d. tnher. Metab. Dis. 15 (1992) 518-525 © SSIEM and KluwerAcademicPublishers. Printed in the Netherlands
Abnor
alities of Hu
an Sex Deter
ination
M. A. FERGUSON-SMITH
Cambridge University, Department of Pathology, Tennis Court Road, Cambridge CBV2 IQP, UK Summary: Cytogenetic and molecular studies in patients with abnormalities of sex determination have been the key to the isolation and investigation of candidates for the primary testis determining factor (TDF). A gene, SRY, isolated from the sex determining region of the Y chromosome within 5 kilobases of the pairing segment boundary, has been characterized recently which fulfils the expectations of TDF. It is expressed in the embryonic gonads at the critical time of differentiation; it is highly conserved among mammals; it has the structure of a transcription regulator; and mutations within its conserved domain are found in 10% of sex-reversed XY females. The murine homologue of this gene has been shown to cause sex reversal in XX embryos following injection of a 14 kb DNA fragment containing SRY into fertilized eggs. However, most XX true hermaphrodites and a proportion of XX sex-reversed males lack SRY despite the presence of testicular differentiation. It is postulated that the constitutive activation of an X-linked gene, TDF-2, normally regulated by SRY, is responsible for male differentiation in these cases. The female phenotype of XY individuals with duplications of Xp may be the result or deletion of disruption of TDF-2.
Sex is determined by the Y chromosome. Males usually have a Y chromosome and females usually do not. In mice and men XXY individuals are male and X0 individuals are female. In Drosophila, on the other hand, XXY individuals are female and X0 individuals are male. This tells us that the primary male determining signal or testis determining factor (TDF) is carried by the Y chromosome in our species, while in Drosophila the primary signal is different and is the ratio between the number of X chromosomes and autosomes. The sex determining system is an important system to study in both species because it provides a marvellous model for the molecular basis of tissue differentiation. It is a particularly good model in man because the sex chromosome dimorphism provides such an obvious starting point. This review therefore starts appropriately with consideration of Figure 1, which shows the human X and Y chromosomes at the pachytene stage of the prophase of the first meiotic division. This reveals that the sex chromosomes pair together only at one end, their short arms. This is important because it allows the two sex chromosomes to segregate evenly into different gametes. If they did not pair, random assortment into different gametes would result in 50% of zygotes having XXY or X0 518
Human S e x Determination
519
Figure 1 The XY sex bivalent at the pachytene stage of first meiosis. The pairing segment is seen at the top with the differential parts of the X (left) and Y (right) presenting a more diffuse appearance. The testis determining factors (TDF) are located on the differential part of the Y chromosome close to the boundary of the pairing segment. TDF is therefore at risk of being transferred to the X, should an abnormal recombination event occur outside the pairing segment. Electron micrograph by A. Chandley, reproduced by permission of the publishers from Essential Medical Genetics (Connor and Ferguson-Smith 1990)
sex chromosomes, leading to offspring with the Klinefelter and Turner syndromes, respectively. It is also important that the sex determining region of the Y is in the non-pairing or differential segment so that crossing over does not occur. Crossing over might, of course, transfer male determinants from the Y to the X, leading to apparent sex reversal in XX and XY individuals. The first step towards the discovery of the nature of the sex determining mechanism was to locate the site of the male determining genes on the Y. Fortunately, karyotypephenotype studies in patients with Y-chromosome aberrations are very helpful. In short, those individuals who have retained the distal part of the Y short arm are male and those who have lost it are female. The key individuals are those with pseudodicentric isochromosomes for the long arm of the Y chromosome. These Y chromosome aberrations have breakpoints in the short arm of the Y and are essentially duplication-deficiency aberrations in which there is variable loss of the terminal part of the short arm. Most of these patients are female, without any evidence of masculinization. However, a few patients are male, and in these the breakpoints occur in the pairing segment distal to the male determining region (Ferguson-Smith et al 1987). The boundary between the pairing and non-pairing region of the Y (the pseudoautosomal boundary) therefore marks the distal limit of
J. Inher. Metab. Dis. 15 (1992)
520
Ferguson-Smith
the male determining region, while the most distal breakpoint in the non-pairing region found in female patients with long-arm isochromosomes of the Y marks the proximal limit of the male determining region. The critical region must contain testis-determining factors. Developmental studies in mouse and rabbit embryos show that the first sign of testis determination occurs in the supporting cell lineage which surround the incoming germ cells (McLaren 1988). Changes in the undifferentiated gonad occur first in the male, with the transformation of the supporting cells into Sertoli cells. The Sertoli cells secrete antiMullerian hormone which causes regression of the female internal genitalia. They also stimulate the interstitial cells to produce testosterone, which is vital for the differentiation of the Wolffian ducts and male external genitalia. The Sertoli cells form tubular structures around the incoming germ cells, inhibit meiosis, and nurture the spermatogonia. Jost (1947) showed that if the undifferentiated fetal gonad were removed before a critical stage of development,.none of this occurred and the internal and external genitalia ducts developed by default along female lines; unilateral castration in male embryos resulted in female differentiation on the same side. The Sertoli cells appear to be the important component in male differentiation. Without a Y chromosome, the supporting cell lineage of the undifferentiated gonad becomes the pre-follicular cells of the ovary. They surround the incoming germ cells which enter meiosis and become primordial follicles. As no anti-Mullerian hormone is produced, the Mullerian ducts develop into uterus tubes and upper vagina. The interstitial cells are unstimulated and Wolllian ducts do not develop. SEX REVERSAL
There are some rare and apparent exceptions to this series of events, and these exceptions have proved important in the search for the primary sex determining genes. A few patients with Klinefelter syndrome and a few patients with pure gonadal dysgenesis have what seem to be normal female and male karyotypes, respectively. They appear to be examples of sex reversal. The XX males are not quite typical of XXY Klinefelter syndrome, in that their height is below average and within the normal female range. They seldom show much intellectual impairment and gynaecomastia seems more common. They are azoospermic and have all the endocrinological features of Klinefelter syndrome, including small testes and characteristic testicular histology. It was noted 25 years ago that a few of these XX males had failed to inherit an X-linked blood group gene from their father. This suggested the hypothesis that the Xg blood group locus on the X might have been exchanged for testis determinants on the Y during an abnormal crossover event in male meiosis (Ferguson-Smith 1966). Once recombinant DNA techniques were applied to test this idea, it became clear that the majority of XX males had Y-specific sequences in their DNA (Guellaen et al 1984) and that these sequences came from the short arm of the Y. In situ hybridization using some of these sequences soon showed that part of the tip of the short arm of the Y had been transferred to the X. Could XY females result from toss of male determinants by the same mechanism? XY females tend to be taller than average women, and their sterility and sexual J. lnher. Metab. Dis. 15 (1992)
Human Sex Determination
521
infertilism is due to ovarian failure. The ovaries are represented by thin streaks of ovarian stroma without follicles, and these remnants are liable to malignancy through the formation of unusual tumours (gonadoblastoma) which often develop into dysgerminomas. Some believe that these tumours develop from XY germ cells in an abnormal environment. DNA studies in these cases only rarely show loss of Yspecific sequences and so it is concluded that most XY females result from mechanisms other than abnormal X - Y interchange. More importantly, no XY female has been shown to carry paternal X-chromosome-specific sequences (such as the Xg blood group locus), which might be expected if there had been abnormal X Y interchange during paternal meiosis. THE SEARCH FOR THE TESTIS DETERMINING FACTOR Studies in sex-reversed XX males and the rare XY females who have loss of Y-specific sequences have been important in efforts to clone candidates for TDF. Early candidates such as the Bkm sequence, which is sex-specific in some reptiles (Jones and Singh 1981) and the H-Y male specific histocompatibility antigen (Wachtel et al 1975) were soon excluded. Bkm sequences are rare and not sex-specific in man, and the H-Y antigen locus maps to the long arm of the human Y chromosome (Simpson et al t987). The first plausible candidate for T D F was described by Page and colleagues (1987). They identified an expressed gene on the short arm of the Y which coded for a zinc finger transcriptional activator gene. This gene, termed ZFY, was present in an XX male and absent in an XY female who had about t40 kilobases deleted from her Y chromosome as a result of an apparently balanced X-autosomal translocation. ZFY was found to have a homologue on the short arm of the X which escaped X-inactivation, apparently excluding the possibility of a gene dosage model for sex determination. Most XX males were shown to have the ZFY gene, which was also found to be highly conserved on the Y chromosome of other mammals. However, almost immediately a Y-positive XX male was identified who lacked ZFY (Ferguson-Smith and Affara 1988). In addition, many XX true hermaphrodites and XX males were found who appeared to have no Y sequences whatsoever. Moreover, ZFY proved to be autosomal among marsupial species that have an X - Y sex determining system. ZFY was finally excluded as a candidate for T D F by the demonstration that its expression is confined to the germ-cell lineage in the mouse and could not be demonstrated in embryonic testes which lacked germ cells. The possibility remains that ZFY has an important rote in germ cell maturation. The search for a better candidate for T D F then concentrated on those XX males that lacked ZFY yet had other more distal Y-specific sequences. Palmer et al (1989) described three XX males with less than 60 kb of the distal end of the differential segment of the Y~ The sex determining region in these patients was further reduced to 35 kb and characterization of cloned sequences from within this region revealed a new gene, SRY (for sex determining region of the Y), which was conserved and Yspecific among a wide range of mammals (Sinclair et al 1990). The SRY gene was found to map within the sex determining region of the mouse Y chromosome. This routine homologue (Sty) was shown to be deleted in an XY female mouse and
J. Inher. Metab. Dis. 15 (t992)
522
Ferguson-Smith
expressed in the developing gonads of male mouse embryos at the critical stage of testis differentiation. SRY appears to code for a transcription factor which shares a DNA-binding motif with the yeast mating-type protein, Mc. The evidence that SRY is the testis determining factor appeared to be clinched by the subsequent observation of de novo mutations in SRY in two XY females with gonadal dysgenesis (Berta et al 1990; Jager et al 1990a). At the time of writing, it now seems that about 10% of XY females may have mutations in the SRY gene. The latest and most compelling finding to confirm the testes determining nature of SRY comes from transgenic studies in the mouse. Koopman and colleagues (1991) have transferred a 14kb fragment including Sry from the male determining region of the mouse into fertilized mouse eggs. Sex-reversed XX male embryos and adult mice with sterile testes and male external genitalia were identified as chromosomally female by the presence of sex chromatin in amnion and the absence of Zfy sequences in their DNA.
SEX REVERSAL IN THE ABSENCE OF SRY
By analogy with the elegant studies of sex differentiation in Drosophila species and in Caenorhabditis elegans (Hodgkin 1990), the genetic control of sex differentiation in mammals is likely to result from a cascade of genes, each gene switching on the next gene in the series. The primary switch in Drosophila and C. elegans is the X chromosome: autosome ratio, whereas in mammals it is clearly a Y-linked factor and almost certainly the transcription factor coded by SRY. It follows that in individuals who lack SRY and in whom testicular differentiation occurs none the less, there must be a gain-of-function mutation later in the sex determination pathway. In fact, testicular differentiation could well result from the constitutive activation of the gene regulated by SRY. Several questions can now be asked: What is the nature of this gene? Where is it expressed? And where is it likely to be located? It seems most likely that the product of this gene may be found in the Sertoli cells of the testes. However, there is no need for the gene to be Y-linked. In fact there are some observations which suggest that it might map to the X chromosome. No doubt the answer wilt come from DNA-binding studies using the product of the SRY gene. Meanwhile, we can look for clues among other abnormalities of human sex determination, in particular the XX true hermaphrodites and Y-negative XX males. The phenotype of XX true hermaphrodites is characteristic. All have stature within the normal female range and the diagnosis may be apparent in infancy or childhood because of ambiguity of the external genitalia. Those cases missed in childhood will usually be apparent in adolescence because of virilization of the external genitalia and development of the mammary glands. There is usually a phallus with perineal hypospadias, a vaginal pouch and variable development of the uterus and Fallopian tubes depending on the nature of the ipsilateral gonad. The most frequent finding is an ovary on one side and a testis on the other, with suppression of the Mullerian ducts on the side of the testis. Bilateral ovotestes are usually associated with uterus and bilateral tubes and in these patients virilization is less than in those with a unilateral testis. J. lnher, Metab. Dis. 15 (1992)
Human Sex Determination
523
True hermaphrodites seem to be of average intelligence and the condition is not usually associated with other developmental malformations. Almost all cases have proved to have a Y-negative XX genotype when tested by PCR for the presence of the Y-pseudoautosomal boundary sequences. There are, however, two important exceptions in the patients described by Sinclair and colleagues (1990) and Jager and colleagues (1990b). Both have Y-specific sequences at the autosomal boundary which include SRY. One case has bilateral ovotestes and the other has an ovary on one side and a testis on the other. It is possible that mosaicism for testis determining and ovary determining gonadal cells has been the result of non-random inactivation of the X chromosome carrying the SRY gene as postulated previously (Ferguson-Smith 1966). The exact breakpoint on the X chromosome may be critical in determining whether or not the Y-specific sequences are subject to inactivation, and this may explain why most XX males are not hermaphrodites. Alternatively, only a small proportion of cells with an active SRY locus may be sufficient to ensure complete male differentiation. Mosaicism involving the Y chromosome has long been known in association with true hermaphroditism. XX/XY chimeras have been described in man (Gartler et al 1962) and artificial XX/XY chimeras have been constructed in the mouse (Tarkowski 1964), both with similar sex differentiation. In addition, true hermaphrodites are known with XX/XXY mosaicism. However, in both XX/XY chimeras and XX/XXY mosaics, it is more usual to find unambiguous male differentiation, presumably because only a small proportion of Y-bearing cells are required to ensure differentiation of a testis. Y-negative XX males, i.e. those that do not have the pseudoautosomal Y boundary sequence, show interesting phenotypic differences when compared to the majority of Y-positive XX males. Most show varying degrees of hypospadias and some have cryptorchidism (Ferguson-Smith et al 1990)- findings which suggest an aetiological reIationship with XX true hermaphrodites rather than with Y-positive XX males. Furthermore, there are a number of rare families in which there are sibships containing Y-negative true hermaphrodites, Y-negative XX males and normal males and females. The pattern of inheritance in these families is consistent with that of an X-linked mutation transmitted either by normal males or by non-manifesting females in which the mutation is subject to X-inactivation. The occurrence of these familial cases is the most important evidence in favour of an X-linked gene for sex determination, TDF-2, normally regulated by SRY. Constitutive activation of the TDF-2 gene will have no effect on an XY individual but may result in the sex reversal of an XX individual. If TDF-2 is subject to Xinactivation, a constitutive mutation in one X chromosome may lead to mosaicism for testis determining and ovary determining cells and, in cases with the appropriate proportions in the gonadal anlage, to true hermaphroditism. Perhaps the only justification for considering such a hypothesis is that it may suggest studies which might lead to an improved understanding of sex determination and differentiation. One consequence of the existence of the postulated TDF-2 is that its deletion might be expected to result in sex reversal in an XY individual. Thus a search for X-linked deletion in XY females could be rewarding. In this respect it is J. Inher. Metab. Dis. 15 (t992)
524
Ferguson-Smith
noteworthy that there are several cases reported in which a chromosome aberration resulting in the duplication of part of the distal end of the X chromosome resulted in female sex differentiation despite the presence of an apparently normal Y chromosome (Bernstein et al 1980; Scherer et al 1989). These aberrations could conceivably disrupt the T D F - 2 locus and thus interfere with the pathway to normal testis differentiation. REFERENCES
Bernstein R, Jenkins T, Dawson Bet al (1980) Female phenotype and multiple abnormalities in sibs with a Y chromosome and partial X chrornosome duplication. J Med Genet 17: 291-300. gerta P, Hawkins JR, Sinclair AH et al (1990) Genetic evidence equating SRY and the testis determining factor. Nature 348:448 450. Connor JM, Ferguson-Smith MA (1990) Essential Medical Genetics, 3rd edn. Oxford: Blackwell Scientific Publications. Ferguson-Smith MA (1966) X-Y chromosomal interchange in the aetiology of true hermaphroditism and of XX Klinefelter's syndrome. Lancet 2: 475-476. Ferguson-Smith MA, Affara NA (1988) Accidental X-Y recombination and the aetiology of XX males and true hermaphrodites. Phil Trans R Soc Lond B 322: 133-144. Ferguson-Smith MA, Affara NA, Magenis RE (1987) Ordering of Y specific sequences by deletion mapping and analysis of X-Y interchange males and females. Development 101 (Suppl.): 41-50. Ferguson-Smith MA, Cooke A, Affara NA, Boyd E, Tolmie JL (1990) Genotype phenotype correlations in XX males and their bearing on current theories of sex determination. Hum Genet 84: 198-202. Gartler SM, Waxman SM, Giblett E (1962) An XX/XY human hermaphrodite resulting from double fertilisation. Proc Natl Acad Sci USA 48: 332-335. Guellaen G, Casanova M, Bishop C et al (1984) Human XX males with Y single copy DNA fragments. Nature 307: 172-173. Hodgkin J (1990) Sex determination compared in Drosophila and Caenorhabditis. Nature 344: 721-728. J~iger RJ, Anvret M, Hall K, Scherer G (1990a) A human XY female with a frame shift mutation in the candidate testis determining gene SRY. Nature 348: 452-454. J~iger RJ, Ebensperger C, Fraccaro M, Scherer G (1990b) A XFY-negative 46,XX true hermaphrodite is positive for the Y pseudoautosomal boundary. Hum Genet 85: 666-668. Jones KW, Singh L (1981) Conserved repeated DNA sequences in vertebrate sex chromosomes. Hum Genet 58: 46-53. Jost A (1947) Recherches sur la differenciation sexuelle de l'embryon de lapin. III. Role des gonades foetales dans la differentiation sexuelle somatique. Arch Anat Micr Morphol Exp 36: 271. Koopman P, Gubbay J, Vivian N, Goodfellow P, Lovell-Badge R (1991) Male development of chromosomally female mice transgenie for Sry. Nature 351: 117-121. McLaren A (1988) Sex determination in mammals. Trends Genet 4: 153-157. Page DC, Mosher R, Simpson EM et al (1987) The sex determining region of the human Y chromosome encodes a finger protein. Cell 51: 1091-1104. Palmer MS, Sinclair AH, Berta P e t al (1989) Genetic evidence that ZFY is not the testisdetermining factor. Nature 342: 937-939. Scherer G, Schempp W, Baccichetti C et al (1989) Duplication of an Xp segment that includes the ZFX locus causes sex inversion in man. Hum Genet 81: 291-294. Simpson E, Chandler P, Ooulmy E, Disteche CM, Ferguson-Smith MA, Page DC (1987) Separation of the genetic loci for the H-Y antigen and for testis determination on human Y chromosome. Nature 326: 876-878.
J. Inher. Metab. Dis. 15 (1992)
Human Sex Determination
525
Sinclair AH, Berta P, Palmer MS et al (1990) A gene from the human sex determining region encodes a protein with homology to a conserved DNA-binding motif. Nature 346: 244. Tarkowski AJ (1964) True hermaphroditism in chimaeric mice. J Embryol Exp MorphoI 12: 735-757. Wachtel S, Ohno S, Koo G, Boyse EA (1975) Possible role for H-Y antigen in the primary determination of sex. Nature 257:235 236.
J. Inher. Metab. Dis. 15 (1992)
Abnor
alities of Hu
an Sex Deter
ination
M. A. FERGUSON-SMITH
Cambridge University, Department of Pathology, Tennis Court Road, Cambridge CBV2 IQP, UK Summary: Cytogenetic and molecular studies in patients with abnormalities of sex determination have been the key to the isolation and investigation of candidates for the primary testis determining factor (TDF). A gene, SRY, isolated from the sex determining region of the Y chromosome within 5 kilobases of the pairing segment boundary, has been characterized recently which fulfils the expectations of TDF. It is expressed in the embryonic gonads at the critical time of differentiation; it is highly conserved among mammals; it has the structure of a transcription regulator; and mutations within its conserved domain are found in 10% of sex-reversed XY females. The murine homologue of this gene has been shown to cause sex reversal in XX embryos following injection of a 14 kb DNA fragment containing SRY into fertilized eggs. However, most XX true hermaphrodites and a proportion of XX sex-reversed males lack SRY despite the presence of testicular differentiation. It is postulated that the constitutive activation of an X-linked gene, TDF-2, normally regulated by SRY, is responsible for male differentiation in these cases. The female phenotype of XY individuals with duplications of Xp may be the result or deletion of disruption of TDF-2.
Sex is determined by the Y chromosome. Males usually have a Y chromosome and females usually do not. In mice and men XXY individuals are male and X0 individuals are female. In Drosophila, on the other hand, XXY individuals are female and X0 individuals are male. This tells us that the primary male determining signal or testis determining factor (TDF) is carried by the Y chromosome in our species, while in Drosophila the primary signal is different and is the ratio between the number of X chromosomes and autosomes. The sex determining system is an important system to study in both species because it provides a marvellous model for the molecular basis of tissue differentiation. It is a particularly good model in man because the sex chromosome dimorphism provides such an obvious starting point. This review therefore starts appropriately with consideration of Figure 1, which shows the human X and Y chromosomes at the pachytene stage of the prophase of the first meiotic division. This reveals that the sex chromosomes pair together only at one end, their short arms. This is important because it allows the two sex chromosomes to segregate evenly into different gametes. If they did not pair, random assortment into different gametes would result in 50% of zygotes having XXY or X0 518
Human S e x Determination
519
Figure 1 The XY sex bivalent at the pachytene stage of first meiosis. The pairing segment is seen at the top with the differential parts of the X (left) and Y (right) presenting a more diffuse appearance. The testis determining factors (TDF) are located on the differential part of the Y chromosome close to the boundary of the pairing segment. TDF is therefore at risk of being transferred to the X, should an abnormal recombination event occur outside the pairing segment. Electron micrograph by A. Chandley, reproduced by permission of the publishers from Essential Medical Genetics (Connor and Ferguson-Smith 1990)
sex chromosomes, leading to offspring with the Klinefelter and Turner syndromes, respectively. It is also important that the sex determining region of the Y is in the non-pairing or differential segment so that crossing over does not occur. Crossing over might, of course, transfer male determinants from the Y to the X, leading to apparent sex reversal in XX and XY individuals. The first step towards the discovery of the nature of the sex determining mechanism was to locate the site of the male determining genes on the Y. Fortunately, karyotypephenotype studies in patients with Y-chromosome aberrations are very helpful. In short, those individuals who have retained the distal part of the Y short arm are male and those who have lost it are female. The key individuals are those with pseudodicentric isochromosomes for the long arm of the Y chromosome. These Y chromosome aberrations have breakpoints in the short arm of the Y and are essentially duplication-deficiency aberrations in which there is variable loss of the terminal part of the short arm. Most of these patients are female, without any evidence of masculinization. However, a few patients are male, and in these the breakpoints occur in the pairing segment distal to the male determining region (Ferguson-Smith et al 1987). The boundary between the pairing and non-pairing region of the Y (the pseudoautosomal boundary) therefore marks the distal limit of
J. Inher. Metab. Dis. 15 (1992)
520
Ferguson-Smith
the male determining region, while the most distal breakpoint in the non-pairing region found in female patients with long-arm isochromosomes of the Y marks the proximal limit of the male determining region. The critical region must contain testis-determining factors. Developmental studies in mouse and rabbit embryos show that the first sign of testis determination occurs in the supporting cell lineage which surround the incoming germ cells (McLaren 1988). Changes in the undifferentiated gonad occur first in the male, with the transformation of the supporting cells into Sertoli cells. The Sertoli cells secrete antiMullerian hormone which causes regression of the female internal genitalia. They also stimulate the interstitial cells to produce testosterone, which is vital for the differentiation of the Wolffian ducts and male external genitalia. The Sertoli cells form tubular structures around the incoming germ cells, inhibit meiosis, and nurture the spermatogonia. Jost (1947) showed that if the undifferentiated fetal gonad were removed before a critical stage of development,.none of this occurred and the internal and external genitalia ducts developed by default along female lines; unilateral castration in male embryos resulted in female differentiation on the same side. The Sertoli cells appear to be the important component in male differentiation. Without a Y chromosome, the supporting cell lineage of the undifferentiated gonad becomes the pre-follicular cells of the ovary. They surround the incoming germ cells which enter meiosis and become primordial follicles. As no anti-Mullerian hormone is produced, the Mullerian ducts develop into uterus tubes and upper vagina. The interstitial cells are unstimulated and Wolllian ducts do not develop. SEX REVERSAL
There are some rare and apparent exceptions to this series of events, and these exceptions have proved important in the search for the primary sex determining genes. A few patients with Klinefelter syndrome and a few patients with pure gonadal dysgenesis have what seem to be normal female and male karyotypes, respectively. They appear to be examples of sex reversal. The XX males are not quite typical of XXY Klinefelter syndrome, in that their height is below average and within the normal female range. They seldom show much intellectual impairment and gynaecomastia seems more common. They are azoospermic and have all the endocrinological features of Klinefelter syndrome, including small testes and characteristic testicular histology. It was noted 25 years ago that a few of these XX males had failed to inherit an X-linked blood group gene from their father. This suggested the hypothesis that the Xg blood group locus on the X might have been exchanged for testis determinants on the Y during an abnormal crossover event in male meiosis (Ferguson-Smith 1966). Once recombinant DNA techniques were applied to test this idea, it became clear that the majority of XX males had Y-specific sequences in their DNA (Guellaen et al 1984) and that these sequences came from the short arm of the Y. In situ hybridization using some of these sequences soon showed that part of the tip of the short arm of the Y had been transferred to the X. Could XY females result from toss of male determinants by the same mechanism? XY females tend to be taller than average women, and their sterility and sexual J. lnher. Metab. Dis. 15 (1992)
Human Sex Determination
521
infertilism is due to ovarian failure. The ovaries are represented by thin streaks of ovarian stroma without follicles, and these remnants are liable to malignancy through the formation of unusual tumours (gonadoblastoma) which often develop into dysgerminomas. Some believe that these tumours develop from XY germ cells in an abnormal environment. DNA studies in these cases only rarely show loss of Yspecific sequences and so it is concluded that most XY females result from mechanisms other than abnormal X - Y interchange. More importantly, no XY female has been shown to carry paternal X-chromosome-specific sequences (such as the Xg blood group locus), which might be expected if there had been abnormal X Y interchange during paternal meiosis. THE SEARCH FOR THE TESTIS DETERMINING FACTOR Studies in sex-reversed XX males and the rare XY females who have loss of Y-specific sequences have been important in efforts to clone candidates for TDF. Early candidates such as the Bkm sequence, which is sex-specific in some reptiles (Jones and Singh 1981) and the H-Y male specific histocompatibility antigen (Wachtel et al 1975) were soon excluded. Bkm sequences are rare and not sex-specific in man, and the H-Y antigen locus maps to the long arm of the human Y chromosome (Simpson et al t987). The first plausible candidate for T D F was described by Page and colleagues (1987). They identified an expressed gene on the short arm of the Y which coded for a zinc finger transcriptional activator gene. This gene, termed ZFY, was present in an XX male and absent in an XY female who had about t40 kilobases deleted from her Y chromosome as a result of an apparently balanced X-autosomal translocation. ZFY was found to have a homologue on the short arm of the X which escaped X-inactivation, apparently excluding the possibility of a gene dosage model for sex determination. Most XX males were shown to have the ZFY gene, which was also found to be highly conserved on the Y chromosome of other mammals. However, almost immediately a Y-positive XX male was identified who lacked ZFY (Ferguson-Smith and Affara 1988). In addition, many XX true hermaphrodites and XX males were found who appeared to have no Y sequences whatsoever. Moreover, ZFY proved to be autosomal among marsupial species that have an X - Y sex determining system. ZFY was finally excluded as a candidate for T D F by the demonstration that its expression is confined to the germ-cell lineage in the mouse and could not be demonstrated in embryonic testes which lacked germ cells. The possibility remains that ZFY has an important rote in germ cell maturation. The search for a better candidate for T D F then concentrated on those XX males that lacked ZFY yet had other more distal Y-specific sequences. Palmer et al (1989) described three XX males with less than 60 kb of the distal end of the differential segment of the Y~ The sex determining region in these patients was further reduced to 35 kb and characterization of cloned sequences from within this region revealed a new gene, SRY (for sex determining region of the Y), which was conserved and Yspecific among a wide range of mammals (Sinclair et al 1990). The SRY gene was found to map within the sex determining region of the mouse Y chromosome. This routine homologue (Sty) was shown to be deleted in an XY female mouse and
J. Inher. Metab. Dis. 15 (t992)
522
Ferguson-Smith
expressed in the developing gonads of male mouse embryos at the critical stage of testis differentiation. SRY appears to code for a transcription factor which shares a DNA-binding motif with the yeast mating-type protein, Mc. The evidence that SRY is the testis determining factor appeared to be clinched by the subsequent observation of de novo mutations in SRY in two XY females with gonadal dysgenesis (Berta et al 1990; Jager et al 1990a). At the time of writing, it now seems that about 10% of XY females may have mutations in the SRY gene. The latest and most compelling finding to confirm the testes determining nature of SRY comes from transgenic studies in the mouse. Koopman and colleagues (1991) have transferred a 14kb fragment including Sry from the male determining region of the mouse into fertilized mouse eggs. Sex-reversed XX male embryos and adult mice with sterile testes and male external genitalia were identified as chromosomally female by the presence of sex chromatin in amnion and the absence of Zfy sequences in their DNA.
SEX REVERSAL IN THE ABSENCE OF SRY
By analogy with the elegant studies of sex differentiation in Drosophila species and in Caenorhabditis elegans (Hodgkin 1990), the genetic control of sex differentiation in mammals is likely to result from a cascade of genes, each gene switching on the next gene in the series. The primary switch in Drosophila and C. elegans is the X chromosome: autosome ratio, whereas in mammals it is clearly a Y-linked factor and almost certainly the transcription factor coded by SRY. It follows that in individuals who lack SRY and in whom testicular differentiation occurs none the less, there must be a gain-of-function mutation later in the sex determination pathway. In fact, testicular differentiation could well result from the constitutive activation of the gene regulated by SRY. Several questions can now be asked: What is the nature of this gene? Where is it expressed? And where is it likely to be located? It seems most likely that the product of this gene may be found in the Sertoli cells of the testes. However, there is no need for the gene to be Y-linked. In fact there are some observations which suggest that it might map to the X chromosome. No doubt the answer wilt come from DNA-binding studies using the product of the SRY gene. Meanwhile, we can look for clues among other abnormalities of human sex determination, in particular the XX true hermaphrodites and Y-negative XX males. The phenotype of XX true hermaphrodites is characteristic. All have stature within the normal female range and the diagnosis may be apparent in infancy or childhood because of ambiguity of the external genitalia. Those cases missed in childhood will usually be apparent in adolescence because of virilization of the external genitalia and development of the mammary glands. There is usually a phallus with perineal hypospadias, a vaginal pouch and variable development of the uterus and Fallopian tubes depending on the nature of the ipsilateral gonad. The most frequent finding is an ovary on one side and a testis on the other, with suppression of the Mullerian ducts on the side of the testis. Bilateral ovotestes are usually associated with uterus and bilateral tubes and in these patients virilization is less than in those with a unilateral testis. J. lnher, Metab. Dis. 15 (1992)
Human Sex Determination
523
True hermaphrodites seem to be of average intelligence and the condition is not usually associated with other developmental malformations. Almost all cases have proved to have a Y-negative XX genotype when tested by PCR for the presence of the Y-pseudoautosomal boundary sequences. There are, however, two important exceptions in the patients described by Sinclair and colleagues (1990) and Jager and colleagues (1990b). Both have Y-specific sequences at the autosomal boundary which include SRY. One case has bilateral ovotestes and the other has an ovary on one side and a testis on the other. It is possible that mosaicism for testis determining and ovary determining gonadal cells has been the result of non-random inactivation of the X chromosome carrying the SRY gene as postulated previously (Ferguson-Smith 1966). The exact breakpoint on the X chromosome may be critical in determining whether or not the Y-specific sequences are subject to inactivation, and this may explain why most XX males are not hermaphrodites. Alternatively, only a small proportion of cells with an active SRY locus may be sufficient to ensure complete male differentiation. Mosaicism involving the Y chromosome has long been known in association with true hermaphroditism. XX/XY chimeras have been described in man (Gartler et al 1962) and artificial XX/XY chimeras have been constructed in the mouse (Tarkowski 1964), both with similar sex differentiation. In addition, true hermaphrodites are known with XX/XXY mosaicism. However, in both XX/XY chimeras and XX/XXY mosaics, it is more usual to find unambiguous male differentiation, presumably because only a small proportion of Y-bearing cells are required to ensure differentiation of a testis. Y-negative XX males, i.e. those that do not have the pseudoautosomal Y boundary sequence, show interesting phenotypic differences when compared to the majority of Y-positive XX males. Most show varying degrees of hypospadias and some have cryptorchidism (Ferguson-Smith et al 1990)- findings which suggest an aetiological reIationship with XX true hermaphrodites rather than with Y-positive XX males. Furthermore, there are a number of rare families in which there are sibships containing Y-negative true hermaphrodites, Y-negative XX males and normal males and females. The pattern of inheritance in these families is consistent with that of an X-linked mutation transmitted either by normal males or by non-manifesting females in which the mutation is subject to X-inactivation. The occurrence of these familial cases is the most important evidence in favour of an X-linked gene for sex determination, TDF-2, normally regulated by SRY. Constitutive activation of the TDF-2 gene will have no effect on an XY individual but may result in the sex reversal of an XX individual. If TDF-2 is subject to Xinactivation, a constitutive mutation in one X chromosome may lead to mosaicism for testis determining and ovary determining cells and, in cases with the appropriate proportions in the gonadal anlage, to true hermaphroditism. Perhaps the only justification for considering such a hypothesis is that it may suggest studies which might lead to an improved understanding of sex determination and differentiation. One consequence of the existence of the postulated TDF-2 is that its deletion might be expected to result in sex reversal in an XY individual. Thus a search for X-linked deletion in XY females could be rewarding. In this respect it is J. Inher. Metab. Dis. 15 (t992)
524
Ferguson-Smith
noteworthy that there are several cases reported in which a chromosome aberration resulting in the duplication of part of the distal end of the X chromosome resulted in female sex differentiation despite the presence of an apparently normal Y chromosome (Bernstein et al 1980; Scherer et al 1989). These aberrations could conceivably disrupt the T D F - 2 locus and thus interfere with the pathway to normal testis differentiation. REFERENCES
Bernstein R, Jenkins T, Dawson Bet al (1980) Female phenotype and multiple abnormalities in sibs with a Y chromosome and partial X chrornosome duplication. J Med Genet 17: 291-300. gerta P, Hawkins JR, Sinclair AH et al (1990) Genetic evidence equating SRY and the testis determining factor. Nature 348:448 450. Connor JM, Ferguson-Smith MA (1990) Essential Medical Genetics, 3rd edn. Oxford: Blackwell Scientific Publications. Ferguson-Smith MA (1966) X-Y chromosomal interchange in the aetiology of true hermaphroditism and of XX Klinefelter's syndrome. Lancet 2: 475-476. Ferguson-Smith MA, Affara NA (1988) Accidental X-Y recombination and the aetiology of XX males and true hermaphrodites. Phil Trans R Soc Lond B 322: 133-144. Ferguson-Smith MA, Affara NA, Magenis RE (1987) Ordering of Y specific sequences by deletion mapping and analysis of X-Y interchange males and females. Development 101 (Suppl.): 41-50. Ferguson-Smith MA, Cooke A, Affara NA, Boyd E, Tolmie JL (1990) Genotype phenotype correlations in XX males and their bearing on current theories of sex determination. Hum Genet 84: 198-202. Gartler SM, Waxman SM, Giblett E (1962) An XX/XY human hermaphrodite resulting from double fertilisation. Proc Natl Acad Sci USA 48: 332-335. Guellaen G, Casanova M, Bishop C et al (1984) Human XX males with Y single copy DNA fragments. Nature 307: 172-173. Hodgkin J (1990) Sex determination compared in Drosophila and Caenorhabditis. Nature 344: 721-728. J~iger RJ, Anvret M, Hall K, Scherer G (1990a) A human XY female with a frame shift mutation in the candidate testis determining gene SRY. Nature 348: 452-454. J~iger RJ, Ebensperger C, Fraccaro M, Scherer G (1990b) A XFY-negative 46,XX true hermaphrodite is positive for the Y pseudoautosomal boundary. Hum Genet 85: 666-668. Jones KW, Singh L (1981) Conserved repeated DNA sequences in vertebrate sex chromosomes. Hum Genet 58: 46-53. Jost A (1947) Recherches sur la differenciation sexuelle de l'embryon de lapin. III. Role des gonades foetales dans la differentiation sexuelle somatique. Arch Anat Micr Morphol Exp 36: 271. Koopman P, Gubbay J, Vivian N, Goodfellow P, Lovell-Badge R (1991) Male development of chromosomally female mice transgenie for Sry. Nature 351: 117-121. McLaren A (1988) Sex determination in mammals. Trends Genet 4: 153-157. Page DC, Mosher R, Simpson EM et al (1987) The sex determining region of the human Y chromosome encodes a finger protein. Cell 51: 1091-1104. Palmer MS, Sinclair AH, Berta P e t al (1989) Genetic evidence that ZFY is not the testisdetermining factor. Nature 342: 937-939. Scherer G, Schempp W, Baccichetti C et al (1989) Duplication of an Xp segment that includes the ZFX locus causes sex inversion in man. Hum Genet 81: 291-294. Simpson E, Chandler P, Ooulmy E, Disteche CM, Ferguson-Smith MA, Page DC (1987) Separation of the genetic loci for the H-Y antigen and for testis determination on human Y chromosome. Nature 326: 876-878.
J. Inher. Metab. Dis. 15 (1992)
Human Sex Determination
525
Sinclair AH, Berta P, Palmer MS et al (1990) A gene from the human sex determining region encodes a protein with homology to a conserved DNA-binding motif. Nature 346: 244. Tarkowski AJ (1964) True hermaphroditism in chimaeric mice. J Embryol Exp MorphoI 12: 735-757. Wachtel S, Ohno S, Koo G, Boyse EA (1975) Possible role for H-Y antigen in the primary determination of sex. Nature 257:235 236.
J. Inher. Metab. Dis. 15 (1992)
