mucgpofPm

Structure of the PCCA Gene and Distribution of MutationsCausing Propionic Acidemia

Eric Campeau,*,1 Lourdes R. Desviat,† Daniel Leclerc,* Xuchu Wu,2 Belen Perez,†Magdalena Ugarte,† and Roy A. Gravel*,2,3

*Departments of Biology, Human Genetics, and Pediatrics, McGill University Health Centre, Montreal, Canada H3H 1P3;

Molecular Genetics and Metabolism 74, 238–247 (2001)doi:10.1006/mgme.2001.3210, available online at http://www.idealibrary.com on

and †Centro de Biologia Molecular “Severo Ochoa,” CSIC-UAM, Madrid, Spain

uly 11

p

g

o

Received May 18, 2001, and in revised form J

Propionyl-CoA carboxylase (PCC, EC 6.4.1.3) is amitochondrial, biotin-dependent enzyme that func-tions in the catabolism of branched-chain aminoacids, fatty acids with odd-numbered chain lengths,and other metabolites. It catalyzes the ATP-depen-dent carboxylation of propionyl-CoA to D-methyl-

alonyl-CoA. PCC is composed of two types of sub-nits, likely as a4b4 or a6b6, with the a subunitontaining the covalently bound biotin prostheticroup. A genetic deficiency of PCC activity causesropionic acidemia, a potentially fatal disease withnset in severe cases in the newborn period. Af-ected patients may have mutations of either theCCA or PCCB gene. In this study, we have deter-ined the structure of the human PCCA gene

which, at the present time, is only partially repre-sented in the databases. Based on reported ESTsand confirmed by RT-PCR, we also redefine thetranslation initiation codon to a position 75 nucle-otides upstream of the currently accepted initiationcodon. We show the distribution of mutations, in-cluding three identified in this study, and renumberall reported mutations to count from the new initi-ation codon. The gene spans more than 360 kb andconsists of 24 exons ranging from 37 to 335 bp inlength. The introns range in size from 104zbp to 66kb. We have also determined the nucleotide se-quence of ;1 kb of the 5*-flanking region upstreamof the ATG translation initiation site. The proximal

1 Present address: Lawrence Berkeley National Laboratory,MS 74-157, 1 Cyclotron Road, Berkeley, CA 94720.

2 Present address: Department of Biochemistry and MolecularBiology, Room 250 Heritage Medical Research Building, Univer-

sity of Calgary, 3330 Hospital Drive NW, Calgary, Alberta, Can-ada T2P 1C1. E-mail: rgravel@ucalgary.ca.

3 To whom reprint requests should be addressed.

2381096-7192/01 $35.00Copyright © 2001 by Academic PressAll rights of reproduction in any form reserved.

, 2001; published online September 17, 2001

400 bp of the 5*-flanking region shows a high G 1 Ccontent (67%) and is part of a putative 1-kb CpGisland that extends into exon 1 and part of intron 1.The putative promoter lacks a TATA box but con-tains two AP-1 sites and a conservatively definedconsensus GC box, the latter characteristic of thecore binding sequence of the Sp1 transcription fac-tor. © 2001 Academic Press

Key Words: propionic acidemia; propionyl-CoAcarboxylase; PCCA gene; gene structure; mutations;

romoter sequence; carboxylase; biotin.

Propionyl-CoA carboxylase (PCC,4 EC 6.4.1.3) is amitochondrial, biotin-dependent enzyme that func-tions in the degradation of isoleucine, valine, methi-onine, threonine, fatty acids with odd-numberedchain lengths, and cholesterol (1). PCC has beenpurified to homogeneity (2,3) and functional enzymehas been expressed in Escherichia coli (4,5). It iscomposed of a and b subunits, likely in an a4b4 ora6b6 structure, with the a subunit containing thecovalently bound biotin prosthetic group. cDNAshave been cloned for the a and b subunits, and thecorresponding PCCA and PCCB genes have beenmapped to chromosomes 13q32 and 3q13.3–q22, re-spectively (6–12). The structure of the human PCCBene has been reported and it contains 15 exons (13).PCC catalyzes the ATP-dependent carboxylation

f propionyl-CoA to D-methylmalonyl-CoA. This is atwo-step reaction that generates carboxybiotin as atransient intermediate (reviewed in 14). In the firstreaction, bicarbonate is transferred to biotin uponhydrolysis of ATP to form carboxybiotin. In the sec-

4 Abbreviation used: PCC, propionyl-CoA carboxylase.

fbst

tapwmpopgyde

F THE

ond, the carboxyl group of carboxybiotin is trans-ferred onto propionyl-CoA. The a subunit is respon-sible for the first half reaction. It contains sequencesrelated to ATP and bicarbonate binding sites withina putative biotin carboxylase domain (15). It alsocontains the biotin attachment site at a lysine resi-due in the sequence Ala-Met-Lys-Met, that is highlyconserved among biotin-dependent carboxylases(16). The biotin site is contained within a 67 aminoacid peptide that forms the carboxy-terminus of theprotein that can be independently expressed andbiotinylated in vitro (17). The b subunit, responsibleor the second half reaction, contains putative car-oxybiotin and propionyl-CoA binding sites. Bothubunits contain mitochondrial leader sequences athe amino-terminus (18,19).

An inherited deficiency of PCC activity results inhe inborn error propionic acidemia (1). Severelyffected patients usually present in the newborneriod with life-threatening metabolic ketoacidosisith vomiting, lethargy, and hypotonia. The diseaseay be fatal in infancy or early childhood. Some

atients are less severely affected and some do welln a protein-restricted diet (20,21). Two classes ofatients are defined, depending on which of the twoenes is affected. This can be assessed by the anal-sis of mRNA or polypeptide expression or can beetermined in cell fusion-based complementationxperiments (22,23).To date some 27 PCCA and over 30 PCCB muta-

tions have been identified. In order to provide agenomic context for characterizing mutations and toprovide information on the promoter, we have deter-mined the structure of the human PCCA gene, ob-tained and confirmed through a combination of in-verse-PCR and sequencing-based approaches, andpresent the physical map of the PCCA gene in thecontext of the public databases. We also redefine thetranslation initiation codon based on reported ESTs,the additional sequence confirmed by RT-PCR, andshow the distribution of mutations, renumbered tofit the newly defined initiation codon, responsible fora subunit defects of PCC.

MATERIALS AND METHODS

Fibroblast cell lines from patients and controlswere from the McGill University Repository for Mu-

STRUCTURE O

tant Human Cell Strains or from the author’s (M.U.)laboratory in Madrid. In some studies, leukocyteDNA was examined. Mutations were identified at

the genomic level and confirmed in two independentPCR and sequencing reactions. Mendelian inheri-tance was confirmed when possible in parental sam-ples.

Several approaches were used to obtain genomicsequences. Some introns were amplified by PCR us-ing oligonucleotides located in adjacent exons asprimers (Table 1). Another approach was to use in-verse PCR to obtain flanking sequences around in-dividual exons. Briefly, genomic DNA was digestedwith a “four-base cutter” restriction enzyme (BfaI,RsaI, MseI, TaqI, NlaIII, HaeIII, or MboI) and reli-gated in order to form small circular DNA molecules.Intronic sequences were selectively amplified by us-ing oligonucleotides facing opposite directions in agiven exon. For both PCR and inverse PCR, ampli-fied products were subcloned into the pCR2.0 orpCR2.1 vector (TA cloning kit, Invitrogen, CA) andsequenced using primers located in the vector (T7,Sp6, M13 reverse primers) using the Sequenase kit(United State Biochemicals). The intronic sequencewas read as far as possible or until the recognitionsequence for the restriction enzyme used to generatethe circular DNA was encountered. PCCA exons11–22 and their surrounding intronic sequenceswere obtained using these procedures. BAC (bacte-rial artificial chromosome) clones 541B17, 364F8(Accession Number AL353697), 382J13, 407D5,411G12, and 509O5 were identified from theRPCI-11 human genomic library by hybridizationscreening with a PCCA cDNA probe, performed atthe Centre for Applied Genomics at the Hospital forSick Children, Toronto (these clones are documentedat http://genome.wustl.edu/cgi-bin/ace/GSCMAPS.cgi). An additional BAC clone, 3071A15 (AccessionNumber AQ165744), was identified through BLASTgenome searches of the GSS database. BAC DNAwas extracted using Magnum Kb-100 columns (Ge-nome Systems, Inc.) and dialyzed using drop-dialy-sis. PCRs were used to map each BAC using oligo-nucleotide primers specified from the a subunitcDNA sequence. Purified BAC DNA was sequencedusing oligonucleotides located in exons by the Cen-tre for Applied Genomics facility. PCCA exons 5–7,9–11, 16, 18, 21, and 23 and their introns weredetermined or confirmed using these strategies.

In an independent approach, BLASTN searches ofpreliminary human genomic DNA sequences in

239PCCA GENE

the NCBI databases (http://www.ncbi.nlm.nih.gov/)were performed using PCCA cDNA sequences orrelevant partial genomic sequences as queries. This

CTCCAGAGCCTGT

identified the remaining flanking intronic sequencesand promoter. For all approaches used to obtainintronic sequences, amplification of the splice junc-tions and direct sequencing of the PCR product wereused to confirm the sequence. In these cases, PCR

TAIntronic Primers Used in the Amplifi

Exon Primer

1 APA1S 59-TGGTC1–30AS 59-GCCCA

2 31–108S 59-CGTTC31–108AS 59-TGAAC

3–4 109–225S 59-GGTCT109–225AS 59-CTGCT

5 AU558 59-CGACT226–339AS 59-ACTCA

6 AU559 59-CAGTG340–393AS 59-TCTATA

7 AU556 59-ATACAAT630 59-ATACA

8 526–562S 59-GACAT526–562AS 59-CTCAA

9 AK432S 59-CCTGGAT629AS 59-GATGG

10 642–744S 59-TGTTAGAU700AS 59-ATTCTA

11 AT632S 59-CCTTTG745–839AS 59-GAATG

12 1604S 59-TAAATT1673AS 59-ATCTTT

13 1656S 59-CATTCG1605AS 59-GGATT

14 1606S 59-CTGTG1661AS 59-TCACTT

15 1600S 59-TGAGA1601AS 59-CCTGC

16 1602S 59-TCGAG1609AS 59-GCTGT

17 1355–1465S 59-GTGAT50AS 59-ATTCTG

18 AU554S 59-ATCCTAAU555AS 59-AGATC

19 1607S 59-GTTCG1608AS 59-GTACT

20 1662S 59-CTGGA1663AS 59-CGACT

21 1674S 59-CCACT1675–2AS 59-TTAAG

22 1592S 59-AATTG1603AS 59-CTGCA

23 1965–2047S 59-TGGCT1965–2047AS 59-CGTTC

24 2048–2109S 59-ATTAGAPA14AS 59-AGAGC

240 CAMPE

products were purified and sequenced using the fmolsequencing kit (Promega) and analyzed on an auto-mated ALF Express System (Amersham Pharmacia

Biotech). Intron sizes were calculated from the DNAsequence data. Extension of the cDNA sequence inthe 59 direction was made by the identification ofESTs in the EST division of GenBank. Transcribedsequence was confirmed by RT-PCR of human fibro-

1n of the 24 Exons of the PCCA Gene

SequenceLength of PCR

product (bp)

CCGGACGGCGT-39 139CGGCACGGCTA-39AACCTAACATCC-39 238

TGTGGAGACGAG-39CATCGGTTTAAAAG-39 334CAAAATTAAGTCC-39AATGATAGGCAGAA-39 363CATGAACTAACTGC-39GAGAAGTGTACTTA-39 335ATCACTACCCACAT-39GGGCTTTTCTTTTA-39 360GTATGTGCGTGCAT-39AATGCCTCATG-39 151TCTGCTGAAAG-39TTTTCCTGACTC-39 340GTAATTACTCCATT-39ACTCTTCTTCTCC-39 301CACAACTCACAATC-39CATGAATCAAAATA-39 317

AATGAATGAAATGCA-39TAGATGATCTATATC-39 385AACACTTTATGGAG-39ACTTCATTTTACCC-39 375ATCCAATAGTTGGC-39GTAAATACCATATG-39 311TTCAATTTACACCC-39ATTCTCTATCCATT-39 321TTTCACATTTAC-39TTTAGAAATCTGTTAT-39 282

TATTCTCATTGG-39TCATAGAACAGTGA-39 364CCACAAGACTGCA-39CATAATAGATGCCC-39 334AACTGAGATCAGCC-39ACATTCAATGGCAG-39 326TTCAAAAGACGAGA-39

TAAACATGACAAGTT-39 362AAGTATAAGGTGAA-39GCTATACTATATTT-39 218AAAGAATGATTTC-39GAATGAATGCTACT-39 318

GCCGAATCGTGCTGT-39CCTAAGTGTTAATG-39 199TCACAGGGCC-39

CTGCCGCCTC-39 470CGCCTATGTGT-39

T AL.

BLEcatio

GCTGGAGGGCTAGACAGTAAACTGACCCTATATGTGATATAAATCC

CATATTGTATTGTGCCCGTATGTTATTACTTCTGAAGGTTAC

AATGGGTAATTTATTA

ACGTAATTTGGTCC

TTGAATCTGCATATAGCCAATAGG

CACTGCTG

TCTAGCAAATACAAAGCAGGTAAAGTATGAGAATATTTAAAGATCATT

AU E

blast RNA. The RNA was extracted using Trizolreagent (Gibco) and reversed-transcribed by randompriming using the Omniscript RT kit (Qiagen). The

T

tXa

p

hese wsequenones ob

line.

resulting cDNA was subjected to PCR using theindicated primers.

Promoter region sequences were obtained fromthe NCBI databases after BLASTN searches usingexon 1 as probe and confirmed after PCR amplifica-tion and sequencing of three overlapping segmentsusing specifically designed primers (see Figure 3).The promoter region was sequenced using the Ther-mosequenase cycle sequencing kit with 7-deaza-dGTP (Amersham Pharmacia Biotech). CpG islandswere located as defined by Gardiner-Garden andFrommer (24). Promoter elements were predictedusing the TRANSFAC database (http://transfac.gbf.de/TRANSFAC (25)) and the program MOTIF atthe Genomenet server (http://www.genome.ad.jp).

RESULTS

Structural Organization of the Human PCCA Gene

FIG. 1. Structural organization of the human PCCA gene. Tositions according to Map Viewer (http://www.ncbi.nlm.nih.gov

represented as vertical lines. Introns are drawn to scale except inlengths, Table 2). Also, the length of introns 12 and 19 is unknownin figure), with marker positions designated by vertical lines. Tsequences. Not shown is SHGC-3131 about 40 kb 59 of exon 1. Thethe gene structure (“Clones”) is a physical map of the genomic cla PCCA cDNA probe. Unknown ends are represented as a dotted

STRUCTURE O

Figure 1 shows the structure of the PCCA gene.he gene is positioned relative to the surrounding

human gene map. The figure also shows the physicalmap of the clones identified in this study. At least 60bp of every intron at intron-exon junctions wereconfirmed by PCR amplification and sequencing ofDNA from control individuals (Accession NumbersAY035786 to AY035808).

Inspection of the EST database revealed addi-tional transcribed sequence upstream of the most 59reported cDNA (18). Five ESTs were identified (Ac-cession Numbers AU076648, AU126846, AU132092,BG428051, and AU099081). All five indicate thepresence of a more upstream in-frame ATG at posi-tion 275 relative to the currently accepted transla-ion initiation codon, as shown in Accession NumberM012283 (18). This provides for an additional 25mino acid residues at the N-terminus of the pro-a

sequence (Accession Number AF385926 for the ad-ditional 59 sequence and deduced amino acid se-quence). While the additional 38 nucleotides (rede-

schematic (“Genes”) shows the distribution of genes and mapsecond schematic (“Exons”) shows the PCCA gene with exons

4, 6, and 21 for which “//” is used to denote large introns (introndicated by “?”. Immediately below the gene is the STS map (“STS”

ere identified in Map Viewer and positioned relative to clonece of WI-7253 overlaps with that of exon 24. The schematic belowtained through BLAST genomic searches or by hybridization to

241PCCA GENE

he top/). Thetrons, as in

F THE

fined as starting at position 238) upstream of thenew initiation codon remain in frame, there are noother ATGs farther upstream in the genomic se-

eab1t

55TC

AU E

quence prior to the next in-frame stop codon (at2183). The additional transcribed sequence wasconfirmed in human fibroblast RNA by RT-PCR us-ing two distinct sense and anti-sense primers, thelatter from exons 1 and 4 to eliminate the possibilityof genomic DNA contamination of the PCR reaction(Fig. 2).

The human PCCA gene spans more than 360 kband consists of 24 exons ranging from 37 to 335 bp inlength (Table 2). All exon-intron splice junctionsfollow the gt/ag rule (26). Other consensus elementsassociated with splice junctions are also found witha high frequency: 17 exons show 59-splice sites thatnd with a G and 11 of them with an AG. Introns 16nd 17 are type 1 introns (interrupting a codon

FIG. 2. RT-PCR of human fibroblast RNA for detection oftranscribed sequence at the 59 end of PCCA gene. (Top) Schematicshowing positions of primers used in PCR. (Bottom) Polyacryl-amide gel electrophoresis of PCR products from lane 1, oligonu-cleotide primers A and P1 (142 bp in length); lane 2, B and P1(133 bp); lane 3, A and P2 (295 bp); lane 4, B and P2 (286 bp).Primer A starts at position, 237 which is one less than the 59limit of the most 59 EST detected in the EST database. A,-TGAGAGGTCAGCAGAGG (from 237 to 221 in exon 1); B,-AGCAGAGGGGCGGTCTGC (228 to 211 of exon 1); P1, 5-CT-CAGGGTCCGCAGCG (from 1105 to 189 of exon 1); P2, 5-GC-CATCTTCTTGCAAG (from 1258 to 1242 of exon 4).

242 CAMPE

etween the first and second nucleotides), introns 9,1, and 18 are type 2 (interrupting a codon betweenhe second and third nucleotides) and the remainder

are type 0 introns (splicing occurring betweencodons). The introns range in size from 104 bp to 66kb, with two (introns 12 and 19) not determined(Fig. 1).

Promoter Region

The nucleotide sequence of ;1000 bp of the 59-flanking region upstream of the previously reportedATG translation initiation site was identified byBLAST searches of the HTGS database (see Acces-sion Numbers AC011573 and AL355338) and con-firmed by PCR amplification and sequencing (Fig.3).This putative promoter region has a high G 1 Ccontent in the first 400 bp (67%) upstream of thetranscribed sequence, lacks a TATA box, but con-tains candidate Sp1 boxes at position 2114 relativeto the proposed translation initiation site and an-other within the 59-untranslated region at 222. TwoAP-1 sites, as well as sequences similar to bindingsites for other transcription factors (Dof, USF,GATA-1), are also found in this region. Other poten-tial binding sites for regulatory elements includeone cAMP-response element (CREB) and one stress-response element (STRE) (Fig. 3).

Novel Mutations and Cryptic Splicing in the PCCAGene

As a result of the availability of genomic sequenceinformation on the PCCA gene, we characterizedmutant alleles corresponding to mutations not ex-pressed at the mRNA level or for whom only DNAsamples had been referred. Three novel mutationswere identified, including two small deletions caus-ing frameshifts and a splice-defect mutation (Fig. 4).Mutation 440delC was identified in one patientwhose second allele is I164T. The frameshift muta-tion was missed by RT-PCR, since it was not ex-pressed in the mRNA (I164T appeared as sole de-tectable mutation). The patient had a latepresentation (mild form) of the disease (27). We sur-mise that I164T is responsible for the mild diseasein this patient. In the second case, 2103delT provedto be homozygous on testing genomic DNA. It intro-duces a stop codon seven residues downstream of themutation, immediately before the region involved inbiotin attachment. This patient had a severe, neo-natal presentation and died shortly after birth, con-

T AL.

sistent with null-type expression due to the homozy-gous frameshift mutation. Finally, a splicemutation, IVS14 1 1G . A, was identified in a

lsS

GAATCT.

patient cell line in which RT-PCR analysis revealedskipping of exons 13 and 14. The second allele hasnot been determined in this patient whose pheno-type was not available.

We and others previously reported the frequentoccurrence of a transcript containing an insertion of84 nucleotides after nt 1284 of the coding sequencein patients with mRNA destabilizing mutations(28,29). The longer transcript proved to be a low-level cryptic splice product in normal individualsthat becomes prominent in RT-PCR products in sit-uations where the level of normal length mRNA isdrastically reduced. We have located the 84 bp se-quence from RP11-466B4 (Accession NumberAL355680) and RP11-364F8 (Accession NumberAL353697) sequences. It occurs in intron 14 and islocated 2753 bp from the donor splice site of exon 14and 1358 bp from exon 15 (Fig.4). Because of itsready detection in patients with vanishing levels ofmRNA due to RNA destabilizing mutations such assplice mutations, frameshifts, small insertions ordeletions, or stop codon mutations, the 84-bp dele-

TAExon-Intron Structure

ExonSize(bp) nt No. 39 splice

1 .143 238–105 CCGCTG2 78 106–183 ttttttgtag CAT GTT3 48 184–231 ctctcttcag ACT TTT4 69 232–300 tttaaaatag GTT ATT5 114 301–414 gtctcctcag GTT CAT6 54 415–468 catttctaag GTA CAT7 132 469–600 ttctccacag GCA GCA8 37 601–637 gcttttacag GAT GCA9 79 638–716 ctgctgttag GC TAC

10 103 717–819 tttttctcag G GAT G11 95 820–914 gctctttcag GTT CTA12 151 915–1065 gtatatgtag C ATT TT13 144 1066–1209 ggccttgcag GTT GAG14 75 1210–1284 gaactttcag GAC CCC15 69 1285–1353 tttctcccag GTC CGA16 76 1354–1429 ttgtttttag CTA ATC17 111 1430–1540 ttacttttag GT GTT18 103 1541–1643 tttccctcag GA CAC19 103 1644–1746 tctacttcag A ATG C20 99 1747–1845 gtcacaacag GTG GAA21 54 1846–1899 tgttttttag TGT CTT22 141 1900–2040 ttcccttaag TAC AAG23 78 2041–2118 tgttttccag GTA GCA24 335 2119–2187 aaaattcaag GTG AAA

STRUCTURE O

tion may prove to be of diagnostic advantage as apreliminary test of such compound mutations. Itmay also serve to determine the identity, PCCA or

PCCB, of the disease gene in the absence of comple-mentation studies.

DISCUSSION

We have characterized the transcription unit ofthe PCCA gene. The gene is in excess of 360 kbong and contains 24 exons. It is found on chromo-ome 13q32 in the region defined by STS markersHGC-3131 to WI-7253, with the former STS ;40

kb upstream of exon 1 and the latter comprisingpart of exon 24 (Fig. 1). During the course of ourstudy, the sequences composing the gene haveappeared in the draft sequence of the human ge-nome. Our studies have confirmed or clarified se-quences and ordered contig segments across thelength of the gene. The gene is flanked by FGF14(FGF homologous factor 4; FHF-4) about 1 Mbpaway on the 59 side (30) and by inferred locus,LOC65698 (similar to ribosomal protein S26),about 10 kb away on the 39 side. The gene islocated within the assembled contig, NT_009932,

2e Human PCCA Gene

Exon 59 splice IntronSize(bp)

. . . . ACC CTG AAG gtgaggagca 1 13658. . . . AAT GAA AAA gtaagtattt 2 8881. . . . GCA TGT CGG gtgagtagaa 3 104. . . . GCT AGT TCT gtaagtatat 4 42917. . . . GCC CAA GCT gtgagtctga 5 2195. . . . AGA TGT TTG gtaagttggt 6 51992

. . . . . GTA GTC AAG gtgagaagct 7 26369

. . . . . AGG GAA ATT gtaagtcctt 8 21717. . . GAG ACC AG gtgagaggct 9 5075. . GAA ATC CAG gttggtacat 10 5855. . . . GCA CCA AG gtaagtctcc 11 4413. AGA CTC CAG gtaacaacaa 12 ?. . . . TAT GCT GAG gtaaaatgaa 13 1321. . . . CTA CCT GGT gtaagtcatt 14 4194

. . . . . ATT TCA AAA gttagtttaa 15 2573. . . . ATT CGA G gtaaaaacaa 16 20653. . . TTC AAA G gtttgtatgc 17 9486. . . AAT TCA AG gtatggtaga 18 28213. . GTG TTC TCG gtgagttttc 19 ?. . . . . ACT GTC CAG gtgagtgttg 20 23521. . . . GGT ACA GTG gtaagtatga 21 66182. . . . GGA GAC GCG gtaagggctg 22 12108

. . . . . ACT GGC ACG gtgagtccct 23 2341. . . . CATTTGTCT

243PCCA GENE

BLEof th

GTC .CTG.GAT.AGA.GTG.CCA.GAAGAA

CCT. .GT. . .GGT.T. . . .CAT.TAC.GTGACA.

ACA. .ATG. .CT. . .

GTTTCT.GTG.

F THE

of the working draft of the International Consor-tium, although only a partial gene structure canbe deduced from this contig.

iaiis tron 1s g it ac

At .360 kb, the PCCA gene is significantly largerthan average. Based on analysis of the human ge-nome, the average gene has been estimated to spanabout 28 kb (31). One possibility is that gene size isrelated to chromosomal location and gene density.Interestingly, chromosome 13 is predicted to be theleast gene-rich chromosome, containing ;5 genesper megabase and has the lowest GC content afterthe Y chromosome (31). This might imply a greaterdispersion of coding sequences than on other chro-

FIG. 3. Nucleotide sequence of ;1 kb of the 59-flanking regiodentified below the sequence and are underlined. Primers (PR3Snd sequencing of the region in three overlapping segments are shn bold and the ATG initiation codon is indicated in uppercase lettn single letter code below the nucleotide sequence starting from thequence from the EST database (at position 238) and the exon-intarts with the A in the initiation codon and nucleotides precedinodons are indicated in uppercase letters.

244 CAMPE

mosomes. However, the human XPG gene, whichmaps to the same region of chromosome 13 at q32.3–q33.1, encodes a protein nearly twice the size of the

PCCA polypeptide but spans only 30 kb (32,33). Themajor difference between the two genes lies in thesize of introns, at least three of which are greaterthan 40 kb in PCCA compared to the largest in XPGat less than 6 kb. Another possibility relates to nat-ural selection. The PCCB gene, which is on chromo-some 3, is also large at over 135 kb with five gapsremaining of indeterminant length ((13) and http://www.ncbi.nlm.nih.gov/AceView/hs.cgi). It has twolarge introns of 35 and 70 kb. Perhaps the multiple

n 1 and the first part of intron 1. Potential control elements areS, PR1S, PR3AS, PR2AS, and PR1AS) used in the amplifications arrows pointing in the orientation of their use. Exon 1 is showne amino acid sequence deduced from the exon sequence is shown

t 59 ATG. The 59 limit of exon 1 is defined by the longest identifiedjunction is identified by two vertical lines. Nucleotide numbering

re indicated by negative numbers. The potential ATG inititation

T AL.

n, exo, PR2own aers. The mos

AU E

functional domains of each subunit were pieced to-gether by exon shuffling, leading to exons imbeddedin larger spans of DNA (34,35). Clearly, the rapidly

amop ir ethnh sertionc

increasing collection of confirmed genes in the hu-man genome will shed much light on the evolution offunctional genetic units.

FIG. 4. Schematic representation of the PCCA gene showincidemia. The left side of the panel, “Changes to Nomenclature” gutations listed as small deletions, followed by the renumbered nu

f Mutations” using the revised nomenclature and aligned with coroposed by Campeau et al. (15). References to mutations and theere: 440delC (27), 2103delT, and IVS14 1 1G . A. The “84-bp inells with very low levels of mRNA (28).

STRUCTURE O

Sequence analysis of the 59-flanking region iden-tified a region compatible with that of a promoter.The proximal 400 bp of the 59-flanking region shows

a high G 1 C content (67%) and is part of a putative1-kb CpG island (24) that extends into exon 1 andpart of intron 1. This region lacks a TATA box but

location of the mutations identified in patients with propionice previous amino acid designation, or nucleotide designations forde and amino acid. The right side of the panel shows the “Locationnding exons or introns. Also shown are the domain designations

ic origins are given in Ugarte et al. (37) except for those identified” is a cryptic exon from intron 15 observed by RT-PCR of patient

245PCCA GENE

g theives thcleotirrespo

F THE

contains candidate Sp1 binding sites, one just be-yond and one within the 59 end of the transcribedsequence. While the cap site has yet to be deter-

dbi

Asbt

iorPtmtmedchsctm(lcmbtm2cfsrf

SRC

AU E

mined, the data suggest that the transcription of thePCCA is mediated by an Sp1-dependent mechanism,as characteristic of housekeeping genes.

We previously reported that the mature N-termi-nus of the PCC a polypeptide occurs at Gly52 (resi-

ue 27 in the old numbering sequence), determinedy Edman sequencing of the N-terminus of HPLC-solated a subunit from human liver PCC (18). Using

the program TRANSPEP of the PCGENE software,we note that either ATG codon (Met1 or Met26)generates a leader peptide with predicted cleavageat the identified site for transport into mitochondria.This suggests that heterogeneity of the N-terminuscannot be excluded formally at this time, sincetranslation from either ATG codon might reslt in thesame mature a subunit. However, we do note thatthe more upstream ATG (Met1) is in excellent con-text for translation inititiation with “A” and “G” atthe critical 23 and 14 positions, while the second

TG (Met26) shows no consensus nucleotides in theurrounding positions (36). Ultimately, it will alsoe necessary to determine the transcription initia-ion site(s) to settle this question.

A significant interest in the PCCA gene is its usen the analysis of mutations in patients with propi-nic acidemia. Ugarte et al. (37) recently summa-ized the identified mutations in the PCCA andCCB genes. In this report, we add three mutations

hat produced chain-terminating or aberrantRNA. Through determination of the structure of

he PCCA gene, it will be possible to evaluate suchutations, in particular splice mutations, by PCR of

xons and flanking sequences. Figure 4 shows theistribution of mutations in the PCCA gene in theontext of the exon-intron structure. The mutationsave been renumbered to count from the more up-tream ATG identified in this study. In addition, thehanges in numbering from the previous nomencla-ure are included as part of the figure (the publishedutations can be tracked through the Ugarte et al.

36)). Eleven of 16 point mutations or in-frame de-etion are found in a large region of the a subunitorresponding to the putative biotin carboxylase do-ain, while two other mutations are found in the

iotin binding domain (15). Several of these muta-ions cluster in exon 13. Significantly, one of the twoutations identified in this study, the frameshift at

103delT (at Ala701), is only 27 residues from thearboxy-terminus. While mRNA was not availableor analysis of the impact of this mutation, previous

246 CAMPE

tudies have shown that truncation of the last 27esidues blocked biotinylation of a carboxyl-terminalragment of the a polypeptide of PCC (17).

ACKNOWLEDGMENTS

The authors thank R. Navarrete for excellent technical assis-tance. A special thanks to Steve Scherer, Barbara Kellam, andJack Huizenga at the Centre for Applied Genomics at the Hospi-tal for Sick Children, Toronto, for clone reagents and technicalassistance. These studies were supported by grants from theCanadian Institutes of Health Research to the CIHR Group inMedical Genetics, the Canadian Genetic Diseases Network, theFundacion Ramon Areces, and the Fondo de Investigaciones

anitarias. E.C. is recipient of a scholarship from the Fonds de laecherche en Sante du Quebec and Fonds pour la Formation dehercheurs et l’Aide a la Recherche Sante.

REFERENCES

1. Fenton WA, Gravel RA, Rosenblatt DS. Disorders of propi-onate and methylmalonate metabolism. In The Metabolic &Molecular Bases of Inherited Disease, 8 ed. (Scriver CR,Beaudet AL, Sly WS, Valle D, Eds.). New York: McGraw-Hill, pp 2165–2193, 2001.

2. Kalousek F, Darigo MD, Rosenberg LE. Isolation and char-acterization of propionyl-CoA carboxylase from normal hu-man liver. Evidence for a protomeric tetramer of nonidenti-cal subunits. J Biol Chem 255:60–65, 1980.

3. Gravel RA, Lam KF, Mahuran D, Kronis A. Purification ofhuman liver propionyl-CoA carboxylase by carbon tetrachlo-ride extraction and monomeric avidin affinity chromatogra-phy. Arch Biochem Biophys 201:669–673, 1980.

4. Kelson TL, Ohura T, Kraus JP. Chaperonin-mediated as-sembly of wild-type and mutant subunits of human propio-nyl-CoA carboxylase expressed in Escherichia coli. Hum MolGenet 5:331–337, 1996.

5. Chloupkova M, Ravn K, Schwartz M, Kraus JP. Changes inthe carboxyl terminus of the beta subunit of human propio-nyl-CoA carboxylase affect the oligomer assembly and catal-ysis: Expression and characterization of seven patient-de-rived mutant forms of PCC in Escherichia coli. Mol GenetMetab 71:623–632, 2000.

6. Lamhonwah AM, Barankiewicz TJ, Willard HF, MahuranDJ, Quan F, Gravel RA. Isolation of cDNA clones coding forthe alpha and beta chains of human propionyl-CoA carbox-ylase: Chromosomal assignments and DNA polymorphismsassociated with PCCA and PCCB genes. Proc Natl Acad SciUSA 83:4864–4868, 1986.

7. Kraus JP, Williamson CL, Firgaira FA, Yang-Feng TL,Munke M, Francke U, Rosenberg LE. Cloning and screeningwith nanogram amounts of immunopurified mRNAs: cDNAcloning and chromosomal mapping of cystathionine beta-synthase and the beta subunit of propionyl-CoA carboxylase.Proc Natl Acad Sci USA 83:2047–2051, 1986.

8. Kraus JP, Firgaira F, Novotny J, Kalousek F, Williams KR,Williamson C, Ohura T, Rosenberg LE. Coding sequence ofthe precursor of the beta subunit of rat propionyl-CoA car-boxylase. Proc Natl Acad Sci USA 83:8049–8053, 1986.

9. Stankovics J, Ledley FD. Cloning of functional alpha propio-nyl CoA carboxylase and correction of enzyme deficiency in

T AL.

pccA fibroblasts. Am J Hum Genet 52:144–151, 1993.

10. Kennerknecht I, Suormala T, Barbi G, Baumgartner ER.The gene coding for the alpha-chain of human propionyl-

F THE

CoA carboxylase maps to chromosome band 13q32. HumGenet 86:238–240, 1990.

11. Ohura T, Ogasawara M, Ikeda H, Narisawa K, Tada K. Themolecular defect in propionic acidemia: Exon skippingcaused by an 8-bp deletion from an intron in the PCCBallele. Hum Genet 92:397–402, 1993.

12. Lamhonwah AM, Leclerc D, Loyer M, Clarizio R, Gravel RA.Correction of the metabolic defect in propionic acidemiafibroblasts by microinjection of a full-length cDNA or RNAtranscript encoding the propionyl-CoA carboxylase beta sub-unit. Genomics 19:500–505, 1994.

13. Rodriguez-Pombo P, Hoenicka J, Muro S, Perez B, Perez-Cerda C, Richard E, Desviat LR, Ugarte M. Human propio-nyl-CoA carboxylase beta subunit gene: Exon-intron defini-tion and mutation spectrum in Spanish and Latin Americanpropionic acidemia patients. Am J Hum Genet 63:360–369,1998.

14. Knowles JR. The mechanism of biotin-dependent enzymes.Annu Rev Biochem 58:195–221, 1989.

15. Campeau E, Dupuis L, Leon-Del-Rio A, Gravel R. Codingsequence mutations in the alpha subunit of propionyl-CoAcarboxylase in patients with propionic acidemia. Mol GenetMetab 67:11–22, 1999.

16. Chapman-Smith A, Cronan JE. Molecular biology of biotinattachment to proteins. J Nutr 129:477S-484S, 1999.

17. Leon-Del-Rio A, Gravel RA. Sequence requirements for thebiotinylation of carboxyl-terminal fragments of human pro-pionyl-CoA carboxylase alpha subunit expressed in Esche-richia coli. J Biol Chem 269:22964–22968, 1994.

18. Lamhonwah AM, Mahuran D, Gravel RA. Human mitochon-drial propionyl-CoA carboxylase: Localization of the N-ter-minus of the pro- and mature alpha chains in the deducedprimary sequence of a full-length cDNA. Nucleic Acids Res17:4396, 1989.

19. Kraus JP, Kalousek F, Rosenberg LE. Biosynthesis andmitochondrial processing of the beta subunit of propionylcoenzyme A carboxylase from rat liver. J Biol Chem 258:7245–7248, 1983.

20. Wolf B, Hsia YE, Sweetman L, Gravel R, Harris DJ, NyhanWL. Propionic acidemia: A clinical update. J Pediatr 99:835–846, 1981.

21. Thomas E. A study of the response to protein-modified dietsfor propionic acidemia in twelve patients. Brain Dev 16(Suppl):58–63, 1994.

22. Gravel RA, Lam KF, Scully KJ, Hsia Y. Genetic complemen-tation of propionyl-CoA carboxylase deficiency in culturedhuman fibroblasts. Am J Hum Genet 29:378–388, 1977.

23. Wolf B, Willard HF, Rosenberg LE. Kinetic analysis of ge-

STRUCTURE O

netic complementation in heterokaryons of propionyl CoAcarboxylase-deficient human fibroblasts. Am J Hum Genet32:16–25, 1980.

24. Gardiner-Garden M, Frommer M. CpG islands in vertebrategenomes. J Mol Biol 196:261–282, 1987.

25. Wingender E, Chen X, Hehl R, Karas H, Liebich I, Matys V,Meinhardt T, Pruss M, Reuter I, Schacherer F. TRANSFAC:An integrated system for gene expression regulation. Nu-cleic Acids Res 28:316–319, 2000.

26. Stephens RM, Schneider TD. Features of spliceosome evo-lution and function inferred from an analysis of the infor-mation at human splice sites. J Mol Biol 228:1124–1136,1992.

27. Perez-Cerda C, Merinero B, Rodriguez-Pombo P, Perez B,Desviat LR, Muro S, Richard E, Garcia MJ, Gangoiti J, RuizSP, Sanz P, Briones P, Ribes A, Martinez-Pardo M, Camp-istol J, Perez M, Lama R, Murga ML, Lema-Garrett T,Verdu A, Ugarte M. Potential relationship between geno-type and clinical outcome in propionic acidaemia patients.Eur J Hum Genet 8:187–194, 2000.

28. Campeau E, Dupuis L, Leclerc D, Gravel RA. Detection of anormally rare transcript in propionic acidemia patients withmRNA destabilizing mutations in the PCCA gene. Hum MolGenet 8:107–113, 1999.

29. Ohura T, Narisawa K, Tada K, Iinuma K. An 84 bp insertionfound in a propionic acidaemia patient is not a disease-causing mutation but a product of cryptic mRNA. J InheritMetab Dis 22:676–677, 1999.

30. Smallwood PM, Munoz-Sanjuan I, Tong P, Macke JP, Hen-dry SH, Gilbert DJ, Copeland NG, Jenkins NA, Nathans J.Fibroblast growth factor (FGF) homologous factors: Newmembers of the FGF family implicated in nervous systemdevelopment. Proc Natl Acad Sci USA 93:9850–9857, 1996.

31. Venter JC, Adams MD, Myers EW, LiPW, Mural RJ, et al.The sequence of the human genome. Science 291:1304–1351, 2001.

32. Takahashi E, Shiomi N, Shiomi T. Precise localization of theexcision repair gene, ERCC5, to human chromosome13q32.3-q33.1 by direct R-banding fluorescence in situ hy-bridization. Jpn J Cancer Res 83:1117–1119, 1992.

33. Emmert S, Schneider TD, Khan SG, Kraemer KH. Thehuman XPG gene: Gene architecture, alternative splicingand single nucleotide polymorphisms. Nucleic Acids Res 29:1443–1452, 2001.

34. Patthy L. Genome evolution and the evolution of exon-shuf-fling—A review. Gene 238:103–114, 1999.

35. de Souza SJ, Long M, Gilbert W. Introns and gene evolution.Genes Cells 1:493–505, 1996.

36. Kozak M. Initiation of translation in prokaryotes and eu-karyotes. Gene 234:187–208, 1999.

37. Ugarte M, Perez-Cerda C, Rodriguez-Pombo P, Desviat LR,

247PCCA GENE

Perez B, Richard E, Muro S, Campeau E, Ohura T, GravelRA. Overview of mutations in the PCCA and PCCB genescausing propionic acidemia. Hum Mutat 14:275–282, 1999.