Katharine Tansley

THE FORMATION OF ROSETTES IN THE RAT RETINA

INTRODUCTION. Several conditions of the eye are known to be associated with the production of rosettes in the retina. Most, if not all, of these seem to be formed by complicated foldings of the outer nuclear layer so that the lumen of the rosette is bounded by the external limiting membrane. Probably, retinal rosettes are most often encountered in the tumour known as glioma retinae. As the name implies, this tumour was at one-time thought to be composed of glial cells and therefore to be of ectodermal origin. Flexner (1891) was the first to show definitely that the growth was really produced from the outer layers of the retina and to suggest the name neuro-epithelioma for this type of tumour. He believed that the processes which are sometimes to be seen protruding into the lumina of the rosettes are really undeveloped rods and cones. Wintersteiner (1897) later enlarged on this conception in a detailed discussion of the condition. Rosettes have also been found in the folded retinae of microphthalmic eyes and occasionally at the edges of colobomata.

Electroretinogram during Development of Hereditary Retinal Degeneration in the Dog

DURING an investigation of the electroretinogram of the dog (Parry, Tansley, and Thomson, 1953), some observations were made on Red Irish Setters suffering from hereditary degeneration of the retina. This form of retinal degeneration is inherited as a Mendelian recessive character and is characterized by degeneration of the rod nuclei at a very early age (Parry, 1953). The degeneration results in night-blindness and loss of the electroretinogram in young animals. Later the dogs become totally blind. In normal puppies the eyes open about 10-15 days after birth but no electroretinogram can be recorded until about 21 days when it is normal in general appearance but with a very small b-wave. From this age on the b-wave increases until it reaches its full size between 40 and 50 days (Fig. 1, overleaf). This finding is in accordance with the results on mice reported by Keeler, Sutcliffe and Chaffee (1928), on rabbits by Demirchoglian and Mirzoian (1953), on man by Zetterstrom (1951), and on frogs by Miuller-Linumroth and Andree (1954). In none of these species can an electroretinogram be recorded from very young eyes. We were interested to discover whether an electroretinogram can ever be obtained from a puppy suffering from this type of hereditary retinal degeneration, and if so, how its appearance and subsequent disappearance is related to the progress of the disease as determined by histological examination. For this purpose we investigated a litter, both parents of which were known from previous matings and other criteria (Parry, 1951) to be affected so that all the offspring were homozygous for the retinal condition. There were six puppies in this litter born 60 days after coitus, of which five were first tested 22 days after birth. None produced an undeniable electroretinogram although two (DO 199 and DO 198) may have had a tiny b-wave (Fig. 2). One of the five died and was not examined histologically while another (DO 201) was killed and sections made of its retina. These showed some degeneration among the rod nuclei but no recognizable differentiation of their outer limbs. The remaining three dogs were tested again at 26 days, when DO 199 and DO 198 had definite electroretinograms and the third,

IODOACETATE POISONING OF THE RAT RETINA II. GLYCOLYSIS IN THE POISONED RETINA

REPEATED intravenous injections of sodium iodoacetate produce a characteristic visual cell degeneration in the monkey, cat, and rabbit (Schubert and Bornschein, 1951; Noell, 1952), the resulting histological picture being very similar to that presented in human retinitis pigmentosa (Noell, 1953). Recently, it has been shown that a similar condition can be produced in the rat by careful injection of two doses of 30 mg. iodoacetate per kg. body weight, an interval of 4 to 5 hrs being allowed between the two injections (Graymore and Tansley, 1958a). Simultaneous administration of sodium malate was shown to decrease the mortality rate of animals treated in this manner, and yet increase the severity of the visual cell damage. In this way it is possible to obtain experimental animals in which the visual cell population is almost completely destroyed. As early as 1924, Warburg and his associates suggested that all the elements of the retina may not make an equal contribution to the general biochemical picture (Warburg, Posener, and Negelein, 1924) and since then there have been indications that the nerve and receptor components may possess essentially different patterns of metabolism (Noell, 1952; Sjostrand, 1953; Strominger and Lowry, 1955; Lowry, Roberts, and Lewis, 1956). Noell (1952), on the basis of the differing effects of anoxia and iodoacetate on the electroretinogram of rabbits, believed that iodoacetate acted on some mechanism which was not dependent on oxygen, and suggested that the specific action of this substance on the visual cells might be due to the fact that these cells possessed a predominantly glycolytic form of metabolism. It was felt that a study of glycolysis in retinae from rats treated in the manner described above might help to elucidate both the mechanism of action of iodoacetate on the retina as well as certain aspects of the differential metabolism of this tissue. A preliminary report of part of this work has been published elsewhere (Graymore and Tansley, 1958b).

IODOACETATE POISONING OF THE RAT RETINA: I. PRODUCTION OF RETINAL DEGENERATION

IT is now well established that retinal degeneration can be produced in several species of laboratory animals by intravenous injection of sodium iodoacetate (Schubert and Bornschein, 1951; Noell, 1952; Berardinis, 1953; Karli, 1954; Rabinovitch, Mota, and Yoneda, 1954; Babel and Ziv, 1956). Thus, Noell (1952) described a destruction of the visual cells in rabbits, cats, and monkeys after repeated doses, and drew attention to the close similarity between the experimental condition and retinitis pigmentosa in man (Noell, 1953). Most workers have used rabbits for these studies, but Rabinovitch and others (1954) recorded the same result in chickens. So far the rat has been found to be an unsatisfactory animal for such experiments; Rabinovitch and his co-workers (1954) failed to obtain retinal degeneration in this species after repeated sub-lethal doses. Noell (1952) has remarked on the difficulty of finding a single sub-lethal dose which would produce histologically recognizable damage to the retina. Most of his studies in rabbits involved the use of repeated doses of 15 to 20 mg./kg. body weight (12 mg./kg. in the cat). In this laboratory, however, it has been found possible to produce histological damage in the rabbit by giving a single injection of40 mg./kg. In order to achieve this without killing the animal it was necessary to administer the dosein at least 10 ml. saline over a period of about 20 minutes (Tansley, 1955). The present observations arose out of attempts to produce retinal degeneration in the rat for metabolic studies. We also failed to achieve this in the rat with a single sub-lethal dose and therefore tried other procedures.