Search

The cellular and synaptic organization of new born mouse cerebellum maintained in organotypic slice cultures was investigated using immunohistochemical and patch-clamp recording approaches. The histological organization of the cultures shared many features with that observed in situ. Purkinje cells were generally arranged in a monolayer surrounded by a molecular-like neuropil made of Purkinje cell dendritic arborizations. Purkinje cell axons ran between clusters of small round cells identified as granule cells by Kv3.1b potassium channel immunolabelling. The terminal varicosities of the Purkinje cells axons enwrapped presumptive neurons of the cerebellar nuclei whereas their recurrent collaterals were in contact with Purkinje cells and other neurons. Granule cell axons established contacts with Purkinje cell somata and dendrites. Parvalbumin and glutamine acid decarboxylase (GAD) immunohistochemistry revealed the presence of presumptive interneurons throughout the culture. The endings of granule cell axons were observed to be in contact with these interneurons. Similarly, interneurons endings were seen close to Purkinje cells and granule cells. Whole cell recordings from Purkinje cell somata showed AMPA receptor-mediated spontaneous excitatory post-synaptic currents (sEPSCs) and GABAA receptor-mediated spontaneous inhibitory post-synaptic currents (sIPSCs). Similar events were recorded from granule cell somata except that in this neuronal type EPSPs have both a NMDA component and an AMPA component. In addition, pharmacological experiments demonstrated a GABAergic control of granule cell activity and a glutamatergic control of GABAergic neurons by granule cells. This study shows that a functional neuronal network is established in such organotypic cultures even in the absence of the two normal excitatory afferents, the mossy fibers and the climbing fibers.

Low-voltage activated (LVA) Ca2+ currents have been characterized in a large variety of neurons including cerebellar Purkinje cells (PCs). This review summarizes and discusses the biophysical, pharmacological properties, as well as the molecular identity of LVA Ca2+ channels described in PCs in various experimental conditions. Putative functional roles for LVA Ca2+ currents include generation of low-threshold Ca2+ spikes (LTS) that underlie burst firing, promotion of intrinsic oscillatory behaviour, Ca2+ entry close to the resting membrane potential and synaptic potentiation. Based on our recent findings on cerebellar rat PCs in slice cultures, this review presents the major evidence demonstrating that LVA Ca2+ channels produce a dendritic initiated LTS with a regulated propagation to the soma. This new role for LVA Ca2+ channels is particularly important in determining firing patterns in PCs.

In isolated chromaffin cells, the high-voltage-activated Ca2+ current, recorded using 5 mM Ca2+ as the divalent charge carrier, exhibits rundown within 10 min, which is delayed for 1 h at least by the addition of 1 mM adenosine 5'-triphosphate (ATP) to the pipette medium. The mechanism of this stabilizing action of ATP has been examined. ATP action is dose dependent; the rundown process, which was delayed at concentrations below 0.4 mM, was totally abolished at higher concentrations. The requirement for ATP was shown to be quite strict: 2 mM inosine 5'-triphosphate (ITP) could not replace ATP, whereas guanosine 5'-triphosphate (GTP) could, but at higher concentrations. This effect of ATP was shown to require the presence of MgCl2 and the liberation of a phosphate group since the ATP analogue 5'-adenylyl-imidodiphosphate (AMP-PNP) could not act as a substitute for ATP, suggesting an action through either adenosine 5'-diphosphate (ADP) or a phosphorylation step. ADP, in the presence of Mg2+ only, could replace ATP in the same concentration range. This effect was shown to be specific to ADP; it was maintained after blocking the pathways which convert ADP into ATP, and could not be mimicked by guanosine 5'-diphosphate (GDP). Similarly, ATP and ADP effects were abolished at an increased internal Ca2+ concentration (pCa 6 instead of pCa 7.7, where pCa = -log10[Ca2+]). Nevertheless, the presence of 1 mM Mg-ADP in the bathing solution did not prevent the rundown of the Ca2+ channels when going to the inside-out patch recording configuration.(ABSTRACT TRUNCATED AT 250 WORDS)

The Ca2+ current recorded by the whole-cell technique in chromaffin cells shows, before the often described rundown, a transient facilitation or runup. Initial current amplitude was 570 +/- 165 pA and then it increased by 49 +/- 23% (n = 19, SD) over 2 +/- 1 min in the absence of adenosine 5'-triphosphate (ATP). In the presence of ATP, this process occurred with the same magnitude but it was slowed in a dose-dependent manner, lasting 17 +/- 2 min with 2 mM ATP (n = 8). Since adenosine 5'-diphosphate (ADP) does not reproduce this ATP effect, a complex series of phosphorylations is likely to intervene and we show that, at least, a cAMP-dependent i.e., cyclic adenosine monophosphate) phosphorylation occurs. Pertussis toxin (PTX) pretreatment yielded an already maximal Ca2+ current (around 1000 pA) at the time of the patch rupture, which only slightly increased thereafter (10%, n = 11). Also, guanosine 5'-diphosphate (GDP) and guanosine 5'-O-(2-thiodiphosphate) (GDP[ beta s]), induced a fast runup, which was absent in the presence of GTP. Furthermore, we show that facilitation does not occur in the presence of dihydrophyridine (DHP) antagonists. Globally, our data suggest that an ATP-dependent phosphorylation stabilizes the inhibitory control exerted by a PTX-sensitive G protein and, as a result, slows down the facilitation of L-type Ca2+ channels. The recruitment of L-type channels can also be facilitated by the application of a DHP agonist or a depolarizing prepulse protocol.l We show that these processes are only effective over a period which parallels the runup and are not additive to it.(ABSTRACT TRUNCATED AT 250 WORDS)

In cultured cerebellar granule cells, we examined the effects of dihydropyridines (DHPs) on K+ currents, using the whole-cell recording configuration of the patch-clamp technique and on Ca(2+)-activated K+ channels ("maxi K+ channels") using outside-out patches. We found that micromolar concentrations of nicardipine, nifedipine, (+) and (-) BAY K 8644, nitrendipine, nisoldipine and (-) nimodipine block 10-60% of macroscopic K+ currents. The most potent of these DHPs was nicardipine and the least potent, (-) BAY K 8644. (+) Nimodipine had no effect on this current. The inhibitory effects of nifedipine and nicardipine were not additive with those of 1 mM tetraethylammonium (TEA). Outside-out recordings of "maxi K+ channels" showed a main conductance of 200 pS (in 77% of the patches) and two subconductance states (in 23% of the patches). Neither nifedipine nor nicardipine affected the main conductance, but decreased the values of the subconductance levels. In 10% of these patches, nicardipine induced a flickering activity of the channel. These findings show that both Ca2+ and K+ channels have DHP-sensitive sites, suggesting similarity in electrostatic binding properties of these channels. Furthermore, cerebellar granule cells may express different subtypes of "maxi K+ channels" having different sensitivities to DHPs. These drugs may provide new tools for the molecular study of K+ channels.

Isolated bovine capillary endothelial cells have been examined for voltage-dependent Ca entry. All cells displayed a low threshold activity, with the main characteristics of a T-type transient current, when examined using whole-cell recording for activation and inactivation and cell-attached conditions or inside-out patches for the elementary conductance (8 pS). 25% of the cells displayed an additional sustained current in 5 mM CaCl2 above -40 mV, which was enhanced by application of BAY K 8644, but almost insensitive to superfusion with nicardipine. Two types of channels (2.8 and 21 pS, in 110 mM BaCl2) were shown to have a BAY K 8644 sensitivity. The large conductance channels were L-type channels. The smaller events were elicited at more hyperpolarized potentials (by some 30 mV). Their mean open time was 16 ms in control conditions. In presence of BAY K 8644, additional long open times were observed (up to 100 ms as compared to 7.8 ms for the time constants of the slow mode of the L-type channel). We refer to these channels as SB channels: of small conductance and sensitive to BAY K 8644. In the presence of nicardipine, SB channels are not noticeably modified, in contrast to the L-type openings which are abolished. Also, SB open times are close to control values when nicardipine is added after a BAY K 8644 application. We suggest that, at physiological concentrations of divalent ions, an SB-type activity is elicited above -40 mV which generates the low threshold sustained current.

Cell-attached patch recordings were used to study calcium channels on the dendritic membrane of rat cerebellar Purkinje cells maintained in culture. Experiments were performed with isotonic BaCl2 (110 mM) in the pipette and isotonic potassium gluconate in the bath to zero the cell membrane potential. Two distinct types of voltage-activated calcium channels were identified. The first one had a small conductance (9 pS), was activated at a low threshold (congruent to -50 mV) and could be inactivated by holding the membrane potential at -30 mV. This channel had the same characteristics as the T channel described in other neuronal preparations. The second type of Ca channel activated at a high threshold (-30 or +10 mV depending on whether BAY K 8644 was added or not to the pipette solution) and was still activatable even when the membrane was held at -40 mV. In the presence of BAY K 8644 this channel had a conductance of 21 pS with long openings. All these characteristics are similar to those of the S (L) Ca channel described in many preparations. The present study is in agreement with our previous experiments on Purkinje dendrites, where we identified low and high threshold Ca currents using the whole-cell configuration. Up to now, no channel corresponding to the N current has been observed but we cannot exclude its presence.

The calcium currents of rat sensory neurones (of the IX and X cranial nerves) grown in culture were studied using whole cell recordings. In cells loaded with CsCl, and bathed in a solution where Na was replaced by choline or Tris, a step depolarization from -80 mV to 0 mV elicited the well-documented sustained Ca current (iCa,s). In contrast, depolarization from -80 mV to -60 mV and up to -20 mV evoked a distinct transient inward current (iCa,t) which could be isolated by imposing an internal pCa 7. It relaxed in about 100 ms and could possibly occur independently of the former current. The transient current was only affected by manipulation of the Ca concentration in the external medium and therefore was considered to be also a transfer of Ca. Ba was shown to act as a substitute with a lower affinity than Ca. The maximal amplitude of this current was in the order of a few hundred pA in Ca 5 mM and Mg 2 mM. Both activation and inactivation occurred in the same voltage span. The underlying event was studied using noise analysis and compared to the Ca transfer occurring during the sustained current as measured in chromaffin cells by Fenwick et al. (1982). We found them to be of similar amplitude.

Using the patch-clamp technique in whole-cell configuration we have investigated the effect of increasing the internal calcium concentration (Cai) from below 10(-8) M to 10(-6) M on the three calcium currents: ICa,T (T for transient), ICa,S (S for sustained), ICa,N (N for neither), recently described in rat sensory neurones. Increasing Cai led to a dose-dependent reduction of the amplitude of ICa,S and, as Cai reached 5 X 10(-7) M ICa,S was nearly abolished. ICa,N is well evidenced from 5 X 10(-10) M to 10(-7) M where its is a large current. Preliminary observations indicate an increase of its inactivation rate following, as expected for a possible Cai dependent-inactivation, the increase of Cai from 5 X 10(-10) M to 10(-7) M. With Ca = 5 X 10(-7) M, all the cells displayed ICa,T and half of the cells in addition ICa,N, but it was of small amplitude. At Cai = 10(-6) M, most of the recorded cells only exhibited ICa,T.

The patch-clamp technique in whole-cell configuration was used to investigate the kinetics of decay of calcium currents in rat sensory neurones. Whole-cell recording permitted control of the internal medium, particularly of the internal free calcium concentration, which was maintained at either 10(-9) M or 10(-6) M using a high concentration of Ca buffer. The inactivation decay of the total Ca current elicited above -10 mV was found to be faster at pCa 6 than at pCa 9. The total current contained three exponential components which were tentatively identified as the three types of Ca currents (IcaT, IcaN and IcaS). Kinetic analyses indicated that the control of the inactivation process by internal Ca results from an effect on both high-threshold Ca currents, IcaN and IcaS. The inactivation kinetics reported in the literature presents a large variability depending on the cell type. We propose that this variability may result from differences in the capacity of those cells to control their internal Ca.

Porcine intermediate lobe (IL) endocrine cells maintained in primary culture have been studied using patch-clamp derived configurations to record unitary activity on outside-out vesicles. Solutions were devised so as to record Cl current in isolation and to fix cytoplasmic Ca concentration [Ca]i between 0.1 microM and 3 microM. Between [Ca]i 0.5 and 1 microM, the chloride permeability was restricted to single events with a small amplitude, that varied linearly with the membrane potential. Mean slope conductance of this chloride channel was 2.5 pS. Single channel analysis yielded two mean open time values of 10 and 55 ms at -80 mV. Relaxations of chloride currents on outside-out patches was examined at different [Ca]i. Relaxation was negligible at 0.15 microM [Ca]i, whereas at higher [Ca]i, the current exhibited relaxation in response to voltage jumps the kinetic of which could be fitted by two exponentials. At 0.5 microM [Ca]i, the fast relaxation time constant was shown to be voltage insensitive with a value of about 10 ms. The slow relaxation time constant had a mean value of 75 ms at -60 mV and increased with membrane depolarization with a twofold change over 120 mV. Another voltage effect was to favour the slow opening mode at the more depolarized potentials: the ratio of fast to slow relaxations being 5:1 at -60 mV as compared to 1:1 at +80 mV). Finally the estimated probability of opening (po) linearly increased with voltage. po displayed a bell-shaped dependence on [Ca]i, so that full activation of the channels was not achieved.

On porcine intermediate lobe (IL) endocrine cells, spontaneously opening chloride channels have been studied and compared to GABA-A activated chloride channels. Elementary currents were recorded mainly from outside-out patches excised from IL cells maintained in culture for 1-4 weeks. Spontaneous inward currents were observed in Cs-loaded cells after replacing Na in the extracellular medium by the impermeant ion choline. This activity, at an internal calcium concentration of 10(-8) M corresponded to a channel for chloride ions with a main conductance level of 26 pS, and substates around 11 pS. The sequence of permeabilities to halides was I greater than Br greater than Cl. These conductance characteristics were common to the GABA-operated channels which also showed a main conductance substate of 23-31 pS. The open time of the 26 pS level mostly encountered in spontaneous activity, was distributed along two modes: one, the most frequent, around 1 ms, and the other around 4 ms. This latter mode was the predominant one observed during GABA and isoguvacine applications but in addition a bursting activity of 19 ms duration was also seen. Specific GABA-A receptor antagonists (bicuculline and SR42641, 1 microM) blocked activity evoked by GABA (1-10 microM), but did not affect spontaneous events. These spontaneous Cl events were only observed in a restricted range of internal Ca concentrations, i.e. between 1 nM and 0.1 microM, and were practically abolished at Cai 1 microM. The GABA-induced activity of Cl channels was also Ca-sensitive, being reduced when Cai reached 1 microM.