Can the Ca2+ hypothesis and the Ca2+-voltage hypothesis for neurotransmitter release be reconciled?

Abstract

It is well established that Ca2+ plays a key role in promoting the physiological depolarization-induced release (DIR) of neurotransmitters from nerve terminals (Ca2+ hypothesis). Yet, evidence has accumulated for the Ca2+-voltage hypothesis, which states that not only is Ca2+ required, but membrane potential as such also plays a pivotal role in promoting DIR. An essential aspect of the Ca2+-voltage hypothesis is that it is depolarization that is responsible for the initiation of release. This assertion seems to be contradicted by recent experiments wherein release was triggered by high concentrations of intracellular Ca2+ in the absence of depolarization [calcium-induced release (CIR)]. Here we show that there is no contradiction between CIR and the Ca2+-voltage hypothesis. Rather, CIR can be looked at as a manifestation of spontaneous release under conditions of high intracellular Ca2+ concentration. Spontaneous release in turn is governed by a subset of the molecular scheme for DIR, under conditions of no depolarization. Prevailing estimates for the intracellular calcium concentration, [Ca2+]i, in physiological DIR rely on experiments under conditions of CIR. Our theory suggests that these estimates are too high, because depolarization is absent in these experiments and [Ca2+]i is held at high levels for an extended period.

Fig 1.

Experimental results obtained for CIR and DIR from refs. and with permission. (A–D) CIR in the calyx of Held from rats. (A) Dependence of the rate of release on [Ca²⁺]_i (19). [Reproduced with permission from ref. (Copyright 2000, AAAS, www.sciencemag.org).] (B) Time course of release at three levels of [Ca²⁺]_i (20). [Reproduced with permission from ref. (Copyright 2000, MacMillan Magazines, Ltd., www.nature.com).] (C) Dependence of the minimal delay on [Ca²⁺]_i (19). [Reproduced with permission from ref. (Copyright 2000, AAAS, www.sciencemag.org).] (D) Dependence of the time to peak release on [Ca²⁺]_i (20). [Reproduced with permission from ref. (Copyright 2000, MacMillan Magazines, Ltd., www.nature.com).] (E–H) DIR. (E) Dependence of the quantal content, expressed as E.p.p. (end plate potential), on extracellular Ca²⁺ concentration, [Ca²⁺]₀ (21). [Reproduced with permission from ref. (Copyright 1967, The Physiological Society).] (F) Time course of release at two levels of [Ca²⁺]₀ (normalized time course is seen on the right) (15). [Reproduced with permission from ref. (Copyright 1980, The Physiological Society).] (G and H) Schematic drawings of the dependence on [Ca²⁺]_i of the minimal delay and time-to-peak release in DIR; see text.

Fig 2.

Simplified kinetic scheme for DIR (12). R_H denotes the high-affinity- and R_L the low-affinity receptor. The state of association between the receptor bound with two molecules of agonist, R^••, and the exocytotic machinery, Ex, is denoted by R ∼ Ex. Depolarization and repolarization are denoted by dep and rep.

Fig 3.

Outline of the processes that mediate the three types of release. (A) DIR. (Left) Molecular scheme of Fig. 2. (Center) Membrane-potential-dependent Ex kinetics. (Right) Kinetics of [Ca²⁺]_i-depolarization-mediated influx and membrane-potential-independent removal. (B) Spontaneous release. (Left) Part of the scheme of Fig. 2 that remains under resting potential. A question mark symbolizes spontaneous dissociation of Ex from REx. (Center and Right) Low steady-state values of Ex and [Ca²⁺]_i. The low values are symbolized by small letters in calcium-mediated processes for forming Ex. (C) CIR. (Left) As in B but with a higher rate of forming Ex (symbolized by a larger dotted-dashed arrow), owing to a large [Ca²⁺]_i (symbolized by large letters). (Center) Ca²⁺-dependent kinetics of free Ex. (Right) Fixed high level of [Ca²⁺]_i. Note that the calcium-mediated processes are common to all three modes of release. For details, see text.

Fig 4.

Simulation of CIR (see supporting information) at various levels of [Ca²⁺]_i. The maximum release rate increases with [Ca²⁺]_i whereas Ex and R ∼ Ex decrease with [Ca²⁺]_i. (A) Time course of release at various levels of [Ca²⁺]_i (Inset). (B–D) Time courses of intermediate states that govern A: Ve(B), Ex(C), and Q(D).

Fig 5.

Simulations of DIR with the same model as for CIR. The thick bars in the time axis represent the 2-ms duration of a high depolarizing pulse, in the range of an action potential. (A) Time course of [Ca²⁺]_i obtained by solving the equations presented in ref. 23). See supporting information. (B) Time course of release that corresponds to the [Ca²⁺]_i profiles seen in A. Higher [Ca²⁺]_i corresponds to higher release rate. (Inset) Each graph is normalized to its own peak.

Fig 6.

Levels of [Ca²⁺]_i required to obtain the same peak release in simulations of CIR and DIR. (A) Solid line shows CIR, at high [Ca²⁺]_i; the level of [Ca²⁺]_i is fixed throughout the simulation (Inset). Dashed line shows DIR, of comparable magnitude, at low [Ca²⁺]_i. (B) Descriptions are the same as in A, except that [Ca²⁺]_i is elevated only briefly.

Fig 7.

Dependence of peak release rate on long pulses of [Ca²⁺]_i for CIR and DIR. To obtain K_d, simulation results were fit to the Hill equation [Ca²⁺]/K + [Ca²⁺]. For DIR, similar results are obtained when [Ca²⁺]_i is maintained at a constant level for 2 ms or for tens of milliseconds.