ABSTRACT
In this report the morphology of the gonads and the growth of the Wolffian and Müllerian duct in foetal mouse true hermaphrodites (16 days p.c.) have been studied and compared to that of normal mice. The ducts from the hermaphrodites were placed in one of five groups according to the proportion of male and female characteristic of the gonad.
When more than 85 % of the gonadal tissue was masculine, the Wolffian ducts showed the same percentage of cells in mitosis (mitotic index, MI) as normal males. The MI of the Wolffian ducts was lower, but constant, if the gonad contained between 0 and 85 % of testicular tissue. The number of Leydig cells in the gonads showed a linear relationship with the percentage of testicular tissue. Apparently, the MI of the Wolffian duct does not increase with increasing ‘maleness’ and with the number of Leydig cells. Four possibilities are put forward to explain the constant level of MI: (1) The Leydig cells of hermaphrodites may be deficient in producing testosterone. (2) The Leydig cells may produce testosterone at a normal rate but the epithelial cells of the Wolffian duct may not respond to increasing levels of testosterone by increasing their mitotic activity. (3) The presence of female gonadal tissue may directly or indirectly inhibit the mitotic activity of the epithelial cells of the Wolffian duct. (4) The epithelial cells of the Wolffian duct may respond to a low threshold level of testosterone, but maximal response is only triggered by a critical higher hormone level present only in group V.
In hermaphrodites, the Wolffian duct attached to a gonad without testicular tissue and without Leydig cells, has a MI which is significantly higher than in normal females. It is suggested that circulating testosterone from the contralateral gonad is responsible for this high MI.
In the Müllerian duct a mitotic index similar to that of the normal females was only found when the gonad contained from 0 to 15 % of testicular tissue. If a gonad contained more than 15 % of testicular tissue, the MI of the attached Müllerian duct was much lower equalizing that of normal males. No influence on the growth of the Müllerian duct could be observed from the contralateral gonad.
INTRODUCTION
The Müllerian and Wolffian ducts are present in the early embryo of both sexes. After the sex of the gonads are distinguishable morphologically the ducts begin to differentiate.
In the male, the Müllerian duct regresses due to the action of the antiMüllerian hormone (AMH), which is probably secreted by the Sertoli cells (Josso, Picard & Tran, 1980). The Wolffian duct is stimulated to grow by testosterone produced by the Leydig cells (Setchell, 1978 for review).
In the female, the processes are quite different. The Müllerian duct grows autonomically, whereas the Wolffian duct regresses due to a lack of stimulation from androgens. The Müllerian duct grows since AMH probably is absent in the female (Pelliniemi & Dym, 1980, for review). It appears that the ducts become phenotypically female if they are not actively induced to become masculine (Jost, 1971).
In order to study varying degrees of influence from testicular tissue on the development and growth of the ducts, we have looked at true hermaphrodites. These are characterized by having ovarian as well as testicular tissue in their gonads. Since all degrees of male-femaleness (from 100% ovarian tissue to 100 % testicular tissue) can be seen, these were suitable to study the interaction between the masculinity of the gonad and the growth and differentiation of the Wolffian and Müllerian ducts. The growth pattern of the ducts, evaluated by the percentage of mitosis, was studied in true hermaphrodites of the foetal mouse and compared to that of normal male and female foetuses.
MATERIALS AND METHODS
The hermaphrodites used in this study were the same as the ones used by Whitten, Beamer & Byskov (1979), and were produced by pairing female mice of different strains with Balb/c-WT males.
Gonads with mesonephric tissue attached were dissected from the foetuses on day 16p. c. The morning when the copulatory plug was found, was defined as day 1 post coitum (p.c.). A total of 48 ducts and gonads from 32 foetuses were studied. In 16 foetuses both gonads and ducts were evaluated, whereas only one gonad and duct from each of the other foetuses could be used.
For comparison eight normal 16-day-old foetuses, four males and four females, of Balb/c strain were used.
The tissue is fixed in Zenker’s solution, processed for paraffin embedding, serially sectioned at 6 μm and stained with PAS and haematoxylin.
Based on the histological preparations both gonads of each hermaphrodite were categorized according to the relative proportions of male and female gonadal tissue. The male part contained testicular cords with nonmeiotic germ cells and the female part lacked testicular cords and had germ cells in meiosis. The male-femaleness of the gonads was quantified by point counting (Gundersen & Østerby, 1980; 1981) of every 10th section starting with the section in which 10 points could be counted. The point density of the grid was a priori selected to give a mean score of 80 per section, which should provide a sufficient precision (Gundersen & Østerby, 1980). Points over ovarian and testicular tissue were counted separately. All point counts of testicular and ovarian tissue of each gonad were added to give a relative evaluation of the size of the gonad, i.e. total point score, as well as a measurement of the degree of maleness. The degree of maleness was calculated as the percentage of points over the testes tissue per total point score. Depending on the percentage of testicular tissue within the gonads they were placed in one of the following five groups:
In each specimen the Wolffian and the Müllerian duct were identified. At least 250 nuclei were counted in the epithelia of the ducts, using every second section, and the number of mitoses were recorded, from which the mitotic index, i.e. the percentage of cells in mitosis (MI) was calculated. However, in three specimens the epithelium of one or the other duct was degenerated to such an extent that only 175 nuclei could be counted. The mean value ±S.E. of the MI for the Wolffian and Müllerian ducts were calculated for each of the five groups as well as for the normal foetuses; these values were plotted against the proportion of male tissue (Fig. 1). The MI for the ducts of the different groups was compared and the differences were evaluated statistically using analysis of varians.
The mitotic index (mean±s.E.) in the Müllerian duct (—○—) and the Wolffian duct (—●—) as a function of male tissue in the gonad of hermaphrodites. In the hermaphrodites the number of duct systems included in each group was, group I: 5, group II: 7, group III: 9, group IV: 12, group V: 15. The mitotic indices of the sex ducts from normal male and females are also shown. The normal male and females included four duct systems each.
The mitotic index (mean±s.E.) in the Müllerian duct (—○—) and the Wolffian duct (—●—) as a function of male tissue in the gonad of hermaphrodites. In the hermaphrodites the number of duct systems included in each group was, group I: 5, group II: 7, group III: 9, group IV: 12, group V: 15. The mitotic indices of the sex ducts from normal male and females are also shown. The normal male and females included four duct systems each.
Before any statistical treatment the data were transformed using x = Arcsin (p = the observed MI). According to Davies (1971) this is an appropriate transformation of the data before statistical analysis.
The number of Leydig cells were counted in every 10th section on the same section which had been used for point counting. The Leydig cells were distinguished from other somatic cells by their PAS-positive-reacting cytoplasm and their spherical nuclei. The total number of Leydig cells per total point score of the gonad was calculated and the mean value ±S.E. of each group was plotted against the percentage of male tissue in that group. Linearity was examined by linear regression analysis.
Further analyses were carried out to determine whether the contralateral gonad influences growth of the ducts on the other side of the foetuses. Influence from the testicular tissue was tested by comparing the MI of the ducts of a normal female with the MI from the ducts of a hermaphrodite, in which one gonad belonged to group I (i.e. 0–15 % male tissue) and the contralateral gonad contained more testicular tissue (group II-V, i.e. 15–100 % male tissue). Influence of the ovarian tissue was tested by comparing the MI from the ducts of a normal male foetus with the MI from the ducts of a hermaphrodite, in which one gonad belonged to group V (i.e. 85-100% male tissue) and the contralateral gonad contained less ovarian tissue (group I-IV, i.e. 0–85 % male tissue).
RESULTS
The MI of Wolffian and Müllerian ducts of all foetuses are shown in Table 1. The mitotic index of the Müllerian and Wolffian ducts as a function of the percentage of testicular tissue in the gonads is shown as a mean ±S.E. in Fig. 1. The gonads varied from normal ovaries (group I) to normal testes (group V) (Fig. 2, 3, 4) and were grouped according to the percentage of testicular tissue in the gonad. The number of observations in each group is also listed in Fig. 1.
Mitotic indices of Wolffian (W) and Müllerian (M) ducts of 16-day p.c. mouse hermaphrodites and normal mouse foetuses

A. Gonad with the sex ducts (Müllerian (M), Wolffian (W)) from a 16-day p.c. mouse hermaphrodite (group I). (× 90). B. Müllerian duct with mitosis, (× 350). C. Wolffian duct degenerating without mitosis, (× 350).
A. Gonad with the sex ducts (Müllerian (M), Wolffian (W)) from a 16-day p.c. mouse hermaphrodite (group V). (× 90). B. A degenerating Müllerian duct, (× 350). C. The Wolffian duct growing well with mitoses, (× 350).
A. Gonad with the sex ducts (Müllerian (M), Wolffian (W)) from a 16-day p.c. mouse hermaphrodite (group IV) with testicular cords and germ cells in meiosis (arrows), (× 90). B. Müllerian duct without mitoses, (× 350). C. Wolffian duct with mitoses, (× 350).
In the Wolffian duct the mitotic index was of the same magnitude in group I, II, III and IV (P = 0·25). The mitotic index in group V is significantly higher than in group IV (P = 0·0001). In the Müllerian duct the mitotic index decreases from group I to a nearly constant low level in group II, III, IV and V (Fig. 1). No linear relationship is found between the MI of the Müllerian duct and the percentage of male tissue in the gonad (correlation coefficient 0·67). A linear relationship exists between the mean value of the relative number of Leydig cells per gonad and the percentage of male tissue found in the gonad (correlation coefficient = 0·983) (Fig. 5). The mean of the relative number of Leydig cells is almost zero in group I and approaches 600 in group V.
The relative number of Leydig cells per gonad (mean±s.E.) in each group as a function of the percentage of male tissue in the hermaphrodite gonad. (Correlations coefficient: 0·983).
By comparing the MI of the ducts from normal male and female foetuses with the hermaphrodites, it was determined whether a gonad influenced the MI of the ducts at the other side of the foetus.
The ovarian tissue at one side of a foetus does not seem to influence the growth of the contralateral ducts with an attached gonad of group V, since the MI of the Müllerian and Wolffian ducts does not differ significantly from the MI of a normal male foetus (Table 1).
In contrast, the testicular tissue seems to influence the growth of the contralateral Wolffian duct belonging to group I. The MI of the Wolffian duct from group I is significantly higher than the MI of normal females (P = 0·02).
The MI of ducts from foetuses where both sides could be counted is shown in Table 2. The relative few data make statistical analysis difficult.
Mitotic indices on the hermaphrodites, where the Wolffian (W) and the Müllerian (M) ducts on both sides of the foetus are counted

However, in foetus number 11 and 16 the two gonads belonging to group I are both determined to be 100 % pure ovaries. The higher MI for the Wolffian duct, compared to that of normal female foetuses, support the suggestion of a stimulation from testosterone produced by the other gonad of the foetus containing testicular tissue.
There is no significant difference between the MI of the Müllerian duct from group I and normal female (P = 0·46) Table 1. This indicates that the testicular tissue has no influence on the growth of the Müllerian duct on the other side of the foetus.
DISCUSSION
This study shows that the growth of the Wolffian duct, expressed as the percentage of cells in mitosis, is not correlated to the number of Leydig cells in the ipsilateral gonad. Since the growth of the Wolffian duct is dependent on testosterone secreted by the Leydig cells, it could be expected that an increase in the number of Leydig cells would result in an increase in MI of the attached Wolffian duct. However, in these foetal mouse hermaphrodites the MI of the ducts is the same in group I to group IV although the relative number of Leydig cells gradually increases from 0 in a ‘pure’ female gonad of group I to about 600 in an 84 % male gonad of group IV. The MI of the Wolffian duct in the normal female mouse foetus is significantly lower. To explain this constant level of MI in group I to IV some possibilities are discussed.
One explanation may be that the Leydig cells of true hermaphrodites are deficient in producing testosterone, so that the number of Leydig cells is not proportional to the testosterone secretion.
Another explanation is that the Leydig cells produce normal amounts of testosterone but that the target of the testosterone, the epithelial cells of the Wolffian duct, respond differently to testosterone, due to their genetic constitution. True hermaphrodites are in fact chromosome mosaics, with a mixture of XX and XY cells (Lyon, 1969; Whitten, 1975; Polani, 1981).
It is also a possibility that the female tissue inhibits the mitotic activity of the epithelial cells of the Wolffian duct, e.g. by aromatizing testosterone or by producing substances which affect the growth of the Wolffian duct. However, such an interaction is likely to be proportional to the amount of female tissue, i.e. the effect should gradually disappear from group I to group IV.
Perhaps the most plausible explanation for the constant level of MI of the Wolffian duct is that the cells respond to a low threshold level of testosterone, but that the maximal response is only triggered by a critical higher hormone level, which is only present in group V.
The relative few data on foetuses where enough cells of the ducts could be counted on both sides, make it difficult to say to what an extent a gonad affects the growth of the ducts on the other side of the foetus. However, it seems that testosterone not only influences the growth of the attached Wolffian duct but also stimulates the contralateral duct to grow. The statistical significance between the MI of normal females and group I hermaphrodites seems to show this. It is further supported by the higher MI of the Wolffian duct from the two 100% pure ovaries, which have gonads on the other side containing testicular tissue.
The mitotic index of the Müllerian duct shows that suppression of growth is obtained when the attached gonad contains more than 15 % of testicular tissue. It appears that more than 15 % of testicular tissue is needed to produce antiMüllerian hormone (AMH) in concentrations which result in full inhibition of growth in the ipsilateral Müllerian duct.
The influence of AMH from one gonad was not expressed in an effect on the contralateral gonad. The MI of the Müllerian ducts of group I is the same as in normal females showing that the testicular tissue on the contralateral side does not influence their growth.
Acknowledgement
The authors wish to thank Mrs. Lene Ahrenst, Mrs. Ingelise Green and Mrs. Dorrit Hansen for excellent technical assistance. This work was supported by The Danish Natural Science Research Council (No. 11-0026), The Danish Medical Research Council (No. 12-2350 and No. 512-2130) and by Nordic Insulin Foundation.