J Med Genet 1996;33:177-183

Hypothesis

Hypomelanosis of Ito and X;autosometranslocations: a unifying hypothesis

Eli Hatchwell

AbstractHypomelanosis of Ito is a sporadic multi-system disorder known to be associatedin many cases with chromosomal mo-saicism. While no particular pattern isgenerally evident for the specific chro-mosomes involved in such patients, a sub-group of female patients exists in whomthe common factor is the presence ofa balanced, constitutional X;autosometranslocation, with a cytogenetic break-point in the pericentromeric region of theX. It is argued here that the phenotype inthese cases results not from the in-terruption ofX linked genes but from thepresence of mosaic functional disomy ofX sequences above the breakpoint.(7Med Genet 1996;33:177-183)

Key words: hypomelanosis of Ito; X;autosome trans-location; functional disomy Xp.

Wessex RegionalGenetics Service,Level G, PrincessAnne Hospital,Coxford Road,SouthamptonS016 SYA, UKE Hatchwell.

Correspondence to:Dr Hatchwell.

Received 21 August 1995Revised version accepted forpublication20 October 1995

Hypomelanosis of Ito (HI) is a sporadic mul-tisystem disorder whose defining feature is apigmentary dysplasia.' It is now clear that thecommon factor in many cases of HI is thepresence of chromosomal mosaicism, althoughthis is not an invariable finding.23 The frequentpresence of chromosomal mosaicism in HImakes sense of the wide variability in clinicalphenotype. In a review of over 70 publishedcases, the only consistent feature was the pres-ence ofhypopigmented skin without precedingvesicular or verrucous stages.4 A total of 74%of the patients, however, had one or moreabnormalities of the central nervous system,eyes, hair, teeth, or musculoskeletal system.4Such clinical variability is common to othermosaic genetic conditions (for example, Pro-teus syndrome5) and presumably reflects thevariable presence of the genetically abnormalcell line within different lineages.The description of a number of females

with balanced, constitutional X;autosometranslocations in association with a phenotypedesignated incontinentia pigmenti (IP 1,McKusick 308300) prompted the suggestionthat a locus for incontinentia pigmenti existsat the common cytogenetic breakpoint inXp 1I.'6-lo Linkage analysis in familial cases ofincontinentia pigmenti, however, failed to con-

firm this cytogenetic localisation and muchevidence now exists for the presence of a locusat Xq28."'-5 As a result of such analyses, fa-

milial incontinentia pigmenti is now designatedIP2 (McKusick 308310). A reappraisal offamilial incontinentia pigmenti (IP2) and thephenotypes described in females with X;auto-some translocations shows that there is verylittle overlap between the two; sporadic in-continentia pigmenti (IP 1) does, however, re-semble HI in the type and distribution of skinlesions, and in the occurrence of dysmorphicfeatures, which are not present in IP2. 16 17 Someof the confusion between HI and incontinentiapigmenti has resulted from the alternative namefor HI, namely incontinentia pigmenti ach-romians. Given the clinical variability of HIalluded to above, there is little to distinguishcases of HI associated solely with autosomesfrom those related to X;autosome trans-locations.

This paper reviews all cases of HI describedin association with a balanced X;autosometranslocation and puts forward a unifying hy-pothesis to explain the presence of this sub-group of patients. Evidence is considered forfunctional disomy ofX sequences as the mainaetiological factor in the appearance of thephenotype in such patients. It is proposed thatthe term incontinentia pigmenti should now bereserved for the familial form which maps toXq28.

HI and X;autosome translocationsIn the rest of this review, females with "IP 1"and X;autosome translocations are referred toas having HI. While HI is clearly heterogeneousin origin, the common factor is thought tobe the presence of chromosomal mosaicism,although mosaic aneuploidy of almost anyautosome or sex chromosome has beendescribed.2 318 19 In addition, it is likely thatmosaicism for single gene defects can also giverise to HI. The existence of a subgroup withbalanced X;autosome translocations and sim-ilar cytogenetic breakpoints on the X requiresexplanation. Table 1 lists all the females de-scribed so far with HI in association with abalanced X;autosome translocation. A thor-ough search ofpublished reports indicated thatHI has not been reported with any of the manyX;autosome translocations described with Xbreakpoints outside the interval Xq 1 3-Xp 11.21.

Examination of table 1 shows the followingpoints. (1) All the breakpoints involve the peri-

177

group.bmj.com on August 8, 2016 - Published by http://jmg.bmj.com/Downloaded from

Table I Summary offemales with IPIIHI in association with balanced X;autosome translocations

Gilgenkrantz Kajii Hodgson Hodgson Cannizzaro Lungarotti Bitoun Koiffmann Penchaszadeh Sybert Steichenet alt et al' aetalet al and Hecht et al" et al et ale etaet et al et alt'

et al'Kayotype 46,X,t(X;9) 46,X,t(X;13) 46,X,t(X;17) 46,X,t(X;9) 46,X,t(X;1O) 46,X,t(X;18) 46,X,t(X;5) 46,X,t(X;10) 46,X,t(X;14) 46,X,t(X;22) 46,X,t(X;1 7)

(pll;q34) (p11.21;ql 2.3) (pll .2;q1l.2) (pll.2;q33.2) (pll;q22) (pll;q23) (pll1.2;q35.2) (pll;1l)matf (p11;q13) (pll.2;ql3.3) (plI3;ql13)

SkinlHypopigmentation + 0 + + 0 + 0 + + + +Hyperpigmentation ? + 0 0 + 0 + - 0 - 0Blistering - + - - + 0 - 0 -

CNS involvementPsychomotor delay + + + + + - + + + + +Seizures + + + + 0 0 0 + + 0 0EEG abnormalities + + + 0 0 - + + 0 0 0Mental retardation + + + + + + + + 0 0 0Hypotonia 0 + 0 + + 0 + + + 0 0Eye anomalies 0 + + + 0 0 - + 0 0 +Teeth anomalies - 0 + + + 0 - + 0 0 0Hair anomalies - 0 0 0 0 0 + - 0 0 0Nail anomalies - 0 0 0 0 0 0 + 0 0 0Cerebral anomalies 0 * 0 0 0 N 0t 0 15Diagnosis made IP IP IP/Ito IP Ito IP Ito IPI Ito Ito 0 ItoSkin biopsy + - + + _ _ _ + 0 0 0Other features/ Bilateral Maternally Non- Plexusinformation retinoblastoma inherited dysmorphic papilloma

translocationX inactivation (blood) >90% Xn 91% Xn 100% Xn 100% Xn 0 0 92% Xn 100% Xn 0 0 0

inactive inactive inactive inactive inactive inactive inboth patientand mother

X inactivation (skin)** >90% Xn 99% Xn 0 0 0 0 0 0 0 0inactive inactive

+ present, - absent, 0 not mentioned. N normal. * Moderate cortical atrophy, slight ventricular dilatation (CT scan). t Cerebral cortical atrophy (air encephalography).t Diffuse corticosubcortical atrophy (CT scan). § Macrocephaly; huge plexus papilloma with enlargement of ventricular system. T The patient's mother carried thesame balanced translocation; the mother had seizures as a child but was well and had no pigmentary anomalies. Xn normal, intact X. ** No mention of whetheraffected/unaffected skin was biopsied (all biopsies cultured and not assayed directly).

centromeric region of the X; while in a majorityit is Xp that is implicated, in three cases Xq isinvolved. It is thus unlikely that the same, singlelocus can be implicated in all these cases, unlessthe cytogenetic localisation is incorrect in anumber of cases. (2) In one case, the X;auto-some translocation was familial, being presentin both mother and daughter. The mother,however, did not have the pigmentary dysplasia.If the breakpoint in this family were in-terrupting a critical locus, it might be expectedthat the mother and daughter would have anidentical phenotype.

The effect of functional disomy ofsequences on the XUntil recently, it has been difficult to gauge thephenotypic effect of functional disomy of theX in females. Females who harbour a du-plication of material on one X chromosomeare almost invariably normal phenotypicallyas they have a severely skewed X;inactivationpattern in favour of the intact X.2022 Suchfemales are, however, at 50% risk of passingon the duplicated chromosome to their sons

who invariably have a severe phenotype. Atleast 15 cases of duplications of the X chro-mosome in a male karyotype have been re-

ported with the duplications, when consideredtogether, covering most of the X chromosome,with the exception of Xpll-Xqll and Xq25-26.20-31 The reported cases have a wide rangeof differing phenotypic abnormalities, althoughfacial dysmorphism, hypotonia, and mental re-

tardation were common, albeit non-specificfeatures. None of these subjects was reportedas having a pigmentary dysplasia reminiscentof HI. This is perhaps unsurprising as suchsubjects are not mosaic. Six people with a

duplication encompassing chromosomal region

Xp21 had sex reversal in addition to the otherclinical features mentioned (that is, these were

46,XY females).2628-30 Evidence now exists forthe presence of a dosage sensitive sex reversallocus (DSS), adjacent to the locus for con-

genital adrenal hypoplasia at Xp21.2932Cases ofX chromosome duplication in males

are thus instructive in that functional disomyof sequences on the X is tolerated, albeit witha severe phenotype, related to the presence offunctional disomy ofX chromosome sequences(there is evidence that the duplicated segmentdoes not undergo X inactivation in such cases

generally23 33). These males, however, are notmosaic and do not easily allow for predictionof the effect of such disomy in a heterozygousfemale with random X inactivation.The description of a subgroup of females

with Turner's syndrome who have a severe

phenotype (that is, compared to most Turnerpatients) in association with a (mosaic) cell thatharbours a ring (X) chromosome is, however,instructive.3"38 There appears to be a goodcorrelation between the presence of a severe

phenotype in such females and the inability ofthe ring to inactivate, either because the Xinactivation centre is absent, or because XISTexpression is defective for reasons that are not

entirely clear.353940 It seems likely that the clin-ical severity in these cases is related to thepresence of functional disomy of sequencespresent on the (active) ring. Of particular in-terest is the fact that a number of such femaleshave been described as possessing a pigmentarydysplasia reminiscent of HI.343741

In addition to the evidence from ring (X)females of the effect of (mosaic) functionaldisomy ofX sequences, another case ofa femalepresenting with HI is instructive (J Hurst, per-sonal communication). This female presentedwith seizures, developmental delay, hypo-

178 Hatchwell

group.bmj.com on August 8, 2016 - Published by http://jmg.bmj.com/Downloaded from

Hypomelanosis of Ito and X;autosome translocations: a unifying hypothesis

dontia, pigmented whorls on the back, anddysmorphism, including hypertelorism and flatnasal bridge with epicanthic folds. Cytogeneticanalysis showed a de novo duplication atXpl1.2 of the region Xpl1.4-+pl1.2. X in-activation analysis in peripheral blood lym-phocytes showed random X inactivation withapproximately a 50:50 pattern, rather than theexpected skewing towards activity of the intactX. It is likely that the phenotype in this femaleresults from the presence of functional disomyof sequences located in Xp 11.2 -+11.4, in that

proportion of cells that have the duplicated Xactive. The recognition ofthis female makes theexistence of other females with submicroscopicduplications in the region likely (see below).While functional mosaicism results from ran-dom X inactivation in this female, it is im-portant to note that in the case of ring(X)females it results from the physical presence orabsence of the ring in different cell lineages.

Mosaic functional disomy of X specificsequences as the most likely mechanismfor HI associated with balancedX;autosome translocationsWhat then is the mechanism whereby HI ismanifested in the cases detailed in table 1? Anumber of possibilities are to be considered.(1) Interruption of a single locus by the breakpointon the X chromosome can explain all the cases

reported so far. This explanation is unlikely to betrue, not only because the reported breakpointsmap on both sides of the centromere but mo-lecular evidence also exists for heterogeneity ofbreakpoints that map to Xp 11, implying thatat least three of the breakpoints span a distanceof 5 Mb, a distance greater than that spanned bythe largest known human gene, dystrophin.42A5(2) Each breakpoint defines a different locus(homologous in structure orfunction or both?) thatcan be implicated in the pathogenesis ofHI in thesefemales. Again, this is unlikely. (A) The presenceof a discordant phenotype in two females withthe same breakpoint (X; 10 translocation intable 1) implies that it is the functional statusof the derived X that is important, rather thanthe precise site of interruption. (B) In the caseof the other breakpoints, it is difficult to becertain that a specific locus interruption can beexcluded. In the case of the X; 17 translocationreported by Hodgson et al,9 however, thereis mounting evidence against a direct locusinterruption. In situ hybridisation analysis haspreviously shown the X breakpoint to be cen-

tromeric and a recent reanalysis, in addition toconfirming the centromeric nature of the Xbreakpoint, has shown the 17 breakpoints alsoto be centromeric, thus making involvement ofa locus on 17 unlikely.46(3) The phenotype results from interruption ofautosomal loci. This possibility is very difficultto exclude but would appear to be unlikely as

a number of different autosomes have now

been described in such cases, and in the case

of the X; 17 translocation, the 17 breakpoint iscentromeric. In addition, no disruption of an

autosomal gene has ever been diagnosed inassociation with a balanced X;autosome trans-

location, probably because most ofthe resultingphenotypes are recessive.47(4) The phenotype results from an unusual patternof X inactivation. This seems the most likelyexplanation and is best exemplified by the casedescribed above in which the phenotype isdiscordant between mother and daughter, des-pite identical karyotypes. In a female with abalanced X;autosome translocation, failure toeradicate cell lines bearing an inactive derivedX will result in both mosaic functional disomyofX sequences and mosaic (variable) functionalmonosomy of autosomal sequences inactivatedby spreading of X inactivation from the XICof derived X. By comparison with the cases ofring (X) females, and the female describedabove with random X inactivation of a du-plicated X chromosome, in both examples ofwhich disomy of the X is the only possiblemechanism, it would seem likely that it is thefunctional disomy that is the major factor inthe females with HI and X;autosome trans-locations. Mosaic functional monosomy ofautosomal sequences may, however, also beplaying a role. While the majority of casesof mosaic aneuploidy described in associationwith HI involve the presence of extra chro-mosomal material, in a small number of casesmosaic monosomy of autosomal sequences hasbeen found.3

Recent analysis of the female with a balancedX; 17 translocation and HI (case 1 of Hodgsonet al') showed randomX inactivation in affected(hypopigmented) skin while X inactivation innormal skin and blood was skewed towardsactivity of the derived X as expected.46 Suchaberrant X inactivation in affected skin wouldresult in functional disomy for all of Xp and avariable portion of 17p in those cells with aninactive derived X.However, the phenotypic overlap between

ring(X) females and those with HI associatedwith balanced X;autosome translocationsmakes disomy the most likely mechanism. VanDyke et at7 found that pigmentary dysplasiawas associated with the presence of a small ring(X) in females with a severe phenotype (inat least 19% of cases). Indeed, the femaledescribed by de Grouchy et al" was thought tohave incontinentia pigmenti; Xpll was againput forward as a region likely to contain thelocus for incontinentia pigmenti as one of thebreakpoints was in this region. It would seempreferable to reclassify this female as having HIassociated with a small ring (X) chromosomethat failed to inactivate (the ring (X) was foundnever to be late labelling).Many ofthe dysmorphic features found in fe-

males with small ring (X) chromosomes are alsofrequent in females with HI associated with X;autosome translocations (tables 2 and 3).

X;autosome translocations and X linkeddisorders: theoretical reasons for IPI/Hlfemales not having classicalincontinentia pigmentiThe discovery of rare females manifesting Xlinked phenotypes in association with a bal-anced, constitutional X;autosome trans-

179

group.bmj.com on August 8, 2016 - Published by http://jmg.bmj.com/Downloaded from

Table 2 List of dysmorphic features reported in females with HI associated with balanced X;autosome translocations, graded according to frequency.Note overlap with ring (X) cases (see table 3).

X; 1 7 X;9 X;13' X;107 X;98* X;185 t X;56 X;JI5° X;146/'t X;223* X; 1 761

High arched palate (4/11) + + + +Broad nasal bridge (3/11) + + +HC>95th centile (3/11) + + +Abnormal ears (3/11) + + +High forehead (3/11) + + +Dental abnormalities (3/11) + +Flat face (2/1 1) +Hypertelorism (2/11) + +Epicanthic folds (2/11) + +Small hands and feet (2/11) + +Brachycephaly (1 / 1) +Prognathism (1/ 1) +Large tongue (1/1 ) +Plagiocephaly (1 / 11) +Carp shaped mouth (1/11) +Syndactyly (1/11) +Short philtrum (1/11 ) +Cleft soft palate (1/ 1) +Hirsutism (1/1 1) +Gum hypertrophy (1/1 1) +Neonatal teeth (1/ 1) +Alopecia (1/ 1) +

* Dysmorphic features noted as being absent. t No mention made of dysmorphic features. With the exception of cases 5, 6, 9, and 10, a blank indicates that therelevant feature was not observed.

location has been of great benefit in thelocalisation and isolation of the loci involvedin such disorders.48 50 The manifestation of therelevant X linked disorder in such females isthought to result from the severe skewing ofXinactivation that is usually observed. While Xinactivation in such females probably occursrandomly, subsequent selection against cellsthat have inactivated the derived X chro-mosome results in a uniform pattern of in-activation in which the intact X is almostinvariably the inactive one. Interruption of theX linked locus under consideration on the de-rived X renders the female functionally nul-lisomic and the phenotype is therefore manifest.While it is usually assumed that cell selectionis primarily directed against cells that have aninactive derived X because of spreading ofinactivation onto the attached autosome, it islikely that functional disomy ofX sequences isan equally potent abnormality against whichselection occurs (fig 1).47

Females with X linked phenotypes resultingfrom balanced X;autosome translocationsmanifest those phenotypes precisely because ofthe severe skewing that renders them "male-like". Thus it would be surprising, a priori,to discover females manifesting an X linkeddominant disorder with male lethality (such asIP2) in association with a balanced X;autosome

Table 3 Dysnmorphic features reported in three fenmales with a pigmentary dysplasiaassociated with the presence of a ring(X) cell line, graded according to frequency.The most common features overlap with the most common features described in theX;autosome cases

Case 3 of Dennis Case 5 of Van de Grouchyet alC' Dyke et alC et al

r(X) r(X) r(X)

Broad nasal bridge (3/3) + + +Hypertelorism (3/3) + + +High arched palate (2/3) + +Abnormal ears (2/3) + +Epicanthic folds (2/3) + +HC>95th centile (1/3) +Syndactyly (1/3) +Heart malformation (1/3) +Polydactyly (1/3) +Alopecia (1/3) +

translocation. In addition, one would predictthat females manifesting X linked disordersin association with X;autosome translocationswould not have a mosaic pattern of expression.For example, the pattern of disease expressionin IP2 is characterised by the presence of ab-normal skin alongside normal skin, the de-marcation being along Blaschko's lines. Such"phenotypic mosaicism" is likely to arise be-cause of random X inactivation during de-velopment; abnormal skin presumably containslarge numbers of cells with the mutationbearing X active while normal skin has thischromosome inactivated. Such phenotypicmosaicism is known to occur in females whoare obligate heterozygotes for a number of Xlinked disorders, including hypohidrotic ec-

....

.

xlr

(.L!l.}, i;I f _ er !e:i.|t| ; t t~~~~~~~~~~~~~lai; T v ..

Figure I This figure illustrates the situation in those cellsthat have failed to inactivate in the usual way. The intactX is active, while the derived X is inactive. Twofunctionally active copies of Xp (that is, of all sequencesdistal to the breakpoint in the pericentromeric region of theX) are therefore present, one on the intact X and one onthe derived autosome. The latter Xp is presumed to remainfunctional as it has been physically disconnected from theXIC of the derived X. In addition to functional disomy ofXp, it is likely that there will be a (variable) monosomy ofautosomal sequences owing to spreading of inactivationfrom the XIC of the inactive derived X.

180 Hatchwell

Ami,

Im

aw4w :,

.w.1, li.1""""-I.-A

I

,ob

group.bmj.com on August 8, 2016 - Published by http://jmg.bmj.com/Downloaded from

Hypomelanosis of Ito and X;autosome translocations: a unifying hypothesis

todermal dysplasia, Menkes disease, and Goltzsyndrome.5' Thus, not only do significantdifferences exist between the phenotypes de-scribed in IP1 and those of "classical" IP (IP2)but there are more fundamental reasons whythe females with IP1 are unlikely to have thesame locus implicated as those with IP2. Thisargument predicts that no female will be de-scribed who has "incontinentia pigmenti" inassociation with an X;autosome translocationinvolving the IP2 locus at Xq28. In addition,it argues against the likelihood that any of theX breakpoints in females with HI associatedwith a balanced X;autosome translocation in-volve a locus that defines a familial disorder.

Specific reasons for IP1/HI breakpointsnot being related to incontinentiapigmentiIt would now seem clear that the phenotypicdifferences between IP1/HI and IP2 are suffi-cient to warrant the designation IP 1 re-dundant.'7 In addition to the differences inevolution of skin abnormalities between thetwo conditions, most of the females with IPl/HI have been noted to be dysmorphic (table2), a feature not present in classical familialincontinentia pigmenti."

Functional disomy ofX sequences indiploid/triploid mixoploidy as amechanism for HI?HI has been described in association with dip-loid/triploid mixoploidy. It is tempting to spec-ulate that functional disomy of X sequencesmay also be implicated in the pathogenesishere. In cells that have more than one X chro-mosome, it is invariably the case that one Xremains active, while all others are inactivated.This is true for diploid cells, irrespective of thenumber of X chromosomes present. Triploidcells, however, appear able to support the ac-tivity of more than one X chromosome. Thus,while only one X is active in 47,XXY cells,69,XXY cells may have two active X chro-

Region of overlapL -- within which disomy

may give rise to HixIc

Xq

Der (A) Der (X)

Xq

r(X)

Figure 2 Diagram illustrating the possible location of a locuslloci that causes the HIphenotype when present in a functionally disomic form. Given that quite a number offemales have now been described who manifest a severe phenotype in association with a

small ring X chromosome, it should be possible to define this region more precisely. Anysuch locus, according to the hypothesis, should map distal to the most "telomeric" IPIX;autosome translocation breakpoint, and proximal to the most "centromeric" of all theXp ring X breakpoints.

mosomes.52 It seems likely that it is the ratioof autosomes to X that determines the numberofX chromosomes that will remain active, witha usual diploid complement of 44 autosomesensuring that only one X remains active. Inthe case of 69,XXX or 69,XXY triploidy, aproportion of cells will harbour more than oneactive X chromosome, with the result that thesecells will be functionally disomic for the wholeof the X, a situation that is likely to be de-leterious to the cell line under consideration.Clearly, however, the presence of an extra setof autosomes cannot be ruled out as beingimplicated in the pathogenesis of HI in thesepeople.

Can specific X linked loci be definedwhich are necessary for development ofHI resulting from functional disomy ofthe X?It is probable that the females ascertained ashaving HI in association with X;autosometranslocations with breakpoints in Xql 3-Xp 1 1represent a small minority of all females withsuch translocations. In a recent review of 122published cases of balanced X;autosome trans-locations in females, the breakpoints in 57 werelisted as lying above XIC.47 Of these, 20 weresaid to be phenotypically normal and it is tobe expected that the true proportion is evenhigher as such phenotypically normal femalesare much less likely to be ascertained. In themajority of cases, the breakpoints are unlikelyto interrupt specific loci and skewed X in-activation will result in a normal phenotype.There is unlikely to be a specific reason forineffective selection against cells bearing aninactive derived X. In their review, Schmidtand Du Sart47 reported that inactivation of thederived X, in at least some tissues, occursin 23% of females with balanced X;autosometranslocations.However, no females have been described

with HI where the X breakpoint lies outsideXql3-Xpll. It is noteworthy that, of 24 fe-males known to manifest Duchenne musculardystrophy in association with a balanced X;autosome translocation that disrupts the DMDlocus, none has been described as having apigmentary dysplasia, although a number haveother abnormalities, including psychomotor re-tardation.53 Thus it seems likely that if specificloci are to be implicated in the pathogenesis ofHI, when present in a functionally disomicstate, then such loci must be located proximalto Xp2 1. In addition, the presence of HI infemales with small ring (X) chromosomes, inwhom functional disomy is almost certainly theaetiological mechanism, argues for the presenceof the implicated locus (loci) within the ring.Thus, the locus (loci) in question may be

mapped by considering the smallest ring (X)found to cause HI and the most distal X break-point that also gives rise to the phenotype (fig2). However, although such analysis may definea minimum region within which such a locus(loci) must reside, it is conceivable that differ-ent loci may be responsible in different cases.

Xp

181

group.bmj.com on August 8, 2016 - Published by http://jmg.bmj.com/Downloaded from

Hatchwell

Why are no X breakpoints describedbelow Xql3?Inactivation of the derived X in a cell bearingan X;autosome translocation with a breakpointbelow Xq 13 will result in functional disomy ofXq sequences. The absence of X breakpointsbelow Xq13 argues against functional disomyof Xq sequences as being important in pro-duction of the HI phenotype. If a locus (loci)on Xp is particularly important in the pro-duction of the HI phenotype (that is, in afunctionally disomic state), then only break-points above the XIC will be ascertained inthis way.

X;autosome translocations and mappingXLMR genesFailure to eradicate cell lines bearing an inactivederived X may happen in some tissues and notothers. Thus, it is likely that some females withmental retardation associated with balancedX;autosome translocations, in whom a pig-mentary dysplasia is absent, may still be man-ifesting their phenotype as a result of themechanism outline above (that is, in those caseswhere the breakpoint lies above XIC). In suchfemales, eradication of cell lines bearing aninactive derived X may have been successfulin skin but not in other tissues, such as CNS. Itis possible that such females will be erroneouslyinterpreted as possessing a translocation thatresults in the interruption of a locus for Xlinked mental retardation (XLMR). This islikely to be a potential hazard in the mappingand isolation of genes implicated in XLMR,particularly where this is non-syndromic. Thecase described by Hodgson et al4 illustratesthe point. A female with a balanced, de novo,constitutional X; 13 translocation was describedas developmentally delayed and dysmorphic.No pigmentary dysplasia was noted but herdysmorphic features, albeit relatively non-spe-cific, were consistent with those seen in thefemales described above. Others have noted anincreased incidence of breakpoints at Xp 11 infemales with mental retardation and X;auto-some translocations and correlated this withXLMR loci thought to map to this region.5

Functional disomy of Xp incytogenetically normal females with HI?It is likely that females exist with sub-microscopic duplications in Xp, who manifestthe phenotype of HI. Indeed, this may be onepossible mechanism to explain the rare reportedcases of familial HI.56 The presence of a sub-microscopic duplication on one X may resultin a variety of phenotypes depending on thepattern of X inactivation in relevant tissues. Aphenotypically normal mother (or one whosimply had a pigmentary dysplasia) might havea severely affected daughter with HI andCNS abnormalities. It is harder to explain fa-milial occurrences of HI by implicating sub-microscopic autosomal abnormalities, as thereis not as much scope for phenotypic variabilityas is afforded by differential X inactivation. Itis suggested that a search for submicroscopic

duplications in proximal Xp may be fruitful infemale cases of HI where cytogenetic analysisis normal, particularly if there are indicationsthat the HI may be familial, albeit with variableseverity.

ConclusionThis review has presented evidence for the roleof functional disomy of Xp in the pathogenesisof HI associated with X;autosome trans-locations with breakpoints in the peri-centromeric region of the X. Previous ideasabout such X breakpoints defining specific lociare likely to be false for reasons stated above.

Further work to map minimal regions ofoverlap found in (A) ring (X) cases and (B)translocation cases may define a locus (loci)that is implicated in the pathogenesis of HIand may lead to the discovery of patients withHI resulting from submicroscopic duplications.

In addition, caution is required when in-terpreting the pathogenesis of mental re-tardation in females with balanced X;autosometranslocations and it should not be assumedthat the X breakpoint necessarily defines a locusimplicated in XLMR. The obvious exceptionswill include females who manifest a well definedphenotype in association with a balancedX;autosome translocation (for example,Menkes syndrome, see above).

Implications for genetic counsellingWhile the majority of females with balancedX;autosome translocations with breakpointsabove XIC are likely to be phenotypically nor-mal, accurate prediction of the phenotype ofany daughters with an identical translocationcannot be made with certainty, as "un-fortunate" X inactivation may result in a severephenotype as illustrated by the maternally in-herited X; 10 translocation detailed in table1. Conversely, where a phenotypically normalmother with a balanced X;autosome trans-location has a son with the same balancedtranslocation, reassurance can be given that thechild will be phenotypically normal althoughmale infertility would be expected.57 In thissituation the presence of a normal phenotypein the mother makes interruption of a specificlocus on either the X or the autosome unlikely,so that there is no specific reason why maleoffspring should have an abnormal phenotype.

I am grateful to Nick Dennis and Pat Jacobs for helpful com-ments. This work was supported by the Wellcome Foundation.

1 Ito M. Incontinentia pigmenti achromians. A singular caseof nevus depigmentosus systematicus bilateralis. Tohoki _7Exp Med (Seoidai) 1952;55 (suppl):57-9.

2 Donnai D, Read AP, McKeown C, Andrews T. Hypo-melanosis of Ito: a manifestation of mosaicism or chime-rism. J Med Genet 1988;25:809-18.

3 Sybert VP, Pagon RA, Donlan M, Bradley CM. Pigmentaryabnormalities and mosaicism for chromosomal aberration:association with clinical features similar to hypomelanosisof Ito. I Pediatr 1990;116:581-6.

4 Takematsu H, Sato S, Igarashi M, Seiji M. Incontinentiapigmenti achromians (Ito). Arch Dermiiatol 1983;119:391-5.

5 Cohen MJ. Proteus syndrome: clinical evidence for somaticmosaicism and selective review. Am _7 Med Genet 1993;47:645-52.

6 Bitoun P, Philippe C, Cherif M, Mulcahy MT, GilgenkrantzS. Incontinentia pigmenti (type 1) and X;5 translocation.Anns- Geniet (Paris) 1992;35:51-4.

182

group.bmj.com on August 8, 2016 - Published by http://jmg.bmj.com/Downloaded from

Hypomelanosis of Ito and X;autosome translocations: a unifying hypothesis

7 Cannizzaro LA, Hecht F. Gene for incontinentia pigmentimaps to band Xp 11 with an (X; 10) (p11 ;q22) trans-location. Clin Genet 1987;32:66-9.

8 Gilgenkrantz S, Tridon P, Pinel BN, Beurey J, Weber M.Translocation (X;9) (p1 1 ;q34) in a girl with incontinentiapigmenti (IP): implications for the regional assignment ofthe IP locus to Xpl I? Ann Genet (Paris) 1985;28:90-2.

9 Hodgson SV, Neville B, Jones RW, Fear C, Bobrow M.Two cases of X/autosome translocation in females withincontinentia pigmenti. Hum Genet 1985;71 :231-4.

10 Kajii T, Tsukahara M, Fukushima Y, Hata A, Matsuo K,Kuroki Y Translocation (X;13) (pll.21;qI2.3) in a girlwith incontinentia pigmenti and bilateral retinoblastoma.Ann Genet (Paris) 1985;28:219-23.

11 Sefiani A, Sinnett D, Abel L, et al. Linkage studies do notconfirm the cytogenetic location of incontinentia pigmention Xpll. Hunm Genet 1988;80:282-6.

12 Sefiani A, Abel L, Heuertz S, et al. The gene for incontinentiapigmenti is assigned to Xq28. Genomics 1989;4:427-9.

13 Sefiani A, M'rad R, Simard L, et al. Linkage relationshipbetween incontinentia pigmenti (IP2) maps to the distalpart of Xq28. Hum Genet 1993:3:273-8.

14 Smahi A, Hyden GC, Peterlin B, et al. The gene for thefamilial form of incontinentia pigmenti (IP2) maps to thedistal part of Xq28. Hum Mol Genet 1994;3:273-8.

15 Harris A, Lankester S, Haan E, et al. The gene for in-continentia pigmenti: failure of linkage studies using DNAprobes to confirm cytogenetic localization. Clin Genet1988;34: 1-6.

16 Landy SJ, Donnai D. Incontinentia pigmenti (Bloch-Sulz-berger syndrome). J7 Med Genet 1993;30:53-9.

17 Sybert VP. Incontinentia pigmenti nomenclature. Am7HumGenet 1994;55:209-1 1.

18 Thomas IT, Frias JL, Cantu ES, Lafer CZ, Flannery DB,Graham JJ. Association of pigmentary anomalies withchromosomal and genetic mosaicism and chimerism. AmJHum Genet 1989;45:193-205.

19 Donnai D, McKeown C, Andrews T, Read AP. Diploid/triploid mixoploidy and hypomelanosis of Ito. Lancet 1986;i: 1443-4.

20 Nielsen KB, Langkjaer F. Inherited partial X chromosomeduplication in a mentally retarded male. 7 Med Genet1982;19:222-4.

21 Schwartz S, Schwartz MF, Panny SR, Peterson CJ, WatersE, Cohen MM. Inherited X-chromosome inverted tandemduplication in a male traced to a grandparental mitoticerror (published erratum appears in Am _J Hum Genet1987;40:297). Am iHum Genet 1986;38:741-50.

22 Vejerslev LO, Rix M, Jespersen B. Inherited tandem du-plication dup(X) (ql31-q212) in a male proband. ClinGenet 1985;27:276-81.

23 Schmidt M, Du SD, Kalitsis P, et al. Duplications of the Xchromosome in males; evidence that most parts of the Xchromosome can be active in two copies. Hum Genet 199 1;86:519-21.

24 Thode A, Partington MW, Yip MY, Chapman C, Rich-ardson VF, Turner G. A new syndrome with mentalretardation, short stature and an Xq duplication. Am JMed Genet 1988;30:239-50.

25 Steinbach P, Horstmann W, Scholz W. Tandem duplicationdup (X) (ql3q22) in a male proband inherited from themother showing mosaicism of X-inactivation. Hum Genet1980;54:309-13.

26 Scherer G, Schempp W, Baccichetti C, et al. Duplicationof an Xp segment that includes the ZFX locus causes sexinversion in man. Hum Genet 1989;81:291-4.

27 Cremers FP, Pfeiffer RA, van de Pol, TJ et al. An interstitialduplication of the X chromosome in a male allows physicalfine mapping of probes from the Xql3-q22 region. HumGenet 1987;77:23-7.

28 Bernstein R, Jenkins T, Dawson B, et al. Female phenotypeand multiple abnormalities in sibs with a Y chromosomeand partial X chromosome duplication: H-Y antigen andXg blood group findings. _JMed Genet 1980;17:291-300.

29 Rao PN, Klinepeter K, Stewart W, Hayworth R, Grubs R,Pettenati MJ. Molecular cytogenetic analysis of a du-plication Xp in a male: further delineation of a possiblesex influencing region on the X chromosome. Hum Genet1994;94: 149-53.

30 Reichenbach H, Holland H, Thamm B, Theile T. Multipleabnormalities in a child with male karyotype due to familialpartial Xp duplication. (German.) Kinderaztliche Praxis1993;61:291-5.

31 Muscatelli F, Verna JM, Philip N, et al. Physical mappingof an Xq-proximal interstitial duplication in a male. HumGenet 1992;88:691-4.

32 Bardoni B, Zanaria E, Guioli S, et al. A dosage sensitivelocus at chromosome Xp21 is involved in male to femalesex reversal. Nature Genet 1994;7:497-501.

33 Muscatelli F, Lena D, Mettei MG, Fontes M. A malewith two contiguous inactivation centers on a single Xchromosome: study of X inactivation and XIST ex-pression. Hum MolGenet 1992;1:115-9.

34 Dennis NR, Collins AL, Crolla JA, Cockwell AE, FisherAM, Jacobs PA. Three patients with ring (X) chromosomesand a severe phenotype. J Med Genet 1993;30:482-6.

35 Cole H, Huang B, Salbert BA, etal. Mental retardation andUllrich-Turner syndrome in cases with 45,X/46,X, + mar:

183

additional support for the loss of the X-inactivation centerhypothesis. Ani 7 Med Genet 1994;52:136-45.

36 Grompe M, Rao N, Elder FF, Caskey CT, Greenberg F.45,X/46,X, + r(X) can have a distinct phenotype differentfrom Ullrich-Turner syndrome. Am i Med Genet 1992;42:39-43.

37 Van Dyke D, Wiktor A, Palmer CG, et al. Ullrich-Turnersyndrome with a small ring X chromosome and presenceofmental retardation. AmJ7Med Genet 1992;43:996-1005.

38 Wolff DJ, Brown CJ, Schwartz S, Duncan AM, Surti U,Willard HF. Small marker X chromosomes lack the Xinactivation center: implications for karyotype/phenotypecorrelations. An _7 Hum Genet 1994;55:87-95.

39 Migeon BR, Luo S, Stasiowski BA, et al. Deficient tran-scription of XIST from tinv ring X chromosomes infemales with severe phenotypes. Proc Natl Acad Sci USA1993;90: 12025-9.

40 Migeon BR, Luo S, Jani M, Jeppesen P. The severe pheno-type of females with tiny ring X chromosomes is associatedwith inability of these chromosomes to undergo X in-activation. Am _7 Hum Genet 1994:55:497-504.

41 de Grouchy J, Turleau C, Doussau DBM, Maroteaux P,Thibaud D. Incontinentia pigmenti (IP) and r(X). Tent-ative mapping of the IP locus to the X juxtacentromericregion. Ann Genet (Paris) 1985;28:86-9.

42 Reed V, Rider S, Maslen GL, et al. A 2-Mb YAC contigencompassing three loci (DXF34, DXS 14, and DXS390)that lie between Xpl 1.2 translocation breakpoints as-sociated with incontinentia pigmenti tvpe 1. Genomtics1994;20:341-6.

43 Gorski JL, Burright EN, Harnden CE, Stein CK, GloverTW, Reyner EL. Localization of DNA sequences to aregion within XpI1.21 between incontinentia pigmenti(IPI) X chromosomal translocation breakpoints. Am]Hum Genet 1991;48:53-64.

44 Hatchwell E, Bitoun P, Gilgenkrantz S, et al. Incontinentiapigmenti translocation breakpoints - further evidence forheterogeneity. Accepted for oral presentation at ChromosomeSpecific Session, Human Gene Mapping Workshop, Kobe,Japan, 1993.

45 Hatchwell E, Bitoun P, Gilgenkrantz S, et al. Further evid-ence for heterogeneity of IP1/HI breakpoints (in pre-paration).

46 Hatchwell E, Robinson D, Crolla JA, Cockwell AE. Xinactivation analysis in a female with hypomelanosis ofIto associated with a balanced X; 17 translocation: evidencefor functional disomy ofXp. JMed Genet 1996;33:216-20.

47 Schmidt M, Du Sart D. Functional disomies of the Xchromosome influence the cell selection and hence the Xinactivation pattern in females with balanced X-autosometranslocations: a review of 122 cases. Am _7 Med Genet1992;42:161-9.

48 Attree 0, Olivos IM, Okabe I, et al. The Lowe's oc-ulocerebrorenal syndrome gene encodes a protein highlyhomologous to inositol polyphosphate-5-phosphatase.Nature 1992;358:239-42.

49 ChellyJ, Tumer Z, Tonnesen T, et al. Isolation ofa candidategene for Menkes disease that encodes a potential heavymetal binding protein. Nature Genet 1993;3: 14-19.

50 Ray PN, Belfall B, Duff C, et al. Cloning of the breakpoint ofan X;21 translocation associated with Duchenne musculardystrophy. Nature 1985;318:672-5.

51 Happle R. Lyonization and the lines of Blaschko. Hunm Genet1985;70:200-6.

52 Weaver DD, Gartler SM. Evidence for two active X chro-mosomes in a human XXY triploid. Humiangenetik 1975;28:39-42.

53 Boyd Y. In: Roses C, ed. Molecular genetics and neurologa:London: Royal Society of Medicine, 1991.

54 Hodgson SV, Barber JC, Dowie A, Dubowitz V. A de novoX; 13 translocation with abnormal phenotype. _JMed Geniet1986;23:477-8.

55 Teboul M, Mujica P, Cherv M, Leotard B, Gilgenkrantz S.Balanced X-autosomal translocation and mental re-tardation. Mapping mental retardation linked to X (ex-cluding fragile X). (French.) _7 Genet Hum 1989;37:179-95.

56 Montagna P, Procaccianti G, Galli G, Ripamonti L, PatriziA, Baruzzi A. Familial hypomelanosis of Ito. Eur Neurol1991:31:345-7.

57 Carpenter NJ. Balanced X;autosome translocations andgonadal dysfunction in females and males. In: SandbergAA, ed. Cytogenetics of the 7tmnnaliati X chromnosomte. PartB. Volume 3. New York: Alan R Liss, 1983:211-24.

58 Lungarotti MS, Martello C, Calabro A, Baldari F, MariottiG. Hypomelanosis of Ito associated with chromosomaltranslocation involving Xp 11. Am T7Med Genet 1991;40:447-8.

59 Koiffmann CP, De SD, Diament A, et al. Incontinentiapigmenti achromians (hypomelanosis of Ito, MIM146150): further evidence of localization at Xp 11. Am]7Med Genet 1993;46:529-33.

60 Penchaszadeh VB, Babu A, Schwartz M, David KL, PopescuS, Rubinstein C. Hypomelanosis of Ito in a girl with a denovo translocation (X;14)(qll;ql3). Am .7 Hum GenetSuppl1989;45:A219.

61 Steichen GE, Trawoger R, Duba HC, Mayr U, Felber S,Utermann G. Hypomelanosis of Ito in a girl with plexuspapilloma and translocation (X;17). HumGenet 1993;90:611-3.

group.bmj.com on August 8, 2016 - Published by http://jmg.bmj.com/Downloaded from

translocations: a unifying hypothesis.Hypomelanosis of Ito and X;autosome

E Hatchwell

doi: 10.1136/jmg.33.3.1771996 33: 177-183 J Med Genet 

http://jmg.bmj.com/content/33/3/177Updated information and services can be found at:

These include:

serviceEmail alerting

box at the top right corner of the online article. Receive free email alerts when new articles cite this article. Sign up in the

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:

group.bmj.com on August 8, 2016 - Published by http://jmg.bmj.com/Downloaded from