ABSTRACT
This paper studies the surface morphology of the developing gonadal ridge in X. laevis between stages 44 and 49 (Nieuwkoop & Faber, 1956). During this period the primordial germ cells (PGCs) move laterally from the dorsal mesentery of the gut to the position of the presumptive gonadal ridge. As they do so the coelomic lining cells lateral to the mesentery differentiate into a specialized, longitudinally orientated band, stretching nearly the full length of the dorsal mesentery on each side. The PGCs migrate beneath this band of cells, which thus becomes the germinal epithelium of the gonadal ridge. We have demonstrated by irradiation experiments that this specialized band of cells can differentiate independently, in the absence of the PGCs.
INTRODUCTION
The early events of gonadal ridge formation in Xenopus, as in other anurans, are known only in outline (Witschi, 1929; Kalt & Gall, 1974; Wylie & Heasman, 1975; Whitington & Dixon, 1975). It is thought, but by no means certain, that the primordial germ cells (PGCs) migrate from the gut along the dorsal mesentery to the posterior body wall. They then migrate laterally to the site of formation of the paired gonadal ridges. The evidence for this sequence of events is largely drawn from work in other species, e.g. the chick (Dubois, 1968) and the mouse (Peters, 1970).
In a previous study (Wylie & Heasman, 1975) the fine structure of PGCs, and the cells with which they associate to form the gonadal ridge, has been described. In this paper the surface morphology of gonadal ridge formation is studied.
MATERIALS AND METHODS
(1) Growth of embryos
Eggs were artificially fertilized to obtain synchronously developing batches of embryos. Embryos were allowed to develop until hatching in Gurdon-modified Barth’s saline (Gurdon, 1968) at 10 x dilution, and thereafter in degassed tapwater at 21 °C.
(2) Preparation for scanning electron microscopy (SEM)
At the desired morphological stage (Nieuwkoop & Faber, 1956) tadpoles were first anaesthetized with low concentrations of MS 222. The abdominal cavity was then carefully opened and, using a hypodermic syringe, the abdominal cavity was washed out with fixative. This prevents the precipitation of coelomic fluid constituents on to the posterior abdominal wall. Two fixatives were found to give good results: Karnovsky’s fluid (Karnovsky, 1965) and 4% glutaraldehyde in 0-2 M phosphate buffer, pH 7-4 (Lofberg, 1974). After fixing overnight the tadpoles were dissected by removing the front and sides of the abdominal cavity, the abdominal contents and the head and tail of the tadpole. A photomicrograph of such dissection is published elsewhere (Wylie & Roos, 1975). The tadpoles were then postfixed in 1% osmium tetroxide, dehydrated through an ethanol series, followed by an ethanol/Freon series, before being critical-point-dried in a Polaron critical-point drying apparatus from liquid CO2. The specimens were mounted on aluminium stubs with UHU glue, coated with gold in a sputter coating unit, and examined in a Cambridge Stereoscan S4-10 scanning electron microscope operated at 10 kV.
(3) Irradiation of fertilized eggs
These were chemically dejellied, using 0·2% papain in 2% cysteine hydrochloride, and placed on a quartz glass strip in a drop of saline over an ultraviolet lamp. The lamp delivers a dose of 518-4 ergs/ mm2/min. A dose of 29000 ergs/mm2 was found to eliminate PGCs from the ensuing embryos. Doses less than this reduced, but did not entirely eliminate, the PGCs.
RESULTS
(1) The median gonadal ridge stage
PGCs were found in the dorsal mesentery, and clumped at the root of the mesentery, between stages 43-45. At this stage the PGCs bulge outwards from the dorsal mesentery, at or near its junction with the posterior body wall (Wylie & Heasman, 1975).
Figure 1 shows a montage of this region on the left-hand side of a tadpole at stage 44. The following observations can be made:
Fig. 1. Scanning electron micrograph montage of stage-44 tadpole showing the junction of the posterior body wall (B) and gut mesentery (M). The PGCs are seen as large bulges (arrowed) beneath the surface epithelium.
Fig. 2. Stage-47/8 tadpole. The free edge of the mesentery (M), where the gut has been removed, is visible. The germ cells are now visible as bulges beneath the newly differentiated germinal epithelium.
Fig. 1. Scanning electron micrograph montage of stage-44 tadpole showing the junction of the posterior body wall (B) and gut mesentery (M). The PGCs are seen as large bulges (arrowed) beneath the surface epithelium.
Fig. 2. Stage-47/8 tadpole. The free edge of the mesentery (M), where the gut has been removed, is visible. The germ cells are now visible as bulges beneath the newly differentiated germinal epithelium.
There are numerous bulges in the dorsal mesentery. These represent PGCs covered over by a thin layer of cytoplasm of the mesentery cells.
The detailed surface morphology of the cells is difficult to see, due to the presence of many small, round projections from the surface of most of the cells. These projections are due to the presence of large yolk granules, which have not been used at this stage, pushing outwards against the surface membranes of the cells. These have been seen many times in material sectioned for light and electron microscopy (Wylie & Heasman, 1975).
The junction of the mesentery and posterior body wall is almost featureless. There is no sign of an indifferent gonadal ridge to which the PGCs are moving, or indeed any organized tissue at all in this region.
(2) The early-paired-gonadal-ridge stage
Figure 2 shows a montage of scanning photomicrographs of the full length of the right side of the dorsal abdominal wall of a tadpole at stage 47-48. The yolk granules have disappeared from the cells in this region now, so that their surfaces can be more easily studied.
The PGCs can now be seen as bulges lateral to the root of the mesentery, in an irregular order down nearly the full length of the abdominal cavity.
The most striking observation is that there is now a highly organized band of somatic cells which covers over the PGCs at the site of the gonadal ridge. These cells are elongated in a cranio-caudal axis and are interdigitated with each other. It has been shown before (Wylie & Heasman, 1975) that the PGCs are inevitably covered at this stage by a layer of somatic cells. However, in conventional cross-sections there is no hint of the complexity and organization of these cells. The somatic cells are presumably the precursors of the germinal epithelium of the gonad at later stages.
Figure 3 shows a rather more detailed view of the gonadal ridge at this stage. Figures 4-6 show PGCs, covered with somatic cells, from progressively more caudal areas of the gonadal ridge shown in Fig. 3. There are distinct cranio-caudal differences in the ridge, with the more cranial regions seemingly more highly organized than those more caudal. We have observed this in all embryos (about 20) of this stage studied.
Montage of stage-47 gonadal ridge to show the surface appearance of the early gonadal ridge cells.
Figs. 4-6. These are higher-power micrographs of areas A, B and C in Fig. 3, to show the cranio-caudal differences in the gonadal ridge at this stage. Fig. 4 is a stereo pair to show the detailed surface morphology of the differentiated gonadal ridge cells.
Fig. 7. Area of gonadal ridge from stage-48 tadpole to show the single cilium in the centre of each coelomic lining cell of the gonadal ridge.
Figs. 4-6. These are higher-power micrographs of areas A, B and C in Fig. 3, to show the cranio-caudal differences in the gonadal ridge at this stage. Fig. 4 is a stereo pair to show the detailed surface morphology of the differentiated gonadal ridge cells.
Fig. 7. Area of gonadal ridge from stage-48 tadpole to show the single cilium in the centre of each coelomic lining cell of the gonadal ridge.
Figure 4 is a stereo pair of scanning micrographs, and shows a typical cranial bulge in the gonadal ridge. Here, many somatic cells are involved in covering the PGCs. They are more obviously spindle-shaped and stand out further from the posterior body wall. At the caudal end (Figs. 5, 6) the somatic cells become progressively more flattened and plate-like. They overlap each other, showing small surface projections resembling microvilli at their boundaries. The caudal end of the ridge does not stand out as far from the posterior body wall.
Three-dimensional inspection of these bulges reveals cilia-like processes on their surfaces. These are shown to greater advantage in a micrograph taken from another tadpole of the same stage (Fig. 7). Each somatic cell of this bulge has one cilium, more or less centrally placed.
The distribution of cells of the gonadal ridge at this stage (Figs. 3-6) suggests that the cranial end is beginning to proliferate more rapidly than the caudal end. This phenomenon is emphasized by studying later stages in ridge formation (Fig. 8). Here the cranial end is seen to be growing much more rapidly than the caudal end. Indeed, the most cranial part seems to be almost an independent body at this stage. This more rapid growth of the cranial end is paradoxical in the light of previous studies (Wylie & Heasman, 1975), which show that the PGCs move laterally to form the gonadal ridge later at the cranial end.
Fig. 8. Low-power SEM to show gonadal ridge of stage-49/50 tadpole. Both ridges are visible since the micrograph was taken from directly above the posterior body wall.
Fig. 9. Stage-48 tadpole, to show the differentiation of the gonadal ridge cells in the absence of PGCs. Note the characteristic rounded mass of cells at the cranial end of the ridge.
Fig. 10. (A) Light micrograph of haematoxylin and eosin-stained 7 μm wax section from stage-49 tadpole. Notice the bulge of somatic cells, devoid of germ cells, at cranial end of the gonadal ridge. (B) Twenty-one μm further caudally, the cranial bulge of somatic cells is being replaced by the main part of the gonadal ridge, containing typical PGCs.
Fig. 8. Low-power SEM to show gonadal ridge of stage-49/50 tadpole. Both ridges are visible since the micrograph was taken from directly above the posterior body wall.
Fig. 9. Stage-48 tadpole, to show the differentiation of the gonadal ridge cells in the absence of PGCs. Note the characteristic rounded mass of cells at the cranial end of the ridge.
Fig. 10. (A) Light micrograph of haematoxylin and eosin-stained 7 μm wax section from stage-49 tadpole. Notice the bulge of somatic cells, devoid of germ cells, at cranial end of the gonadal ridge. (B) Twenty-one μm further caudally, the cranial bulge of somatic cells is being replaced by the main part of the gonadal ridge, containing typical PGCs.
These observations raise two important questions concerning the specialized band of somatic cells:
their origin,
whether the PGCs have a role in the organization or differentiation of these cells to form the gonadal ridge.
With respect to the first question, we think it most likely that the somatic cells are present before the PGCs arrive, as squamous epithelial cells of the coelomic lining. Light and electron microscopy of cross-sections demonstrate the presence of coelomic epithelial cells at the junction of the mesentery and posterior body wall. We think that these become organized concomitantly with the arrival of the PGCs, to form the gonadal ridge.
We were able to test the second question by seeing whether the somatic cells were capable of differentiating and organizing themselves in the absence of PGCs. Newly fertilized eggs were irradiated at the vegetal poles with 29000 ergs/ mm2 of u.v. light. The embryos were allowed to develop and examined for the presence of PGCs. This is a well-established method for causing sterility in some anurans by preventing the early differentiation of the PGCs (Bounoure, Aubry & Auck, 1954; Smith, 1966). Figure 9 shows an embryo so treated, examined at stage 48. The specialized strip of somatic cells has formed normally, but there are none of the bulges characteristic of PGCs beneath it. The bulge at the cranial end of the ridge was found, on re-embedding and sectioning, to contain no PGCs. The strip of specialized somatic cells is therefore capable of differentiating without any influence from the PGCs.
All the irradiated specimens examined (20) showed a disorganized mass of somatic cells at the cranial end of the sterile gonadal ridge (Fig. 9). On careful examination these were also found in unirradiated controls (Fig. 2). In histological sections of controls, these cranial bulges were usually found to be devoid of PGCs (Fig. 10A). In sections immediately caudal to these bulges (Fig. 10B) PGCs reappear. These bulges stick out for some distance into the coelomic cavity and often appear to be appendages to the main part of the gonadal ridge (Fig. 11). The significance of this rapidly growing cranial part of the gonadal ridge, which is usually sterile in control embryos, is uncertain. It may be the homologue of Bidder’s organ in the toad (Witschi, 1933).
SEM stereo pair of the cranial end of the gonadal ridge from a stage-49 tadpole. This stereo pair shows the difference in surface morphology between the cranial bulge of cells and the main part of the gonadal ridge.
CONCLUSIONS
We have shown in this study that the PGCs move laterally from the root of the mesentery to form the paired gonadal ridges by stage 47-48 under our conditions. Concomitant with this movement, a population of somatic cells, possibly arising from the coelomic lining in this area, differentiate to form a highly organized band of tissue, which completely covers the PGCs. These are probably the precursor cells of the germinal epithelium of the developing gonad. The somatic cells can differentiate by themselves without any influence from the PGCs, which come to lie beneath them.
The appearance of these somatic cells raises the question of whether they play any role in the organization of the gonadal ridge. Two obvious possibilities present themselves:
the PGCs are moved passively from the median gonadal ridge to their final, more lateral position, by the morphogenetic movements undergone by the sheet of coelomic lining cells, as they condense to form a narrow strip of gonadal ridge cells.
the somatic cells exert a chemotactic effect on the PGCs, which actively migrate laterally towards them.
The latter hypothesis is most likely, given the evidence from mouse (Blandau, White & Rumery, 1963; Peters, 1970) and chick (Dubois, 1968) embryos. It remains to be seen, however, whether the large PGCs of Xenopus embryos are capable of independent locomotion, and if so, whether the cells of the coelomic lining destined to form the germinal epithelium play any role in attracting them.
ACKNOWLEDGEMENTS
We are indebted to Mrs M. Reynolds and Mrs P. Beveridge for technical assistance, and to Mr R. F. Moss and Miss D. Bailey for photographic assistance.