Decline in hippocampal theta activity during cessation of locomotor approach sequences: amplitude leads frequency and relates to instrumental behavior.

Hippocampal theta frequency and amplitude decrease as locomotor approach slows and the goal is reached. This study compared the declines of these theta parameters and related them to behavioral events. Theta activity was recorded with bipolar electrodes spanning cornu Ammon, sector 1 or cornu Ammon, sectors 2/3 cell layers of the dorsal hippocampus in 12 rats trained to approach and depress a treadle which exposed a milk dipper. Behavioral events were identified using a video capture system (20-ms sampling) synchronized to the hippocampal recording system (10-ms sampling). Peri-event averages of theta activity were made around the initial paw contact with the treadle, the presentation of the dipper, and the first lick at the dipper. Phase relationships between averaged hippocampal slow wave activity and behavioral events occasionally were found but they were inconsistent. In averages of both amplitude and frequency, times of minimum were less variable around paw contact indicating that compared with reward presentation and consummatory behavior, it more closely related to the processes determining the declines. Theta amplitude declined more rapidly than frequency and reached an earlier minimum in averages around initial paw contact and dipper presentation. Mean amplitude minimum occurred after the paw contact at 159 ms but the decline of frequency continued into the licking bout with its minimum occurring at 343 ms. The findings indicate that during the termination of approach locomotion, the amplitude of hippocampal theta activity is closely related to specific expected sensorimotor events.

Hippocampal theta activity and behavioral sequences in a reward-directed approach locomotor task.

Hippocampal rhythmic slow wave activity (theta) has been implicated in the processing of stimuli associated with movement. This study determined whether the theta rhythm showed phase relationships or changes in amplitude and frequency with the onset of stimuli and behavioral sequences in a skilled locomotor approach task. Rats with bipolar electrodes spanning CA1 approached a stall, turned to enter it, approached and depressed a treadle, waited 1.35 s, and approached a milk reward located forward either to the right or to the left. Auditory cues indicated the location of the reward during the waiting period and at the reward onset. A video capture system (20-ms sampling) was synchronized to the hippocampal recording system (10-ms sampling). Behavioral events identified by motion analysis were used to generate averages of hippocampal slow wave activity, theta peak amplitudes, and intervals between peaks. Theta activity at 8-10 Hz was almost continuous during the behavioral sequences. Phase relations with stimuli or movement onsets occurred infrequently and were not consistent across the four subjects. Theta peak amplitude and frequency decreased as the rat slowed locomotion in the stall and reached the treadle. Onset of locomotion directed to a reward location occurred on a positive peak of averaged theta activity. When locomotion had short latencies, increases in theta frequency appeared after the onset but, when it had longer latencies, frequency increases appeared 200 ms before onset. The results indicate that the execution of instrumental movement modulates both theta amplitude and frequency, and that the preparation for locomotion modulates theta frequency.

Brainstem regions with neuronal activity patterns correlated with priming of locomotor stepping in the anesthetized rat.

Locomotor stimulation in the perifornical hypothalamus produces a transient facilitation of subsequent locomotion, a priming effect, such that stepping to a second train of stimulation occurs with a shorter latency of onset and increased amplitude. Neurons responsible for the initiation of this facilitated stepping presumably respond to locomotor stimulation with a similar priming effect, i.e. either a shorter latency or a larger change in activity rate. This study used anesthetized rats (urethane, 800mg/kg) to compare brainstem regions in terms of the relative rates of occurrence of single neurons that showed both specific responses to locomotor stimulation and also priming effects. Specific responses were characterized by a progressive increase in activity prior to the first step (a Type I pattern). In that they co-varied in time with the increased probability of stepping onset, Type I responses were more specific than Type II responses, which peaked early in the stimulation train several seconds before the onset of stepping. Regions with high proportions of neurons showing Type I responses and priming effects included the anterior dorsal tegmentum lateral to the central gray, the oral pontine reticular nucleus and the medial gigantocellular nucleus. Few Type I neurons showed a modulation of activity related to the step cycle. Type I primed neurons were uncommon in the cuneiform and the pedunculopontine regions, but neurons showing other patterns (decreases and antidromic responses) were relatively prevalent there. The ventral tegmental area was generally unresponsive. The results indicate that stepping elicited by perifornical stimulation in the anesthetized rat is mediated by circuits that differ at midbrain levels from the circuits implicated in other types of locomotion. Two regions, the anterior dorsal tegmentum and the oral pontine reticular nucleus, warrant further attention to determine their possible roles in the initiation of locomotion.

Priming pattern determines the correlation between hippocampal theta activity and locomotor stepping elicited by stimulation in anesthetized rats.

The after-effects of locomotor stimulation are a transient facilitation of locomotor initiation (the priming effect), and a transient increase in hippocampal rhythmic slow activity in the 3-6 Hz band of the theta range. The similar time course of the two effects suggests that hippocampal 3-6 Hz activity may be linked to the excitability of locomotor initiation. This study tested the hypothesis that power in the 3-6 Hz band that is present prior to stimulation would predict the magnitude of elicited stepping. Stimulation electrodes were implanted in 15 locomotor sites of 10 anesthetized rats (urethane, 800 mg/kg). Hindlimb stepping was elicited by a single control train of electrical stimulation presented once every 62 s. On test trials, a test train at the same intensity followed the control train at varying control/test intervals (15-36 s) to assess the priming effect on stepping. The priming pattern determined whether hippocampal 3-6 Hz power predicted the amount of stepping to be elicited by a stimulation train. Positive correlation (0.47>r>0.22) was found for seven out of eight sites showing positive priming effects. Correlation was absent for three other sites that showed non-significant priming effects and were mixed for four sites that showed negative effects. Sites with positive priming patterns, compared to sites with inconsistent or negative priming patterns, had similar trends in post-stimulation 3-6 Hz power, smaller increases in 6-8 Hz power during the control train and lower 1-3 Hz power during the periods immediately before the control stimulation. For six of 15 sites, regardless of the priming pattern, 1-3 Hz power was inversely related to subsequent stepping, and in three cases provided an independent predictor of stepping. Stimulation at two sites produced discrete episodes of post-stimulation stepping. In one of these cases, a 0.5-Hz increase in peak frequency of hippocampal activity preceded stepping. The results show that the association between hippocampal 3-6 Hz activity and the excitability of locomotor initiation is sufficiently specific to allow prediction of the magnitude of stepping by the prior levels of 3-6 Hz power. However, the occurrence of negative priming effects during prominent 3-6 Hz activity indicates that other factors determine the actual stepping and they can suppress the correlation between theta activity and subsequent locomotion.

Hippocampal theta activity and facilitated locomotor stepping produced by GABA injections in the midbrain raphe region.

Inactivation of neurons in the midbrain raphe region produces increases in locomotor activity, and it appears that they function to suppress locomotion. Inactivation of neurons there also produces hippocampal slow wave (theta) activity and it appears that they also function to inhibit rhythmic activity in the hippocampus. We determined whether the degree of association between the two effects was consistent with the operation of a single mechanism. Stimulation electrodes were implanted into locomotor sites of the hypothalamus of 34 urethane-anesthetized rats. Hindlimb stepping was elicited by 5.12-s trains of perifornical electrical stimulation presented once per minute. Hippocampal theta activity was recorded across the CA1 layer of the dorsal hippocampus. GABA injections were used to locate raphe sites at which neuronal inactivation influenced stepping and hippocampal activity. A glass pipette (80-microm tip) was inserted into the midbrain, and injections of GABA (50-100 mg/0.1-0.2 microl) were made in 70 sites in the midbrain. Injections at 34 sites facilitated stimulation-elicited stepping, and at 17 sites, they also produced intertrial stepping. Facilitating injections, but not ineffective or suppressive injections, increased the mean peak frequency of hippocampal activity, and increased power in the 4-5 Hz band during the period that preceded the stimulation trains, but did not change the 5-6 Hz activity produced during the stimulation trains. Priming locomotor stimulation which also facilitated stepping produced generally similar increases in pre-stimulation peak frequency and 4-Hz power. The magnitudes of the increases in stepping and 4-Hz power were uncorrelated. The increase in 4-Hz power appeared earlier than the increase in stepping in 18 of 34 cases, and later in 11 cases; no increases in 4-Hz power were apparent in five cases. The results indicate that pre-locomotor 4-Hz hippocampal activity in the urethane-anesthetized rat is loosely coupled with facilitated locomotor initiation. Neurons in the midbrain raphe region appear to suppress both processes, but the low degree of association between the magnitudes and onset times of increases in stepping and hippocampal 4-Hz power indicate the operation of multiple mechanisms.

Locomotion and head scanning initiated by hypothalamic stimulation are inversely related.

Stimulation in the hypothalamus elicits locomotor stepping. Before stepping is initiated, head scanning movements occur. We determined the relationships between the latency of locomotor initiation and the number, extent and direction of the head scanning movements. Chronic stimulation electrodes were stereotaxically implanted in and around the hypothalamus of 29 rats. Under awake conditions, 38 locomotor sites were tested in a runway apparatus. Behaviors occurring between the onset of stimulation and the first step were recorded on videotape. Points on the rat were digitized at sampling rate of 6 Hz to produce measures of head angles in the vertical, horizontal, and sagittal planes. The priming paradigm was used with a current selected for each site that was minimally sufficient to produce reliable stepping. In trials at approximately 1-min intervals, a 5-s train of stimulation (the control) was followed by a second train (the test) delivered 5-20 s later. Initiation latency on control trains was strongly correlated with head movement measures. Vertical and lateral head movements were independent of one another. Together, their frequency and extent accounted for 85% of the variance in locomotor initiation latencies. In effective priming trials, when locomotor initiation latencies were reduced on the test train, the frequency and extent of vertical and lateral head movements were also reduced. In non-effective priming trials, when latencies were not reduced, head movements were not reduced. Head scanning and locomotor initiation reflect reciprocal processes.

GABA injected into the anterior dorsal tegmentum (ADT) of the midbrain blocks stepping initiated by stimulation of the hypothalamus.

Previous work showed that the activity rates of certain neurons in the anterior dorsal tegmentum (ADT) of the midbrain correlated with the onset of stepping elicited by hypothalamic stimulation. This study determined if reversible inactivation of the ADT would block locomotion elicited by hypothalamic stimulation of anesthetized rats (urethane, 800 mg/kg). GABA (concentrations 0.25-1.0 mg/microl in saline) were injected in 52 sites in 21 rats. GABA at volumes of 0.1 or 0.2 microl blocked hindlimb stepping in 18 cases. Locomotor blocks occurred within 5 min of the injection, and typically recovered within 10-20 min. The effective blocking sites were clustered around the interstitial nucleus of the medial longitudinal fasciculus. Sites more dorsal and more anterior were not as effective as sites in and ventral to this nucleus. The data are consistent with a role for the ADT of the midbrain in locomotor initiation.

Coordination of neck and hindlimb extension during the initiation of locomotion elicited by hypothalamic stimulation.

We tested the hypothesis that during the initiation of stepping elicited by hypothalamic stimulation, hindlimb extension was coordinated with head extension in the sagittal plane. Chronic stimulation electrodes (monopolar stainless-steel, 125 microm diameter) were implanted bilaterally into the perifornical hypothalamus of anaesthetized rats (N = 15) under stereotaxic control. Under freely moving and awake conditions, 18 sites which reliably elicited forward locomotion at a latency of approximately 3 s were tested in a videotaping session. The locomotor stimulation was a constant current train of 5 s duration composed of biphasic pulses at 40-50 Hz. The videotape records were digitized at a sampling rate of 6 Hz for seven points on the rat: Nose, pinnae, midpoint of inter-pinnae line, right forepaw, right hindpaw and base of tail. A characteristic pattern of coordinated movements preceded, by approximately 0.5 s, the execution of the first locomotor step. The pattern included a movement of the pelvis in the anterior or superior direction that was produced by hindlimb extension and an extension of the neck forward along the sagittal plane. There was considerable flexibility in this pattern, but it was invariant to the extent that it occurred at a variety of latencies and after several types of head movements. Associated with the coordinated extensions of the neck and hindlimbs was a lowering of the head angle which had a more variable time course. These data indicate that there is significant coupling between the systems that produce hindlimb extension and control head position when the rat prepares to step.

Priming of locomotor initiation by electrical stimulation in the hypothalamus and preoptic region in the anesthetized rat.

Electrical stimulation at a locomotor site can prime (i.e., shorten the latency to initiate) stepping elicited by subsequent stimulation of the same or a different site. We tested for the priming effect in representative sites along the medial forebrain bundle, and determined if its magnitude showed regional differences. Rats (n = 20) were anesthetized with Nembutal and held in a stereotaxic apparatus over a wheel. Stepping was detected by accelerometers attached to the hindlimbs. Priming and test trains of stimulation (0.5-ms cathodal pulses, 50 Hz, 25-75 microA, 7-9-s train duration) separated by 20 s were delivered every 90 s. When the priming and test stimulations were applied to the same site, the priming effects were similar along the entire extent of the medial forebrain bundle. When the priming and test sites were different, the priming effect depended on their relative positions. Anterior stimulation primed posterior sites at magnitude comparable to those produced by stimulating the same posterior site. Posterior stimulation primed anterior sites at a level half of that produced by stimulation of the same anterior site. This pattern was found for priming and test sites that were ipsilateral and contralateral. Priming is a general and robust phenomenon with properties that may be useful for studying locomotor initiation pathways.

Neural activity in the midbrain correlated with hindlimb extension initiated by locomotor stimulation of the hypothalamus of the anesthetized rat.

Midbrain neuronal activity that correlated with the initiation of locomotion produced by hypothalamic stimulation was studied. Locomotion was elicited by electrical stimulation in the perifornical hypothalamus of 59 rats anesthetized with Nembutal. The first hindlimb extension indexed stepping onset. Single and multiple neurons were recorded ipsilateral to the stimulation site at 2230 sites in the anterior and posterior midbrain. To classify responses, activity patterns averaged around stimulation onset and around the extension onset were examined. Responses with specific correlations to extension onset were Type I; responses not specifically related to the extension onset were Type II. In the anterior midbrain, 6% of sites were Type I and 8% were Type II. The larger Type I responses were frequent in the anterior tegmentum near the central gray. The relative frequency of Type I patterns in the posterior ventrolateral tegmentum was similar. Other regions showed relatively more Type II responses; they included the ventral tegmental area, and the regions near the superior cerebellar peduncle and the posterior central gray. Regional population profiles showed that during the initiation of locomotion, neurons in the posterior peribrachial region responded early and neurons in the anterior dorsal and the posterior ventrolateral tegmentum responded later. The initiation-related activity of Type I neurons in the anterior and posterior midbrain tegmentum suggest that they warrant further study for a role in locomotor initiation.

Preoptic and hypothalamic neurons and the initiation of locomotion in the anesthetized rat.

Despite its insensate condition and apparent motoric depression, the anesthetized rat can provide useful information about the systems involved in locomotor initiation. The preparation appears to be particularly appropriate for the study of the appetitive locomotor systems and may be more limited for the study of the circuits involved in exploratory and defensive locomotion. In the anesthetized rat, pharmacological evidence indicates that the preoptic basal forebrain contains neurons which initiate locomotor stepping. Mapping with low levels of electrical stimulation indicates, but does not prove, that a region centered in the lateral preoptic area might be the location of these neurons. Several lines of evidence indicate that locomotor stepping elicited by electrical stimulation of the hypothalamus is mediated by neurons in the perifornical and lateral hypothalamus. Locomotor effects of hypothalamic stimulation persist in the absence of descending fibers of passage from the ipsilateral preoptic locomotor regions but are severely impaired by kainic acid lesions in the area of stimulation. Injections of glutamate into the perifornical and lateral hypothalamus elicit locomotor stepping at short latencies. Anatomical evidence suggests that the two regions are components of a network for appetitive locomotion. The recognition that multiple systems initiate locomotion both clarifies and complicates the study of locomotion. It provides a framework that incorporates disparate findings but it also underscores the need for increased attention to behavioral issues in studies of locomotor circuitry.

Double dissociation of spatial and object visual memory: evidence from selective interference in intact human subjects.

A functional dissociation of the spatial and object visual systems was produced by selective interference in intact young adults. Subjects were instructed to remember the location of a dot in a spatial memory test, and the form of an object memory test. As predicted by current notions of dissociable visual systems in the primate, spatial memory was selectively impaired by a movement discrimination spatial task, whereas object memory was selectively impaired by a color discrimination object task.

Microstimulation mapping of the basal forebrain in the anesthetized rat: the "preoptic locomotor region".

Previous studies have indicated that the basal forebrain at the level of the preoptic area contains neurons which participate in the initiation of locomotion. This study attempted to localize those neurons by mapping sites at which 25- and 50-microA stimulation (50 Hz, 0.5 ms cathodal pulses, 10-s trains) initiated hindlimb stepping. Anesthetized rats were held in a stereotaxic apparatus supported by a sling so that stepping movements rotated a wheel. Anesthesia was maintained by periodic injections of Nembutal (7 mg/kg) supplemented by lidocaine injections. Stimulation was applied through 50-70-microns diameter pipettes filled with 2 M NaCl at approximately 1600 sites in the basal forebrain, adjacent thalamus, and striatum. A circumscribed grouping of 25-microA locomotor sites, centered in the lateral preoptic area, defined the preoptic locomotor region. It extended into the ventral bed nucleus of the stria terminalis, the lateral part of the medial preoptic area, the anterior hypothalamic area, the medial and rostral parts of the ventral pallidum, medial substantia innominata, and the horizontal limb of the diagonal band. This general region is known to project to the midbrain locomotor region and the ventral tegmental area; it is proposed to initiate locomotion in service of primary motivational systems. Among the structures generally negative for locomotor sites were the dorsal and ventral striata, septal complex, bed nucleus of stria terminalis, and lateral ventral pallidum and substantia innominata. These findings indicate that low current stimulation applied to a circumscribed area centered in the lateral preoptic area produces locomotor stepping in the anesthetized rat. Whether the activated elements in this preoptic locomotor region are cells or fibers is not yet known. The degree of localization afforded by these findings indicates that the areas that are most likely to contain the mediating elements are quite limited in extent.

Locomotor sites mapped with low current stimulation in intact and kainic acid damaged hypothalamus of anesthetized rats.

To determine whether local neurons mediated the locomotor effects of electrical stimulation of the lateral hypothalamus, kainic acid injections (0.5-1.25 micrograms), intended to destroy neural somata as opposed to fibers of passage, were made unilaterally in the tuberal-posterior hypothalamus of 22 rats. The area of lesion and its contralateral homolog were mapped for locomotor stepping sites in Nembutal-anesthetized rats mounted in a stereotaxic apparatus such that locomotor stepping rotated a wheel. Stimulation (25 and 50 microA, 50 Hz, 0.5-ms cathodal pulses, 10-s trains) was delivered through 50-80 microns glass pipettes filled with 2 M saline. Contralateral to the lesion, locomotor stepping sites were common in the perifornical lateral and medial hypothalamus and less dense in the zona incerta. On the side of the kainic-acid lesion, locomotor sites were generally absent in the central part of the damaged area. If they did appear within the area of lesion, they tended to be near the border with intact tissue. In a few cases, locomotor stepping sites were found centrally located in the lesion amidst widespread loss of somata. In four rats, additional maps of anterior locomotor regions in the preoptic area ipsilateral to the lesion suggested that their descending fibers were largely spared by the kainic lesions. Local neurons appear to be major contributors to the locomotion elicited by electrical stimulation of the lateral hypothalamus, but fibers of passage may also participate.

Locomotor stepping initiated by glutamate injections into the hypothalamus of the anesthetized rat.

Glutamate (50 mM, 50 nl) injected into the tuberal and posterior hypothalamus was tested for capacity to elicit locomotor stepping. Rats (n = 23) were anesthetized with Nembutal and suspended by a sling in a stereotaxic apparatus such that locomotor stepping rotated a wheel. In 61 of 275 sites tested, stepping was initiated by glutamate injections within 60 s. Positive sites were widespread and contained in the lateral hypothalamus, the perifornical area, the dorsomedial nucleus, and the zona incerta. The perifornical and lateral hypothalamic sites were most likely to have locomotor responses and the shortest latencies. These findings indicate that selective activation of hypothalamic neurons as opposed to fibers of passage can initiate locomotion.

Initiation and execution of locomotion elicited by diencephalic stimulation: regional differences in response to nembutal.

At moderate levels of Nembutal, within the anesthetic range, locomotor stepping can be elicited by brain stimulation. We determined if Nembutal (7, 14 and 28 mg/kg) had different effects on locomotion elicited by stimulation at different brain regions. Two regions were compared: the medial forebrain bundle (MFB, 13 sites) and the areas medial and dorsal to it (MED/DORSAL, 20 sites). Locomotion was produced by electrical stimulation (50 microA, 0.5 msec pulses, 10 to 160 Hz) of unrestrained rats in a rotary runway. The latency to initiate locomotion and the time to complete 1 revolution of the rotary were measured. With no drug, MFB locomotion was initiated sooner but took longer to complete than MED/DORSAL locomotion. Nembutal at 7 mg/kg did not affect initiation of MFB or MED/DORSAL locomotion. Nembutal at 14 mg/kg shortened MFB initiations, but this dose prolonged MED/DORSAL initiations. Initiations with both types of sites were blocked with 28 mg/kg. The 7 and 14 mg/kg doses prolonged the locomotor completion times of the MFB sites but not of the MED/DORSAL sites. The results indicate that the response to Nembutal differs qualitatively for locomotion elicited by stimulation of the MFB and locomotion elicited by stimulation of the medial and dorsal hypothalamus. The mechanisms underlying the difference remain to be elucidated; they may relate to nonlocomotor behaviors also elicited by stimulation or to the motivational states reflected in those behaviors.

Latency to initiate locomotion elicited by stimulation of the diencephalon positively correlates in awake and anesthetized rats.

Locomotor stepping can be elicited by brain stimulation at various diencephalic sites under moderate levels of Nembutal. This study determined if locomotor initiation measured under anesthesia provides a valid measure of the intersite factors which determine initiation in the awake condition. We compared the latencies to initiate locomotor stepping elicited by electrical stimulation (50 microA, 0.5-msec pulses, 10 to 160 Hz) by rats tested while awake and unrestrained in a rotary runway or anesthetized and held in a stereotaxic apparatus. In the latter tests, initial anesthesia was provided by Nembutal (25 mg/kg) and 2% halothane and maintenance anesthesia was provided by 7 mg/kg as needed and local injections of lidocaine. For 30 sites in 16 rats, average locomotor initiation latency in the awake condition and the shortest latencies in the anesthetized condition were positively correlated (r = .78). Locomotion at sites with long latencies in the awake condition was frequently blocked in the anesthetized condition, but sites with short latencies were rarely blocked. The results indicate that the shortest locomotor latencies in the anesthetized condition approximate the latencies measured in the awake condition. It is concluded that the anesthetized condition can provide valid initiation measures, but sites with long latencies in the awake condition are prone to depression under anesthesia.

Locomotor stepping elicited by electrical stimulation of the hypothalamus persists after lesion of descending fibers of passage.

Locomotion initiated by electrical stimulation of the lateral hypothalamus could be due to activation of local neurons or of fibers of passage descending from locomotor regions in the basal forebrain. This study mapped hypothalamic sites for electrically elicited locomotion in six rats with electrolytic lesions of the ipsilateral basal forebrain sources of descending fibers of passage. For mapping, anesthetized rats were held in a stereotaxic apparatus supported by a sling so that stepping movements rotated a wheel. Anesthesia was maintained by periodic injections of Nembutal (7 mg/kg) supplemented by lidocaine injections. Stimulation (25 and 50 microA, 50 Hz, 0.5 msec cathodal pulses, 10 sec trains) was applied through 50-80 microns diameter pipettes filled with 2 M saline. In all cases, locomotor stepping could be elicited by stimulation in sites ipsilateral to the lesion at currents of 50 microA or less. In the one case in which 25-microA sites were not found in the lateral hypothalamus, the lesion extended caudally to within 1 mm of the stimulation sites. These findings do not exclude a locomotor role for fibers of passage but they suggest that activation of lateral hypothalamic neurons is sufficient to initiate locomotion.

Midbrain areas required for locomotion initiated by electrical stimulation of the lateral hypothalamus in the anesthetized rat.

Locomotor stepping in the Nembutal-anesthetized rat was elicited by electrical stimulation of either of two sites in the right or left posterolateral hypothalamus. Essential midbrain loci were identified by reversibly blocking the elicited locomotion through local injections of the anesthetic procaine (15%, 0.5 microliter). Two types of critical midbrain sites were found. At ipsilateral block sites (n = 21), procaine blocked only that locomotion elicited by ipsilateral stimulation. These sites could be along the course of a direct descending ipsilateral pathway although a possible bidirectional pathway is not to be excluded. At bilateral block sites (n = 21), procaine blocked locomotion elicited by both ipsilateral and contralateral stimulation. These sites could be involved in functions prerequisite for the initiation of locomotion or in the generation of the stepping pattern. Procaine injections in 35 sites had no effect on locomotion. Ipsilateral and bilateral block sites were intermixed and generally located in regions ventral to the midbrain central gray: chiefly the anterior ventromedial midbrain, the pontis oralis nucleus and the pedunculopontine nucleus. Negative sites were located in both the dorsal and ventral midbrain. Ipsilateral block sites were relatively prevalent in the anterior midbrain, indicating that the locomotor initiation signals are lateralized at this level. Bilateral block sites were more prevalent in the posterior levels, suggesting that the initiation signals are proximal to, or interact with, circuits that have a bilateral influence on locomotion.