J Med Genet 1992; 29: 563-567

-hexosaminidase splice site mutation has a highfrequency among non-Jewish Tay-Sachs diseasecarriers from the British Isles

Eleanor C Landels, Peter M Green, Ian H Ellis, Anthony H Fensom, Martin Bobrow

AbstractIn the course of defining mutations caus-ing Tay-Sachs disease (TSD) in non-Jew-ish patients and carriers from the BritishIsles, we identified a guanine to adeninechange (also previously described) in theobligatory GT sequence of the donorsplice site at the 5' end of intron 9 of thehexosaminidase a peptide gene. Of 24unrelated mutant chromosomes from 20non-Jewish subjects (15 TSD carriers,four TSD patients, and one TSD fetus),five had mutations common in the Ash-kenazi Jewish community, and 10 had theintron 9 splice site mutation. This is anunexpected result considering the di-verse origin of the population of the Brit-ish Isles. This mutation was not found in28 control UK subjects or 11 Jewish car-riers of known TSD mutations. Beforeattempting detection of unknown muta-tions, non-Jewish TSD carriers from theBritish Isles should be screened for theintron 9 donor splice site mutation aswell as those mutations which predomi-nate in the Jewish community.

Paediatric ResearchUnit, Division ofMedical andMolecular Genetics,United Medical andDental Schools ofGuy's and StThomas's Hospitals,8th Floor, Guy'sHospital Tower,London SE1 9RT.E C LandelsP M GreenI H EllisA H FensomM Bobrow

Correspondence toMs Landels.

Received 8 November 1991.Revised version accepted30 January 1992.

Tay-Sachs disease (TSD) is one of the GM2gangliosidoses, a group of neurodegenerativedisorders of autosomal recessive inheritance.In the 'classic' infantile form of TSD, onset ofsymptoms occurs at about 6 months and deathby 4 years of age. The symptoms are caused byaccumulation of the ganglioside GM2 in neur-

onal cells owing to deficient activity of thelysosomal enzyme which metabolises it, 1-

hexosaminidase A (hex A).' The frequency ofTSD carriers in the general population isapproximately 1/200, but is much higher incertain ethnic groups (for example, 1/27 inAshkenazi Jews), and carrier screening pro-grammes involving assay of serum hex A havebeen implemented in order to reduce incidenceof the disease.Hex A is composed of a and peptide

chains, coded for by genes on chromosomes 15and 5, respectively; mutations in the a peptide(HEXA) gene are the cause ofTSD. This genehas been cloned2' and several mutations caus-

ing TSD have been described. The first ofthese were found in ethnic groups having a

high incidence of TSD, namely the French-Canadian population (nearly 3% of whom arecarriers, predominantly of a 7-6 kb deletion),45and the Ashkenazi Jews (3% of whom areTSD carriers, predominantly of a 4 base pairinsertion in exon 116 and a mutation of the

intron 12 donor splice site79). More recently, aphenylalanine deletion has been described infive out of eight obligate TSD carriers fromthe Moroccan Jewish population.'0 Many dif-ferent mutations, all individually rare, havealso been described in patients from low inci-dence populations, some of which have beenreviewed by Neufeld."1We have previously assessed the frequency

of known TSD causing mutations in 41 UKAshkenazi Jewish TSD carriers'2; 33 had theexon 11 insert mutation and eight had theintron 12 splice site mutation. In the presentstudy we have attempted to identify the muta-tions in non-Jewish TSD patients and carriersfrom the UK and Eire using amplification bypolymerase chain reaction (PCR)"3 and chem-ical mismatch detection'4'5 followed by directsequencing'6 of amplified products. Thesetechniques have already been used successfullyto detect new mutations causing other geneticdiseases, for example haemophilia A'7 and B'5and Duchenne muscular dystrophy,'8 andmore recently Tay-Sachs disease. 9

MethodsThe diagnosis of TSD, or carrier status, wasestablished by enzyme assay of fibroblast sam-ples, chorionic villi sample and/or serum andleucocyte samples using previously describedmethods.20 Genomic DNA of TSD patients,carriers, and 28 UK non-Jewish control sub-jects was extracted from EDTA blood, fibro-blasts, or from cultured chorionic villi cellsusing established techniques.2' RNA wasextracted from fibroblasts or EDTA bloodusing the acid-guanidinium thiocyanate-phenol-chloroform method22 modified (byusing sodium acetate at pH 5-2 instead ofpH 4) to coprecipitate DNA with the RNA.

All of the TSD carriers (as defined by en-zyme assay) were tested for the exon 11 insertand intron 12 donor splice site mutations(which predominate in Jewish TSD carriers)using methods described elsewhere'2; subjectsnegative for these were screened for unknownmutations using chemical mismatch detection.The HEXA gene was amplified from genomicDNA by PCR in nine sections from exon 2 tothe 3' untranslated region with primersdesigned from the published cDNA sequence2(sequences available on request) and synthe-sised using a PCR-mate DNA synthesiser(Applied Biosystems). Approximately 50 nggenomic DNA were subjected to PCR amplifi-cation in a 50 gil reaction volume containing1 x PCR buffer (67 mmol/l Tris-HCl, pH 8-8,

563

on 20 April 2018 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.29.8.563 on 1 August 1992. D

ownloaded from

Landels, Green, Ellis, Fensom, Bobrow

Table 1 Results of screening for the intron 9 splice site mutation and the exon 11insert and intron 12 splice site mutations.

No positive for mutation

No of Intron 9 'Jewish' mutationsunrelatedchromosomes Source Exon 11 Intron 12

1! Obligate carrier parents 4 4 0(n= 11)

I Carrier on enzyme screening 0 0 0plus family history (n = 1)

3 Carriers on enzyme 0 0 1screening but no familyhistory (n= 3)

7 TSD patients (n=4*) 5 0 02 TSD fetus (n= 1) 1 0 0

Total 24 20 10 4 1

* Each TSD patient has two unrelated chromosomes except for one whose parents are highlyconsanguineous and therefore represents one unrelated chromosome.

166 mmol/l ammonium sulphate, 67 mmol/lmagnesium chloride, 10 mmol/l ,B-mercap-toethanol, 170 gg/ml bovine serum albumin(Sigma)), 0 5 mmol/l dNTP (Pharmacia),300 ng of each primer, and 2 5 units Taqpolymerase (Cetus). Taq polymerase wasadded after an initial denaturation step of 94'Cfor five minutes, followed by 30 cycles of 60seconds at 93'C, 30 seconds at 64'C, 90seconds at 72'C (plus an additional eightseconds per cycle); lastly the reaction wasincubated at 72'C for five minutes to ensurecomplete polymerisation of all products. Afterpurification with Geneclean II (Bio 101), thePCR products were subjected to the chemicalmismatch method with osmium tetroxide andhydroxylamine.'145 The regions of mutation orpolymorphism indicated by any mismatchbands were sequenced directly'6 using the en-zyme Sequenase (United States BiochemicalCorporation).The mutation detected using the chemical

mismatch method (on PCR products amplifiedwith primers Hexa8F and Hexol OR) wasscreened for by sequencing or restriction en-zyme digestion since it fortuitously destroys aMaeII cutting site. Purified PCR products(amplified using the primers Hexa9F andHexalOR) were digested with MaeII (Boehr-inger Mannheim Biochemica) in the bufferrecommended by the manufacturer. PurifiedPCR products of a 736 base pair (bp) section ofthe factor IX gene'6 containing a controlMaeII site were added to show whether diges-tion had been complete.

In order to determine effects of the mutationon mRNA, RNA extracted from fibroblasts ofa TSD patient homozygous for the mutationand normal RNA was subjected to 'duplex'reverse transcription and PCR amplification.Two sets of primers were used, one setdesigned from the published HEXA genecDNA sequence2 (HEXa7F and HEXab) anda second set designed from the published 3peptide (HEXB) gene cDNA sequence23(HEX08F and HEXib). Approximately 100,200, 400, and 800 ng RNA were incubatedwith 50 ng each of HEXab and HEXIb in avolume of 7-5 pl at 65'C for 10 minutes andthen transferred to an ice bath. Then 200 unitsMolony Murine Leukaemia Virus reversetranscriptase (RT) in RT buffer (both fromGibco-BRL) were added together with 25

units RNase inhibitor (Boehringer MannheimBiochemica) and 0-025 mmol/l dNTP (Phar-macia), to give a final volume of 20 pd. Thismixture was incubated at 42'C for one hourthen transferred to ice, after which 450 ng eachof primers HEXab and HEXpb and 500 ngeach of primers HEXa7F and HEXP8F wereadded. After heating to 94'C for five minutes, apremix containing 5 units Taq polymerase and5 pl 10 x PCR buffer (see above) were added togive a final volume of 50 RI. This mixture wasthen subjected to 23 cycles of PCR amplifica-tion, each cycle being one minute at 93'C, oneminute at 65'C, and seven minutes at 72'C,lastly followed by five minutes at 72'C. Elec-trophoresis of 8 p1 of the reaction products wasperformed in a 1% agarose (Ultrapure, Gibco-BRL) minigel with ethidium bromide. Reversetranscription and PCR amplification were alsocarried out with only one set of primers, speci-fic for the HEXA gene (Hexa6F andHexoxl3R). Approximately 200 to 300 ngRNA, and 30 cycles of PCR amplification,were used.

SEQUENCES OF PRIMERS USED FOLLOWED BYcDNA NUCLEOTIDE POSITIONSAmplification from genomic DNA:Hexa8F: 5'-CTGGATTACTGACTCC-

TTGC-3' (812-821).Hexa9F: 5'-GAAGTCCAACCCAGAGA-

TCC-3' (987-1006).HexalOR: 5'-AACACCTCCTGCCACA-

CCAC-3' (1130-1111).Reverse transcription and PCR amplificationfrom RNA:Hexa7F: 5'-GTCACCCACATCTACAC-

AGC-3' (688-707).Hexab: 5'-AAGGGGCTGCGTCCCCT-

GGC-3' (1679-1660).Hexp8F: 5'-CAGAAAGACCTCCTGA-

CTCC-3' (904-923).Hexpb: 5'-TAGTACAGATTGCTGTG-

GCC-3' (1708-1689).Hexa6F: 5'-TGGCATCTGGTAGATG-

ATCC-3' (607-626).Hexal3R: 5'-AGCAACTCACAGCGGA-

AGTG-3' (1523-1504).

ResultsTwenty TSD carriers or patients from the UKand Eire, of no known Jewish ancestry (table1), were screened for the mutations whichpredominate in the Jewish community; fourhad the more frequent mutation, a 4 bp inser-tion into exon 11, and one had the less frequentintron 12 donor splice site mutation. Theremaining 15 carriers and patients weresearched for unknown mutations; the results ofchemical mismatch of a section of the HEXAgene spanning exon 8 to exon 10 in eightsubjects (seven obligate TSD carriers, five UKand two Turkish, and one patient affected withjuvenile onset TSD) are shown in fig 1. Thelarger mismatch band (> 500 bp) appears, byvirtue of its size, to be positioned well withinan intron and therefore indicates a changewhich will not affect the coding sequence. The

564

on 20 April 2018 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.29.8.563 on 1 August 1992. D

ownloaded from

,B-hexosaminidase splice site mutation has a high frequency among non-Jewish Tay-Sachs disease carriers from the British Isles

1 2 3 4 5 6 7 8

FllI length 1619 bp---

500 bp

357 bp-

Figure 1 Chemical mismatch detection withhydroxylamine in a section of the HEXA gene spanningexons 8 to 10. Lanes 1, 2, 3, 4, 6: parents of non-JewishUK TSD patients; lane 5: juvenile onset UK TSDpatient; lanes 7, 8: parents of Turkish TSD patient.(Lane 4 was lost; however this subject does show the Gto A change in fig 2.).

region indicated by the 357 bp mismatch bandwas sequenced in the same eight subjects (fig2). The first base of intron 9 was changed fromguanine (G) to adenine (A) in those patientswho have the mismatch band. All five subjectsshown in fig 2 to have the mutation were

heterozygous for it since they also had thenormal G base at this position. The G to Amutation alters the obligatory GT ofthe intron9 donor splice site. (The donor splice site isnext to the 3' end of an exon, and is essentialfor 'splicing out' of introns to form themRNA.)Nineteen unrelated mutant chromosomes

from 15 UK or Eire TSD carriers and patientswho were negative for the common 'Jewish'mutations were tested for the intron 9 donorsplice site mutation using sequencing or diges-tion with MaeII restriction enzyme or both; 10were found to be positive (table 1). The par-

ents of four TSD patients included in table 1were also tested for the intron 9 donor splicesite mutation (table 2). Twenty-eight non-

Jewish controls and 12 Jewish carriers of otherknown mutations causing TSD were tested forthe intron 9 splice site mutation; all were

negative. Two Turkish TSD carriers were

found to be negative for the intron 9 donorsplice mutation (lanes 7 and 8 on figs 1 and 2).The results of reverse transcription and

PCR amplification ofRNA are shown in fig 3.In 'duplex' reactions with two sets of primers(one set specific to the HEXA gene, the otherspecific to the HEXB gene), normal RNA(lanes 2 to 5) gives amplification products ofthe correct size from both the HEXA andHEXB genes. RNA from a TSD patienthomozygous for the intron 9 splice site muta-tion (lanes 5 to 9) gives the correct size PCRproduct from the HEXB gene but no producthas amplified from the HEXA gene. In reac-

tions using one set of primers (specific to the

HEXA gene), RNA from a late infantile TSDpatient negative for the intron 9 splice sitemutation (lane 1 1) shows the correct size PCRproduct. In contrast, the PCR product of theTSD patient homozygous for the intron 9mutation (lane 12) is barely visible.

DiscussionOver 20 mutations causing TSD have nowbeen described (some reviewed by Neufeld"),most being of low frequency or confined to onefamily. The exceptions to this are mutationsfound at high frequency within geneticallycircumscribed populations such as the French-Canadians45 and the Ashkenazi Jews.'9 Theintron 9 donor splice site mutation describedin this study, which has also been detected byAkli et al,19 appears to be very frequent amongnon-Jewish TSD patients and carriers fromthe British Isles. Of 19 unrelated chromo-somes which had already been shown not tocarry either of the two common 'Jewish' muta-tions, 53% were found to be positive for theintron 9 donor splice site mutation. This wasan unexpected result since the populationstudied is not a genetic isolate; carriers of themutation include people of English, Scottish,Welsh, and Irish origin.The nature of the mutation itself indicates

that it is highly likely to be the cause of disease,since it is a change in the first nucleotide ofintron 9, the donor splice site. Splice sites areknown to have invariant sequence24 (GT)within their consensus sequence, any changeleading to incorrect splicing of the mRNA andthe production of aberrant, unstable mRNAspecies.25 This has already been shown to bethe case for the intron 12 donor splice sitemutation,26 causing TSD in the AshkenaziJewish community, an intron 5 donor splicesite in a Tunisian TSD patient,27 and introns 2and 4 donor splice site mutations in twoFrench TSD patients.'9The G to A change in the first base of intron

9 is also unlikely to be a polymorphism since(1) it was not found in 28 UK non-Jewishcontrols, (2) it was not found in 12 healthyJewish carriers of known TSD mutations, and(3) if this change is simply a very frequentpolymorphism then some non-TSD subjectscould be expected to be homozygous; the onlyhomozygotes found were two TSD patients(table 2). The G to A change has occurred at aCpG dinucleotide, frequent sites of mutationowing to spontaneous deamination of methyl-cytosine to thymidine.28 Since this is a G to Achange it must have arisen in the antisensestrand (rather than the sense strand whichwould result in C to T).We would speculate that mutation of the

intron 9 donor splice site would cause exon 9 tobe lost during mRNA splicing. However, thecoding sequence would continue to be in frameso that the protein produced would not beseverely truncated but simply reduced by 29amino acids (those coded for by exon 9). Pre-liminary studies using duplex reverse tran-scription and PCR amplification of RNA of apatient homozygous for the mutation showed

565

on 20 April 2018 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.29.8.563 on 1 August 1992. D

ownloaded from

Landels, Green, Ellis, Fensom, Bobrow

I 2`3X`35 6I7 8 -M 3 X r 8 - ;

ntro ,

C'

A_.;- -

_

Ai

Figure 2 Genomic DNA sequence of region indicated by 357bp mismatch band in fig 1. Lanes as in fig 1.

Table 2 Intron 9 mutations in non-Jewish TSD patients and parents.

Presence of intron 9 mutation

Mother Father Patient

Patient 1 (infantile) + + + +Patient 2 (infantile) - + + -Patient 3 (infantile) + NT + +Patient 4 (juvenile) + - + -

NT = not tested; however the parents of patient 3 are consanguineous.

Di priex

1 2 3 4` r ? 8f7

i91 bp-4EXA)

17 ) H.EXA

805 bp

Figure 3 Reverse transcription and PCR amplification. Lanes 1-9: duplex reactionusing HEXA and HEXB specific primers; lanes 10-12: reaction using only HEXAspecific primers. Lane 1: standard size marker; lanes 2-5: control normal RNA(approximately 100, 200, 400, and 800 ng RNA respectively); lanes 6-9: RNA ofTSD patient homozygous for the intron 9 donor splice mutation (approximately 100,200, 400, and 800 ng RNA respectively); lane 10: standard size marker; lane 11:RNA of TSD patient negative for the intron 9 splice mutation; lane 12: RNA ofTSD patient homozygous for the intron 9 splice mutation.

severe disruption of the splicing process; whilea PCR product of the correct size amplifiedfrom the HEXB gene, no PCR product, ofnormal or reduced size, was obtained from theHEXA gene. Reverse transcription and PCRamplification using only one set of primers,specific for the HEXA gene (from exon 6 toexon 13), produced only a trace ofnormal sizedproduct. It may be that incorrectly splicedmRNA species are formed but are unstable;further studies will be needed to determineprecisely the effects of this mutation on HEXAmRNA splicing.The high frequency of this mutation within

a population of diverse origin indicates that itmay be a very ancient mutation, or that it arosein a subgroup which has since mixed with the'general population'. We tentatively suggestthat this subgroup may have been of Celticorigin because the intron 9 splice site mutationwas found at higher frequency in unrelatedchromosomes of Irish, Scottish, or Welsh ori-gin compared to English (63% compared to31%). There is already evidence that threedifferent TSD mutations have reached a highfrequency in two populations, the AshkenaziJews'9 and the French-Canadians.45 The re-

sults of this study suggest that yet anotherdifferent mutation at the same locus was alsorelatively frequent in a third population,before recent population admixture. If so, this

may add force to the argument that someselective agent historically conferred a positiveadvantage on carriers of TSD mutations.Of the 24 unrelated chromosomes tested in

this study, the mutations of nine remainunknown; it will be interesting to see howmany other mutations are present in the popu-lation. It seems unlikely that there will be oneother very frequent mutation as in the Ashke-nazim, as the population of the British Isles isvery diverse in origin. The situation may re-semble that of cystic fibrosis where there is onepredominant mutation, AF508, which ac-counts for approximately 74% of UK cysticfibrosis carriers, and several other rare muta-

29tions.Carrier screening and prenatal diagnosis of

Tay-Sachs disease can be done very effectivelyby enzyme assay of hex A.' Rapid detection ofnew mutations causing TSD using PCRamplification, chemical mismatch, andsequencing will be useful in confirmation ofborderline diagnoses and the development ofpreimplantation diagnosis. The present workshows that TSD families from the non-Jewishpopulation of the British Isles should bescreened for the intron 9 donor splice muta-tion, as well as the mutations which predomi-nate in the Ashkenazi Jewish community,before more intense studies are undertaken todelineate their mutations.

We thank the British Tay-Sachs Foundationfor generous financial support, Tina Slade forinvaluable help with cell culture, and the manyphysicians who have sent us samples fromtheir patients. The work of the PaediatricResearch Unit is supported by the GenerationTrust and the Spastics Society.

1 Sandhoff K, Conzelmann E, Neufeld E, Kaback M, SuzukiK. The G gangliosidoses. In: Scriver CR, Beaudet AL,Sly WS, Valle D, eds. The metabolic basis of inheriteddisease. 6th ed. New York: McGraw-Hill, 1989:1807-42.

2 Myerowitz R, Piekarz R, Neufeld E, Shows TB, Suzuki K.Human [-hexosaminidase a chain: coding sequence andhomology with the [-chain. Proc Natl Acad Sci USA1985;82:7830-4.

3 Proia RL, Soravia E. Organisation of the gene encoding thehuman [-hexosaminidase a-chain. J Biol Chem1987;262:5677-8 1.

4 Myerowitz R, Hogikyan ND. A deletion involving Alusequences in the 3-hexosaminidase a-chain gene ofFrench Canadians with Tay-Sachs disease. J Biol Chem1987;262: 15396-9.

5 Hechtman P, Kaplan F, Bayleran J, et al. More than onemutant allele causes infantile Tay-Sachs disease inFrench-Canadians. Am J Hum Genet 1990;47:815-22.

6 Myerowitz R, Costigan FC. The major defect in AshkenaziJews with Tay-Sachs disease is an insertion in the gene forthe a-chain of 3-hexosaminidase. J Biol Chem1988;263: 18587-9.

7 Myerowitz R. Splice junction mutation in some AshkenaziJews with Tay-Sachs disease: evidence against a singledefect within this ethnic group. Proc Natl Acad Sci USA1988;85:3955-9.

566

.- t,%,qW.

*A*4* 4i

-1:.- dwa.,,- rs,

on 20 April 2018 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.29.8.563 on 1 August 1992. D

ownloaded from

fi-hexosaminidase splice site mutation has a high frequency among non-Jewish Tay-Sachs disease carriers from the British Isles

8 Arpai E, Dunbrille-Ross A, Maler T, et al. Identification ofan altered splice site in Ashkenazi Tay-Sachs disease.Nature 1988;333:85-6.

9 Ohno K, Suzuki K. A splicing defect due to an exon-intronjunctional mutation results in abnormal 3-hexosamini-dase a-chain mRNAs in Ashkenazi Jewish patients withTay-Sachs disease. Biochem Biophys Res Commun1988;153:463-9.

10 Navon R, Proia RL. Tay-Sachs disease in Moroccan Jews:deletion of a phenylalanine in the a-subunit of J-hexosa-minidase. Am J Hum Genet 1991;48:412-19.

11 Neufeld E. Lysosomal storage diseases. Annu Rev Biochem1991;60:257-80.

12 Landels EC, Ellis IH, Fensom AH, Green PM, Bobrow M.Frequency of the Tay-Sachs disease splice and insertionmutations in the UK Ashkenazi Jewish population. J7 MedGenet 1991;28:177-80.

13 Saiki RK, Gelfand DH, Stoffel S, et al. Primer-directedenzymatic amplification of DNA with a thermostableDNA polymerase. Science 1988;239:487-91.

14 Cotton RGH, Rodrigues NR, Campbell RD. Reactivity ofcytosine and thymine in single-base pair mismatches withhydroxylamine and osmium tetroxide and its applicationto the study of mutations. Proc Natl Acad Sci USA1988;85:4397-401.

15 Montandon AJ, Green PM, Gianelli F, Bentley DR. Directdetection of point mutations by mismatch analysis: appli-cation to haemophilia B. Nucleic Acids Res 1989;17:3347-58.

16 Green PM, Bentley DR, Reuben SM, Nilsson IM, GianelliF. Molecular pathology of haemophilia B. EMBO J1989;8:1067-72.

17 Naylor JA, Green PM, Montandon AJ, Rizza CR, GianelliF. Detection of three novel mutations in two haemophiliaA patients by rapid screening of whole essential region offactor VIII gene. Lancet 1991;337:635-9.

18 Roberts RG, Montandon AJ, Bobrow M, Bentley DR.Detection of novel genetic markers by mismatch analysis.Nucleic Acids Res 1989;17:5961-71.

19 Akli S, Chelly J, Lacorte JM, Poenaru L, Kahn A. Sevennovel Tay-Sachs disease mutations detected by chemicalmismatch cleavage of PCR-amplified cDNA fragments.Genomics 199 1;Il :124-34.

20 Landels EC, Ellis IH, Bobrow M, Fensom AH. Tay-Sachsdisease heterozygote detection: use of a centrifugal ana-lyser for automation of hexosaminidase assays with twodifferent artificial substrates. J Med Genet 1991;28:101-9.

21 Sambrook J, Fritsch EF, Maniatis T. Molecular cloning; alaboratory manual. 2nd ed. New York: Cold Spring Har-bor Laboratory Press, 1989.

22 Chomczynski P, Sacchi N. Single-step method of RNAisolation by acid guanidinium thiocyanate-phenol-chloro-form extraction. Anal Biochem 1987;162:156-9.

23 Neote K, Bapat B, Dumbrille-Ross A, et al. Characteriza-tion of the human HEXB gene encoding lysosomal f3-hexosaminidase. Genomics 1988;3:279-86.

24 Shapiro MB, Senapathy P. RNA splice junctions of dif-ferent classes of eukaryotes: sequence statistics and func-tional implications in gene expression. Nucleic Acids Res1987;15:7155-74.

25 Treisman R, Orkin SH, Maniatis T. Specific transcriptionand RNA splicing defects in five cloned ,-thalassaemiagenes. Nature 1983;302:591-6.

26 Ohno K, Suzuki K. Multiple abnormal 0-hexosaminidasea-chain mRNAs in a compound-heterozygous AshkenaziJewish patient with Tay-Sachs disease. J Biol Chem1988;34: 18563-7.

27 Akli S, Chelly J, Mezard C, Gandy S, Khan A, Poenaru L.A "G" to "A" mutation at position - 1 of a 5' splice sitein a late infantile form of Tay-Sachs disease. J Biol Chem1990;265:7324-30.

28 Coulandre C, Miller JH, Farabaugh PJ, Gilbert W. Mole-cular basis of base substitution hotspots in E coli. Nature1978;274:775-80.

29 The Cystic Fibrosis Genetic Analysis Consortium. World-wide survey of the dF508 mutation-report from thecystic fibrosis genetic analysis consortium. Am J HumGenet 1990;47:354-9.

567

on 20 April 2018 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.29.8.563 on 1 August 1992. D

ownloaded from