Human Molecular Genetics

Figure 3. Purification of expressed MUT. Aliquots from several purification steps were electrophoresed and stained with Coomassie blue. Lanes 1 and 5: protein standards; lane 2: soluble cell-free extract; lane 3: pooled active fractions eluted from a Cibacron blue column with 1.1 M potassium phosphate; lane 4: pooled active fractions eluted at ~150 kDa from a Superose 12 column. The migration position of authentic MUT is indicated (Mutase).

In the third approach, a fusion protein (TRXMUT), with thioredoxin at the mature MUT protein NH₂-terminus, was produced in pTRXMUT-transformed E.coli GI698 cells. Contrary to expectation, improved solubility of TRXMUT was not observed; at a post-induction temperature of 27^oC, nearly all of the fusion protein was present in the inclusion bodies (not shown). Analogous to mature protein expression, however, at 12^oC, the majority of TRXMUT is soluble (Fig. 1 , lanes 10 and 11) and, surprisingly, it is active (Table 1 ). Therefore, addition of an extraneous 12 kDa polypeptide at the NH₂-terminus has no major effect on the basic structure of the dimeric enzyme. The activity of TRXMUT in the crude extract is comparable with that of the expressed mature protein (Table 1 ). Because the expression level of the fusion protein is several times higher than that of mature protein (Figure 1 , compare lanes 5 and 10), however, the intrinsic specific activity of TRXMUT must be proportionally lower; alternatively, a significant fraction of the soluble TRXMUT could be inactive. Cleavage of TRXMUT by enterokinase at a specific target sequence localized between thioredoxin and the NH₂-terminus of the mature MUT is necessary to yield fully active, mature human MUT (not shown). Cleavage was not very efficient, however, and, when higher proteinase concentrations and longer reaction times were used, degradation of the mature MUT protein was also observed (not shown).

Purification of human MUT expressed in E.coli

To purify mature human MUT expressed in E.coli, a simplified purification scheme was developed, based on that used for purifying the enzyme from human liver (17 ). Ammonium sulfate precipitation (35-75% saturation) of a sonic extract was followed by chromatography on a Cibacron blue matrix. Because MUT is present as apoenzyme in the bacterial cells, it binds tightly to this pseudoaffinity resin and elutes at high ionic strength, with only a few minor contaminants (Fig. 3 , lane 3). Gel filtration chromatography on Superose 12 removed these and yielded pure protein (Fig. 3 , lane 4). AdoCbl was added to form holoenzyme prior to this step to improve the enzyme's stability. The final specific activity, based on several different methods of protein concentration determination, was 23-26 U/mg, somewhat higher than that achieved by purification from human liver (17 ) and lower than that reported after purification from a yeast expression system (18 ).

Table 2 Previous biochemical findings in mut^- fibroblast lines from patients with methylmalonic acidemia, mutations uncovered in these lines and kinetic parameters of the mutant MUT enzymes expressed in E.coli

Yale cell	Previous biochemical data		Mutation		Expressed MUT kinetic data
line number	Km (AdoCbl) (M)	Vmax (% wt)	Nucleotide	Amino acid	Km (AdoCbl) (M)	Vmax (% wt)
Normal (4)	6.25*10-8	100			5.0*10-8	100
378	1.3*10-5	1	G1182A	R369H	1.3*10-5	1
394	1.9*10-5	5	T767A	Y231N	9.0*10-6	18
515	~2*10-6	~5	G357T	G94V	1.4*10-5	0.2
	3.9*10-5	35	G2226T	G717V	2.0*10-5	~60-90
550	9*10-5	43	G2226T	G717V	2.0*10-5	~60-90
589	2*10-6	0.4	A2109G	H678R	1.8*10-6	0.5
601	3.5*10-5	2	G1943A	G623R	4.5*10-5	3

The biochemical data previously obtained on extracts of these cell lines are summarized from refs 6 and 7; V_max values are presented as a percentage of the V_max of the enzyme in normal fibroblast extracts under these conditions (1700 pmol/min/mg of total extract protein). Kinetic parameters for MUT in extracts from four normal cell lines were averaged (7). Mutations in MUT cDNA from the cell lines indicated were determined by direct sequencing after PCR amplification of mRNA, as described. The nucleotide changes found and the resulting amino acid changes are indicated. Mutant MUT enzymes were expressed and partially purified by ion exchange chromatography as described in Materials and Methods. Concentrations of mutant enzymes in the preparations were estimated by comparison of stained bands on SDS-polyacrylamide gels with bands containing known amounts of pure wild-type MUT. Enzyme assays were carried out in duplicate with a range of AdoCbl concentrations spanning the K_m for each mutant enzyme. Double reciprocal plots (1/v versus 1/[AdoCbl]) were used to estimate the kinetic constants; V_max values are presented as a percentage of wild-type V_max under these assay conditions.

Point mutations in MUT cDNA of patients with the mut- form of methylmalonic acidemia

Six point mutations have been uncovered in MUT cDNAs of six patients in the mut^- complementation group by direct sequencing of PCR products (Table 2 ). One patient (Yale cell line 515) was found to be doubly heterozygous for two mutations, with missense changes in both the NH₂-terminal (G94V) and COOH-terminal (G717V) part of the protein. Because these regions were amplified separately, we have assumed that they represent independent alleles, each affected in only one of the two positions, to account for the disease state. The other five patients each appear to be homozygous for single, different missense mutations, at least at the cDNA level, and one of them (cell line 550) has in homozygous state the mutation producing the G717V change, present in heterozygous form in patient 515. These findings are consistent with previously published biochemical data obtained on cell-free extracts of human fibroblasts of these patients (7 ), which showed that each patient's residual enzyme activity was unique with respect to both the K_m for AdoCbl and apparent V_max . Two of the mutations (G623R and G717V) have been reported previously in other patients (12 ,13 ), G717V occurring several times. In contrast to previous reports (19 ), the six mutations are spread throughout the MUT sequence, with only three of them in the COOH-terminal part of protein, previously suggested to be involved in co-factor interaction (10 ) based on homology to the cobalamin-binding domain of methionine synthase (see Discussion). In two cases, conservative substitutions are apparently disease-causing (R369H and H678R). Of the other cases, three involve substitutions for glycine, which has no side chain-twice to branched side chain valine (G94V, G717V), once to basic and bulky side chain arginine (G623R). The fourth is a non-conservative substitution for a residue with an aromatic side chain, Y231N.

Biochemical characterization of individual mutant MUT proteins

Mutant forms of human MUT, carrying the individual putative disease-causing missense mutations, were expressed in E.coli and partially purified by ion exchange chromatography to permit estimation of the kinetic constants of enzymes with low residual activity. (Cibacron blue chromatography was ineffective, probably reflecting the co-factor binding deficiency of these variants.) The kinetic parameters determined for these enzyme forms are summarized in the last two columns of Table 2 . By comparison with the previous biochemical data in the second and third columns (Table 2 ), it can be seen that these results correspond reasonably well to the earlier findings, determined in cell-free extracts of patient fibroblasts (7 ). All of these mutations considerably increase the K_m for AdoCbl (40-900 times), while V_max values vary from 0.2% of wild-type activity to nearly 100%, in the case of G717V. Even for this last protein, MUT activity at physiological AdoCbl concentration is very low, because its affinity for co-factor is 400 times less than that of the normal human MUT.

Several additional points of comparison with the previous fibroblast data should be noted. First, the relative V_max for the protein with the G717V mutation is variable, because it appears to be unstable in the absence of AdoCbl and can be partially inactivated during preparation (not shown). With the most rapid isolation procedures, its activity approaches 100% of wild-type. This correlates well with the data of Willard and Rosenberg that MUT activity in extracts of line 550 was thermolabile, particularly in the absence of added AdoCbl (7 ). Second, the G623R mutation, previously suggested to lead to the mut⁰ phenotype on the basis of a transient transfection assay (13 ), actually produces the mut^- phenotype, both in vivo (cell line 601) and in vitro (Table 2 ). Finally, there is no simple explanation for the lower of the two apparent K_m (AdoCbl) values determined in extracts of the doubly heterozygous cell line 515 (ref. 7 and Table 2 ), at least on the basis of the K_m values determined for the isolated mutations, G94V and G717V, present in this cell line; the apparent residual activity corresponding to this K_m is also unexplained (see Fig. 2A in ref. 7 ). These complex kinetics could reflect interaction between the mutant alleles in a heterodimeric enzyme molecule, leading to interallelic complementation in a doubly heterozygous patient (see below).

DISCUSSION

We have developed two efficient systems for expression of human MUT in E.coli. MUT activity in cell-free extracts from the system producing the mature form of the protein is 0.2-0.3 U/mg, i.e. 10-15 times higher than in human liver homogenate (17 ). After purification, its specific activity was 23-26 U/mg, indicating that the mature protein in the cell-free bacterial extract comprises ~1% of total protein. This compares favorably with the MUT activity achieved in the yeast expression system, which was about three times higher than that in liver in the reports from one group (9 ,20 ) and somewhat higher, but still <0.5% of total protein, in studies by another group (18 ). To achieve this MUT activity in E.coli, an unusually low post-induction temperature (12^oC) is necessary, on the physiological edge of bacterial growth. It is not clear, however, what the reason is for the dramatic shift in the distribution of MUT protein between soluble and insoluble fractions, depending on the post-induction temperature. Both improved spontaneous folding, as a consequence of slower MUT production, or folding assistance by a cold-shock chaperone protein are possible explanations. The GroEL/GroES pair of heat-shock proteins was ineffective in assisting MUT folding in vivo, consistent with results in vitro with purified components (16 ).

The high residual activity of the TRXMUT fusion protein was initially very surprising, because the addition of the 12 kDa thioredoxin protein at the MUT NH₂-terminus might have been expected to compromise protein function. After examining the crystal structure of the highly homologous MUT from P.shermanii (4 ), however, this observation seems more explicable. While the penultimate NH₂-terminal domain appears to have an important role in subunit-subunit interaction, the immediate NH₂-terminus is free on the surface of the molecule and might be extended without much steric interference with the rest of MUT. Even with this fusion protein, the potentially rapid and stable folding of the thioredoxin domain was insufficient to drive MUT folding at 27^oC, and low temperature growth after induction was necessary to achieve solubility. Because cleavage to remove the thioredoxin was inefficient and resulted in MUT cleavage as well, the fusion protein was not characterized further.

Because the active [alpha]-subunit of P.shermanii MUT is 60% identical (74% similar) at the amino acid level to human MUT and aligns with only three minor gaps (one, two and five amino acids), mapping the mut^- mutations onto the crystal structure of the bacterial protein is helpful in accounting for the decreased affinity for co-factor of the mutant enzymes. On the other hand, because the [beta]-subunit of P.shermanii MUT is much less conserved, to the point of lacking both substrate- and AdoCbl-binding sites, any effects of mutations on subunit-subunit interactions may be harder to rationalize structurally. Not surprisingly, none of the six mut^- mutations affects a residue homologous to one of those in direct contact with the cobalamin in the crystal structure of the P.shermanii MUT heterodimer (4 ), although all affect residues that are identical between the two homologs. Substitutions at contact residues, like the previously described G630E (8 ) and G703R (13 ), would be more likely to result in enzymes displaying a mut⁰ phenotype (10 ). Three of the detected mutations (G623R, H678R and G717V) are localized in the COOH-terminal domain responsible for interacting with the lower face of the corrin ring of the cofactor, including the dimethylbenzimidazole side chain. This domain of human MUT is also homologous to the cobalamin-binding fragment of E.coli methionine synthase, whose crystal structure has also been solved (5 ). Gly717 is conserved in P.shermanii MUT and in the methionine synthase fragment and is localized in a turn at the edge of a [beta]-sheet, remote from the binding site of the corrin ring. Interestingly, this is a `pure' cofactor-binding mutation, affecting only the K_m for AdoCbl, and not the V_max. Replacement of glycine, without a side chain, with valine, contributing a branched side chain, would restrict the flexibility of the [alpha]-carbon backbone at this position and might displace various neighboring structural elements. The structural changes clearly affect protein stability as well as affinity for AdoCbl, because this mutant enzyme is uniquely unstable, both in crude extracts and after partial purification.

One of the other two changes in COOH-terminal residues, G623R, has been described previously (13 ) and appears to disrupt the functionally essential `histidine loop' region (4 ,21 ) and, therefore, affects both kinetic parameters, K_m and V_max. Gly623 is in the neighborhood of His627, the residue whose side chain occupies the bottom ligand position of the cobalt in the bound cofactor, and lies between Lys621 and Asp625, the other two members of the hydrogen-bonding triad (Lys621-Asp625-His627) that positions the His side chain (4 ,21 ). Glycine at this position probably contributes, as a result of backbone flexibility, to the conformation adopted by this loop following cofactor binding (10 ,21 ). We found residual activity in expressed MUT with the G623R substitution, despite its earlier description as having a mut⁰ phenotype (13 ). The combination of a high K_m (the highest among this group) with low residual activity probably caused a very weak response to growth in OH-Cbl in the doubly heterozygous cell line where it was initially found and in transiently transfected fibroblasts (13 ), so the activity went unnoticed. The other mutated residue in this region, His678, is not close to the histidine loop in the P.shermanii structure, but lies at the start of a long [alpha]-helix that forms the outside of the dimethylbenzimidazole-binding pocket. The H678R change may subtly affect the overall structure of this domain; alternatively, it might affect the positioning of Leu674, which interacts with Asp625 of the triad to stabilize further the positioning of His627 (21 ).

The other three changes detected in these mut^- cell lines (G94V, Y231N and R369H) are located well outside of the COOH-terminal cobalamin-binding domain of human MUT. Proteins with these mutations have similar K_ms for AdoCbl (0.9-1.4*10^-5 M), but somewhat different V_max values (0.2-18% of wild-type). All three residues are conserved in the [alpha]-subunit of P.shermanii MUT, and Gly94 and Arg369 are even conserved in the less-homologous, inactive [beta]-subunit of the heterodimeric bacterial enzyme. Based on the P.shermanii structure, Tyr231 is at the end of an element whose residues have several interactions with side chains of the corrin, but does not itself contact the cofactor. Importantly, it appears to have a stacking interaction with the side chain of the conserved Phe225 and, thus, probably contributes to positioning Val227 and Arg228, both of which bind side chains of the corrin. The Y231N mutation probably disrupts this interaction and destabilizes the positioning of the adjacent residues. Arg369 is not adjacent to the corrin or residues that interact with it, so it is difficult to provide a simple rationalization for the effect of the R369H mutation. The side chain of this residue participates in a hydrogen bond with a residue in an adjacent structural element, so subtle positioning effects on the barrel domain, which binds to the top of the corrin ring, may be responsible.

The G94V mutation, discovered in heterozygous state in Yale cell line 515, is most interesting. This residue lies just beyond the COOH-terminal end of the NH₂-terminal domain of MUT, which forms, at least in the P.shermanii structure, a long, partially helical element laid across the surface of the other subunit. Both subunits in the heterodimer share this feature, and it would appear to contribute significantly to subunit-subunit interactions. In addition, this region contains residues which interact with the substrate, methylmalonyl-CoA. In particular, Tyr96 binds to one of the phosphate residues in the CoA moiety. The precise structural effect of the G94V change on this region is not obvious, except that reduced flexibility at this point could influence the position of the Tyr96 side chain, with obvious consequences for substrate binding, for example. How this accounts for its effects on cofactor binding, however, remains unknown.

MUT activity in extracts of line 515 showed complex kinetics relative to AdoCbl concentration (ref. 7 and Table 2 ), which were interpreted originally to reflect co-dominant expression of two mutant alleles in a doubly heterozygous cell line (7 ). The kinetic data for the two expressed mutant proteins (Table 2 ) make this simple explanation unlikely, however. The extremely low residual MUT activity of the G94V allele (0.2%) could not have been detected in fibroblast extracts on the background of the almost fully active G717V allele, which has a similar K_m for AdoCbl. One possible explanation for the appearance of a new kinetic species in a dimeric enzyme such as MUT is the formation of a heterodimer between the two mutant alleles, yielding a protein with enhanced activity, i.e. interallelic complementation. Several reports have documented interallelic complementation between certain mut cell lines upon polyethylene glycol-mediated fusion in culture, in particular between a line with an R93H mutation and lines with any of several changes in the COOH-terminal domain, including G717V (8 ,12 ,13 ). Because one allele in cell line 515 bears the G717V change and the other has a missense change at residue 94, adjacent to the known complementing R93H mutation, it seems possible that the kinetic data on extracts from this line reflect the natural occurrence, in cultured cells and presumably in the patient in vivo, of interallelic complementation, unknown for MUT alleles at the time of the original publication. Therefore, we would suggest that the part of the AdoCbl saturation curve, characterized by a lower K_m value for AdoCbl (~2 *10^-6 M) and significant residual activity (~5%), in extracts of the 515 cell line (7 ) was due to the contribution of the heteroallelic enzyme dimer, with kinetic constants different from those of either of the homoallelic mutant proteins.

The in vivo consequences of such interallelic complementation are difficult to predict, although a milder course/later onset of disease or increased responsiveness to cobalamin supplementation might have been anticipated. The limited clinical data on this patient (22 ) indicated onset in the newborn period (8 days), although she recovered with vigorous intervention. At the time of the case report (~18 months of age), she was developing normally on a low protein diet, although below the 10th percentile in height and weight, and had experienced only one additional episode of metabolic decompensation. She continued to excrete abnormally large amounts of methylmalonic acid and was reported to be unresponsive clinically and biochemically to pharmacologic doses of cobalamin. This presentation is typical for patients of this class; thus, there is no evidence in this patient for a unique phenotypic effect of interallelic complemetation.

Although complementation studies to test this hypothesis in cultured cells are not possible at this time, as no cell line homozygous for G94V is available, it might be feasible to carry out analogous studies in E.coli by co-expressing the mutant alleles. The need for precise control of the relative expression of the alleles would complicate this system, however, suggesting that it might be more informative to study the problem in vitro with purified mutant proteins. The interaction of G94V MUT with both G717V and other mutants with defects in the COOH-terminal domain could validate the hypothesis of interallelic complementation and establish the mechanism by which it occurs in this enzyme.

Table 3 Sequences of oligonucleotides used for amplification and introduction of mutations

Oligo name	Seq. pos.	Sequence
mut-1A (for)	5-21	5'-TCTCCCACAGCGGAGTC
mut-1B (rev)	1202-1183	5'-CCATTGCTTCTATTGCAGTA
mut-2A (for)	1122-1140	5'-CACACTGTCAGACATCTGG
mut-2B (rev)	2492-2474	5'-CTTATAGCATGACACCAGG
mut-ex2 (for)	176-226	5'-CATGCACCAGCAACAGCCCCTTCACCCAGAATGGGCTGCCCT GGCTAAAAAGCAG<
mut-ex3 (for)	173-226	5'-GTACTACACCAGCAACAGCCCCTTCACCCAGAATGGGCTGCC
mut-rev (rev)	226-176	5'-CTGCTTTTTAGCCAGGGCAGCCCATTCTGGGTGAAGGGGCTG TTGCTGGTG
m357for (for)	354-374	5'-GTGTACCATATCCTACCATGT
m357rev (rev)	376-354	5'-ATACATGGTAGGATATGGTACAC

for = forward (sense); rev = reverse. Oligonucleotides mut-1A, mut-1B, mut-2A and mut-2B were used for amplification of cDNAs prior to sequencing; the others were used in expression constructs as outlined in Materials and Methods. The sequence positions (Seq. pos.) are numbered according to the human MUT cDNA sequence in GenBank, accession No. M65131.

MATERIALS AND METHODS

Preparation of pmMUTand pTRXMUT constructs

Full-length human MUT cDNA, cloned in the pGEM-7Zf(+) vector, was used as the source of the normal human MUT cDNA sequence (23 ). Two constructs for the expression of human MUT cDNA in E.coli were prepared. One (pmMUT),generating the mature form of the enzyme with Met instead of Leu at the NH₂-terminus, was prepared by a three-part ligation: the 2.15 kbp PvuII-BclI fragment of human MUT cDNA; a synthetic fragment, prepared by hybridization of mut-ex2 and mut-rev oligonucleotides (Table 3 ), encoding the NH₂-terminal part of the protein; and the BamHI-NcoI backbone of the expression vector, pET11d (New England Biolabs). The other construct (pTRXMUT),producing a fusion protein with thioredoxin at the mature protein NH₂-terminus of MUT, was prepared by a similar three-part ligation: the PvuII-BclI fragment of human MUT cDNA; a synthetic fragment prepared by hybridization of mut-ex3 and mut-rev oligonucleotides (Table 3 ); and the Asp718-BamHI backbone of the expression vector pTrxFus (ThioFusion Expression system, Invitrogen).

Expression of human MUT in E.coli and enzyme assay

Escherichia coli BL21(DE3) strain, transformed by pmMUT, was grown on LB medium with ampicillin (100 [mu]g/ml) at 37^oC to OD = 0.7 at 550 nm, induced with 0.4 or 1.0 mM IPTG and cultivated for 1 h at 37^oC or for 24-26 h at 12^oC. The E.coli GI698 strain, transformed by pTRXMUT, was grown on RM medium with ampicillin (100 [mu]g/ml) at 27^oC to OD = 0.6, induced with 0.1 mg/ml tryptophan and cultivated at 12^oC for 22 h.

Samples of each culture were centrifuged briefly to recover the bacteria, sonically disrupted in 50 mM Tris/phosphate, pH 7.4, 1 mM EDTA, and centrifuged 5 min in a microfuge to separate soluble and insoluble fractions. Aliquots of each were electrophoresed on SDS-10% polyacrylamide gels, and the proteins were either stained with Coomassie blue or transferred electrophoretically to a PVDF membrane (Millipore). The membrane was blocked according to the manufacturer's protocol, incubated with rabbit anti-MUT (human) antiserum (1/3000), then developed with the ECL Kit (Amersham). Rat liver mitochondrial matrix served as a source of native MUT (17 ).

MUT activity was determined by the permanganate oxidation assay, under conditions described previously (17 ). The AdoCbl concentration in the reaction mixture was 0.04 mM, except when varied in the kinetics experiments. One unit (U) of activity is defined as that amount producing 1 [mu]mol of succinyl-CoA/min at 37^oC. Kinetic parameters were calculated from double-reciprocal plots, with V_max values relative to wild-type based on amounts of MUT protein detected on Coomassie-stained SDS-polyacrylamide gels of partially purified proteins.

Purification of human MUT expressed in E.coli

Harvested bacterial cells were resuspended in 50 mM Tris/phosphate, pH 7.4, 1 mM EDTA, disrupted by sonication, and centrifuged for 15 min at 15 000 g. The cell-free extract was fractionated by ammonium sulfate precipitation, and the 35-75% fraction was dialyzed against 20 mM potassium phosphate, pH 7.4, 1 mM EDTA. Normal human MUT was purified from this fraction by affinity chromatography on a Cibacron blue column (Econo-Pac Blue Cartridge, Bio-Rad). Most of the bacterial proteins did not bind to the column or were eluted by 0.3 and 0.4 M potassium phosphate, pH 7.4, and active human mutase was recovered by increasing the potassium phosphate concentration to 1.1 M. This fraction was purified further by gel filtration on a Sepharose 12 column (Pharmacia) in 50 mM potassium phosphate, pH 7.4, 1 mM EDTA.

Affinity chromatography on Cibacron blue is not an effective purification procedure for the mutant forms of human MUT studied here, because their affinity for the pseudoaffinity ligand on the resin is apparently considerably decreased, similarly to their reduced affinity for cobalamin. Ion exchange chromatography on a HiTrap Q column (Pharmacia) was used for partial purification of these mutant MUT enzymes for kinetic experiments. Cell-free extracts were fractionated with ammonium sulfate as above, the dialyzed 35-75% fraction was applied to the column in 50 mM Tris/phosphate, pH 7.4, 1 mM EDTA, and the column was eluted with a KCl gradient in the same buffer. The MUT proteins were eluted at 150 mM KCl. After this step, the mutant proteins were visible as distinct, well-resolved bands on SDS-polyacrylamide gels, allowing quantitative comparison with standards of known protein concentration. The MUT content achieved after the HiTrap Q column was ~5% of total protein, and the protein was sufficiently pure for determination of enzyme kinetics, even in mutant forms with extremely low V_max values (0.1-0.2% of wild-type).

Identification of missense changes and introduction of mutations into the MUT expression vector

mRNA was isolated from fibroblast cell lines using the Micro-FastTrack kit (Invitrogen), and first strand cDNA was synthesized from oligo(dT) or random primers with SuperScript II reverse transcriptase (Gibco-BRL). The MUT sequence was amplified from these templates with the Expand High Fidelity PCR System (Boehringer), using two pairs of specific primers (Table 3 ) to cover the whole human MUT cDNA sequence with a small overlap, and directly sequenced from sets of specific primers (Sequenase PCR Product Sequencing Kit, USB-Amersham). For five of the lines, only a single sequence was detected, with the respective point mutations reported in Table 2 . For both the 5' and 3' partial cDNA amplification products of line 515, however, two bands of about equal intensity were apparent at one position in gels of different sequencing reactions, implying heterozygosity at each position.

Amplified regions with missense changes were introduced into pmMUTby using unique restriction enzyme sites, except in the case of the G357T mutation, producing G94V, where a 22 bp PmlI-AccI fragment was replaced by a synthetic fragment prepared by annealing the m357for and m357rev oligonucleotides (Table 3 ). All replaced regions were sequenced to prove identity with the consensus MUT sequence, except for the introduced mutations.

ACKNOWLEDGEMENTS

The authors thank Adelle M. Hack for cell culture assistance and Philip R. Evans, MRC Laboratory of Molecular Biology, Cambridge, for the structural coordinates of P.shermanii MUT, Protein Data Bank ID code 1REQ. This work was supported in part by a grant from the National Institutes of Health (DK12579).

REFERENCES

1 Fenton, W.A. and Rosenberg, L.E. (1995) Disorders of propionate and methylmalonate metabolism. In Scriver, C.R., Beaudet, A.L., Sly, W.M. and Valle, D. (eds), The Metabolic and Molecular Bases of Inherited Disease. McGraw-Hill, New York, pp. 1423-1450.

2 Fenton, W.A., Hack, A.M., Helfgott, D. and Rosenberg, L.E. (1984) Biogenesis of the mitochondrial enzyme methylmalonyl CoA mutase: synthesis and processing of a precursor in a cell-free system and in cultured cells. J. Biol. Chem., 259, 6616-6621. MEDLINE Abstract

3 Marsh, E.N., McKie, N., Davis, N.K. and Leadlay, P.F. (1989) Cloning and structural characterization of the genes coding for adenosylcobalamin-dependent methylmalonyl-CoA mutase from Propionibacterium shermanii. Biochem. J., 260, 345-352.

4 Mancia, F., Keep, N.H., Nakagawa, A., Leadlay, P.F., McSweeney, S., Rasmussen, B., Bösecke, P., Diat, O. and Evans, P.R. (1996) How coenzyme B₁₂ radicals are generated: the crystal structure of methylmalonyl-coenzyme A mutase at 2 Å resolution. Structure, 4, 339-350. MEDLINE Abstract

5 Drennan, C.L., Huang, S., Drummond, J.T., Matthews, R.G. and Ludwig, M.L. (1994) How a protein binds B₁₂: a 3.0 Å X-ray structure of B₁₂ binding domains of methionine synthase. Science, 266, 1669-1674. MEDLINE Abstract

6 Willard, H.F. and Rosenberg, L.E. (1977) Inherited deficiencies of human methylmalonyl CoA mutase activity: reduced affinity of mutant apoenzyme for adenosylcobalamin. Biochem. Biophys. Res. Commun., 78, 927-934. MEDLINE Abstract

7 Willard, H.F. and Rosenberg, L.E. (1980) Inherited methylmalonyl CoA mutase apoenzyme deficiency in human fibroblasts: evidence for allelic heterogeneity, genetic compounds, and codominant expression. J. Clin. Invest., 65, 690-698. MEDLINE Abstract

8 Crane, A.M. and Ledley, F.D. (1994) Clustering of mutations in methylmalonyl CoA mutase associated with mut^- methylmalonic acidemia. Am. J. Hum. Genet., 55, 42-50. MEDLINE Abstract

9 Crane, A.M., Jansen, R., Andrews, E.R. and Ledley, F.D. (1992) Cloning and expression of a mutant methylmalonyl coenzyme A mutase with altered cobalamin affinity that causes mut^- methylmalonic aciduria. J. Clin. Invest., 89, 385-391. MEDLINE Abstract

10 Drennan, C.L., Matthews, R.G., Rosenblatt, D.S., Ledley, F.D., Fenton, W.A. and Ludwig, M.L. (1996) Molecular basis for dysfunction of some mutant forms of methylmalonyl-CoA mutase: deductions from the structure of methionine synthase. Proc. Natl Acad. Sci. USA, 93, 5550-5555. MEDLINE Abstract

11 Ogasawara, M., Matsubara, Y., Mikami, H. and Narisawa, K. (1994) Identification of two novel mutations in the methylmalonyl CoA mutase gene with decreased levels of mutant mRNA in methylmalonic acidemia. Hum. Mol. Genet., 3, 867-872. MEDLINE Abstract

12 Raff, M.L., Crane, A.M., Jansen, R., Ledley, F.D. and Rosenblatt, D.S. (1994) Genetic characterization of a MUT locus mutation discriminating heterogeneity in mut⁰ and mut^- methylmalonic aciduria by intra-allelic complementation. J. Clin. Invest., 87, 203-207.

13 Qureshi, A.A., Crane, A.M., Matiaszuk, N.V., Resvani, I., Ledley, F.D. and Rosenblatt, D.S. (1994) Cloning and expression of mutations demonstrating intragenic complementation in mut⁰ methylmalonic aciduria. J. Clin. Invest., 93, 1812-1819. MEDLINE Abstract

14 Jansen, R. and Ledley, F.D. (1990) Heterozygous mutations at the mut locus in fibroblasts with mut⁰ methylmalonic acidemia identified by polymerase-chain-reaction cDNA cloning. Am. J. Hum. Genet., 47, 808-814.

15 Goloubinoff, P., Gatenby, A.A. and Lorimer, G.H. (1989) GroE heat-shock proteins promote assembly of foreign prokaryotic ribulose bisphosphate carboxylase oligomers in Escherichia coli. Nature, 337, 44-47.

16 Weissman, J.S., Hohl, C.M., Kovalenko, O., Kashi, Y., Chen, S., Braig, K., Saibil, H.R., Fenton, W.A. and Horwich, A.L. (1995) Mechanism of GroEL action: productive release of polypeptide from a sequestered position under GroES. Cell, 83, 577-87. MEDLINE Abstract

17 Fenton, W.A., Hack, A.M., Willard, H.F., Gertler, A. and Rosenberg, L.E. (1982) Purification and properties of methylmalonyl coenzyme A mutase from human liver. Arch. Biochem. Biophys., 214, 815-823. MEDLINE Abstract

18 Taoka, S., Padmakumar, R., Lai, M., Liu, H. and Banerjee, R. (1994) Inhibition of the human methylmalonyl-CoA mutase by various CoA-esters. J. Biol. Chem., 269, 31630-31634. MEDLINE Abstract

19 Ledley, F.D. and Rosenblatt, D.S. (1997) Mutations in mut methylmalonic acidemia: clinical and enzymatic correlations. Hum. Mutat., 9, 1-6. MEDLINE Abstract

20 Andrews, E., Jansen, R., Crane, A.M., Cholin, S., McDonnell, D. and Ledley, F.D. (1993) Expression of recombinant human methylmalonyl-CoA mutase in primary mut fibroblasts and Saccharomyces cerevisiae. Biochem. Med. Metab. Biol., 50, 135-144. MEDLINE Abstract

21 Ludwig, M.L., Drennan, C.L. and Matthews, R.G. (1996) The reactivity of B₁₂ cofactors: the proteins make a difference. Structure, 4, 505-512. MEDLINE Abstract

22 Whelan, D.T., Ryan, E., Spate, M., Morris, M., Hurley, R.M. and Hill, R. (1979) Methylmalonic acidemia: 6 years' clinical experience with two variants unresponsive to vitamin B₁₂ therapy. Can. Med. Assoc. J., 120, 1230-1235. MEDLINE Abstract

23 Jansen, R., Kalousek, F., Fenton, W., Rosenberg, L.E. and Ledley, F.D. (1989) Cloning of full-length methylmalonyl-CoA mutase from a cDNA library using the polymerase chain reaction. Genomics 4, 198-205. MEDLINE Abstract

---

*To whom correspondence should be addressed. Tel: +1 203 785 7095; Fax: +1 203 785 3404; Email: wayne.fenton@yale.edu

⁺Present address: Institute of Microbiology, Academy of Sciences of the Czech Republic, Vídenská 1083, 14220 Prague, Czech Republic

-->

---

This page is maintained by OUP admin. Last updated Wed Aug 13 15:52:16 BST 1997. Part of the OUP Journals World Wide Web service. Copyright Oxford University Press, 1996

CiteULike    Connotea    Del.icio.us    What's this?

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (15) Request Permissions Google Scholar Articles by Janata, J. Articles by Fenton, W. A. Search for Related Content PubMed PubMed Citation Articles by Janata, J. Articles by Fenton, W. A. Social Bookmarking          What's this?