The rôle of tissue organization in the differentiation of embryonic chick neural retina

IT is generally recognized that the micro-environment to which the undifferentiated vertebrate blastomere is subjected during the course of development determines which portions of that cell’s genome will be expressed. The nature of the stimuli provided by the micro-environment is known in only a few cases. Fell & Mellanby (1951) showed that the presence of vitamin A in the nutrient medium on which embryonic chick skin was cultured would suppress the development of keratinizing epithelial cells and elicit the production of mucous secreting cells. In another well known case (Wilde, 1956) phenylalanine added to the culture medium containing urodele ventral ectoderm stimulated these cells to differentiate into dendritic melanocytes.

It has been demonstrated in a number of species that a minimal critical size of tissue or groups of cells is necessary for development and growth to occur (Chalkley, 1945; Peebles, 1897; Muchmore, 1957; Child, 1907; Grobstein, 1955). It thus becomes important to examine the extent to which closely associated groups of developing cells exert mutual influences upon one another in determining the nature of the microenvironment of each individual cell.

In order to attack the problem of cellular differentiation experimentally, a developing system should be used which contains as few cell types as possible, shows a definite and easily recognizable or measurable characteristic (e.g., pigment, a hormone, etc.) and which is known to be under genetic control. By these criteria the retinal layers of the developing chick eye provide suitable experimental material. The neural and pigmented layers are easily separable from one another and the genetic control of the melanosome of the pigmented layer has been established (Moyer, 1961) for mice. Presumably the same principles would apply to the chick.

The technique of tissue dissociation (Moscona & Moscona, 1952) in conjunction with tissue culture methods provides a means of determining the extent to which closely associated developing cells exert mutual influences upon one another. With this system it is possible to alter the micro-environment of developing cells and perhaps thereby elicit metabolic activities not ordinarily exhibited by these particular cells.

Neural retinas were excised from chicks after 6 to 16 days of incubation at 37·5°C. The eggs were obtained from white leghorns of a variety known as Kimber 44 and were purchased from Florida State Hatcheries, Inc. Embryos were decapitated and the eyes removed as quickly as possible and placed in sterile Tyrode’s solution containing 10 per cent horse serum and 100 units of penicillin-G plus 0·1 mg. streptomycin sulfate per ml. An incision was then made around the circumference of the eye approximately halfway between the periphery of the iris and the equator. The pecten was removed and the neural retina transferred to fresh Tyrode’s solution. The neural retinas thus obtained were freed of all adhering fragments of pigmented retina and unwanted cell types, particularly mesenchyme.

Small pieces of the neural retinas were cut into fragments approximately 1·0 mm³ and explanted into Eagles’ minimal essential medium containing 10 per cent, horse serum and allowed to float therein unattached to the vessel’s wall. These will be referred to as intact fragment cultures. Other neural retinas were dissociated into suspensions of single cells by treatment with an enzyme mixture of elastase (prepared according to Rinaldini, 1959) and pancreatin (Nutritional Biochemicals Co.) at concentrations of 3·0 mg. and 2·0 mg./ml., respectively. This empirically determined mixture was found to dissociate the neural retina very effectively. Lilly Number 44 Pancreatin may be substituted for the Nutritional Biochemicals Company product. The enzyme mixture was prepared by dissolving the dried preparations in calcium- and magnesium-free Tyrode’s solution at a pH of 7·6 and 7·8. This enzyme mixture was sterilized by passing it through a millipore filter mounted in a Swinney adapter. The neural retinas were prewashed three times with calcium- and magnesium-free Tyrode’s solution and then suspended in the enzyme mixture at 37°C. (2·0 ml. of enzyme mixture effectively dissociated 100 mg. wet weight of tissue at any age). Gentle agitation dissociated this tissue into a suspension of single cells. The duration of enzyme treatment required for complete dissociation into single cells depended upon the age of the embryo from which the tissue was obtained. This varied from 10 to 20 min. for retinas from 8-day chicks to over 3 hr. in the case of neural retinas obtained from chicks incubated for 17 days.

The dissociated cells were centrifuged at 200 to 500 r.p.m. in the standard clinical table model International centrifuge for 2 to 5 min. The packed cells were then washed twice with calcium- and magnesium-free Tyrode’s solution containing 0·25 per cent, horse serum. Steinberg (1961) found that a small amount of serum in the wash medium will prevent cell damage to a considerable extent. This point will be taken up again in the discussion. The packed cell volume was determined after the final wash by placing the cell suspension in a Hopkins thrombocytocrit tube or other appropriate vessel and centrifuging in the standard International table model clinical centrifuge at 500 r.p.m. for 3 min. The packed cell volume was then read from the graduated stem at the base of the tube into which the cells had been packed. In our hands this has been found to be not only a more consistent measure of the actual amount of cellular material but, on theoretical grounds, to constitute more accurate determination than counting cells in a haemocytometer. The final cell suspension was made up in Eagle’s minimal essential medium with 5 per cent, horse serum, but no antibiotics except in those experiments in which the effects of serum were to be determined. The suspensions were then seeded on to flying cover-slips, placed in small (60 mm. diameter) plastic disposable Petri dishes, covered with Eagle’s minimal essential medium and 10 per cent, horse serum, and incubated at 37°C. in an atmosphere of 4 -5 per cent. CO₂/95·5 per cent, air at 100 per cent, humidity. Cells were examined in the living state by phase contrast and bright field microscopy and tests for cell viability were made by the method of Hanks and Wallace (1958). Tissues were fixed in Bouin’s fixative or 7·0 M paratoluene sulfonic acid and stained with iron hematoxylin and eosin or Feulgen and fast green (Plate, Fig. C).

Mechanical dissociation of the neural retinas into cell suspensions was achieved by forcing the tissue through stainless steel wire gauze (100 meshes/in.), then through a finer stainless steel gauze (230 meshes/in.), and finally the suspension was forced through nylon bolting cloth with a pore size of 25 μ.

Intact fragments of neural retina showed a small but definite halo of outgrowth within 24 hr. after explanation into tissue culture. Fragments larger than 1 to 2 mm³ usually rounded up into vesicles as in the Plate, Fig. A. Although the degree of histogenesis is somewhat less than that in a normal neural retina in situ (see Coulombre, 1955), nevertheless the formation of the typical retinal layers is apparent. The degree of histogenesis seen in Fig. B is more advanced than at the time of explanation, and fairly normal in appearance, despite some degeneration in the inner nuclear and ganglion cell layers. The loosely associated cells seen within the vesicle are quite similar in appearance to cells found in the vitreous humour during normal development.

In contrast to the foregoing, dispersed cell populations of neural retina in culture did not manifest the characteristics exhibited by cells within the organized framework of a tissue mass; cells with axons are conspicuously absent. After 5| hr. in culture, some of the cells have assumed a distinctly different appearance from freshly explanted cells in that a new cytoplasmic structure appeared. Pigmented granules, reminiscent of the melanosomes seen in the pigmented retina and in melanocytes, were synthesized by dissociated neural retina cells as early as 6 hr. after explantation into culture. After 30 hr. in culture virtually all the cells exhibited these new organelles or ectopic melanosomes (Plate, Figs. D and E).

Histochemical tests were performed on the dispersed cells containing these ectopic melanosomes to ascertain whether the pigment was indeed melanin. The pigment was not affected by 6 N HC1 even after 48 hr. of exposure, but 4 N KOH dissolved or degraded the pigment within 30 hr. Alkaline digestion caused a graded change in color from black or dark brown to a reddish hue, then to tan, and finally the granules became entirely devoid of color. The ectopic melanosomes were effectively bleached by 10 per cent H₂O₂ within 24 hr., and also by Lugol’s iodine followed by sodium thiosulfate treatment. Since lipofuschins also exhibit these properties, but in contrast to melanin also fluoresce when irradiated with ultraviolet light, the ectopic melanosomes were examined for fluorescence with ultraviolet irradiation using the Zeiss fluorescence microscope. No fluorescence was seen, thus ruling out the possibility that the pigment was lipofuschin (Pearse, 1960).

It has been demonstrated that tryptophan and phenylthiourea inhibit melanogenesis in amphibian and chick tissues that ordinarily synthesize melanin (Wilde, 1955, 1956; Markert, 1948; Lynn, 1948). These compounds were added to dispersed cell cultures at a final concentration of 1·0 M. Tryptophan caused a pronounced delay of 15 hr. or more in the appearance of pigmented granules in comparison to untreated control cultures; however, the cultures eventually became pigmented. Phenylthiourea failed to repress completely the appearance of the ectopic melanosome, but the intensity of their pigmentation was considerably diminished. The addition of equimolar quantities of phenylalanine to the culture medium reversed the effects of phenylthiourea and tryptophan. While these studies with histochemical and biochemical inhibitors has not critically proven that the pigment seen in the ectopic melanosome is melanin, the evidence strongly suggests that this is the case.

The effect of the age of the embryo from which the neural retinas were obtained was investigated. As might be expected the retinas from older embryos were more difficult to dissociate. Neural retinas from 17-day embryos required over 3 hr. of treatment with the enzyme mixture at 37°C. for dissociation, whereas retinas taken from 5- or 6-day embryos (these were the youngest studied) required less than 15 min. of this treatment to achieve complete dissociation. Retinas removed from embryos incubated for 13 days and 10 days fell between these two extremes with regard to the ease of dissociation; retinas from 13-day chicks required 2 hr. of treatment as opposed to that from 10-day chicks which dissociated within 45 min.

Paradoxically the relationship between the age of the dissociated retinas explanted and the appearance of ectopic melanosomes was the reverse of what one might logically assume to occur. Retinas were taken from embryos of various ages and dissociated as just described. The cell suspensions were all adjusted to contain 0·05 c.c. of packed cells per c.c. of cell suspension; 0·1 c.c. of these suspensions were then spread on flying cover-slips measuring 10 mm. ×20 mm. so that the final concentration was 2·5 × 10⁻⁵ c.c. of packed cells per mm². The coverglasses were placed in small Petri dishes (60 mm. diameter) and covered with 1·9 c.c. of Eagle’s minimal essential medium plus 5 per cent, horse serum. Within 18 hr. the retina cells from the youngest embryos (6 days) had formed a monolayer over the cover-glass. Very faintly hued and quite tiny (less than 1μ diameter) granules were seen in about 30 per cent, of the cells. Cultures explanted from 10-day embryos were very similar to this. Cells explanted from 13-day embryos, on the other hand, did not form a coherent monolayer and the percentage of cells bearing pigment granules, as well as the intensity of hue and size of the granules themselves, were all noticeably increased. Retina cells explanted as dispersed cell populations from 13-, 14- and 17-day-old embryos remained as isolated individuals and small aggregates of several cells. After 18 hr. virtually all of these cells contained ectopic melanosomes that were from 1 to 2 in diameter and definitely more intensely pigmented than those of cells taken from younger embryos.

These cultures were maintained for a total of 5 days without changing the nutrient medium. The percentage of cells bearing pigment granules and the size and intensity of the pigmentation continued to increase slowly in all cultures until the third day after which no further change occurred. The Hanks’ eosin test for viability was applied at the end of 5 days and less than 1 per cent, of the cells observed proved to be nonviable by this criterion.

Intact fragments of neural retina were treated with the dissociating enzyme mixture as described, except that mechanical agitation was held to a minimum. The tissue, although obviously loosened by this procedure, was explanted into culture virtually in its intact condition. Such fragments developed normal retina histological patterns and failed to manifest ectopic melanosomes. Neural retina was dissociated as previously described by means of the dissociating enzyme mixture and then reaggregated into masses of cells by centrifugation at about 100 g. for 2 min. This was done after successively longer periods of maintenance as dispersed cells in suspension. The resultant cell masses were explanted into culture and failed to develop ectopic melanosomes unless the cells were held in suspension for an hour or more. Moreover they reorganized themselves into rosettes and vesicles as described by Moscona (1957). Finally, neural retinas mechanically dissociated into cell suspensions by forcing the tissue through stainless steel mesh and nylon bolting cloth and never exposed to any dissociation enzyme mixture, nevertheless, formed ectopic pigment granules just as did the cells dissociated by enzyme treatment.

Quantitative studies were undertaken to examine the extent to which cells of the differentiating neural retina exert mutual influences upon one another. Neural retina cell suspensions, taken from embryos at 10 days of incubation, when seeded on flying cover-slips at a concentration of 7 × 10⁻⁴ c.c. of packed cells per mm² attachment surface, grew quite readily and formed monolayers. No pigment formation was observed. When such suspensions were seeded out in the same manner at a concentration of 5× 10⁻⁶ c.c. of packed cells per mm² attachment surface, the cells remain isolated and virtually all developed ectopic melanosomes. This system has proved to be reproducible in over 150 experiments. At concentrations between these extremes clumps of cells were observed which in some cases exhibited ‘rosette’ or pallisade organization. These clumps or aggregates were quite frequently interconnected by cytoplasmic processes, presumably axons. In some cases bipolar-like cells were seen connecting two clumps of cells.

In order to ascertain whether the appearance of ectopic melanosomes is an early manifestation of cell damage or approaching cell death, a viability test was employed (Hanks & Wallace, 1958). Considerably less than 1 per cent, of the cells took up the stain, indicating a high degree of viability.

A working hypothesis to account for the production of ectopic melanosomes under dispersed culture conditions was developed. Aberrant or ectopic melanosomes only occur in dispersed cell populations and then only when the cells are sufficiently separated from one another. Under these conditions a considerably larger proportion of the cells’ surface area per unit of volume is exposed to the culture medium and therefore substances may enter and leave the cell with much greater facility. It is proposed that either : (a) some stimulating factor diffuses from the medium into the cell and activates the gene dependent machinery for the expression of tyrosinase, or (b) some inhibiting factor produced by the cell which represses the manifestation of the tyrosinase gene diffuses out of the cell and its intracellular concentration falls below the level necessary to repress the ectopic melanosomes.

The first alternative was tested by enriching the culture medium with various compounds that might be expected to stimulate melanin production, i.e., phenylalanine (Wilde, 1956). These tests were carried out at cell concentrations that repressed ectopic melanosome formation. The results were inconclusive until the effect of varying the concentration of horse serum in the culture medium was considered. When the cell concentration was held constant at 7 × 10⁻⁵ c.c. of packed cells per mm² attachment surface, 1 per cent, or more horse serum in the culture medium allowed axon formation and repressed the appearance of ectopic melanosomes. At a concentration of 0·001 per cent, horse serum or less, no axon formation occurred and ectopic melanosomes were seen in virtually every cell examined. Furthermore, culture medium containing 5 per cent, horse serum, when boiled for 10 min. and passed through an HA millipore filter to remove the proteins, allowed ectopic melanosomes to appear and prevented the formation of cytoplasmic processes. Hypothesis (a) was consequently tentatively discarded on the basis of these findings.

The second alternative, namely, that ectopic melanosome formation is the result of intracellular inhibitors leaking out of the cell, was tested in two types of experiment. In the first, homogenates and extracts of neural retina were added to the culture medium. In the second, medium in which neural retinas had been repeatedly cultured so as to enrich it with the postulated inhibitor were employed. Neither procedure prevented the appearance of ectopic melanosomes.

The primary issue of this investigation is the question of whether the fundamental unit of cyto-differentiation is the individual cell or a group of organized interdependent cells. It has been shown that small fragments of neural retina are capable of normal development to a considerable extent when isolated and cultured in vitro. If this were the only evidence available an erroneous conclusion could easily be drawn, namely, that the developmental fate of individual cells had been fixed at the time of explantation. There are two lines of evidence, however, which refute this interpretation. Moscona (1957,1960) has shown that reaggregated masses of previously dissociated embryonic chick neural retina will form a lentoid structure when cultured under crowded conditions. Under conditions where crowding did not obtain, he demonstrated that these reaggregates form ‘rosettes’ and that the subsequent development of these is characteristic of the neural retina. It has been demonstrated in the present study that, when the embryonic chick neural retina is dissociated and explanted into culture as a dispersed cell population, the cells exhibit a quite unnatural course of differentiation, namely, the production of melanosomes.

On the basis of histochemical evidence presented, it is justifiable to conclude that the pigmented granules contain melanin. The failure of tryptophan to repress the appearance of the ectopic melanosomes completely and indefinitely may be explained on the basis of its being a normal metabolite which may well have been metabolized by the cells within 20 hr. (The fact that equimolar amounts of phenylalanine added to the culture medium completely reversed the effects of phenylthiourea and tryptophan supports the identification of the pigment as melanin.) It may therefore be concluded that the tyrosinase gene has expressed itself in a cell in which this would never have occurred under normal conditions. Although this gene is never normally expressed in cells of the neural retina there is a pathological condition in which melanization has been observed. This disease, retinitis pigmentosa, is under genetic control (Stern, 1949; Guttes, 1953) and can be inherited either as a sex linked recessive or as an autosomal recessive. Apparently two different genes produce the same pathological manifestations.

The mechanism by which the micro-environment of the dispersed populations of embryonic neural retina cell allows the tyrosinase gene to be expressed is not clear. Since the experiments in which the culture medium was enriched with various metabolites, especially phenylalanine, failed to bring about the appearance of ectopic melanosomes, and conversely, because the serum dilution, cell concentration/mm² diminutions, and boiled serum experiments did evoke melanosome appearance, it may be assumed that some intracellular inhibitor of the tyrosinase locus diffuses out of the cell and thereby allows the synthesis of melanin.

The dispersal of a cell population may be a non-specific stimulus which merely damages the cell membrane and allows the inhibitor to escape. This possibility is supported to some extent by the serum dilution and boiled serum experiments, since even a small amount of serum in the medium will protect against cytolysis (Steinberg, 1961). Moreover, the failure actually to demonstrate the proposed inhibitor by enriching the culture medium with retina homogenates and extracts does not entirely disprove this working hypothesis, since such an inhibitor might be unstable.

Further attempts to elucidate the mechanism of the ectopic melanosome phenomenon are in progress using antibiotics, the specific sites of action of which are known, namely, Actinomycin-D, which inhibits messenger RNA synthesis (Goldberg, Rabinowitz & Reich, 1962) and puromycin, which inhibits protein synthesis (Yarmolinsky & De la Haba, 1960).

1. On a cultivé la rétine nerveuse d’embryons de poulet de 6 à 14 jours d’incubation sous la forme soit de fragments intacts, soit de cellules dispersées en une suspension de cellules isolées à la suite d’une dissociation enzymatique. En 10 à 24 heures, 95 à 100 % des cellules cultivées sous forme de populations de cellules dispersées se sont transformées en cellules pigmentées rappelant celles de la rétine pigmentaire, mais non pas identiques à elles. Cette transformation n’a pas eu lieu dans des fragments intacts de rétine nerveuse cultivés de la même manière.

2. On a étudié expérimentalement la nature du mécanisme par lequel se produit cette transformation. Les résultats indiquent qu’un inhibiteur instable de la tyrosinase pourrait être responsable de la différenciation inhabituelle de ces cellules.

The author wishes to express his gratitude to Dr C. L. Markert for his advice and encouragement during the course of this study. He wishes also to extend his thanks to Dr Konrad Keck, Dr Malcolm S. Steinberg and Dr James G. Wilson for their many helpful suggestions and criticisms. The author is especially indebted to Miss Elizabeth A. Ingalls for her very capable technical assistance during parts of this investigation.

The research was supported by a grant from the National Institutes of Health—NIH NB 04497 and by contract AT (30-l)-2194 from the Atomic Energy Commission. This paper is based in part on a dissertation presented to the Graduate School of Johns Hopkins University for the Degree of Doctor of Philosophy, November, 1961.

OP—Outer plexiform layer

IN—Inner nuclear layer

IP—Inner plexiform layer

NF—Nerve fiber layer

ILM—Inner limiting membrane

PG—Pigmented granule

RC—Developing rod and cone cells

ON—outer nuclear layer G—Ganglion layer

VH—This material strongly resembles the vitreous humor seen in sections of the embryonic chick eye stained with iron hematoxylin

OLM—Outer limiting membrane