FRAXE and mental retardation

16Med Genet 1995;32:162-169

Original articles

FRAXE and mental retardation

Centre for MedicalGenetics, Cytogeneticsand MolecularGenetics, Women'sand Children'sHospital, Adelaide5006, AustraliaJ C MulleyS YuA DonnellyA K GedeonP W L YuE BakerR I RichardsG R Sutherland

Department ofGenetics, University ofAdelaide, AustraliaA DonnellyA K Gedeon

Department ofGenetics andEpidemiology,Women's andChildren's Hospital,Adelaide, AustraliaE A Haan

Department ofPsychology, La TrobeUniversity, Melbourne3083, AustraliaD Z LoeschD A Hay

NH & MRCSchizophreniaResearch Unit, MentalHealth ResearchInstitute of Victoria,AustraliaD A Hay

Unidad de GeneticaHumana, Centro deBioquimica y GeneticaClinica, ConjuntoResidencial, Apartado61, 30100 Espinardo,Murcia, SpainP CarbonellI L6pezG GloverI Gabarr6n

Genetics Service,Princess MargaretHospital for Childrenand Irrabeena, 53 OrdStreet, Perth 6005,AustraliaA Hockey

Molecular GeneticsGroup, Institute ofMolecular Medicine,John RadcliffeHospital, Oxford, OX39DU, UKS J L KnightK E Davies

Correspondence to:Dr Mulley.Received 26 July 1994Revised version accepted forpublication 28 October 1994

J C Mulley, S Yu, D Z Loesch, D A Hay, A Donnelly, A K Gedeon, P Carbonell,I Lopez, G Glover, I Gabarron, P W L Yu, E Baker, E A Haan, A Hockey,S J L Knight, K E Davies, R I Richards, G R Sutherland

AbstractMental impairment and instability of theCCG repeat at FRAXE is described in sixkindreds. Cosegregation of FRAXA andFRAXE was found within one of thesekindreds. Cytogenetic expression ofFRAXE was shown to skip a generationwhen associated with a reduction in sizeof the CCG expansion when transmittedthrough a male; however, in general,transmission occurred through femalesand copy number increased from one gen-eration to the next. In these respects thebehaviour of FRAXE paralleled that ofFRAXA. A relationship between FRAXEand non-specific mental impairment isstrongly suggested by the occurrence inthese families of more mentally impairedmale and female carriers, after removalof index cases, than could reasonably beexpected by chance.

(Jt Med Genet 1995;32:162-169)

Cytogenetic detection of the fragile X(FRAXA) has been confused by the discoveryof two nearby rare folate sensitive fragile sites.These were designated FRAXE12 and FRAXF.'FRAXE sequences have been isolated ap-proximately 600 kb distal to FRAXA4 andFRAXF sequences have been isolated furtherdistal.5 These three fragile sites cannot bedifferentiated by conventional cytogenetic ana-lysis but were initially delineated by fluor-escence in situ hybridisation (FISH) usingDNA probes which separate them. Direct mo-lecular analysis of specific restriction fragmentsnow enables definitive diagnosis by Southernanalysis for all three fragile sites.The fragile sites FRAXE and FRAXF have

recently been shown to have cytogenetic andmolecular features in common withFRAXA."'6 The mutations responsible for thecytogenetic expression of FRAXA, FRAXE,and FRAXF are amplifications of (CCG)n tri-nucleotide repeats; in the case of FRAXA, theCCG repeat is adjacent to the CpG island atthe 5' end of the FMR1 gene.7 AmplificationofCCG to a critical size atFRAXA andFRAXEhas been shown to be associated with hy-permethylation at nearby CpG islands. The

FMR1 gene is usually silenced (presumably byhypermethylation of the CpG island) when thetrinucleotide copy number exceeds ap-proximately 230 copies,8 causing fragile X syn-drome. The findings of Knight et al49 suggestthat a similar mechanism may inactivate a genenear the CpG island associated with FRAXEand thereby cause mental retardation.To investigate the relationship between

FRAXE and mental retardation we describethe phenotypic, cytogenetic, and molecular fea-tures in six FRAXE families collected fromfour centres. One of these families was as-certained for FRAXE in an unbiased fashion asFRAXE was found accidentally during familyfollow up after diagnosis of fragile X (FRAXA)syndrome in the proband. The mode of as-certainment for the remaining FRAXE familieswas by cytogenetic detection and molecularconfirmation ofFRAXE in a mentally impairedfamily member. Detailed phenotypic in-vestigations and molecular delineation of car-riers of FRAXA and FRAXE was carried outin the family segregating with both fragile sites,FRAXE and FRAXA.

Methods and clinical dataDELINEATION OF FRAGILE SITES BY FISHThe fragile sites in the first four families weredelineated as FRAXE by FISH before the mo-lecular characterisation of FRAXE. This wasachieved using methods as in Sutherland andBaker. '

LOCALISATION OF DXS1123 BY FISHThe marker DXS 1 123 is closely linked(1 2 cM) and distal to FRAXA'° and musttherefore be closely linked to FRAXE. FISHwas carried out using the lambda clone DNAof DXS 1 123 using methods previously de-scribed1 in order to orientate DXS 1123 phys-ically in relation to FRAXE. FISH establishedthe physical location of the AC dinucleotiderepeat marker DXS 1 123 to be distal toFRAXE. Thus, the closest flanking markers forindirectly tracking the segregation of FRAXEthrough families are the AC dinucleotide repeatmarkers FRAXAC 1 and FRAXAC2 within theFMR1 locus proximally and DXS 1 123 distally.

162

group.bmj.com on July 14, 2011 - Published by jmg.bmj.comDownloaded from

http://jmg.bmj.com/

http://group.bmj.com/

163FRAXE and mental retardation

FRAXA and FRAXE

1 2

/5° 26% 0%1-2-1;81 0% NT

1 2 3 411 -Il

0% 3%0-3NT 0.2NT

111 z 1 *2 3 6

0% 0% 20% 24% 0%,3% 1%,3%0.580%o 0 490o/ 0 0% 0.050% 0.050%o 0.7100%o

Family 5

FRAXE1 2

0% 0%0.0 0.370%

612%

0.8100%.X 30%

1.0100%

FRAXE2

11

5III

39% 0% 9%1.5-2.1'0 0-0 0.7, 15-1.9NT

j1 (® Obligate carrier for FRAXE

* * Intellectual disability

NT Not tested

Families studied showing percentage of cytogenetic expression offragile sites (two values indicate repeated testing) above the a in kb for FRAXE. Theestimated degree of methylation as a percentage, for the FRAXE associated CpG island only, is given as a superscript to the FRAXE A value. Thefragile site expression in family 1 relates to FRAXE for L and III. 6, FRAXE or FRAXA for H.13 and FRAXA for III.3, III. 4, and III.5. The avalues for FRAXA for the relevant family members are given separately in the text. Families 2 to 6 have only FRAXE.

CYTOGENETIC ANALYSIS

The fragile site FRAXE was induced fromlymphocyte cultures as previously described.'The percentage of cells expressing the fragilesite in distal Xq is shown in the figure.

SOUTHERN ANALYSISDirect molecular diagnosis of the amp-

lifications at FRAXA and FRAXE was carriedout by standard Southem analysis as previouslydescribed using the probe pfxa3 specific forFRAXA" and the probe OxE20 specific forFRAXE.4 The FRAXE CCG amplification (A)was determined from the HindIlI fragment

(normal size 5-2 kb). Assessment of methyl-ation status of the CpG island adjacent toFRAXE was initially attempted by HindIII/SacII double digestion (normal fragment size18 kb). However, an additional SacII re-

striction site omitted from the original re-

striction map4 occurs immediately proximal tothe OxE20 probe between the OxE20 sequenceand the CCG repeat. This restriction site pre-vented detection of incremental fragmentswhen unmethylated (where there were smallCCG amplifications). Methylation was re-

assessed by HindIIIINotI double digestion (nor-mal fragment size 2-8 kb). Methylation statusof the CpG island adjacent to FRAXE for

Family 1 Family 2 FRAXE

IV

Family 3

11

21%08-2 *8100"Y'

Family 4 FRAXE

11

III

Family 6

11

III

1111

FRAXE1 2


http://jmg.bmj.com/


164

carrier males and carrier females was estimatedsimply as the relative intensity of the expandedHindIII fragment and the expanded HindIIVNotI fragment. Absence of an expandedHindIII/NotI fragment in a FRAXE carrier in-dicated complete methylation. Blots ofFRAXEcarriers were then reprobed with pfxa3, as a

control probe, to exclude incomplete digestionas an explanation for apparent methlyation ofthe FRAXE CpG island. Presence of a singleHindIII/NotI fragment of 2x4 kb excluded thepossibility of partial digestion. Reprobing theoriginal HindIII/SacII blots onFRAXE subjectswith pfxa3 gave normal fragments of size 2-4 kbindicating that all FRAXE carriers were un-

methylated across the FRAXA CpG island.The FRAXA amplification in family 1 was

determined from the PstI fragment (normal size1 0 kb) and the methylation status of the CpGisland adjacent to FRAXA was assessed byEcoRI/EagI double digestion (normal size ofthe EcoRI fragment 5 2 kb and normal size ofthe EcoRIPEagI fragment 2-8 kb).

PCR ANALYSISThe trinucleotide repeat polymorphism as-

sociated with FRAXE was genotyped as copynumber.4

TESTING FOR COGNITIVE ABILITIES IN FAMILY 1

A comprehensive battery of testing was ad-ministered to family 1 to measure specific abil-ities" but only data on Peabody PictureVocabulary Test (PPVT), Block Design (BD),Digit Span (DS), and Similarities (Sim) are

reported here. Block Design and Similaritieswere subtests from the Wechsler Pre-Schooland Primary Scale of Intelligence or the BritishAbility Scales, depending on the person's men-tal age. The Digit Span task was that in theWechsler Intelligence Scale for Children-Re-vised. All scores have been scaled to a uniformmean of 100 and standard deviation of 15.PPVT and Block Design have been combinedto give an overall IQ score, using the methoddescribed by Sattler'3 for deriving short formIQs.

ResultsDESCRIPTION OF FAMILIES

The families described were ascertained byreferral of a mentally impaired family memberfor routine cytogenetic analysis, leading to thedetection of a fragile site in distal Xq. Families2 and 3 were excluded as FRAXA familiesby Southern analysis using the probe pfxa3,classified as FRAXE families by FISH, andsubsequently confirmed as FRAXE by directanalysis when the OxE20 probe was isolated.Families 4, 5, and 6 were excluded as FRAXAand classified as FRAXE by direct Southernanalysis using the probes pfxa3 and OxE20. Infamily 1, originally ascertained by the presenceof fragile X (FRAXA) syndrome in a familymember, the cosegregation of FRAXE in ad-dition to FRAXA was inferred when 1.2 ratherthan I. 1 had an amplified fragment detectable

Mulley et al

with the probe pfxa3 which is specific forFRAXA. I.1 only, in this generation, was theone with cytogenetic expression of a fragile site.FISH was then used to classify the fragile sitein I.1 as FRAXE. The pattern of segregationof both FRAXA and FRAXE in family 1 wassubsequently confirmed by direct moleculardiagnosis using the probes pfxa3 for FRAXAand OxE20 for FRAXE. The figure shows theFRAXE amplification with methylation status.The pattern of transmission of instability at theFRAXE locus for the first three pedigrees, asdetermined by Southern analysis, was identicalto that predetermined by the closely linkedPCR markers flanking FRAXE (data notshown). PCR analysis ofFRAXE expressed ascopy number in all pedigrees (data not shown)was consistent with results determined usingthe probe OxE20; however, absence of a PCRproduct as a criterion for diagnosis must beused with caution as it could result from eitherexpansion of the CCG repeat or the failure ofa normal allele to PCR amplify for unknowntechnical reasons. Closely linked flankingmarkers such as FRAXAC1, FRAXAC2, andDXS 1 123, and the FRAXE polymorphismitself, represent a backup to diagnosis in theevent of ambiguity in the interpretation ofSouthern analysis.

FAMILY 1 (FIGURE)The lower functioning subjects III.3, III.4, andIII.5 carry only FRAXA and have fragile Xsyndrome. The somatically heterogeneousFRAXA expansion is represented by amp-lifications between 1x7 and 2-0 kb for III.3, of1 1 and 2-4 kb for III.4, and 0-6 and 1 1 kb forIII.5 These people do not carry FRAXE and areirrelevant for the investigation of an associationbetween FRAXE and mental impairment. II.3carries both FRAXA (A=06) and FRAXE(A=0 2) and, therefore, the relative con-tributions to mental impairment from FRAXEand FRAXA cannot be delineated. II.1 andII.3 both have small FRAXE amplifications(A = 0 3 and 0 2 respectively), but II. 1 performsconsiderably better on cognitive tests than II.3who has both fragile sites (table 1). The mem-bers of this family with mental impairmentassociated with the presence of FRAXE areI.1, III.1, and III.6. III.2 has a small FRAXEamplification (A=0-4 kb), and does not showintellectual deficits.

Family 1 is unique since both fragile sites,FRAXA and FRAXE, are segregating. FRAXEwas transmitted from I.1 who had a smear ofcompletely methylated high molecular weightfragments (A= 1 2-1 8). The CCG expansionwas reduced when transmitted to each daugh-ter. FRAXE was cytogenetically undectable inthe daughter II. 1 who had a A value of 0 3 kb.A fragile site was cytogenetically detectable ata low level in the other daughter, I.3, who hada contraction to a A value of 0-2 kb. Whetherthe fragile site in this daughter was FRAXA orFRAXE could not determined by the con-ventional cytogenetic analysis.There was little expansion when FRAXE was

passed on from II.1, to III.1 and III.2, who


http://jmg.bmj.com/


Table 1 Summary of clinical status and intellectual performance in all members offamily 1

No in pedigree Age at Education and Intellectual status Mental and neurological status Physical (minor) anomaliesand sex examination occupation

PPVT BD DS Sim IQ

1.1 (FRAXE) 65 Never good at school; 63 20 65 50 53 Several hospital treated Large head, prominentMale various unqualified "nervous breakdowns"; takes eyebrows, narrow lower face,

jobs tranquillisers because of large sticking out ears,anxiety; explosive speech Dupuytren contracturewith bad stutter bilaterally

IL.1 (FRAXE) 35 Civil Servant after 92 76 95 103 80 Panic attack in public Unusually short halluxFemale Technical College transport bilaterallyIII.I (FRAXE) 22 Attended special 60 52 55 48 49 Psychiatric diagnosis of Irregular teeth, high arched andFemale school; no occupation schizophrenia confirmed by narrow palate, moderate joint

SCID-R Structured Clinical laxityInterview for DSMIII-R.14Bad stutter, handmannerisms

III.2 (FRAXE) 19 Apprenticeship for 93 116 70 103 104 NAD Moderately high arched palateMale carpentryIII.6 (FRAXE) 10 Difficulties in learning 75 78 95 83 73 NAD Hypertelorism, joint laxity,Female and concentration; unusually short hallux bilaterally

works as salesman'sassistant

II.3 (FRAXE & 37 Difficulties in 90 54 110 93 68 NAD Long chin, dysmorphic ears,FRAXA) Female mathematics during 8 shortened 5th metacarpal bone

years' schooling; on right, unusually short hallux,occasional clerical job enlarged thyroid gland

III.3 (FRAXA) 19 Special school, 45 20 65 27 33* Hand mannerisms, Antimongoloid slant withMale sheltered workshop perseverations, dyslalia hypertelorism, long narrow face,

long bulbous nose, upperprognathism, short philtrum,facial asymmetry, irregular teeth,moderately high arched palate,joint laxity, clinodactyly offingers 2 and 5

III.4 (FRAXA) 17 Learning difficulties, 98 49 85 65 76 Very clumsy movements, Epicanthic folds, long bulbousFemale technical college poor motor coordination; nose, long chin, short philtrum,

(cooking) not extreme short sightedness high narrow palate, joint laxitycompleted; nooccupation

III.5 (FRAXA) 14 Learning difficulties at 89 46 105 81 68 Emotionally unstable, panic Facial asymmetry, high andFemale school; works as a attacks at night narrow palate

waitressI.2 (FRAXA) 61 Occasional cleaning 91 54 80 57 68 Emotionally unstable NAD (palate and teeth notFemale jobs examined)II.4 Male 38 Part time carpentry job 84 83 85 93 83 NAD Moderately high arched palate

*Approximate assessment because of very low performance level

had A values of 0 5 kb and 0 4 kb respectively.Methylation was approximately 80% and 90%respectively for III.1 and III.2. Absence of aCCG PCR product for III.2 excluded copynumber mosaicism in lymphocytes involvingcopies within the normal range. There was anexpansion when FRAXE was passed on to III.6from II.3. The A value increased from 0-2 kbto 0 7 kb. The CpG island was fully methylatedin III.6. The female FRAXE carriers in thisfamily show no consistent correlation betweengenotypic and phenotypic status. For example,both III. 1 and III.6 are mentally impaired. III.6is completely methylated at the FRAXE CpGisland but III.1 has approximately the samemethylation status as her unaffected brotherIII.2. These observations could be explainedby somatic mosaicism and differential tissueexpression but this is not testable. There doesappear to be some correspondence betweenFRAXE genotype and clinical phenotype inthis family for the two male carriers I.1 andIII.2, since the first one with the larger A valueand complete methylation is the mentally im-paired person, whereas the second one with amuch smaller A value is unimpaired. I. 1 has afully methylated CpG island whereas methyl-ation of the CpG island for III.2 is incomplete.

INTELLECTUAL PERFORMANCE IN MEMBERS OFFAMILY 1The clinical features and results of psy-chometric assessments of this family are sum-

marised in table 1. The most outstanding fea-ture in some subjects who carry FRAXE, butnot FRAXA, is borderline or mild intellectualdeficit. This concerns verbal and non-verbalabilities, combined with psychiatric problemssuch as ill-defined "nervous breakdowns", re-quiring hospital treatment in I.1 or schizo-phrenia in III.1. This woman (III.1) reportedauditory hallucinations starting at the age of 8.The main theme was a male voice telling herto "get rid of' or "kill" her mother. She hasresponded to Trifluoperazine HCG (Stelazine)but has residual symptoms including a very flataffect. Both I. 1 and III. 1 also expressed speechproblems in the form of a stutter. Physicalfeatures included various minor limb mal-formations, high arched palate, and some laxityof joints. The remaining families describedbelow did not undergo the detailed psy-chometric and psychiatric examinations as de-scribed for family 1.

FAMILY 2This is an extension of family 2 of Sutherlandand Baker.' Direct molecular diagnosis con-firmed the identity ofthe fragile site as FRAXE.The proband IV. 1 is a boy with no dysmorphicfeatures. He had mild mental impairment andattends special school. IV.2 completed the firstyear of secondary schooling and after a difficultsecond year left school to find employment.He is described by his mother as the mostintelligent of her three sons. IV.3 attends a

FRAXE and mental retardation 165


http://jmg.bmj.com/


Mulley et al

normal secondary school, is below average inhis school work, and receives remedial teachingwith reading. III.4 has normal intelligence andsatisfactorily completed three years of sec-ondary schooling. III.3 has low normal in-telligence, is illiterate, and did not undertakesecondary schooling. He is employed as a truckdriver. Possible explanations for his sons' in-tellectual disability need to take into accounthis own intellectual difficulties. II.2 has normalintelligence and before her retirement had em-ployment as a cleaner. III.5 appears to be ofnormal intelligence and has worked as a gardnerfor 20 years.

In family 2, FRAXE was transmitted fromII.2 to III.4 and III.5, and from III.4 to herthree sons. Only IV.1, the index case in thisfamily, was clearly mentally impaired on thebasis of clinical impression. His A value was1-3. Partial methylation was detected only forIV.2; the remaining tested FRAXE subjectsshowed full methylation. Obviously, no clearcorrelation can be seen between FRAXE frag-ment size, or methylation status, with mentalimpairment from the male carriers III.5, IV.2,and IV.3; but it must be emphasised that thisfamily did not undergo the detailed psy-chometric testing reported for family 1 andtherefore existence of some correlation cannotbe excluded. IV. 1 is clearly mentally impairedand IV.3 receives remedial teaching for reading.III.5 is a key person who appears to be normal,but comprehensive testing is essential to de-termine his true intellectual status.

FAMILY 3FISH and direct molecular diagnosis showedthat the fragile site in this family was FRAXE.The 18 year old proband II.1 is mildly in-tellectually handicapped with some definite be-havioural difficulties. He could read and writein a limited way with very good eye-hand co-ordination but little understanding of some ofthe material. Physically he was ofaverage heightand strong build with a square face and smallishears, prominent upper incisors, and high nar-row palate. His eyes had an antimongoloidslant with hypotelorism and epicanthic folds.The nipples were low placed and the handsand feet relatively small. Head circumferencewas 62 cm. Speech abnormality was his mostcharacteristic finding with constant verbal re-petitive utterances. He also had a number ofhand mannerisms with posturing and patting;if these were stopped, he became confusedand aggressive. Speech abnormality and handmannerisms were similar to those described forsome affected members of family 1 in table 1.His behaviour in the hostel has been describedas sexually overactive, intrusive, and aggressive,but he is responding to behavioural man-agement. Testes were above the 90th centilebut not grossly enlarged. There was no jointlaxity or skin hyperextensibility, the skin wasnot unusually soft, and there was no gyn-aecomastia. The sister I.2 who is positive cyto-genetically appears by clinical impression to benormal in all respects, as is the obligate carriermother, 1.2.

Family 3 consists of three FRAXE subjects.FRAXE was transmitted from I.2 to her daugh-ter II.2, with the amplification increased from0 3 kb to 0 8 kb and fully methylated in II.2.The only obviously mentally impaired personwas the index case II.1 with a value of 10 kband fully methylated CpG island. The valuefor II.1 (affected) in family 3 was similar tothat for IV.2 (unaffected) in family 2. However,the latter was only 50% methylated whereasthe former was fully methylated.

FAMILY 4Direct molecular diagnosis excluded FRAXAbut indicated the presence of FRAXE in thisfamily. The index case III.1 was an 8 year oldintellectually disabled boy with multiple spiderangiomata. FRAXE was transmitted to him byhis mother I.2 who received it from her motherI.2. Both the mother and the grandmotherappeared intellectually normal. During trans-mission the amplification steadily increased insize consistent with its behaviour during trans-mission through the other families.

FAMILY 5This kindred has two mentally retarded males,the proband III.7 and his cousin III.11. Theproband has moderate mental retardation butthe degree ofmental retardation for III. 1 1, whohas not been personally examined, remainsunspecified. The amplifications in both boysare almost fully methylated. II.2, II.4, II.6,III.2, and III.6 are confirmed female carrierswith the amplification fully methylated in III.2and III.6. The amplification has expandedwhen passed through the females in generationII to males and females in generation III. Theorigin of the mutation carried by II.2, II.4, andII.6 is paternal, given that their mother (I.2)was inferred to be heterozygous for 15 and19 copies of the FRAXE CCG repeat, andassuming that the pedigree as given is correct.The CCG copy number for I.2 could be in-ferred because the copy number for II.2, II.4,and II.6 was 15, 19, and 15 respectively. Hence,the maternal genotype was 15, 19 and thepaternally derived allele of II.2, II.4, and II.6failed to amplify because ofthe CCG expansioninherited from I. 1.

FAMILY 6Affected status was based on clinical im-pression. The proband III. 1 is severely retardedand his brother III.5 is moderately retarded.Their mother II.2 is moderately retarded andthe aunt II.3 is dull. The remaining aunts anduncles II.6, II.8, and II.10 are moderate toborderline in mental retardation status. Al-though the grandmother is cytogeneticallyFRAXE negative, she is the likely carrier giventhat her son II.10 is cytogenetically positive(although his fragile site was not confirmedas FRAXE by molecular testing). All of thementally impaired family members carry a fra-gile site and this was confirmed by molecularanalysis as FRAXE for II.2, II.3, III.1, andIII.5.

166


http://jmg.bmj.com/



Table 2 Contraction ofCCG repeat size during paternal transmission

Father a Daughter A Reference

FRAXA 0 8-1 1 kb 0-2-3 kb (2 daughters) 11Not stated Not stated 201-01-5 kb 0-20-3 kb (3 daughters) 21

FRAXE 08 kb 04 kb (1 daughter) 40-65 kb 0 35-04 kb (2 daughters) 41-2-1-8 kb 0-20-3 kb (2 daughters) Present study

DiscussionDetailed characterisation by in situ hy-bridisation using probes in Xq28 initially es-tablished the existence of FRAXE in the firstthree families, including the one described bySutherland and Baker' (family 2) and the oneoriginally reported by Loesch et al'5 (family 1).This diagnosis was confirmed by detection ofunstable sequences using the OxE20 probe. Itis worth noting that the fragile site describedby Romain and Chapman6 and reported bySutherland and Baker' in their family 1 was

excluded as FRAXE using the OxE20 probe.A localisation distal to FRAXE was shown forthe fragile site in this family by FISH. Thisfamily was then confirmed to have FRAXF bydirect molecular analysis.5 Both FRAXE andFRAXA were found to be segregating in family1, which could have easily confounded geno-type-phenotype inferences in the absence ofmolecular diagnosis to distinguish these fragilesites. Webb et al'6 reported I. 1 from this familyin the context offragileX syndrome and Loeschet all'5 carried out detailed psychological as-

sessment (in their family A). Cosegregation ofinstability at the FRAXA locus with instabilityat the FRAXE or FRAXF loci is a possibilityto consider when characterising families forfragile sites in distal Xq.

Reduction in the size of the FRAXE CCGexpansion was documented in family 1 duringtransmission from the affected male I.1 to histwo daughters. This behaviour of trinucleotiderepeats from the upper end of the observed sizedistribution of expansions may be sex specificsince this observation parallels the findings inFRAXA"7 and another FRAXE family (table2) and for myotonic dystrophy.'8'1' The generallack of increase in copy number during thetransmission of the FRAXA CCG repeatthrough males, including regression of the lar-ger copy numbers, is consistent with the ob-servation that FRAXA is not usually cyto-genetically penetrant in daughters of trans-mitting males. The same might be predictedto apply to FRAXE given the cytogenetic andmolecular similarities between the two fragilesites. Although there are limited data for thetransmission of FRAXA or FRAXE throughmales with large A values, the available in-formation (table 2) is suggestive of regressionof large A values for both FRAXA and FRAXEduring transmission through a male. Thismechanism can explain why the cytogeneticexpression of FRAXA and FRAXE can skipgenerations and in particular why the daughtersof FRAXA and FRAXE males do not havecytogenetic expression of these fragile sites.This may prolong the presence of the unstableFRAXA mutation in the population by pro-

tecting it from natural selection, supplementingthe input into the population from recurrentmutation of chromosomes with trinucleotiderepeats at the upper end of the normal un-expanded range in copy number.22 The processmay be similar for FRAXE.

Certain similarities, as well as differences,also appear between clinical symptomatologyofFRAXA and FRAXE. Psychiatric problems,and speech defect in the form of dyslalia orstutter such as encountered in FRAXE subjectsin families 1 and 3, may also occur in FRAXAsubjects, but rarely in such a degree and formas in I. 1 and III. 1 in family 1. However, noneof the four male carriers of FRAXE in family2 had any of the discernible speech, be-havioural, or psychiatric problems observed insome members of the other families. At leasttwo of the four male carriers did have learningdifficulties and this may reflect phenotypic vari-ability (as observed for fragile X syndrome)and, as such, family 2 may fit into a differentportion of the clinical spectrum.Data on intellectual scores for family 1 in

FRAXE and FRAXA people listed in table1 show that intellectual performance is, onaverage, lower in FRAXA than in FRAXEmembers and, consistent with what has beenreported earlier, it concerns specific deficitssuch as on Block Design and Similaritiesscores. 12 23 For example, FRAXA membersusually have reasonable vocabulary as meas-ured by the PPVT but are much poorer on theverbal concept skills assessed in the Similaritiessubtest where similarities between two andthree words have to be identified. In FRAXEthis discrepancy appears to be much less andall FRAXE subjects in family 1 are char-acterised by somewhat lower performanceacross all the abilities tested. Although this smallnumber of subjects does not allow for anyformal statistical comparisons, these data maybecome important if similar trends can beshown in other FRAXE families where theresults of psychometric testing, rather than de-scriptive or general IQ data, are available.Among physical features, minor limb mal-formation encountered in the FRAXE mem-bers of family 1 are not uncommon inFRAXA.24 However, none of the FRAXE car-riers in this family had facial features suggestiveof the fragile X syndrome.

Factors which blur a genotype-phenotyperelationship for FRAXE, already an issue forfragile X syndrome, will need to be recognisedand taken into account. The first of these is alack of proper clinical and psychological as-sessment. If psychometric assessment had notbeen used with family 1 then subjects such asIII.6 (IQ= 73) or I.3 (IQ= 69) may have beenmisclassified as normal on the basis of generalclinical impression alone. Assessment by clin-ical impression alone was carried out for themembers in the remaining families. The tent-ative classification of III.5, IV.2, and IV.3 infamily 2 as normal might therefore be con-sidered provisional in the absence of formalpsychometric assessment.The second factor affecting genotype-pheno-

type relationship could be copy number mo-

167


http://jmg.bmj.com/


Mulley et al

saicism first reported by Knight et al.49 Peoplehaving multiple fragments, some premutationsand some full mutations (to use terminologyfirst applied to FRAXA and fragile X syn-drome), may be less severely affected. Knightet a/49 found affected FRAXE males to have Avalues greater than 650 base pairs, similar tothe A value separating premutations from fullmutations in fragile X (FRAXA) syndrome.The A value of04 kb for the male III.2 in family1 of the present study was 90% methylated,suggesting that it is close to the FRAXE pre-mutation-full mutation boundary. A A value of10 for the male IL.1 in family 3 was fullymethylated, suggesting that it is within theFRAXE full mutation range. Degree of methyl-ation is a less reliable indicator for females.Random inactivation of the female X chro-mosomes leads to an expression of 50% methyl-ation in non-carrier females; however, randomX inactivation occurring early in embryogenesiscan skew the proportion of active and inactiveX chromosomes. Although methylation statusis given for females, no conclusions are madefrom these data.The phenomenon which could also blur a

genotype-phenotype relationship is methyl-ation mosaicism." This is the incompletemethylation of fragments with A valueswithin the accepted full mutation range, whichare normally methylated and associated with aclinical phenotype. For fragile X syndrome thishas been observed at the lower end of the fullmutation range for A values up to 1 2 kb.The effect on phenotype of copy number mo-saicism and methylation mosaicism has beendocumented for the fragile X.2126 The A valueof1 1 for IV.2 in family 2 of the present studywas associated with 50% methylation, sug-gesting that this might be a methylation mosaicsimilar to cases documented for FRAXA. Inaddition,III.2 in family 1 andIII.7 andIII.11in family 5 were not fully methylated, butmethylation was at least 90%.

Phenotypic expression may be modulatedfurther by between tissue mosaicism. Avalues were determined from lymphocytes, butcorresponding A values have not been de-termined in neural tissue. Discrepant pheno-typic expression between III.1 and III.2 infamily 1, both with similar A and methylationvalues, could have such an explanation. Thecritical tissue for phenotypic expression of de-velopmental delay or mental impairment isundoubtedly not blood. Tissue specific differ-ences in amplification, between lymphocytesand muscle, have been observed in myotonicdystrophy.2728

Finally, a genotype-phenotype relationshipmay be confounded by the occurrence of men-tally retarded subjects unrelated to the fragilesite. These phenocopies have been documentedin fragile X syndrome families.1125 We suggestthat the mental retardation of the schizophrenicFRAXE carrier 111.1 in family 1 could be ofthis type. The early onset of her auditory hal-lucinations distinguishes her from any of theother intellectually disabled people in thesefamilies. She carries only a small FRAXE amp-lification and unless she carries extensive be-

tween tissue mosaicism she is unlikely to beaffected by the FRAXE site since males with asimilar size amplification in lymphocyte DNAare clearly unaffected.The nature of the clinical relationship be-

tween FRAXE and non-specific mental re-tardation will be difficult to establish, incomparison with fragile X syndrome, given thatthe proposed phenotype is milder, manifestedprimarily as learning problems, and in need offormal quantification for confirmation. Never-theless, these pedigrees (figure) show thatwith the exclusion of index cases there are morementally impaired male and female carriersin these pedigrees than could reasonably beexpected by chance. FRAXE and FRAXF arefrequent enough among families ascertained byconventional cytogenetic analysis as having afragile site in distal Xq such that molecularconfirmation of which fragile site is involvedmust be carried out before genetic counsellingof family members.

Note added in proofA recent report29 indicates that a FRAXE fullmutation can be passed from father to daughter,so an affected male can have an affecteddaughter.

We thank the families described for their cooperation, and MrsNola Marshall for conducting the SCID-R Structured ClinicalInterview. We also thank the National Health and MedicalResearch Council of Australia, Fondo de Investigaciones San-itirias de la SS Grant Number 93/0004-00 for support, andthe Medical Research Council in the UK for the release of theOxE20 probe and forfinancialsupport. The research in Adelaidewas supported in part by an IRS award to GRS from the HNMI.

1 Sutherland GR, BakerE. Characterisation of a new rarefragile site easily confused with the fragile X. Hum MolGenet 1992;1:11-3.

2 Flynn GA, Hirst MC, Knight SJL, et al. Identification ofthe FRAXE fragile site in two families ascertained for Xlinked mental retardation.Jf Med Genet 1993;30:97-1 00.

3 Hirst MC, Bamicoat A, Flynn G, et al.The identificationof a third fragile site, FRAXF, in Xq27-q28 distal to bothFRAXA and FRAXE. Hum Mol Genet 1993;2:197-200.

4 Knight SJL, Flannery AV, Hirst MC, et al. Trinucleotiderepeat amplification and hypermethylation of a CpG islandin FRAXE mental retardation. Cell 1993;74:127-34.

5 Parrish JE,Oostra BA, Verkerk AJMH, et al. Isolation of aGCC repeat showing expansion in FRAXF, a fragile sitedistal to FRAXA and FRAXE. Nature Genet 1994;8:229-34.

6 Romain DR, Chapman CJ. Fragile site Xq27.3 in a familywithout mental retardation. Clin Genet 1992;41:33-5.

7 Verkerk AJMH, Pieretti M, Sutcliffe JS, et al. Identificationof a gene (FMR-1) containing a CGG repeat coincidentwith a breakpoint cluster region exhibiting length variationin fragile X syndrome. Cell 1991;65:905-14.

8 Pieretti M, Zhang F, Fu YH, et al. Absence of expressionof the FMR-1 gene in fragile X syndrome. Cell 1991;66:817-22.

9 Knight SJL, Voelckel MA, Hirst MC, Flannery AV, MonclaA, DaviesKE. Triplet repeat expansion at the FRAXElocus and X-linked mild mental handicap. Am J HumGenet 1994;55:81-6.

10 Donnelly A, Kozman H, Gedeon AK, et al. A linkage mapof microsatellite markers on the human X chromosome.Genomics 1994;20:363-70.

11 Mulley JC, YuS, Gedeon AK, et al. Experience with directmolecular diagnosis of fragile X. J Med Genet 1992;29:368-74.

12 Theobald T, Hay DA,JudgeJ. Individual variation andspecific cognitive deficits in the fragile X syndrome. AmJMed Genet 1987;28:1-11.

13 Sattler JM. The assessment of children. 3rd ed. San Diego:JM Sattler, 1988.

14 Spitzer RL, Williams JBW. Structured clinical interview forDSM-III-R. New York: New York State PsychiatricIn-stitute, 1985.

15 Loesch DZ, Hay DA, Sutherland GR, Halliday J,JudgeC, Webb GC. Phenotypic variation in male-transmittedfragile X: genetic inferences. Am J Med Genet 1987;27:401-17.

16 Webb GC, Rogers JG, Pitt DB, Halliday J, Theobald T.Transmission of fragile (X)(q27) site from a male. Lancet1981;ii:1231-2.

168


http://jmg.bmj.com/



17 Reyniers E, Vits L, De Boulle K, et al. The full mutationin the FMR-1 gene of male fragile X patients is absent intheir sperm. Nature Genet 1993;4:143-6.

18 Mulley JC, Staples A, Donnelly A, et al. Explanation forexclusive maternal origin for the congenital form of myo-tonic dystrophy. Lancet 1993;341:236-7.

19 Brook JD. Retreat of the triplet repeat? Nature Genet 1993;3:279-81.

20 Willems PJ, Van Roy B, De Boulle K, et al. Segregation ofthe fragile X mutation from an affected male to his normaldaughter. Hum Mol Genet 1992;1:511-15.

21 Hori T, Yamauchi M, Seki N, et al. Heritable unstable DNAsequences and hypermethylation associated with fragile Xsyndrome in Japanese families. Clin Genet 1993;43:34-8.

22 Richards RI, Holman K, Friend K, et al. Evidence offounderchromosomes in fragile X syndrome. Nature Genet 1992;1:257-60.

23 Crowe S, Hay DA. Neuropsychological dimensions of thefragile X syndrome: a test of hemispheric dysfunctionhypothesis. Neuropsychologia 1990;28:9-16.

24 Loesch DZ, Huggins R, Hay DA, Gedeon AK, MulleyJC, Sutherland GR. Genotype-phenotype relationships infragile X: a family study. Am J7 Hum Genet 1993;53:1064-73.

25 Rousseau F, Heitz D, Biancalana V, et al. Direct diagnosisby DNA analysis of the fragile X syndrome of mentalretardation. N Engl Med 199 1;325:1673-81.

26 Fu YH, Kuhl DPA, Pizzuti A, et al. Variation of the CGGrepeat at the fragile X site results in genetic instability:resolution ofthe Sherman paradox. Cell 199 1;67: 1047-58.

27 Anvret M, Ahlberg G, Grandell U, Hedberg B, Johnson K,Edstrom L. Larger expansions of the CTG repeat inmuscle compared to lymphocytes from patients with myo-tonic dystrophy. Hum Mol Genet 1993;2:1397-400.

28 Ashizawa T, Dubel JR, Harati Y Somatic instability ofCTGrepeat in myotonic dystrophy. Neurology 1993;43:2674-8.

29 Hamel BCJ, Smits APT, de Graaff E, et al. Segregation ofFRAXE in a large family: clinical, psychometric, cyto-genetic and molecular data. Am Hum Genet 1994;55:923-31.

169


http://jmg.bmj.com/


doi: 10.1136/jmg.32.3.162 1995 32: 162-169J Med Genet

J C Mulley, S Yu, D Z Loesch, et al. FRAXE and mental retardation.

http://jmg.bmj.com/content/32/3/162Updated information and services can be found at:

These include:

References http://jmg.bmj.com/content/32/3/162#related-urls

Article cited in:

serviceEmail alerting

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

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:


http://jmg.bmj.com/content/32/3/162

http://jmg.bmj.com/content/32/3/162#related-urls

http://group.bmj.com/group/rights-licensing/permissions

http://journals.bmj.com/cgi/reprintform

http://group.bmj.com/subscribe/

http://jmg.bmj.com/


Documents

FRAXE and mental retardation