Cerebellar Ataxia Caused by Type II Unipolar Brush Cell Dysfunction in the Asic5 Knockout Mouse

Abstract

Unipolar brush cells (UBCs) are excitatory granular layer interneurons in the vestibulocerebellum. Here we assessed motor coordination and balance to investigate if deletion of acid-sensing ion channel 5 (Asic5), which is richly expressed in type II UBCs, is sufficient to cause ataxia. The possible cellular mechanism underpinning ataxia in this global Asic5 knockout model was elaborated using brain slice electrophysiology. Asic5 deletion impaired motor performance and decreased intrinsic UBC excitability, reducing spontaneous action potential firing by slowing maximum depolarization rate. Reduced intrinsic excitability in UBCs was partially compensated by suppression of the magnitude and duration of delayed hyperpolarizing K⁺ currents triggered by glutamate. Glutamate typically stimulates burst firing subsequent to this hyperpolarization in normal type II UBCs. Burst firing frequency was elevated in knockout type II UBCs because it was initiated from a more depolarized potential compared to normal cells. Findings indicate that Asic5 is important for type II UBC activity and that loss of Asic5 contributes to impaired movement, likely, at least in part, due to altered temporal processing of vestibular input.

Figure 1

Mature Asic5 KO mice exhibit ataxic behavior. Summary graphs of mean fall speeds (A) and fall times (B) for trained mature wild type (black bars) and Asic5 KO (gray bars) mice as quantified using an accelerating rotarod test. Balance beam test: Summary graphs comparing the mean number of slips per step taken (C), steps taken per crossing (D), and total time taken to cross (E) for trained mature littermate control (black bars) and Asic5 KO (gray bars) mice during a full run on an elevated horizontal balance beam. Summary data for both experiments from 10 different male littermate (83.0 ± 3.0 days old, 27.1 ± 0.5 g) and 10 different male Asic5 KO (90.5 ± 0.5 days, 25.8 ± 0.6 g) mice where each mouse was tested once in triplicate. *Significantly different using a two tailed unpaired t-test.

Figure 2

Weanling Asic5 KO mice exhibit ataxic behavior. Summary graphs of mean fall speeds (A) and fall times (B) for trained wild type (black bars) and Asic5 KO (gray bars) mice as quantified using an accelerating rotarod test. Summary data from 12 different male littermate (24.9 ± 0.2 days old, 11.2 ± 0.6 g) and 15 different male Asic5 KO (24.8 ± 0.4 days old, 11.7 ± 0.4) mice where each mouse was tested once in triplicate. Balance beam test: Summary graphs comparing the mean number of slips per step taken (C), steps taken per crossing (D), and total time taken to cross (E) for trained littermate control (black bars) and Asic5 KO (gray bars) mice during a full run on an elevated horizontal balance beam. Summary data from 13 different male littermate (28.2 ± 1.2 days old, 14.3 ± 0.8 g) and 14 different male Asic5 KO (27.8 ± 1.8 days old, 14.3 ± 1.2 g) mice where each mouse was tested once in triplicate. *Significantly different using a two tailed unpaired t-test.

Figure 3

Deletion of Asic5 does not affect type II UBC density, size or capacitance. (A) Representative Z-projected confocal images of mGluR1α positive type II UBCs immuno-labeled with anti-mGluR1α (red) antibody in lobule X of cerebellar cryosections prepared from control (left) and Asic5 KO (right) mice. Summary graphs of mGluR1α- and GFP-positive type II UBC mean densities (B) and circumferences (C), respectively, in cerebellar sections from wild type (black bars) and Asic5 KO (gray bars) mice. Summary data for type II UBC density from 9 (3 mice, mean age P60) control and 15 (3 mice mean age P60) knockout cryosections, respectively. Summary data for type II UBC circumference from 53 littermate (5 different cryosectons from 2 different animals) and 58 knockout (7 different cryosections from 2 different animals) type II UBCs. (D) Summary graph of GFP-positive type II UBC capacitance in cerebellar sections from wild type (black bars; n = 9, 7 slices, 4 animals, P17 age) and Asic5 KO (gray bars; n = 9, 8 slices, 4 animals, P16.8 age) mice. No significant difference was found between wild type and Asic5 KO for any of these measurements using a two tailed unpaired t-test.

Figure 4

Asic5 influences the inherent excitability of type II UBCs. (A) Representative traces of typical action potential trains in current-clamped type II UBCs in vestibulocerebellar slices from wild type (top) and Asic5 KO (bottom) mice evoked by 200 msec 40 pA supra-threshold current injections. Summary graphs of mean resting membrane potentials (B) and membrane resistances (C) in current-clamped type II UBCs from wild type (black bars, n = 32, 15–20 slices, 5 animals, mean age P19) and Asic5 KO (gray bars, n = 28, 12–15 slices, 4 animals, mean age P18) mice. No value is significantly different. Summary graphs of mean firing frequencies (D) and interspike intervals (E) for action potentials evoked by current injections in current-clamped type II UBCs in vestibulocerebellar slices from wild type (black circles, n = 32, 5 animals) and Asic5 KO (gray boxes, n = 28, 4 animals) mice. Inset in E shows an enlarged area for significantly different values for interspike intervals evoked by current injections greater than 65 pA. Significantly different at *P < 0.05 and **P < 0.01 with t-test.

Figure 5

Deletion of Asic5 slows the maximum depolarization rate of type II UBCs. Summary graphs of maximum depolarization (A) and repolarization (B) rates of first action potentials evoked by current injections in current-clamped wild type (n = 32, 15–20 slices, 5 animals, mean age P19) and Asic5 KO (n = 28, 12–15 slices, 4 animals, mean age P18) type II UBCs. *Significantly different using a t-test. (C) Overlay of representative first action potentials from current-clamped type II UBCs injected with 70 pA from wild type (black) and Asic5 KO (gray) mice. Summary graphs of mean thresholds (D), amplitudes (E), and half-widths (F) for first action potentials in type II UBCs evoked by 10 pA supra-threshold current injections in wild type (n = 13, 5 animals) and Asic5 KO (n = 8, 3 animals) mice. No value is significantly different.

Figure 6

The magnitude and presence of hyperpolarizing, slow outward K⁺ currents in type II UBCs treated with glutamate are decreased in the Asic5 KO mice. (A) Shown here is mean current as a function of time evoked by a train of 10 mM glutamate puffs from a compilation of raw current traces from a representative voltage-clamped type II UBC in a typical vestibulocerebellum brain slice from a wild type mouse. Currents evoked by application of glutamate were biphasic in type II UBCs with a fast depolarizing inward current seen initially followed by a slower developing hyperpolarizing outward current. Summary graphs of glutamate-evoked inward (B) and outward (C) current amplitudes, and outward current durations (D) in voltage-clamped type II UBCs from wild type (n = 32, 15–25 slices, 5 animals, mean age P20) and Asic5 KO (n = 55, 21–35 slices, 7 animals, mean age P19) mice. Summary data from experiments similar to that in A. *Significantly different compared to wild type with a t-test. (E) Percentage of type II UBCs from wild type (n = 32, 5 animals) and Asic5 KO (n = 55, 7 animals) mice having no outward currents in response to treatment with 10 mM glutamate. Data from experiments are similar to A. *Significantly different compared to wild type with a χ²-test.

Figure 7

Deletion of Asic5 reduces spontaneous firing but elevates glutamate-evoked burst firing frequency in type II UBCs. Representative gap-free traces of spontaneous and evoked action potential trains in typical current-clamped type II UBCs before and after treatment with 10 mM glutamate (arrow) in vestibulocerebellar slices from wild type (A; n = 9, 6 slices, 4 animals, mean age P16.5) and Asic5 KO (B; n = 10, 6 slices, 3 animals, mean age P16)) mice. The period during initial exposure to glutamate is shown at an expanded scale below. Representative instantaneous firing frequency histograms over the duration of such experiments from the same cells as in A and B (C,D, respectively). The point at which glutamate was added is shown with an arrow. Summary graphs of membrane potential (E) and duration (F) of hyperpolarization following treatment with glutamate in wild type (black bars) and Asic5 KO (gray bars) type II UBCs. Data from experiments identical to that shown in A and B. Significantly different at *P < 0.05 with t-test. Summary graphs of mean firing frequency for spontaneous (G) and burst (H) action potential firing, and then spontaneous firing again (I) before and after exposure to glutamate for wild type (black bars) and Asic5 KO (gray bars) type II UBCs. Summarized data from experiments identical to those shown in (A,B). Spontaneous firing prior to exposure to glutamate was quantified during the 0–11 second window, burst firing after exposure to glutamate was quantified during the 12–20 second window, and the return to spontaneous firing quantified over the 60–80 second window. Significantly different at *P < 0.05 with t-test.