There are four major sex chromosome abnormalities. Otherwise unassisted, infertility is practically inevitable in XXY Klinefelter syndrome and almost always in 45,X Turner syndrome and its variants. The other two conditions, XXX and XYY, apparently have little effect on fertility and, furthermore, are not discernibly associated with any increased risk for chromosomally abnormal offspring. Mosaic forms need to be considered on their own merits.
BIOLOGY
We need briefly to consider why X chromosome aneuploidy is associated with so little phenotypic abnormality, compared with autosomal imbalance. The important factor is dosage compensation (see also Chapter 5). Only one X in each cell needs to be fully active, and in the cells of the 46,XX female, one X chromosome is genetically inactivated (lyonization). Thus potentially detrimental effects of an X chromosomal imbalance are mitigated (although not exactly cancelled out) by inactivating a supernumerary X, or by not inactivating a sole remaining X, as the case may be.
X-Inactivation
X chromosome inactivation occurs early in blastogenesis. The process (reviewed in Avner and Heard, 2001) is controlled by the X chromosome inactivation center (XIC). The XIC works by counting the X chromosomes relative to ploidy, so that only one remains active per diploid chromosome set. Elements that direct the process include XIST and Tsix, both acting as mRNA molecules. Tsix, transcribed from the antisense DNA strand of the XIST gene, is expressed in the very early embryo and may control XIST expression at the onset of X inactivation. XIST is transcribed only from the inactive X. This transcript is the primary effector of X inactivation, and it coats the chromatin of its own chromosome (that is, in cis). Inactivation having been established, the susceptible genes on the X (the considerable majority) are now silenced, and the chromosome takes on the characteristic pattern of late timing of replication in the cell cycle. Inactivation is maintained by DNA methylation of the inactive X and accumulation on it of a novel histone variant (macroH2A). In contrast, on the X chromosome destined to be the active one, hypermethylation of DXPas34, a locus sited downstream of XIST, may be the means whereby XIST transcription is prevented. A point to be noted is this: a der(X) chromosome without an XIC locus cannot be inactivated. In males with a single X chromosome that is abnormal and has two copies of the XIST gene due to X duplication, the whole chromosome remains active and imparts a functional X disomy for the duplicated segment (Apacik et al., 1996). This illustrates the point that a second copy of the XIST gene has its inactivating effect only when it is on a different X chromosome.
An X chromosome–controlling element (XCE) affects the choice of X chromosome to be inactivated in any cell. The XCE locus is polymorphic and some alleles can push the chromosome on which they are located more strongly towards inactivation than do other alleles. This can result in skewing of inactivation, but rarely more than 70:30. Once the choice is made, the inactivation pattern remains fixed in all descendants of that cell. An exception is that, during oogenesis, reactivation restores two genetically active X chromosomes. As for the trophoblast, which is to contribute to the placenta, X inactivation is nonrandom and specifically affects the paternal X chromosome (Belmont, 1996).
Certain parts of the X chromosome are not subject to inactivation. The small pseudoautosomal regions (PAR) at the tips of the X (PAR1 at Xpter and PAR2 at Xqter) remain genetically active on both X chromosomes in females and likewise on the X and the Y in males (Rappold, 1993) (Fig. 5-1). About 15% of genes outside PAR1 and PAR2, some of which have homologs on the Y chromosome and which are mostly on Xp, escape inactivation (Carrel et al., 1999) (Fig. 5-1). Thus, in the normal state these genes function disomically in both the 46,XX female and in the 46,XY male. With these genes, it is likely to be the haploinsufficient state that is, of itself, the basis of the clinical abnormality in Turner syndrome. One well-documented locus in this respect in the PAR1 region is SHOX (short stature homeobox), which accounts for much of the height deficit (Ross et al., 2001). Some X-borne “oogenesis genes,” not necessarily subject to inactivation, may need to be present in the disomic state in order to function effectively (Disteche, 1995). Haploinsufficiency may similarly be the cause of the testicular abnormality in some azoospermic males with deletions in proximal Yq and having a Turner-like somatotype (Barbaux et al., 1995), and in Turner-like neonatal lymphedema with Yq disruption (Erickson et al., 1995).
The conceptus with an X chromosome complement in excess of the normal 46,XX or 46,XY accommodates to this imbalance by inactivating any additional X chromosome. This is nearly successful in the 47,XXX female and the 47,XXY male, in whom there is apparently normal in utero survival and a relatively mild postnatal phenotype (see p. 425). The fact that some loci are not subject to inactivation and may therefore function in the disomic (XXY), trisomic (XXX), or even quintasomic (49,XXXXX) states is likely the predominant reason for the phenotypic abnormalities associated with these karyotypes.
In females with abnormal X chromosomes, the pattern of X inactivation is usually nonrandom, particularly when the imbalance due to the abnormality is large. In the 46,X,abn(X) karyotype, with one normal X and one abnormal X, the abnormal X [abn(X)] is characteristically inactive. However, if the abnormality is a very small deletion or duplication, the inactivation pattern can be random. In the case of the X-autosome translocation heterozygote, the normal X is usually, although not invariably, inactive (Chapter 5).
Laboratory Tests for X-Inactivation
Three laboratory procedures require mention. First, is the Barr body test (Fig. 12-1). In the early years of cytogenetics, this test was a clinically important means of demonstrating the presence of the inactive X chromosome in interphase nuclei of buccal mucosal or other cells, and it does provide a nice visual illustration of the concept of X-inactivation. Second is the use of late-labeling techniques. Inactive X chromosomes replicate their DNA later in the cell cycle than the other chromosomes. Exploiting this fact, these techniques work on the principle of adding DNA precursor substances (e.g., BrdU, tritiated thymidine) late in the cell cycle and then identifying the precursor cytogenetically (Fig. 12-2). Third, analyzing the pattern of X chromosome methylation with molecular methodology can show whether inactivation is random or nonrandom. A useful assay is methylation-specific polymerase chain reaction (PCR) based on the androgen receptor gene, located at Xq13 (or any other gene with a convenient polymorphism). A skewed pattern, with one X mostly methylated and the other mostly not, is indicative of nonrandom inactivation (Kubota et al., 1999).
|
|
|
Figure 12-1. Buccal mucosal cells from (a) a 45,X female, with no Barr body present; (b) a 46,XX female showing the inactive X as a Barr body (arrow); (c) a 47,XXX female showing two Barr bodies (arrows); and (d) a 48,XXXX female with three Barr bodies (arrows). |
MEIOSIS
Meiosis proceeds differently in each of the various sex chromosome abnormalities, and each warrants separate consideration.
XXX
On theoretical grounds, one might expect the three X chromosomes to display 2:1 segregation, with the production of equal numbers of X and XX ova. But this is not the case. No discernible increased risk for chromosomally abnormal offspring of these women has been demonstrated. Apparently, only normal ova, with a single X, are regularly produced. It may be that the extra X is lost before meiosis occurs (Neri, 1984), with meiosis then proceeding as in the normal XX female. A few instances of premature ovarian failure (POF) in 47,XXX women are on record, including some even in adolescence (Holland, 2001). But since XXX and POF are both fairly common, cause and effect remain uncertain.
XXY, and XXY Mosaic States
Barring medical intervention, infertility is almost inevitable in Klinefelter syndrome, although some remarkable exceptions exist. Terzoli et al. (1992), for example, report an XXY man who fathered a daughter, with paternity testing confirming fatherhood, and they quote two other such cases. Undetected XY/XXY mosaicism could account for some of these cases. Bergère et al. (2002) showed both XY and XXY cell populations in testicular biopsies from three of four men who were nonmosaic 47,XXY on blood karyotyping and FISH analysis. These three men had small numbers of sperm identified in the biopsied tissue (one went on to have a child by in vitro fertilization [IVF]). Several workers have karyotyped sperm from XXY men, and all find an excess, albeit not a large one, of 24,XX and 24,XY sperm. Possibly, these XY and XX sperm come from XXY spermatogonial stem cells. Alternatively, the abnormal gonadal environment may of itself predispose to gonosomal nondisjunction in the XY tissue, and from that stance autosomal segregation may also be vulnerable (Bergère et al., 2002; and see Chapter 24).
|
|
|
Figure 12-2. Partial metaphases showing X inactivation: (a) a normal X chromosome, (b) an isochromosome of X long arm, (c) an X with a short arm deletion, and (d) a ring X. BrdU had been added for the last 6 hours of culturing. The inactive chromosomes, replicating at this late time in the cell cycle, incorporate BrdU extensively, and thus are palely stained. The active X stains darkly. |
Kitamura et al. (2000) found sperm suitable for ICSI (intracytoplasmic sperm injection) in the ejaculates of only four out of 52 men with 47,XXY Klinefelter syndrome. Testicular aspiration of sperm (the needle entering the lumen of the seminiferous tubule) and its use via ICSI may be a means to overcome the natural infertility, always provided that there are sperm present to be aspirated. By 2002, the total of successful pregnancies achieved numbered 38, with the outcomes being 32 karyotypically normal neonates, two karyotypically normal pregnancy losses, one healthy unkaryotyped neonate, and one XXY fetus (Tachdjian et al., 2003). The success rate is low: Levron et al. (2000) treated 20 XXY patients, from eight of whom testicular sperm was obtained, and with four couples finally achieving pregnancy (two multiple, two single).1 The observations of XX and XY sperm mentioned above suggest a chromosomal risk for children conceived at IVF, and one example is known, as noted above. In a triplet pregnancy, the three karyotypes at chorionic villus sampling (CVS) were 46,XX, 46,XY, and 47,XXY. Fetal reduction was done at 14 weeks, leaving XX and XY twins (Ron-El et al., 2000).
As for mosaic states, Giltay et al. (2000) have studied men presenting with severe oligospermia, among whom there were cases of XY/XXY and XY/XXY/XXXY. Applying FISH to sperm analysis and probing for chromosomes 18, X, and Y, the aneuploidy rate was somewhat increased compared to that in a normal population, although in fact it was similar to that for a group of normal 46,XY men with oligoasthenoteratospermia and who were ICSI candidates. It may be that the sperm abnormalities reflect the testicular defect per se, rather than being a direct consequence of the XXY constitution. Studies in the XXY mouse support this interpretation (Mroz et al., 1999).
XYY
From early meiotic studies, it was concluded that the extra Y was eliminated before the spermatocyte formed, with an X–Y bivalent usually seen at diakinesis, and more recent studies support this concept (Chandley et al., 1976; Shi and Martin, 2000a; Blanco et al., 2001). However, FISH analyses of sperm, enabling hundreds of cells to be analyzed, have shown a very small increased fraction of 24,YY spermatozoa in the ejaculate of XYY men (Table 12-1). Thus it appears that although the vast majority of XYY spermatocytes lose the extra Y before entering meiosis, a very few XYY primary spermatocytes are able to slip though and produce YY (and XY) spermatozoa. These cytogenetic findings parallel the observation that XYY men have no discernible increase in risk to have children with a sex chromosome aneuploidy. A true increased risk of a fraction of a per cent could be distinguished only with the greatest difficulty when the background population risk is of a similar order of magnitude. As for the autosomes, no convincing case exists for any increased risk for aneuploidy in the children of men with 47,XYY (Shi and Martin, 2000a).
45,X Turner Syndrome
The great majority of women with 45,X Turner syndrome (TS) are infertile and do not spontaneously menstruate or develop secondary sexual characteristics. The ovaries initially appear to be normal but begin to degenerate in midfetal life. Oocytes undergo apoptosis and disappear at an accelerated rate and in most cases are gone by the age of 2 years: “the menopause occurs before the menarche” (Federman, 1987; Modi et al., 2003). Spontaneous menstruation is uncommon but recorded (Lippe, 1991; Hovatta, 1999). Completed pregnancy in women with an apparent 45,X karyotype is very rare: there were only 25 cases documented in three recent reviews (Magee et al., 1998b; Tarani et al., 1998; Schwack and Schindler, 2000). In a Danish study based on a national TS register, none of 200 45,X women achieved a natural pregnancy (one had twins by ovum donation) (Birkebaek et al., 2002).
|
Table 12.1. Rates of 24,YY Sperm from Four 47,XYY Men, Using FISH Analysis, Compared with Control(s) from the Same Laboratory |
||||||||||||||||||||||||
|
||||||||||||||||||||||||
What is the explanation for fertility in these cases? An obvious point to consider is gonadal mosaicism, with a 46,XX cell line in the ovary. This has often been suggested but rarely proven (Birkebaek et al., 2002). Jacobs et al. (1997) undertook a systematic search in 84 subjects with TS whose standard blood karyotype was 45,X, with molecular testing of blood and of a second tissue (buccal cells), and found only two cases of X/XX mosaicism. One very thorough study is that reported in Magee et al. (1998b), concerning a 45,X woman who had had seven pregnancies, five miscarrying, one producing a healthy male, and the last terminated following demonstration of fetal cystic hygroma and a 45,X karyotype on amniocentesis. Biopsies of skin, uterus, and ovary at subsequent gynecological surgery all gave a 45,X karyotype, but molecular testing showed two alleles in ovarian DNA, indicating the presence of occult 46,XX tissue. A subtler consideration is whether a pure 45,X oocyte could proceed through meiosis I, even though the sex chromosome has no homolog with which to pair. As Blumenthal and Allanson (1997) comment, there have been put forward “a myriad of theories to account for the variability of ovarian function in TS, but, as yet, there are insufficient data to yield any conclusions.” The parental origin of the single X chromosome does not appear to correlate with a number of clinical indices (Tsezou et al., 1999), although the question of its relationship to the behavioral and personality phenotype has provided an intriguing controversy (p. 329).
X/XX, X/XX/XXX, and X/XXX Mosaicism Turner Syndrome
The relative fractions of the various karyotypes are listed in Table 12-2. For practical purposes, one should make a distinction between those mosaic women who display to some extent a TS phenotype and in whom the fraction of 45,X cells is substantial, and those of normal phenotype, who have only a low proportion (the latter noted separately below). The risk for chromosomally abnormal offspring hypothetically depends upon the degree and, crucially, the distribution of the 45,X cell line. If the gonad contains 45,X cells—in other words, if there is somatic-gonadal mosaicism—and perhaps if these abnormal cells' survival is enabled by support from surrounding 46,XX oogonia, a true increased risk may exist, although it is difficult to assess the extent to which the ascertainment in published reports has been biased (Tarani et al., 1998; Sybert, 2002). In Sybert's (2002) review, mothers with X/XX or X/XX/XXX mosaicism had had 44 live births, of which 7 (five sex chromosome aneuploidy, two trisomy 21) were chromosomally abnormal. In the Danish survey of Birkebaek et al. (2002), 27 out of 78 women with either of these mosaic states had had at least one child; and one was 45,X/46,XY with ambiguous genitalia. Uehara et al. (1999c) record the exceptional circumstance of a woman with 45,X/46,XX having had three monosomic X pregnancies, all showing fetal hydrops; she also had a normal son.
The variability of phenotype according to the degree of mosaicism is well illustrated in the report of Lespinasse et al. (1998), who studied monozygous (but not identical) triplets with 45,X/46,XX mosaicism. One child with typical TS had only 6% 45,X cells on blood karyotyping but 99% of fibroblast analysis. One sister with only mild features to suggest TS had 43% of fibroblasts with 45,X, and the third sister, of normal phenotype, had 3%. Presumably, the mosaicism existed from a very early stage, and the three-way division of the 45,X/46,XX blastocyst, or (if marginally later) of the inner cell mass, happened to cut across an asymmetric disposition of normal/monosomic cells.
The argument that all 45,X patients are really X/XX mosaic has been a long one, and is not yet settled (Jacobs et al., 1997; Quilter et al., 1998; Chu, 1999; Fernández-García et al., 2000). The existence of an XX cell line in early embryonic life, subsequently lost, has been an attractive theory to explain the tiny fraction (0.5%) of 45,X survivors at birth compared to the majority that abort. On conservative criteria, about 10% may be the fraction of TS subjects testing 45,X on routine blood analysis who are actually X/XX mosaics (Jacobs et al., 1997). Uematsu et al. (2002) suggest that the typical scenario to produce TS is a paternal meiotic error producing an abnormal gonosome, which is subsequently lost mitotically to produce the 45,X lineage.
|
Table 12.2. Relative Frequencies of Turner Syndrome Karyotypes |
||||||||||||||||||||||||
|
||||||||||||||||||||||||
Low-Level 45,X/46,XX Mosaicism in Phenotypically Normal Women
This category is to be distinguished from that of TS due to 45,X/46,XX mosaicism discussed above, and is likely to be without reproductive consequence. Horsman et al. (1987) studied 103 women who had suffered repeat (2 or more) miscarriages and found that 16% of women had varying types and degrees of X aneuploidy (Table 12-3), but none had more than 10% of cells aneuploid. All of those proceeding to skin fibroblast culture karyotyped 46,XX on this tissue. The total fraction of X aneuploid cells (1.64%) from all the women did not differ significantly from the fraction of 1.78% in an age-matched control population. Nowinski et al. (1990) came to a similar conclusion in a similar study. The observation may relate to the greater propensity for the X chromosome to undergo nondisjunction, compared with the autosomes (Marshall et al., 1996; Zijno et al., 1996).2 The case for an association with infertility is not, however, closed: Morel et al. (2002) observed that women presenting for ICSI are more likely than a control group to have two or more 45,X cells on a count of up to 50 lymphocytes.
|
Table 12.3. Occasional X Aneuploidy in Women with Recurrent (2 or more) Miscarriages |
||||||||||||||||
|
||||||||||||||||
X,abn(X) Turner Syndrome Variants
Fertility may be retained in some TS variants resulting from an effective partial X monosomy. Deletion of Xp, deletion of Xq, isochromosome Xq, and large ring X are the major categories (Fig. 12-3). In these cases, mosaicism with 45,X or 46,XX cell lines is common. Presumably a partial synapsis occurs at meiosis in the 46,X,abn(X) oocytes, with the intact segment of the abnormal X pairing with the homologous region of the normal X. A 1:1 segregation would be expected, with equal frequencies of gametes carrying either the normal X or the abnormal X, from that part of the gonad tissue containing the abn(X).
X Deletions
Mother–daughter transmission of a del(Xp) and of a del(Xq) are on record (Wandstrat et al., 2000). Palka et al. (1994) describe an apparently non-mosaic 45,X woman who had an abnormal child with an interstitial Xp deletion, del(X)(p22.2::p11.3). Upon restudy, the mother herself had one 46,X,del(X) out of 450 cells, allowing the presumption of a so-matic–gonadal mosaicism. In a more direct demonstration of gonadal mosaicism, Varela et al. (1991) studied a woman with TS and normal menstruation and who had a 46, X,del(X)(p21) daughter. They showed 5/100 cells with 46,X,del(X)(p21) in one ovary, while all cells from the other ovary, fibroblasts, and lymphocytes were 45,X. Gonadal function can vary in a family, as Zinn et al. (1997) show for a familial del(X)(p21.2). The 45,X/46,X,del(X) mother had three pregnancies, including one miscarriage, and had normal menses until age 39. Her two daughters were both 46,X,del(X).The elder one was amenorrheic at age 15, while the younger had spontaneous menarche at age 14½, with regular cycles 1 year later. A very similar family is on record in Adachi et al. (2000). It is to be noted that some X chromosomes with deletions may actually be combined del(Xp)/dup(Xq), or vice versa (Giglio et al., 2000). Simpson and Rajkovic (1999) have summarized the recorded data with respect to terminal X chromosome deletions, relating the functional ovarian phenotypes, and their summary diagram is reproduced in Figure 12-4.
|
|
|
Figure 12-3. Some sex chromosome complements: (a) normal female XX and normal male XY; (b) X and XXX females; (c) XXY and XYY males; (d and e) abnormal chromosomes from females with a ring X, an isochromosome of X long arm, an X short arm deletion, and an X long arm deletion. |
What is the identity of the maternal X chromosome in the child of a woman who herself has a Turner variant? Is it the normal X represented in the mother's blood cells with karyotype 45,X or 46,X,abn(X)? Or is there an occult 46,XX line in the ovary that could contribute either of its X chromosomes to an ovum? James et al. (1997) were able to demonstrate the latter in the case of a short (147 cm) woman whose pubertal development had been normal, and whose routine blood karyotype on two occasions was reported as 45,X. In fact, she had a minor cell line in blood (2/150 cells) and in skin fibroblasts (17/100) with a pseudoisodicentric X, and her formal karyotype was 45,X/46,X,psu idic(Xq)(qter → p21::p21 → qter). She had a normal son and daughter, and each had an X chromosome deriving from a cryptic second maternal X chromosome. Thus, presumably, her karyotype was actually 45,X/46,XX[ovary]/ 46,X,psu idic(Xq). A similar scenario with an occult 46,XX ovarian complement might apply to some other cases of fertility in X,abn(X) women.
|
|
|
Figure 12-4. Terminal Xp and Xq deletions and the associated ovarian functional phenotype, according to position of the breakpoint. (From Simpson, J. L. and Rajkovic, A. Ovarian differentiation and gonadal failure, Am. J. Med. Genet. 89, 186–200, © 1999 Am. J. Med. Genet., courtesy J. L. Simpson, and with the permis-sion of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.) |
An X deletion may be sufficiently unremarkable in its phenotypic effect in the female that it is not suspected in the mother, but only discovered fortuitously at prenatal diagnosis. Examples are noted in Wandstrat et al. (2000) of an Xp deletion, del(X)(p11.4), and in Brown et al. (2001) of an Xq deletion, del(X)(q22q26). In the latter case, the mother was having amniocentesis because of her age (39 years); her menarche had been at 12 years, she had had no difficulty in conceiving, she was 178 cm tall, and “of at least normal intelligence.” The parents decided to continue the pregnancy, and a normal girl was born. Both hers and her mother's del(Xq) showed complete inactivation, and presumably this was the basis of the normal somatic phenotypes. A single case report exists of a girl with severe hemizygous expression for an X-linked condition (adrenoleukodystrophy), the mutant gene inherited on her maternal X chromosome, and the mutation then “exposed” due to a second-hit event on her paternally inherited X, a small de novo deletion at Xq27.2–qter, which removed the normal allele (Hershkovitz et al., 2002).
Genes for height are presumed to reside in distal Xp, and the SHOX gene is the classic example (see above). Short stature is a frequent observation in del(Xp) females, and is often the reason for cytogenetic investigation. Wandstrat et al. (2000) point out the bias of ascertainment in this respect, and record their own cases of del(X)(p11.4), noted above, and another of del(X)(p22), both having normal stature.
Ring X Chromosome
Blumenthal and Allanson (1997) record a woman with mosaic ring X Turner syndrome, 45,X/46,X,r(X), who was amenorrheic until being given hormone replacement therapy. She had three pregnancies: a healthy 46,XY son, a 12-week miscarriage, and a healthy daughter with the same 45,X/46,r(X) karyotype. The latter was presumably 46,X,r(X) at conception, with postzygotic loss of the r(X) in some tissue.
Other such reports exist (Uehara et al., 1997). A rather different case is that in Matsuo et al. (2000), in which a mother and daughter were 45,X/46,X,r(X)(p22.3q28), the ratios of X:r(X) being 97:3 in the mother and 73:27 in the daughter. The ring comprised almost a complete X, but small distal Xp and Xq segments were deleted. The two X chromosomes were randomly inactivated, and consequently, presumably some “brain genes” would have been functionally nullisomic in those cells having the normal X inactivated. Thus, mental function in the mother, and more so in the daughter, was compromised. A male with r(X) is almost unknown, but Ellison et al. (2002) describe transmission from a nonmosaic 46,X,r(X) mother to her nonmosaic 46,Y,r(X) son, with mother and son both being short-statured. As expected, very little material had been lost in the tip-to-tip fusion, with breakpoints within and beyond the Xp and Xq PARs, respectively.
The tiny ring X syndrome due to absence of XIST with a functional X disomy is a quite different clinical entity that is typically associated with severe mental retardation, and is discussed on p. 431. Some tiny XIST-lacking ring X chromosomes can yet be associated with only a Turner phenotype, likely reflecting different characteristics or tissue distributions of the abnormal chromosome (Turner et al., 2000).
45,X/46,XY or 45,X/46,X,mar(Y)
A fraction of females with TS have mosaicism with a Y-containing line, either an intact or an abnormal Y chromosome, the Y line often being at a very low level (at least in peripheral blood). A structurally abnormal Y chromosome typically comprises two full copies of the functional Y euchromatin (Robinson et al., 1999a). This group has an increased risk for gonadoblastoma of around 10% (Gravholt et al., 2000; Yorifuji et al., 2001). If an abnormal Y is not seen on standard karyotyping in TS, it is most unlikely there is a cryptic cell line. Jacobs et al. (1997) examined 165 such TS subjects using very sensitive molecular methodology, and found not one case of cryptic Y chromosome mosaicism. Nishi et al. (2002a,b) and Hall (2002) note that some molecular techniques can lead to overinterpretation of Y-sequences.
45,X/46,XY and 45,X/47,XYY Mosaicism in the Male
As well as a TS phenotype as just mentioned, X/XY mosaicism is occasionally found in males presenting with hypogonadism (Telvi et al., 1999). The maleness presumably reflects the fact that the gonad contained XY cells which were able to induce effective testicular differentiation with consequent androgenizing capacity. Infertility is typical. Reddy and Sulcova (1998a) did testicular biopsy on an X/XY man and demonstrated absence of spermatogenesis; about half of the supporting cells showed a Y-signal on FISH. One X[10]/XY[90] man with moderate oligoasthenoteratozoospermia showed a two- to threefold rate for XY disomy and 18 disomy in sperm (using 18 as a representative autosome) (Newberg et al., 1998). In contrast, a man with 45,X/47,XYY mosaicism reported in Dale et al. (2002) showed normal gonosomal complements in 99.9% of sperm, and this may be due to the selective destruction of abnormal spermatids. This man had presented with infertility due to oligospermia; a normal 46,XY pregnancy was achieved with intracytoplasmic sperm injection.
X Microdeletions
Deletions of the X chromosome can range in size upwards from a single DNA base. Indeed, mutations that are small deletions appear to be more common for genes on this chromosome than for the autosomes. Here we will confine our discussion to deletions that affect more than a single gene and are demonstrated by molecular cytogenetic methodology—microdeletions.
These microdeletions can be transmitted from a 46,X,del(X) mother, who may be phenotypically normal, to a 46,Y,del(X) male conceptus. The male fetuses will be nullisomic for loci within the region of the deletion. Viability may be possible, but the absence of loci will lead to a contiguous gene syndrome. For example, an Xp22.3 deletion may remove the steroid sulfatase gene, causing ichthyosis; the Kallmann gene, causing hypogonadism and anosmia (inability to smell); the CDPX (X-linked chondrodysplasia punctata) gene(s), causing a specific skeletal dysplasia; and the OA1 (ocular albinism-1) gene (Meindl et al., 1993). Another well-described contiguous gene syndrome is the variable combination of Duchenne muscular dystrophy, retinitis pigmentosa, adrenal hypoplasia, glycerol kinase deficiency, and mental retardation, due to microdeletion within Xp21. The deletion is demonstrable by FISH, enabling accurate recognition of the carrier state in the female (Worley et al., 1995).3
A rare case is that of a nucleolar organizing region (NOR) translocation onto the long arm of an X chromosome, with a small segment of distal Xq being deleted in the process (Chen et al., 2000b). The heterozygous mother in this family was normal and had a normal menstrual history, with the “Xqs” (that is, satellited Xq) discovered only at prenatal diagnosis. Her son, with 46,Xqs,Y, was abnormal, and this presumably reflected nullisomy for a segment within Xq28. A different example concerns an interstitial insertion of an NOR into Xq11.2, which may have disrupted a nervous system locus, with spastic paraplegia observed in the hemizygous males (Tamagaki et al., 2000; and see p. 168).
X Duplications
In the rare case of the abnormal X having a duplication of X material, 1:1 segregation in the female heterozygote would be expected. The basis of the rearrangement can be a direct duplication, an inverted duplication, or an isodicentric chromosome (James et al., 1997; Shapira et al., 1997a; Matsuo et al., 1999; Kokalj Vokac et al., 2002). The typical karyotype is 46,X,dup(X), although mosaic forms are on record, such as 46,XX/46,X,dup(X) and 45,X/46,X,iso dic(X). Offspring inheriting the abn(X) are hemizygous males and heterozygous females.
Hemizygous sons in whom the abn(X) is genetically active have a functional partial X disomy and are of abnormal phenotype. If the duplication includes Xp21.1–21.2, the phenotype may include sex reversal, with ovarian formation and female or ambiguous genitalia, due to disomic expression of a gene SRVX (Rao et al., 1994). Bernstein et al. (1980) record such as case: a mother with the karyotype 46,X,dup(X)(p21pter) had a severely retarded and malformed daughter, with a female internal genital tract and karyotype 46,Y,dup(X). Another daughter and the grandmother were 46,X,dup(X), and a subsequent pregnancy was 46,Y,dup(X) at amniocentesis. As for long arm duplications, Goodman et al. (1998) studied three families with a duplication of the small distal segment Xq27–qter, the segment being joined on to the tip of the short arm; the abnormal chromosome is described as dup(X)(qter–q27.2::p22.33–qter). Males with distal Xq imbalance are very abnormal, and a phenotype resembling Prader-Willi syndrome with dup(X)(q27.2qter) has been reported (Akiyama et al., 2001; Lammer et al., 2001).
The phenotype in heterozygous daughters is less predictable. If the rule of selective lyonization holds, the abn(X) is consistently inactivated, and normality should be expected. Thus, Apacik et al. (1996) report a normal grandmother and a mother with a duplication of Xq12–q13.3 due to an inverted insertion. In them, the abn(X) was preferentially inactivated. The family came to attention only when the mother had two retarded and dysmorphic sons with 46,dup(X),Y. The mothers in the family of Goodman et al. (1998) mentioned above were also normal, typically showing preferential inactivation of the dup(X). Exceptions exist which make it difficult to give precise advice. If the lyonization rule fails—and in this setting of a dup(X) it may—and the normal X is preferentially inactivated, or if inactivation is random, a functional partial pure or mosaic X disomy exists, and this is associated with an abnormal phenotype of variable degree (Schwartz et al., 1986). Tihy et al. (1999) describe an abnormal female infant who had a direct duplication of Xq22.1–q25 (in this case, a de novo abnormality), with the dup(X) being consistently the inactive X. A normal child might have been predicted if this karyotype had been found at prenatal diagnosis. There is always the question of the inactivation pattern in unstudied tissues as a possible explanation for the apparent incongruity. In a somewhat similar case involving an Xp duplication, Kokalj Vokac et al. (2002) showed consistent inactivation of the dup(X), except for the intriguing observation of early replication within the actual breakpoint region. Equally, Armstrong et al. (2003) describe an abnormal child with 46,X,dup(X)(q22.3q26) in whom the dup(X) was preferentially inactivated, but with a part of the duplicated segment apparently escaping inactivation. They propose functional disomy restricted to this small part to have been the cause of the observed anomalies.
Some rearrangements with a combined dup/del of Xp/Xq, or vice versa, can be labeled “X-X translocations,” although the mechanics of formation may not always involve a true translocation process. In any event, this type of X rearrangement can legitimately be mentioned here. Some patients present with a Turner-like phenotype, others just with a picture of ovarian hypofunction. X-X translocations are also discussed on p. 119.
Microduplication. A very small duplication (a microduplication) can lead to a Mendelian disorder, presumably due to a dosage effect of the gene in question (cf. Charcot-Marie-Tooth disease, p. 291). Pelizaeus-Merzbacher syndrome due to a dup(X)(q22) is described on p. 292. Another even more rare neurological condition is periventricular nodular heterotopia (PNH), in which there are nests of gray matter cells remaining in the internal substance of the brain adjacent to the ventricles (thus, periventricular) that have failed to migrate out to their proper position in the cerebral cortex. Typically, PNH is due to mutation in the filamin-1 gene at Xq28. Fink et al. (1997) report a boy with PNH who had a distal Xq duplication, a very subtle cytogenetic observation, confirmed on FISH with a probe only a few megabases from the Xq telomere. His mother carried the same dup(X)(q28). In this case, the disease in the child may have been due to a dosage effect, with two functional copies of the filamin1 gene.
Sex Chromosome Polysomy
The 48,XXXX female characteristically has diminished ovarian function, and fertility in pure XXXX is on record in only one case (ascertained through a Down syndrome child) (Gardner et al., 1973b). Sterility is presumably invariable in XXXY and XXYY males, who have a further sex chromosome superadded upon the Klinefelter karyotype (Linden et al., 1995).
Y Chromosome Abnormality
Y chromosome deletions associated with male infertility are discussed in Chapter 21 (p. 351). A single family is recorded with an inter-arm insertional Yq duplication, presumed transmitted from a normal father to two normal sons. The wife of one had presented with two miscarriages, which may or may not have been related (Engelen et al., 2003).
GENETIC COUNSELING
Many of the gonosomal disorders are associated with infertility, or at least subfertility. Some present a phenotype of relatively mild abnormality. Whether or not prenatal diagnosis is chosen among those who are able to achieve pregnancy may depend on the parents' perception of the seriousness of the potential abnormal outcome. Their decision may well also be influenced by how difficult it was to achieve the pregnancy.
XXX
XXX mothers have no discernibly increased risk of bearing chromosomally abnormal children. A theoretical increased risk for children with an X aneuploidy has not been demonstrated in practice. Despite reports of chromosomally abnormal children born to XXX women, it should be emphasized, as did Dewhurst and Neri in 1978 and 1984, respectively, that when biased ascertainment is taken into account, no excess of abnormal offspring has been reported. Near-silence subsequently in the literature on this issue suggests at least a rarity of abnormal pregnancy outcomes. An additional risk estimate of 1% for a chromosomally abnormal child may be reasonable. A possibility of premature ovarian failure with 47,XXX can be brought to the attention of these women, which may assist in decisions about the timing of childbearing.
XXY
These men will hardly ever father children without recourse to IVF. The early data are certainly small, but one may propose a preliminary risk figure of about 3% for a sex chromosomal abnormality. Sperm chromosome studies indicate that in addition to this small increased risk of gonosomal aneuploidy, autosomal aneuploidy might also be implicated, although an actual case is yet to be observed. Since IVF is in any event needed, preimplantation genetic diagnosis, if available, would be the preferable prenatal procedure of choice.
XYY
To our knowledge, there is no report of a discernibly increased risk for the XYY male to have chromosomally abnormal children. A slight increase in gonosomal imbalances in sperm (Table 12-1) might nevertheless lead some to choose prenatal diagnosis.
45,X Turner Syndrome
Natural fertility is very rare. However, a 45,X woman who has spontaneous menses may possibly be fertile. Endocrine and ultrasound studies may clarify whether ovulation is occurring or likely to occur (Mazzanti et al., 1997; Paoloni-Giacobino et al., 2000a). Any period of fertility is likely to be short-lived; thus, a woman with 45,X TS who wishes to have a child should not delay in trying for a pregnancy.
Tarani et al. (1998) reviewed the literature on pregnancy outcome in (apparently) nonmosaic TS women. In all, 15 45,X women had 26 recorded pregnancies, with 9 miscarriages and 16 completed pregnancies. From these 16 pregnancies there were 13 normal children (81%), 2 stillborn, and 1 with Down syndrome. While the miscarriage rate is high, it may be that the maternal gynecology was more contributory than were fetal factors. The numbers of pregnancies proceeding through to childbirth are small, and a normality rate of 81% gives merely a broad indication. There may have been selective reporting in the literature of those with abnormality (although it is true that any natural pregnancy in a 45,X woman might warrant publication).
For the great majority of TS patients with ovarian failure, ovum donation with IVF may be one route to achieve childbearing (Hovatta, 1999). A related donor (mother, sister) would have obvious attraction. Foudila et al. (1999) report their experience with 18 women with TS, and although the rates of embryo transfer were similar to those of other women with primary ovarian failure, the miscarriage rate was high (40%); this may have been due to uterine factors. Any genetic risk to the TS patient bearing children via ovum donation is due to that of the biological parents. The improving methodology of ovum storage offers the possibility of maternal donation well ahead of the time of potential use (Gook et al., 2001), and the mother may well find this a personally satisfying thing to do. Anticipating possible artificial fertility, girls should have hormone treatment from the age of 10–12 years, to avoid uterine hypoplasia (Leclercq et al., 1992).
Mosaic 45,X Turner Syndrome
Women with 45,X mosaicism and a TS phenotype presumably carry the 45,X cell line in much of the soma and gonad. Categories include X/XX, X/XXX, and X/XX/XXX. Ovarian function is often intact, although premature failure is common (Blair et al., 2001; Sybert, 2002). There is apparently an increased risk for X monosomy in a child of theirs, and this is consonant with theoretical expectation. Normal cells in the gonad may provide support for monosomic cells that otherwise would not have survived. The upper limit of the risk may be about 15%. Whether other chromosomal defects (trisomy 21) or nonchromosomal malformations (spina bifida, heart defects, cleft palate) recorded in this setting are consequential to the aneuploidy or simply coincidental is an open question (Tarani et al., 1998). Some of these mothers might never have come to a chromosome study if they had not suffered an unfortunate pregnancy outcome.
Low-Level 45,X/46,XX Mosaicism in Phenotypically Normal Women
This category is, for practical purposes, to be distinguished from that of mosaic 45,X Turner syndrome. The discovery of a low level (a single-digit percentage) of 45,X cells in a woman presenting no phenotype traits of TS is not to be overinterpreted, neither is a reproductive risk to be exaggerated. Indeed, no such risk may apply (Horsman et al., 1987; Nowinski et al., 1990). Loss of an X chromosome is a normal concomitant of aging.2
X,abn(X) Turner Syndrome Variant
Not infrequently, women with incomplete Turner phenotypes due to a 46,X,abn(X) karyotype have normal secondary sexual development, and fertility is likely or indeed proven. The majority involve deletions of Xp or Xq (or both, due to a ring chromosome). The deletions may be of quite substantial size, with a phenotypic range from partial TS through minor menstrual abnormality. Premature ovarian failure is likely, thus practical advice for these women is that childbearing should be embarked upon earlier rather than latter. Assuming 1:1 segregation (a fair assumption), the deleted X will be transmitted in 50% of ova. Tarani et al.'s (1998) review included TS mothers with Xp deletions and ring X chromosomes, most of whose children (5 out of 6) were abnormal, with the same type of TS. It might be guessed that this weighting toward abnormality was due to publication bias.
If the ovum containing a deleted X meets an X-bearing sperm, a conceptus with the same karyotype as the mother results. If the ovum meets a Y-bearing sperm, a zygote with partial nullisomy X results, and will end in early abortion. In theory, viable offspring are in the ratio of 1:1:<1 of chromosomally normal males, normal females, and X,abn(X) females. Mosaicism in the mother may well be reflected in mosaicism in the daughter. Uehara et al. (1997) record a 45,X/46,X,r(X) mother who transmitted the ring chromosome to her child; the ring was present in 4% of the mother's cells and in 34% of the daughter's. A similar case is reported in Blumenthal and Allanson (1997); in this instance, the mother had been on hormone replacement therapy.
X/XY Male
Infertility is probable. If there is any sperm production, an IVF pregnancy might be possible. Given a possible increased risk for aneuploidy, gonosomal or autosomal (Newberg et al., 1998), preimplantation genetic diagnosis, if available, is to be considered in that setting.
X Microdeletion
The carrier can be fertile, and 1:1 segregation with respect to the normal X and the abn(X) is to be expected. This implies a 50% risk of having a son affected with the full contiguous gene syndrome, or a daughter who might display a partial phenotype. There may be an increased fetal loss rate with the 46,Y,del(X) karyotype. Prenatal diagnosis, if chosen, should be a molecular genetic exercise.
X Duplication
The female carrier has a 50% risk of transmitting the dup(X). If the pregnancy proceeds to live birth, a son would be abnormal due to X disomy. A daughter is not necessarily protected by selective inactivation, and thus phenotypic abnormality is possible.
Sex Chromosome Polysomy
Many XXXX women are of low–normal or borderline intelligence, and the questions of fertility and genetic risk may well be raised. In fact, it appears that sterility is usual. XXXY and XXYY men are undoubtedly sterile.
Notes
1. A tale of perseverance is told in Rosenlund et al. (2002). An azoospermic man with non-mosaic 47,XXY Klinefelter syndrome and his wife came to IVF; a few spermatozoa were obtained by testicular needling and used for ICSI, but with no fertilization of the two oocytes available; another oocyte retrieval produced 11 eggs but another testis biopsy got no sperm. Frozen sperm from the first biopsy produced five embryos, two of which failed to transfer, and the other three were cultured to blastocyst stage and frozen; and finally, one thawed blastocyst was transferred, a pregnancy resulted, amniocentesis 46,XY, and a healthy boy born.
2. Loss of one X (or one Y) to give an occasional 45,X cell is a normal characteristic of aging in the 46,XX female (and the 46,XY male) (Guttenbach et al., 1995). After age 44, 14% of women have at least one of 15–30 cells with gain of an X chromosome, and 21% have at least one cell with X loss (Nowinski et al., 1990). However, note the comment on p. 349 with respect to the study of Morel et al. (2002b), in which women presenting for ICSI were in fact more likely to have low-level 45,X/46,XX mosaicism than controls.
3. This approach can even be used to identify carriers for a single gene condition, if the molecular defect is an intragenic deletion. In a woman at risk to be a Duchenne carrier, and in whose family an intragenic dystrophin deletion has been shown in an affected boy, cosmids from the dystrophin gene can be used as FISH probes, and the presence of a signal on only one of her X chromosomes would confirm heterozygosity (Voskova-Goldman et al., 1997).
