Development of Embryonic Mouse Gonads Transferred to the Spleen: Effects of Transplantation Combined with Genotypic Autonomy

Knowledge of the time and manner of origin of genetically induced defects may well be sought by classical methods of experimental embryology, involving transplantations between genotypes, providing the effects on a tissue of residence in a host of different genotype is not obscured by generalized reactions to transplantation. Such experimental transplantation is reported here, applied to analysis of the pleiotropic effects of the deleterious alleles at the fP-locus in the mouse (Russell, review, 1955).

Through extensive investigation it has become apparent that the three major types of defect associated with these alleles are already established at birth. In contrast to their normal (ww) littermates, new-borns of the genotypes with which this paper will largely be concerned (W^VW^V and WW) are severely anaemic (Russell & Fondai, 1951), their gonads are almost totally devoid of germ-cells (Coulombre & Russell, 1954), and their hair follicles lack melanoblasts (Silvers, 1953). It is further known that these same genotypes are deficient in bloodforming capacity at least as early as days of embryonic life (Borghese, 1952; Attfield, 1951). The erythrocyte levels of both W^VW^V and WW are measurably lower than those of littermates at the 14th day of embryonic life (Russell, Fondai, & Smith, 1950). Conventional histological methods have not given critical evidence concerning the relative numbers of germ-cells in normal ww and potentially sterile W^VW^V and WW at 12 days of embryonic life, although differences have been observed in 16-day embryos.

One reasonable hypothesis for the anaemia-germ-cell pleiotropic effects of these genes attributes the lack of germ-cells in ovaries and testes of new-born WW and W^VW^V individuals to a suppression of their pre-natal development or survival caused by the anaemia present in the same embryos. It is essential to this theory that anaemia be present before the germ-cell defect has reached an irreversible maximum expression.

A second possible hypothesis suggests that the germ-cell defect develops by local gene-action completely independent of the embryonic anaemia. Time of first appearance and final determination of each defect is, of course, less essential to this interpretation.

These opposing theories may be susceptible to test by observations of the nature of development of germ-cells in gonads from embryos of normal and defective genotypes, transplanted at stages prior to final determination, to a neutral site with a rich blood-supply. One necessity for such analysis is a site which sustains development sufficiently normally to allow interpretation of genotypic differences. Certainty of interpretation will also be greatly enhanced if identification of genotype of embryos can be made at the time of operation. With methods available for this study, the earliest stage at which PF-genotype identification could be attempted from observation of living embryos was days. The present paper is a report of observations of gonads from new-borns and 12–16-day embryos transplanted to the spleen of normal adult castrate male hosts.

The receiving tissue must have a rich blood-supply, must be very different from the gonad in histology, so that all implanted tissue can be recognized, and must be capable of supporting reasonably normal development of gonads of both sexes. Since the spleen has a rich blood-supply and is completely different in its histology from any gonadal tissue, it presents many advantages as a transplant site. Spleens of castrate mice have been used extensively as host sites for adult ovaries (Furth & Sobel, 1947; Li & Gardner, 1947; Klein, 1952; Hummel, 1954; and Gardner, 1955), largely because long periods of splenic residence lead to ovarian tumour formation. Ovarian function is maintained on retransplantation to the ovarian capsule, however, after short periods of splenic residence (Little, Hummel, Eddy, & Rupple, 1951). Ovaries of new-born mice transplanted to castrate spleen become abnormal after prolonged residence, with defects first observed after 4 weeks (Guthrie, 1954, 1955). The evidence of alteration of ovaries made imperative the selection of a short term for the experiment, and tests of the effects of short terms of splenic residence upon the development of ovaries and testes of normal new-borns. Preliminary trials of the effects of 2 weeks’ splenic residence on gonads taken from normal new-borns indicated that although transplant to the spleen for 2 weeks caused some reduction in the spermatogenic activity of the testis, otherwise normal development was obtained in both sexes.

For all subsequent experiments the spleen of 4–8 week C57BL/6 males, castrated 1–2 weeks before the operation, was selected as host site.

All donor animals were embryos and new-borns of the various genotypes segregating from W^vw x W^vw and Ww x Ww crosses. They were isogenic with C57BL/6, and thus acceptable to the selected hosts. In all cases, identification of donor genotype was made at the time of operation rather than on histological study, to avoid possibility of circular reasoning.

The age of embryos was determined within 8 hours by plugs from timed matings, and postpartum matings were avoided. In operations using new-borns and 16-day embryos, normal and W^VW^V or WW genotypes could be distinguished by inspection (W^VW^V and WW appear much paler), and sex could be determined at the time of operation by position and shape of gonads. In 14-day embryos it was not always possible to establish the sex of transferred gonads at the time of operation, and at 12 and 13 days no attempt at such identification was made. Hence sex of gonads transferred at early stages could only be inferred from results.

The genotype was difficult to determine in 12–13-day embryos. In 13-day embryo litters segregating for WW, slightly less than (11 of 52) were recorded at operation as having noticeably pale livers. It was assumed these were of the extreme anaemic WW genotype. All of the pale-liver embryos found in any uterus were used as donors, along with an equal number of presumed wnor Ww donors, selected at random from among the embryos with bright red livers. The W genotype of embryos in crosses segregating for W^vI w were identified with the help of a linked gene, lx, whose phenotype has been described as early as the 12th day of embryonic life (Carter, 1954). Both W^v and lx are on the third chromosome, 18 units apart (Carter, 1951>; Russell, unpublished data). An isogenic strain carrying both of these genes on one third chromosome was developed by repeated backcrosses of the double heterozygote to C57BL/6. On this C57BL/6 background practically all Lxlx adults have extra and frequently misshapen digits on the inner side of one or both rear feet, and all lxlx individuals have much more extreme abnormalities of hind legs and feet (Carter, 1951a). In -13-day embryos Carter (1954) has described a narrowing of the leg combined with an overgrowth of the inner side of the rear feet in presumed homozygous lxlx, and a slight enlargement of the inner side of the rear feet of Lxlx heterozygotes. In 12-day embryos Carter’s clearest evidence of homozygosity for lxlx was a narrowing of the rear leg, and change of its angle with the body although he also observed some enlargement of the inner side of the feet. One hundred and eight 13-day embryos from matings of W^vwLxlx parents classified according to the presence and extent of abnormality of the rear feet gave 32 completely normal (presumed LxLx), 48 with mild abnormalities of one foot (presumed Lxlx), 6 with mild to marked abnormalities of both feet (doubt ful, either Lxlx or Ixlx), and 22 with extreme abnormalities of both feet (presumed Ixlx), a reasonable approximation of the expected 1:2:1 ratio. Recognition of the 12-day Ixlx phenotype was less certain. Eight of 54 12-day embryos from W^vwLxlx matings were classified as Ixlx on the basis of narrowing of the leg and abnormal shape of the foot. Since in these crosses approximately 80 per cent, of individuals homozygous for Ixlx are also homozygous for W^VW^V, the appearance of rear feet and legs offers the best method at present available for distinguishing normal and potentially sterile embryos at the time of operation. Crossovers and difficulty in Ixlx identification are recognized as possible sources of error. From any one uterus all of the presumed W^VW^V embryos and an equal number of normal embryos were selected as donors. Wherever possible, the normals had completely normal feet (wwLxLx) but W^vwLxlx donors were occasionally used when ww embryos were not available in the litter.

Pregnant females were anaesthetized with nembutal and the uterus exposed. Beginning at the anterior end of the uterus, the section containing a single embryo was removed, the uterine wall and embryonic membranes cut, and the exposed embryo transferred to 0·85 per cent, saline solution for examination of feet and liver colour. Those selected as donors were spread on filter paper with the ventral surface exposed. After removal of the ventral abdominal organs, the gonads, usually with some adjacent tissue, were placed into the tip of a No. 19 hypodermic needle with a plunger inserted to make it serve as a trocar. The host was anaesthetized, its spleen temporarily exteriorized, a nick cut in its surface with scissors, and the donor tissue extruded from the needle-trocar into the body of the spleen.

The transplanted gonads were allowed to grow in the spleen to the time equivalent to 14 days post-natal (14 days for new-borns, 21 days for 13-day embryos). The hosts were then sacrificed and the spleen examined. In many cases the region of implanted tissue was externally visible; such regions were saved and sectioned serially. If no implant was visible, the entire spleen was sectioned serially. Examination of histological sections formed the basis for quantitative and qualitative evaluation of the results of these gonad transfers. All types of tissue recognizable in the implants were recorded for each case.

If ovarian tissue was observed, the total number of follicles was established by examination of all sections, with counts of nuclei of ova. The stage of development of the most advanced follicles found in each transplant was also recorded.

If testicular tubules were observed, the proportion containing spermatogenic cells was tested by Chalkley sampling (Chalkley, 1943). One section in each slide row (approximately 5 per cent, of all sections) was classified. In each test-section the tubules falling under four pointers were classified for presence or absence of spermatogenic cells. No detailed attempt was made to classify stages of spermatogenesis observed, nor to compare relative numbers of seminiferous cells per functioning tubule section. To avoid sampling error arising from variability in the number of transplanted tubules observed, the index of spermatogenic activity for each genotype was based on proportion of functioning tubule-sections in a combination of all pertinent testis transplants rather than on presence or absence of functioning tubules in a given implant.

Variables other than genotype of the donor contributed considerably to chance of recovery of pertinent tissue. The chance of recovering any transplanted tissue varied with the age of the donor at transfer. Tissue implants were found in 77 per cent. (153/200) of the recipient spleens, but the percentage was lower with 12-day donors (21 / 32 or 66 per cent.) than with 16-day and new-born donors (50/60 or 83 per cent.). Other types Of tissue were found in many implants (Table 1), including kidneys with glomeruli and collecting tubules, intestinal epithelium, skin (often with hair follicles), bone and included marrow, and muscle. As would be expected from the position of the transferred organs, kidney-tissue appeared very frequently, especially when donors were of ages (12 and 13 days) when gonad and kidney are very closely associated. Other extraneous tissue types (muscle, skin, bone, cartilage) were especially frequent in transfers from early stages. Large ducts or cysts were frequently found, more commonly with ovaries than with testes. Their appearance was suggestive of oviducts. The probability of recovering gonad tissue also varied with the developmental stage of the donor. In transfers from 12to 14-day embryos, testes, all showing excellent differentiation of tubular structure, were recovered much more frequently than ovaries. A probable explanation, in line with previous reports of operations involving embryonic gonads (Everett, 1943) is that the non-germ-cell tissue of embryonic testes survives better in the spleen than does that of ovaries. In those cases where female germ-cells from ovaries of fertile genotypes were recovered, they had differentiated more normally in the spleen than had male germ-cells in transplanted testes. When recovered after 18–22 days in the spleen, ovaries from 16to 12-day normal ovaries (Plate 2, fig. 5) contained many follicles, all of which had developed to stages corresponding to those seen in a normal 14-day post-natal ovary (Plate 1, fig. 1) (Coulombre & Russell, 1954). By contrast, the number of spermatogenic cells in functioning tubules in a testicular transplant never approached the level in the normal 14-day post-natal testis. In place of the typical 4–5 layers of seminiferous cells, with numerous primary and secondary spermatocytes and meiotic division plates (Plate 1, fig. 2) tubules found in transplants usually showed a single incomplete layer of cells, spermatogonia and/or spermatocytes (Plate 2, fig. 6). Testicular tubules survive easily, but their spermatogenesis is diminished in the spleen.

The gonads of 14-day post-natal ww individuals (Plate 1, figs. 1, 2) are strikingly different in appearance and germ-cell content from those of 14-day W^VW^V individuals (Plate 1, figs. 3, 4). Regardless of the developmental stage at transfer, ovaries from potentially fertile genotypes (ww, W^vw, Ww) recovered from the spleen at the post operative time equivalent of 14-days post-natal contained more ova than did ovaries from potentially sterile genotypes (WW, W^VW^V)

In a series of transfers of ovaries from normal and anaemic new-borns (Table 2) ovarian tissue was recovered from 9 of 15 ww ovaries after 2 weeks’ residence in the spleen of adult castrate ww males. All contained large numbers of follicles, the most advanced showing a many layered granulosa, frequently with coalescing antra (stages 5 and 6, Coulombre & Russell, 1954). Ovarian tissue was recovered from only 1 of 6 W^VW^V new-born ovaries transferred to the spleen of ww, and this one lacked ovarian follicles. It is interesting to note that 6 of 10 ovaries transferred from normal ww new-borns into anaemic W^VW^V castrate male spleens developed normally, all with large numbers of follicles, the most advanced at stage 5 or 6. Although it was more difficult to obtain successful ovarian transplants from early (12–13-day) embryo donors, in those cases where ovarian tissue was recovered the number of ova was higher in transfers from potentially fertile genotypes (Table 3; Plate 2, figs. 5,7). The numbers of follicles observed in ovaries recovered from transfers of 13-day normals approximated numbers recovered from transfers of new-born normal ovaries (Table 3). The most advanced follicles observed were at stages 5 and 6.

In testicular transplants, the proportion of functioning tubules was uniformly higher in transplants from potentially fertile than from potentially sterile genotypes. Although in transfers from normal genotypes residence in the spleen inhibited spermatogenesis to some extent, this was not sufficient to prevent expression of the genotypic difference. All of six testes from ww new-borns transferred to castrate ww spleens were recovered 14 days later (Table 2). Slightly more than half of the tubules (253/425) showed spermatogenic cells. Two testes were transferred to the spleen W^VW^V castrates, where they developed well, showing a high proportion (64/85) of tubules with spermatogenesis. The proportion of such tubules was much lower (65/201) in two transplants of testes from W^VW^Vnew-borns. Similar differences between implants from normal and potentially sterile donors were found with transplants made at early embryonic stages (Table 3; Plate 2, figs. 6,8).

The proportion of tubules with spermatogenesis was high in 8 testes recovered from transfers from 13-day ww embryos (256 / 338) and in 4 testes recovered from transfers from 12-day ww embryos (239/274). The proportion was much lower in 9 testes recovered from W^VW^V 13-day embryos (44/373), in 2 from W^VW^V 12-day embryos (9/45), and in 4 from WW 13-day embryos (10/163). Similarly, the total number of follicles was much higher in 5 ovaries recovered from transfers from 13-day ww embryos (118–13) and in one ovary from transfers from day ww embryos (29) than in 2 ovaries recovered from transfers from W^VW^V day embryos (0, 3) or in one ovary recovered from transfer from a 13-day WW embryo (0). These latter ovarian transplants were identified as ovarian by the swirling arrangement of stromal cords.

Thus it appears that explantation of gonads from a potentially anaemic sterile genotype to a site with a rich blood-supply at mid-embryonic stages does not alter the genotypic pattern of germ-cell development. In the age range tested, from the 12th day of embryonic life to birth, the time of removal from the natural (anaemic) milieu and transfer to the abundant blood-supply provided in the host spleen had no effect on the degree of germ-cell deficiency observed in gonads of either sex.

In all analyses of pleiotropism or studies of gene-action involving transplantation between differently affected genotypes or explants to tissue culture, possible non-specific effects of the operations involved must be taken into consideration. Hence it has appeared pertinent in this study to present evidence on effects of the operation and of splenic residence on subsequent development of embryonic gonads. From the results it appears that the trauma of operation has reduced the chance of recovering ovarian tissue, but has not seriously affected differentiation in successful transplants. In contrast, testicular tubules appear to have survived well in the spleen, but the multiplication and / or differentiation of their contained spermatogenic cells was somewhat inhibited in this foreign milieu.

In spite of these limitations, it is clear that in transplants of both ovaries and testes there are fewer germ-cells if the donor is of the potentially sterile PF IF or W^VW^V genotype than if it is of the normal fertile (ww) genotype. This agrees with the findings of Borghese (1955 and personal communication) who explanted tissues from 12-day embryos from Ww x Ww matings into tissue culture and observed genotypic autonomy of development. These findings are also in complete accord with the report of Mintz & Russell (1955) that the numbers of primordial germ-cells (not yet in the gonad) found in PF-series defective genotypes at the 11th day of embryonic life correspond to the known differences in numbers of gonadal germ-cells at birth in the same PF-series genotypes.

Thus this analysis of pleiotropic relations by means of transplantation between genotypes has been one of several sources of information indicating that the germ-cell defect in W^VW^V and WW genotypes has reached full expression at the earliest stage (⁠-day embryo) at which evidences of blood-forming defect have been identified in the same genotypes.

Since no gonads were transplanted from defective genotypes at stages prior to the full development of their germ-cell abnormality, no critical test of the alternative hypotheses suggested in the introduction has been possible. However, the time of possible influence of embryonic anaemia on initial development of germ-cell defect has been limited to the period before the 12th day of embryonic life.

The inter-genotype transfers have, nevertheless, contributed other important information. The extremely small numbers of germ-cells observed in splenic transplants of gonads from WW and W^VW^V embryos after 18-22 days in ww adult hosts indicates no differential influence favouring increased proliferations of germ-cells during this period, resulting from differences in blood-level of surrounding tissues, or from any other differences associated with IF-series genotypes.

This work has been supported in part by a grant to the Jackson Laboratory from the United States Atomic Energy Commission, in part by a grant from the American Cancer Society upon recommendation of the Committee on Growth of the National Research Council, and in part by a research grant (C-1074) from the National Cancer Institute of the National Institutes of Health, Public Health Service.