JN Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

J Neurophysiol 77: 3003-3012, 1997;
0022-3077/97 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Uchino, Y.
Right arrow Articles by Graf, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Uchino, Y.
Right arrow Articles by Graf, W.

The Journal of Neurophysiology Vol. 77 No. 6 June 1997, pp. 3003-3012
Copyright ©1997 by the American Physiological Society

Sacculocollic Reflex Arcs in Cats

Y. Uchino1, H. Sato1, M. Sasaki1, M. Imagawa1, H. Ikegami2, N. Isu3, and W. Graf4

1 Department of Physiology and 2 Department of Orthopedic Surgery, Tokyo Medical College, Tokyo 160; 3 Department of Information and Knowledge Engineering, Faculty of Engineering, Tottori University, Tottori 680, Japan; and 4 Laboratoire de Physiologie de la Perception et de l'Action, Centre National de la Recherche Scientifique, Collège de France, Paris Cedex 05, France

    ABSTRACT
Abstract
Introduction
Methods
Results
Discussion
References

Uchino, Y., H. Sato, M. Sasaki, M. Imagawa, H. Ikegami, N. Isu, and W. Graf. Sacculocollic reflex arcs in cats. J. Neurophysiol. 77: 3003-3012, 1997. Neuronal connections and pathways underlying sacculocollic reflexes were studied by intracellular recordings from neck extensor and flexor motoneurons in decerebrate cat. Bipolar electrodes were placed within the left saccular nerve, whereas other branches of the vestibular nerve were removed in the inner ear. To prevent spread of stimulus current to other branches of the vestibular nerve, the saccular nerve and the electrodes were covered with warm semisolid paraffin-Vaseline mixture. Saccular nerve stimulation evoked disynaptic (1.8-3.0 ms) excitatory postsynaptic potentials (EPSPs) in ipsilateral neck extensor motoneurons and di- or trisynaptic (1.8-4.0 ms) EPSPs in contralateral neck extensor motoneurons, and di- and trisynaptic (1.7-3.6 ms) inhibitory postsynaptic potentials (IPSPs) in ipsilateral neck flexor motoneurons and trisynaptic (2.7-4.0 ms) IPSPs in contralateral neck flexor motoneurons. Ipsilateral inputs were about twice as strong as contralateral ones to both extensor and flexor motoneurons. To determine the pathways mediating this connectivity, the lateral part of the spinal cord containing the ipsilateral lateral vestibulospinal tract (i-LVST) or the central part of the spinal cord containing the medial vestibulospinal tracts (MVSTs) and possibly reticulospinal fibers (RSTs) were transected at the caudal end of the C1 segment. Subsequent renewed intracellular recordings following sacculus nerve stimulation indicated that the pathway from the saccular nerve to the ipsilateral neck extensor motoneurons projects though the i-LVST, whereas the pathways to the contralateral neck extensors and to the bilateral neck flexor motoneurons descend in the MVSTs/RSTs. Our data show that sacculo-neck reflex connections display a qualitatively bilaterally symmetrical innervation pattern with excitatory connections to both neck extensor motoneuron pools, and inhibitory connections to both neck flexor motoneuron pools. This bilateral organization contrasts with the unilateral innervation scheme of the utriculus system. These results suggest a different symmetry plane along which sacculus postural reflexes are organized, thus supplementing the reference planes of the utriculus system and allowing the gravistatic system to represent all three translational spatial degrees of freedom. We furthermore suggest that the sacculocollic reflex plays an important role in maintaining the relative position of the head and the body against the vertical linear acceleration of gravity.

    INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References

Vestibulocollic reflexes originate from the semicircular canals and from the two otolithic gravistatic receptors, i.e., the utriculus and the sacculus. The adequate stimuli for the otolith systems are horizontal and vertical linear accelerations, respectively, that induce shear forces acting on the utricular and saccular sensory epithelia. The resulting neuronal activity is transmitted by vestibular afferents to second-order vestibular neurons in the vestibular nuclei, and from there to spinal motoneurons to elicit particular vestibulocollic reflexes, among other reflex behaviors. Head and body posture in physical space are thus maintained and controlled by activating the appropriate neck and trunk muscles (Fernández and Goldberg 1976a; Graf et al. 1992; Schor and Miller 1981; Wilson and Melvill-Jones 1979).

These three-neuron arcs of the vestibulocollic reflexes provide the fastest links between the vestibular receptor cells and neck muscles. There is now considerable detailed information available about particular receptor-effector relationships. The classical semicircular canal-neck muscle relationships (Wilson and Maeda 1974) have been further elaborated and specified (Shinoda et al. 1994; Sugiuchi et al. 1992a, 1995), and utriculo-neck connections have been examined with a newly developed technique allowing selective utricular nerve stimulation (Sasaki et al. 1991). In the present study we consider otolith reflexes acting on the head-neck system, with focus on sacculus projections.

Utricular stimulation induces predominantly disynaptic excitatory postsynaptic potentials (EPSPs) in ipsilateral extensor and flexor neck motoneurons, whereas inhibitory postsynaptic potentials (IPSPs) are evoked in contralateral neck extensor and flexor motoneurons (Bolton et al. 1992; Ikegami et al. 1994). These utriculo-neck links play a role in maintaining an upright head-neck posture on the trunk during horizontal linear accelerations (e.g., fore-aft, left-right). During left-right movement, for instance, neck extensors and flexors can cooperate to tilt the head-neck ensemble sideways (Baker et al. 1985; Ikegami et al. 1994).

Historically, the sacculus system has received little attention compared with the wealth of information available about the semicircular canal and the utriculus systems, but it is all of equal importance. Vertical linear acceleration detection, which activates neck muscles, is presumably due to stimulation of sacculus receptors (Borel and Lacour 1992; Lacour et al. 1987; Watt 1976; Xerri et al. 1987). These latter observations in cats were obtained by analyzing behavioral and electromyographic data, and suggested the existence of as yet undescribed neuronal connections between the sacculus and neck motoneurons. Because little information is available about the neuronal organization conveying saccular information to neck motoneurons (see, e.g., Wilson et al. 1977), we examined the sacculocollic relationships and the pathways mediating them with a newly developed technique of electrical stimulation of the saccular nerve that incurs minimum current spread to the other vestibular nerves. The present study was also intended to elucidate the functional role of the sacculus for spatial coordination of postural control mechanisms.

Preliminary results have been reported previously (Sato et al. 1994; Uchino et al. 1994b; Wilson et al. 1995).

    METHODS
Abstract
Introduction
Methods
Results
Discussion
References

Successful experiments were performed in 19 adult cats. All animal care and experimental procedures conformed with guidelines stipulated by the Physiological Society of Japan (Guiding Principles for the Care and Use of Animals in the Field of Physiological Sciences) and the Council of the American Physiological Society.

Animals were initially anesthetized with ketamine hydrochloride (Ketalar, Parke-Davis; 15-20 mg/kg im), followed by halothane-nitrous oxide inhalation after a tracheotomy had been performed. At the last stage of surgery, cats were decerebrated at the intercollicular level and prepared for recording without anesthesia after the typical signs of decerebration had been observed. The animals were subsequently paralyzed by intravenous administration of pancuronium bromide (Mioblock, Organon; 0.25-0.5 mg·kg-1·h-1) and ventilated artificially. Mean arterial pressure was monitored routinely from the femoral artery. When necessary, a 5-10% glucose solution was administered by intravenous infusion to maintain systolic arterial pressure above 100 mmHg. Body temperature was kept at 37.5°C.

Stimulating electrodes were prepared in advance for selective saccular nerve and dorsal root stimulation. Saccular nerve stimulation was effected with a bipolar tungsten electrode (insulated except for ~1 mm at the tip; interelectrode distance ~0.8 mm). For dorsal root and selective neck muscle nerve stimulation, bipolar silver electrodes and cuff electrodes were used.

The left inner ear was opened via a ventrolateral approach to expose the utricular nerve and the ampullary nerves of the anterior, horizontal, and posterior semicircular canals (Suzuki et al. 1969). The nerves were resected with a cutting tool specially prepared from a razor blade and removed (Fig. 1A). Cut ends were covered by bone wax. The sacculus and its nerve were then visualized by carefully scraping away the overlaying bone. The bipolar stimulating electrode was inserted into the saccular nerve after the inner ear had been drained of fluid with the use of small cotton twigs. The electrode was fixed in place and anchored to the occipital bone with dental cement. To avoid drying out of the saccular nerve and to reduce current spread, the area was bathed in warm semisolid paraffin-Vaseline. Cathodal current pulses 200 µs in duration were applied to the saccular nerve at a rate of 2-2.5 Hz.


View larger version (58K):
[in this window]
[in a new window]
  		FIG. 1. A: illustration of selective stimulation of saccular (S) nerve from a ventrolateral aspect of left inner ear. Nerves of anterior (A), horizontal (H), and posterior (P) semicircular canals and of utriculus (U) were removed; saccular nerve was preserved. Semicircular canal ampullary structures and saccular nerve are outlined by dashed lines. Bipolar electrodes are placed on saccular nerve. B and C: topographic relationships of lateral and medial vestibulospinal tracts (LVST and MVST, respectively; solid dark areas) within spinal cord at C1/C2 level and extent of spinal cord transections. MVSTs form a midline structure ventral to central canal. LVSTs are found bilaterally near ventral surface of spinal cord. In addition, spinal cord region containing ipsilaterally and contralaterally descending reticulospinal fibers (RSTs) is outlined (horizontal hatching). Ba and Bb: extent of 4 lesions involving ipsilateral (i-) LVST (diagonally hatched areas). Ca and Cb: extent of 4 lesions involving MVST and/or reticulospinal tracts (diagonally hatched areas).

In the neck region, laminectomies were performed on cervical vertebrae 1-5. The bilateral C2 and C3 dorsal rami (DR) and the C2 ventral rami innervating the longus capitis (LC) muscles were dissected. The C2 and C3 DR nerve roots innervate mainly the biventer cervicis and complexus muscles (neck extensors). Bipolar silver electrodes were placed on the cut ends of the C2 and C3 DR of both sides. The bilateral C2 nerve branches innervating the LC muscles (a neck flexor) were prepared for stimulation with the use of cuff electrodes. The nerves were stimulated during the experiment with cathodal current pulses 150 µs in duration for antidromic identification of extensor and flexor motoneurons.

Access to the vestibular complex was gained by resecting the portion of the occipital bone that overlays the fourth ventricle, and by removing part of the cerebellum. Extracellular orthodromic field potentials following saccular nerve stimulation were recorded with glass micropipettes (2 M NaCl saturated with Fast Green FCF; resistance 0.8-2.0 MOmega ) in the left vestibular nuclei. Recorded potentials were averaged 15-20 times.

Glass micropipettes containing 2 M potassium citrate (resistance 4-8 MOmega ) were used for intracellular recordings from neck motoneurons in the already exposed C2 and C3 segments. Depolarizing or hyperpolarizing currents were passed through the recording electrode into some neurons to verify the nature of a recorded potential, i.e., whether it was an EPSP or an IPSP. Only data from motoneurons recorded with resting membrane potentials of at least -35 mV were selected for analysis [range -35 to -75 mV; -44.6 ± 9.1 mV, mean ± SD (n = 157)].

The pathways mediating sacculo-neck motoneuron reflexes were studied by transecting the extent of the classically defined ipsilateral lateral vestibulospinal tract (LVST) and both the ipsi- and contralateral medial vestibulospinal tract (MVST) in nine cats. The locations and trajectories of the LVST and the MVST (Fig. 1, B and C) in the spinal cord were taken from the descriptions of Nyberg-Hansen (1964) and Petras (1967). Transections (Fig. 1, B and C) were performed in the caudal part of the C1 segment or in the rostral part of the C2 segment with the use of a razor blade specifically prepared for this purpose. Although the LVST could be lesioned unilaterally in a reliable and reproducable fashion, MVST sectionings always involved the pathways on both sides of the spinal cord (Fig. 1C). In the latter case, the location, inaccessability, and configuration of the tracts left no other choice. MVST lesions most likely also involved severing reticulospinal tract fibers (RST; Fig. 1, Ca and Cb) (Mitani et al. 1988). Thisfact is taken into consideration in the RESULTS and DISCUSSION sections. It should be noted, however, that RST damage associated with the LVST lesions (Fig. 1, Ba and Bb) was of no consequence because in such a case only a direct vestibulospinal, i.e., monosynaptic projection was implicated. Reconstruction of lesion sites from the histological sections indicated that the intended transections always covered the classically defined tract areas without overtly damaging the pathways that were meant to be left intact. Reconstructions of eight of the nine transection sites are illustrated in Fig. 1, Ba and Bb and Ca and Cb.

At the beginning of each experimental session, a field potential analysis was performed from the extracellular recordings in the vestibular nuclei to assess the validity of a given preparation for selective stimulation. In cases where the outcome was satisfactory (see RESULTS), synaptic potentials were subsequently recorded intracellularly in identified cervical cord motoneurons in intact and in vestibulospinal-tract-lesioned preparations.

Recording sites of field potentials in the vestibular nuclei were marked with the Fast Green FCF dye in some cats. At the end of each experiment, a lethal dose of the anesthetic was administered and the brain stem and upper spinal cord were removed. After fixation in 10% Formalin, the brains were cut into 100-µm serial transverse sections on a freezing microtome and processed for Nissl stain to identify the location of the recording sites.

    RESULTS
Abstract
Introduction
Methods
Results
Discussion
References

Synaptic potentials were analyzed in a total of 157 neck motoneurons, of which 102 were recorded in intact animals and 55 in spinal-cord-transected animals (n = 9). All lateralities are referenced to the sacculus stimulation side.

Field potentials in the vestibular nuclei

Stimulation of the saccular nerve evoked the classic small positive-negative wave and N1 potential (Fig. 2, inset) in the ipsilateral vestibular nuclear complex. The latency of the foot of the N1 potential was 0.8-1.1 ms (0.95 ± 0.08 ms; n = 19); that of the peak was 1.2-2.7 ms (1.35 ± 0.29 ms; n = 19), signaling activation of second-order vestibular neurons by primary vestibular neurons. In earlier, similar experiments (see DISCUSSION), Wilson et al. (1978) reported latencies of 0.7-1.9 ms (1.1 ± 0.3 ms; n = 4) for the foot of the N1 potential, and 1.2-2.5 ms (1.5 ± 0.5 ms; n = 4) for the peak. The typically observed monosynaptic latencies of 0.7-1.2 ms for EPSPs in second-order vestibular neurons following primary nerve stimulation would superimpose on these N1 potentials (cf. Wilson and Melvill Jones 1979). Thus we defined synaptic potentials in neck motoneurons with latencies of 1.8-3.0 ms as disynaptic. The field potentials were observed in the rostral part of the descending vestibular nucleus and to some degree in the ventral part of the lateral vestibular nucleus.


View larger version (17K):
[in this window]
[in a new window]
  		FIG. 2. Stimulus-response curves of N1 potential. Amplitude of N1 field potential (ordinate) is plotted against stimulus intensity (abscissa) for 3 different prototypical preparations (types A, B, and C). For experiment, only stimulus intensities that were within plateau levels of type A and B preparations, i.e., 19 cases, were admitted. Type C preparations were considered not to have sufficient stimulus separation and were discarded. Inset: typical field potential (averaged) recorded in vestibular nuclei with incoming volley of saccular afferents positive/negative (P/N) and a monosynaptic N1 field potential elicited in vestibular nuclei by saccular nerve stimulus. Triangle: saccular nerve stimulus. In 9 A-type preparations, latencies ofP/N and N1 potentials ranged from 0.6 to 0.7 ms (0.62 ± 0.04 ms,mean ± SD; n = 9) and from 0.9 to 1.1 ms (0.97 ± 0.70 ms; n = 9), respectively, at stimulus intensities of 2 × threshold for N1 potential (N1T). Stimulus strengths needed to evoke 80% of maximal N1 potential amplitude were between 22 µA (1.5 × N1T) and 46 µA (3.7 × N1T). Average of these stimulus intensities was 35.1 ± 9.1 µA (2.4 ± 0.8 × N1T; n = 9). Maximal N1 potential amplitudes, i.e., plateau levels, were attained at stimulus intensities between 35 µA (1.7 × N1T) and 70 µA (7.5 × N1T), corresponding to a mean of 49.4 ± 10.7 µA (3.8 ± 2.2 × N1T; n = 9). Responses from 10 cats were classified as type B. In these cases, latencies of initial P/N potentials ranged from 0.5 to 0.7 ms (0.61 ± 0.07 ms; n = 10) and latencies of N1 potentials ranged from 0.8 to 1.0 ms (0.92 ± 0.08 ms; n = 10). N1T was found between 6 and 40 µA (18.7 ± 8.4 µA; n = 10). N1 potentials of type A responses reached 80% of maximal amplitude with weaker stimulus intensities than those of type B.

Because the three ampullary nerves and the utricular nerve run near the saccular nerve in the inner ear, there is always a possibility that stimulus current applied to the saccular nerve might spread to these nerves with increasing stimulus intensity. This depends on the quality of the inner ear surgery and the stimulation electrode placement. The viability of a given preparation was therefore tested by recording N1 field potential amplitudes in relation to applied stimulus intensity. We distinguished between three prototypical stimulus-response curves of N1 potentials in respective preparations, which we termed A, B, and C types (Fig. 2).

In the illustrated example of a typical A-type preparation (Fig. 2), the threshold for the N1 potential (N1T) was 8 µA. As stimulus intensity was increased to ~20 µA (2.5 × N1T), the amplitude grew rapidly to 600 µV, i.e., 86% of the maximal recorded amplitude. The N1 amplitude then gradually leveled to a plateau between a stimulus intensity of ~6 and 12.5 × N1T (i.e., 100 µA).

Stimulus-response A-type curves were observed in nine cats. In these cases the N1T ranged from 8 to 30 µA(16.9 ± 8.0 µA; n = 9). In the second response category, type B (10 cases), only a short plateau was observed within a stimulus intensity range of 45 µA (5.5 × N1T) to 60 µA (7.3 × N1T; Fig. 2). Additional increases in N1 amplitudepresumably indicated current spread to other branches of the vestibular nerve (Fig. 2). Latency and stimulus strength parameters were not different from those observed in type A preparations.

We interpreted these findings as follows. At stimulus intensities evoking N1 potential amplitudes below the plateau level, not all saccular nerve fibers have as yet been recruited by the stimulus current. At higher stimulus intensities comprising the plateau level of the response curve, all available saccular nerve fibers would be recruited, but the stimulation site would still be confined to the individual nerve. At higher stimulus intensities, current spread to other vestibular nerve branches would occur, recruiting more input fibers to the second-order vestibular neurons, thus increasing again the N1 field potential amplitude. Consequently, stimulus intensities were kept at or slightly below plateau level (e.g., at points b and c as illustrated in Fig. 3A).


View larger version (13K):
[in this window]
[in a new window]
  		FIG. 3. A: stimulus-response curve of N1 potential evoked by saccular nerve stimulation in 1 type A preparation. B: excitatory postsynaptic potentials (EPSPs) recorded from an i-dorsal rami (DR) motoneuron in same preparation following stimulation of saccular nerve at intensities indicated at points a-d in A (stimulus intensities: a, 25 µA; b, 30 µA; c, 40 µA; d, 60 µA). Top traces: averaged intracellular records. Bottom traces: averaged extracellular records.

As stimulus intensity increased, the amplitude of the EPSP also increased. An example is shown in Fig. 3, Ba and Bb. The EPSP amplitudes at 3 × N1T in the recorded neurons ranged from 126 µV to 1.3 mV (average 494 ± 387 µV;n = 14).

In C-type preparations, no indication for the point of selective saccular nerve stimulation could be obtained, i.e., there was a complete absence of any stimulus intensity-N1 potential amplitude plateau. In these cases the experiment was not further pursued, and consequently, only data from A- and B-type preparations exist and were entered into this study.

Synaptic potentials in neck motoneurons following saccular nerve stimulation

INTACT PREPARATION. Synaptic potentials, either of excitatory or inhibitory nature, following sacculus stimulation could be recorded in all identified motoneurons. In most cases, these were either simple EPSPs or IPSPs (Fig. 4, 1, 4, 7, and 10, and Fig. 5). As a rule, synaptic potentials in ipsilateral motoneurons had larger amplitudes than those recorded in contralateral ones (Fig. 4, 1 and 4, 7 and 10), and required less stimulus strength to evoke a maximal response. On average, EPSPs in ipsilateral (i-) DR motoneurons had an amplitude of 501 µV; those in c-DR motoneurons had an amplitude of 230 µV.


View larger version (16K):
[in this window]
[in a new window]
  		FIG. 4. Typical synaptic potentials (averaged) in neck motoneurons following saccular nerve stimulation in intact and in lesioned preparations (i-LVST and MVST transections). Stimulus intensities for various records were as follows: 1, 150 µA (9.0 × N1T); 2, 40 µA (2.7 × N1T); 3, 100 µA (5.0 × N1T); 4, 40 µA (2.7 × N1T); 5, 30 µA (2.0 × N1T); 6, 50 µA (6.3 × N1T); 7, 25 µA (3.6 × N1T); 8, 70 µA (3.9 × N1T); 9, 50 µA (2.8 × N1T); 10, 40 µA (8.0 × N1T); 11, 50 µA (2.8 × N1T); 12, 100 µA (4.3 × N1T). Top and bottom traces in each record: intracellular and extracellular recordings, respectively. Note different amplitude calibrations. i-, ipsilateral; c-, contralateral; DR, dorsal rami; LC, longus capitis.


View larger version (23K):
[in this window]
[in a new window]
  		FIG. 5. Latency of synaptic potentials in i-DR, c-DR, i-LC, and c-LC motoneurons in intact and in lesioned preparations (i-LVST cut/MVST cut). Different symbols identify various observed synaptic responses and stimulus conditions. Open circles: EPSPs in type A and B preparations. Solid circles: inhibitory postsynaptic potentials (IPSPs) in type A and B preparations. Open triangles: EPSP/IPSP complexes in type B preparation. Solid triangles: IPSP/EPSP complexes in type B preparations. Asterisks: recordings in identified motoneurons without elicitation of synaptic responses, indicating thus positive effect of respective pathway transection. Note that saccular input to i-DR was affected by i-LVST lesion, and that to c-DR, i-LC, and c-LC by MVST lesion.

DR motoneurons. Intracellular recordings were performed on 29 i-DR motoneurons. Latencies of the EPSPs ranged from 1.8 to 3.0 ms (2.2 ± 0.3 ms; n = 27; Figs. 4, 1, and5, i-DR), suggesting a disynaptic pathway (Bolton et al. 1992; Ikegami et al. 1994; Uchino et al. 1994a; Wilson and Maeda 1974). Simple EPSPs were evoked in almost all of the recorded i-DR motoneurons (27 of 29). Only in two cases of type B preparations were EPSP/IPSP complexes observed (Fig. 5).

Synaptic potentials were also analyzed in 29 c-DR motoneurons. The typical response in c-DR motoneurons was an EPSP with disynaptic or trisynaptic latency (Fig. 4, 4). All synaptic potentials recorded from c-DR motoneurons in type A preparations, and from the majority in type B preparations, were simple EPSPs (Fig. 5); the remaining examples were IPSP/EPSP or EPSP/IPSP complexes (see also DISCUSSION). Latencies ranged from 1.8 to 4.0 ms, indicating di- and trisynaptic connectivities (Fig. 5, c-DR). Most EPSP thresholds, tested with double or occasional triple shocks, were below 2.0 x N1T, similar to the situation found in i-DR motoneurons. The amplitude of the evoked EPSPs in c-DR motoneurons at 3 × N1T stimulation ranged from 66 to 340 µV (230 ± 102 µV; n = 12).

LC motoneurons. Intracellular recordings were performed on 27 i-LC motoneurons. Typical responses were simple IPSPs (Fig. 47), whose latencies ranged from 1.7 to 3.6 ms. Thus the greater part of these connections could be considered disynaptic (Fig. 5, i-LC).

Synaptic potentials following saccular nerve stimulation were also recorded in 17 c-LC motoneurons. Typically, these responded with longer-latency IPSPs than did i-LC motoneurons, i.e., 2.7-4.0 ms (3.2 ± 0.4 ms; n = 14), suggesting that this pathway may be largely trisynaptic (Figs. 410 and 5, c-LC). Simple IPSPs were recorded from all c-LC motoneurons in type A preparations, and from the majority in type B preparations. In the remaining cases, EPSP/IPSP complexes and one IPSP/EPSP complex were present. Some motoneurons required double shocks to evoke responses at weak stimulus intensities near N1T.

LESIONED PREPARATIONS: VESTIBULOSPINAL TRACT TRANSECTIONS. After confirming the basic pattern of sacculo-neck motoneuron relationships, either the i-LVST (Fig. 1B) or the MVSTs/RSTs (Fig. 1C) were transected in a given preparation. Subsequently, synaptic responses to saccular nerve stimulation were recorded again from DR and LC motoneurons.

Pathway transections always covered the intended lesion site (Fig. 1, B and C). The electrophysiological results (Fig. 4, middle and bottom rows, and Fig. 5) indicate that the pathway that was meant to be left intact in a given preparation was transmitting the appropriate information to a given motoneuron population, and thus had remained viable despite the extensive primary and collateral damage within the neighboring structures (see Fig. 1, B and C).

Transection of the i-LVST abolished all disynaptic EPSPs in the eight tested i-DR motoneurons even when triple shocks were used (Fig. 42) at a stimulus intensity of 40 µA (i.e., ~2.7 × N1T; Fig. 5, i-DR). On the other hand, transectionof the MVSTs/RSTs had no effect on these excitatory potentials: disynaptic EPSPs could still be recorded in all tested i-DR motoneurons (n = 8; Fig. 43 and Fig. 5, i-DR). Thedata thus suggest that the disynaptic sacculo-neck connections to ipsilateral neck extensors project via the i-LVST.

Lesioning the i-LVST had no effect on the synaptic responses of c-DR motoneurons: simple disynaptic EPSPs or EPSPs in combination with IPSPs could still be demonstrated in all eight recorded c-DR motoneurons (Fig. 45 and Fig. 5, c-DR). After transection of the MVSTs/RSTs, excitatory synaptic potentials were absent in all 12 recorded c-DR motoneurons (Fig. 46 and Fig. 5, c-DR). Somewhat unexpectedly, IPSPs were still present in three c-DR motoneurons of one B-type preparation (Fig. 5, c-DR) (see DISCUSSION). Thus we conclude that the sacculo-neck connectivity to contralateral neck extensors projects largely through the MVSTs, but may also involve other pathways.

Transection of the i-LVST had no effect on the transmission of saccular information to any of the subsequently recorded LC motoneuron: IPSPs could be elicited in all i-LC (Fig. 48 and Fig. 5) and c-LC (Fig. 411 and Fig. 5) motoneurons, although in type B preparations IPSPs were preceded by an EPSP in three c-LC motoneurons after the transection. After the MVSTs/RSTs had been cut, sacculus-elicited potentials were entirely absent in all tested LC motoneurons (Fig. 4, 9 and 12, and Fig. 5). The data thus indicate that sacculus-related postural control information to the ipsilateral and contralateral neck flexor LC is transmitted via the MVSTs and/or reticulospinal tracts.

    DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References

Organization of sacculo-neck motoneuron pathways

One of the remarkable features of the sacculo-neck motoneuron relationship is its qualitatively bilaterally symmetrical organization to neck flexors and extensors (Fig. 6). The primary vestibular neurons coming from the sacculus contact second-order vestibular neurons of excitatory and inhibitory nature located in the vestibular nuclear complex (descending and ventral part of the lateral vestibular nucleus). Potentials for our initial field potential analyses were also recorded in these areas, i.e., in locations similar to those observed in previous investigations (Hwang and Poon 1975; Wilson et al. 1977, 1978). Through these relay neurons, the sacculus of one side sends excitatory input to ipsilateral and contralateral neck extensors and inhibitory commands to ipsilateral and contralateral neck flexors. Three-neuron arc connections are established mainly with the ipsilateral extensors (excitation) and flexors (inhibition), but also with contralateral extensors (excitation). The latter also receive additional excitatory input via a trisynaptic connectivity by insertion of a spinal interneuron, or possibly via reticulospinal neurons. Inhibitory input may also arrive at ipsilateral flexors via an interneuron connection. The inhibitory input to contralateral flexors occurs exclusively via a trisynaptic pathway. In such a case, the reticulospinal and/or interneuron link may also serve to change the synaptic quality of the original excitatory input. Except for the excitatory disynaptic projection to ipsilateral neck extensors, which makes use of the ipsilateral vestibulospinal tract, all pathways travel in the MVSTsand/or reticulospinal tracts. Our estimation and comparison of the amplitudes of the postsynaptic potentials indicate, furthermore, that the ipsilaterally projecting pathways provide a more powerful input to their respective motoneurons than the contralaterally projecting ones. On average, EPSP amplitudes in i-DR motoneurons were about twice those of c-DR motoneurons.


View larger version (14K):
[in this window]
[in a new window]
  		FIG. 6. Schematic diagram of disynaptic and trisynaptic excitatory (A) and inhibitory (B) sacculo-neck motoneuron connections. Saccular nerve contacts excitatory (open circle ) and inhibitory (bullet ) 2nd-order vestibular neurons. These project directly or indirectly (via spinal cord interneurons and/or reticulospinal neurons) to neck extensor and flexor moroneurons. Thick lines: strong connections. Thin lines: weaker connections. Locations of excitatory and inhibitory interneurons have not been identified, nor have midline crossing points within MVST and/or reticulospinal tract. Only ipsilateral excitatory pathway to neck extensors projects through i-LVST. Note qualitatively bilaterally symmetrical innervation pattern of sacculo-neck reflexes.

The location of excitatory and inhibitory interneurons is speculative. They could reside in the vestibular nuclei and/or in the spinal cord, or could be constituted by reticulospinal neurons. Future experiments, furthermore, will have to determine whether the sacculus information that reaches the contralateral motoneurons is transmitted via a shared pathway, as illustrated schematically in Fig. 6, or whether private channels exist for the three-neuron arc and the interneuron connectivities. In addition, the axonal crossings from the ipsilateral to the contralateral side need to be specified.

Role of the sacculocollic reflex

The role of the sacculus for equilibrium function and postural control reflexes has been elusive since the beginning of systematic vestibular research (e.g., de Kleijn and Magnus 1921a,b; Magnus and de Kleijn 1932) largely because of the difficulties in obtaining reliable and repeatable lesion and stimulation experiments (cf. Graf et al. 1992; Kanesada et al. 1989). Our study made use of a combination of elaborate inner ear surgery and rigid stimulation criteria to ensure selective activation of sacculus-related postural circuits.

Our results show that, in essence, stimulation of one sacculus activates neck extensors bilaterally and inhibits neck flexors bilaterally, as demonstrated by di- and trisynaptic EPSPs and IPSPs in the respective motoneurons. This innervation pattern is quite different from that of the other otolith receptor, the utriculus (Bolton et al. 1992; Ikegami et al. 1994), and only comparable with that of the vertical semicircular canals (Table 1). One utriculus projects to neck extensors and flexors in a unilateral fashion, i.e., the ipsilateral extensors and flexors receive excitatory input in the form of disynaptic EPSPs whereas the contralateral extensors and flexors receive inhibitory input in the form of trisynaptic IPSPs. This differential organization of otolith output accounts for the fundamentally different symptoms following ablation of individual statoreceptors (see below). As in the sacculus, vertical canal innervation of neck motoneurons also shows a bilaterally symmetrical organization (Table 1). For example, extensor motoneurons on both sides receive disynaptic EPSPs from an anterior canal and disynaptic IPSPs from a posterior canal (cf. Bolton et al. 1992; Fukushima et al. 1979; Ikegami et al. 1994; Isu et al. 1988; Shinoda et al. 1994; Sugiuchi et al. 1992a; Uchino et al. 1990; Wilson and Maeda 1974).

 
View this table:
[in this window] [in a new window]
  		TABLE 1. Innervation pattern of neck extensor and neck flexor motoneurons originating from labyrinth receptors

Saccular nerve stimulation selectivity

One of the criteria for a meaningful analysis of sacculo-neck motoneuron relationships was to ensure unequivocally the selectivity of the demonstrated synaptic qualities, i.e., to eliminate false positive projections due to current spread to other vestibular nerve branches. To that end, the viability of each preparation was tested at the beginning of each experiment with a field potential analysis in the vestibular nuclei. We concluded that in an ideal situation the saccular nerve would be stimulated without current spread to the other branches of the vestibular nerve, if the current intensity-N1 potential graph showed a plateau, as in type A preparations. This consideration is supported by the comparison of observed synaptic potentials in neck motoneurons to a given sacculus stimulus with available published responses following stimulation of individual branches of the vestibular nerve (Table 1).

By contrast, type B responses require additional evaluation and interpretation. A priori, the input pattern of sacculus-related information in neck motoneurons recorded in type B preparations was identical to that of type A preparations, except for some compound responses in 9.8% of all recordings in intact and 12.7% in lesioned preparations. Several explanations can be envisaged for the appearance of these compound responses (10.8% of all analyzed synaptic potentials). One explanation concerns the inherently weaker connectivity to contralateral neck motoneurons. Postsynaptic potentials in contralateral motoneurons were often smaller in amplitude than in ipsilateral motoneurons, and thus somewhat stronger stimuli had to be employed occasionally to examine synaptic potentials in contralateral motoneurons. Although stimulus intensities were still within our criterion range of the plateau of the amplitude of N1 potentials, there could have been an increased risk of current spread to other vestibular nerve branches due to the narrow plateau in type B preparations.

In these isolated cases we may have to assume current spread to the posterior canal nerves as a source for some of the observed compound responses. The reason for this probability is based in the anatomic configuration of inner ear innervation. The saccular and posterior canal nerves comprise the posterior ramus of labyrinthine nerves, the utricular nerve, and the other semicircular canal nerves form the anterior ramus. Saccular and posterior canal nerves run together over a considerable distance immediately outside the inner ear until they join the anterior ramus. The two rami then compose the vestibular nerve proper. In addition, the anterior ramus is separated from the posterior ramus by bone. Because selective stimulation of the posterior semicircular canal nerve evokes disynaptic IPSPs in extensor motoneurons and disynaptic EPSPs in flexor motoneurons (Table 1), this innervation pattern coincides with the synaptic quality of the respective IPSP or EPSP of the compound response. Thus we may assume that the observed compound responses evoked with disynaptic latencies in c-DR and c-LC motoneurons of type B preparations were originating from posterior canal input and were not idiosyncratic to a particular sacculus connectivity. In light of this interpretation and because of the other unequivocal synaptic responses in A-type and other B-type preparations, we decided to enter the principal synaptic quality of the discussed compound responses seen in the neurons in question into the data roster, i.e., EPSPs for ipsilateral and contralateral extensors and IPSPs for contralateral flexors.

Minimalization of current spread and contextual data evaluation was also important in light of slight differencesbetween our present results and those of a previous study(Wilson et al. 1977). In that study, chronically prepared animals with long survival times were used. In such cases, sprouting and the formation of new connectivities may have already occurred. Furthermore, Wilson et al. (1977) stimulated with stimulus strengths of 60-500 µA (typically 200 µA), whereas in our study stimulus intensities of ~40 µA were used. These significantly higher values in the earlier study clearly compromised stimulation selectivity.

Vestibulospinal pathways

Transection of either the i-LVST or the MVST/RST with subsequent renewed recordings of synaptic potentials in neck motoneurons provided the necessary information about the principal sacculocollic pathways. Unequivocal results concerned ipsilateral extensor, and ipsilateral and contralateral flexor motoneurons. Our data indicate that the i-DR motoneurons receive their excitatory input through the i-LVST, the i-LC, and c-LC motoneurons via the MVSTs/RSTs. Synaptic potentials could no longer be recorded in either of these motoneurons after the corresponding tracts had been transected.

The innervation pattern involving c-DR motoneurons is not as clear. Although the MVSTs/RSTs transmit the major portion of sacculus-related information to the c-DR motoneurons, they do not contain all relevant pathways, because saccular nerve stimulation did not elicit any postsynaptic potentials in most (12 of 15) of the tested c-DR motoneurons after the MVSTs/RSTs had been severed. In three exceptional cases, IPSP and IPSP/EPSP complexes remained after MVST/RST transection. These may follow additional pathways involving vestibulospinal neurons with widespread bilateral terminations in the upper cervical spinal cord (Donevan et al. 1990), which arise from axons traveling in several funiculi such as the ventromedial, ventrolateral, lateral, dorsolateral, and dorsal funiculus. These funiculi could transmit the respective sacculus information. The dorsal funiculus of the contralateral side, however, can be excluded as a possible pathway, because it was always transected together with the MVSTs in our preparations.

Trisynaptic pathways

A large source of trisynaptic information arriving at contralateral neck extensor and flexor motoneurons may be reticulospinal neurons (Fukushima et al. 1979; Peterson 1979; Wilson and Peterson 1982). Our midline lesion involving the MVSTs most likely also severed reticulospinal neuron axons (see Mitani et al. 1988). Differentiating betweenthe two pathways with the employed techniques was not possible.

We also have to assume that interneurons mediating polysynaptic activation of neck motoneurons from the sacculus could be located in the spinal cord. Commissural interneurons in the upper cervical spinal cord project to the contralateral ventral horn (Bolton et al. 1991) and to contralateral neck motoneurons (Sugiuchi et al. 1992b). However, in the present study, we did not examine the location of interneurons that mediate the observed polysynaptic responses from second-order vestibular neurons to neck motoneurons.

Behavioral correlates

Electromyographic activity of extensor neck muscles during sinusoidal vertical linear accelerations in alert cats was characterized by two typical patterns depending on stimulus frequency (Lacour et al. 1987). Neck muscle responses were composed of a second harmonic of the stimuli in the low frequency range (0.05-0.25 Hz), i.e., electromyographic activity increased during the upward and the downward segment of vertical translation. By contrast, neck extensor responses were reported to be in phase with acceleration in the higher frequency range (0.25-1 Hz), i.e., maximal electromyographic activity coincided with the peak of downward acceleration only. This reported change in the low-frequency dynamics of neck muscle activity is difficult to explain exclusively on the basis of the di- and trisynaptic circuits of the sacculocollic reflex arcs described in the present study. Thus we suggest that additional polysynaptic pathways involving brain stem and cerebellar circuits may be contributing to the described reflex behavior.

The postural symptom following unilateral saccular lesion is a slight side tilt of head and neck away from the side of the lesion (Graf et al. 1992; see also Kanesada et al. 1989). This lateral deviation of the head-neck ensemble is much less pronounced than that following a complete hemilabyrinthectomy, or even an isolated unilateral otolith extirpation (Schaefer and Meyer 1974; de Waele et al. 1989). The sacculus lesion symptom can now be explained on the basis of our present results. Saccular inputs evoke EPSPs in bilateral neck extensor motoneurons and IPSPs in bilateral flexor motoneurons, however with a more powerful projection to the ipsilateral side (Fig. 6, thick axons). Thus a unilateral saccular lesion would only cause a slight differential in tonic resting activity between the both sides. For instance, a right-side lesion would leave functionally intact only the illustrated connectivity of Fig. 6. The left-side extensor tonus together with removal of inhibition from flexor motoneurons (circuits previously under control of the right sacculus) would provoke the observed side tilt to the left, i.e., away from the lesion. A bilateral lesion of the sacculus system has no obvious static postural consequences.

By contrast, the pronounced side tilt after hemilabyrinthectomy (Schaefer and Meyer 1974; de Waele et al. 1989) is largely due to a unilateral loss of utricular inputs that do not exhibit bilateral projections of the same synaptic qualities, i.e., utricular inputs elicit EPSPs in ipsilateral neck extensor and flexor motoneurons and IPSPs in contralateral extensor and flexor motoneurons (Bolton et al. 1992; Ikegami et al. 1994). Thus the lesion symptoms, in essence, reflect this bilaterally asymmetric projection pattern of the utriculomotor system.

The above synaptic connectivites, however, do not explain the classical hemilabyrithectomy lesion symptoms per se, which are the perceptual consequences of a sensory lesion and the resulting acute compensatory righting reflex manifestations. In essence, a lesioned animal attempts to regain its equilibrium zero setpoint by rotating in the opposite direction, thus trying to eliminate or reduce an apparent postural disturbance.

Otolith reference frames

The adequate stimulus to stimulate optimally the receptor cells in the sacculus is a linear vertical acceleration based on electrophysiological results (Fernández and Goldberg 1976a,b) and on anatomic and geometric considerations (see, e.g., Lewis et al. 1985): the sacculus maculae are oriented vertically, that of the utriculus system horizontally (cf., Spoendlin 1966).

Taking together the lesion results and the otolith-to-neck motoneuron connectivities, we suggest that different symmetry planes apply to the two systems, in a similar fashion as in the semicircular canal system (Table 1). Clearly, the major sensory-to-motor output channel of the utriculus system concerns movement in the left-right and also in the anterior-posterior direction, as suggested by the respective directional sensitivity vector distributions (Fernández and Goldberg 1976a; Loe et al. 1973). It thus detects and compensates, e.g., for side tilts and flexion-extension displacements of the head and/or the body. In the former case, the utriculi of both sides would operate as a left-right differentiating system; in the latter case, the two utriculi would provide the same sensory output, working as a fore-aft movement detector. The sacculus system, by contrast, seems to be principally controlling postural movements occurring in an up-down (and to a certain degree, also anterior-posterior) direction (Fernández and Goldberg 1976a,b), thus being mainly concerned with vertical linear accelerations.

Conclusions

The present study demonstrates the presence of excitatory connections from the sacculus to the bilateral neck extensor muscle motoneurons, and inhibitory connections to the bilateral neck flexor muscle motoneurons. The system thus seems to be organized in a push-pull fashion, with extensor muscles fulfilling an antigravity function during vertical upward accelerations or when the sacculus senses the omnipresent gravity acceleration in a terrestrial life situation. During a possible vertical downward acceleration, flexor activation would be produced by inhibition of the inhibitory circuits to the flexor motoneurons. These reciprocal connections control the excitability of neck motoneurons in a sacculus-specific plane of reference that is different from that of the utriculus. The sacculus system thus supplements the reference planes of the utriculus system, and thereby all three translational spatial degrees of freedom are represented in the gravistatic system. Individually, the sacculocollic reflex plays an important role in maintaining the relative position of the head and of the body against vertical linear postural disturbances and against gravity.

    ACKNOWLEDGEMENTS

  We thank K. Takayama for secretarial assistance. This study was carried out as a part of Space Utilization Frontiers Joint Research Projects promoted by NASDA.

    FOOTNOTES

  Address for reprint requests: Y. Uchino, Dept. of Physiology, Tokyo Medical College, 6-1-1 Shinjuku, Shinjuku-ku, Tokyo 160, Japan.

  Received 16 November 1995; accepted in final form 20 February 1997.

    REFERENCES
Abstract
Introduction
Methods
Results
Discussion
References
0022-3077/97 $5.00 Copyright ©1997 The American Physiological Society



This article has been cited by other articles:


Home page
 NeurologyHome page
K. Liao, J. Wagner, A. Joshi, I. Estrovich, M. F. Walker, M. Strupp, and R. J. Leigh
Why do patients with PSP fall?: Evidence for abnormal otolith responses
Neurology, March 4, 2008; 70(10): 802 - 809.
[Abstract] [Full Text] [PDF]

Home page
 J. Physiol.Home page
F. Deriu, E. Ortu, S. Capobianco, E. Giaconi, F. Melis, E. Aiello, J. C. Rothwell, and E. Tolu
Origin of sound-evoked EMG responses in human masseter muscles
J. Physiol., April 1, 2007; 580(1): 195 - 209.
[Abstract] [Full Text] [PDF]

Home page
 NeurologyHome page
M. S. Welgampola and J. G. Colebatch
Characteristics and clinical applications of vestibular-evoked myogenic potentials
Neurology, May 24, 2005; 64(10): 1682 - 1688.
[Abstract] [Full Text] [PDF]

Home page
 J. Neurophysiol.Home page
F. Deriu, E. Tolu, and J. C. Rothwell
A Sound-Evoked Vestibulomasseteric Reflex in Healthy Humans
J Neurophysiol, May 1, 2005; 93(5): 2739 - 2751.
[Abstract] [Full Text] [PDF]

Home page