Dopamine neuron glutamate cotransmission: frequency-dependent modulation in the mesoventromedial projection.

Mesoventromedial dopamine neurons projecting from the medial ventral tegmental area to the ventromedial shell of the nucleus accumbens play a role in attributing incentive salience to environmental stimuli that predict important events, and appear to be particularly sensitive to the effects of psychostimulant drugs. Despite the observation that these dopamine neurons make up almost the entire complement of neurons in the projection, stimulating their cell bodies evokes a fast glutamatergic response in accumbens neurons. This is apparently due to dopamine neuron glutamate cotransmission, suggested by the extensive coexpression of vesicular glutamate transporter 2 (VGLUT2) in the neurons. To examine the interplay between the dopamine and glutamate signals, we used acute quasi-horizontal brain slices made from DAT-YFP mice in which the intact mesoventromedial projection can be visualized. Under current clamp, when dopamine neurons were stimulated repeatedly, dopamine neuron glutamate transmission showed dopamine-mediated facilitation, solely at higher, burst-firing frequencies. Facilitation was diminished under voltage clamp and flipped to inhibition by intracellular Cs(+) or GDPbetaS, indicating that it was mediated postsynaptically. Postsynaptic facilitation was D1 mediated, required activation of NMDA receptors and closure of voltage gated K(+)-channels. When postsynaptic facilitation was blocked, D2-mediated presynaptic inhibition became apparent. These counterbalanced pre- and postsynaptic actions determine the frequency dependence of dopamine modulation; at lower firing frequencies dopamine modulation is not apparent, while at burst firing frequency postsynaptic facilitation dominates and dopamine becomes facilitatory. Dopamine neuron glutamate cotransmission may play an important role in encoding the incentive salience value of conditioned stimuli that activate goal-directed behaviors, and may be an important subtract for enduring drug-seeking behaviors.

Glutamate is a cotransmitter in ventral midbrain dopamine neurons.

Interactions between apparently separate dopaminergic and glutamatergic pathways figure prominently in the pathophysiology of Parkinson's Disease. So it is surprising that the ventral midbrain dopamine neurons, which give rise to the dopaminergic pathway, may themselves also be glutamatergic. We have addressed this idea in both rat and monkey brain and found that most ventral midbrain dopamine neurons exhibit glutamate immunoreactivity. We used postnatal cell culture to examine ventral midbrain dopamine neurons more closely. In vitro most dopamine neurons exhibit glutamate immunoreactivity, as well as immunoreactivity for phosphate-activated glutaminase, the enzyme principally responsible for the synthesis of neurotransmitter glutamate; inhibition of glutaminase reduces glutamate staining. In single cell microcultures, dopamine neurons make both dopaminergic and glutamatergic synaptic varicosities. Stimulation of individual dopamine neurons evokes a fast excitatory synaptic response mediated by glutamate; it also evokes dopamine release that inhibits the excitatory response via presynaptic D2 receptors. Thus, dopamine neurons appear to exert rapid synaptic actions via their glutamatergic synapses and slower modulatory actions via their dopaminergic synapses, including possibly inhibition of their own glutamatergic synapses. So, in the setting of dopamine neuron demise, there will be a loss of both dopaminergic and glutamatergic inputs to the striatum; furthermore, glutamate released by dopamine neurons may contribute to an excitotoxic cascade and the death of neighboring dopamine neurons.

Mesoaccumbens dopamine neuron synapses reconstructed in vitro are glutamatergic.

The mesoaccumbens projection, formed by ventral tegmental area dopamine neurons synapsing on nucleus accumbens gamma-aminobutyric acid neurons, has been implicated in the pathogenesis of schizophrenia and drug addiction. Despite intensive study, the nature of the signal conveyed by dopamine neurons has not been fully resolved. In addition to several slower, dopamine-mediated, modulatory actions, several lines of evidence suggest that dopamine neurons have fast excitatory actions. To test this, we placed dopamine neurons together with accumbens neurons in microcultures. Surprisingly, most dopamine neurons made excitatory recurrent connections (autapses), which provided a basis for their identification; accumbens gamma-aminobutyric acid neurons were identified by their distinctive size. In 75% of mesoaccumbens cell pairs, stimulation of the dopamine neuron evoked a glutamate-mediated, excitatory synaptic response in the accumbens neuron. Immunostaining revealed dopamine neuron varicosities that were predominantly dopaminergic, ones that were predominantly glutamatergic, and ones that were both dopaminergic and glutamatergic. Despite close appositions of both glutamatergic and dopaminergic varicosities to the dendrites of accumbens neurons, only glutamatergic synaptic responses were seen. In the majority of cell pairs, pharmacologic activation of D2-type dopamine receptors inhibited glutamatergic responses, presumably via immunocytochemically-visualized presynaptic D2 receptors. In some cell pairs, the evoked autaptic and synaptic responses were discordant, suggesting that D2 receptors may be differentially trafficked to different presynaptic varicosities.Thus, dopamine neurons appear to mediate both slow dopaminergic and fast glutamatergic actions via separate sets of synapses. Together with evidence for glutamate cotransmission in serotonergic raphe neurons and noradrenergic locus coeruleus neurons, these results add a new dimension to monoamine neuron signaling that may have important implications for neuropsychiatric disorders.

Dale's principle and glutamate corelease from ventral midbrain dopamine neurons.

While direct application of dopamine modulates postsynaptic activity, electrical stimulation of dopamine neurons typically evokes excitation. Most of this excitation appears to be due to activation of collateral pathways; however, several lines of evidence have suggested that there is a monosynaptic component due to glutamate corelease by dopamine neurons. Recently, more direct evidence obtained in culture has shown that ventral midbrain dopamine neurons release both dopamine and glutamate. Moreover, they appear to do so from separate release sites, calling into question recent modifications of Dale's Principle. The neurochemical phenotype of a given synapse may be determined by subcellular neurotransmitter levels, uptake, or storage. However, the relationship between dopamine and glutamate release from dopamine neuron synapses in the intact brain--and the mechanisms involved--has yet to be resolved.

D1- and D2-like dopamine receptors are co-localized on the presynaptic varicosities of striatal and nucleus accumbens neurons in vitro.

The neuromodulatory actions of dopamine in the striatum and nucleus accumbens are likely to depend on the distribution of dopamine receptors on individual postsynaptic cells. To address this, we have visualized D1- and D2-like receptors on living medium-spiny GABAergic neurons in cultures from the striatum and nucleus accumbens using receptor antagonist fluoroprobes. We labeled D1-like receptors with rhodamine-SCH23390, D2-like receptors with rhodamine-N-(p-aminophenethyl)spiperone and synaptic sites with K+-stimulated uptake of the activity-dependent endocytic tracer FM-143. The fluoroprobes were applied in sequence to assess co-localization. We found that D1- or D2-like receptors were present on about two-thirds of the cells, and co-localized on 22+/-3% (mean +/- S.E.M.) of striatal and 38+/-6% of nucleus accumbens cells. On either D1 or D2 labeled cells, postsynaptic labeling continuously outlined the cell body membrane and extended to proximal dendrites, but not axons. About two-thirds of synaptic varicosities showed D1 or D2 labeling. D1- and D2-like receptors were co-localized on 21+/-4% of striatal and 27+/-3% of nucleus accumbens varicosities. Presynaptic labeling was typically more intense than postsynaptic labeling. The distribution of presynaptic dopamine receptors contrasted with that of postsynaptic GABA(A) receptors, which were clustered in longer patches on neighboring postsynaptic membranes. The extensive presence of D1- and D2-like receptors on presynaptic varicosities of medium-spiny neurons suggests that the receptors are likely to play an important and interacting role in the presynaptic modulation of inhibitory synaptic transmission in the striatum and nucleus accumbens. The significant overlap in labeling suggests that D1-D2 interactions, which occur at the level of individual postsynaptic cells, the circuit level and the systems level, may also be mediated at the presynaptic level. Finally, the ability to visualize dopamine, as well as GABA(A), receptors on the individual synapses of living neurons now makes possible physiological studies of individual mesolimbic system synapses with known receptor expression.

Dopamine neurons make glutamatergic synapses in vitro.

Interactions between dopamine and glutamate play prominent roles in memory, addiction, and schizophrenia. Several lines of evidence have suggested that the ventral midbrain dopamine neurons that give rise to the major CNS dopaminergic projections may also be glutamatergic. To examine this possibility, we double immunostained ventral midbrain sections from rat and monkey for the dopamine-synthetic enzyme tyrosine hydroxylase and for glutamate; we found that most dopamine neurons immunostained for glutamate, both in rat and monkey. We then used postnatal cell culture to examine individual dopamine neurons. Again, most dopamine neurons immunostained for glutamate; they were also immunoreactive for phosphate-activated glutaminase, the major source of neurotransmitter glutamate. Inhibition of glutaminase reduced glutamate staining. In single-cell microculture, dopamine neurons gave rise to varicosities immunoreactive for both tyrosine hydroxylase and glutamate and others immunoreactive mainly for glutamate, which were found near the cell body. At the ultrastructural level, dopamine neurons formed occasional dopaminergic varicosities with symmetric synaptic specializations, but they more commonly formed nondopaminergic varicosities with asymmetric synaptic specializations. Stimulation of individual dopamine neurons evoked a fast glutamatergic autaptic EPSC that showed presynaptic inhibition caused by concomitant dopamine release. Thus, dopamine neurons may exert rapid synaptic actions via their glutamatergic synapses and slower modulatory actions via their dopaminergic synapses. Together with evidence for glutamate cotransmission in serotonergic raphe neurons and noradrenergic locus coeruleus neurons, the present results suggest that glutamatergic cotransmission may be the rule for central monoaminergic neurons.

Visualization of D1 dopamine receptors on living nucleus accumbens neurons and their colocalization with D2 receptors.

To examine the substrate for dopamine (DA) synaptic action in the nucleus accumbens (nAcc), we visualized the cellular and subcellular distribution of DA receptors on postnatal nAcc neurons in culture using fluoroprobe derivatives of DA receptor ligands. Previously, we have shown that rhodamine-N-(p-aminophenethyl)-spiperone (NAPS) (10 nM), a derivative of the D2 antagonist spiperone, labels D2-like receptors on living nAcc neurons. We now show that rhodamine-Sch-23390 (30 nM), a derivative of the D1 antagonist, labels D1-like receptors. Putative specific membrane labeling reached a plateau after about 20 min. Labeling was stereospecific, as it was unaffected by competition with (-)-butaclamol, but blocked with (+)-butaclamol. We found that 52 +/- 7% of nAcc medium-sized neurons showed D1 labeling, which extended onto the dendrites. Labeling was also seen on presynaptic terminals, often abutting D1-positive and D1-negative cell bodies, consistent with a presynaptic modulatory role for D1 receptors. Larger neurons, which may be GABAergic or cholinergic interneurons, were also labeled. By sequential labeling first with rhodamine-Sch-23390 and then rhodamine-NAPS, we found that 38 +/- 6% of medium-sized neurons express both D1- and D2-like receptors, indicating that D1-D2 interactions may occur at the level of single postsynaptic neurons.

Reserpine inhibits amphetamine action in ventral midbrain culture.

Although amphetamine releases catecholamines from isolated secretory vesicles, a number of in vivo experiments have indicated that the vesicular amine transport blocker reserpine does not block amphetamine-induced release. To address this paradox, we examined the effect of reserpine on amphetamine-induced dopamine release from postnatal ventral midbrain neurons in culture. These cultures provide a preparation in which intracellular, extracellular, and releasable dopamine pools can be measured simultaneously. We found that 1 microM reserpine for 90 min reduced stimulation-dependent dopamine release by > 95%. In parallel, reserpine reduced amphetamine-induced dopamine release by > 95% compared with cells not exposed to reserpine or by 75% compared with reserpine-treated cultures. This shows that amphetamine acts principally by redistributing dopamine from synaptic vesicles to the cytosol.

Visualization of antipsychotic drug binding to living mesolimbic neurons reveals D2 receptor, acidotropic, and lipophilic components.

To examine the binding of antipsychotic drugs to living neurons, we applied fluoroprobe derivatives of the D2 antagonist spiperone to mesolimbic system neurons in postnatal culture. We found that rhodamine-N-(p-aminophenethyl)spiperone (rhodamine-NAPS) stereospecifically labeled the plasma membranes of 38 +/- 6% of ventral tegmental area neurons, 22 +/- 7% of which were dopaminergic, and 50 +/- 6% of medium-sized putatively GABAergic nucleus accumbens neurons, with a time constant of approximately 8 min. In contrast, the BODIPY derivative of NAPS rapidly labeled intracellular sites in all neurons in a punctate pattern, consistent with acidotropic uptake. Native antipsychotics also show acidotropic uptake, which we visualized by their displacement of the fluorescent weak base vital dye acridine orange from acidic intracellular compartments. We found that acidotropic uptake correlated best with the partition coefficients of the drugs. With a time constant of 23 min, rhodamine-NAPS labeled all neurons in a pattern suggestive of lipophilic solvation. Thus, initially rhodamine-NAPS makes possible visualization of D2 receptors on living neurons; however, acidotropic uptake and lipophilic solvation obscure receptor labeling and may account for time-dependent factors in the action of antipsychotic drugs, as well as affect their use as radioreceptor ligands.

Amphetamine redistributes dopamine from synaptic vesicles to the cytosol and promotes reverse transport.

Whether amphetamine acts principally at the plasma membrane or at synaptic vesicles is controversial. We find that d-amphetamine injection into the Planorbis giant dopamine neuron causes robust dopamine release, demonstrating that specific amphetamine uptake is not required. Arguing for action at vesicles, whole-cell capillary electrophoresis of single Planorbis dopamine neurons shows that amphetamine reduces vesicular dopamine, while amphetamine reduces quantal dopamine release from PC12 cells by > 50% per vesicle. Intracellular injection of dopamine into the Planorbis dopamine neuron produces rapid nomifensine-sensitive release, showing that an increased substrate concentration gradient is sufficient to induce release. These experiments indicate that amphetamine acts at the vesicular level where it redistributes dopamine to the cytosol, promoting reverse transport, and dopamine release.

GABA synapses formed in vitro by local axon collaterals of nucleus accumbens neurons.

GABAergic medium-spiny neuron axons not only form the principal projections of the nucleus accumbens (nAcc) but also branch locally in a dense network overlapping their own dendrites, suggesting that their recurrent synapses mediate the major information processing functions of the nAcc. We used postnatal nAcc cultures to study these synapses individually. In culture, as in the intact nAcc, medium-spiny neurons account for over 95% of the cells and are GABAergic. Strikingly, these neurons showed a spike afterhyperpolarization (AHP) that was largely blocked by the GABAA antagonist bicuculline. The bicuculline-sensitive AHP occurred without or with latency, and met criteria for monosynapticity; consistent with this, dye fills showed the presence of recurrent axons and a low incidence of dye coupling. Blockade of Ca2+ influx eliminated this autaptic PSP, while TTX almost completely eliminated it, indicating that it is due to exocytic GABA release principally at axodendritic contacts. While blocking GABAB receptors had no direct effect on the autaptic PSP, activating these receptors with baclofen produced presynaptic inhibition, as well as directly mediated hyperpolarization; together, these actions increased the signal-to-noise ratio in the cellular response to synaptic inputs. Bicuculline also increased the signal-to-noise ratio; in addition, it induced burst firing and depolarization inactivation. In contrast, the indirect GABA agonist flurazepam and the GABA uptake blocker nipecotic acid each enhanced autaptic PSPs. Since autapses formed in vitro appear to be functionally equivalent to synapses between neighboring medium-spiny neurons that receive similar inputs, these results bear on the function of intrinsic GABA synapses in the intact nAcc. Thus, intrinsic GABA synapses are likely to regulate the signal-to-noise ratio in nAcc information processing and may be important targets for the modulatory actions of endogenous neurotransmitters and drugs.

Methamphetamine neurotoxicity involves vacuolation of endocytic organelles and dopamine-dependent intracellular oxidative stress.

Methamphetamine (MA) produces selective degeneration of dopamine (DA) neuron terminals without cell body loss. While excitatory amino acids (EAAs) contribute to MA toxicity, terminal loss is not characteristic of excitotoxic lesions nor is excitotoxicity selective for DA fibers; rather, EAAs may modulate MA-induced DA turnover, suggesting that DA-dependent events play a key role in MA neurotoxicity. To examine this possibility, we used postnatal ventral midbrain DA neuron cultures maintained under continuous EAA blockade. As in vivo, MA caused neurite degeneration but minimal cell death. We found that MA is a vacuologenic weak base that induces swelling of endocytic compartments; MA also induces blebbing of the plasma membrane. However, these morphological changes occurred in MA-treated cultures lacking DA neurons. Therefore, while collapse of endosomal and lysosomal pH gradients and vacuolation may contribute to MA neurotoxicity, this does not explain selective DA terminal degeneration. Alternatively, MA could exert its neurotoxic effects by collapsing synaptic vesicle proton gradients and redistributing DA from synaptic vesicles to the cytoplasm. This could cause the formation of DA-derived free radicals and reactive metabolites. To test whether MA induces oxidative stress within living DA neurons, we used 2,7-dichlorofluorescin diacetate (DCF), an indicator of intracellular hydroperoxide production. MA dramatically increased the number of DCF-labeled cells in ventral midbrain cultures, which contain about 30% DA neurons, but not in nucleus accumbens cultures, which do not contain DA neurons. In the DA neuron cultures, intracellular DDF labeling was localized to axonal varicosities, blebs, and endocytic organelles. These results suggest that MA redistributes DA from the reducing environment within synaptic vesicles to extravesicular oxidizing environments, thus generating oxygen radicals and reactive metabolites within DA neurons that may trigger selective DA terminal loss.

Amphetamine and other weak bases act to promote reverse transport of dopamine in ventral midbrain neurons.

Amphetamine-like psychostimulants are thought to produce rewarding effects by increasing dopamine levels at mesolimbic synapses. Paradoxically, dopamine uptake blockers, which generally increase extracellular dopamine, inhibit amphetamine-induced dopamine overflow. This effect could be due to either inhibition of amphetamine uptake or inhibition of dopamine efflux through the transporter (reverse transport). We used weak bases and dopamine uptake blockers in ventral midbrain neuron cultures to separate the effects on blockade of amphetamine uptake from reverse transport of dopamine. Amphetamine, ammonium chloride, tributylamine, and monensin, at concentrations that produce similar reductions in acidic pH gradients, increased dopamine release. This effect was inhibited by uptake blockers. Although in the case of amphetamine the inhibition of release could have been due to blockade of amphetamine uptake, inhibition also occurred with weak bases that are not transporter substrates. This suggests that reduction of vesicular pH gradients increases cytoplasmic dopamine which in turn promotes reverse transport. Consistent with this model, extracellular 3,4-dihydroxyphenylacetic acid was increased by ammonium chloride and monensin, as would be expected with elevated cytoplasmic dopamine levels. These findings extend the weak base mechanism of amphetamine action, in which amphetamine reduces vesicular pH gradients resulting in increased cytoplasmic dopamine that promotes reverse transport.

Identified postnatal mesolimbic dopamine neurons in culture: morphology and electrophysiology.

To examine the intrinsic properties of postnatal mesolimbic dopamine (DA) neurons, we dissociated the ventral tegmental area (VTA) from postnatal rats, enriched for DA neurons by microdissection or gradient purification, and grew the cells in culture. In these cultures, up to 50% of neurons were dopaminergic. DA neurons resembled their in vivo counterparts in soma shapes, and in showing two levels of tyrosine hydroxylase (TH) expression, axodendritic differentiation, two sizes of synaptic vesicles, nest-like synaptic arrangements with non-DA cells, and synaptic specializations. Electrophysiologically, however, they could not be distinguished from non-DA cells, which could be consistent with heterogeneity in cell properties. To examine a functional subset of VTA DA neurons, we retrogradely labeled VTA neurons projecting to the nucleus accumbens. These mesoaccumbens neurons were 86% TH positive, 56% cholecystokinin positive, and 0% neurotensin positive; they also displayed the soma shapes characteristic of DA neurons more generally and two levels of TH expression. Like their in vivo counterparts, mesoaccumbens cells generally fired single broad spikes that were triggered by slow depolarizations and had robust spike afterhyperpolarizations, low- and high-threshold Ca2+ spikes, rapid accommodation of firing, time-dependent anomalous rectification, and hyperpolarizing autoreceptor responses. Strikingly, the expression of these active properties did not change with time in culture. Mesoaccumbens DA cells could be identified by a distinctive subset of properties that made up an electrophysiological signature; however, unlike their in vivo counterparts, they were less often spontaneously active and never fired in bursts. These results suggest that most DA cell properties are intrinsic to the cells, including a significant heterogeneity that is maintained in postnatal culture; their level and mode of activity, however, appear to require afferent input. Culturing identified postnatal VTA DA neurons now makes possible examination of the impact of their individual properties on synaptic function.

Amphetamine and other psychostimulants reduce pH gradients in midbrain dopaminergic neurons and chromaffin granules: a mechanism of action.

Rewarding properties of psychostimulants result from reduced uptake and/or increased release of dopamine at mesolimbic synapses. As exemplified by cocaine, many psychostimulants act by binding to the dopamine uptake transporter. However, this does not explain the action of other psychostimulants, including amphetamine. As most psychostimulants are weak bases and dopamine uptake into synaptic vesicles uses an interior-acidic pH gradient, we examined the possibility that psychostimulants might inhibit acidification. Pharmacologically relevant concentrations of amphetamine as well as cocaine and phencyclidine rapidly reduced pH gradients in cultured midbrain dopaminergic neurons. To examine direct effects on vesicles, we used chromaffin granules. The three psychostimulants, as well as fenfluramine, imipramine, and tyramine, reduced the pH gradient, resulting in reduced uptake and increased release of neurotransmitter. Inhibition of acidification by psychoactive amines contributes to their pharmacology and may provide a principal molecular mechanism of action of amphetamine.

Experiments on the use of DAMP to study retina and cultured neurons.

Immunocytochemical localization of DAMP, a reagent used to detect low pH intracellular compartments, was studied in cultured neurons from rat hippocampus and in frog retinas. We find that DAMP is more sharply localized and that the immunocytochemical reaction is stronger when horseradish peroxidase or other proteins are included in the medium used to administer DAMP to the cells. A likely explanation is that the proteins enter acidified endocytic compartments and there provide sites to which DAMP molecules can be attached during fixation.

Giant Aplysia neuron R2 has distal dendrites: evidence for protein sorting and a second spike initiation site.

Because of its anatomy, the neuron R2 of Aplysia has been used to study how proteins are distributed to their appropriate destinations within the cell. The R2 cell body resides in the abdominal ganglion, while its axons terminate on glands in the skin. Using intracellular injection of HRP and intraxonal recordings, we found that R2 has a dendritic (receptive) arborization in the pleural ganglion. The structure of these dendrites was examined after injecting the soma with 3H-L-fucose, thereby labeling glycoproteins that are transported to all regions of the cell. Light- and electron-microscope autoradiography show that the openings to the dendrites are not on the periphery, but are suspended inside the axon by glial cell infoldings. All of the organelles seen in the axon are found in the dendrites, including 2 types of vesicles. Neither the axon nor the dendrites contain ribosomes. Thus, R2 has 3 functionally distinct regions--cell body, dendrites, presynaptic terminals--that are separated from each other by at least 4 cm. This implies that pre- and postsynaptic proteins made in the cell body are transported along the axon to the pleural ganglion, where they are sorted. To investigate this idea, we exposed the abdominal ganglion to 35S-methionine to label R2's proteins. Analyses by SDS-PAGE of the rapidly transported labeled proteins from R2 consistently showed a 78 kDa band that accumulated in the pleural ganglion and did not move into the peripheral nerves. This then is a putative dendritic constituent.

Synaptic plasticity in vitro: cell culture of identified Aplysia neurons mediating short-term habituation and sensitization.

The gill withdrawal reflex of the marine mollusk, Aplysia californica, shows habituation and sensitization, two simple forms of learning. In order to extend the cellular studies on synaptic plasticity underlying the changes in the reflex behavior, and to explore further the development of synaptic plasticity during synapse formation, we have sought to establish the neural circuit of the gill withdrawal reflex in vitro. We report here the reconstruction of the elementary gill withdrawal circuit in cell culture and find that the cells show short-term homosynaptic depression and heterosynaptic facilitation, the cellular mechanisms of habituation and sensitization, respectively.

Development of plastic mechanisms related to learning at identified chemical synaptic connections in Aplysia.

In the marine snail Aplysia californica learned changes in behavior have been traced to alterations in synaptic efficacy. With the ability to raise the animals in the laboratory, we have explored the development of four types of plastic mechanisms at identified synapses: post-tetanic potentiation and pre-synaptic inhibition, which do not as yet have known behavioral functions, and homosynaptic depression, the cellular mechanism of short-term habituation, and heterosynaptic facilitation, the basis of short-term sensitization. Homosynaptic depression and pre-synaptic inhibition are present early in juvenile life. In contrast, post-tetanic potentiation and heterosynaptic facilitation appear only later, after a discrete interval. The step-wise ontogeny of synaptic plastic mechanisms in Aplysia parallels the gradual emergence of behavior in successive developmental stages. Interestingly, senescence reverses the developmental sequence for habituation and sensitization mechanisms. To pursue further an understanding of the relationship between synapse formation and plasticity underlying learning it will be necessary to extend these studies to dissociated cell culture where mechanisms can be explored on the molecular level. Cell culture may also permit examination of the development of cellular mechanisms underlying classical and operant conditioning which may clarify differences between associative and non-associative mechanisms.