Elementary mechanisms producing facilitation of Cav2.1 (P/Q-type) channels

Abstract

The regulation of Ca(V)2.1 (P/Q-type) channels by calmodulin (CaM) showcases the powerful Ca(2+) decoding capabilities of CaM in complex with the family of Ca(V)1-2 Ca(2+) channels. Throughout this family, CaM does not simply exert a binary on/off regulatory effect; rather, Ca(2+) binding to either the C- or N-terminal lobe of CaM alone can selectively trigger a distinct form of channel modulation. Additionally, Ca(2+) binding to the C-terminal lobe triggers regulation that appears preferentially responsive to local Ca(2+) influx through the channel to which CaM is attached (local Ca(2+) preference), whereas Ca(2+) binding to the N-terminal lobe triggers modulation that favors activation via Ca(2+) entry through channels at a distance (global Ca(2+) preference). Ca(V)2.1 channels fully exemplify these features; Ca(2+) binding to the C-terminal lobe induces Ca(2+)-dependent facilitation of opening (CDF), whereas the N-terminal lobe yields Ca(2+)-dependent inactivation of opening (CDI). In mitigation of these interesting indications, support for this local/global Ca(2+) selectivity has been based upon indirect inferences from macroscopic recordings of numerous channels. Nagging uncertainty has also remained as to whether CDF represents a relief of basal inhibition of channel open probability (P(o)) in the presence of external Ca(2+), or an actual enhancement of P(o) over a normal baseline seen with Ba(2+) as the charge carrier. To address these issues, we undertake the first extensive single-channel analysis of Ca(V)2.1 channels with Ca(2+) as charge carrier. A key outcome is that CDF persists at this level, while CDI is entirely lacking. This result directly upholds the local/global Ca(2+) preference of the lobes of CaM, because only a local (but not global) Ca(2+) signal is here present. Furthermore, direct single-channel determinations of P(o) and kinetic simulations demonstrate that CDF represents a genuine enhancement of open probability, without appreciable change of activation kinetics. This enhanced-opening mechanism suggests that the CDF evoked during action-potential trains would produce not only larger, but longer-lasting Ca(2+) responses, an outcome with potential ramifications for short-term synaptic plasticity.

Figure 1.

Ca²⁺ regulation of Ca_V2.1 (P/Q-type) Ca²⁺ channels, as seen from macroscopic currents. Currents are from HEK293 cells transfected with recombinant P/Q-type channels (α_1A/β_2a/α₂δ), here and throughout. (A) Evidence of Ca²⁺ regulation during stimulation by action-potential waveforms (APW). Top, 100-Hz APW train, with APWs derived from action potentials recorded in the calyx of Held (Borst et al., 1995; Patil et al., 1998). The single APW on the extreme left is displayed with an expanded timebase, compared with APWs on the right. Middle, corresponding evoked Ca²⁺ currents. Over the first several stimuli, Ca²⁺ currents increase in amplitude, according to a Ca²⁺-dependent facilitation (CDF) process. With continued repetitive stimulation, the current amplitudes decline by a Ca²⁺-dependent inactivation (CDI) process. The first response is reproduced on the far left, using a fast time base. Bottom, Ba²⁺ currents evoked in the same cell by the identical APW stimuli. Little facilitation or inactivation is apparent. Vertical bars, 1 nA. (B) CDF as characterized by a rectangular voltage-pulse protocol. Top, voltage pulse paradigms. To resolve CDF while minimizing CDI, test pulse depolarizations (to +5 mV) are comparatively short (50 ms). Test pulse depolarization is delivered either without a preceding voltage prepulse (gray segment), or with a brief voltage prepulse (to +20 mV). Middle, corresponding evoked Ca²⁺ currents. Without a prepulse, test pulse depolarization initially evokes rapid activation to a lower level, marked L. Following this, Ca²⁺ current (gray trace) slowly increases to a higher level H, according to CDF. When test pulse depolarization follows a voltage prepulse, Ca²⁺ current (black trace) activates immediately to the higher level, because channels have already undergone CDF in the prior prepulse. Bottom, Ba²⁺ currents evoked identically in the same cell show little indication of facilitation. Unless noted otherwise, the scale bar pertains to Ca²⁺ currents here and subsequently throughout, while Ba²⁺ currents are scaled downward by 2–3× to optimize visual comparison of kinetics. (C) Prolonged 1-s depolarization to near the peak of I-V relations (+10 mV) produces faster decay of Ca²⁺ versus Ba²⁺ current, indicative of CDI.

Figure 2.

Single P/Q-type channels exhibit CDF, but not CDI. (A) Unitary Ca²⁺ currents during modified voltage prepulse protocol. Top, test pulse depolarizations (+20 mV) without (left) and with (right) a voltage prepulse (30 mV). Middle, single-channel records from an exemplar patch containing one channel. Dashed lines indicate the closed (top, at zero level) and open (bottom) levels of unitary current. Gray rectangle signifies period during which a sparser pattern of gating predominates. Bottom, ensemble average currents after normalization to represent P _o (gray traces), as averaged from n = 5 patches. Smooth black curves are fitted by eye. Upper dashed line indicates zero-current level. Middle and lower dashed lines are analogous to L and H levels in Fig. 1 B. Traces show hallmarks of CDF, as present in macroscopic data (Fig. 2 B). Extreme bottom, exemplar whole-cell Ca²⁺ current evoked by +10-mV depolarization, illustrating clearly detectable CDI over the time period observed in single-channel recordings. Absence of such decay in ensemble average currents indicates lack of CDI at the single-channel level. (B) Unitary Ba²⁺ currents observed under an identical protocol; identical format as in A. Middle, exemplar single-channel records from a patch containing one P/Q-type channel. Calibration bar identical to that in A. Bottom, ensemble average currents normalized to P _o, averaged from n = 5 patches. Note the absence of a sparser pattern of gating during the corresponding period denoted by a gray rectangle.

Figure 3.

Potential elementary mechanisms underlying macroscopic CDF. (A) State diagram representation of mechanisms. For the enhanced-opening mechanism (left), the normal mode of gating (states within white box) features rapid activation with a lower P _o after first opening, whereas the facilitated mode (states within gray box) supports nearly identical activation with a higher P _o after first opening. Ca²⁺ binding to CaM drives the transition from normal to facilitated modes, thereby producing CDF. C represents a closed conformation, and O represents an open conformation. For the accelerated-activation mechanism (right), the normal mode exhibits a delayed time to first opening, but an appreciable P _o thereafter. The facilitated mode sports fast first opening, but the same P _o thereafter. (B) Cartoons of expected single-channel activity for the enhanced-opening (left) and accelerated-activation (right) mechanisms. Top, anticipated activity for test pulse depolarization without prepulse. Bottom, expected activity for test pulse with prepulse. Labels denote gating mode associated with indicated segments of activity; arrow marks transition between modes. Either scenario (left or right) would give rise to the well-studied macroscopic profile of CDF, schematized in the center of the panel. (C) Expected distinctions between the two mechanisms, when probed with macroscopic tail current analysis (e.g., Fig. 4). The enhanced-opening mechanism (left) predicts a hump-shaped relative P _o curve as determined from tail current measurements. This distinctive shape arises from the smaller plateau level of steady-state P _o curves for the normal (black dashed curve) versus facilitated modes of gating (gray curve), the weighted average of which produces the measured relative P _o curve (solid black). The relevant weighting factor is F_fac, the fraction of channels in the facilitated mode at the end of the test depolarization just before capturing tail currents. The accelerated-activation mechanism predicts a measured relative P _o curve (solid black) that lacks an overshoot hump. For short 20-ms depolarizations before tail current measurement, the corresponding P _o curves for normal (gray) and facilitated modes of gating (black dashed) are simply voltage shifted with respect to each other, such that their weighted average cannot produce an overshoot. (D and E) Expected single-channel outcomes for enhanced-opening (left) and accelerated-activation (right) mechanisms. FL is the probability that a first opening occurs before time t in the test pulse depolarization (D), whereas P _oo is the probability of finding a channel open with a delay t after a channel is known to have had a first opening (E). The enhanced-opening mechanism predicts no prepulse dependence of FL (D, left), but a characteristic prepulse dependence of P _oo (E, left). The accelerated-activation mechanism will show just the opposite profile (D and E, right) (see text for further details).

Figure 4.

Macroscopic tail current activation curves favor the enhanced-opening mechanism of CDF. (A) Exemplar tail current records. Top, voltage protocol, with parenthetical values pertaining to Ba²⁺ currents shown below. Middle, Ca²⁺ tail currents. Bottom, Ba²⁺ tail currents. 5 mM Ca²⁺ or Ba²⁺ as charge carrier. Outward currents were clipped for clarity. (B) Top, mean tail current activation curves for Ba²⁺ currents (open circles, averaged from n = 5 cells) and Ca²⁺ currents (filled circles, averaged from n = 5 cells). Tail currents have been normalized by amplitudes recorded after the most extreme depolarizations shown. Ba²⁺ voltages have been shifted 12 mV in the depolarizing direction, to account for the surface charge shift between Ca²⁺ and Ba²⁺ currents. Smooth curve fits by eye. Bottom, difference of smooth fits to Ca²⁺ and Ba²⁺ data above, furnishes a rough indication of presumed F _fac function (Fig. 3 C).

Figure 5.

In depth single-channel analysis supports a nearly exclusive enhanced-opening mechanism. For all panels, the left column pertains to mean statistical profiles for unitary Ca²⁺ currents through Ca_V2.1 channels (EF_a splice variant), whereas the right column concerns mean statistics for unitary Ba²⁺ currents through the same type of channels. (A) FL functions measured with (black curve) and without (gray curve) a preceding prepulse. There is no appreciable prepulse dependence. Ca²⁺ curves averaged from n = 5 patches; Ba²⁺ curves from a different n = 4 patches. (B) P _oo functions measured with (black curve) and without (gray curve) a prepulse. Smooth curves fit by eye. The Ca²⁺ relation obtained without a prepulse exhibits the characteristic “dip” expected for an enhanced opening mechanism. Ca²⁺ data averaged from n = 5 patches. Ba²⁺ data averaged from a different n = 5 patches. (C) Top, difference of mean P _oo relations in B (no prepulse P _oo–prepulse P _oo). Bottom, time integral of difference relations above (ms units, shown at 20× amplification). Top left, difference relation for Ca²⁺ currents fit by eye with single exponential function with a time constant of 20 ms.

Figure 6.

Unitary Ca²⁺ currents through the EF_b splice variant of P/Q-type channels (A–C). (A) Exemplar unitary Ca²⁺ current records and ensemble averages, using an identical protocol and display format as in Fig. 2 A. Exemplar records are from a patch with a single channel. Ensemble averages are derived from n = 5 patches. (B and C) In depth single-channel analysis of these EF_b splice variant channels, indicating channel trapping within the normal mode of gating. Format identical to that in Fig. 5 (A–C). All data averaged from n = 5 patches. (D and E) Open-time histogram analysis for various experimental configurations, shown as P{opening > t} versus t on a log–log plot. Black curves pertain to data obtained after a prepulse (pre), whereas gray curves concern data recorded without a prepulse (no pre). (D) Open-time histograms for unitary Ba²⁺ currents through EF_a channels, averaged from n = 5 patches. t _1/2 = 0.295 ms (no prepulse); t _1/2 = 0.270 ms (prepulse). (E) Open-time histograms for unitary Ca²⁺ currents through EF_b channels, averaged from n = 4. t _1/2 = 0.249 ms (no prepulse); t _1/2 = 0.252 ms (prepulse). (F) Open-time histograms for unitary Ca²⁺ currents through EF_a channels, averaged from n = 5 patches. t _1/2 = 0.410 ms (no prepulse); t _1/2 = 0.385 ms (prepulse). (G) Superposition of open-time histograms (with prepulse) reproduced from D–F. Unitary Ca²⁺ currents through EF_a channels support the longest open times.

Figure 7.

Resolving dual dimensions of APW-induced current augmentation, as predicted by the enhanced-opening mechanism of CDF. (A) Alignment of first (dashed black curve) and maximal Ba²⁺ current responses (solid black curve) from APW train experiment in Fig. 1 A. Upward scaling of the first response, to match the amplitude of the maximal response, yields a scaled waveform (gray curve) that superimposes the maximal response (black solid curve). (B) Alignment of first (dashed black curve) and maximal Ca²⁺ current responses (solid black curve) from APW train experiment in Fig. 1 A. Upward scaling of the first response yields a scaled waveform (gray curve) that decays earlier than the maximal response (black solid curve). Hence, CDF yields Ca²⁺ responses that are both larger and longer lasting, as predicted by the enhanced opening mechanism. The accelerated-activation mechanism would predict only an increase in amplitude, without change in longevity.

Figure 8.

Explicit kinetic simulations of the enhanced-opening mechanism furnish a global quantitative explanation for multiple single-channel datasets. (A) Simulations of unitary Ba²⁺ activity through channels bearing the EF_a splice variant. Far left, kinetic layout and rate-constant parameters for transitions within the normal mode of gating at a step potential of +20 mV (white rectangle). Numerical values are in units of ms⁻¹. Parameters and states shown in black are best determined, being doubly constrained by P _o and P _oo functions. Those displayed in gray impart subtle improvements to the overall fit, but are less well constrained. Mean experimental P _o (center left), P _oo (center right), and FL (far right) are reproduced from earlier data figures as gray traces (prepulse data only). The initial conditions of the P _o and FL simulations featured 20% of available channels in the leftmost closed state, and 80% of available channels in the immediately adjacent closed state. Superimposed upon these plots are the kinetic simulations of the enhanced-opening mechanism (black curves), with parameters on the far left. (B) Simulations of unitary Ca²⁺ activity through EF_b channels, using identical parameters as in A, except that the initial conditions of P _o and FL simulations featured 15% of available channels in the leftmost closed state, and 85% of available channels in the immediately adjacent closed state. Format as in A. (C) Simulations of unitary Ca²⁺ activity through EF_a channels, after a prepulse. Format as in A, except that these data help constrain the facilitated mode parameters (gray rectangle) in relative isolation. The main difference of the facilitated mode, as compared with the normal mode (A and B), concerns only the transitions between the final closed and open states (shown in bold). Subtle differences in the exit transitions from the two leftmost closed states afford modest improvements of the fit to FL data, while voltage inactivation is left unchanged. The initial conditions of P _o and FL simulations featured 15% of available channels in the leftmost closed state of the facilitated mode, and 85% of available channels in the immediately adjacent closed state. (D) Simulations of unitary Ca²⁺ activity through EF_a channels, without a prepulse. Format as in A. The facilitated mode parameters are maintained exactly as in C. The normal mode parameters are identical to those used above (A and B), except for minor differences in the exit rates for the two leftmost closed states. The latter differences yield small improvements to the fit of FL. For P _o, P _oo, and FL simulations, the initial fraction of available channels in the facilitated mode was 35%, with 65% in the normal mode. The initial fraction of channels in the facilitated mode probably has to do with the comparatively short 6-s repetition interval used in single-channel recordings, producing some cumulative CDF. The recovery time course from CDF is ∼0.5 s (DeMaria et al., 2001). Additionally, for P _o and FL simulations, the initial configuration was for 15% of channels in either mode to reside in the leftmost closed state, and for the remaining 85% of channels in either mode to reside in the immediately adjacent closed state. The transition from normal to facilitated modes occurs via a simple cooperative transition, as detailed in the Materials and methods.