The main subunits of voltage-dependent calcium channels have four domains (I-IV), each comprising six transmembrane regions (S1-S6). Calcium channels are classified into low-voltage activated (LVA) and high-voltage activated (HVA) channels. Ca_V3 (T type) channels are low-voltage activated, while Ca_V1 (L type) and Ca_V2 (N, P/Q and R type) channels are high-voltage activated. The aim of this study is to investigate the molecular basis for the difference in voltage-dependence of activation between LVA and HVA channels.

For this, cDNA chimeras were made between a LVA channel, Ca_V3.1 (α_1G), and a HVA channel, Ca_V1.2 (α_1C), by using overlap extension PCR and standard molecular biological methods. Referring to the four domains of Ca_V1.2 wild-type as CCCC and the wild-type Ca_V3.1 as GGGG, four chimeras (CGGG, GCGG, GGCG and GGGC) were constructed. cRNAs for these constructs were injected into Xenopus oocytes, together with α₂/δ and β subunits. Two-electrode voltage-clamp recordings were made 2-3 days after injection. To obtain current voltage (I/V) curves, depolarising pulses were applied (0.1 Hz, 500 ms duration) from a holding potential of -80 mV, with Ba²⁺ as charge carrier. The voltage for half maximal activation (V_0.5) and slope parameter (k) were obtained from Boltzmann fits.

For wild-type channels, as expected, Ca_V3.1 activated at more negative potentials (V_0.5 -40.8 ± 2.5 mV, n = 11, mean ± S.E.M.) than for the high-voltage activated Ca_V1.2 (V_0.5 6.9 ± 0.9 mV, n = 10). The three chimeras CGGG, GGCG and GGGC activated at high voltages with I/V curves like Ca_V1.2, and indeed the voltages for half-maximal activation for these chimeras were not significantly different from wild-type Ca_V1.2 (V_0.5 5.2 ± 0.5 mV, n = 8, CGGG; 4.6 ± 0.9 mV, n = 9, GGCG; and 5.8 ± 1.0 mV, n = 7, GGGC, Student’s t test). Furthermore, the slope parameters k for these chimeras were also not significantly different from wild Ca_V1.2. However, chimera GCGG activated at low voltages with I/V curves like Ca_V3.1, with a value for V_0.5 -40.0 ± 1.3 mV, n = 9) that was not significantly different from that for Ca_V3.1. The k value for GCGG was also not significantly different from that for Ca_V3.1.

According to current ideas of channel gating, following depolarisation, S4 segments from all four domains must move before the channel can be opened. Our data for three of our chimeras are consistent with this: replacing domains I, III and IV of a LVA channel with the corresponding domain of a HVA channel led to a HVA channel. However, swapping domain II did not alter the voltage dependence of activation, indicating that domain II somehow plays a different role in channel gating from the other calcium channel domains.

We thank the BBSRC for support.