Diffusion of Ca2+ from Small Boutons en Passant into the Axon Shapes AP-Evoked Ca2+ Transients

Abstract

Not only the amplitude but also the time course of a presynaptic Ca²⁺ transient determine multiple aspects of synaptic transmission. In small bouton-type synapses, the mechanisms underlying the Ca²⁺ decay kinetics have not been fully investigated. Here, factors that shape an action-potential-evoked Ca²⁺ transient were quantitatively studied in synaptic boutons of neocortical layer 5 pyramidal neurons. Ca²⁺ transients were measured with different concentrations of fluorescent Ca²⁺ indicators and analyzed based on a single-compartment model. We found a small endogenous Ca²⁺-binding ratio (7 ± 2) and a high activity of Ca²⁺ transporters (0.64 ± 0.03 ms^-1), both of which enable rapid clearance of Ca²⁺ from the boutons. However, contrary to predictions of the single-compartment model, the decay time course of the measured Ca²⁺ transients was biexponential and became prolonged during repetitive stimulation. Measurements of [Ca²⁺]_i along the adjoining axon, together with an experimentally constrained model, showed that the initial fast decay of the Ca²⁺ transients predominantly arose from the diffusion of Ca²⁺ from the boutons into the axon. Therefore, for small boutons en passant, factors like terminal volume, axon diameter, and the concentration of mobile Ca²⁺-binding molecules are critical determinants of Ca²⁺ dynamics and thus Ca²⁺-dependent processes, including short-term synaptic plasticity.

Figure 1

Presynaptic Ca²⁺ imaging. (A) A cell visualized using Alexa 568 is shown. Synaptic boutons (arrowheads) were identified along its axon collaterals. (B) A magnified view of a bouton (inside square in (A)) showing the baseline fluorescence of OGB-1 is given. (C) The top shows an AP evoked by a square current injection at the soma (2 nA, 2 ms). In the bottom right, the acquisition of line scans (dashed line in (B)) revealed an increase in OGB-1 fluorescence after the AP. The bottom left shows a Gaussian fit to the first line scan in the time series. Black bar indicates ±2 SD around the peak.

Figure 2

Analysis of Ca²⁺ transients based on the single-compartment model. (A) The average transients evoked by single APs, measured with OGB-1 or OGB-6F (n = 11 − 29), are shown. Shaded areas represent mean ± SD. Each decay phase was fitted with a single exponential function (dashed) and the sum of two exponential functions (solid). The insets show an expanded view of the initial decay. (B–D) A⁻¹, τ_slow, and the time integral $\left(\sum A_i{\tau }_i\right)$ are shown plotted against ${{\kappa }^{\prime }}_{\text{D}}$ or ${\left({{\kappa }^{\prime }}_{\text{D}}\right)}_{\text{slow}}$. Each data point represents the average value for a different concentration of OGB-1 or OGB-6F (filled or open circles). Plots were fitted with a weighted regression line. Dashed line indicates no correlation.

Figure 3

Blockade of SERCA prolonged the slow decay. (A) Ca²⁺ transients measured with OGB-6F (100 μM) before (black) and after addition of CPA (gray) are shown. Each trace is an average of five repeats, fitted with the sum of two exponential functions. (B–E) The peak amplitudes of Δ[Ca²⁺]_i, τ_fast, A_fast, and τ_slow before and after CPA addition are shown. n = 15 in (B) and 12 in (C–E).

Figure 4

The initial decay slowed down during repetitive stimulation. (A) The left shows ΔF/F₀ of fluo-4FF (500 μM) after an AP (black) and 5 APs evoked at 50 Hz (gray). Each trace is an average of two repeats. The right shows the transient evoked by a single AP, which was peak-scaled to match that evoked by the fifth AP in the train. Each decay time course was fitted with the sum of two exponential functions. (B–D) τ_fast, A_fast, and τ_slow after a single AP versus the fifth AP are shown. (E) ΔF/F₀ per AP during repetitive stimulation is shown plotted against the AP number. n = 13.

Figure 5

Diffusion of Ca²⁺ into the axon. (A) The top shows the baseline fluorescence of OGB-1 (80 μM) in a bouton. In the bottom, scanning OGB-1 fluorescence along the axon (arrows) revealed an increase in [Ca²⁺]_i after an AP (arrowhead). (B) Δ[Ca²⁺]_i at different locations from the bouton in (A) is shown. Each trace is an average of eight pixels and 10 repeats. The transient measured at 0 μm was fitted with two decaying exponential functions, whereas those along the axon were fitted with a rising and a decaying exponential function. (C) Similar to (B), but each trace is an average of Δ[Ca²⁺]_i from 11 boutons or the corresponding axonal segments (shaded areas indicate mean ± SD). (D and E) Time to peak or Δ[Ca²⁺]_i versus axonal distance. Filled circles indicate peak amplitude, open circles indicate the amplitude at 5 ms after the AP onset (A₅), and triangles indicate the difference between them. Plots were fitted with linear regression or an exponential function. Dashed line indicates no correlation.

Figure 6

Simulations of AP-evoked Ca²⁺ transients. (A) Dimensions of the modeled bouton are shown. Gray areas represent locations of Ca²⁺ influx. The top shows simulated Ca²⁺ current at the bouton. (B) The rate of decay versus time or Δ[Ca²⁺]_i (inset) is shown, with or without a 0.2-μm-thick axon, and 450 μM OGB-6F in both cases (solid or dashed trace, and filled or open circles, respectively). (C) The average Ca²⁺ transients measured with different concentrations of OGB-1 or OGB-6F (black) were superimposed with simulated transients (gray). (D and E) Analysis based on the single-compartment model is shown. Simulation results, obtained with an axon diameter of 0.2 (filled circles), 0.1 or 0.4 μm (dashed lines), were superimposed with experimental data (open circles with solid line indicating linear regression). (F) The relative contribution of diffusion to the initial rate of decay versus bouton diameter in the absence of exogenous buffers is shown. The axon diameter was 0.2 (filled) or 0.4 μm (open). Error bars were determined based on the uncertainty of the rate constants of Ca²⁺ removal mechanisms (Modeling). The inset shows simulated transients with a bouton diameter of 1 μm and no added buffers. The axon diameter was 0.2 (solid) or 0.4 μm (dashed).