Cerebellar synapse properties and cerebellum-dependent motor and non-motor performance in Dp71-null mice

Abstract

Recent emphasis has been placed on the role that cerebellar dysfunctions could have in the genesis of cognitive deficits in Duchenne muscular dystrophy (DMD). However, relevant genotype-phenotype analyses are missing to define whether cerebellar defects underlie the severe cases of intellectual deficiency that have been associated with genetic loss of the smallest product of the dmd gene, the Dp71 dystrophin. To determine for the first time whether Dp71 loss could affect cerebellar physiology and functions, we have used patch-clamp electrophysiological recordings in acute cerebellar slices and a cerebellum-dependent behavioral test battery addressing cerebellum-dependent motor and non-motor functions in Dp71-null transgenic mice. We found that Dp71 deficiency selectively enhances excitatory transmission at glutamatergic synapses formed by climbing fibers (CFs) on Purkinje neurons, but not at those formed by parallel fibers. Altered basal neurotransmission at CFs was associated with impairments in synaptic plasticity and clustering of the scaffolding postsynaptic density protein PSD-95. At the behavioral level, Dp71-null mice showed some improvements in motor coordination and were unimpaired for muscle force, static and dynamic equilibrium, motivation in high-motor demand and synchronization learning. Dp71-null mice displayed altered strategies in goal-oriented navigation tasks, however, suggesting a deficit in the cerebellum-dependent processing of the procedural components of spatial learning, which could contribute to the visuospatial deficits identified in this model. In all, the observed deficits suggest that Dp71 loss alters cerebellar synapse function and cerebellum-dependent navigation strategies without being detrimental for motor functions.

Keywords: Cerebellum; Cognitive deficit; Dp71; Dystrophin; Glutamatergic transmission; Motor coordination; Purkinje neuron.

Fig. 1.

Glutamate transmission at PF-Purkinje neuron synapses in WT and Dp71-null mice. (A) Left: Representative superimposed EPSCs recorded in Purkinje neurons at increasing electrical stimulation of PFs (PF-EPSCs) in WT (filled circles) and Dp71-null mice (open circles). Right: Input-output relations for WT (n=7 cells, 5 mice) and Dp71-null mice (n=9 cells, 5 mice); P>0.45. (B) Left: PF-EPSCs evoked by two stimuli separated by 30 ms induced PPF in both WT (filled circles) and Dp71-null mice (open circles). Right: mean values of PPF calculated at different ISI (30, 50, 70, 90 and 200 ms) for WT (n=9 cells, 6 mice) and Dp71-null mice (n=8 cells, 5 mice; P>0.08). Holding potential was −70 mV with gabazine 5 µM in the extracellular solution.

Fig. 2.

Purkinje neurons from Dp71-null mice present larger CF-EPSCs. (A) Representative ‘all or none’ CF-EPSCs from a Purkinje neuron in WT (filled circles) and Dp71-null mice (open circles). The graph shows CF-EPSC mean amplitudes as a function of stimulus intensity for n=16 cells in WT (11 mice) and n=15 cells in Dp71-null mice (11 mice). (B) Distribution and mean peak values of CF-EPSCs from WT (n=16 cells) and Dp71-null mice (n=15 cells; P=0.01). (C) Left: CF-EPSCs are mainly mediated by activation of AMPA/Kaïnate receptors, as they are strongly inhibited by NBQX (10 µM). The remaining current is partially abolished by further addition of APV (50 µM) indicating a role of NMDA receptors. Right: Individual values of the NMDA receptor component of CF-EPSC in WT (n=7 cells, 5 mice) and Dp71-null mice (n=6 cells, 4 mice; P=0.029). (D) TBOA (100 µM) significantly increases the decay time of CF-EPSC in both WT (n=8 cells, 4 mice; P=0.01) and in Dp71-null (n=7 cells, 3 mice; P=0.02) mice and no statistical difference was observed in the two populations (P=0.71). (E) Top: CF-EPSCs evoked by two stimuli separated by 30 ms show a PPD in both WT and Dp71-null mice. Bottom: Time course of recovery from PPD is similar between WT and Dp71-null mice. Mean PPD values are from n=9 Purkinje cells in both WT and Dp71-null mice. ISIs were fixed at 30, 50, 70, 90, 120, 150, 180, 200 and 500 ms (P>0.43 for each interval). Holding potential was −20 mV with gabazine 5 µM in the extracellular solution.

Fig. 3.

Tetanic stimulation of CF-PC synapses reveals long-term plasticity alterations in Dp71-null mice. (A) Top: Representative complex spikes recorded in a WT (left) and Dp71-null (right) PC following CF stimulation at 0.033 Hz. Note the initial fast depolarizing component followed by smaller and slower depolarizing peaks (spikelets). Dashed lines indicate the parameters characterizing the complex spike: the depolarized plateau potential after the last spikelet (ΔV_plateau) and the time interval between the stimulus artifact and the peak of the first spikelet (latency). Single and averaged values from WT (n=8 cells, 3 mice) and Dp71-null (n=6 cells, 2 mice) mice are reported below. P>0.5 for all. (B) Top: CF-evoked complex spikes before (t=−5 min) and after CF stimulation at 5 Hz for 30 s (t=25 min) in WT and mutant mice. Bottom: Time course of the amplitude of the first spikelet in WT (n=7 cells, 3 mice) and Dp71-null mice (n=6 cells, 2 mice). The arrow shows the moment of CF tetanic stimulation. Each data point represents the average of two successive responses evoked at 0.033 Hz.

Fig. 4.

Excitatory synapse organization. (A) Sample confocal images taken in the cerebellar molecular layer in WT and Dp71-null mice. Immunoreactive clusters of PSD-95 can be seen in the proximal dendritic area of PCs (≤60 µm from neuron soma), as the majority of CF synapses occupy this territory on the PC dendritic tree. Scale bar: 20 µm. (B) Density of PSD-95 clusters in the Purkinje neuron proximal dendritic area, normalized to 10,000 µm² of tissue. (C) Cluster sizes were analyzed in a cumulative frequency curve. The distribution of PSD-95 clusters (0.05-2 µm²) presented a leftward shift in Dp71-null mice (arrow, P<0.005, n=3 mice per genotype). The lines show that 80% of clusters in Dp71-null mice were ≤0.220 µm², while 80% were ≤0.269 µm² in WT mice.

Fig. 5.

Muscle force, coordination and motivation. (A) Force expressed in arbitrary units (A.U.) averaged from three successive trials in the grip strength test (9 WT and 7 Dp71-null mice). (B,C) Performance during the three trials of the wire suspension test (18 WT, 15 Dp71-null mice) expressed as the reflex latency to touch the wire with one hind paw (B) and the best score (C) according to the motor performance scale (Aruga et al., 2004). (D) Latency to fall from the unstable platform in light (mean of three successive trials) and dark conditions (one trial, 24 h later) (n=10 per genotype). (E-G) Exploration in the hole board showing the number of stumbles (E), number of nose pokes in holes (F) and number of entries in the virtual central area of the apparatus (G). *P<0.01, one-way ANOVA. (H) Freezing latency in two successive training days (1, 2) in the tail suspension test (13 WT, 14 Dp71-null mice). (I) Freezing latency in two successive training days (1, 2) in the forced swim test. (J) Duration of freezing, climbing and floating in two successive training days (1, 2) in the forced swim test (n=14 per genotype).

Fig. 6.

Motor synchronization learning and navigation strategies. (A,B) Motor synchronization learning in the rotarod test. In all experiments, performance was quantified as the latency to fall from the rod. In a first protocol (A), mouse equilibrium was evaluated following placement of the mouse on the non-rotating rod (speed=0), and then on the rod rotating at a constant low speed (4 rpm) and during accelerating speed ramps (4-40 rpm) during 4 days (18 WT and 15 Dp71-null mice). In a second protocol (B,C), motor learning was first assessed during accelerating speed ramps during 3 days (B) and then in eight successive trials, each at a single constant speed as indicated on the x-axis (C) (n=10 per genotype). (D) Escape latency during 3-day training in the visible-platform protocol in a water maze (10 WT, 11 Dp71-null mice). (E) Percentage time spent in the Whishaw's corridor during the first training day in the visible-platform protocol. (F) Escape latency during 7-day training in the non-visible-platform protocol with fixed departure locations in a water maze (n=12 per genotype). (G) Distance swum in the target quadrant before finding the platform during day 2. Significant genotype effects: *P<0.05, **P<0.01, ***P<0.001. ^#P<0.05, significant genotype×day interaction.