gamma-Aminobutyric acid receptors in brain postsynaptic densities

Role of actin in anchoring postsynaptic receptors in cultured hippocampal neurons: differential attachment of NMDA versus AMPA receptors

We used actin-perturbing agents and detergent extraction of primary hippocampal cultures to test directly the role of the actin cytoskeleton in localizing GABAA receptors, AMPA- and NMDA-type glutamate receptors, and potential anchoring proteins at postsynaptic sites. Excitatory postsynaptic sites on dendritic spines contained a high concentration of F-actin that was resistant to cytochalasin D but could be depolymerized using the novel compound latrunculin A. Depolymerization of F-actin led to a 40% decrease in both the number of synaptic NMDA receptor (NMDAR1) clusters and the number of AMPA receptor (GluR1)-labeled spines. The nonsynaptic NMDA receptors appeared to remain clustered and to coalesce in cell bodies. alpha-Actinin-2, which binds both actin and NMDA receptors, dissociated from the receptor clusters, but PSD-95 remained associated with both the synaptic and nonsynaptic receptor clusters, consistent with a proposed cross-linking function. AMPA receptors behaved differently; on GABAergic neurons, the clusters redistributed to nonsynaptic sites, whereas on pyramidal neurons, many of the clusters appeared to disperse. Furthermore, in control neurons, AMPA receptors were detergent extractable from pyramidal cell spines, whereas AMPA receptors on GABAergic neurons and NMDA receptors were unextractable. GABAA receptors were not dependent on F-actin for the maintenance or synaptic localization of clusters. These results indicate fundamental differences in the mechanisms of receptor anchoring at postsynaptic sites, both regarding the anchoring of a single receptor (the AMPA receptor) in pyramidal cells versus GABAergic interneurons and regarding the anchoring of different receptors (AMPA vs NMDA receptors) at a single class of postsynaptic sites on pyramidal cell dendritic spines.

Brain plasticity and behavior

Brain plasticity refers to the brain's ability to change structure and function. Experience is a major stimulant of brain plasticity in animal species as diverse as insects and humans. It is now clear that experience produces multiple, dissociable changes in the brain including increases in dendritic length, increases (or decreases) in spine density, synapse formation, increased glial activity, and altered metabolic activity. These anatomical changes are correlated with behavioral differences between subjects with and without the changes. Experience-dependent changes in neurons are affected by various factors including aging, gonadal hormones, trophic factors, stress, and brain pathology. We discuss the important role that changes in dendritic arborization play in brain plasticity and behavior, and we consider these changes in the context of changing intrinsic circuitry of the cortex in processes such as learning.

Role of actin in the organisation of brain postsynaptic densities

Brain synaptic junctions are marked by a prominent dense-staining structure, the postsynaptic density (PSD), embedded in the postsynaptic membrane. Isolated PSDs contain a complex mixture of proteins among which the most abundant are the alpha subunit of calcium/calmodulin-dependent kinase II (CaMK II alpha) the membrane cytoskeletal proteins actin and spectrin and receptors for both excitatory and inhibitory neurotransmitters. We have investigated the relationship of these proteins to the junctional structure by extracting isolated PSDs with lithium diiodosalicylate (LIS). This selectively solubilized actin and spectrin while other prominent PSD proteins, such as CaMK II alpha, the AMPA- and NMDA-type glutamate receptors and GABA receptors, were not extracted at all. Electron microscopy revealed that LIS treatment caused some fragmentation of PSDs but that their basic lattice-like structure remained intact. These observations suggest that PSD structure is organised at two levels; a core component containing CaMK II alpha and neurotransmitter receptors which we have previously described as the postsynaptic junctional lattice and a peripheral actin-associated component that draws the lattice components together into the complete PSD structure.

Localization of a 82 kDa protein in postsynaptic density and its association with cytoskeletons

A fraction of synaptic junctional complex (SJC) was prepared from rat synaptosomes and served as antigen material to produce monoclonal antibodies (Mab) for examining the component proteins of the SJC. An antibody, Mab SJ-8, was obtained, which recognized a protein with a molecular weight of 82,000 Da in the SJC preparation by immunoblot analysis. The immunohistochemical localization of the 82 kDa protein was studied with the rat cerebellum. Mab SJ-8 labeled the peripheral areas of the Purkinje and granule cells. Small punctate areas were also stained in the molecular layer with SJ-8. Intracellular localization of the protein was examined with rat brain synaptosomes. Immunoelectron microscopy demonstrated that Mab SJ-8 strongly labeled the postsynaptic density (PSD) and also a fibrous network spreading out of it. However, the antibody did not label the pre- or post-synaptic membrane or the cleft material.

Differential scanning calorimetry characterization of rabbit brain membrane fractions

Various membrane fractions have been prepared from rabbit brain synaptosomes by centrifugation on discontinuous sucrose gradients after osmotic shock. These fractions were characterized by electron microscopy (E.M.), SDS-PAGE and GABA binding. The fractions were then studied by differential scanning calorimetry (DSC). The calorimetric results indicate that all the fractions studied show thermal transitions at around 60 degrees C which correspond to the "melting" of membrane structures. An additional transition at higher temperature (82 degrees C) seems to be associated with an enrichment in central myelin fragments.

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.

Mechanisms of synaptic plasticity. Changes in postsynaptic densities and glutamate receptors in chicken forebrain during maturation

We have shown that the synapse maturation phase of synaptogenesis is a model for synaptic plasticity that can be particularly well-studied in chicken forebrain because for most forebrain synapses, the maturation changes occur slowly and are temporally well-separated from the synapse formation phase. We have used the synapse maturation phase of neuronal development in chicken forebrain to investigate the possible link between changes in the morphology and biochemical composition of the postsynaptic density (PSD) and the functional properties of glutamate receptors overlying the PSD. Morphometric studies of PSDs in forebrains and superior cervical ganglia of chickens and rats have shown that the morphological features of synapse maturation are characteristic of a synaptic type, but that the rate at which these changes occur can vary between types of synapses within one animal and between synapses of the same type in different species. We have investigated, during maturation in the chicken forebrain, the properties of the N-methyl-D-aspartate (NMDA) subtype of the glutamate receptors, which are concentrated in the junctional membranes overlying thick PSDs in the adult. There was no change in the number of NMDA receptors during maturation, but there was an increase in the rate of NMDA-stimulated uptake of 45Ca2+ into brain prisms. This functional change was not seen with the other ionotropic subtypes of the glutamate receptor and was NMDA receptor-mediated. The functional change also correlated with the increase in thickness of the PSD during maturation that has previously been shown to be due to an increase in the amount of PSD associated Ca(2+)-calmodulin stimulated protein kinase II (CaM-PK II). Our results provide strong circumstantial evidence for the regulation of NMDA receptors by the PSD and implicate changing local concentrations of CaM-PK II in this process. The results also indicate some of the ways in which properties of existing synapses can be modified by changes at the molecular level.

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.

Neurochemistry of GABAergic system in cerebral cortex chronically exposed to bromide in vivo

Neurochemical correlates of GABAergic synaptic transmission [binding, uptake, metabolism, and tissue content of gamma-aminobutyric acid (GABA)] were investigated in the cortex of rats that had been given 27 mM bromide in drinking water for periods of time ranging from 1 day to 1 month. No effect of bromide on any of the parameters was found and it is concluded that chronic administration of bromide has no profound effect on GABAergic inhibitory system in the rat cortex.

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)

Characterization of Ca2+/calmodulin-dependent protein kinase associated with rat cerebral synaptic junction: substrate specificity and effect of autophosphorylation

The Ca2+/calmodulin (CaM)-dependent protein kinase associated with rat cerebral synaptic junction (SJ) was characterized, using the SJ fraction as the enzyme preparation, to clarify the functional significance of the enzyme in situ. The protein kinase was greatly activated in the presence of micromolar concentrations of both Ca2+ and calmodulin (EC50 for Ca2+, 1.0 microM; that for CaM, 100 nM). The Km for ATP was 150 microM. SJ proteins were phosphorylated without a lag time, and the phosphorylation reached its maximum within 2-10 min at 25 degrees C. The endogenous substrates consisted of four major (160K, 120K, 60K, and 51K Mr) and 10 minor proteins. Compared with the endogenous substrate phosphorylation, the phosphorylation of exogenously added proteins (myosin light chains from chicken muscle, casein, arginine-rich histone, microtubule-associated protein-2, tau-protein, and tubulin) was weak, although they are expected to be good substrates for the soluble form of the Ca2+/CaM-dependent protein kinase. Autophosphorylation of the enzyme in SJ inhibited its activity and did not alter the subcellular distribution of the enzyme.

GABA receptor binding in rat cerebral cortex and superior cervical ganglion in the absence of GABAergic synapses

High affinity binding of [3H]gamma-aminobutyric acid (GABA) to membrane-containing homogenates prepared from rat cerebral cortex (neonatal or adult) or from adult rat superior cervical ganglion (SCG) was tested for sensitivity to baclofen, bicuculline and several other analogues of GABA. Since no major differences were found between the tissues free of GABAergic synapses (SCG and neonatal cortex) and GABAergic synapse-rich adult rat cortex, it is suggested that the presence of GABAergic synapses does not play a decisive role in the determination of pharmacological sensitivities of high-affinity GABA receptors.

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)

The postsynaptic density: a possible role in long-lasting effects in the central nervous system

A theory is proposed that biochemical changes at the synapse that occur as a result of stimulation of specific neuronal circuits can lead to long-term changes only if alterations occur in synaptic structures in these circuits. The main synaptic structure that is thought to undergo this alteration is the postsynaptic density (PSD). There are many reports in the literature of overall structural changes at the synapse, including the PSD, resulting from various neuronal stimuli. These structural changes are here envisaged to include those of concentration and conformation of PSD proteins, changes that could alter the neural physiology of dendritic spines and even that of the presynaptic terminal.

Selective association of N-methyl aspartate and quisqualate types of L-glutamate receptor with brain postsynaptic densities

Recognition sites for the excitatory neurotransmitter, L-glutamate, were studied in synaptic plasma membranes and postsynaptic densities (PSDs) isolated from rat brains. The results demonstrate (i) that L-glutamate binding sites may be resolved into three distinct subtypes (categories A1, A2, and A4), each corresponding to an electrophysiologically identified receptor class, and (ii) that the N-methyl aspartate (A1) and quisqualate (A2) receptor types are selectively associated with PSDs. L-[3H]Glutamate bound to an apparently homogeneous population of sites in PSDs with a Kd of 3.39 X 10(-7) M and a Bmax (maximum number of binding sites) of 6.1 pmol/mg of protein. Inhibition studies demonstrated that these sites could be resolved into two distinct subtypes. N-Methyl aspartate maximally inhibited 58% of PSD-located L-glutamate binding sites with a Ki of 7.2 X 10(-6) M (the A1 site), and quisqualate inhibited 42% with a Ki of 1.1 X 10(-6) M (the A2 site); the effects of both substances were additive. Experiments with a range of acidic amino acid analogues indicated that the ligand selectivities of these two binding sites conformed to those of the N-methyl D-aspartate and quisqualate receptor classes defined electrophysiologically. The Cl--dependent population of L-glutamate binding sites (the A4 site), which predominates in synaptic membranes, was absent from isolated PSDs.

Characterization of Na+-independent GABA and flunitrazepam binding sites in preparations of synaptic membranes and postsynaptic densities isolated from canine cerebral cortex and cerebellum

The binding of [3H]GABA and [3H]flunitrazepam was performed with synaptic membranes and post-synaptic densities (PSDs) isolated from canine cerebral cortex and cerebellum. Two GABA binding sites were found with cerebral cortex membranes but only one with cerebellar membranes. PSDs isolated from these showed only single binding sites, with cerebellar PSDs exhibiting lower KD values and a larger concentration of sites than did cerebral cortex PSDs. In the case of flunitrazepam, only one binding site was found for all four preparations, with cerebellar PSDs having twice the concentration of sites of cerebral PSDs. Photoaffinity labeling of the flunitrazepam receptor in PSDs resulted in the binding to a 51,000 Mr protein in both cases, with cerebellar PSDs again showing an increased concentration over that found in cerebral cortex PSDs. Based on this work, and on earlier work of ourselves and of others, we conclude that both populations of isolated PSDs contain inhibitory sites, but that the intact PSDs in both preparations are derived from Gray type I, probably excitatory, synapses, and that the inhibitory sites are found in the broken-up material in the PSD fractions which are derived from Gray type II, probably inhibitory, synapses.

L-glutamate receptor heterogeneity: labeling of distinct receptor sub-types using radioligand binding techniques

This study demonstrates (1) that L-[3H]glutamate labels 3 distinct binding sites (types A1, A2 and A4) in isolated rat brain membranes and (2) that only the N-methyl-aspartate (A1) and quisqualate (A2) receptor classes are associated with the postsynaptic density (PSD). L-[3H]glutamate bound to PSDs with Kd 339 nM and Bmax 6 X 1 pmol/mg protein. These sites were resolved into 2 distinct sub-types on the basis of inhibition studies. N-Methyl-aspartate maximally inhibited 57% of PSD-located L-glutamate binding sites (the A1 site) and quisqualate 43% (the A2 site); the effects of both substances were additive. The ligand selectivities of these 2 sites indicated their identity with the N-methyl-D-aspartate and quisqualate receptor classes defined electrophysiologically. The C1--dependent population of L-glutamate binding sites (the A4 site) which predominates in synaptic membranes was absent from PSDs.

Surface antigens of brain synapses: identification of minor proteins using polyclonal antisera

Antigenic proteins of brain synaptic plasma membranes (SPM) and postsynaptic densities (PSD) were characterized using antisera raised against SPM. Immunostaining of brain sections showed that the antigens were restricted to synapses, and electron microscopy revealed staining at both presynaptic terminals and PSDs. In primary brain cell cultures the antisera were also neuron-specific but the antigens were distributed throughout the entire neuronal plasma membrane, suggesting that some restrictive influence present in whole tissue is absent when neurons are grown dispersed. The antigenic proteins with which these antisera react were identified using SDS gel immunoblots. SPM and PSD differed from one another in their characteristic antigenic proteins. Comparison with amido-black stained gel blots showed that in both cases most of these did not correspond to known abundant proteins of SPM or PSDs revealed by conventional biochemical techniques. None of the antigens revealed by the polyclonal antisera were detected by any of a large series of monoclonal antibodies against SPM.

Synaptic junctions isolated from cerebellum and forebrain: comparisons of morphological and molecular properties

Synaptic junction (SJ) fractions have been isolated from rat cerebellum which are similar to forebrain SJs on the basis of morphology and enrichment of synaptic structures. The polypeptide compositions of SJ fractions were analyzed by one- and two-dimensional SDS-gel electrophoresis and stained by the silver technique. Equivalent numbers of proteins that possess similar relative mobilities (Mr), isoelectric points and staining intensities were present in cerebellum and forebrain synaptic fractions. A few prominent differences were observed between cerebellum and forebrain synaptic fractions; cerebellum SJs contained a 240 K protein that was not detected in the forebrain and the 52,000 K, major PSDp protein was present in forebrain SJs in amounts that are approximately 5-fold greater than in cerebellum SJ fractions. The identity of the cerebellum mPSDp was verified by electrophoretic mobility, peptide fingerprinting and [125I]calmodulin binding. Differences between various synaptic fractions in mannose containing glycoproteins were examined by the binding of [125I]concanavalin A (Con A) to gels. On the basis of apparent molecular weights, the glycoproteins in forebrain and cerebellum SPMs were very similar. In contrast, however, the prominent glycoproteins that reside in the postsynaptic junctional membrane of forebrain SJs were undetectable in SJ fractions isolated from cerebellum. SJ fractions from cerebellum contained their own distinct group of Con A binding glycoproteins. SPM and SJ fractions from forebrain and cerebellum were examined for receptors for excitatory (aspartate, glutamate and kainic acid) and inhibitory (GABA) neurotransmitters and the benzodiazepine analog flunitrazepam. On the basis of relative receptor contents, SJ fractions isolated from either brain region were qualitatively similar and bound significant amounts of excitatory and inhibitory transmitters. These findings support the notion that SJs from cerebellum contain a distinct class of synaptic elements that are in large part derived from asymmetric, type I synapses.

Biochemical and morphological comparison of postsynaptic densities prepared from rat, hamster, and monkey brains by phase partitioning

A new procedure for the preparation of postsynaptic densities (PSDs) is described. A synaptic membrane fraction was homogenized in an aqueous two-phase polymer system containing Poly(ethylene glycol) 6000 (5% wt/wt) and Dextran T500 (6% wt/wt) containing 1% 1-o-n-octyl-beta-D-glucoside. Following a brief centrifugation to separate the phases, highly purified PSDs banded at the interface of the two phases. Using this procedure PSDs have been isolated from rat and hamster cerebral cortex and from the frontal cortex, cerebral cortex, hippocampus, and pooled caudate/putamen regions of Macaca mulatta Rhesus monkeys. The isolated PSDs appeared as curved bars when sectioned or as discs when viewed en face in the electron microscope. The hamster PSDs were associated with large numbers of small rod-like structures 4.5 nm thick and 28 nm long. Similar structures were present, although in fewer numbers, in the rat and monkey preparations. Polyacrylamide gel electrophoresis showed that the PSDs contained a complex population of proteins with major components having molecular weights of 180,000, 130,000, 110,000, 94,000, 65,000, 60,000, and 51,000. Reaction of polyacrylamide gels with 125I-concanavalin A (Con A) identified two major (apparent Mr 180,000 and 130,000) and three minor (apparent Mr 230,000, 145,000, and 110,000) Con A-binding glycoproteins in the PSD fractions. Although some quantitative variation between species and brain regions was apparent, the overall protein and glycoprotein composition was similar for all fractions studied.

An analysis of GABA receptor changes in the discrete regions of mouse brain after acute and chronic treatments with morphine

The effects of morphine on the affinity and distribution of GABA receptors in the mouse regions (striatum, medulla, diencephalon, cortex, and cerebellum) were investigated in relation to: (a) acute administration, (b) chronic administration (tolerance), (c) precipitated withdrawal by naloxone, an opiate antagonist, and (d) abrupt withdrawal for 8 and 24 h. The alterations in the affinity as reflected by the dissociation constant (KD) and the number of receptors (Bmax) in the synaptic membranes obtained from controls and various treatments were determined by radioligand binding assay using [3H]muscimol as a ligand. Significant changes were observed in striatum, medulla, and diencephalon, whereas other regions including whole brain exhibited marginal changes. In general the number of GABA receptors increased after tolerance development, which upon abrupt withdrawal returned to control levels except in the case of naloxone-induced precipitated withdrawal. The affinity changes in different regions were diverse in nature and were not evident in the whole brain membranes. These results indicate that: (as) the regional alterations in the affinity and distribution of GABA receptors may play a role in the induction, maintenance, and regression of morphine tolerance; (b) abrupt withdrawal and antagonist precipitated withdrawal affect the GABA system differently, (c) chronic morphine treatment appears to influence the GABA receptors in the cerebellum, a region generally known for its lack of opiate receptors.

Comparison of centrifugation and filtration assays of ligand binding: do multiple GABA receptive sites exist?

Filtration and microcentrifugation procedures for the separation of bound and free ligand are compared for the assay of GABA receptors. The 3H-GABA and 3H-muscimol are employed as ligands in saturation analyses of SPM and JC preparations. A direct comparison of the two methods, applied to two membrane preparations and two ligands, reveals that nonspecific binding is consistently higher with the microcentrifugation procedure. Analysis of the binding data yields essentially the same constants in either case; however, the filtration assay provides a better estimate in all cases.

Regularity and differentiation within the structure of brain postsynaptic densities

Postsynaptic densities (PSDs) isolated from rat forebrain were examined in samples prepared for electron microscopy by negative staining. Two structurally distinct areas could be distinguished within each individual PSD, a peripheral zone composed of a planar array of spherical subunits with a mean diameter of 18 nm and one or more islands of fine granular material enclosed by the subunits. These enclosures correspond in distribution and size to 'holes' which are known to occur in PSDs in intact brain tissue. The spherical subunits can occur singly in the isolated PSD preparations showing that the structure of each individual subunit does not depend upon its interaction with others. These observations indicate that the PSD structure has both regular and differentiated features. These are considered in relation to the possible roles of the PSD in defining the postsynaptic locus and mediating the postsynaptic events of synaptic transmission.

Cytoskeletal organization at the postsynaptic complex

Postsynaptic densities and the adjacent cytoskeleton were examined in deep-etched, unfixed slices of guinea pig anteroventral cochlear nucleus. The postsynaptic density seen in conventional thin sections corresponds to a meshwork of 4-nm filaments associated with intramembrane particles at the postsynaptic active zone of inhibitory as well as excitatory synapses. These filaments intermesh with a lattice of 8- to 9-nm microfilaments, tentatively identified as F-actin, that is concentrated under the postsynaptic density. We postulate that the meshwork of 4-nm filaments anchors receptors to the adjacent microfilament lattice; this extended postsynaptic complex may limit the mobility of receptors and help maintain the curvature of the postsynaptic membrane.