Developmental changes in calcium channel types mediating synaptic transmission in rat auditory brainstem

Abstract

1. Calcium channel blockers were tested on excitatory postsynaptic currents (EPSCs) at the synapse formed by the calyx of Held on the principal cells in the medial nucleus of trapezoid body (MNTB) in brainstem slices of 4- to 14-day-old rats.

2. At postnatal day 4-9 (P4-9), EPSCs were irreversibly suppressed by the P/Q-type Ca²⁺ channel blocker ω-agatoxin-IVA (ω-Aga-IVA, 200 nM) and also by the N-type Ca²⁺ channel blocker ω-conotoxin GVIA (ω-CgTx, 2 μM). A small fraction of EPSCs was resistant to both toxins but abolished by Cd²⁺ (100 μM).

3. After P7, the ω-CgTx-sensitive EPSC fraction diminished and eventually disappeared after P10. Concomitantly the fraction insensitive to both toxins decreased and became undetectable after P10.

4. In contrast, the ω-Aga-IVA-sensitive EPSC fraction increased with development and became predominant after P10. All through the developmental period examined, the L-type Ca²⁺ channel blocker nicardipine (10 μM) had no effect.

5. We conclude that presynaptic Ca²⁺ channel types triggering transmitter release undergo developmental switching during the early postnatal period.

Pharmacological studies using subtype-specific Ca²⁺ channel blockers have revealed that multiple types of Ca²⁺ channel mediate mammalian central synaptic transmission (Takahashi & Momiyama, 1993; Luebke, Dunlap & Turner, 1993). However, at the synapse formed by the calyx of Held in the brainstem, the P-type Ca²⁺ channel blocker ω-agatoxin-IVA (ω-Aga-IVA) almost completely abolished excitatory postsynaptic currents (EPSCs) as well as presynaptic Ca²⁺ currents, whereas the N-type Ca²⁺ channel blocker ω-conotoxin GVIA (ω-CgTx) or the L-type channel blocker dihydropyridine (DHP) had no effect (Takahashi, Forsythe, Tsujimoto, Barnes-Davies & Onodera, 1996b; Forsythe, Tsujimoto, Barnes-Davies, Cuttle & Takahashi, 1998). The 50 % inhibitory dose (IC₅₀) of ω-Aga-IVA on the presynaptic Ca²⁺ current was less than 5 nM, indicating that the Ca²⁺ channel was P-type rather than Q-type (Forsythe et al. 1998). While these results were obtained from 10- to 18-day-old rats, we have noticed that EPSCs in younger animals could be partially blocked by ω-CgTx (see also Wu, Borst & Sakmann, 1997). Thus it is possible that presynaptic Ca²⁺ channel types involved in synaptic transmission change with development. To address this possibility, we have systematically examined the effect of type-specific Ca²⁺ channel blockers on EPSCs at the calyx-medial nucleus of the trapezoid body (MNTB) synapse at different postnatal days. Our results indicate that multiple types of Ca²⁺ channel are involved in auditory synaptic transmission during the early postnatal period and that the responsible type become unique as animals mature.

METHODS

Preparation and recording methods were as previously described (Forsythe & Barnes-Davies, 1993; Takahashi et al. 1996b). Briefly, Wistar rats, 4-14 days old, were decapitated under halothane anaesthesia and transverse slices (250 μm thickness) of superior olivary complex were prepared. Each slice was placed in a recording chamber and superfused with artificial cerebrospinal fluid (ACSF) containing (mM): NaCl, 120; KCl, 2.5; NaHCO₃, 26; glucose, 10; NaH₂PO₄, 1.25; CaCl₂, 2; MgCl₂,1; myo-inositol, 3; sodium pyruvate, 2; ascorbic acid, 0.5 (pH 7.4 when bubbled with 5 % CO₂ and 95 % O₂). The principal neurones in the medial nucleus of trapezoid body (MNTB) were visually identified with a × 40 water immersion objective (Zeiss) attached to an upright microscope (Axioskop, Zeiss). The perfusate routinely contained bicuculline methiodide (10 μM; Sigma) and strychnine hydrochloride (0.5 μM; Sigma) to block inhibitory synaptic responses. The pipette solution contained (mM): potassium gluconate, 97.5; KCl, 32.5; Hepes, 10; EGTA, 5; MgCl₂, 1.0 (pH 7.4, adjusted with KOH). N-(2,6-diethylphenylcarbamoylmethyl)-triethyl-ammonium bromide (QX314; 5 mM; Research Biochemicals International) was routinely included in the internal solution to suppress action potential generation. The electrode resistance was 4-7 MΩ. Series resistance was 10-20 MΩ and monitored throughout the recordings. Cells were voltage clamped at a holding potential of -70 mV. The liquid junction potential between the pipette and ACSF was not corrected. Recordings were made at room temperature (24-28°C).

EPSCs were evoked in MNTB principal neurones at 0.05-0.1 Hz using a bipolar electrode positioned half-way between the mid-line and the MNTB (Forsythe & Barnes-Davies, 1993). Before formation of a gigaohm seal, the principal neurone receiving an intact excitatory input was identified from an orthodromic spike recorded with an extracellular patch pipette. Experiments were made on EPSCs which were evoked in an all-or-none manner with a suprathreshold stimulus and had amplitudes larger than 1 nA at -70 mV (Forsythe & Barnes-Davies, 1993). Drugs were applied by switching between perfusion lines with magnetic valves. Synthetic ω-Aga-IVA (200 nM, Peptide Institute) and ω-CgTx GVIA (2 μM, Peptide Institute) were dissolved in oxygenated ACSF containing cytochrome C (0.1 mg ml⁻¹, Sigma) just before bath application. Nicardipine (10 μM, Sigma) was diluted from a 10 mM stock solution in dimethyl sulphoxide (DMSO, 0.1 % final concentration). Records were low-pass filtered at 5 kHz and digitized at 10 kHz by a LM-12 interface (Dagan Corporation). Values in the text and figures are given as means ±s.e.m., and significance of difference was evaluated by Steel7s multiple comparison test with 0.05 taken as the level of significance.

RESULTS

In auditory brainstem, axonal projections from cochlear nuclei to target nuclei are present before birth but established at around P5 (Kandler & Friauf, 1993). At P3, most EPSCs recorded from MNTB neurones were less than 200 pA in amplitude (see also, Chuhma & Ohmori, 1998). At P4, EPSCs larger than 1 nA could be observed although infrequently. At P6, such large EPSCs were readily evoked. Experiments were made on such EPSCs, which almost certainly arose at the calyx of Held (Forsythe & Barnes-Davies, 1993).

In P6 pups, the N-type Ca²⁺ channel blocker ω-CgTx applied at the saturating concentration (2 μM, Takahashi & Momiyama, 1993) partially blocked EPSCs recorded from MNTB neurones (Fig. 1A). The blocking effect of ω-CgTx appeared irreversible since EPSCs remained suppressed after washing out the toxin for at least 30 min. The relative magnitude of block by ω-CgTx was 33.9 ± 2.2 % (mean ±s.e.m., n= 6 cells). The remaining fraction of EPSCs was largely blocked by the P/Q-type Ca²⁺ channel blocker ω-Aga-IVA (200 nM, Fig. 1A). When ω-Aga-IVA was applied first (Fig. 1B), EPSCs were blocked by 91.5 ± 1.3 % (n= 6). The remaining fraction was further diminished by ω-CgTx. After applications of both toxins, a small fraction of EPSCs remained (4.38 ± 1.5 %, n= 7, Fig. 1A and B) and this fraction was further abolished by Cd²⁺ (100 μM). The L-type Ca²⁺ channel blocker DHP (nicardipine, 10 μM) had no effect on EPSCs (99.9 ± 1.1 %, n= 5). These results indicate that EPSCs are mediated by multiple types of Ca²⁺ channels at P6. An apparent overlap in the fraction suppressed by ω-CgTx and that by ω-Aga-IVA may be due to the power relationship between presynaptic Ca²⁺ current and postsynaptic response (Dodge & Rahamimoff, 1967; Takahashi & Momiyama, 1993; Takahashi et al. 1996b).

Figure 1. Effects of Ca²⁺ channel blockers on the evoked EPSCs at the calyx-MNTB synapses in a 6-day-old rat.

Effects of ω-CgTx (2 μM), ω-Aga-IVA (200 nM) and Cd²⁺ (100 μM) (A and B) on the amplitude of EPSCs. EPSCs were evoked at 0.05 Hz. Blockers were applied during the periods indicated by open bars. Recordings in the right panels show averaged EPSCs (8 events) sampled from different periods (a-d) superimposed. Dotted lines here and in subsequent figures indicate the mean amplitude of EPSCs before drug application.

At P8, the pharmacological feature of EPSCs was qualitatively the same as at P6 (Fig. 2A). However, the fractional block produced by each blocker changed significantly. The ω-CgTx-sensitive EPSC fraction was reduced to about one-third of that at P6 (to 10.8 ± 2.0 %, n= 5, significantly less with P < 0.01). Apparently the fraction which was insensitive to both toxins was also reduced (2.26 ± 0.79 %, n= 8), whereas the ω-Aga-IVA-sensitive EPSC fraction increased to 94.1 ± 1.5 % (n= 5).

Figure 2. Effects of ω-CgTx and ω-Aga-IVA on the evoked EPSCs at the calyx-MNTB synapses in an 8- and a 10-day-old rat.

ω-CgTx (2 μM) reduced the amplitude of EPSCs by 19 % at P8 (A) but had no effect at P10 (B). After application of ω-CgTx and ω-Aga-IVA (200 nM), a small fraction remained at P8 (A) but not at P10 (B). EPSCs were evoked at 0.05 Hz in A and 0.1 Hz in B. Recordings in the right panels show averaged EPSCs (8 events) sampled from different periods (a-d) superimposed.

At P10, ω-CgTx no longer suppressed EPSCs, whereas ω-Aga-IVA almost completely blocked them (Fig. 2B). The toxin-insensitive component was not detectable (< 0.3 %) at this age. As at P6, nicardipine (10 μM) had no effect (100 ± 0.3 %, n= 6, see also Takahashi et al. 1996b; Forsythe et al. 1998). Thus at P10, EPSCs appear to be mediated solely by ω-Aga-IVA-sensitive Ca²⁺ channels.

Figure 3 illustrates developmental changes of the fraction of EPSCs sensitive to ω-CgTx (A), ω-Aga-IVA (B) or those insensitive to both toxins but sensitive to Cd²⁺ (C). The magnitude of the ω-CgTx-sensitive fraction was similar between P4 and P7 but dramatically decreased thereafter until P10 (Fig. 3A). In contrast, the ω-Aga-IVA-sensitive fraction increased with a similar time course (Fig. 3B). Concomitantly, the toxin-insensitive, Cd²⁺-sensitive, fraction diminished (Fig. 3C). Hence in animals older than 10 days, EPSCs are mediated exclusively by ω-Aga-IVA-sensitive Ca²⁺ channels.

Figure 3. Summarized effects of ω-CgTx and ω-Aga-IVA on the evoked EPSCs at different postnatal ages.

Remaining fractions of EPSCs after application of ω-CgTx (2 μM, A), ω-Aga-IVA (200 nM, B) or both toxins (C) plotted against postnatal age. Data points and error bars indicate means and s.e.m. of remaining fractions relative to control (n= 5-8 each at P6-14, n= 3 at P4). Control amplitude of EPSCs were 1.71 ± 0.086 nA (n= 3) at P4, 2.32 ± 0.29 nA (n= 12) at P6, 2.78 ± 0.28 nA (n= 11) at P7, 2.47 ± 0.22 nA (n= 10) at P8, 2.77 ± 0.23 nA (n= 11) at P9, 2.95 ± 0.47 (n= 11) at P10, 2.65 ± 0.35 nA (n= 11) at P11 and 3.42 ± 0.53 nA (n= 10) at P14.

DISCUSSION

We have demonstrated that EPSCs are sensitive to both ω-CgTx and ω-Aga-IVA at the immature calyx-MNTB synapse and that ω-CgTx sensitivity is gradually diminished after P7 and eventually lost at P10. At the adult mice, end-plate currents are exclusively sensitive to ω-Aga-IVA (Katz, Ferro, Weisz & Uchitel, 1996). However, it has been recently found that they are also sensitive to ω-CgTx before P3 (M. D. R. Siri & O. D. Uchitel, personal communication). Thus the situation is somewhat similar to that at the calyx of Held. Developmental changes in Ca²⁺ channel types have been reported also in rat hippocampal neuronal somata, where both the low-voltage- and high-voltage-activated Ca²⁺ currents are present during the first two postnatal weeks, but only the latter remains after 1 month (Thompson & Wong, 1991). In embryonic chick ciliary ganglia, somatic Ca²⁺ currents are sensitive to both ω-CgTx and DHP but become insensitive to the latter with development (White, Crumling & Meriney, 1997). Similarly, transmitter release at the ganglia is sensitive to both drugs in embryo but DHP sensitivity is lost after hatching (Gray, Bruses & Pilar, 1992). Thus, developmental changes in presynaptic Ca²⁺ channel types may be a general feature at many types of synapse.

Our pharmacological results suggest that N-type Ca²⁺ channels are expressed at the calyx of Held only transiently in the early postnatal period. In fact, Ca²⁺ currents directly recorded from the giant calyceal preterminal were insensitive to ω-CgTx in animals older than P10 (Takahashi et al. 1996b; Forsythe et al. 1998), whereas they were sensitive to both ω-CgTx as well as to ω-Aga-IVA in the earlier period (P7, our unpublished observation; see also Wu et al. 1997). Thus N-type Ca²⁺ channels expressed at an immature preterminal must be downregulated in expression or become non-functional as the animals mature. Since Ca²⁺ channel types are also differentially expressed between somata and axon terminals of the same type of neurones (Fisher & Bourque, 1995; see also Momiyama & Takahashi, 1994), there might be a type-specific sorting mechanism which differentiates Ca²⁺ channel type both temporally and spatially.

During postnatal development, nicotinic acetylcholine receptors, glycine receptors and NMDA receptors undergo subunit switching, thereby establishing fast kinetics of synaptic transmission (Mishina et al. 1986; Takahashi, Momiyama, Hirai, Hishinuma & Akagi, 1992; Takahashi et al. 1996a). While the changes in these postsynaptic receptor subunits take several weeks to 1 month, the changes observed here in presynaptic Ca²⁺ channels types were completed within several days. In auditory brainstem, after axonal projections from ventral cochlear nuclei to target MNTB are established at around P5, the calyx of Held undergoes remodelling until P14 (Kandler & Friauf, 1993). At around P7-10, the calyx loses filopodia-like processes (in gerbil; Kil, Kageyama, Semple & Kitzes, 1995) and changes from spoon-shaped to a digitiform structure by around P14 (in rat; Kandler & Friauf, 1993). Unlike other synapses, neither aberrant projections nor synaptic eliminations have been observed at the calyx throughout the whole developmental period in rat (Kandler & Friauf, 1993, but see Kuwabara, DiCaprio & Zook, 1991 for mice). Since EPSCs lose their ω-CgTx sensitivity at P7-10, N-type Ca²⁺ channels might be eliminated in association with the retraction of filopodia-like structure. Interestingly, at the neuromuscular junction of immature mice, with aberrant multiple innervation, end-plate currents are sensitive to ω-CgTx (M. D. R. Siri & O. D. Uchitel, personal communication), whereas the reinnervated adult neuromuscular junction after denervation acquires DHP sensitivity as well as sproutings of nerve terminals (Katz et al. 1996).

At the calyx of Held, it has recently been reported that the synaptic efficacy evaluated by the coefficient of variation of EPSCs increases with development and becomes maximal at P9 concomitantly with an increase in the sensitivity of transmitter release to external Ca²⁺ concentration, thereby establishing a high fidelity transmission (Chuhma & Ohmori, 1998). It has been reported also that the slope of the power relationship between presynaptic Ca²⁺ currents and EPSCs for the ω-Aga-IVA-sensitive components is higher than ω-CgTx-sensitive ones (Wu et al. 1997). Our present results taken together with these reports suggest that the apparently higher Ca²⁺ sensitivity at the more mature synapse may be due, at least in part, to the changes of presynaptic Ca²⁺ channel type from ω-CgTx sensitive to ω-Aga-IVA sensitive. Thus the developmental change in Ca²⁺ channel type may contribute to the high fidelity synaptic transmission at the mature auditory brainstem synapse.