Clinical genetics in neurological disease · clinical genetic andmoleculargeneticpractice in the management of the major inherited neurological diseases. Weconsider diagnostic issues

J7ournal ofNeurology, Neurosurgery, and Psychiatry 1994;57:7-15

NEUROLOGICAL MANAGEMENT

Clinical genetics in neurological disease

John C MacMillan, Peter S Harper

Inherited neurological disorders have tradi-tionally engendered feelings of therapeuticnihilism in the medical community: not onlyare these conditions often difficult to diag-nose with any certainty, they are also per-ceived as "untreatable" whether or not a firmdiagnosis has been reached. This negativeresponse is due in part to a lack of knowledgeconcerning the underlying pathological basisfor these disorders.The 1980s and early 1990s have seen a

rapid increase in our understanding of themolecular mechanisms behind many of thesedebilitating conditions, largely due toadvances in recombinant DNA technology.This article aims to summarise the role ofclinical genetic and molecular genetic practicein the management of the major inheritedneurological diseases. We consider diagnosticissues as they pertain to the individual case-that is, the specific disorder affecting thepatient and its likely prognosis-and coun-selling issues for the family, such as the iden-tification of individuals who may developsymptoms in later life (presymptomatic test-ing) or risks for the inheritance of the disor-der by a future family member (prenataltesting and carrier detection). To date therehave not been equivalent advances in ourtherapeutic options in the management ofthese conditions although experimental pro-tocols have been initiated with this in mind,some of which show promising results in ani-mal models.The disorders and the relevant molecular

genetic analyses under consideration are dis-cussed in two groups: those for which one ormore genetic loci have been identified but thegene(s) not yet cloned; and those for whichthe genes have been isolated and the relevantproduct characterised. We begin by reviewingthe available data from epidemiological andfamily studies on the frequency of thesedisorders.

Institute ofMedicalGenetics, UniversityofWales College ofMedicine, Heath Park,CardiffCF4 4XN, UKJ C MacMillanP S HarperCorrespondence to:Dr MacMillan

EpidemiologyIndividual genetic neurological diseases arenot common but as a group they account fora significant burden of disability in a generallyyounger age group than affected by otherneurological diseases. In a review of reportspublished throughout the world, Emery' esti-mated a population prevalence for the inher-ited neuromuscular diseases of over 1 in3000. The authors' study in southeast Wales2suggests a minimum prevalence for the majorinherited neurological disorders of over 58per 100 000 population (approximately 1 in1700 population). Table 1 shows the disor-ders and the prevalence of symptomatic indi-viduals in the southeast Wales population.

These figures do not take into accountasymptomatic gene carriers who may developclinical disease in the future; nor do theyallow for the unaffected but "at-risk" familymembers who are also likely to request infor-mation on the disease (this latter group aremore likely to seek guidance from a geneticsunit but a not insignificant number may beseen initially by the neurologist dealing withthe index case).

Linkage analysisDISORDERS LOCALISED TO SPECIFICCHROMOSOMAL SITES BUT NOT YET CLONEDThe number of disorders in this groupdecreases monthly as a result of the continualisolation of new genes for specific diseases.The technology used in "linkage analysis"merits a brief description as it needs to beappreciated that there are limitations to itsusefulness in many clinical situations.

Linkage analysis has been the first stage inthe localisation of a gene responsible formany genetic disorders. It makes use ofrestriction fragment length polymorphisms(RFLPs). These are lengths of DNA thathave been cut at each end at a specific site

Table 1 Prevalence of major, single-gene neurological disorders in southeast Wales

Disorder Prevalence (/100 000) 95% confidence interval

Hereditary motor and sensory neuropathy types I,JI,JII and V 12-9 10-7-15-4Myotonic dystrophy 7-1 5-5-9-1Duchenne muscular dystrophy 9-6* 7-0-12-9Becker muscular dystrophy 5.0* 3-2-7-6Facioscapulohumeral muscular dystrophy 2-9 1-9-4-1Spinal muscular atrophy (all types) 1-3 0-7-2-2Huntington's disease 8-4 6-6-10-5Tuberous sclerosis 1-6 0-9-2-6Von Hippel Lindau disease 0-6 0-2-1-4Hereditary spastic paraplegia 3-4 2-3-4-8Neurofibromatosis type 1 13-3 11 1-158

*Per 100 000 males.

7

on March 29, 2021 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.1.7 on 1 January 1994. D

ownloaded from

http://jnnp.bmj.com/

MacMillan, Harper

recognised by a bacterial restriction endonu-clease enzyme. Different enzymes recogniseand cleave at different DNA sequence motifsof variable numbers of base pairs. The greatvariability in human DNA results in differ-ences in the length of sequence betweencleavage sites in unrelated individuals. Withina family, however, these anonymous DNAsegments (RFLPs) are inherited in amendelian fashion and can be trackedthrough a family pedigree using standard lab-oratory techniques. Botstein et al3 first sug-gested the use of these in the construction ofgenetic linkage maps for the localisation ofdisease genes to particular segments of chro-mosomes and they have superseded the bloodgroup polymorphisms used in early studies.

In a family linkage analysis, each memberof the family is assessed for the presence ofthe disease of interest and their status(RFLP) for the marker locus is ascertained. Ifthe disease locus and the marker locus areclose together on the same chromosome(linked) then independent assortment atmeiosis will be rare and both traits will beinherited in the offspring. The closer they areto each other the less likely it is that they willseparate during the pairing of homologouschromosomes at meiosis-in other words,that a recombination will occur. The recom-bination fraction, known as theta (0) gives anindication of the genetic distance between thetwo loci. Genetic distance is expressed incentiMorgans where 1 cM is defined as a seg-ment of chromosome with a 1% chance ofrecombination per meiosis4 and this is equiva-lent to a 0 of 0.01. A recombination fractionof 0 5 means either that the two loci are ondifferent chromosomes or that they are farapart on the same chromosome.The successful application (accuracy) of

linkage analysis in the clinical situation isdependent on several factors: it is crucial thatthe diagnosis is correct; and the disease musthave been localised (mapped to a segment ofchromosome) with close polymorphic mark-ers available. Affected individuals should beheterozygous for the marker polymorphisms("informative"), there should not (ideally) begenetic heterogeneity (more than one diseaselocus for the clinical phenotype), and thefamily relationships, especially paternity,must be clearly established. The disease sta-tus for each individual typed must be accu-rate (especially where the disease is of lateonset and "unaffected" individuals are crucialto the analysis). The figure shows a familytyped for a hypothetical two-allele polymor-phism known to be linked to the disease ofinterest. Individual I2 is affected by the disor-der and is heterozygous for the polymorphism(she types 2-1); her unaffected spouse ishomozygous (2-2) as is her unaffected sonIIL. Individual 13 is clinically affected and hasinherited allele 1 from his mother, and hassubsequently transmitted this allele to indi-vidual III,. This three-generation situation isideally suited to linkage analysis as phase isknown-that is, the chromosome carrying themutant gene can be identified-and affected

i

LLanePerson

4..

i -

Figure The lanes in the autoradiograph contain DNAfrom the individuals indicated in the pedigree. The DNAhas been digested using a restriction endonuclease enzyme,subjected to electrophoresis on an agarose gel andtransferred to a nylon membrane. The DNA probe ofinterest has been labelled with radioisotope and incubatedovernight with the membrane. The membrane has beenstored at -200C adjacent to radiographic film for sevendays. Bands are assigned numbers with the largest banddesignated 1.

individuals are heterozygous for the polymor-phisms of interest. It is possible to predict forindividual III3, who may be too young for thedisease to have manifested, the likelihood thatthey will develop the disease from a knowl-edge of their genotype. In this situation thepresence of allele 1 indicates a high risk, theactual value of which would depend on therecombination fraction between the markerand the disease. In general it is unlikely that amarker would be used whose rate of recombi-nation with the disease was greater than 3%.In the example shown the actual value wouldbe 97%, using a marker with 3% recombina-tion, as it is a fully informative situation. Thisapproach can also be used to increase thereproductive options for couples at risk bypermitting prenatal diagnosis through linkageanalysis and, where affected, termination ofan affected pregnancy. Individual III, in thefigure could represent a conceptus fromwhich chorionic villus sampling would pro-vide appropriate tissue for analysis in earlypregnancy.

Specific disordersTable 2 lists those diseases linked to specificchromosomes, but with the gene not yet iso-lated, for which family linkage studies canplay a useful role in patient and family man-agement. It is not intended to be a compre-hensive list of all mapped neurogeneticdiseases; the interested reader should consultsources such as the bimonthly listings inNeuromuscular disorders (Pergamon Press),the summaries of the Human GenomeMapping Project annual meetings,

8

t.7---,



j.com/

J Neurol N


ownloaded from



Table 2 Chromosomal localisations for major genetic neurological disorders for which thegene has notyet been cloned

Disorder Localisation(s)

Emery-Dreifuss muscular dystrophy Xq27 3-28Facioscapulohumeral* muscular dystrophy 4q35Familial spastic paraplegia (one dominant type) 14qFriedreich's ataxia 9ql3-21Limb-girdle muscular dystrophy (dominant) 5q22-3-31-3Spinal muscular atrophy recessive types I, II and III 5ql 1-2-13*3Tuberous sclerosis (see also table 3) 9q34

Gene locations are identified by the chromosome number followed by an indication of whicharm (p = short, q = long) and the sub-band.*But see below for comments on the molecular rearrangement reported in this disorder.

McKusick's Mendelian inheritance in man(published and "online" versions) and peri-odic reviews.5

It may be possible to provide presympto-matic testing for these disorders for relativesat risk if the family structure is suitable butthe presence of genetic heterogeneity, as intuberous sclerosis, can make even this verydifficult in all but the largest families. VariousBayesian approaches have been proposed totake such additional factors into accountwhen counselling individuals in these situa-tions.6

Although the gene for facioscapulohumeral(FSH) muscular dystrophy has not yet beencloned, it is possible to make diagnostic useof molecular DNA analysis where it isthought that an affected individual representsa new dominant mutation for the disorder.This paradoxical situation arises as the resultof the presence of detectable rearrangementsassociated with a series of 3-2 kb tandemrepeat elements in the subtelomeric region ofthe long arm of chromosome 4.7 (1 kb =1000 base pairs.) The DNA probe pl3E-1 1(locus D4F104S1) detects bands over 50 kbin size after EcoRI digestion in normal indi-viduals and a band less than 35 kb in lengthin individuals with FSH dystrophy. In mostcases thought to represent new mutations, itis possible to show that both parents possesstwo normal-sized alleles whereas the affectedindividual has one normal-sized and onerearranged fragment.Some disorders, mostly rare, have not yet

been mapped to a specific chromosome(s)and genetic management is confined to coun-selling families on the inheritance and natureof the disease.

Mutation analaysisOnce a gene associated with a particular dis-ease has been isolated it becomes possible totest DNA samples from affected individualsfor mutations within that gene. These maytake the form of point mutations that substi-tute alternative nucleic acid bases and resultin changes in the amino-acid sequence (mis-sense mutations) or introduce stop codonswith premature truncation of the translationproduct (nonsense mutations). Other muta-tions may involve the insertion or deletion ofa number of base pairs (from one to manythousands) resulting in disruption of the nor-mal function of the gene. Occasionally dele-

tions or duplications of whole chromosomesegments are found in association with spe-cific disorders. In addition to these "classical"types of mutation, we have seen in the pasttwo years, the identification of a new class ofmutation that is often unstable in its trans-mission between generations. This is the trin-ucleotide repeat, the principal characteristicof which is instability, giving varying molecu-lar and clinical changes in the differentaffected members of a single kindred, as dis-cussed more fully below.

Mutation analysis has the advantage (overlinkage analysis) that it is not dependent onfamily structure and can therefore be appliedto an isolated case of a disorder suspected tobe genetic. A variety of laboratory techniquesare used in the detection of mutations8 and itmay take several weeks for a specific defect tobe identified, depending on whether themethod is in regular service use or still at theresearch stage.

DISORDERS FOR WHICH THE RESPONSIBLEGENE HAS BEEN CHARACTERISEDTable 3 lists some of those disorders forwhich the gene responsible has been identi-fied and mutation analysis is feasible (thespecial situation for facioscapulohumeralmuscular dystrophy has been mentionedabove). The number of disorders in this cate-gory is increasing rapidly and individualgenetic units cannot be expected to provideevery analysis. In the United Kingdom, theclinician primarily responsible for the individ-ual patient's care should consult their localgenetics unit for information on which molec-ular services are available locally and thosewhich can be obtained elsewhere (usuallyfrom other genetic laboratories in theNational Health Service). It should bestressed that a comprehensive genetic coun-selling service should be available to all indi-viduals on whom a molecular genetic analysisis undertaken. This is particularly relevantwhere specific mutation analysis is requestedbut the individual has not been aware that thedisorder may be hereditary. The analyses mayhave significant implications for siblings oroffspring. Examples of these difficulties aregiven below in respect of specific disorders.

In several disorders listed, a single, or verysmall number, of distinct mutations in thesame gene are responsible for the same phe-notype: thus in Leber's hereditary optic neu-ropathy (LHON) a point mutation at basepair 11 778 (in the gene for subunit 4 of thereduced nicotinamide adenine dinucleotide(NADH) dehydrogenase (ND4) complex inthe respiratory chain) of the mitochondrialgenome accounts for 50% of cases, whereasother causative mutations giving an identicalclinical picture occur at base pairs 3460 (inND1), 4160 (ND1) and 15 257 (cytochromeb).9 Mutation analysis in this situation there-fore requires careful planning and interpreta-tion. Recent work has suggested that somemitochondrial mutations may be synergistic-for example, mutations at base pairs 13 708,15 257, 15 812 and 5244-and the accumu-

9



j.com/

J Neurol N


ownloaded from


MacMillan, Harper

Table 3 Cloned genes, their products, functions and mutations

Disease Chromsonal Gene Mutationslocalisation product Function Mutation identified (%/6)

DMB/BMD Xp21 Dystrophin Cytoskeletal protein Deletions, duplications 50Point mutations

Neurofibromatosis type 1 17ql 1-2 Neurofibromin Tumour suppressor Deletions 20Point mutations

Neurofibromatosis type 2 22ql 1 Merlin Tumour suppressor Deletions 50Point mutations

Tuberous sclerosis (one type) 16pl3 Tuberin Tumour suppressor Deletions UncertainMyotonic dystrophy l9q13 2 Myotonin Protein kinase Expanded trinucleotide

repeat sequence 100Huntington's disease 4pl6-3 Huntingtin Unknown Expanded trinucleotide

repeat sequence 99Charcot-Marie-Tooth disease

type 1A 17pll 2 PMP-22 Myelin protein Duplication


X mental retardation syndrome associatedwith a cytogenetically detectable fragile site atXq27T3 (FRAXA), the expanded sequencelies 5' (upstream) to the coding sequence. Itappears that in this disorder there is a criticaldegree of expansion beyond which there issilencing of gene expression'6 possibly due tohypermethylation within the gene'7 and it isthis deficiency that is responsible for the dis-ease phenotype. In the fragile X syndromewith a fragile site at Xq28 (FRAXE), there isamplification of a GCC repeat with associ-ated hypermethylation of the neighbouringCpG island (a region associated with anactive gene) implying a similar disease mech-anism to that seen in FRAXA.18 In myotonicdystrophy it has been shown that the level ofmyotonin protein kinase messenger mRNA isreduced when the mutant allele has less than200 repeats and very low when there are 800repeats.'9 In congenital cases it has beenreported that not only is no mutant mRNAdetectable but also that there is a reducedlevel of the normal paternal mRNA.20Another group has, however, reported anincrease in the level of mutant mRNA in con-genital cases.2' This conflict needs to beresolved.A specific characteristic of this type of

mutation is its instability; the length of therepeat sequence tends to increase as it under-goes successive meiotic passages (as it passesfrom one generation to the next). Inmyotonic dystrophy this successive expansionaccounts for the long debated phenomenonof anticipation whereby the clinical manifes-tations (phenotype) become more severe insubsequent generations. It does not, however,account for the observation that congenitalcases of myotonic dystrophy are always bornof affected mothers.

These trinucleotide repeat amplificationsare quantitative mutations and have been thesubject of several studies examining correla-tions between degree of amplification and theage at onset of symptoms.22 The questionarises as to whether it is possible to predictthe age at disease onset and severity of thedisease phenotype knowing the degree ofexpansion in the DNA sequence.2' We wouldcaution against attempting this in a specificclinical situation. Although there is a cleardistinction between the degree of expansionin congenital cases of myotonic dystrophyand late onset/minimal manifesting individu-als, such as cataract only, there is consider-able overlap in other situations, even withinthe same family.

Phenotype prediction on the basis ofdegree of sequence expansion inHuntington's disease is fraught with evenmore difficulties: we have shown that theindividual variation in age at onset for a givenrepeat number can range over a span of20-30 years24 and, in addition, that the size ofthe repeat sequence on the normal paternalallele has an effect in maternally transmitteddisease.25 This may account for the largervariability in age at onset in sibships with dis-ease of maternal compared with paternal ori-

gin. As with any inherited disorder, the dis-ease-specific mutation is acting within theindividual's unique (with the exception ofidentical twins) genetic background whichmay contain several modifying loci. It couldwell be argued that the lack of a clinically use-ful correlation has potential benefits to anindividual requesting presymptomatic testing:the psychological well being of an individualand his or her family would be unlikely to behelped by foreknowledge of a precise age atonset of this devastating disorder.

HEREDITARY MOTOR AND SENSORYNEUROPATHY TYPE IIn considering the applicability of moleculargenetic testing, the case of type I hereditarymotor and sensory neuropathy (HMSN I)illustrates some of both the benefits and limi-tations of using direct mutational analysis inthe clinical situation. HMSN type I (alsoknown as Charcot-Marie-Tooth disease type1) is an autosomal dominant condition withextreme variability in phenotype. The mani-festations range from significantly reducedperipheral nerve conduction velocity (medianmotor conduction velocity

MacMillan, Harper

sis). One situation that has been clarified withmutation analysis for CMT1/HMSN I is thecase of the isolated, affected individual whomay be considered to have the autosomalrecessive form of HMSN I. Hoogendijk et aP9have shown that in nine out of 10 isolatedcases of HMSN I (in whom both parentswere clinically and electrophysiologically nor-mal), the disorder had arisen from a de novoduplication occurring in one of the parentalgerm cells. The risk to subsequent offspringof such an affected individual is clearly 50%(and not "minimal" as may previously havebeen counselled) whereas the risk to siblingsof such patients would be minimal (this maybe modified in some cases by gonadalmosaicism in the parent) rather than the pre-viously counselled 25%. Counselling of aduplication-negative isolated male caseremains difficult in view of the possibility thathe has the X-linked form of CMT. Hayasakaet alP0 have recently reported the identificationof point mutations in the gene for Po, anotherintegral membrane protein of peripheral ner-vous system myelin, in individuals withCMTlb. Where this complexity of situationoccurs there is much to be said for the con-tinued application of standard clinical screen-ing protocols (measurement of motor nerveconduction velocities and sensory nerveaction potentials) for individuals at risk inaddition to direct molecular analysis.

Considerations when requestingmutation analysisIn myotonic dystrophy, Huntington's diseaseand Charcot-Marie-Tooth type 1 (HMSN I)the success of mutation analysis makes ittempting to expand the range of the clinicalpicture that prompts the dispatch of a bloodsample for "DNA analysis". We have alreadystated our view that, before any sample istaken, the patient should be made aware ofthe potential genetic implications of a "posi-tive" result. This may not cause undue anxi-ety if the individual believes other familymembers show similar signs already, or he orshe perceives the disorder not to be undulyserious.

Three disorders merit closer examinationin considering the clinical spectrum thatshould prompt a "DNA analysis":Huntington's disease; Alzheimer's disease;and prion dementia. All may manifest withprogressive dementia with or without addi-tional features. Should we therefore screen allcases of dementia for the specific mutationsassociated with these disorders? Several prob-lems arise in considering this.

Firstly, there is the question of consent.Can the patients be made aware of the impli-cations of a positive result or should theirfamilies make the decision on their behalf?After all it is they (the family) who will bemost affected by such a result.

Secondly, the sensitivity of the analysisshould be considered. In those situationswhere we believe this to be approaching

100% (Huntington's disease) the family canbe reassured by a negative result. In familialAlzheimer's disease, however, there is geneticheterogeneity, and the available molecularanalysis (looking for point mutations in theAPP gene) would screen only one geneaccounting for a small proportion of cases.3'A negative result would not therefore excludea diagnosis of familial Alzheimer's disease.Screening DNA from patients with dementiafor mutations in the prion protein gene asso-ciated with Gerstmann-Straussler syndromeand familial Creutzfeldt-Jakob disease would,superficially, seem reasonable in view of thereported cases where dementia is the onlymanifestation.'2The implications for the siblings and off-

spring of a patient with an identifiable prionmutation are, however, less clear. Whereas anasymptomatic gene carrier for Huntington'sdisease has approaching a 100% likelihood ofdeveloping clinical manifestations (barringother causes of death), there are currentlyonly limited data with which to counsel anasymptomatic carrier of a prion gene muta-tion; Collinge et al give an empiric risk of95%." Any reservations on the application ofthese mutation analyses in a dementedpatient would be removed should an effectivetreatment become available.

There have been no reservations, however,in applying molecular analyses of the dys-trophin gene and its product, despite theknowledge that a positive result identifies anuntreatable disorder. Hoffman et al34 haveshown that at least 10% of women with amyopathy and raised plasma creatine kinaseconcentration have an identifiable abnormal-ity in the dystrophin gene or the dystrophinprotein, or both. The perceived benefit thatthese women and their families can now becounselled on the appropriate risks of recur-rences would appear to have been justifica-tion enough for testing.The availability of analyses for the trinu-

cleotide repeat amplification in the androgenreceptor gene in bulbospinal neuronopathy35and mutations in the copper/zinc superoxidedismutase gene (SOD 1)36 in some familieswith autosomal dominant motor neuron dis-ease, makes feasible the screening of isolatedindividuals (especially males) with motorneuron disease. We are of the opinion that,whereas screening for the androgen receptorgene trinucleotide repeat expansion is valu-able (a normal result excludes Kennedy's dis-ease), SOD1 analysis in the absence of asignificant family history is not an appropriateuse of laboratory services.

Table 4 shows the mutation analyses thatwe consider should be available to neurolo-gists as part of a regional molecular geneticservice and the clinical situations in which wefeel the request is appropriate. Fragile X (typeA) mutation analysis is usually requested bypaedialricians when faced with a child withdevelopmental delay but could be added tothis list. It would however, add considerablyto the workload and would require appropri-ate additional staff.

12



j.com/

J Neurol N


ownloaded from



Table 4 Clinical indications for specified mutation analyses

Mutation analysis Clinical indication

Dystrophin gene "Limb girdle" dystrophy with raised CK*Raised neonatal creatine kinase

CMT1A duplication Demyelinating neuropathyMyotonin-kinase CTG repeat Muscular dystrophy with myotoniaIT1 5 CAG repeat Adult onset chorea

Juvenile rigid syndromeLHON mutations Acute bilateral optic atrophyMELAS/MERRF mutationst Myoclonic epilepsy, proximal myopathy with appropriate

histology, young stroke victims

*Ethically approved research protocol.tOn DNA extracted from blood and muscle.LHON = Leber's hereditory optic neuropathy; MELAS = mitochondrial myopathy,encephalopathy, lactic acidosis, stroke-like episodes; MERRF = myoclonus epilepsy withragged-red fibres.

The issue of which cases should be referredto a clinical geneticist in conjunction with therequest for a specific mutation analysis shouldbe addressed locally, taking into considera-tion the available staffing levels (see below)and the specific interests of the clinicians.Access to diagnostic molecular servicesshould not be limited by the availability of aclinical geneticist, although it should be astandard of acceptable practice that theappropriateness of the request for mutationanalysis should be discussed with a geneticscolleague whenever possible.Once a positive result has been generated

the clinician is duty bound to transmit thisinformation to the patient and, with his or herconsent, to the family. Clinical neurologistswill often find this a most useful time toengage the help of their clinical genetics col-leagues, if they were not approached initiallyto advise on what molecular test may havebeen applicable. Extended family counsellingon the nature and implications of a specificdiagnosis is often time consuming and mayinvolve other health professionals such as spe-cialist genetics nurses and clinical psycholo-gists if presymptomatic testing is requested bysiblings or offspring.

SITUATIONS WHERE MUTATION ANALYSIS ISINAPPROPRIATEWe are also aware of the situations where amolecular DNA test should not be under-taken. The most obvious situation is wherethe patient does not want it. This is no differ-ent from an individual declining a test forHIV carrier status. Most other situationsoccur where presymptomatic testing isrequested and these will usually be directedtowards the geneticist. Practising neurologistsshould be aware of these situations, however,as family members may approach them ini-tially, following the diagnosis of a heritabledisorder in the patient. Occasionally, an indi-vidual at risk for an inherited disorder mayapproach a neurologist with symptoms unre-lated to the genetic disease. In this situation itwould not be appropriate to offer specificmutation analysis. What is less clear is themerit of mutation analysis in the presence ofvague symptoms which may be diseaserelated, such as clumsiness or forgetfulness,where there is a family history ofHuntington's disease. The identification of adisease-specific mutation in this situation may

not offer a definitive explanation for the indi-vidual's symptoms.Many inherited neurological disorders have

onset in mid to later life and presymptomatictesting of children for these is not appropri-ate.37 This would change should effectivetherapy to prevent clinical disease becomeavailable. Individuals should not be underpressure from third parties, such as insurancecompanies, to test and those requesting test-ing should be made aware of the possibleinterests of such parties in the outcome. Itmay be useful to consider what impact a posi-tive test would have on one's own family andsocial situation to gain an appreciation of thepossible consequences for the individual inquestion.

Susceptibility genesThis article has focused on those neurologicaldisorders whose aetiology is primarily "singlegene" and where advances in our knowledgehave followed the isolation of that gene. Inmany of the most prevalent neurological dis-orders (multiple sclerosis, idiopathic epilepsy,stroke) there may be a significant geneticcomponent, as recognised from familial clus-tering of cases or from twin studies, butwhere true mendelian inheritance is not seen.Attention is now focusing on the identifica-tion of "susceptibility genes" which may pre-dispose an individual to the development ofthese conditions. Harding et aP8 have pro-posed that mitochondrial loci may be impli-cated in the susceptibility to develop multiplesclerosis. It has long been recognised thatthere is a higher frequency of specific HLAhaplotypes in individuals with multiple scle-rosis, suggesting additional immunologicalfactors (reviewed in reference 39). It is notknown how relevant any single susceptibilitylocus will turn out to be, nor how amenableto modification this predisposition will be inan individual.We are in no doubt, however, that individ-

uals with these disorders and their familieswill in future expect to receive counselling onthe genetic contribution to disease suscepti-bility and advice on strategies for modifyingtheir risks. We are equally convinced thatsociety will require considerable education onthe relevance of these factors, if individualswho have been identified to be at increasedrisk are not to be discriminated against ineducation, employment, and insurance.Legislative intervention may be necessary toavoid the creation of a genetically stigmatisedsubpopulation.

Applying molecular analysisSAMPLESMost analyses are carried out on DNA iso-lated from peripheral blood lymphocytescollected in containers with EDTA as an anti-coagulant. A sample of 20 ml of bloodprovides sufficient DNA for all analyses.Where polymerase chain reaction technologyis used, considerably less is required, so that

13 on M

arch 29, 2021 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.1.7 on 1 January 1994. Dow

nloaded from


MacMillan, Harper

the dried blood spot on a Guthrie card maybe adequate. Some mutations may be tissuespecific (such as some mitochondrial genemutations) and the appropriate sample willbe required if a false-negative result is to beavoided. Transportation of blood using thefastest available postal service (approximatelytwo days delay) is adequate provided thatarrangements have been made with the recip-ient laboratory to expect the sample. Tissuesuch as muscle biopsy specimens may requiresnap freezing in liquid nitrogen, especially ifRNA analysis is required, and again thisshould be discussed with the unit undertakingthe analysis.

SERVICESThe availability of a comprehensive moleculargenetic laboratory service is essential if one isto make optimal clinical use of the advancesoutlined in this review. The pace at whichmutation analysis is becoming feasible forgenetic disorders in general, precludes anysingle diagnostic unit from providing such aservice and many laboratories have imple-mented consortium arrangements to enablethe widest possible range of services to beavailable. Single gene neurological disordersdo, however, account for 54% of all clinicalreferrals to regional genetic units40 and itwould seem appropriate that a core molecularneurogenetic diagnostic service should beavailable at a regional level.We feel that the minimal provision should

include molecular analyses for the trinu-cleotide repeat expansions associated withmyotonic dystrophy and Huntington's dis-ease, the CMT1A duplication, a deletionscreening protocol for Duchenne and Beckermuscular dystrophies, a core mitochondrialmutation screen (see table 4) and linkageanalysis for the autosomal recessive spinalmuscular atrophies of childhood. The work-load generated at a regional level should beable to be handled by two trained laboratoryscientific officers working alongside col-leagues carrying out other analyses.Additional analyses will generally be availableby negotiation with other units. TheLaboratory and National ConsortiumDirectory, compiled by the ClinicalMolecular Genetics Society, lists by diseaseand centre the services available throughoutthe United Kingdom. This publication alsoindicates the target times by which a resultshould be available (these vary by disease andcentre).The provision of a clinical service to com-

plement the molecular analyses is essential.As we discussed above, it will often be neces-sary to undertake extended family counsellingwhen a genetic diagnosis is made. Most dis-tricts will have access to the services of a clini-cal geneticist and a specialist genetic nurse,who will be able to provide such a service. Wefeel that, in addition, there should be, at aregional level, the provision of a clinical spe-cialist with training in both clinical neurologyand clinical and molecular genetics. Thisindividual would be able to offer an optimal

clinical service for patients and colleagues andprovide informed interpretation of relevantmolecular DNA analyses.

ConclusionsMolecular genetic studies have begun to elu-cidate the pathobiological mechanisms under-lying many of the inherited neurologicaldisorders. Improved diagnostic accuracyenables a more accurate prognosis to be givento the patient and facilitates appropriategenetic counselling for the family unit. It islikely that more effective therapy will followfor some of these disorders but the prospectof a "cure", as perceived by the patient,remains a matter of speculation. This shouldnot prevent individuals and their familiesbenefiting from the advances made to date.

We wish to thank Dr M Upadhyaya for useful discussions onthe molecular genetics of facioscapulohumeral musculardystrophy, S Cochran, unpublished data on CMTX, Dr LLazarou for the autoradiograph used in the figure and Mr PDavies for discussions on the feasibility of the estimatedlaboratory workload. JC MacMillan is supported through agrant from the Muscular Dystrophy Group of Great Britainand Northern Ireland.

1 Emery AEH. Population frequencies of inherited neuro-muscular diseases-a world survey. Neuromusc Disord1991;1: 19-30.

2 MacMillan JC, Harper PS. Single gene neurological disor-ders in South Wales; an epidemiological study. AnnNeurol 1991;30:411-4.

3 Botstein D, White RL, Skolnick M, Davis RW.Construction of a genetic linkage map in man usingrestriction fragment length polymorphisms. Am Jf HumGenet 1980;32:314-31.

4 Haldane JBS. The combination of linkage values and thecalculation of distances between loci of linked factors. JGenet 1919;8:299-309.

5 Harding AE. The DNA laboratory and neurological prac-tice. J Neurol Neurosurg Psychiatry 1993;56:229-33.

6 Daniels RJ, Suthers GK, Morrison KE, et al. Prenatal pre-diction of spinal muscular atrophy. JT Med Genet1992;29: 165-70.

7 Wijurenga C, Frants R, Hewitt J, et al. Molecular geneticsof facioscapulohumeral muscular dystrophy. NeuromuscDTsord, 1993, in press.

8 Rossiter BJF, Caskey CT. Molecular studies of humangenetic disease. FASEBJ_ 1991;5:21-7.

9 Shoffner JM, Wallace DC. Mitochondrial genetics:Principles and practice. Am J Hum Genet 1992;51:1179-86.

10 Brown MD, Voljavec AS, Lott MT, MacDonald I,Wallace DC. Leber's hereditary optic neuropathy: Amodel for mitochondrial neurodegenerative disease.FASEB J 1992;6:2791-9.

11 Koenig M, Beggs AH, Moyer M, et al. The molecularbasis for Duchenne versus Becker muscular dystrophy:Correlation of severity with type of deletion. Am J HumGenet 1989;45:498-506.

12 Bradley DM, Parsons EP, Clarke AJ. Experience withscreening newborns for Duchenne muscular dystrophyin Wales. BMJ 1993;306:357-60.

13 Matsumura K, Tome FMS, Collin H, et al. Deficiency ofthe 50K dystrophin-associated glycoprotein in severechildhood autosomal recessive muscular dystrophy.Nature 1992;359:320-2.

14 Matsumura K, Nonaka I, Campbell KP. Abnormalexpression of dystrophin-associated proteins inFukuyama-type congenital muscular dystrophy. Lancet1993;341:521-2.

15 Upadhyaya M, Shen M, Cherryson A, et al. Analysis ofmutations at the neurofibromatosis 1 (NFI) locus. HumMol Genet 1993;1:735-40.

16 Pieretti M, Zhang F, Fu YH, et al. Absence of expressionof the FMR-1 gene in fragile X syndrome. Cell199 1;66:817-22.

17 Scott Hansen R, Gartler SM, Scott CR, Chen S, LairdCD. Methylation analysis of CGG sites in the CpGisland of the human FMR1 gene. Hum Mol Genet1992;1:571-8.

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

19 Fu YH, Friedman DL, Richards S, et al. Decreasedexpression of myotonin-protein kinase messanger RNAand protein in adult form of myotonic dystrophy.

14 on M


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



Science 1993;260:235-8.20 Hofmann-Radvanyi H, Lavedan C, Rabes J-P, et al.

Myotonic dystrophy: absence of CTG enlargedtranscript in congenital forms, and low expression of thenormal allele. Hum Mol Genet 1993;2:1263-6.

21 Sabouri LA, Mahadevan MS, Narang M, Lee DSC, SurhLC, Korneluk RG. Effect of the myonic dystrophy(DM) mutation on mRNA levels of the DM gene.Nature Genet 1993;4:233-8.

22 Harley HG, Rundle SA, MacMillan JC, et al. Size of theunstable CTG repeat sequence in relation to phenotypeand parental transmission in myotonic dystrophy. Am 7Hum Genet 1993;52:1164-74.

23 Suthers GK, Huson SM, Davies K. Instability versus pre-dictability: the molecular diagnosis of myotonic dystro-phy. J Med Genet 1992;29:761-5.

24 MacMillan JC, Snell RG, Tyler A, et al. Molecular analy-sis and clinical correlations of the Huntington's diseasemutation. Lancet 1993;342:954-8.

25 Snell R, MacMillan JC, Cheadle JP, et al. Relationshipbetween trinucleotide repeat expansion and phenotypicvariation in Huntington's disease. Nature Genet 1993;4:393-7.

26 MacMillan JC, Upadhyaya M, Harper PS. Charcot-Marie-Tooth disease type la (CMT1A): evidence fortrisomy of the region pl 12 of chromosome 17 in southWales families. J Med Genet 1992;29:12-3.

27 Suter U, Welcher AA, Ozcelik T, et al. Trembler mousecarries a point mutation in a myelin gene. Nature 1992;356:241-4.

28 Valentijn LJ, Baas F, Wolterman RA, et al. Identical pointmutations of PMP-22 in Trembler-J mouse andCharcot-Marie-Tooth disease type 1A. Nature Genet1992;2:288-9 1.

29 Hoogendijk JE, Hensels GW, Gabreels-Festen AAWM, etal. De-novo mutation in hereditary motor and sensoryneuropathy type I. Lancet 1992;339:1081-2.

30 Hyasaka K, Takada G, Ionasescu V. Mutation of the

myelin Po gene in Charcot-Marie-Tooth neuropathytype 1 B. Hum Mol Genet 1993;2:1369-72.

31 Fidani L, Rooke K, Chartier-Harlin M-C, et al. Screeningfor mutations in the open reading frame and promoterof the ,B-amyloid precursor protein gene in familialAlzheimer's disease: identification of a further familywith APP717 Val-Ile. Hum Mol Genet 1992;1:165-8.

32 Collinge J, Harding AE, Owen F, et al. Diagnosis ofGerstmann-Straussler syndrome in familial dementiawith prion protein gene analysis. Lancet 1989;334: 15-7.

33 Collinge J, Poulter M, Davis MB, et al. Presymptomaticdetection or exclusion of prion protein gene defects infamilies with inherited prion diseases. Am J Hum Gent1991;49:1351-4.

34 Hoffman EP, Arahata K, Minetti C, et al.Dystrophinopathy in isolated cases of myopathy infemales. Neurology 1992;42:967-75.

35 La Spada AR, Wilson M, Lubahn DB, Harding AE,Fischbeck KH. Androgen receptor gene mutations in X-linked spinal and bulbar muscular atrophy. Nature199 1;352:77-9.

36 Rosen DR, Siddique T, Patterson D, et al. Mutations inCu/Zn superoxide dismutase gene are associated withfamilial amyotrophic lateral sclerosis. Nature 1993;362:59-62.

37 Harper PS, Clarke A. Should we test children for "adult"genetic diseases? Lancet 1990;335:1205-6.

38 Harding AE, Sweeny MG, Miller DH, et al. Occurrenceof a multiple sclerosis-like illness in women who have aLeber's hereditary optic neuropathy mitochondrialDNA mutation. Brain 1992;115:979-89.

39 Jersild C, Grunnet N. Demyelinating disease. In Thegenetic basis of common diseases. King RA, Rotter JI,Motulsky AG, eds. Oxford: Oxford University Press,1992; 753-74.

40 Rona RJ, Swan AV, Wilson OM, et al. DNA probetechnology: implications for service planning in Britain.Clin Genet 1992;42:186-95.

15 on M


http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

Documents

Clinical genetics in neurological disease · clinical genetic andmoleculargeneticpractice in the management of the major inherited neurological diseases. Weconsider diagnostic issues