The Superior Colliculus and Head Movements in the Cat
The Superior Colliculus and Head Movements in the Cat
Abstract and Keywords
There seems to be little doubt that the superior colliculus of the cat plays a crucial role in head movement. The proof comes largely from a single kind of experiment in which the superior colliculus is stimulated electrically and occasionally chemically. The movements that stimulation generates have long been regarded as close to natural movement. Some believe that stimulating the tectum and protectum led to the animals executing turning movements, and gave to them the term, “visual-grasp reflex.” The experiments by Apter were on anesthetized cats. In this case, crystals of strychnine were place on the exposed superior colliculus. The application of the crystals first caused an increase in the visual evoked potential.
There seems to be little doubt that the superior colliculus of the cat plays a significant role in head movement. The evidence comes largely from a single type of experiment in which the superior colliculus is stimulated, usually electrically (Hess et al., 1946; Schaefer, 1970; Harris, 1980; Roucoux et al., 1980b) and occasionally chemically (Apter, 1946). The movements that such stimulation produces have long been regarded as close to natural movement. Hess et al. (1946) reported that stimulating the tectum and pretectum led to the animals executing turning movements as if to capture an object in movement. Hess et al. (1946), reinforcing what was regarded as the naturalness of these movements, gave to them the name “visual-grasp reflex.” The experiments by Apter (1946) were on anesthetized cats. In this instance, crystals of strychnine were placed on the exposed superior colliculus. The application of the crystals first caused an increase in the visual evoked potential. Then a movement of the eyes was observed directed so as to foveate the flash. Finally body movements commenced, starting with movements of the head. To these early experiments must be added many subsequent experiments, including those by Schaefer (1970), Syka and Radii-Weiss (1971), Roucoux et al. (1980b), and Harris (1980). All these experiments, whether on the conscious or decorticate cat, agree on one point: that electrical stimulation in the superior colliculus of the cat leads to head movement, even at current strengths as low as 2 μA.
Experimental observations on the effects of collicular lesions on head movement in the cat are few. In their seminal experiments, Sprague and Meikle (1965) reported that unilateral collicular lesions led to the appearance of a head tilt. Apart from that there was no deficit that could be specifically attributed to an interference with head motor control. Bilateral ablation of the superior colliculus produced a far more serious deficit, because then the cat was unable to lift its head above the horizontal.
Physiologic Role of the Superior Colliculus in Head Movement
In general, the physiologic role of the superior colliculus has most commonly been seen in light of the concerns of the visual physiologist and the oculomotor physiologist, and rarely from the concerns of the motor physiologist. It should be no surprise, therefore, that most studies of the role of the superior colliculus have been related directly or indirectly to matters connected with visual function. Only occasionally has the superior colliculus been studied by motor physiologists with an interest confined to the organization and control of head movement.
At first glance a motor physiologist would regard the superior colliculus as a very unlikely brain site to play a major role in head movement. As Grantyn and Berthoz (1988) have pointed out and as the literature has repeatedly stressed, there are no direct anatomic connections from the superior colliculus to neck motoneurons. The most direct pathway from the superior colliculus to the spinal cord, the tectospinal pathway, is relatively sparse (Nyberg-Hansen, 1964b; Petras, 1967) and only capable of producing small synaptic effects (Anderson et al., 1971). The major output from the superior colliculus to neck motoneurons is via the reticulospinal system (Anderson et al., 1971; Peterson et al., 1971; Grantyn and Berthoz, 1988), a system that shares with the vestibulospinal (Wilson and Yoshida, 1969a) and the corticospinal (Alstermark et al., 1985) systems the possibility of exerting potent effects on neck motoneurons.
Grantyn and Berthoz (1988) have tried to resolve this and other incongruities by proposing that the tectospinal and reticulospinal systems be considered as a functional entity. One major observation that has led to this viewpoint is that recordings of neuronal activity in the conscious cat have not identified a specific role for the superior colliculus in head movement. There is even abundant evidence that the firing patterns of output neurons of the superior colliculus are not necessarily related to the execution of head movement (Munoz and Guitton, 1985; Grantyn and Berthoz, 1988). The situation is further confused by the fact that although electrical stimulation of the superior colliculus in the cat produces head movements, other movements that have been described include avoidance and defense reactions (Spiegel et al., 1954; Schaefer, 1970).
The evidence from the study of head movements in response to electrical stimulation is basically consistent. However, the movements of the head elicited by electrical stimulation are generally found to be closely integrated with eye movement and therefore part of a gaze change. Indeed, the movements of head and eyes are coordinated, and head movement associated with gaze only seems necessary to compensate for limited movement of the eye within the orbit. If the cat eye cannot move more than 25° in the orbit (Guitton et al., 1980) then the role of head movements elicited from the superior colliculus is seen as simply to extend the range of gaze (Guitton, 1988).
Such a view of head movement has been extended by Keshner et al. (1989), who, on the basis of experiments by Outerbridge and Melvill Jones (1971) and Goldberg and Peterson (1986), concluded that “motions of the head relative to the trunk, however, are primarily directed toward orienting and stabilizing the position of the eyes and head in space.” This statement probably would not win support among all head motor physiologists. There are those who would point out that head movements in the cat engaging in natural behavior include a range of movements in which manifestly “orienting and stabilizing … the eyes and head in space” is not the major concern. Such movements must have been seen by every cat owner. They include bunting, a movement in which the head is butted against an object and that man selfishly accepts as a sign of affection. There are also the many head movements that are involved in socializing and grooming activity and in prey catching, carrying, and eating (Leyhausen, 1979)—movements that are essential in the life of the cat and in which head stabilization is not the primary concern.
Somatosensory Input and Function of the Superior Colliculus
Most consider the function of the superior colliculus to lie in sensory-motor integration (Sparks, 1986; Grantyn and Berthoz, 1988). (p.290) This hypothesis is generally considered only in terms of responses to visual and auditory function. However, there is a nonvisual sensory modality that has significant input to the superior colliculus and that is too often left out of serious consideration. This is the somatosensory input. The existence of a somatosensory input to the superior colliculus has been known for many years (Jassik-Gerschenfeld, 1966; Stein and Arigbede, 1972; Gordon, 1973; Abrahams and Rose, 1975a, 1975b; Stein et al., 1976; Meredith and Stein, 1986). A major characteristic of this input in the cat is a relatively crude somatotopy (Stein et al., 1976; Abrahams et al., 1988a) that does not have the precision observed in other somatotopically organized regions, such as the dorsal column nuclei. Receptive fields are commonly large, often including half of the body (Stein et al., 1976; Abrahams et al., 1988a), and points are often “discordant,” lying in a region that does not fit with the somatotopy (Stein et al., 1976). The situation is further complicated by the fact that although many neurons of the superior colliculus receiving somatosensory input may be activated by trivial stimuli such as hair movements (Stein et al., 1976), the majority of these neurons require for their activation sharp taps that activate both deep and superficial receptors and set up vibrations that can activate distant receptors (Nagata and Kruger, 1979; Abrahams et al., 1988a). Furthermore, unlike most other regions concerned with somatosensory sensation, there is no clear separation of cutaneous and deep input to the superior colliculus (Abrahams and Rose, 1975a, 1975b; Abrahams et al., 1988a).
One other characteristic of the somatosensory input to the superior colliculus is the considerable representation from the cutaneous receptors of the face (Gordon, 1973; Stein et al., 1976). This source of input is not easy to reconcile with theories of collicular function that stress foveation. It is difficult to stimulate cutaneously the face of a cat without the stimulus first being foveated. It is also not possible for the surface of the face itself to be foveated. There are no conceivable movements of the head or the eyes of a cat that can bring the skin of the face into view. However, if the object stimulating the face is relatively stationary then a fixed sequence of events is often seen. There is first an appropriate withdrawal movement of the head followed by those movements of the head and eyes necessary to bring the object into view and into focus. However, if an unexpected facial stimulus can be presented prior to foveation, what is seen is an uncomplicated aversion movement of the head to withdraw from the stimulus.
This has been documented by my colleagues, Drs. Loeb and Richmond, who have closely examined the movements of the cat head following facial tactile stimulation. They have observed unrestrained cats instrumented to record neck muscle electromyographic activity and with movement recorded with orthogonally oriented video cameras. In the inexperienced animal it was found that a light tap to the nose does cause the animal to withdraw its head. The cat then maintains fixation on the object for some time. However, as the animal becomes more experienced and the stimulus is repeated, the animal ceases to pay attention to the object being used to tap the face. Now there is only withdrawal of the head each time the nose is tapped or the object is brought in close proximity to the face. What is now found is that tactile stimulation of the face initiates a reflex head movement of aversion.
There are reasons to suppose that this reflex movement of head aversion to facial stimulation may be a candidate in considering function in the superior colliculus. The abundant facial input to the superior colliculus from the face is characterized by a latency to the superior colliculus that can be as brief as 2 ms (Fig. 45–1), with many responses appearing within 5 ms of stimulation (Abrahams et al., 1988a). This input originates in part from nerves that supply the specialized sensitive cutaneous receptors of the face, the vibrissae and rhinarium (Andres and von During, 1973; Gott-schaldtetal., 1973; Montagna al., 1975; Abrahams et al., 1987). Furthermore, the same input from the trigeminal system that is
Role of the Superior Colliculus in the Trigeminal-Neck Reflex
To begin the analysis, the reflex activation of neck motoneurons by branches of the trigeminal nerve was examined in ketamine/chloralose-anesthetized cats. Does the electrical activity in neck motoneurons that is elicited by infraorbital nerve stimulation have the necessary characteristics to be considered as an electrical analogue of head aversion movements? The reflex excitation of neck motoneurons by trigeminal stimulation was described by Sumino and Nozaki (1977) and called by them the trigemino-neck reflex. For the full expression of the reflex high-threshold nociceptive afferents had to be activated. This finding is at odds with our behavioral data, which implicate low-threshold mechanoreceptors as important receptors for the head aversion reflex. However, our reexamination of the trigemino-neck reflex has shown that the reflex can be produced by low-threshold stimuli when trains of electrical pulses to branches of the trigeminal nerve are used, a finding consistent with the reflex originating with mechanoreceptors served by large myelinated fibers. Furthermore, examination of the motoneurons activated in the reflex showed (p.291)
In subsequent experiments, the trigemino-neck reflex was examined to determine whether the integrity of the superior colliculus was necessary for the elaboration of the reflex. Recordings were made of the trigemino-neck reflex from four neck muscle nerves. Recording electrodes were placed on one biventer cervicis and one splenius nerve on each side of the cat. Sixteen consecutive trigemino-neck reflexes elicited by electrical stimulation of both infraorbital nerves at ten times threshold were then averaged. This averaged reflex was found to be quite stable and could be used for quantitative analysis of effects on the reflex. To do this, the reflex was full-wave rectified and the area of the reflex was computed. This provided a quantitative test situation against which factors influencing the trigemino-neck reflex could be measured.
After establishing the stability of the trigemino-neck reflex, Ringer's solution that had been frozen and crushed was applied to the exposed surface of the superior colliculus. Cooling the superior colliculus in this way causes a temperature fall to about 20°C within a few minutes that reaches 3 mm into the structure. This temperature fall is sufficient to block synaptic action (Andersen et al., 1972; Bénita and Condé, 1972) and thus can be used to reversibly block transmission in the superior colliculus. Such cooling was found to lead to significant effects on the trigemino-neck reflex in the 14 animals examined. Almost always, cooling led to a decrease in the computed size of the reflex, and the reflex returned to close to normal values on rewarming.
These results are consistent with some role for the superior colliculus in the elaboration of the trigemino-neck reflex. To determine if the superior colliculus is essential for the reflex, the structure was ablated in the same animals after allowing time for the trigemino-neck reflex to recover from cooling. In many animals ablation led to changes in amplitude, latency, and duration of the reflex as well as in the overall size of the reflex. However, these changes were transient, and in general the reflex returned to values close to control after about 20 minutes (Fig. 45–2). Only in one of 14 experiments did ablation produce an enduring effect. The ablation experiments therefore do not support the hypothesis that the facial input to the superior colliculus is concerned with the expression of the trigemino-neck reflex. If indeed the trigemino-neck reflex is the electrical analogue of head aversion to facial stimulation then it seems unlikely that the superior colliculus is concerned with head aversion movements following facial stimulation.
The results of our experiment still leave open the question of a suitable hypothesis that will explain the existence of a substantial facial input to the superior colliculus that has ample access to output neurons and thus can cause head movements. Perhaps the answer must first await a more careful analysis of the nature of the sensory input from the face to the superior colliculus. It may be no accident that it is stimulation of the infraorbital nerve that is most effective in eliciting a trigemino-neck reflex (Sumino et al., 1981). The infraorbital nerve supplies the two very specialized cutaneous receptor systems of the cat face: the vibrissae, with their exquisite mechanoreceptive sensitivity (Gottschaldt et al., 1973) even for objects approaching the face, and the rhinarium, a structure in the cat that, as in all fur-bearing animals, has a receptor organization closely resembling that of the ridged fingertips of primates (Montagna et al., 1975; Abrahams et al., 1987). Frequently in exploratory movements of the cat, the rhinarium is moved over the object, an example perhaps of active touch, in which movement of cutaneous receptors over an object is used to improve tactile sensitivity. Maybe it is in this kind of activity, in which sensory and motor systems must be totally integrated, that a role for the superior colliculus in head movement will be found.