INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Adult vestibular plasticity can be achieved by different experimental approaches, most prominently by a peripheral lesion, e.g., unilateral labyrinthectomy or neurectomy. After a peripheral section of N. VIII the ganglion cells survive and electric stimulation of the sectioned nerve branch evokes monosynaptic field potentials in the vestibular nuclei similar to those recorded in control animals even two months after the lesion (Kunkel and Dieringer 1994). Postoperatively, several neural changes were documented with various methods in different species (Dieringer 1995; Smith and Curthoys 1989). However, detailed studies concerning the spatial or functional specificity of these rearrangements within the central vestibular system akin to the functional reorganization of somatosensory maps at cortical and subcortical levels (O'Leary et al. 1994) are so far absent. To investigate a possible expansion of signals from intact vestibular afferent fibers onto second-order vestibular neurons (2°VN) with silenced vestibular afferent inputs, we sectioned the frog's anterior branch of N. VIII on one side. Thereby, the afferent inputs from utricle, horizontal, and anterior vertical semicircular canals were eliminated, whereas the afferent inputs from saccule, lagena, and posterior vertical semicircular canal, and from the auditory organs, remained intact. Two months after this lesion we recorded from 2°VN and studied the convergence of monosynaptic inputs from different vestibular nerve branches. Convergence of afferent signals from different canals is rare in controls, since most (about 90%) of the 2°VN receive their monosynaptic vestibular excitation from only one of the three ipsilateral semicircular canal nerves in pigeon (Wilson and Felpel 1972), cat (Kasahara and Uchino 1974), and frog (Straka et al. 1997).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In deeply anesthetized grass frogs (Rana temporaria; 0.1% MS-222) the otic capsule was opened, and the ramus anterior (RA) of N. VIII was sectioned under visual control distal to the entry of the saccular nerve branch. Two months after this lesion chronic animals were reanesthetized and perfused transcardially with iced Ringer solution. The brain together with the attached N. VIII and their branches to the labyrinthine endorgans were removed and prepared for in vitro experimentation (see Straka et al. 1997). The posterior vertical semicircular canal (PC) and the RA nerve branches were electrically stimulated on either side with short pulses via suction electrodes. Extra- and intracellular records in the vestibular nuclei ipsilateral to the side of stimulation were obtained with glass pipettes (2 M sodium chloride: 1-3 M; or 2 M potassium acetate and 0.3 M potassium chloride: 90-120 M). The stimulus intensity was limited to a value five times above the threshold of the first postsynaptic negativity (N₁) in the evoked field potential (Precht et al. 1974). This threshold was determined at the beginning of each experiment, and absolute values (between 2 and 3.5 µA) were similar between different experiments. Field potentials were searched in depth tracts with bilaterally symmetrical coordinates throughout the entire rostro-caudal extent of the vestibular nuclear complex. Our observations were restricted to monosynaptic excitatory postsynaptic potentials (EPSPs) in 2°VN with membrane potentials of at least 40 mV and to crossed early responses in the abducens nerves. About equal numbers of neurons were recorded on a given side of the brain stem per animal. Single sweeps of the evoked responses were digitized, stored on a computer, averaged 15-20 times, and analyzed off-line. EPSPs from 2°VN were analyzed after electronic subtraction of the extracellular field potential recorded in the vicinity of the neuron. Abducens nerve responses evoked by contralateral PC nerve stimulation were analyzed by calculating the area covered by the positive response over the first 20 ms (in mV × ms). Statistical differences were calculated according to the Mann-Whitney U test (unpaired parameters) or to the Wilcoxan signed rank test (paired parameters).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Field potentials in the vestibular nuclei evoked by stimulation of the ipsilateral PC or RA nerve consisted of a presynaptic (N₀) and one (N₁) or more postsynaptic negativities (Fig. 1A) (Precht et al. 1974). The latencies of N₁ responses were shorter following RA nerve stimulation (2.64 ± 0.30 ms; mean ± SD, n = 27) than following PC nerve stimulation (3.41 ± 0.46 ms; n = 27) because of a shorter distance between the stimulation and the recording site (Straka et al. 1997). These latencies characterized monosynaptic EPSPs in intracellular records (shaded bars in Fig. 2). PC nerve evoked N₁ potentials on the operated side were significantly larger in amplitude (Fig. 1, A and B) than PC nerve evoked N₁ potentials recorded at corresponding sites on the intact side (Fig. 1, A and B). Following RA nerve stimulation, very similar amplitudes of N₁ potentials and depth profiles were recorded in controls and on either side of operated frogs.

View larger version (17K): [in this window] [in a new window]   Fig. 1. Posterior canal (PC) nerve evoked afferent field potentials in the vestibular nuclei. A: PC nerve evoked pre- (N0) and postsynaptic (N1) field potentials recorded 65 days after the transection of the ramus anterior of N. VIII. B: depth profiles of PC nerve evoked N1 potentials of controls (n = 6) and of operated frogs recorded on the intact (n = 6) or on the operated side (n = 6). Traces in A are the averages from 20-24 single sweeps and were recorded 0.7 mm caudal to the caudal end of the entry of N. VIII at a depth of 0.5 mm. The shaded vertical bar in A shows the average ± SDs of the N1 onset latencies.

View larger version (39K): [in this window] [in a new window]   Fig. 2. Convergence of afferent inputs from different vestibular endorgans in intact and operated frogs. A-C: monosynaptic response components were evoked in some 2nd-order (2°) vestibular neurons only after stimulation of the posterior canal (PC) nerve (A), in others only after stimulation of the ramus anterior (RA) of N. VIII (B) and in a 3rd group after stimulation of the PC as well as of the RA nerve (C). D: the percentages of these 2°PC neurons, of 2°RA neurons and of 2°RA + PC neurons differed in part significantly between controls and operated frogs as between the intact and the operated side of operated frogs (#P  0.0001; *P  0.01). Closed arrow heads in A-C for stimulus onset and shaded vertical bars for the average ± SDs of monosynaptic onset latencies following PC or RA nerve stimulation. Each trace in A-C represents an average of 30 single sweeps.

Intracellular records were collected from more than 700 identified 2°VN (Table 1). On the operated side of chronic frogs, significantly more 2°VN received monosynaptic EPSPs following PC nerve stimulation than on the intact side or in controls (Table 1). The number of 2°VN that responded with monosynaptic EPSPs after RA nerve stimulation was larger than that after PC nerve stimulation, but similar values were encountered on either side of operated frogs and in controls (Table 1). According to the convergence of afferent canal inputs on 2°VN, three subgroups of 2°VN were differentiated: monosynaptic EPSPs in 2°PC neurons were evoked by PC but not by RA nerve stimulation (Fig. 2A), monosynaptic EPSPs in 2°RA neurons were evoked by RA but not by PC nerve stimulation (Fig. 2B), and monosynaptic EPSPs in 2°RA + PC neurons were evoked by RA as well as by PC nerve stimulation (Fig. 2C). The percentage of 2°RA + PC neurons was significantly increased on the operated side (Fig. 2D) when compared with data from controls or from the intact side of operated frogs. Interestingly, the percentage of 2°RA neurons was significantly reduced on the same side (Fig. 2D). The decrease in the number of 2°RA neurons on the operated side (about 24%) corresponded to the increase in the number of 2°RA + PC neurons (about 21%) on the same side.

Long-lasting excitatory abducens nerve responses were recorded on either side with suction electrodes following stimulation of the contralateral PC nerve. The onset latency of responses on the intact side (5.88 ± 0.48 ms; n = 6) was significantly shorter (P  0.05) than on the operated side (6.67 ± 0.50 ms; n = 6) or in controls (7.53 ± 1.27 ms; n = 8). Abducens responses (area of the 1st 20 ms) recorded on the intact side were significantly (P  0.05) larger (1.40 ± 0.19 mV × ms; n = 6) than the corresponding response components recorded on the operated side (0.79 ± 0.21 mV × ms; n = 6) or in controls (0.62 ± 0.40 mV × ms; n = 8).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The significant increase in the amplitude of the PC nerve evoked N₁ field potentials on the operated side corresponds with the increased number of 2°VN that exhibited monosynaptic EPSPs following PC nerve stimulation on this side. Apparently, PC nerve afferent inputs expanded on the operated side onto some of those 2°VN that were vestibularly deprived after RA nerve section. Consistent with this interpretation is the fact that the percentage of true 2°PC neurons, i.e., of 2°VN with a monosynaptic vestibular afferent input exclusively from the PC nerve, had not changed in operated frogs. Rather, the number of 2°RA + PC neurons increased, and the number of 2°RA neurons decreased on the operated side by about the same amount. This shift in the relative size of these two subpopulations strongly supports the notion of an expansion of PC nerve afferent inputs onto vestibularly deprived ipsilateral 2°VN. A direct consequence of such an expansion may be seen in the increased responsiveness of the abducens nerve on the intact side after PC nerve stimulation on the operated side. More efficient excitatory inputs and an earlier recruitment of action potentials could explain shorter onset latencies as observed for abducens nerve responses on the intact side.

Functional reorganization of somatosensory maps after nerve injury includes an expansion of signals from intact peripheral afferents into the territory of deprived afferents and an activation of neurons that had not responded to this input before the lesion (O'Leary et al. 1994). As for the somatosensory system, axonal or dendritic sprouting of intact afferent fibers and the formation of new synaptic contacts or a postoperative increase in the efficacy of already existing but silent terminals are possible mechanisms for the expansion of vestibular signals on the operated side. Activity-dependent interaxonal competition between vestibular afferent fibers for functional synaptic contacts, as discussed for lesion-induced reactive synaptogenesis in the somatosensory system (O'Leary et al. 1994), is a possible trigger for these changes. Accordingly, the terminals of PC nerve afferent fibers were expected to have an activity-related competitive advantage compared with injured RA nerve afferent fibers with the result that functionally new synaptic contacts emerged. However, the expansion of PC nerve afferent signals on the operated side was neither paralleled by an obvious reduction in the number of 2°RA neurons nor by a decrease in the amplitude of RA nerve evoked field potentials on this side. Both results speak against a degeneration of the axotomized afferent fibers and for the survival of their synaptic contacts. A similar observation was made after unilateral labyrinthectomy (Kunkel and Dieringer 1994). The same competitive mechanism, assumed to be responsible for the expansion of vestibular afferent signals after RA nerve section, might also account for the expansion of ascending spinal projections and for the amplification of vestibular commissural signals in frogs after unilateral labyrinthectomy (see Dieringer 1995). In fact, preliminary data from this study indicate that the synaptic efficacy of commissural fibers terminating on the operated side was increased two months after a RA nerve section as well. The time course of this increase after RA nerve section is so far unknown but should be as delayed in its onset as it is after labyrinthectomy (Kunkel and Dieringer 1994), provided both increases share a common mechanism.