How the 'slow' Ca(2+) buffer parvalbumin affects transmitter release in nanodomain-coupling regimes

Abstract

Parvalbumin is thought to act in a manner similar to EGTA, but how a slow Ca(2+) buffer affects nanodomain-coupling regimes at GABAergic synapses is unclear. Direct measurements of parvalbumin concentration and paired recordings in rodent hippocampus and cerebellum revealed that parvalbumin affects synaptic dynamics only when expressed at high levels. Modeling suggests that, in high concentrations, parvalbumin may exert BAPTA-like effects, modulating nanodomain coupling via competition with local saturation of endogenous fixed buffers.

Figure 1

Cell-specific differences in somatic parvalbumin concentration correlate with synapse-specific effects on synaptic dynamics. (a) Confocal stack maximum projection of the dentate gyrus and the CA3 region, showing a granule cell filled with biocytin and 30 μM recombinant parvalbumin and several surrounding cells. (b) Expanded views, with the right and bottom panels corresponding to the rectangle in a. The left two panels show biocytin labeling, the top right panel shows parvalbumin immunoreactivity and the bottom right panel is an overlay. Scale bars represent 100 μm (top left) and 25 μm (bottom left and both right panels). (c) Analysis of parvalbumin concentration in cerebellar basket cells. The image is a confocal stack maximum projection of the molecular layer, Purkinje cell layer and granule cell layer of the cerebellum, showing a cell filled with biocytin and 30 μM recombinant parvalbumin and several surrounding cells. (d) Expanded view corresponding to the rectangle in c. Top, biocytin labeling. Middle, parvalbumin immunoreactivity. Bottom, overlay. (e,f) Histograms of the somatic parvalbumin (PV) concentration in 64 hippocampal basket cells (e) and 53 cerebellar basket cells (f), each in a representative slice. Insets, mean parvalbumin concentration and coefficient of variation of 6 and 5 slices, respectively. Error bars indicate s.e.m. (g,h) Short-term dynamics during a 50-Hz train of 10 action potentials in paired recordings between basket cells and synaptically connected Purkinje cells in cerebellar slices from Pvalb^+/+ (g) and Pvalb^−/− mice (h). Red traces, presynaptic action potentials evoked by brief current pulses; black traces, average IPSCs. (i) Amplitude of the i^th IPSC in the train measured from the preceding baseline and divided by the amplitude of the first IPSC. Data are from 9 (Pvalb^+/+, black) and 8 (Pvalb^−/−, red) pairs, respectively. (j) Relative contribution of synchronous release (Syn), asynchronous release during the train (Asyn 1), and asynchronous release after the train (Asyn 2) in Pvalb^+/+ (black) and Pvalb^−/− synapses (red). 50-Hz trains of ten presynaptic action potentials were used in all experiments. Error bars in i and j indicate s.e.m. (k,l) Rescue of Pvalb deletion by acute presynaptic infusion of 500 μM recombinant parvalbumin (rPV). Shown is a recording from a cerebellar basket cell–Purkinje cell synapse from a Pvalb^−/− mouse 0–6 min after break in (k) and 35–41 min after break in (l).

Figure 2

Parvalbumin may have fast effects and regulate synaptic dynamics by shunting of local buffer saturation. (a,b) Ca²⁺ transients following a single action potential 20 nm (a) and 200 nm (b), respectively, from the point source. Inset in a shows expansion of the peak region. Concentrations: BAPTA (100 μM), EGTA (100 μM) and parvalbumin (50 μM protein concentration, corresponding to 100 μM buffer concentration, as parvalbumin has two Ca²⁺-binding sites). Illustrations on top schematically depict the binding scheme of parvalbumin and the binding schemes for BAPTA and EGTA (buffer). (c–e) Parvalbumin has specific regeneration properties. (f,g) Spatiotemporal profile of concentration of free fixed buffer (FB, f) and parvalbumin (g) in a mixed fixed buffer and parvalbumin scenario following a single action potential (50 μM parvalbumin protein concentration; 100 μM fixed buffer). (h) Facilitation of Ca²⁺ transient at 20, 50 and 200 nm from the Ca²⁺ channels. Two action potentials were simulated at 20-ms intervals. Ca²⁺ transients are shown normalized to the peak of the first Ca²⁺ transient and horizontal lines indicate facilitation. Bottom right, expansion of the peak region of the second Ca²⁺ transient at the three distances. (i) Multiple-pulse facilitation (MPF) of the Ca²⁺ transient 50 nm from the Ca²⁺ channels, plotted against action potential number. (j) MPF of the Ca²⁺ transient for the tenth action potential plotted against source-sensor distance. Insets in i and j show the corresponding MPF of transmitter release, predicted from a power relation. Longer line segments indicate higher parvalbumin concentrations.