The Respiratory Centre in the Tench (Tinca Tinca L.): I. The Effects of Brain Transection On Respiration

Though the concept of a respiratory centre is a very old one, investigations of such a centre, in the vertebrate animals at least, have been confined almost entirely to the mammals. The centre is usually thought of as a group of neurones, situated largely in the bulb, which is responsible for all the respiratory integration. Sensory messages, from stretch and gas-tension receptors, impinge on these neurones and affect the rhythmic activity they produce. These changes in turn, when transmitted to the respiratory effectors via connexions in the cord and the spinal motor neurones, cause modifications of the respiratory movements. The exact anatomical location of such a centre has not been universally agreed on, however, and even its existence has been doubted by some authors. Liljestrand (1953), for example, points out that the possibility of respiratory integration occurring at the spinal level has been almost completely disregarded. In the mammal one of the factors complicating any investigation of the site of respiratory integration is that the major sensory input of the respiratory complex goes into the medulla whilst the motor supply to the respiratory muscles comes from the spinal cord. Consequently, a considerable portion of the central nervous system has to be intact for respiration to continue. In the teleost fish a more compact arrangement appears to exist, with the important respiratory pathways, both afferent and efferent, being carried largely in the Vth and Vllth cranial nerves. It would be interesting, therefore, to see whether integrating neurones exist outside the direct connexions between the sensory and motor nuclei of these cranial nerves.

The technique of making transactions is one of the simplest means of delimiting the respiratory areas of the brain and it has been used a great deal in investigations of the mammalian respiratory centre (Lumsden, 1923; Stella, 1938; Hoff & Breckenridge, 1949). One of the principal limitations of the method is that it is possible for respiratory breakdown to occur when parts of the brain, other than those directly concerned in respiratory integration, are removed. It has been demonstrated that systems existing in the brain stem can affect the performance of quite separately co-ordinated activities, such as the simple reflexes (Magoun, 1944). A system of this sort, though its removal or damage might cause modification of respiration, would not be considered to be part of a respiratory centre. In spite of this difficulty the transection technique is still a valuable method for delimiting those parts of the brain which can co-ordinate the normal breathing movements.

A few transection experiments have been done on elasmobranchs. Hyde (1904) showed that the respiratory centre in the skate is located in the medulla. She claimed that the centre in these forms is segmentally arranged, the units associated with the sensory and motor areas of the VHth, IXth and Xth cranial nerves being capable of independent rhythmic activity when separated by transverse cuts. Springer (1928), working on dogfish, was unable to confirm the segmental independence and found that the respiratory region occupied a much greater area of the medulla. No investigations of this sort have been done on teleosts, though it is clear from some of the results of the early workers (see review by Healey, 1957) that damage to the medulla stops the respiratory movements.

The experiments were done on fifty-six tench, the majority of which were 15-20 cm. in length and 45-60 g. in weight. In a few later experiments, in which the labyrinth reflexes were tested, slightly larger fish were used because mirrors had to be attached to their eyes. The fish were fixed in a holder (Fig. 1) which has proved to be satisfactory in several different types of experiment on fish respiration. The holder consists basically of three elements, the trunk clamp A and the head clamps B and C. The trunk clamp is made of wood, grooves being cut out in the two halves to match the body of the fish. One half of the clamp is fixed to the base plate and the other half is adjustable on the two bolts. The head clamps B and C are retained on the fixed half of the trunk clamp. The clamp C consists of two $18$ in. diameter rods in the ends of which are cut V-shaped notches of a size suitable to fit on to the bony supra-orbital ridges of the fish. The rod fixing on to the left supra-orbital ridge is looped over the cranium and slides on the rod of the right side as the diagram shows. This arrangement leaves the left side of the head free from any obstruction which might interfere with recording apparatus. The operative parts of clamp B are the two pieces of metal which can slide apart on their supporting rods in much the same way as a surgeon’s retractor. In this way the ends of these metal claws can be made to grip the bone of the skull at the edges of the hole made to expose the brain. Both clamps B and C are independently adjustable so that different sizes of fish can be accommodated in the holder. When all three clamps are used the head of the fish can be rigidly fixed, although the respiratory movements are not interfered with in any way. In many experiments, including most of those described in this paper, it was not necessary to hold the head of the fish quite so firmly and in these cases the clamp B was not used.

The holder was fastened to the bottom of a Perspex tank which had a water capacity of 1000 ml., the water being constantly aerated throughout an experiment. All the experiments were done at room temperature (18–20° C.). The movements of the lower jaw and of an opercular flap were recorded on a smoked drum using very fight levers so that the movements were not visibly impeded. As far as possible the records were taken before and after the operation to expose the brain, and then after recovery from each transection so that a clear picture of the normal pattern and the subsequent modifications was obtained.

During the operation to expose the brain, and when subsequently the brain transections were being made, the fish were deeply anaesthetized in o·5−1·0% urethane solution. They were allowed to recover to a lighter level of anaesthesia (0·2 % urethane approx.) when the recordings of the movements were taken. The transections themselves were made with a cataract knife or with mounted razor blade fragments. A Marconi MME 3 cautery was also used to produce the lesions in some cases, though the results were not noticeably different from those obtained by cutting. After transection there was usually a period of shock when the fish did not breathe and during this period water was passed over the gills from a cannula inserted into the buccal cavity. When respiration had ceased permanently as the result of a brain transection the gills were perfused continuously until the end of the experiment.

The experiments involved transections of the nervous system in both the midbrain and the spinal cord—posterior medulla regions. In the figures which show the results, the brain is drawn as seen from the dorsal surface and the transection levels are represented by the transverse lines. The order of the lettering in the diagrams represents the sequence in which the sections were made at the various levels. The parts of the brain involved in these vertical transections can be seen in Fig. 2 which shows the brain from a dorso-lateral aspect.

Sections through the optic region lobe, though causing a considerable loss of blood in some cases, had very little effect on the respiratory rhythm after the initial shock period. Certainly the variations were not outside those that normally occurred in the intact animal which had been deeply anaesthetized and then allowed to recover. Transections were made down to the level of the anterior border of the cerebellum, with surprisingly little change in the respiratory pattern (Fig. 3 a, b). Section at a level lower than this became impossible without damaging the Vth and VIIth cranial nerves, but it was possible to remove the cerebellum completely (Fig. 3c) without affecting respiration. Attempts were made to remove, by suction, the more dorsal parts of the medulla beneath the cerebellum and these were always followed by a considerable change in the respiratory pattern. Usually the breathing movements stopped, but on two occasions rhythmic movements were produced even though lesions were made in this way approximately to the level of the Vth motor nuclei.

Transections in the region of the spinal cord and posterior parts of the medulla were more easily performed without causing undue haemorrhage. High sections of the spinal cord in tench never caused respiratory breakdown. In the mammal such sections disrupt important connexions between the medulla and the respiratory motor neurones and the breathing movements cease. In the teleost fish the only connexions which are disrupted by section of the cord or medulla are those to the spino-occipital efferents and this means that the hypoglossal musculature (m. sternohyoideus in the tench) can no longer function in respiration. This failure had very little effect on the breathing movements as a whole, other muscles being able to participate in opening the mouth (Fig. 4b). Transections of the medulla at the level of the obex (Fig. 4c) and, further forward, in the middle of the exposed part of the IVth ventricle (Fig. 4d) were also possible without affecting the ability of the respiratory complex to produce rhythmic activity. The sections in the region of the IVth ventricle did result in modifications of the breathing rhythm in most cases and in some animals the breathing movements ceased altogether. However, the critical level of transection, after which normal breathing movements ceased in all cases, was at the level of the posterior border of the facial lobe (Fig. 5b). After such a section a considerable change occurred in the movement pattern ; the usual rhythmic activity disappeared completely and the mouth (and to a lesser extent the operculum) usually made quivering movements. This type of movement continued to some extent with higher sections (Fig. 5c), though recovery was often slow and frequently no movements were produced at all.

A very striking feature of the movement pattern, produced after transections had been made in these regions of the posterior medulla, was the appearance of prolonged (5–10 sec.) opercular abductions recurring rhythmically every $12$ to 2 min. (Fig. 4c, dFig. 5b−d). The movements of the mouth during these periods of opercular abduction varied somewhat in different individuals. If rhythmic respiratory movements had not been stopped by the transection then usually the mouth stopped moving in the closed position (Fig. 4 c, d). If, however, the normal breathing had ceased as a result of a posterior facial lobe transection and the mouth was making quivering movements then there was no change in the mouth movements in some individuals, whilst in others the opercular abduction was accompanied by an increase in the intensity of the mouth quivering. With still higher transection the mouth again closed during the opercular transection (Fig. 5 c). The slow rhythm persisted even after transection at the posterior border of the cerebellum (Fig. 5 d).

Because of their position on the brain stem the IXth and Xth cranial nerves were, of necessity, progressively removed during the hind-brain transections. It was possible therefore that the breakdown of respiration or the appearance of the slow abductions (or possibly both) was due to damage to the nerves rather than to the brain itself. Indeed, Powers & Clark (1942) had concluded that in teleost fish the IXth and Xth cranial nerves, particularly the former, were of fundamental importance in the initiation of the respiratory rhythm. These authors suggested that afferent volleys in the nerves converted tonic activity in the respiratory centre to the normal rhythmic pattern. To decide which of the possible explanations was correct, experiments were done in which the IXth and Xth cranial nerves were sectioned before any brain transections were made. The IXth and Xth nerves were approached at their origin from the brain stem via holes made in the skull in the region of the labyrinth. The nerve roots were lifted up on glass hooks and then cut with scissors or knife. This ensured that no fragments were left intact. The only effect that sections of these nerves had on the breathing movements was to increase their amplitude ; the normal respiratory rhythm continued and the prolonged opercular abductions did not appear (Fig. 6b). The amplitude of the movements returned to the normal level over a period of 1-2 hr. Transections at a higher level than the spinal cord did result in the appearance of the abductions (Fig. 6c), and even higher transections stopped breathing in the same animal (Fig. 6d). It is very likely therefore that the activity seen after transection of the posterior medulla is due to the damage caused to the brain itself and not to the removal of any sensory or motor components carried in the IXth and Xth cranial nerves.

It was of interest to see whether the slow rhythmic abductions of the opérenla, like the respiratory movements themselves, were produced by activity in the medulla, or whether they were the result of activity in the higher centres such as the optic tectum. In experiments in which the whole of the fore- and midbrains and the cerebellum were removed it was found that a subsequent transection at the facial lobe level would result in the appearance of the slow rhythm (Fig. 3e). The neurones instrumental in the production of this activity must He in the anterior part of the medulla.

In the mammal it has been shown that certain areas of the brain reticular formation contain neurones having a facilitator or suppressor effect on reflex activity in general (Magoun, 1950). Hoff & Breckenridge (1949, 1954) have proposed that the inspiratory cramps, ensuing after section of the brain at the pontine level, are caused by removal of a generalized suppressor area of the brain stem and are not due to removal of a pneumotaxic centre as proposed by Lumsden (1923) and Pitts (1946). Liljestrand (1953) has extended this concept and suggests that the regions within the medial reticular formation of the bulb, long accepted as the site of the respiratory centre, are themselves generalized facilitator areas. Similarly, the nature of the neural mechanism, which is situated in the facial lobe region of the fish brain and whose removal causes such serious breakdown of the normal breathing rhythm, is not obvious. It may be a vital part of the respiratory centre itself or it may be part of a more generalized system in the reticular formation.

Since any interference with such generalized suppressor or facilitator areas should result in the modification of all reflexes, it should be possible to decide between the possibifities outlined above. The static vestibular-eye reflexes were chosen as being the most suitable for experiments on fish. It must be noted that investigations of the brain-stem reticular formation of the mammal have been largely restricted to examination of its influences on cortical activity or on motor activity from the spinal cord. However, there seems to be no reason to suppose that the eye muscle motor neurones, or the intemeurones having synaptic contact with them, are immune from these influences. One reflex at least, the mammalian blink reflex, which is mediated by neurones within the brain, can be suppressed by bulbar stimulation (Magoun, 1944).

In the present experiments the fish was rotated on its long axis from the normal upright position through about 60° (in 10° steps) to a position of right eye down, left eye up. Mirrors were fixed to the left eye and to the body by means of rubber solution and light levers were used to measure both the angle through which the body was rotated and the angle of the eye deflexion. The deflexions were measured before and after the spinal cord and medulla had been transected at various levels, several measurements being made at each level. There was considerable variation between individuals even with intact central nervous systems. One of the six animals tested showed an angular displacement of 22° ± 7° of the eye relative to the trunk, when the trunk was rotated through 6o°. This was the smallest deflexion measured, and, at the other extreme, one fish showed an eye displacement of 34° ± 3° for the same trunk rotation. After stopping respiratory movements with a transection through the IVth ventricle region, the eye deflexions in each individual were found to be both qualitatively and quantitatively the same as before, provided an adequate supply of water was maintained to the gills. A complicating factor which made the measurement of eye deflexion more difficult after transection at any level was the occurrence of a lot of eye movement particularly in the horizontal plane. The resting position of the eye from which these excursions were made was still obvious and only when it was in this position were the measurements made. Transection caused no enhancement or inhibition of this one reflex which was tested.

The fact that the slow rhythmic abductions can occur concurrently with a normal respiratory rhythm and can then persist when normal breathing has failed demonstrates that the nervous elements producing the two are largely independent. This slow rhythm cannot therefore be considered as a development of the primary respiratory rhythm as is suggested for gasping respiration in the medullary preparation in the mammal (Brodie & Borison, 1957). It must be an expression either of another pre-existing activity, probably much modified by the effects of transection, or of an entirely new pattern of nervous activity. The evidence favours the former of these two possibilities, since the rhythmically occurring cough is an example of such a slow rhythm in the intact animal and examination does reveal some similarities between the cough and the slow abduction. The frequency range over which the two occur is roughly the same, though the cough in the intact animal is usually more frequent than the slow abduction. Furthermore, after low transections when normal breathing is continued, it is sometimes possible to see transitional states between the normal cough and the prolonged opercular abduction (Fig. 3a−c). Finally, it is perhaps significant that, during the initial part of the cough, the operculum opens whilst the mouth closes (Hughes & Shelton, 1958) and the same attitude is usually adopted by these two structures in the prolonged activity after transection. It is suggested, therefore, that the slow abductions represent the activity of neurones situated beneath the cerebellum in the anterior part of the medulla, and concerned in the intact animal with co-ordination of rhythmic coughing movements. The normal, brief cough is not produced unless the lower levels of the medulla are intact, however, and some part of the integrative mechanism necessary for normal activity must be situated here. Moreover, this posterior part of the mechanism is independent of input from the IXth and Xth cranial nerves and so is entirely central in location (Fig. 6b). It is interesting to note in passing that the cough does persist after section of the IXth and Xth cranial nerves, although it is usually thought of as a reflex action initiated by foreign matter on the gills.

Though the transection experiments do not allow an exact anatomical locus to be given to the respiratory neurones, it is clear that these neurones must be contained within the medulla between the transection levels having little or no effect on the respiratory movements. They occur, therefore, below the region where the Vth and Vllth cranial nerves emerge from the brain. A more exact rostral limit to the respiratory neurones cannot be given by this method because these nerves must remain intact for respiration to continue. The caudal limits, on the other hand, can be set more exactly. The experiments on the vestibular-eye reflex show that all reflex activity is not affected by brain transection at the facial lobe level. It is unlikely, therefore, that respiratory failure is due to damage to a generalized suppressor or facilitator area of the medulla. It is also unlikely that this failure is the result of direct injury to the tissue of the brain, causing for example massed injury discharges in neurones unrelated to respiration. Similar injury effects must have been caused by the more posterior brain sections, some of which were very near the critical level, and yet these had no fundamental effect on the breathing rhythm. Application of procaine to the cut surface of a brain, transected at the facial lobe level, had no effect on the random quivering movements of the mouth and opercula until it was present in sufficient concentration to act as a general anaesthetic, when all movements stopped. Furthermore, no indication of the normal rhythm was ever seen, though a fish was kept alive up to 3 hr. after facial lobe transections. During this time an effect due to an injury discharge should have disappeared.

In this case, therefore, respiratory failure after brain transection at the posterior border of the facial lobe is apparently due to removal of part of the system directly involved in respiratory integration. There is very little interference with the sensory or motor pathways of the Vth and Vllth cranial nerves as the nuclei of these nerves are situated in the anterior part of the medulla. The descending ramus of the Vth cranial nerve, and a sensory ramus of the Vllth cranial nerve ending in the large facial lobe, are the only components extending back into the region of the medulla involved in these transections. The descending ramus of the Vth runs back to a secondary gustatory nucleus and is sectioned at lower levels than those stopping the breathing movements. It can also be shown that transections causing respiratory failure need not involve the facial lobe. Furthermore, the respiratory neurones or connexions situated at this critical level are not part of an essential reflex involving the IXth and Xth cranial nerves, as the experiments have shown. Therefore, as direct sensory and motor pathways are not involved, transection in the facial lobe region removes neurones which are part of an intermediate integrating system between the sensory and motor nuclei of the cranial nerves involved in respiration.

These neurones are essential in the production of the normal respiratory rhythm and so are part of what could properly be called a respiratory centre. It is unlikely that the whole of such a centre is removed by transection at the facial lobe level. The centre probably consists of a large number of interacting neurones and removal of a relatively small number of these would be sufficient to cause respiratory failure. The work on mammals would suggest that the neurones are situated in the reticular formation of the medulla. However, the site of the neurones and their extension within the region of the medulla delimited by the transections are problems which can be solved only by the use of other techniques.

I wish to record my thanks to Prof. O. E. Lowenstein, F.R.S., and to Dr G. M. Hughes for valuable discussion and encouragement. Some of the work described was carried out in the Department of Zoology, Cambridge, while the author held a Junior Research Grant from the Department of Scientific and Industrial Research.