Molecular Phenotypes in Cultured Maple Syrup Urine Disease Cells

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 264, No. 6, Issue of February 25, pp. 3448-3453, 1989 Printed in U. S. A.

Molecular Phenotypes in Cultured Maple Syrup Urine Disease Cells COMPLETE Ela cDNA SEQUENCE AND mRNA AND SUBUNIT CONTENTS OF THE HUMAN BRANCHED CHAIN a-KETO ACID DEHYDROGENASE COMPLEX*

(Received for publication, June 17, 1988)

Charles W. Fisher, Jacinta L. Chuang, Thomas A. Griffin$, Kim S. Lau, Rody P. Cox, and David T. Chuangt From the Genetics Division, Department of Medicine, Veterans Administration Medical Center, Cleveland, Ohio 44106 and the Department of Biochemistry, the University of Texas Southwestern Medical Center, Dallas, Texas 75235

The activity of the branched-chain a-keto acid dehydrogenase complex is deficient in patients with the inherited maple syrup urine disease (MSUD). To elucidate the molecular basis of this metabolic disorder, we have isolated three overlapping cDNA clones encoding the E10 subunit of the human enzyme complex. The composite human Ela cDNA consists of 1783 base pairs encoding the entire human Ela subunit of 400 amino acids with calculated M, = 45,552. The human Ela and the previously isolated human Ez cDNAs were used as probes in Northern blot analysis with cultured fibroblasts and lymphoblasts from seven unrelated MSUD patients. The results along with those of West- ern blotting have revealed five distinct molecular phenotypes according to mRNA and protein-subunit contents. These consist of type I, where the levels of Ela mRNA and Ela and El@ subunits are normal in cells, but E1 activity is deficient; Type 11, where the Ela mRNA is present in normal quantity, whereas the contents of Ela and El@ subunits are reduced; Type 111, where the level of Ela mRNA is markedly reduced with a concomitant loss of Ela and El@ subunits; Type IV, where the contents of both E2 mRNA and E2 subunits are markedly reduced; and Type V, where the E2 mRNA is normally expressed, but the E2 subunit is markedly reduced or completely absent. Type V in- cludes thiamin-responsive (WG-34) and certain classical MSUD cells. These molecular phenotypes have demonstrated the complexity of MSUD and identified the affected gene in different patients for further characterization.

The mammalian branched-chain a-keto acid dehydrogen-

* This work was supported by Grants DK 26758 and DK 37373 from the National Institutes of Health, Grant 1-796 from the March of Dimes Birth Defects Foundation, and by the Veterans Administra- tion. A preliminary account of this work was presented at the 72nd Meeting of the Federation of American Societies for Experimental Biology, May 1-5, 1988, Las Vegas, NV. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in thispaper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) 504474.

$ Medical Scientist Trainee supported by National Institutes of Health Grant GM 07250 awarded to Case Western Reserve Univer- sity, Cleveland, OH 44106.

§ Current address: Dept. of Biochemistry, The University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75235. To whom correspondence should be sent.

ase complex catalyzes oxidative decarboxylation of the a-keto acids derived from leucine, isoleucine, and valine through transamination (1). The enzyme complex from mammalian tissues consists of three catalytic components: a branched chain a-keto acid decarboxylase or El,’ a dihydrolipoyl transacylase or E2, and a dihydrolipoyl dehydrogenase or E3 (2-5). The enzyme complex also contains two regulatory enzymes: a specific kinase (6-8) and a specific phosphatase (9, 10) that modulate the activity of the complex through a phosphorylation-dephosphorylation cycle.

The mitochondrial branched-chain a-keto acid dehydrogenase complex is structurally and mechanistically analogous to pyruvate and a-ketoglutarate dehydrogenase complexes (1). The organization of a-keto acid dehydrogenase complexes is common in that each complex contains a structural core consisting of E2 subunits, to which molecules of El, E3, the kinase, and presumably the phosphatase are attached through noncovalent interactions (2,9,11). The size of the E2 subunit from bovine liver is 46,518 daltons as calculated from the amino acid composition deduced from a cloned cDNA (12). However, the E2 subunit migrates anomalously on SDS-polyacrylamide gel as a 52,000-dalton species (2-5). The El component of the bovine kidney branched-chain a-keto acid dehydrogenase complex consists of a and @ subunits with M, = 46,500 and 36,500, respectively (2). The bovine E3 component is a dimer consisting of identical subunits ( M , = 55,000) and is common among a-keto acid dehydrogenase complexes (1). The Ea-binding domain of the transsuccinylase core of Esch- erichia coli a-ketoglutarate dehydrogenase complex has re- cently been identified (13).

The activity of branched-chain a-keto acid dehydrogenase complex is deficient in patients with maple syrup urine disease (MSUD) (14). A metabolic block at this step results in dra- matic increases in plasma and urinary concentrations of branched chain amino and a-keto acids in MSUD patients (15, 16). The accumulated a-keto acids have devastating clinical consequences including mental retardation, ketoaci- dosis, and possible death. To date five different clinical phenotypes have been recognized for MSUD, i.e. classical, inter- mittent, intermediate, thiamin-responsive, and E3 deficiency (15). The classification is based on the rapidity of onset, severity of the disease, tolerance for dietary protein, response to thiamin supplements, and results of enzyme assay. The etiology of MSUD is heterogeneous as the branched-chain a- keto acid dehydrogenase complex is encoded by at least six structural genes. A mutation at any of these genetic loci could

The abbreviations used are: El, branched chain a-keto acid decarboxylase; EP, dihydrolipoyl transacylase; Ea, dihydrolipoyl dehydrogenase; MSUD, maple syrup urine disease; SDS, sodium dodecyl sulfate; kb, kilobase pair(s); bp, base pair(s).

3448

Molecular Phenotypes in Maple Syrup Urine Disease 3449

result in a dysfunction of the enzyme complex and conse- quently the MSUD phenotype. We (17-20) and others (21, 22) have shown that the El activity of the branched-chain a- keto acid dehydrogenase complex is deficient in several classical MSUD patients that were studied. The Ez polypeptide was shown to be absent in cultured cells derived from other classical MSUD patients (22, 23). A defect in the E3 component has also been reported to result in a combined deficiency in the three a-keto acid dehydrogenase complexes (24-26). Despite these findings, the molecular mechanisms underlying enzyme deficiency in MSUD are mostly unknown.

In the present paper, we report cloning of the cDNA encoding the entire Ela subunit of human branched-chain a-keto acid dehydrogenase complex. The Ela and E? (27) cDNAs along with polyclonal antibodies were used as probes to meas- ure mRNA and subunit contents of the enzyme complex in cultured cells from normal individuals and MSUD patients. This approach has allowed us to detect five distinct molecular phenotypes for MSUD based on mRNA and protein contents. The results have identified the affected gene in these MSUD patients for further characterization at the molecular level.

EXPERIMENTAL PROCEDURES

Cell Lines-Fibroblasts of classical and intermediate MSUD patients were kindly provided by the following physicians (names and institutions in parentheses): P.K. of the Mennonite kindred (Dr. Angelo DiGeorge, St. Christopher's Hospital, Philadelphia); F.J. (Dr. Frederico Mariona, Hutzel Hospital, Detroit); Lo and A.L. (Dr. Sey- mour Packman, University of California, San Francisco); Ech (Dr. Juan Sotos, Children's Hospital, Columbus, OH). Lymphoblasts of a classical patient (GM-1366) were obtained from the National Insti- tute of General Medical Sciences Human Genetic Mutant Cell Re- pository, Camden, N.J. Fibroblasts of a thiamin-responsive MSUD patient (WG-34) (28) were obtained from the Repository for Mutant Human Cell Strains, Montreal, Canada.

Cell Culture-Fibroblasts were grown in monolayer culture in Waymouth's medium containing 10% (v/v) fetal calf serum (GIBCO) and antibiotics (penicillin, 50 units; streptomycin, 50 pg; and kana- mycin, 30 pg/ml) as described previously (29). Both control and patient's cells were grown to confluency and harvested by trypsini- zation 4-7 days after the last medium change. Lymphoblasts were grown at 37 "C in 5% COZ RPMl 1640 medium (GIBCO) supple- mented with 1% (v/v) of 200 mM glutamine solution and 15% (v/v) fetal calf serum.

Enzyme Assays-The total activity of the dephosphorylated branched-chain a-keto acid dehydrogenase complex or dephosphorylated E, was assayed radiochemically using a-chloroisocaproate- treated cells (27, 29). For the El assay, 2,6-dichlorophenolindophenol was used as an electron acceptor (27). E3 activity was assayed spec- trophotometrically as described previously (17, 29). Protein was determined by the method of Lowry et al. (30).

Screening of u t 1 1 Expression Libraries-Bovine ,!?,-antiserum (rabbit) was used as a probe to screen a Xgtll library of human liver cDNA (a generous gift of Dr. Savio Woo, Baylor College of Medicine). The procedure for screening was described previously (27, 31) except that the goat-anti-rabbit-peroxidase conjugate (ICN) instead of lZ5I- protein A was used to probe the bound antibody (32). Phage DNA of positive clones was prepared by the plate lysate method (33). Hybrid- ization of the phage insert with a bovine Ela cDNA (20) was carried out according to the method of Southern (34). To screen a second Xgtll library of human fetal liver cDNA (Clontech), a 0.9-kb human K a cDNA (hEla-1) isolated from Woo's library (this work) or the 5'-end EcoRI fragment (351 bp) of a bovine Ela cDNA (bEla-l) (20) was radiolabeled (27) and used as a probe. The procedure for screening has been previously described (27).

Nucleotide Sequencing of Human Ela cDNAs-The 0.9-kb fragment (hI3,a-l) of human Ela cDNA was excised from 1% agarose gel, extracted with phenol (33), and subcloned into the EcoRI site of Bluescript SK- (Stratagene, San Diego). The subcloned DNA was digested with EcoRI and PstI, electrophoresed, and the restriction fragments were subcloned after separation by electrophoresis into the appropriate sites of Bluescript SK-. Single stranded DNA was rescued by superinfection with R408 helper phage and the nucleotide sequence determined by the dideoxynucleotide chain termination method (35)

using Sequenase (United States Biochemical Corp., Cleveland, OH). To analyze the 1.5-kb human Ela cDNA (hEla-2), digested phage insert DNA was isolated from 1% agarose gel and subcloned into the EcoRI site of Bluescript SK-. The subcloned hE1a-2 was mapped by restriction enzyme analysis and deletions were constructed by the method of Henikoff (36). Deleted inserts were characterized by double-strand sequence analysis (37).

Northern Blot Analysis"Poly(A)+ RNA was prepared from fibroblasts (18 X T-150 flasks) or lymphoblasts (10 X T-25 flasks) according to the method of Chirgwin et al. (38). The steps for denaturation, electrophoresis, and hybridization were as described previously (20, 27). A mixture of nick-translated human Ela (hE1a-1) (0.9 kb) and Ez (hEz-l) (1.6 kb) cDNAs was used as probe. The blots were washed three times with 2 X SSC (1 X SSC = 15 mM sodium citrate, pH 7.0, 150 mM NaCl), 0.1% SDS at 65 "C (15 min each wash), and three times with 0.2 X SSC, 0.1% SDS also at 65 "C (15 min each wash). Autoradiography was carried out at -70 "C 24-48 h with Kodak X- Omat AR film and an intensifying screen.

Production and Affinity Purification of Ela or E,@ Antibody-Ela and EIP subunits were separated by subjecting the highly purified bovine branched chain a-keto acid dehydrogenase complex to SDS- polyacrylamide gel electrophoresis. Individual Ela or El@ subunit was recovered by electroelution as described previously (20). The isolated Ela (120 pg) or E,@ (80 pg) subunit in complete Freund's adjuvant was used as antigen to raise antiserum in rabbits (20). High titers were obtained after the fourth booster injection with incomplete Freund's adjuvant.

To purify the antibody by affinity chromatography, the isolated Ela or ElO subunit was immobilized on a mixture of Affi-Gel 10 and 15 (Bio-Rad) according to the manufacturer's procedure. Rabbit antiserum to the Ela (20) or the El@ subunit was passed through the Affi-Gel affinity column. After washing, the specific IgGs were eluted from the column as described previously (27).

Western Blot Analysis-Fibroblasts or lymphoblasts (14 X lo6 cells) were disrupted by a freeze-thaw procedure. The pellet was suspended in 30 mM potassium phosphate buffer, pH 7.5, containing 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine and 10 p~ leupep- tin to inhibit proteases and homogenized by sonic oscillation. The extract was subjected to Western blotting (300 pg of protein/lane) as describedpreviously (20,32), except that the protein was electrotrans- ferred to a polyvinylidene difluoride (Millipore) instead of nitrocel- lulose membrane. The probe used was a combination of affinity- purified anti-Ela (1:300 dilution) and anti-&@ (1:30 dilution) IgGs or the E2 antiserum (27) (1:500 dilution). IgGs bound to the specific antigens were detected with lZ5I-protein A (20).

RESULTS

Isolation of cDNA Clones for the Human Ela Subunit-To probe the Ela mRNA in normal and MSUD cells, we screened for cDNA clones encoding the human Ela subunit. A human Ela cDNA (hE1a-1) of 0.9 kb in size was initially isolated from a hgtll library (kindly provided by Savio Woo) by the use of rabbit serum against bovine El as a probe. The identity of the hEla-1 clone was established by its hybridization with the bovine Ela cDNA' (20) and nucleotide sequence analysis which showed the published sequence flanking the two phosphorylation sites (39, 40). The 0.9-kb human Ela cDNA was then used as a probe to screen a second Xgtll library prepared from human fetal liver (Clontech). The screening of 5 X lo5 phage produced 82 positive clones. This result indicates 0.02% abundance of the Ela cDNA in the human fetal liver library. The longest human Ela cDNA (hE1a-2) isolated from the Clontech library is 1.5 kb in length. Fig. 1 shows restriction maps and sequence analysis strategies for the two human Ela clones. Nucleotide sequence analysis of the longer hEla-2 clone has revealed that this cDNA is 1550 bp in length with an open reading frame encoding 378 amino acids (Fig. 2). An alignment with the bovine Ela cDNA (20) indicates that the 1.5-kb human Ela cDNA encodes the segment between Ile- 23 and carboxyl terminus of the Ela subunit including the sequence around the two phosphoserine residue3 (underlined residues 288-311). No presequence or the immediate amino- terminal region is encoded by the hE1a-2 cDNA. The 3'-end

3450 Molecular Phenotypes in Maple Syrup Urine Disease

hEl a-4

5 ' v E

vv S A

v P

v P

3'

L hEla-2

""

A.Acc I E.Eco R I P.Pst I S.Srna I

" , -"+ P !a-

t" -" hEla-1

0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 I Kb

FIG. 1. Restriction maps and sequencing strategies of human Ela cDNAs. Solid boxes and lines depict coding and noncoding regions, respectively, of hEla-4 (0.2 kb), hEla-2 (1.5 kb), and hE1a-l (0.9 kb) cDNAs. Vertical arrowheads point to various restriction sites on the inserts. Ordered 5' + 3' serial deletions of both strands in the Bluescript SK- were produced with exonuclease I11 (36). Horizontal arrows indicate orientations and regions of the inserts analyzed by the double-strand sequencing method (37).

of the human Ela-2 clone consists of a 416-bp noncoding region followed by a 20-bp poly(A)+ tail.

The 0.9-kb human Ela cDNA (hE1a-l) begins at Asn-256 (Fig. 2) and continues through the 3'-end with a poly(A)+ tail of 57 bp (not shown). A single nucleotide difference from the sequence of the 1.5-kb human Ela (hEla-2) clone is located in the noncoding region (G + A at 1403, Fig. 2). No variations in the nucleotide sequence were observed in the coding region between the two clones.

To clone the immediate amino-terminal sequence (residues 1-22) of the human E1a subunit, the 5'-end EcoRI fragment (ordinate 1-351) of a bovine Ela cDNA (20) generated via a cleavage at its internal EcoRI site was used as a probe to rescreen the above 82 (hEla-l) positive clones from the Clon- tech libary. A third clone, hElol-4, which contained a 196-bp EcoRI fragment was isolated by this method (Fig. 1). The hEla-4 cDNA encodes the amino-terminal sequence (residues 1-22) of the mature human Ela subunit, and a presequence of 43 residues without an initiation methionine. Thus the composite human Ela cDNA sequence derived from clones hEla-l, -2, and -4 consists of 1783 bp encoding the entire human Ela subunit of 400 amino acids with a calculated molecular mass of 45,552 daltons (Fig. 2). This value is in good agreement with the one (Mr = 46,000) determined by SDS-polyacrylamide gel electrophoresis for the human Ela subunit (41). The deduced amino-terminal sequence (underlined residues 1-32, Fig. 2) of human Ela is identical to that determined for the bovine Ela subunit (20). An internal EcoRI site is located at ordinate 191 in the coding region (Fig. 2), equivalent to the one present in the bovine Ela cDNA (20).

Northern and Western Blottings with Normal and MSUD Fibroblasts-To study MSUD at the mRNA level, Northern blotting was carried out with poly(A)+ fractions prepared from cultured fibroblasts from normal and MSUD patients. A mixture of nicked-translated human Ela (0.9 kb, this work) and human EP (27) cDNAs were used as probes. Fig. 3 shows that Ez (4.6 kb) and Ela (2.0 kb) mRNAs are present in normal size and abundance in fibroblasts from classical MSUD patients P.K., A.L., F.J., and Ech, as well as thiamin- responsive patient WG-34 (28). The level of El mRNA is significantly reduced in fibroblasts from intermediate patient

110 AspLeuVaIYh~GlvCloTvrArgGlll~Gl~V~1LevHeLTvrArgAspTyrProLeuCluLeuPhenefAiaClnCvsTyrClyAsn CACCTCGTGTITGCCCAGTACCGGGAGGCAGGTGTGCTGATGTATCG~ACTACCCCCT~MCTATTCATGGCCCAGTGCTAT~CAAC 538

! x 1 3 0

?on S s r C i u C l y A r p A l a H i s A l a G l v P h r * s n P h e l \ l a i i l l ~ ~ ~ ~ ACTGACCGGGACGCCCATCCCGGC~CMC~CGCTGCCACAC~TGAGTGCC~:CATCATCTTCTTCTGCC~~CMTGGCTACGCCATC 808

? I n 22n

270 2 L Q SerThrProThrSerGluCInTyrAr~ClyAspGI~lleAlaAlaArgGlyProGlyTyrciylieMetSerIleAr~ValArpGlyAirn TCCACGCCCACCTCTGAGCACTATCGCCGCGATGGCATTGCAGCACGA~CCCC~GTAT~CATCATGTCMTCCGCGTGGATGGTMT 898

? m

A s p V a l P h ~ A l ~ V ~ 1 T y r A ~ ~ ~ ~ ~ h ~ L y ~ G ~ u * ~ ~ A ~ ~ A ~ ~ A ~ ~ A ~ ~ V ~ l A l a G i ~ r n C i n P r o P h e L e u l l e G l u * l ~ e t T h r T v r GAI'GTCTTTGCCGTATACMCGCCAC~GGAGGCCCGACGGCG~CTGTGGCAGAGMCCAGCCCT~CTCATCGA~CCATGACCTAC 988

260 2 7 0 280

290 I,

AreIieClyH~sHinSerThrSerAr;pArpSerSe~~laTyrAirpSerValA~~GluYal*snIyr'lrpA~pH~sYrolle 700 ;t x n

ACCATCGGGCACCACAGCACCAGTGACGACAG~CAGCGTACCCCTCGGTGGATGAGGTC~TTACTGGGAT~CAGGACCACCCCATC 1078

120 110 760 S e r A r ~ L e ~ r g H i s ' ~ y r L r u L e u S e r C l n C l y T r p T r p * s p C l u G l ~ G ~ ~ G l ~ G ~ ~ L ~ ~ A ~ ~ ' ~ ~ ~ A ~ g ~ y ~ ~ ~ ~ ~ S ~ ~ A ~ ~ A ~ ~ L y ~ V ~ ~ M ~ ~ T C C C G G C T G C G G C R C T A T C T G C T C A C C C M ~ C T G G T G G G 1168

3 5 0 ~ : l u * l a P h r G l u C l n A l a C l u * ~ ~ L ? ~ P ~ ~ L ~ ~ P ~ ~ A ~ ~ P ~ ~ A ~ ~ L ~ ~ L ~ ~ P h ~ S ~ ~ A ~ p V ~ l T y r G l n C l ~ e t P r o A l a G l n L e ~ r g GAGGCCTTTGAGCAGGCCGAGC~AAGCCCiUACCMCCC~MCCTGCTCTTCTCAGACGTGTATCAGGAGATGCCCGCCCAGCTCCCC 1258

360 770

180 300 A00 L y s C l n C i n C l u S e r L e ~ l ~ A ~ g ~ ~ ~ L ~ ~ G l ~ T h ~ T y ~ G l ~ G l ~ ~ ~ T ~ ~ P ~ ~ L ~ u * s p H i ~ P h ~ A ~ ~ L y ~ * * * MGCAGCAGGACTCTCTGCCCCGCCACCTGCAGACCTACG~GAGCACTACCCACTffiATCACTTCGATMGTGAGACCTGCTCAGCCCA 1 3 U

CI:CCCACCCATCCTCAGCTACCCCGAGAGGTAGCCCCACTCTM~~AGCAGG~GACCTGACAGCACACCACTGTC~CCCCAGTCAG I438

CTCCCTCTRhllhTACTCAGCGGC~AGGGCGGCTGCCACTCTTCACCCCTGCT~CTCCCGGCTGTT~.CA~GTCAGG~ACAGCATCTCCA 1528

GCAGTTGCTGAI:CCTCCGTCAGCCCCCTCTTCACCTGTTG~ACAGTGCCTTCTCCCA~GGCT~GTGAGGGCACATTCAGGACTAGM 1618

GCCCCTCTGGCCATGGGGT~~ACATGCCACCTCAGCCTGTCC~CTTGCGCAGGTGCGAGTGGCCAGCAGAGGTCACG~CTGCAT 1708

C T C T G C G C C T G G C T C T C T I \ U U A l U U V L h A h A h A M i l l U A 1181

FIG. 2. Composite nucleotide and deduced amino acid sequences of the complete human Ela cDNA. The numbers on the far right correspond to ordinates of the last nucleotide in each row. The numbers aboue the sequence refer to the positions of amino acid corresponding to the deduced bovine E1a sequence (20). The amino- terminal sequence (residues 1-32) determined experimentally with bovine &a subunit (20) and the published sequence (resides 288- 311) (39, 40) around the 2 phosphoserine (*) residues are underlined. The internal EcoRI site (ordinates 191-196) and the polyadenylation signal, AATAAA (at ordinates 1499-1504), are also underlined. The stop codon (***) is located at ordinates 1128-1130.

Lo, whereas the content of EP mRNA is normal in Lo cells. The minor bands in addition to Ez and Ela mRNA appeared to be nonspecific as their intensities varied depending on the washing conditions.

The levels of specific protein subunits (E1a, El@, and E p )

of the branched-chain a-keto acid dehydrogenase complex in cultured fibroblasts were measured by Western blot analysis. After electrotransfer, resolved proteins from normal and MSUD fibroblasts were blotted with either a mixture of affinity-purified anti-E1a and anti-EIP or anti-EZ antibodies. Fig. 4 (Panel A ) shows that E1a and El@ subunits are present in normal size and abundance in fibroblasts from classical patients F.J., A.L., and Ech, and slightly reduced in cells from thiamin-responsive patient WG-34. Both Ela and EIP subunits are nearly absent in cells from the intermediate patient Lo, whereas the levels of both subunits are reduced in cells from the Mennonite patient (P.K.). The Ez subunit exists in normal size and quantity in fibroblasts from patients F.J., P.K., and Lo (Fig. 4, Panel B ) . By contrast, the Ez subunit is completely absent in cells from patient Ech and markedly reduced in cells from thiamin-responsive patient WG-34 and classical patient A.L.


( 2 . O K b ) E l a -

NORTHERN

FIG. 3. Northern blot analysis with poly(A)+ RNA isolated from normal and maple syrup-urine disease (MSUD) fibroblasts. Poly(A)' RNA ( 5 rgllane) was resolved by electrophoresis on 0.8% agarose gel followed by Northern blot analysis. A mixed probe consisting of nick-translated human E2 (hE2-1) (1.6 kb) and human E,n (hEln-l) (0.9 kb) was used. The sizes of human E2 (4.6 kb) and E,a (2.0 kb) were determined previously (20, 27). P.K. (Mennonite), A.L., F.J., and Ech are classical MSUD patients. Lo is an intermediate patient. WG-34 is a thiamin-responsive patient (28). The two A.L. lanes represent results from two different experiments.

E l a - E I P -

0

B

E 2 -

FIG. 4. Western blot analysis with cel lular extracts from norma l and MSUD fibroblasts. Cellular extracts from cultured fibroblasts were subjected to Western blot analysis (300 pg/lane) with either a combination of affinity-purified anti-Eln and anti-EI/3 (Panel A ) or anti-& (Panel R ) antibodies as a probe. Details are described under "Experimental Procedures." The sizes of bovine Eln, E& and E2 markers (far left lanes) determined by SDS-polyacrylamide gel electrophoresis are 46,500, 36,500, and 52,000 daltons, respectively (2-5). F.J., P.K. (Mennonite), Lo, Ech, and A.L. are classical patients. WG-34 is a thiamin-responsive patient. BCKD, branched-chain n- keto acid dehydrogenase.

Characterization of an E2 Mutant-The E2 subunit of the branched-chain a-keto acid dehydrogenase complex was reported to be absent in a lymphoblastoid cell line GM-1366 (21). To elucidate the mechanism for this deficiency, we measured mRNA and subunit contents of the branched-chain a-keto acid dehydrogenase complex in normal and GM-1366 lymphoblasts. Fig. 5 (upper panel) shows that the content of E2 mRNA is markedly lower in GM-1366 lymphoblasts than in normal cells, whereas the level of Ela mRNA appears to be normal in GM-1366 cells. The reduced E2 mRNA level is consistent with the absence of E2 subunit in GM-1366 lymphoblasts (Fig. 5, lower panel). The Ela subunit is present in normal abundance in GM-1366 lymphoblasts.

Measurements of El Actiuity in MSUD Cells-Table I shows

E l a - - - E 2

WESTERN FIG. 5. mRNA and subuni t contents of the branched chain

a-keto acid dehydrogenase complex in &-deficient MSUD lymphoblasts (GM-1366). Northern blot analysis (for mRNA) was carried out with either human Eln (hEta-1) or human E2 (hR-1) cDNA as a probe. For Western blotting (protein), affinity-purified anti-E,n, or combined rabbit antisera to native and SDS-denatured E2 (bovine) (27) were used as a probe. See "Experimental Procedures" for details.

TABLE I Activities of E, and E, components in cultured fibroblasts from

normals and MSUD patients Harvested fibroblasts were treated with 1 mM a-chloroisocaproate

at 37 "C for 15 min. The treated cells were spun, resuspended in Krebs buffer and freeze-thaw disrupted for assays for the total E, activity. See "Experimental Procedures" for details.

Cell line Activity

Overall E , (X IO?) E3

nmol C02/min/mg protein Normal 0.09-0.1s 7.7-9.5 P. K. 0 0.1 57.1 F. J. 0 0 69.6 WG-34 0.02 3.9 52.6 Lo 0 0.06 50.4

37.3-51.3

the activity levels of the branched chain a-keto acid dehydrogenase complex and El and Es components in cultured fibroblasts from MSUD patients P.K., F.J., and Lo and the thiamin-responsive patient WG-34. The results indicate that the overall and El activities are absent in P.K., F.J., and Lo cells and markedly reduced in WG-34 cells (20-40% residual activity). As a control, Ea activity is normal in cells from these patients.

DISCUSSION

The present communication describes the cloning of complete human E I a cDNA and different mRNA and subunit contents of the branched-chain a-keto acid dehydrogenase complex in MSUD cells. This approach has allowed us to classify MSUD based on its molecular phenotypes (Table 11). These comprise: Type I, where the levels of Ela mRNA and EIa and El@ subunits are normal, but E, activity is deficient as observed with F.J. cells; Type 11, where the Ela mRNA is present in normal quantity, whereas the levels of Ela and El@ subunits are reduced (P.K.); Type 111, where the level of E1a

3452 Molecular Phenotypes in Maple Syrup Urine Disease

TABLE I1 Molecular phenotypes in cultured cells from different MSUD patients

Symbols: +, present in normal size and abundance; -, absent or much reduced in abundance; ND, not determined.

Cell MSUD Molecular Elol K P Ez line phenotype phenotype mRNA Subunit mRNA Subunit Subunit

Normal None None + + + + + F. J. Classical I + + + + +

I1 P. K. Classical + - - + + I11 Lo Intermediate - - - + +

GM-1366 IV Classical + + ND Ech Classical V + + + + A. L. Classical V + + + + WG-34" Thiamin-responsive V + + + +

- - - - -

The original thiamin-responsive MSUD patient described by C. Scriver (28).

mRNA is markedly reduced with a concomitant loss of E l a and EIP subunits (Lo); Type IV, where the levels of both E2 mRNA and E2 subunit are markedly reduced (GM-1366); and Type V, where the E2 mRNA is normally expressed, however, the EP subunit is absent or markedly reduced in cells (Ech, A.L., and thiamin-responsive WG-34). The above five molecular phenotypes are derived from results of Northern and Western blotting with cells from only seven MSUD patients. We anticipate that new molecular phenotypes will emerge as more MSUD cell lines are studied. There appears to be no correlation between clinical (15) and molecular phenotypes since multiple molecular phenotypes are observed within classical MSUD.

The existence of different molecular phenotypes supports the genetic heterogeneity in MSUD demonstrated previously by complementation studies (42, 43). The multiple molecular phenotypes probably reflect the large number of the genes encoding the branched-chain a-keto acid dehydrogenase complex. The results have thus identified the genetic locus affected in different MSUD patients. Although specific mutations leading to these phenotypes remain to be elucidated, the following mechanisms can be speculated. In Type I MSUD (F.J.) (Table 11), a structural mutation in the E l a or EIP subunit is likely. The presence of a mutant El protein is consistent with a reduced affinity of the branched chain a- keto acid dehydrogenase complex for a-ketoisovalerate in fibroblasts from F.J.2 On the other hand, a defect in the expression of E l a and E2 genes is the probable cause for Type I11 (Lo) and Type IV (GM-1366) MSUD, respectively. The levels of mRNA and subunit are concomitantly reduced or absent in these cells. The absence of both Ela and EIP subunits in Lo (Type 111) is of interest if one assumes the mutation involves only the expression of the E l a gene. I t is speculated that the EIP subunit is normally expressed, but is rapidly degraded because of its failure to assemble into a stable a2P2 structure with Ela. The situation appears to be similar to that observed with patients with primary lactic acidosis. Both Ela and El@ subunits of the pyruvate dehydrogenase complex are absent in cells from these patients (44, 45).

The mechanisms for Type I1 and Type V are presently uncertain. The results obtained with P.K. cells (Type 11) (Table 11) are consistent with a structural mutation in either E l a or EIP subunit. Alternatively, the expression of E$ gene may be deficient in Type I1 resulting in a reduced level of the EIP subunit. In either case, both E l a and EIP chains may become unstable as a result of improper folding and assembly similar to Type 111. P.K. is a member of the Mennonite kindred where the prevalence of MSUD is 1 in 175 because

* T. A. Griffin and D. T. Chuang, manuscript in preparation.

of consanguinity. It is believed the same mutation is trans- mitted within the Mennonite families. As for Type V, the absence of E2 subunit in MSUD has been described previously (22, 23). Nonetheless, the normal expression of E2 mRNA with no detectable E2 subunit in cells as observed in Ech (Fig. 4B) is of interest. This may have arisen from a frame-shift or non-sense mutation, or the synthesis of an unstable protein. Actual sequence determination of the Type V transcript, e.g. by the polymerase chain reaction (46), is in progress to sub- stantiate these possibilities.

The classification of the thiamin-responsive (WG-34) patient under type V requires specific comments. The patient was originally described by Scriver et al. (28) to have re- sponded favorably to oral thiamin treatment at a dosage of 10 mg/day. On this regimen, her branched-chain amino acid levels in plasma returned to normal without restriction of dietary proteins. Intact fibroblasts from WG-34 exhibited 40% of the normal decarboxylation rate when incubated with a- [1-14C] ketoisovalerate (46). Although E l a and El@ subunits in WG-34 cells are somewhat diminished, the marked reduc- tion in the E2 subunit content would indicate that the E2 locus is affected (Fig. 4). Since the level of E2 mRNA is normal while the protein content is low, a structural mutation is the likely cause for the synthesis of an unstable El. We have shown previously that the K,,, value of the branched- chain a-keto acid dehydrogenase complex for thiamin pyrophosphate is elevated in WG-34 cells compared to normal cells (47). The apparent deficiency in EP is therefore unex- pected, because the E1a subunit contains the binding site for thiamin pyrophosphate (48). One possible explanation is that the presence of a normal E2 is essential for the efficient binding of thiamin pyrophosphate to El . It has been shown with the chicken glycine cleavage system that the addition of the lipoate-bearing H-protein causes a conformation change on the pyridoxal phosphate containing P-protein and significantly increases the affinity of the latter enzyme component for substrate glycine (49). Whether a similar mechanism exists in the branched-chain a-keto chain dehydrogenase complex will have to be studied. Alternatively, Heffelfinger et al. (50) have shown that the E1a subunit is protected against chymotryptic digestion by added thiamin pyrophosphate and a-ketoisocaproate and suggested that these reagents stabilize the residual branched-chain a-keto acid dehydrogenase complex in thiamin-responsive MSUD in uiuo.

The putative amino terminus of the human Ela subunit is Ser, as deduced from a complete alignment with the determined amino terminal sequence (residues 1-32) of the bovine Ela subunit (20). However, the amino terminus of the rat Ela subunit was determined to be Phe (48). We have proposed that the presence of Pro (residue 2) in the rat Ela subunit


appears to have resulted in an alternate cleavage at the bond between Gln (-1) and Phe (+l), compared to bovine or human Elm subunit, to produce the amino-terminal Phe for the rat Ela peptide (20). Comparison of the entire mature human E l m sequence (this work) with corresponding rat (51) and bovine (20) Ela sequences discloses that there is 95 and 96% identity, respectively, at the amino acid level. Homologies in the presequence region is significantly reduced to 60% between human and bovine, and 70% between human and rat pre-Ela subunits. The region between residues 287 and 319 (Fig. 2) containing phosphorylation sites 1 and 2 (39, 40) is highly conserved (>99%) among all three species. There is only a single amino acid substitution, i.e. LeuP4’ (human) or LeuP5’ (rat) to Metz4’ (bovine) in the entire region spanning 133 amino acids. The conservation of primary structure is apparently essential for the folding of this region into a correct conformation to be recognized by the specific kinase.

While the present paper was in review, Zhang et al. (51) reported the cloning of a partial human E1a cDNA (1552 bp), which is similar to clone hEla-2 (1550 bp) of this work. The only difference between the two human Ela cDNAs is a base change from C (ordinate 738, this work) to A (ordinate 632) (52). This results in an alteration of Ala (this work) to Asp (52) in the amino acid. Moreover, both studies show that the Ela mRNA is present in normal size and abundance in thiamin-responsive WG-34 cells.

In summary, the present study has demonstrated the diver- sity and complexity of MSUD at mRNA and protein levels. The molecular phenotype associated with each patient will help develop rational approaches to the management of this metabolic disorder. The isolation of a complete EIa cDNA of the human branched-chain a-keto acid dehydrogenase complex will permit further investigation of possible structural mutations in Type I MSUD at the molecular level. Cloning of the human Ela and EP genes is in progress.

Acknowledgments-We thank Yoshie Hervey for assistance in cell culture, Teiko Kimura for help in Northern blotting, and Lily Ku for the word processing of the manuscript.

REFERENCES 1. Reed, L. J., Pettit, F. M., Yeaman, S. J., Teague, W. M., and

Bleile, D. M. (1980) in Enzyme Regulation and Mechanism of Action (Mildner, P., and Ries, B., eds) Vol. 60, pp. 47-56, Pergamon Press, New York

2. Pettit, F. H., Yeaman, S. J., and Reed, L. J . (1978) Proc. Natl. Acad. Sci. U. S. A. 7 5 , 4881-4885

3. Heffelfinger, S. C., Sewell, E. T., and Danner, D. J. (1983) Biochemistry 22,5519-5522

4. Chuang, D. T., Hu, C.-W. C., Ku, L. S., Niu, W.-L., Myers, D. E., and Cox, R. P. (1984) J. Biol. Chem. 259,9277-9284

5. Ono, K., Hakozaki, M., Nishimaki, H., and Kochi, H. (1987) Biochem. Med. Metab. Biol. 3 7 , 133-141

6. Odessey, R. (1982) Biochem. J . 2 0 4 , 353-356 7. Lau, K. S., Fatania, H. R., and Randle, P. J. (1982) FEBS Lett.

8. Paxton, R., and Harris, R. A. (1982) J. Biol. Chem. 2 5 7 , 14433-

9. Damuni, Z., Merryfield, M. L., Humphreys, J. S., and Reed, L. J.

10. Fatania, H. R., Patston, P. A., and Randle, P. J. (1983) FEBS

144,57-62

14439

(1984) Proc. Natl. Acad. Sci. U. S. A. 81,4335-4338

Lett. 158 , 234-238 11. Cook, K. G., Bradford, A. P., and Yeaman, S. J. (1985) Biochem.

J . 2 2 5 , 731-735 12. Griffin, T. A., Lau, K. S., and Chuang, D. T. (1988) J. Biol. Chem.

13. Packman, L. C., and Perham, R. N. (1986) FEBS Lett 206,193-

14. Dancis, J., Jansen, V., Hutzler, J., and Levitz, M. (1963) Biochim.

15. Tanaka, K., and Rosenberg, L. E. (1983) in The Metabolic Basis

263,14008-14014

198

Biophys. Acta 77,523-524

of Inherited Disease (Stanbury, J. B., Wyngaarden, J. B., Fred- rickson, D. S., Goldstein, J. L., and Brown, M. S., eds) 5th Ed., pp. 440-473, McGraw-Hill Book Co., New York

16. Snyderman, S. E., Goldstein, F., Sansaricq, C., and Norton, P. M. (1984) Pediatr. Res. 18,851-853

17. Chuang, D. T., Niu, W.-L., and Cox, R. P. (1981) Biochem. J.

18. Chuang, D. T., Ku, L. S., Kerr, D. S., and Cox, R. P. (1982) Am. J . Hum. Genet. 34,416-424

19. Gonzales-Rios, M. C., Chuang, D. T., Cox, R. P., Schmidt, K., Knopf, K., and Packman, S. (1985) Clin. Genet. 2 7 , 153-159

20. Hu, C.-W. C., Lau, K. S., Griffin, T. A., Chuang, J. L., Fisher, C. W., Cox, R. P., and Chuang, D. T. (1988) J. Biol. Chem. 2 6 3 , 9007-9014

21. Rudiger, H. W., Langenbeck, U., Schulge-Schencking, M., and Goedde, H. W. (1972) Humungenetik 14 , 257-263

22. Indo, Y., Kitano, A., Endo, F., Akaboshi, I., and Matsuda, I. (1987) J. Clin. Inuest. 80,63-70

23. Danner, D. J., Armstrong, N., Heffelfinger, E. T., Sewell, J. H.,

860 Priest, J. H., and Elsas, L. J. (1985) J. Clin. Znuest. 7 5 , 858-

24. Haworth, J. C., Perry, T. L., Blass, J . P., Hansen, G., and Urquhart, N. (1976) Pediatrics 5 8 , 564-572

25. Robinson. B. H.. Taylor. J., and Sherwood. W. G. (1977) Pediatr.

200,59-67

Res. 11; 119811202 . .

26. Robinson. B. H.. Tavlor. J.. Kahler. S. G.. and Kirkman. H. N. (1981) Eur. J. Ped&.’ 136,35-39

27. Lau, K. S., Griffin, T. A., Hu, C.-W. C., and Chuang, D. T. (1988) Biochemistry 27,1972-1981

28. Scriver, C. R., Clow, C. L., MacKenzie, S., and Delvin, E. (1971) Lancet i, 310-312

29. Chuang, D. T., and Cox, R. P. (1988) Methods Enzymol. 166 , 135-146

30. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Bwl. Chem. 193 , 265-275

31. Huynh, T. V., Young, R. A., and Davis, R. W. (1985) in DNA Cloning: A Practical Approach (Glover, D. M., ed) Vol. I, pp. 49-78, IRL Press Ltd., Oxford

32. Hu, C.-W. C., Griffin, T. A., Lau, K. S., Cox, R. P., and Chuang, D. T. (1986) J. Biol. Chem. 2 6 1 , 343-349

33. Maniatis, T., Fritsch, E., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

34. Southern, E. (1975) J. Mol. Biol. 98,503-517 35. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl.

Acad. Sci. U. S. A. 74,5463-5467 36. 37. 38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48. 49.

50.

51.

52.

Henikoff, S. (1984) Gene (Amst.) 28,351-359 Chen, E. Y., and Seeburg, P. H. (1985) DNA (N. Y.) 4,165-170 Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter,

W. J. (1979) Biochemistry 18,5294-5299 Cook, K. G., Bradford, A. P., Yeaman, S. J., Aitken, A., Fearnley,

I. M., and Walker, J. E. (1984) Eur. J. Biochem. 145,587-591 Paxton, R., Kuntz, M., and Harris, R. A. (1986) Arch. Biochem.

Biophys. 2 4 4 , 187-201 Ono, K., Hakozaki, M., Nishimaki, H., and Kochi, H. (1987)

Biochem. Med. Metab. Biol. 3 7 , 133-141 Lyon, L. B., Cox, R. P., and Dancis, J. (1973) Nature 2 4 3 , 533-

535 Singh, S., Willers, I., and Goedde, H. W. (1977) Clin. Genet. 11,

277 Ho, L., Hu, C.-W. C., Packman, S., and Patel, M. S. (1986) J.

Clin. Znuest. 7 8 , 844-847 Wicking, C. A., Scholem, R. D., Hunt, S. M., and Brown, G. K.

(1986) Bwchem. J . 239,89-96 Emburry, S. H., Scharf, S. J., Saiki, R. K., Gholson, M. A.,

Golbus, M., Arnheim, N., and Erlick, H. (1987) N. Engl. J . Med. 316,656-661

Chuang, D. T., Ku, L. S., and Cox, R. P. (1982) Proc. Natl. Acad. Sci. U. S. A. 79 , 3300-3304

Stepp, L. R., and Reed, L. J. (1985) Biochemistry 2 4 , 7187-7191 Hiraga, K., and Kikuchi, G. (1980) J. Biol. Chem. 2 5 5 , 116’71-

Heffelfinger, S. C., Sewell, E. T., Elsas, L. J., and Danner, D. J.

Zhang, B., Kuntz, M. J., Goodwin, G. W., Harris, R. A., and

Zhang, B., Crabb, D., and Harris, R. H. (1988) Gene (Amst.) 6 9 ,

11676

(1984) Am. J . Hum. Genet. 36,802-807

Crabb, D. W. (1987) J. Biol. Chem. 2 6 2 , 15220-15224

159-164

Documents

Molecular Phenotypes in Cultured Maple Syrup Urine Disease Cells