A quantitative measurement of the dependence of short-term synaptic enhancement on presynaptic residual calcium

We simultaneously measured presynaptic free calcium ion concentration ([Ca²⁺](i)) and synaptic strength at the crayfish claw opener neuromuscular junction (nmj) under a variety of experimental conditions. Our experiments were designed both to test the hypothesis that elevated [Ca²⁺](i) is necessary and sufficient for the induction of a form of synaptic enhancement that persists for several seconds after tetanic stimulation-augmentation-and to determine the quantitative relationship between elevated [Ca²⁺](i) and this enhancement. Action potential trains increased [Ca²⁺](i) and enhanced transmission. During the decay phase of synaptic enhancement known as augmentation (time constant of decay ≃7 sec at 20°C with <200 μM fura-2 in terminals), [Ca²⁺](i) was elevated 700 nM or less above rest and an essentially linear relationship between [Ca²⁺](i) and enhancement was observed. Introduction of exogenous Ca²⁺ buffers into the presynaptic terminal slowed the buildup and recovery kinetics of both [Ca²⁺](i) and the component of synaptic enhancement corresponding to augmentation. The slope of the relationship relating Δ[Ca²⁺](i) to augmentation was not changed. The time course of augmentation and recovery of [Ca²⁺](i) remained correlated as the temperature of the preparation was changed from about 10°C to 20°C, but the quantitative relationship of enhancement to [Ca²⁺](i) was increased more than two- to threefold. During moderate frequency trains of action potentials, a slowly developing component of the total synaptic enhancement was approximately linearly related to residual [Ca²⁺](i) measured with fura-2. The quantitative relationship between [Ca²⁺](i) and this component of synaptic enhancement during trains was the same as that during synaptic augmentation after trains. Bath application of the calcium ionophore A23187 increased [Ca²⁺](i) and resulted in synaptic enhancement. The quantitative relationship between increased [Ca²⁺](i) and synaptic enhancement was similar to that seen after augmentation-inducing action potential trains. Blocking Na⁺/K⁺ pump with ouabain also increased [Ca²⁺](i), possibly through reversal or slowing of Na⁺/Ca²⁺ exchange. With increased [Ca²⁺](i) of 300 nM or less above basal, the relationship between enhancement and [Ca²⁺](i) produced by ouabain application was comparable to that seen during synaptic augmentation after a train. Our results in conjunction with previous work (Delaney et al., 1989) support the hypothesis that elevated [Ca²⁺](i) of about 1 μM or less is a necessary and sufficient component for the production of augmentation and a longer-lasting form of activity-dependent synaptic enhancement, posttetanic potentiation (ptp). We conclude that the differing time constants of augmentation and ptp largely reflect multicomponent recovery kinetics of [Ca²⁺](i) that occur following the stimulus conditions that produce these two forms of synaptic enhancement. Our experiments also suggest that shorter-lasting facilitation of release, termed F1 and F2 facilitation, which decays within a second of the offset of stimulation, is not significantly activated by the increases in [Ca²⁺](i) that are sufficient to produce augmentation and ptp. Evidence is presented that at higher [Ca²⁺](i) other facilitatory processes (possibly corresponding to F1 or F2 facilitation) contribute to the enhancement of both spontaneous and action potential-evoked release.