Letters to the Editor J Med Genet 1999;36:927–928 Leigh syndrome transmitted by uniparental disomy of chromosome 9 EDITOR—Severe, isolated, and generalised deficiency of complex IV (cytochrome c oxidase, COX) can result in Leigh syndrome (LS) (MIM 256000), an early onset mito- chondrial disorder characterised by rapidly progressive, symmetrical degeneration of the brain stem, diencephalon, and basal ganglia. 12 SURF-1, a gene located on chromo- some 9q34, has recently been identified as the gene responsible for numerous cases of LS COX . 34 SURF-1 associated LS COX is usually inherited as an auto- somal recessive trait. We report here a homozygous loss of function mutation of SURF-1 in two monozygotic LS COX female twins, owing to uniparental disomy of two almost identical maternal chromosomes 9. The probands were born to non-consanguineous parents at 33 weeks of gestation by caesarean section. The mother was 46 years old. The pregnancy was uneventful until the 24th week, when persistent uterine contractions ensued. Two older sibs of the probands are alive and healthy.The family history was negative for neurological or metabolic disorders. Birth weight and body length were <3rd centile and the Apgar score was 1-8 in both patients. Growth rate and psychomotor development were regular during the first 8 months of life. During the following months the patients developed a rapidly progressive clinical syndrome characterised by failure to thrive, psychomotor regression, hypotonia, ophthalmoparesis, mild bilateral optic atrophy, and ataxia. At 18 months both patients had mild lactic aci- dosis. MRI showed symmetrical paramedian lesions in the mesencephalon and brain stem, as typically found in LS. Both patients died of respiratory failure in the third year of life. Needle muscle biopsies performed at 24 months of age showed a diVuse reduction of the histoenzymatic reaction to COX. Biochemically, COX activity in muscle homoge- nates was 12.1 nmol/min/mg in one patient and 3.6 nmol/ min/mg in the second (normal values 68 ± 20), while the activities of the other respiratory complexes were all normal. The COX defect was also detected in cultured fibroblasts of one patient (0.4 nmol/min/mg, normal value 25 ± 11), but this assay was not performed in the second patient. Specific activities of the respiratory complexes in a muscle homogenate of the mother were all normal. Automated sequence analysis of the nine exons of the SURF-1 gene in the probands showed the presence of a previously reported 3 homozygous frameshift mutation (751C>T). This mutation destroys a BsiWI restriction site, which is present in the wild type gene. BsiWI RFLP analy- sis showed a heterozygous mutation in the mother, while no mutation was detected in the probands’ father, sister, or brother (fig 1). A de novo mutation in the paternal chromosome 9 identical to the mutation carried by the mother was considered unlikely. Non-paternity was excluded by linkage analysis with numerous microsatellite markers. To test the hypothesis of chromosome 9 specific paternal non-contribution, we then analysed three STSs (D9S1831, D9S1826, and D9S158), flanking the SURF-1 locus at 9q34. All three markers showed the presence of one maternal allele only, while the paternal allele was con- sistently absent. To verify whether the paternal non- contribution was the result of a microdeletion at 9q34, the cosmid P117B6, which contains the SURF-1 gene, 5 was used as a probe in FISH experiments 36 on metaphases from one proband. The probe detected two comparable signals on both chromosome 9 homologues (fig 2). These results excluded the presence of a deletion in a paternal chromosome, suggesting instead a mechanism of uniparen- tal disomy (UPD) of two maternal chromosomes. To test this hypothesis, additional microsatellites distributed along the whole of chromosome 9 were analysed for a total of 22 markers (fig 1). With the exception of two small regions (D9S288-D9S286, and D9S167-D9S283-D9S287, see fig 1) the alleles were all homozygous; in 15 instances the obligate contribution of the paternal allele was unequivo- cally missing. In particular, homozygosity was detected for 10/10 markers encompassing the SURF-1 locus, in the interval defined by markers D9S1831-D9S158. We conclude that loss of the contribution of a second normal SURF-1 allele has led to the manifestation of LS in our patients. UPD is defined as the exceptional inheritance of a pair of chromosomes from one parent only, as the result of gamete complementation, chromosome loss in trisomy, or duplica- tion in monosomy. In isodisomy, the uniparental pair is a duplicate of the same chromosome DNA template, and causes an increased risk of a recessive disorder by reduction to homozygosity. 7 In our patients the presence of two small heterozygous regions can be explained as the result of two crossing over events in otherwise identical maternal chromosomes 9. These data indicate that the double maternal contribution was the result of a non-disjunction Figure 1 (Top) Mutation specific RFLP analysis of exons 6+7 of the SURF-1 gene, amplified as described in reference 3. After digestion with BsiWI, the 389 bp wild type fragment is cut into two 351 and 38 bp fragments, while an intact fragment is obtained in the presence of the 751C>T mutation. (Middle) Pedigree of the family. (Bottom) Haplotype reconstruction of chromosome 9 specific microsatellite markers. The list and order of the markers along chromosome 9 are also indicated. 9p markers D9S288 148 144 148 146 144 152 146 152 146 152 146 152 D9S286 172 150 172 146 150 150 146 150 146 150 146 150 D9S285 111 109 111 111 109 91 111 111 111 111 111 91 D9S157 231 241 231 241 179 173 241 231 241 241 241 241 241 231 D9S171 179 173 140 138 179 173 173 173 173 173 173 179 D9S161 140 136 283 299 140 138 138 138 138 138 D9S1817 283 301 212 220 283 299 299 299 299 299 138 134 D9S273 212 212 266 278 212 220 220 220 220 220 299 304 D9S175 266 266 324 314 266 278 278 278 278 278 220 212 D9S167 324 320 324 314 314 324 314 324 278 264 D9S283 100 96 100 98 100 98 98 100 98 100 314 324 D9S287 294 294 294 296 294 294 296 294 296 294 98 100 D9S1831 253 257 253 263 257 253 263 263 263 263 296 294 D9S1847 192 166 192 190 166 190 188 188 188 188 263 253 D9S1861 187 193 187 185 193 185 181 181 181 181 188 190 D9S186 242 246 242 250 246 242 250 250 250 250 181 185 D9S125 125 133 125 123 133 123 123 123 123 123 242 250 D9S1793 180 180 180 180 180 180 182 182 182 182 123 123 182 180 D9S164 84 98 84 84 98 84 88 88 88 88 88 84 D9S1818 153 157 153 151 157 151 157 157 157 157 157 151 D9S1826 226 226 226 222 226 224 224 224 224 224 224 222 D9S158 339 339 339 343 339 345 345 345 345 345 345 343 9q Cen 389 bp 351 bp Letters 927 on October 3, 2021 by guest. Protected by copyright. http://jmg.bmj.com/ J Med Genet: first published as 10.1136/jmg.36.12.931 on 1 December 1999. Downloaded from
Transcript
Letters to the Editor
J Med Genet 1999;36:927–928
Leigh syndrome transmitted byuniparental disomy of chromosome 9
EDITOR—Severe, isolated, and generalised deficiency ofcomplex IV (cytochrome c oxidase, COX) can result inLeigh syndrome (LS) (MIM 256000), an early onset mito-chondrial disorder characterised by rapidly progressive,symmetrical degeneration of the brain stem, diencephalon,and basal ganglia.1 2 SURF-1, a gene located on chromo-some 9q34, has recently been identified as the generesponsible for numerous cases of LSCOX.3 4
SURF-1 associated LSCOX is usually inherited as an auto-somal recessive trait. We report here a homozygous loss offunction mutation of SURF-1 in two monozygotic LSCOX
female twins, owing to uniparental disomy of two almostidentical maternal chromosomes 9.
The probands were born to non-consanguineous parentsat 33 weeks of gestation by caesarean section. The motherwas 46 years old. The pregnancy was uneventful until the24th week, when persistent uterine contractions ensued.Two older sibs of the probands are alive and healthy. Thefamily history was negative for neurological or metabolicdisorders. Birth weight and body length were <3rd centileand the Apgar score was 1-8 in both patients. Growth rateand psychomotor development were regular during thefirst 8 months of life. During the following months thepatients developed a rapidly progressive clinical syndromecharacterised by failure to thrive, psychomotor regression,hypotonia, ophthalmoparesis, mild bilateral optic atrophy,
and ataxia. At 18 months both patients had mild lactic aci-dosis. MRI showed symmetrical paramedian lesions in themesencephalon and brain stem, as typically found in LS.Both patients died of respiratory failure in the third year oflife.
Needle muscle biopsies performed at 24 months of ageshowed a diVuse reduction of the histoenzymatic reactionto COX. Biochemically, COX activity in muscle homoge-nates was 12.1 nmol/min/mg in one patient and 3.6 nmol/min/mg in the second (normal values 68 ± 20), while theactivities of the other respiratory complexes were allnormal. The COX defect was also detected in culturedfibroblasts of one patient (0.4 nmol/min/mg, normal value25 ± 11), but this assay was not performed in the secondpatient. Specific activities of the respiratory complexes in amuscle homogenate of the mother were all normal.
Automated sequence analysis of the nine exons of theSURF-1 gene in the probands showed the presence of apreviously reported3 homozygous frameshift mutation(751C>T). This mutation destroys a BsiWI restriction site,which is present in the wild type gene. BsiWI RFLP analy-sis showed a heterozygous mutation in the mother, whileno mutation was detected in the probands’ father, sister, orbrother (fig 1). A de novo mutation in the paternalchromosome 9 identical to the mutation carried by themother was considered unlikely. Non-paternity wasexcluded by linkage analysis with numerous microsatellitemarkers. To test the hypothesis of chromosome 9 specificpaternal non-contribution, we then analysed three STSs(D9S1831, D9S1826, and D9S158), flanking the SURF-1locus at 9q34. All three markers showed the presence ofone maternal allele only, while the paternal allele was con-sistently absent. To verify whether the paternal non-contribution was the result of a microdeletion at 9q34, thecosmid P117B6, which contains the SURF-1 gene,5 wasused as a probe in FISH experiments3 6 on metaphasesfrom one proband. The probe detected two comparablesignals on both chromosome 9 homologues (fig 2). Theseresults excluded the presence of a deletion in a paternalchromosome, suggesting instead a mechanism of uniparen-tal disomy (UPD) of two maternal chromosomes. To testthis hypothesis, additional microsatellites distributed alongthe whole of chromosome 9 were analysed for a total of 22markers (fig 1). With the exception of two small regions(D9S288-D9S286, and D9S167-D9S283-D9S287, see fig1) the alleles were all homozygous; in 15 instances theobligate contribution of the paternal allele was unequivo-cally missing. In particular, homozygosity was detected for10/10 markers encompassing the SURF-1 locus, in theinterval defined by markers D9S1831-D9S158. Weconclude that loss of the contribution of a second normalSURF-1 allele has led to the manifestation of LS in ourpatients.
UPD is defined as the exceptional inheritance of a pair ofchromosomes from one parent only, as the result of gametecomplementation, chromosome loss in trisomy, or duplica-tion in monosomy. In isodisomy, the uniparental pair is aduplicate of the same chromosome DNA template, andcauses an increased risk of a recessive disorder by reductionto homozygosity.7 In our patients the presence of two smallheterozygous regions can be explained as the result of twocrossing over events in otherwise identical maternalchromosomes 9. These data indicate that the doublematernal contribution was the result of a non-disjunction
Figure 1 (Top) Mutation specific RFLP analysis of exons 6+7 of theSURF-1 gene, amplified as described in reference 3. After digestion withBsiWI, the 389 bp wild type fragment is cut into two 351 and 38 bpfragments, while an intact fragment is obtained in the presence of the751C>T mutation. (Middle) Pedigree of the family. (Bottom) Haplotypereconstruction of chromosome 9 specific microsatellite markers. The list andorder of the markers along chromosome 9 are also indicated.
which occurred at the second meiotic division, with main-tenance of euploidy in the zygote by elimination of thepaternal contribution. The age of the mother (46 years)could have favoured the non-disjunction event in ourpatients, as it is known that the risk of such an eventincreases with maternal age.
In addition, the haplotype reconstruction showedevidence of two recombinant events in the probands’ sibs,close to the SURF-1 locus (fig 1). The brother shares withthe probands the maternal allele for D9S186, but not thatfor D9S1861; the sister shares with the probands thematernal alleles for D9S1826 and D9S158, but not that forD9S1818. Since both sibs are homozygous wild type forSURF-1, these recombination events indicate that the dis-ease locus is contained within the interval between markersD9S1861 and D9S1826.3
UPD may also cause functional balance disruption ofimprinted genes.7 The existence of imprinted genes onchromosome 9 is controversial, but it seems unlikely.8–10
Our patients did not show gross dysmorphic features ormalformations apart from LS. With the limitations becauseof the brief survival and severe phenotype, this observationsuggests that chromosome 9 does not contain maternallyimprinted genes crucial for embryonic development.
We are indebted to Ms Barbara Geehan for revising the manuscript and to MrFranco Carrara and Mr Matteo Granatiero for technical assistance. This workwas supported by Fondazione Telethon-Italy (grant No 767 and grant NoE672), Ministero dell’Università e della Ricerca Scientifica e Tecnologica,Associazione Italiana Ricerca sul Cancro (AIRC), Ricerca Finalizzata Ministerodella Sanità No ICS 120.2/RF96.348, and European Union Human Capital andMobility network grant “Mitochondrial Biogenesis in Development andDisease”.
VALERIA TIRANTIELEONORA LAMANTEA
GRAZIELLA UZIELMASSIMO ZEVIANI
Divisione de Biochimica e Genetica, Istituto Nazionale Neurologico“Carlo Besta”, Via Celoria 11, 20133 Milano, Italy
PAOLO GASPARINICasa Sollievo della SoVerenza, S Giovanni Rotondo, Italy
ROSALIA MARZELLAMARIANO ROCCHI
Università Statale di Bari, Bari, Italy
MIKE FRIEDImperial Cancer Research Fund, UK
1 Rahman S, Blok RB, Dahl HH, et al. Leigh syndrome: clinical features andbiochemical and DNA abnormalities. Ann Neurol 1996;39:343-51.
2 Zeviani M, Bertagnolio B, Uziel G. Neurological presentations ofmitochondrial diseases. J Inherit Metab Dis 1996;19:504-20.
3 Tiranti V, Hoertnagel K, Carrozzo R, et al. Mutations of SURF-1 in Leighdisease associated with cytochrome c oxidase deficiency. Am J Hum Genet1998;63:1609-21.
4 Zhu Z, Yao J, Johns T, et al. SURF1, encoding a factor involved in the bio-genesis of cytochrome c oxidase, is mutated in Leigh syndrome. Nat Genet1998;20:337-43.
5 Duhig T, Rurberg C, Mor O, Fried M. The human surfeit locus. Genomics1998;52:72-8.
6 Lichter P, Tang Chang CJ, Call K, et al. High resolution mapping of humanchromosomes 11 by in situ hybridization with cosmid clones. Science 1990;247:64-9.
7 Engel E. Uniparental disomies in unselected populations. Am J Hum Genet1998;63:962-6.
8 Morison IM, Reeve AE. A catalogue of imprinted genes and parent-of-origin eVects in humans and animals. Hum Mol Genet 1998;7:1599-609.
9 Sulisalo T, Makitie O, Sistonen P, et al. Uniparental disomy in cartilage-hairhypoplasia. Eur J Hum Genet 1997;5:35-42.
10 Willatt LR, Davison BC, Goudie D, et al. A male with trisomy 9 mosaicismand maternal uniparental disomy for chromosome 9 in the euploid cell line.J Med Genet 1992;29:742-7.
J Med Genet 1999;36:928–932
A case of Williams syndrome with alarge, visible cytogenetic deletion
EDITOR—Williams syndrome (WS) is generally character-ised by mental deficiency, gregarious personality, dysmor-phic facies, supravalvular aortic stenosis (SVAS), and idio-pathic infantile hypercalcaemia. Patients with WS showallelic loss of STX1A,1 elastin (ELN),2 3 and LIMK1,4 withmost exhibiting a submicroscopic deletion at 7q11.23,detectable by FISH.3 5 The common deletion size is about1.5 Mb.6 Previous studies have shown that WS patientshave consistent deletion sizes and share common proximaland distal breakpoints.7 8 Here we report a patient who hasa large, atypical, visible chromosomal deletion of 7q11.2and features consistent with, and in addition to, those typi-cally seen in Williams syndrome.
The patient was originally referred to the genetics clinicat 5 months of age for evaluation of global developmentaldelay and dysmorphic features. She was delivered at 37
weeks’ gestation by caesarean section weighing 2350g(<5th centile). The initial course included a history of poorfeeding in the newborn period. Clinical examinationshowed macrocephaly, cutaneous haemangioma, andcraniofacial features consisting of a large anterior fonta-nelle, frontal bossing, depressed nasal bridge, cup shapedears, hypertelorism, and prominent lips (fig 1A). Neuro-logical examination showed generalised hypotonia withheel cord and hamstring tightness. CT scan of the head andrenal ultrasound were normal. Because of a grade III/VIsystolic murmur, echocardiogram was performed, whichshowed a slightly thickened aortic valve. Cytogeneticanalysis showed a 46,XX karyotype.
Re-evaluation at 4 years of age, showed short stature (90cm, <3rd centile), continued significant developmentaldelay, and coarsened facial features with stellate irides.Repeat echocardiogram showed moderately severe supra-valvular aortic stenosis. Cardiac catheterisation confirmedthese findings without involvement of the leaflets or the restof the ascending aorta, and no evidence of aorticinsuYciency. Ophthalmological examination showed bilat-eral exotropia and hyperopia requiring corrective lenses
Figure 2 FISH of DAPI stained metaphases from a fibroblast cell cultureof a patient. FISH was performed as described in references 3 and 6. Redsignals correspond to cosmid P117B6, which contains the entire SURF-1gene.
and very grey optic discs with surrounding peripapillaryretinal pigment epithelial changes. Serum calcium levelswere raised once (10.8 mg/dl, normal range 8.5-10.5mg/dl), but have since been in the normal range. FISHanalysis for the elastin locus on chromosome 7q11.23showed a deletion consistent with Williams syndrome3 andrepeat cytogenetic analysis showed a visible deletion inband 7q11.2 (46,XX,del(7)(q11.1q11.23)) (fig 2).
Follow up at 61⁄2 years of age showed weight 19.8 kg(40th centile), height 107 cm (5th centile), and headcircumference 52.5 cm (70th centile), with other notablefindings that included prominent supraorbital ridging,periorbital fullness, stellate pattern to her irides, cupshaped ears in normal position, hyperplastic gums, promi-nent, full lips, long philtrum, and broad nose withanteverted nostrils (fig 1B). She had fifth finger clinodac-tyly and brachydactyly. Examination of her skin showed ahaemangioma in the midline lumbosacral region, whichhad reportedly once extended from her occiput to her but-tocks. She had a hoarse, raspy voice and frequent drooling.Developmentally, she functioned in the severe mentalretardation range. She showed significant delays incommunication; expressive language was severely delayedwith rare speech (a few words) and limited sign languageand receptive language at <1 year of age. She had atypicalbehaviour including diminished interest in social interac-
tion with others, self-injurious behaviour, intermittentstereotypic behaviour, and sleep disturbance.
The patient had a sister (aged 8 years) and a brother(aged 10 years) who were healthy with normal develop-ment. There were no other family members with mentalretardation, short stature, or birth defects. The parentswere non-consanguineous. Blood samples were obtainedfrom both parents and the proband for additional molecu-lar studies.
In order to identify a FISH probe for the gene encodingthe á2/ä subunits of the L type voltage dependent skeletalmuscle calcium channel (CACNL2A), a PCR primer set(exnCa-A: 5'-CGGTGAGTGCTAAGACCTGAATG-3',exnCa-B: 5'-CAGCCCTCATAGATGTCAGTAGG-3')was designed from exon sequence obtained from theEMBL database under accession number Z28599. Primerswere used at a final concentration of 1.5 µmol/l in a 20 µlreaction. The amplification was performed with an anneal-ing temperature of 60°C for 30 cycles. Total human DNA(25 ng) was used as a template. The resulting 311 bpproduct was electrophoresed on a 1% low melting pointagarose gel. The gel fragment was excised and diluted 1:3with sterile water and was labelled with 32P-dCTP bystandard methodology. The probe was hybridised to highdensity filters arrayed with clones from a human totalgenomic P1 library. Four positive coordinates were
Figure 1 Front view of the patient at 22 months (A) and 61⁄2 years of age (B). The child has characteristic Williams syndrome facies including frontalbossing, prominent supraorbital ridging, periorbital fullness, stellate pattern to her irides, cup shaped ears in normal position, prominent, full lips, longphiltrum, and a broad nose with anteverted nostrils.
submitted to the Baylor College of Medicine YAC core andstreaks, representing 12 P1 clones per coordinate, werereceived. For each, single colonies were tested by PCRusing the exnCa primer pair. Clones from coordinate101F1 were found to be positive for the primer sequence.The positive clone was grown and DNA was preparedusing the Qiagen Plasmid Midi Kit following the amendedinstructions for P1 clones distributed by Qiagen. A cosmidsize FISH probe was made by subcloning the P1 cloneusing the SuperCos 1 Cosmid Vector Kit available fromStratagene. A clone (CaE-1) specific to the exnCa primerpair was isolated by testing subclones by PCR as describedabove.
A cosmid containing the D7S849 locus, known to belinked to the CACNL2A locus,9 but not contained in theprevious P1 clone, was subcloned from a yeast artificialchromosome (YAC) isolated from the CEPH Mark I YAClibrary by the Baylor College of Medicine YAC core usingthe D7S849 primer set. YAC DNA was prepared by astandard caesium chloride protocol and subcloned usingthe SuperCos I Cosmid Vector Kit. Human clones wereidentified by hybridisation of colony lifts with 25 ng of totalhuman DNA radiolabelled with 32P. Human clones werethen screened by PCR with D7S849. A single clone, posi-tive for D7S849, was identified (CaE-2) and DNA wasisolated using a Qiagen Plasmid Midi Kit and was used (20ng/µl) as a FISH probe against patient metaphase chromo-somes.
Additional probes used in the FISH analyses included acosmid containing the 5' end of the elastin gene(cELN272),3 7 a cosmid containing the full sequence ofLIMK1,4 7 and a bacterial artificial chromosome clone(BAC) containing the STX1A gene (BAC 137N23;Research Genetics, Huntsville, AL).
All FISH probes were labelled by nick translation withdigoxigenin-dUTP and detected with anti-digoxigeninconjugated to rhodamine. Either a biotin labelled chromo-some 7 alpha satellite centromere probe or a digoxigeninlabelled chromosome 7q telomere probe (Oncor, Inc,Gaithersburg, MD) was used as a control to identify thechromosomes 7. The centromere probe was detected usingavidin conjugated to fluorescein isothiocyanate (FITC).Slides were counterstained with DAPI. FISH analyses wereperformed as recently described.7
This patient was deleted for all the FISH probes tested,including the CACNL2A gene and the locus D7S849(fig 3).
DNA was extracted from peripheral blood from thepatient and each parent using standard methodology. Poly-morphic dinucleotide repeat markers were used to detectdeletions and determine the parental origin of the deletionas previously described.3 7 A deletion was evident when theproband failed to inherit an allele from one of the parents.The following loci were examined (listed centromeric totelomeric): D7S672, D7S1816, D7S489U, D7S2476,ELN, LIMK1, D7S613, D7S2472, D7S1870, D7S489L,D7S849, D7S675, D7S699, D7S440, and D7S634.
The patient was deleted for markers D7S489U (centro-meric) to D7S440 (telomeric) and uninformative forD7S634. The centromeric breakpoint was the same as seenin classical WS patients.7 The patient’s distal deletionbreakpoint was telomeric to the classical breakpoint(D7S1870), beyond the marker D7S440, with the exactdistal breakpoint undetermined.
WS presents as a remarkable collection of features withsignificant phenotypic variability among patients. Variabil-ity in the phenotype could be the result of diVerent sizeddeletions around ELN or the variation in gene content orgene activity of the hemizygous alleles on the non-deletedchromosome. Our previous studies have shown the size ofthe deletions in the majority of WS patients studied to beconsistent between the markers D7S489U and D7S1870.7
The current patient represents a rare exception.The gene encoding the á2/ä subunits of the L type volt-
age dependent skeletal muscle calcium channel(CACNL2A) was mapped to 7q21-q22.10 In addition, aform of malignant hyperthermia susceptibility (MHS) hasbeen linked to CACNL2A by analysis of a (CA)n repeatpolymorphism, D7S849, mapping to 7q11.23-q21.1.9
Given the occurrence of hypercalcaemia and the reports ofmasseter spasm and sudden death during surgical proce-dures in WS patients11 and the mapping of CACNL2Anear the WS critical region on chromosome 7, Mammi et
Figure 2 GTG banded partial karyotype of the patient. The normalchromosome 7 is shown on the left and the deleted chromosome 7 is shownon the right (arrow).
Figure 3 Fluorescence in situ hybridisation using cosmid CaE-1 to theCACNL2A locus. A 7q telomeric probe is used as a control. Two signalsare seen on the normal chromosome 7. Only the control signal is present onthe deleted chromosome 7 (arrow), indicating a deletion of the CACNL2Alocus.
al12 investigated the inclusion of this gene in microdele-tions in WS. Although CACNL2A was excluded from thecommon critical region in the WS patients studied,12 thislocus was deleted in the current case of a visible deletionof 7q11.2. It is not known if this patient would be suscep-tible to malignant hyperthermia, but caution should beexercised if anaesthesia is ever necessary.
To date, this is the only case of a visible deletion of7q11.2 in which the extent of the deletion has been delin-eated by molecular methods. The present deletion appearsto extend distally beyond D7S849 and CACNL2A. Thispatient has the characteristic craniofacial and cardiovas-cular abnormalities, including supravalvular aortic steno-sis, described in patients with Williams syndrome. By thescoring system developed by Preus,13 this patient’s scorewas +9.75, well within the Williams syndrome range(+12.59 ± 4.18). She has typical complications reported inpatients with Williams syndrome, such as esotropia,hyperopia, enamel hypoplasia, microdontia, early hyper-calcaemia, chronic otitis, short stature, and feeding prob-lems. She has not developed scoliosis, kyphosis, orcontractures. In addition, she has some atypical findingsnot seen in Williams syndrome including macrocephaly,retinal problems, severe mental retardation, and minimalspeech. She also has had a history of seizures (petit maltype) not often described in patients with Williamssyndrome. Her additional features are probably the resultof hemizygosity for genes outside the classical WS deletionregion.
In examining previously published reviews of cases of 7qdeletion, there is considerable clinical variation.14–19 In anattempt to determine if specific phenotypic features areassociated with proximal or distal deletions of 7q, deletionsof 7q were grouped into terminal deletions14–17 and intersti-tial deletions.20–22 Interstitial deletions of 7q have beendivided into three categories based on the region involved:(1) cen→q21/q22 (proximal), (2) q21→q31/32 (interme-diate), and (3) q32→q34 (distal).17 Although correlationsbetween phenotypes and deletions are diYcult to establishowing to the variable breakpoints, there are many commonbut non-specific findings reported in patients withproximal deletions or rearrangements of 7q, including lowbirth weight, mental retardation, microcephaly, growthretardation, early feeding problems in infancy, abnormalEEG/seizures, hypotonia, and abnormal skull shape.23–25
Comparison of the phenotypes between WS and proximaldeletion of 7q shows many common features, such as men-tal retardation, developmental delay, growth retardation,low birth weight, cardiovascular defects, eye abnormalities,facial dysmorphism and clinodactyly. However, the uniquecognitive profile seen in WS patients, significant deficits inmotor skills and impaired visuospatial recognition,26 27
loquaciousness, sociability, weak adaptive skills, depend-ency, hyperactivity, distractibility, inattention, and limitedperseverance,27–31 is not seen in patients with largedeletions, including our patient, perhaps because of moresevere mental retardation. Additionally, seizures arecommon among visible deletion patients, but not generallyseen as a feature in WS. This finding suggests that thegenes responsible for seizures are outside the common WSdeletion. The proximal breakpoint of this case is the sameas the proximal breakpoint of the critical deletion region ofWilliams syndrome, indicating, perhaps, a common mech-anism for the deletion in this case and the classicaldeletion.32 33
The authors thank the family for their participation, Dr M Keating (Universityof Utah) for the elastin containing cosmid, and Dr M Tassabehji (St Mary’sHospital, UK) for the LIMK1 containing cosmid. This research was supportedin part by an American Heart Association grant in aid (LGS) and NIH grantRO3 HD 35112 (LGS).
YUAN-QING WUELIZABETH NICKERSON
LISA G SHAFFERDepartment of Molecular and Human Genetics,Baylor College of Medicine, One Baylor Plaza,Room 15E, Houston, TX 77030, USA
KIM KEPPLER-NOREUILANN MUILENBURG
Department of Pediatrics, Division of Medical Genetics,University of Iowa Hospitals & Clinics,Iowa City, IA, USA
Correspondence to: Dr ShaVer
1 Osborne LR, Soder S, Shi XM, et al. Hemizygous deletion of the syntaxin1A gene in individuals with Williams syndrome. Am J Hum Genet1997;61:449-52.
2 Ewart AK, Morris CA, Atkinson D, et al. Hemizygosity at the elastin locusin a developmental disorder, Williams syndrome. Nat Genet 1993;5:11-16.
3 Nickerson E, Greenberg F, Keating MT, McCaskill C, ShaVer LG.Deletions of the elastin gene at 7q11.23 occur in ∼90% of patients withWilliams syndrome. Am J Hum Genet 1995;56:1156-61.
4 Tassabehji M, Metcalfe K, Fergusson WD, et al. LIM-kinase deleted in Wil-liams syndrome. Nat Genet 1996;13:272-3.
5 Lowery MC, Morris CA, Ewart A, et al. Strong correlation of elastindeletions, detected by FISH, with Williams syndrome: evaluation of 235patients. Am J Hum Genet 1995;57:49-53.
6 Meng X, Lu X, Li Z, et al. Complete physical map of the common deletionregion in Williams syndrome and identification and characterization ofthree novel genes. Hum Genet 1998;103:590-9.
7 Wu YQ, Sutton VR, Nickerson E, et al. Delineation of the common criticalregion in Williams syndrome and clinical correlation of growth, heartdefects, ethnicity, and parental origin. Am J Med Genet 1998;78:82-9.
8 Pérez Jurado LA, Peoples R, Kaplan P, Hamel BC, Francke U. Moleculardefinition of the chromosome 7 deletion in Williams syndrome and parent-of-origin eVects on growth. Am J Hum Genet 1996;59:781-92.
9 Iles DE, Lehmann-Horn F, Scherer SW, et al. Localization of the geneencoding the á2/ä-subunits of the L-type voltage-dependent calcium chan-nel to chromosome 7q and analysis of the segregation of flanking markersin malignant hyperthermia susceptible families. Hum Mol Genet 1994;3:969-75.
10 Powers PA, Scherer SW, Tsui LC, Gregg RG, Hogan K. Localization of thegene encoding the á2/ä subunit (CACNL2A) of the human skeletal musclevoltage-dependent Ca++ channel to chromosome 7q21-q22 by somatic cellhybrid analysis. Genomics 1994;19:192-3.
11 Patel J, Harrison MJ. Williams syndrome: masseter spasm during anaesthe-sia. Anaesthesia 1991;46:115-16.
12 Mammi I, Iles DE, Smeets D, Clementi M, Tenconi R. Anesthesiologicproblems in Williams syndrome: the CACNL2A locus is not involved. HumGenet 1996;98:317-20.
13 Preus M. The Williams syndrome: objective definition and diagnosis. ClinGenet 1984;25:422-8.
14 Nistrup Madson H, Lundsteen C, Steinrud J. A case of partial deletion ofthe long arm of chromosome 7 (7q34 leads to 7qter). Dan Med Bull 1983;30:14-16.
15 Schwartz S, Meekins J, Panny SR, Sun CC, Cohen MM. Brief clinicalreport: cebocephaly-holoprosencephaly in a newborn girl with a terminal7q deletion [46, XX, del (7) (pter leads to q32:)]. Am J Med Genet1983;15:141-4.
16 Reynolds JD, Golden WL, Zhang Y, Hiles DA. Ocular abnormalities in ter-minal deletion of the long arm of chromosome seven. J Pediatr OphthalmolStrabismus 1984;21:28-32.
17 Young RS, Weaver DD, Kukolich MK, et al. Terminal and interstitial dele-tions of the long arm of chromosome 7: a review with five new cases. Am JMed Genet 1984;17:437-50.
18 Gibson J, Ellis PM, Forsyth JS. Interstitial deletion of chromosome 7: a casereport and review of the literature. Clin Genet 1982;22:256-65.
19 PfeiVer RA. Interstitial deletion of a chromosome 7 (q11.2q22.1) in a childwith splithand/splitfoot malformation. Ann Genet 1984;27:45-8.
20 Baeteman MA, Philip N, Mattei MG, Mattei JF. Clinical, chromosomal andenzymatic studies in four cases of rearrangements of chromosome 7. ClinGenet 1985;27:564-9.
21 Fryns JP, Kleczkowska A, Limbos C, Vandecasseye W, Van den Berghe H.Centric fission of chromosome 7 with 47, XX, del (7) (pter→cen::q21→qter)+ cen fr karyotype in a mother and proximal 7q deletion in twomalformed newborns. Ann Genet 1985;28:248-50.
22 Frydman M, Steinberger J, Shabtai F, Steinherz R. Interstitial 7q deletion[46, XY, del (7) (pter→cen::q112→qter)] in a retarded quadriplegic boywith normal beta glucuronidase. Am J Med Genet 1986;25:245-9.
23 McElveen C, Carvajal MV, Moscatello D, Towner J, Lacassie Y.Ectrodactyly and proximal/intermediate interstitial deletion 7q. Am J MedGenet 1995;56:1-5.
24 Zackowski JL, RaVel LJ, Blank CA, Schwartz S. Proximal interstitialdeletion of 7q: a case report and review of the literature. Am J Med Genet1990;36:328-32.
25 Gillar PJ, Kaye CI, Ryan SG, Moore CM. Proximal 7q interstitial deletionin a severely mentally retarded and mildly abnormal infant. Am J MedGenet 1992;44:138-41.
26 Udwin O, Yule W. A cognitive and behavioural phenotype in Williams syn-drome. J Clin Exp Neuropsychol 1991;13:232-44.
27 Chapman CA, du Plessis A, Pober BR. Neurologic findings in children andadults with Williams syndrome. J Child Neurol 1996;11:63-5.
28 Arnold R, Yule M, Martin N. The psychological characteristics of infantilehypercalcaemia: a preliminary investigation. Dev Med Child Neurol1985;27:49-59.
29 Udwin O. A survey of adults with Williams syndrome and idiopathic infan-tile hypercalcemia. Dev Med Child Neurol 1990;32:129-41.
30 Udwin O, Yule W. Expressive language of children with Williams syndrome.Am J Med Genet Suppl 1990;6:108-14.
31 Tomc SA, Williamson NK, Pauli RM. Temperament in Williams syndrome.Am J Med Genet 1990;36:345-52.
32 Pérez Jurado LA, Wang YK, Peoples R, Coloma A, Cruces J, Francke U. Aduplicated gene in the breakpoint regions of the 7q11.23 Williams-Beurensyndrome deletion encodes the initiator binding protein TFII-I and BAP-135, a phosphorylation target of BTK. Hum Mol Genet 1998;7:325-34.
33 Osborne LR, Herbrick JA, Greavette T, Heng HH, Tsui LC, Scherer SW.PMS2-related genes flank the rearrangement breakpoint associated withWilliams syndrome and other diseases on human chromosome 7. Genomics1997;45:402-6.
J Med Genet 1999;36:932–934
First molecular evidence for a de novomutation in RS1 (XLRS1) associatedwith X linked juvenile retinoschisis
EDITOR—Juvenile retinoschisis (RS, OMIM 312700) is anX linked recessive vitreoretinal disorder that variablyaVects visual acuity because of microcystic degeneration ofthe central retina.1 2 In approximately 50% of aVectedmales, peripheral schisis may also occur. Major sightthreatening complications include vitreal haemorrhages,retinal detachment, and neovascular glaucoma.3
Recently, the gene underlying RS, designated RS1 (alsocalled XLRS1), was positionally cloned4 and more than 80diVerent mutations covering a wide mutational spectrum,including intragenic deletions, splice site, frameshift, non-sense, and missense mutations, were identified.4–7 Interest-ingly, missense mutations mainly cluster in exons 4 to 6 ofthe RS1 gene known to encode a highly conserved discoi-din domain thought to be involved in cell-cell interactionson membrane surfaces.8
The high recurrence rate of some of the RS1 mutations(for example, Glu72Lys in more than 34 patients from dif-ferent ethnic backgrounds) suggests a significant de novomutation rate in RS.8 In this report, we provide the firstmolecular evidence of a de novo RS1 mutation(Pro203Leu) in a Greek family. The Pro203Leu mutationis present in two brothers diagnosed with severe features ofRS at the ages of 9 and 5 years, respectively. We show thatthe mother is a heterozygous carrier while neither of thematernal grandparents carry the Pro203Leu mutation.Haplotyping data from several polymorphic DNA lociflanking the RS1 gene confirm paternity and strongly sug-gest that the Pro203Leu mutation originated on the Xchromosome of the maternal grandfather.
Two brothers were referred to one of the authors (BL)presenting with unclassified vitreoretinal degeneration inboth eyes. By history, retinal detachment had beendiagnosed in the right eye in the older (III.1) at the age of9 months. At the age of 9 years, best corrected visual acu-ity was 20/200 in the right eye (RE) and 20/40 in the lefteye (LE). Fundoscopy showed a bullous peripheral schisisand a flat schisis at the entire posterior pole with inner leafhole formation in the RE. In the LE, a macular schisis withmarked vitreous veils could be seen. Electroretinogram(ERG) recordings corresponding to the ISCEV Standardwere consistent with the diagnosis of RS, that is, rodresponse was unrecordable in the RE and residual in theLE, there was a negative maximal response, and anunrecordable cone response in both eyes.
In the younger brother (III.2), bullous cyst-like retinalchanges in both eyes had been diagnosed at the age of 1year. Four years later fundoscopy showed a bullous retinaldetachment in the inferotemporal retina of the RE includ-ing the macula with some cystic changes in the area of theinferior temporal vascular arcade. In the LE, only pigmen-tary abnormalities and whitish subretinal deposits consist-ent with a collapsed schisis could be seen. Best correctedvisual acuity was light perception RE and 2/100 LE.Because of severe nystagmus and reduced compliance, theERG was not recorded.
Fundus examination and ERG were normal in themother (II.1) and maternal grandfather (I.2).
Genomic DNA from the members of the Greek familywas extracted using standard techniques. Haplotyping wasdone using microsatellite markers 207F/R (DXS207),389gt, 418F/R (DXS418), and RX324 (DXS443) closelyflanking the RS locus (table 1).9–11 Microsatellite marker389gt was identified in PAC clone dJ389A20 as a (CA)30
dinucleotide repeat located 50 kb distal to DXS418(genomic sequence available at http://www.sanger.ac.uk/).The repeat sequences were PCR amplified in the presence of32P-dCTP (3000 Ci/mmol) using flanking oligonucleotideprimers and conditions as given in the references (table 1).To confirm paternity, an additional two highly polymorphicmicrosatellite markers at the ATM locus on 11q23 and theBRCA1 locus on 17q21 (D17S855) were used (table 1).
For mutational analysis, the six exons of the RS1 genewere PCR amplified from genomic DNA of patients III.1and III.2 with intronic oligonucleotide primers flanking therespective coding exons and amplification conditions asdescribed previously.4 Mutation detection was done bysingle stranded conformational analysis (SSCA). Amplifi-cation of the coding exons was carried out with Taqpolymerase (Gibco BRL) in a 25 µl volume in 1 × PCRbuVer supplied by the manufacturer. PCR products wereelectrophoretically separated on a 6% non-denaturingpolyacrylamide gel with or without 5% glycerol at 4°C.DNA fragments showing aberrant mobility shifts as well asthe corresponding maternal and grandparental PCR prod-ucts were directly sequenced using the Thermo Sequenaseradiolabelled terminator cycle sequencing kit (Amersham,Life Science).
Prescreening by SSCA of the six coding exons of the RS1gene showed a similar aberrant band shift in exon 6 in thetwo brothers III.1 and III.2 (fig 1 and data not shown).Direct sequencing of PCR products identified a C to Ttransition at nucleotide position 608 of the cDNA. This ispredicted to result in a proline to leucine substitution atcodon 203 (fig 1). Subsequently, sequencing of RS1 exon 6was performed in the mother, II.1, as well as in both mater-
Table 1 Polymorphic microsatellite markers used in the study
Name Locus Primer sequence 1 (5'—3') Primer sequence 2 (5'—3') Reference
nal grandparents, I.1 and I.2. The mother was heterozygousfor the Pro203Leu mutation while neither of the maternalgrandparents showed a mutational change (fig 1).
To confirm paternity, genotyping with polymorphic mark-ers ATM/in45 and D17S855 localised within the ATM andthe BRCA1 gene, respectively, were performed. Segregationof allelic markers is consistent with the grandfather, I.2, beingthe father of II.1 (fig 1). In addition, haplotype analysis wasdone with markers closely flanking the RS1 locus. Ahaplotype could be constructed (DXS207-389gt-DXS418-DXS443: 1-2-2-2) which is shared by the carrier mother,II.1, and her two sons, III.1 and III.2, and therefore should beassociated with the Pro203Leu mutation (fig 1). The grand-father also carried the 1-2-2-2 haplotype indicating that thePro203Leu mutation occurred on this haplotype.
Here, we describe a de novo missense mutation in theRS1 gene, Pro203Leu, that is associated with severefeatures of X linked juvenile retinoschisis (RS) in twoGreek brothers. Although the mother is a heterozygouscarrier neither of the maternal grandparents have thePro203Leu mutation. Haplotype analysis with polymor-phic markers closely flanking the RS1 locus providesstrong evidence that the Pro203Leu mutation occurred denovo on the X chromosome of the maternal grandfather.
The proline residue at codon 203 of the RS1 protein ispart of the evolutionarily highly conserved discoidindomain that is thought to be involved in cell-cellinteraction on membrane surfaces.8 Without exception, theproline residue is retained at this particular positionthroughout evolution in all proteins containing the discoi-din motif.4 7 In addition, Pro203Leu mutations have inde-pendently been identified in aVected subjects of threefamilial RS cases of French and Dutch origin but were notfound on 100 additional normal X chromosomes.7
Together, this strongly suggests that, rather than apolymorphism, the Pro203Leu mutation represents anamino acid change that should severely aVect protein func-tion and therefore should be responsible for the RS pheno-type in the two Greek brothers.
Haplotype analysis has shown that the maternalgrandfather of the two Greek RS patients carries the hap-lotype that becomes disease associated in his daughter andhis two grandsons. This provides strong evidence that thePro203Leu mutation is in fact a de novo event. It should bepointed out that the Pro203Leu mutation occurred at aCpG dinucleotide (codon 203: CCG to CTG) which, ifmethylated at the genomic level, is known to be frequentlyinvolved in C→T transitions.16 We cannot exclude that theunaVected grandfather is a mosaic for the Pro203Leumutation with the mutant genotype being present in one ormore tissues, excluding the ocular tissues but including aprecursor of the germ cells. Assuming such a situation inthe grandfather, the mutation could be transferred to hisdaughter and would then be perceived as a de novo germi-nal mutation.
Besides the Greek family, we were able to analyse thesegregation of RS1 mutations in another four pedigreeswhere RS occurred in a single generation of large families.There was no further evidence of de novo events in theextended families. However, considering the small numberof families tested, the present study supports an earliernotion that the new mutation rate in RS may besignificant.7 Further segregation analyses in multigenera-tion families with “sporadic” or only a few cases of RS willbe required to estimate more accurately the frequency of denovo mutations in X linked juvenile retinoschisis.
We thank the patients and their family for their kind cooperation. This work wassupported by the Deutsche Forschungsgemeinschaft (We 1259/5-3 and Lo 457/3-1).
ANDREA GEHRIGBERNHARD H F WEBER
Institut für Humangenetik, Biozentrum, Universität Würzburg, AmHubland, 97074 Würzburg, Germany
BIRGIT LORENZMONIKA ANDRASSI
Abteilung für Kinderophthalmologie, Strabismologie und Ophthalmogenetik,Klinikum der Universität, 93042 Regensburg, Germany
Correspondence to: Dr Weber
Figure 1 Analysis of a three generation Greek family with two cases of X linked juvenile retinoschisis (III.1 and III.2). Polymorphic markers at the ATMand the BRCA1 locus (D17S855) were used to confirm paternity. Haplotype analysis was performed using microsatellite markers DXS207, 389gt,DXS418, and DXS443 that closely flank the RS1 gene on Xp22.2. The order of markers is from telomere to centromere. Haplotypes associated with thePro203Leu mutation in exon 6 of the RS1 gene are boxed. Note that the grandfather, I.2, shares the disease associated haplotype with his two grandsonsIII.1 and III.2 but does not carry the Pro203Leu mutation.
1 Condon GP, Brownstein S, Wang N, Kearns AF, Ewing CC. Congenitalhereditary (juvenile X-linked) retinoschisis: histological and ultrastructuralfindings in three eyes. Arch Ophthalmol 1986;104:576-83.
2 George NDL, Yates JRW, Moore AT. X-linked retinoschisis. Br J Ophthalmol1995;79:697-702.
3 Deutman AF. Vitreoretinal dystrophies. In: Krill A, Archer D, eds.Hereditary retinal and choroidal diseases. New York: Harper and Row, 1977:1043-108.
4 Sauer GS, Gehrig A, Warneke-Wittstock R, et al. Positional cloning of thegene associated with X-linked juvenile retinoschisis. Nat Genet 1997;17:164-70.
5 Hotta Y, Fujiki K, Hayakawa M, et al. Japanese juvenile retinoschisis iscaused by mutations of the XLRS1 gene. Hum Genet 1998;103:142-4.
6 Rodriguez IR, Mazuruk K, Jaworski C, Iwata F, Moreira EF, Kaiser-KupferMI. Novel mutations in the XLRS1 gene may be caused by early Okazakifragment sequence replacement. Invest Ophthalmol Vis Sci 1998;39:1736-9.
7 The Retinoschisis Consortium. Functional implications of the spectrum ofmutations found in 234 cases with X-linked juvenile retinoschisis (XLRS).Hum Mol Genet 1998;7:1185-92.
8 Baumgartner S, Hofmann K, Chiquet-Ehrismann R, Bucher P. The discoi-din domain family revisited: new members from prokaryotes and ahomology-based fold prediction. Protein Sci 1998;7:1626-31.
9 Pawar H, Bingham EL, Hiriyanna K, Segal M, Richards JE, Sieving PA.X-linked juvenile retinoschisis: localization between (DXS1195, DXS418)and AFM291wf5 on a single YAC. Hum Hered 1996;46:329-35.
10 Van de Vosse E, Bergen AAB, Meershoek EJ, et al. An Xp22.1-p22.2 YACcontig encompassing the disease loci for RS, KFSD, CLS, HYP and RP15:refined localization of RS. Eur J Hum Genet 1996;4:101-4.
11 Huopaniemi L, Rantala A, Tahvanainen E, de la Chapelle A, Alitalo T.Linkage disequilibrium and physical mapping of X linked juvenileretinoschisis. Am J Hum Genet 1997;60:1139-49.
12 Oudet C, Weber C, Kaplan J, et al. Characterization of a highly polymorphicmicrosatellite at the DXS207 locus: confirmation of very close linkage tothe retinoschisis disease gene. J Med Genet 1993;30:300-3.
13 Van de Vosse E, Booms PFM, Vossen RHAM, Wapenaar MC, Van OmmenGJ, Den Dunnen JT. A CA-repeat polymorphism near DXS418 (P122).Hum Mol Genet 1993;2:2202.
14 Browne D, Barker D, Litt M. Dinucleotide repeat polymorphisms at theDXS365, DXS443 and DXS451 loci. Hum Mol Genet 1992;1:213.
15 Dib C, Fauré S, Fizames C, et al. A comprehensive genetic map of thehuman genome based on 5,264 microsatellites. Nature 1996;380:152-4.
16 Duncan B, Miller J. Mutagenic deamination of cytosine residues in DNA.Nature 1980;287:560-1.
J Med Genet 1999;36:934–935
Pathogenicity of homoplasmicmitochondrial DNA mutation andnuclear gene involvement
EDITOR—Seneca et al1 reported a homoplasmic deletion ofa T nucleotide in a 5T stretch (15 940-15 944 base pairs(bp)) of mitochondrial DNA (mtDNA) in two familiesassociated with clinical and pathological findings of mtcytopathy. Although this deletion was homoplasmic anddid not fulfil the classical criteria of pathological mutation,Seneca et al1 suggested that it was pathological, as theycould not identify any other heteroplasmic mutations,deletions, or duplications in tRNA genes of mtDNA inthese patients. However, this mutation was present not onlyin aVected patients but also in asymptomatic relatives inboth families. Therefore, this mutation does not cosegre-gate with the disease. It is diYcult to confirm whetherhomoplasmic mutations are pathological, as was recentlyindicated by Chinnery et al.2 3 There are currently no con-crete criteria to determine what kind of homoplasmicmtDNA abnormalities are pathological. Maternal inherit-ance is an important characteristic to confirm their patho-genicity, which, however, was not significant in these twofamilies. The mode of inheritance of this deletion isdiYcult to confirm, as it is currently unknown whether thesingle nucleotide deletion is inherited maternally likemtDNA point mutations. It is possible that it is inheritedautosomal dominantly like mtDNA deletions.4 In suchcases, cosegregation of the mutation in aVected familymembers is important to determine its pathogenicity. Apopulation based association study is another method forconfirming a significant role of homoplasmic or heteroplas-mic mtDNA mutations. The association should also beconfirmed by other studies on the same and diVerent eth-nic groups.
By directly sequencing a mutation hot spot of mtDNA(3130-3423 bp) from 30 patients with type 2 diabetes mel-litus (DM), we identified a G3316A homoplasmicmutation.5 The prevalence of this mutation was signifi-cantly higher in patients with glucose intolerance than inthose with normal glucose tolerance.5 This missense muta-tion in the ND-1 gene, which substitutes alanine for threo-nine, was present at an increased frequency in patients withtype 2 DM compared with non-diabetic subjects in otherstudies in Japanese6 or European7 populations. The samemutation was also identified in a patient through screeningpatients with hypertrophic cardiomyopathy, suggesting a
role of this homoplasmic mutation in the development ofmt cytopathy (manuscript in preparation). Althoughhomoplasmic mtDNA mutations do not fulfil the classicalcriteria for pathogenicity, another recent study indicatedthat homoplasmic mutations are significantly associatedwith type 2 DM (p=0.0011, 0.0457, 0.0194).8 These find-ings suggest that the homoplasmic mutations are also ofpathological importance in mt cytopathy. Investigations onLeber’s hereditary optic neuropathy (LHON) suggest arole of the nuclear gene in the pathogenesis of clinicalsymptoms of mt cytopathy. Previous investigations, how-ever, failed to identify any nuclear gene abnormalities inpatients with mt cytopathy.9 We consider that homoplasmicmutations are also important in the development of mtcytopathy, as nuclear DNA may be involved in itspathogenesis.
Concerning the A3243G mutation, we suspect thatnuclear gene abnormalities may be responsible for the dif-ferent clinical phenotypes of type 2 DM or MELAS (mito-chondrial encephalomyopathy, lactic acidosis, and stroke-like episodes) associated with the same A3243Gmutation.10 Recent investigations indicate that nuclearencoded gene mutations are associated with Leighsyndrome.11 12 These observations highlight theimportance of nuclear gene-mtDNA interaction in thepathogenesis of mt dysfunction. They also suggest thathomoplasmic mtDNA mutations are important in devel-oping mt cytopathy, such as mt myopathies, diabetes mel-litus, or cardiomyopathies. Although diYculties exist inconfirming a pathogenic role of homoplasmic mtDNAmutations, some homoplasmic mutations are probablyassociated with mt dysfunction causing mt cytopathy. Wepropose that investigations of mtDNA abnormalities inpatients with mt dysfunction should include homoplasmicmutations which cosegregate with clinical or pathologicalmanifestations of mt cytopathy or are present with anincreased frequency in aVected patients.
MASATO ODAWARAHISATAKA MAKI
NOBUHIRO YAMADA
Institute of Clinical Medicine, University of Tsukuba, 1-1-1, Tennodai,Tsukuba City, 305-8575 Japan
1 Seneca S, Lissens W, Liebaers I, et al. Pitfalls in the diagnosis of mtDNAmutations. J Med Genet 1998;35:963-4.
2 Chinnery PF, Turnbull DM, Howell N, Andrews RM. Mitochondrial DNAmutations and pathogenicity. J Med Genet 1998;35:701-2.
3 Albin RL. Fuch’s corneal dystrophy in a patient with mitochondrial DNAmutations. J Med Genet 1998;35:258-9.
4 Odawara M, Yamashita K. Idiopathic dilated cardiomyopathy. N Engl J Med1995;332:1385.
5 Odawara M, Sasaki K, Yamashita K. A G-to-A substitution at nucleotideposition 3316 in mitochondrial DNA is associated with Japanese
non-insulin-dependent diabetes mellitus. Biochem Biophys Res Commun1996;227:147-51.
6 Nakagawa Y, Ikegami H, Yamato E, et al. A new mitochondrial DNA muta-tion associated with non-insulin-dependent diabetes mellitus. Biochem Bio-phys Res Commun 1995;209:664-8.
7 McCarthy M, Cassell P, Tran T, et al. Evaluation of the importance ofmaternal history of diabetes and of mitochondrial variation in the develop-ment of NIDDM. Diabet Med 1996;13:420-8.
8 Tawata M, Ohtaka M, Iwase E, Ikegishi Y, Aida K, Onaya T. Newmitochondrial DNA homoplasmic mutations associated with Japanesepatients with type 2 diabetes. Diabetes 1998;47:276-7.
9 Odawara M, Yamashita K. Mitochondrial DNA and disease. N Engl J Med1996;334:270-1.
10 Odawara M, Yamashita K. Are MELAS and diabetes mellitus caused solelyby the same mutation at np 3243 of the mitochondrial gene? Diabetologia1995;38:1488-90.
11 LoeVen J, Smeitink J, Triepels R, et al. The first nuclear-encoded complex Imutation in a patient with Leigh syndrome. Am J Hum Genet1998;63:1598-608.
12 Tiranti V, Hoertnagel K, Carrozzo R, et al. Mutations of SURF-1 in Leighdisease associated with cytochrome c oxidase deficiency. Am J Hum Genet1998;63:1609-21.
J Med Genet 1999;36:935–937
Identification and clinical presentationof â thalassaemia mutations in theeastern region of Saudi Arabia
EDITOR—The autosomal recessive disease â thalassaemia isa common single gene disorder that poses a serious healthproblem in many parts of the world. According to theHuman Gene Mutation Database (http://www.uwcm.ac.uk/uwcm/mg/hgmd.html) and the â-GlobinGene Server (http://globin.cse.psu.edu) about 300 se-quence variants in the â globin gene have been identifiedup to the present. Mutations in the â globin gene have beenfound at carrier frequency rates ranging from 1% in someareas of Saudi Arabia to 15% in others.1 Both â° and â+
thalassemia have been reported.2 Studies on the molecularpathogenesis of â thalassaemia have shown that the muta-tions encountered in Arab countries close to the Mediter-ranean basin are the same as those reported in other Medi-terranean populations.3 In the Gulf region, in SaudiArabia, UAE, and Iraq, the Asian pattern of mutationsseems to be prevalent.4 5 The precise genetic changesprevalent in the diVerent regions of the large country ofSaudi Arabia and analysis of the genotype/phenotype rela-tionship of the disease in Saudi patients still remain inad-equately studied.
The present study aimed to investigate the mutationalpattern of the â globin gene and to explore the relationshipbetween these mutations and disease presentation in agroup of patients with â thalassaemia major from the east-ern region of Saudi Arabia. For this purpose, 31 childrendiagnosed with â thalassaemia major who over the past twoyears had regularly attended the paediatric clinics of QatifCentral Hospital or Dammam Maternity and ChildrenHospital were selected. Within this group of patients therewere four pairs of sibs and one pair of first cousins. Thewhole â globin gene of all patients was amplified usingstandard PCR techniques and six specially designed diVer-
ent primers for amplification and sequencing. Nucleotidesequencing was performed by electroinjection of the PCRproducts into an automatic capillary ABI Prism GeneticAnalyzer type 310 (Perkin-Elmer, USA).
Results of the nucleotide sequencing have enabled accu-rate identification of disease causing mutations both in thehomozygous and heterozygous states in each of the 31patients diagnosed with â thalassaemia major. In total,eight disease causing mutations were detected:CD39(C→T),6 IVS-1 3'end-25bp,7 −2CD8(−AA),8
IVS-2+1(G→A),9 −1CD44(−C),10 +1CD8/9(+G),11
IVS-1+5(G→C),11 and IVS-1+5(G→T)12 that comprisedallele frequencies of 32.1%, 22.6%, 15.1%, 15.1%, 7.5%,3.8%, 1.9%, and 1.9%, respectively. An overall âthalassaemia detection rate of 100% was achieved (tables 1and 2), thus reflecting the eYciency of the technique. Theaccuracy of the genetic analysis has a special diagnosticimportance in view of the fact that certain haemoglobin-opathies, for example, sickle cell anaemia, spherocytosis,and autoimmune haemolytic disease, are known toproduce symptoms that mimic those of â thalassaemiamajor. The silent polymorphic mutationCD2(CAC→CAT) was very frequent among this group ofchildren as it was encountered in 16 out of 31 patients withâ thalassaemia major. On the other hand, the polymor-phisms IVS-2+26 (T→G) and IVS-2+74(G→T) were lesscommon and detected in only two patients each (Nos 29and 37 and 17 and 45, respectively). A fourth polymor-phism IVS-2+666(T→C) was also detected in two patients(Nos 17 and 45). These sequence variants are not shown intable 2.
The spectrum of mutations identified here confirms thenotion that, for historical reasons, there is an overlapbetween Mediterranean and Asian mutations in SaudiArabia.1 Five of the eight mutations detected here havebeen reported to exist among patients from other parts ofthe country, albeit with diVerent allele frequencies.1 Wereport here the novel identification in the Saudi populationof three mutations: the Turkish frameshift mutation-2CD8(-AA),8 the Kurdish frameshift mutation
Table 1 â globin gene mutations in patients with â thalassaemia major from the eastern region of Saudi Arabia
Mutation Type of mutation Nucleotide changeType ofthalassaemia
Independent alleles
Origin of mutationOriginalreferenceNumber Frequency
CD39(C→T) Nonsense Substitution C→T â° thalassaemia 17 32.1% Mediterranean, European 6IVS-1 3’ end-25bp Splice junction frameshift Deletion of 25 bp â° thalassaemia 12 22.6% Asian, Indian 7−2CD8(−AA) Frameshift Deletion of 2A â° thalassaemia 8 15.1% Turkish 8IVS-2+1(G→A) Splice junction change Substitution G→A â° thalassaemia 8 15.1% Mediterranean, Tunisian,
African-American9
−1CD44(−C) Frameshift Deletion of C â° thalassaemia 4 7.5% Kurdish 10+1CD8/9(+G) Frameshift Insertion of G â° thalassaemia 2 3.8% Asian, Indian 11IVS-1+5(G→C) Consensus change Substitution G→C â+ thalassaemia 1 1.9% Indian, Chinese, Melanesian 11IVS-1+5(G→T) Consensus change Substitution G→T â+ thalassaemia 1 1.9% Mediterranean, Blacks 12
Calculation of allele frequency:There are 31 patients with â thalassaemia major.No of alleles = 31×2 = 62 alleles.There are 4 pairs of sibs; hence 62−8 = 54 alleles.Subtract 1 allele for first cousin = 53 independent alleles.Detection rate = 53/53×100 = 100%.
-1CD44(-C),10 and the Mediterranean/Black consensusmutation IVS-1+5(G→T).12 It is highly unlikely that thesemutations arose independently in the Saudi population.On the contrary, there is historical evidence that theArabian, Turkish, and Kurdish populations have interactedin the past through trade and the introduction of Islam inTurkey. The introduction of the Mediterranean/Blackmutation in Saudi Arabia can be explained by a similarhistorical reason, that is, gene flow facilitated by trade andexpansion of Islam in the Mediterranean region and inAfrica. Similarly, evidence of a common ethnic origin ofthe cystic fibrosis mutation 3120+1G→A in Arab, African,Greek, and African-American populations has recentlybeen reported.13 However, we have detected the rare AsianIndian mutation +1CD8/9(+G) in the compound hetero-zygous state in two patients. It is likely that the Indianmutation may have been introduced to this region by geneflow since the mutation has recently been identified inpatients from the geographically close island of Bahrain,14
which is a known historical centre for international tradebetween Arabia, India, and Asia. Similarly, we have foundthat the splice junction frameshift mutation IVS-13'end-25bp, first reported in Asia and India, was thesecond most prominent mutation with an allele frequencyof 22.6% in this group of patients. This mutation hasrecently been reported as the most common mutation inBahrain with the highest allele frequency of 40%.14 There-fore, all these findings provide genetic evidence that theeastern coast of Saudi Arabia, unlike the inner desert partsof the Arabian Peninsula, was particularly prone to geneflow from Turkey, Iran, and the Indian subcontinent. Themost frequent mutation in the â globin gene detected inthis study is the Mediterranean CD39(C→T) mutationwith an allele frequency of 32.1%. This mutation hasrecently been reported to have an even higher allelefrequency in Libya, where the mutation, together withIVS-1+6 (T→C) and IVS-1+110 (G→A), comprised anallele frequency of more than 90% in Libyan patients.15
The mutation IVS-1+5(G→C) is a known Asian mutationand has been detected in only one patient in this study.Interestingly, and by contrast, this mutation has recentlybeen found to have the highest allele frequency of 61% inpatients from the neighbouring Sultanate of Oman.16
The clinical presentation of â thalassaemia major in thisgroup of children was mostly severe. Most patients sufferedfrom failure to thrive leading to delayed puberty, increasedplasma volume, pallor, lethargy, haemochromatosis,hepatosplenomegaly, and jaundice. Less common presen-tations of the disease noticed in this region of the countryincluded cardiomegaly, repeated infections (such as pneu-monia, peritonitis, and meningitis), deformity of the facialbones and teeth, osteoporosis, liver cirrhosis, ascites, anddiabetes mellitus. In the clinical management of patientswith â thalassaemia, blood transfusion comes second afterbone marrow transplantation as the most delicate,laborious, and costly management of the disease. There-
fore, for the sake of simplification, the presentation of thedisease in this study is classified as severe, moderate, ormild if the number of annual blood transfusions needed ismore than six, between six and three, or between two andnone, respectively. Hence, 25 patients (81%) presentedwith a severe form, four patients (13%) with a moderateform, and only two patients (6%) presented with a mildform of the disease. The genotype/phenotype analysis indi-cated that the mutations CD39(C→T) and -1CD44(-C) inthe homozygous state were consistently associated withsevere presentation of â thalassaemia major. However weobserved variability of disease presentation from severe tomoderate and mild caused by the mutations IVS-2+1(G→A), IVS-1 3'end-25bp and -2CD8(-AA) in thehomozygous state. It is possible that the phenotypic varia-tion in patients with the same genotype may well find itsbasis in the number of active á globin genes. Geneticanalysis of the á globin gene has not been performed in thisstudy. In two patients who are carriers of the mutationsIVS-1+5(G→C) and IVS-1+5(G→T), known to cause â+
thalassaemia, the mutations apparently did not confer pro-tection against the adverse phenotypic expression of thedisease. A possible explanation is that the two mutationsare encountered here in the compound heterozygous statewith CD39(C→T), which is known to cause a severe formof the disease.
In summary, we report the first identification of threemutations, the Turkish frameshift mutation -2CD8(-AA),the Kurdish frameshift mutation -1CD44(-C), and theMediterranean/Black mutation IVS-1+5(G→T) in pa-tients with â thalassaemia major from the eastern region ofSaudi Arabia. Documentation of the spectrum of âthalassaemia mutations could facilitate national screeningand educational programmes which would be importantwith respect to the problem of the haemoglobinopathies inthis region.
We are grateful to Dr Samia Flimban of the Maternity and Children Hospital inDammam for supporting this study. Our thanks are extended to Miss MichaelaFinsel for her assistance in the preparation of the manuscript. We gratefullyacknowledge the financial sponsorship of this work by the Alexander vonHumboldt-Stiftung, Bonn (Bad Godesberg), Germany.
EL-HARITH A EL-HARITHWOLFGANG KÜHNAU
JÖRG SCHMIDTKEMANFRED STUHRMANN
Institute of Human Genetics, Medical School of Hannover, Carl-NeubergStrasse 1, D-30623 Hannover, Germany
ZAKI NASSERALLAH
Paediatric Department, Qatif Central Hospital, Qatif, Saudi Arabia
ABDALLAH AL-SHAHRI
Paediatric Department, Maternity and Children Hospital, Dammam,Saudi Arabia
1 El-Hazmi MAF, Warsy AS. Genetic disorders among Arab populations.Saudi Med J 1996;17:108-23.
2 El-Hazmi MAF. The distribution and nature of haemoglobinopathies inArabia. In: Winter WP, ed. Haemoglobin variants in human populations. BocaRaton, Florida: CRC Press, 1987:65-77.
Table 2 The genotype, mean clinical indices, and disease presentation of Saudi children diagnosed with â thalassaemia major
Patient No GenotypeMean HbA2 (%)
Mean age atdiagnosis (y)
Mean annual No ofblood transfusions Disease presentation
3 Hussein R, Temtamy SA, El-Beshlawy A, et al. Molecular characterization ofâ-thalassemia in Egyptians. Hum Mutat 1993;2:48-52.
4 Varawalla NY, Old JM, Sarkar R, et al. The spectrum of â-thalassemia muta-tions in the Indian sub-continent: the basis for prenatal diagnosis. Br JHaematol 1991;78:242-7.
5 Quaife R, Al-Gazali L, Abbes S, et al. The spectrum of â-thalassemia muta-tions in the UAE national population. J Med Genet 1994;31:59-61.
6 Trecartin RF, Liebhaber SA, Chang JC, et al. â°-thalassemia in Sardinia iscaused by a nonsense mutation. J Clin Invest 1981;68:1012-17.
7 Orkin SH, Sexton JP, GoV SC, Kazazian HH, Jr. Inactivation of an acceptorRNA splice site by a short deletion in â-thalassemia. J Biol Chem 1983;258:7249-51.
8 Orkin SH, GoV SC. Nonsense and frameshift mutation in â-thalassemiadetected in cloned â-globin genes. J Biol Chem 1981;256:9782-4.
9 Chibani J, Vidaud M, Duquesnoy P, et al. The peculiar spectrum ofâ-thalassemia genes in Tunisia. Hum Genet 1988;73:190-2.
10 Kinniburgh AJ, Maquat LE, Schedl T, et al. mRNA-deficient â-thalassemiaresults from a single nucleotide deletion. Nucleic Acids Res 1982;10:5421-7.
11 Kazazian HH Jr, Orkin SH, Antonarakis SE, et al. Molecular characteriza-tion of seven â-thalassemia mutations in Asian Indians. EMBO J1984;3:593-6.
12 Atweh GF, Wong C, Reed R, et al. A new mutation in IVS-1 of the humanâ-globin gene causing â-thalassemia due to abnormal splicing. Blood 1987;70:147-51.
13 Doerk T, El-Harith EA, Stuhrmann M, et al. Evidence for a common ethnicorigin of cystic fibrosis mutation 3120+1G→A in diverse populations. AmJ Hum Genet 1998;63:656-62.
14 Jasim N, Merghoub T, Pascaud O, et al. Molecular basis of â-thalassemiaand G-6-PD deficiency in Bahrain. Proceedings of the symposium on geneticdiseases in Arab population - a wealth of information, King Saud University,Riyadh, 23-24 November 1997:39.
15 Marwan MM, Felice AE. Molecular analysis of â-thalassemia, sickle cellanaemia and Hb S-â-thalassemia in patients from Libya. Proceedings of thesymposium on genetic diseases in Arab population - a wealth of information, KingSaud University, Riyadh, 23-24 November 1997:44.
16 Krishnamoorthy R, Darr S, Hussein HM, Merghoub T . Recent moleculardata on the spectrum of â-thalassemia mutations in Oman: a particularmention about Hb Dhofar. Proceedings of the symposium on genetic diseases inArab population - a wealth of information, King Saud University, Riyadh,23-24 November 1997:58.
J Med Genet 1999;36:937–938
Rapid screening for the most commonâ thalassaemia mutations in south eastAsia by PCR based restrictionfragment length polymorphism analysis(PCR-RFLP)
EDITOR—The heterogeneity of the molecular lesions whichunderlie the failure of erythropoietic cells to synthesisenormal haemoglobin in â thalassaemia1 is a complicatingfactor in its molecular diagnosis. However, although morethan 100 diVerent mutations have been identified, mostlysingle base substitutions or small deletions and insertionsin the â globin gene, in many populations the bulk of âthalassaemia is caused by a population specific spectrum ofonly a small number of mutations.1 The strategy for thedetection of mutations in patients, therefore, normallyinvolves screening in the first instance for a small numberof mutations that are the most common for the populationconcerned.
Two of the most commonly used screening proceduresare dot blot or reverse dot blot hybridisation2 and theamplification refractory mutation system (ARMS).3 Whilethese procedures are satisfactory in the hands of the moreexperienced specialised laboratories, the exacting condi-tions required for the performance of allele specifichybridisation or PCR amplification steps have made themless reproducible in less advanced laboratories.4 5 In southeast Asia, where in some regions the frequency of âthalassaemia can be as high as 10%,6 7 many laboratories inmedium sized provincial hospitals and universities are suit-ably equipped for routine molecular biology laboratory
testing, but do not have the skill to comply with the strictconditions demanded by the above procedures, a situationwhich is perhaps shared by many other countries in whichâ thalassaemia is common. We have endeavoured,therefore, to develop a more robust and accuratealternative method for the detection of â thalassaemiamutations in south east Asia, based on restriction endonu-cleases that recognise naturally occurring or PCR gener-ated restriction sites associated with the â thalassaemiamutations. The approach is widely used in the detection ofpathological mutations in mitochondrial DNA8 and hasalso been applied in the molecular diagnosis of âthalassaemia.9
A strategy has been devised to allow rapid detection ofnine of the most common mutations in south east Asiawhich requires the amplification of two segments only ofthe â globin gene. The mutations are at positions IVS-1nt5, IVS-1 nt1, codon 26, codon 15, codon 17, codon 19,codon 30, IVS-1 nt2, and codon 41-42 of the â globin gene(fig 1 (top), table 1), which together account for around70-90% of â thalassaemia mutations in most populationsof south east Asia.6 7 10 The PCR amplifications use primersets TLF62028-TLR62320 and TLF62392-TLR62703.Primer TLR62320 includes a G at the position equivalentto nt8 of intron 1 instead of the normal A to create aGCTAGC site for Cac8I in the presence of the IVS-1 nt5G>C mutation. A Cac8I site is also created in the presenceof the IVS-1 nt2 T>C mutation which represents less than1% of the â thalassaemia alleles in Indonesia. Themutations G to T at IVS-1 nt1, G to A at codon 26, and Ato G at codon 19 abolish the natural occurring sites forBslI, MnlI, and MaeII, respectively. The mutations G to Aat codon 15, A to T at codon 17, and G to C at codon 30create sites for SfcI, BfaI, and Bsp1286I, respectively. Thedetection of the 4 bp deletion of codon 41-42 is essentially
Table 1 PCR primers and restriction endonucleases used in the detection procedure
Region Primer sets* Mutation Restriction endonuclease
*The primers used were: TLF62028-5'ACCTCACCCTGTGGAGCCAC3' (common C in Old et al3); TLR62320-5'CTATTGGTCTCCTTAAACCTGTCTTGTAACCTTGCTA3'; TLF62392-5'TATTTTCCCACCCTTAGGCTGCTGGTGGTCTACCCTT GGACCCAGAGGTC3'; TLR62703-5' CCCCTTCCTATGACATGAACTTAA 3'; modifications in nucleotide sequence are indicated by bold letters.†Bold numbers indicate fragments of distinguishable sizes.
as described by Chang et al9; a C has been introduced at thesecond position of codon 41 in the sequence of primerTLF62392, which together with the 4 base deletion createsa TCGA site for TaqI.
As shown in fig 1(A-D) the method produces unambigu-ous results. Thus, for the detection of the IVS-1 nt5 muta-tion, the normal allele (indicated by a 293 bp undigestedproduct) can be distinguished readily from the mutantallele (a 257 bp digested fragment). Heterozygosity couldbe easily detected (fig 1A). Similarly, the 122 bp fragment
of the mutant allele could be distinguished from the 60 and62 bp fragments of the normal allele among the digestionfragments of MnlI in the detection of the HbE codon 26mutation (fig 1B). Definitive electrophoretic patterns werealso obtained in the detection of IVS-1 nt1 (fig 1C), codon41-42 (fig 1D), codon 15 (fig 1E), codon 19 (fig 1F), andIVS-1 nt2 (fig 1G) mutations. Similar results wereobtained in the detection of codon 17 and codon 30 muta-tions (data not shown). In all the above cases, thePCR-RFLP results agreed with those of ARMS and wereconfirmed by DNA sequencing.
The detection procedure described here does not involvean allele specific hybridisation or an allele specific PCRamplification step and is thus less prone to non-specificreactions which could lead to false positive/negative results.We have applied our new procedure to the detection of theunderlying mutations in a number of diYcult samples inwhich ARMS gave ambiguous results. In all cases thePCR-RFLP method proved to be more reliable and gavedefinitive identification of the underlying mutation, asconfirmed by DNA sequencing data (data not shown). Thenine mutations included here account for 70% of âthalassaemia alleles in Thailand,6 90% in Malaysia,6 53%in India,6 and 68-90% in Indonesia7 10 depending on theethnic population. The procedure, therefore, is suitable asthe front line screening for the molecular diagnosis of âthalassaemia in south east and perhaps also in eastern andsouthern Asia.
This work was supported by the RUT grant No IIIo/13/I/-/IPD from theNational Research Council (Indonesia) to Iswari Setianingsih, and grant in aidsfrom PT Krakatau Steel and PT Inti through the Agency for Strategic Industries(Indonesia).
PATCHARIN PRAMOONJAGOALIDA HARAHAP
RATNA AGUNG TAUFANIISWARI SETIANINGSIH
SANGKOT MARZUKI
Eijkman Institute for Molecular Biology, Jl Diponegoro 69, Jakarta10430, Indonesia
ALIDA HARAHAPDepartment of Clinical Pathology, University of Indonesia, Salemba 6,Jakarta 10430, Indonesia
1 Weatherall DJ. The thalassaemias. In: Stamatoyannopoulos G, NienhuisAW, Majerus PW, Varmus H,eds. The molecular basis of blood diseases. 2nded. Philadelphia: Saunders, 1994:157.
2 Rady MS, BaYco M, Khalifa AS, et al. Identification of Mediterraneanbeta-thalassemia mutations by reverse dot-blot in Italians and Egyptians.Hemoglobin 1997;21:59-69.
3 Old JM, Varawalla NY, Weatherall DJ. The rapid detection and prenataldiagnosis of â-thalassaemia in the Asian Indian and Cypriot populations inthe UK. Lancet 1990;336:834-7.
4 Orou A, Fechner B, Utermann G, Menzel HJ. Allele-specific competitiveblocker PCR: a one-step method with applicability to pool screening. HumMutat 1995;6:163-9.
5 Chang JG, Lu JM, Huang JM, Chen JE, Liu HJ, Chang CP. Rapid diagno-sis of â-thalassemia by mutagenically separated polymerase chain reaction(MS-PCR) and its application to prenatal diagnosis. Br J Haematol1995;91:602-7.
6 Thein SL, Winichagoon P, Hesketh C, et al. The molecular basis ofâ-thalassemia in Thailand: application to prenatal diagnosis. Am J HumGenet 1990;47:367-75.
7 Setianingsih I, Williamson R, Marzuki S, Harahap A, Tamam M, Forrest S.Molecular basis of â-thalassemia in Indonesia: application to prenatal diag-nosis. Mol Diag 1998;3:11-20.
8 Lertrit P, Noer AS, Jean-Francois MJ, et al. A new disease-related mutationfor mitochondrial encephalopathy lactic acidosis and strokelike episodes(MELAS) syndrome aVects the ND4 subunit of the respiratory complex I.Am J Hum Genet 1992;51:457-68.
9 Chang JG, Chen PH, Chiou SS, Lee LS, Perng LI, Liu TC. Rapid diagno-sis of beta-thalassemia mutations in Chinese by naturally and amplifiedcreated restriction sites. Blood 1992;80:2092-6.
10 Lie Injo LE, Cai SP, Wahidiyat I, et al. â-thalassemia mutations in Indonesiaand their linkage to â-haplotypes. Am J Hum Genet 1989;45:971-5.
Figure 1 Detection of the common â thalassaemia mutations of southeast Asia by PCR-RFLP. (Top) Detection strategy; see table 1 for details.(Bottom) Results for the detection of (A) IVS-1 nt5, (B) codon 26, (C)IVS-1 nt1, (D) codon 41/42, (E) codon 15, (F) codon 19, and (G)IVS-1 nt2 mutations; lane a, uncut PCR product; lane b, patienthomozygous for the â thalassaemia mutation; lane c, heterozygote for the âthalassaemia mutation; and lane d, normal subject.
Clinical and molecular findings in apatient with a deletion on the long armof chromosome 12
EDITOR—Reports of congenital abnormalities resultingfrom deletions on the long arm of chromosome 12 arerare.1–4 A number of genes have been mapped to 12q, whichinclude a gene for Noonan syndrome (NS) in somefamilies5 and the gene for insulin-like growth factor I(IGF-I).6 The patient presented here was referred forassessment regarding the diagnosis of NS in view of hershort stature, dysmorphic facies, and developmental delayand was found to have a de novo interstitial deletion onchromosome 12q. Despite some of the similar clinicalfindings to NS in this patient, molecular analysis showedthat the deletion mapped approximately 17 cM centro-meric to the critical region for NS. To our knowledge, therehave been no similar cases of interstitial deletion reportedon chromosome 12q.
The patient, a 15 year old girl was the first born child toa 23 year old mother and an unrelated 21 year old father.Her younger brother, aged 11 years, was fit and well. Thefamily history was unremarkable. The pregnancy was nor-mal. She was born at term as an emergency caesarean sec-tion for fetal distress. Her birth weight was 2860 g. She wasadmitted to the special care baby unit for 24 hours forobservation. The neonatal period was complicated bypyloric stenosis for which surgery was carried out on day16. During the postoperative period, she developed severecandidiasis and lactose intolerance. At the age of 1 year shewas challenged with milk and seemed to have recoveredfrom the intolerance. Developmentally she was able to situnsupported at the age of 8 months and walked independ-ently at the age of 2 years. Her speech was delayed and shewas unable to say two words together until the age of 3years. She has had no visual or hearing problems. Herheight has always been below the 3rd centile. At the age of12 years, just as she was entering puberty, she wasdiagnosed as having growth hormone deficiency. Her peakgrowth hormone level on glucagon stimulation test was 7mU/l. She was started on treatment with growth hormoneand she showed a good response with sustained growthover a three year period. She had no known cardiac abnor-malities. She suVered from recurrent nose bleeds but therewas no other history of bleeding problems. Physical exam-ination at the age of 15 years indicated that her height andweight were just below the 3rd centile and her headcircumference was on the 3rd centile. She had pectusexcavatum, but her nipples were not widely spaced. Shehad bilateral ptosis, hypertelorism, and low set and poste-riorly rotated ears (fig 1). Her hair was thin, fine, andsparse. Her posterior hairline was trident and low. Her neckwas short with mild webbing. Cardiovascular examinationshowed a grade II ejection systolic murmur at the secondleft intercostal space. Echocardiogram showed normaldimensions and arrangements of the cardiac chambers andheart valves. The velocity of blood flow in the proximalpulmonary artery was slightly increased and this wasthought to be the origin of the systolic murmur.Neurological examination was normal, but she hadrelatively poor coordination. Coagulation studies, includ-ing intrinsic clotting factor assays, were normal. Growthhormone secretion was normal with a peak response of40.7 mU/l during the insulin tolerance test. Basal levels ofIGF-I were normal (IGF-I 371 ng/ml, normal range70-420 ng/ml).
Chromosome analysis by GTG banding was carried outon peripheral lymphocyte cultures. This showed aninterstitial deletion of chromosome 12 (fig 2), karyotype46,XX,del(12)(q21.2q23.2). Parental chromosomes werenormal. FISH studies were not performed.
Venous blood was sampled from the patient, her normalbrother, and her unaVected parents and genomic DNA wasisolated. Seventeen microsatellite markers (D12S291,D12S96, D12S90, D12S335, D12S313, D12S80,D12S92, D12S337, D12S88, D12S81, D12S82,D12S101, D12S218, D12S346, IGF-I, D12S78, D12S84)of chromosome 12q were analysed. PCR amplification of50 ng genomic DNA was performed in a 15 µl reactionmixture containing 30 ng of each primer, 200 µmol/ldNTPs, 50 mmol/l KCl, 1.5 mmol/l MgCl2, 10 mmol/lTris-HCl, 0.01% gelatin, 0.1% Triton X-100, and 0.1 U
Figure 1 Photographs of the patient at the age of 15 years showing(A) hypertelorism and ptosis and (B) low set, posteriorly rotated ears.(Photographs reproduced with permission.)
Figure 2 Chromosomes 12 of the patient showing her normal (left) anddeleted (right) chromosomes 12 and an ideogram of chromosome 12(centre). The breakpoints are indicated by arrows on the ideogram.
SuperTaq DNA polymerase (HT Biotechnology Ltd). In allreactions, 30 cycles (one minute at 94°C, two minutes at55°C, and two minutes at 72°C with a final seven minuteelongation step) were carried out in an automated thermalcycler (Hybaid). The amplified products were separated byelectrophoresis on 10% polyacrylamide gels. The gels weresilver stained and dried. HaploinsuYcieny of markersD12S337, D12S88, D12S82, and D12S101 was found.Haplotype analysis indicated that the deletion was paternalin origin and mapped between markers D12S313 andD12S218. The deletion represented a minimum region ofabout 18 cM positioned approximately 17 cM proximal tothe critical region for NS (fig 3).
We have reported a girl with short stature, dysmorphicfacies, developmental delay, and a de novo interstitialdeletion on chromosome 12q. Interstitial deletions of chro-mosome 12q are rare. There have only been three previouscases reported to our knowledge with deletions on chromo-some 12q, a male with a deletion (12)(q13.3q21.1),1 afemale with a deletion (12)(q15q21.2),2 and a male with adeletion (12)(q12q13.12).4 A female with a derivative chro-mosome 9 and a recombinant chromosome 12 resultingfrom a balanced complex rearrangement involving chromo-somes 8, 9, and 12 was suspected of having a submicroscopicdeletion of chromosome 12q.3
Common characteristics between our case and thereported cases were hypertelorism,1 3 4 apparently low setears,1–4 posteriorly rotated ears,1 3 sparse, fine hair,2 4 devel-opmental delay,1–4 growth retardation, and pectus deform-
ity (table 1).3 4 From the UK cytogenetic databases, a malepatient with a deletion on chromosome 12,del(12)(q21.33q24.1), had been seen by the clinicalgeneticists in Edinburgh. He was found to have markeddevelopmental delay, short stature, hydrocephalus, promi-nent eyes, and long, thin fingers. He did not have any facialfeatures of NS and no congenital heart defect (Dr DavidFitzpatrick, Department of Clinical Genetics, WesternGeneral Hospital, Edinburgh, personal communication).Isolated deficiency of IGF-I, the gene for which has beenmapped to chromosome 12q22-24.1,6 has been found inpygmies of the Central African Republic.8 In view of ourpatient’s short stature, dysmorphic facies, and increasedpulmonary valve blood flow, it was suspected that the dele-tion might involve either the IGF-I gene or the criticalregion for NS on chromosome 12q. Analysis using micro-satellite markers showed that the deletion was proximal tothe critical region for NS and also proximal to the IGF-Igene (fig 3), making either scenario unlikely. Some of thegenes which lie in the area deleted are shown in fig 3. Thepatient reported by Tonoki et al4 had features similar tothose found in NS and the authors suggested that thedeleted region in their patient may contain a gene for NS.Similarly, the deleted segment in our patient may also con-tain a gene for NS. Since the deletions in these patients donot overlap, it would mean that three genes responsible forNS would have to be situated on chromosome 12q, a con-siderable distance apart. This seems an unlikely event.Although the deletion in our patient lies outside the critical
Figure 3 Results of analysis of microsatellite markers in the patient showing (A) genetic map of chromosome 12q with results of analysis of microsatellitemarkers (right) and their distance apart (left),7 and (B) haplotype in the patient and her family; the thick black bar represents area deleted. The names ofthe genes in the area deleted are given in Krauter et al.7 DEL=deleted, U=uninformative, HET=heterozygous. IFNg=interferon gamma, RAP1B=rasrelated protein, ATP2B1=human PMCA1, DCN=decorin, BTG1=B cell translocation gene 1, HAL=histidine ammonia lyase, IGF-I=insulin-like growthfactor I.
region for NS, it is possible that genes in this region mayregulate genes in the NS critical region. Interestingly, ourpatient has a relatively mild clinical phenotype despite thesize of the deletion. This suggests that the region deleted inthis patient may only contain a small number of criticaldevelopmental genes. Alternatively, the eVects of the genesdeleted in this patient may be compensated for by the cor-responding genes present on the non-deleted chromosome12.
We thank the family for all their help with this study. We are most grateful for thefinancial support from the Birth Defects Foundation and the British HeartFoundation.
ANGELA F BRADYMADIHA M ELSAWIC RUTH JAMIESON
KAREN MARKSSTEVE JEFFERY
MICHAEL A PATTON
Medical Genetics Unit, St George’s Hospital Medical School, LondonSW17 ORE, UK
LILY MURTAZA
Department of Child Health, St John’s Hospital, Chelmsford CM2 9BG,UK
MARTIN O SAVAGE
Paediatric Endocrinology Section, St Bartholomew’s Hospital, London,EC1A 7BE, UK
Correspondence to: Dr Brady, Kennedy-Galton Centre, Northwick Park andSt Mark’s NHS Trust, Watford Road, Harrow, Middlesex HA1 3UJ, UK
1 Meinecke P, Meinecke R. Multiple malformation syndrome including cleftlip and palate and cardiac abnormalities due to an interstitial deletion ofchromosome 12q. J Med Genet 1987;24:187.
3 Masuno M, Asano J, Yasuda K, Kondo T, Orii T. Balanced complexrearrangement involving chromosomes 8, 9 and 12 in a normal mother,derivative chromosome 9 with recombinant chromosome 12 in her daugh-ter with minor anomalies. Am J Med Genet 1993;45:65-7.
4 Tonoki H, Saito S, Kobayashi K. Patient with del(12)(q12q13.12) manifest-ing abnormalities compatible with Noonan syndrome. Am J Med Genet1998;75:416-18.
5 Jamieson CR, van der Burgt I, Brady AF, et al. Mapping a gene for Noonansyndrome to the long arm of chromosome 12. Nat Genet 1994;8:357-60.
6 Morton CC, Byers MG, Nakai H, Bell GI, Shows TB. Human genes forinsulin-like growth factors I and II and epidermal growth factor are locatedon 12q22-24.1, 11p15, and 4q25-q27, respectively. Cytogenet Cell Genet1986;41:245-9.
7 Krauter K, Montgomery K, Yoon SJ, et al. A second-generation YAC contigmap of human chromosome 12. Nature 1995;337:321-33.
8 Merimee TJ, Zapf J, Froesch ER. Dwarfism in the pygmy: an isolated defi-ciency of insulin-like growth factor I. N Engl J Med 1981;305:965-8.
J Med Genet 1999;36:941–943
Amelogenesis imperfecta, sensorineuralhearing loss, and Beau’s lines: a secondcase report of Heimler’s syndrome
EDITOR—Hearing loss owing to genetic causes has areported prevalence of 1 in 1000 births and among these15-30% are associated with other abnormalities, althoughonly a small number are associated with oral and dentaldisorders.1 Heimler et al2 reported two sibs with acombination of sensorineural hearing loss, amelogenesisimperfecta, and nail abnormalities (McKusick No234580). We describe a further case here and extend thephenotypic spectrum of this syndrome.
A 12 year old girl presented with a combination of uni-lateral sensorineural deafness and amelogenesis imper-
fecta. She was born at term following a normal pregnancyand had a birth weight of 3200 g (50th centile). There wasno consanguinity within the family; her mother was aged25 and her father was aged 29 at the time of conception.She had no major illnesses in the first years of life and allher developmental milestones were achieved within normallimits. At the age of 7, it was discovered that she had uni-lateral hearing loss and subsequent investigation showedthat she had profound sensorineural deafness on the left,hearing on the right being normal. At the time her motherfelt that she may have had reduced hearing in the left earfor about two years before the diagnosis was made. At theage of 8 years, there was evidence of extensive enamel dis-coloration in her permanent dentition and amelogenesisimperfecta was diagnosed. Her primary dentition wasreported as having erupted on time and the remaining pri-mary teeth had a normal appearance. There was no historyof intellectual impairment and she was doing well in main-
Table 1 Clinical features reported in patients with interstitial deletions of chromosome 12q
Clinical feature Meinecke and Meinecke1 Watson et al2 Masuno et al3 Tonoki et al4 Present patient
stream education. Her family history was unremarkable;her parents were unrelated and there was no history ofsensorineural deafness or enamel defects on either side ofthe family. She has one younger brother who is unaffected.
On examination she was on the 10th centile for heightand the 25th centile for weight. She had a fair complexion(fig 1) and examination of her nails showed multiple trans-verse grooves (Beau’s lines) on several fingers and toes (fig2A, B). There was no evidence of leuconychia. Examina-tion of her teeth showed dental overcrowding and extensiveenamel discoloration (fig 3). Ground sections of perma-nent premolars, extracted for orthodontic purposes,showed dentine and enamel of normal thickness. It seemslikely therefore that the amelogenesis imperfecta is ahypomineralised rather than a hypoplastic variant.
The combination of sensorineural hearing loss, amelo-genesis imperfecta, and nail abnormalities was firstreported by Heimler et al.2 They suggested that thesyndrome could be the result of a single gene aVectingderivatives of the ectodermal tissue because the abnormali-ties described have a common embryological origin in theectoderm. This theory is made more probable by thereporting of a second case here.
Heimler et al2 described two sibs who both had profoundbilateral sensorineural hearing loss, amelogenesis imper-fecta of the permanent dentition, and Beau’s lines. Inherit-ance of this syndrome was postulated to be autosomalrecessive and our case does not allow us to draw any furtherconclusions on this. To date, no other cases of Heimler’ssyndrome have been identified (A Heimler, personal com-munication). The case described here had unilateralsensorineural hearing loss, a hypomineralised form ofamelogenesis imperfecta, and Beau’s lines and it is likely
that this is a similar association to the originally describedsyndrome, which extends the phenotypic spectrum (table1).
It is interesting that the hearing loss was diagnosed at arelatively late age and the mother’s observations suggestthat it had previously been normal. In the original report,deafness was diagnosed in the second sib at the age of 21⁄2years, hearing having been normal for the first two years. Itthus seems that the hearing loss in this condition is notcongenital.
Amelogenesis imperfecta is classified according to thepredominant clinical and radiographic appearance of theenamel defect and on the mode of inheritance of the trait.3
The amelogenesis imperfecta described here is unusual inthat only the permanent teeth are aVected, which makesnon-genetic causes less likely. The diagnosis of hypoplasticamelogenesis imperfecta in the original case report was
Table 1 A comparison of phenotypes between the originally described cases and a new case of Heimler’s syndrome
Original case (sib 1) Original case (sib 2) Case reported here
Sensorineural hearing loss Bilateral, diagnosed at 18 months Bilateral, diagnosed at 21⁄2 y Unilateral, diagnosed at 7 yPrimary dentition Normal Normal NormalPermanent dentition Enamel hypoplasia Enamel hypoplasia Enamel hypomineralisationNail abnormalities Punctate leuconychia, Beau’s lines Punctate leuconychia, Beau’s lines Beau’s lines
Figure 1 The 12 year old girl with amelogenesis imperfecta, unilateralhearing loss, and nail abnormalities.
Figure 2 (A, B) Transverse grooves (Beau’s lines) on finger nails inindex case.
Figure 3 Permanent dentition in index case illustrating extensive enameldiscoloration of incisors, canines, and premolars.
made on radiographic evidence and a histopathologicalexamination was not performed. The hypomineralisedamelogenesis imperfecta in our case illustrates the variationin phenotype that can occur in this condition.
Beau’s lines are transverse lines across the nails that canarise from severe illnesses, such as sepsis, AIDS, or bullouspemphigoid, and can also be physiological in menstruatingwomen.4 They can be quite subtle as in this case where theyonly became apparent once nail varnish had been removed.Our patient had no evidence of punctate leuconychia whichwas described in the original family, which could suggest thatthis feature may have been an incidental finding in the origi-nal report, as it is a common normal variant. Our case alsoemphasises the importance of careful examination of thenails in patients with a combination of impaired hearing,which may only be unilateral, and dental pathology.
MARC TISCHKOWITZCATHERINE CLENAGHAN
SALLY DAVIES
Institute of Medical Genetics, University Hospital of Wales, Heath Park,CardiV CF4 4XW, UK
LINDSAY HUNTERJOHN POTTS
University of Wales College of Medicine Dental School, Heath Park,CardiV CF4 4XY, UK
SENNO VERHOEF
MGC Department of Clinical Genetics EE2422, Erasmus UniversityRotterdam and Academic Hospital Rotterdam Dijkzigt, PO Box 1738,3000 DR Rotterdam, The Netherlands
1 Gorlin RG, Cohen MM. Epidemiology, etiology and genetic patterns. In:Gorlin RG, Toriello HV, Cohen MM, eds. Hereditary hearing loss and itssyndromes. Chapter 3. Oxford: Oxford University Press, 1995:9-21.
2 Heimler A, Fox JE, Hershey JE, Crespi P. Sensorineural hearing loss, enamelhypoplasia and nail abnormalities in sibs. Am J Med Genet 1991;39:192-5.
3 Witkop CJ Jr. Amelogenesis imperfecta, detinogenesis imperfecta and den-tin dysplasia revisited: problems in classification. J Oral Pathol 1988;17:547-73.
4 Baran R, Dawber RPR. Physical signs. In: Baran R, Dawber RPR, eds. Dis-eases of the nails and their management. Oxford: Blackwell Scientific Publica-tions, 1994:50-1.
J Med Genet 1999;36:943–944
PTEN and LKB1 genes in laryngealtumours
EDITOR—PTEN, a tumour suppressor gene located inchromosome 10q23, is homologous to tyrosine and dualspecificity phosphatases with a high degree of substratespecificity. This enzymatic activity is necessary for PTEN/MMAC1 tumour suppressor function.1–4 Mutations of thisgene have been identified in many glioma, glioblastoma,prostate, kidney, and breast carcinoma cell lines and in pri-mary tumours including gliomas, and breast, thyroid, andkidney carcinomas.1 2 5 Germline mutations of the PTENgene underlie Cowden disease, an autosomal dominantdisorder associated with an increased risk of breast andthyroid cancer and possibly endometrial malignancy.Benign tumours of the intestine (hamartomas) and skin(such as trichilemmomas) also occur.6 7
LKB1, a candidate tumour suppressor gene, encodes aserine/threonine kinase which is highly homologous (84%)to Xenopus serine/threonine kinase XEEK1.8 Germlinemutations in LKB1 have been associated with Peutz-Jeghers syndrome (PJS).9 10 Jenne et al10 report that theirsuccess in identifying the gene by analysing only two can-didate sequences was based on strong linkage disequilib-rium. This is surprising, as while linkage disequilibrium isa powerful tool in disease gene identification, no linkagedisequilibrium has been reported in PJS and the patientstypically display diVerent mutations.9 10 PJS is character-ised by hamartomatous intestinal polyposis, mucocutane-ous pigmentation, and increased risk of cancer of multipleorgan systems.6 LKB1 germline mutations are typically ofinactivating nature, often causing truncation of the proteinproduct.9–12 Genes involved in hereditary cancer syndromesare often targets of somatic mutations. In the case of theLKB1 gene, several studies, with the exception of one studyon colorectal cancer,13 have reported a low frequency ofLKB1 somatic mutations in colorectal, testicular, breast,and gastric cancers.14–16
The aetiology of laryngeal cancer is considered to bemultifactorial and genetic alterations are likely to play arole in it. The PTEN and LKB1 genes have been recentlycharacterised and appear to cause somewhat similarphenotypes when mutated in the germline.6 Because oftheir tumour suppressor function, both PTEN and LKB1
are possible candidates as genes which could be involved inlaryngeal tumorigenesis. To test this hypothesis, we evalu-ated the role of PTEN and LKB1 somatic mutations inmalignant and premalignant laryngeal tumours.
This study was based on laryngeal tumour specimenscollected from 16 patients treated at Helsinki UniversityCentral Hospital, Department of Otorhinolaryngology andHead and Neck Surgery. Thirteen of these patients hadlaryngeal carcinoma and three had laryngeal papilloma. Ofthe patients with papilloma, one had juvenile onset and twohad adult onset disease. One of these adult onset patientshad a malignant transformation of his laryngeal tumourinto epidermoid carcinoma (table 1). The tumour sampleswere resected during operations and fresh frozen. A 5 µmsection of each tumour was stained with haematoxylin toensure that the sample contained at least 60% of tumourtissue. DNA was extracted from the laryngeal tumoursamples with phenol-chloroform according to the standardprocedure.
The nine exons of the PTEN gene were sequenced bymeans of nested primers designed within the flankingintronic sequence. PCR conditions and primers have beendescribed previously.7 17 After PCR, 5 µl of each ampliconwas run on a 2% agarose gel to verify the specificity of thePCR reaction. The rest was purified with QIAquick PCRpurification kit (QIAGEN GmbH, Hilder, Germany).
Table 1 Clinical data of the patients
Patients Age Sex Smoking Diagnosis* TNM classification
1 59 M Yes Laryngeal carcinoma T3N1M02 81 M Yes Laryngeal carcinoma T3N0M03 70 M Yes Laryngeal carcinoma T3N0M04 85 M Yes Laryngeal carcinoma T4N0M05 71 M Yes Laryngeal carcinoma T2N0M06 47 M No Laryngeal carcinoma T2N2M17 65 M Yes Laryngeal
carcinoma†T3N0M0
8 73 M Yes Laryngeal carcinoma T2N0M09 75 M Yes Laryngeal carcinoma T2N0M010 61 M Yes Laryngeal carcinoma T4N0M011 62 M Yes Laryngeal carcinoma T4N0M012 76 M Yes Laryngeal carcinoma T3N1M013 61 M Yes Laryngeal carcinoma T3N0M014 62 M Yes Laryngeal
Direct sequencing of PCR products was performed withthe ABI PRISM Dye Terminator or ABI PRISMdRhodamine cycle sequencing kits (Perkin-Elmer AppliedBiosytems Division, Forster City, CA). Cycle sequencingproducts were analysed on an Applied Biosytems model377 DNA sequencer (PE/ABI) or 310 Genetic Analyzer(PE/ABI).
Mutation analysis of the LKB1 gene was performed bysingle strand conformation polymorphism (SSCP) analy-sis. Primer sequences and PCR conditions used in thisstudy have been described previously.18 After PCR, 5 µl ofeach reaction product was mixed with 5 µl of denaturingloading buVer (98% formamide, 20 mmol/l EDTA, 0.05%bromphenol blue, and 0.05% xylene cyanol), denatured forfive minutes at 94°C, and subjected to electrophoresis on0.4 mm × 30 cm × 45 cm gels containing 0.6 × mutationdetection enhancement solution (AT Biochem, Malvern,PA) and 0.6 × TBE buVer. Electrophoresis was conductedat 4 W overnight. PCR fragments were visualised in gels bysilver staining. In order to validate the SSCP results,amplicon of exon 1 in the LKB1 gene from all samples weredirectly sequenced (as described above).
To determine whether the PTEN and LKB1 genes aremutated in laryngeal tumours, we screened 16 laryngealtumour samples by genomic sequencing and SSCP assay,respectively. We performed direct sequencing of ampliconfor PTEN mutation analysis; no mutations in the codingsequences and exon/intron boundaries were found. Toscreen for LKB1 mutations, SSCP assay was applied inconjunction with PCR. DiVerence in mobility pattern onthe SSCP assay usually suggests a variant, the nature ofwhich should be examined by sequencing. In our handsSSCP analysis of LKB1 has always shown knownmutations and it is reasonable to assume a 70% to 80%sensitivity, which is commonly reported for the method.19
To validate the SSCP results, exon 1 was directlysequenced from each sample. No mutations in exon 1 andthe respective exon/intron boundaries were observed. Themutation detection methods used in this study would nothave detected some mutation types, such as large genomicdeletions or alterations in the promoter region. Yet thecomplete absence of changes indicates that somatic muta-tions of PTEN and LKB1 are not frequent in laryngealtumours.
The molecular events which induce laryngeal tumori-genesis, especially laryngeal carcinogenesis, are not wellcharacterised. Proto-oncogenes seem to be the target of therisk factors (cigarette smoking, alcohol abuse, ionisingradiation, and human papillomavirus infection) that arecommonly considered to be associated with laryngealsquamous cell carcinoma. Several tumour suppressorgenes have been shown to play important roles in humantumours, including head and neck cancers. Mutations ofthe p53 gene are frequent events in primary squamous cellcarcinomas of the head and neck as well as in SCCHN celllines. While many cancer genes are associated withlaryngeal tumorigenesis,20 our study suggests that PTENand LKB1 are not among them.
The authors are grateful to Annukka Nyholm, Tiiu Arumäe, and Irina Suoma-lainen for excellent technical assistance. This work was supported by a grantfrom Helsinki University Central Hospital Research Fund.
REN WEI CHEN
Department of Virology, Haartman Institute, University of Helsinki,Finland
EGLE AVIZIENYTESTINA ROTH
Department of Medical Genetics, Haartman Institute, University ofHelsinki, Finland
ILMO LEIVO
Department of Pathology, Haartman Institute, University of Helsinki,Finland
ANTTI A MÄKITIELEENA-MAIJA AALTONEN
Department of Otorhinolaryngology, Helsinki University Central Hospital,Finland
LAURI A AALTONEN
Department of Medical Genetics, Haartman Institute, University ofHelsinki, PO Box 21, FIN-00014 Helsinki, Finland
1 Li J,Yen C, Liaw D, et al. PTEN, a putative protein tyrosine phosphatase genemutated in human brain, breast, and prostate cancer. Science1997;275:1943-7.
2 Steck PA, Pershouse MA, Jasser SA, et al. Identification of a candidatetumour suppressor gene, MMAC1, at chromosome 10q23.3 that is mutatedin multiple advanced cancers. Nat Genet 1997;15:356-62.
3 Stambolic V, Suzuki A, de la Pompa JL, et al. Negative regulation ofPKB/Akt-dependent cell survival by the tumor suppressor PTEN. Cell1998;95:29-39.
4 Myers MP, Stolarov JP, Eng C, et al. P-TEN, the tumor suppressor fromhuman chromosome 10q23, is a dual-specificity phosphatase. Proc NatlAcad Sci USA 1997;94:9052-7.
5 Dahia PLM, Marsh DJ, Zheng Z, et al. Somatic deletions and mutations inthe Cowden disease gene, PTEN, in sporadic thyroid tumors. Cancer Res1997;57:1710-13.
6 Phillips RKS, Spigelman AD, Thompson JPS, eds. Familial adenomatouspolyposis and other polyposis syndromes. London: Edvard Arnold, 1994.
7 Liaw D, Marsh DJ, Li J, Dahia PLM, et al. Germline mutations of the PTENgene in Cowden disease, an inherited breast and thyroid cancer syndrome.Nat Genet 1997;16:64-7.
8 Su JY, Erikson E, Maller JL. Cloning and characterization of a novel serine/threonine protein kinase expressed in early Xenopus embryos. J Biol Chem1996;271:14430-7.
9 Hemminki A, Markie D, Tomlinson I, et al. A serine/threonine kinase genedefective in Peutz-Jeghers syndrome. Nature 1998;391:184-7.
10 Jenne DE, Reimann H, Nezu J, et al. Peutz-Jeghers syndrome is caused bymutations in a novel serine threonine kinase. Nat Genet 1998;18:38-43.
11 Mehenni H, Gehrig C, Nezu J, et al. Loss of LKB1 kinase activity in Peutz-Jeghers syndrome, and evidence for allelic and locus heterogeneity. Am JHum Genet 1998;63:1641-50.
12 Ylikorkala A, Avizienyte E, Tomlinson IPM, et al. Mutations and impairedfunction of LKB1 in familial and non-familial Peutz-Jeghers syndrome anda sporadic testicular cancer. Hum Mol Genet 1999;8:45-51.
13 Dong SM, Kim KM, Kim SY, et al. Frequent somatic mutations in serine/threonine kinase 11/Peutz-Jeghers syndrome gene in left-sided coloncancer. Cancer Res 1998;58:3787-90.
14 Avizienyte E, Roth S, Loukola A, et al. Somatic mutations in LKB1 are rarein sporadic colorectal and testicular tumors. Cancer Res 1998;58:2087-90.
15 Bignell GR, Barfoot R, Seal S, Collins N, Warren W, Stratton M. Low fre-quency of somatic mutations in the LKB1/Peutz-Jeghers syndrome gene insporadic breast cancer. Cancer Res 1998;58:1384-7.
16 Park WS, Moon YW, Yang YM, et al. Mutations of the STK11 gene in spo-radic gastric carcinoma. Int J Oncol 1998;13:601-4.
17 Marsh DJ, Roth S, Lunetta KL, et al. Exclusion of PTEN and 10q22-24 asthe susceptibility locus for juvenile polyposis syndrome. Cancer Res1997;57:5017-21.
18 Avizienyte E, Loukola A, Roth S, et al. LKB1 somatic mutations in sporadictumors. Am J Pathol 1999;154:677-81.
19 Jordanova A, Kalaydjieva L, Savov A, et al. SSCP analysis: a blind sensitivitytrial. Hum Mutat 1997;10:65-70.
20 Califano J, van der Riet P, Westra W, et al. Genetic progression model forhead and neck cancer: implications for field cancerization. Cancer Res1996;56:2488-92.
