|Home | About | Journals | Submit | Contact Us | Français|
We report five mutations, three of them novel, responsible for maple syrup urine disease in four unrelated Cypriot families. The five children studied are the first cases of classic maple syrup urine disease to be reported among Cypriots. The first novel mutation identified is a single-base deletion in exon 6 of the Elα gene (c.718delG), which leads to a frameshift after Ala240 and to a stop codon 89 residues further downstream. The other two novel mutations identified are in the Elβ subunit: a two-base deletion in exon 6, c.662_663delCC, which leads to a frameshift after Ala221 and creates a stop codon 17 residues further downstream, as well as a splice mutation, IVS3[+3]delA, which results in the skipping of exon 3. The two known mutations identified are in the Elα gene: the G>C transversion at the 3′-splice acceptor site, (IVS5-1G>C), which results in the deletion of the entire exon 6, and the missense mutation in exon 5 (c.632C>T), which corresponds to a p.Thr211Met substitution. The p.Thr211Met substitution is located in a potassium-ion pocket in the E1 component required for stability of the bound cofactor thiamine diphosphate. The mutant E1 protein harboring the p.Thr211Met substitution was shown unable to bind thiamine diphosphate, leading to undetectable E1 activity.
Maple syrup urine disease (MSUD; OMIM 248600) is a genetically heterogeneous metabolic disorder of pan ethnic distribution inherited in an autosomal recessive manner. The worldwide frequency based on data from neonatal screening is approximately 1 in 185,000 newborns (Chuang and Shih, 2001).
MSUD results from a deficiency in one of the three catalytic components of the mitochondrial branched chain α-keto acid dehydrogenase (BCKD) complex: the branched chain α-keto acid decarboxylase or E1 (EC 126.96.36.199.), the dihydrolipoamide branched chain transacylase or E2 (EC 188.8.131.52), and the dihydrolipoyl dehydrogenase or E3 (EC 184.108.40.206.). The E1 component consists of two α and two β subunits encoded by the BCKDHA gene on chromosome 19 and the BCKDHB gene on chromosome 6, respectively. The E2 subunit is encoded by the DBT gene on chromosome 1. The E3 subunit is encoded by the DLD gene on chromosome 7 and is a common component of three mitochondrial multienzyme complexes. As a result, E3-deficient MSUD presents as combined deficiencies of the BCKD, pyruvate dehydrogenase, and α-ketoglutarate dehydrogenase complexes (Munnich et al., 1982). Apart from the three catalytic components, the BCKD complex contains two regulatory enzymes, the BCKD kinase and the BCKD phosphatase (Chuang and Shih, 2001). The E2 subunit forms the 24-meric cubic core of the complex; to this core multiple copies of E1 and E3 as well as the BCKD kinase and the BCKD phosphatase are attached through noncovalent bonds (Pettit et al., 1978; Chuang, 1998; Ævarsson et al., 2000; Chuang et al., 2001; Harris et al., 2001). The metabolic block in the BCKD complex, which occurs in MSUD results in the accumulation of the branched chain amino acids (BCAA) leucine, isoleucine, and valine and the corresponding branched chain α-keto acids. The elevated branched chain α-keto acid concentrations are responsible for severe clinical consequences, including often-fatal ketoacidosis, mental retardation, and neurological impairment (Chuang and Shih, 2001).
There are presently five known MSUD clinical phenotypes based on clinical presentation, tolerance to dietary protein, and response to thiamine and enzyme levels, that is, classic, intermediate, intermittent, thiamine-responsive, and E3-deficient forms (Chuang and Shih, 2001). Classic MSUD has a neonatal onset of encephalopathy and is the most severe and most common form accounting for 75% of MSUD patients. Symptoms include poor feeding, lethargy, seizures, and coma. The majority of untreated classic patients die within the first few months of life from recurrent metabolic crisis and neurologic deterioration. Treatment involves both long-term dietary management and aggressive intervention during acute metabolic decompensation.
The genomic changes that impair BCKD activity can occur in any of the catalytic components of the complex, but both alleles at a single gene locus must harbor mutations (Chuang and Shih, 2001; Henneke et al., 2003; Rodriguez et al., 2006). More than 107 disease-causing mutations in all four molecular phenotypes of MSUD have been identified (HMGD-Cardiff mutation database) of which 36 (34%) are in the E1α subunit, 31 (29%) in the E1β subunit, 26 (24%) in the E2 subunit, and 14 (13%) in the E3 subunit. Many of these mutations have been characterized, for example, the p.Tyr438Asn substitution in the E1α, which is present in Mennonite patients with classic MSUD, was shown to impede E1 assembly resulting in the preferential degradation of E1β (Fisher et al., 1991). The crystallographic studies (Ævarsson et al., 2000) showed that the p.Tyr438Asn mutation disrupts the E1α-E1β subunit interface, preventing the dimerization of heterodimeric intermediates into a functional E1 component.
In the present report, we describe the first five Cypriot patients with classic MSUD. All disease alleles have been characterized. Two known mutations in the Elα subunit, one novel mutation in the Elα subunit, and two novel mutations in the E1β subunit have been identified. The c.632C>T mutation in exon 6 of the E1α subunit has been biochemically characterized.
The present study includes all known cases of MSUD that have been diagnosed in Cyprus, a country with a population of 780,000 and a birth rate of approximately 10,000 per year. The families gave informed consent for this study. All four families are of Greek Cypriot origin with no consanguinity, and unrelated between them. Basic clinical and biochemical data are summarized in Table 1.
Lymphoblastoid and/or fibroblast cell lines were established from the patients or their parents according to standard procedures (Miller and Lipman, 1973).
DNA was extracted from peripheral blood leukocytes or cell cultures using standard procedures (Miller et al., 1988). Total RNA was extracted from cultured lymphoblasts or fibroblasts using the RNeasy kit (Cat. No. 74104) by Qiagen (Chatsworth, CA). To synthesize the first-strand E1α cDNA and E1β cDNA from total RNA, the Omniscript™ reversed transcriptase from Qiagen (Cat. No. 205111) was used. The cDNAs obtained were used as template for PCR amplification using standard protocols and the Taq DNA polymerase (Qiagen).
The procedure for every fragment started with a 3-min denaturation step at 94°C, followed by 36 cycles of 60s denaturation, 30s annealing time at 55°C, and 60s extension at 72°C. After the last cycle, the samples were incubated for 5min at 72°C for final extension. PCR products were sequenced using an ABI Prism™ model 377 automated DNA sequencer (Applied Biosystems, Foster City, CA). The sequences of the primers used to synthesize and sequence the E1α cDNA and E1β cDNA were previously reported (Chuang et al., 2004).
Western blot analysis was performed on homogenates from cultured lymphoblasts using previously published procedures (Chuang and Chuang, 2000). After washing, the blots were incubated for 1h with a secondary antibody, rabbit IgG-horseradish peroxidase conjugated (Biorad Cat. No. 172-1019, Hercules, CA), at a dilution of 1:50,000. The bound antibodies were detected by Pierce SuperSignal Substrate (Cat. No. 34080; Pierce, Rockford, IL).
The Altered Site in vitro mutagenesis system (Promega, Madison, WI) was used to introduce the p.Thr211Met (ACG to ATG) mutation into the cDNA for the mature E1α subunit. Detailed protocols for the mutant vector construction and subsequent mutagenesis were described previously (Chuang et al., 1995). Briefly, oligonucleotides for the desired mutations and the β-lactamase repair primer were annealed to the single-stranded form of pAlter-α or pAlter-βvector. After the second strand synthesis and two rounds of ampicillin selection, clones harboring the correct mutations were isolated for plasmid preparation. DNA segments containing the mutations were used for cassette replacements of the expression vector pHis-TEV-E1 for wild-type E1, which contained a tobacco-etch virus protease cleavage site after the 6xHis tag (Chuang et al., 1995).
Both wild-type and mutant recombinant His6-tagged E1 heterotetramer were efficiently expressed in Escherichia coli strain CG-712 (ESts) by cotransformation of the pGroESL plasmid overproducing chaperonins GroEL and GroES as described previously (Wynn et al., 1992, 1998). Wild-type and mutant His6-tagged E1 heterotetramers were isolated from cell lysates using a Ni2+-NTA-derivatized Sepharose CL-6B column (Qiagen) as also described previously (Chuang et al., 2000). E1 proteins were further purified on a Superdex-200 gel filtration column (2.6×60cm) connected to an FPLC system (Amersham Pharmacia Biotech, now GE Healthcare, Piscataway, NJ). The column buffer consisted of 50mM potassium phosphate (pH 7.5), 250mM KCl, 10% (v/v) glycerol, 5mM dithioerythritol, 1mM benzamidine, and 1mM phenylmethylsulfonyl fluoride. E1 activity during purification was assayed spectrophotometrically (see below). Protein concentrations were determined using the Coomassie Plus protein reagent from Pierce with absorbance read at 595nm. Alternatively, during enzyme purification, protein concentrations were determined by the direct measurement of absorbance at 280nm using a calculated molar extinction coefficient of 1.15cm−1 mg−1 mL for the α2β2 heterotetramer.
The residual activity of the BCKD complex in lymphoblastoid cell lines or fibroblasts from MSUD patients was assayed based on the rate of decarboxylation of α-keto (1-14C) isovalerate (KIV) using intact cells as described previously (Chuang et al., 2000).
The BCKD complex was reconstituted with E1, lipoylated E2 (lip-E2), and E3 at a molar ratio of 12:1:55 as described previously (Chuang et al., 2000). The oxidative decarboxylation of KIV catalyzed by the reconstituted complex was monitored by the increase of absorbance at 340nm.
Thiamine diphosphate (ThDP)–mediated decarboxylation by the isolated E1 component was assayed at 25°C with KIV as substrate in the presence of an artificial electron acceptor 2,6-dichlorophenolindophenol (DCPIP) as described previously (Wynn et al., 1992). The oxidation of enamine–ThDP (RCOH=ThDP) by DCPIP converts enamine–ThDP to a free acid (RCOOH), which promotes the turnover of enamine–ThDP with concomitant release of the product acid from E1. Reduction of DCPIP was monitored by the decrease in absorbance at 600nm.
Steady-state fluorescence quenching upon ThDP binding to wild-type and p.T211M-α mutant E1 was measured using an LS50 B luminescence spectrometer (PerkinElmer Life Sciences, Waltham, MA) in the photon counting mode (Hennig et al., 1966). Fluorescence intensities were recorded at 25°C using a 3-mL quartz cuvette at an excitation wavelength of 290nm and an emission wavelength of 335nm. Slit widths were set at 5nm for excitation and 10nm for emission. A 290nm cut-off emission filter was installed to reduce light scattering effects. Protein concentrations for E1 and ThDP (A235 nm=11,300M−1 cm−1, pH>7.0) were determined spectrophotometrically. The concentration for all protein samples was 0.25μM (as heterotetramers) in 50mM HEPES buffer, pH 7.5, 200mM KCl, and 1mM MgCl2. Fluorescence readings were corrected for dilution and inner filter effects using Equation 1 (Lakowicz et al., 1983).
where Fcorr is the corrected fluorescence intensity value, Fobs is the experimentally measured fluorescence intensity, Vo is the initial volume of the sample, V is the volume after adding ThDP, d is the path length of the cuvette, Aex is the absorption of the sample at the excitation wavelength, and Aem is the absorption of the sample at the emission wavelength. Three different readings were taken and averaged, and the experiment was repeated three times (n=3). The binding data were fitted by nonlinear regression using the program Prism 4 (GraphPad Software, San Diego, CA) according to Equation 2 for a bimolecular reaction (Nemeria et al., 2001).
where ΔF is the corrected fluorescence change, Fo is the fluorescence intensity before the addition of ThDP, ΔFmax is the maximal fluorescence change, Kd is the dissociation constant, and [ThDP] is the concentration of ThDP in the cuvette. The parameters determined by the fitting procedure were ΔFmax and Kd.
In this study, we present the clinical, biochemical, and molecular characteristics of the first five reported cases of MSUD in Cyprus. All five patients, from four unrelated families, were of the classic phenotype with age at presentation ranging from 9 days to 1.5 months. The diagnosis was based on their clinical features, elevated branched chain amino acids/organic acid levels, and/or residual activities of the BCKD complex (Table 1). Patient 1 developed bilateral cataract at the age of 17 months, a finding that has only been reported twice before for MSUD patients (le Roux et al., 2006; Bellini et al., 2007).
Five mutations were found to be responsible for MSUD in the Cypriot families, three of which are novel (Table 2). The first patient was found to be homozygous for a previously described mutation in the E1α subunit (IVS5-1G>C), which results in the deletion of the entire exon 6 in the mRNA. This mutation was first identified in a Muslim family from Jordan (Chuang and Chuang, 2000). Western blot analysis showed the absence of both the E1α and the E1β subunit of the protein (Fig. 1). The absence of the E1α protein is expected. Interestingly, the E1β subunit that does not carry a mutation is also absent, a finding similar to that observed with the p.Tyr438Asn missense mutation in the E1α subunit of the Mennonite classic MSUD patients (Fisher et al., 1989, 1991; Chuang et al., 1994). The absence of the E1β subunit due to mutations in the E1α subunit likely results from the fact that the unassembled E1β subunit is largely insoluble and is degraded in the cell. We have shown that in the chaperonin-assisted E1 assembly pathway, the E1β subunit maintains its solubility through assembly with the E1α subunit as a heterodimeric intermediate, which subsequently dimerizes into an active heterotetramer (Chuang et al., 1999; Wynn et al., 2000). This thesis is corroborated by our previous studies that showed that transfection of the wild-type E1α subunit into the E1α-affected MSUD cells restores the assembly and presence of the normal E1β subunit in cultured cells from Mennonite MSUD patients (Fisher et al., 1991; Koyata et al., 1993). A total of five prenatal diagnoses were performed for this family (family I), and all five fetuses were found to be affected (four of the five prenatal diagnoses were performed both by DNA analysis and by measurement of the activity of the BCKD complex activity in chorionic villi).
Since no biological samples were available from the MSUD patients of the second family, we performed mutational analysis on the parents. The father was found to carry a novel mutation in E1β subunit, IVS3[+3]delA, which affects the splicing mechanism. As a result, the entire exon 3 is skipped in the mRNA (Fig. 2). Sequencing of the E1α cDNA and the E1β cDNA in the mother showed no mutations. After the identification of the mutation in the E1β subunit in the father, exons of the E1β subunit from the mother were sequenced using genomic DNA. A novel frameshift mutation, c.662_663delCC, was identified in exon 6. This mutation leads to a frameshift after Ala221 and creates a stop codon 17 residues further downstream of the sequence (Table 2). The reason that this mutation was not identified in the cDNA was probably due to the production of an unstable mutant mRNA, which could not possibly be amplified by RT-PCR. Since the mother was heterozygous for this specific mutation, one was only able to sequence the cDNA from the normal allele. Extensive carrier screening in this family identified nine carriers, three with the E1β c.662_663delCC mutation and six with the E1β IVS3[+3]delA mutation.
Analysis of the cDNA of the proband of the third MSUD family revealed a missense mutation in exon 5 of the E1α subunit, replacing a threonine by a methionine residue at position 211 of the precursor sequence. The p.Thr211Met mutation (Thr166Met according to the mature sequence) was reported earlier by us in a homozygous Thai classic MSUD patient (Ævarsson et al., 2000; Chuang and Shih, 2001). The recombinant mutant E1 protein carrying the p.Thr211Met substitution shows no detectable activity of E1-catalyzed decarboxylation and no overall activity of the BCKD complex when reconstituted with this E1 variant (Table 3). The results establish that the p.Thr211Met missense mutation is a disease-causing mutation for classic MSUD.
Amino acid sequence analysis using the Ensemble database showed that p.Thr211 is invariant across species, which suggests that a substitution by methionine is likely to have profound effects on the structure of the protein. The conserved p.Thr211 is one of the amino acid ligands (others are p.Ser206, p.Ser207, p.Leu209, and p.Gln212) in the potassium-binding pocket of the E1α subunit, which maintains a loop conformation through interactions between main chain oxygen atoms and the potassium ion (Fig. 3A). This loop conformation carries Leu209, an essential hydrophobic determinant that stabilizes the aminopyrimidine ring of the bound cofactor ThDP. The introduction of a bulky methionine residue likely impedes the access of the potassium ion to the binding pocket. This could also lead to a collapse of the potassium-binding pocket and the resultant absence of bound ThDP in the p.Thr211Met E1 variant. This hypothesis was corroborated by ThDP-binding studies based on tryptophan fluorescence quenching, which shows nonmeasurable binding of the ThDP cofactor by the mutant protein (Fig. 3B). Under the same experimental conditions, wild-type E1 shows a hyperbolic-responsive curve for fluorescence quenching as a function of increasing ThDP concentrations, with a dissociation constant (Kd) of 1.3μM for ThDP. Thus, the inability to bind the cofactor essential for the decarboxylation reaction explains the absence of overall activity of the BCKD complex caused by the p.Thr211Met mutation. These results provide the biochemical mechanism underlying the classic MSUD phenotype caused by the p.Thr211Met mutation in the E1α subunit. Moreover, studies of this missense mutation substantiate the critical role that the residue p.Thr211 of the E1α subunit plays in fostering cofactor ThDP binding by the E1 component of the BCKD complex.
The proband of family III appeared to be homozygous for the p.Thr211Met mutation. However, when genomic DNA from this patient was examined, the mutation was found to be present in the heterozygous state. Sequencing of the genomic DNA from the parents showed the mother to be a carrier for this mutation but the father to be normal. This result suggested that the father's mutation did not allow the production of a stable mutant mRNA, and therefore the paternal mutation could not be detected in the proband's cDNA, making the p.Thr211Met mutation appear homozygous. Subsequent sequencing of the genomic DNA of the proband revealed a novel single base deletion in exon 6, c.718delG, which leads to a frameshift after Ala240 and creates a stop codon further downstream of the amino acid sequence (Table 2). The presence of the c.718delG mutation was confirmed in the father's E1β genomic DNA.
The proband of the fourth family was found to be a compound heterozygote for the IVS5-1G>C mutation found in family I and the novel c.718delG mutation found in family III (Table 2). It is interesting that five different mutations were found in the four families without evidence for a founder effect. The most frequent mutation, the IVS5-1G>C (deletion of exon 6), found in three out of eight alleles, might prove to be common in the Middle East since it has previously been reported in a family originating from Jordan (Nemeria et al., 2001 plus personal communication). Population screening for this mutation in Cyprus and other neighboring countries might prove rewarding in this respect.
This work was supported in part by the Ministry of Health of the Republic of Cyprus, a training grant from the Cyprus Human Resources Authority (to T.G.) and Grant DK26758 from the National Institutes of Health (to D.T.C).
The URLs for data presented in this article are as follows:
No competing financial interests exist.