Untitled Molecular and Biochemical Basis of Intermediate Maple Syrup Urine Disease Occurrence of Homozygous G245R and F364C Mutations at the Ela Locus of Hispanic-Mexican Patients Jacinta L. Chuang,* James R. Davie,* Jeffrey M. Chinsky,* R. Max Wynn,* Rody P. Cox,* and David T. Chuang * *Departments of Biochemistry and Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas 75235; and tDivision of Human Genetics, Department of Pediatrics, University of Maryland School of Medicine, Baltimore, Maryland 21201 Abstract Introduction Maple syrup urine disease (MSUD) is caused by a deficiency of the mitochondrial branched-chain a-keto acid dehydro- genase (BCKAD) complex. The multienzyme complex com- prises five enzyme components, including the El decar- boxylase with a heterotetrameric (a242) structure. Four unrelated Hispanic-Mexican MSUD patients with the inter- mediate clinical phenotype were diagnosed 7 to 22 mo after birth during evaluation for developmental delay. Three of the four patients were found homozygous for G to A transi- tion at base 895 (exon 7) of the Ela locus, which changes Gly-245 to Arg (G245R) in that subunit. The remaining patient was homozygous for T to G transversion at base 1,253 in the Ela gene, which converts Phe-364 to Cys (F364C) in the gene product. Transfection studies in Ela- deficient lymphoblasts indicate that both G245R and F364C mutant Ela subunits were unable to significantly reconsti- tute BCKAD activity. Western blotting showed that both mutant Ela subunits in transfected cells failed to efficiently rescue the normal E1l8 through assembly. The putative as- sembly defect was confirmed by pulse-chase labeling of El subunits in a chaperone-augmented bacterial overex- pression system. The kinetics of initial assembly of the G245R Ela subunit with the normal E1,8 was shown to be slower than the normal Ela. No detectable assembly of the F364C Ela subunit with normal E18 was observed during the 2 h chase. Small amounts of recombinant mutant El proteins were produced after 15 h induction with isopropyl thiogalactoside and exhibited very low or no El activity. Our study establishes that G245R and F364C mutations in the Elta subunit disrupt both the El heterotetrameric assembly and function of the BCKAD complex. Moreover, the results suggest that the G245R mutant Elta allele may be important in the Hispanic-Mexican population. (J. Clin. Invest. 1995. 95:954-963.) Key words: branched-chain a- keto acid dehydrogenase complex * El dysfunction * pulse- chase labeling , defective heterotetrameric assembly * mo- lecular chaperones Address correspondence to David T. Chuang, Ph.D., Department of Biochemistry, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75235-9038. Phone: 214-648-2457; FAX: 214-648-8856; e-mail: COX02@UTSW.SWMED.EDU. Received for publication 18 July 1994 and in revised form 18 No- vember 1994. Maple syrup urine disease (MSUD)' or branched-chain a-keto- nuria is an autosomal recessively inherited deficiency in the mitochondrial branched-chain a-keto acid dehydrogenase (BCKAD) complex. The biochemical basis of this disease is the inability to metabolize branched-chain a-keto acids (BCKAs) derived from the essential branched-chained amino acids (BCAAs) leucine, isoleucine, and valine. The elevated BCAAs and BCKAs may have severe clinical consequences including ketoacidosis, mental retardation, and neurological impairment. Variations in clinical presentations have led to the classification of MSUD into five clinical phenotypes, i.e., classic, intermedi- ate, intermittent, thiamine-responsive, and E3-deficient (1, 2). The classic form, which comprises 75% of MSUD patients, is manifested within the first 2 wk of life with poor feeding, leth- argy, seizures, coma, and death if untreated. Intermediate MSUD is associated with elevated BCAAs and BCKAs, with progressive mental retardation and developmental delay without a history of catastrophic illness. The diagnosis is usually delayed for many months. An intermittent form of MSUD may have normal levels of BCAAs, normal intelligence and development until a stress, e.g., infection, precipitates decompensation with ketoacidosis and neurologic symptoms, which are usually re- versed with dietary treatment. Thiamine-responsive MSUD is similar to the intermediate phenotype but responds to pharmaco- logic doses of thiamine with normalization of BCAAs. The E3- deficient MSUD is caused by defects in the dehydrogenase (E3) component of the BCKAD complex that is common to the pyruvate and a-ketoglutarate dehydrogenase complexes. Pa- tients with E3 deficiency have dysfunction of all three enzyme complexes, and patients usually die in infancy with severe lactic acidosis (3). The mammalian BCKAD complex is a mitochondrial multi- enzyme complex, which catalyzes the oxidative decarboxylation of the three BCKAs derived from BCAAs (4, 5). The enzyme complex is organized around a cubic core comprising 24 identi- cal dihydrolipoyl transacylase (E2) subunits. The other enzyme components attached to the E2 core include a branched-chain a-keto acid decarboxylase (El), a dihydrolipoamide dehydro- genase (E3), a specific kinase and a specific phosphatase. The latter two enzyme components are responsible for the regulation of BCKAD complex by phosphorylation (inactivation)/de- 1. Abbreviations used in this paper: BCAA, branched-chain amino acid; BCKA, branched-chain a-keto acid; BCKAD, branched-chain a-keto acid dehydrogenase; El, branched-chain a-keto acid decarboxylase; E2, dihydrolipoyl transacylase; E3, dihydrolipoyl dehydrogenase; Hsp, heat- shock protein; IPTG, isopropyl thiogalactoside; MSUD, maple syrup urine disease; SSCP, single-stranded conformational polymorphism; TEV, tobacco etch virus; TPP, thiamine pyrophosphate. 954 Chuang et al. J. Clin. Invest. © The American Society for Clinical Investigation, Inc. 0021-9738/95/02/0954/10 $2.00 Volume 95, March 1995, 954-963 Table L Clinical Presentations of Intermediate MSUD Patients Plasma leucine Dietary leucine* tolerance Developmental status (assessed Patient Age at diagnosis Presenting symptoms AM mg/kg per day age/chronological age) A. G. 12 mo Developmental delay, seizure disorder 2572 40-50 6.0 yr/8.0 yr J. B. 16 mo Developmental delay, seizure disorder 1225 10-50 8.0-9.0 yr/15 yr P. R. 7 mo Developmental delay 1218t 50 20 mo/27 mo B. S. 22 mo Developmental delay, seizure disorder 3347 70 12-18 mo/5 yr * Total dietary leucine tolerated to maintain plasma leucine concentrations of 200-400 MM (normal range 50-130 MM). tLevel obtained on partially restricted protein intake (1 gfkg/day). A. G., J. B., P. R., and B. S. are intermediate MSUD patients. phosphorylation (activation) (6). The El component is a heter- otetramer consisting of two Ela (47,000 Mr) and two El/I (37,000 Mr) subunits. The E3 component, a flavoprotein, is a homodimer. Both kinase and phosphatase are products of single genes. Multiple copies of the six distinct subunits of the BCKAD complex are synthesized in the cytosol and imported into the mitochondria (7). In the matrix space of this organelle, mature subunits are assembled into a multienzyme complex of 4-5 x 106 daltons. MSUD is genetically heterogeneous, and mutations in Ela, El1,, E2 and E3 loci have been identified (1). Some of these mutations have been characterized at the molecular and bio- chemical level. For example, the Y393N substitution in Ela, which is present in Mennonite patients with classic MSUD, was shown to impede El assembly resulting in the preferential degradation of E1l/ (8, 9). However, these studies were con- fined to classic MSUD patients, and little is known about the variant forms of MSUD. In the present report, we describe the clinical and molecular characterization of four MSUD patients with an intermediate phenotype. These patients, all of Hispanic- Mexican origin, were identified many months after birth during evaluation for developmental delay, without acute neurologic symptoms. Molecular studies show that three of the four unre- lated patients are homozygous for a missense mutation in the Ela locus. The remaining patient is homozygous for a second mutation also in the Ela gene. Both mutations produce inactive or slightly active BCKAD complex when the mutant peptide was transiently expressed in Ela-deficient lymphoblasts. More- over, by pulse-chase labeling of El subunits synthesized in Escherichia coli, we show that these mutations in Ela alter the kinetics of assembly with El/. Methods Case reports. A.G., a male Mexican infant, appeared normal during the first few months of life except for mild feeding problems. Although development seemed normal during the first 5-6 mo of age, he was unable to sit without support until 11 mo of age. During infancy, he sustained several 30-s tonic seizures associated with fever. He presented at 12 mo of age for evaluation of his developmental delay and possible seizure disorder. Physical exam was significant for relative microcephaly (1Oth percentile compared with length and weight, 75th and 25th percen- tiles, respectively), mild hypotonia with cerebellar ataxia. MSUD was diagnosed on the basis of elevated serum BCAAs and alloisoleucine (Table I). After initiation of BCAA-restricted diet, neurologic evalua- tion showed normal tone, reflexes, and cerebellar function. J.B. was a 5-lb male born to Mexican parents of small stature (< 10th percentile). During infancy, he had formula intolerance and exhibited significant developmental delay, manifested by inability to sit, stand, or crawl by 17 mo old. He had two brief tonic/clonic seizures at 4 mo, which were easily controlled with phenobarbital. At 17 mo, examination was significant for developmental delay; head circumfer- ence, length, and weight at the third percentile; and generalized hypoto- nia. Analysis of serum amino acids demonstrated elevated BCAAs and alloisoleucine, suggesting MSUD. Improved development occurred dur- ing dietary treatment, but hyperkinesis and psychomotor delay persisted. P.R., a Mexican female, demonstrated excessive irritability, poorly tolerated her formula, and showed developmental delay with poor head control, inability to roll over and poor social interactions during the first 4-5 mo of life. Qualitative analysis of serum amino acids showed elevated amino acids including BCAAs. She was placed on 1 g protein/ kg per day diet with improvement in her reactions to auditory and visual stimuli, decreased ataxia and a weight gain. At 7 mo, physical examination revealed length 75th, weight 50th, and head circumference 75th percentiles with generalized hypotonia and developmental delay. Quantitative serum amino acids revealed elevated BCAAs and alloiso- leucine. Strict dietary treatment resulted in further improvement in her development (Table I). B.S. was a 6 lb 8 oz female born to a 14 yr-old-Mexican primi- gravida. Although irritable as an infant she had a good appetite and tolerated standard formula. During her first year of life she was hospital- ized four times with pneumonia. At 9 mo, developmental delay was noted and an EEG was normal. Between 12 and 22 mo, three uncompli- cated tonic/clonic seizures occurred and were controlled with phenobar- bital. At 20 mo, she was readmitted to the hospital with an infection and evaluated for her developmental abnormalities, being unable to stand or control head movements. Physical examination showed length 30th, weight 60th, and head circumference 50th percentiles with diffuse hypotonia, brisk reflexes, and poor motor control. Psychomotor func- tioning was at a 3-4-mo level. An EEG showed a disorganized pattern with paroxysmal spikes, and a CT scan revealed increased white matter lucency consistent with deficient myelinization. A diagnosis of MSUD was made by metabolic studies (elevated serum BCAAs and alloisoleu- cine), however, compliance with therapy was poor, and her most ad- vanced milestones are independent walking with a broad-based gait, and a few words (c 10) were learned. During periods of appropriate BCAA restriction, her motor and developmental skills demonstrated marked improvement. Cell cultures and other cell lines. Fibroblast cultures were derived from skin biopsies and grown in monolayer culture as previously de- scribed (10). Intermediate MSUD patients A.G., J.B., P.R., and B.S. and classic MSUD patient P.V. were provided by Drs. Arthur Beaudet and Edward McCabe (Baylor College of Medicine, Houston, TX). GM- 649 was purchased from the Human Genetic Mutant Cell Repository (Camden, NJ). B.F. was a classic MSUD patient of African-American descent provided by Dr. David Valle (John F. Kennedy Institute, Balti- more). K.U. is a classic Caucasian MSUD patient provided by Dr. Selma Snyderman (New York University Medical Center, New York). L.C. is a classic MSUD patient obtained from the Cell Repository at McGill Medical Center (Montreal, Canada). H.H. is a classic Mennonite MSUD patient provided by Dr. Marc Yudkoff (Childrens Hospital of Intermediate Maple Syrup Urine Disease 955 Philadelphia, PA). S.B. is a classic MSUD patient provided by Dr. Lisa Dahl (University of Minnesota Medical Center, Minneapolis, MN). Classic MSUD fetal fibroblasts, A-1, were obtained from Dr. Rivka Carmi (Saroka Medical Center, Beer Sheva, Israel). Lymphoblasts were grown in RPMI-1640 medium containing 15% heat-inactivated fetal calf serum as described previously (8). Single-stranded conformational polymorphism (SSCP) analysis. All 9 exons of the human Ela gene were amplified by PCR using primers for the flanking intronic sequences ( 12). The PCR reaction was carried out in Tris-HCl pH 9.0 (10 mM), MgCl2 (1 mM), KCl (50 mM), 0.1% Triton X-100, dNTPs (0.2 gLM each), Taq DNA polymerase (2.5 U) and 1 pM of each sense and antisense primer. To the reaction mixture, 10 ,uCi of 32P-dCTP and 200 ng of genomic DNA as template was added, and the reaction cycled 35 times at 94°C for 30 s, 55°C for 30 s, and 72°C for 1 min. 3 .1d of 32P-labeled PCR product were diluted with 27 1.l of a solution containing 95% formamide, 20 mM EDTA, 0.05% bromophenol blue, 0.05% xylene cyanol FF and 37 mM methyl mercury hydroxide. Samples were then heated at 100°C for 2 min and chilled on ice. Electrophoresis with 3 1.d per lane of each sample applied was carried out on a 0.5x MDET gel (J. T. Baker Inc., Phillipsburg, NJ) at constant 6 watts for 18-20 h in lx TBE buffer. The gels were dried on filter paper and exposed to Kodak X-OMAT film. Expression and site-directed mutagenesis of Ela cDNA. The con- struction of the EBO vector for normal human Ela and methods for transient expression and selection of transfected cells have been pre- viously described (8). The EBO-pLpp parental plasmid (11) contains OriP as the origin of replication. The plasmid also harbors the EBNA- 1 sequence encoding Epstein-Barr virus nuclear antigen and the hygro- mycin phosphotransferase sequence that confers antibiotic resistance. The presence of OriP and EBNA-1 sequences allows efficient episomal replication of the plasmid in Epstein-Barr virus transformed lymphoblas- toid cells. Nucleotide changes were carried out using the Altered Site" in vitro mutagenesis system from Promega (Madison, WI). The chim- eric bh Ela cDNA containing the bovine mitochondrial targeting and the mature human sequences (8) was cloned into pAlter- 1 vector, which contained a mutation in the ampicillin-resistant gene rendering the trans- formed cell susceptible to the antibiotic. Single strand DNA was pre- pared from the pAlter-bh Ela plasmid. Mutagenic oligonucleotides con- taining the mutation and the repair oligonucleotide for the ampicillin- resistant gene were annealed simultaneously to the single stranded DNA. T4 DNA polymerase and ligase were added to synthesize an ampicillin- resistant mutant strand. After two rounds of ampicillin selection, first in the repair minus strain of E. coli (BMH71-18 mut S) then in JM- 109 cells, colonies were picked for plasmid DNA preparation. DNA sequencing was performed to confirm the mutation. DNA fragment containing the mutation were excised from pAlter vector by restriction enzyme digestion and ligated into the EBO expression vector (8). Construction of pHisT-El prokaryotic expression vector. The 5' portion of the mature human Ela cDNA sequence was amplified from the pMAL-c2-hEla expression vector (9) using an internal 22-mer antisense primer with sequence 5 '-GTAACAGATGTCGACCCTGTT- 3 ', and a 49-mer sense primer with sequence 5 '-GGCTCTAGACTC- GAGAATCTTTATTTtcaatcatctctggatgacaagc-3' to yield a 601-bp product. The sense primer adds exogenous sequence (shown in upper case) to the 5 ' terminus of the mature E1Ca open-reading frame (shown in lower case). This exogenous sequence includes an Xbal restriction site (shown in bold) followed by sequence encoding the first six amino acids specific for the tobacco etch virus (TEV) protease cleavage (shown underlined). The seventh required amino acid for the TEV protease cleavage is supplied by the amino-terminal serine of the mature Ela sequence. To generate the pHisT-Ela expression vector, the 601-bp amplifica- tion product was cut with XbaI and Narl to yield a 454-bp fragment encoding the TEV cleavage site and the 5' portion of the Ela open- reading frame. The pEBO-hbEla expression vector (8) was cleaved with NarI and XhoI to yield a 1064-bp fragment that encodes the 3' portion of the mature E1a open-reading frame. Both fragments were ligated into the pTrcHisB expression vector (Invitrogen) digested with NheI and XhoI to yield the pHisT-Ela expression vector. To generate the pHisT-E1 expression vector, a BamHI-ScaI fragment (2,555 bp) comprising the trc promoter and the mature human El/i open-reading frame was isolated from expression vector pKK-hEl,B (13) and ligated into the corresponding sites in the host pHisT-Ela expression vector, yielding the pHisT-El expression vector. Mutant Ela variants of the pHisT-hEl expression vector were constructed identically, except for the substitution of 1064-bp NarI-XhoI fragments isolated from variant pEBO-hbEla plasmids harboring the desired Ela mutation. Pulse-chase labeling for kinetics of Ela and EJ,3 assembly. E. coli strain CG-712 and the plasmid pGroESL (for overexpression of chaperonins GroEL and GroES) (14) were received as kind gifts from Dr. Anthony Gatenby of DuPont Experimental Station (Wilmington, DE). CG-712 cells containing the pGroESL plasmid were transformed with pHisT-hEl expression vectors carrying the normal mature El/ cDNA and either normal or mutant His-tagged mature Ela cDNA se- quences (15). Cells were grown at 42°C to an O.D.595 of 0.80 in C- broth minimal media (16) supplemented with 40 ,uM TPP, 50 Hg/ml carbenicillin, and 50 jig/ml chloramphenicol. Cells were pelleted and resuspended in one-fifth original volume of the same media without antibiotics, and allowed to recover with shaking for 5 min at 37°C. Cells were subsequently induced with 2 mM isopropyl thiogalactoside (IPTG) for 5 min, pulsed with 50 pCi/ml of 35S-Cys/35S-Met (ICN Radiochemi- cals, Costa Mesa, CA) for 1 min, and chased with three volumes of the same media (without antibiotics) supplemented with 8 mg/mi each of nonradioactive L-cysteine and L-methionine. At specified time points following the chase, cell samples ( 1.8 ml) were taken and quickly frozen in liquid N2. Thawed samples were lysed by sonication, and supernatants after microcentrifugation were treated batchwise with an excess (15 Ail) of Ni2+-NTA (nitrilotriacetic acid) resin from Qiagen (Chatsworth, CA). The resin was washed batchwise three times (total volume: 2.4 ml) with 15 mM imidazole in 100 mM potassium phosphate (pH 7.5) containing 2 mM MgCl2, 0.1 mM EDTA, 0.1 mM EGTA, 0.2 mM TPP and 2 mM mercaptoethanol. Bound (His)6-tagged Ela and assembled untagged El,3 polypeptides were eluted with 30 1.l of Laemmli SDS- sample buffer (17) containing 50 mM EDTA. Eluted labeled polypep- tides were analyzed by SDS-PAGE and fluorography. Other procedures. The amplification of genomic DNA, the flanking intronic primers, the methods for subcloning and sequencing techniques used in this study were previously described (12, 18). Northern and Western blotting were performed as described elsewhere (8, 18, 19). Assays of decarboxylation rates using a-keto [1-I4C]isovalerate were performed on intact fibroblasts and untransfected and transfected lymphoblasts as previously described ( 10). The activity of recombinant El was assayed radiochemically using a-keto[1-'4C]isovalerate as a substrate and 2,6-dichlorophenolindophenol as an artificial electron ac- ceptor ( 10). Results Clinical phenotype of intermediate MSUD. As described in Methods, the patients studied in the present report are all of Hispanic-Mexican origin. Their clinical phenotype conforms to that accepted as intermediate MSUD. The patients presented 7-22 mo after birth for evaluation of developmental delay and a mild seizure disorder (Table I). The developmental status of the patients showed significant retardation when their achieve- ment level was compared to their age, (Table I, right column). MSUD was not suspected until a metabolic screen revealed marked elevation of the BCAAs and BCKAs and the presence of alloisoleucine. Plasma leucine in the four patients ranged from 1,218-3,347 MM as also shown in Table I. Dietary restric- tion of BCAAs ameliorates further neurological impairment, and allowed developmental progress to occur (See Methods). To confirm that the above intermediate MSUD patients are deficient in the decarboxylation of the a-keto acids, intact cell 956 Chuang et al. Table I. Rate of Decarboxylation of a-Keto Acids by Intact Normal and MSUD Fibroblasts Decarboxylation rate Percent normal Cell line Pyruvate KIV activity with KIV Normal 0.411 0.177 100 Classic MSUD 0.382 0 0 A. G. 0.308 0.007 4.0 J. B. 0.333 0.007 4.0 P. R. 0.489 0.010 5.6 B. S. 0.399 0 0 Intact cells from fibroblast cultures were assayed in triplicate at 2 mM for pyruvate and a-ketoisovalerate (KIV). Specific radioactivities were 1,275 and 750 cpm/nmol for [1-'4C]pyruvate and a-keto[1-_4C]- isovalerate, respectively. For each assay, 1 X 106 cells (0.25 mg of protein) were incubated at 37°C for 10 min with labeled pyruvate and 80 min with a-ketoisovalerate. A. G., J. B., P. R., and B. S. are intermediate MSUD patients. 1 2 3 4 5 6 Ela _ _ Figure 2. Northern blot analysis of poly A+-en- riched RNA from normal and MSUD fibroblasts. The 1.5-kb human Ela and the murine E1,f cDNA's were radiola- beled with [a-32P] using the Megaprime DNA-la- beling system (Amers- ham Corp., Arlington Heights, IL) (19). Each lane contained 5 ig poly(A)+ RNA except lane 1, where 1.5 ,ug was applied. Lane 1, normal subject; Lane 2, A-1. (classic MSUD of Israeli origin); Lane 3, B.S.; Lane 4, P.R.; Lane 5, A.G.; and Lane 6, J.B. assays with cultured fibroblasts were carried out. Table II shows that three of the intermediate patients (A.G., J.B., and P.R.) had 4 to 5.6% of residual activity compared to normal. These activities are in the range reported for other intermediate MSUD patients ( 1, 2). The fourth patient, B.S., had no detectable activ- ity similar to classic MSUD. The decarboxylation rate with pyruvate was normal in all patients and served as a control for cell viability (Table II). The intact cell results establish that the metabolic block in the intermediate MSUD patients is at the decarboxylation of BCKAs. Western and Northern blot analysis of normal and mutant cells. Since the BCKAD complex is a heteromultimeric protein, it was necessary to identify which subunit was affected in these individuals with MSUD. Westem blotting was carried out using lysates prepared from normal and MSUD fibroblasts. Fig. 1 4bb to v#t§O+stt§+ * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...........1wtlf 1 a * * * * * , ~~~~~~~~~~~~~~~~~~~~~~~~. ......... .. .~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... .'",,'" .....:: :.............. :-2..;.': 4- E2 -*- Ela Figure 1. Western blotting of lysates prepared from normal and MSUD fibroblasts. All lysates were applied to a 10% SDS polyacrylamide gel at 200 tg protein per lane. Separated proteins were electrotransferred to polyvinylidene difluoride membranes. The filters were probed with either E2 (upper panel) or Elra (lower panel) antibodies radiolabeled by coupling with 251I-protein A. MSUD patients B.S., A.G., J.B. and P.R. were of the intermediate phenotype. Patients S.B., L.C., H.H., and B.F. exhibited classic MSUD. (lower panel) shows that the ElIa subunit is markedly reduced in cells from three intermediate patients A.G., J.B., and P.R. compared with normal. The Ela subunit is barely detectable in the fourth intermediate patient B.S. and a classic patient L.C. In contrast, the Ela subunit is present in normal amounts in another classic MSUD patient S.B. The E2 subunit (Fig. 1, upperpanel ) is present in normal amounts in all MSUD patients except S.B. The results establish that S.B. is affected at the E2 locus. On the other hand, the decreased abundance of Ela subunit in the five MSUD patients studied (A.G., J.B., P.R., B.S., L.C., H.H. and B.F.) only indicate that either Ela or E1,6 locus is affected. It has been previously shown by Westem blotting that reduced level of either of the two subunits of El (Ela or E1,3) caused by a mutation reduces the level of the other normal subunit (20-22). Northem blotting was performed using poly(A)+ RNA pre- pared from normal, a classic MSUD patient from Israel (A-1) and the four intermediate MSUD subjects (B.S., P.R., A.G. and J.B.) from Mexico. The Ela mRNA is present in normal amounts in the four intermediate MSUD patients (lanes 3-6, Fig. 2, upper panel), and is absent in the classic MSUD patient (lane 2). The E1/6 mRNA is present in all MSUD patients (Fig. 2, lower panel). The reduced signal of El/3 mRNA in normal (lane 1, lower panel) was due to the lower amount of poly (A) + RNA applied (1.5 jug) compared with the other lanes (5 1.tg). The results show that the mutation of the Ela or E1,6 locus did not affect transcription and stability of either mRNA in the intermediate MSUD, and that reduction in the immunoreactive Ela subunit is a posttranscriptional event. Identification of mutations in intermediate MSUD. Pre- viously we identified an 8-bp deletion in exon 7 of the Ela gene of an MSUD patient GM-649, which abolishes a SmaI restriction site ( 18). This change in restriction pattem was used to screen other MSUD patients. Exon 7 was amplified from genomic DNA of 35 known MSUD subjects and digested with SmaI. The digested samples were analyzed on 2% agarose gels as shown in Fig. 3. Undigested amplified exon 7 from a normal subject was 240-bp in size (lane 2). Digestion of the normal amplified genomic DNA with SmaI produced two smaller 169- and 71-bp fragments (lane 3). Digestion of the amplified exon 7 from GM-649, who is heterozygous for 8-bp deletion, resulted Intermediate Maple Syrup Urine Disease 957 1 2 3 4 5 6 7 8 9 A.G. AC G T A C G T A C G T 240 bp 169 bp - 71 bp Figure 3. Restriction enzyme analysis of amplified genomic DNA from normal subject and MSUD patients. Exon 7 of the Ela gene was ampli- fied by PCR and digested with SmaI restriction enzyme. The digest was separated on 2% agarose gel and stained with ethidium bromide. Lane 1, 123-bp ladder markers; Lane 2, undigested DNA; lane 3, normal digested DNA; lane 4, GM-649, who is heterozygous for an 8-bp dele- tion that abolishes the SmaI site in one allele; lanes 5, 6, and 7 are MSUD patients A.G., J.B., and P.R., respectively, who are affected at the SmaI site; lanes 8 and 9 are MSUD patients S.B. (intermediate) and P.B. (classic), respectively, who are not affected at the SmaI site. in three fragments (lane 4). An uncut 240-bp fragment is caused by the absence of the SmaI site secondary to the 8-bp deletion. The other two smaller species are contributed by the other allele, which contains a normal SmaI site in exon 7. Interestingly, amplified DNA from exon 7 of three of the Mexican patients with intermediate MSUD (A.G., J.B. and P.R.) was not cut by SmaI (lanes 5-7). Two other MSUD patients of Mexican ori- gin, B.S. (intermediate) and P.V. (classic), shows a normal restriction pattemn of 169-and 71-bp fragments (lanes 8 and 9), which was also observed in 29 other MSUD patients (data not shown). The absence of the SmaI site in exon 7 in these intermediate patients, A.G., J.B., and P.R., led us to initially speculate that these patients were homozygous for the 8-bp deletion at the Ela locus. To confirm this, amplified exon 7 from these patients was subcloned into Bluescript plasmid and sequenced. To our surprise, a homozygous G to A transition at base 895 of the Ela gene (12), instead of an 8-bp deletion, was found at the SmaI site (Fig. 4). The G to A conversion accounts for the abolishment of the SmaI site, and a Gly-245 to Arg substitution (G245R) in the Ela subunit. Since intermediate patient B.S. showed a normal SmaI di- gestion pattemn, the mutation in this patient does not involve this restriction site. To determine the mutation in B.S., all nine exons of the Ela gene were amplified using flanking intronic primers. Each amplified exon was subjected to analysis by sin- gle-stranded conformational polymorphism (SSCP) (23). As shown in Fig. 5 (the upper bands), amplified exon 9 from B.S. (lane 2) has a distinct shift in mobility, compared to a normal (lane 1) and two other classic MSUD patients B.F. (lane 3) and H.H. (lane 4) that also had normal mobility. Lanes 5 and 6 are two patients (K.U. and B.A.), who are heterozygous for the Y393N mutation in exon 9, which resulted in a doublet in the upper band as indicated by the arrows. The results of SSCP indicated that intermediate patient B.S. is homozygous for a mutation in exon 9. Direct sequencing of amplified exon 9 using 32P-end labeled primers was carried out with a normal and three MSUD patients (B.S., B.F., and K.U.). To facilitate detection of the mutation, the sequencing samples XC) Figure 4. Nucleotide sequencing showing the G to A mutation in exon 7 of the Ela gene of intermediate MSUD patients A.G., J.B., and P.R. PCR amplified exon 7 from the three patients were subcloned into pBS plasmid vector (Stratagene, La Jolla, CA) and sequenced with the anti- sense primer 5'-GTCAGTGCTGTGGGGATGCT-3'. The boxed nucle- otide depicts the homozygous G to A transition in the three intermediate MSUD patients. The base change (underlined) abolishes the SmaI site by altering the recognition sequence from CCCGGG to CCCAGG. were applied to a polyacrylamide gel according to the four nucleotide groups A, C, G and T (Fig. 6). Intermediate patient B.S. shows a distinct homozygous T to G transversion at base 1253 of the Ela gene (12) (see column G and T). The mutation results in a Phe-364 to Cys substitution (F364C) in the Ela subunit. Characterization of the G245R and F364C mutations. The 1 2 3 4 5 6 _. -._. Figure 5. SSCP analysis of amplified exon 9 of the Ela gene. Genomic DNA's were prepared from normal and MSUD fibroblasts. The region encompassing the entire coding region of exon 9 of the Ela gene was am- plified using 32P-dCTP as a labeled nucleotide. The primers used were: 5'- TAGCCTGCCCAC- TGCCCCATGT-3' (in- tron 8) (sense) and 5'- TCTCGGGGTA- CCTGAGGATGG-3' (bases 1408-1388) (antisense). SSCP anal- ysis was carried as de- scribed under Methods. Lane 1, normal; lane 2, B.S. (intermediate MSUD patient); lanes 3 and 4, B.F. and H.H., respectively, who are classic patients; lanes 5 and 6, K.U. and B.A., respectively, who are heterozygous for the Y393N mutation in exon 9. The arrows indicate the doublet in lanes 5 and 6. 958 Chuang et al. J.B. P.R. A C G T 7 F 1-m 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 CZ i_- ~T to G c *.= _. w t........ :.* ..... ; *: :. }:. ... ...... :: . .. -d,"i- .... . .I . . ' ;. .. ::.. . .. *R Figure 6. Direct sequencing of amplified exon 9 DNA fragment. Se- quencing samples of normal and MSUD patients were applied in groups according to the nucleotides A, C, G, and T. Under each nucleotide, lane 1, normal; lane 2, intermediate patient B.S.; lanes 3 and 4, classic MSUD patients B.F. and K.U., respectively. The T to G transversion in B.S. is indicated by an arrow. G245R and F364C mutations were introduced into the normal Ela cDNA by site-directed mutagenesis, and the mutations were confirmed by DNA sequencing. The entire Ela cDNA containing either mutation was inserted into the EBO expression vector for transient expression studies. The plasmids carrying the normal and the mutant Ela cDNAs were transfected into Ela deficient lymphoblasts from a Mennonite classic MSUD patient (LMK). LMK cells are homozygous for the Y393N substitution in the Ela subunit (24). Intact untransfected and transfected lymphoblasts were used to measure the rate of decar- boxylation of a-keto [1- '4C] isovalerate. Fig. 7 shows that un- transfected normal lymphoblasts have substantial (100%) en- zyme activity, (lane 1), and that LMK cells transfected with the EBO vector without insert (lane 2) exhibited trace or no decarboxylase activity. Transfection of LMK cells with vector containing the normal Ela cDNA (lane 3) restored decarboxyl- ation activity to 72% of normal. LMK cells transfected with vector containing the G245R mutation in Ela cDNA produced a slight but reproducible (5%) decarboxylation of the a-keto acid. Transfection of LMK cells with vector containing the Ela cDNA with the F364C substitution was unable to decarboxylate a-ketoisovalerate. The results establish that the G245R and F364C alterations produce defective Ela polypeptides, and are the cause of the intermediate MSUD phenotype. Lysates prepared from untransfected normal lymphoblasts and from transfected LMK lymphoblasts were analyzed by Westem blotting (Fig. 8). Specific antibodies against either Ela or E1lf subunits were used as probes. LMK cells un- transfected or transfected with EBO vectors without inserts showed reduced levels of the Ela subunit, compared to normal 1 2 3 4 5 Figure 7. Decarboxylation of a-keto [1- 4C ] isovalerate by untransfected normal lymphoblasts and transfected Ela-deficient lymphoblasts (LMK). The EBO vector without insert (lane 2) or carrying nonnal (lane 3), G245R (lane 4) or F364C (lane 5) Ela cDNAs was transfected into LMK cells as described previously (8). Untransfected normal and hygromycin-selected transfected LMK cells were analyzed for decarboxylation activity by the intact cell assay using a-keto [1- 14C] isovalerate as substrate (10). The results are expressed as percent of activity in normal untransfected cells (100% or 0.288 nmol C02/ min/mg protein) (lane 1). Each bar represents mean+SEM (n = 6). untransfected cells (Fig. 8, upper panel). LMK cells transfected with vector harboring normal Ela cDNA increased the ElB subunit to a level similar to normal, but with the appearance of a smaller proteolytic fragment (Ela*). (Fig. 8, upper panel). A similar increase in El a subunit was observed in LMK cells transfected with the five mutant ElB cDNAs (Fig. 8, upper panel). G245R and F364C in Ela are the subjects of the current study. N222S and R220W in Ela encoded by exon 6 are muta- tions in the putative TPP-binding pocket. T265R in Ela en- coded by exon 7 is a mutation in the putative subunit interaction region (Chuang, J. L., D. T. Chuang, and R. P. Cox, unpublished results). The LMK cells transfected with vector having a nor- mal, N222S or R220W Ela cDNAs restored the E1f6 subunit to approximately half the normal level (Fig. 8, lower panel). The results indicates that mutant Ela subunits containing the N222S or R220W mutation are able to assemble with endoge- nous E1/6, similar to normal Ela. The mutant El proteins con- taining these mutations have markedly reduced El enzymatic activity (data not shown). LMK cells untransfected or transfected with the EBO vector without inserts showed trace amounts of the E1,6 subunit (Fig. 8, lower panel). Transfection of LMK cells with vector containing the G245R Ela cDNA restored Elf3 to - 20-30% of that transfected with normal Ela cDNA. In LMK cells transfected with the F364C and T265R cDNA, the El/3 subunit remained in trace amounts, indicating the absence of assembly between the mutant Ela and normal E1,6 (Fig. 8, lower panel). Altered kinetics of El assembly caused by G245R and F364C mutations. The reduced ability of transfected G245R Intermediate Maple Syrup Urine Disease 959 .1im.. lig-ii'll ......bkokl.-I -.01. 10. NO 6"'M"ioiiiiii: Transfection host: Y393N A 1os v oN #x{S¢+ kS wEk ,b CI. t -R| 3 1 -__E1clc NORMAL e' "_ _wvw-w -.o l( "%- -w -E {x -, 0 - _ElfA G245R __-El_t --E1[3.. _*I Figure 8. Western blotting of lysates from normal and transfected Ela deficient LMK lymphoblasts. LMK cells were transfected with EBO vector without insert or carrying normal, N222S, R220W, G245R, T265R, or F364C mutant Ela cDNAs (see text for these mutations). The lysates were prepared from untransfected normal and transfected cell and were subjected to SDS-PAGE, with each lane containing 200 ,ig protein. Proteins were electrotransferred to polyvinylidene difluoride membranes. The filters were probed with either Ela (upper panel) or El1,. (lower panel) antibodies radiolabeled by coupling with '25I-pro- tein A. Ela* represents proteolytic products. and F364C Ela to restore El,lI in LMK cells compared with normal Ela, strongly suggests that, in addition to causing en- zyme dysfunction, these mutations impair El assembly. To in- vestigate the kinetics of Ela and E1/ assembly, we have devel- oped a method for pulse-chase labeling of El subunits in an E. coli host. ES ts E. coli cells were doubly transformed with a His- tagged El expression vector carrying either normal or mutant His-tagged Ela cDNA and normal untagged Elp cDNA, as well as a second vector pGroESL that over-expresses bacterial chaperonin proteins GroEL and GroES. The cells were heat- shocked at 42°C for 4 h, followed by induction at 37°C with IPTG for 5 min. The cells were pulsed with [35S]cysteine/ [35S]methionine for 1 min and chased with the corresponding nonradioactive amino acids. Samples were taken at different intervals from 2 min to 120 min, and lysates purified by Ni2+- NTA affinity chromatography. The eluted radioactive peptides were analyzed by SDS-PAGE and autoradiograms obtained by fluorography. Since El/ is untagged, its copurification with His-tagged Ela is an indication of assembly of the two polypep- tides. Fig. 9 A shows that the assembly of radioactively labeled E13 with labeled normal Ela occurs as early as 10 min after the chase, approaching a plateau in 40 min. The level of the normal labeled His-tagged Ela remains relatively constant dur- ing the 2 h chase. The assembly of normal Elp with the labeled G245R Ela appears to be slower than normal Ela. The level of associated E118 is clearly much lower than that observed in the normal Ela, and reached a plateau at 60 min after the chase (Fig. 9 B). In contrast, there was no significant assembly of .-N4 4"N o _m - El (X -El ,B 2 10 20 30 40 60 120 Time, min Figure 9. Pulse-chase labeling autoradiogram showing the kinetics of assembly of normal and mutant His-tagged Ela with untagged El,B. The bacterial expression system containing overexpressed GroEL and GroES and the pulse-chase of El subunits are described under Methods. Cell lysates containing 35S-labeled polypeptides were bound to Ni2+- NTA resin, washed and eluted with an SDS sample buffer (16). The eluted labeled polypeptides were analyzed by SDS-PAGE followed by fluorography. Assembly at different time points (2-120 min) of the untagged El/I with the His-tagged normal or mutant Ela resulted in the copurification of both subunits as shown in the autoradiograms. (A) Normal His-tagged Ela; (B) G245R His-tagged Ela; (C) F364C His- tagged ElIa. labeled E1l6 with His-tagged mutant F364C Ela during the 2 h chase. This was indicated by the absence of the copurified El/I?, while the signal of His-tagged Ela remained relatively constant (Fig. 9 C). These results indicate that both G245R and F364C mutations reduce the rate of assembly of mutant Ela with normal El,B. Reduced or undetectable activity of mutant El proteins. Prolonged growth of E. coli with IPTG induction at 37°C for 15 h resulted in the production of reduced but significant amounts of assembled mutant His-tagged El proteins carrying G245R or F364C mutations in Ela, compared with normal. Both normal and mutant His-tagged El proteins were purified by Ni2" NTA affinity column and eluted with a 15 to 250 mM imidazole gradient (pH 7.5). The final yield was 50 mg/l cul- ture for normal El protein, 8 mg/l culture for G245R mutant El and 6 mg/l culture for F364C mutant El. Activity of normal and mutant El proteins was assayed by a radiochemical method using a-keto [1-I4C] isovalerate as a substrate and 2,6-dichloro- 960 Chuang et al. C F364C phenolindolphenol as an artificial electron acceptor. Recombi- nant normal El protein showed specific activity of 87.3±13.7 nmol C02/min/mg protein (n = 3) and G245R mutant El pro- tein exhibited specific activity of 2.32 ± 0.14 nmol C02/min/ mg protein (n = 3). F364C mutant El protein had no measur- able activity. The specific activity of G245R mutant El protein was 2.7% of normal. Discussion This is the first detailed molecular and biochemical characteriza- tion of intermediate MSUD. Previous studies were predomi- nantly of the classic form of the disease because of its preva- lence among MSUD patients and the severity of its presentation. On the other hand, intermediate MSUD is milder and often later in presentation, and as a result is understudied. It is of interest that BCKAD activity of intermediate patient B.S. is at the level of classic MSUD, but her phenotype is distinctly of the interme- diate type. Other studies have also noted a poor correlation of residual BCKAD complex activity with the clinical phenotype (25). These findings suggest that the individual's genotype may modulate the severity of disease. Alternatively, the enzyme ac- tivity in a patient's hepatocytes may not be accurately reflected in fibroblast or lymphoblast cultures. The BCKAD complex is 30-50% dephosphorylated or active in fibroblasts (10, 26), whereas it is nearly 100% dephosphorylated or active in the liver (27). In addition, activities of the totally dephosphorylated BCKAD complex in several tissues (liver, kidney, and heart) are higher than in fibroblasts (28). The presence of residual activity in various tissues of intermediate MSUD patients may be sufficient under certain physiological conditions to amelio- rate clinical manifestations. For example, in response to changes in dietary protein or increased catabolism (e.g., starvation), levels of hepatic BCKAD subunits increase a few fold (Chin- sky, J. M., unpublished results) (29). Unfortunately, tissues other than fibroblasts and lymphoblasts from these intermediate patients were unavailable for testing. It should be pointed out that patient B.S. had the most severe neurological damage and the highest plasma leucine levels of the four intermediate MSUD patients described. However, she was also diagnosed later than the other three intermediate patients. The results un- derscore the importance of newborn screening and early dietary therapy for intermediate MSUD. It is noteworthy that all of the three intermediate MSUD patients that are homozygous for the G245R mutation are of Hispanic-Mexican origin. Homozygosity in MSUD is relatively rare except for the Y393N Mennonite mutation where the Founder effect is responsible (30). Analysis of 31 non-Mexican patients showed that these subjects did not have the G245R mutation. Although relatedness and consanguinity were denied by their families, two of the intermediate patients, who are homozygous for the G245R mutation, were from the Monterey area of Mexico and the third was from Southern Mexico. It is possible that the three apparently unrelated families inherited this mutation from a common ancestor. More intermediate MSUD families of Hispanic origin need to be studied to deter- mine the prevalence of the G245R allele in the Mexican popula- tion. There is a high frequency of MSUD in Spain ( 1/12,000 to 1/50,000) (31) compared with the estimated worldwide prevalence of 1 / 180,000 (32). The increased incidence of MSUD in Spain suggests a possible ethnic source for these mutations. The fourth Mexican patient is homozygous for the F364C mutation. She is the progeny of a 14-yr-old primigravida, so the significance of this homozygosity is uncertain. Functional analysis of MSUD mutations cannot be carried out by overexpression of the affected subunit in a normal host, where multiple copies of all six distinct polypeptides are pres- ent. In the present study, we used an Ela-deficient lymphoblasts host from a Mennonite patient homozygous for the Y393N mutation in the Ela subunit (24). Transfection with a replica- tion-competent EBO vector carrying the full-length normal EBIa cDNA resulted in overexpression of the Ela subunit. However, only a near normal range of decarboxylation activity is restored in the host cells (Fig. 7). The data suggest that the assembly of recombinant normal Ela with endogenous normal E1,3 sub- units is limiting as will be discussed later. A partial (5%) resto- ration of decarboxylation activity was observed, when the G245R Ela cDNA was transfected into LMK cells. This rate is similar to the residual (4.0-5.6%) activity in cultured fibro- blasts from the intermediate MSUD patients, who are homozy- gous for this mutation. The lack of restoration of decarboxyl- ation activity by transfection with F364C EBIa cDNA also agrees with the absence of decarboxylation measured with cultured fibroblasts from the homozygous patient B.S. The concordance between the transfection results and assays of patient's cultured fibroblasts indicates the fidelity of the EBO expression system. The LMK host cells used in transfection studies contain reduced levels of the Y393N mutant Ela subunit and trace amounts of the normal E 1/3 subunit (8). The homozygous Y393N Ela mutation was shown to impede assembly of that subunit with normal E1/ subunits. Unassembled Ela and in particular El,6 subunits are not stable and are degraded in cells (20, 22). It is perplexing that transfection of LMK cells with normal Ela cDNA restores the El/3 subunit to only 50% of normal level (Fig. 8.). Since the EBO vector overexpresses normal Ela subunits, one would expect that the El/3 level is restored to normal through assembly with recombinant Ela. Competition between the endogenous Y393N and the trans- fected normal Eia subunits for assembly with E1p is unlikely because the host mutant Ela subunit is assembly-incompetent. A plausible explanation is that only a portion of the transfected normal Ela subunit in the LMK cell is properly folded, and capable of assembly with EI/3. Recent studies have firmly estab- lished that molecular chaperones, most of them heat-shock pro- teins (Hsp), play an essential role in the biogenesis of mito- chondrial proteins (33). In mitochondria of transfected LMK cells, chaperone proteins Hsp6O and HsplO may be limiting, resulting in reduced levels of assembly-competent recombinant Ela subunits. The current model for chaperone-mediated mito- chondrial import is that the nascent polypeptides, while still associated with the ribosome, bind to the cytosolic Hsp7O to maintain the unfolded and import-competent state. Once im- ported into the mitochondrial matrix, the unfolded mature poly- peptide initially interacts with mitochondrial Hsp7O, and is then transferred to the Hsp6O/HsplO complex of this organelle. The latter complex provides a scaffold for proper folding of the mature polypeptide to occur through multiple cycles of ATP hydrolysis-dependent reactions (33, 34). The pulse-chase labeling studies in E. coli represents a novel approach to measure the initial rates of Ela and E11, subunit assembly. Development of this method is based on the thesis that the bacterial cells are analogous to the mitochondrial matrix in the folding and assembly of mature mitochondrial polypep- tides. The molecular chaperones that participate in the biogene- Intermediate Maple Syrup Urine Disease 961 sis of mitochondrial proteins are conserved in E. coli (33). We have shown previously that overexpression of the bacterial chaperones GroEL and GroES, which are homologues of Hsp6O and HsplO, respectively, in ESts E. coli (CG-712) increased the specific activity of El by 500-fold (15). Expression and assembly of El in wild-type E. coli is very low in the absence of overexpressed chaperones (9). The overproduction of GroEL and GroES was achieved through cotransformation with the pGroESL plasmid. The latter plasmid encodes these chaperones under the regulation of a heat-shock sensitive element and an IPTG-inducible promoter. The double-transformation overex- pression is adapted in the pulse-chase labeling studies with the pHisT-hEl plasmid that encodes both His-tagged Ela and untagged E16 as the second vector. The bacterial overex- pression system may not completely duplicate the situation in mammalian cells. Nonetheless, it offers a useful method to dis- sect, under efficient ex vivo conditions, the effect of certain natural human mutations on chaperone-mediated folding and assembly of mammalian proteins. Isolation of the His-tagged Ela/untagged E1,6 subunit com- plex by nickel affinity column chromatography represents the association of the pulse-labeled nascent chains (Fig. 9). No attempt was made to distinguish assembly intermediates from the final a2/62 heterotetramers. Using this optimal overex- pression system in E. coli, we were able to demonstrate altered kinetics of El assembly in intermediate MSUD. The slower than normal kinetics observed with the G245R mutation confirm the partial assembly defect deduced from the transfection stud- ies. The absence of significant assembly of the His-tagged mu- tant F364C Ela subunit with the normal E1,6 subunit during the 2 h chase supports the severe assembly incompetence associated with this mutation. The crystal structure of El has not been determined, but the conserved Gly-245 is in the putative subunit interaction region based on sequence alignment of Ela subunits (35). Chou-Fasman prediction (36) indicates that Gly-245 in the Ela subunit occurs at a turn preceding a Pro at position 246. The G245R substitution may alter the conformation of the turn, thereby adversely affecting the binding of the Ela to E1/6 subunit. Phe-364 is situated in the carboxy-terminal region encoded by exon 9 of the Ela gene. This segment is also highly conserved between E1Ca subunits of different a-keto acid dehydrogenase complexes, however, no function has been as- signed to this carboxy-terminal region. We have shown that, in addition to F364C, two other known mutations Y393N (9, 24) and Y368C (18) encoded by exon 9 were also defective in El assembly. These findings strongly suggest that the exon 9 en- coded carboxy-terminal residues are critical for Ela and E1l3 subunit interaction. Finally, sustained growth ( 15 h) of transformed E. coli with IPTG induction results in production of reduced but significant amounts of G245R and F364C mutant El proteins compared with normal. The data indicate that the individual mutant Ela and normal E1,6 subunits are stable in E. coli cotransformed with chaperonins GroEL and GroES, and eventually assemble into mutant heterotetramers. The low or undetectable activity of recombinant G245R and F354C mutant El proteins provides direct evidence that both mutations in Ela disrupt the catalytic function of the assembled heterotetramers, in addition to the slow rates of assembly. The impaired El activity is consistent with the reduced or absent decarboxylation of BCKA observed in patients' fibroblasts (Table II) and in the transfection studies (Fig. 7). Moreover, the deficiency of El activity is apparently responsible for dysfunction of the BCKAD complex in the intermediate MSUD patients. Acknowledaments This work was supported by grants DK-26758 and DK-37373 from the National Institutes of Health, grant 1-1149 from the March of Dimes Birth Defects Foundation, and a grant from the Life and Health Insur- ance Medical Research Fund. J. R. D. is a Medical Scientist Trainee supported by grant 5-P32GM-08014 from the National Institutes of Health and by the Perot Family Foundation. References 1. Chuang, D. T., and V. E. Shih. 1994. Disorders of branched chain amino acid and keto acid metabolism. In The Metabolic and Molecular Basis of Inherited Disease, 7th edition. C. R. Scriver, A. L. Beaudet, W. S. Sly, and D. Valle, editors. McGraw-Hill Inc., New York. 1239-1278. 2. Danner, D. J., and L. J. Elsas. 1989. Disorders of branched-chain amino acid and ketoacid metabolism. In The Metabolic Basis of Inherited Disease, 6th edition. C. R. Scriver, A. L. Beaudet, W. S. Sly, and D. Valle, editors. McGraw- Hill Inc., New York. 671-692. 3. Taylor, J., B. H. Robinson, and W. G. Sherwood. 1978. A defect in branched-chain amino acid metabolism in a patient with congenital lactic acidosis due to dihydrolipoyl dehydrogenase deficiency. Pediatr. Res. 12:60-62. 4. Pettit, F. H., S. J. Yeaman, and L. J. Reed. 1978. Purification and character- ization of branched-chain a-keto acid dehydrogenase complex of bovine kidney. Proc. Natl. Acad. Sci. USA. 75:4881-4885. 5. Yeaman, S. J. 1989. The 2-oxo acid dehydrogenase complexes: recent advances. Biochem. J. 247:625-632. 6. Harris, R. A., R. Paxton, S. M. Powell, G. W. Goodwin, M. J. Kuntz, and A. C. Han. 1986. Regulation of branched-chain a-ketoacid dehydrogenase com- plex by covalent modification. Adv. Enzyme Regul. 25:219-237. 7. Lindsay, J. G. 1989. Targeting of 2-oxo acid dehydrogenase complexes to the mitochondrion. Ann. N.Y. Acad. Sci. 573:254-266. 8. Fisher, C. R., J. L. Chuang, R. P. Cox, C. W. Fisher, R. A. Star, and D. T. Chuang. 1991. Maple syrup urine disease in Mennonites. Evidence that the Y393N mutation in Ela impedes assembly of the El component of branched-chain a- keto acid dehydrogenase complex. J. Clin. Invest. 88:1034-1037. 9. Davie, J. R., R. M. Wynn, R. P. Cox, and D. T. Chuang. 1992. Expression and assembly of a functional El component (a2,/2) of mammalian branched- chain a-ketoacid dehydrogenase complex in Escherichia coli. J. Biol. Chem. 267:16601-16606. 10. Chuang, D. T., and R. P. Cox. 1988. Enzyme assays with mutant cell lines of maple syrup urine disease. Methods Enzymol. 166:135-146. 11. Margolskee, R. F., P. Kavathas, and P. Berg. 1988. Epstein-Barr virus shuttle vector for stable episomal replication of cDNA expression libraries in human cells. Mol. Cell. Biol. 8:2837-2847. 12. Chuang, J. L., R. P. Cox, and D. T. Chuang. 1993. Characterization of the promoter-regulatory region and structural organization of the Ela gene (BCKDHA) of human branched-chain a-keto acid dehydrogenase complex. J. Biol. Chem. 268:8309-8316. 13. Wynn, R. M., J. L. Chuang, J. R. Davie, C. W. Fisher, M. A. Hale, R. P. Cox and D. T. Chuang. 1992. Cloning and expression in Escherichia coli of mature E1l# subunit of bovine mitochondrial branched-chain a-keto acid dehydrogenase complex. Mapping of the El/I-binding region on E2. J. Biol. Chem. 267:1881- 1889. 14. Goloubinoff, P., A. A. Gatenby, and G. H. Lorimer. 1989. GroE heat- shock proteins promote assembly of foreign prokaryotic ribulose bisphosphate carboxylase oligomers in Escherichia coli. Nature (Lond.). 337: 44-47. 15. Wynn, R. M., J. R. Davie, R. P. Cox, and D. T. Chuang. 1992. Chaperonins GroEL and GroES promote assembly of heterotetramers (aZ62) of mammalian mitochondrial branched-chain a-keto acid decarboxylase in Escherichia coli. J. Biol. Chem. 267:12400-12403. 16. Guzman-Verduzco, L.-M., and Y. M. Kupersztoch. 1987. Fusion of Esche- richia coli heat-stable enterotoxin and heat-labile enterotoxin ,/ subunit. J. Bacte- riol. 169:5201-5208. 17. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond.). 227:680-685. 18. Chuang, J. L., C. R. Fisher, R. P. Cox, and D. T. Chuang. 1994. Molecular basis of maple syrup urine disease: Novel mutations at the Ela locus that impairs El (a2/2) assembly or decrease steady-state Ela mRNA levels of branched-chain a-keto acid dehydrogenase complex. Am. J. Hum. Genet. 55:297-304. 19. Chinsky, J. M., and P. A. Costeas. 1993. Molecular cloning and analysis of the expression of the E1/I subunit of branched chain a-ketoacid dehydrogenase in mice. Biochem. Biophys. Acta. 1216:499-503. 962 Chuang et al. 20. Fisher, C. W., J. L. Chuang, T. A. Griffin, K. S. Lau, R. P. Cox, and D. T. Chuang. 1989. 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. J. Biol. Chem. 264:3448- 3453. 21. Zhang, B., H. J. Edenberg, D. W. Crabb, and R. A. Harris. 1989. Evidence for both a regulatory mutation and a structural mutation in a family with maple syrup urine disease. J. Clin. Invest. 83:1425-1429. 22. Nobukuni, Y., H. Mitsubuchi, I. Akaboshi, Y. Indo, F. Indo, A. Yoshioka, and I. Matsuda. 1991. Maple syrup urine disease. Complete defect of the E1f6 subunit of the branched chain a-ketoacid dehydrogenase complex due to a deletion of an l1-bp repeat sequence which encodes a mitochondrial targeting leader peptide in a family with the disease. J. Clin. Invest. 87:1862-1866. 23. Orita, M., L. H. Iwahana, H. Kanazawa, K. Hayashi, and T. Seikiya. 1989. Detection of polymorphisms of human DNA by gel electrophoresis as single- strand conformation polymorphisms. Proc. Natl. Acad. Sci. USA. 86:2766-2770. 24. Fisher, C. R., C. W. Fisher, D. T. Chuang, and R. P. Cox. 1991. Occurrence of a Tyr393-Asn (Y393N) mutation in the Ela gene of the branched-chain a- keto acid dehydrogenase complex in maple syrup urine disease patients from a Mennonite population. Am. J. Hum. Genet. 49:429-434. 25. Gonzales-Rios, M. C., D. T. Chuang, R. P. Cox, K. Schmidt, K. Knopf, and S. A. Packman. 1985. A distinct variant of intermediate maple syrup urine disease. Clin. Genet. 27:153-159. 26. Tashima, K., Y. Kuroda, I. Yokota, E. Naito, M. Ito, T. Watanabe, E. Takeda, and M. Miyao. 1985. Activation of branched-chain a-ketoacid dehydroge- nase complex by a-chloroisocaproate in normal and enzyme-deficient fibroblasts. Clin. Chim. Acta. 147:103-108. 27. Gillim, S. E., R. Paxton, G. A. Cook, and R. A. Harris. 1983. Activity state of the branched chain a-ketoacid dehydrogenase complex in heart, liver and kidney of normal, fasted, diabetic and protein-starved rats. Biochem. Biophys. Res. Commun. 111:74-81. 28. Harper, A. E., R. H. Miller, and K. P. Block. 1984. Branched-chain amino acid metabolism. Ann. Rev. Nutr. 4:409-454. 29. Zhao, Y., K. M. Popov, Y. Shimomura, N. Y. Kedishvili, J. Jaskiewicz, M. J. Kuntz, J. Kain, B. Zhang, and R. A. Harris. 1994. Effect of dietary protein on the liver content and subunit composition of the branched-chain a-ketoacid dehydrogenase complex. Arch. Biochem. Biophys. 308:446-453. 30. Marshall, L., and A. DiGeorge. 1981. Maple syrup urine disease in the old order Mennonites. Am. J. Hum. Genet. Suppl. 33:139A. 31. Fraga, J. M.. and J. R. Alonso-Femandez. 1987. Neonatal screening pro- grammes in Spain: 1982-1986. In Advances in Neonatal Screening. B. L. Therrell, editor. Elsevier Science Publishers, Amsterdam. 487-488. 32. Naylor, E. W. 1980. Newbom screening for maple syrup urine disease (branched-chain ketoaciduria). In Neonatal Screening for Inbom Errors of Metab- olism. H. Bickel, R. Guthrie, and G. Hamersen, editors. Springer-Verlag, Berlin. 19-28. 33. Hartl, F. U., and J. Martin. 1992. Protein folding in the cell: The role of molecular chaperones Hsp7O and Hsp6O. Annu. Rev. Biophys. Biomol. Struct. 21:293-322. 34. Martin, J., M. Mayhew, T. Langer, and F. U. Hartl. 1993. The reaction cycle of GroEL and GroES in chaperonin-assisted protein folding. Nature (Lond.). 366:228-233. 35. Wexler, I. D., S. G. Hemalatha, and M. S. Patel. 1991. Sequence conserva- tion in the a and ,3 subunits of pyruvate dehydrogenase and its similarity to branched-chain a-keto acid dehydrogenase. FEBS (Fed. Eor. Biochem. Soc.) Lett. 282:209-213. 36. Chou, P. Y., and G. D. Fasman. 1978. Prediction of the secondary structure of proteins from their amino acid sequence. Adv. Enzvmol. 47:45-148. Intermediate Maple Syrup Urine Disease 963