Thrombin causes pseudopod detachment via a pathway involving cytosolic phospholipase A2 and 12/15-lipoxygenase products.

Thrombin causes rapid pseudopod detachment and shortening in Dunning rat prostatic carcinoma (MAT-Lu) cells. As seen by interference reflection microscopy and by immunofluorescence analysis with antibodies to paxillin and talin, the primary event is disassembly of adhesion sites. Biochemically, thrombin is a potent activator of cytosolic phospholipase A2 and increases eicosanoid production in these cells. The pseudopod effects are blocked by lipoxygenase (but not cyclooxygenase) inhibitors. Arachidonic acid and 12(S)-hydroxyeicosatetraenoic acid or 15(S)-hydroxyeicosatetraenoic acid mimic the thrombin effect. We conclude that in certain cancer cells, thrombin is a pseudopod repellent that exerts its effect via a cascade involving cytosolic phospholipase A2, 12/15-lipoxygenase, and 12(S)- and/or 15(S)-hydroxyeicosatetraenoic acid.

Thrombin-induced growth cone collapse: involvement of phospholipase A(2) and eicosanoid generation.

The studies presented here explore intracellular signals resulting from the action of repellents on growth cones. Growth cone challenge with thrombin or thrombin receptor-activating peptide (TRAP) triggers collapse via a receptor-mediated process. The results indicate that this involves activation of cytosolic phospholipase A(2) (PLA(2)) and eicosanoid synthesis. The collapse response to repellents targets at least two functional units of the growth cone, the actin cytoskeleton and substratum adhesion sites. We show in a cell-free assay that thrombin and TRAP cause the detachment of isolated growth cones from laminin. Biochemical analyses of isolated growth cones reveal that thrombin and TRAP stimulate cytosolic PLA(2) but not phospholipase C. In addition, thrombin stimulates synthesis of 12- and 15-hydroxyeicosatetraenoic acid (HETE) from the released arachidonic acid via a lipoxygenase (LO) pathway. A selective LO inhibitor blocks 12/15-HETE synthesis in growth cones and inhibits thrombin-induced growth cone collapse. Exogenously applied 12(S)-HETE mimics the thrombin effect and induces growth cone collapse in culture. These observations indicate that thrombin-induced growth cone collapse occurs by a mechanism that involves the activation of cytosolic PLA(2) and the generation of 12/15-HETE.

Anatomical distribution of glycoprotein 93 (gp93) on nerve fibers during rat brain development.

Recent studies have implicated glycoconjugates on the membrane of growth cones as the necessary markers and intermediaries for axonal recognition, axonal motility, and pathway development. One such glycoconjugate, glycoprotein 93 (gp93), has been characterized, but the relative distribution of gp93 has yet to be described for the embryonic brain. In this study, the anatomical distribution of gp93 has been analyzed at embryonic day 15 (E15) and E18, and on postnatal day 3 in the rat by using a polyclonal gp93 antibody. Furthermore, fetal brain tissue transplanted into the adult rat eye has been tested for gp93 immunoreactivity, since central noradrenergic neurons in brainstem transplants are known to provide a continuous source of growing axons, even in adult tissue. In general, a greater abundance of gp93 immunoreactivity is apparent in the earlier embryonic stages (E15 and E18), whereas less is seen in the postnatal brain. The regions showing unique dispersal patterns of gp93 are the neuroepithelium, cerebral cortex, septo-hippocampal pathways, brainstem, and midbrain. This study has therefore focused on these areas and found implications for gp93 distribution appearing in the early development of specific neuronal pathways. Moreover, axons stain densely for gp93 within brain tissue transplants. The presence of gp93 in areas of extensive axonal outgrowth in the normal brain and in transplants suggests that this antibody is used as an early marker for axonal growth. Furthermore, gp93 might be used to map normal development in order to improve our understanding of diseases arising from developmental abnormalities.

SRC binding to the cytoskeleton, triggered by growth cone attachment to laminin, is protein tyrosine phosphatase-dependent.

The interaction of the non-receptor tyrosine kinase, Src, with the cytoskeleton of adhesion sites was studied in nerve growth cones isolated from fetal rat brain. Of particular interest was the role of protein tyrosine phosphatases in the regulation of Src-cytoskeleton binding. Growth cones were found to contain a high level of protein tryrosine phosphatase activity, most of it membrane-associated and forming large, multimeric and wheat germ agglutinin-binding complexes. The receptor tyrosine phosphatase PTPalpha seems to be the most prevalent species among the membrane-associated enzymes. As seen by immunofluorescence, PTPalpha is present throughout the plasmalemma of the growth cone including filopodia, and it forms a punctate pattern consistent with that of integrin beta1. For adhesion site analysis, isolated growth cones were either plated onto the neurite growth substratum, laminin, or kept in suspension. Plating growth cones on laminin triggered an 8-fold increase in Src binding to the adherent cytoskeleton. This effect was blocked completely with the protein tyrosine phosphatase inhibitor, vanadate. Growth cone plating also increased the association with adhesion sites of tyrosine phosphatase activity (14-fold) and of PTPalpha immunoreactivity (6-fold). Vanadate blocked the enzyme activity but not the recruitment of PTPalpha to the adhesion sites. In conjunction with our previous results on growth cones, these data suggest that integrin binding to laminin triggers the recruitment of PTPalpha (and perhaps other protein tyrosine phosphatases) to adhesion sites, resulting in de-phosphorylation of Src's tyr 527. As a result Src unfolds, becomes kinase-active, and its SH2 domain can bind to an adhesion site protein. This implies a critical role for protein tyrosine phosphatase activity in the earliest phases of adhesion site assembly.

Expression and distribution of IGF-1 receptors containing a beta-subunit variant (betagc) in developing neurons.

Betagc is a beta-subunit variant of the insulin-like growth factor-1 (IGF-1) receptor highly enriched in growth cone membranes prepared by subcellular fractionation of fetal rat brain (). The present study is focused on the expression and on the cellular and subcellular distribution of betagc in developing neurons and differentiating PC12 cells. In the developing cerebral cortex and, at least at early stages, in cultured primary neurons, betagc expression was found to be correlated with neurite outgrowth. In PC12 cells betagc expression was nerve growth factor (NGF)-dependent and also paralleled neurite outgrowth. In contrast, beta-subunits of the insulin receptor and/or of other IGF-1 receptors ("betaP5"; detected with antibody AbP5) were downregulated as betagc expression increased. Immunofluorescence studies confirmed the enrichment of betagc at growth cones and demonstrated morphologically its spatial separation from betaP5, which is confined to the perikaryon. At the growth cone, betagc colocalizes and associates in a proximal region with microtubules, but it seems independent of the more peripheral microfilaments. Some betagc immunoreactivity is detected in the perinuclear region of PC12 cells, most likely the Golgi complex and its vicinity. betagc seems to emerge from the periphery of this structure in an apparently vesicular compartment distinct from that carrying synaptophysin to the growth cones. The facts that (1) betagc expression is correlated closely with neurite outgrowth, that (2) it is regulated in PC12 cells by a neurotrophin, NGF, and that (3) betagc is concentrated in the proximal growth cone region raise new questions regarding a possible role of IGF-1 receptors containing betagc in the regulation of neurite growth.

Characterization of phospholipase A2 activity enriched in the nerve growth cone.

Nerve growth cones isolated from fetal rat brain exhibit in their cytosol a robust level of phospholipase A2 activity hydrolyzing phosphatidylinositol (PI) and phosphatidylethanolamine (PE) but not phosphatidylcholine (PC). Western blot analysis with an antibody to the well-characterized cytosolic phospholipase A2 (mol wt 85,000) reveals only trace amounts of this PC- and PE-selective enzyme in growth cones. By gel filtration on Superose 12, growth cone phospholipase A2 activity elutes essentially as two peaks of high molecular mass, at approximately 65 kDa and at well over 100 kDa. Anion exchange chromatography completely separates a PI-selective from a PE-selective activity, indicating the presence of two different, apparently monoselective phospholipase A2 species. The PI-selective enzyme, the predominant phospholipase A2 activity in whole growth cones, is enriched greatly in these structures relative to their parent fractions from fetal brain. This phospholipase A2 is resistant to reducing agents and is found in the cytosol as well as membrane-associated in the presence of Ca2+. However, its catalytic activity is Ca(2+)-independent regardless of whether the enzyme is associated with pure substrate or mixed-lipid growth cone vesicles. The PE-selective phospholipase A2 in growth cones was studied in less detail but shares with the PI-selective enzyme several properties, including intracellular localization, the existence of cytosolic and membrane-associated forms, and Ca2+ independence. Our data indicate growth cones contain two high-molecular-weight forms of phospholipase A2 that share many properties with known, Ca(2+)-independent cytosolic phospholipase A2 species but that appear to be monoselective for PI and PE, respectively. In particular, the PI-selective enzyme may represent a new member of the growing family of cytoplasmic phospholipases A2. The enrichment of the PI-selective phospholipase A2 in growth cones suggests it plays a major role in the regulation of growth cone function.

Axonal origin and purity of growth cones isolated from fetal rat brain.

The investigation of the molecular properties of nerve growth cones depends to a significant degree on their isolation from fetal brain in the form of 'growth cone particles' (GCPs). The availability of markers for developing axons and dendrites, as well as glial cells, has made it possible to characterize the GCP fraction in much greater detail than before and to optimize its yield. Marker analyses show that a member of the N-CAM family (5B4-CAM), synaptophysin, and especially GAP-43 and non-phosphorylated tau, are enriched in the GCP fraction. In contrast, MAP2 and, particularly, glial fibrillary acidic protein and vimentin are fractionated away from GCPs. Furthermore, GCP yield can be doubled relative to the original procedure, without compromising purity, by raising the sucrose concentration of the fractionation gradient's uppermost layer. The results indicate that GCPs are highly purified growth cone fragments with very little glial contamination, and that they are primarily of axonal origin.

Na+ channel changes in the growth cone and developing nerve terminal.

Previous saxitoxin binding studies indicated two forms of the sodium channel in the fetal rat brain; a low-affinity precursor located in an internal membrane compartment, present exclusively in growth cones and a high-affinity mature form present in the plasmalemma of growth cones and characteristic of synapses. This raises the questions (1) of the presence or absence of the beta2 subunit in these channel forms and (2) of the developmental regulation of the the beta2 subunit. Antibodies against the alpha and beta2 channel subunits were used to probe Western blots of subcellular fractions from rat brains at embryonic day 18 (E18), pups at postnatal (P) days 7-25, and adults, as well as purified sodium channels from adult brain. In both synaptosomes and the purified sodium channel the beta2 antibody recognized the expected band at 38 kDa under reducing conditions. However, in contrast to the alpha subunit, this band was absent at E18 and became apparent only from P7 onwards. At the earlier time intervals a very prominent immunoreactive band of unknown identity was evident at 260--300kDa, which declined in intensity concomitant with the appearance of the 38 kDa beta2 band. These data indicate that beta2 subunits are regulated independently from alpha subunits, are absent in differentiating neurons, and hence are not necessary for insertion of the sodium channel into the plasmalemma, at least during early development of the neuron.

Insulin-like growth factor I receptors of fetal brain are enriched in nerve growth cones and contain a beta-subunit variant.

Nerve growth cones isolated from fetal rat brain are highly enriched in a 97-kDa glycoprotein, termed beta gc, that comigrates with the beta subunit of the IGF-I receptor upon two-dimensional PAGE and is disulfide-linked to this receptor's alpha subunit. Antibodies prepared to a conserved domain shared by the insulin and IGF-I receptor beta subunits (AbP2) or to beta gc were used to study receptor distribution further. Subcellular fractionation of the fetal brain segregated most AbP2 immunoreactivity away from growth cones, whereas most beta gc immunoreactivity copurified with growth cones. Experiments involving ligand-activated receptor autophosphorylation confirmed the concentration of IGF-I but not of insulin receptors in growth cone fractions. These results indicate the enrichment of IGF-I receptors in (presumably axonal) growth cones of the differentiating neuron. Furthermore, the segregation of beta gc from AbP2 immunoreactivity suggests that such neurons express an immunochemically distinct variant of the IGF-I receptor beta subunit at the growth cone.

Growth cone enrichment and cytoskeletal association of non-receptor tyrosine kinases.

Fetal rat brain (E18) expresses at least three c-src-like, membrane-associated non-receptor tyrosine kinases: c-src, fyn, and lyn. c-src and fyn are the most abundant and are highly enriched in a subcellular fraction of nerve growth cones (GCPs). To study the cytoskeletal association of these tyrosine kinases, Triton X-100-resistant fractions were prepared from GCPs. All three non-receptor tyrosine kinases are associated with the cytoskeleton to a significant degree with the relative affinities: fyn > c-src > lyn. The binding is sensitive to ionic strength and to phosphotyrosine, but not to phosphoserine or phosphothreonine. To investigate the regulation of this association we used phosphatase inhibitors to increase phosphotyrosine levels in GCPs. This resulted in the release of c-src from the cytoskeleton. Under these conditions tyrosine phosphorylation was increased selectively in released c-src and primarily on tyrosine 527. Cytoskeletally bound c-src had a higher specific kinase activity than Triton X-100-soluble c-src. These findings indicate that src family members interact in a regulated manner with the cytoskeleton in non-transformed cells. This regulation is explained by a model in which c-src binds to the cytoskeleton via its SH2 domain and is released when phosphorylated tyrosine-527 binds to this domain intramolecularly, inhibiting kinase activity.

Characterization of gp93, a novel, highly heterogeneous glycoprotein present in growth cone membranes.

gp93 was first described in growth cones from fetal rat brain as a 90-97-kDa glycoprotein family that binds wheat-germ agglutinin and consists of at least 12 different isoelectric variants (pl range approximately 4.9-6.4). Of particular interest is that different sets of gp93 variants are expressed in growth cones isolated from different brain regions. The preparation of a polyclonal antibody to gp93 allowed further characterization of this glycoprotein. The carbohydrate groups of gp93 were partially characterized by digestion with different glycosidases. The results indicate that most or all oligosaccharide units are N-linked (asparagine-linked) and contain sialic acid. Two-dimensional polyacrylamide gel electrophoresis and western blot with anti-gp93 show that deglycosylated gp93 is an only slightly heterogeneous polypeptide of 66 kDa, indicating that gp93 heterogeneity is due, primarily or exclusively, to differential glycosylation. Analysis of the tissue distribution in fetal rat showed gp93 to be highly enriched in the brain. Immunoblots and immunostaining of cross sections of developing cerebellum revealed that gp93 is developmentally regulated in this tissue, associated primarily with growing parallel fibers and Purkinje dendrites. Immunostaining of neurons in culture shows significant amounts of gp93 in elongating neurites and growth cones. Our results indicate that gp93 is a developmentally regulated glycoprotein of the brain that is most prominent in growth cones and growing neurites and that appears to be glycosylated differentially by different neurons.

A chloride channel reconstituted from fetal rat brain growth cones.

Chloride channels were reconstituted into planar lipid bilayers isolated from a preparation of growth cone particles (GCPs) isolated from fetal rat brain. One type of channel was predominantly seen and some of its biophysical and pharmacological properties were studied. The single channel i-V relationship was curvilinear with a chord conductance of 75 pS at +30 mV in symmetric 200 mM NaCl solutions buffered with phosphate. The channel was inactivated by depolarization, and this inactivation was reversed rapidly upon returning to -25 mV. The Cl- channel was significantly permeant to Na+ ions (PNa/PCl = 0.26), and the relative halide permeabilities were determined to be: I(1.92) > Br(1.73) > Cl(1.0) > F(0.34). The channel was inhibited by the common stilbene compounds (DIDS, SITS, DNDS), as well as by Zn2+ ions and an indanyloxyacetic acid derivative. A developmental role for the GCP Cl- channel is suggested by the observation that adult rat brain synaptosomal membranes were nearly devoid of this type of Cl- channel.

Arachidonic acid turnover and phospholipase A2 activity in neuronal growth cones.

We analyzed de novo synthesis and local turnover of phospholipids in the growing neuron and the isolated nerve growth cone. The metabolism of phosphatidylinositol (PI) was studied with regard to the incorporation of saturated and unsaturated fatty acids and inositol. A comparison of de novo phospholipid synthesis in the intact neuron (whole brain, cell cultures) versus local turnover in isolated growth cone particles (GCPs) from fetal rat brain revealed different incorporation patterns and, in particular, high arachidonic acid (AA) turnover in PI of GCPs. These observations, together with elevated levels of free AA (2.5% of total AA content) in GCPs, demonstrate the predominance of acylation/deacylation in the sn-2 position of PI. GCP phospholipase A2 (PLA2) activity was demonstrated using [3H]-or [14C]AA-phosphatidylcholine (PC) or -PI as the substrate in vitro and GCPs or a cytosolic GCP extract as the source of enzyme. In contrast to PC, which is hydrolyzed very slowly, PI is a very good GCP PLA2 substrate. PLA2 activity is much higher in GCPs than that of phospholipase C, as demonstrated by the comparison of AA and inositol turnover, by the low levels of 1,2-diacylglycerol generated by GCPs, and by the resistance of AA release to treatment of GCPs with RHC-80267, a specific inhibitor of diacylglycerol lipase. The predominance of PLA2 activity in GCPs raises questions regarding its regulation and the functional roles of PI metabolites, especially lysocompounds, in growth cones.

Sites of plasmalemmal expansion in growth cones.

Studies on plasmalemmal expansion in isolated nerve growth cones identified large, clear vesicles characteristically found in growth cones as the plasmalemmal precursor. The present article examines these plasmalemmal precursor vesicles (PPVs) in greater detail in the intact cell. (a) Pulse-chase experiments with the phospholipid precursor, [3H]glycerol, followed by radioautographic analysis show that PPVs in distal neurites and growth cones are labeled prior to equilibration of the label with the plasmalemma. (b) Pulse-chase experiments with lectin-ferritin conjugates demonstrate that PPVs are not endocytotic, that they contain lectin receptors, and that, during growth, patches of lectin receptors appear on the plasmalemma covering PPV clusters. (c) Freeze-fracture studies show that this plasmalemma shares with PPVs a paucity of intramembrane particles. (d) Lectin labeling experiments and freeze-fracture analysis demonstrate, furthermore, that the plasmalemma forms a network of invaginations at the base of PPV clusters. (e) Correlative studies indicate that the refractive 'vacuoles' seen in growth cones by phase-contrast light microscopy correspond to the PPV clusters seen at the ultrastructural level. These results confirm the identity of the plasmalemmal precursor in the intact cell and demonstrate that PPV clusters form distinct, dynamic organelles specialized for plasmalemmal expansion in the growth cone.

A complex consisting of pp60c-src/pp60c-srcN and a 38 kDa protein is highly enriched in growth cones from differentiated SH-SY5Y neuroblastoma cells.

Nerve growth cones of primary neurons are highly enriched in the proto-oncogene product pp60c-src. In order to investigate this molecule further in growing neuronal cells, growth cone and cell body fractions were prepared from human SH-SY5Y neuroblastoma cells differentiated neuronally in vitro under the influence of phorbol ester. The fractions were characterized ultrastructurally and by biochemical criteria. The neuronal (pp60c-srcN) and the fibroblastic (pp60c-src) forms of pp60src are slightly enriched and activated in the growth cones relative to the perikarya. Immunoprecipitates of pp60src from differentiated SH-SY5Y growth cones contain at least four phosphoproteins in addition to pp60src. One of these, pp38, migrates as a 100-140 kDa complex with pp60src under non-reducing conditions of gel electrophoresis. The pp38/pp60src complex is not easily detected in non-differentiated SH-SY5Y cells or perikarya of differentiated SH-SY5Y cells, but it is highly enriched in the growth cone preparation. These data suggest that growth-cone pp60src exists in a disulfide-linked oligomeric complex. The complex appears to be assembled only in the cell periphery and may be dependent upon neuronal differentiation.

Plasmalemmal insertion and modification of sodium channels at the nerve growth cone.

We have characterized voltage-dependent sodium channels in growth cones (GCPs) isolated from fetal rat brain using saxitoxin and TTX binding as well as recordings from channels reconstituted into lipid bilayer membranes. Both high- and low-affinity binding sites are present in GCP membranes. However, the two binding sites are segregated largely or completely, with the high-affinity binding sites in the plasmalemma, and the low-affinity sites in an internal membrane compartment. Plasmalemmal insertion of these internal sites can be triggered by high-potassium depolarization and depends on a metalloendoprotease-requiring mechanism. These observations indicate that a precursor-product relationship exists between the internal and external sodium channels of the growth cone, and therefore suggest that channel externalization causes conversion of low-affinity to high-affinity saxitoxin receptors. This conversion may represent a step of channel capacitation.

A developmentally regulated plasmalemmal antigen present in synaptosomes but not in growth cones.

Monoclonal antibody 2L4 was generated against rat synaptosomes but does not cross-react with nerve growth cones. Expression of the 2L4 antigen is developmentally regulated in a manner that is, in part, the opposite of the expression of the 5B4-CAM antigen, a marker of neuronal outgrowth belonging to the N-CAM family. While 5B4-CAM appears and increases during sprouting and then decreases to reach low levels in the adult, the 2L4 antigen appears only late in development, when neuronal outgrowth ceases, at or around the time of synaptogenesis. Once expressed, the antigen is found on the entire plasmalemmal surface of the neuron, but seems to be enriched at synaptic endings, at least of some neuron types. Biochemical analyses involving blotting of non-denaturing gels and immunoaffinity chromatography identify the antigen as a pair of polypeptides with similar, basic isoelectric points. These polypeptides form a somewhat diffuse, probably glycosylated band at 67 kDa and may be part of a hetero-oligomeric complex. The localization, biochemical, and developmental results suggest that the 2L4 antigen is a plasmalemmal marker of maturing and/or mature neurons whose expression may be triggered by synaptogenesis.

Variable membrane glycoproteins in different growth cone populations.

The question of whether growth cones generated by different neurons contain distinctive membrane glycoproteins was examined. Growth cone particles (GCPs) were isolated from specific regions of fetal or early postnatal brain, and their membrane proteins were analyzed by 2D gel electrophoresis and Western blotting, using WGA as a probe. These blots were compared to those generated by synaptosomes from adult brain. The patterns reveal a number of WGA-binding glycoproteins that are uniformly present in these subcellular fractions and others that are found in GCPs from selected brain regions only. The results indicate, therefore, substantial pattern diversity for the different, restricted growth cone populations. Some of the WGA-binding glycoproteins seen in GCPs disappear with increasing age and are absent from synaptosomes, while others seem to become more prominent. One of the glycoprotein complexes present in all GCP and synaptosome fractions analyzed is gp93. It has an apparent molecular weight of 90-97 kDa and exhibits unusually high heterogeneity in GCPs from whole fetal brain. The gp93 complex covers a pI range from about 4.9 to about 6.4 and consists of at least 12 different species, probably isoelectric variants. In GCPs from different brain regions, the sets of gp93 species observed are different and characteristic. Neuraminidase digestion shifts the gp93 pattern to a more neutral pI but simplifies it only partially, indicating that variable sialic acid content explains the molecular diversity to some extent. Thus, gp93 is a glycoprotein complex whose members are expressed and/or posttranslationally processed differentially in different growth cone populations. Such a glycoprotein family may be involved in selective cell-cell recognition.

5B4-CAM expression parallels neurite outgrowth and synaptogenesis in the developing rat brain.

In order to study whether 5B4-CAM expression parallels neurite outgrowth and synaptogenesis, a monoclonal antibody, 5B4, was used, which recognizes both fetal (185-250 kD) and adult (140 kD, 180 kD) forms of the neural cell adhesion molecule (N-CAM), to identify and localize the antigen in rat tissue during developmental ages P1 through P31 and in adults between P60 and 2 years of age. A ubiquitous pattern of intense immunolabelling was detected during the earliest stages of development. 5B4-CAM expression paralleled process outgrowth and the early stages of synaptogenesis in the cerebral cortex, hippocampal formation, and cerebellum. In the adult, immunoreactivity was generally less intense, but the cerebral cortex and hippocampal and cerebellar molecular layers, all areas implicated in learning-associated plasticity, retained substantial immunoreactivity. The inner one-third of the dentate gyrus molecular layer, an area implicated in axonal sprouting and reactive synaptogenesis, was particularly intensely labelled. Evidence from this work suggests that 5B4-CAM expression may be useful in monitoring neurite outgrowth and the early stages of synapse formation during development and possibly axonal sprouting and reactive synaptogenesis in the adult.

N-type and L-type calcium channels are present in nerve growth cones. Numbers increase on synaptogenesis.

We have demonstrated the presence of both N- and L-type calcium channels in growth cone and other subcellular fractions of fetal rat brain, using the ligands omega-conotoxin GVIA for N-type channels and nitrendipine for L-type channels. The N-type channels seem to be distributed evenly throughout the perikaryon, neurite shaft and growing tip of the neurons. In contrast, the L-type channels appear to have a lower density in the growth cone than on the rest of the neuron. These observations apply at least within the limitations of cell fractionation technology. We have also studied both calcium channel subtypes in rat brain synaptosomal membranes. In both adult and fetal fractions there are approximately 6 times more N-type than L-type channels. Synaptosomal membranes contain more N- and L-type channels than any of the fetal subfractions, indicating that there is a substantial increase in calcium channel numbers upon synaptogenesis.