons under reducing and denaturing conditions in both the CTX-SM and CTX-PSD preparations, similar to that (28 kdaltons) reported for rat CTX-SM fractions. Prior phosphorylation of isolated CTX-PSD had no effect on apamin binding, nor did apamin binding influence subsequent phosphorylation of CTX-PSD. Calmodulin, an intrinsic PSD protein, may not play a role in apamin binding to PSD, since addition of calmodulin, or removal of the calmodulin by EGTA treatment, resulted in no change in the binding capacity of the PSD. The apamin binding protein seems to be bound quite firmly in the CTX-PSD fraction since treatments with 0.5% deoxycholate, 1% N-lauroyl sarcosinate, 4 M guanidine-HCl, pH 7.0, 0.5 M KCl and 1.0 M KCl, could only remove the apamin-receptor complexes from CTX-PSD by 40, 55, 52, 12 and 15%, respectively. These results contrast with the findings that the two detergents mentioned solubilize 80-93% of the receptor from synaptosomal or synaptic membrane fractions, indicating that a good deal of the receptor in these fractions is membrane-bound and not connected to the PSD.(ABSTRACT TRUNCATED AT 400 WORDS)

Ontogenetic changes in the cyclic adenosine 3',5'-monophosphate-stimulatable phosphorylation of cat visual cortex proteins, particularly of microtubule-associated protein 2 (MAP 2): effects of normal and dark rearing and of the exposure to light.

Based on a theory that a norepinephrine-stimulated cascade of events resulting in an increase of intracellular cyclic adenosine 3',5'-monophosphate (cAMP) modulates the state of plasticity for the receptive field property of visual cortical neurons, we have followed the ontogenetic changes in cAMP-stimulated phosphorylation of proteins in whole homogenates obtained from developing visual cortices of cats. In vitro phosphorylation was assayed with and without cAMP and the cAMP-dependent protein kinase, and the phosphoproteins separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis were counted for 32P incorporated from [gamma-32P]ATP. It was found that the regulatory subunits of the cAMP-dependent protein kinase are present and fully active by birth, whereas the synapsin content increases at a rate concomitant with synaptogenesis. These ontogenetic developments are not influenced by dark rearing (DR) from birth, a procedure which postpones the onset of the critical period (CP) for plasticity. By contrast, the cAMP-stimulatable phosphorylation of microtubule-associated protein 2 (MAP 2), which under normal rearing conditions increases from birth to the second month, is strongly modulated by the presence of light in the environment. After DR for various periods, kittens were subsequently exposed to light so as to trigger the onset of the CP that had been postponed. A few hours of light were sufficient to cause a large increase in the in vitro phosphorylation of MAP 2. This effect is not observed in the auditory cortex or the lateral geniculate nucleus of the same animals, or in the visual cortex of normally reared cats which were then dark reared in adulthood. But this effect was seen in the visual cortices of cats following 5 months of DR from birth, animals which by chronological age have passed the CP, presumably because the onset of the CP was extended by the DR procedure. The cAMP-dependent phosphorylation of MAP 2 (and its dephosphorylation) may be an important factor for determining the state of plasticity in the CP through its affecting the dendritic cytoskeletal organization involving tubulin and actin.