American Journal of Medical Genetics 107:181±189 (2002)

Microsatellite Analysis in Turner Syndrome: ParentalOrigin of X Chromosomes and Possible Mechanism ofFormation of Abnormal Chromosomes

Nancy Monroy,1 Marisol LoÂpez,2 Alicia Cervantes,1 Diana GarcõÂa-Cruz,3 Gildardo Zafra,4

Sonia Canu n,3 Juan Carlos Zenteno,1 and Susana Kofman-Alfaro1*1Servicio de Genetica, Hospital General de Mexico/Facultad de Medicina, UNAM, Mexico, D.F, Mexico2Depto. Sistemas Biologicos, Universidad Autonoma Metropolitana-Xochimilco, Mexico, D.F., Mexico3Genetica, Centro de Investigacion Biomedica de Occidente, IMSS, Guadalajara, Jalisco, Mexico4Genetica, Hospital Espanol, Mexico, D.F., Mexico5Genetica, Hospital General M. Gea Gonzelez, Mexico, D.F., Mexico

Turner syndrome is a chromosomal disor-der in which all or part of one X chromosomeis missing. The meiotic or mitotic origin ofmost cases remains unknown due to thedif®culty in detecting hidden mosaicism andto the lack of meiotic segregation studies.We analyzed 15 Turner patients, 10 with a45,X whereas the rest had a second cell linewith abnormal X-chromosomes: a pseudodi-centric, an isochromosome, one large andone small ring, and the last with a long armdeletion. Our aims were: to detect X crypticmosaicism in patients with a 45,X constitu-tion; to determine the parental origin of theabnormality; to infer the zygotic origin ofthe karyotype and to suggest the timing andmechanism of the error(s) leading to theformation of abnormal X chromosomes frommaternal origin. Molecular investigationdid not revealed heterozygosity for anymicrosatellite, excluding X mosaicism inthe 45,X cases. Parental origin of the singleX chromosome was maternal in 90% of thesepatients. Three of the structurally abnormalXs were maternally derived whereas theother two were paternal. These resultsallowed us to corroborate breakpoints inthese abnormal X chromosomes and suggestthat the pseudodicentric chromosome origi-nated from post-zygotic sister chromatidexchange, whereas the Xq deleted chromo-

some probably arose after a recombinationevent during maternal meiosis.ß 2001 Wiley-Liss, Inc.

KEY WORDS: Turner syndrome; hidden Xmosaicism; parental origin;abnormal X chromosomes

INTRODUCTION

Turner syndrome (TS) is a common cytogeneticabnormality, with an estimated frequency of 1±2%among all clinically recognized pregnancies [Hassoldet al., 1988; Jacobs et al., 1990]. It is characterized byshort stature, gonadal dysgenesis and various somaticstigmata. Slightly more than half of the patients withTurner syndrome are monosomic for the X chromosomein peripheral blood lymphocytes [Loughlin et al., 1991;Jacobs et al., 1997]. The remaining cases have astructurally abnormal X or Y or are mosaics with asecond cell line containing a normal or abnormal sexchromosome [Hall and Gilchrist, 1990; Gicquel et al.,1992; Larsen et al., 1995; Jacobs et al., 1997].

Several molecular studies have been carried out todetect cryptic mosaicism in cytogenetically 45,X cases.Special attention has been given to identify a secondcell line containing Y-derived material because itcorrelates with 7±10% risk of developing gonadoblas-toma, requiring preventive removal of the gonads[Gravholt et al., 2000]. The frequency of Y mosaicismreported varies considerably from 0±61%, mainly dueto methodological differences [Larsen et al., 1995;FernaÂndez et al., 1996; LoÂpez et al., 1998]. In contrast,relatively few approaches have been made to detect Xcryptic mosaicism in 45,X individuals and the fre-quencies reported also differ noticeably from 0±75%[Hassold et al., 1988; Mathur et al., 1991; Hassold et al.,1992; Larsen et al., 1995; Jacobs et al., 1997; Yorifujiet al., 1997; MartõÂnez-Pasarell et al., 1999a; Nazarenkoet al., 1999; FernaÂndez-GarcõÂa et al., 2000].

Grant sponsor: CONACYT; Grant number: G-28494 M.

*Correspondence to: Susana Kofman-Alfaro, Servicio deGeneÂtica, Hospital General de MeÂxico, Dr. Balmis 148, MeÂxico,D.F. 06726, MeÂxico. E-mail: skofman@servidor.unam.mx

Received 1 March 2001; Accepted 31 August 2001

ß 2001 Wiley-Liss, Inc.DOI 10.1002/ajmg.10113

Molecular evidence has demonstrated that 70±80%of 45,X patients retain the maternal X chromosome[Hassold et al., 1988; Jacobs et al., 1990; 1997; Larsenet al., 1995; MartõÂnez-Pasarell et al., 1999a,c]. It hasbeen proposed that most live-born 45,X individualsarise from a mitotic error in an early zygote resulting inmosaics [Hook and Warburton, 1983; Cockwell et al.,1991; Mathur et al., 1991; Held et al., 1992]. If this werethe case, it would mean that the etiology of apparentlynon-mosaic TS patients is generally the post-zygoticloss of a sex chromosome. If mitotic non-disjunctionoccurs with equal frequency in XX or XY zygotes, thenthe probability that the XM would be retained would betwice than the XP, taking into account the non-viabilityof 45,Y products [Mathur et al., 1991; Jacobs et al.,1997]. It has also been proposed that the excess of 45,Xproducts that retain the XM could be also due to theincreased likelihood of the XP being lost at the pro-nuclear stage, a mechanism responsible for many45,XM mice [Chandley, 1991]. Particular susceptibilityfor XY non-disjunction during meiosis I has been shownin the analysis of sex chromosome aneuploidy in spermfrom fathers of TS patients [MartõÂnez-Pasarell et al.,1999b] and in FISH studies in sperm of normal men,suggesting an effect of paternal age [Grif®n et al.,1995]. This XY susceptibility to segregation errorscould be related to the small region of pairing be-tween both sex chromosomes. In fact, a correlationbetween reduced recombination and non-disjunctionhas been established in sex chromosome trisomies,47,XXX and 47,XXY of paternal and maternal origin[Hassold et al., 2000].

The most frequent structural abnormality present inTurner patients is i(Xq), in mosaic or non-mosaic form.This abnormal chromosome could be originated duringmale or female meiosis with a similar frequency [Lorda-SaÂnchez et al., 1991; Wolff et al., 1996]. Approximately15% of TS patients are mosaics with 45,X and a secondcell line with a ring X chromosome. It appears that twobreaks rearrangements involving a single X, whetherthe breaks are on the same arm (deletions) or oppositearms (rings), occur predominately during paternalmeiosis as in the rest of the chromosomes probablydue to major liability [Callen et al., 1991; Jacobs et al.,1997].

In this study, we carried out a molecular analysis in15 TS patients using X microsatellites. Ten cases werecytogenetically 45,X and the rest had a second cell linewith an abnormal X chromosome. The aims of ourinvestigation were: 1) to detect X cryptic mosaicism inpatients with a 45,X constitution; 2) to determine theparental origin of the abnormality; 3) to infer thezygotic origin of the karyotype; and 4) to suggestthe timing and mechanism of the error(s) leading tothe formation of abnormal X chromosomes.

MATERIALS AND METHODS

Subjects

Fifteen unrelated TS patients of Mexican mestizoethnic origin and their parents were studied. Clinicaland cytogenetic data are shown in Table I. Chromoso-

mal analysis was carried out in 100 metaphases withGTG banding from peripheral blood leukocytes in allpatients. Ten cases were 45,X whereas the rest weremosaics with a second cell line with an abnormalX chromosome: one isochromosome, pseudodicentric a-large and a-small rings, and one with a long armdeletion. Figure 1 shows partial karyotypes fromPatients 11±15. CBG banding was undertaken in thosecases with abnormal X chromosomes to determinewhether they were mono- or dicentric. Clinical andcytogenetic data of Patient 15 will be reported else-where [Mesa-Cornejo et al., in press]. In all cases,Y mosaicism was previously excluded by PCR ampli®-cation of four Y-speci®c sequences [LoÂpez et al., 1998].

Molecular Analysis

DNA was isolated from blood samples from allpatients and their parents by standard techniques[Sambrook et al., 1989]. In Patient 10 no blood samplefrom the father was available. Parental origin of Xchromosomes was determined by PCR ampli®cationusing (CA/GT)n X chromosome microsatellites (p22.2±q28) from panel 28 ABI-PRISM Linkage Mapping SetVer 2.0-PE containing 18 polymorphic markers (Fig. 2).In 45,X patients, at least ®ve markers were ampli®edincluding two located in the short arm, one pericen-tromeric and two of the long arm to exclude cryptic Xmosaicism and to determine the parental origin. Incases with a structurally abnormal X chromosome, themarkers were chosen according to cytogenetic data.PCR products were analyzed on an ABI 310 automatedsequencer with GENESCAN 3.1 software. Markeralleles frequencies were previously calculated from 30independent chromosomes in Mexican population (datanot shown).

RESULTS

At least ®ve informative markers were enough todetermine the parental origin of X chromosomes in allpatients. In Patient 10, in whom the father was notavailable, the origin of the single X chromosome wasmaternal because the proband and her mother sharedan allele at each of the ®ve loci tested.

None of the ten 45,X patients was heterozygous forany microsatellite, excluding a second cell line with ahidden X mosaicism. The single X chromosome wasmaternally derived in 90% of these cases (Table II).

In the ®ve mosaic cases, the structurally abnormal Xchromosome were maternal in three patients [(45,X/46,X,psu dic(X)(p11.2), 45,X/46,X,r(X)(p22.2q28) and45,X/46,X,del(X)(q23)], whereas the other two [45,X/46,X,i(Xq) and 45,X/46,X,r(X)(p21q13)] were paternal(Table III). None of these patients had uniparentaldisomy.

The results were used to corroborate breakpoints inthe ®ve cases with an abnormal X chromosome. In twoof these patients, one with a pseudodicentric chromo-some and one with Xq deletion, both of maternal origin,additional members of their families were studied toestablish the haplotypes of maternal X chromosomes(Figs. 3, 4). These data were analyzed to infer the

182 Monroy et al.

recombination events during maternal meiosis thatcould be involved in the mechanism of formation of theaberrant chromosomes.

DISCUSSION

In this paper we describe a molecular investigation in15 TS patients with different karyotypes to detect Xhidden mosaicism, determine the parental origin of theabnormality, infer the zygotic origin of the karyotypeand suggest the mechanisms of formation of abnormalX chromosomes.

Several attempts have been made to detect cryptic Xchromosome mosaicism in TS, and it may be present in0±75% of patients cytogenetically diagnosed as 45,X[Hassold et al., 1988; Mathur et al., 1991; Hassoldet al., 1992; Larsen et al., 1995; Jacobs et al., 1997;Yorifuji et al., 1997; MartõÂnez-Pasarell et al., 1999a;Nazarenko et al., 1999; FernaÂndez-GarcõÂa et al., 2000].In our 45,X patients, the absence of heterozygosity for Xmarkers excluded cryptic mosaicism. This could beeither due to the small size of our sample (10 subjects),the use of a single tissue, or because the previouscytogenetic analysis had already discarded them.Although our data differs from those previouslyreported, we assumed that these patients are non-mosaics in peripheral blood because we employed a

Fig. 1. Partial karyotypes from Patients 11±15 showing the normaland abnormal X chromosomes in each case.

TABLE I. Clinical and Cytogenetic Data in 15 TS Patients

Case Age (years) Height (m) Cytogenetic analysis Clinical features

1 18 1.29 45,X Low hairline, short and wide neck, multiple nevi, widethorax, bilateral camptodactyly

2 7 1.08 45,X Webbed neck, low posterior hairline, short metacarpals andmetatarsals

3 6/12 0.62 45,X Lymphedema of hands and feet, loose folds of skin over theback of the neck

4 2 days 0.49 45,X Hygroma of the neck, lymphedema of hands and feet,ungueal hypoplasia of the feet

5 19 1.34 45,X Cleft palate, low hairline, webbed neck, teletelia6 9/12 0.77 45,X Lymphedema of hands and feet, short and wide neck, wide

thorax, teletelia7 11 1.21 45,X Congenital lymphedema, low hairline, short and wide neck,

teletelia, inverted nipples, multiple nevi, cubitus valgus,shortened 4th metatarsal

8 34 1.37 45,X Low hairline, clinodactyly of both 5th toes, multiple nevi9 16 1.37 45,X Short and wide neck, wide thorax, teletelia, inverted

nipples, cubitus valgus, shortened 4th metacarpal,hypoplastic labia, hypertrophia of the clitoris, multiplenevi, left kidney mal rotation

10 18 1.46 45,X Short neck, wide thorax, teletelia, cubitus valgus, shor-tened 4th and 5th metacarpals and metatarsals, multiplenevi

11 23 1.41 45,X[57]/46,X,i(Xq)[43] Low hairline, short and wide neck, cubitus valgus,shortened 4th metacarpals and metatarsals, multiplenevi

12 7 1.13 45,X[45]/46,X, psudic(X)(p11.2)[55]

Teletelia, low implantation of the ®rst digits, clinodactyly ofthe 5th ®nger

13 5 1.00 45,X[75] /46,X,r(X)(p22.2q28)[25]

Wide thorax, teletelia, hyperkinesia

14 10 1.16 45,X[32]/46,X,r(X)(p21q13)[68]

Short and wide neck, cubitus valgus, shortened 4thmetacarpal, multiple nevi, inverted supernumerarynipples, hypoplastic labia

15 22 1.44 45,X[20]/46,X,del(X)(q23)[80] Shield-like thorax, cubitus valgus, hypoplastic labia,multiple nevi

Microsatellite Analysis in Turner Syndrome 183

very sensitive method of PCR ampli®cation and CArepeat examination. Although no formal measurementof sensitivity was undertaken, by using at least ®veinformative polymorphic markers in each case weincreased the chance of detecting low level X-mosaicism[Larsen et al., 1995; MartõÂnez-Pasarell et al., 1999a]. Itis also important to note that capillary electrophoresisis an analytical tool with extremely high sensitivity (ofthe order of yoctomoles) with laser-induced ¯uorescentdetection [Righetti and Gel®, 1997]. We cannot, how-ever, exclude a small Xq marker chromosome becausean Xq pericentromeric microsatellite was not included.

It has been speculated that most live-born 45,Xindividuals arise from a mitotic error in an early zygote,and hence are actually mosaics [Hook and Warburton,

Fig. 2. Localization of the X-chromosome polymorphic markers used.

TA

BL

EII

.R

esu

lts

ofX

Mark

ers

an

dP

are

nta

lage

in45,X

Pati

ents

Case

Kary

otyp

e

Xch

rom

osom

em

ark

ers

use

dP

are

nta

lage

(Yea

rs)

DX

S-

1060

DX

S-

8051

DX

S-

987

DX

S-

1226

DX

S-

1068

DX

S-

991

DX

S-

986

DX

S-

1106

DX

S-

8055

DX

S-

1073

Fath

erM

oth

erP

are

nta

lor

igin

ofsi

ngle

Xch

rom

osom

e

145,X

MN

DN

DN

DU

MM

MM

U47

44

Mate

rnal

245,X

UN

DP

ND

UP

PU

PP

32

33

Pate

rnal

345,X

UN

DM

ND

UM

MM

MU

19

18

Mate

rnal

445,X

UU

MN

DM

MN

DM

ND

M34

36

Mate

rnal

545,X

UM

MU

ND

MM

ND

ND

M35

33

Mate

rnal

645,X

UM

UM

ND

MM

ND

ND

M28

33

Mate

rnal

745,X

ND

MM

MN

DM

ND

MM

ND

32

29

Mate

rnal

845,X

ND

UM

MN

DM

ND

MM

ND

35

19

Mate

rnal

945,X

ND

MM

ND

ND

MM

ND

MN

D15

26

Mate

rnal

10

45,X

ND

MM

MN

DM

ND

MM

ND

20

16

Mate

rnal

X�

29.7

X�

28.7

M,

mate

rnal

all

ele;

P,

pate

rnal

all

ele;

U,

un

info

rmati

ve;

ND

,n

ond

eter

min

ed.

184 Monroy et al.

1983; Held et al., 1992]. The exact mechanism by whichmosaicism arises remains unclear. Recently, Leonovaand Hanson [1999] demonstrated that 20 mosaicTurner patients were heterozygous for the androgenreceptor gene, con®rming that in most 45,X/46,XXcases, mosaicism arises through loss of the X chromo-somes in some cell lines of originally 46,XX zygotes,rather than through ``rescue'' of 45,X conceptuses by amitotic non-disjunction during early embryogenesis.Thus, if the original zygotes were 45,XM and thesubjects were cryptic mosaics they would have unipar-ental isodisomy (UPD). In our 10 45,X individuals wecannot exclude UPD, but according to Jacobs et al.[1997] and Leonova and Hanson [1999] this is a highlyimprobable event. It has also been suggested that UPDmay contribute to mental retardation in Turnerpatients [Yorifuji et al., 1998] but none of our 45,Xcases exhibited this anomaly.

We determined the parental origin in all the 45,Xpatients. In nine of them the single X was maternal andpaternal in the remaining case. These data are inagreement with other reported series [Hassold et al.,1988, 1992; Jacobs et al., 1990, 1997; Villamar et al.,1990; Mathur et al., 1991; Cockwell et al., 1991;Loughlin et al., 1991; Lorda-SaÂnchez et al., 1992; Chuet al., 1994; Collins et al., 1994; Larsen et al., 1995;Yorifuji et al., 1997; MartõÂnez-Pasarell et al., 1999a;Tsezou et al., 1999]. To explain the high frequency ofmaternal X chromosome it has been proposed that 45,Xproducts retaining the maternal X chromosome couldbe due to the increased loss of the paternal X at thepronuclear stage [Chandley, 1991]. In a recent study, alow frequency of cryptic mosaicism (2.3%) and a lowproportion of normal cells (< 10%) were found in mostTurner syndrome patients. They concluded thatalthough some abnormalities may arise at the pro-nuclear stage or during the ®rst zygotic division, mostof them result from paternal errors during gametogen-esis [Jacobs et al., 1997]. Our data are in agreementwith this proposal, suggesting that the majority of 45,XTS patients are non-mosaics, and resulted from anerror during spermatogenesis. Particular susceptibilityfor XY non-disjunction during meiosis I has also beenobserved in the father's sperm of 45,XM patients[MartõÂnez-Pasarell et al., 1999b,c], and in FISH studiesin sperm of normal men suggesting a paternal ageeffect [Grif®n et al., 1995].

For Jacobs et al. [1997], however, parental age doesnot appear to be an important etiologic factor in theorigin of 45,X individuals. Our data in 10 45,X patientscon®rm this hypothesis (Table II). In our cases withstructurally X abnormalities, however, parental agewas slightly higher than the average age of the 45,Xparents (Table III). These data are in agreement withprevious reports where a signi®cantly higher maternalage was found for maternal isochromosomes(31.2�7.5) versus those paternally derived (24.2� 3)[Lorda-SaÂnchez et al., 1991].

Structural abnormalities of the X chromosomeappear equally likely to occur in either parent [Lorda-SaÂnchez et al., 1991; Chu et al., 1994; Collins et al.,1994; Dalton et al., 1998; Tsezou et al., 1999]. In our ®ve

TA

BL

EII

I.R

esu

lts

ofX

Mark

ers

an

dP

are

nta

lA

ge

inP

ati

ents

Wit

hA

bn

orm

al

XC

hro

mos

omes

Case

Kary

otyp

e

Xm

ark

ers

use

dP

are

nta

lage

(yea

rs)

Ori

gin

ofabn

orm

al

XD

XS

1060

DX

S8051

DX

S987

DX

S1226

DX

S993

DX

S991

DX

S986

DX

S1001

DS

X1073

Fath

erM

oth

er

11

45,X

/46,X

,I(X

q)

UU

M/-

M/-

ND

M/-

M/P

ND

M/P

40

33

Pate

rnal

12

45,X

/46,X

,p

sud

ic(X

)(p

11.2

)N

DP

/-P

/-P

/-P

/M?

P/M

P/M

ND

ND

33

33

Mate

rnal

13

45,X

/46,X

,r(X

)(p

22.2

q28)

P/-

P/-

P/M

UP

/MP

/MN

DP

/MP

/M28

31

Mate

rnal

14

45,X

/46,X

,r(X

)(p

21q13)

UM

/-M

/-M

/-U

M/P

M/-

M/-

U26

24

Pate

rnal

DX

S8051

DX

S1226

DX

S993

DX

S986

DX

S990

DX

S1106

DX

S8055

DX

S1227

DX

S8043

15

45,X

/46,X

,del

(X)(

q23)

P/M

P/M

P/M

P/M

P/M

P/M

P/-

P/-

P/-

45

43

Mate

rnal

X�

34.4

X�

32.8

M,

mate

rnal

all

ele;

P,

pate

rnal

all

ele;

U,

un

info

rmati

ve;

ND

,n

ond

eter

min

ed.

Xch

rom

osom

eall

eles

inth

e®

rst

cell

lin

e/all

eles

inth

ese

con

dce

llli

ne.

Microsatellite Analysis in Turner Syndrome 185

Fig. 3. Pedigree and haplotype analysis of Patient 12. The haplotypes of I.1 and I.2 were deduced. *Meiotic recombination.

Fig. 4. Pedigree and haplotype analysis of Patient 15. *Meiotic recombination.ss

186 Monroy et al.

patients, three abnormal chromosomes were of mater-nal origin and two paternal (Table III). Some reportsindicate that the frequency of breaks occurring duringpaternal meiosis is higher than in maternal; however asthe probability of recombination errors between bothmaternal X chromosomes is higher these could origi-nate equal proportions in each gender [Jacobs et al.,1997]. Xq isochromosomes could result either fromcentromere misdivision or sister-chromatid exchange,but as the majority of the isochromosomes are dicentric,this last mechanism is the most probable [Harbisonet al., 1988; Hassold et al., 1988; Jacobs et al., 1990,1997]. Heterozygosity of the two Xq arms has beendocumented, suggesting meiotic recombination as analternative mechanism [Lorda-SaÂnchez et al., 1991].The majority of dicentric X chromosomes are pseudodi-centric, with only one active centromere [Sullivan andWillard, 1998; Therman et al., 1986]. It is well knownthat some isochromosomes may be mitotically unstableand could be easily lost giving rise to a 45,X cell line[Wolff et al., 1996; Dalton et al., 1998]. In Patient 12(Fig. 3), the pseudodicentric chromosome shows onlyone maternal allele at DXS991 but because DXS993was an uninformative marker we cannot concludewhether the patient is homozygous for DXS993 or ifthe maternal allele was lost during the formation of theabnormal chromosome. A recombination event wasobserved at DXS991 that probably coincides with thebreakpoint for the rearrangement of the pseudodi-centric. Alternatively, the breakpoint could lie aboveDXS991 or even above DXS993. As it is shown in Fig. 5,this abnormal chromosome was originated in the ®rstpost-zygotic division by sister chromatids exchangewith posterior loss of the unstable dicentric chromo-some. In Case 11, the isochromosome was paternallyderived, and it was apparently monocentric by cytoge-netics; furthermore the patient did not show a paternalallele at DXS991 in Xp11.1.

In Case 15, with the Xq deleted chromosome, abreakpoint in Xq23 was cytogenetically observed. Thepatient only showed one allele at DXS1106 in Xq22indicating homozygosity for this polymorphism (Fig. 4).These data suggest a recombination event in DXS1106,so the breakpoint must lie between this marker andDXS8055 in Xq23. The abnormal X chromosomeprobably originated during the recombination event inmaternal meiosis as it is shown in Figure 6. A post-zygotic event cannot, however, be excluded.

Ring chromosomes are considered to arise fromchromosome breaks occurring on either side of thecentromere and subsequent rejoining of the brokenends. In our patients with ring X chromosomes (Cases13 and 14) the use of X polymorphic markers con®rmedthe breakpoints cytogenetically observed. None of thesesubjects exhibited a severe phenotype, althoughPatient 13 had a large ring chromosome compared tothe very tiny ring chromosome observed in Patient 14.It had been assumed that small ring (X) chromosomeslacking the XIST gene at Xq 13.2 were associated with asevere phenotype including mental retardation, facialdysmorphism and congenital abnormalities [Van Dykeet al., 1992; Dennis et al., 1993; McGinniss et al., 1997],although mental retardation was observed in twopatients with ring (X) chromosomes that included anintact XIST locus [Migeon et al., 2000]. Conversely, anunexpected mild phenotype was found in six patientswith ring (X) lacking this gene [Turner et al., 2000].According to our molecular data, none of the two caseswith ring X chromosomes had uniparental disomy.Usually, large ring (X) chromosomes result in a mildphenotype, but in mosaic Turner cases the presence ofXIST in abnormal X chromosomes was shown to resultin a low IQ [Collins et al., 1994] or even in mentalretardation [Matsuo et al., 2000]. Interestingly, thelarge ring X chromosome in our patient has similarbreakpoints as the one previously reported in a mother

Fig. 5. Schematic representation of the possible origin of the pseudodi-centric X chromosome. A: The interstitial recombination between the Xchromosomes (white and black) during maternal meiosis is indicated by asmall arrow. B, C: Results from maternal meiosis I and II respectively, withrecombined region in Xp shown in white. D: Paternal X chromosome (gray)and recombined maternal X chromosome (black and white) during the ®rstpost-zygotic division in the propositi. Dotted arrows show breakpoints in Xpwith an interchange between sister chromatids followed by fusion. E: Cellline with paternal X chromosome and pseudodicentric maternal Xchromosome. F: Final result of post-zygotic mitosis, the loss of the unstabledicentric chromosome could originate the 45,X cell line.

Fig. 6. Schematic representation of the origin of the Xq23 deletion. A:Meiotic recombination between the maternal X chromosomes resulting inbreakage and loss of a chromosomal segment. B: X chromosomes after ®rstmaternal meiosis showing an Xq chromatid deletion, and (C) X chromo-somes from second maternal meiosis, including the Xq deleted chromo-some.

Microsatellite Analysis in Turner Syndrome 187

and daughter with mental retardation [Matsuo et al.,2000]. The mild phenotype of Patient 13 was probablydue to a skewed pattern of X-inactivation, as opposed tothe random X-inactivation that resulted in mentalretardation in the familiar case [Matsuo et al., 2000].Unfortunately, we were not able to analyze themechanism of formation in this maternally derivedring because the rest of the family was not available.The absence of a severe phenotype in Patient 14 with asmall ring (X) chromosome probably indicates thepresence of an active XIST resulting in the lack offunctional disomy of a relatively large amount of X-euchromatin (Xp21-q13) [Yorifuji et al., 1998; Turneret al., 2000].

REFERENCES

Callen DF, Eyre HJ, Ringenbergs ML, Freemantle CJ, Woodroffe P, HaanEA. 1991. Chromosomal origin of small ring marker chromosomes inman: characterization by molecular genetics. Am J Hum Genet 48:769±782.

Chandley AC. 1991. On the parental origin of the novo mutation in man. JMed Genet 28:217±223.

Chu CE, Donaldson MDC, Kelnar CJH, Smail PJ, Greene SA, PatersonWF, Connor JM. 1994. Possible role of imprinting in the Turnerphenotype. J Med Genet 31:840±842.

Cockwell A, MacKenzie M, Youings S, Jacobs P. 1991. A cytogenetic andmolecular study of a series of 45,X fetuses and their parents. J MedGenet 28:151±155.

Collins AL, Cockwell AE, Jacobs PA, Dennis NR. 1994. A comparison of theclinical and cytogenetic ®ndings in nine patients with a ring(X) cell lineand 16 45,X patients. J Med Genet 31:528±533.

Dalton P, Coppin B, James R, Skuse D, Jacobs P. 1998. Three patients witha 45,X/46,X,psu dic (Xp) karyotype. J Med Genet 35:519±524.

Dennis NR, Collins AL, Crolla JA, Cockwell AE, Fisher AM, Jacobs PA.1993. Three patients with ring (X) chromosomes and a severephenotype. J Med Genet 30:482±486.

FernaÂndez R, MeÂndez J, PaÂsaro E. 1996. Turner syndrome: a study ofchromosomal mosaicism. Hum Genet 98:29±35.

FernaÂndez-GarcõÂa R, GarcõÂa-Doval S, Costoya S, PaÂsaro E. 2000. Analysisof sex chromosome aneuploidy in 41 patients with Turner syndrome: astudy of ``hidden'' mosaicism. Clin Genet 58:201±208.

Gicquel C, Cabrol S, Schneid F, Le Bouc Y. 1992. Molecular diagnosis ofTurner syndrome. J Med Genet 29:547±551.

Gravholt CH, Fedder J, Naeraa RW, Muller J. 2000. Occurrence ofgonadoblastoma in females with Turner syndrome and Y chromosomematerial: a population study. J Clin Endocrinol Metab 85:199±202.

Grif®n DK, Abruzzo MA, Millie EA, Sheean LA, Feingold E, Sherman SL,Hassold TJ. 1995. Non-disjunction in human sperm: evidence for aneffect of increasing paternal age. Hum Mol Genet 4:2227±2232.

Hall J, Gilchrist D. 1990. Turner syndrome and its variants. Pediatr ClinNorth Am 37:1421±1440.

Harbison M, Hassold T, Kobrryn C, Jacobs PA. 1988. Molecular studies ofthe parental origin and nature of human X isochromosomes. CytogenetCell Genet 47:217±222.

Hassold T, Benham F, Leppert M. 1988. Cytogenetic and molecularanalysis of sex-chromosome monosomy. Am J Hum Genet 42:534±541.

Hassold T, Pettay D, Robinson A, Uchida I. 1992. Molecular studies ofparental origin and mosaicism in 45,X conceptuses. Hum Genet89:647±652.

Hassold T, Sherman S, Hunt P. 2000. Counting cross-overs: characterizingmeiotic recombination in mammals. Hum Mol Genet 9:2409±2419.

Held KR, Keber S, Kaminsky E, Singh S, Goetz P, Seemanova E, GoeddeHW. 1992. Mosaicism in 45,X Turner syndrome: does survival in earlypregnancy depend on the presence of two sex chromosomes? HumGenet 88:288±294.

Hook EB, Warburton D. 1983. The distribution of chromosomal genotypesassociated with Turner syndrome: livebirth prevalence rates and

evidence for diminished fetal mortality and severity in genotypesassociated with structural X abnormalities or mosaicism. Hum Genet64:24±27.

Jacobs PA, Betts PR, Cockwell AE, Crolla JA, Mackenzie MJ, Robinson DO,Youings SA. 1990. A cytogenetic and molecular reappraisal of aseries of patients with Turner syndrome. Ann Hum Genet 54:209±223.

Jacobs P, Dalton P, James R, Mosse K, Power M, Robinson D, Skuse D.1997. Turner syndrome: a cytogenetic and molecular study. Ann HumGenet 61:471±483.

LoÂpez M, Canto C, Aguinaga M, Torres L, Cervantes A, Alfaro G, MeÂndezJP, Kofman-Alfaro S. 1998. Frequency of Y chromosomal material inMexican patients with Ullrich-Turner syndrome. Am J Med Genet76:120±124.

Lorda-Sanchez I, Binkert F, Maechler M, Schinzel A. 1992. Molecularstudy of 45, X conceptuses: Correlation with clinical ®ndings. Am J MedGenet 42:487±490.

Larsen T, Gravinoit CH, Tillebeck A, Larsen H, Jensen MB, Nielsen J,Friedrich U. 1995. Parental origin of the X chromosome, X chromosomemosaicism and screening for ``hidden '' Y chromosome in 45,X Turnersyndrome ascertained cytogenetically. Clin Genet 48:6±11.

Leonova J, Hanson C. 1999. A study of 45,X/46,XX mosaicism in Turnersyndrome females: a novel primer pair for the (CAG)n repeat within theandrogen receptor gene. Hereditas 131:87±92.

Lorda-SaÂnchez I, Binkert F, Maechler M, Schinzel A. 1991. A molecularstudy of X isochromosomes: parental origin, centromeric structure, andmechanisms of formation. Am J Hum Genet 49:1034±1040.

Loughlin SAR, Redha A, McIver J, Boyd E, Carothers A, Connor JM. 1991.Analysis of the origin of Turner syndrome using polymorphic DNAprobes. J Med Genet 28:156±158.

MartõÂnez-Pasarell O, Templado C, Egozcue J, Vicens-Calvet E, NogueÂs C.1999a. PCR protocol to detect parental origin and hidden mosaicism insex chromosome aneuploidies. Horm Res 51:248±252.

MartõÂnez-Pasarell O, NogueÂs C, Bosch M, Egozcue J, Templado C. 1999b.Analysis of sex chromosome aneuploidy in sperm from fathers of Turnersyndrome patients. Hum Genet 104:345±349.

MartõÂnez-Pasarell O, Templado C, Vicens-Calvet E, Egozcue J, NogueÂs C.1999c. Paternal sex chromosome aneuploidy as a possible origin ofTurner syndrome in monozygotic twins: case report. Hum Reprod14:2735±2738.

Mathur A, Stekol L, Schatz D, MacLaren NK, Scott ML, Lippe B. 1991. Theparental origin of the single X chromosome in Turner syndrome: lack ofcorrelation with parental age or clinical phenotype. Am J Hum Genet48:682±686.

Matsuo M, Muroya K, Nanao K, Hasegawa Y, Terasaki H, Kosaki K, OgataT. 2000. Mother and daughter with 45,X/46,X,r(X)(p22.3q28) andmental retardation: analysis of the X inactivation patterns. Am JMed Genet 91:267±272.

McGinniss MJ, Brown DH, Burke LW, Mascarello JT, Jones MC. 1997.Ring chromosome X in a child with manifestations of Kabuki syndrome.Am J Med Genet 70:37±42.

Mesa-Cornejo VM, GarcõÂa-Cruz D, Monroy-Jaramillo N, VaÂzquez AI,DaÂvalos NO, Galaviz C, Kofman S. Del Xq23 in a mosaic Turnerfemale: molecular and cytogenetic studies. Ann Genet (in press).

Migeon BR, Ausems M, Giltay J, Hasley-Royster C, Kazi E, Lydon TJ,Engelen JJM, Raymond GV. 2000. Severe phenotypes associated withinactive ring X chromosomes. Am J Med genet 93:52±57.

Nazarenko SA, Timoshevsky VA, Sukhanova NN. 1999. High frequency oftissue-speci®c mosaicism in Turner syndrome patients. Clin Genet56:59±65.

Righetti PG, Gel® C. 1997. Capillary electrophoresis of DNA for moleculardiagnostics. Electrophoresis 18:1709±1714.

Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular cloning: a laboratorymanual, 2nd ed. Cold Spring Harbor, NY: Cold Spring HarborLaboratory Press.

Sullivan BA, Willard HF. 1998. Stable dicentric X chromosomes with twofunctional centromeres. Nat Genet 20:227±228.

Therman E, Trunca C, Kuhn EM, Sarto GE. 1986. Dicentric chromosomesand the inactivation of the centromere. Hum Genet 72:191±195.

Tsezou A, Hadjiathanasiou C, Gourgiotis D, Galla A, Kavazarakis E,Pasparaki A, Kapsetaki M, Sismani C, Theodoridis C, Patsalis PC,Moschonas N, Kitsiou S. 1999. Molecular genetics of Turner syndrome:

188 Monroy et al.

correlation with clinical phenotype and response to growth hormonetherapy. Clin Genet 56:441±446.

Turner C, Dennis NR, Skuse DH, Jacobs PA. 2000. Seven ring (X)chromosomes lacking the XIST locus, six with an unexpectedly mildphenotype. Hum Genet 106:93±100.

Van Dyke DL, Wiktor A, Palmer CG, Miller DA, Witt M, Babu VR,Worsham MJ, Roberson JR, Weiss L. 1992. Ullrich-Turner syndromewith a small ring X chromosome and presence of mental retardation.Am J Med Genet 43:996±1005.

Villamar M, FernaÂndez E, Ayuso C, Ramos C, BenõÂtez J. 1990. Study of theparental origin of sexual aneuploidy in ten families using RFLPs. AnnGenet 33:29±31.

Wolff DJ, Miller AP, Van Dyke DL, Schwartz S, Willard HF. 1996.Molecular de®nition of breakpoints associated with human Xqisochromosomes: implications for mechanism of formation. Am J HumGenet 58:154±160.

Yorifuji T, Muroi J, Kawai M, Sasaki H, Momoi T, Furusho K. 1997. PCR-based detection of mosaicism in Turner syndrome patients. Hum Genet99:62±65.

Yorifuji T, Muroi J, Kawai M, Uematsu A, Sasaki H, Momoi T, Kaji M,Yamanaka C, Furusho K. 1998. Uniparental and functional X disomyin Turner syndrome patients with unexplained mental retardationand X derived marker chromosomes. J Med Genet 35:539±544.

Microsatellite Analysis in Turner Syndrome 189