Faculty Bibliography

We have used a focal stimulation method to study neurotransmission at synapses onto hippocampal pyramidal neurons in cultures derived from neonatal rats. Single functional boutons were visualized by activity-dependent preloading with the fluorescent dye FM1-43, then focally stimulated by localized application of elevated K+/Ca2+ solution via a puffer pipette, while postsynaptic currents were recorded under whole cell voltage clamp (Liu and Tsien, 1995). This paper gives a detailed description of the main properties of this experimental system and of information it has provided about fundamental properties of hippocampal synapses. Most of the experiments focused on excitatory postsynaptic currents (EPSCs), but preliminary recordings of inhibitory events (IPSCs) are also reported here. The unitary EPSCs at individual synapses varied greatly in amplitude, but were relatively uniform in their time course. The frequency of the synaptic events was greatly reduced by lowering the external Ca2+ concentration or by application of baclofen, a GABAB receptor agonist. Frequent repetitive stimulation produced a decline in the incidence of EPSCs that was readily reversed upon rest. We attribute the decline to exhaustion of a pool of available vesicles; typically, recovery proceeded with a time constant of approximately 40 sec (23 degrees C), and involved a vesicular pool capable of generating approximately 90 EPSCs without recycling. While synaptic currents were broadly distributed in amplitude (Bekkers et al., 1990), this distribution was remarkably similar at multiple synapses on a given postsynaptic neuron. The median synaptic current amplitude varied 4-fold across different cells, decreasing markedly with increasingly dense synaptic innervation. The implications of these results for cellular signal processing and quantal analysis are discussed

Ca2+ channels display remarkable selectivity and permeability, traditionally attributed to multiple, discrete Ca2+ binding sites lining the pore. Each of the four pore-forming segments of Ca2+ channel alpha 1 subunits contains a glutamate residue that contributes to high-affinity Ca2+ interactions. Replacement of all four P-region glutamates with glutamine or alanine abolished micromolar Ca2+ block of monovalent current without revealing any additional independent high-affinity Ca2+ binding site. Pairwise replacements of the four glutamates excluded the hypothesis that they form two independent high-affinity sites. Systematic alterations of side-chain length, charge, and polarity by glutamate replacement with aspartate, glutamine, or alanine weakened the Ca2+ interaction, with considerable asymmetry from one repeat to another. The P-region glutamate in repeat I was unusual in its sensitivity to aspartate replacement but not glutamine substitution. While all four glutamates cooperate in supporting high-affinity interactions with single Ca2+ ions, they also influence the interaction between multiple divalent cations

Four different types of Ca2+ channel alpha 1 subunits, representing the major classes of voltage-gated Ca2+ channels, were individually coexpressed along with alpha 2/delta and beta 2b subunits in Xenopus oocytes. These subunits (and the encoded channel types and major tissues of origin) included alpha 1C (L-type, cardiac), alpha 1B (N-type, central nervous system), alpha 1A (P/Q-type, central nervous system), and alpha 1E (most likely R-type, central nervous system). Divalent cation currents through these channels (5 mM Ba2+) were evaluated with the two-microelectrode voltage-clamp technique. The expressed channels were compared with regard to their responses to a structurally novel, nondihydropyridine compound, mibefradil (Ro 40-5967). In the micromolar concentration range, this drug exerted clear inhibitory effects on each of the four channel types, reducing divalent cation current at all test potentials, with the non-L-type channels being more sensitive to inhibition than the L-type channels under fixed experimental conditions. For all channel types, mibefradil was a much more effective inhibitor at more depolarized holding potentials, suggesting tighter binding of the drug to the inactivated state than to the resting state. The difference in apparent affinities of resting and inactivated states of the channels, calculated based on a modulated receptor hypothesis, was 30-70-fold. In addition, the time course of decay of Ca2+ channel current was accelerated in the presence of drug, consistent with open channel block. The effect of increasing stimulation frequency was tested for L-type channels and was found to greatly enhance the degree of inhibition by mibefradil, consistent with promotion of block by channel opening and inactivation. Allowing for state-dependent interactions, the drug concentrations found to block L-, N-, Q-, and R-type channels by 50% are at least 10-fold higher than half-blocking levels previously reported for T-type channels in vascular smooth muscle cells under similar experimental conditions. This may help explain the ability of the drug to spare working myocardium (strongly negative resting potential, dominance of L-type channels in their resting state) while reducing contraction in blood vessels (presumably involving T-type channels or partially inactivated L-type channels). Thus, mibefradil is a new addition to the family of nonselective organic Ca2+ channel inhibitors, as exemplified by bepridil and fluspirilene, and may prove useful as an experimental tool for studying diverse physiological events initiated by Ca2+ influx. It complements classes of drugs with relatively selective effects on L-type channels, as exemplified by nifedipine and diltiazem

Long-term potentiation has previously been studied with electrophysiological techniques that do not readily separate presynaptic and postsynaptic contributions. Changes in exocytotic-endocytotic cycling have now been monitored at synapses between cultured rat hippocampal neurons by measuring the differential uptake of antibodies that recognize the intraluminal domain of the synaptic vesicle protein synaptotagmin. Vesicular cycling increased markedly during glutamate-induced long-term potentiation. The degree of potentiation was heterogeneous, appearing greater at synapses at which the initial extent of vesicular turnover was low. Thus, changes in presynaptic activity were visualized directly and the spatial distribution of potentiation could be determined at the level of single synaptic boutons

Synaptic transmission between individual presynaptic terminals and postsynaptic dendrites is a fundamental element of communication among central nervous system neurons. Yet little is known about evoked neurotransmission at the level of single presynaptic boutons. Here we describe key functional characteristics of individual presynaptic boutons of hippocampal neurons in culture. Excitatory postsynaptic currents (e.p.s.cs) were evoked by localized application of elevated K+/Ca2+ solution to single functional boutons, visually identified by staining with the vital dye FM1-43 (refs 6, 7). Frequent repetitive stimulation produced a decline in the incidence of e.p.s.cs as the pool of releasable vesicles was exhausted; typically, recovery proceeded with a time constant of about 40 s (23 degrees C), and involved a vesicular pool capable of generating about 90 e.p.s.cs without recycling. At individual synapses, synaptic currents were broadly distributed in amplitude, but this distribution was remarkably similar at multiple synapses on a given postsynaptic neuron. The average size of synaptic currents and of responses to focal glutamate application varied fourfold across different cells, decreasing markedly with increasingly dense synaptic innervation. This raises the possibility of a very effective mechanism for coordinating synaptic strength at multiple sites throughout the dendritic tree

The diversity of Ca2+ channel types in rat cerebellar granule neurons was investigated with whole-cell recordings (5 mM external Ba2+). Contributions of five different high-voltage-activated Ca2+ channel current components were distinguished pharmacologically. Nimodipine-sensitive L-type current and omega-CTx-GVIA-sensitive N-type current contributed 15 and 20% of the total current, respectively. The bulk of the remaining current (46%) was inhibited by omega-Aga-IVA. The current blocked by this toxin was further subdivided into two components, P-type and Q-type, on the basis of differences in their inactivation kinetics and sensitivity to omega-Aga-IVA. P-Type current was noninactivating during 0.1 sec depolarizations, half-blocked at about 1-3 nM omega-Aga-IVA, and contributed approximately 11% of the total current; Q-type current was prominently inactivating, half-blocked at approximately 90 nM omega-Aga-IVA, and comprised 35% of the total current. Both P- and Q-type currents were potently inhibited by the Conus magus toxin omega-CTx-MVIIC. A current component resistant to all of the aforementioned blockers (R-type) displayed more rapid inactivation than the other components and constituted 19% of the total current. The Q-type current, the largest of the current components in the granule neurons, resembles currents that can be generated in Xenopus oocytes by expression of cloned alpha 1A subunits