5 Med Genet 1998;35:456-462

A silent mutation, C924T (G308G), in theLi CAM gene results in X linked hydrocephalus(HSAS)

Yang-Zhu Du, Cheryl Dickerson, Arthur S Aylsworth, Charles E Schwartz

AbstractThe Li cell adhesion molecule (LICAM) isa neuronal gene involved in the develop-ment of the nervous system. Mutations inLICAM are known to cause several clini-cally overlapping X linked mental retarda-tion conditions: X linked hydrocephalus(HSAS), MASA syndrome (mental retar-dation, aphasia, shuffling gait, adductedthumbs), spastic paraplegia type I(SPGl), and X linked agenesis of thecorpus callosum (ACC). In an analysis ofa family with HSAS, we identified a C-Ttransition (C924T) in exon 8 that wasinitially thought to have no effect on theprotein sequence as the alteration affectedthe third base of a codon (G308G). Exten-sive analysis of the other 27 exons showedno other alteration. A review of thesequence surrounding position 924 indi-cated that the C-*T transition created apotential 5' splice site consensus se-quence, which would result in an in framedeletion of 69 bp from exon 8 and 23 aminoacids of the LICAM protein. RT-PCR ofthe RNA from an affected male fetus andsubsequent sequence analysis confirmedthe use of the new splice site. This is thefirst report of a silent nucleotide substitu-tion in LICAM giving rise to an alterationat the protein level. Furthermore, it showsthat as mutation analysis plays an evermore important role in human genetics,the identification of a synonymous basechange should not be routinely discountedas a neutral polymorphism.(7Med Genet 1998;35:456-462)

Keywords: silent mutation; LI CAM; restriction endo-nuclease fingerprinting (REF); cryptic splice site

Rosenthal et al' were the first to presentevidence for the involvement of a mutation ofthe Li cell adhesion molecule (LlCAM) in Xlinked hydrocephalus. The finding was consist-ent with linkage analyses which placed bothLiCAM and X linked hydrocephalus resultingfrom stenosis of the aqueduct of Sylvius(HSAS) in Xq28,2 and studies that had impli-cated LiCAM in neurite outgrowth.4 Subse-quently, mutations in this neuronal glycopro-tein were found to cause other disordersclinically related to X linked hydrocephalus,such as MASA syndrome (mental retardation,aphasia, shuffling gait, adducted thumbs), Xlinked spastic paraplegia type I (SPG1), and Xlinked agenesis of the corpus callosum

(ACC).5 6 At present, there are over 50LiCAM mutations listed on the LI MutationWeb Page.7 Approximately 75% of the muta-tions are contained within 19 exons that repre-sent 65% of the coding sequence! Thesemutations include frameshifts, missense muta-tions, stop codons, and alterations in splice sitejunctions. In almost every case, a mutationappears limited to a single family. In a fewinstances, the same mutation will give rise tomore than one of the above clinically relatedgenetic disorders, so that one can observe bothinter- and intrafamilial variability in familieswith L1CAM mutations.616As part of our interest in X linked mental

retardation, we have been investigating familieswith HSAS and MASA. As expected, we haveobserved the range of mutations in Li CAMmentioned above.' However, in one family witha history of HSAS over four generations, wewere unable to detect any mutation other thana presumed silent alteration at position 924 inwhich a C-iT transition affected the thirdposition in a glycine codon (G308G). Wepresent evidence that this neutral mutationalevent is, in fact, the cause of HSAS as a resultof the creation of an alternate 5' splice site.

Materials and methodsCLINICAL FINDINGS

Case reportThe proband (IV. 1) had severe congenitalhydrocephalus associated with apparent aque-ductal stenosis; he received a shunt at 4 days ofage. The remaining cortex was very thin. Hishead circumference was within the normalrange at 3 months and was almost 1 SD belowthe mean at 4 years 8 months. At the age of 7years, a head MRI showed agenesis of the cor-pus callosum and hydrocephalus. His handswere held tightly clenched with the thumbs andthe third finger overlapping the second finger.All fingers could be straightened passivelyexcept the index fingers which were held tightlycontracted in the palms. The thumbs could beextended more easily than the index fingers.Therefore, he appeared to have adducted/contracted index fingers rather than theadducted thumb position that is sometimesassociated with HSAS.

FAMILY HISTORYFour maternal great uncles died at birth or inearly infancy with congenital hydrocephalus;they were brothers of the maternal grand-mother (fig 1A). Clinical information is notavailable, but they were described by the family

J C Self ResearchInstitute ofHumanGenetics, GreenwoodGenetic Center, OneGregor Mendel Circle,Greenwood, SC 29646,USAY-Z DuC E Schwartz

Women's Institute,Division of ClinicalGenetics, CarolinasMedical Center,Charlotte, NC, USAC Dickerson

Division of Geneticsand Metabolism,Department ofPediatrics and theNeuroscience Center,University of NorthCarolina at ChapelHill, Chapel Hill, NC,USAA S Aylsworth

Correspondence to:Dr Schwartz.

Received 4 September 1997Revised version accepted forpublication17 November 1997

456

on Novem

ber 9, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.35.6.456 on 1 June 1998. D

ownloaded from

Silent mutation in LlCAM and hydrocephalus

-0_.l.

"I

-I*

v'04-T*.:

E,...I

_1 _ __

Figure 1 (A) Pedigree of kindred K8715 with HSAS. (B) HphI analysis of the C924T alternation in kindred K8715.The unaffected chromosome contains a 117 bpfragment. A chromosome with the C924T alteration contains an HphI sitein the 117 bp fragment which upon digestion gives rise to two fragments of 79 bp and 38 bp, respectively. Notation abovethe lane gives the location in the pedigree in (A).

as looking like the proband. This family is fam-ily 7 in the report by Willems et al.2

GENOMIC DNA

Blood or amniocytes were obtained withinformed consent. Genomic DNA was isolatedfrom peripheral blood using a high saltmethod." DNA from amniocytes was preparedusing routine procedures. '2

RESTRICTION ENDONUCLEASE FINGERPRINTING(REF)Exons 7-10 were amplified as a 1088 bpfragment using the following primers: ex7-1 Of:5'CTGGGGTGGAGGGAAGAGTG3', ex7-1 Or: 5'CCAGTGGGTGCAGGGACAGA3'.The PCR reaction mixture contained 10mmol/l Tris-HCl, pH 8.3, 50 mmol/l KC1, 1.5mmol/l MgCl2, 50 [tmol/l of each dATP, dCTP,dGTP, dTTP, 1 ,umol/l of each primer, 2.5 U of

Taq DNA polymerase (Boehringer-Mannheim), 0.55 ptg TaqI polymerase startantibody (Clontech), and 125 ng genomicDNA in a 50 jtl total volume. The initial dena-turing temperature was 94°C for five minutes.This was followed by 35 cycles of denaturationat 94°C for 30 seconds, annealing at 70°C forone minute, and extension at 72°C for 1.5minutes, followed by a final extension at 72°Cfor 10 minutes. The PCR products werechecked by 1.5% agarose gel electrophoresisbefore doing REF.A total of 100 ng of the exons 7-10 PCR

product were digested separately by fourdifferent restriction endonucleases in a totalvolume of 10 [tl. The four digestion reactionscontained 4 U HhaI in NEB buffer 4, or 4 UDdeI in NEB buffer 3, or 4 U Hinfl, or 4 UMspI in NEB buffer 2, and 0.4 U CIAP at 37°Cfor 12 hours.

457

go

T

/

z

on Novem

ber 9, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.35.6.456 on 1 June 1998. D

ownloaded from

Du, Dickerson, Aylsworth, et al

=.. ...; m ....... _ . !.. ,,lE,.; .t

:: .t:.

.: :S

,rL jIL

.....

t: : t,,o'=

.;, .

__r.. ..* iX:::::: W* :. . . .. :

g.W

After digestion, all the endonucleases were

inactivated at 96°C for 10 minutes. The severalreactions were mixed together and reheated at

96°C for five minutes. Digested fragments (10ng) were incubated with 6 ptCi y-32P ATP and 1U T4 kinase in a buffer of 70 mmol/l Tris-HCl,pH 7.6, 10 mmol/l MgC12, 5 mmol/l DTT at

37°C for three hours. After the labellingreaction, 10 gl stop buffer (95% formamide, 10mmol/l NaOH, 0.25% bromophenol blue,0.25% xylene cyanol) were added to each tubeand 4 pl were loaded onto either a 6%acrylamide gel or a MDE (FMC) gel sepa-rately. The 6% acrylamide gel contained 10%glycerol and 0.6 x TBE and was run at 8 W,4°C for 24 hours. The MDE gel was preparedaccording to the FMC instruction manual andrun at 8 W at room temperature for 12 hours.The gels were dried and exposed to Biomax

film (Kodak, Rochester, NY, USA) at -80°Cfor 1.5 hours.

SEQUENCING OF PCR PRODUCTSPCR products (50 !tl) were run on 1% agarosegels containing 1 x TBE. The DNA bands werecut out and DNA was purified using QIAquickgel extraction kit (Qiagen). DNA sequencingwas performed using thefmolDNA sequencingsystem (Promega). a-33P dATP or a-35S dATPwere used in sequencing reactions and 3 ptl ofsequencing reaction was run on 6% polyacryl-amide gel containing 7 mol/l urea at 85 W forthree to six hours.

DETECTION OF HphI POLYMORPHISM

Template DNA (100 ng) from peripheralblood was PCR amplified in a 40 pl volumeusing primers G10 and GI1 (23), which flank

458

.0: .:

on Novem

ber 9, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.35.6.456 on 1 June 1998. D

ownloaded from

Silent mutation in LlCAM and hydrocephalus

ANormal

Exon 8 Exon 9

Genomic sequence

Protein sequence

-TGATGGCGAGT- GTGGAGgtatgg- ---9y/

/

// \ ---/ \ /

--QLLKVGEEDDGEYRCLAENSLGSARHAYYVTVEMPYWLHKPQSHL

- 23aa

BK8715

Splice site

Genomic sequence

Protein sequence

Splice site

Exon 8 69bp 1 / Exon 9

==--GATGAT--GGAGd gtatgg -C---.00-

----LLKVGEEDD(--)AAPYWLHKPQSHL

Figure 3 Schematic representation ofexon 8, intron 8, exon 9 organisation in LlCAM. (A) Normal genomic sequencenear the exon 8/intron 8 and intron 81exon 9 boundaries. Protein sequence of the LlCAM gene corresponding to this codingregion is given below the genomic sequence. (B) Genomic sequencefor same region in the X chromosome in K8715 whichcarries the C924T alteration. The predicted amino acid sequence for the altered LlCAM protein is given below the genomicsequence. The new 5' splice site is indicated by an arrow. In both (A) and (B), position 924 is indicated by *.

exon 8 of the L1CAM gene. The reaction con-

tained 10 mmol/l Tris-HCl, pH 8.3, 50 mmol/lKCl, 1.5 mmolIl MgCl2, 50 gmol/l of eachdATP, dCTP, dGTP, dTTP, 1 gmol/l of eachprimer, 2 U Taq DNA polymerase, and 0.44 jgTaqI Polymerase Start antibody. The initialdenaturing was performed at 94°C for fiveminutes and then followed by 35 cycles of94°C for 30 seconds, 68°C for 30 seconds, and72°C for one minute; the final extension was at72°C for seven minutes. PCR product (16 il)was digested using 8 U HphI in a 24 gl totalvolume containing 50 mmol/l potassium ac-

etate, 20 mmol/l Tris-acetate, 10 mmol/l mag-nesium acetate, 1 mmol/l dithiothteitol (pH7.9) at 37°C for 12 hours. The digestedsamples were run on a mini non-denaturing gelcontaining 8% acrylamide and 15% glycerol at170 volts and room temperature for one hour.The gel was stained in a 0.5 ,ug/ml ethidiumbromide solution for five minutes and a gelpicture was taken using an AlphaImager 2000(Alpha Innotech Corporation, California).

CELL CULTURE AND RNA ISOLATION

Amniocytes were maintained inMEM (Sigma)supplemented with 20% fetal bovine serumand 1% penicillin/streptomycin (Gibco). Whenthe number of the cultured cells reached 5 mil-lion, isolation of the total RNA was accom-

plished using Rneasy mini kit (Qiagen).

CDNA PREPARATION AND RT-PCRThe total RNA was treated with DNase Ibefore the reverse transcription (RT) reaction.The RT reaction was performed using 5 jg

DNase I treated total RNA containing 20mmol/l Tris-HCl (pH 8.4), 50 mmol/l KCI, 2.5mmol/l MgCl2, 0.5 mmol/l of each dNTP, 0.01

mol/l DDT, 250 ng random hexamer primers,and 200 units of SuperScript II reversetranscriptase (Gibco BRL). The reactionconditions were according to the manufactur-er's instructions.cDNA mix (2 gl) from the RT reaction was

used for RT-PCR in a 50 gl volume. The PCRreaction contained 10 mmol/l Tris-HCl (pH8.3), 1.5 mmol/l MgCl2, 50 mmol/l KCI, 0.2mmol/l each dATP, dCTP, dGTP, dTTP, 1jmol/l primers of L5-lOf (5'CCTC-CGACAACCACTCAGACTACA3') and L5-lOr (5'CGGTTGCGGGCCTCACATT3'),0.66 gg Taq polymerase start antibody, and 3 UTaq polymerase (Boehringer Mannheim). ThePCR conditions were as follows: denature at90°C for five minutes followed by 35 cycles ofdenature at 94°C for 30 seconds, annealing at62°C for one minute and extension at 72°C fortwo minutes; the final extension was at 72°Cfor 10 minutes. The PCR products were run on1.2% agarose gel containing 0.5 x TBE bufferand 0.5 ,ug/ml ethidium bromide.

ResultsMUTATION ANALYSIS

The Li CAM gene was screened for mutationsusing DNA from the proband in family K87 15(IV. 1, fig 1A). Restriction endonuclease finger-print (REF) analysis was used for 19 of the 28exons.8 13 This method involved amplifying the19 exons in five clustered groups: 5-6, 7-10,11-14, 16-18, 20-24.8 REF analysis showed a

fingerprint pattern consistent with an altera-tion in exon 8 (fig 2A). Sequence analysis ofexon 8 detected a C-*T transition at position924 (fig 2B). This base substitution occurred inthe third base in codon 308 which encodes a

glycine, thereby creating a neutral mutation

/

459

on Novem

ber 9, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.35.6.456 on 1 June 1998. D

ownloaded from

460 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Du,Dickerson, Aylsworth, et al

~~

Figure 4 RT-PGR analysis of exons 6-10 of the LlCAM cDNA from amniocytes from the affected fetus (IV3,figlJA).(A) Agarose gel electrophoresis of RT-PCR products generatedfrom a control (lane 1), the fetus (lane 3), and a cDNA

library (lane 5), using pri.mers gi.ven i.n Materials and methods. Lane 7 is the PCR product generated using genomic DNA

as a template. Lane 8 is a product generated using primers and RNA supplied with the SuperScript II reverse transcriptase

kit (Gibco BRL) which was used as a positive control for the RT-PCR reactions. (B) Sequence analysis of the RT§PCRproduct from a control sample (left panel) and the affected fetus (right panel). The arrow in the right panel indicates the

locati'on of the 69 bp deletion in the mRNA from the affected fetus.

(G308G). The base substitution, however, didcreate a new Hphl site. Analysis of the familyshowed that the presence of the site segregatedwith the HSAS (fig 1 B). This was consistentwith the previous haplotype analysis (data notshown).The remaining nine exons (1-4, 19, 25-28)

with their flanking intronic sequences wereexamined using SSCP. No sequence alterationswere detected after duplicate analyses (data notshown).

THE C924T (G308G) ALTERATION AS A MUTATIONWVhile the. family was still being investigated,111.2 became pregnant. Amniocentesis showedthe fetus to be a male (IV.3, fig IA). Analysis ofDNA isolated from the amniocytes indicatedthe X chromosome of the fetus carried the atrisk haplotype (data not shown) and the HphIrestriction site (fig 1 B). Ultrasound at 23weeks' gestation indicated borderline to mildventriculomegaly. Physical examination of thefetus after delivery at 24 weeks showed relative

460

on Novem

ber 9, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.35.6.456 on 1 June 1998. D

ownloaded from

Silent mutation in LlCAM and hydrocephalus

macrocephaly and clasped thumbs. Necropsywas refused. Based on these clinical featuresand the absence of any other identifiedsequence alteration in the LiCAM gene, theDNA sequence encompassing the C924Talteration was re-examined. Re-examination ofthe sequence showed that the C-*T transitionappeared to create a potential 5' splice site 69bp upstream from the normal exon 8/intron 8splice site (fig 3). The new consensus sequence(TGgtgagt) had a consensus value of 0.88while the normal splice site (GGgtatgg) had avalue of 0.76.14 In theory, the new 5' splice sitewould possibly result in 69 bp being eliminatedfrom the mRNA (fig 3B). Furthermore, theabsence of the HphI polymorphism in 200 Xchromosomes indicated that this alteration wasnot a common polymorphism in the normalpopulation.

Based on these findings and assumptions,RT-PCR was performed on total RNA isolatedfrom amniocytes of IV3 using primers de-signed to amplify exons 6-10 in LiCAMmRNA. The results, presented in fig 4A,showed a band estimated to be about 559 bpwhile the band amplified from control Li CAMmRNA contained the expected band of 628 bp.Thus, 69 bp were apparently missing from theLi CAM transcript of IV.3 as predicted. Directsequencing of the amplified fragment from IV.3showed that the deletion involved the last 69 bpof exon 8 (fig 4B).

DiscussionSequence analysis of exon 8 of LI CAM in afamily with HSAS indicated a single differencebetween the proband's exon and the normalLiCAM exon 8, a thymine substituted for acytidine (C924T). Both the normal codon 308(GGC) and the proband's codon 308 (GGT)encode for glycine. This silent nucleotide sub-stitution was not found to be a common variantin the normal population. In fact, this neutralmutation resulted in the activation of a newsplice site 69 bp 5' to the normal exon 8/intron8 donor splice site (fig 3B), and segregatedexclusively with HSAS in affected subjects andcarriers in the family. The alternatively splicedmRNA codes for a LICAM protein in which23 amino acids (residues 308-330) are missingfrom the third Ig domain. 15 This effectivelyeliminates the third near 3 sheet of the domain,undoubtedly severely affecting the activity ofthe LICAM protein and its ability to induceneurone-neurone adhesion.6 17

This mutation is notable for two reasons.First, the mutation differs from other splice sitemutations in LI CAM and other genes that giverise to a pathogenic phenotype as it neitherdestroys the function of a normal splice site norcreates a new splice site within an intron.6 18 Areview of published reports found that mRNAsplicing mutations account for 15% of all pointmutations causing human genetic disease.'8 Ofthese, 87% (88/101) are point mutationswithin either the 5' or 3' splice site consensussequence. The remaining point mutations(13%) lead to creation of novel splice sites.18Thus, the creation of a novel splice site result-ing from the C924T transition in exon 8 in our

family is representative of a very rare muta-tional event that accounts for about 2% of allpoint mutations.The second feature of the C924T mutation

in LICAM is perhaps more important. Thepoint mutation in codon 308 in our family is aclassic case of a neutral or silent mutation sincethe normal codon (GGC) and the variantcodon (GGT) both encode for glycine. To ourknowledge, there have been at least six otherreports of silent mutations giving rise to apathogenic phenotype. Goldsmith et al'8 re-ported a T-*A substitution in codon 24 of thef globin gene that did not alter the amino acidas both normal and variant codon encoded forglycine. Using a gene expression system, theauthors proved that the substitution causedabnormal mRNA splicing in a patient with Pthalassaemia."9 The second report was of anA-*G substitution in the third position of alysine codon of the factor IX gene in a patientwith haemophilia B.20 Although the alteredsequence had a splice donor consensus value of1.00, the authors did not prove that alternatemRNA splicing occurred in their patient. As noother mutation was found, the authors pre-sumed the silent A-*G substitution to bepathogenic.20 Baumbach et a12' reported a neu-tral mutation (C766T) in the third position ofcodon 236 of the growth hormone receptor(GH-R) gene. The mutation was shown to cre-ate a cryptic splice site 63 bp 5' to the normalexon 7 splice site, thereby eliminating 21 aminoacids from the GH-R. Homozygosity for thismutation results in Laron's syndrome in agenetic isolate in the Bahamas.2' A neutral G toA transition in the third position of the alanineat codon 344 (A344A) in FGFR2 (fibroblastgrowth factor receptor II) has been observed infour unrelated Crouzon syndrome patients.Reardon et ar-2 thought this base substitutionwould result in a new acceptor splice site. Jabset a!'3 suggested the mutation could createeither an acceptor or donor splice site.Recently, Li et al,24 using RT-PCR and RNAisolated from cell cultures of their affected sub-ject, showed that the A344A mutation createda new donor splice site. This finding wasconfirmed independently by Del Gatto andBrethnach25 using cloned genes in culturedHeLa cells. Richard and Beckman26 found asilent C-*T substitution in codon 624 of thecalpain gene (CANP3) in a homozygouspatient from one family with limb girdle mus-cular dystrophy (LGMD2A). Analyses ofillegitimate transcripts showed an abnormalproduct which, upon sequencing, was found tobe missing the last 44 bases of exon 16. Thus,this neutral mutation results in the gain of anew donor splice site. Lastly, a synonymoussubstitution (G2142A; S649S) in exon I1 pre-sumably created a new acceptor splice site giv-ing rise to a truncated RET proto-oncogeneprotein in members of a family with Hirsch-sprung disease.2Most certainly the C924T mutation in

LiCAM would have appeared as an alteredtranscript using RT-PCR. Indeed, at least twoneutral mutations have been detected in thismanner, both involving the XNP gene in two

461

on Novem

ber 9, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.35.6.456 on 1 June 1998. D

ownloaded from

Du, Dickerson, Aylsworth, et al

unrelated male patients with the ATR-Xsyndrome.28 29 In neither paper, however, wasthe "neutral" nature of the mutation stressedsince the method of detection involved RT-PCR. Unfortunately, in many instances, use ofgenomic DNA rather than mRNA is bettersuited for mutational analysis. For one thing, acell line is not always available for a deadproband while DNA may have been obtainedat some time before death. Second, not allgenes are expressed in lymphocytes which arethe most readily accessible source ofmRNA.30 31 Finally, some genes are quite large,making it difficult if not problematical to obtainfull length mRNA for analysis.The silent mutation in Li CAM that was

identified in our family with HSAS hasenormous implications for mutation screeningin any gene. The concept of a synonymous orneutral mutation should be altered when one isscreening a candidate gene for disease causingmutations. A neutral base substitution can

clearly have a profound effect on mRNA splic-ing or even codon usage, even though thenucleotide replacement is irrelevant with re-

spect to the amino acid encoded by the codon.Thus, it would be wise to remember the wordsof Sherlock Holmes: "We must fall back uponthe old axiom that when all other contingenciesfail, whatever remains, however improbable,must be the truth."32

We thank the members of the family for their cooperation andinterest in our research on LICAM. Retecher Nelson preparedthe RNA from the amniocytes and Chang-Yan Zou prepared thegenomic DNA. Karen Buchanan typed the manuscript. Thiswork was supported, in part, by a grant from the SCDepartment of Disabilities and Special Needs.

1 Rosenthal A, Jouet M, Kenwrick S. Aberrant splicing ofneural cell adhesion molecule L 1I mRNA in a family withX-linked hydrocephalus. Nat Genet 1992;2:107-12.

2 Willems P, Dijkstra I, Van der Auwera B, et al. Assignment ofX-linked hydrocephalus to Xq28 by linkage analysis.Genomics 1990;8:367-70.

3 Jouet M, Feldman E, Yates J, et al. Refining the genetic loca-tion of the gene for X linked hydrocephalus within Xq28. JMed Genet 1993;30:214-17.

4 Rathjen F, Schachner M. Immunocytological and biochemi-cal characterization of a new neuronal cell surface compo-nent (Li antigen) which is involved in cell adhesion.EMBO3 1984;3:1-10.

5 Jouet M, Rosenthal A, Armstrong G, et al. X-linked spasticparaplegia (SPG1), MASA syndrome and X-linked hydro-cephalus result from mutations in the Li gene. Nat Genet1994;7:402-7.

6 Vits L, Van Camp G, Coucke P, et al. MASA syndrome is

due to mutations in the neural cell adhesion gene L1CAM.Nat Genet 1994;7:408-13.

7 Van Camp G, Fransen E, Raes G, Willems P. Li mutationweb page. http://alt-www.uia.ac.be/u/dnalab/1 l.html 1996.

8 Du YZ, Srivastava AK, Schwartz CE. Multiple exon screen-

ing using restriction endonuclease fingerprinting (REF):detection of six novel mutations in the LI cell adhesionmolecular (LiCAM) gene. Hum Mutat 1998;11:222-30.

9 Fransen E, Schrander-Stumpel C, Vits L, Coucke P, VanCamp G, Willems P. X-linked hydrocephalus and MASA

syndrome present in one family are due to single missensemutation in exon 28 of the LI CAM gene. Hum Mol Getnet1994;3:2255-6.

10 Fryns J, Spaepen A, Cassiman JJ, Van den Berghe H. Xlinked complicated spastic paraplegia, MASA syndrome,and X linked hydrocephalus owing to congenital stenosis ofthe aqueduct of Sylvius: variable expression of the same

mutation at Xq28._Med Genet 1991;28:429-31.11 Schwartz CE, Ulmer J, Brown A, Pancoast I, Goodman HO,

Stevenson RE. Allan-Herndon syndrome. II. Linkage toDNA markers in Xq21. AmJ Hum Genet 1990;47:454-8.

12 Patterson M, Schwartz C, Bell M, et al. Physical mappingstudies on the human X chromosome in the region Xq27-Xqter. Genomics 1987;1:297-306.

13 Liu Q, Sommer S. Restriction endonuclease fingerprinting(REF): a sensitive method for screening mutations in long,contiguous segments of DNA. BioTech 1995;18:402-7.

14 Cooper D, Krawczak M. Are splice sites involved inmutations representative of the "average" splice site? In:Human gene mutation. UK: BIOS Scientific PublishersLimited, 1995:224.

15 Hlavin M, Lemmon V. Molecular structure and functionaltesting of human L1CAM: an interspecies comparison.Genomics 1990;11:416-23.

16 Bateman A, Jouet M, Macfarlane J, Du J, Kenwrick S, Cho-thia C. Outline structure of the human Li cell adhesionmolecule and the sites where mutations cause neurologicaldisorders. EMBOJ 1996;15:6050-9.

17 Takechi T, Tohyama J, Kurashige T, et al. A deletion of fivenucleotides in the L1CAM gene in a Japanese family withX-linked hydrocephalus. Hum Genet 1997;3:353-6.

18 Krawczak M, Reiss J, Cooper D. The mutational spectrumof single base-pair substitutions in mRNA splice junctionsof human genes: causes and consequences. Humi Geniet1992;90:41-54.

19 Goldsmith M, Humphries R, Ley T, Cline A, Kantor J,Nienhuis A. "Silent" nucleotide substitution in a ,B-thalassemia globin gene activates splice site in codingsequence RNA. Proc Natl Acad Sci USA 1983;80:2318-22.

20 Koeberl D, Bottema C, Ketterling R, Bridge P, Lillicrap D,Sommer S. Mutations causing hemophilia B: directestimate of the underlying rates of spontaneous germ-linetransitions, transversions, and deletions in a human gene.Am7Hunm Genet 1990;47:202-17.

21 Baumbach L, Schiavi A, Bartlett R, et al. Clinical, biochemi-cal, and molecular investigations of a genetic isolate ofgrowth hormone insensitivity (Laron's syndrome). _7 ClinEndocrinol Metab 1997;82:444-51.

22 Reardon W, Winter RM, Rutland P, Pulleyn LJ, Jones BM,Malcolm S. Mutations in the FGFR-2 gene cause Crouzonsyndrome. Nat Genet 1994;8:98-103.

23 Jabs EW, Li X, Scott AF, et al. Jackson-Weiss and Crouzonsyndromes are allelic with mutations in FGFR-2. Nat Genet1994;8:275-9.

24 Li X, Park WJ, Pyeritz RE, Jabs EW. Effect on splicing of a

silent FGFR2 mutation in Crouzon syndrome. Nat Genet1995;9:232-3.

25 Del Gatto F, Brethnach R. A Crouzon syndrome synony-mous mutation activates a 5' splice site within the IIIC exon

of the FGFR2 gene. Genomnics 1995;27:558-9.26 Richard I, Beckmann JS. How neutral are synonymous

codon mutations? Nat Genet 1995;10:259.27 Salomon R, Attie T, Pelet A, et al. Germline mutations of the

RET ligand GDNF are not sufficient to cause Hirsch-sprung disease. Nat Genet 1996;14:345-7.

28 Picketts D, Higgs D, Bachoo S, Blake D, Quarrell 0,Gibbons R. ATRX encodes a novel member of the SNF2family of proteins: mutations point to a common

mechanism underlying the ATR-X syndrome. Hum MolGenet 1996;5:1899-907.

29 Villard L, Lossi A, Cardoso C, et al. Determination of thegenomic structure of the XNP/ATRX gene, encoding a

potential zinc finger-helicase. Genomzics (in press).30 Gecz J, Gedeon A, Sutherland G, Mulley J. Identification of

the gene FMR2, associated with FRAXE mental retarda-tion. Nat Genet 1996;13:105-8.

31 Gu Y, Shen Y, Gibbs R, Nelson D. Identification of FMR2,a novel gene associated with the FRAXE CCG repeat andCpG island. Nat Genet 1996;13:109-13.

32 Conan Doyle A. The adventure of the Bruce-Partingtonplans. In: The works of Sir Arthur Conan Doyle, complete andunabridged. London: Longmeadow Press, 1984:803.

462

on Novem

ber 9, 2021 by guest. Protected by copyright.

http://jmg.bm

j.com/

J Med G

enet: first published as 10.1136/jmg.35.6.456 on 1 June 1998. D

ownloaded from