Development 107, 95-105 (1989)Printed in Great Britain © The Company of Biologists Limited 1989

95

Development and fertility of ovaries in the B6.YDOM sex-reversed female

mouse

TERUKO TAKETO-HOSOTANI1*, YUTAKA NISHIOKA2, CLAUDE M. NAGAMINE3,

IRMA VILLALPANDO4 and HORACIO MERCHANT-LARIOS4

1 Urology Research Laboratory, McCill University, Royal Victoria Hospital, Montreal, Canada H3A 1A1

^Department of Biology, McGill University, Montreal, Canada H3A 1B13Howard Hughes Medical Institute, University of California, San Francisco, CA 94143, USA4Instituto de Investigaciones Biom&dicas, Universidad Nacional Autdnoma de Mexico, Mexico DF, Mexico 04510

*To whom reprint requests/correspondence should be addressed.

Summary

When the Y chromosome of Mus musculus domesticus(YDOM) was introduced onto the C57BL/6 (B6) mousebackground, half of the XY progeny (B6.YDOM) devel-oped bilateral ovaries and female internal and externalgenitalia. We examined the fertility of the B6.YDOM sex-reversed female mouse. The chromosomal sex of theindividual mouse was identified by dot hybridizationwith mouse Y chromosome-specific DNA probes. Theresults indicated that all XY females lacked regularestrous cyclicity although most were able to mate andovulate after treatment with gonadotropins. When theyhad been ovariectomized and grafted with ovaries fromthe XX female litter mate, they initiated estrous cycli-city. Reciprocally, the XX female that had received XYovarian grafts did not resume estrous cyclicity.

Development of the XY ovary was morphologicallycomparable to the XX ovary until 16 day of gestation(d.g.), when most germ cells had reached the zygotene orpachytene stage of meiotic prophase. However, by the

day of delivery (19 or 20 d.g.), no oocyte remained in themedullary cords of the XY ovary. In the control XXovary, the first generation of follicles developed in themedullary region, and 5A-3/J-hydroxysteroid dehydro-genase (3/J-HSDH) activity appeared first in the stromalcells around growing follicles by 10 days after birth. Incontrast, in the XY ovary, follicles were not formed inthe medullary region, and 3/3-HSDH activity appearedin epithelial cells of the oocyte-free medullary cords.Primordial follicles in the cortex region continued devel-opment in both the XX and XY ovaries.

These results suggest that the XY female is infertiledue to a defect inside the XY ovary. The prenatal loss ofoocytes in the medullary cords may be a key eventleading to abnormal endocrine function, and thereby,the absence of estrous cyclicity.

Key words: sex-reversal, XY ovary, oocyte, mouse.

Introduction

It has been generally accepted that the presence of theY chromosome determines the development of testes,and its absence results in the development of ovaries inmammals. Page and others have recently isolated aDNA fragment of human Y chromosome (ZFY), whichis conservatively present on the Y chromosome of mostmammalian species examined (Page et al. 1987). Theyhave shown that the ZFY sequence is translocated ontoeither the X chromosome or autosomes of most XXmale patients, translocated onto the X chromosome ofthe XX sex-reversed (Sxr) male mouse, and deletedfrom the Y chromosome of some XY female patients.These results suggest that the ZFY region contains the

gene responsible for primary testis-determination(called TDF in human or Tdy in the mouse).

On the other hand, Eicher and others have reportedthat, when the Y chromosome from Mus musculusdomesticus (or poschiavinus) is placed onto theC57BL/6J (B6) mouse background, all XY progeny(B6.Y M) fail to develop normal testes during fetallife, and some develop bilateral ovaries and the femalephenotype (Eicher et al. 1982; Eicher & Washburn,1983). These findings suggest that the mere presence ofthe Y chromosome is not sufficient to induce testiculardifferentiation. Eicher and others have postulated thatthe Tdy gene from M. m. domesticus (TdyDOM) needsto interact with another dominant autosomal gene(Tda-1) to impose testicular development, and that the

96 T. Taketo-Hosotani and others

TdyDOM gene cannot properly interact with recessiveautosomal genes {tda-1) from the B6 mouse strain. Wehave previously reported that the tda-1 -related sex-reversal can be transferred to other mouse strains, butthe effect is less than that seen when the YDOM

chromosome is placed onto the B6 background (Naga-mine etal. 19876). The mechanism of the B6.YDOM sex-reversal remains to be investigated.

It was of interest to determine if the B6.YDOM XYsex-reversed female was fertile. Eicher and others havereported that the XY female is infertile with rareexceptions; all the females that had been placed withnormal males mated; however, only one female pro-duced a litter (Eicher et al. 1982). They explained thatsterility is caused by rapid depletion of oocytes in theXY female between 4 and 8 weeks of age. The cause ofoocyte loss in the XY female was not discussed. In thepresent study, we investigated the infertility of theB6.YDOM XY female mouse.

Materials and methods

Preparation of B6. YDOM miceMale mice carrying the B6 genetic background and the Ychromosome from M. m. domesticus (N9-N13 backcrossinggenerations) were prepared as previously described (Naga-mine et al. 1987a). For matings, each B6.YDOM male mouse(50-180 days old) was housed with three B6 female mice(50-100 days old) overnight, and the presence or absence ofcopulation plugs was examined next morning. The day ofcopulation was defined as 0 day of gestation (d.g.). The day ofdelivery was defined as 0 day postpartum (pp.). The B6 maleand female mice were puchased from Charle's River (Tor-onto) .

Determination of chromosomal sexThe chromosomal sex of individual mice was determined bydot hybridization with mouse Y chromosome-specific DNAprobes according to the methods described previously(Nishioka & Lamothe, 1986; Nishioka, 1988). Briefly, tissuefrom individual fetuses or the tail tip of neonatal mice(50-100 mg wet weight) was homogenized and incubated withproteinase K (Boehringer Mannheim, W. Germany) and SDSat 60°C for 30min. The total DNA was extracted with phenolmixture, followed by chloroform and isoamylalcohol. Thenthe DNA was precipitated with ethanol in the presence ofNaCl. This procedure yielded 50-500ng DNA/sample. Aftertreatment with NaOH, 2 and 5 jig DNA of each sample wereapplied to hybridization transfer membranes (NEN Res.,Boston) and air-dried. The membrane sheet was hybridizedwith 32P-labelled mouse Y-specific DNA probe AC11 or145SC5, lxK^ctsmin"1 in 10ml hybridization solution over-night at 42 °C. The DNA probe was labelled with 32P using anick translation kit from BRL (Bethesda, ML). After washingfour times in OlxSSC at 45°C, the membrane sheet wasexposed to an X-ray film. The loading of DNA in representa-tive samples was examined by hybridization with non-specificDNA probe R17 (Nishioka, 1989).

FertilityEach female offspring (50-120 days pp.) was housed with a B6male mouse (50-150 days pp.) which was known to be fertile.The presence or absence of copulation plugs was examinedbetween 8:00 and 11:00a.m. every morning up to 7 days. If

copulation plugs were not found, mating was repeated twomore times. The female mouse that had copulated was housedsingly until the expected delivery day (19 or 20 d.g.). Ifpregnancy was not confirmed after 12d.g., this process was'repeated two more times.

Estrous cyclicityFemale offspring (70-90 or 110-150 days pp.) were housedsingly at least one week prior to examinations. Smears wereobtained from the vagina daily between 8:00 and 11:00a.m.for 3-4 weeks. The dried smears on microscope slides werefixed in absolute methanol, and stained in 4% Giemsasolution. Smears were then classified as to the stage of estrouscyclicity according to the criteria described by Nelson et al.(1982).

Transplantation of ovariesBilateral ovaries were removed from two female mice (XX orXY) (30-60 days pp.) at a time, and transplanted beneath thekidney capsule of the reciprocal female mouse (e.g. the rightovary from one mouse into the right kidney of the othermouse). After 2-3 weeks post-transplantation, smears wereobtained to determine the estrous cyclicity as describedabove. The female mice that did not show cyclicity weresacrificed and the ovarian grafts examined with light mi-croscopy. The female mice that showed clear or ambiguousestrous cyclicity were kept until 4 months pp., when smearswere taken again. After these experiments, females weresacrificed, and the condition of ovariectomy examined. Thedata were discarded when any residue of the original ovarywas found.

Induction of ovulationFemale offspring (25-65 days pp.) were injected intraperito-neally with 5i.u. pregnant mare's serum gonadotropin(PMSG) between 2:00 and 4:00p.m. on the first day, 5i.u.human chorionic gonadotropin (hCG) between 6:00 and8:00 p.m. on the third day, and sacrificed between 8:00 and10:00a.m. on the fourth day. (Gonadotropins were purchasedfrom SIGMA Chemicals.) The ovaries and oviducts wereremoved, and examined under a stereo microscope withtransillumination. Ova were flushed out from the oviduct andcounted.

Gross anatomy, light and electron microscopyBilateral gonads were dissected from offspring between12 d.g. and 3 months pp., and examined for gonadal structuresunder a stereomicroscope with transillumination. Specimenswere fixed in Karnovsky's solution (Karnovsky, 1965), post-fixed in OsC>4, and embedded in Epon. Semithin sections(1 fim thickness) were stained with toluidine blue and exam-ined with a light microscope. Serial sections from, at least, 5ovaries of each stage were examined. Ultrathin sections werestained with uranyl acetate and lead citrate, and examinedunder a Zeiss EM9 electron microscope.

Histochemical staining for J/3-HSDHFresh tissues were embedded in OCT medium (Tissue-Tek) inliquid nitrogen, and serial sections of 10 JHTI thickness were cutwith a cryostat. After air-drying on microscope slides over-night, sections were incubated in Levy's solution using dehyd-roepiandrosterone as a substrate (Levy et al. 1959).

Organ culture and testosterone determinationBilateral gonads were dissected from fetal offspring of 14 or 16d.g. and cultured individually for 3 days as described before

Fertility of B6. YDOM sex-reversed female mice 97

(Taketo & Koide, 1981). The basic culture medium wasEagle's minimum essential medium containing 10% heat-inactivated horse serum, 50i.u. ml"1 penicillin G sodium, and50^gmr' streptomycin sulfate. (All media and chemicalswere purchased from G1BCO, New York.) The culturemedium was changed every 24 h, and the spent medium wasstored at -20°C. After culture, the explants were fixed inBouin's solution and embedded in paraffin. All serial sectionsof 5;Um thickness were stained with hematoxylin and eosin.The fifth section from each end and three more sections,evenly distributed in between, were selected and photo-graphed with a light microscope. The ratio of testicular area towhole gonad was estimated by cutting and weighing thephotographs of the gonadal area. The concentration oftestosterone in the spent medium during the first day ofculture was determined by radioimmunoassay using the kitfrom Radioassay Systems Lab (Carson, California) withoutextraction of the sample. The cross-reactivity with dehydro-testosterone and androstenedione was 3-4 and 0-52%, re-spectively. The lowest concentration of assay for testosteronewas 0-1 ngml"1.

Statistical analysisAll data were analyzed statistically and evaluated using eitherChi-square test or Student's r-test.

Results

Dot hybridization with mouse Y chromosome-specificDNA probesAn example of dot hybridization is shown in Fig. 1. Asa control, total DNA was extracted from B6 male andfemale fetuses of 14 d.g. The DNA from offspring ofthe B6 female x B6.Y M male cross gave all- or-none-type spots at an intensity comparable to B6 male orfemale controls. The DNA from offspring that pos-sessed mono- or bilateral ovotestes (or testes) gaveexclusively positive results.

1 2 3 4 5 6 7 8 9 10

11 12 13 14 15 16 17 18 19 20

Fig. 1. Dot hybridization with a mouse Y-specific DNAprobe AC 11. The DNA from individual mice atconcentrations of 2-0 jug (top and third rows) and 5-0 fig(second and bottom rows) were applied onto themembrane. No. 1, B6 male, no. 2, B6 female, no. 3-20,offspring from the B6 female X B6.YDOM male cross. No. 5,7, & 8 were phenotypical males and others were females.

Table 1. Sex ratio of Fj progeny from the B6female X B6. YD male cross at various

developmental stages

Age

13 d.g.14 d.g.*16 d.g.18 d.g.0-7 d.pp.12-16 d.pp.21 d.pp.

No. ofmice

examined

18414619

10869

120

XX

bilateralovaries

(%)

44445053555150

bilateralovaries

(%)

22122832192924

XY

testicularcomponents

(%)

33442215272026

Total 411 50-6 23-0 26-3

* The distribution is significantly different from that of the total(Chi-square test, /»<001).

Sex ratio of offspring at various developmental stagesThe chromosomal sex and the gonadal sex of offspringare summarized in Table 1. Testes were characterizedby the presence of seminiferous tubules, which wereclearly distinguishable from other structures with trans-illumination or in histological sections. Ovaries wereidentified by the absence of seminiferous tubules, or thepresence of oocytes and follicles in histological sections.

At 12 d.g., 94% of 48 fetal progeny possessedsexually undifferentiated gonadal primordia, which hadovarian appearance under a dissecting microscope (datanot included in Table 1). At 14d.g., the largest percent-age of progeny possessed monolateral or bilateralovotestes (Table 1). After 16d.g., the ratio of XXfemales, XY females (with bilateral ovaries), and XYmales (with testicular components) was 1:0-5:0-5throughout development.

FertilitySix out of nine XY females that had been placed withB6 males mated as evidenced by the presence ofcopulation plugs (Table 2). None of them were preg-nant on 12 day after copulation. In contrast, 95 % of XXfemales mated with B6 males, and most of them carriedthe pregnancy to full term.

Estrous cyclicityAll XX females of 70-90 days pp. showed regularestrous cyclicity (3-5 days cycle) (Fig. 2). In contrast,

Table 2. Fertility of female progeny from the B6female x B6. YDOAi male cross

Chromosomalsex

No. of miceexamined

No. ofmice withcopulation

plugsNo. of

pregnancy

XX3OT

289

266

240*

* Significantly lower than the control XX female (Chi-square test,P< 0-005).

98 T. Taketo-Hosotani and others

(6) (5) (20) (8)20 -i

lOOn

80-

I 40-E

20-

XX XY XX/XX* XX/XY* XY/XX*Chromosomal sex

f j cyclicity+ | | | PD g | PVC ^ PD/PVC

Fig. 2. Estrous cyclicity of the female progeny from the B6female x B6.YD°M male cross. *, the XX and XY femalehad been ovariectomized and grafted with ovaries from theXX and XY female (the chromosomal sex of the ovariangraft/that of the host is indicated). Cyclicity +, regularestrous cyclicity; PD, persistent diestrous; PVC, persistentvaginal cornification; PD/PVC, shifted from PD to PVC,and then to PD. The number in parentheses indicates thenumber of mice examined in each group.

no XY females of the same age showed regular estrouscyclicity, and all stayed at persistent diestrous stage. Allof 13 older XY females (110-150 days pp.) were inpersistent vaginal cornification stage.

Estrous cyclicity after transplantation of ovariesAll XY females that had been ovariectomized andgrafted with ovaries from the XX female showedregular estrous cyclicity (Fig. 2). They continuedestrous cyclicity until, at least, 120 days pp. (data notshown). In contrast, of the seven XX females that hadbeen ovariectomized and grafted with ovaries from theXY female, three stayed in persistent diestrous (PD)stage, three in persistent vaginal cornification (PVC)stage, and one shifted between PVC and PD stages(Fig. 2). No degenerative change was observed in theXY ovarian graft when histologically examined. Allcontrol XX females that had been grafted with XXovaries showed regular estrous cyclicity.

Induction of ovulationIn the XX female, the number of ova released aftergonadotropin treatment was the highest at 25 days pp.,decreasing to the lowest level at 45 days pp., andincreasing again after 55 days pp. (Fig. 3). No signifi-cant difference was observed between the right and theleft ovaries (data not shown). In the XY female, muchsmaller numbers of ova were released, reaching themaximum (3 ova/ovary) at 35 days pp. (Fig. 3). At 55days pp. or later, no ovum was produced by the XYfemale with only a few exceptions.

B

(20)

(10)*

(30)T

^ \

(22)*

( 3 4 )

(9)*

(18)

—T

- i 1 8 ) *

(32)

A

(7)*

-°- XX-•-XY

25 ±3 35 ±3 45 + 3 55 ±2 65 + 4Age (days)

Fig. 3. Number of ova released from each ovary of thefemale progeny after treatment with gonadotropins. Thebar indicates the S.E. of the mean value. The number inparentheses indicates the total number of ovaries examined(including left and right hand sides) in each group.*, significant difference from the control XX female withStudent's t test (P<0-001).

Morphological observationsFig. 4 shows a representative XY ovotestis at 14 d.g.Sex cords had been dissociated from the surface epi-thelium by formation of tunica albuginea, and differen-tiated into testis cords in the mid-portion. Inside thetestis cord, germ cells, either in mitotic cell cycle orarrested at the prespermatogonia stage, were enclosedtogether with fetal Sertoli cells in the basement mem-brane (Fig. 5). In the cranial and caudal poles, on theother hand, sex cords remained attached to the surfaceepithelium (Fig. 4). Some germ cells had entered mei-otic prophase while others were in mitotic cell cycle(Fig. 6). These structures are characteristics of ovariandifferentiation. The location of testicular and ovariansex cords was consistent in all ovotestes although theratio of two structures varied among individual gonads.

The XY ovary was morphologically indistinguishablefrom the XX ovary between 12 and 16 d.g. At 16 d.g.,both ovaries contained abundant germ cells, most ofwhich had reached the zygotene or pachytene stage ofmeiotic prophase and distributed all over the ovary(Fig. 7, 8).

Between 17 and 19 d.g., many oocytes progressed tothe diplotene stage in the medullary cords of the controlXX ovary (Fig. 9). In contrast, all oocytes degeneratedin the medullary cords of the XY ovary while manyoocytes remained in the cortex region (Fig. 10, 11).Under the electron microscope, synaptonemal com-plexes were often seen in degenerating oocytes of theXY ovary (Fig. 11). Epithelial cells in the medullarycords were ultrastructurally indistinguishable fromthose in the cortex region.

During the second week after birth, follicles hadinitiated growth in the medullary region of the controlXX ovary and they were often undergoing atresia(Fig. 12, 13). In contrast, the XY ovary containedprimordial and growing follicles only in the cortexregion (Fig. 14). The medullary region was occupied by

Fertility of B6. YDOM sex-reversed female mice 99

Fig. 4. An XY ovotestis with the adjacent mesonephros of the B6.YDOM progeny at 14d.g. Sex cords have differentiatedinto testis cords (ts) in the mid area. The tunica albuginea (ta) has developed well beneath the surface epithelium. In thecranial and caudal poles, sex cords remain attached to the surface epithelium, characteristics of ovarian differentiation (ov).Note the distribution of stromal cells (st) in the center of the gonad and the mesonephros. md, mesonephric ducts; mt,mesonephric tubules. The bar indicates 0-1 mm.Fig. 5. Testicular region of the ovotestis shown in Fig. 4. The testis cord is composed of fetal Sertoli cells (sc) and germcells. Some germ cells are arrested at the prcspermatogonial stage (ps). Note thick layers of stromal cells (st) around thebasement membrane (bm) of the testis cord. Magnification, same as Fig. 6.Fig. 6. Ovarian region in the caudal pole of the ovotestis shown in Fig. 4. Germ cells, in either mitotic cell cycle (mt) ormeiotic prophase (mi), are packed with epithelial cells (ep) in sex cords surrounded by basement membranes (bm). Stromalcells (st) are scarce in these regions. The bar indicates 01mm.

remnants of sterile sex cords surrounded by stromaltissue (Fig. 14, 15). Around one month after birth, themedullary cords were no longer distinguishable in the

XY ovary since the medullary region became occupiedby large follicles of the cortex origin (data not shown).The number of follicles in the XY ovary was smaller

100 T. Taketo-Hosotani and others

•J4

• V v ^ '«• ' • '

md

9*•;

Fertility of B6. Y sex-reversed female mice 101

Figs 7, 8. Part of an XX and an XY ovary, respectively, atL6d.g. Most germ cells in the sex cord (see arrowheads)have entered meiotic prophase. The bar indicates 0- L mm.Fig. 9. An XX ovary at 18 d.g. Germ cells in meioticprophase are abundant in both the medullary (md) andcortex (cr) regions. The bar indicates 0- L mm.

Fig. 10. An XY ovary at 18d.g. The center of medullaryregion (md) is devoid of oocytes. Oocytes near the cortex(cr) region are undergoing degeneration (see arrowheads).Many oocytes are seen in the cortex region. Magnification,same as Fig. 9.

Fig. 11. Electron micrograph of the medullar region of the XY ovary shown in Fig. 10. Two oocytes (indicated with *) areundergoing degeneration. Note synaptonemal complexes in one of these oocytes (indicated by arrowheads). Epithelial cells(ep) around oocytes show characteristics of undifferentiated ovarian cells. A thin basal lamina (bl) separates the epithelialsex cord from stromal cells (st). bv, Blood vessels. Magnification, X45000.

102 T. Taketo-Hosotani and others

12

Fig. 12. An XX ovary at 12 days pp. The section was cut through the center region including the rete ovary (ro). Follicleshave initiated growth in the medullary region (md). Many are undergoing atresia (see arrowheads). The cortex region (cr) isoccupied by primordial follicles. The bar indicates 0-1 mm.Fig. 13. Part of the XX ovary shown in Fig. 12. Follicles in the medullary region, formed with cuboidal follicular cells (fc)and large oocytes (oo), are undergoing atresia. The rete ovary (ro) is formed by compacted epithelial cells (indicated byarrowheads), which occasionally continue to a lumen (indicated with *). The bar indicates 0-1 mm.Fig. 14. An XY ovary at 11 days pp. The medullary region (md) is completely devoid of follicles, and occupied by remnantsof sex cords. Primordial and growing follicles are seen in the cortex region (cr). Magnification, same as Fig. 12.Fig. 15. The medullary region of the XY ovary shown in Fig. 14. Sex cords are composed of epithelial cells (ep) enclosed inbasement membranes (bm). No germ cells are seen inside the sex cord. A follicle near the cortex region is undergoingatresia (indicated with *). Magnification, same as Fig. 13.

than that in the XX ovary. Around 2 months pp., only afew follicles were seen in the XY ovary whereas manyfollicles at various stages were present in the control XXovary.

3/3-HSDHIn the control XX ovary, the 3/3-HSDH activity ap-peared first in stromal cells around growing follicles inthe medullary region around 10 days pp. (Fig. 16). Incontrast, the XY ovary acquired the 3/3-HSDH activityin epithelial cells of the oocyte-free medullary cords,but not in the interstitial cells in the medullary region

(Fig. 17). Weak activity was also seen in stromal cellsaround growing follicles in the cortex region.

Development of fetal gonads and testosteroneproduction in vitroWhen gonads were dissected from fetal offspring of16d.g. and cultured for 3 days, they showed morpho-logical changes comparable to those of the newbornoffspring, i.e. loss of oocytes from the medullary cordsof the XY ovary (data not shown). The amount oftestosterone secreted from gonadal explants of 14 d.g.was positively correlated to the ratio of testicularstructures (Fig. 18). Testosterone production from the

Fertility of B6. Y sex-reversed female mice 103

md

cr cr

16 17

Fig. 16. An XX ovary at 10 days pp. stained for 3/3-HSDH. The enzymatic activity is localized in stromal cells aroundgrowing follicles in the medullary region (md) in contrast to the cortex region (cr). See Fig. 12 for details of structures. Thebar indicates 0-1 mm.Fig. 17. An XY ovary at 10 days pp. stained for 3/3-HSDH. The enzymatic activity is localized in epithelial cells forming thesex cords (see arrowheads) in the medullary region (md). Weak staining is also seen around growing follicles in the cortexregion (cr). See Fig. 14 for details of structures. Magnification, same as Fig. 16.

201• B6 testes

D B6.YDOMovotestesor ovaries

0-4 0-6Testicular area/gonad

0-8 1-0

Fig. 18. The relation between the amount of testosteronesecreted by XY ovotestes, XY ovaries, and B6 XY testes of14d.g. and the ratio of testicular structures in whole gonad.*, ten XY ovaries without testicular components (testiculararea/gonad = 0-0) were examined.

control B6 testis was proportional to that from theB6.YDOM ovotestis. None of the XY ovaries (testiculararea/gonad = 0) produced testosterone throughout theculture period (ten ovaries were examined). Similarresults were obtained with explants of 16 d.g.

Discussion

In the present study, the chromosomal sex was deter-mined by dot hybridization with Y-specific DNAprobes. We assume that the hybridization with DNAprobe ACll or 145SC5 indicates the presence of the Y

chromosome. We have previously found that the DNAprobe AC11- and 145SC5-related sequences are re-peated several hundred times over a wide range in themouse Y chromosome (Nishioka & Lamothe, 1986;Nishioka, 1988; Nishioka, unpublished observations).If translocation had occurred, it would have involved alarge segment of the Y chromosome. In a previousreport by Eicher and others of cytogenetic studies, thepresence of the Y chromosome was confirmed whileXO cells or chromosomal abnormalities were not foundin the B6.YDOM female mouse (Eicher et al. 1982).

Thus determining the chromosomal sex, we haveshown that when the B6.YDOM male mouse was crossedwith B6 females, half of the XY progeny developedbilateral ovaries and female internal and external geni-talia whereas the other half developed testicular com-ponents (bilateral testes or hermaphrodites). This resultis consistent with a previous report (Eicher & Wash-burn, 1983). The mature XY female copulated nor-mally, but none of them became pregnant. AlthoughEicher et al. (1982) reported one exceptional fertile XYfemale, we could not confirm their results.

When the XY female had been ovariectomized andgrafted with XX ovaries, regular estrous cyclicity wasinitiated. In contrast, all XX females that had beenovariectomized and grafted with XY ovaries stoppedestrous cyclicity. These results suggest that the pituitaryof the XY female can respond to normal ovarian signalsto induce estrous cyclicity. Therefore, the XY female isinfertile due to a defect inside the XY ovary. This resultdoes not exclude the possibility that an abnormalityoutside the gonad results in the defect of the XY ovaryduring development. Oocytes of the XY female couldbe made to ovulate after injection with PMSG followedby hCG (Eicher & Washburn, 1986; present results). It

104 T. Taketo-Hosotani and others

is, therefore, likely that one of the possible causes of theinfertility is abnormal endocrine function of the XYovary.

We examined the morphological development of theXY ovary. In previous studies, we investigated only XYovaries from hermaphrodites, which possessed contra-lateral ovotestes or testes (Nagamine et al. 1987a). Inthe present study, we ruled out the possible influence oftesticular tissue by examining the XY female withbilateral ovaries. The sex ratio of offspring (XX fe-males: XY females: XY males or hermaphrodites)remained constant throughout the development after 16d.g. At 14 d.g., although a larger number of theprogeny contained testicular tissue, 21 % of the XYprogeny possessed no trace of testicular components.Since testosterone production was positively correlatedwith the area of testicular structures, it is unlikely thatsome cells had differentiated into testicular cell typeswithout characteristic organization. Accordingly, it isreasonable to assume that some XY females had neverpossessed testicular components during development.Hence, the infertility of the XY ovary could be at-tributed to an abnormality common to all XY ovaries.

The XY ovary was morphologically indistinguishablefrom the XX ovary until 16 d.g., when all germ cells hadentered meiotic prophase and were distributed in sexcords over the whole gonad. Between 16 and 19 d.g.,most oocytes in medullary cords of the XY ovarydegenerated. Since synaptonemal complexes were seenin many degenerating oocytes, the degeneration ap-peared to occur in mid to late pachytene stages. In thenormal XX ovary, the first set of oocytes reached thediplotene stage and induced follicle formation in themedullary region. These follicles underwent atresia,and left steriodogenic cells in the interstitium (Mer-chant-Larios, 1984; present observations). The role ofthe steroidogenic cells derived from the first generationof atretic follicles is not fully understood. Since the XYovary lost oocytes in the medullary region prenatally, itnever developed follicles in this region. Instead, epi-thelial cells forming sterile medullary cords acquiredthe 3/3-HSDH activity while stromal cells remainednegative. We have previously reported that in W/Wv

mutant mice or busulphan-treated rats, the oocytes-freesex cords remain negative for 3/3-HSDH (Merchant-Larios & Centeno, 1981; Merchant-Larios, 1976). It isworth noting that the XX and XY epithelial cells of themedullary cords appear to respond differently to theloss of oocytes. On the other hand, we observed that inthe testicular area of the B6.YDOM ovotestis, only theinterstitial cells became positive for the 3/3-HSDHactivity (data not shown). Hence, the XY epithelialcells of the oocyte-free medullary cords are also differ-ent from the XY epithelial cells forming the seminifer-ous tubules.

Degeneration of oocytes is common during normalovarian development (Baker, 1963; Speed, 1988). How-ever, the XY female lost far more oocytes than thecontrol XX female. It has been reported that the XOfemale mouse also loses many oocytes after late pachy-tene stage (Burgoyne & Baker, 1985). Most degener-

ating oocytes appear to be located in the cortex regionof the XO ovary shown in this paper while occasionaloocytes are seen in the medullary region. Since the XOfemale mouse is usually fertile (Lyon & Hawker, 1973),the infertility of the B6.YD°M female cannot beexplained by a single dose of the X chromosome. It isalso difficult to attribute the oocyte loss to the geneticelement of XY oocytes because the fate of the XYoocyte is very different in two regions. It is more likelythat the presence of the Y chromosome in the epithelialcells disturbs the normal ovarian development of themedullary region. Although epithelial cells in the med-ullary and cortex regions are similar at the ultrastructur-al level, the interaction of epithelial cells with oocytes isclearly different between these two regions.

From the present study, we postulate that the pre-cocious death of oocytes in the medullary cords of XYovaries prevents follicular formation from these cords,leading to abnormal differentiation of steroidogeniccells. This altered developmental pathway may beresponsible for abnormal endocrine functions of the XYovary. Therefore, in the XX ovary, the growth andatresia of follicles from the medullary cords may becritical for the development of normal ovarian func-tions.

We are grateful to Dr J. F. Nelson for valuable discussions,and Jose Guadalupe Baltazar and Jamilah P. Saeed forexcellent technical assistance. This study was supported inpart by grants from the Medical Research Council of Canada(MA-9740 to T.T. and MT-6809 to Y.N.).

References

BAKER, T. G. (1963). A quantitative and cytological study of germcells in human ovaries. Proc. R. Soc. London, B. 158, 417-433.

BURGOYNE, P. S. & BAKER, T. G. (1985). Perinatal oocyte loss inXO mice and its implications for the aetiology of gonadaldysgenesis in XO women. J. Reprod. Fert. 75. 633-645.

EICHER, E. M. & WASHBURN, L. L. (1983). Inherited sex reversalin mice: identification of a new primary sex-determining gene. J.exp. Zool. 228, 297-304.

EICHER, E. M. & WASHBURN, L. L. (1986). Genetic control ofprimary sex determination in mice. Ann. Rev. Genet. 20,327-360.

EICHER, E. M.. WASHBURN, L. L., WHITNEY, J. B. & MORROW, K.

E. (1982). Mus poschiavinus Y chromosome in the C57BL/6Jmurine genome causes sex reversal. Science 217, 535-537.

KARNOVSKY, M. J. (1965). A formaldehyde-glutaraldehyde fixativeof high osmolarity for use in electron microscopy. J. Cell Biol.27, 137A.

LEVY, H., DEANE, H. W. & RUBIN, B. L. (1959). Visualization of

steroid-3/3-ol-dehydrogenase activity in tissues of intact andhypophysectomized rats. Endocrinology 65, 932-943.

LYON. M. F. & HAWKER, S. G. (1973). Reproductive lifespan inirradiated and unirradiated XO mice. Genet. Res. 21, 185-194.

MERCHANT-LARIOS, H. (1976). The role of germ cells in themorphogenesis and cytodifferentiation of the rat ovary. InProgress in Differentiation Research, (eds. N. Miiller-Be>at, C.Rosenfeld, D. Tavin & D. Viza), pp. 453-462. Amsterdam:North-Holland.

MERCHANT-LARIOS, H. (1984). Germ and somatic cell interactionsduring gonadal morphogenesis. In Uhrastructure ofReproduction, (eds. J. Van Blerkom & P. M. Motta), pp. 19-30.Netherlands: Martinus Nijhoff Publisher.

MERCHANT-LARIOS, H. & CENTENO, B. (1981). Morphogenesis of

Fertility of B6.YDOM sex-reversed female mice 105

the ovary from the sterile W/Wv mouse. In Advances inMorphology of Cells and Tissues, (eds. A. Viclrio & C. Galina),pp. 383-392. New York: Alan R. Liss.

NACAMINE, C. M., TAKETO, T. & Koo, G. C. (1987a).

Morphological development of the mouse gonad in lda-l sexreversal. Differentiation 33, 214-222.

NAGAMINE, C. M., TAKETO, T. & Koo, G. C. (19876). Studies onthe genetics of ldu-l sex reversal in the mouse. Differentiation 33,223-231.

NELSON, J. F., FELICIO, L. S., RANDALL, P. K., SIMS, C. & FINCH,

C. E. (1982). A longitudinal study of estrous cyclicity in agingC57BL/6L mice: I. Cycle frequency, length and vaginal cytology.Biol. Reprod. 27, 327-339.

NISHIOKA, Y. (1988). Application of Y chromosomal repetitivesequences to scxing mouse embryos. Teratology 38, 181-185.

NISHIOKA, Y. (1989). Genome comparison in the genus Mus: astudy with Bl, MIF, centromeric and Y chromosomal repetitivesequences. Cytogenet. Cell Genet, (in press).

NISHIOKA, Y. & LAMOTHE, E. (1986). Isolation and characterizationof a mouse Y chromosomal repetitive sequence. Genetics 113,417-432.

PAGE, D. C , MOSHER, R., SIMPSON, E. M., FISHER, E. M. C ,

MARDON, G., POLLACK, J., MCGILLIVRAY, B., DE LA CHAPELLE,

A. & BROWN, L. G. (1987). The sex-determining region of thehuman Y chromosome encodes a finger protein. Cell 24,1091-1104.

SPEED, R. M. (1988). The possible role of meiotic pairinganomalies in the atresia of human fetal oocytes. Hum. Genet. 78,260-266.

TAKETO, T. & KOIDE, S. S. (1981). In vitro development of testisand ovary from indifferent fetal mouse gonads. Devi Biol. 84,61-66.

(Accepted 14 June 1989)