HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Izawa, Y. Articles by Shinoda, Y. Search for Related Content PubMed PubMed Citation Articles by Izawa, Y. Articles by Shinoda, Y.

Neural Organization of the Pathways From the Superior Colliculus to Trochlear Motoneurons

Department of Systems Neurophysiology, Graduate School of Medicine, Tokyo Medical and Dental University, Yushima, Tokyo, Japan

Submitted 9 October 2006; accepted in final form 23 February 2007

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
The neural organization of the pathways from the superior colliculus (SC) to trochlear motoneurons was analyzed in anesthetized cats using intracellular recording and transneuronal labeling techniques. Stimulation of the ipsilateral or contralateral SC evoked excitation and inhibition in trochlear motoneurons with latencies of 1.1–2.3 and 1.1–3.8 ms, respectively, suggesting that the earliest components of excitation and inhibition were disynaptic. A midline section between the two SCs revealed that ipsi- and contralateral SC stimulation evoked disynaptic excitation and inhibition in trochlear motoneurons, respectively. Premotor neurons labeled transneuronally after application of wheat germ agglutinin-conjugated horseradish peroxidase into the trochlear nerve were mainly distributed ipsilaterally in the Forel's field H (FFH) and bilaterally in the interstitial nucleus of Cajal (INC). Consequently, we investigated these two likely intermediaries between the SC and trochlear nucleus electrophysiologically. Stimulation of the FFH evoked ipsilateral mono- and disynaptic excitation and contralateral disynaptic inhibition in trochlear motoneurons. Preconditioning stimulation of the ipsilateral SC facilitated FFH-evoked monosynaptic excitation. Stimulation of the INC evoked ipsilateral monosynaptic excitation and inhibition, and contralateral monosynaptic inhibition in trochlear motoneurons. Preconditioning stimulation of the contralateral SC facilitated contralateral INC-evoked monosynaptic inhibition. These results revealed a reciprocal input pattern from the SCs to vertical ocular motoneurons in the saccadic system; trochlear motoneurons received disynaptic excitation from the ipsilateral SC via ipsilateral FFH neurons and disynaptic inhibition from the contralateral SC via contralateral INC neurons. These inhibitory INC neurons were considered to be a counterpart of inhibitory burst neurons in the horizontal saccadic system.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
The superior colliculus (SC) plays a prominent role in controlling eye and head movements. The neural organization of the pathways from the SC to motoneurons in the horizontal oculomotor system has been extensively analyzed (see reviews in Grantyn and Moschovakis 2004; Scudder et al. 2002; Sparks 1999). Electrical stimulation of the SC evokes saccades with a contraversive horizontal component (Guitton et al. 1980; Robinson 1972), and neurons in the deeper layers of the SC show a burst of spikes before contraversive saccades (Schiller and Körner 1971; Wurtz and Goldberg 1972). Medium-lead burst neurons (MLBNs) in the paramedian reticular formation that burst before and during horizontal saccades include both excitatory neurons (Cohen and Henn 1972; Keller 1974; Luschei and Fuchs 1972) and inhibitory neurons (Hikosaka and Kawakami 1977; Yoshida et al. 1982). MLBNs in the paramedian reticular formation rostral and caudal to the abducens nucleus are activated monosynaptically from the contralateral SC in alert cats (Chimoto et al. 1996). The shortest excitatory pathway from the contralateral SC to abducens motoneurons is a disynaptic circuit via the ipsilateral paramedian pontine reticular formation (PPRF), whereas the shortest inhibitory pathway from the ipsilateral SC to abducens motoneurons is a disynaptic circuit via the contralateral paramedian pontomedullary reticular formation (Izawa et al. 1999). In contrast, the connections between the SC and motoneurons in the vertical oculomotor system have not yet been analyzed in detail. The trochlear nucleus, which supplies the contralateral superior oblique muscle, is separate from the oculomotor nucleus (Warwick 1953) and provides a convenient model for investigation of the vertical gaze system, as its major pulling direction is downward with lesser lateral and intortion components (Fuchs and Luschei 1971; Tokumasu et al. 1965). The effects of stimulation of the SC on trochlear motoneurons were first examined by Precht et al. (1974). However, consistent postsynaptic potentials (PSPs) were not detected in trochlear motoneurons in this study, even after strong and multiple shock stimulation of the SC, in contrast to the presence of clear SC-evoked PSPs in abducens motoneurons in the same preparations. Since this report, to our knowledge, inputs from the SC to trochlear motoneurons have not been investigated.

The best-studied premotor neurons in the vertical oculomotor system are found in the Forel's field H (FFH) (Graybiel 1977), the homologue of the rostral interstitial nucleus of the medial longitudinal fasciculus (riMLF) in the monkey (Büttner-Ennever and Büttner 1978) and in the interstitial nucleus of Cajal (INC) and adjacent mesencephalic reticular formations (MRF). Schwindt et al. (1974) examined the effects of stimulation of the latter area on trochlear motoneurons and found that stimulation of the INC evoked a monosynaptic excitatory and inhibitory postsynaptic potential (EPSP-IPSP) sequence in trochlear motoneurons on both sides. Later, Nakao and Shiraishi (1985) reported direct synaptic connections between the FFH and trochlear motoneurons; FFH stimulation evoked monosynaptic EPSP-IPSPs or EPSPs in trochlear motoneurons on both sides. The FFH and INC contain MLBNs that discharge most vigorously for saccades with vertical components and have either up- or downward on-directions (Büttner et al. 1977; King and Fuchs 1979; Moschovakis et al. 1991a,b; Nakao et al. 1988). In addition to MLBNs, neurons that exhibit a burst-tonic discharge pattern are also found in the INC and appear to provide position signals for vertical and torsional gaze (Chimoto et al. 1999; Fukushima et al. 1990; King et al. 1981). Furthermore, the FFH and INC contain neurons that project to the spinal cord that may be related to gaze shifts (Fukushima et al. 1981; Isa and Sasaki 2002). In addition to the discharge patterns of neurons in the FFH and INC, anatomical experiments in the cat suggested that the FFH/riMLF contains overlapping populations of neurons that are immunoreactive to putative inhibitory and excitatory neurotransmitters (Spencer and Wang 1996). In spite of the wealth of knowledge that has been obtained about premotor neurons in the vertical saccadic system, the input-output organization of the FFH and INC in the pathways from the SC to vertical ocular motoneurons is not fully understood.

We performed the present study to determine the neural organization of circuits connecting the SC to trochlear motoneurons by using electrophysiological and morphological techniques in anesthetized cats. In trochlear motoneurons, intracellular responses were examined following electrical stimulation of the SC, FFH, and INC. To locate last-order premotor neurons terminating on trochlear motoneurons, we labeled them transneuronally with wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP). The results showed that a reciprocal input pattern from the SCs to trochlear motoneurons existed in the vertical saccadic system; the shortest excitatory pathway from the ipsilateral SC to trochlear motoneurons was a disynaptic one via the ipsilateral FFH, whereas the shortest inhibitory pathway from the contralateral SC to trochlear motoneurons was a disynaptic one via the contralateral INC. These results were previously reported briefly (Izawa et al. 1997).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Experiments were performed in 19 cats weighing 2.4–4.4 kg. The surgery and animal care conformed to the "Guide for the Care and Use of Laboratory Animals" (Institute of Laboratory Animal Resources, National Research Council 1996), and also to the "Guiding Principles for the Care and Use of Animals in the Field of Physiological Sciences" (The Physiological Society of Japan, revised in 2001). All surgical and experimental protocols were approved by the Committee for Animal Experimentation of Tokyo Medical and Dental University. The animals were initially anesthetized with an intramuscular injection of ketamine hydrochloride (Ketalar, Parke-Davis; 25 mg/kg) followed by an intravenous injection of -chloralose (40–45 mg/kg initial dose, supplemented with an additional dose of 10–25 mg/kg as required). The body temperature was kept between 37.5 and 39.0°C by a heating pad. Heart rate was monitored constantly by electrocardiogram. The trochlear nerve was detached from the superior oblique muscle, and its peripheral end was mounted on a bipolar hook electrode for electrical stimulation or to record nerve discharge. One silver ball electrode was placed on the oval window and the other on the round window for stimulation of the primary vestibular nerves on each side, and single or double shock stimuli of 0.2 ms in duration were delivered between them at a maximum intensity of 500 µA (Shinoda and Yoshida 1974; Shinoda et al. 1986a, 1992). The bone over the parietal and occipital cortex was removed and the cerebral cortex overlying the SC was aspirated uni- or bilaterally to introduce stimulating or recording electrodes into the SC. Three pairs of monopolar stimulating electrodes (epoxy-coated acupuncture needle electrodes; diameter, 0.3 mm) or four concentric bipolar stimulating electrodes (0.1 mm ID; 0.3 mm OD; interelectrode distance, 0.5 mm) were positioned in the intermediate or deep layer (1.5–2.0 mm deep from the surface) of the SC on either side. Stimulation sites in the SC were determined by monitoring nerve discharge of the trochlear nerve that was elicited by stimulating the deeper layers of the SC. Two to 10 negative pulses of 0.2 ms in duration were applied at an interval of 1.5 ms to evoke nerve discharge in trochlear nerves at a maximum of 100 µA by use of a constant current generator. Similar concentric stimulating electrodes were placed bilaterally at A7.0–7.5 in the FFH and also bilaterally at A5.0–5.5 in the INC stereotaxically. The final depths of the electrode tips were determined by monitoring negative field potentials or unit spikes evoked by stimulation of the ipsilateral SC. The vermis overlying the fourth ventricle was aspirated for intracellular recording from trochlear motoneurons. In three experiments, muscimol, which was used as a solution in saline with a concentration of 1 µg/µl, was injected with a Hamilton syringe into the FFH contralateral to the recorded trochlear nerve. Muscimol was injected in steps of 1 µl at 10-min intervals to give a total volume of 2.0–4.0 µl, and its effect on SC-evoked trochlear nerve discharge was examined. To examine axonal projections of single tectal output neurons, a concentric stimulating electrode of the same type as that used to stimulate the SC was placed in the FFH on each side, another electrode was placed in the contralateral PPRF, and two other electrodes were placed in the descending medial longitudinal fasciculus (MLF) at the level of the obex (Verhaart 1964) on the side contralateral to the SC in two animals. During recording, the animals were paralyzed by the intravenous administration of pancuronium bromide (Mioblock, Organon, The Netherlands) and artificially ventilated with the end-tidal CO₂ held at 37 mmHg.

Glass microelectrodes for intracellular recording were filled with 3 or 0.4 M KCl and had a resistance of 8–15 M. A tungsten electrode insulated in a glass microelectrode was used to record extracellular spikes in the SC (Shinoda et al. 1976, 1986b). Negative pulses of 0.2-ms duration were delivered at 100–500 µA to stimulate the SC, FFH, and INC, at <200 µA (usually <100 µA) to stimulate the PPRF, and at a maximum of 500 µA to stimulate the MLF. At the end of each experiment, the positions of the stimulating electrodes were marked by passing negative currents of 20 µA for 20 s. The animals were then deeply anesthetized with pentobarbital sodium (45 mg/kg, Nembutal, Abbott, Switzerland) and perfused with 2 l of saline followed by 2 l of 10% formalin solution. Stimulated sites in the SC, FFH, INC, PPRF, and MLF were histologically confirmed in serial transverse sections stained with thionin. In two experiments in which the PPRF was stimulated, we confirmed that the stimulating electrode was located in the PPRF where last-order interneurons were labeled retrogradely by injecting HRP into the ipsilateral abducens nucleus. This procedure has been described by Izawa et al. (1999).

To determine the location of last-order premotor neurons terminating on trochlear motoneurons, 2% WGA-HRP (Toyobo) in 0.05 M Tris-HCl buffer (pH 8.6) was injected into the trochlear nerve for transneuronal labeling in three animals. For the first two animals, after 4–6 days, the brain was removed and serial transverse sections of 50 or 75 µm were reacted to reveal the presence of HRP by use of the tetramethyl benzidine method (Mesulam 1978) following previously described procedures (Sugiuchi et al. 1995). For the last animal, 12.5% dextran-biotin (Molecular Probes) was also injected into the SC contralateral to the WGA-HRP-injected trochlear nerve to confirm that axon terminals of tectofugal neurons terminate on last-order premotor neurons, which, in turn, terminate on trochlear motoneurons. After 4–6 days, the brain was removed and serial transverse sections of 75 µm thickness were treated for WGA-HRP and biotin according to the double-labeling procedure described previously (Izawa et al. 1999).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
To determine the neural pathways from the SC to trochlear motoneurons, intracellular potentials were recorded from 151 trochlear motoneurons in the cat, and the effects of electrical stimulation of the SC, FFH and INC on these neurons were examined. All lateralities in this study are described with reference to the recording side.

Effects of stimulation of the SC on trochlear motoneurons

Trochlear motoneurons were identified by their antidromic responses to stimulation of the trochlear nerve in the contralateral orbit (Fig. 1B). Of the 151 trochlear motoneurons examined, SC-evoked PSPs were identified in 89 trochlear motoneurons, and therefore these motoneurons were used for later analysis. Their resting membrane potentials ranged from –40 to –76 mV (–56 ± 15 (SD) mV). In every motoneuron examined, we always examined vestibular inputs. This procedure helped differentiate EPSPs from reversed IPSPs due to spontaneous diffusion of Cl^– into a penetrated cell because we used KCl-filled micropipettes. Stimulation of the ipsilateral vestibular nerve evoked hyperpolarization (Fig. 1Ca), and stimulation of the contralateral vestibular nerve evoked depolarization in every trochlear motoneuron (Fig. 1Da) (Baker and Berthoz 1974; Highstein 1973; Precht and Baker 1972). In the same motoneuron, single shock stimuli applied to the SC on either side usually did not evoke any response, and double or triple shock stimuli were required to evoke PSPs. Stimulation of the ipsilateral SC evoked a depolarization (Fig. 1Ea), which was followed by a hyperpolarization in some cases. Stimulation of the contralateral SC evoked a hyperpolarization (Fig. 1Fa) that was followed or preceded by a depolarization in some motoneurons. Latencies and amplitudes of these PSPs fluctuated, suggesting that these responses were induced polysynaptically. To determine the nature of these depolarizing and hyperpolarizing PSPs, we injected Cl^– into the cell or passed hyperpolarizing or depolarizing currents through a recording microelectrode. To check the effects of the intracellular injection of Cl^– or the passage of current through a motoneuron membrane, PSPs evoked by stimulation of the contra- and ipsilateral primary vestibular nerves were compared before and after the injection of Cl^– or before and during the passage of current. Injection of Cl^– into the cell did not affect the polarity of the depolarizations evoked by stimulation of the contralateral vestibular nerve (Fig. 1Db), whereas the hyperpolarizations evoked by stimulation of the ipsilateral vestibular nerve were reversed in a depolarizing direction (Fig. 1Cb). Therefore the depolarizations evoked by contralateral vestibular stimulation were considered to be EPSPs, and the hyperpolarizations evoked by ipsilateral vestibular stimulation were considered to be IPSPs (Baker and Berthoz 1974; Coombs et al. 1955; Eccles 1964). Under the same conditions as those with which the vestibular-evoked disynaptic IPSPs were reversed after Cl^– injection, the polarity of the depolarizations evoked by stimulation of the ipsilateral SC was not affected (Fig. 1Eb), whereas the hyperpolarizations evoked by stimulation of the contralateral SC were reversed to depolarizing potentials (Fig. 1Fb). Therefore the depolarizations evoked by ipsilateral SC stimulation were considered to be EPSPs, and the hyperpolarizations evoked by contralateral SC stimulation were considered to be IPSPs (Coombs et al. 1955; Eccles 1964). Because these PSPs were usually evoked by the second or third stimulus, their latencies were measured from the first effective stimulus, which was most often the second stimulus. One additional SC stimulus more than the first effective stimulus usually decreased such latencies of the PSPs by 0.1–0.3 ms. The onsets of EPSPs and IPSPs were determined by superimposing the EPSPs and IPSPs on field potentials recorded just outside a cell. When the IPSPs followed EPSPs, the onsets of the IPSPs were obscure and were determined by superimposing the reversed IPSPs on the hyperpolarizing IPSPs, and the point of divergence of these two IPSPs was considered to be the onset of the IPSPs. Ipsilateral SC stimulation evoked EPSPs at latencies of 1.1–2.3 ms (1.6 ± 0.3 (SD) ms, n = 72; Fig. 2Aa), and IPSPs at latencies of 1.5–3.9 ms (2.7 ± 0.7 ms, n = 13; Fig. 2Ab). Contralateral SC stimulation evoked IPSPs at latencies of 1.1–3.8 ms (1.8 ± 0.6 ms, n = 50; Fig. 2Ba), and EPSPs at latencies of 1.2–2.7 ms (1.9 ± 0.4 ms, n = 17; Fig. 2Bb).

View larger version (28K): [in this window] [in a new window]   FIG. 1. Synaptic inputs from the superior colliculus (SC) to a trochlear motoneuron. A: schematic diagram of the experimental setup. Intracellular potentials were recorded in a trochlear motoneuron in the trochlear nucleus (IV Nucl.). Stimulating electrodes were placed in the SCs and primary vestibular nerves (VIII n.) on both sides, and in the contralateral trochlear nerve. SO, superior oblique muscle; VIII Nucl., vestibular nucleus. B: antidromic spikes of a trochlear motoneuron (SO MN) evoked by stimulation of the trochlear nerve at 40 µA. Bottom 2 traces: juxtacellular field potentials recorded just outside the penetrated cell. C and D: postsynaptic potentials (PSPs) in the trochlear motoneuron evoked by stimulation of the ipsilateral (C) and contralateral vestibular nerves at 500 µA (D) before (a) and after Cl– injection into the cell (b). E and F: PSPs in the trochlear motoneuron evoked by stimulation of the ipsilateral (E) and contralateral SCs at 500 µA (F) before (a) and after Cl– injection into the cell (b). Cc, Dc, Ec and Fc, juxtacellular field potentials. Calibrations in D apply to C, and calibrations in F apply to E. Upward deflection in all potential traces in this and the following figures represents positivity if not stated otherwise.

View larger version (28K): [in this window] [in a new window]   FIG. 2. Latency histograms of PSPs recorded in trochlear motoneurons. A: latency histograms of excitatory PSPs (EPSPs; a) and inhibitory PSPs (IPSPs; b) evoked by ipsilateral SC stimulation. B: latency histograms of IPSPs (a) and EPSPs (b) evoked by contralateral SC stimulation. C: latency histograms of the 1st (a) and 2nd components of EPSPs (b), and IPSPs (c) evoked by ipsilateral Forel's field H (FFH) stimulation. D: latency histograms of IPSPs (a) and EPSPs (b) evoked by contralateral FFH stimulation. E: latency histograms of EPSPs (a) and IPSPs (b) evoked by ipsilateral stimulation of the interstitial nucleus of Cajal (INC). F: latency histograms of IPSPs (a) and EPSPs (b) evoked by contralateral INC stimulation. Abscissas: latencies of PSPs. Ordinates: number of cells recorded.

View larger version (28K): [in this window] [in a new window]   FIG. 3. Pattern of synaptic inputs from the SCs on both sides to trochlear motoneurons. A: experimental setup. Three pairs of medial and lateral monopolar electrodes were arranged rostrocaudally within the ipsilateral (1–3) and contralateral (4–6) SC. Stimulus currents were passed through each pair of electrodes, and lateral electrodes were cathodal for the ipsilateral SC and anodal for the contralateral SC. The retinotopic map is superimposed (McIlwain 1983). B and C: IPSPs and EPSPs in a trochlear motoneuron evoked by stimulation of the ipsilateral (B) and contralateral vestibular nerve (500 µA) (C), respectively. D and E: double shock stimulation (1.5-ms interval) of the ipsilateral (D) and contralateral SC at 500 µA (E). Same neuron as in B and C. Numbers correspond to stimulation sites in the SC shown in A. F and G: effects of stimulus polarity of the ipsilateral (F) and contralateral SC (G) on another trochlear motoneuron. a: same polarity as in A. Cathodes were lateral (Lt. SC) and medial electrodes (Rt. SC). b: opposite polarity. Bottom traces in individual records indicate juxtacellular field potentials. Calibrations in B apply to C–G.

View larger version (45K): [in this window] [in a new window]   FIG. 4. Spatial distribution of effective stimulation sites in the SC for inducing synaptic potentials in 2 (A and B) trochlear motoneurons. A: PSPs elicited from the ipsilateral SC. B: PSPs elicited from the contralateral SC. a: dorsal view of stimulation sites in the SC. Four concentric bipolar electrodes were placed rostrocaudally and mediolaterally. Their approximate positions in the retinotopic afferent map are shown as in Fig. 3. b: IPSPs evoked by stimulation of the ipsilateral vestibular nerve (500 µA). c: EPSPs evoked by stimulation of the contralateral vestibular nerve (500 µA). d: PSPs evoked by stimulation of the SC at 500 µA. Numbers 1–4 and 5–8 correspond to those in Aa and Ba, respectively. Middle traces in B, b—d, show records while passing hyperpolarizing currents (7 nA). Bottom traces in individual records indicate juxtacellular field potentials.

Next we attempted to determine the location of interneurons that mediated excitation and inhibition from the SC to trochlear motoneurons. Last-order premotor neurons terminating on trochlear motoneurons were identified by transneuronal labeling after the injection of WGA-HRP into the trochlear nerve. Retrogradely labeled neurons were darkly stained in the trochlear nucleus contralateral to the injected trochlear nerve. Lightly stained neurons were considered to be transneuronally labeled on the assumption that the WGA-HRP they contained was present because their axon terminals contacted retrogradely labeled trochlear motoneurons. These cells were mainly distributed in the FFH on the same side as the stained trochlear motoneurons and were densely distributed in the INC on both sides (Fig. 5). These findings are consistent with the observations that have been made using transneuronal tracers in the monkey (Horn and Büttner-Ennever 1998). Transneuronally labeled neurons were also observed in the nucleus of the posterior commissure mainly ipsilateral to retrogradely labeled trochlear motoneurons and bilaterally in the vestibular nuclei. Based on these anatomical data, we analyzed the supranuclear pathways that mediate excitation and inhibition from the SC to trochlear motoneurons electrophysiologically.

View larger version (37K): [in this window] [in a new window]   FIG. 5. Distribution of transneuronally labeled last-order premotor neurons terminating on trochlear motoneurons. Wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP) was injected into the left trochlear nerve. Sections at the mesencephalic level are arranged from the rostral (A) to the caudal order (J). Labeled neurons observed in 3 consecutive 50-µm-thick sections are plotted in representative transverse sections at 300-µm intervals at the level of the FFH and INC (A–G), and the trochlear nucleus (H–J). Large dots, transneuronally labeled neurons; small dots, retrogradely labeled trochlear motoneurons. III Nucl., oculomotor nucleus; III n., oculomotor nerve; CP, cerebral peduncle; IC, inferior colliculus; NPC, nucleus of posterior commissure; OT, optic tract; PAG, periaqueductal gray; RF, retroflex bundle; RN, red nucleus; SN, substantia nigra; ST, subthalamic nucleus. Lt. and Rt., left and right.

We examined the effects of blocking FFH activity on the trochlear nerve discharge recorded following SC stimulation. Lateralities in these experiments are described with reference to the side of the recording in the trochlear nerve, which is contralateral to its corresponding cell bodies. Stimulation of the ipsilateral primary vestibular afferents evoked large responses in the trochlear nerve (Fig. 6Bc), which confirmed the viability of trochlear nerve recordings. In contrast, stimulation of the contralateral vestibular nerve did not evoke any response (Fig. 6Bd) due to type I inhibition as shown in Figs. 1C and 3B. Stimulation of the SC with single or double pulses did not evoke any trochlear nerve discharge, and multiple stimulus pulses were always required to evoke visible nerve responses. Multiple stimulation of the contralateral SC evoked a large response in the trochlear nerve (Fig. 6Bb), but multiple stimulation of the ipsilateral SC evoked responses with much smaller amplitudes (Fig. 6Ba). After these responses were recorded as controls, muscimol solution (3 µl; concentration, 1 µg/µl) was injected into the FFH contralateral to the trochlear nerve that was being recorded from (Fig. 6A). Responses of the trochlear nerve evoked by stimulation of the ipsilateral (Fig. 6Ca) and contralateral SC (Fig. 6Cb) gradually decreased and disappeared 5 min after muscimol injection. However, even after SC-evoked responses had disappeared, a large response was still evoked by vestibular stimulation, although its amplitude was slightly decreased (Fig. 6Cc). Similar results obtained in three experiments indicated that the response of the trochlear nerve evoked by stimulation of the SC on either side was mediated via the contralateral FFH (the FFH ipsilateral to cell bodies of the corresponding trochlear nerve).

View larger version (19K): [in this window] [in a new window]   FIG. 6. Effects of muscimol injection into the FFH on trochlear nerve discharge. A: experimental setup. Nerve discharge was recorded from the left trochlear nerve. Muscimol solution (3 µl; concentration, 1 µg/µl) was injected into the right FFH, which was contralateral to the recording site. B and C: discharge of the left trochlear nerve evoked by stimulation of the left (a) and right SC (b), and the left (c) and right vestibular nerve (d) before (B) and after injection of muscimol into the right FFH (C). Stimulation of the right SC evoked much larger nerve discharge of the trochlear nerve than that of the left SC in control. Note that trochlear nerve discharge evoked by stimulation of the right and left SC disappeared after muscimol injection into the right FFH, whereas large nerve responses were still evoked by stimulation of the left vestibular nerve. , the points of divergence of the traces before and after muscimol injection. Stimulus intensities: B, a and b, and C, a and b, 100 µA; B, c and d, and C, c and d, 500 µA.

To examine the projection of tectofugal neurons to the FFH and their spike conduction times, extracellular spikes were recorded in the SC, and tectofugal neurons projecting to the FFH and/or the PPRF were sought in the SC following stimulation of the FFH and the PPRF. Figure 7 shows examples of tectofugal neurons activated from the ipsilateral FFH. These neurons were further examined to determine whether they also sent their axons to the contralateral PPRF and spinal cord. For the neuron in Fig. 7B, extracellular spikes were evoked by stimulation of the ipsilateral FFH at a fixed latency of 1.0 ms. These spikes were regarded as antidromic because they had a fixed latency even at threshold and followed double shock stimuli at 0.7-ms intervals in half of the trials (Fig. 7Bb) and at 0.8-ms intervals in all trials (Fig. 7Bc). This neuron could not be activated by stimulation of the contralateral FFH (Fig. 7Bd), the contralateral PPRF (Fig. 7Be) or the descending MLF at the level of the obex (Fig. 7Bf), even with a stimulus intensity of 500 µA. For the neuron in Fig. 7D, spikes were evoked by stimulation of the ipsilateral FFH at a fixed latency of 0.6 ms at a threshold intensity of 225 µA (Fig. 7Da) and followed double shock stimuli (300 µA) at an interval of 0.6 ms (Fig. 7Db). This neuron was also activated by stimulation of the contralateral PPRF at a latency of 0.7 ms (Fig. 7Dc) and the contralateral descending MLF at a latency of 0.6 ms (Fig. 7De) but was not activated by stimulation of the contralateral FFH (not illustrated). Whenever spikes were evoked from both the PPRF and the descending MLF, we examined the possibility that the activation was due to current spread from the PPRF stimulation site to a stem axon descending as far as the obex by using a spike collision test that was carried out between spikes activated from the contralateral PPRF and MLF (Shinoda et al. 1976, 1977) (Fig. 7D, g–o). Using the values obtained in this spike collision test, the conduction time was calculated along an axon collateral from a stimulation site in the PPRF to a branching point of a stem axon descending as far as the obex, using an equation reported by Shinoda et al. (1976). To be brief, assume that a stem axon of a neuron is divided into two branches projecting to sites A and B, respectively. When this neuron is antidromically activated from sites A and B, the latencies of extracellular spikes evoked from sites A (L_A) and B (L_B) may be calculated as follows

where X_A and X_B are the conduction times between the branching point of the stem axon and the stimulation sites A and B, respectively; X_S is the conduction time between the branching point of the stem axon and its cell body. Using the value obtained in a spike collision test (I_AB), the conduction times of X_B and X_S may be calculated as follows (Shinoda et al. 1976)

where I_AB is the maximal conditioning-test interval when the spikes evoked from site B are blocked by the spikes evoked from site A; R_B is the refractory period measured by double shock stimuli at site B. Using this equation, spike collision tests were performed between spikes evoked from different pairs of stimulation sites (Fig. 7D, g–i, j–l, and m–o). The calculated conduction times along individual parts of the axon are shown in Fig. 7C. For example, the conduction time along the axon collateral in the PPRF from its stem axon was 0.1 ms, indicating that an axon collateral was stimulated in the PPRF rather than its stem axon due to current spread. If a stem axon had been activated by current spread from the PPRF, this conduction time along the axon collateral should have been zero (see the details of the calculation of spike conduction time along an axon collateral and its associated problems in Shinoda et al. 1976, 1977, 1986b). In a similar way, neurons were considered to have an axon collateral to the PPRF when conduction times along their axon collaterals were 0.1 ms. The stimulating electrode tips were histologically confirmed to be located in the PPRF (Izawa et al. 1999). The latencies of antidromic spikes were 0.5–2.9 ms from the ipsilateral FFH (1.0 ± 0.4 ms, n = 107; Fig. 7F), 0.5–2.8 ms from the contralateral PPRF (1.2 ± 0.5 ms, n = 132; Fig. 7G), and 0.5–2.7 ms from the MLF (1.1 ± 0.4 ms, n = 92; Fig. 7H). These tectofugal neurons were recorded at depths of 1.5–3.0 mm from the SC surface. Of the 159 SC neurons that were antidromically activated from the FFH or the PPRF, 19 were antidromically activated only from the FFH, 29 were activated from both the FFH and the PPRF, and 51 were activated from the FFH, the PPRF and the MLF. Of the 114 tectofugal neurons with ascending branches, only 17 (15%) were activated antidromically from the contralateral FFH, and 10 of these (59%) were activated from the FFH on both sides. Based on a comparison of the antidromic latencies of tectofugal neurons projecting to the FFH (Fig. 7F) and the latencies of PSPs evoked by SC stimulation (Fig. 2, A and B), the earliest components of both the EPSPs and IPSPs evoked in trochlear motoneurons by ipsi- and contralateral SC stimulation were considered to be disynaptic.

View larger version (46K): [in this window] [in a new window]   FIG. 7. Antidromic activation of tectofugal neurons in the SC from the FFH, the paramedian pontine reticular formation (PPRF) and the descending medial longitudinal fasciculus (MLF). A: experimental setup. Extracellular spikes were recorded from neurons in the right SC. Stimulating electrodes were placed in the FFH on both sides, the left PPRF at a depth of 2.0 mm from the ventricular surface and 1 mm rostral to the abducens nucleus, and in the left descending MLF near the obex. SO, superior olive; IO, inferior olive; VI n, abducens nerve; XII, hypoglossal nucleus. B: a tectofugal neuron projecting to the ipsilateral FFH. a: antidromic spikes evoked by stimulation of the right FFH at 300 µA (spike threshold, 200 µA). Spikes followed double shock stimuli at 0.7-ms intervals in half of the trials (b) but followed them at 0.8-ms intervals in all trials (c). This neuron did not respond to stimulation of the left FFH (d), the left PPRF (e), or the left descending MLF at 500 µA (f). C: branching pattern of a tectoreticulospinal neuron revealed by spike collision experiment in D. Numbers attached represent conduction times along axon collaterals calculated based on the values obtained by spike collision experiment (see the details of calculation in the text). D: tectoreticulospinal neuron in C projecting to the ipsilateral FFH and the contralateral PPRF and spinal cord. a: stimulation of the right FFH at 225 µA (spike threshold). b: double shock stimulation of the right FFH (300 µA) at 0.6-ms intervals. The latency of spikes evoked at 300 µA was 0.5 ms. c: PPRF stimulation at 70 µA (spike threshold). Latency of antidromic spikes, 0.7 ms. Stimulation at 1.5 times the threshold evoked spikes at a latency of 0.6 ms. d: double-shock PPRF stimulation (120 µA) at a 0.6-ms interval. e: MLF stimulation at 350 µA (spike threshold). f: double-shock MLF stimulation (450 µA) at a 0.6-ms interval. Latency of antidromic spikes, 0.6 ms. Note that partial spikes appeared in b, d, and f instead of full spikes (a, c, and e) because of an initial segment-soma-dendritic (IS-SD) blockage of antidromic spikes. g–o: spike collision experiment. Stimulus intensities: FFH, 300 µA; PPRF, 120 µA; MLF, 450 µA. g–i: MLF stimulation preceded FFH stimulation by 1.6 (g), 1.0 (h), and 0.9 ms (i). j–l: PPRF stimulation preceded FFH stimulation by 1.6 (j), 1.1 (k), and 1.0 ms (l). m–o: PPRF stimulation preceded MLF stimulation by 1.6 (m), 0.8 (n), and 0.7 ms (o). This neuron did not respond to stimulation of the left FFH at 500 µA. E: extracellular spikes of an FFH neuron evoked by stimulation of the ipsilateral SC and trochlear nucleus. a: single shock stimulation of the ipsilateral SC at 230 µA (spike threshold). b: temporal facilitation induced by double shock stimulation of the SC at 220 µA, indicating that these spikes were orthodromic. c: antidromic responses evoked by stimulation of the ipsilateral trochlear nucleus at 320 µA. F–J: latency histograms of antidromic spikes of tectofugal neurons evoked by ipsilateral FFH stimulation (F), antidromic spikes of tectofugal neurons evoked by contralateral PPRF stimulation (G), antidromic spikes of tectofugal neurons evoked by contralateral MLF stimulation (H), ipsilateral SC-evoked orthodromic spikes in neurons of the FFH (I), and ipsilateral trochlear nucleus-evoked antidromic spikes in neurons of the FFH (J).

Effects of stimulation of the FFH on trochlear motoneurons

The results so far showed that trochlear motoneurons were excited through the ipsilateral FFH from the ipsilateral SC. To confirm the monosynaptic connection between the FFH and trochlear motoneurons, we recorded intracellular potentials from trochlear motoneurons while stimulating the FFH on either side. Stimulation of the ipsilateral FFH evoked depolarization in trochlear motoneurons even with a single shock stimulus (Fig. 8Ba). Intracellular injection of Cl^– (Fig. 8Bb) or the passage of a hyperpolarizing current did not change the polarity of the depolarization evoked from the ipsilateral FFH, indicating that the depolarization was an EPSP (Coombs et al. 1955; Eccles 1964). However, late components of the depolarization became larger after Cl^– injection (Fig. 8Bb), indicating that IPSPs were involved in the later part of the depolarization (Fig. 8Ba). Latencies of EPSPs evoked by ipsilateral FFH stimulation ranged from 0.6 to 1.7 ms (1.0 ± 0.2 ms, n = 68; Fig. 2Ca), and latencies of IPSPs evoked by ipsilateral FFH stimulation ranged from 0.9 to 2.4 ms (1.6 ± 0.4 ms, n = 12; Fig. 2Cc). When the stimulus intensities were increased from 100 to 500 µA, EPSPs increased in size and their latencies decreased by 0.2–0.3 ms (Fig. 8D, a–d). Double shock stimulation of the ipsilateral FFH did not produce temporal facilitation of the early EPSPs (Fig. 8E, a and b), indicating that these EPSPs were monosynaptic. Because the latencies of trochlear nucleus-activated antidromic spikes in FFH neurons were 0.3–1.0 ms (Fig. 7J), the latencies of the early components of FFH-evoked EPSPs were considered to be monosynaptic. In some motoneurons, however, the depolarization had a second component (arrow in Fig. 8Fa) and double shock stimulation of the FFH facilitated this second component. The latencies of the second components of these EPSPs ranged from 1.3 to 2.3 ms (1.7 ± 0.3 ms, n = 8; Fig. 2Cb). Because these latencies were 0.8 ms longer than those of the early components of the EPSPs, the second components of the EPSPs were considered to be disynaptic from the ipsilateral FFH.

View larger version (32K): [in this window] [in a new window]   FIG. 8. Synaptic inputs from the FFH to trochlear motoneurons. A: photomicrograph of the frontal section of the midbrain showing the stimulation sites in the FFH on both sides (arrowheads). B and C: PSPs in a trochlear motoneuron evoked by stimulation of the ipsilateral (B) and contralateral FFH at 500 µA (C) before (Ba, C, a and b) and after Cl– injection into the cell (Bb and Cc). Bc and Cd: juxtacellular field potentials. D: effects of stimulus intensity on FFH-evoked PSPs in a trochlear motoneuron. Stimulus intensities: a, 100 µA; b, 200 µA; c, 300 µA; d and e, 500 µA. e, juxtacellular field potentials. E: effects of stimulus number on ipsilateral FFH-evoked PSPs in a trochlear motoneuron. Single (a) and double shock stimuli at 500 µA (b and c). c, juxtacellular field potentials. Dashed line, algebraic sum of the EPSPs evoked by single shock stimulation. Note that the dashed line is occluded by the solid lines of intracellular records, indicating the absence of temporal facilitation. F: 2 components of EPSPs evoked by stimulation of the ipsilateral FFH. Arrow indicates the 2nd component of the EPSPs. b, juxtacellular field potentials. G: spatial facilitation of ipsilateral FFH-evoked monosynaptic EPSPs by preconditioning stimulation of the ipsilateral SC. Records were obtained from the same trochlear motoneuron as in D. a: stimulation of the ipsilateral SC at almost the threshold for evoking disynaptic EPSPs (400 µA). b: small monosynaptic EPSPs evoked by test FFH stimulation at 150 µA. c: preconditioning SC stimulation facilitated FFH-evoked monosynaptic EPSPs, when the SC stimuli preceded the FFH stimuli by 2.0 ms. Dashed line, algebraic sum of the response in G, a and b; vertical dashed line indicates the onset of EPSPs. Bottom traces in individual records, juxtacellular field potentials. d: time course of the facilitation of FFH-evoked EPSPs produced by preconditioning SC stimulation. Abscissa: interval between the preconditioning stimuli of the SC and the test FFH stimuli. Ordinate: means and SDs of the difference in amplitude between the control and conditioned FFH-evoked EPSPs.

To confirm the disynaptic excitatory pathway from the SC to trochlear motoneurons morphologically, we injected WGA-HRP into the left trochlear nerve and dextran-biotin into the right SC in the same animal. Labeled tectofugal axon terminals were observed on the cell bodies and proximal dendrites of many transneuronally labeled FFH neurons contralateral to the injected left trochlear nerve. In the section shown in Fig. 9A, 4 () of 12 transneuronally labeled neurons (both and ) in the right FFH had labeled tectofugal axon terminals on them. Figure 9B shows a photomicrograph of such axon terminals of tectofugal fibers that made contact with the cell body and proximal dendrites of a transneuronally labeled premotor neuron in the FFH indicated by in Fig. 9A. These morphological data confirm that the shortest main pathway from the ipsilateral SC to trochlear motoneurons is disynaptic via last-order premotor neurons in the ipsilateral FFH.

View larger version (44K): [in this window] [in a new window]   FIG. 9. Morphological evidence of the disynaptic excitatory pathway from the SC to trochlear motoneurons. A: distribution of transneuronally labeled premotor neurons in the FFH following transneuronal labeling from WGA-HRP injections of the left trochlear nerve with () and without () associated boutons from dextran-biotin-labeled right tectofugal neurons on a transverse section (75 µm thick). , transneuronally labeled neuron in the right FFH with synaptic contacts of tectofugal axon terminals shown in B. B: photomicrograph of close associations, suggesting synaptic contact, between dextran-biotin-labeled tectofugal axon terminals and the cell body and proximal dendrites (arrowheads) of a transneuronally labeled premotor neuron in the FFH.

Effects of stimulation of the INC on trochlear motoneurons

To determine the pathway that mediates inhibition from the contralateral SC to trochlear motoneurons, synaptic inputs from the INC to trochlear motoneurons were examined by recording intracellular potentials from them while stimulating the INC on either side. Stimulation of the ipsilateral INC evoked depolarization at a latency of 0.8 ms in a trochlear motoneuron, even with a single shock stimulus (Fig. 10A). The amplitudes of the depolarizations increased and their latencies decreased by 0.2–0.3 ms when the stimulus intensities were increased from 50 to 500 µA (Fig. 10A, a–e). In contrast, single shock stimulation of the contralateral INC evoked hyperpolarization at a latency of 1.0 ms in the same trochlear motoneuron (Fig. 10Ba). Double shock stimulation of the contralateral INC did not produce temporal facilitation of the hyperpolarization (dashed line in Fig. 10Bb), suggesting that this hyperpolarization was monosynaptic. Intracellular injection of Cl^– or the passage of a hyperpolarizing current reversed the hyperpolarization evoked from the contralateral INC (Fig. 10Ba) in a depolarizing direction (Fig. 10Bc), although the polarity of the depolarization evoked from the ipsilateral INC remained unchanged (Fig. 10Af), indicating that the depolarization was an EPSP and the hyperpolarization was an IPSP (Coombs et al. 1955; Eccles 1964). The latencies of EPSPs evoked by ipsilateral INC stimulation ranged from 0.6 to 1.5 ms (0.9 ± 0.2 ms, n = 51; Fig. 2Ea), whereas those of IPSPs evoked by contralateral INC stimulation ranged from 0.9 to 1.8 ms (1.2 ± 0.2 ms, n = 55; Fig. 2Fa), indicating that the earliest components of both EPSPs and IPSPs were monosynaptic. In 29 trochlear motoneurons, stimulation of the ipsilateral INC evoked IPSPs following EPSPs (Figs. 2Eb and 10Cb). These IPSPs were confirmed by their reversal to the depolarizing direction in response to passing hyperpolarizing current (Fig. 10Ca). Double shock stimulation of the ipsilateral INC did not produce temporal facilitation of the IPSPs (dashed line in Fig. 10Cc), suggesting that these IPSPs were monosynaptic. The latencies of the IPSPs evoked by stimulation of the INC on the ipsilateral side were shorter than those on the contralateral side by 0.1–0.3 ms. This finding suggested that ipsilateral INC stimulation might activate passing fibers originating from inhibitory neurons in the contralateral INC because injection of a tracer into the INC labeled many neurons in the contralateral INC (Kokkoroyannis et al. 1996; Onodera 1997). In contrast, stimulation of the contralateral INC evoked an EPSP-IPSP complex in only five trochlear motoneurons (Fig. 2Fb).

View larger version (37K): [in this window] [in a new window]   FIG. 10. Synaptic inputs from the INC observed in trochlear motoneurons. A: EPSPs in a trochlear motoneuron evoked by stimulation of the ipsilateral INC. Stimulus intensities: a, 50 µA; b, 100 µA; c, 200 µA; d, 300 µA; e, 500 µA. f: while passing a hyperpolarizing current (4 nA) at a stimulus intensity of 200 µA. g: juxtacellular field potentials at 200 µA. B: IPSPs evoked by stimulation of the contralateral INC at 500 µA in the same trochlear motoneuron as in A. a and b show single and double shock stimulation, respectively. Dashed line, algebraic sum of the IPSPs evoked by single shock stimulation. Note that the dashed line is occluded by the solid lines of intracellular records, indicating the absence of temporal facilitation. c: while passing a hyperpolarizing current (4 nA). d: juxtacellular field potentials. C: IPSPs following EPSPs evoked by ipsilateral INC stimulation (500 µA) in another trochlear motoneuron. a: while passing a hyperpolarizing current (4 nA). b: control records. c: double shock stimulation. Dashed line, algebraic sum of the PSPs evoked by single shock stimulation. Note that the dashed line is occluded by the solid lines of intracellular records, indicating the absence of temporal facilitation. d: juxtacellular field potentials. D: spatial facilitation of INC-evoked monosynaptic IPSPs by preconditioning stimulation of the FFH on the same side. a: stimulation of the contralateral FFH just subthreshold for evoking disynaptic IPSPs (400 µA). b: test stimulation of the contralateral INC evoked small monosynaptic IPSPs at 100 µA. c: preconditioning FFH stimulation (a) facilitated INC-evoked monosynaptic IPSPs (b). The FFH stimuli preceded the INC stimuli by 1.0 ms. Dashed line, algebraic sum of the responses in D, a and b. d: time course of facilitation of INC-evoked IPSPs produced by preconditioning FFH stimulation. Abscissa: interval between the preconditioning FFH stimuli and the test INC stimuli. Ordinate: means and SDs of the difference in amplitude between the control and conditioned INC-evoked IPSPs. E: spatial facilitation of INC-evoked monosynaptic IPSPs by preconditioning stimulation of the SC on the same side. Records were obtained from the same trochlear motoneuron as in A and B. a: stimulation of the contralateral SC just subthreshold for evoking disynaptic IPSPs (400 µA). b: test stimulation of the contralateral INC evoked small monosynaptic IPSPs at 450 µA. c: preconditioning SC stimulation facilitated INC-evoked monosynaptic IPSPs. The INC stimuli followed the second SC stimuli by 1.0 ms. Dashed line indicates the algebraic sum of the responses in E, a and b. Vertical dashed lines in D and E show the onsets of IPSPs; bottom traces in individual records in D and E show juxtacellular field potentials. F: photomicrograph of the frontal section of the midbrain showing the stimulation sites in the INC on both sides (arrowheads).

To determine whether disynaptic inhibition from the contralateral SC to a trochlear motoneuron is mediated by the contralateral INC, we examined the interaction between IPSPs evoked by stimulation of the contralateral SC (Fig. 10Ea) and the contralateral INC (Fig. 10Eb) in a preconditioning-test paradigm (Fig. 10Ec). Combined stimulation of the contralateral SC and the contralateral INC evoked much larger IPSPs in the trochlear motoneuron (Fig. 10Ec) than the algebraic sum of individual responses evoked by preconditioning SC and test INC stimulation (dashed line in Fig. 10Ec). This facilitation of INC-induced monosynaptic IPSPs by the preconditioning SC stimulation indicated that the disynaptic inhibition from the contralateral SC was mediated via the contralateral INC. Similar facilitation was observed in all of the five trochlear motoneurons tested. In addition, disynaptic IPSPs evoked by stimulation of the contralateral FFH were facilitated by preconditioning stimulation of the contralateral SC in three trochlear motoneurons. Facilitation occurred when the FFH stimuli were given at 0–1.5 ms after the SC stimuli. Facilitation that occurred at 0 ms may be attributed to the convergence of inputs from the SC and FFH onto common INC neurons terminating on trochlear motoneurons on the opposite side. In addition, the facilitation that occurred at 1.0–1.5 ms might be attributed to the tectal excitation of FFH excitatory neurons that caused disynaptic inhibition on trochlear motoneurons (see Figs. 2Da and 8C), suggesting the existence of a trisynaptic inhibitory pathway from the contralateral SC to trochlear motoneurons mediated via the contralateral FFH and INC.

Effects of the midline section between the SCs on SC-evoked PSPs in trochlear motoneurons

As previously described, stimulation of the ipsilateral and contralateral SC evoked mainly EPSPs and IPSPs in trochlear motoneurons, respectively (Fig. 1). However, stimulation of the SC evoked an EPSP-IPSP complex in some trochlear motoneurons; i.e., stimulation of the ipsilateral SC evoked IPSPs after the EPSPs, and stimulation of the contralateral SC evoked EPSPs prior to the IPSPs. The earliest latencies of such IPSPs and EPSPs were in the disynaptic range, but the later latencies might be in the trisynaptic range (Fig. 2, Ab and Bb). As previously mentioned in relation to Fig. 7, the projection from the SC to the FFH was mainly ipsilateral in agreement with the previous findings (Grantyn and Grantyn 1982; Kawamura et al. 1974; Moschovakis and Karabelas 1985). In addition, the projection from the FFH to the trochlear nucleus was also mainly ipsilateral because last-order premotor neurons in the FFH terminating on trochlear motoneurons identified by transneuronal transport of WGA-HRP injected in the trochlear nerve were mainly distributed ipsilaterally (Fig. 5). As shown in Fig. 10, INC neurons exert, for the most part, excitation on ipsilateral trochlear motoneurons and inhibition on contralateral trochlear motoneurons. Therefore if excitatory FFH neurons project to the INC on the opposite side and activate the excitatory and inhibitory INC neurons there, this pathway from the FFH to the INC can explain excitation of contralateral trochlear motoneurons and inhibition of ipsilateral trochlear motoneurons by the SC. However, this pathway seems highly unlikely because the injection of biocytin into the INC revealed that retrogradely labeled cell bodies in the riMLF were located mainly ipsilaterally and few cells were observed contralaterally (Kokkoroyannis et al. 1996). On the other hand, there is a commissural connection between the two SCs (Behan and Kime 1996; Edwards 1977; Grantyn and Grantyn 1982; Kawamura et al. 1974; Moschovakis and Karabelas 1985; Munoz and Istvan 1998; Olivier et al. 1998). This commissural connection was considered to be inhibitory (Maeda et al. 1981; Moschovakis and Karabelas 1985). However, our recent study showed that strong commissural excitation exists in the rostral SC, and that tectal output neurons projecting to the ipsilateral FFH provide commissural axon collaterals to the contralateral SC (Takahashi et al. 2005). Therefore the most likely explanation for the excitation of trochlear motoneurons by the contralateral SC and inhibition by the ipsilateral SC is that this activity is due to direct activation of commissural axon collaterals by axon reflex or to commissural synaptic excitation of tectal output neurons that project to the FFH and/or INC. To examine this possibility, a midline section was made between the SCs on both sides and the effects of stimulation of the SC on either side were investigated in 28 trochlear motoneurons in three cats. The midline section extended for 4.0–5.5 mm rostrocaudally, included the rostral half of the SC, and was 4–5 mm deep from the surface of the SC (Fig. 11D). In the control, stimulation of the ipsilateral SC evoked an EPSP-IPSP complex (Fig. 11Bc), and stimulation of the contralateral SC also evoked an EPSP-IPSP complex in a trochlear motoneuron (Fig. 11Bd). After the midline section, stimulation of the ipsilateral SC evoked EPSPs but not IPSPs (Fig. 11Cc), and stimulation of the contralateral SC evoked IPSPs but not EPSPs in the trochlear motoneuron (Fig. 11Cd). Similar tectal inputs were observed in 26 of the 28 trochlear motoneurons tested. Stimulation of the ipsilateral and contralateral vestibular nerves evoked IPSPs and EPSPs in the same trochlear motoneurons in the control case (Fig. 11B, a and b) and after midline section (Fig. 11C, a and b). These vestibular-evoked PSPs in the same trochlear motoneurons assured that IPSPs were not reversed to depolarizations due to the spontaneous diffusion of Cl^–. Therefore this midline-section experiment showed that excitation of contralateral trochlear motoneurons and inhibition of ipsilateral trochlear motoneurons evoked by stimulation of the SC were mediated by tectal output neurons in the SC opposite to the one stimulated, i.e., tectal output neurons opposite to the stimulated SC were antidromically activated through their commissural collaterals or synaptically activated by stimulation of excitatory commissural neurons. These tectal output neurons in turn orthodromically activated excitatory neurons in the FFH and inhibitory neurons in the INC on the same side as the nonstimulated SC. After the midline section between the SCs, the latencies of SC-evoked PSPs were almost equal to or shorter than those with the normal commissural connection between the SCs, and the thresholds for evoking PSPs in trochlear motoneurons by stimulation of the SC tended to be lower than those in the control. These changes may be ascribed to the abolition of commissural inhibition between the SCs (Munoz and Istvan 1998; Takahashi et al. 2005) by midline section, although experiments were performed under anesthesia. In the remaining 2 of the 28 trochlear motoneurons tested, 1 neuron displayed an EPSP-IPSP complex at a long latency by stimulation of the ipsilateral SC and the other neuron displayed EPSPs by stimulation of the contralateral SC. These tectal inputs might be due to some bilaterality in the projection of the SC to the FFH and in the projection of the FFH to the trochlear nucleus.

View larger version (28K): [in this window] [in a new window]   FIG. 11. Effects of a midline section between the 2 SCs on SC-evoked PSPs in trochlear motoneurons. A: experimental setup. A midline section was made between the SCs on both sides as shown in D. B: control records in a trochlear motoneuron before sectioning. a and b: IPSPs evoked by stimulation of the ipsilateral vestibular nerve (a) and EPSPs with following spikes evoked by stimulation of the contralateral vestibular nerve (b), respectively. The peaks of these PSPs and spikes are clipped at this scale. c and d: an EPSP-IPSP complex evoked by stimulation of the ipsilateral SC (c) and contralateral SC (d), respectively. C: PSPs in a trochlear motoneuron after midline sectioning. IPSPs (a) and EPSPs (b) evoked by stimulation of the ipsi- and contralateral vestibular nerve, respectively, were not affected by midline sectioning. Note that IPSPs evoked by stimulation of the ipsilateral SC (c) and EPSPs evoked by stimulation of the contralateral SC disappeared after midline sectioning (d), respectively. Stimulus intensities: 500 µA for both vestibular and SC stimulation before and after midline sectioning. D: photomicrograph of the transverse section of the midbrain showing the parasagittal section between the 2 SCs near the midline (arrowheads), and the locations of stimulating electrodes in the SCs (small arrowheads).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

View larger version (17K): [in this window] [in a new window]   FIG. 12. Summary diagram of neural connections from the SC to trochlear motoneurons. Connections indicated by thick lines are the most direct disynaptic pathways and those indicated by thin lines are trisynaptic pathways. Open and filled neurons indicate excitatory and inhibitory neurons, respectively.