Heterozygosity mapping by quantitative ... - Semantic Scholar

2 downloads 0 Views 174KB Size Report
Jul 28, 2005 - BACKGROUND: Deletions of Xq chromosome are reported for a number of familial conditions exhibiting prema- ture ovarian failure (POF) and ...
Human Reproduction Vol.21, No.2 pp. 529–535, 2006

doi:10.1093/humrep/dei356

Advance Access publication October 20, 2005.

Heterozygosity mapping by quantitative fluorescent PCR reveals an interstitial deletion in Xq26.2–q28 associated with ovarian dysfunction Giorgia Fimiani, Carmela Laperuta, Geppino Falco, Valerio Ventruto, Michele D’Urso, Matilde Valeria Ursini and Maria Giuseppina Miano1 Institute of Genetics and Biophysics ‘Adriano Buzzati Traverso’, CNR 80131, Napoli, Italy 1

To whom correspondence should be addressed at: IGB-ABT CNR, Via Pietro Castellino 111, 80131 Napoli, Italy. E-mail: [email protected]

BACKGROUND: Deletions of Xq chromosome are reported for a number of familial conditions exhibiting premature ovarian failure (POF) and early menopause (EM). METHODS AND RESULTS: We describe the inheritance of an interstitial deletion of the long arm of the X chromosome associated with either POF or EM in the same family. Cytogenetic studies and heterozygosity mapping by quantitative fluorescent PCR revealed a 46,X,del(X)(q26.2–q28) karyotype in a POF female, in her EM mother, and also in her aborted fetus with severe cardiopathy. Applying a microsatellite approach, we have narrowed the extension of an identical interstitial deletion located between DXS1187 and DXS1073. These data, in line with other mapped deletions, single out the proximal Xq28 as the region most frequently involved in ovarian failure. We also propose that other factors may influence the phenotypic effect of this alteration. Indeed, skewed X inactivation has been ascertained in EM and POF to be associated with different X haplotypes. CONCLUSION: Our analysis indicates that Xq26.2–q28 deletion is responsible for gonad dysgenesis in a family with EM/POF. The dissimilar deletion penetrance may be due to epigenetic modifications of other X genes that can contribute to human reproduction, highlighting that ovarian failure should be considered as a multifactorial disease. Key words: heterozygosity mapping/premature ovarian failure syndrome/quantitative fluorescent PCR/XCI/Xq26.2–Xq28

Introduction Premature ovarian failure (POF) and early menopause (EM) conditions are present in a broad spectrum of gonad dysgenesis, from a complete cessation of ovarian function to an intermittent follicle maturation failure. They are often present in members of the same family in association with alterations of the Xq chromosome (Vegetti et al., 2000; Zinn, 2001). This intra-familial variability may represent a different degree of expression of the same genetic disease explained by genetic heterogeneity and/or by different environmental factors. The complex spectrum of these conditions is also enhanced by the huge repertoire of POF with loss of different genome portions that may cause defects in oogenesis or impairment of follicle maturation (Schlessinger et al., 2002). Moreover, POF patients are a fascinating disease group because those carrying a structural abnormality of the X chromosome exhibit extremely skewed XCI (X chromosome inactivation) due to cell selection (Sato et al., 2004). During the past few years, accurate mapping of X chromosome anomalies has been successfully achieved by quantitative fluorescent PCR (QF-PCR) amplification of chromosome-specific short tandem repeats (STR)

(Donaghue et al., 2003). More precisely defined region(s) on Xq have been mapped in various POF diseases (Maraschio et al., 1996; Marozzi et al., 2000) although neither a clear relationship between the position of breaks within this region nor the severity of POF has been established. Findings derived from investigation of X/autosome balanced translocations and Xq terminal deletions have allowed identification of two independent loci within the Xq arm that seem to be involved in ovarian function (Marozzi et al., 2000). These loci are located at Xq26–q28 (POF1) and at Xq13.3–q22 (POF2) respectively. Which genes cause POF conditions is not completely understood. In the POF2 locus, translocation mapping studies have revealed three different genes interrupted in families with POF, DIAPH2 (Diaphanous Drosophilae homologue 2; Sala et al., 1997), DACH2 (Drosophilae dachshund homologue 2; Prueitt et al., 2002; Bione et al., 2004) and POF1B (Bione et al., 2004). In the POF1 locus no disease gene has been ascertained, although the association between FMR1 premutations and POF disease has suggested that this gene acts as a POF risk factor with an estimated relative risk of ∼20% (Murray et al., 2000; Sullivan et al., 2005). Because these disorders are highly frequent

© The Author 2005. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved. For Permissions, please email: [email protected]

529

G.Fimiani et al.

in the female population (1%, Sato et al., 2004), finding POF genes may improve our knowledge of reproductive capacity, allowing the prediction of impending menopause and the implementation of strategies to advance conception. In the present study, we report on a female with POF and her mother with EM with the genotype 46,X,del(Xq26.2–q28). We have constructed a detailed map of a cytogenetically visible deletion of Xq present in both females, using QF-PCR and microsatellite analysis, to assess the X chromosome DNA dosage and to determine accurately the extension of the monosomy. Differences in the X chromosome inactivation patterns have been evaluated in association with EM/POF conditions. Materials and methods Subjects According to common protocol, POF and EM are defined as hypergonadotrophic amenorrhoea with FSH measurements ≥20 IU, LH >15 IU and cessation of menses for a duration of ≥6 months, at age ≤40 years (POF), or between the ages of 41 and 44 years (EM) (Anasti, 1998; Vegetti et al., 2000). A 26 year old Italian woman, II:2, was referred because of periods of amenorrhoea alternating with regular menstruation. She had menarche at 10 years, was of normal height (160 cm) and had none of the Turner syndrome traits. She was pregnant with a malformed female fetus, and therapeutic abortion was decided at 14 weeks. Tissue samples of the abortion were unavailable for histological analysis and we cannot report the phenotype accurately. The mother of II:2, I:1, had menarche at 12 years and had another daughter (II:1) with regular menstruation. At 42 years of age, I:1 entered menopause. Both female subjects, I:1 with EM and II:2 with POF, underwent complete clinical assessment, which included medical analysis, and gynaecological and obstetric history considering the outcome of pregnancies. Chromosomal analysis Peripheral blood samples were obtained from the two probands of the family DG (I:1 and II:2). A cell culture of the fetus biopsy was also obtained (III:1). Tissue culturing was carried out using the standard protocol. Metaphase chromosome spreads were prepared from phytohaemagglutinin-stimulated lymphocytes, and the chromosomes were GTG-banded by standard methods (Lemieux et al., 1990). Routine resolution banding techniques were also performed on metaphase chromosomes. More than 30 cells were analysed for each karyotype. QF-PCR and genotyping DNA was extracted from peripheral blood lymphocytes by high salt extraction (Miller et al., 1988). The DNA concentration was measured with a spectrophotometer, and samples were diluted with distilled water to a concentration of 20 ng/μl. DNA samples were analysed in QF-PCR protocol (Donaghue et al., 2003). This method is based on a PCR amplification that incorporates a fluorochrome into the samples, which are then visualized and quantified using an automated DNA sequencer. The marker set was based on Marshfield Screening Set 11 and consisted of 15 polymorphic microsatellites (STR) with an average spacing of 1 cM and an average heterozygosity of 0.75. Four additional markers, B_53a12/4, B_53a12/5, B_53a12/15 and B_XY_22, were developed ad hoc and used to increase the information content for the distal Xq28–XqPar region. The genetic frequency of the new STR markers was provided from male equimolar DNA pools of

530

normal large multiethnic CEPH (Foundation Jean Dausset-Centre d’Etude du Polymorphisme Humain) families (data unpublished). A heterozygosity index was established and marker positions were verified using human public databases (University of California, Santa Cruz (UCSC) http://genome.ucsc.edu/, GDB http://www.gdb.org/, Table I). Either the forward or the reverse oligonucleotide primer was labelled with 6-FAM, HEX or NED fluorescent dyes (Amersham Biosciences, Buckinghamshire, UK). PCR were performed in a 10 μl volume containing 20 ng of template DNA, 0.8 μmol of each oligonucleotide primer, 200 mmol of each dNTP, and 0.4 IU of AmpliTaq Gold (Amersham Biosciences, Buckinghamshire, UK), in 1×PCR buffer with 1–2 mmol MgCl2 (Applied Biosystems, Foster City CA, USA). DNA was initially denatured at 94°C for 7 min and was then subjected to 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s, followed by a final extension step of 30 min at 72°C. Twenty-four to 26 PCR cycles were carried out as standard practice for QF-PCR analysis. The internal size standard TAMRA 350 was added to 2 μl of PCR products and 24 μl of formamide and then was submitted to capillary electrophoresis in an automatic sequencer (MEGABace1000 Amersham Biosciences, Buckinghamshire, UK). Alleles were assigned with Genetic Profiler 2.0 software (Amersham Biosciences, Buckinghamshire, UK). X inactivation The primers we used flank a sequence containing a highly polymorphic CAG repeat and two HpaII sites in the AR gene (Fusco et al., 2004). When the undigested control DNA is used as the PCR template, both alleles become amplified, whereas using the HpaIIdigested DNA as template, only the inactive allele, which is methylated and therefore resistant to restriction digestion with HpaII, will be amplified. When DNA obtained from the control male is digested with HpaII, no PCR product is obtained. In a sample obtained from the female control, if the X chromosome is subject to random inactivation, both alleles should be present, whereas in the case of skewed inactivation, only the allele that is preferentially inactivated will be amplified. XCI degree threshold patterns are classified as random (XCI ≤ 70%), non-random (70% ≥ XCI ≤ 80%) or skewed (XCI ≥ 90%).

Results Abnormal X determining A 26 year old woman (D.G. II:2) with POF was referred to our laboratory for genetic counselling at 14 weeks of gestation. Abnormal ultrasound findings revealed a fetus (III:1) with cystic hygromas and congenital complex cardiopathy with one ventriculus (Figure 1). After a therapeutic abortion, conventional cytogenetic procedures were performed on the fetus sample (III:1), the POF proband (II:2) and her mother, reported as EM female (I:1). Chromosome metaphases were obtained from peripheral blood lymphocyte cultures of both females and the fetus biopsy. GTG staining showed the presence of a microscopically visible deletion of the long arm of the X chromosome (Xq) in the three samples in which the karyotype designation was 46X, del(X)(q26–qter) (Figure 2). The finding of the abnormal X chromosome in the two females with ovarian dysfunction is completely in line with the literature data that Xq deletions are responsible for the POF phenotypes (Marozzi et al., 2000). We did not have the clinical and molecular evidence to

Xq26.2–q28 deletion mapping in EM/POF family

Table I. Primers for the heterozygosity mapping defining the early menopause/premature ovarian failure deletion Marker

Primer sequence

Heterozygote

Xq band

Position (Mb)

DXS1047

F: 5′-CCGGCTACAAGTGATGTCTA-3′ R: 5′-CCTAGGTAACATAGTGAGACCTTG-3′ F: 5′-CAGCTACTCAATGAAAAGCC-3′ R: 5′-AGAGGAGTTGCAACCCAGA-3′ F: 5′-GATCTTCTGTTCTTATTTTTAAAT-3′ R: 5′-CCGAGAGGACTGTTAAGAG-3′ F: 5′-GCCATTGTAAAATAAAATTCAG-3′ R: 5′-ATGGTCTTGAGTCACTGTCT-3′ F: 5′-GTTGCCAACTGCTGGAACG-3′R: 5′-TGTGGTGCAGGGAAGCC-3′ F: 5′-CCTACGCATGTGGCTC-3′ R: 5′-ATTAATGGCTTAGAGTACTTTTTCA-3′ F: 5′-CTTGCACTTGCTGTGG-3′ R: 5′-AGCTGTAGAGTTGAGGAATG-3′ F: 5′-AGTTCTCAGAAACATTTGGTTAGGC-3′ R: 5′-AATTATTGGCAAAGAGTACAGGCAG-3′ F: 5′-CACATTCAGGTTCCACAGG-3′ R: 5′-CAAGATCCAGGCAAAAGTC-3′ F: 5′-AACAGTCATTGTAGGCATCG-3′ R: 5′-GAATTGCCAGTCATCCC-3′ F: 5′-AGCACCCAGTAAGAGACTGAAC-3′ R: 5′-CCTCAGTGGCAACCACTCAAG-3′ F: 5′-GCTTGAAGTGTCCATGAGGTATC-3′ R: 5′-AGAAGCTGATGTGCTCCCTG-3′ F: 5′-AGCACATGGTATAATGAACCTCCACG-3′ R: 5′-CAGTGTGAGTAGCATGCTAGCATTTG-3′ F: 5′-GGCTGACTCCAGAGGC-3′ R: 5′-CCGAGTTATTACAAAGAAGCAC-3′ F: 5′-GTGAATTCATCATATGTGATTTCC-3′R: 5′-ACTAGGCGACTAATACAGTGGTGC-3′ F: 5′-GGAGATAGGAATGATGGA-3′ R: 5′-AGTTTGAGGATGGTTTGTT-3′ F: 5′-TTTTGTAGGGTTTTGTTTTG-3′ R: 5′-ATGTACCCCAGAACTCTTTA-3′ F:5′-ATGGGAGTGGGTTTACAAT-3′ F:5′-TCAGAAATCCACCCACTAAA-3′ F:5′-TAATGAGAGTTGGAAAGAAA-3′ F:5′-CCCATTGTTGCTACTTGAGA-3′

0.81

Xq25

128.7

0.75

Xq26.2

130.6

0.79

Xq26.2

132.9

0.78

Xq26.3

135.8

0.84

Xq27.1

137.9

0.72

Xq27.2

139.8

0.69

Xq27.3

141.8

0.81

Xq27.3

143.6

0.76

Xq28

147.2

0.62

Xq28

149.2

0.81

Xq28

149.6

0.69

Xq28

151.5

0.84

Xq28

151.9

0.80

Xq28

153.2

0.75

Xq28

154.3

0.40

Xq28

154.4

0.30

Xq28

154.4

0.30

Xq28

154.5

0.86

XqPar

154.6

DXS1187 DXS1254 DXS8094 DXS1192 DXS1205 DXS8106 DXS8043 DXS8091 DXS8069 DXS1684 DXS8061 DXS15 DXS1073 DXS1108 B_53a12/4a B_53a12/5

a

B_53a12/15a B_XY_22

a

a New markers designed on Xq28 sequence [University of California, Santa Cruz (UCSC) Human Genome Release]. Frequency among Caucasians. For the new short tandem repeats, allele frequencies and H-score have been determined in the CEPH (Foundation Jean Dausset-Centre d’Etude du Polymorphisme Humain) families. F = forward; R = reverse.

ascribe the fetus malformation to the Xq26–q28 deletion. Significantly, we cannot exclude the possibility that the severe symptoms are associated with an X inactivation mechanism. QF-PCR and STR mapping Detailed deletion mapping was performed by QF-PCR, a technique used in the accurate diagnosis of the sex chromosome status and its imbalances (Donaghue et al., 2003). The aim of this analysis was to assess the fine mapping of the Xq deletion choosing polymorphic STR within the genetic region involved in the visible cytogenetic alteration, in order to detect the amount of fluorescent activity produced by the X chromosome markers (Table I). In a public database, we chose 15 informative STR with a high heterozygosity index among the tri/tetra/penta/hexanucleotide repeat markers already characterized in the Xq26–Xqter region. We enriched the density of polymorphic tags in distal Xq28 with the integration of four new polymorphic STR developed in our laboratory (B_53a12/4, B_53a12/5, B_53a12/15, B_XY_22; Table I). Using the selected markers, we genotyped in QF-PCR

reactions the DNA of female 46,X,del(Xq26.2–qter) (II:2 Figure 3A) and of two control samples, CEPH 1347M02F (XX) and 1347C16M (XY). In this case, the peaks of fluorescent activity of an X chromosomal marker outside the deleted region of the POF female were identical to the activity of the same marker in a normal female (ratio 1:1). Moreover, the peaks of fluorescent activity of an X chromosomal marker inside the deletion corresponded to one allele and were identical to the activity of the same marker in a normal hemizygous male (ratio 1:1). According to the hemizygosity of each STR in the CEPH control male (1347C16M) with known genotypes, the 1:1 ratio was detected for the contiguous DNA polymorphisms DXS1187–DXS1254–DXS8094–DXS1192–DXS1205–DXS8106– DXS8043–DXS8091–DXS8069–DXS1684–DXS8061–DXS15– DXS1073 (Figure 3B), documenting that the tested female with the deletion was monoallelic for those markers. The heterozygosity correlation between the POF female and the CEPH female control (1347M02F) established that the POF locus was bi-allelic for the markers DXS1047 and DXS1108, flanking the mono-allelic DXS1187 and DXS1073 that are the proximal 531

G.Fimiani et al.

Evaluation of the POF1 deletions

I:1 Height: 160cm Age at menarche: 12 years Age at menopause: 43 years Karyotype: 46,X,del(X)(q26.2q28) XCI: X1 98%; X2 2%

II:2 Height: 160cm Age at menarche: 10 years Age at oligomenorrhea: 26 years Karyotype: 46,X,del(X)(q26.2-q28) XCI: X1 1%; X2 99%

I:1

II:1

II:2

II:3

III:1

III:1 Therapeutic abortion at the 14th week Cystic hygromas; congenital complex cardiopathy with one ventriculus Karyotype: 46,X,del(X)(q26.2-q28) XCI: not determined

Figure 1. Pedigree data of individuals I:1 (early menopause, EM), II:2 (premature ovarian failure, POF) and III:1 (abortion). In each box, the clinical features, the karyotype and the X chromosome inactivation pattern are summarized. EM and POF probands are indicated by arrows.

and distal STR to the deletion respectively (Figure 3B). The deletion is therefore interstitial, located in a region of ∼22 Mb from Xq26.2 to Xq28. Based on the estimated physical distances reported in the X-chromosome consensus map (UCSC Release May 2004 at http://genome.ucsc.edu/; Figure 4) the proximal and distal breakpoints are, respectively, ∼130600 and ∼153200 kb from ptel.

By comparing other POF deletions in the region with our findings, we have reviewed the DNA loss located outside the POF2 locus (Xq13.3–q22) (Table II). Xq interstitial and terminal deletions including POF1 (Xq26–q28) were found from q23 to q28 (Marozzi et al., 2000); from q25 to qter (Davison et al., 1998); from q22.3 to q27 (Vegetti et al., 2000); from q27 to q28 (Rossetti et al., 2004); and from q27.3 to q28 (Eggermann et al., 2005). A comparison of these findings suggests that the minimal POF1 region could be located between DXS1200 (Eggermann et al., 2005) and DXS15 (Rossetti et al., 2004) reducing the critical region roughly to 5 Mb (UCSC Release May 2004; Figure 4). The absence of any correlation between the deletion size and the type of ovarian failure is in agreement with the hypothesis that a restricted Xq region is responsible for the POF/EM phenotype. X inactivation status XCI analysis was performed in the probands, I:1 with EM and II:2 with POF, and in the CEPH DNA controls, 1347M02F (female) and 1347C16M (male). Using the androgen receptor (AR) fluorescent assay that analyses the methylation status of CpG islands of the AR gene, we found that I:1 and II:2 females were extremely skewed (XCI degree threshold ≥90%), with different X haplotypes (Figure 1). In particular, female I:1 showed the degree threshold X1 98%; X2 2%, while her daughter II:2 showed X1 1%; X2 99%. Considering the evidence that POF Xq deletions have different lengths, these results are compatible with the hypothesis that haplo-insufficiency is the molecular mechanism for the POF phenotype. The association

Figure 2. Conventional GTG-banding analysis of the II:2 POF female carrying a karyotype 46,X,del(X)(q26-qter). The arrow identifies the deleted X chromosome.

532

Xq26.2–q28 deletion mapping in EM/POF family

Figure 3. (A) Representative microsatellite results. Electrophoretograms show quantitative fluorescent (QF)-PCR products of a marker outside the premature ovarian failure (POF)/early menopause (EM) deletion (DXS1047) and two markers falling on the POF/EM deletion (DXS1187 and DXS8069). We have genotyped the normal CEPH (Foundation Jean Dausset-Centre d’Etude du Polymorphisme Humain) female (b) and the normal CEPH male (c) with the known genotype for each marker and the POF female II:2 (a). First lane: DXS1047 is heterozygous in the POF female (1:1) and the CEPH female (1:1); peak size in nt: (a) 196/224; (b) 196/224; (c) 224/Y. Second lane: DXS1187 is homozygous in the POF female (1: –) and the CEPH male (1: Y); peak size in nt: (a) 126/–; (b) 118/126; (c) 126/Y. Third lane: DXS8069 is homozygous in the POF female (1: –) and the CEPH male (1: Y); peak size in nt: (a) 168/– ; (b) 168/174; (c) 170/Y. (B) Heterozygosity mapping by QF-PCR. Genotype of X chromosomes in the POF female (II:2) with 15 different alleles spacing with an average of 1–2 Mb and spanning between Xq25 (DXS1047) and Xq28 regions (DXS1108).

Table II. Types of DNA loss in POF1 locus Patient

Cytogenetic band

Deleted markersa

Length(Mb)b

Phenotype

References

C.s.1 L.L. M.O. No. 3 No. 5

Xq25–Xqter Xq23–Xq28 Xq26.1–Xqter Xq22.3–Xq27 Xq26.3–Xqter

DXS6928–Xqter DXS1220–ALD HIGM1–Xqter ND ND

121.0–154.8 114.3–152.5 135.4–154.8 ND ND

Davison et al. (1998) Marozzi et al. (2000) Marozzi et al. (2000) Vegetti et al. (2000) Vegetti et al. (2000)

No. 12 K.F. D.G.

Xq27–Xq28 Xq27.3–Xq28 Xq26.2–Xq28

DXS8106–DXS15 DXS1200–DXS1684 DXS1187–DXS1073

141.8–151.9 145.3–149.6 130.6–153.2

Complete cessation of ovarian function Complete cessation of ovarian function Intermittent follicle maturation failure Complete cessation of ovarian function Complete cessation of ovarian function and early menopause Complete cessation of ovarian function Complete cessation of ovarian function Intermittent follicle maturation failure and early menopause

Rossetti et al. (2004) Eggermann et al. (2005) This report

a

Markers delimiting the X monosomy. The order of markers and physical distances were obtained from the X chromosome map from UCSC genome browser (http://www.ucsc.edu). ND = not determined. b

with different X haplotypes could instead explain the variable penetrance of the Xq26.2–Xq28 deletion. If so, both XCI and POF result from an underlying mutation that is selected against random XCI of ovarian development genes. Discussion Pedigrees with EM and POF often segregate with Xq monosomy. They are classified as the result of a variable expression of the same genetic disease causing oocyte maturation failure (Vegetti et al., 2000). In an Italian family showing both EM

and POF conditions, we established the inheritance of an identical Xq26.2–q28 deletion in association with different XCI haplotypes. By comparing classical cytogenetic studies with the QF-PCR heterozygosity mapping method, we ascertained a hemizygous fragment with a single peak for the contiguous markers DXS1187–DXS1073, and a double peak for the flanking STR DXS1047 and DXS1108. The QF-PCR approach was successfully applied to narrow down the karyotyped Xq deletion using new polymorphic markers developed ad hoc, which could also facilitate the identification of thePOF1 gene, as well as the fine mapping of other Xq28 anomalies. In this way, we 533

G.Fimiani et al.

DXS8043

143,63

DXS8043

DXS8028

143,84

DXS8028

DXS1200

145,35

DXS1200

found that the two ovarian dysfunctions, EM and POF, are associated with an identical interstitial deletion of 22 Mb of DNA located in the POF1 locus, corroborating the genetic role that this portion of the long arm of the X chromosome could play in human reproduction (Figure 4; Marozzi et al., 2000). The EM/POF deletion is rather large and can be tentatively restricted to 5 Mb considering the previously published findings of Eggermann et al. (2005) that described a POF deletion located in the DXS1200–DXS1684 interval which is the minimal critical region so far described. Although this locus contains several putative candidate genes, the genetic mechanism that produces ovarian failure is still controversial. One of the mechanisms might involve a gene dosage effect, such as the lack of expression of genes normally escaping X inactivation (haploinsufficiency). Alternatively, we hypothesize that an X chromosome carrying deletions determines mispairing at meiosis (Schlessinger et al., 2002). In this respect, the fine mapping of the inherited Xq deletion in POF families might be informative in the definition of the minimal region required for such a proposed mispairing mechanism. Furthermore, it is noteworthy that the identification of a chromosome abnormality within an individual may segregate in other family members who have the same defect. One possible explanation for the differences in the penetrance of the Xq26.2– Xq28 deletion in EM/POF conditions could result from a different inactivation status of the X chromosome. Skewed XCI is frequently observed in women with POF (Sato et al., 2004) and the abnormal X is frequently inactivated in carriers of X-linked deletions (Leppig and Disteche, 2001). The abnormalities are generally well tolerated because of the preferential inactivation of the abnormal X, which can restore, at least in part, a balanced genetic make-up. However, the description of some abnormal phenotypes can be ascribed to failed or partial X inactivation and/or incomplete selection in favour of cells with a normal balance of gene expression. Our study confirms and extends these considerations. In fact, the observation that the EM mother and the POF daughter carrying the Xq26.2–Xq28 deletion are skewed for different X haplotypes, suggests that a direct role for epigenetic factors in determining dissimilar ovarian dysfunction penetrance exists. Alternatively, we cannot exclude the epigenetic contribution of modifying loci unlinked to the Xq to the female fertility. Thus, our findings highlight the pivotal role that the proximal Xq28 gene content could play in determining human menopausal age, and also underlines that other factors, such as epigenetic modifications, may affect ovulation.

Xcen

SLITRK2

FMR1

DXS8091

147,20

DXS8069

149,20

DXS1684

149,69

DXS8061

151,59

DXS15

151,99

DXS1073

153,29

DXS1108

154,32

DXS8091 FMR2 IDS DXS8069 MTM1 MTMR1 CD99L2 DXS1684 HMGB3 GPR50 HMGB3 TMG3 FATE CNGA2 MAGEA4 GABRE MAGEA5 MAGEA10 GABRA5 GABRQ MAGEA6 TRG3 MAGEA2 MAGEA12 MAGEA2 TRG3 MAGEA3 CETN2 NSDL DXS8061 ZNF185 PNMA5 Q8N624 MAGEA1 ZNF275 DXS15 TREX2 BGN ATP2B3 DUSP9 SLC6AB BCAP31 ABCD1 PLXNB3 STK23 IDH3G SSR4 L1CAM AVPR2 ARHGAP4 ARD1 RENBP HCFC1 IRAK1 MECP2 OPN1LW OPN1MW TKTL1 FLNA EMD DNASE1L1 TAZ ATP6AP1 GDI1 XAP5 PLXNA3 UBL4 SLC10A3 FAM3A G6PD NEMO Q8IWN6 CTAG1 DXS1073 CTAG1 CTAG2 GAB3 DKC1 MPP1 MTCP1 C61A VBP1 RAB39B CLIC2 F8A TMLHE DXS1108

Acknowledgments

Xtel

q27

q28

Figure 4. Evaluation of POF1 deletions. At the top the transcription map in the Xq27–Xq28 bands is shown according to the GenomeBrowser at University of California, Santa Cruz (http://genome.ucsc.edu/). The genomic map is not to scale. At the bottom the POF1 critical deletion region is shown with the location of the deleted (white circle) and the non-deleted (black circle) markers in relation to the POF monosomy reported by Eggermann et al. (2005), and to the POF/EM deletion described in this study.

534

We thank the family for its cooperation and the Gene Expression Core Laboratory of BioGeM for the technical support in the microsatellite electrophoresis. This project was supported by a FIRB grant from MIUR-Italy.

References Anasti JN (1998) Premature ovarian failure: an update. Fertil Steril 70,1–15. Bione S, Rizzolio F, Sala C, Ricotti R, Goegan M, Manzini MC, Battaglia R, Marozzi A, Vegetti W, Dalpra L et al (2004) Mutation analysis of two candidate genes for premature ovarian failure, DACH2 and POF1B. Hum Reprod 19,2759–2766.

Xq26.2–q28 deletion mapping in EM/POF family Davison RM, Quilter CR, Webb J, Murray A, Fisher AM, Valentine A, Serhal P and Conway GS (1998) A familial case of X chromosome deletion ascertained by cytogenetic screening of women with premature ovarian failure. Hum Reprod 13,3039–3041. Donaghue C, Roberts A, Mann K and Ogilvie C M (2003) Development and targeted application of a rapid QF-PCR test for sex chromosome imbalance. Prenat Diagn 23,201–210. Eggermann T, Meschede D, Schuler H, Palm S, Glaser D, Horsthemke B, Eggermann K, Haverkamp F and Zerres K (2005) Premature ovarian failure associated with a small terminal Xq deletion: narrowing the POF1 region down to Xq27.2/Xq27.3-qter. Clin Genet 67,434–437. Fusco F, Bardaro T, Fimiani G, Mercadante V, Miano MG, Falco G, Israel A, Courtois G, D’Urso M and Ursini MV (2004) Molecular analysis of the genetic defect in a large cohort of IP patients and identification of novel NEMO mutations interfering with NF-kappaB activation. Hum Mol Genet 15,1763–1773. Lemieux N, Drouin R, Richer CL (1990) High-resolution dynamic and morphological G-bandings (GBG and GTG): a comparative study. Hum Genet 85,261–266. Leppig KA and Disteche CM (2001) Ring X and other structural X chromosome abnormalities: X inactivation and phenotype. Semin Reprod Med 19,147–157. Maraschio P, Tupler R, Barbierato L, Dainotti E, Larizza D, Bernardi F, Hoeller H, Garau A and Tiepolo L (1996) An analysis of Xq deletions. Hum Genet 97,375–381. Marozzi A, Manfredini E, Tibiletti MG, Furlan D, Villa N, Vegetti W, Crosignani PG, Ginelli E, Meneveri R and Dalpra L (2000) Molecular definition of Xq common-deleted region in patients affected by premature ovarian failure. Hum Genet 107,304–311. Miller SA, Dykes DD and Polesky HF (1988) A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 16,1215.

Murray A, Ennis S, MacSwiney F, Webb J and Morton NE (2000) Reproductive and menstrual history of females with fragile X expansions. Eur J Hum Genet 8,247–252. Prueitt RL, Chen H, Barnes RI and Zinn AR (2002) Most X;autosome translocations associated with premature ovarian failure do not interrupt X-linked genes. Cytogenet Genome Res 97,32–38. Rossetti F, Rizzolio F, Pramparo T, Sala C, Bione S, Bernardi F, Goegan M, Zuffardi O and Toniolo D (2004) A susceptibility gene for premature ovarian failure (POF) maps to proximal Xq28. Eur J Hum Genet 12,829–834. Sala C, Arrigo G, Torri G, Martinazzi F, Riva P, Larizza L, Philippe C, Jonveaux P, Sloan F, Labella T and Toniolo D (1997) Eleven X chromosome breakpoints associated with premature ovarian failure (POF) map to a 15Mb YAC contig spanning Xq21. Genomics 15,123–131. Sato K, Uehara S, Hashiyada M, Nabeshima H, Sugawara J, Terada Y, Yaegashi N and Okamura K (2004) Genetic significance of skewed Xchromosome inactivation in premature ovarian failure. Am J Med Genet A 15,240–244. Schlessinger D, Herrera L, Crisponi L, Mumm S, Percesepe A, Pellegrini M, Pilia G and Forabosco A (2002) Genes and translocations involved in POF. Am J Med Genet 15,328–333. Sullivan AK, Marcus M, Epstein MP, Allen EG, Anido AE, Paquin JJ, YadavShah M and Sherman SL (2005) Association of FMR1 repeat size with ovarian dysfunction. Hum Reprod 20,402–412. Vegetti W, Marozzi A, Manfredini E, Testa G, Alagna F, Nicolosi A, Caliari I, Taborelli M, Tibiletti MG, Dalpra L and Crosignani PG (2000) Premature ovarian failure. Mol Cell Endocrinol 30,53–57. Zinn AR (2001) The X chromosome and the ovary. J Soc Gynecol Invest 8,S34–36. Submitted on July 28, 2005; accepted on September 26, 2005

535