ABSTRACT
The behaviour of the egg pigment was studied by histological analysis of wild-type and ‘rusty’ embryos and tadpoles of Xenopus laevis as well as by experimental procedures.
The histological analysis of the wild-type animals showed that the various tissues, notably the skin, neural tube, alimentary system and cement gland go through progressive stages of egg pigment migration and concentration at the apical ends of the cells. In the ‘rusty’ mutants the migration and concentration of pigment occur to a slight extent only, the majority of the pigment granules remaining dispersed.
The experiments (tail cultures, squashes of cement gland mucus and of meconium) showed that in wild-type animals the pigment, after migration and concentration, is eliminated from the cells by expulsion. In ‘rusty’ animals, this expulsion does not take place.
Parabiotic tadpoles of a ‘rusty’ wild-type combination possess a coloration corresponding to their genotype. Ectodermal grafts performed at the neurula stage between ‘rusty’ and wild-type embryos develop according to their origin.
The amount of egg pigment found in wild-type and ‘rusty’ tadpoles, and the exceptional case of the cement gland are discussed.
It is concluded that the behaviour of the egg pigment is an active cell-specific process, and that the pigment is eliminated by expulsion. The non-elimination of the egg pigment in the ‘rusty’ mutant, accounting for its characteristic colour, appears to be due to a failure of the expulsion mechanism.
INTRODUCTION
The pigmentation of the amphibian egg is due to the presence of pigment granules which are produced during the later stages of oogenesis. This pigment which we shall call egg pigment is found mainly in the cortex of the egg, the vegetative surface being less pigmented than the animal one. The role of the egg pigment in embryogenesis is still unknown but, like any pigmentation, it certainly has a protective function. Many authors have also considered the possibility of a relation with yolk digestion but as yet no definite role of the pigment in the breakdown of the yolk platelets has been established (Barth & Barth, 1954; Flickinger, 1956; Nass, 1962; Karasaki, 1963; Denis, 1964; Lanzavecchia, 1966).
The behaviour of the egg pigment in Xenopus laevis has not been studied extensively. Nieuwkoop & Faber (1956) mentioned that the pigment disappears from the tissues in the developing embryo. Pigment granules have been observed in the cerebrospinal fluid by Adam (1954), Komnick (1961), Millot & Lynn (1966) and Kordylewski (1969).
We have been led to study the behaviour of the egg pigment in Xenopus after the discovery of a mutation, ‘rusty’, which affects this behaviour but does not otherwise interfere with the normal development of the tadpoles or frogs. The mutant exhibits a reddish brown colour which is due to the persistence of egg pigment in various tissues at times when it has already disappeared from the tissues of the wild-type tadpoles. This persistence of egg pigment is particularly obvious in the tail fin (Fig. 1). The origin and heredity of this mutation have already been described (Uehlinger & Droin, 1969). The comparison of wild-type and ‘rusty’ tadpoles facilitated the interpretation of the results.
Tadpoles of Xenopus laevis at stage 44. The transparent tail fin of the wild-type animal (right) is scarcely visible on the light background, whereas the pigmented tail fin of the ‘rusty’ tadpole (left) is distinct. (Reprinted from the 20th annual report of the Société Suisse de Génétique 1969.) × 7 5.
Tadpoles of Xenopus laevis at stage 44. The transparent tail fin of the wild-type animal (right) is scarcely visible on the light background, whereas the pigmented tail fin of the ‘rusty’ tadpole (left) is distinct. (Reprinted from the 20th annual report of the Société Suisse de Génétique 1969.) × 7 5.
MATERIALS AND METHODS
Wild-type embryos of Xenopus were obtained from our laboratory stock and the ‘rusty’ embryos from two different homozygous ry/ry strains. Opposition and fertilization were induced by injection of gonadotropic hormones. The embryos were reared at room temperature until they reached the required stages. They were then fixed in Zenker’s solution, cut transversely at 7 μm and stained with Mayer’s haemalun. The staging follows the normal table of Nieuwkoop & Faber (1956).
To test the behaviour of the egg pigment the following methods were devised using normal and ‘rusty’ tadpoles.
Tails of stage 39 embryos were cut off and cultured in a slide provided with a perforation which was closed in the lower side with a coverslip sealed on with paraffin wax. The tails were cultured at room temperature in normal Niu & Twitty solution (Flickinger, 1949). Except for the medium, no sterility precautions were taken. Twenty-four to forty-eight hours later the tails were removed and the coverslip was detached from the culture slide and placed upside-down on a normal slide to facilitate microscopic inspection.
The secretion of the cement gland was collected from 20 tadpoles of stages 37–40. These were kept at room temperature in a small dish filled with aquarium water to a height of 1 cm. At these stages the tadpoles adhere to the surface by strands of mucus secreted by the cement gland. If one places a coverslip on the water surface, 24 h later it is covered with mucus strands. The coverslip is then removed and placed on a normal slide.
The meconium (the intestinal waste-product of the embryo) of stage 46–47 tadpoles was collected from Petri dishes containing clear water in which the tadpoles were kept without food for a period of 24 h. The fragments of meconium were put on a slide and squashed with a coverslip.
To test the specificity of the egg pigment parabiosis and grafting experiments were performed on neurulae (stages 20–22) in normal Niu & Twitty solution.
OBSERVATIONS
The following description is concerned with the characteristic features of the egg pigment in several tissues. As melanophores develop normally in ‘rusty’ animals, they will not be mentioned.
The various tissues chosen for analysis are derived from all three germ layers, each showing particular aspects of egg pigment behaviour: skin, neural tube, cement gland, alimentary canal, notochord and muscles. The observations concerning the amount of pigment granules are based on rough approximations established by comparing the sections.
The general condition of the egg pigment from the egg until stage 24 of normal and ‘rusty’ embryos
The pigment is found throughout the whole egg as small granules of variable size, which are easily recognizable in stained and unstained tissues as small round black dots. A large proportion of the pigment is concentrated in the cortical layer, the remainder being dispersed throughout the cytoplasm. During cleavage the cortical pigment becomes located mainly in the prospective ectoderm. After gastrulation, a layer of pigment-bearing ectodermal cells covers the whole embryo. Inside these cells, as well as inside the cells of the mesoderm and endoderm, the pigment granules are dispersed between the numerous yolk platelets, sometimes accumulating around the nucleus. The quantity of egg pigment remains approximately constant during these early stages in all the tissues, except for the cement gland, which from stage 15 onwards exhibits a progressive accumulation of black pigment.
The condition of the egg pigment in various tissues from stage 24 onwards Skin
Normal individuals
At stage 24 the thick outer epithelial cell-layer bears a large number of granules dispersed throughout the cells. Some of them are found to be accumulating near the external cell-wall. In the thinner sensorial cell-layer, the granules are less numerous. At stage 29–30 in the epithelial layer, the pigment condensation against the external cell-walls becomes obvious. At stage 35–36, this condensation continues. In some cells of the sensorial layer one can now also see a fine layer of pigment granules against the apical cell-wall. By stage 39 all the pigment is accumulated against the cell-walls while some cells in the crest of the tail-fin appear devoid of pigment. At stage 41 when the cells have become more flattened and stretched, one begins to notice a distinct decrease in the quantity of pigment granules, many cells being already devoid of them. Some cells show a protruding surface containing a cloud of pigment granules while others seem to extrude the granules into the exterior. The disappearance of pigment progresses both rostrally and dorso-ventrally, the cells of the tail and back already having lost their egg pigment while the majority of the ventral cells are still retaining it. By stage 45 no more pigment granules can be found (Fig. 2).
‘Rusty’ individuals
Until stage 35–36 no significant difference in egg pigment behaviour can be observed between normal and ‘rusty’ animals except that the concentration of pigment against the external cell-wall is less marked. From stage 39 to stage 45 all cells still contain many dispersed pigment granules showing very little concentration at the external cell-walls (Fig. 3). At stage 48 and even after stage 50 a few isolated granules can be found in the epithelium.
Neural tube
Normal individuals
After the closure of the neural folds (stage 21) irregular groups of pigment granules are found throughout the whole neural tube. By stage 24 they have started to aggregate against the wall of the lumen. At stage 29–30 some granules are still dispersed between the yolk platelets but most of them form an irregular black border around the lumen. Some epithelial cells have ruptured walls with pigment pouring out into the neural lumen. A few granules are observed in the cerebrospinal fluid, sometimes forming clusters in the IVth ventricle; this becomes much more obvious by stage 32 when pigment granules can be found throughout the whole canal. At stage 35–36 the egg pigment is almost completely gone from the lumen of the canal and from the neural tissue, except in the tail tip where the elimination of the pigment is slightly delayed. At stage 39 a few pigment granules can still be found but by stage 48 they have completely disappeared (Fig. 4).
+/+ neural tube at stage 41, showing a few isolated pigment granules against the wall of the lumen. On the roof of the neural tube, two heavily pigmented melanophores. × 464.
Neural tube ry/ry stage 41, showing patches of aggregated pigment granules. On the roof of the neural tube some heavily pigmented melanophores. ×464.
‘Rusty’ individuals
Prior to stage 29–30 no differences can be observed as compared with normals. At these stages and until stage 32 the pigment is less strongly condensed against the inner wall of the neural tube and no granules can be observed in the lumen. After stage 32 the amount of pigment granules becomes slowly diluted as the cells multiply. No pigment granules are found in the lumen (Fig. 5).
Cement gland
Normal individuals
The nearly fully differentiated gland consists of very elongated cells bearing an enormous quantity of pigment granules which are more concentrated than in any other tissue. Only the area surrounding the nucleus, in the basal part of the cell, is free of pigment. By stage 29–30 the gland is fully differentiated and mucus secretion has begun. The cells contain even more pigment than before except at their apical ends where it has disappeared. By stage 35–36 the amount of pigment has diminished. In the basal parts of the cells some vacuoles have appeared becoming more numerous as the cement gland begins to show signs of degeneration at stage 41 (Fig. 6). Isolated pigment granules can occasionally be found in the vacuoles. At stage 52 the gland has completely disappeared.
‘Rusty’ individuals
There is no obvious difference as compared with normals until stage 39. From this stage until stage 45 there is always a concentration of pigment in the basal parts of the cells (Fig. 7). During stages 45 to 50 the gland degenerates in an unorganized manner, due to the presence of large pigment clusters.
The alimentary system
Normal individuals
At stage 29–30 the epithelium surrounding the pharyngeal cavity is studded with pigment granules, many of which are concentrated against the wall of the cavity. During stages 35 to 39 some pigment granules can be seen inside the cavity accumulating against its edges. From stage 39 to stage 41 crusts of pigment peel off, forming large clusters of pigment granules inside the cavity (Fig. 8). By stage 45 no more pigment granules are present in the pharynx.
Pharynx of +/+ stage 39 showing a large cluster of pigment inside the cavity. ×464.
With the beginning of the intestinal coiling, pigment granules are found dispersed throughout the whole endoderm mass. As the intestine continues to differentiate (stages 45–47) the pigment moves toward the apical walls of the cells, thus becoming more and more condensed around the lumen. By stage 46–47,the most differentiated parts of the intestine have lost most of their pigment which has passed into the meconium except for the rectum whose epithelium loses its pigment at a much earlier stage (stage 41). In the other parts of the intestine the layer of pigment becomes progressively thinner from stage 48 onwards and finally disappears at stage 52.
‘Rusty’ individuals
Until stage 41 the pigment appears to be more dispersed in the pharyngeal and intestinal epithelium than in normal animals but some condensation against the walls of the cavities can be observed (Fig. 9). No pigment granules are ever observed in the pharyngeal cavity but some rare granules can be found in the meconium. Isolated pigment granules can be observed even at stage 55 in the epithelium throughout the whole alimentary canal.
Notochord and muscles
Normal and ‘rusty’ individuals
Pigment granules are dispersed throughout these tissues showing no obvious condensation.
As vacuolization proceeds in the notochord (stage 29–30) the pigment becomes attached to the cytoplasmic strands and the corners of the cell junctions.
From stage 39 and stage 45 onwards (notochord and muscles respectively) the pigment gradually diminishes, but some isolated granules persist until stage 52.
EXPERIMENTAL RESULTS
(1) Pigment elimination
To test the hypothesis that the pigment is eliminated by some sort of excretion, the following experiments were performed.
(a) Skin
The elimination of pigment from the skin was tested by amputating tails of stage 39 embryos and culturing them. Twenty-four hours later the normal tails have lost their pigmentation and become transparent. The coverslip at the bottom of the culture slide is covered with isolated or clustered pigment granules, their amount being proportional to the number of tails cultured (Fig. 10).
Pigment granules found on the coverslip after culturing a stage 39 +/ + tail for 24 h. × 528.
The culturing of ‘rusty’ tails shows that the epithelial cells of the ‘rusty’ skin do not excrete any pigment. After 24–48 h of culturing (stage 39) the skin becomes transparent but it still contains pigment and exhibits the reddish brown colour typical of ‘rusty’ larvae. The coverslip at the bottom of the culture slide is devoid of pigment granules. Occasionally whole epithelial cells break away as a result of damage inflicted at the time of amputation, liberating a few pigment granules on the coverslip.
(b) Cement gland
The mucus strands secreted by the cement glands of normal tadpoles which were collected on the coverslip are full of black pigment granules as well as a rich microflora (Fig. 11).
Pigment granules attached to mucus strands secreted by the cement gland of a +/+ stage 40 animal. ×464.
The same test failed to show pigment granules in the mucus strands of ‘rusty’ tadpoles.
(c) Intestine
The elimination of pigment from the intestine was tested by taking samples of meconium excreted by normal tadpoles of stage 46–47. Besides abundant yolk debris, the samples contain a large quantity of black pigment granules (Fig. 12).
Meconium of +/+ stage 46-47 containing a large amount of pigment granules. ×183.
Squash preparations of meconium of ‘rusty’ tadpoles were found to contain some black pigment granules. Their number, however, is very small as compared with the amount present in the meconium of normal tadpoles.
(2) Specificity of pigment behaviour
In order to determine whether the factor controlling the pigment behaviour is body-specific, tissue-specific or cell-specific, parabiosis and grafting experiments were carried out.
(a) Parabiosis
Parabiosis was performed between +/ + and ry/ry embryos at stages 20–22. The parabionts were fixed at stages 41–45. Some + / + to + / + and ry/ry to ry/ry pairs served as controls.
All parabionts developed according to their own genotype. In the + / + to ry/ry combination, the skin of the +/+ individual is devoid of pigment while thé skin of its ry/ry partner is full of pigment granules. Morphologically and histologically the junction is easily recognizable (Fig. 13). The brain and the neural tube are without pigment in the + / + partner (Fig. 14), whereas many granules occupy the ry/ry brain (Fig. 15). The same holds for the pharynx and the rectum. However, pigment granules are present throughout the intestinal tract which is partly common to the two partners. Since the pigment does not normally disappear from the intestine of the + / + before stage 47, the different parts cannot be allocated to the + / + or ry/ry parabionts. The muscles and notochord of the two partners still bear pigment granules which is in agreement with our observations on individual + / + and ry/ry tadpoles.
Junction of the skin in a ry/ry to +/+ parabiotic combination at stage 45. The skin of the ‘rusty’ partner contains pigment granules, × 656.
The brain of a + / + parabiont without any pigment granules. On the roof of the brain, two melanophores. × 464.
(b) Grafts
Pieces of ventral ectoderm were grafted from rusty donors on to + /+ host embryos and conversely at stage 20–21.
The .+ / + host larvae develop normally until stage 40. When the yolk granules in the skin begin to disappear, the skin becomes colourless all over the body except in the ventral graft region, which retains a typical ‘rusty’ coloration.
The outlines of the ‘rusty’ grafts are sharply delimited. On the other hand, the ‘rusty’ host larvae develop their typical colour except in the region of the + / + grafts which becomes colourless.
The histological sections show that in those cases in which smooth healing has taken place, the boundary between the host and the graft periderm can be easily recognized. On one side of the boundary the cells are loaded with pigment granules while the cells of the other side are completely devoid of pigment. Figure 16 shows the junction of a +/+ graft with the skin of a ‘rusty’ host. When healing is less perfect some intermingling of host and graft cells takes place, the ry/ry cells always being characteristically full of pigment and the +1 + cells free of pigment.
DISCUSSION
The pigment granules which are abundant in the embryonic skin of most amphibian species have their origin in the oocyte. This pigment called embryonic pigment by Nieuwkoop & Faber (1956) is synthesized in the growing oocyte. According to Balinsky & Devis (1963) the synthesis begins when the yolk platelet precursors become recognizable in the subcortical cytoplasm of oocytes having a diameter of Ca. 300 μm and ceases by the time they attain a diameter of 500 μm. The ultrastructure of the pigment granules has been described by several authors (Dollander, 1954, 1956; Wischnitzer, 1957, 1965, 1966; Wartenberg & Schmidt, 1961; Wartenberg, 1962; Balinsky & Devis, 1963; Karasaki, 1963; Eppig, 1970). However, the mechanism underlying the formation of the egg pigment still awaits complete elucidation; this also holds for the question whether the egg pigment granule is identical with the melanin granule of the melanocyte (Wilde, 1961; McCurdy, 1969; Eppig, 1970).
The amount of egg pigment synthesized in the oocyte is determined by the genotype of the oocyte itself. It has often been observed that the intensity of pigmentation of the eggs laid by any one female is constant but that it varies from one female to another. Blackler & Fischberg (personal communication) grafted primordial germ cells from one Xenopus laevis embryo to another. Host embryos which developed into females spawned donor and host eggs which differed in pigmentation. This experiment in which the Oxford nucleolar mutant (Elsdale, Fischberg & Smith, 1958) was used as marker, clearly showed that the oocyte genotype and not the maternal ovary determines the amount of egg pigmentation. It has further been observed that a phenotypically normal female heterozygous for the ‘rusty’ mutant will lay eggs of uniform pigmentation. When fertilized by a homozygous ry/ry male, 50 % of the eggs will give rise to phenotypically normal embryos and 50% to ‘rusty’ mutants (Uehlinger & Droin, 1969). The normal and ‘rusty’ embryos of this spawning initially must contain the same amount of egg pigment, since it was found by Balinsky & Devis (1963) that pigment formation is restiicted to a certain period during oogenesis. Moreover, our observations on various developmental stages of normal animals indicate that there is probably no renewed synthesis of pigment granules in the embryo. Thus it seems more logical to name this pigment which is present in the embryo ‘egg pigment’ rather than ‘embryonic pigment’.
The only exception in this respect seems to be the cement gland which, during its differentiation, accumulates a larger amount of pigment. Nieuwkoop & Faber (1956) mention a ‘concentration’ of embryonic pigment in the differentiating cement gland. However it seems unlikely that the pigment granules move from one cell into another. The experiments of Holtfreter (1943) who turned parts of the embryo insideout showed that the pigment granules move towards the outer surface of the cell without leaving it. Moreover, in poorly differentiated cells of the oral ‘sucker’ of Hyla regilia embryos treated with actinomycin D, Eakin (1964) observed a reduction in the amount of pigment. This suggests that most of the pigment granules present in the cement gland are synthesized during the differentiation of its cells so that in this case the name ‘embryonic pigment’ is acceptable.
Several authors have observed a decrease in the number of pigment granules during anuran development (Elias, 1937; Sung, 1962; Karasaki, 1963; Millot & Lynn, 1966). In Xenopus laevis, Nieuwkoop & Faber (1956) pointed out that the embryonic pigment in the skin disappears rather fast during a definite period of development (stages 41–44). Adam (1954) observed cells filled with pigment granules floating in the cerebrospinal fluid. Kordylewski (1969) made the same observation but, unlike Adam, he believes that they contain only pigment granules originating from the neural tissue and that there is no additional pigment resulting from synthesis in these cells. Komnick (1961) also described the excretion of the egg pigment into the cerebrospinal canal by the breaking open of the apical walls of the neuro-epithelial cells and the pouring out of the egg pigment together with some yolk platelets. He states that this is the means by which the embryonic brain can clear itself from excess pigment. Our observations are in accordance with his findings but it seems that it is not only an excess but the totality of the egg pigment which is expelled from the neural tissue.
The similarity between pigment movements in neural tissue and in the majority of all other tissues has led us to assume that egg pigment is eliminated by expulsion, as was also suggested by Kordylewski (1969). Our three experiments devised to test this (tail cultures, squashes of cement gland mucus and meconium) confirm the expulsion hypothesis. Additional evidence for pigment expulsion has been provided by H. Kobel (personal communication) in Xenopus mülleri. In this species, the hatching takes place at a later stage than in X. laevis, i.e. after the embryo has normally lost its egg pigment. The intact egg membranes of X. mülleri with the tadpoles inside were observed to have a definite brownish colour; after puncturing the membranes to free the tadpoles, a cloud of brownish material burst into the surrounding medium. Inspection under the microscope revealed that this brown substance consisted of egg pigment granules.
In conclusion we can define three different components in the behaviour of the egg pigment: migration, concentration at the apical border of the cell and expulsion into the exterior or into a lumen. These three components are characteristic and time specific for nearly all the tissues.
In tissues which are not in contact with the exterior, either directly or by a lumen, such as the notochord or the muscles, pigment granules were observed at much later stages than in the other tissues. In these cases, the granules remain more or less dispersed, are neither concentrated nor expulsed, and their density decreases with time in the growing tissues.
An indirect confirmation of the occurrence of the egg pigment movements in wild-type tadpoles is provided by the abnormal behaviour of the pigment in the ‘rusty’ mutants. In embryos homozygous for ‘rusty’, migration and concentration of the pigment do occur but to a much lesser extent. Elimination never takes place (except to a very small extent in the intestine), as was evident from our observations and confirmed by our experiments. Thus the ‘rusty’ colour is the result of the non-elimination of pigment. However, in later stages, the pigment concentration gradually decreases; as the tadpoles grow, the granules are distributed passively among the dividing cells and the ‘rusty’ colour subsequently disappears.
With regard to the yolk platelets, we have observed that the disappearance of the yolk platelets in normal animals occurs shortly before that of the pigment granules. In ‘rusty’ animals no difference in the behaviour of yolk platelets has been observed.
The parabiosis and grafting experiments show that the action of the ‘rusty’ factor is not only tissue-specific but also cell-specific. No influence of ‘nonrusty’ host tissues on the ‘rusty’ effect can be observed. The mechanism controlling the elimination of the egg pigment, the nature of which is still unknown, is thus located within the cells. The sudden onset and speed of the elimination indicates an active process. Since the pigment granules are of maternal origin, the expulsion controlling factor cannot be connected with the pigment granules themselves but must be related to the genotype of the cells. The lack of expulsion observed in the ‘rusty’ mutants could result from a mechanism of inhibition. Further investigations will be necessary to elucidate this problem.
RESUME
L’étude histologique de têtards de Xenopus laevis, sauvages et ‘rusty’, et une série d’expériences ont permis d’analyser le comportement du pigment de l’oeuf.
L’analyse histologique des têtards sauvages révèle que, dans différents tissus tels que la peau, le tube nerveux, la papille et le tube digestif, le pigment présente un mouvement de migration puis de concentration au bord apical des cellules. Chez les mutants ‘rusty’, ces mouvements sont moins prononcés, la majorité des granules de pigment reste dispersée.
Les cultures de queue, les ‘squashes’ du mucus de la papille et de méconium montrent que, chez les têtards sauvages, après migration et concentration, il y a élimination du pigment par expulsion. Cette expulsion n’a pas lieu chez les têtards ‘rusty’.
Dans les cas de parabioses et de greffes effectuées au stade neurula entre têtards sauvages et ‘rusty’, les parabiontes et les greffes se développent selon leur génotype respectif.
C’est un processus actif, spécifique de la cellule qui contrôle le comportement du pigment de l’oeuf. Ce dernier est éliminé par expulsion. La non-élimination du pigment chez les têtards ‘rusty’, leur donnant leur couleur caractéristique, semble résulter d’un défaut du mécanisme d’expulsion.
ACKNOWLEDGEMENTS
We are indebted to Professor M. Fischberg, in whose Department the material was put at our disposal, for his advice and criticism, to Dr J. Faber (Utrecht) for valuable discussions and to the Library of the Hubrecht Laboratory, International Embryological Institute, for bibliographical assistance. We wish to thank Misses M. van Schaik, C. Voll and H. Jutte for technical help, and Miss M. Maye for the photographic work.
This work was supported by the Fonds national suisse de la Recherche scientifique (requêtes no. 4411 and 3.60.68).