Recent studies of visually elicited orienting in the frog Rana pipiens suggest that tectofugal signals important in this behavior relay in the midbrain tegmentum before descending to the spinal cord. They also suggest that the high degree of topographic organization displayed by the retinotectal projection may be less characteristic of other tectal afferent and efferent pathways. To explore these possibilities, we have studied patterns of retrograde and anterograde labelling following multiple and single injections of horseradish peroxidase into the tectum. We have found that the midbrain tegmentum is a major terminal zone for tectal efferent projections. Our material also provided a description of the boundaries of other structures which project to and receive input from the tectum. With this background, we studied topographic organization by analyzing for each structure the distribution of labelling following multiple injections, and comparing it with the label distribution following single injections at tectal loci with known visual field input. Multiple injections produced patchy anterograde and retrograde labelling in the nucleus isthmi, with the number of patches corresponding to the number of tectal sites injected. Single injections produced labelling in restricted regions of the nucleus isthmi, the location of which varied systematically with the location of the tectal injection site. In all other structures studied, labelling was more evenly distributed following multiple injections. In none of these structures could we detect systematic variations in the location of labelling associated with variations in the location of single tectal injection sites, and the labelling following single injections was frequently coextensive with that following multiple injections. We also found no evidence that there exist structures which project to or receive input from particular tectal regions and not others. We conclude that there exist adequate neuroanatomical substrates for a tectotegmentospinal pathway believed to be important for visually elicited orienting in the frog. We also conclude that a high degree of topographic organization is more the exception than the rule in considering tectal connections generally in the frog. Topographic organization was readily apparent in connections related to the nucleus isthmi but not in connections related to any other nonretinal structure.

Electrical stimulation of the optic tectum in many vertebrate species elicits eye, head or body orienting movements in the direction of the receptive field location recorded at the site of stimulation; in the barn owl, tectal stimulation produces short latency saccadic head movements (du Lac and Knudsen 1990). However, the barn owl, like other avians, lacks a direct projection from the tectum to the spinal cord, implying that less direct connections underlie tectally mediated head movements. In order to determine the pathways by which the tectum gains access to spinal cord circuitry, we searched for overlap regions between tectal efferent projections and the locations of cells afferent to the spinal cord. Tectal efferent pathways and terminal fields were revealed by anterograde labeling using horseradish peroxidase (HRP) or tritiated amino acids injected into the optic tectum. Cells afferent to the spinal cord were identified by means of retrograde labeling using HRP, rhodamine, or rhodamine-coupled latex beads injected into the cervical spinal cord. A comparison of results from the anterograde and retrograde labeling experiments demonstrated several areas of overlap. All of the cell groups that both received heavy tectal input and contained a high proportion of cells projecting to the spinal cord were located in the medial half of the midbrain and rhombencephalic tegmentum, and included the red nucleus, the interstitial nucleus of Cajal, the medial reticular formation, the nucleus reticularis pontis giganto-cellularis, and the nucleus reticularis pontis oralis. All of these cell groups receive their tectal input from the medial efferent pathway, one of three major output pathways from the tectum. The other two output pathways (the rostral and the caudal) project to regions containing no more than a few scattered cells that are afferent to the spinal cord. Based on these data and on the functions of homologous cell groups in other vertebrates, we hypothesize that the medial efferent pathway and its brainstem target nuclei are primarily responsible for tectally mediated orienting head movements in the barn owl.

The hypothesis that sound localization and gaze control are mediated in parallel in the midbrain and forebrain was tested in the barn owl. The midbrain pathway for gaze control was interrupted by reversible inactivation (muscimol injection) or lesion of the optic tectum. Auditory input to the forebrain was disrupted by reversible inactivation or lesion of the primary thalamic auditory nucleus, nucleus ovoidalis (homolog of the medial geniculate nucleus). Barn owls were trained to orient their gaze toward auditory or visual stimuli presented from random locations in a darkened sound chamber. Auditory and visual test stimuli were brief so that the stimulus was over before the orienting response was completed. The accuracy and kinetics of the orienting responses were measured with a search coil attached to the head. Unilateral inactivation of the optic tectum had immediate and long-lasting effects on auditory orienting behavior. The owls failed to respond on a high percentage of trials when the auditory test stimulus was located on the side contralateral to the inactivated tectum. When they did respond, the response was usually (but not always) short of the target, and the latency of the response was abnormally long. When the auditory stimulus was located on the side ipsilateral to the inactivated tectum, responses were reliable and accurate, and the latency of responses was shorter than normal. In a tectally lesioned animal, response probability and latency to contralateral sounds returned to normal within 2 weeks, but the increase in response error (due to undershooting) persisted for at least 12 weeks. Despite abnormalities in the response, all of the owls were capable of localizing and orienting to contralateral auditory stimuli on some trials with the optic tectum inactivated or lesioned. This was not true for contralateral visual stimuli. Immediately following tectal inactivation, the owls exhibited complete neglect for visual stimuli located more than 20 degrees to the contralateral side (i.e., beyond the edge of the visual field of the ipsilateral eye). In the tectally lesioned animal, this neglect diminished with time. Unilateral inactivation of nucleus ovoidalis had different effects in three owls. Response error to contralateral sound sources increased for one owl and decreased for two; response error to ipsilateral sources did not change significantly for any. The probability of response to ipsilateral (but not contralateral) stimuli decreased for one owl. The latency of response to ipsilateral (but not contralateral) stimuli increased for one and decreased for another. All of the owls, however, routinely localized and oriented toward ipsilateral and contralateral auditory stimuli with nucleus ovoidalis inactivated.(ABSTRACT TRUNCATED AT 400 WORDS)

To generate behaviour, the brain must transform sensory information into signals that are appropriate to control movement. Sensory and motor coordinate frames are fundamentally different, however: sensory coordinates are based on the spatiotemporal patterns of activity arising from the various sense organs, whereas motor coordinates are based on the pulling directions of muscles or groups of muscles. Results from psychophysical experiments suggest that in the process of transforming sensory information into motor control signals, the brain encodes movements in abstract or extrinsic coordinate frames, that is ones not closely related to the geometry of the sensory apparatus or of the skeletomusculature. Here we show that an abstract code underlies movements of the head by the barn owl. Specifically, the data show that subsequent to the retinotopic code for space in the optic tectum yet before the motor neuron code for muscle tensions there exists a code for head movement in which upward, downward, leftward and rightward components of movement are controlled by four functionally distinct neural circuits. Such independent coding of orthogonal components of movement may be a common intermediate step in the transformation of sensation into behaviour.

The size and direction of orienting movements are represented systematically as a motor map in the optic tectum of the barn owl (du Lac and Knudsen, 1990). The optic tectum projects to several distinct regions in the medial brainstem tegmentum, which in turn project to the spinal cord (Masino and Knudsen, 1992). This study explores the hypothesis that a fundamental transformation in the neural representation of orienting movements takes place in the brainstem tegmentum. Head movements evoked by electrical microstimulation in the brainstem tegmentum of the alert barn owl were cataloged and the sites of stimulation were reconstructed histologically. Movements elicited from the brainstem tegmentum were categorized into one of six different classes: saccadic head rotations, head translations, facial movements, vocalizations, limb movements, and twitches. Saccadic head rotations could be further subdivided into two general categories: fixed-direction saccades and goal-directed saccades. Fixed-direction saccades, those whose direction was independent of initial head position, were elicited from the midbrain tegmentum. Goal-directed saccades, those whose direction changed with initial head position, were elicited from the central rhombencephalic reticular formation and from the efferent pathway of the cerebellum. Particular attention was paid to sites from which fixed-direction saccadic movements were elicited, as these movements appeared to represent components of orienting movements. Microstimulation in the medial midbrain tegmentum elicited fixed-direction saccades in one of six directions: rightward, leftward, upward, downward, clockwise roll, and counterclockwise roll. Stimulation in and around the interstitial nucleus of Cajal (InC; a complete list of anatomical abbreviations is given in the Appendix) produced ipsiversive horizontal saccades. Stimulation in the ventral InC and near the dorsal and medial edges of the red nucleus produced upward saccades. Stimulation in the reticular formation near the lateral edge of the red nucleus produced downward saccades. Stimulation in the ventromedial central gray produced ipsiversive roll saccades. The metrics and kinetics of fixed-direction saccades, but not their directions, could be influenced by stimulation parameters. As such, direction was an invariant property of the circuits being activated, whereas movement latency, duration, velocity, and size each demonstrated dependencies on stimulus amplitude, frequency, and duration. The data demonstrate directly that at the level of the midbrain tegmentum there exists a three-dimensional Cartesian representation of head-orienting movements such that horizontal, vertical, and roll components of movement are encoded by anatomically distinct neural circuits.(ABSTRACT TRUNCATED AT 400 WORDS)