a strategy genetic counselling - jmg.bmj.com · 07medgenet 1992; 29: 602-607 reviewarticle...

07 Med Genet 1992; 29: 602-607

REVIEW ARTICLE

A new strategy for the genetic counselling ofdiseases of marked mutational heterogeneity:haemophilia B as a model

F Giannelli, S Saad, A J Montandon, D R Bentley, P M Green

A number of human inherited diseases are ofsuch high mutational heterogeneity that eachfamily may be expected to carry a mutation ofindependent origin. This wealth of differentnatural mutants offers excellent opportunitiesto study the molecular biology of the diseases,and to define the features essential to thefunction of the relevant genes. Conversely,such heterogeneity complicates the provisionof carrier and prenatal diagnoses by DNAbased methods and, hence, hinders progress ingenetic counselling.The diagnostic challenge presented by

diseases of high mutational heterogeneitygenerally has been met using indirect proced-ures based on the cosegregation of the detri-mental mutation and intra- or perigenic linkedmarkers. This has several drawbacks, namely:(1) it is cumbersome, as each diagnosis re-quires the analysis of a family group and, ifpaternal markers are used, also the confirma-tion of paternity; (2) it fails in a significantproportion of families because of lack of eitherinformative markers or positive family historyor key family members; (3) its accuracy islimited by the probability of meiotic re-combination between the marker(s) and thedetrimental mutation; and (4) it provides noinformation on the cause and hence on themolecular genetics of the disease.

In order to overcome the above limitationsand to take advantage of the existing wealth ofnatural mutants, it is necessary to developrapid procedures for the identification of genedefects.'We have pursued the establishment of such

methods and have applied them, in the firstinstance, to haemophilia B.24 This is a sexlinked, recessive, haemorrhagic disorderowing to defects in coagulation factor IX.Before modern therapy the genetic fitness ofmales with haemophilia B was approximately0 55 and therefore the renewal rate of haemo-philia B genes in the population pool was of theorder of 1/6 per generation.6 This ensures thatpatients in distinct families usually carry mu-tations of independent origin.For a number of different reasons we con-

sidered haemophilia B an ideal model fordeveloping strategies to meet the diagnosticchallenge and take advantage of the opportuni-ties presented by diseases of high mutational

heterogeneity. In haemophilia B there is a realneed for DNA based diagnoses because Xchromosome inactivation results in a very widespectrum of factor IX values in carriers who,thus, often show values within the normalrange.7 Furthermore, prenatal diagnosis byfactor IX assays requires fetoscopy for fetalblood sampling at the 20th week of pregnancy8and therefore it involves a procedure with asmall but significant risk to the fetus and of lowacceptance. Several practical and scientificconsiderations also contribute to indicate thathaemophilia B is an ideal model. The factor IXgene is of average size and complexity, namely34 kb and 8 exons9' 10 (figure). The incidence ofthe disease is high but not overwhelminglyhigh (1/30 000 males)."I In the UK and othercountries, good records exist on each affectedsubject and his family, and this facilitates de-tection of genotype/phenotype correlationsand also genetic investigations. Both the ac-tivity and plasma concentration of factor IXcan be measured and it is therefore possible todistinguish mutations that affect productionand stability of factor IX from those that affectenzymatic activity.'2 Finally the domains offactor IX (figure) are homologous to thoseof large families of proteins and, in addition,the factor IX gene is thought to derive fromthe same ancestral gene as factors VII andX and protein C.'3 This is very importantbecause comparison between homologuesallows identification of amino acid residues orprotein regions that are conserved and there-fore probably of functional importance. Cluesas to the possible consequences of factor IXmutations can also be gained from informationavailable on the three dimensional structure ofsome of its homologues and, conversely, thewealth of natural mutants available for factorIX may increase understanding of the struc-ture/function relationships of the factor IXhomologues since natural mutants for thesegenes are harder to find.

Development of new strategyIn order to characterise haemophilia B muta-tions rapidly, we isolated all the essential re-gions of the gene (promoter, exons, and RNAsplicing signals) by PCR amplification.3 Suchregions were initially sequenced directly3 and

Division of Medicaland MolecularGenetics, UnitedMedical and DentalSchools of Guy's andSt Thomas'sHospitals, 7th Floor,Guy's Hospital Tower,London SE1 9RT.F GiannelliS SaadA J MontandonD R BentleyP M Green

Correspondence toProfessor Giannelli

602

on August 7, 2020 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.29.9.602 on 1 Septem

ber 1992. Dow

nloaded from

http://jmg.bmj.com/

A new strategy for the genetic counselling of diseases of marked mutational heterogeneity: haemophilia B as a model

1 V 20 30 ktb

e

. _ ....... ....GF- EG_F _.

.-_.:_---

--46 -18 1 39 47 85 128145

Pre Pro: GLA H EypF EGFtype B type A

--_-K-

41._5

g hf

180

Catal ytic

Activation

Diagram offactor IX gene and protein domains. Top: gene showing exons (a-h) as black boxes. Bottom: factor IXdomains. Numbers indicate amino acid residues at domain boundaries. Shaded areas are cleaved away either beforesecretion (Pre, Pro) or during activation offactor IX (Activation). Dotted lines indicate protein segment coded byeach exon. Pre =prepeptide, needed for transport to endoplasmic reticulum and secretion. Pro = propeptide, neededfor gamma carboxylation of 12 glutamates in gla domain. GLA = region with 11 gamma carboxylated glutamates(gla) conferring Ca"+ binding and affinity for phospholipidic membranes. H= hydrophobic stack, a short segmentcontributing the last gla residue and hydrophobic residues. EGF-type B = epidermal growth factor like domaincontaining a high affinity Ca"+ binding site important for factor IX conformation. EGF-type A = other epidermalgrowth factor like domain lacking high affinity Ca"+ binding site. Activation= peptide cleaved during activation.Catalytic = serine protease domain homologous to other proteases of the family of trypsin and chymotrypsin.

shortly afterwards screened for sequence var-iation by mismatch analysis in heteroduplexesformed by wild type and complementary testDNA strands,14 so that sequencing could belimited only to regions containing anomalies.4This reduced the time to characterise a haemo-philia B mutation to four to five person work-ing days. It thus became possible to makecarrier and prenatal diagnoses of haemophiliaB by the direct detection of the gene defect inevery case. This abolishes the need for cum-bersome family studies and paternity testing,increases the expected diagnostic success from60% to virtually 100%, and indicates the rootcause of the disease without increasing the costof diagnosis. Furthermore, once the haemo-philia B mutation of a subject has been identi-fied, carrier and prenatal diagnoses for thegenetic counselling of his blood relatives canbe done for generation after generation byexamining only the region that is affected inthe index case. This reduces by at least tenfoldthe cost of such diagnoses and cuts the waitingfor results down to two days.

In order to maximise the benefits derivingfrom the above approach and optimise theprovision of carrier and prenatal diagnoses andgenetic counselling of haemophilia B families,we have proposed to characterise the haemo-philia B mutation of a patient or, when neces-sary, an obligate carrier from each affected UKfamily and to construct a national database torecord such mutational information, togetherwith the haematological and pedigree informa-tion pertinent to each index patient. Wethought that such work would advance under-standing of the molecular biology of the dis-ease and possibly help in the development ofbetter treatment while creating a central re-source that could ensure virtually 100% dia-gnostic success, increase the accuracy, speed,and economy of diagnostic services, and bothextend and improve genetic counselling.

In developing the rationale for the abovestrategy, we reasoned that the preliminary sys-tematic characterisation of mutations in indexpatients would not only rapidly advance the

understanding of the disease but also ensure asound theoretical basis for the provision ofdiagnostic services for the following reasons.The analysis of a large number of patients

verifies the efficiency and reliabity of the tech-nical procedures; it also yields data that help inassessing the functional consequences of anyobserved sequence change and shows correla-tions between mutations and phenotypes, andhence offers prognostic information of value togenetic counselling. Although, usually, dia-gnoses based on direct gene analysis are notdependent on the accuracy of the functionalinterpretation of any observed sequencechange, such interpretations are importantwhen testing the collateral and ascendant rela-tives of isolated ('sporadic') patients, since inthis case the failure to detect the causal muta-tion and the attribution of this role to a trivialchange could result in erroneous positive dia-gnoses. The detrimental nature of some muta-tions is immediately obvious but that of othersmust be inferred on the basis of indirect con-siderations such as: the independent occur-rence of the mutation in two or more patients;the demonstration that the mutation has arisende novo in the patient or his mother; and theidentification of factor IX residues that are ofparticular functional significance. Such infor-mation obviously accrues as more and morepatients are analysed and the mutational pro-file of haemophilia B becomes better defined.Furthermore, analysis of all the essential re-gions of the gene will also roughly indicatehow frequently neutral mutations occur in theexamined region of the factor IX gene, sincethese will generally accompany detrimentalmutations in the patients.We were induced to propose the construc-

tion of a national database also by the beliefthat this is the best way to inform even themost remote clinical centre of the new oppor-tunities for diagnosis and genetic counsellingand to offer such centres access to 'state of theart' technology, so that the highest standard ofservice might be attained throughout thecountry.

603



g.bmj.com

/J M



ber 1992. Dow

nloaded from

http://jmg.bmj.com/

Giannelli, Saad, Montandon, Bentley, Green

Finally, we were aware that the individualityof the mutations in different families ensures

the automatic self-verification of mutationaldata. This is because whenever the mutationidentified in the index patient is found in one

of his relatives the later result confirms theaccuracy of the first.

Ethical considerations arise whenever a

population register is created. However, in thehaemophilias there is already a precedentbecause a UK register of haemophilia A and Bwas set up several years ago in order to

optimise administration and monitoring oftreatment. The management of such a registeris such that each haemophilia centre controlsthe release of any information it provides.Similarly, in our proposal, the referring phys-icians will control the release of any informa-tion pertinent to their patients and will remaintotally in charge of the contacts with thepatients and their relatives. In particular, theywill ensure informed consent from the patientsor other subjects collaborating with the projectand will control the use of the database, withrespect to the families they refer. Such use may

entail either simply answering diagnosticenquiries from people who are aware of theirrisks or, in a more active mode, contactingrelatives of the index patients to inform themof their risk and of the availability of accurate

carrier and prenatal diagnoses.

Progress in the implementationof the strategyThe feasibility of the above strategy was testedby a pilot study on a representative sample ofthe Swedish haemophilia B population repre-

senting 60% of the total.'5 Such a study hasshown that our procedure could detect themutation in each one of the 45 families ana-

lysed (table 1) and that full characterisation ofthe mutation entails four to five person days'work. Forty-one patients had only one se-

quence change in the essential regions of thefactor IX gene, one had both a detrimental anda neutral mutation, and three had gross dele-tions.'5'6 A few distinct families had identicalmutations that could be shown to be of inde-pendent origin.17 These always occurred at

CpG sites and identified hotspots of muta-

tions. Thirty-eight (84%) of the 45 mutationsidentified could definitely be considered thecause of the disease. Of these 38 mutations, 18caused absence or gross disruption of factor IXsynthesis (three gross deletions, 12 frameshiftsand nonsense mutations, and three obliter-ations of splice site signals) and 20 were singleamino acid substitutions. The latter could beconsidered clearly detrimental because theyaffected residues essential for the maturationor activation of factor IX (Arg-4, Arg,45), for itsstability (Cys,32, Cys336), or because they repre-

sented changes that have occurred indepen-

Table I Haemophilia B mutations in the families registered with the Malmo haemophilia centre.

Mutation name FIX:C FIX:Ag Nuc change Nuc pos aa change Origin

Malmo 33 3 2 G-*A 122 Donor splice (a) GrandfatherMalmo 40 1 30 C-+T 6364 Arg-4-+Trp OldMalmo 20 2 27 C-+T 6364 Arg-4-_Trp GrandparentsMalmo 19 2,3 36 C T 6364 Arg-4-.Trp OldMalmo 6 < 1 30 C T 6364 Arg-4-Trp OldMalmo 8 <1 <01 2bp del 6398-9 f/s Glu8, STOP 14 GrandparentsMalmo 4 < 1 <0 1 CG-T 6460 Arg29-+STOP NKMalmo 9 < 1 <0 1 lbp del 6466 f/s Val3l, STOP 57 NKMalmo 10 <1 02 10bp del 6665-75 Acceptor splice (c) GrandparentsMalmo 27 19 108 G- T 10395 Gly48-+Val MotherMalmo 21 12 52 CG-T 10415 Pro55-+Ser NKMalmo 22 20 CG-T 10416 Pro55-_Leu NKMalmo 11 <1 <1 G- C 17668 Acceptor splice (e) GrandfatherMalmo 35 21 14 G-+A 17736 VallO7-silent OldMalmo 42 20 24 G--A 17736 VallO7-silent OldMalmo 37 15 24 G- A 17736 VallO7-silent OldMalmo 7 < 1 <0-1 C-+T 17761 Argll6-.STOP GrandfatherMalmo 12 <1 <0-1 G-oT 20375 Cysl32-4Phe OldMalmo 44 lbp del 20398 f/s ThrI40, STOP 156 GrandparentsMalmo 23 7 148 G-.A 20414 Argl45-.His OldMalmo 36 7 110 G-+A 20414 Argl45-eHis OldMalmo 17 7 91 G--A 20414 Argl45-+His OldMalmo 32 8 115 G-_A 20414 ArgI45-+His NKMalmo 38 Mild 86 G-.A 20414 Argl45-+His OldMalmo 43 15 115 G- A 20414 Argl45-His NKMalmo 13 < 1 0 5 lbp del 20510 f/s Aspl77, STOP 198 NKMalmo 5 < 1 <0-1 G- A 20561 Trpl94-+STOP OldMalmo 39 4 4 T-.C 30100 Ile216-*Thr OldMalmo 30 15 G-_A 30150 Ala233-+Thr NKMalmo 29 17 12 G-+A 30150 Ala233-'Thr OldMalmo 31 11 15 G- A 30150 Ala233-.Thr OldMalmo 28 22 G-+A 30150 Ala233-.Thr OldMalmo 14 <1 <0-1 C-.T 30863 Arg248- STOP GrandfatherMalmo 15 <1 <01 C-*T 30863 Arg248-+STOP GrandfatherMalmo 3 < 1 <0-1 C-+T 30863 Arg248-+STOP OldMalmo 41 <1 6 C-+T 30875 Arg252-STOP MotherMalmo 7 < 1 <0-1 CG-T 30890 His257-+Tyr Old (neutral change)Malmo 1 <1 <0-1 8bp del 30950-7 f/s Asp276, STOP 288 OldMalmo 25 4 15 C-+T 31008 Thr296-*Met NKMalmo 18 < 1 0 34 G-+A 31035 Gly3O5S-Asp GrandfatherMalmo 26 3 4 T- G 31041 Val307-+Gly NKMalmo 24 < 1 < 1 G-.A 31128 Cys336-.Tyr OldMalmo 16 15 13 C-*A 31248 Thr376-+Asn OldMalmo 34 < 1 <0-1 Del Del MotherMalmo 2 < 1 <0-1 Del Del OldMalmo 45* 2 Del Del NK

FIX:C and FIX:Ag factor IX coagulation activity and protein as percent of average normal values. Nuc = nucleotide; pos =nucleotide number according toYoshitake et all'; aa change= amino acid change or other consequence ofmutation on protein; amino acid numbers as in Anson et al.9 (a), (c), (e) refer to exon adjacentto mutated splice site; Del= complete gene deletion; NK = patients without family history where origin of mutation could not be determined; * = female patientdescribed by Kling et al."6 Data from Green et al," Kling et al,'0 and Montandon et al."

604



g.bmj.com

/J M



ber 1992. Dow

nloaded from

http://jmg.bmj.com/


dently in two or more families (see Pro55,Ala233), or because they were de novo mutations(Glyu-+Val, Gly305-+Asp). Four other aminoacid substitutions could also be consideredprobable detrimental mutations because theywere non-conservative changes at residuesconserved among the factor IX homologues(Ile2l6 -+Thr, Thr296-+Met, Val307-+ Gly,Thr376-+Asn). Only one mutation that gener-ated a new potential splice site (ValI07 silent)was of uncertain functional significance, butwas observed in three allegedly unrelatedfamilies.

Interesting genotype/phenotype correla-tions were identified. For example, patientsproducing factor IX of low specific activityhad single amino acid substitutions while thosewith defects in the synthesis or stability offactor IX had a wide spectrum of mutationsincluding amino acid substitutions, some ofwhich clearly involved amino acids importantto the stability of factor IX, such as cytosinesforming essential disulphide bridges. Ofgreater importance for genetic counselling wasthe confirmation that mutations causing grossphysical or functional loss of factor IX codinginformation predispose to the development ofinhibiting antibodies against therapeutic factorIX, as this is a very serious clinical complica-tion.'819 The empirical risk of this complica-tion suggested by our data was 1/5 for dele-tions, frameshifts, and stop codons, andvirtually 0 for the others.'5 The Swedish studyalso showed that over 50% of the families (23/44) had a single affected male and no familyhistory. In 13 of these families ascendant rela-tives were available for investigation of theorign of the mutation. In three cases the muta-tion had occurred in the patient's mother andin 10 in the maternal grandparents. Six of thelatter families had informative haplotypes ofDNA markers and in each family the mutationcould be shown to be of grandparental origin.20Furthermore, since we had looked at a repre-sentative population sample we could attemptto estimate directly the overall haemophilia Bmutation rate and also the sex specific mutationrates. This suggested an overall mutation rate of4-1 x 10-6 in keeping with indirect estimates5and a higher mutation rate in the male.2'

After the successful completion of the pilotstudy we have begun the construction of theUK national database. In this country thereare just over 1000 patients with haemophilia Band these probably derive from 500 to 600families. In the last 18 months we have col-lected samples from 220 unrelated patients andwe have characterised the mutation of 125. Inonly one of the 170 patients examined so far(Swedish and British) have we failed to find a

mutation in the essential regions of the gene,and this clearly indicates that our screeningprocedure is capable of detecting all types ofmutations. The phenotype/genotype correla-tions observed in the Swedish population areconfirmed and extended (table 2) and so is ourknowledge of the factor IX mutational hot-spots. In fact we have been able to estimatedirectly the rate of mutation at such hotspots.

Table 2 Mutations and haematological classification ofSwedish and British patients examined so far.

I II III Inh NK

Promoter 1Gross deletion 2 4 1Splice site 6 2New splice site 4 1Frameshift 8 2 6Nonsense 12 3 8Missense/aa del 18 16 23 39Total 18 16 55 9 58 (156)

Class I = low FIX:C and normal FIX:Ag; class II = low FIX:Cand FIX:Ag reduced but not as low as FIX:C; class III= lowFIX:C and correspondingly low FIX:Ag; Inh= inhibitorpatients; NK = not known.

This is 1-05 x 10-' transitions per CpG site pergamete per generation.22A substantial step forward has been made in

the definition of the mutational profile of hae-mophilia B. In particular our work and that ofothers worldwide has led to the characterisa-tion of approximately 600 mutations.'223 Mostof these are distinct small changes or 'point'mutations. In keeping with the data shown intable 2 some of these affect the promoter andresult in factor IX deficits that show a cleartendency to improve with age and especiallywith pubertal maturation.2s26 Others causegross defects in factor IX coding (frameshift,stop codons, and splicing defects) but themajority result in single amino acid substitu-tions. So far, 175 different single amino acidsubstitutions and five single amino acid dele-tions have been reported as the probable ordefinite cause of the disease. The amino acidsubstitutions affect 124 different residues ofthe factor IX protein. Analysis of the factor IXhomologues or of the factor IX of differentmammalian species has indicated that most ofthe above residues are conserved both in thefactor IX homologues and the factor IX ofdifferent mammalian species, while a smallproportion is conserved only in the factor IXof different mammals. Different substitutionshave been found at 39 sites (table 3). At 14 ofthese sites the degree of discrepancy betweenthe side chains of wild type and mutant aminoacids appear to correlate with the severity ofthe factor IX defect (table 3A) while at 17 sitesit does not (table 3B). This seems to suggestthat at some positions the detrimental effect isthe result principally of the nature of thesubstituent amino acid while at others it is not,presumably because there is a specific require-ment for the wild type amino acid. For muta-tions at the remaining eight sites there isinsufficient phenotypic information or otherfactors may contribute to the phenotype (forexample, effect of mutation on splicing).A few neutral amino acid substitutions have

also been detected, which accompanied detri-mental mutations. For example, in the familydescribed by Montandon et aP' an isolatedpatient showed a stop codon and a potentialamino acid substitution distal to the stopcodon. The patient's maternal grandfather,who was haematologically and clinically nor-mal, had the amino acid substitution but, ofcourse, not the stop codon. This had occurredde novo in the grandfather's germline and waspresent in the patient's mother.

605



g.bmj.com

/J M



ber 1992. Dow

nloaded from

http://jmg.bmj.com/

Giannelli, Saad, Montandon, Bentley, Green

Table 3 Comparison of phenotypes in patients with different substitutions of the sameamino acid.

Mutation Severity Haematological class

A Gla27-_Lys Severe IIGla27-*Val Severe IIIGly59_- Ser Mild ?Gly59-_Val Severe ?Gly6O0-Ser Mild IIGly6O-Asp Severe IIIGly6O-Arg Severe IIIArgl45-*Cys Severe IIArgl45-_His Moderate IArgl45-_Leu Moderate I/IIVall82-_Leu Mild IVall82-_Phe Severe IVall82-_Ala Mild IIle216-_Thr Moderate IIIIle216-_Met Mild IIIGlu245-_Lys Severe IGlu245-Val Severe IIAla291 -_Pro Severe IIIAla291 -_Thr Moderate IIIVal307-_Ala Mild IIVal307-_Gly Severe IIISer308-*Asn Moderate ?Ser308-*Arg Severe ?Val328-_Phe Severe/moderate IIIVal328-Ile Normal NormalVal331 -+del Severe IIIVal331 -+Ala Moderate IThr380_- Ser Moderate IIIThr380_-Ile Severe IIIThr380-_Pro Severe ?Ala390-_Val Severe IAla390-_Glu Severe II

B Arg-4-_Trp Severe IIArg-4--Leu Severe IIArg-4-_Gln Severe I/IICysl8-_Arg Severe ICysl8-Tyr Severe ?Cys23-_Arg Severe IICys23-_Tyr Severe IIGly48-.Arg Mild IGly48-*Val Mild IPro55-_Ala Mild IIPro55-.Ser Mild IIPro55-_Leu Mild ?Cys56-*Arg Severe IIICys56-.Ser Severe IIICys56-_Tyr Severe IIIAsp64-_Asn Severe/moderate IAsp64-_Gly Moderate ICysl32-*Arg Severe IIICysl32-Phe Severe IIICysl32-_Tyr Severe ?Argl8O-Trp Severe IArgl8O-Gly Severe IArgl8O-Gln Severe IArgl8O-Pro Severe ?Cys2O6_- Ser Severe IIICys2O6-.Tyr Severe ?Cys222-_Trp Severe ?Cys222-Ser Severe ?Ile270-_Phe Severe IIIIle270-_Thr Severe IIIGly309- Ser Severe I/IGly309-*Val Severe I/IICys336-_Arg Severe IIICys336-Tyr Severe IIIGly363-_Val Severe IGly363-*Glu Severe I/IIGly363-Arg Severe IGly363-_Ala Severe IAsp364-*His Severe IAsp364-Asn Severe ?Asp364-_Val Severe ISer365-+Gly Severe ISer365-_Ile Severe ?Ser365-_Arg Severe I

C Thr38-_Arg Severe ?Thr38-_Ile Moderate ?Asp47-_Asn Severe III (may affect splicing)Asp47-+Gly Mild IAsp47-_Glu Severe(l) I

Mild(1) IAsn92-_His Severe IIIAsn92-*Lys Severe ?Cys95_-Tyr Severe IIICys95-_Trp Severe ?Gly311 _-Arg Severe IGly3l 1 -_Glu Severe IArg333-.Gly Mild(1) I

Severe(1) IArg333-Gln Severe IArg333-+Leu Severe IMet348_-Ile Severe IMet348-_Val Severe IIPro368-Thr Severe IPro368-Leu Severe/moderate ?

We can now confidently expect that, in thenext few years, the mutational profile of hae-mophilia B will be well defined. This willclearly identify the structural features that are

important to factor IX and provide furtherdata to assess the functional consequences ofany change found in the patients' genes.

Our work to date also confirms that thestrategy proposed above attains virtually 100%success as well as accurate and fast results incarrier and prenatal diagnosis. We havereceived 91 unsolicited requests for carrierdiagnoses and all have been successfully dealtwith. These have usually involved situationsthat could not have been helped by indirectDNA diagnostic procedures based on DNAmarkers. These are: families with a singleaffected member, subjects whose affected rela-tives were not available, or female patientswithout a family history of haemophilia B.

Eight prenatal diagnoses were requested on

blood relatives of patients in the database.These were generally at short notice and one

involved a person living abroad at the time ofneed. Success was again achieved in each case.

Such diagnostic work, as well as investiga-tion of genetically interesting families, has alsoresulted in the automatic verification of themutation data in 50% of the index cases listedin our UK database. This clearly indicates thatfull use of the database will result in the rapidverification of much of the mutational infor-mation it contains.

ConclusionsWe have shown that our methods for the rapiddetection of gene defects are reliable andeconomical enough to be applied to largepopulations of haemophilia B patients. Thesemethods are indeed applicable to larger andmore complex genes, as illustrated by our workon haemophilia A and Duchenne musculardystrophy2829 where the use of leaky RNAtranscription has been used to reduce the com-plexity of the task. The availability of rapidmutation detection methods is allowing theconstruction of a national database of all hae-mophilia B mutations. This information,together with pedigree and haematologicaldata, will form a critical resource for the provi-sion of carrier and prenatal diagnoses in vir-tually every family with haemophilia B. Thework for the construction of the database hasalready advanced our knowledge of the mole-cular genetics of the disease and is creating a

solid scientific framework on which to baseprecise genetic counselling, not only because itallows more accurate interpretation of thefunctional consequences of observed sequencechanges, but also because it provides prognos-tic information that enhances the quality ofadvice that can be given. Requests for carrierand prenatal diagnosis from many familieswhere the indirect RFLP procedure hadfailed, or could not be applied, underline theusefulness of the new approach. This shouldeventually allow the highest standard of ser-vice to be attained uniformly throughout thecountry.

Disease is defined as severe (<3% FIX:C), moderate (3-10% FIX:G), and mild (11-30%FIX:C). Haematological class is as defined in table 2: the seriousness of the effects of mutationsincreases in the order mild-+severe and haematological class I--III. Group A have effects thatcorrelate with the difference in the substituents; group B have the same phenotype regardless ofthe substituent amino acid; in group C there are either insufficient data or other factors areexpected to affect the phenotype as specified in brackets. Data from Giannelli et al.23

606



g.bmj.com

/J M



ber 1992. Dow

nloaded from

http://jmg.bmj.com/


This work is supported by Action Research.We are also grateful for support from theSpastics Society and the Generation Trust.We are indebted to all the patients and clini-cians of the UK haemophilia centres for theircollaboration. We are particularly gratefulto Professor I M Nilsson, Dr R L Jung, andS Kling for their collaboration during the pilotstudy in Sweden. We are also grateful toProfessor M Bobrow for reading the manuscriptand to Adrienne Knight for secretarial help.

1 Giannelli F. The identification of haemophilia B mutations.In: Peeters H, ed. Protides of the biological fluids. Vol 35.Oxford: Pergamon Press, 1987:29-32.

2 Tsang TC, Bentley DR, Mibashan RS, Giannelli F. Afactor IX mutation, verified by direct genomic sequenc-ing, causes haemophilia B by a novel mechanism. EMBOJ 1988;7:3009-15.

3 Green PM, Bentley DR, Mibashan RS, Nilsson IM, Gian-nelli F. Molecular pathology of haemophilia B. EMBO J1989;8:1067-72

4 Montandon AJ, Green PM, Giannelli F, Bentley DR.Direct detection of point mutations by mismatch analysis:application to haemophilia B. Nucleic Acids Res1989;17:3347-58.

5 Ferrari N, Rizza CA. Estimation of genetic risks of carrier-ship for possible carriers of Christmas disease (haemophi-lia B). Braz J Genet 1986;9:87-99.

6 Haldane JBS. The rate of spontaneous mutation in a humangene. J Genet 1935;31:317-26.

7 Giannelli F. The contribution of gene specific probes to thegenetic counselling of families segregating for haemophi-lia B. In: Ciavarella NL, Ruggieri ZM, Zimmerman TS,eds. Factor VIII/von Willebrand factor - biological andclinical advances. Milan: Wichtig Editore, 1986:159-74.

8 Mibashan RS, Rodeck CH, Nicolaides KH, Warren R,Pembury ME, Giannelli F. Phenotypic and DNA dia-gnosis of fetal bleeding disorders. In: Ciavarella NL,Ruggeri ZM, Zimmerman TS, eds. Factor VIII/vonWillebrandfactor - biological and clinical advances. Milan:Wichtig Editore, 1986:175-89.

9 Anson DS, Choo KH, Rees DJG, et al. The gene structureof human anti-haemophilia factor IX. EMBO J1984;3:1053-60.

10 Yoshitake S, Schach BG, Foster DC, Davie EW, KurachiK. Nucleotide sequence of the gene for human factor IX(antihemophilic factor B). Biochemistry 1985;24:3736-50.

11 Rizza CR, Spooner RJD. Treatment of haemophilia andrelated disorders in Britain and Northem Ireland during1976-80: report on behalf of directors of haemophiliacentres in the United Kingdom. BMJ 1983;286:929-33.

12 Green PM, Montandon AJ, Bentley DR, Giannelli F.Genetics and molecular biology of haemophilias A and B.Blood Coag Fibrin 1991;2:539-65.

13 O'Hara PJ, Grant FJ, Haldman BA, et al. Nucleotide

sequence of the gene coding for human factor VII, avitamin K-dependent protein participating in blood coa-gulation. Proc Natl Acad Sci USA 1987;84:5158-62.

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 Green PM, Montandon AJ, Ljung R, et al. Haemophilia Bmutations in a complete Swedish population sample. Atest of new strategy for the genetic counselling of diseaseswith high mutational heterogeneity. Br J Haematol1991;78:390-7.

16 Kling S, Coffey AJ, Ljung R, et al. Moderate haemophilia Bin a female carrier caused by preferential inactivationof the paternal X chromosome. Eur J7 Haematol1991;47:257-61.

17 Green PM, Montandon AJ, Ljung R, Nilsson IM, Gian-nelli F. Haplotype analysis of identical factor IX mutantsusing PCR. Thromb Haemostas 1992;67:66-9.

18 Giannelli F, Choo KH, Rees DJG, Boyd, Y, Rizza CR.Gene deletions in a patient with haemophilia B and antifactor IX antibodies. Nature 1983;303:181-2.

19 Giannelli F, Brownlee GG. Cause of the 'inhibitor' pheno-type in the haemophilias. Nature 1986;320:196.

20 Kling S, Ljung R, Sjorin E, et al. Origin of mutation insporadic cases of haemophilia B. Eur J Haematol1992;48: 142-5.

21 Montandon AJ, Green PM, Bentley DR, et al. Directestimate of the haemophilia B (factor IX deficiency)mutation rate and of the sex-specific mutation rates inSweden. Hum Genet 1992;89:319-22.

22 Green PM, Montandon AJ, Bentley DR, Ljung R, NilssonIM, Giannelli F. The incidence and distribution ofCpG--aTpG transitions in the coagulation factor IX gene.A fresh look at CpG mutational hotspots. Nucleic AcidsRes 1990;18:3227-31.

23 Giannelli F, Green PM, High KA, et al. Haemophilia B:database of point mutations and short additions anddeletions - third edition. Nucleic Acids Res 1992;20:2027-63.

24 Reitsma PH, Bertina RM, Ploos van Amstel JK, RiemansA, Briet E. The putative factor IX gene promoter inhemophilia B Leyden. Blood 1988;72:1074-6.

25 Reitsma PH, Mandalaki T, Kasper CK, Bertina RM, BrietE. Two novel point mutations correlate with an altereddevelopmental expression of blood coagulation factor IX(hemophilia B Leyden phenotype). Blood 1989;73:743-6.

26 Crossley M, Brownlee GG. Disruption of a C/EBP bindingsite in the factor IX promoter is associated with haemo-philia B. Nature 1990;345:444-6.

27 Montandon AJ, Green PM, Bentley DR, Ljung R, NilssonIM, Giannelli F. Two factor IX mutations in the familyof an isolated haemophilia B patient. Direct carrier dia-gnosis by amplification mismatch detection (AMD). HumGenet 1990;85:200-4.

28 Naylor JA, Green PM, Montandon AJ, Rizza CR, GiannelliF. Detection of three novel mutations in two haemophiliaA patients by rapidly screening whole essential regions ofthe factor VIII gene. Lancet 1991;337:635-9.

29 Roberts RG, Bobrow M, Bentley DR. Point mutationsin the dystrophin gene. Proc Natl Acad Sci USA1992;89:2331-5.

607



g.bmj.com

/J M



ber 1992. Dow

nloaded from

http://jmg.bmj.com/

a strategy genetic counselling - jmg.bmj.com · 07medgenet 1992; 29: 602-607 reviewarticle...

Documents