Consomic C57BL/6 males, carrying either the Mus musculus musculus-derived C57BL/6 Y chromosome or the Mus musculus domesticus-derived Poschiavinus Y chromosome, were outcrossed to females of the inbred strains C3H/Bi and CXBH/By and to females of the random bred strain MF1/Ola. In a study at 12.5 days post coitum, gonads of XYCS7 and XYPOS fetuses were assessed for the presence of testicular cords. It was found that XYPOS fetuses had a later onset of testicular development than XY57 fetuses. Limb development, which was monitored as a measure of overall development, was unaffected by the strain of Y present. These data were supported by a longitudinal study in which the increased growth rate of the testes relative to undifferentiated gonads, was also shown to be delayed in XYPOS fetuses. The extent of the delay was estimated to be approximately 14 h. It is concluded that this delay in the onset of testicular differentiation must be caused by differences between the two Y-chromosome types, most probably allelic differences in the testis determinant Tdy.

In 1982, Eicher et al. described a case of XY sex reversal in mice, resulting from backcrossing the Mus musculus domesticus-deiived Poschiavinus Y chromosome onto a C57BL6/J background. Adult C57BL/6-Ypos mice show a range of sexual phenotypes, including fully sex-reversed XY females, true hermaphrodites (i.e. they have both ovarian and testicular tissue) and males; although the males have small testes and are of poor fertility. Observations on fetal litters from this cross have revealed that the XY individuals at 16.5 days post coitum (dpc) have either ovaries, or ovotestes with a varying ratio of ovarian to testicular tissue. It seems clear that most ovotestes go on to form testes in adult life, the regression of the ovarian component accounting for the reduced testicular size.

Further studies have shown that a number of other (but not all) domesticus-derived Y chromosomes generate some XY females when placed on the C57BL/6 background (Eicher et al. 1982; Nagamine et al. 1987a; Biddle and Nishioka, 1988). Eicher and Washbum (1986) have suggested that the ‘domesticus’ Y carries a Tdy allele which is later-acting than that on the ‘musculus’-derived Y chromosome of the C57BL/6 inbred strain and that this delay sometimes enables the process of ovary determination to pre-empt Y action. To explain the requirement for a C57BL/6 background, Eicher and Washbum (1983, 1986) proposed that C57BL/6 carries a recessive autosomal allele (Tda-7b) which is in some way incompatible with the ‘domesticus’ Y chromosome.

There are two possible models for how Tda-1 is involved in this scheme (Eicher and Washbum, 1983). (1) This autosomal gene may be one element in the Tdy-initiated cascade of genes involved in testicular differentiation. (2) A second interpretation supported by Burgoyne (1988), proposes that it is an ovary determining gene, with the C57BL/6 allele acting earlier than those of other inbred strains. When this is brought together with the late-acting ‘domesticus’ Tdy allele, a ‘timing-mismatch’ may occur such that ovary determination may pre-empt testis determination. It is not hard to imagine how minor differences in the timing of expression of these two key genes could lead to the range of gonadal phenotypes observed in these mice. The ‘timing mismatch’ concept is discussed more fully in Burgoyne and Palmer (1991).

The aim of the present study was to test directly the hypothesis that the Poschiavinus Y chromosome is later acting than the C57BL/6 Y. Since C57BL/6 XYPOS fetuses develop ovaries or ovotestes, rather than testes, the effect of the Poschiavinus Y on the timing of testicular development was assessed in F, hybrids in which the paternal parent was C57BL/6 or C57BL/6-YPOS. In these F1 hybrids all the XY fetuses develop testes.

Experiment 1

C57BL/6Mcl-Ypos stud males were produced by backcrossing C57BL/6J-YPOS males to C57BL/6Mcl females for more than 10 generations. Inbred C3H/Bi females were mated to stud C57BL/6Mcl-YC57 and to C57BL/6Mcl-YPOS males in a room in which the dark period was 7pm to 5am. Pregnant mothers were killed at 12 days 15 h post coitum (pc) (mating was assumed to have taken place at 12:00 midnight) and the fetuses placed into Hepes-buffered Eagle’s minimum essential medium (EMEM, ICN Flow Ltd). Amniotic membranes were removed for sex chromatin analysis. Each membrane was placed near the top of a conical-bottomed centrifuge tube and flushed down into the tube with 3:1 methanokglacial acetic acid fixative. Fetuses were killed by decapitation, the hind limbs staged (see below) and the gonads removed. The gonads were photographed under EMEM using a Wild M400 Photomakroskop. Photographic prints of male and female gonads from both crosses were randomized and scored blind for the presence of testicular cords.

The hind limbs were staged according to the developmental series of McLaren and Buehr (1990), which covers the period spanning sexual differentiation in the mouse. This scheme was subdivided to provide maximum sensitivity and extended to cover the later stages found at this time point and the stages of the subsequent two days studied in experiment 2 (see Fig. 1).

Fig. 1.

Hind limb stages of fetuses between 11.5 and 14.5 days post coitum. E, early; L, late. In general, stages 4 and 5 were found during the 12th day of pregnancy, 5 to L7 during the 13th, L7 to 9 during the 14th and 9 to L10 during the 15th. Bar, 1 mm.

Fig. 1.

Hind limb stages of fetuses between 11.5 and 14.5 days post coitum. E, early; L, late. In general, stages 4 and 5 were found during the 12th day of pregnancy, 5 to L7 during the 13th, L7 to 9 during the 14th and 9 to L10 during the 15th. Bar, 1 mm.

The amniotic membranes were processed using an adaptation of the method of Evans et al. (1912) for yolk sac preparations (Burgoyne et al. 1983). Tubes containing the amniotic membranes were centrifuged briefly at 1000 revs min−1 and excess fixative decanted off (being careful to retain the amnion). The tubes were inverted on a paper towel and tapped on the bench to bring the amnion 1 or 2 cm away from the base of the tube. A single drop of 60% glacial acetic acid in distilled water was applied directly onto the amnion in order to dissociate the cells. After 60s, the action of the 60 % acetic acid was blocked by the addition of approximately 1 ml of fixative. The cells were then centrifuged at 1000 revs min−1 for 5 min. Excess fix was decanted off and the tube inverted for a few seconds to allow the tube to drain, without letting the cells dry completely. When the tube was righted again, the cells were resuspended in the small amount of fixative that ran down to the bottom from the sides of the tube. If this process was judged correctly, the cells could be concentrated into a single drop of fluid. This single drop was spotted onto a clean glass slide. When the cells had air-dried, one drop of 1% aqueous toluidine blue was spotted onto the cells and a coverslip pressed on. These preparations were scored immediately for the presence or absence of the sex chromatin (see Fig. 2).

Fig. 2.

Disaggregated amnion cells (A) positive and (B) negative for sex chromatin (arrows). Bar, 10 μm.

Fig. 2.

Disaggregated amnion cells (A) positive and (B) negative for sex chromatin (arrows). Bar, 10 μm.

Experiment 2

Consomic C57BL/6Mcl-YC57 and C57BL/6Mcl-YPOS stud males were mated to females of the inbred strain CXBH/By and to females of the random-bred albino stock MFl/Ola. Once again the dark period was 7pm to 5am and mating was assumed to have taken place at 12:00 midnight. In order to study the gonadal growth rates in the fetuses produced by these crosses, pregnant females were killed at time points between 11.5 and 14.5 dpc. Amniotic membranes were fixed for sexing and the hind limbs were scored as before.

Gonads were dissected out and scored under the dissecting microscope for the presence of testicular cords. The length and breadth of each gonad was measured to provide continuous variables to study the relative growth rates. However, we found that the length measurements were highly variable at the earlier stages due to difficulties in delineating the ends of the gonads. The breadth was better defined and showed a relatively smooth exponential growth curve. This was therefore chosen as the representative parameter for gonadal growth. Breadth was measured by gently holding the mesonephros with forcepts so that the gonad was uppermost. The width across the widest part was measured using a graticule fitted in the eyepiece of the dissecting microscope.

Experiment 1. Transverse study: fetal litters examined at 12 days 15 h pc

9 litters of fetuses produced from the cross C3H × C57BL/6-YC57 were found to contain 19 individuals typed as sex-chromatin negative. In the C3H ×C57BL/ 6-Ypos cross, a total of 8 Utters yielded 15 sexchromatin-negative individuals. These sex-chromatinnegative fetuses are assumed to be XY although in rare instances they could be XO. All but 2 of the XY057 males were scored as having visible testis cords whereas none of the XYPOS fetuses showed any sign of testicular differentiation at this time point. The probability of this occurring by chance (χ2=26.8, 1 degree of freedom) is P = 0.00014.

When these data are plotted against hind limb stage (see Fig. 3), individuals that are at the same stage of limb development show a clear difference in testicular development dependent upon the source of the Y chromosome. A comparison of the distributions of the hind limb stages demonstrates that, as far as the sensitivity of this measure will allow, there is no significant difference between the two male populations with respect to overall fetal development (χ2=3.01 4df. P=0.56).

Fig. 3.

Frequency of sex-chromatin-negative individuals with and without visible testis cords at 12 days 15 h post coitum, plotted against hind limb stage. Solid bars, C3H ×C57BL/6-YC57 F1; open bars, C3H ×C57Bl/6-Ypos F1.

Fig. 3.

Frequency of sex-chromatin-negative individuals with and without visible testis cords at 12 days 15 h post coitum, plotted against hind limb stage. Solid bars, C3H ×C57BL/6-YC57 F1; open bars, C3H ×C57Bl/6-Ypos F1.

Experiment 2. Longitudinal study

The delay in the onset of testicular differentiation is also found in the XY fetuses produced from crosses of the consomic C57BL/6-YC57 and C57BL/6-YPOS males with CXBH/By or MF1/Ola females (see Fig. 4). These data also show that all the XYPOS fetuses in both crosses develop testes from late 12.5 dpc onwards and no signs of XY sex reversal were found in a total of 47 sex-chromatin-negative individuals between 13.5 and 14.5 dpc.

Fig. 4.

Percentage of sex-chromatin-negative individuals with and without visible testicular differentiation at a range of hind limb stages. Solid bars, XYC57 fetuses; open bars, XYPOS fetuses. (A) F1 males produced from CXBH outcross; 33 XYPO5 and 28 XYC57 fetuses. (B) F1 males from MF1 outcross; 28 XYPOS and 56 XYC57 fetuses. Missing bars indicate that no sex-chromatin-negative individuals were found with that hind limb stage. The dip in the XYC57 data at E7 in the MF1 cross is caused by a single individual. It is possible that this fetus was mistyped or may have been a rare XO.

Fig. 4.

Percentage of sex-chromatin-negative individuals with and without visible testicular differentiation at a range of hind limb stages. Solid bars, XYC57 fetuses; open bars, XYPOS fetuses. (A) F1 males produced from CXBH outcross; 33 XYPO5 and 28 XYC57 fetuses. (B) F1 males from MF1 outcross; 28 XYPOS and 56 XYC57 fetuses. Missing bars indicate that no sex-chromatin-negative individuals were found with that hind limb stage. The dip in the XYC57 data at E7 in the MF1 cross is caused by a single individual. It is possible that this fetus was mistyped or may have been a rare XO.

The gonadal growth rate data are shown in Fig. 5. The mean gonad breadth measurements are log-transformed to improve the linearity of the resulting regression fines, decreasing the residual variance. Regression lines and their associated errors were computed using litter means of gonadal breadth for each sex, weighted according to the number of females or males in the litter. For the males, only the data between 12.5 and 14.5 dpc are included to be as certain as possible that the points fall during the phase of increased growth that is characteristic of testes. Some outliers have been removed using a test for outliers in regression analysis. Regression lines of gonadal growth in XYC57 and XYPOS fetuses were compared for both the CXBH and the MF1 outcross using an analysis of covariance. This test compares separately the variance from the mean due to the elevation, the slope and the residual error.

Fig. 5.

Regression lines of gonadal growth in males and females between 11.5 and 14.5 days post coitum. Closed circles, litters fathered by C57BL/6-YC57 males; open circles, litters fathered by C57BL/6-Ypos males. Each point is the litter mean (for that sex) of the individual gonadal breadth means. Regression data are compared in each case with respect to the Y-chromosome-type of the father using an analysis of covariance. The residual variances are not significantly different in any of the four tests. (A) Females produced from CXBH outcross. Difference between slopes, P = 0.59; between elevations, P = 0.86. Data are pooled to produce one regression line. (B) Males produced from CXBH outcross. Difference between slopes, P = 0.67; between elevations, P = 0.00009. The combined slope is used in plotting the two regression lines. (C) Females produced from MFI outcross. Difference between slopes, P = 0.321; between elevations, P = 0.286. Data are pooled to produce one regression line. (D) Males produced from MFI outcross. Difference between slopes, P = 0.203; between elevations, P = 0.00126. The combined slope is used in plotting the two regression lines.

Fig. 5.

Regression lines of gonadal growth in males and females between 11.5 and 14.5 days post coitum. Closed circles, litters fathered by C57BL/6-YC57 males; open circles, litters fathered by C57BL/6-Ypos males. Each point is the litter mean (for that sex) of the individual gonadal breadth means. Regression data are compared in each case with respect to the Y-chromosome-type of the father using an analysis of covariance. The residual variances are not significantly different in any of the four tests. (A) Females produced from CXBH outcross. Difference between slopes, P = 0.59; between elevations, P = 0.86. Data are pooled to produce one regression line. (B) Males produced from CXBH outcross. Difference between slopes, P = 0.67; between elevations, P = 0.00009. The combined slope is used in plotting the two regression lines. (C) Females produced from MFI outcross. Difference between slopes, P = 0.321; between elevations, P = 0.286. Data are pooled to produce one regression line. (D) Males produced from MFI outcross. Difference between slopes, P = 0.203; between elevations, P = 0.00126. The combined slope is used in plotting the two regression lines.

The regression lines for the XX fetuses produced by the two consomic fathers were also compared for each outcross. These data (see Fig. 5 legend) demonstrate that, for both the MF1 cross and the CXBH cross, the females produced by the two consomic males show no significant difference in the residual variance, the variance due to the slope or the variance due to the elevations. This is to be expected since there is no genetic difference between the F1 females produced by each paired cross. It is therefore justified’to pool these two populations within each outcross.

A comparison of the regression lines for the males (XVC57versus XYPOS), on the other hand, although revealing no significant difference in residual variance or slope, shows a highly significant difference between elevations in both crosses (see Fig. 5 legend). This supports the conclusions of experiment 1 and shows that the late onset of testis cord formation in XYPOS fetuses is correlated with a delayed onset of increased growth.

Both experiments demonstrate that the Poschiavinus Y chromosome causes a later onset of testicular development than the C57BL/6 Y chromosome. The simplest explanation for this disparity is a difference between the two Tdy alleles. This may be caused by differences in the regulatory elements or the structural sequence of Tdy. This is currently being tested by examining the structure and expression of the Poschiavinus Sry since it is now established that Sry is Tdy (Koopman et al. 1991).

These findings provide an explanation for the results of Nagamine et al. (1987b) who found that when C57BL/6 males with a domesticus-derived Y chromosome are outcrossed to various inbred females, a proportion of the F1 male fetuses produced show ‘abnormalities’ of testicular development. In the light of results presented here, it seems likely that these ‘abnormalities’ are due to a delay in testicular cord formation caused by a late-acting Tdy allele.

From the growth rate data, it is possible to model the growth kinetics of male and female gonads and thereby estimate the timing difference between the action of the two Y chromosomes (see Fig. 6). If Tdy is the first gene to influence male and female gonads differentially, it is logical that prior to Tdy action the growth rates of the indifferent XX and XY gonads should be the same. After Tdy expression, however, the gonadal growth of XY fetuses increases dramatically but the gonads of XX fetuses continue to grow at a slower rate. The gonads of XY fetuses with delayed Y action would be expected to continue growing at the same rate as female gonads until expression of the delayed Tdy has occurred. Therefore, to estimate the timing difference it is necessary to find where the XYC57 and XYPOS testicular growth curves depart from the XX curve. Calculations based on this model show that thepoints of intersection for the CXBH outcross are: XYC57 fetuses, 269 h post coitum (hpc); XYPOS fetuses, 283 hpc; and for the MF1 outcross XYC57, 270 hpc; XYPOS, 284 hpc. The estimate for the timing difference between the action of the two Y chromosomes is therefore 14 h. The estimated points of intersection should be regarded with caution since they result from extrapolations that may misrepresent the true shape of the curves. In spite of this caution, the timing for the C57BL/6 Y (early on the 12th day) fits well with the timing of expression of Sry which begins to be expressed during the 11th day, peaks on the 12th and declines during the 13th day (Gubbay et al. 1990; Koopman et al. 1990).

Fig. 6.

Model of gonadal growth in females and in males carrying either the YC57 or the YPOS chromosome. The initial single line represents the growth rate of the indifferent XX or XY gonads. After expression of either of the two Tdy alleles the testicular growth rate increases and departs from that of the female gonad. The testicular growth rate is the same in YC57-bearing and YPOS-bearing fetuses, only the time of departure differs. The points of intersection are calculated from the regression lines shown in Fig. 5. For two lines with slopes m1 and m2 and elevations C1 and c2, at the point of intersection X= (c2 – c1) ÷ (m1 – m2).

Fig. 6.

Model of gonadal growth in females and in males carrying either the YC57 or the YPOS chromosome. The initial single line represents the growth rate of the indifferent XX or XY gonads. After expression of either of the two Tdy alleles the testicular growth rate increases and departs from that of the female gonad. The testicular growth rate is the same in YC57-bearing and YPOS-bearing fetuses, only the time of departure differs. The points of intersection are calculated from the regression lines shown in Fig. 5. For two lines with slopes m1 and m2 and elevations C1 and c2, at the point of intersection X= (c2 – c1) ÷ (m1 – m2).

XY fetuses have been shown to have an overall developmental advantage over their XX siblings (Seller and Perkins-Cole, 1987), males being about 1| hours ahead of females (Burgoyne et al. unpublished). This developmental advantage is now known to be a consequence of a retarding effect of the paternally imprinted X chromosome in XX fetuses, rather than an accelerating effect of the Y chromosome (Thornhill and Burgoyne, unpublished). Because it is not Y-linked, this developmental advantage of XY fetuses should not affect the estimate for the difference in timing of YPOS and YC57 action. However, it does complicate the interpretation of XX versus XY gonadal size differences.

Measurements of gonadal volume in rat fetuses have shown that XY gonads are larger than XX gonads prior to the formation of testis cords (Lindh, 1961; Mittwoch et al. 1969). This led Mittwoch (1969, 1989) to propose that an early growth advantage of XY gonads, rather than a single gene ‘switch’, forms the basis of the sexdetermining mechanism in mammals. However, there is no previous evidence to show that these ‘within litter’ differences between XX and XY gonadal volume are not simply a manifestation of the fact that males are ahead of females. Our data, on the other hand, indicate that the increased growth, characteristic of testes, precedes testis cord formation. Nevertheless, this increased growth is after the onset of Sry expression so it remains a moot point whether the increased gonadal size is simply a consequence of Sry expression, perhaps caused by the differentiation and growth of Sertoli cell precursors (Jost et al. 1973; Magre and Jost, 1980), or whether it is a requirement for determining the gonad as a testis.

The results presented here appear to be contradicted by a recent study on the onset of Mullerian inhibitor, AMH (or MIS), production in fetal testes (Taketo et al. 1991). The authors compared the onset of AMH production in SJL XYdom fetuses with control SJL XYC57 fetuses and found no difference. It is possible that the assay system used and time intervals of 24 h, were not sensitive enough to detect a timing difference of only 14 h.

The delay in action of the Poschiavinus Y chromosome supports, in part, the hypothesis that sex-reversal in C57BL/6-Ypos individuals is caused by a developmental mismatch involving a late-acting Tdy allele in conjunction with an early-acting ovarian program. It remains to be seen whether the other component of C57BL/6-YPOS sex reversal, the recessive C57BL/6 allele of Tda-1, is a relatively early-acting gene in the ovary-determining pathway.

We thank Dr Eva Eicher for providing the C57BL/6J-YPOS mice and Dr Costas Goutis for help with the statistical analysis.

Biddle
,
F. G.
and
Nishioka
,
Y.
(
1988
).
Assays of testis development in the mouse distinguish three classes of domesticus-type Y chromosome
.
Genome
30
,
870
878
.
Burgoyne
,
P. S.
(
1988
).
Role of mammalian Y chromosome in sex determination
.
Phil. Trans. Roy. Soc. Land. B
.
322
,
63
72
.
Burgoyne
,
P. S.
,
Evans
,
E. P.
and
Holland
,
K.
(
1983
).
XO monosomy is associated with reduced birthweight and lowered weight gain in the mouse
.
J Reprod. Fert
.
68
,
381
385
.
Burgoyne
,
P. S.
and
Palmer
,
S. J.
(
1991
).
The genetics of XY sex reversal in the mouse and other mammals
.
Semin. Devi.Biol
.
2 (in press
).
Eicher
,
E. M.
and
Washburn
,
L. L.
(
1983
).
Inherited sex reversal in mice: identification of a new primary sex-determining gene
.
J. exp. Zool
.
228
,
297
304
.
Eicher
,
E. M.
and
Washburn
,
L. L.
(
1986
).
Genetic control of, primary sex determination in mice
.
Ann. Rev. Genet
.
20
,
327
360
.
Eicher
,
E M.
,
Washburn
,
L. L.
,
Whitney
,
J. B.
and
Morrow
,
K. E.
(
1982
).
Mus poschiavinus Y chromosome in the C57BL/6J murine genome causes sex reversal
.
Science
217
,
535
537
.
Evans
,
E. P.
,
Burtenshaw
,
M. D.
and
Ford
,
C. E.
(
1972
).
Chromosomes of mouse embryos and newborn young: preparations from membranes and tail tips
.
Stain Technol
.
47
,
229
234
.
Gubbay
,
J.
,
Collignon
,
J.
,
Koopman
,
P.
,
Capel
,
B.
,
Economou
,
A.
,
Munsterberg
,
A.
,
Vivian
,
N.
,
Goodfellow
,
P.
and
Lovell-Badge
,
R.
(
1990
).
A gene mapping to the sex-, determining region of the mouse Y chromosome is a member of a novel family of embryonically expressed genes
.
Nature
346
,
245
250
.
Jost
,
A.
,
Vicier
,
B.
,
Prepin
,
J.
and
Perchellet
,
J. P.
(
1973
).
Studies on sex differentiation in mammals
.
Rec. Prog. Harm. Res
.
29
,
1
41
.
Koopman
,
P.
,
Gubbay
,
J.
,
Vivian
,
N.
,
Goodfellow
,
P.
and
Lovell-Badge
,
R.
(
1991
).
Male development of chromosomally female mice transgenic for Sry
.
Nature
351
,
117
121
.
Koopman
,
P.
,
Munsterberg
,
A.
,
Capel
,
B.
,
Vivian
,
N.
and
Lovell-Badge
,
R.
(
1990
).
Expression of a candidate sexdetermining gene during mouse testis differentiation
.
Nature
348
,
450
452
.
Lindh
,
J.
(
1961
).
Quantitative aspects of prenatal gonad growth in albino rat and golden hamster studied by morphologic and experimental methods. Thesis, Dept of Anatomy, University of Lund
.
Magre
,
S.
and
Jost
,
A.
(
1980
).
The initial phases of testicular organogenesis in the rat. An electron microscopy study
.
Arch. Anat. microsc. Morph, exp
.
69
,
297
318
.
McLaren
,
A.
and
Buehr
,
M.
(
1990
).
Development of mouse germ cells in cultures of fetal gonads
.
Cell Diff. Dev
.
31
,
185
195
.
Mittwoch
,
U.
(
1969
).
Do genes determine sex?
Nature
221
,
446
448
.
Mittwoch
,
U.
(
1989
).
Sex differentiation in mammals and tempo of growth: probabilities vs switches
.
J. theor. Biol
.
137
,
445
455
.
Mittwoch
,
U.
,
Delhanty
,
J. D. A.
and
Beck
,
F.
(
1969
).
Growth of differentiating testes and ovaries
.
Nature
224
,
1323
1325
.
Nagamine
,
C. M.
,
Taketo
,
T.
and
Koo
,
G. C.
(
1987a
).
Morphological development of the mouse gonad in Tda-1 XY sex reversal
.
Differentiation
33
,
214
222
.
Nagamine
,
C. M.
,
Taketo
,
T.
and
Koo
,
G. C.
(
1987b
).
Studies on the genetics of Tda-1 XY sex reversal in the mouse
.
Differentiation
33
,
223
231
.
Seller
,
M. J.
and
Perkins-Cole
,
K. J.
(
1987
).
Sex differences in mouse embryonic development at neurulation
.
J. Reprod. Fert
.
79
,
159
161
.
Taketo
,
T.
,
Saeed
,
J.
,
Nishioka
,
Y.
and
Donahoe
,
P. K.
(
1991
).
Delay of testicular differentiation in the B6.YDOM ovotestis demonstrated by immunocytochemical staining for Mullerian inhibiting substance
.
Devl Biol. (in press)
.