Electrically evoked dendritic pH transients in rat cerebellar Purkinje cells

Abstract

Our aim was to test the hypothesis that depolarization-induced intracellular pH (pH(i)) shifts in restricted regions (dendrites) of mammalian neurones might be larger and faster than those previously reported from the cell soma. We used confocal imaging of the pH-sensitive dye, HPTS, to measure pH changes in both the soma and dendrites of whole-cell patch-clamped rat cerebellar Purkinje cells. In the absence of added CO(2)-HCO(3)(-), depolarization to +20 mV for 1 s caused large (approximately 0.14 pH units) and fast dendritic acid shifts, whilst the somatic acidifications were significantly smaller (approximately 0.06 pH units) and slower. The pH(i) shifts were smaller in the presence of 5 % CO(2)-25 mM HCO(3)(-)-buffered saline (approximately 0.08 pH units in the dendrites and approximately 0.03 pH units in the soma), although a clear spatiotemporal heterogeneity remained. Acetazolamide (50 microM) doubled the size of the dendritic acid shifts in the presence of CO(2)-HCO(3)(-), indicating carbonic anhydrase activity. Removal of extracellular calcium or addition of the calcium channel blocker lanthanum (0.5 mM) inhibited the depolarization-evoked acid shifts. We investigated more physiological pH(i) changes by evoking modest bursts of action potentials (approximately 10 s duration) in CO(2)-HCO(3)(-)-buffered saline. Such neuronal firing induced an acidification of approximately 0.11 pH units in the fine dendritic regions, but only approximately 0.03 pH units in the soma. There was considerable variation in the size of the pH(i) shifts between cells, with dendritic acid shifts as large as 0.2-0.3 pH units following a 10 s burst of action potentials in some Purkinje cells. We postulate that these large dendritic pH(i) changes (pH microdomains) might act as important signals in synaptic function.

Figure 1. Regional intracellular pH (pH_i) shifts in cerebellar Purkinje neurones evoked by 1 s depolarization

Ai, greyscale confocal image of 8-hydroxypyrene-1,3,6-trisulphonic acid (HPTS) fluorescence from a voltage-clamped Purkinje cell, and four regions of interest (ROIs) from which measurements were made. Aii, plots of membrane potential (E_m) and HPTS fluorescence shifts (F/F₀), from the ROIs, on 1 s depolarization to +20, +40 and 0 mV in Hepes-buffered saline. Bi, greyscale confocal image of HPTS fluorescence from another voltage-clamped Purkinje cell, and ROIs from which measurements were made. The saturated signal in the centre of the soma (white) was excluded. Bii, E_m and HPTS fluorescence shifts, on 1 s depolarization to +20 and +40 mV in the presence of 5 % CO₂–25 mm HCO₃⁻. Ci, plot of mean HPTS fluorescence shift in CO₂–HCO₃⁻ against the mean shift in the absence of added CO₂–HCO₃⁻ for the four cellular regions following 1 s depolarization to +20 mV. Cii, mean HPTS fluorescence shift in each region, normalized to that across the whole cell during a 1 s depolarization to +20 mV in the presence and absence of added CO₂–HCO₃⁻. Ciii, mean data of time to peak pH_i shift in each neuronal region, during a 1 s depolarization to +20 mV in the presence and absence of added CO₂–HCO₃⁻.

Figure 2. Calibration of F/F₀ HPTS transients in different cellular regions

A, graph showing the relationship between the measured pH of 12 calibration (140 mm KCl, 10 mm Hepes, 0.5 mm HPTS) solutions and the value calculated from the relative falls (10 ± 1.9 %; mean ±s.d.) in the single-wavelength HPTS fluorescence (straight line shows least-squares fit where; slope = 1.01 and regression coefficient = 1.00). The calculation was based on two assumptions; first, that the most alkaline solution had a pH of 8.00 and second, a pK of 7.10 for HPTS. B, the relationship between the calculated pH transient size and the starting pH_i for two sizes of fluorescence shifts. The blue lines (mean ± 1 s.d.) show the pH shift seen in the soma following a 1 s depolarization (5.2 ± 1.0 %), whilst the red lines (mean ± 1 s.d.) show the pH shift seen in the tertiary dendrites under the same conditions (12.5 ± 1.6 %). C, five curves showing the difference in the size of the depolarization-evoked pH shifts between the dendrites and the soma over a range of different assumed starting soma pH values. Each curve is plotted for a different dendritic starting pH with an F/F₀ shift of 5.2 % for the soma and 12.5 % for the dendritic data. The vertical dashed line shows a soma pH of 7.3 (the patch-pipette solution). At a dendritic pH of ≈6.5 (•), a soma pH of 7.5 is required to make the regional depolarization-evoked pH transients equal in size. The black curve shows the difference in size between the regional pH transients assuming no steady-state pH_i gradients throughout the cell.

Figure 3. The effects of carbonic anhydrase inhibition on depolarization-induced acid shifts in Purkinje cell dendrites

Ai, greyscale confocal image of a cerebellar Purkinje cell loaded with 500 μM HPTS via the patch pipette, and ROI (white box) from which fluorescence data are plotted. Aii, membrane potential, and F/F₀ HPTS data from the selected region during three 1 s depolarizations to +20 mV. Addition of 50 μM acetazolamide to the 5 % CO₂–25 mm HCO₃⁻-buffered saline reversibly potentiated the size of the acid shift. Bi, greyscale confocal image and ROI (white box) from a different Purkinje cell loaded with 500 μM HPTS, superfused with Hepes-buffered saline. Bii, addition of 50 μM acetazolamide to the Hepes-buffered saline did not potentiate the size of the acid shifts evoked by 1 s depolarizations to +20 mV.

Figure 4. The effect of inhibiting calcium entry on the size of depolarization-evoked acid shifts in Purkinje cell dendrites

Ai, greyscale confocal image of HPTS fluorescence from a voltage-clamped Purkinje cell showing the ROI (white circle). Aii, the effect of 1 s depolarization on HPTS F/F₀ signal before, during and after the removal of extracellular calcium in the presence of CO₂–HCO₃⁻. Aiii, time course of the declining F/F₀ shift induced by extracellular calcium removal in three separate cells. The grey lines indicate the period of calcium removal. The x-axis is broken to allow the plots to be aligned both during calcium washout and calcium readdition. Bi, confocal image of Purkinje cell with region indicated from which fluorescence was plotted. Bii, the irreversible effect of 0.5 mm lanthanum on the depolarization-induced F/F₀ fall. This was performed in Hepes-buffered saline as lanthanum precipitated in the presence of added HCO₃⁻. C, mean data showing the effect of calcium removal (filled bars, n = 3) and lanthanum addition (open bars, n = 6) on the size of depolarization-induced acid shifts (* P < 0.05, †P < 0.01 when compared to control: paired t test).

Figure 5. Regional acid shifts evoked by action potential (AP) firing

Ai, Confocal fluorescence image of a patch-clamped Purkinje cell and ROIs from which fluorescence has been plotted in Aii. Aii, an extract from an experiment showing the pH-sensitive fluorescence shift caused by a ≈10 s burst of APs, evoked by a 240 pA depolarizing current. The AP firing rate was consistent throughout the burst. Bi, confocal fluorescence image of a patch-clamped Purkinje neurone (soma region with saturating fluorescence) and dendritic ROIs from which fluorescence has been plotted in Bii. Bii, an extract from an experiment in which a 230 pA depolarizing current, for ≈ 10 s elicited a burst of APs showing adaptation. The fine dendritic regions show a biphasic pH_i shift.

Figure 6. Local differences in activity-induced acid shifts within fine dendritic regions of the Purkinje cell arborization

A, confocal image of a whole-cell patch-clamped Purkinje cell loaded with 500 μM HPTS (optical slice ≈4 μm thick), and image showing ROIs from which F/F₀ HPTS data were plotted. B, membrane potential (E_m) trace recorded at the cell soma. Depolarization was evoked by positive current injection (≈250 pA). C, F/F₀ data for the soma, primary, secondary and tertiary dendritic regions. D, F/F₀ data for two fine dendritic ROIs during the same period bursts of APs. The cerebellar slice was superfused with CO₂–HCO₃⁻-buffered saline containing 20 μM picrotoxin.