Human Genetics

, Volume 126, Issue 3, pp 363–373

X chromosome inactivation in clinical practice

Authors

    • Department of Medical Genetics, Oslo University Hospital, Rikshospitalet and Faculty Division RikshospitaletUniversity of Oslo
Review Article

DOI: 10.1007/s00439-009-0670-5

Cite this article as:
Ørstavik, K.H. Hum Genet (2009) 126: 363. doi:10.1007/s00439-009-0670-5

Abstract

X chromosome inactivation (XCI) is the transcriptional silencing of the majority of genes on one of the two X chromosomes in mammalian females. Females are, therefore, mosaics for two cell lines, one with the maternal X and one with the paternal X as the active chromosome. The relative proportion of the two cell lines, the X inactivation pattern, may be analyzed by simple assays in DNA from available tissues. This review focuses on medical issues related to XCI in X-linked disorders, and on the value of X inactivation analysis in clinical practice.

Introduction

X chromosome inactivation (XCI) is the transcriptional silencing of one of the two X chromosomes in female mammalians. Females are functional mosaics for two cell lines, one with the maternal X and one with the paternal X as the active chromosome. This mosaicism has important implications for the phenotypic expression of X-linked diseases (Puck and Willard 1998; Brown and Robinson 2000; Lyon 2002; van den Veyver 2001; Ørstavik 2006; Migeon 2006. Although XCI analysis is widely used, indications and limitations for the use in clinical practice have received little attention. This review focuses on medical aspects of XCI, and XCI analysis as a diagnostic tool in clinical practice. Barbara Migeon’s (2007) recent book has been a useful resource and an inspiration in writing this article.

X chromosome inactivation

Inactivation of one of the two X chromosomes occurs in early embryonic life, about the time of implantation, and is random and permanent for all descendants of a cell. However, not all of the more than one thousand genes on the X chromosome are inactivated. Genes that escape X inactivation are spread throughout the X chromosome, although the majority are located on the short arm, especially in the pseudoautosomal region. Escape from X inactivation may be incomplete, and may vary between tissues and between individuals (Carrel et al. 1999). Extensive investigation of the expression of X-linked genes showed that 15% escaped X inactivation to some degree. For an additional 10%, X inactivation varied between females (Carrel and Willard 2005). This variation in gene expression between females may explain a portion of the phenotypic variation in females in X-linked disorders (Lyon 2005).

X inactivation analysis

Assays of XCI may be performed by direct expression analysis that requires RNA, or indirect DNA-based methylation analysis. Differential methylation analysis at the androgen-receptor gene (AR) is widely used. Heterozygosity in the CAG repeat polymorphism in exon 1 of AR distinguishes the two X chromosomes from each other in ~90% of females (Allen et al. 1992).

DNA-based XCI analysis may be performed on any tissue. Cells with high mitotic activity, such as peripheral blood cells, have a more skewed pattern than cells with lower mitotic activity (Knudsen et al. 2007). The correlation of XCI between tissues ranges from 0.5 to 0.8 in the same individual (Gale et al. 1994; Sharp et al. 2000; Knudsen et al. 2007; Bittel et al. 2008; Bolduc et al. 2008). This correlation permits XCI analysis in alternative tissues when the appropriate tissue is not available. However, XCI in one tissue cannot be assumed to be the representative for another tissue. Most studies on XCI in the general population and in patients have been performed in DNA from peripheral blood cells.

DNA from lymphoblastoid cell lines has recently been shown to exhibit a higher degree of skewed XCI than DNA from uncultured peripheral blood cells from the same individual, probably secondary to clonal selection (Plagnol et al. 2008). XCI assays should, therefore, be performed on DNA from uncultured cells.

Skewed X inactivation

Inactivation of one of the two X chromosomes is random, and the distribution of the resulting two cell lines is dependent on the number of cells present when inactivation takes place. Skewed XCI is a marked deviation from a 50:50 ratio and is arbitrarily defined, often as preferential inactivation of either the maternally or paternally inherited X chromosome in 75–80% or more of cells. Extreme skewing of X inactivation is the preferential inactivation of one X chromosome in 90–95% of cells.

Skewed X inactivation in the general female population

The prevalence of skewed XCI in normal females varies between studies (Table 1). This may be due to differences in tissues examined, age of subjects, and methods of analysis. Migeon (2007) examined X-linked glucose-6-phosphate dehydrogenase (G6PD) protein variants directly in cells from heterozygote young females, and found that 5% had extremely skewed XCI, which is in agreement with the findings of Amos-Landgraf et al. (2006) and Bolduc et al. (2008) using methylation analysis.
Table 1

Skewed XCI in normal females

Age (years)

Subjects

Tissue

Skewed XCI (%)

References

≥75%

≥80%

≥90%

Newborn

590

Blood

 

5

0.5

Amos-Landgraf et al. (2006)

Newborn

450

Blood

14

 

3

Bolduc et al. (2008)

 

Buccal swab

8

 

0.5

 

0–19

74

Blood

34

 

11

Hatakeyama et al. (2004)

20–39

90

Blood

41

 

10

Hatakeyama et al. (2004)

40–59

89

Blood

48

 

14

Hatakeyama et al. (2004)

60+

97

Blood

60

 

29

Hatakeyama et al. (2004)

<25

121

Blood

28

 

7

Sharp et al. (2000)

≤32

229

Blood

11

 

3

Busque et al. (1996)

Adult

415

Blood

 

14

4

Amos-Landgraf et al. (2006)

Childbearing age

450

Blood

28

 

5

Bolduc et al. (2008)

 

Buccal swab

14

 

1

 

Age unknown

205

Blood

 

9

3

Plenge et al. (2002)

≥60

139

Blood

48

32

16

Sharp et al. (2000)

≥73

142

Blood

 

27

8

Kristiansen et al. (2005)

101

33

Blood

 

67

18

Christensen et al. (2000)

Skewed XCI is more common in blood DNA from elderly females (Busque et al. 1996; Christensen et al. 2000; Hatakeyama et al. 2004). This age-related skewing seems to start around age 55 years, continues to increase until age 100 years (Kristiansen et al. 2005), and has been confirmed in longitudinal studies (Sandovici et al. 2004). However, in a recent large study, mothers had a skewed XCI twice as often as their newborn daughters, indicating that an effect of age on skewing of XCI may start at birth (Bolduc et al. 2008). A significant, but less pronounced, age effect has also been reported in DNA from buccal swab cells (Knudsen et al. 2007; Bolduc et al. 2008).

The reasons for age-related skewing are not clear. A stochastic clonal loss may occur (Gale et al. 1997). However, the preferential inactivation of the same X chromosome in blood and buccal swab cells (Knudsen et al. 2007; Bolduc et al. 2008), and a genetic effect on the age-related skewing (Vickers et al. 2001; Kristiansen et al. 2005) make selection seem more likely. Age-related skewing is probably due to complex mechanisms involving both stochastic and genetic factors.

Since 8% of females aged 73 years or older; and 18% of centenarians have skewed XCI ≥ 90%, many females in this age group are no longer mosaics (Table 1). The significance of this lack of mosaicism for longevity is not clear (Christensen et al. 2001). Debut of some X-linked disorders, such as sideroblastic anemia (Cazzola et al. 2000) and G6PD deficiency (Au et al. 2004, 2006) in the elderly has been attributed to age-related skewing.

Interestingly, Swierczek et al. (2008) found no increased skewing in elderly females with a novel transcriptionally based assay for XCI analysis. The implications of their findings remain to be determined.

Causes of skewed X inactivation

Chance

X inactivation is usually a stochastic event. Because the number of cells at the time of X inactivation is small, some females will have skewed XCI due to chance alone.

Genetic factors

Several families have been reported where skewed XCI seems to be an isolated dominant trait (Naumova et al. 1996). There are also families with X-linked disorders, such as hemophilia, where several carrier females in the same family have a severe phenotype associated with skewed XCI (Bicocchi et al. 2005; Renault et al. 2007). In other families both carrier and non-carrier females have skewed XCI (Ørstavik et al. 1999; Tanner et al. 1999). This may be due to a genetic factor not associated with the disorder in the family. A mutation in the X inactivation-specific transcript gene (XIST) may be a rare cause of familial skewed X inactivation (Plenge et al. 1997). Twin studies on young adult females have confirmed a genetic influence on X inactivation, with a heritability of 0.6 (Kristiansen et al. 2005). However, a recent large study of XCI in 450 neonates and their mothers revealed no correlation between XCI in the mother–child pairs (Bolduc et al. 2008). This could indicate that most of the skewing of XCI in adults is acquired secondarily (Minks et al. 2008).

Selection

Differential cell growth and/or survival rates may lead to skewed XCI. Skewing may be present early, or manifest with time leading to X-linked disease late in life. In general, selection will favor cells in which the normal X is the active X chromosome. However, in carriers of adrenoleukodystrophy, the X chromosome carrying a mutation in the ABCD1 gene is the preferentially active chromosome. This can result in disease in adult females (Migeon et al.1981; Maier et al. 2002).

The mechanisms underlying selection have not been extensively investigated. The skewed XCI in mouse carriers of Wiskott–Aldrich syndrome (WAS) was related to a migration defect of hematopoietic stem cells (Lacout et al. 2003). The skewed XCI in mice heterozygous for a null mutation in the ortholog of the human X-linked α-thalassemia/mental retardation syndrome (ATR-X) gene was caused by multiple cell selection events at different times in different tissues, i.e. by apparently independent tissue-specific events (Muers et al. 2007).

Females are mosaics

X-linked disorders have traditionally been classified as recessive or dominant. However, because of the mosaicism, X-linked disorders have a dominant phenotype at the cellular level in females (Migeon 2007), and disease expression varies widely. Dobyns et al. (2004) analyzed penetrance and severity in males and females in 32 X-linked disorders. Penetrance had a predominantly bimodal distribution in females. However, an intermediate penetrance was found in many disorders. The authors argued, therefore, that “X-linked inheritance” is a more appropriate term.

The phenotype in X-linked disorders is dependent on the way cells migrate from the blastocyst where XCI is complete to form the various organs (Sun and Tsao 2008). The phenotype may manifest as areas of affected and unaffected skin, often following the lines of Blaschko, in female carriers of X-linked disorders affecting the skin (Happle 2006). However, more frequently the two cell types are intermingled. This may lead to interaction between cells, i.e. sharing of gene products, interference between cells, competition between cell lines, and selection of cells at the chromosome and gene level (Migeon 2007). In most cases, cellular mosaicism is an advantage, and skewed XCI can be considered a “unique form of gene therapy” (Migeon 2007). One exception is the X-linked craniofrontonasal syndrome, which is consistently more severe in females than in males, and where XCI is random (Twigg et al. 2004). A patchy loss of the gene product ephrin-B1 in heterozygous females may disturb the boundary formation in the developing coronal suture (Migeon 2007).

Lethality is higher in males than in females in all age groups, from fetuses to centenarians. Functional X chromosome mosacism in females may theoretically contribute to lower mortality rates (Migeon 2007).

X inactivation and phenotype in carriers of X-linked disorders

X chromosome inactivation has an important phenotypic impact on female carriers of X-linked disorders. However, the clinical use of XCI analysis requires knowledge of the relationship between XCI and phenotype. There are few studies where the relationship between XCI and phenotype has been systematically investigated. One difficulty is how to define the phenotype; another difficulty is the rarity of many X-linked disorders. In the following paragraphs, some X-linked disorders where X inactivation and phenotype have been investigated will be discussed.

X-linked disorders usually associated with random XCI and variable phenotype in carriers

In this group of disorders, selection is not likely to take place. Carrier females are usually healthy and have random XCI. Carriers may be mildy, or in rare cases, severely affected. Extremely skewed XCI in peripheral blood cells has been reported in severely affected females for a large number of X-linked disorders, probably as a chance occurrence. Alternatively, heterozygosity for a second, undiagnosed X-linked disorder may result in skewed XCI and a severe phenotype (Migeon and Haisley-Royster 1998).

Although skewed XCI may explain a phenotype, a consistent relationship between phenotype and XCI in blood cells could not be confirmed for several X-linked disorders such as hemophilia A and B (Ørstavik et al. 2000), Fabry disease (Maier et al. 2006) and myotubular myopathy (Kristiansen et al. 2003) (Table 2). One explanation may be that the most appropriate tissue is not available for analysis. This was the case in a family with ornithine transcarbamylase deficiency, where skewed XCI was found in liver, but not in blood cells of severely affected females (Yorifuji et al. 1998).
Table 2

X-linked disorders usually associated with random XCI and variable phenotype in carriers

Disease

Gene

Phenotype

Tissue

Phenotype and XCI

References

Duchenne muscular dystrophy

DMD

Clinical

Blood

Some relationship

Pegoraro et al. (1995)

Clinical

Blood

Some relationship

Yoshioka et al. (1998)

Clinical

Blood and muscle

No clear relationship

Matthews et al. (1995)

Serum creatine kinase

Blood

No relationship

Sumita et al. (1998)

Fabry disease

GLA

Clinical and biochemical

Blood

No relationship

Maier et al. (2006)

Hemophilia A

F8

Plasma concentration of factor VIII

Blood

No relationship

Ørstavik et al. (2000)

Hemophilia B

F9

Plasma concentration of factor IX

Blood

No relationship

Ørstavik et al. (2000)

Hypohidrotic ectodermal dysplasia

EDA1

Clinical

Blood and saliva

Some relationship

Lexner et al. (2008)

Myotubular myopathy

MTM1

Clinical

Blood

No relationship

Kristiansen et al. (2003)

Nephrogenic diabetes insipidus

V2R

Clinical

Blood

Some relationship

Satoh et al. (2008)

Rare cases of severely affected female carriers of Duchenne muscular dystrophy have been associated with extreme skewing of XCI (Burn et al. 1986). However, female carriers are usually asymptomatic or mildly affected. It has been proposed that manifesting carriers with a skewed XCI had a higher risk of developing moderate to severe muscular dystrophy (Pegoraro et al. 1995). This was supported by Yoshioka et al. (1998). However, a clear relationship between XCI and clinical phenotype was not found in the study of Matthews et al. (1995), where non-manifesting and some manifesting carriers had identical patterns of XCI. XCI in muscle was random in all manifesting carriers investigated. Furthermore, no correlation was found between XCI and phenotype measured as serum creatine kinase levels (Sumita et al. 1998).

Female heterozygotes for the fragile X syndrome full mutation are often affected. It seems that more than 50% of the neurons have to express the fragile X mental retardation protein to give a normal phenotype (Oostra and Willemsen 2002). XCI is of importance for the clinical phenotype. In females with the full mutation, a significant relation was found between IQ and the proportion of normal FMR1 alleles on the active X chromosome in peripheral blood cells (de Vries et al. 1996).

X-linked disorders usually associated with skewed XCI and normal phenotype in carriers

In many severe X-linked disorders female carriers are usually asymptomatic and have extremely skewed XCI, probably because of selectively mediated favorable skewing (Table 3). Examples are ATR-X syndrome, WAS, dyskeratosis congenita, X-linked agammaglobulinemia, severe combined immunodeficiency, and the MECP2 duplication syndrome. In these disorders, a skewed XCI in the mother of a possibly affected male may be diagnostically helpful. Affected females are very rare but have been reported in ATR-X syndrome (Wada et al. 2005; Badens et al. 2006), and in WAS (Parolini et al. 1998; Lutskiy et al. 2002; Andreu et al. 2003).
Table 3

X-linked disorders usually associated with skewed XCI and normal phenotype in carriers

Disease

Gene

Tissue

XCI

References

Alpha-thalassemia/mental retardation syndrome (ATR-X)

ATR-X

Blood, saliva and hair roots

Skewed

Gibbons et al. (1992)

Barth syndrome

TAZ

Blood and skin fibroblasts

Skewed, not totally penetrant

Ørstavik et al. (1998)

Dyskeratosis congenita

DKC1

Blood, granulocytes, and mononuclear cells

Skewed

Ferraris et al. (1997)

 

Blood

Skewed

Devriendt et al. (1997)

 

Blood

Skewed

Vulliamy et al. (1997)

Dystonia–deafness–optic neuronopathy syndrome

DDP

Blood

Skewed, not totally penetrant

Ørstavik et al. (1996);

Plenge et al. (1999)

Hypohidrotic ectodermal dysplasia with immunodeficiency

NEMO

Blood

Skewed in two of three carriers

Ørstavik et al. (2006)

Agammaglobulinemia

BTK

Blood (B cells)

Skewed

Li et al. (1998)

 

Blood (B cells)

Skewed

Moschese et al. (2000)

Severe combined immunodeficiency syndrome

IL2RG

Blood (T and NK cells)

Skewed

Li et al. (1998)

MECP2 duplication

MECP2

Blood

Skewed

Van Esch et al. (2005)

Wiskott–Aldrich syndrome

WASP

Blood

Skewed

Fearon et al. (1988)

 

Fibroblasts

Random

 

Reduced penetrance for extremely skewed XCI has been reported for several disorders, such as the dystonia–deafness–optic neuronopathy syndrome (DDP) (Ørstavik et al. 1996; Plenge et al. 1999), Barth syndrome (Ørstavik et al. 1998), and hypohidrotic ectodermal dysplasia with immunodeficiency (Ørstavik et al. 2006). One reason for the skewed XCI in some but not all carrier cases may be the variability in the expression of X-linked genes among females (Carrel and Willard 2005).

X-linked disorders with reduced viability in males

Several X-linked disorders are lethal in males, and have a variable phenotype in females (Franco and Ballabio 2006) (Table 4). In females with Rett syndrome disease severity varies and XCI is usually random. It is likely that XCI has an influence on phenotype since the very rare healthy mutation carriers have favorably extremely skewed XCI (Dayer et al. 2007). The influence of XCI on phenotype in Rett syndrome has been extensively investigated, and there seems to be a tendency towards a milder phenotype in females with skewed XCI (Huppke et al. 2006; Chahrour and Zoghbi 2007). The influence of XCI may also be dependant on the mutation and affected functional domain of the MECP2 gene (Archer et al. 2007).
Table 4

X-linked disorders with reduced viability in males and affected phenotype in females

Disorder

Gene

Tissue

XCI

References

Focal dermal hypoplasia (Goltz)

PORCN

Blood

Random

Leoyklang et al. (2008)

Angioma serpiginosum

PORCN

Blood

Skewed

Blinkenberg et al. (2008)

Incontinentia pigmenti

NEMO

Blood

Skewed and random

Fusco et al. (2004)

Microphthalmia with linear skin-defects syndrome (MIDAS)

MLS

Blood

Skewed

Franco and Ballabio (2006)

OPD spectrum

 Otopalatodigital syndrome

FLN1

Blood

Skewed

Robertson et al. (2003)

 Melnick–Needles syndrome

FLN1

Blood and buccal swab

Skewed

Kristiansen et al. (2002)

 Frontometaphyseal dysplasia

FLN1

Blood

Skewed

Robertson et al. (2006)

Orofaciodigital syndrome 1

OFD1

Blood

Random and skewed

Thauvin-Robinet et al. (2006)

Rett syndrome

MECP2

Blood

Random

Chahrour and Zoghbi 2007

Rett-like syndrome

CDKL5

Blood

Random

Bahi-Buisson et al. 2008

Females with incontinentia pigmenti have a skewed XCI, and XCI is more likely to modulate the phenotype in patients with hypomorphic mutations (Fusco et al. 2004). In the otopalatodigital spectrum disorders, otopalatodigital syndrome, Melnick–Needles syndrome, and frontometaphyseal dysplasia, all caused by mutations in the FLN1 gene, females are affected and have favorably skewed XCI (Kristiansen et al. 2002; Robertson et al. 2003, 2006). XCI could not explain the great phenotypic variability in one family with Melnick–Needles syndrome (Kristiansen et al. 2002).

Mental retardation

A significant proportion of mentally retarded males have X-linked mental retardation (XLMR). Female carriers of some of the XLMR syndromes have skewed XCI, probably as a result of selection. Systematic studies of families with XLMR revealed skewed XCI in all carriers in three of 19 (Raynaud et al. 2000) and four of 20 families (Plenge et al. 2002). Skewed XCI in the mother of an affected male may, therefore, indicate the presence of one of the XLMR syndromes. In the XLMR syndrome, Börjeson–Forssman–Lehmann syndrome, female carriers are often affected and skewed XCI has been reported (Turner et al. 2004).

Recurrent spontaneous abortions

Females with recurrent spontaneous abortions (RSAs) have increased frequency of skewed XCI (Lanasa et al. 1999; Sangha et al. 1999; Robinson et al. 2001; Beever et al. 2003). One explanation could be loss of male fetuses in female carriers of undiagnosed X-linked lethal disorders. Furthermore, an excess of trisomic losses has been reported in women with RSA and skewed XCI, compared with women with RSA and random XCI (Beever et al. 2003; Bretherick et al. 2005). Skewed XCI was more strongly associated with a trisomic pregnancy than with RSA in general. This finding could theoretically be explained by a reduction in the size of the follicular pool. However, maternal XCI analysis was not considered indicated in the investigation of fetal trisomies (Bretherick et al. 2005).

Monozygotic twins

Female monozygotic (MZ) twins may differ in the expression of an X-linked disorder. Several MZ twin pairs have been reported where one twin is healthy and the co-twin severely affected with an X-linked disorder. This discordance has been attributed to skewed XCI in the affected twin in several cases, such as Duchenne muscular dystrophy (Burn et al. 1986; Richards et al. 1990; Lupski et al. 1991; Tiberio 1994), Hunter disease (Winchester et al. 1992), Fabry disease (Marguery et al. 1993; Redonnet-Vernhet et al. 1996) and hemophilia A (Bennett et al. 2008). It has been speculated that skewed X inactivation of the inner cell mass may lead to MZ twinning (Nance 1990; Lubinsky and Hall 1991). However, no difference in XCI pattern has been found between MZ twins and dizygotic twins or singletons (Monteiro et al. 1998; Kristiansen et al. 2005).

XCI and chromosome anomalies

In aneuploidies involving the X chromosome, all extra X chromosomes are inactivated, leaving only one active X chromosome. Structural abnormalities limited to the X chromosome in females are generally tolerated because of the preferential inactivation of the abnormal X chromosome. Random XCI should, therefore, be associated with a more abnormal phenotype (Leppig and Disteche 2001). However, in a study of X inactivation in 12 females with various X chromosome rearrangements, the XCI pattern was not a consistent predictor of prognosis (Schluth et al. 2007).

In balanced X;autosome translocations, the normal X chromosome is usually inactivated and the phenotype may be normal. XCI pattern was the major predictor of the clinical phenotype in 122 females with a balanced X;autosome translocation, with a more normal phenotype in females with a skewed XCI (Schmidt and Du Sart 1992). In unbalanced translocations, the abnormal X is inactivated, the inactivation spreads to the autosomal region, and the phenotype is often abnormal.

Females with severe X-linked disorders usually expressed only in males, may have a cryptic chromosomal rearrangement disrupting a putative gene. A girl with severe hypohidrotic ectodermal dysplasia had extremely skewed XCI. This finding lead to the detection of a cryptic balanced X;9 translocation that disrupted the EDA1 gene (Ørstavik et al. 2007). A similar case was reported in a female with the hyper IgM immunodeficiency syndrome (Imai et al. 2006). The identification of skewed XCI may, therefore, be helpful in the diagnosis of a cryptic X;autosome rearrangement.

XCI analysis as a tool in clinical practice

There are many reports of XCI in females, who carry mutations for X-linked disorders. Often the focus is on skewed XCI as an explanation for an affected phenotype. However, when considering XCI as a tool in clinical practice, the following should be kept in mind.

Method

Different methods of XCI analysis may give different results. XCI analysis is usually performed by indirect, i.e. by methylation analysis, not by direct expression analysis. A recent comparison of XCI using both methylation analysis and expression analysis in elderly females revealed increased skewing of XCI with methylation analysis only.

Skewed X inactivation

The definition of skewed XCI varies between studies.

Age of subjects

The frequency of skewed XCI increases after the age of ~55 years. The finding of a skewed XCI in elderly females may, therefore, be a consequence of advanced age and not related to the clinical situation.

Appropriate tissue

X chromosome inactivation analysis is usually performed on DNA from easily available tissues, such as peripheral blood cells or buccal swab epithelium. These tissues may not be appropriate for the disorder in question.

Small numbers of investigated patients

Skewed XCI in female carriers of X-linked disorders has often been reported in single cases or very small patient series. Larger series are needed.

Selection bias

Many studies may be compromised by selection bias. Carriers who have had children have been preferentially investigated and may not be representative for all carriers.

Different mutations

The influence of XCI on the phenotype may vary according to the mutation in the carrier. A relationship between XCI and phenotype may, therefore, be missed when carriers with different mutations are pooled before analysis.

Heterogeneity of inactivation

At 10% of X-linked loci inactivation varies between females. This heterogeneity may explain some phenotypic variation, but may also complicate the interpretation of results of XCI analysis.

Conclusion

Research on XCI has had a great impact on our understanding of X-linked inheritance and X-linked disorders. In a clinical setting, however, analysis of XCI currently appears to be of value in a limited number of situations. Examples of such situations are:

Females who are severely affected with a suspected or verified X-linked disorder in which clinical manifestations are generally limited to males:
  • Extremely skewed XCI may contribute to confirmation of the clinical diagnosis.

  • Random XCI does not rule out the presence of an X-linked disorder.

Mothers of undiagnosed affected males where X-linked inheritance is suspected:
  • Extremely skewed XCI is supportive of the presence of an X-linked disorder.

  • Random XCI does not rule out the presence of an X-linked disorder.

Females with suspected X chromosome anomalies:
  • Skewed XCI may point to a cryptic X chromosome anomaly.

Females with verified X chromosome anomalies:
  • X chromosome inactivation pattern may contribute to information on phenotypic consequences.

Monozygotic female twins discordant for a clinical phenotype:
  • In cases of a verified X-linked diagnosis, a discordant XCI phenotype may explain why the twins are discordant for the clinical phenotype. In cases of a suspected X-linked diagnosis, a discordant XCI phenotype may support the clinical suspicion of an underlying X-linked disorder.

X chromosome inactivation analysis will usually not be helpful diagnostically in females with mild symptoms suggestive of a possible X-linked disorder. The finding of a skewed XCI may be irrelevant for the clinical problem at hand, and may create a new problem, the possibility that the individual is a carrier of an unrelated and unknown X-linked disorder.

It seems likely that new methods for the analysis of XCI based on the direct expression analysis will prove more useful in the future.

Acknowledgments

I am thankful to Trine Prescott for critical reading of the manuscript.

Conflict of interest statement

None.

Copyright information

© Springer-Verlag 2009