Role of the flocculus in mediating vestibular nucleus neuron plasticity during vestibular compensation in the rat

Abstract

We investigated the role of the cerebellar flocculus in mediating the adaptive changes that occur in the intrinsic properties of brainstem medial vestibular nucleus (MVN) neurons during vestibular compensation. Ipsi-lesional, but not contra-lesional, flocculectomy prevented the compensatory increase in intrinsic excitability (CIE) that normally occurs in the de-afferented MVN neurons within 4 h after unilateral labyrinthectomy (UL). Flocculectomy did not, however, prevent the down-regulation of efficacy of GABA receptors that also occurs in these neurons after UL, indicating that these responses of the MVN neurons to deafferentation are discrete, parallel processes. CIE was also abolished by intra-floccular microinjection of the metabotropic glutamate receptor (mGluR) antagonist AIDA, and the protein kinase C inhibitor bisindolymaleimide I (BIS-I). The serene-threonine kinase inhibitor H-7 had no effect when microinjected at the time of de-afferentation, but abolished CIE if microinjected 2 h later. These cellular effects are in line with the recently reported retardatory effects of BIS-I and H-7 on behavioural recovery after UL. They demonstrate that the increase in intrinsic excitability in MVN neurons during vestibular compensation is cerebellum dependent, and requires mGluR activation and protein phosphorylation in cerebellar cortex. Furthermore, microinjection of the glucocorticoid receptor (GR) antagonist RU38486 into the ipsi-lesional flocculus also abolished CIE in MVN neurons. Thus an important site for glucocorticoids in facilitating vestibular compensation is within the cerebellar cortex. These observations ascribe functional significance to the high levels of GR and 11-beta-HSD Type 1 expression in cerebellum.

Figure 1. Ipsi-lesional flocculectomy abolishes the increase in intrinsic excitability of MVN neurons during compensation

A, photomicrograph of the brain from a representative left-flocculectomised rat (ventral view). Note that the left flocculus-paraflocculus lobe (FL) is absent. B, mean resting in vitro discharge rates (± s.e.m.) of rostral ipsi-lesional MVN neurons in slices that were prepared from animals that had undergone either 4 or 48 h of vestibular compensation after unilateral labyrinthectomy (UL_4h, UL_48h), in combination with either bilateral flocculectomy (bFX), contra-lesional flocculectomy (cFX) or ipsi-lesional flocculectomy (iFX). The number of cells recorded in each experimental condition is shown at the base of the columns. In this and following figures, asterisks indicate where the mean in vitro discharge rate of the MVN cells is significantly elevated compared to control (P < 0.05, Mann-Whitney rank sum test).

Figure 2. Ipsi-lesional flocculectomy does not prevent adaptive changes in the functional efficacy of GABA receptors in MVN neurons during compensation

A, firing rate histograms showing the dose-dependent inhibition of the spontaneous discharge of two MVN neurons by bath-application of muscimol (upper panel) and baclofen (lower panel). The bars above the data indicate the 60 s period of application of the agonists. The inhibitory response was measured as the maximal inhibition of firing rate expressed as a percentage of the resting discharge rate for each cell. B, dose-response relationships for the inhibitory response to muscimol (left panel) and baclofen (right panel), of MVN neurons in control slices (○) and in slices prepared from animals that underwent iFX+UL 4 h earlier (•). Data are means and s.e.m.

Figure 3. Inhibition of cerebellar plasticity in ipsi-lesional flocculus prevents the increase in intrinsic excitability of MVN neurons during compensation

Mean in vitro discharge rates (±s.e.m.) of rostral ipsi-lesional MVN neurons in slices prepared from animals that were unilaterally labyrinthectomised under urethane anaesthesia, and which remained anaesthetised for 4 h post-UL. The number of cells recorded in each experimental condition is shown at the base of the columns. A, in animals treated with dexamethasone i.p. at the time of UL (UL_4h+ Dex), the mean in vitro discharge rate of the MVN neurons was significantly elevated compared to that in control MVN slices, while in slices from animals that did not receive dexamethasone i.p. (UL_4h - Dex) the mean in vitro discharge rate is not different from control. B, effects of intrafloccular micro-injection of the metabotropic glutamate receptor antagonist AIDA, the PKC inhibitor BIS-I and the serene-threonine kinase inhibitor H-7 at the time of UL, on the mean in vitro discharge rates of MVN neurons in slices prepared 4 h post-UL. In each of these groups dexamethasone was given i.p. at the time of UL. Intrafloccular micro-injection of AIDA and BIS-I abolished the expected increase in intrinsic excitability of the MVN neurons (AIDA_F, BIS-I_F groups). H-7 had no effect when micro-injected at the time of UL (H-7_F group), but did abolish the increase in intrinsic excitability when micro-injected 2 h post-UL (H-7_F2h group). C, ffects of intra-floccular microinjection of dexamethasone (Dex_F group) and the GR antagonist RU38486 (RU_F group). The Dex_F group did not receive dexamethasone i.p.; instead dexamethasone was micro-injected into the ipsi-lesional floccular lobe at the time of UL. The RU_F group received dexamethasone i.p. at the time of UL, but RU38486 was micro-injected into the ipsi-lesional flocculus to block GR.