identification and characterization of hydroxymethylbilane synthase mutations causing acute...

Identification and Characterization ofHydroxymethylbilane Synthase Mutations CausingAcute Intermittent Porphyria: Evidence for anAncestral Founder of the Common G111R Mutation

A. De Siervi,1,2 M. V. Rossetti,1,2 V. E. Parera,2 K. H. Astrin,1 G. I. Aizencang,1 I. A. Glass,1A. M. del C. Batlle,2 and R. J. Desnick1*1Department of Human Genetics, Mount Sinai School of Medicine, New York, New York2Centro de Investigaciones sobre Porfirinas y Porfirias, University of Buenos Aires, Buenos Aires, Argentina

Acute intermittent porphyria (AIP), themost common hepatic porphyria, resultsfrom the half-normal activity of hydroxy-methylbilane synthase (HMB-synthase;EC 4.3.1.8), the third enzyme in the hemebiosynthetic pathway. Because life-threat-ening acute neurologic attacks of this auto-somal dominant disease are triggered byvarious ecogenic factors (e.g., certain drugs,hormones, alcohol, and starvation), effortshave been directed to identify and counselpresymptomatic heterozygotes in affectedfamilies to avoid the precipitating factors.Thus, to determine the nature of the muta-tions causing AIP in 26 unrelated enzyme-confirmed patients from Argentina, a long-range polymerase chain reaction methodwas developed to amplify the entire 10-kbgene in two fragments for efficient cyclesequencing and mutation detection. Eightnew mutations were identified includ-ing two missense mutations (Q34P andG335S), four small deletions (728delCT,815delAGGA, 948delA, and 985del12), asingle base insertion (666insA), and a splicesite mutation (IVS12+1). In addition, fivepreviously reported mutations (G111R,R173W, Q204X, R201W, and 913insC) weredetected. Notably, G111R was identified in

12 of the 26 (46%) presumably unrelated pro-positi; however, haplotype analysis with in-tragenic and flanking markers indicated anancestral founder. Expression of the twonew missense mutations (Q34P and G335S)in E. coli resulted in 2.5% or less of the nor-mal expressed enzyme, confirming their de-fective function. Thus, eight new and fivepreviously reported HMB-synthase muta-tions, including a common lesion, were de-tected, permitting accurate identificationand counseling of presymptomatic carriersin these 26 unrelated Argentinean AIP fami-lies with this dominant porphyria. Am. J.Med. Genet. 86:366–375, 1999.© 1999 Wiley-Liss, Inc.

KEY WORDS: porphyria; acute intermit-tent porphyria; hydroxy-methylbilane synthase; HMB-synthase deficiency; muta-tion analysis; ArgentinianAIP patients

INTRODUCTION

Acute intermittent porphyria (AIP) is an autosomaldominant inborn error of heme biosynthesis resultingfrom the half-normal activity of hydroxymethylbilanesynthase (EC 4.3.1.8; HMB-synthase) [Kappas et al.,1995; McGovern et al., 1996]. This enzyme, also knownas porphobilinogen deaminase or uroporphyrinogen Isynthase, catalyzes the head-to-tail condensation offour molecules of porphobilinogen (PBG) to form thelinear tetrapyrrole, hydroxymethylbilane. HMB-synthase is encoded by a single gene localized to thechromosomal region 11q24.1 → q24.2 [Namba et al.,1991]. The cDNA [Raich et al., 1986; Grandchamp etal., 1987] and entire 10-kb gene have been sequenced,including the 58regulatory, 38 untranslated, and in-tronic regions [Yoo et al., 1993], and contains 15 exons

Contract grant sponsor: National Institutes of Health; Contractgrant numbers: 5 R01 DK26824, 5 M01 R00071, 5 P30 HD28822;Contract grant sponsor: Argentinean National Research Council;Contract grant number: 00509108/97; Contract grant sponsor:Science and Technology Agency of Argentina; Contract grantnumber: PMT-PICT002b/97; Contract grant sponsor: Universityof Buenos Aires; Contract grant number: EX032/95-98.

*Correspondence to: R.J. Desnick, M.D., Ph.D., Professor andChairman, Department of Human Genetics, Box 1498, MountSinai School of Medicine, Fifth Avenue at 100th Street, NewYork, NY 10029. E-mail: [email protected]

Received 1 April 1999; Accepted 9 June 1999

American Journal of Medical Genetics 86:366–375 (1999)

© 1999 Wiley-Liss, Inc.

and two distinct promoters that generate housekeepingand erythroid-specific transcripts by alternative splic-ing [Chretien et al., 1988; Grandchamp et al., 1987].The housekeeping promoter is in the 58 flanking regionand its transcript contains exons 1 and 3–15 while theerythroid-specific promoter is in intron 1 and its tran-script is encoded by exons 2–15. The housekeepingtranscript is expressed in all tissues and its promoterhas certain aspects characteristic of housekeeping pro-moters, while the intron 1 promoter is active in ery-throid tissues and has regulatory elements similar tothose of the b-globin gene promoter [Grandchamp etal., 1987; Mignotte et al., 1989]. cDNAs encoding the44-kD housekeeping and 42-kD erythroid-specific iso-enzymes, which differ only in their NH2-terminalamino acid sequences, have been isolated and charac-terized [Chretien et al., 1988].

AIP is the most common acute hepatic porphyria andoccurs in all races and ethnic groups [Elder, 1993]. Twomajor subtypes of AIP have been delineated. In classi-cal AIP (∼90% of the AIP families), both the housekeep-ing and erythroid-specific enzymes have half-normalactivities, whereas in variant AIP (representing ∼5% ofAIP families) the housekeeping enzyme has half-normal activity, while the erythroid-specific enzyme isexpressed at normal levels [Mustajoki, 1981]. To date,the molecular lesions in three unrelated families withvariant AIP have been identified [Chen et al., 1994;Grandchamp et al., 1989a, 1989b]. Clinical onset of thedisease typically occurs during or after puberty and ischaracterized by intermittent attacks of neurologicaldysfunction, abdominal pain and other gastrointestinalcomplaints, hypertension, tachycardia, and various pe-ripheral and central nervous system manifestations.Expression of the disease is highly variable, deter-mined in part by environmental, metabolic, and hor-monal factors that induce hepatic d-aminolevulinic acidsynthase activity, the first and rate-limiting enzyme ofheme biosynthesis, thereby increasing the productionof the porphyrin precursors, d-aminolevulinic acid(ALA), and PBG [Kappas et al., 1995; McGovern et al.,1996]. The half-normal hepatic HMB-synthase activityin AIP patients is insufficient to convert all the newlysynthesized PBG to HMB so that the accumulation ofPBG and ALA are associated with the acute, life-threatening symptoms of the disease.

Identification of presymptomatic heterozygotes infamilies with affected individuals and the counseling ofthese at-risk relatives to avoid specific precipitatingfactors are the primary forms of medical managementfor this disease. Whereas symptomatic heterozygotesexcrete increased levels of ALA and PBG, and can beidentified easily by measurement of these urinary por-phyrin precursors, asymptomatic heterozygotes usu-ally have normal levels of urinary ALA and PBG, andtheir diagnosis has been based on their levels of eryth-rocyte HMB-synthase enzyme activity. However, theenzyme assay is only about 80% accurate in identifyingasymptomatic carriers, primarily due to the significantoverlap between high heterozygote and low normal val-ues [Lamon et al., 1978; Bottomley et al., 1981; McCollet al., 1982; Bonaıti-Pellie et al., 1984; Pierach et al.,1987; Puy et al., 1997]. Moreover, identification of

asymptomatic heterozygotes for variant AIP is not pos-sible as they have normal erythrocyte enzyme activi-ties [Mustajoki, 1981; Mustajoki and Desnick, 1985;Mustajoki and Tenhunen, 1985]. In view of the diffi-culties with biochemical diagnosis, investigators haveturned to molecular techniques to identify specific mu-tations in the HMB-synthase gene for accurate diagno-sis of affected members in AIP families.

Although many HMB-synthase mutations causingAIP have been detected (for reviews, see Astrin andDesnick [1994], Kappas et al. [1995], and Grandchamp[1998]) most have been “private,” occurring in a singleor only a few unrelated AIP families. These findingshave emphasized the molecular heterogeneity of themutations causing AIP. To date, only two common mu-tations, R116W and W198X, have been detected, withR116W found in families from northern Sweden (pre-sumably due to a founder effect) and W198X in severalapparently unrelated Dutch AIP families [Lee and An-vret, 1991; Gu et al., 1993b]. The presence of such alarge number of private mutations requires that theHMB-synthase gene in each new family be sequencedto identify the specific mutation.

In this communication, we report the development ofa long-range polymerase chain reaction (PCR) proce-dure to amplify the entire 10-kb HMB-synthase gene intwo fragments, and the application of this method toefficiently identify the mutations in 26 unrelated AIPpropositi from Argentina. Eight new mutations andfive previously reported lesions were detected, andthe novel missense mutations were expressed. Ofnote, 12 presumably unrelated index cases (46%)had the G111R mutation, and haplotype analysis indi-cated that this frequent mutation was from a commonfounder.

MATERIALS AND METHODS

Patients and Enzyme Assay

The diagnosis of AIP was confirmed in each proposi-tus in Argentina by determining their erythroid HMB-synthase activity [Batlle et al., 1978] and urinary ALA,PBG, and porphyrins [Mauzerall and Granick, 1956;Seubert and Seubert, 1983] (Table I). Peripheral bloodsamples were collected with informed consent from 26unrelated Argentinean AIP patients and their rela-tives. Lymphoid cell lines were established and main-tained as described previously [Anderson and Gusella,1984]. Genomic DNA was extracted either from lym-phoid cell lines or peripheral blood using the PuregeneDNA isolation kit (Gentra Systems, Minneapolis, MN).

Long-Range PCR

The entire HMB-synthase gene (10 kb) was ampli-fied from genomic DNA in two fragments, fragment 1including the promoter region through intron 3 (prim-ers LR1 and LR2; 4.5-kb PCR product) and fragment 2,exon 2 through exon 15 (primers LR3 and LR4; 5.5-kbPCR product; Table II). Using the GeneAmp XL PCRKit (Perkin Elmer, Foster City, CA), an initial dena-turation was performed at 94°C for 1 min, and thenamplification was carried out in a PCR Minicycler

Novel Mutations Causing AIP 367

(M.J. Research, Watertown, MA) as follows for ampli-fication of both fragments 1 and 2 in a single or sepa-rate reaction(s). The first 16 cycles were performedwith denaturation at 94°C for 30 sec and annealing andextension at 67°C for 5 min. The second 12 cycles werecarried out with denaturation at 94°C for 30 sec, an-nealing and extension at 67°C for 5 min and 15 sec,respectively, with 15-sec increments each additionalcycle. The final extension step was performed at 72°Cfor 10 min. A portion of the PCR products was analyzedby agarose gel electrophoresis to determine that thelong-range reactions were successful, and to identifyany gross gene rearrangements. The PCR productswere purified with the QIAquick PCR Purification Kit(QIAGEN Inc., Santa Clarita, CA).

Sequencing Reactions

Each exon and the flanking intronic regions weresequenced using the Amplicycley Sequencing Kit (Per-

kin Elmer) according to the manufacturer’s instruc-tions using the sense and antisense primers listed inTable II and the following modifications. The cycle se-quencing reaction was performed (30 cycles) with a de-naturation step of 30 sec at 95°C, annealing 30 sec at60°C, and extension 60 sec at 72°C. After a denatur-ation step of 3 min at 94°C, a portion of the PCR reac-tion was loaded in a 4% acrylamide solution (19:1) con-taining 7 M urea gel. The gels were dried, placedagainst Kodak X-OMAT film for 24 to 48 hr, and thendeveloped.

Prokaryotic Expression ofHMB-Synthase Mutations

The normal and mutant HMB-synthase alleles wereexpressed in Escherichia coli using the pKK233-2 vec-tor (Pharmacia Biotech Inc., Piscataway, NJ) as de-scribed previously [Chen et al., 1994; Tsai et al., 1988].

TABLE I. Biochemical Findings in the Argentinean AIP Patients*

FamilyPatient orrelation Sex

Age atdiagnosis

Urinary HMB-synthaseactivityALA PBG Porphyrins

1 Patient F 40 6.2 11 563 32.62 Patient F 20 9.8 80.7 678 43.03 Patient F 31 8.2 66.6 500 32.0

Mother F 63 1.6 1.9 197 27.54 Patient F 33 3.9 13.4 551 42.9

Sister F 33 1.3 7.7 170 41.35 Patient F 48 5.6 44.9 3,107 36.76 Patient F 49 1.3 1.8 127 21.17 Patient F 37 20.1 65.6 849 67.38 Patient F 29 1.9 1.7 125 35.5

Sister F 27 4.1 9.1 770 41.29 Patient F 37 2.9 8.6 881 41.3

Mother F 62 1.6 5.0 125 36.410 Patient F 21 17.7 93.3 1,286 39.611 Patient F 35 21.6 74.1 1,195 62.712 Patient F 35 39.0 183 1,525 37.813 Patient F 39 4.6 19.9 624 48.7

Sister F 48 2.7 8.7 437 37.714 Patient F 43 12.8 99.8 875 64.415 Patient F 25 12.3 102 1,538 31.4

Sister F 27 5.5 30.2 1,001 21.216 Patient F 43 6.5 46.1 866 41.917 Patient F 34 18.2 56.7 3,913 57.018 Patient F 22 8.8 34.5 2,176 54.8

Brother M 33 2.2 1.5 100 40.919 Patient F 53 5.7 41.4 339 35.9

Son M 25 1.1 1.0 23 37.0Grandchild F 13 1.1 1.0 52 45.2

20 Patient F 23 36.7 182 6,831 29.2Mother F 48 1.9 3.4 218 36.2Brother M 22 2.3 4.1 180 39.5

21 Patient F 35 6.3 43.1 828 31.5Father M 61 0.9 1.7 130 25.9

22 Patient F 42 4.5 15.8 1,502 48.0Daughter F 26 0.6 1.4 60 33.7Daughter F 23 1.9 0.7 35 37.6

23 Patient F 49 9.8 68.9 465 71.1Daughter F 24 6.8 30.1 691 34.5

24 Patient F 27 7.8 37.1 687 24.425 Patient F 31 33.6 224 4,067 49.526 Patient F 42 6.4 30.8 690 38.7

Nephew M 12 3.1 1.7 123 44.6Nephew M 17 1.7 1.2 65 39.2

*Normal values: Urinary ALA, 2–4 mg/24 hr; urinary PBG, 1–2 mg/24 hr; urinary porphyrins, 20–250 mg/24 hr;HMB-synthase mean ± 1 SD, 81.5 ± 12 U/ml erythrocytes (female) and 73.1 ± 13.6 U/ml erythrocytes (male).

368 De Siervi et al.

To introduce each of the mutations into the normalpKK-HMB-synthase (pKK-HMBS) expression con-struct, site-directed mutagenesis was performed[Warner et al., 1992] with the mutation being intro-duced into the construct by PCR amplification.

For the Q34P mutation, sense and antisenseprimers, MV101 (58-AAGATGAGAGTGATTCGCGTG-GGTACCCGCAAGAGCCAGCTTGCTCGCATACCG-ACGGAC-38 containing the mutation in bold, andunderlined Kpn I site), and MV 672 (58CACAGCATA-CATGCATTCCTCAGGGTGCAG-38 containing an un-derlined Nsi I site) were used to generate the 596-bpamplification product. The reaction was performed in afinal volume of 100 ml containing 10× Tris-HCl buffer,pH 8.3 (10 ml), 1 mM of MgCl2, 200 mM of each dNTP,100 ng of genomic DNA, 25 pmol of each primer, and2.5 U of Taq polymerase (Life Technologies). PCR am-plification was performed with an initial denaturationat 95°C for 3 min and then 30 cycles of denaturation at95°C for 30 sec, annealing at 65°C for 1 min, and ex-tension at 72°C for 20 sec, followed by a final extensionstep at 72°C for 5 min. The amplification product wasdigested with Kpn I and Nsi I, and the Kpn I/Nsi Ifragment, after purification with the Wizard PCRPreps DNA Purification System (Promega, Madison,WI), was ligated as a cassette into the correspondingsites in pKK-HMBS, generating the mutant constructpKK-HMBS-Q34P.

For the G335S mutation, sense and antisense prim-ers MV1030 (58-ATCCAGCATGTTTTTGGCTCCTTT-GCTCAGCAACAAGTTGGCCAGGCTGATGCTCAA-GTTCTGGGCAGCCAA-38 containing the mutation inbold, and an underlined Blp I site) and MV472 (58-CGAAGAGCAGCCCAGCTGCAGAGAAAGTTC-38)

were used to generate the 611-bp amplification prod-uct. The reaction was performed as described aboveand the product was digested with Blp I and BstX I.The Blp I/BstX I fragment was purified as above andligated as a cassette into the corresponding sitesin pKK-HMBS, generating the mutant constructpKK-HMBS-G335S. The entire coding region of eachexpression construct was sequenced to confirm its au-thenticity. After transfection, bacterial growth, andisopropylthiogalactoside (IPTG) induction, the ex-pressed HMB-synthase activity was assayed as previ-ously described [Chen et al., 1994; Tsai et al., 1988].

Detection of the Common G111R Mutation

To detect the G111R mutation, two PCR reactionswere used to amplify specifically the mutant or normalallele. For the normal allele, the sense primer AD76and antisense primer AD7N (Table II) were used inPCR reactions in a final volume of 100 ml containing10× Tris-HCl buffer, pH 8.3 (10 ml), 1 mM of MgCl2, 800mM of each dNTP, 100 ng of genomic DNA, 25 pmol ofeach primer, and 2.5 U of Taq polymerase (Life Tech-nologies, Gaithersburg, MD). For the mutant allele, thesense primer AD76 and a mismatched antisenseprimer AD7M (Table II) with the underlined mutationat the 38 end were used. PCR amplification was per-formed with an initial denaturation step at 95°C for 3min followed by 30 cycles of denaturation at 95°C for 30sec, annealing at 65°C for 30 sec, and extension at 72°Cfor 20 sec, and then a final elongation step at 72°C for5 min. A portion of each PCR product was electropho-resed in a 2% agarose gel and stained with ethidiumbromide. The size of the PCR product was 191 bp.

TABLE II. Primers for PCR Amplification and Sequencing*

Primer Oligonucleotide sequence

For long-range PCRLR1 (sense) 58-TGCTCCCACTTCAGTTACTTGTCTTTA-38LR2 (antisense) 58-GACGCCCATCTCTAAACCTAATCAGGAC-38LR3 (sense) 58-AAGGGACCAGCCTTGGAGTATTTCCCCACTC-38LR4 (antisense) 58-CAAGGATAGAAGGGCGGTTGAGGTGTGC-38

For direct sequencingExon 1 58-GAGACCAGGAGTCAGACTGT-38Exon 2/3 58-CCCACTGACAACTGCCTTGGTCAAG-38Exon 4 58-CCTAACCTGTGACAGTCT-38Exon 5/6 58-AGACCTAGCATACTAGGG-38Exon 7 58-AGGGTCAGGCCCCAAAGGGAAAGG-38Exon 8 58-CGAGAGAATAGAGGTGAT-38Exon 9 58-TTGTCTTTTTCCTTGGCTGC-38Exon 10 58-GGGAAAGACAGACTCAGGCAGAG-38Exon 11 58-CGGTAGCATCCCAAGGTCT-38Exon 12 58-TAAGAAATCTTCCCTGC-38Exon 13 58-CAGTGATGTCCTCAGGTCTG-38Exon 14 58-ATCCCAGGTTTCTAGGTAG-38Exon 15 58-AGACCATGCAGGCTACCATC-38

58-CGTGACCTGTCGTCGTTG-38For rapid detection of mutation G111R

AD76 58-CCCTAGGCTCCACCACTGAAG-38AD7M 58-GCAAGACTCTTACTTGCAGATGGCTCT-38AD7N 58-GCAAGACTCTTACTTGCAGATGGCTCC-38

*LR1 and LR2, primers used for amplification of promoter region through intron 3 (4.5 kb), and LR3 and LR4,exon 2 through exon 15 (5.5 kb). The others primers were used in sequencing or for rapid detection of the G111Rmutation.


Haplotype Analysis

To investigate the genetic ancestry of the commonG111R mutation, intragenic and flanking polymor-phisms were determined in available families. The in-tragenetic polymorphisms evaluated were: 1345g/a and1500t/c in intron 1, 3581a/g in intron 3, 6479G/T inexon 10, and 7064c/a in intron 10 [Yoo et al., 1993] aswell as the four microsatellite markers flanking theHMB-synthase gene in the chromosomal region11q24.1-q24.2 (Center for Medical Genetics, Marsh-field Medical Research Foundation; http://www.marshmed.org/genetics) [Gelb et al., 1995]. The orderof the flanking markers was (centromeric to telomeric)D11S4111 (localized at 108.59 cM)–D11S1998 (113.13cM)–HMB-synthase gene (113.40-115.53 cM)–D11S4132 (116.07 cM)–D11S4089 (119.07 cM). Theprimer sets used to amplify these upstream and down-stream flanking makers were purchased from Re-search Genetics (Huntsville, AL). The allele sizes andfrequencies of each marker were obtained from theCEPH Genotype Database (http://www.cephb.fr/cephdb/). In addition, the intragenic markers 3581 a/gin intron 3 and 6479 G/T in exon 10 [Yoo et al., 1993]were examined in all AIP index-cases.

RESULTS

Identification of AIP Mutations by Long-RangePCR and Cycle Sequencing

To determine the nature of the mutations in theHMB-synthase gene from 26 unrelated biochemicallydiagnosed Argentinean AIP patients, a long-range PCRmethod was developed to amplify the entire 10-kb genein two fragments, one containing exons 1 through 3 andthe other intron 2 through exon 15 (Fig. 1). Each exonwas then cycled sequenced and eight new mutationsand five previously reported lesions were identified.

The eight new mutations included two missense(Q34P and G335S), four small deletions 728DCT,815DAGGA, 948DA, and 985D12), one insertion(666insA), and a splice site mutation (IVS12+1) (Table

III, Fig. 2). To characterize further the two missensemutations, pKK-HMBS expression vectors containingthe Q34P and G335S lesions were constructed, ex-pressed in E. coli, and the enzymatic activity of therespective mutant protein was determined. As shownin Table IV, the Q34P and G335S enzymes each hadless than 3% of the mean enzymatic activity expressedby the normal allele. The IVS12+1 (Fig. 2C) resulted inexon 12 skipping and deletion of the 120-bp exon (datanot shown).

The four new small deletions and the new insertionwere each identified in a single family. The 728DCTdeletion in exon 12 (Fig. 2D) caused a frameshift re-sulting in premature termination at residue 250. Dele-tion 815DAGGA was in exon 13 (Fig. 2E) and resultedin a premature termination at amino acid 278. The948DA deletion (Fig. 2F) occurred in exon 15 andcaused premature truncation of the encoded polypep-tide at residue 344. Finally, the 985D12 (Fig. 2G)caused the deletion of four amino acids (leucine, twoalanines and a glutamine) but did not alter the rest ofthe amino acids. An insertion of a single A (666insA)exon 12 (Fig. 2H) produced a frameshift and stop ter-mination codon at position 250.

In addition, five previously reported mutations weredetected including three missense mutations (G111R,R173W, and R201W) [Chen et al., 1994; Lundin et al.,1994; Mgone et al., 1994; Kauppinen et al., 1995;Schreiber et al., 1995], a nonsense mutation (Q204X)[Mgone et al., 1994], and a single base insertion913insC [Puy et al., 1996]. R173W in exon 10 was pres-ent in two unrelated AIP patients, while R201W andQ204X, both in exon 10, were each found in a singlefamily. The 913insC in exon 15 was also identified onlyin one family. Of particular interest was the findingthat the G111R mutation, originally identified in twounrelated European patients [Gu et al., 1993a; Puy etal., 1997], was found in 12 presumably unrelated Ar-gentinean AIP patients. Using an allele-specific assay,carriers in AIP families with the G111R mutation werereadily detected (Fig. 3).

Analysis of HMB-Synthase Polymorphisms inthe AIP Patients

To determine if the AIP patients with the G111Rmutation were related, four intragenic [Yoo et al.,1993] and two centromeric and two telomeric flankingpolymorphic markers were analyzed. As shown inTable V, all seven of the G111R propositi studied had acommon haplotype for all intragenic and flanking poly-morphic sites. It should be noted that two of the intra-genic polymorphisms, 3581a/g in intron 3 and 6479G/Tin exon 10 [Yoo et al., 1993], were analyzed in all of theArgentinean AIP propositi and affected relatives. Thefrequency of the 3581a polymorphism in the 64 ana-lyzed alleles was 0.56, similar to that (0.59) in theNorth American Caucasian population [Yoo et al.,1993], while the frequency of the 6479G polymorphismwas 0.81 (74 Argentinean alleles analyzed), which wasslightly higher than that (0.65) reported for EuropeanCaucasians [Gu et al., 1991] or than that (0.69) for

Fig. 1. Agarose gel showing long-range PCR product of the HMB-synthase gene. The entire 10-kb gene was amplified in two fragments usingthe primers and PCR described in Materials and Methods, one fragment of4.5 kb containing exons 1 through 3 (lanes 1 and 2) and the other fragmentof 5.5 kb containing intron 2 through exon 15 (lane 3). Marker is LambdaDNA-HindIII Digest (New England BioLabs, Beverly, MA).


North American Caucasians [Yoo et al., 1993]. Thus, itwas not surprising that the common forms of theseintragenic markers were frequent among the Argentin-ean AIP patients.

In contrast, haplotype analysis of the flanking mark-ers D11S4111, D11S1998, D11S4132, and D11S4089showed that all seven tested G111R patients sharedone common allele for each marker (Table V), indicat-ing their common ancestry. Different haplotypes forthe flanking markers were observed in four Argentin-ean AIP patients with other HMB-synthase mutations.

DISCUSSION

In AIP, the acute neurologic attacks are precipitatedby various ecogenic factors including certain drugs andhormones, alcohol, starvation or dieting, etc. [Kappaset al., 1995; McGovern et al., 1996]. These ecogenicfactors, directly or indirectly, cause the depletion of thehepatic free heme pool and induce the overexpressionof 5-aminolevulinate synthase, the first enzyme in theheme biosynthetic pathway. As a result, large quanti-ties of ALA and porphobilinogen are synthesized. Inpatients with AIP, the half-normal activity of HMB-synthase results in the accumulation of these porphy-rin precursors which leads to the biochemical patho-physiology of the life-threatening acute attacks. Thus,efforts have been directed to identify and counselasymptomatic gene carriers in affected families toavoid potential precipitating factors. With the isolationand characterization of the human HMB-synthasecDNA and genomic sequences [Raich et al., 1986;Grandchamp et al., 1987; Yoo et al., 1993], it becamepossible to identify accurately gene carriers of thisdominant porphyria. Of note, most of the HMB-synthase mutations identified to date in this 10-kbgene have been private, emphasizing the molecularheterogeneity of this disease [Astrin and Desnick,1994; Kappas et al., 1995; Grandchamp, 1998]. Thus, tofacilitate the molecular analysis of AIP propositi, a va-

riety of mutation detection techniques has been em-ployed to identify mutations in reverse-transcribedcDNAs or PCR-amplified genomic DNAs, including de-naturing gradient gel electrophoresis (e.g., Bourgeoiset al. [1992] and Gu et al. [1992]), single-strand confor-mational polymorphism analysis (SSCP; e.g., Kaup-pinen [1992]), heteroduplex analysis [Schreiber et al.,1995], and direct sequencing (e.g., Chen et al. [1994]).

In our experience, the amplification and sequencingof all the exons and adjacent intron/exon boundarieshas proved most effective for mutation detection, be-cause the entire coding region was sequenced, so thereis no question concerning the nature of the causativemutation. In addition, the expression of all missensemutations is necessary to eliminate rare polymor-phisms that encoded functional enzymes. To optimizethe rapidity of this procedure and to minimize its cost,a long-range PCR method was developed to amplify thepromoter region in a single fragment including thehousekeeping and erythroid-specific promoters, and toamplify the entire coding region (exons 3 through 15) ina second fragment. These genomic fragments can beamplified individually or in a multiplex reaction, andused directly for manual or automated sequencing.Moreover, electrophoresis of these PCR products woulddetect gross (>50 bp) insertions and deletions.

Using the long-range PCR techniques and cycle se-quencing, the HMB-synthase mutations in each of the26 unrelated Argentinean AIP probands were detectedincluding eight new mutations (two missense, foursmall deletions, a single base insertion, and a splicesite mutation) and five previously reported lesions(G111R, R173W, Q204X, R201W, and 913insC) [Chenet al., 1994; Gu et al., 1993a; Kauppinen et al., 1995;Lundin et al., 1994; Mgone et al., 1994; Puy et al., 1996,1997; Schreiber et al., 1995]. All of these mutations,except for G111R, (see below), were identified in onlyone or two unrelated AIP families. However, mutationG111R, previously identified in two unrelated Euro-pean AIP patients [Gu et al., 1993a; Puy et al., 1997],

TABLE III. HMB-Synthase Mutations in the Argentinean AIP Patients

AIPfamilya Exon/intron Mutation Nucleotide change (nt) Amino acid change Previous report

1, 2 E4 Q34P 100 A→C 34Q→P3–14 E7 G111R 331 G→A 111G→R Gu et al., 199315, 16 E10 R173W 517 C→T 173R→W Mgone et al., 199417 R201W 601 C→T 201R→W Chen et al., 199418 Q204X 610 C→T 204Q→X Mgone et al., 199419 E12 665insA Insert A after 665 222V→GSASPGQG

HLGCGGCAARSRDSASLHRX

20 728 D CT Delete CT at 728, 729 244L→ASLHRX21 I12 IVS12+1 g→c at 771 + 1 58 splice site mutation;

deletion of exon 1222 E13 815 D4 Delete AGGA at 815–818 272K→MGECTX23 E15 913insC Insert C after 913 305H→PX Puy et al., 199624 948 DA Delete A at 948 317I→ASLLVTFHE

GPSWLPRTWRSAWPTCCX

25 985D12 Delete TTGGCTGCCCAGat 985–996

Inframe deletion of329LAAQ

26 G335S 1003 G→A 335G→S

aBased on Table I.


was identified in 12 unrelated Argentinean patients.Although the 12 families with the G111R mutationcould not identify a common ancestor, the possibilitythat they were related to a common ancestral “founder”was evaluated by haplotype analysis of polymorphicmarkers within and flanking the gene (see below).

To date, only two other HMB-synthase mutationshave been identified as common in specific ethnicgroups. W198X was reported in AIP patients fromnorthern Sweden [Lee and Anvret, 1991] and R116W[Gu et al., 1993b] was found in 15 of 49 Dutch AIPfamilies. The high frequency of W198X presumablywas due to a founder effect, whereas the high frequencyof R116W in the Dutch population is unclear, becausethere was no common ancestor among the Dutch AIPpatients for at least four generations [Gu et al., 1993b].However, haplotype analysis to determine if these pa-tients were distantly related was not performed.Within this context, it was surprising that almost halfof the 26 unrelated Argentinean AIP patients had theG111R mutation, the highest frequency of a mutationamong unrelated AIP patients identified to date. To

facilitate the future detection of this common lesion, aPCR assay was developed using primers specific foreither the normal or mutant G111R allele (Fig. 3).

Efforts were directed to determine if the AIP patientswith mutation G111R have a common ancestral origin.Detailed pedigrees were unavailable since relatives ofmany of the patients were deceased or the patientsthemselves had limited knowledge of their families’ an-cestries. Six families with the G111R mutation werereportedly of Spanish and/or Italian origin while othersclaimed French, Arab, or Irish origins. Since theG111R mutation occurred at a CpG dinucleotide, it wasalso possible that the mutation originated severaltimes independently. However, analysis of four intra-genic and four flanking DNA polymorphic markers in-dicated that all tested patients (7 of the 12) with theG112R mutation had at least one common allele for allintragenic and flanking markers (Table V). Argentin-ean AIP patients with different mutations had allelesfor the flanking markers different than the propositiwith the G111R mutation. These results strongly sug-gest but do not prove conclusively that the individuals

Fig. 2. Partial sequencing gels showing: (A) The A-to-C transversion of nt 101 in exon 4 predicting a glutamine-to-proline substitution at residue 34(Q34P); (B) the G-to-A transition of nt 1003 in exon 15 causing a glycine-to-serine substitution at amino acid 335 (G335S); (C) g→c substitution at position771 + 1 producing exon 12 skipping; small deletions and an insertion; (D) 728delCT; (E) 815delAGGA; (F) 948delA; (G) 985delTTGGCTG; and (H)666insA. All deletions and insertion alter the reading frame and predicted a stop codon at positions 747, 833, 1028, 1028, and 747, respectively.


with the G111R mutation were related because theirmutations are on a common genetic background. Thus,these data suggest that all of the G111R patients had,at some time in the past, a common ancestor possibly ofSpanish or perhaps Italian ancestry, the two major Eu-ropean countries from which people immigrated to Ar-gentina. There are no reports of mutation analysis ofAIP patients from either Spain or Italy so it is un-known whether the G111R mutation has been found inAIP patients from these countries, although it is rec-ognized that a single founder who was an early settlerin Argentina could have dispersed the mutation overseveral generations within the Argentinean popula-tion. Use of the rapid method reported in this paper fordetection of the G111R mutation in newly discoveredcases of AIP in Argentina, Italy, and Spain may lead toadditional information about the origin of this muta-tion.

Of the newly identified mutations, seven occurred inthe coding region and one in the IVS12 splice site con-sensus sequence, all causing the classical form of AIP.Only two of the new mutations were missense (Q34Pand G335S) and neither occurred at a CpG nucleotide,the mutational hot spot [Barker et al., 1984]. Of note,three mutations (666insA, 728D1CT, and IVS12+1) oc-curred in or adjacent to exon 12 and three (948DA,985D12, and G335S) in exon 15, both exons where pre-viously reported mutations were concentrated (e.g.,Mgone et al. [1993], Astrin and Desnick [1994], andGrandchamp [1998]). In addition, three previously re-

ported mutations (R173W, R201W, and Q204X) oc-curred in exon 10, another hot spot exon for AIP mu-tations [Gu et al., 1992]. Of note, the new deletions(728DCT, 815D4, 948DA, and 985D12) increased thepercentage of gene rearrangements of the HMB-synthase gene to 27%, indicating that these lesions areimportant molecular defects causing AIP.

Comparison of the amino acid sequence of HMB-synthase from different species ranging from bacteriato humans has identified 58 invariant residues (18% ofthe sequence) [Brownlie et al., 1994]. These conservedamino acids are clustered in the hydrophobic core of themolecule, at the active site and in other conformation-ally important locations, indicating their structuraland/or functional importance. The E. coli HMB-synthase has been crystallized to 1.9 Å resolution[Louie et al., 1992] and the crystal structure wassolved, demonstrating three domains, each comprisingb-sheets, a-helical secondary structure, and a discretehydrophobic core. A dipyrromethane cofactor was cova-lently bound to the sulfur atom of invariant cysteine-242 in domain 3 and was in an extensive active sitecavity formed between structurally related domains Iand II. Since the E. coli and human HMB-synthaseamino acid sequences have 35% homology and morethan 70% similarity, it is possible to infer the structure/function relationships for certain human HMB-synthase mutations [Brownlie et al., 1994]. For ex-ample, the Q34P missense mutation involved a con-served glutamine at the active site cleft, whichnormally is involved in hydrogen bonding with serine96 and arginine 195 [Brownlie et al., 1994]. This bulkybasic amino acid was replaced with a rigid neutral hy-drophobic proline that introduces a turn that maycause a significant secondary structure alteration bydisruption of the b-strand, consistent with its very lowactivity when expressed in E. coli (Table IV). TheG335S missense mutation in exon 15 involved the sub-stitution of an uncharged glycine, present in both E.coli and humans, by an uncharged serine residue. Thislesion was located in the a23 region of domain 3 of theenzyme and the effects of the mutation are unclear.The glycine at position 111 that is mutated to arginineis located on the protein’s surface but the effect of themutation is unknown [Brownlie et al., 1994].

The deletions (with the exception of 985D12), inser-tion, and IVS12 splice site mutations resulted in frame-shift mutations leading to the incorporation of differentresidues and premature truncation causing the loss ofenzymatic activity. Previously, a different IVS12 mu-tation IVS12+1g→a was described that deleted the en-tire exon 12 and produced an enzyme protein that was

Fig. 3. Rapid detection of G111R mutation by ASO PCR. Lanes 1 and2: amplified portion of exon 7 or normal individuals (N/N); lanes 3 to 6:amplified portion of exon 7 of the HMB-synthase gene of individuals het-erozygous for the G111R mutation (M/N). Primers used to amplify lanes 1,3, and 5 are AD76 and AD7N in Table II (N); primers used to amplify lanes2, 4, and 6 are AD6 and AD7M (M). Marker is fX174-Hae III Digest(New England BioLabs). See Materials and Methods for details of PCRreactions.

TABLE IV. Expression of HMB-Synthase Mutations in E. coli

Construct

HMB-synthase activity (U/mg/hr) Percentage of meannormal expressed

activityMeana Rangea

JM109 0.20 0.08–0.41 0pKK-HMBS 140 132–152 100pKK-HMBS-Q34P 1.78 0.71–3.24 1.3pKK-HMBS-G335S 3.47 0.72–5.21 2.5

aMean and range based on results of six independent experiments.


enzymatically inactive [Daimon et al., 1993; Grand-champ et al., 1989c]. It is assumed that the IVS12+1g→c

would have the same effect as the IVS12+1g→a lesion.In summary, eight new mutations and five previ-

ously identified mutations, including the common mu-tation G111R causing AIP, were identified in Argentin-ean AIP patients. Identification of these mutationshighlights the molecular heterogeneity underlyingAIP, permits the precise diagnosis of asymptomaticheterozygotes in these Argentinian families, and pro-vides information for future structure-function studiesof the human enzyme.

ACKNOWLEDGMENTS

We thank Dr. H. Muramatsu, Mrs. B. Riccilo deAprea, and Lic. L. Dato for their valuable help with thepatients, Mr. Raman Reddy for his expert technicalassistance, and Dr. George Diaz for his help with thehaplotyping. This work was supported in part bygrants from the National Institutes of Health, includ-ing a research grant (5 R01 DK26824), a grant (5 M01RR00071) for the Mount Sinai General Clinical Re-search Center, and a grant (5 P30 HD28822) for theMount Sinai Child Health Research Center, and bygrants from the Argentinean National Research Coun-cil (00509108/97), the Science and Technology Agency(PMT-PICT002b/97), and the University of BuenosAires (EX032/95-98). Alcira Marıa del Carmen Batlle,Marıa Victoria Rossetti and Victoria Estela Parera areSuperior, Independent, and Associate Researchers, re-spectively, in the Career of Scientific Researcher Pro-gram at the Argentinean National Research Council.Adriana De Siervi is a fellow from the same Council.This work represents part of the doctoral thesis sub-mitted by A. De Siervi to the University of BuenosAires.

REFERENCES

Anderson MA, Gusella JF. 1984. Use of cyclosporin A in establishing Ep-stein-Barr virus-transformed human lymphoblastoid cell lines. In Vitro(Rockville) 20:856–858.

Astrin KH, Desnick RJ. 1994. Molecular basis of acute intermittent por-

phyria: Mutations and polymorphisms in the human hydroxymethyl-bilane synthase gene. Hum Mutat 4:243–252.

Barker DF, Schafer M, White R. 1984. Restriction sites containing CpGshow a higher frequency of polymorphism in human DNA. Cell 36:131–138.

Batlle AM, Wider de Xifra EA, Stella AM. 1978. A simple method formeasuring erythrocyte porphobilinogenase, and its use in the diagnosisof acute intermittent porphyria. Int J Biochem 9:871–875.

Bonaıti-Pellie C, Phung L, Nordmann Y. 1984. Recurrence risk estimationof acute intermittent porphyria based on analysis of porphobilinogendeaminase activity: A Bayesian approach. Am J Med Genet 19:755–762.

Bottomley SS, Bonkowsky HL, Kreimer-Birnbaum M. 1981. The diagnosisof acute intermittent porphyria. Usefulness and limitations of theerythrocyte uroporphyrinogen I synthase assay. Am J Clin Pathol 76:133–139.

Bourgeois F, Gu XF, Deybach JC, Te Velde MP, de Rooij F, Nordmann Y,Grandchamp B. 1992. Denaturing gradient gel electrophoresis forrapid detection of latent carriers of a subtype of acute intermittentporphyria with normal erythrocyte porphobilinogen deaminase activ-ity. Clin Chem 38:93–95.

Brownlie PD, Lambert R, Louie GV, Jordan PM, Blundell TL, Warren MJ,Cooper JB, Wood SP. 1994. The three-dimensional structures of mu-tants of porphobilinogen deaminase: Toward an understanding of thestructural basis of acute intermittent porphyria. Protein Sci 3:1644–1650.

Chen CH, Astrin KH, Lee G, Anderson KE, Desnick RJ. 1994. Acute in-termittent porphyria: Identification and expression of exonic mutationsin the hydroxymethylbilane synthase gene. An initiation codon mis-sense mutation in the housekeeping transcript causes “variant acuteintermittent porphyria” with normal expression of the erythroid-specific enzyme. J Clin Invest 94:1927–1937.

Chretien S, Dubart A, Beaupain D, Raich N, Grandchamp B, Rosa J,Goossens M, Romeo PH. 1988. Alternative transcription and splicing ofthe human porphobilinogen deaminase gene result either in tissue-specific or in housekeeping expression. Proc Natl Acad Sci USA 85:6–10.

Daimon M, Yamatani K, Igarashi M, Fukase N, Ogawa A, Tominaga M,Sasaki H. 1993. Acute intermittent porphyria caused by a G to C mu-tation in exon 12 of the porphobilinogen deaminase gene that results inexon skipping. Hum Genet 92:549–553.

Elder GH. 1993. Molecular genetics of disorders of haem biosynthesis. JClin Pathol 46:977–981.

Gelb BD, Edelson JH, Desnick RJ. 1995. Linkage of pycnodysostosis tochromosome 1q21 by homozygosity mapping. Nat Genet 10:235–237.

Grandchamp B. 1998. Acute intermittent porphyria. Semin Liver Dis 18:17–24.

Grandchamp B, De Verneuil H, Beaumont C, Chretien S, Walter O, Nord-mann Y. 1987. Tissue-specific expression of porphobilinogen deami-nase. Two isoenzymes from a single gene. Eur J Biochem 162:105–110.

Grandchamp B, Picat C, Kauppinen R, Mignotte V, Peltonen L, Mustajoki

TABLE V. Intragenic and Flanking Polymorphisms of the HMB-Synthase Gene in the Argentinean AIP Patients

Polymorphismsa

Family Mutation

Intragenic Flankingb

1345g/aIntron 1

3581 a/gIntron 3

6479G/TExon 10

7064c/aIntron 10

D11S4111108.59 cM

D11S1998113.13 cM

D11S4132116.07 cM

D11S4089119.07 cM

3 G111R G/G aa GT C/A 3, 4 2, 3 8, 10 3, 44 G111R G/G ag GG C/A 1, 4 2, 4 8, 10 3, 45 G111R G/G aa GG C/A 5, 4 2, 4 2, 10 3, 76 G111R G/G aa GG C/A 1, 4 2, 3 1, 10 3, 27 G111R G/G aa GT C/A 1, 4 2, 3 4, 10 3, 2

12 G111R G/G aa GG C/A — 2, 4 8, 10 —13 G111R G/G aa GG C/A 1, 4 2, 1 4, 10 3, 418 Q204X — ag GG CA 3, 3 3, 3 4, 8 2, 419 666insA — aa GT CA 3, 5 3, 4 9, 9 3, 321 IVS12+1 — aa GT CC — 4, 4 1, 1 —23 913insC — aa GG CA 2, 6 3, 3 5, 11 7, 7

aCommon alleles in G111R patients are shown in bold.bOrder of flanking markers surrounding HMB-synthase gene is centromere–D11S411 (108.59 cM)–D11S1998 (113.3 cM)–HMBS (113.4–115.5 cM)–D11S4132 (116.07 cM)–D11S4089 (119.07 cM)–telomere.


P, Romeo PH, Goossens M, Nordmann Y. 1989a. Molecular analysis ofacute intermittent porphyria in a Finnish family with normal erythro-cyte porphobilinogen deaminase. Eur J Clin Invest 19:415–418.

Grandchamp B, Picat C, Mignotte V, Wilson JH, Te Velde K, Sandkuyl L,Romeo PH, Goossens M, Nordmann Y. 1989b. Tissue-specific splicingmutation in acute intermittent porphyria. Proc Natl Acad Sci USA86:661–664.

Grandchamp B, Picat C, de Rooij F, Beaumont C, Wilson P, Deybach JC,Nordmann Y. 1989c. A point mutation G→A in exon 12 of the porpho-bilinogen deaminase gene results in exon skipping and is responsiblefor acute intermittent porphyria. Nucleic Acids Res 17:6637–6649.

Gu XR, Lee JS, Delfau MH, Grandchamp B. 1991. PCR detection of G/Tpolymorphism at exon 10 of the porphobilinogen gene. Nucleic AcidsRes 18:1966.

Gu XF, de Rooij F, Voortman G, Te Velde K, Nordmann Y, Grandchamp B.1992. High frequency of mutations in exon 10 of the porphobilinogendeaminase gene in patients with a CRIM-positive subtype of acuteintermittent porphyria. Am J Hum Genet 51:660–665.

Gu XF, de Rooij F, de Baar E, Bruyland M, Lissens W, Nordmann Y,Grandchamp B. 1993a. Two novel mutations of the porphobilinogendeaminase gene in acute intermittent porphyria. Hum Mol Genet 2:1735–1736.

Gu XF, de Rooij F, Lee JS, Te Velde K, Deybach JC, Nordmann Y, Grand-champ B. 1993b. High prevalence of a point mutation in the porphobi-linogen deaminase gene in Dutch patients with acute intermittent por-phyria. Hum Genet 91:128–130.

Kappas A, Sassa S, Galbraith RA, Nordmann Y. 1995. The porphyrias. In:Scriver CS, Beaudet AL, Sly WS, editors. Metabolic and molecularbases of inherited disease, 7th ed. New York: McGraw-Hill. p 2103–2160.

Kauppinen R. 1992. Single-strand conformation polymorphism (SSCP)analysis applied to the diagnosis of acute intermittent porphyria. MolCell Probes 6:527–530.

Kauppinen R, Mustajoki S, Pihlaja H, Peltonen L, Mustajoki P. 1995.Acute intermittent porphyria in Finland: 19 mutations in the porpho-bilinogen deaminase gene. Hum Mol Genet 4:215–222.

Lamon JM, Frykholm BC, Bennett M, Tschudy DP. 1978. Prevention ofacute porphyric attacks by intravenous haematin. Lancet 2:492–494.

Lee JS, Anvret M. 1991. Identification of the most common mutationwithin the porphobilinogen deaminase gene in Swedish patients withacute intermittent porphyria. Proc Natl Acad Sci USA 88:10912–10905.

Louie GV, Brownlie PD, Lambert R, Cooper JB, Blundell TL, Wood SP,Warren MJ, Woodcock SC, Jordan PM. 1992. Structure of porphobi-linogen deaminase reveals a flexible multidomain polymerase with asingle catalytic site. Nature 359:33–39.

Lundin G, Wedell A, Thunell S, Anvret M. 1994. Two new mutations in theporphobilinogen deaminase gene and a screening method using PCRamplification of specific alleles. Hum Genet 93:59–62.

Mauzerall D, Granick S. 1956. The occurrence and determination of d-ami-nolaevulinic acid and porphobilinogen in urine. J Biol Chem 219:435–446.

McColl KE, Moore MR, Thompson GG, Goldberg A. 1982. Screening forlatent acute intermittent porphyria: The value of measuring both leu-cocyte delta-aminolaevulinic acid synthase and erythrocyte uroporphy-rinogen-1-synthase activities. J Med Genet 19:271–276.

McGovern MM, Anderson KE, Astrin KH, Desnick RJ. 1996. Inherited

porphyrias. In: Rimoin DL, Connor JM, Pyeritz RE, editors. Emery andRimoin’s principles and practice of medical genetics. New York: Chur-chill Livingston. p 2009–2037.

Mgone CS, Lanyon WG, Moore MR, Louie GV, Connor JM. 1993. Detectionof a high mutation frequency in exon 12 of the porphobilinogen deam-inase gene in patients with acute intermittent porphyria. Hum Genet92:619–622.

Mgone CS, Lanyon WG, Moore MR, Louie GV, Connor JM. 1994. Identifi-cation of five novel mutations in the porphobilinogen deaminase gene.Hum Mol Genet 3:809–811.

Mignotte V, Eleouet JF, Raich N, Romeo PH. 1989. Cis- and trans-actingelements involved in the regulation of the erythroid promoter of thehuman porphobilinogen deaminase gene. Proc Natl Acad Sci USA 86:6548–6552.

Mustajoki P. 1981. Normal erythrocyte uroporphyrinogen I synthase in akindred with acute intermittent porphyria. Ann Intern Med 95:162–166.

Mustajoki P, Desnick RJ. 1985. Genetic heterogeneity in acute intermit-tent porphyria: Characterization and frequency of porphobilinogen de-aminase mutations in Finland. Br Med J 291:505–509.

Mustajoki P, Tenhunen R. 1985. Variant of acute intermittent porphyriawith normal erythrocyte uroporphyrinogen-I-synthase activity. Eur JClin Invest 15:281–284.

Namba H, Narahara K, Tsuji K, Yokoyama Y, Seino Y. 1991. Assignmentof human porphobilinogen deaminase to 11q24.1→q24.2 by in situ hy-bridization and gene dosage studies. Cytogenet Cell Genet 57:105–108.

Pierach CA, Weimer MK, Cardinal RA, Bossenmaier IC, Bloomer JR. 1987.Red blood cell porphobilinogen deaminase in the evaluation of acuteintermittent porphyria. JAMA 257:60–61.

Puy H, Deybach JC, Lamoril J, Robreau AM, Nordmann Y. 1996. Detectionof four novel mutations in the porphobilinogen deaminase gene inFrench Caucasian patients with acute intermittent porphyria. HumHered 46:177–180.

Puy H, Deybach JC, Lamoril J, Robreau AM, Da Silva V, Gouya L, Grand-champ B, Nordmann Y. 1997. Molecular epidemiology and diagnosis ofPBG deaminase gene defects in acute intermittent porphyria. Am JHum Genet 60:1373–1383.

Raich N, Romeo PH, Dubart A, Beaupain D, Cohen-Solal M, Goossens M.1986. Molecular cloning and complete primary sequence of humanerythrocyte porphobilinogen deaminase. Nucleic Acids Res 14:5955–5968.

Schreiber WE, Fong F, Nassar BA, Jamani A. 1995. Heteroduplex analysisdetects frameshift and point mutations in patients with acute inter-mittent porphyria. Hum Genet 96:161–166.

Seubert A, Seubert S. 1983. Curso sobre porfirinas y porfirias. Etiopato-genia, clinica y tratamiento (Parte practica). Buenos Aires: CIPYP,FCEyN (UBA) and CONICET. p 1–10.

Tsai SF, Bishop DF, Desnick RJ. 1988. Human uroporphyrinogen III syn-thase: Molecular cloning, nucleotide sequence, and expression of a fulllength cDNA. Proc Natl Acad Sci USA 85:7049–7053.

Warner CA, Yoo HW, Roberts AG, Desnick RJ. 1992. Congenital erythro-poietic porphyria: Identification and expression of exonic mutations inthe uroporphyrinogen III synthase gene. J Clin Invest 89:693–700.

Yoo HW, Warner CA, Chen CH, Desnick RJ. 1993. Hydroxymethylbilanesynthase: Complete genomic sequence and amplifiable polymorphismsin the human gene. Genomics 15:21–29.


identification and characterization of hydroxymethylbilane synthase mutations causing acute...

Documents