J. Clin Pathol 1993;46:977-981;;

Leaders:;:::::;::;::;::;::;::;::;::;: ;:; ;;;;;;;;;;:;;

Molecular genetics of disorders of haembiosynthesis

G H Elder

IntroductionHaem is synthesised in all animal cells fromsuccinyl CoA and glycine by a sequence ofreactions catalysed by eight enzymes.1Absence of any one of these enzymes isincompatible with life because haem is anessential component of respiratory proteins.Partial deficiencies occur, however, and causedisease (table 1).12 The activity of 5-amino-laevulinate (ALA) synthase, the first and ratecontrolling enzyme of the pathway, isdecreased in the bone marrow in hereditarysideroblastic anaemias,3 while deficiency ofeach of the subsequent enzymes produces aparticular type of porphyria (table 1) .1 2 Theenzyme defect in each porphyria is inherited(table 1), except in type I porphyria cutaneatarda which appears to be an acquired dis-order. The clinical features and laboratorydiagnosis of the porphyrias have beenreviewed.' 2

Genes or cDNAs have now been isolatedand characterised for all but two of thehuman enzymes (table 2). These advanceshave been followed rapidly by identificationof mutations in the corresponding diseasesand by the application of DNA techniques tothe management of families with these condi-tions. This review summarises current infor-mation about the molecular genetics of eachdisorder.

Departent ofMedicalBiochemistry,University of WalesCollege ofMedicine,Cardiff CF4 4XNG H ElderCorrespondence to:Professor G H Elder

Accepted for publication17 March 1993

Hereditary sideroblastic anaemiasThe hereditary sideroblastic anaemias are a

mixed group of disorders in which the mostfrequent pattern of inheritance indicates link-age to the X chromosome. Bone marrow

ALA synthase activity is usually decreased.'Pyridoxal phosphate is an essential cofactorfor this enzyme and patients may respond totreatment with pyridoxine.3Mammals have two ALA synthase genes

(table 2). The X chromosome gene encodesALAS2 which is expressed only in erythroidcells while that on chromosome 3 encodes a

housekeeping isoenzyme (ALAS 1) that isfound in all tissues.F6 The C-terminal portion

of ALAS2 (exons 5-11) has 73% sequenceidentity with ALAS1 and contains the cata-lytic domain. Exons 2-4 of the ALAS2 gene

encode the N-terminal signal sequence thatdirects mitochondrial import and an inter-vening section that undergoes differentialpre-mRNA splicing to generate two isoforms(table 2).6 The functional importance of thispattern of processing has not been estab-lished.6 Expression of the ALAS2 gene isregulated at transcriptional and translationallevels. Its promoter region, in common withthose of the erythroid PBGD,2' ferro-

chelatase,20 and ,B globin genes,2' containssequences that interact with the trans-actingerythroid specific factors, GATA-1 and NF-E2, and regulate transcription during ery-throid differentiation.5 The 5' untranslatedregion (UTR) of ALAS2 mRNA, unlikeALAS 1 mRNA, contains a sequence struc-ture motif that closely resembles the ironresponsive elements (IREs) found in the 5'UTR of ferritin mRNA and the 3' UTR oftransferrin receptor mRNA.i22 Iron deficiencyleads to high affinity binding of an IRE-bind-ing protein to these elements with consequentinhibition of translation of ALAS2 and fer-ritin mRNAs and stabilisation of transferrinreceptor mRNA,22 a mechanism that allowsco-ordination of protoporphyrin formationwith iron supply. The presequences of bothALAS isoenzymes also contain a conserved,haem regulatory motif which is involved ininhibition by haem of their transport into

24mitochondria.4The discovery that ALAS2 maps to the X

chromosome prompted a search for muta-tions in this gene in sideroblastic anaemia. AT-÷A transition that converts isoleucine toasparagine has recently been identified in aman with severe, pyridoxine responsivesideroblastic anaemia.25 This mutation is in ahighly conserved region of exon 9 that is closeto the putative pyridoxal phosphate bindinglysine residue and thus may impair binding ofthe cofactor. This man had no family historyof anaemia but a second mutation in theALAS2 gene has now been found in a familywith X linked pyridoxine responsive sidero-blastic anaemia.26 Because the low level ofexpression of ALAS 1 in erythroid cells isunlikely to be sufficient by itself for erythro-poiesis,5 other inherited ALAS2 defects insideroblastic anaemia are similarly likely toresult from point mutations that modify butdo not abolish enzyme activity. Not allfamilies with X linked sideroblastic anaemiahave defects in the ALAS2 gene. Combineduse of a highly polymorphic dinucleotiderepeat in intron 7 of the gene and other Xchromosome polymorphic markers in linkagestudies has shown that there are at least twoX loci for sideroblastic anaemia.26 Overall,these studies suggest that analysis of theALAS2 gene is likely to improve both diagno-sis and classification of these anaemias andmay even reveal a role for somatic mutationof this gene in some acquired forms.

Autosomal dominant porphyriasThe enzyme defects in most porphyrias areinherited in an autosomal dominant manner(table 1). The clinical penetrance of these

977

on July 20, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.46.11.977 on 1 N

ovember 1993. D

ownloaded from

Elder

Table 1 Disorders caused by partial deficiencies ofhaem biosynthetic enzymes

Disorder Enzyme deficiency Inhenitance

1 Sideroblastic anaemia ALA synthase (AIAS2) X-linked2 PBG deficiency porphyria PBG synthase (PBGS) AR3 Acute intermittent porphyria PBG deaminase (PBGD) AD4 Congenital erythropoietic porphyria Uroporphyrinogen synthase (UROS) AR5 Porphyria cutanea tarda Uroporphyrinogen decarboxylase (UROD) see text6 Hereditary coproporphyria Coproporphyrinogen oxidase (COPROX) AD7 Variegate porphyria Protoporphyrinogen oxidase (PROTOX) AD8 Erythropoietic protoporphyria Ferrochelatase (FC) AD

AR: autosomal recessive; AD: autosomal dominant; PBG: porphobilinogen.

disorders, however, is low; in each conditionmore than 80% of those who inherit thegenetic defect never have symptoms and mostof these have no detectable biochemicalabnormality apart from enzyme deficiency.The frequency of the genes for these condi-tions in the population is sufficiently high forhomozygous forms of each disorder to occurwithout consanguinity127 and for occasionalco-inheritance of two separate porphyrias.28

ACUTE INTERM=ITENT PORPHYRIAAcute intermittent porphyria (AIP) is thecommonest of the acute hepatic porphyriasand at present is the only one that has beeninvestigated at the DNA level. It is charac-terised by life-threatening acute neurovisceralattacks that are frequently precipitated bydrugs, calorie restriction, or alcohol.' Detec-tion of asymptomatic gene carriers so thatthey can be advised to avoid knownprecipitants of acute attacks is an importantpart of the management ofAIP families.' 2AIP is caused by porphobilinogen deami-

nase (PBGD) deficiency. Enzyme activity isclose to 50% of normal, reflecting expressionof the normal gene allelic to the mutant gene.As with ALA synthase, there are separateerythroid and housekeeping isoenzymes ofPBG deaminase (table 2)9; current evidencesuggests that there are no tissue specificisoenzymes for the other enzymes of thepathway. In contrast to ALA synthase, how-ever, PBGD is encoded by a single gene(table 2) which is transcribed from separateerythroid specific and housekeeping pro-moters.910 The erythroid promoter lies inintron 1 and transcription is initiated 5' toexon 2. Translation of erythroid mRNA startsin exon 3 so that the erythroid isoenzymecontains sequence encoded by exons 4-15and a 3' section of exon 3. RNA for thehousekeeping isoenzyme is transcribed from apromoter 5' to exon 1 and then spliced to

exclude exon 2. Translation then proceedsfrom a start codon in exon 1 so that thehousekeeping isoenzyme is 17 amino acidslonger at the N-terminus than the erythroidform. Mutations that affect both isoenzymeswould thus be predicted to occur in exons3-15 or in intronic sequences that determinethe structure or level of expression of thecorresponding mRNA.

Three subtypes of AIP can be distin-guished by measurement of erythrocytePBGD: an uncommon (less than 5% offamilies) form in which erythrocyte PBGD isnormal with only the ubiquitous isoenzymebeing defective; a form in which the productof the mutant allele cross-reacts immuno-logically with antiserum to normal enzymebut has absent or substantially impaired cata-lytic activity (CRIM positive); and CRIMnegative AIP in which immunoreactivity andcatalytic activity are decreased in parallel.Mutations have been identified in all threesubtypes. In the non-erythroid subtype twomutations that affect the splicing of exon 1during pre-mRNA processing and thusimpair the formation of mRNA for the ubiq-uitous isoenzyme only have been found inseven of 10 unrelated patients.29 About 15%of unrelated patients with AIP have theCRIM positive subtype. Studies from theNetherlands and France suggest that about75% of these patients have mutations thatlead to replacement of either of two con-

served arginine residues (R167 and R173) inexon 10 by glutamine or tryptophan with one

mutation (G-+A at base position 500;R167Q) being more frequent than theothers.3031 These mutations severely impairbut do not abolish enzyme activity.32 Inter-estingly, the only two unrelated cases ofhomozygous AIP that have been reportedwere both compound heterozygotes for threeof the CRIM positive exon 10 mutations,3suggesting that their frequency may be higherthan the prevalence of overt CRIM positiveAIP would indicate. The structure of PBGdeaminase is highly conserved with over 45%amino acid sequence identity between theEscherichia coli and human enzymes. X-rayanalysis of the crystal structure of the E colienzyme shows that the arginine residues cor-

responding to these two residues in exon 10of human PBGD and a third one in exon 3,which is the site of another CRIM positivemutation (R26H) in AIP,34 form salt bridgeswith the acidic side chains of the dipyrro-

Table 2 Molecular genetics ofhuman haem biosynthesiscDNA in base pairs

Enzyme Gene (length of ORF) Expression Chromosome References

ALAS1 1920(168) U 3p21 45ALAS2 22kb, 11 exons 1746(147) E Xpll-21 46

1635(147) EPBGS 990 U 9q34 7PBGD 1Okb, 15 exons 1032 E llq 81011

1083 U 24-1-q24-2 9 10UROS 45kb 795 U 10q25-2-q26-3 12-14UROD 3kb, 10 exons 1101 U 1p34 1516COPROX 9 17PROTOX 14 18FC 45kb, 11 exons 1269(162) U 18q21-3 1920

ORF: open reading frame. Number of bases encoding N-terminal signal peptides for mitochondrial import are shown inparentheses.U: ubiquitous; E: erythroid cells only.

978

on July 20, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.46.11.977 on 1 N

ovember 1993. D

ownloaded from

Molecular genetics of disorders ofhaem biosynthesis

methane cofactor that lies within the cleftbetween domains 1 and 2.5 CRRIM positiveAIP thus seems to result from mutation ofresidues that are both important for catalysisand located deep within the active site cleft sothat their substitution does not alter surfaceepitopes.The commonest subtype, CRIM negative

AIP, seems to be more heterogeneous thanthe others. At least 20 base substitutions,insertions, or deletions that produce mis-sense, nonsense, or splice defective mutationshave been identified.3'3-2 Most have beenfound in only one or two families but a non-sense mutation in exon 10 (Q198 stop) iscommon in Sweden, where it seems to havespread through a founder effect from a familyoriginating in Lapland and to explain the highfrequency of overt AIP in this country (1 in

1500 in the north).40 About one third ofDutch families with AIP share an Ri 16Qmutation, again suggesting a founder effect.42DNA analysis has an important practical

application in AIP for the identification ofasymptomatic gene carriers. Several stud-ies43-46 have shown its superiority for thispurpose over conventional biochemicaltechniques, which have a number of draw-backs.2 At present, the following would seem

a reasonable strategy for detecting carriersby DNA techniques. For families in whichthe mutation is known, methods basedon polymerase chain reaction (PCR) fortheir detection will give unequivocalassignments.2 3434042 Similarly, patients withthe non-erythroid or CRIM positive subtypes,in which certain mutations are common, orfrom areas where one mutation is frequent,can be rapidly screened for the presence ofthese mutations.29 30 For those families inwhich the mutation is unknown, which atpresent means most, two approaches are

possible: gene tracking using linkage tointragenic polymorphisms43-46 or identifi-cation of the mutation in the probandfollowed by direct screening of relatives.3'Seven intragenic two allele polymorphismshave been identified in the PBGD gene(table 3).3 43454748 Three are in complete link-age disequilibrium and there is partial linkagedisequilibrium between others so that the fullnumber of haplotypes is not observed.4549Most patients, however, are heterozygous atone allele or more and study of their familiesis therefore potentially informative. The maindisadvantage is that it is often impossible toestablish linkage because not enoughunequivocally affected relatives are availablefor investigation. An alternative but more

laborious approach is to identify the mutationin genomic DNA or cDNA from the familyunder investigation by using a rapid scanningmethod, such as denaturing gradient gelelectrophoresis (DGGE), to locate the muta-tion, followed by direct sequencing.""DGGE of seven exons and their flankingsequences after PCR amplification identifiedthe mutations in 18 of 43 families, suggestingthat most mutations would be detected ifinvestigation was extended to all the exons.

At present, gene tracking is likely to be thefirst choice for most laboratories, but the

Table 3 Human PBG deaminase gene: intragenictwo-allele polymorphisms

Polymorphism Site FrequencyC/T exon 1 0-63(C)MspI intron 1 0 40PstI intron 1 0-60ApaLI, SnoI intron 1 0-40BstNI, ScrFI, intron 1 0-55G/A intron 3 -0-60(G)G/T exon 10 0-65(G)

For RFLPs, the frequency is that of the uncut allele.MspI, PstI, and ApaLI RFLPs are in complete linkage dis-equilibrium.

position may change once the full extent ofthe molecular heterogeneity of AIP has beenestablished and as improved methods forrapid screening for large numbers of knownmutations become available.

OTHER AUTOSOMAL DOMINANT PORPHYRIASPorphyria cutanea tarda is a cutaneousporphyria that results from decreased activityof uroporphyrinogen decarboxylase (UROD)in the liver. Its pathogenesis is complex butone form (type II), which accounts for about20% of cases, is associated with autosomaldominant inheritance of UROD deficiency inall tissues.' More severe UROD deficiency(activity less than 25% of normal) charac-terises the much rarer, presumed homozy-gous form of type II porphyriac cutaneatarda, hepatoerythropoietic porphyria (HEP),which is biochemically heterogeneous.50 Fivepoint mutations and a deletion of the URODgene have now been identified in patientswith inherited deficiencies of this enzyme.5"Contrary to the anticipated relationship,mutations identified in patients with HEPhave not yet been found in those presentingwith type 11."' For example, the G281E muta-tion, which decreases enzyme stability52 andwas present in one Tunisian and five Spanishfamilies with HEP, was not detected in 12type II porphyria cutanea tarda families fromSpain (Roberts AG, Elder GH, unpublishedobservations). The UROD defect in type IIporphyria cutanea tarda is CRIM negativeand it may be that most of the causativemutations are too severe for homozygotesto survive. Garey et al53 found that five of22 North American families with type IIporphyria cutanea tarda had a splice sitemutation producing deletion of exon 6 frommRNA and encoding an inactive, unstabletruncated protein. This mutation was notpresent in 25 European families (Roberts AG,Elder GH, unpublished observations), whichtogether with other studies of the frequencyof mutant uroporphyrinogen decarboxy-lases,5' suggests extensive molecular hetero-geneity in this group of disorders.

Progress has also been made in under-standing the molecular basis of erythropoieticprotoporphyria (EPP). This disorder is char-acterised by severe photosensitivity; in a fewpatients accumulation of protoporphyrin inhepatocytes leads to irreversible liver failure.'In most families EPP is inherited as an auto-somal dominant trait with low penetrance,27but other modes of inheritance have beenpostulated.54 Animal models of the disease,whether occurring naturally in cattle,55 or

979

on July 20, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.46.11.977 on 1 N

ovember 1993. D

ownloaded from

Elder

produced by ethylnitrosourea mutagenesis inmice,56 are inherited as autosomal recessivetraits. Accurate identification of carriers isimportant if families are to be counselledadequately.EPP results from deficiency of ferro-

chelatase (table 1). The activity of thisenzyme is less than the half-normal thatwould be expected for an autosomal domi-nant disorder.' 27 Functional ferrochelatasemay be a homodimer57; if this is so, inter-action between normal and mutant subunitsmight decrease activity by greater than50%.7 58 Five point mutations that cause EPPhave recently been identified in the fer-rochelatase gene. One patient with the rarehomozygous form of EPP was a compoundheterozygote for two different missense muta-tions.59 In three other families autosomaldominant inheritance of mutations causingaberrant splicing606' or an amino acid substi-tution leading to decreased activity58 werefound. It thus seems likely that carrier detec-tion in EPP will be complicated by extensivemolecular heterogeneity. Linkage analysis inEPP families, however, should be facilitatedby the recent discovery of a dinucleotiderepeat polymorphism in intron 2 of thegene.62 At present, there is no evidence thatparticular genotypes are associated with thepropensity to develop liver disease.

Autosomal recessive porphyriasTwo types of porphyria are inherited in anautosomal recessive pattern: PBGS(5-amino-laevulinate dehydratase) deficiency porphyriaand congenital erythropoietic porphyria(CEP) (table 1).

Patients from the four unrelated familieswith PBGS deficiency porphyria that havebeen described show substantial phenotypicvariation, ranging from failure to thrive ininfancy through the onset of acute attacks ofporphyria soon after puberty, to subacutepolyneuropathy at the age of 63.63 Patientswith the infantile and teenage onset formshave been shown to be compound heterozy-gotes for four different missense mutations inthe PBGS gene63 64; expression studies showthat three of these substantially decreaseenzyme activity, but the other, which waspresent in a teenage patient,64 was less severe,suggesting some correlation between genotype andphenotype in this very rare condition.

Congenital erythropoietic porphyria is theleast common of the main types of porphyriaapart from PBGS deficiency. Typically, itpresents in early infancy with severe skinlesions that progress to photomutilationwhich are accompanied by splenomegaly andhaemolytic anaemia of variable severity.There is also a less severe form with onsetafter childhood.'27 There are no reports ofboth forms occurring in the same familywhich suggests that they may be produced bydifferent genotypes. Congenital erythropoieticporphyria is caused by decreased activity ofUROS (table 1). Eight mutations have beenidentified in the UROS gene or cDNA inpatients with congenital erythropoietic por-phyria; one of these (C73R) accounts forabout 20% of congenital erythropoietic

porphyria alleles.65-7 Most patients are com-pound heterozygotes and no clear relation-ship between genotypes and the two differentphenotypes has yet been established. Homo-zygotes for the C73R mutation, however, areseverely affected while patients with a T228Mmutation of one allele seem to have lesssevere disease.67 The infantile form of con-genital erythropoietic porphyria is probablythe only type of porphyria for which prenataldiagnosis should be available. DNA analysisshould allow this to be achieved with greaterprecision than is currently possible.

The futureSeveral important problems in the pathogene-sis and management of inherited disorders ofhaem biosynthesis now seem likely to besolved by the application of recombinantDNA methods. In the acute hepatic por-phyrias, the way towards much more accuratedetection of gene carriers is already clear fromstudies on AIP; similar progress for variegateporphyria and hereditary coproporphyriashould follow once the relevant genes havebeen cloned. The influence of genetic factorsin determining phenotypic variation and dif-ferences between families in the penetrance ofthe autosomal dominant porphyrias remainsto be explored. For example, genes thatdirectly or indirectly influence the inducibilityof ALAS 1 in the liver may be importantdeterminants of susceptibility to acute por-phyria. Finally, transgenic animal experi-ments may help to solve outstandingproblems, such as the mechanism of the acuteporphyric attack, and lead to new treatments.

1 Kappas A, Sassa S, Galbraith RA, Nordmann Y. The por-phyrias. In: Scriver C, Beaudet A, Sly W. Valle D, eds.The metabolic basis of inherited disease. Ann ClinBiochem 1990;27:395-412.

3 Bottomley SS. The sideroblastic anemias. In: Lee RG,Bithell TC, Foester J, Athens JW, Lufkens JN, eds.Wintrobe's clinical haematology. Philadelphia: Lea &Febiger, 1993:852-71.

4 Bishop DF. Two different genes encode o-aminolevulinatesynthase in humans: nucleotide sequences of cDNAs forthe housekeeping and erythroid genes. Nucleic Acids Res1990;18:7187-8.

5 Cox TC, Bawden MJ, Martin A, May BK. Human ery-throid 5-aminolevulinate synthase: promoter analysisand identification of an iron-responsive element in themRNA EMBO J' 1992;10: 1891-902.

6 Conboy JG, Cox TC, Bottomley SS, et al. Human ery-throid 5-aminolevulinate synthase. J Biol Chem 1992;267:18753-8.

7 Wetmur JG, Bishop D, Cantelmo C, Desnick RJ. Humanb-aminolevulinate dehydratase: nucleotide sequence of afull-length cDNA clone. Proc Nad Acad Sci USA 1986;83:7703-7.

8 Raich N, Romeo PH, Dubart A, et al. Molecular cloningand complete primary sequence of human erythrocyteporphobilinogen deaminase. Nucleic Acids Res 1986;14:5955-68.

9 Grandchamp B, De Verneuil H, Beaumont C, et al.Tissue-specific expression of porphobilinogen deami-nase. Two isoenzymes from a single gene. EurI Biochem1987;162: 105-10.

10 Chretien S, Dubart A, Beaupain D, et al. Alternative tran-scription and splicing of the human porphobilinogendeaminase gene result either in tissue-specific or inhousekeeping expression. Proc Nad Acad Sci USA 1988;85:6-10.

11 Namba H, Narahara K, Tsuji Y, Yokoyama Y, Seino Y.Assignment of human porphobilinogen deaminase to1lq24 1-+q24-2 by in situ hybridisation and genedosage studies. Cytogenet Cell Genet 1991;57:105-8.

12 Tsai S-F, Bishop D, Desnick R. Human uroporphyrino-gen III synthase: Molecular cloning, nucleotidesequence and expression of a full-length cDNA. ProcNal Acad Sci USA 1988;85:7049-53.

13 Warner CA, Yoo HW, Tsai S-F, et al. Congenital erythro-poietic porphyria: characterization of the genomicstructure and identification of mutations in the uropor-phyrinogen Ill synthase gene. Am J Hum Genet 1990;47(supplement):321A.

14 Astrin KH, Warner CA, Han-Wook Yoo, et al. Regional

980

on July 20, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.46.11.977 on 1 N

ovember 1993. D

ownloaded from

Molecular genetics of disorders ofhaem biosynthesis

assignment of the human uroporphyrinogen m synthase(UROS) gene to chromosome 10q25 2-q-26 3. HumGenet 1991;87:18-22.

15 Romeo P-H, Raich N, Dubart A, et al. Molecular cloningand nucleotide sequence of a complete human uropor-phyrinogen decarboxylase cDNA. J7 Biol Chem 1986;261:9825-31.

16 Romana M, Dubart A, Beaupain D, et al. Structure of thegene for human uroporphyrinogen decarboxylase.Nucleic Acids Res 1987;15:7343-55.

17 Grandchamp B, Weil D, Nordmann YN, et al.Assignment of the human coproporphyrinogen oxidaseto chromosome 9. Hum Genet 1983;64:180-3.

18 Bissbort S, Hitzeroth HW, Wentzel DP du, et al. Linkagebetween the variegate porphyria (VP) and the alpha-l-antitrypsin (PI) genes on human chromosome 14. HumGenet 1988;79:289-90.

19 Nakahashi Y, Taketani S, Okuda M, et al. Molecularcloning and sequence of cDNA encoding human ferro-chelatase. Biochem Biophys Res Comm 1990;173:748-55.

20 Taketani S, Inazawa J, Nakahashi Y, et al. Structure of thehuman ferrochelatase gene. Eur J Biochem 1992;105:217-22.

21 Orkin H. Globin gene regulation and switching: circa1990. CeU 1990;63:665-72.

22 Danekar T, Stripecke R, Gray NK, et al. Identification ofa novel iron-responsive element in murine and humanerythroid -aminolevulinic acid synthase mRNA.EMBOJ3 1991;10:1903-91.

23 Haile D, Rouault TA, Harford JB, et al. Cellular regula-tion of the iron-responsive element binding protein:Disassembly of the cubane iron-sulfur cluster results inhigh-affinity RNA binding. Proc Natl Acad Sci USA1992;89:11735-9.

24 Lathrop JT, Timko MP. Regulation by heme of mito-chondrial protein transport through a conserved aminoacid motif. Science 1993;259:522-5.

25 Cotter PD, Baumann M, Bishop DF. Enzymatic defect in"X-linked" sideroblastic anaemia: molecular evidencefor erythroid -aminolevulinate synthase deficiency. ProcNaedAcad Sci USA 1992;89:4028-32.

26 Cox TC, Kozman HM, Raskind W, et al. Identification ofa highly polymorphic marker within intron 7 of theALAS2 gene and suggestion of at least two loci for X-linked sideroblastic anemia. Hum Mol Genet 1992;1:639-41.

27 Nordmann Y, Deybach JC. Human hereditary porphyrias.In: Dailey HA, eds. Biosynthesis of heme and chorophylls.New York: McGraw-Hill, 1990:491-542.

28 Doss M. New dual form of porphyria. Lancet 1988;i:945-6.

29 Bourgeois F, Gu X-F, Deybach JC, et al. Denaturing gra-dient gel electrophoresis for rapid detection of latentcarriers of a subtype of acute intermittent porphyriawith normal erythrocyte porphobilinogen deaminaseactivity. Clin Chem 1992;38:93-5.

30 Gu X-F, Rooij F de, Voortman G, et al. High frequency ofmutations in exon 10 of the porphobilinogen deaminasegene in patients with a CRIM-positive subtype of acuteintermittent porphyria. Am J? Hum Genet 1992;51:660-5.

31 Gu X-F, de Rooji F, Vooruman G, et al. Detection ofmutations responsible for acute intermittent porphyriausing denaturing gradient gel electrophoresis. HumGenet (in press).

32 Delfau MH, Picat C, de Rooij FWM, et al. Two differentpoint G to A mutations in exon 10 of the porphobilino-gen deaminase gene are responsible for acute intermit-tent porphyria. Jf Clin Invest 1990;86:1511-6.

33 Llewellyn DH, Smyth SJ, Elder GH, et al. Homozygousacute intermittent plorphyria caused by adjacent basetransitions in the same codon of the porphobilinogendeaminase gene. Hum Genet 1992;89:97-8.

34 Llewellyn DH, Whatley S, Elder GH. Acute intermittentporphyria caused by an arginine to histidine substitution(R26H) in the co-factor binding cleft of porphobilino-gen deaminase. Hum Mol Genet 1993;2:1315-6.

35 Louie GV, Brownlie PD, Lambert R, et al. Structure ofporphobilinogen deaminase reveals a flexible multi-domain polymerase with a single catalytic site. Nature1992;359:33-40.

36 Grandchamp B, Picat C, de Rooij FWM, et al. A pointmutation G/A in exon 12 of the PBG deaminase generesults in exon skipping and is responsible for acuteintermittent porphyria. Nucleic Acids Res 1989;27:6637-49.

37 Llewellyn DH, Urquhart A, Scobie G, et al. Molecularanalysis of acute intermittent porphyria. Biochem SocTrans 1988;16:799-80.

38 Scobie GA, Llewellyn DH, Urquhart AJ, et al. Acuteintermittent porphyria caused by a C-+T mutation thatproduces a stop codon in the porphobilinogen deami-nase gene. Hum Genet 1990;85:631-4.

39 Delfau MH, Picat C, De Rooij F, et al. Molecular hetero-geneity of acute intermittent porphyria: identificationof four additional mutations resulting in the CRIM-negative subtype of the disease. Am J7 Hum Genet 1991;49:421-8.

40 Lee J-S, Anvret M. Identification of the most commonmutation within the porphobilinogen deaminase gene inSwedish patients with acute intermittent porphyria. ProcNatlAcad Sci USA 1991;88:10912-5.

41 Mgone CS, Lanyon WG, Moore MR, Connor JM.Detection of seven point mutations in the porphobilino-

gen deaminase gene in patients with acute intermittentporphyria by direct sequencing of in vitro amplifiedcDNA. Hum Genet 1992;90:12-6.

42 Gu X-F, de Rooij F, Lee JS, et al. High prevalence of apoint mutation in the porphobilinogen deaminase genein Dutch patients with acute intermittent porphyria.Hum Genet 1993;91:128-30.

43 Llewellyn DH, Elder GH, Kalsheker NA, et al. DNApolymorphism of human porphobilinogen deaminasegene in acute intermittent porphyria. Lancet 1987;ii:706-8.

44 Grandchamp BN, Picat C, Mignotte V, et al. Tissue-spe-cific splicing mutation in acute intermittent porphyria.Proc Nate Acad Sci USA 1989;86:661-4.

45 Lee J-S, Lundin G, Lannfelt L, et al. Genetic heterogene-ity of the porphobilinogen deaminase gene in Swedishfamilies with acute intermittent porphyria. Hum Genet1991;87:484-8.

46 Kauppinen R, Peltonen L, Palotie A, Mustajoki P. RFLPanalysis of three different types of acute intermittentporphyria. Hum Genet 1990;85:160-4.

47 Gu XF, Lee J-S, Delfau MH, Grandchamp B. PCRdetection of a polymorphism at exon 10 of the porpho-bilinogen deaminase gene (PBG-D). Nucleic Acids Res1991;19: 1966.

48 Picat C, Bourgeois F, Grandchamp B. PCR detection of aC/T polymorphism in exon 1 of the porphobilinogendeaminase gene (PBGD). Nucleic Acids Res 1991;19:5099.

49 Scobie GA, Urquhart AJ, Elder GH, et al. Linkagedisequilibrium between DNA polymorphisms within theporphobilinogen deaminase gene. Hum Genet 1990;85:157-9.

50 K6sz6 F, Elder GH, Roberts A, Simon N. Uroporphy-rinogen decarboxylase deficiency in hepatoerythro-poietic porphyria: further evidence for geneticheterogeneity. BrjDermatol 1990;122:365-70.

51 Vemeuil H de, Bourgeois F, Rooij F de, et al.Characterization of a new mutation (R292G) and adeletion at the human uroporphyrinogen decarboxylaselocus in two patients with hepatoerythropoietic por-phyria. Hum Genet 1992;89:548-52.

52 Verneuil H de, Grandchamp B, Beaumont C, et al.Uroporphyrinogen decarboxylase structural mutant(Gly 281-+ Glu) in a case or porphyria. Science 1986;234:732-4.

53 Garey JR, Harrison LM, Franklin KF, et al.Uroporphyrinogen decarboxylase: a splice site mutationcauses the deletion of exon 6 in multiple families withporphyria cutanea tarda. J7 Clin Invest 1990;86: 1416-22.

54 Norris PG, Nunn AV, Hawk JLM. Genetic heterogeneityin erythropoietic protoporphyria; a study of the enzy-matic defect in nine affected families. Invest Dermatol1990;95:260-3.

55 Straka JG, Hill HD, Krickava JM, et al. Immunochemicalstudies of ferrochelatase protein: characterization of thenormal and mutant protein in bovine and human proto-porphyria. Am Jf Hum Genet 199 1;48:72-8.

56 Tutois S, Montagutelli X, Da Silva V, et al. Erythropoieticprotoporphyria in the house mouse. A recessive inher-ited ferrochelatase deficiency with anemia, photosensi-tivity and liver disease. Jf Clin Invest 1991;88:1730-6.

57 Straka JG, Bloomer JR, Kempner ES. The functional sizeof ferrochelatase determined in situ by radiation inacti-vation. Jf Biol Chem 1991 ;266:24637-41.

58 Brenner DA, Didier JM, Frazier F, et al. A moleculardefect in human protoporphyria. Am Jf Hum Genet1992;50:1203-10.

59 Lamoril J, Boulechfar S, Verneuil de H, et al. Humanerythropoietic protoporphyria: two point mutations inthe ferrochelatase gene. Biochem Biophys Res Comm1991;18:281-5.

60 Nakahashi Y, Fujita H, Taketani S, et al. The moleculardefect of ferrochelatase in a patient with erythropoieticprotoporphyria. Proc NadAcad Sci USA 1992;89:281-5.

61 Sarkany RPE, Whitcombe DM, Cox TM. Erythropoieticprotoporphyria: cloning, sequencing and chromosomallocalization of the ferrochelatase gene. Abnormal RNAsplicing in an affected family. Br Jf Dermatol 1992;127(Suppl 40):19.

62 Whitcombe DM, Cox TM. Dinucleotide repeat polymor-phism at the locus for human ferrochelatas (FECH).EMBO Jf (in press).

63 Plewinska M, Thunell S, Holmberg L, et al. 8-Aminolevulinate dehydratase deficient porphyria:identification of the molecular lesions in aseverely affected homozygote. Am J7 Genet 1991;49:167-74.

64 Ishida N, Fujita H, Fukuda Y, et al. Cloning and expres-sion of the defective genes from a patient with b-aminolevulinate dehydratase porphyria. 7 Clin Invest1992;89: 1431-7.

65 Deybach JC, Verneuil H de, Boulechfar S, et al. Pointmutations in the uroporphyrinogen III synthase gene incongenital erythropoietic porphyria. Blood 1990:75:1763-5.

66 Warner CA, Yoo H-W, Roberts AG, Desnick RJ.Congenital erythropoietic porphyria: identification andexpression of exonic mutations in the uroporphyrinogenm synthase gene. 7 Clin Invest 1992;89:693-700.

67 Boulechfar S, DaSilva V, Deybach J-C, et al.Heterogeneity of mutations in the uroporphyrinogen IIIsynthase gene in congenital erythropoietic porphyria.Hum Genet 1992;88:320-4.

981

on July 20, 2021 by guest. Protected by copyright.

http://jcp.bmj.com

/J C

lin Pathol: first published as 10.1136/jcp.46.11.977 on 1 N

ovember 1993. D

ownloaded from