Defects in human methionine synthase in cblG patients
Defects in human methionine synthase in cblG patientsSumedha Gulati1, Priscilla Baker2, Yunan N. Li3, Brian Fowler4, Warren Kruger3,*, Lawrence C. Brody2,* and Ruma Banerjee1,*
1Biochemistry Department, University of Nebraska, Lincoln, NE 68588-0664, USA, 2Laboratory of Gene Transfer, National Center for Human Genome Research, NIH, Bethesda, MD 20892-4442, USA, 3Division of Population Science, Fox Chase Cancer Center, 7701 Burholme Ave., Philadelphia, PA, 19046, USA and 4Universitäts-Kinderklinik, Postfach, CH-4005 Basel, Switzerland
Received September 27, 1996;Revised and Accepted October 11, 1996
Inborn errors resulting in isolated functional methionine synthase deficiency fall into two complementation groups, cblG and cblE. Using biochemical approaches we demonstrate that one cblG patient has greatly reduced levels of methionine synthase while in another, the enzyme is specifically impaired in the reductive activation cycle. The biochemical data suggested that low levels of methionine synthase activity in the first patient may result from mutations in the catalytic domains of the enzyme, reduced transcription, or generation of unstable message or protein. Using Northern analysis, we demonstrate that the molecular basis for the biochemical phenotype in this patient is associated with greatly diminished steady-state levels of methionine synthase mRNA. The biochemical data on the second patient cell line implicated mutations specific to reductive activation, a function that is housed in the C-terminal AdoMet-binding domain and the intermediate B12-binding domain, in the highly homologous bacterial enzyme. We have detected two mutations in a compound heterozygous state, one that results in conversion of a conserved proline (1173) to a leucine residue and the other a deletion of an isoleucine residue (881). The crystal structure of the C-terminal domain of the Escherichia coli MS predicts that the Pro to Leu mutation could disrupt activation since it is embedded in a sequence that makes direct contacts with the bound AdoMet. Deletion of isoleucine in the B12-binding domain would result in shortening of a [beta]-sheet. Our data provide the first evidence for mutations in the methionine synthase gene being culpable for the cblG phenotype. In addition, they suggest directly that mutations in methionine synthase can lead to elevated homocysteine, implicated both in neural tube defects and in cardiovascular diseases.
Cobalamin is an organometallic cofactor required by only two known enzymes in mammals (1 ,2 ). Intracellular cobalamin metabolism is complex, compartmentalized, and requires the interplay of a variety of gene products. In the mitochondrion, the processing pathway leads to adenosylcobalamin (AdoCbl), whereas the cytoplasmic route leads to methylcobalamin (MeCbl). In both compartments, the component steps have been characterized largely at the genetic level by complementation analyses (3 ). In cblC, D and F groups both AdoCbl and MeCbl are defective. However, in cblE and cblG, MeCbl is specifically lowered, and this results from an apparent functional deficiency of methionine synthase (4 ,5 ). Both clinically and biochemically, cblG and E patients have similar phenotypes, displaying megaloblastic anemia, homocystinuria, and hypomethioninemia, together indicative of a block in the biosynthetic pathway unique to MeCbl formation. Based on biochemical studies, it has been postulated that cblG patients may have mutations in MS whereas the cblE patients have defects in one of the two redox proteins that activate MS (4 ). However, direct evidence from mutational analysis of the respective genes has been lacking.
The cblG phenotype is inherited as an autosomal recessive disorder (5 ). We have employed a combination of biochemical and molecular approaches to characterize two cblG cell lines. The first, cblG 79/96, is from a patient with severe neurological dysfunction and homocystinuria but no megaloblastic anemia (6 ), while the second, cblG WG1892, is from a patient with mental retardation, macrocytic anemia, and homocystinuria. Both presented with their symptoms within the first four months of life.
Heterogeneity in cblG patients is displayed at the biochemical level (7 ) and has been categorized broadly into two subgroups. In the first subclass, cblGs are indistinguishable from control cell lines in the standard MS activity assays. In the second cblG subclass, the catalytic activity of MS is markedly reduced (Gulati, Rosenblatt and Banerjee, manuscript in preparation). We have measured MS activity in fibroblast cell lines using a variety of reductants and AdoMet, required for reductive methylation of oxidized and inactive enzyme (Scheme 1). When artificial reductants (such as titanium citrate or DTT) employed in the standard assay are replaced by the physiological reducing agent, NADPH, activity of MS is ~50% in cblG WG1892 compared with that in a control cell line. This information led us to suspect that cblG WG1892, which is impaired in reductive activation but not in catalysis, may have mutations in the activation domain located in the carboxy-terminal half of the protein. We have recently cloned, sequenced and mapped the human cDNA encoding MS (8 ). Using SSCP analysis and sequencing, we have detected a single base pair change resulting in a Pro1173 to Leu mutation, and a 3 bp deletion resulting in loss of Ile881 in the activation domain of MS of patient WG1892. In addition, we demonstrate that the low catalytic activity observed in the cblG 79/96 patient is due to greatly reduced steady-state levels of MS mRNA. Our data provide evidence of the MS gene being directly involved in the pathogenesis of the cblG phenotype.Scheme 1. Postulated reaction mechanism of methionine synthase. The different reductants that are employed in the in vitro assay systems are indicated.
Homocystinuria is a common clinical characteristic presented by cblG patients. Severe homocystinuria has been previously shown to result from mutations in both cystathionine [beta]-synthase and methylenetetrahydrofolate reductase, two enzymes involved either directly or peripherally, in homocysteine metabolism (9 ,10 ). Our study presents the first evidence that MS, an enzyme central to this metabolic circuit, is an additional genetic target for defects leading to homocystinuria.
Two cblG cell lines, WG1892 and 79/96, have been characterized biochemically. MS in the cblG WG1892 cell line is competent in catalysis showing an almost two-fold higher specific activity versus a control fibroblast cell line (Table 1 ). However, an impairment in the reductive activation cycle is unmasked when the physiological NADPH-driven reducing system is employed to activate MS, with approximately 50% of the activity seen in the patient versus a control cell line. In contrast to cblG WG1892, cblG 79/96 exhibits less than 1% of the activity seen with control cell lines under all conditions employed, including high concentrations of reducing agents (data not shown).
Steady-state levels of MS in cblG and control cell lines were compared by Northern analysis. As shown in Figure 1 , MS mRNA levels in cblG WG1892 (lane 3) were comparable to, and perhaps even slightly higher than in the controls (lanes 1 and 4), but barely detectable in cblG 79/96 (lane 2) even after prolonged exposure. In contrast, the GAPDH and Huntington's disease gene mRNA levels were comparable in all cell lines, ruling out lack of mRNA integrity as an explanation for the extremely low level of MS-specific RNA in the cblG 79/96 patient cell line.
We have employed two complementary methods to detect mutations in cblG WG1892. In the first approach, SSCP analysis was used to detect single base changes in the MS cDNA. Seventeen overlapping primer pairs spanning the MS ORF were employed covering 100% of the coding sequence. In the second approach, MS cDNA from cblG WG1892 was amplified in two fragments spanning the 5'- and 3'-halves respectively by RT-PCR (see Table 2 and Materials and Methods for primer descriptions). The resulting products were subcloned, and the insert containing the 3'-half of the gene was sequenced.
Several variant conformational polymorphisms were observed in one of the amplicons shown in Figure 2 . One of the control cell lines (lane 2) is homozygous, whereas the other three [representing one control (lane 1) and two cblG patients, cblG WG1892 (lane 3) and cblG WG1765 (lane 4)] are heterozygous. Anomalously migrating bands were excised from the gel and sequenced. Two of the mutations are CpG changes, consistent with methylation-derived transitions. The heterozygosity in the control and one of the cblG cell lines (WG1765), which exhibit identical banding patterns, is due to two silent mutations: 3862C to T and 3778C to A (nucleotide numbering is based on the DNA sequence in reference 8 that contains 286 bp upstream of the start codon). Both these changes occur in the wobble position of the respective codons. The anomalously migrating pair of bands in cblG WG1892 is due to a 3804C to T transition, which results in the codon for Pro1173 being changed to Leu. The complete cDNA has been analyzed by the SSCP method, and only the Pro -> Leu mutation described here has been detected in this patient. A third likely polymorphism was detected by SSCP analysis at the 5' end of the cDNA (data not shown). This resulted from a G -> A transition at nucleotide 468, and resulted in Arg61 changing to Lys.
Pro1173 (in the human sequence) is conserved in the human, C.elegans and E.coli methionine synthases (8 ). Recently, the crystal structure of the AdoMet binding domain of the E.coli MS has been determined (11 ). Two residues, arginine and alanine, that immediately flank the mutated proline make direct contacts to AdoMet. The [beta] carbon of Pro1135 (corresponding to Pro1173 in the human sequence) is in Van der Waal's contact with C2 of the adenine ring of AdoMet (Fig. 5 A). Replacement of a rigid proline by a flexible leucine in this position at the active site is expected to be detrimental.
A number of lines of evidence have pointed to impairments in MS as the basic defect in cblG patients. MS catalyzes a transmethylation reaction in which successive methyl group transfers from CH3-H4folate to cob(I)alamin to homocysteine lead to the products, H4folate and methionine (Scheme 1). Due to the superreactivity of the intermediate, cob(I)alamin, oxidation competes with remethylation, and over time, the enzyme accrues in an oxidized and inactive state. For reentry into the catalytic cycle, a source of reducing equivalents and AdoMet, as a methyl group donor, are required. Under in vitro assay conditions, the requirement for reducing equivalents is met by DTT/OHCbl or titanium citrate. The physiological reducing system can utilize NADPH as the ultimate source of electrons, and with the participation of two redox proteins, can activate MS (13 ). This activity can be exploited under in vitro assay conditions using cell extracts, where MS and the two redox-active proteins are all present (13 ).
We have employed a combination of these assay methods to localize defects that are specific to the reductive activation cycle. MS in one of the cell lines, cblG WG1892, has such a biochemical phenotype, being impaired in reductive activation but robust in catalysis (Table 1 ). These data suggested that, at the molecular level, mutations might be present in the 3'-half of the gene encoding the B12- and AdoMet-binding domains involved in reductive activation. CblG 79/96 on the other hand displays very low levels of MS activity. In this cell line, one or more mutations affecting catalysis, transcription, or stability of either the message or the protein could explain the reduced enzyme activity.
Northern analysis revealed a greatly diminished level of MS mRNA in cblG 79/96 accounting for its biochemical phenotype. In contrast, the level of MS mRNA in cblG WG1892 is comparable to that of control cell lines (Fig. 1 ). In this patient, two mutations were found in trans in the activation domains of the enzyme, consistent with the biochemical data. A single base transition mutation resulting in a Pro1173 to Leu change in the AdoMet-binding domain was detected by SSCP analysis. The second mutation, a 3 bp deletion resulting in loss of Ile881 in the B12-binding domain was detected by nucleotide sequence analysis. The primers (P9-S and P10-A) spanning this region resulted in a 331 nucleotide-long amplicon. It is likely that the relatively long length of the PCR product precluded detection of this mutation by SSCP analysis.
Based on the high degree of homology between the human and E.coli methionine synthases [55% identical (8 )], we predict that the modular organization of the two enzymes will be similar. The E.coli enzyme has been shown to be a three-domain protein, with the substrates, cobalamin, and AdoMet binding to the N-terminal, middle, and C-terminal domains respectively (14 ,15 ). The crystal structures of the cobalamin (12 ) and AdoMet (11 ) binding domains of the bacterial protein have been determined.While the proline residue is conserved in the E.coli,C.elegans and human MS sequences, Ile881 in the human sequence is substituted by valines in the other two.
The homologous proline residue in the bacterial MS is in Van der Waal's contact with the bound AdoMet (11 ). Its role, along with two proximal prolines (two and five residues downstream), is postulated to be to stiffen the loop thereby fixing the orientation of the backbone (11 ). The flanking and conserved arginine and alanine residues make direct docking interactions that have been identified as being important for AdoMet binding. The three dimensional structure of this domain provides a structural rationale for the detrimental effect of this mutation and provocative evidence for the involvement of the AdoMet-binding domain in electron transfer from the proximal redox partner to MS.
Isoleucine at position 881 in the human sequence corresponds to Val855 in the E.coli sequence and resides in the cobalamin-binding domain. The three dimensional structure of this domain (12 ) reveals that Val855 is in the middle of a six residue long [beta]-sheet (II[beta]5), which together with the adjoining [alpha] helix (II[alpha]5) forms one of the walls around the buried dimethylbenzimidazole moiety of the bound cobalamin (Fig. 5 B). Helix, II[alpha]5, then leads into helix II[alpha]6, which connects to the AdoMet binding domain. Deletion of the corresponding Val855 would lead to contraction of the [beta]-sheet potentially creating structural deformations in the protein wall cupping the cofactor tail. This could either impair cofactor binding or affect domain rearrangement that is expected to occur as the protein switches from the catalytic to the activation mode. It is interesting that a mutation in this region that is distal to the corrin ring is associated with a biochemical phenotype in which reductive activation is specifically impaired. It is possible that the [Delta]Ile881 mutation results in inactive enzyme, and that the biochemical phenotype of the cblG WG1892 patient cell line stems entirely from the Pro1173 to Leu mutation. Characterization of the recombinant mutant human enzymes in the future will permit direct evaluation of this issue.
Mutations in methylenetetrahydrofolate reductase and cystathionine [beta]-synthase, both involved in homocysteine metabolism, have been described (9 ,10 ). These mutations result in homocystinuria in homozygous affected patients. Similarly, dysfunction of MS has been implicated in homocystinuria inherited as an autosomal recessive inborn error of metabolism. Our results now furnish direct evidence for the complicity of MS in homocystinuria.
Fibroblast cell line cblG WG1892 was purchased from the Repository for Mutant Human Cells, Montreal Children's Hospital. Control cell lines and cblG 79/76 were available in our laboratories. The control DNAs were a subset of those described previously (16 ). Cultures were maintained in Eagle's minimum essential medium plus nonesssential amino acids (Sigma), and supplemented with 10% (v/v) fetal bovine serum (Calbiochem). Cells were grown to confluency on ten 150 mm petri dishes, and harvested using a cell scraper. The cells were pelleted at 1200 r.p.m. for 15 min, washed 2 times with phosphate buffered saline and frozen at -80oC.
Approximately 108 cells were resuspended in 1 ml of 100 mM phosphate buffer, pH 7.4, containing 0.25 M sucrose. The cell membranes were disrupted by sonication on ice using the miniprobe (power output setting of 4) and the following cycle: 20 s pulse, 1 min pause, repeated five times. The homogenate was centrifuged at 4oC, at 12 000 * g for 60 min. The resulting supernatant was aliquoted and stored at -80oC. The in vitro anaerobic assays were performed essentially as described previously (17 ) at 37oC for 75 min. When the physiological activation system was employed to measure MS activity, 1 mM concentration of NADPH was used.
RNA was isolated from approximately 106 cells from control and patient cell lines using RNAStat-60, following the procedure recommended by the vendor (Tel-Test `B' Inc, Texas). The RNA was quantitated spectrophotometrically, and ~10 [mu]g RNA/lane was loaded on a 1% formaldehyde agarose gel. The RNA was transferred from the gel to a Hybond N+ membrane (Amersham). A 2.2 kb BglI fragment from the coding region of the human MS was end-labeled with [32P] [gamma]-dCTP (3000 Ci/mol) and 3 ml (containing ~106 counts/ml) of Church and Gilbert hybridization solution and used to probe the northern blot. Hybridization was conducted overnight at 65oC, and the membrane was washed three times with 2* SSC/10% sodium dodecyl sulfate at the same temperature. For hybridization to control probes the membrane was later stripped by boiling in 0.1* SSC, 0.1% SDS for 10 min, and reprobed with human GAPDH cDNA and a 670 bp fragment from Huntington's disease gene cDNA.
RNA isolated as described above was employed as template for first strand cDNA synthesis. The reaction mixture containing 1 [mu]g of RNA in DEPC-treated water (total volume 13 [mu]l) and 1 [mu]l of 1 [mu]g/[mu]l of oligo dt (from Invitrogen) was mixed and heated at 65oC for 10 min, followed by incubation on ice for 2 min. Then the following solutions were added: 2 [mu]l of 10* synthesis buffer (Gibco BRL), 2 [mu]l of 0.1 M DTT, and 1 [mu]l of 10 mM dNTP (Amersham). The sample was mixed by vortexing, spun down and incubated at 42oC for 2 min before 1 [mu]l of Reverse Transcriptase (BRL) was added. Incubation was continued at 42oC for 30 min, the sample was centrifuged and placed on ice. RNaseH (1 [mu]l) was added and the mixture was incubated for 10 min at 42oC. The sample was cooled on ice, then stored at -20oC.
First strand material was amplified for SSCP. Seventeen primer pairs were employed to completely span the open reading frame. The sequences of these primers are available from the authors on request. Primers P1-S and P2-A generated an amplicon harboring the C3804 -> T mutation. Radiolabeled dCTP ([alpha]-dCTP, 3000 Ci/mol) was added to the PCR mixture. The following temperature cycling conditions were employed using the touch down program: 95oC for 2 min followed by 94oC for 30 s, annealing/extension at 67oC for 40 s, with the annealing temperature decreasing by 1.1oC/cycle for 12 cycles. The next 35 cycles used the following conditions: 94oC for 30 s, 54oC for 30 s, 72oC for 30 s.
PCR products were diluted ten-fold with loading buffer (95% formamide, 10 mM NaOH, 0.25% bromophenol blue) prior to use. The heat denatured PCR products (95oC for 5 min followed by incubation on ice for 3 min) were separated on a 0.5* MDE (FMC) sequencing gel at ambient temperature. Bands of interest were excised from the sequencing gel and reamplified using the conditions described above. The PCR products were purified using the Wizard PCR purification system (Promega) and subjected to fluorescent nucleotide sequencing on an ABI 377 Automated Sequencer.
RT-PCR was carried out using Advantage Klen Taq DNA Polymerase (Clontech Inc.) from first strand cDNAs of two controls, HS-68 and MRC-5, and two clbG cell lines, WG1892 and 79/96. The amplifications produced ~2 kb DNA fragments that spanned the 5' half and 3' halves of the gene respectively, and overlapped by 265 bp. The primer sets used for these two PCR reactions were P3-S/P4-A and P5-S/P6-A respectively. Since the enzymatic assay data suggested that mutations in cblG WG1892 are likely to be in the C-terminal reductive activation domains, clones containing the 3' half of the gene were further studied. Semi-nested PCR was performed using 1 [mu]l of 100-fold diluted PCR product containing the 3' half of the gene by using 3NGSP1/P6-A [Table 2 and Table in reference (8 ) for primer description] and the amplified DNA fragments were cloned into the T/A cloning vector, pCR2.1 (Invitrogen Inc.). Eight clones were then used as templates for PCR amplification of a 800 bp fragment (using primers P15F/ P6-A, Table 2 ; ref. 8 ) containing the C3804 -> T transition mutation (Pro1173 -> Leu) previously detected by SSCP in a heterozygotic state. To distinguish the known allele from the unknown allele the 800 bp fragments were digested with MspI and separated on a 2% agarose gel. Three clones that did not have the C3804 -> T mutation were completely sequenced on both strands using the automated sequencing facility at Fox Chase Cancer Center.
To confirm Ile881 deletion in genomic DNA and to examine its prevalence in control samples, a mutation specific primer assay system was utilized. Two primers flanking the mutation site, P-11S and P12-S (Table 2 ) were synthesized. A third primer, specific for the mutant allele, P13-A, was also synthesized. This primer is designed to be complementary to the mutant sequence and will contain a mismatched 3' terminus when annealed to wild type DNA. Genomic DNA was PCR amplified in a reaction containing all three primers. Reaction conditions were as previously described except that 5-fold less of the flanking mutation specific primer, P13-A was added. The outside primer pair amplifies a 95 bp fragment in all samples. If the mutant sequence is present an additional fragment of 79 bp is observed. PCR products were resolved on 10% acrylamide minigels in 0.5* TBE buffer. The control cell lines employed in this study have been described previously (16 ).
We are very grateful to Profs. Martha Ludwig and Rowena Matthews at the University of Michigan for sharing the coordinates of the AdoMet binding domain of MS prior to publication. We acknowledge Tzanko Dukov at the University of Nebraska for his help with structure analysis, Christianne Robbins at the NCHGR core facility for performing nucleotide sequencing, Dan Tagle at NCHGR for the Huntington's Disease gene fragment, and Dr E. A. Kvittingen at Rikshospitalet, Oslo, for making fibroblast cell line 79/96 available to Dr Fowler. This work was supported by the National Institutes of Health (DK45776 to R.B), Center for Biotechnology, UNL (to R.B) and the Pew Charitable Trust (to Y.L. and W.K.).
1 Mellman, I. S., Youngdahl-Turner, P., Willard, H. F., and Rosenberg, L. E. (1977) Intracellular binding of radioactive hydroxocobalamin to cobalamin-dependent apoenzyme in rat liver. Proc. Natl. Acad. Sci. USA74, 916-920.MEDLINE Abstract
2 Kolhouse, J. F., and Allen, R. H. (1977) Recognition of two intracellular cobalamin binding proteins and their recognition as methylmalonyl-CoA mutase and methionine synthetase. Proc. Natl. Acad. Sci. USA74, 921-925.MEDLINE Abstract
3 Shevell, M. I., and Rosenblatt, D. S. (1992) The neurology of cobalamin. Can. J. Neurol. Sci.19, 472-486.MEDLINE Abstract
4 Watkins, D., and Rosenblatt, D. S. (1988) Genetic heterogeneity among patients with methylcobalamin deficiency. J. Clin. Invest.81, 1690-1694.MEDLINE Abstract
5 Watkins, D., and Rosenblatt, D. S. (1989) Functional methionine synthase deficiency (cblE and cblG): clinical and biochemical heterogeneity. Am. J. Med. Genet.34, 427-434.MEDLINE Abstract
6 Kvittingen, E. A., Spangen, S., and Fowler, B. (1991) In 29th Society for the Study of Inborn Errors of Metabolism, London, Abstract No. P37 (ISBN 1-870617-03-7).
7 Sillaots, S. L., Hall, C. A., Hurteloup, V., and Rosenblatt, D. S. (1992) Heterogeneity in cblG: differential retention of cobalamin on methionine synthase. Biochem. Med. Metabolic Biol.47, 242-249.
8 Li, Y. N., Gulati, S., Baker, P. J., Brody, L. C., Banerjee, R., and Kruger, W. D. (1996) Cloning, mapping and RNA analysis of the human methionine synthase gene. Hum. Mol. Genet. 5,1851-1858.MEDLINE Abstract
9 Frosst, P., Blom, J. J., Milos, R., Goyette, P., Sheppard, C. A., Matthews, R. G., Boers, G. J. H., den Heijer, M., Kluijtmans, L. A. J., van der Heuvel, L. P., and Rozen, R. (1995) A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase [letter]. Nature Genet.10, 111-113.MEDLINE Abstract
10 Kraus, J. P. (1994) Molecular basis of phenotype expression in homocystinuria. J. Inher. Metab. Dis.17, 383-390.MEDLINE Abstract
11 Dixon, M. M., Huang, S., Matthews, R. G., and Ludwig, M. L. (1996) The activation domain of cobalamin-dependent methionine synthase: a novel fold for AdoMet binding. Structure, in press.MEDLINE Abstract
12 Drennan, C. L., Huang, S., Drummond, J. T., Matthews, R., and Ludwig, M. L. (1994) How a protein binds B12: a 3Å X-ray structure of B12-binding domains of methionine synthase. Science266, 1669-1674.MEDLINE Abstract
13 Chen, Z., and Banerjee, R. (1996) Characterization of the Human and Porcine Methionine Synthases and their Redox Partners. Homocysteine (Rosenberg, I., Ed.), Kluywer, in press, New York.
14 Banerjee, R. V., Johnston, N. L., Sobeski, J. K., Datta, P., and Matthews, R. G. (1989) Cloning and sequence analysis of the Escherichia coli metH gene encoding cobalamin-dependent methionine synthase and isolation of a tryptic fragment containing the cobalamin-binding domain. J. Biol. Chem.264, 13888-13895.MEDLINE Abstract
15 Drummond, J., Huang, S., Blumenthal, R. M., and Matthews, R. G. (1993) Assignment of enzymatic function to specific protein regions of cobalamin-dependent methionine synthase from Escherichia coli. Biochemistry32, 9290-9295.MEDLINE Abstract
16 Struewing, J. P., Abelovich, D., Peretz, T., Avishai, N., Kaback, M. M., Collins, F. S., and Brody, L. C. (1995) The carrier frequency of the BRCA1 185delAG mutation is approximately 1 percent in Ashkenazi Jewish individuals. Nature Genet.11, 198-200.MEDLINE Abstract
17 Banerjee, R., Chen, Z., and Gulati, S. (1997) Methionine synthase from pig liver. Methods Enzymol. in press.&form=6&uid=97394655&Dopt=r">MEDLINE Abstract
NOTE ADDED IN PROOFThe Ile881 deletion mutation in WG1892 has been found independently by Dr R.Rozen, Montreal Children's Hospital (personal communication; Hum. Mol. Genet. 5, 1867-1874).
*To whom correspondence should be addressed
This page is maintained by OUP admin. Last updated Sun Jan 12 13:36:46 GMT 1997. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
