Mol Diagn 2005; 9 (3): 157-162ORIGINAL RESEARCH ARTICLE 1084-8592/05/0003-0157/$34.95/0

© 2005 Adis Data Information BV. All rights reserved.

Skewed X Inactivation of the Normal Allele inFully Mutated Female Carriers Determines theLevels of FMRP in Blood and the FragileX PhenotypeRaquel Martinez,1,2 Victoria Bonilla-Henao,1,2 Antonio Jimenez,1,2 Miguel Lucas,1,2 Carmen Vega,3

Inmaculada Ramos,3 Francisco Sobrino1 and Elizabeth Pintado1,2

1 Departamento de Bioquimica Medica y Biologia Molecular, Facultad de Medicina, Universidad de Sevilla, Seville, Spain2 Servicio de Biologia Molecular, Hospital Universitario Virgen Macarena, Seville, Spain3 Departamento de Pediatria, Hospital Universitario Virgen Macarena, Seville, Spain

Background: The variable phenotype in female carriers of a full mutation is explained in part by non-randomAbstractX-chromosome inactivation. The molecular diagnosis of fragile X syndrome is based on the resolution of thenumber of CGG triplet repeats and the methylation status of a critical CpG in the fragile X mental retardationgene (FMR1) promoter. Neighboring CpGs in the FMR1 promoter are supposed to be equally methylated orunmethylated.Method: Southern blot analysis was performed with double digestion, either with EcoRI/EagI or with HindIII/SacII. The EagI restriction site was studied by sequencing. The fragile X encoded protein (FMRP) was detectedin white blood cells by Western blot. The fragile X phenotype was evaluated by specific clinical examinations.Results: Within one family we found three female carriers of a full mutation and a different degree ofmethylation of the normal allele that correlated with the levels of FMRP in blood and the fragile X phenotype.Complete methylation at the EagI CpG target (but partially methylated SacII CpG site) was associated withextremely skewed X inactivation (confirmed by analysis of the methylation status at the PGK locus), undetect-able FMRP in blood, and a male-like phenotype.Conclusions: In fully mutated female carriers the methylation status at the EagI restriction site correlates withthe levels of FMRP in blood and the fragile X phenotype. Neighboring CpG sequences in the FMR1 promotercan be differentially methylated, which should be taken into consideration for molecular diagnosis.

Fragile X syndrome is the most common cause of hereditary analysis with double digestion using a methylation-sensitive en-zyme, cutting at the specific CpG in the promoter region.[8,9]mental retardation, with an estimated incidence of 1 in 4000 males

and 1 in 6000 females.[1,2] The molecular basis of fragile X Fragile X syndrome as an X-linked disease is manifested differ-syndrome in nearly all cases is an expansion of the CGG triplet ently in males than in females. Fragile X boys present, in general,repeats within the 5′ untranslated region of the fragile X mental mild to moderate mental retardation, although there are numerousretardation gene (FMR1). Expansion beyond the approximately reports of male subjects with learning disabilities but not retarda-200 repeats (normal range 50–54 repeats) usually results in tion. This milder phenotype appears to be related to FMRP pro-hypermethylation of the FMR1 promoter region, with a concomi- duction arising from transcription of unmethylated alleles.[9,10] Thetant loss of FMR1 transcription and the absence of the encoded spectrum of phenotypes in female carriers with a full mutation isprotein (FMRP), an RNA-binding protein.[3-7] Molecular diagnosis very wide, ranging from mental retardation with almost all of theof fragile X syndrome is generally performed by Southern blot physical features described in fragile X males to subtle learning

158 Martinez et al.

II-P

a

c

b

II-P II-1

I-1

II-2

III-1

III-1

II-1 II-2 I-1

III-1

FMRP 80 kD

Ponceau red

II-1 II-2 C

5.2 kb

2.8 kb

Fig. 1. DNA analysis, levels of the fragile X encoded protein (FMRP) in blood, and the pedigree of the fragile X family studied. (a) Southern blot of thefragile X mental retardation (FMR1) gene using EcoRI and EagI restriction enzymes. DNA from peripheral white blood cells was digested to completionwith the indicated enzymes and hybridized with the probe StB12.3. Individual II-P is a non-fragile X male exhibiting a 2.8 kb fragment. Patients III-1, II-1,and II-2 are females carrying a full mutation, and patient I-1 is a premutated female carrier. The normal allele of patient III-1 inherited from her father wascompletely methylated. (b) Pedigree of the family under study. The gray tone used for patient II-2 indicates a milder phenotype than that observed inpatient III-1. (c) Western blot analysis of protein extracts from white blood cells shows that FMRP was undetectable in patient III-1, whereas patient II-1expressed almost normal levels and patient II-2 significantly lower levels than her sister (II-1). C indicates the FMRP level of a T-cell line used as a control.Ponceau red shows equal protein loading. Southern and Western blottings were repeated three times, obtaining similar results.

disabilities without any remarkable physical involvement.[11] The Methodsmild phenotype observed among females with the full mutation is

explained in part by the presence of two chromosomes, only one of Patients

which is active in each cell.[12] Although it is commonly believedThe pedigree of the family in this study is shown in figure 1b.

that the initiation of X inactivation is random, a severe phenotype The proband, patient III-1, was born with a birthweight of 3000gfor several recessive X-linked diseases has been observed in by normal delivery. At the age of 12 months she was referred for

delayed psychomotor development, as a result of which she fol-females with extremely skewed X inactivation of the normallowed a stimulation program. At the age of 18 months she wasallele.[13-16] In addition, it has been recently reported that thesubmitted for fragile X molecular testing based on her clinical and

severity of clinical signs of the fragile X tremor/ataxia syndromefamily history. Molecular analysis was performed on all available

(FXTAS) in premutated carrier sisters correlated with the molecu- members of the family. Complete examination of their physicallar pattern of X inactivation favoring higher expression of the and psychological characteristics was also undertaken, including

an ophthalmologic and otorhinolaryngologic report and a study ofpremutation allele.[17]

their cardiac function. This research was performed with protocolsIn this study we report the correlation between the degree ofapproved by the local institutional review board and with the

inactivation of the non-fragile X chromosome, the FMRP level ininformed consent of the participants and/or their guardians.

blood, and the fragile X phenotype in three female carriers with

full mutation. In addition, we show that the EagI and the SacII Southern Analysis

restriction sites of the FMR1 promoter can be differentially methy-DNA was extracted from blood according to standard proce-

lated, illustrating the potential for misinterpretation of molecular dures. 10μg of genomic DNA was double digested with thediagnostic tests for fragile X syndrome when using SacII restric- restriction enzymes EcoRI and EagI at 37°C overnight. The EagItion enzyme in Southern blot analysis. is an enzyme sensitive to methylation. The restriction fragments

© 2005 Adis Data Information BV. All rights reserved. Mol Diagn 2005; 9 (3)

Skewed X Inactivation and FMRP Levels in Fragile X Syndrome 159

were separated by electrophoresis on 0.8% agarose gel and hybrid- trifuged for 60 minutes at 100 000g to obtain the ribosome pro-ized with the radiolabel StB12.3 probe as previously described.[18] teins. The pellet was resuspended in a hypertonic buffer containingThe same protocol was followed, but with the restriction enzymes 20 mmol/L Hepes pH 7.9, 10 mmol/L KCl, 420 mmol/L NaCl,HindIII and SacII, the latter being sensitive to methylation. The 20% glycerol, 1 mmol/L EDTA, 1 mmol/L DTT, 0.2% NP40, andpercentage of methylation of the normal allele was determined by the protease inhibitors mentioned above. The protein concentra-scanning densitometry analysis of the bands using Scion Image tion was determined with a Bradford protein assay reagent.software (Frederick, MD, USA). The location of the restriction site 20–40μg of the lysates were loaded and resolved on 7.5% SDS-of each enzyme in the FMR1 promoter is represented by the PAGE transferred onto poly-vinylidene difluoride (PVDF) mem-diagram in figure 2a. brane (Amersham Biosciences, Piscataway, NJ, USA) and sub-

jected to immunodetection using a 1 : 5000 dilution of primaryWestern Analysis FMRP antibody (Chemicon®, Temecula, CA, USA) and a

1 : 10 000 dilution for the secondary mouse antibody (Promega,FMRP expression in white blood cells was analyzed as previ- Barcelona, Spain) and enhanced with a chemiluminescence detec-

ously described.[19] Cells were homogenized in a cold lysis hypo- tion system (Amersham Biosciences, Barcelona, Spain).tonic buffer containing 20 mmol/L Hepes pH 7.9, 10 mmol/L KCl,10% glycerol, 1 mmol/L EDTA, 1 mmol/L DTT, 0.2% NP40, X Inactivation Analysisprotease inhibitors (0.01% PMSF, 1 mmol/L Na3VO4, 10 mg/Laprotinin, and 10 mg/L leupeptin), and 1.5 mg/L RNase-DNase To determine if the observed methylation patterns of the FMR1free for 30 minutes at 4ºC. The homogenate was centrifuged for 5 gene were representative of skewed X inactivation, we performedminutes at 13 000 rpm and the supernatant recollected and cen- additional studies at the androgen receptor (AR) and phosphoglyc-

NruIEagl

BssHIISacII

CpGIsland

II-P

5.2 kb

2.7 kb

III-1 II-1

Exon 1EcoRI/HindIII

a

b

EcoRI/HindIII

(CGG)n StB12.3

Fig. 2. (a) Diagram of the fragile X mental retardation (FMR1) gene and neighboring regions and location of different restriction enzymes used formolecular diagnosis of fragile X syndrome.[9] (b) Southern blot using HindIII and SacII restriction enzymes. Non-fragile X individuals will show 2.7 kb and5.2 kb fragments in females and a 2.7 kb fragment in males. Individual II-P had a 2.7 kb fragment. Patient III-1 had methylated 70% of the normal allele andpatient II-1 had an almost completely unmethylated normal allele. Similar results were obtained in three different blots.

© 2005 Adis Data Information BV. All rights reserved. Mol Diagn 2005; 9 (3)

160 Martinez et al.

erate kinase (PGK) loci.[20,21] The AR locus on the X chromosome Molecular analysis of the FMR1 gene showed that patient III-1contains two methylation-sensitive HpaII restriction sites and is had a full mutation inherited from her mother (II-1) and complete90% polymorphic in Caucasian females for varying allele sizes. methylation (100%) of the normal allele inherited from her fatherHowever, the females in our study were homozygous and there- (II-P). The molecular analysis of the rest of the family indicatedfore uninformative. For the PGK locus, samples were processed that the maternal grandmother (I-1) had a premutated allele thatfor preliminary determination of the PGK BstXI polymorphism. expanded when it was transmitted to both of her daughters (II-1Genomic DNA (0.1μg) was subjected to PCR in a total volume of and II-2) [figure 1a]. Subject II-1 had 10% of her blood cells with20μL containing 1 × PCR buffer, including 1.5 mmol/L MgCl2, the normal FMR1 allele methylated, unlike her sister (II-2) who0.2 mmol/L dNTP, 0.5 μmol/L forward primer (5′-AGC TGG showed 50% inactivation of the normal allele. Analysis of theACG TTA AAG GGA AGC-3′), 0.5 μmol/L reverse primer (5′- CGG repeats at the FMR1 gene alleles from the non-fragile XTAG TCC TGA AGT TAA ATC AAC-3′), and 1U DNA chromosome (by PCR amplification) indicates that sisters II-1 andpolymerase. After an initial incubation of the samples for 5 min- II-2 had each inherited 29 triplet repeats from their father. Patientutes at 94ºC, reactions were cycled for 1 minute at 94ºC, 1 minute III-1 had inherited an allele of 28 triplet repeats from her father (II-at 56ºC, and 1 minute at 72ºC for 33 cycles. Aliquots of the PCR P), and her grandmother (I-1) showed an allele of 25 triplet repeatsproducts (5μL) were digested with 30U of the restriction en- (not shown).donuclease BstXI for 12 hours at 37ºC. The resultant DNA frag- FMRP in white blood cells analyzed by Western blot showedments were electrophoresed in 2% agarose gels and stained with that patient III-1, in whom the normal FMR1 allele was totallyethidium bromide. To obtain the methylation status at the PGK inactivated, had no detectable FMRP in her blood, while herlocus, DNA samples (0.2μg) were incubated for 12 hours at 37ºC mother (II-1) [with unmethylated normal FMR1 in most of herin a total volume of 20μL using 30U of methylation-sensitive cells] had FMRP levels slightly lower than a control from a humanrestriction endonuclease HpaII to completely digest non-methylat- T-cell line (figure 1c). By contrast, patient II-2, with 50% of theed (active) alleles. After the predigestion, 5μL of DNA samples normal FMR1 allele methylated, showed a decreased level ofwere amplified by PCR using the same conditions as described FMRP compared with her sister (figure 1c). To determine if theabove. Aliquots of the PCR products (5μL) were digested with observed 100% methylation at the EagI site of the FMR1 geneBstXI, subjected to agarose gel electrophoresis, and stained with promoter in patient III-1 was really representative of extremelyethidium bromide. skewed X inactivation, we performed additional studies at the AR

and at the PGK loci. Although the family was homozygous for theAR locus and therefore uninformative, the PGK locus clearlyResultsshows the extremely skewed X inactivation in patient III-1 (figure3b, c, and d). This further supports that extreme X-chromosomeThe family history showed that patient III-1 had a granduncleinactivation is responsible for the absence of FMRP in blood and(not shown) and an aunt (patient II-2) from the maternal side withthe male-like fragile X phenotype in this patient. The mutation inmild mental retardation. Both parents (individuals II-P and II-1)the X inactivation transcript gene minimal promoter previouslyhad normal cognitive abilities. Patient III-1 had physical traits ofassociated with skewed X-chromosome inactivation in two Euro-fragile X, such as a long narrow face, prominent ears, and hyper-pean families[22] was not present in any member of the family inextensible finger joints. The remarkable joint flexibility remindedthis study (not shown).us of the laxity observed in Marfan syndrome. Patient III-1 also

presented with hand-flapping, short attention span, tactile defen- Southern blot analysis of patient III-1 and her parents was alsosiveness, shyness, and poor eye contact. Standardized tests could performed using the restriction enzymes HindIII and SacII, thenot be used because of the absence of empathy. Patients II-1 and latter being sensitive to methylation and commonly used forII-2 showed a different percentage of methylation of the normal fragile X diagnosis.[9] The location of the restriction site of eachallele, which was also in agreement with their cognitive perform- enzyme in the FMR1 promoter is represented by the diagram inance. Patient II-1 had >90% active non-fragile X allele (figure 1 figure 2a. Surprisingly, the pattern of methylation observed inand figure 2) and normal intelligence, whereas patient II-2 had patient III-1 with these enzymes differed from the one obtained50% active normal allele (figure 1) and presented mild physical with the restriction enzymes EcoRI and EagI (see figure 1a).traits and mental retardation (she currently attends a special pro- Although a clear skewed inactivation of the normal allele (70% ofgram for the mentally impaired). They did not present any ophthal- the normal allele inactivated) was also observed, the SacII restric-mological or otorhinolaryngological problems and an echocardi- tion site was not completely methylated, indicating a differentogram revealed no obvious abnormalities. methylation status at the two CpG sequences of the FMR1 promot-

© 2005 Adis Data Information BV. All rights reserved. Mol Diagn 2005; 9 (3)

Skewed X Inactivation and FMRP Levels in Fragile X Syndrome 161

differential activation/inactivation of the fragile X allele. A severephenotype for several recessive X-linked diseases has been ob-served in females with extremely skewed X inactivation of thenormal allele.[13-16] The marked psychological and physical fragileX phenotype observed in patient III-1, comparable to that ob-served in fragile X boys of the same age, is consistent with theextremely skewed X inactivation and the absence of FMRP inblood. Patients II-1 and II-2 showed a different percentage ofmethylation of the normal allele, which correlates with the levelsof FMRP in their blood and their cognitive performance. Theseresults agree with the hypothesis that in female carriers with fullmutations, the different phenotypes exhibited are mainly due tounequal X-chromosome inactivation.

Conclusion

The molecular diagnosis of fragile X syndrome is based on thedetection of the magnitude of the expansion and the methylationstatus of the FMR1 promoter. Southern blot analysis using methyl-ation-sensitive restriction enzymes, such as EagI or SacII, pro-vides data about both the size of CGG repeats and the methylationof specific sites, considered crucial for the normal expression ofthe FMR1 gene. Although there has been a report of variable CpGmethylation in hypermethylated epigenotypes obtained from asingle individual,[23] it is commonly believed that the neighboringCpGs at the EagI and SacII restriction sites are equally methylatedor unmethylated.[9] The fact that FMRP was undetected in theblood of patient III-1 suggests a more critical role of the CpGmethylation status at the EagI than at the SacII restriction site inorder to allow FMR1 expression. Therefore, we recommend usingthe EagI methylation-sensitive restriction enzyme for moleculardiagnosis of fragile X syndrome.

Acknowledgment

AfaI

II-P

ND

530 bp

a

b

c

d

433 bp

530 bp

433 bp

285 bp AR

D/BstXI

D/HpaII

ND and D/BstXI

D ND D ND D

ND D ND D ND D

III-1 II-2

II-P

II-P III-1 II-2

III-1 II-2

HpaII HpaII

433530

BstXI

Exon 1

Fig. 3. X-chromosome inactivation analysis. (a) Schematic diagram show-ing endonuclease sites, locations of PCR primers, and sizes of the ampli-fied and digested fragments of the PGK gene.[21] (b) PCR of the PGK locusfrom subjects II-P, III-1, and II-2 shows the expected 530 bp product (non-digested [ND]). Digestion with BstXI after PCR of the PGK locus indicatesthat subject II-P had an allele of 530 bp, patient II-2 was homozygous forthe 433 bp allele, and patient III-1 was heterozygous and therefore bothalleles – the 530 bp inherited from her father and the 433 bp inherited fromher mother – were present. (c) Samples were either digested (D) or NDwith the methylation-sensitive restriction enzyme HpaII. After PCR of thePGK locus, all samples (D and ND) were digested with the restrictionenzyme BstXI. In patient III-1 there is a complete disappearance of thematernal 433 bp allele in the sample previously digested with HpaII. Asexpected, the 530 bp allele of subject II-P also disappeared in the HpaIIdigested sample. Patient II-2 was homozygous for this locus and nochange was observed after HpaII digestion. (d) PCR from the androgenreceptor (AR) locus shows that the three members of the family studiedwere homozygous and therefore uninformative. These experiments wererepeated three times, obtaining similar results.

We would like to express our gratitude to Dr J. Lopez-Barneo for helpfuler. The proper activity of SacII was confirmed, since complete advice during the preparation of this manuscript. We thank Dr J.L. Mandel for

the StB12.3 probe. This work was supported by grants from Servicio Andaluzdigestion was observed in the DNA samples from the patient’sde Salud Exp:80/03 (to E. Pintado) and Exp:22/03 (to M. Lucas). R. Martinez,parents. We excluded a point mutation or a micro deletion at theV. Bonilla-Henao and A. Jimenez were supported by fellowships from

EagI restriction site by digestion after PCR of the FMR1 promoter Servicio Andaluz de Salud, Seville, Spain.and by sequencing (not shown).

ReferencesDiscussion1. Bell MV, Hirst MC, Nakahori Y, et al. Physical mapping across the fragile X:

hypermethylation and clinical expression of the fragile X syndrome. Cell 1991;The fragile X phenotype, in both males or females, is supposed64: 861-6

to be dependent on the methylation status of the FMR1 promot- 2. O’Donnell WT, Warren ST. A decade of molecular studies of fragile X syndrome.Annu Rev Neurosci 2002; 25: 315-38er.[9,10] In female carriers of a full mutation, the fragile X pheno-

3. Fu YH, Kuhl DP, Pizzuti A, et al. Variation of the CGG repeat at the fragile X sitetype comprises a wide spectrum (from normal to full fragile X results in genetic instability: resolution of the Sherman paradox. Cell 1991; 67:characteristics), and the most accepted explanation for this is the 1047-58

© 2005 Adis Data Information BV. All rights reserved. Mol Diagn 2005; 9 (3)

162 Martinez et al.

4. Siomi H, Siomi MC, Nussbaum RL, et al. The protein product of the fragile X 16. Heine-Suner D, Torres-Juan L, Morla M, et al. Fragile-X syndrome and skewed X-gene, FMR1, has characteristics of an RNA-binding protein. Cell 1993; 74: chromosome inactivation within a family: a female member with complete291-8 inactivation of the functional X chromosome. Am J Med Genet 2003; 122A:

5. Feng Y, Gutekunst CA, Eberhart DE, et al. Fragile X mental retardation protein: 108-14nucleocytoplasmic shuttling and association with somatodendritic ribosomes.

17. Berry-Kravis E, Potanos K, Weinberg D, et al. Fragile X-associated tremor/ataxiaJ Neurosci 1997; 17: 1539-47syndrome in sisters related to X-inactivation. Ann Neurol 2005; 57: 144-76. Jin P, Warren ST. Understanding the molecular basis of fragile X syndrome. Hum

18. Hallmayer J, Pintado E, Lotspeich L, et al. Molecular analysis and test of linkageMol Genet 2000; 9: 901-8between the FMR-1 gene and infantile autism in multiplex families. Am J Hum7. Weiler IJ, Spangler CC, Klintsova AY, et al. Fragile X mental retardation protein is

necessary for neurotransmitter-activated protein translation at synapses. Proc Genet 1994; 55: 951-9Natl Acad Sci U S A 2004; 101: 17504-9 19. Bonilla V, Sobrino F, Lucas M, et al. Epstein-Barr virus transformation of human

8. Rousseau F, Heitz D, Biancalana V, et al. Direct diagnosis by DNA analysis of the lymphoblastoid cells from patients with fragile X syndrome induces variablefragile X syndrome of mental retardation. N Engl J Med 1991; 325: 1673-81

changes on CGG repeats size and promoter methylation. Mol Diagn 2003; 7:9. Tarleton J, Kenneson A, Taylor AK, et al. A single base alteration in the CGG

163-7repeat region of FMR1: possible effects on gene expression and phenotype.

20. La Spada AR, Wilson EM, Lubahn DB, et al. Androgen receptor gene mutations inJ Med Genet 2002; 39: 196-200X-linked spinal and bulbar muscular atrophy. Nature 1991; 352: 77-910. Hagerman RJ, Hull CE, Safanda JF, et al. High functioning fragile X males:

demonstration of an unmethylated fully expanded FMR-1 mutation associated 21. Uehara S, Tamura M, Nata M, et al. X-chromosome inactivation in the humanwith protein expression. Am J Med Genet 1994; 51: 298-308 trophoblast of early pregnancy. J Hum Genet 2000; 45: 119-26

11. Hagerman RJ. Physical and behavioural phenotype In: Hagerman RJ, Cronister22. Plenge RM, Hendrich BD, Schwartz C, et al. A promoter mutation in the XIST

MS, editors. Fragile X syndrome. Baltimore (MD): The Johns Hopkins Univer-gene in two unrelated families with skewed X-chromosome inactivation. Natsity Press, 1996: 3-87Genet (1997); 17: 353-35612. Abrams MT, Reiss AL, Freund LS, et al. Molecular-neurobehavioral associations

in females with the fragile X full mutation. Am J Med Genet 1994; 51: 317-27 23. Reinard S, Kajimura TM, Brown TW, et al. Epigenetic variation illustrated by

DNA methylation patterns of the fragile X gene FMR1. Hum Mol Genet 1997;13. Chan V, Chan VW, Yip B, et al. Hemophilia B in a female carrier due to skewedinactivation of the normal X-chromosome. Am J Hematol 1998; 58: 72-6 6: 1791-801

14. Plenge RM, Stevenson RA, Lubs HA, et al. Skewed X-chromosome inactivation isa common feature of X-linked mental retardation disorders. Am J Hum Genet2002; 71: 168-73 Correspondence and offprints: Dr Elizabeth Pintado, Departamento de Bio-

15. Anderson-Cohen M, Holland SM, Kuhns DB, et al. Severe phenotype of chronic quimica Medica y Biologia Molecular, Facultad de Medicina, Avda.granulomatous disease presenting in a female with a de novo mutation in

Sanchez Pizjuan, Sevilla, 4, E-41009, Spain.gp91-phox and a non familial, extremely skewed X chromosome inactivation.Clin Immunol 2003; 109: 308-17 E-mail: elizabet@us.es

© 2005 Adis Data Information BV. All rights reserved. Mol Diagn 2005; 9 (3)