Mesonephric agenesis was achieved by microsurgical excision of the left Wolffian duct and the underlying intermediate mesoderm of different regions between somites 16 and 23 in chickens after 50–52 h of incubation (stage 14 HH). Quail-chick chimaeras were produced by transplantation of corresponding quail tissue in the region of somites 18–21.
A morphometrical analysis of the mesonephric and gonadal area in cross sections shows that the intermediate mesoderm from somites 16 to 23 develops into the mesonephros. A partial agenesis of the mesonephros brought about by removal of the intermediate mesoderm at the level of somites 18 to 21 at stage 14 leads to a mean reduction of the gonadal volume of 37·8 % compared to the volume of the untreated side at stage 30. Transplantation of quail intermediate mesoderm in this region of the excision results in development of a hybrid mesonephros. Consequently, the gonads are invaded and colonized by quail cells mobilized from mesonephric corpuscles examined at stage 30, 35 and 36.
These results are discussed in terms of the origin of the gonadal stroma during this developmental period; they show that in the region from the third to the sixth segment the ventromedial part of the differentiating mesonephros participates in the contribution of stromal cells to the gonad.
The development of the gonads of vertebrates starts with the formation of the germinal epithelium, bilateral thickenings of the coelomic epithelium that appear ventral to the developing mesonephroi. The proliferation of both germinal and ‘nongerminal’ cells leads to the formation of distinct gonadal primordia which are identical in both sexes, the indifferent gonad (see for review Zuckerman & Weir, 1977). The data on the origin of the gonadal stroma in sexually indifferent gonads are at variance. Considered by Callebaut (1976), Dodd (1977), Kopp & Bertrand (1978), Fargeix, Didier & Didier (1981), Merchant-Larios & Villalpando (1981) and Popova & Scheib (1981) to derive from the coelomic epithelium, the stroma of the gonads was suggested by others to arise either from the mesonephric blastema (Dang & Fouquet, 1979; Carlon, Pizant & Stahl, 1983) or from the differentiated mesonephros (Witschi, 1914, 1951; Swift, 1916; Upadhyay, Luciani & Zamboni, 1979; Zamboni, Bezard & Mauléon, 1979).
In order to ascertain the dependence of the gonad on the intermediate mesoderm or the mesonephros in chick embryos, we excised the intermediate mesoderm in the presumptive mesonephric area at day two. Following subsequent incubation, we morphometrically analysed the indifferent chick gonad at day 7 (end of the indifferent stage). To obtain further information about gonad-contributing cells of the mesonephros, we transplanted corresponding quail tissue in the presumptive gonadal area at day 2 and examined the chick gonads at days 7, 9 and 10.
MATERIALS AND METHODS
Microsurgery to cause mesonephric agenesis
Fertilized eggs of the domestic fowl (Gallus domesticus, strain White Leghorn) were acquired from a local supplier and incubated at 37·8°C ± 1°C, 70 % to 80 % humidity.
After 50–52 h of incubation the position of the embryo was marked by candling and the surface of each egg was sterilized by swabbing it with 70 % ethanol. A hole was pricked in the blunt end of the egg, a procedure that leads to air being forced out of the air space and which results in the embryo dropping away from the surface.
Operations were performed under aseptic conditions with the aid of a stereomicroscope, the egg being supported horizontally in a cradle of cellulose. After cutting a window (approximately 0·7 × 0·5 cm) in the shell and its subjacent shell membrane at the marked area, the chick embryo was floated up to the level of the window by infusion of Locke solution (Hara, 1971). The stages of development were determined according to the criteria of Hamburger & Hamilton (1951). We used embryos at stage 14.
For microsurgical procedures, electrolytically pointed tungsten needles were used (Dossel, 1958). After slitting the vitelline membrane, the left intermediate mesoderm and, for technical reasons, the Wolffian duct were excised over a length of three to four somites at alternating levels between somites 16 and 23 (Fig. 1). After the microsurgical procedure, 1 to 2 ml albumen were removed from the egg; it then was sealed with Leukosilk® and subsequently incubated for 5 days.
The operated and normal embryos were killed, fixed in Serra’s fluid containing 60% propanol, 30 % formaldehyde solution and 10 % glacial acetic acid and dehydrated with graded propanol solutions. The 7 μm paraffin-embedded serial sections were stained with haematoxylin and eosin. With the aid of a computer-linked planimeter (Interactive Image Analysis System, IBASI and IBASII, Kontron), the cross-sectional areas of the gonads and the mesonephroi were measured. The levels of the points of measurement were elicited by using the spinal ganglia as a means of orientation.
In the region of the mesonephros at the beginning and end of each spinal ganglion the circumference of the mesonephros was marked by a digitizer (Fig. 2). From this, IB AS calculated the cross-sectional areas in μm2. In the region of the gonad, measurements in the middle of each spinal ganglion and at the midpoint between two ganglia were added.
in each embryo. The percentage deviation of all points of measurement of one organ was then used to calculate the arithmetical mean of the difference between the operated and the normal side.
Fertilized eggs of the Japanese quail (Coturnix coturnix japonica), delivered by a local supplier, were incubated together with the chick eggs. The chick embryos were operated at the level of somites 18 to 21 at stage 14HH as described above. For the quail, whose development is precocious relative to chicks although it does not differ by more than a few hours until the third day of incubation, the stages were determined by analogy with the chick.
Quail embryos were removed from the egg, placed on agar and covered with Locke solution. Left intermediate mesoderm and Wolffian duct were cut from the quail donor with fine tungsten needles at the level of somites 18 to 21 and stained lightly with Nile blue sulphate impregnated in agar. These were then transferred by means of a Spemann micropipette from donor to host embryo and manoeuvred into place with tungsten needles (Fig. 1). In order to ensure a proper fit and good adhesion between host and graft tissues, excess Locke solution was carefully withdrawn with a micropipette. After removal of 1 to 2 ml albumen the sealed egg was incubated for 5, 7 and 8 days. Chimaeras at HH stages 30, 35 and 36 were processed for histological examination: they were fixed in Serra’s fluid by transcardiac perfusion and subsequent immersion. After paraffin embedding the specimens were serially sectioned at 7 μm. As the interphase nuclei of quail cells are characterized by a large mass of nucleolus-associated heterochromatic DNA, the sections were stained for DNA (Feulgen & Rossenbeck, 1924) and counterstained with lightgreen, or they were stained with haematoxylin and eosin following acid hydrolysis (Hutson & Donahoe, 1984).
Morphometry of normal embryos
Therefore we chose the left side for operations.
Morphometry of operated embryos
After excision of the different segments of the intermediate mesoderm and the Wolffian duct at day 2 (stage 14), a partial dysgenesis or agenesis of the mesonephros was noted at day 7 (stage 30).
In accordance with the different regions where the defects of the mesonephros appeared, the specimens were divided into three groups (Table 1): operations in the region of somites to 19 cause defects in the cranial third of the mesonephros (group I). Defects of the middle third (group II) are evoked by operations in the region of somites 18 to 21 and exhibit a maximum between somites 18·5 and 21·5 (Fig. 4). Cessation of development of the caudal third of the mesonephros (group III) results following removal of the intermediate mesoderm and the Wolffian duct in the zone of somites 20 to 23.
Reduction of the mesonephros alone does not influence the size of the gonads in all three groups. But a partial agenesis of the mesonephros in certain regions causes a reduction of the gonad. Mesonephric tissue from the region of the caudal third of the mesonephros has no influence on the gonad (Table 1).
In the region of the upper three segments (ganglia 15–17), an agenesis of the mesonephros resulted in a reduction of the gonad in only one of the five studied cases. If the agenesis in the two studied cases extends further caudally into the region of ganglion 18, an effect on the gonads is observed (Table 2), namely a mean reduction of 17·5 % to an average of 82·5 % as compared to the right side.
A marked influence on the size of the gonads is attributable to the absence of the mesonephros in the region of ganglia 18 to 20. In nine cases studied, a mean reduction of 37·8 % to an average of 62·2 % as compared with the opposite side results (Fig. 4). Agenesis of the mesonephros commencing from the region of ganglion 20 and further caudad has no more influence on the gonad (Table 2). The region of the reduction of the gonad is always observed caudally from the corresponding defect of the mesonephros (Fig. 4). Only in one case did diminution of the gonad appear at the same level as the defect of the mesonephros, but the maximum diminution nevertheless occurred further caudad.
Interspecific quail-chick chimaeras
43 out of 175 embryos operated in this way developed after subsequent incubation. Of these 43 specimens, 27 contained quail tissue. 21 of the latter were fixed on day 7, three on day 9 and three on day 10.
In all the above experimental cases, healing was very good. Macroscopically examined, they showed no alteration of form and bilateral symmetry in the development of the body wall; but 14 of the 43 embryos revealed a dysgenesis of the left upper limb.
Although the quail grafts were originally of the same size as the dissected chicken tissue, the expanding lesion in the developing chick mesonephros is never entirely filled by quail mesonephric tissue. After postoperative incubation for at least 5 days there is a good connection between host and donor mesonephros at the cranial end of the transplanted tissue, but at the caudal end a distinct gap exists. In each case, in the region of the graft, mesonephric tubuli consisting of cells with typical quail nuclei are found. Also hybrid tubuli are demonstrable, one part of the tubulus wall containing chick cells, the other part formed by quail cells; no intermingling of the different cells occurs (Fig. 5). If the quail graft develops only tubuli and no other mesonephric structures, no quail cells are discovered in the gonad.
The quail tissue is also able to form mesonephric stromal cells and to develop a Bowman’s capsule. Either the Bowman’s capsules circumscribe an empty space or they contain hybrid glomeruli. These glomeruli consist of chick-derived cells with oblong, flat nuclei which we consider to be endothelial cells of the glomerular loop; they also contain cells with oval or round nuclei carrying the quail marker, which we assume to be podocytes (Fig. 6).
In nearly all chimaeras the mesonephroi are composed of alternating chick and quail tissue; in many cases the quail mesonephric tissue occupies the dorsal or lateral parts of the organs. These dorsolateral sections of the mesonephroi do not contribute to the gonad, as no quail cells reach the gonad in such situations.
Thus, to form the gonad, the quail graft has to develop mesonephric corpuscles or stroma, and this in the ventromedial section of the mesonephros lying in juxtaposition to the gonad.
In 7-day-old embryos, quail cells egressing from the Bowman’s capsules of the quail type had colonized the indifferent gonad in three cases. Outside the Bowman’s capsule the migrating cells form very delicate trabeculae (Fig. 7) which, in the serial planes of sections, can be followed all the way into the gonad. There, the colonizing quail cells do not disperse randomly throughout the inner core of the gonad and do not intermingle with chicken stromal cells. They occupy a distinct part of the gonadal stroma and in no case do they show contact or intermingle with the stratified surface epithelium. A band-like space beneath the latter approximately twice as wide as the epithelium is formed by chick cells (Fig. 8). Hence, the quail cells establish no association with the primordial germ cells in the deep layer of the epithelium. Germ cells arranged between the inner core and the surface epithelium are enclosed by chick and quail stromal cells; the chick cells form a crescent round the germ cells at the side apposed to the epithelium; the other side is covered by quail cells (Fig. 9).
In a 9- and 10-day-old ovary, two structural components are demonstrable: an outer cortex and an inner medulla. The medulla occupying the major portion of the ovary is partially formed by quail cells. Chick cells constituting the other part of the medulla hardly intermingle with the quail cells. Most of the germ cells lie in pairs or small groups in the middle or inner part of the cortex. These germ cells are closely surrounded by chick cells. Also, the few germ cells lying in the deep medullary part toward the hilus, constituted of quail tissue, are surrounded by chick cells (Fig. 10).
In a 10-day-old testis, germ cells show an even distribution throughout the medulla, indicating the formation of male cords. Although in this case only a few quail cells colonize the gonad in small groups, they come into contact with the germ cells. These germ cells are consequently surrounded by cells of the chick as well as those of the quail type (Fig. 11).
The data indicate that the cranial third of the mesonephros develops from the intermediate mesoderm of the region of somite 16 to the beginning of somite 19. The middle third is formed by material belonging to the region extending from somite 18 to the beginning of somite 21, and the caudal third of the mesonephros originates in the intermediate mesoderm of the region comprising somites 21 to 23 (Fig. 12). As shown in Fig. 12, mesonephric tissue of the region extending from the end of ganglion 17 to the beginning of ganglion 20 (black area) contributes to the gonad in each case. Mesonephric material of two small regions (hatched areas) adjacent to the cranial and caudal end of this region has influence on the gonad only in certain cases. In contradiction to authors who consider the presumptive gonadal area to be at the level of somites 9 to 14 (Romanoff, 1960) or at the level of somites 20 to 26 (Willier, 1933; Grünwald, 1937) or at the level of somites 24 to 29 (Dantschakoff, 1931) or spread over a more extended region from somites 13 to 22 (Didier, Fargeix & Bergeaud, 1980), our results of the extirpation and transplantation experiments suggest a presumptive gonadal area of the intermediate mesoderm extending from the beginning of somite 17 to the beginning of somite 21.
The diminution of the gonadal volume was caused by excision of the intermediate mesoderm; thus on the basis of our experiments alone it cannot be decided whether this reduction of the chicken gonad is due to loss of the mesonephric blastema or to loss of the differentiated mesonephros. However, Bishop-Calame (1966), Popova & Scheib (1981) and Merchant-Larios, Popova & Reyss-Brion (1984) observed a reduction of the gonad following interruption of the Wolffian duct. Since they did not excise the mesonephric blastema, but prevented mesonephric development, we attribute the diminution of the gonadal volume in their and our experiments to the missing contribution of a differentiated mesonephros. This is confirmed by the transplantation experiments. As these show, the gonads are colonized by cells originating in mesonephric corpuscles.
Several studies concerning gonadal development in mammals also postulate or show a mesonephric origin of the somatic cells of the gonad (Witschi, 1951; Wartenberg, 1978; Fraedrich, 1979; Kinsky, 1979; Upadhyay etal. 1979; Zamboni et al. 1979). These cells originate in developing as well as in regressing or disintegrating structures of the mesonephros. In accordance with the assumption that mesonephric degeneration in chickens does not begin earlier than the 11th day of incubation (Romanoff, 1960), in our examination (days 7, 9 and 10) the chickens exhibited no signs of degeneration of the mesonephros. Thus in the chicken the segregation of cells of the mesonephros takes place during the differentiating phase at these stages of development.
In rabbit (Kinsky, 1979), sheep, Macaca mulatta (Zamboni etal. 1980) and man (Wartenberg, 1978) the gonad is colonized by mesonephric cells of the cranial third of the organ. This seems to be different in mammals from in chickens. Neither the cranial fifth of the mesonephros nor the caudal half is observed to contribute to the gonad. Only a complete agenesis of the mesonephros over a region of more than three segments successive to the cranial fifth causes a reduction of the gonad (Fig. 12). Likewise, the quail mesonephric tissue has to appear in the latter region of the mesonephros in order to form gonadal stroma. Quail structures further craniad or caudad have no connection with the gonad.
During the operations, we took pains to avoid injuring both the dorsal aorta lying directly ventral to the intermediate mesoderm and the coelomic epithelium medial of the latter. For this reason we could not always be sure that this ventromedial part of the intermediate mesoderm was excised in toto. We suppose this incomplete removal to be the reason why in many extirpation experiments no partial agenesis but a partially reduced mesonephros resulted in the operated region. If only scant mesonephric tissue appeared in the gonad-contributing area of the mesonephros, no reduction of the gonad was observed. We consider that this ventromedial part of the intermediate mesoderm contributes to the formation of the ventromedial section of the mesonephros and, as the transplantation experiments show, it is only this ventromedial section of the mesonephros that contributes to the gonadal stroma.
An agenesis of the mesonephros restricted to two or three segments never resulted in agenesis of the gonad. In the same way, there was no case in which quail cells originating from quail mesonephric corpuscles formed the entire gonadal stroma at any given level. On the one hand it is possible that the persisting cranial and caudal parts of the chicken mesonephros are extensive enough to restore the gonadal stroma and act as a regulator; for technical reasons we are not able to excise more extended parts of the intermediate mesoderm to eliminate such a regulative mechanism.
On the other hand there still remains the question of a contribution of the coelomic epithelium to the gonadal stroma. With this method we cannot exclude it, because we are not able to decide whether the chick cells in the gonad derive from persisting parts of the chick mesonephros or from the chick coelomic epithelium. This method only proves the mesonephros to participate, at least, in the gonadal stroma.
The reduction of the gonad was found almost exclusively caudal to the corresponding agenesis of the mesonephros (Fig. 4). According to observations in the rabbit (Kinsky, 1979), in sheep (Zamboni et al. 1979), in the Macaca fascicularis (Dang & Fouquet, 1979) and in man (Wartenberg, 1978), confirmed by the transplantation experiments, the contribution of the mesonephros to the gonads would seem to proceed in a craniocaudal direction.
The relationship of the gonadal stromal cells of mesonephric origin to the germ cells seems to differ in male and female gonads. In contradiction to findings in sheep (Zamboni et al. 1979), they do not come into contact with the germ cells in chicken ovaries at this early stage of development. Male germ cells are surrounded by mesonephric quail cells and chicken cells at the same time. This result may correspond with the postulated dual Sertoli cell system in man and mammals (Wartenberg, 1978,1979); the possibility needs to be explored in further studies.
This work was supported by Deutsche Forschungsgemeinschaft.