- Email: [email protected]
Cloning, expression, and purification of 5,10-methenyltetrahydrofolate synthetase from Mus musculus
Protein Expression and PuriWcation 35 (2004) 276–283 www.elsevier.com/locate/yprep
Cloning, expression, and puriWcation of 5,10-methenyltetrahydrofolate synthetase from Mus musculus Montserrat C. Anguera,a Xiaowen Liu,b and Patrick J. Stovera,b,¤ a
Cornell University, Graduate Field of Biochemistry, Molecular and Cellular Biology, Ithaca, NY 14853, USA b Division of Nutritional Sciences, Cornell University, 315 Savage Hall, Ithaca, NY 14853, USA Received 28 October 2003, and in revised form 16 December 2003
Abstract Folate metabolism is necessary for the biosyntheses of purine nucleotides and thymidylate and for the synthesis of S-adenosylmethionine, a cofactor required for cellular methylation reactions and a precursor of spermidine and spermine syntheses. Disruption of folate metabolism is associated with several pathologies and developmental anomalies including cancer and neural tube defects. The enzyme 5,10-methenyltetrahydrofolate synthetase (MTHFS, EC 6.3.3.2) catalyzes the ATP-dependent conversion of 5-formyltetrahydrofolate to 5,10-methenyltetrahydrofolate, and has been shown to aVect intracellular folate concentrations by accelerating folate degradation. Mammalian MTHFS proteins described to date are not stable and no recombinant mammalian MTHFS protein has been successfully expressed in Escherichia coli. The three-dimensional structure of MTHFS has not been solved. The cDNA coding for Mus musculus MTHFS was isolated and expressed in E. coli with a hexa-histidine tag. Milligram quantities of recombinant mouse MTHFS were puriWed using metal aYnity chromatography and the protein was stabilized with Tween 20. Mouse MTHFS has a molecular mass of 23 kDa and is 84% identical in amino acid sequence to the human enzyme. Activity assays conWrmed the functionality of the recombinant protein, with Km D 5 M for (6S)-5-formyltetrahydrofolate and Km D 769 M for Mg–ATP. This is the Wrst example of a mammalian form of MTHFS expressed in E. coli that yielded suYcient quantities of stable puriWed protein to allow for detailed characterization of its three-dimensional structure and kinetic properties. 2004 Elsevier Inc. All rights reserved. Keywords: MTHFS; Folate; Homocysteine
The reduced tetrahydrofolates (THF)1 are cofactors that carry one-carbon units at three oxidation states ranging from formate to methanol [1]. Folate-dependent one-carbon metabolism is necessary for the syntheses of purines (supplies the carbon-2 and carbon-8 of the purine ring) and thymidylate (methylation of deoxyuridylate to thymidylate), and also for remethylation of homocysteine to methionine [2]. These reactions are critical for DNA synthesis and maintenance of genomic integrity [3], and disruption of folate metabolism
¤
Corresponding author. Fax: 1-607-255-9751. E-mail address: [email protected] (P.J. Stover). 1 Abbreviations used: THF, tetrahydrofolate; MTHFS, 5,10-methenyltetrahydrofolate synthetase; MES, 2-(4-morpholino)ethanesulfonic acid; Hepes, 2-[4-(2-hydroxyethyl)-1-piperazine]ethanesulfonic acid; ATP, adenosine triphosphate; IPTG, isopropyl--D-thiogalactopyranoside. 1046-5928/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2004.02.010
increases risk for developmental anomalies, cancer, and cardiovascular disease [4–6]. The biologically active THF derivatives contain a polyglutamate peptide consisting of Wve to eight glutamate residues connected by peptide bonds [2]. Folate polyglutamates function as one-carbon donators and acceptors in reactions known collectively as one-carbon metabolism, which occurs in both the cytoplasm and mitochondria [7]. 5-FormylTHF is the only reduced folate derivate that does not participate directly as a cofactor in one-carbon metabolism, but it can inhibit various folate-dependent enzymes [8–10]. 5-FormylTHF is the most chemically stable form of reduced folate, and it is administered clinically to patients to elevate intracellular folate levels [11]. Because of its stability, it has been suggested that 5formylTHF may function as the primary storage form of folate within the cell [12]. Studies have shown that dormant cells such as Neurospora crassa spores and
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
soybeans have high levels of 5-formylTHF [11,13]. The enzyme 5,10-methenyltetrahydrofolate synthetase (MTHFS; EC 6.3.3.2), which is the only enzyme that utilizes 5-formylTHF as a substrate, catalyzes the following irreversible reaction: 5-formylTHF C Mg–ATP!5, 10-methenylTHF C Mg–ADP C Pi MTHFS activity is found in many organisms such as bacteria, plants, and mammals [14], suggesting that it plays an important role in regulating folate metabolism. Recently, we demonstrated that elevations in MTHFS expression increase rates of folate turnover or degradation [15]. MCF-7 and SH-SY5Y neuroblastoma cells expressing the human MTHFS cDNA exhibit severe folate deWciency, increased rates of folate turnover, and elevated concentrations of folate degradation products in culture medium relative to control cells [15]. These data suggest that MTHFS may have a second activity that catalyzes the degradation of folate cofactors. To determine if MTHFS catalyzes folate degradation in vitro, milligram quantities of stable MTHFS protein are required. EVorts to express mammalian MTHFS proteins in bacteria have not been successful because of low yields and lack of protein stability [16]. In this study, we report the cloning, expression, and puriWcation of recombinant Mus musculus MTHFS. The expression and puriWcation conditions were optimized to yield milligram quantities of pure and functional protein that remained stable after puriWcation. This is the only mammalian form of MTHFS that has been successfully expressed in Escherichia coli and puriWed in milligram quantities, allowing for characterization of the recombinant protein including the determination of its threedimensional structure by NMR.
Materials and methods Materials Restriction enzymes were purchased from New England Biolabs. Mes, Hepes, Tris, and ATP were purchased from Sigma. (6S)-5-FormylTHF and (6R,S)-5formylTHF were a generous gift from Eprova AG. [15N]ammonium chloride and [13C]D-glucose were purchased from Isotech. All other materials were of high quality and obtained from various commercial vendors. cDNA cloning of MTHFS from M. musculus and construction of the MTHFS expression vector The murine MTHFS cDNA was generated by RTPCR using the high Wdelity Pfu polymerase (Stratagene). RNA was isolated from mouse liver (129/Sv strain) using the Total RNA Isolation Kit (Purgene) and con-
277
verted to cDNA using a polyT primer. The cDNA was used as the template to amplify the 30 end of the mouse MTHFS cDNA by PCR using degenerate primers based on human MTHFS sequence 50 GTCCCGCGTACT GAGCCAGAAG 30 (forward primer) and oligo(dT) (reverse primer). The PCR conditions used were 95 °C for 45 s, 42 °C for 1 min, and 72 °C for 1 min, for a total of 30 cycles. The PCR products were sequenced from two independent PCRs and found to be identical. This partial mouse MTHFS sequence was used to determine the sequence of the entire MTHFS open reading frame by 50 RACE following the manufacturer’s instructions (Gibco). Two primers were derived based on the 30 end of the partial mouse MTHFS cDNA sequence, the reverse primer (50 GAGCAGTAAGTGTAGCCGTCG TG 30) and the nested reverse primer (50 ATAAGGC CTCCTCCCGAACAT 30). Once the 50 sequence was determined, the mouse MTHFS open reading frame was ampliWed using the forward primer 50 GCGCCATATG GCAGACAGGGATGCCAC 30 (NdeI site in italics) and the reverse primer 50 GCGCAAGCTTTCAGAAG TCAGGAATTCCCC 30 (HindIII site in italics). Three independent PCR products were sequenced and found to be identical. The products were cloned into the pCR2.1 vector (Invitrogen), creating clone pCRMTHFS, and transformed into E. coli TOP10 cells (Invitrogen). Individual colonies were selected for DNA sequencing analyses to verify both the integrity of the NdeI and HindIII cloning sites, and the MTHFS cDNA sequence. The pCR-MTHFS vector was digested with NdeI and HindIII and ligated into the corresponding sites of the pET28a (+) expression vector (Novagen) inframe with an N-terminal hexa-histidine tag. DNA sequence analysis and amino acid alignment Alignment of the mouse, human, rabbit, Arabidopsis, Caenorhabditis elegans, and Saccharomyces cerevisiae MTHFS amino acid sequences was performed using the Pole Bio-Informatique Lyonnais Network Protein Sequence Analysis server (available at http://npsapbil. ibcp.fr/) using the Clustal W algorithm [17]. DNA sequencing of all constructs was performed at Cornell University BioResource Center. Expression of recombinant mouse MTHFS in E. coli cells Mouse MTHFS cDNA in the pET28a vector was used to transform E. coli BL21 Star (DE3) cells (Invitrogen). A single colony was used to inoculate 5 mL LB containing 50 g/mL kanamycin and the culture was agitated at 225 rpm overnight at 37 °C. The cells were pelleted by centrifugation, then resuspended in LB media, and used to inoculate a 1 L of SuperBroth medium (Bio101). The culture was incubated at 37 °C until OD550 nm D 0.8, then protein expression was induced
278
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
using 1.5mM isopropyl--D-thiogalactopyranoside (IPTG), and the incubation temperature was lowered to 15 °C for 20 h. Cells were harvested by centrifugation at 7000 rpm at 4 °C and cell pellets were stored at ¡80 °C. Expression of 15N and 13C labeled mouse MTHFS protein using M9 minimal media A conXuent 5 mL culture of E. coli BL21 Star cells expressing mouse MTHFS cDNA was used to inoculate a 50 mL starter culture containing LB supplemented with 50 g/mL kanamycin. The 50 mL culture was grown at 37 °C for 3 h, then pelleted by centrifugation, and resuspended in 1 mL of 2£ M9 media. Resuspended cells were used to inoculate 750 mL of M9 media containing 2£ M9 salts, 8 g D-glucose (or [13C]D-glucose), 2 g [15N]ammonium chloride, 5 mM magnesium sulfate, and 50 g/mL kanamycin. Cells were grown at 37 °C to OD600 nm D 1.0, then protein expression was induced using 1.5 mM IPTG, and the incubation temperature was lowered to 15 °C for 20 h. Cells were harvested by centrifugation at 7000 rpm at 4 °C and cell pellets were stored at ¡80 °C. Cell lysis and puriWcation of mouse MTHFS Frozen cell pellets were resuspended on ice in 10 mM Tris, pH 8.0, 100 mM NaCl, and 100 M PMSF, and the cells were lysed by four passages through a French press. Cell debris was pelleted by centrifugation at 15,000 rpm for 15 min at 4 °C. The supernatant was added to a column containing Talon metal aYnity resin (Clontech) equilibrated with wash buVer (10 mM Tris, pH 7.0, 100 mM NaCl, and 12.5 mM imidazole). The column was washed with 5 column volumes of wash buVer. The protein was eluted using 10 mM Tris, pH 7.0, 100 mM NaCl, and 125 mM imidazole. The absorbance at 280 nm was determined for all 10 fractions collected, and fractions containing the eluted protein (fractions 1–8) were pooled together and concentrated to 2–3 mL using a Centricon-8 (Millipore). Cleavage of the fusion protein was initiated by addition of 5 U of biotinylated thrombin (Novagen) to the eluted protein fractions, followed by incubation for 2 h at room temperature. Thrombin was removed using streptavidin–agarose beads and spin columns (Thrombin Cleavage Capture kit, Novagen). The puriWed protein was dialyzed overnight at 4 °C against 2 L of 50 mM Tris, pH 8.0. For 15N and 13C labeled MTHFS, the dialysis buVer was Sorenson’s Phosphate (67 mM), pH 6.0. The purity of the recombinant MTHFS protein was determined from an SDS– PAGE gel and analyzed using ChemiImager 4400 from Alpha Innotech (San Leandro, CA). The concentrated puriWed protein was aliquoted and stored at ¡80 °C. Protein concentrations were determined by the Lowry method.
MTHFS activity assay MTHFS activity was determined by monitoring the increase in absorbance at 355 nm, which is due to the formation of the product 5,10-methenylTHF (360 D 25,100 M¡1 cm¡1). For determination of speciWc activity values during the puriWcation, protein samples were added to a cuvette containing 100 mM Mes, pH 6.0, 100 M (6S)-5-formylTHF, and 1 mM Mg–ATP. Kinetic parameters of both substrates for MTHFS were obtained using saturating concentrations of one substrate with a range of concentrations of the second substrate in 100 mM Mes buVer, pH 6.0. Km values were derived from Lineweaver–Burk plots, and activity assays were performed at 28 °C with readings taken in quadruplicate.
Results and discussion Cloning of the mouse MTHFS open reading frame and sequence comparisons Total RNA from a 129/Sv mouse liver was isolated and converted to cDNA. The 30 end of the mouse MTHFS cDNA sequence was ampliWed by PCR using a high Wdelity thermostable polymerase and degenerate primers based on the human MTHFS cDNA sequence [16]. The 30 sequence derived from the PCR product was used to obtain the murine MTHFS cDNA by 50 RACE. PCR ampliWcation of the MTHFS open reading frame yielded a 612 nucleotide DNA sequence coding for the mouse form of MTHFS, which encoded a 203 amino acid protein with a predicted molecular mass of 23,210 Da. After verifying the nucleotide sequence of three independently derived clones, the mouse MTHFS cDNA was cloned in-frame with an N-terminal hexahistidine tag into a pET28 expression vector. This MTHFS cDNA is nearly identical to a C57/Bl/6 cDNA sequence obtained from embryo tissue [GenBank Accession No. NM026829 (nucleotide) and GenBank Accession No. Q9D110 (protein)], but the two sequences diVer by one nucleotide (T298C) that results in an amino acid substitution (S98P). The human MTHFS protein sequence has a proline residue at position 98 [16], indicating that either the database sequence for C57/Bl/6 mouse MTHFS (GenBank Accession No. NM026829) is incorrect or that the mouse MTHFS protein is polymorphic. To determine the localization of the mouse MTHFS protein and putative substrate binding motifs, the MTHFS amino acid sequence alignment comparisons and cellular localization predictions were obtained using several software programs. Mouse MTHFS was predicted to be cytoplasmic determined by the PSORTII program (http://psort.nibb.ac.jp/). Analysis of MTHFS
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
amino acid sequence from S. cerevisiae using the same program also predicted a cytoplasmic protein [14]. Human and rabbit liver MTHFS activity is found primarily in the cytoplasm [18,19], but one study detected about 15% of total human MTHFS activity in mitochondria from human liver [20]. Mitochondria isolated from rabbit liver do not contain MTHFS activity [19]. Fig. 1 contains the amino acid sequence alignment, using the Cluster W algorithm, for the mouse, human, rabbit, S. cerevisiae, C. elegans, and Arabidopsis forms of MTHFS. The protein sequence of mouse MTHFS is
279
84% identical with the human sequence and 76% identical with the rabbit protein. The mouse MTHFS sequence is 36, 28, and 26% identical compared to the C. elegans, S. cerevisiae, and Arabidopsis sequences, respectively. There is 8.8% sequence identity among all six protein sequences of MTHFS (Fig. 1). Carbodiimideactivated 5-formylTHF(Glu)1 and 5-formylTHF(Glu)5 covalently attach to lysine-68 of rabbit liver MTHFS, indicating that this residue is involved in a salt bridge with the -carboxyl group of the Wrst glutamate residue of the folate substrate [19]. A lysine residue at position
Fig. 1. Alignment of MTHFS protein sequences from mouse, human (GenBank Accession No. AAC41945), rabbit (GenBank Accession No. A53688), S. cerevisiae (GenBank Accession No. P40099), C. elegans (GenBank Accession No. 492692), and Arabidopsis (GenBank Accession No. 568284). The symbol (*) beneath the amino acid residue indicates identity, the symbol (:) indicates a conservative substitution, and the symbol (.) indicates a non-conservative substitution. The sequences were aligned by the Pole Bio-Informatique Lyonnais Network Protein Sequence Analysis server (available at http://npsa-pbil.ibcp.fr/) using the Clustal W algorithm [17]. The mouse, human, and rabbit proteins are 70% identical and have 16% conservative substitutions. All six sequences are 8.8% identical.
280
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
68 is present in MTHFS sequences from mouse, human, rabbit, C.elegans, and S.cerevisiae, and MTHFS from Arabidopsis contains an arginine residue at this position, which could also interact with the glutamate moiety (Fig. 1). The peptide region spanning residues 60–72 of mouse, human, and rabbit liver MTHFS contains alternating Arg and Lys residues that may interact with the negatively charged polyglutamate portion of 5-formylTHF(Glu)n [19]. The human MTHFS enzyme contains a putative folate binding signature sequence (S159-L-L-PX-X-X-W171) [16,21], and these residues are nearly identical in the mouse and rabbit MTHFS proteins (Fig. 1). Analysis of the mouse MTHFS amino acid sequence using the PROSITE database failed to identify any binding motifs, however the MOTIF sequence analysis tool searching the BLOCKS database (http://motif. genome.ad.jp) revealed a dihydropteroate synthase motif in region 186–207, which may correspond to the pterin binding site of the folate substrate of mouse MTHFS. MTHFS is known to bind Mg–ATP [13,19,22,23], and there is a probable ATP-binding site motif at positions 209–217 that is similar to the ATP/GTP-binding site motif (P-loop; [AG]-x-x-x-x-G-K-[ST]). This region contains four glycine residues that would allow Xexibility for the positively charged lysine and arginine residues to interact with the negatively charged phosphate groups of ATP. The last residue in this putative ATP-binding motif is a glycine residue instead of a serine or threonine, which is similar to the ATP-binding sequence for adenylate kinase (ExPASy documentation). The area spanning residues 209–217 displays high sequence identity among the six MTHFS sequences compared in Fig. 1, suggesting that this area is important for binding the ATP substrate. ConWrmation of these possible 5-formylTHF and Mg–ATP-binding motifs requires the three-dimensional structure of mouse MTHFS. Expression and puriWcation of recombinant mouse MTHFS Previous attempts to express the recombinant human MTHFS in E. coli resulted in very low yields of functional protein that required a multi-step puriWcation procedure (unpublished results and [16]). The purpose of this study was to determine if the murine MTHFS protein exhibited greater stability as a recombinant protein than the human protein, and if milligram amounts of protein could be obtained and puriWed with one column. E. coli BL21 Star cells were transformed with the pET28a vector containing the mouse MTHFS cDNA and transformants were screened for optimal protein expression. Initially, 1 L cultures were grown in SuperBroth media, induced with IPTG, and incubated at 25 °C for 5 h. However, only 1–2 mg of puriWed protein was obtained per liter of culture. By lowering the induction temperature to 15 °C and extending the induction time
Fig. 2. SDS–PAGE analyses of recombinant mouse MTHFS protein during puriWcation. Recombinant mouse MTHFS protein, containing an N-terminal hexa-histidine tag, was expressed in E. coli and puriWed using a metal aYnity column. (A) Cells were lysed using a French press, and the soluble and insoluble fractions (25–50 g) were run on a 12% SDS–PAGE gel and stained with Coomassie brilliant blue. Lane 1, Precision Plus Protein Standards (BioRad); lane 2, the uninduced culture; lane 3, the insoluble pellet after cell lysis; and lane 4, the soluble fraction of the cell lysate (supernatant). (B) SDS–PAGE gel (12%) showing the fractions from a typical puriWcation and stained with Coomassie brilliant blue. Lane 1, Precision Plus Protein Standards (BioRad); lane 2, 50 g of soluble supernatant following cell lysis; lane 3, 50 l sample of the column wash; lane 4, 50 l sample of elution fraction 1; lane 5, 20 g sample of elution fraction 2; lane 6, 20 g sample of pooled fractions of eluted protein treatment with thrombin; and lane 7, 10 g sample of puriWed protein after overnight dialysis.
to 20 h, the yield of puriWed protein increased to 8 mg/L of culture. Analysis of the soluble and insoluble fractions of the cell lysate from E. coli expressing mouse MTHFS indicated that the majority of recombinant MTHFS protein was present as inclusion bodies (Fig. 2A). SDS–PAGE analyses of uninduced and induced cell lysates indicated that the recombinant mouse MTHFS protein was primarily present in the insoluble fraction (Fig. 2A). Refolding attempts using sodium deoxycholate and a
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
Protein Refolding Kit (Novagen) were unsuccessful because the protein rapidly denatured during the Wnal dialysis. However, there were suYcient quantities of protein present in the soluble fraction to allow for protein puriWcation and characterization. Cells were lysed using a French press and the soluble MTHFS protein was puriWed using metal aYnity chromatography with Trisbased buVers. The presence of Hepes in the puriWcation buVers resulted in the rapid denaturation of the protein during the Wnal dialysis step. Recombinant mouse MTHFS was puriWed to greater than 98% purity using a single metal aYnity column. Samples from a typical puriWcation of mouse MTHFS were run on a 12% SDS– PAGE gel and are shown in Fig. 2B. MTHFS activity assays of the total protein, total activity, and speciWc activity values for the initial and Wnal steps in a typical puriWcation are shown in Table 1. The recombinant protein was eluted from the column using 125 mM imidazole, and the second elution fraction contained the majority of MTHFS protein (Fig. 2B). The eluted fractions were pooled, and the hexa-histidine tag was removed with thrombin treatment. Complete removal of the hexa-histidine tag occurred without degradation of the puriWed protein (Fig. 2B, lanes 6–7). Removal of the hexa-histidine tag did not aVect the activity nor stability of recombinant mouse MTHFS because the protein remained stable and active after a 22 h dialysis (Table 1 and Fig. 2B). We cannot account for the increased MTHFS activity in the puriWed protein compared to the cell lysate (Table 1), although inhibitory molecules may have been present in the crude cell lysates. Image densitometry analysis of the SDS–PAGE gel in Fig. 2B indicated that the purity of recombinant mouse MTHFS protein was greater than 98% and other puriWcations yielded protein with similar purity. The puriWed protein was concentrated to 20–30 mg/mL without denaturation and the addition of 0.1% Tween 20 stabilized the protein through multiple rounds of freeze–thawing with no changes in speciWc activity (data not shown). The literature reports that MTHFS proteins from rabbit, S. cerevisiae, and Lactobacillus casei are stabilized by the presence of non-ionic detergents because of the hydrophobicity of the MTHFS protein [14,19,24]. The recombinant MTHFS enzyme is not stable after several rounds of freeze–thawing, but it can be stabilized with
281
the presence of either substrate, (6S)-5-formylTHF or Mg–ATP. Expression and puriWcation of mouse MTHFS using minimal media The recombinant MTHFS protein was expressed in minimal media and puriWed using the protocols described previously to determine if NMR could be used to solve the three-dimensional structure of mouse MTHFS protein. A 750 ml culture of E. coli BL21 cells expressing recombinant mouse MTHFS was grown in M9 minimal medium. The culture reached an OD600 nm of 1.0 in approximately 6.5 h, whereas cultures grown in SuperBroth achieved an OD600 nm of 1.0 in approximately 3 h (data not shown). The Wnal yields of puriWed recombinant MTHFS protein expressed in E. coli grown in diVerent culture media are summarized in Table 2. Typical puriWcations of mouse MTHFS from cultures grown using SuperBroth medium yielded between 6 and 8 mg of puriWed protein per liter of culture. Expression of recombinant mouse MTHFS in M9 minimal media containing [15N]ammonium chloride yielded an average of 2 mg of puriWed protein per liter of culture, and media containing both [15N]ammonium chloride and [13C]Dglucose also yielded 2 mg puriWed protein per liter of culture. Therefore, milligram quantities of 15N and 13C labeled recombinant mouse MTHFS protein can be obtained and used for NMR structural determination. Single and double labeled mouse MTHFS protein was Table 2 PuriWed protein yields of recombinant mouse MTHFS grown in diVerent culture media Culture media
PuriWed protein yield (mg puriWed protein/L culture media)
SuperBroth 2£ M9 2£ M9 with [15N]ammonium chloride 2£ M9 with [15N]ammonium chloride and [13C]glucose
7.0 8.3 1.7 2.2
Escherichia coli expressing mouse MTHFS was cultured in diVerent bacterial growth media, then puriWed as described in the Materials and methods section. Values represent the average protein quantities obtained from three individual preparations.
Table 1 Summary of recombinant mouse MTHFS puriWcation Step
Total protein (mg)
Total activity (U) £ 1000
SpeciWc activity (U/g) £ 10,000
PuriWcation fold factor
Cell lysate (supernatant) PuriWed protein (without His tag)
720 5.5
4.1 5.6
0.13 22.2
1.0 171
The values for total protein, total activity, speciWc activity, and the puriWcation fold factor for a typical puriWcation of mouse MTHFS are shown below. Protein concentrations were determined using the Lowry method. The activity values represent the average of triplicate measurements from a representative preparation. Assays were performed at 24 °C using 100 M (6S)-5-formylTHF, 1 mM Mg–ATP, and 100 mM Mes, pH 6.0. One unit (U) corresponds to 1 mol 5,10-methenylTHF/min.
282
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
Table 3 Comparison of kinetic parameters for recombinant mouse MTHFS and other forms of MTHFS Species
(6R,S) 5-CHO-THF Km (M)
Mg–ATP Km (M)
Turnover number (min¡1)
Reference
Mouse Human Rabbit
5.0 4.4 8.0 33
769 20 330 43
980 1000 300 —a
Present report [20] [25] [14]
S. cerevisiae L. casei
0.6
1.0
88
[24]
The Km values of the substrates for MTHFS and the turnover numbers for various forms of MTHFS from diVerent species are shown. The mouse, human, rabbit, and L. casei enzymes were assayed using buVers at pH 6.0 and S. cerevisiae MTHFS was assayed at pH 7.0. a Value not reported.
puriWed using the metal aYnity procedure described previously, and either 5-formylTHF or Mg–ATP was added to the puriWed protein to aVect stability. Biochemical characterization of recombinant mouse MTHFS The molecular mass of puriWed recombinant mouse MTHFS containing a hexa-histidine tag is approximately 25 kDa as determined by SDS–PAGE gels (Fig. 2B, lane 5). Analysis by MALDI-TOF mass spectrometry of recombinant mouse MTHFS (with the hexahisitidine tag) indicated that the molecular mass is 25,221 Da (data not shown). The predicted molecular mass of mouse MTHFS without the hexa-histidine tag is 23,210 Da, and this value agrees with values obtained by SDS–PAGE of mouse MTHFS following thrombin treatment (Fig. 2B, lanes 6 and 7). The Michaelis–Menten constants for the MTHFS-catalyzed formation of 5,10-methenylTHF were determined using a spectrophotometer. Both Km and kcat values for recombinant mouse MTHFS were determined using Lineweaver–Burk plots and compared to known values for MTHFS proteins from other species (Table 3). The Km of 5-formylTHF for mouse MTHFS is similar to the Km values for the human, rabbit, and L. casei enzymes, and it is 6-fold lower than the Km for MTHFS from S. cerevisiae (Table 3). The Km of Mg–ATP for mouse MTHFS is between 2.3and 700-fold higher than the values reported for the human, rabbit, S. cerevisiae, and L. casei forms of MTHFS (Table 3). Recombinant mouse MTHFS has a turnover number of 980 min¡1, similar to the human form of the enzyme (Table 3). This is consistent with the high amino acid sequence identity (84%) found between mouse and human MTHFS, indicating that the three-dimensional structures for these two proteins may be similar.
Conclusion Here, we report a method for expression and puriWcation of MTHFS from M. musculus. This is the Wrst example of a mammalian form of MTHFS that has been successfully and stably expressed in E. coli with a hexahistidine tag allowing for rapid puriWcation. Although
the majority of recombinant protein is present as inclusion bodies, it is possible to purify 2–8 mg of MTHFS per liter of culture medium from the soluble fraction of the cell lysate. The recombinant MTHFS protein can be stabilized during multiple freeze–thawing using Tween 20 or the addition of either substrate. The expression and puriWcation procedures presented here yield suYcient amounts of unlabeled and 15N and 13C labeled protein to allow for further biochemical characterization and solution of the three-dimensional structure by NMR, which are currently in progress. The three-dimensional structure of MTHFS is not known, yet its elucidation will be very important in understanding the multiple roles that this enzyme plays in folate metabolism and folate homeostasis. Because of the high sequence similarity between the mammalian forms of MTHFS, the threedimensional structure of mouse MTHFS can be used to model the structure for the human form of this enzyme. Acknowledgment This work was supported by PHS HD35687 to P.J.S. References [1] C. Wagner, Biochemical role of folate in cellular metabolism, in: L.B. Bailey (Ed.), Folate in Health and Disease, Marcel Dekker, New York, 1995, pp. 23–42. [2] B. Shane, Folate chemistry and metabolism, in: L.B. Bailey (Ed.), Folate in Health and Disease, Marcel Dekker, New York, 1995, pp. 1–22. [3] P.J. Stover, C. Garza, Bringing individuality to public health recommendations, J. Nutr. 132 (2002) S2476–S2480. [4] S. Friso, S. Choi, D. Girelli, et al., A common mutation in the 5,10methylenetetrahydrofolate reductase gene aVects genomic DNA methylation through an interaction with folate status, Proc. Natl. Acad. Sci. USA 99 (2002) 5606–5611. [5] A. Fleming, A.J. Copp, Embryonic folate metabolism and mouse neural tube defects, Science (Washington, DC) 280 (1998) 2107– 2109. [6] P.A. Jones, S.B. Baylin, The fundamental role of epigenetic events in cancer, Nat. Rev. Genet. 3 (2002) 415–428. [7] C.K. Barlowe, D.R. Appling, In vitro evidence for the involvement of mitochondrial folate metabolism in the supply of cytoplasmic one-carbon units, BioFactors 1 (1988) 171–176. [8] P. Stover, V. Schirch, 5-Formyltetrahydrofolate polyglutamates are slow tight binding inhibitors of serine hydroxymethyltransferase, J. Biol. Chem. 266 (1991) 1543–1550.
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283 [9] P. Stover, V. Schirch, The metabolic role of leucovorin, Trends Biochem. Sci. 18 (1993) 102–106. [10] R. Bertrand, J. Jolivet, Methenyltetrahydrofolate synthetase prevents the inhibition of phosphoribosyl 5-aminoimidazole 4-carboxamide ribonucleotide formyltransferase by 5-formyltetrahydrofolate polyglutamates, J. Biol. Chem. 264 (1989) 8843–8846. [11] E. Chu, J.L. Grem, P.G. Johnston, C.J. Allegra, New concepts for the development and use of antifolates, Stem Cells 14 (1996) 41– 46. [12] H. Kruschwitz, D. McDonald, E.A. Cossins, V. Schirch, J. Biol. Chem. 269 (1994) 28757–28763. [13] S. Roje, M.T. Janave, M.J. Ziemak, A.D. Hanson, Cloning and characterization of mitochondrial 5-formyltetrahydrofolate cycloligase from higher plants, J. Biol. Chem. 277 (2002) 42748–42754. [14] W.B. Holmes, D.R. Appling, Cloning and characterization of methenyltetrahydrofolate synthetase from Saccharomyces cerevisiae, J. Biol. Chem. 277 (2002) 20205–20213. [15] M.C. Anguera, J.R. Suh, H. Ghandour, I.M. Nasrallah, J. Selhub, P.J. Stover, Methenyltetrahydrofolate synthetase regulates folate turnover and accumulation, J. Biol. Chem. 278 (2003) 29856– 29862. [16] A. Dayan, R. Bertrand, M. Beauchemin, et al., Cloning and characterization of the human 5,10-methenyltetrahydrofolate synthetase-encoding cDNA, Gene 165 (1995) 307–311. [17] C. Combet, C. Blanchet, C. Geourjon, G. Deléage, G. D. TIBS 25 (2000) 147–150.
283
[18] S. Girgis, J.R. Suh, J. Jolivet, P.J. Stover, 5-Formyltetrahydrofolate regulates homocysteine remethylation in human neuroblastoma, J. Biol. Chem. 272 (1997) 4729–4734. [19] B. Maras, P. Stover, S. Valiante, D. Barra, V. Schirch, Primary structure and tetrahydropteroylglutamate binding site of rabbit liver cytosolic 5,10-methenyltetrahydrofolate synthetase, J. Biol. Chem. 269 (1994) 18429–18433. [20] R. Bertrand, M. Beauchemin, A. Dayan, M. Ouimet, J. Jolivet, IdentiWcation and characterization of human mitochondrial methenyltetrahydrofolate synthetase activity, Biochim. Biophys. Acta 1266 (1995) 245–249. [21] R.J. Cook, R.S. Lloyd, C. Wagner, Isolation and characterization of cDNA clones for rat liver 10- formyltetrahydrofolate dehydrogenase, J. Biol. Chem. 266 (1991) 4965–4973. [22] J. Jolivet, A. Dayan, M. Beauchemin, D. Chahla, A. Mamo, R. Bertrand, Biochemical and molecular studies of human methenyltetrahydrofolate synthetase, Oncologist 1 (1996) 248–254. [23] D.R. Appling, Compartmentation of folate-mediated one-carbon metabolism in eukaryotes, FASEB J. 5 (1991) 2645–2651. [24] C.E. Grimshaw, G.B. Henderson, G.G. Soppe, G. Hansen, E.J. Mathur, F.M. Huennekens, PuriWcation and properties of 5,10methenyltetrahydrofolate synthetase from Lactobacillus casei, J. Biol. Chem. 259 (1984) 2728–2733. [25] S. Hopkins, V. Schirch, 5,10-Methenyltetrahydrofolate synthetase. PuriWcation and properties of the enzyme from rabbit liver, J. Biol. Chem. 259 (1984) 5618–5622.
Cloning, expression, and puriWcation of 5,10-methenyltetrahydrofolate synthetase from Mus musculus Montserrat C. Anguera,a Xiaowen Liu,b and Patrick J. Stovera,b,¤ a
Cornell University, Graduate Field of Biochemistry, Molecular and Cellular Biology, Ithaca, NY 14853, USA b Division of Nutritional Sciences, Cornell University, 315 Savage Hall, Ithaca, NY 14853, USA Received 28 October 2003, and in revised form 16 December 2003
Abstract Folate metabolism is necessary for the biosyntheses of purine nucleotides and thymidylate and for the synthesis of S-adenosylmethionine, a cofactor required for cellular methylation reactions and a precursor of spermidine and spermine syntheses. Disruption of folate metabolism is associated with several pathologies and developmental anomalies including cancer and neural tube defects. The enzyme 5,10-methenyltetrahydrofolate synthetase (MTHFS, EC 6.3.3.2) catalyzes the ATP-dependent conversion of 5-formyltetrahydrofolate to 5,10-methenyltetrahydrofolate, and has been shown to aVect intracellular folate concentrations by accelerating folate degradation. Mammalian MTHFS proteins described to date are not stable and no recombinant mammalian MTHFS protein has been successfully expressed in Escherichia coli. The three-dimensional structure of MTHFS has not been solved. The cDNA coding for Mus musculus MTHFS was isolated and expressed in E. coli with a hexa-histidine tag. Milligram quantities of recombinant mouse MTHFS were puriWed using metal aYnity chromatography and the protein was stabilized with Tween 20. Mouse MTHFS has a molecular mass of 23 kDa and is 84% identical in amino acid sequence to the human enzyme. Activity assays conWrmed the functionality of the recombinant protein, with Km D 5 M for (6S)-5-formyltetrahydrofolate and Km D 769 M for Mg–ATP. This is the Wrst example of a mammalian form of MTHFS expressed in E. coli that yielded suYcient quantities of stable puriWed protein to allow for detailed characterization of its three-dimensional structure and kinetic properties. 2004 Elsevier Inc. All rights reserved. Keywords: MTHFS; Folate; Homocysteine
The reduced tetrahydrofolates (THF)1 are cofactors that carry one-carbon units at three oxidation states ranging from formate to methanol [1]. Folate-dependent one-carbon metabolism is necessary for the syntheses of purines (supplies the carbon-2 and carbon-8 of the purine ring) and thymidylate (methylation of deoxyuridylate to thymidylate), and also for remethylation of homocysteine to methionine [2]. These reactions are critical for DNA synthesis and maintenance of genomic integrity [3], and disruption of folate metabolism
¤
Corresponding author. Fax: 1-607-255-9751. E-mail address: [email protected] (P.J. Stover). 1 Abbreviations used: THF, tetrahydrofolate; MTHFS, 5,10-methenyltetrahydrofolate synthetase; MES, 2-(4-morpholino)ethanesulfonic acid; Hepes, 2-[4-(2-hydroxyethyl)-1-piperazine]ethanesulfonic acid; ATP, adenosine triphosphate; IPTG, isopropyl--D-thiogalactopyranoside. 1046-5928/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2004.02.010
increases risk for developmental anomalies, cancer, and cardiovascular disease [4–6]. The biologically active THF derivatives contain a polyglutamate peptide consisting of Wve to eight glutamate residues connected by peptide bonds [2]. Folate polyglutamates function as one-carbon donators and acceptors in reactions known collectively as one-carbon metabolism, which occurs in both the cytoplasm and mitochondria [7]. 5-FormylTHF is the only reduced folate derivate that does not participate directly as a cofactor in one-carbon metabolism, but it can inhibit various folate-dependent enzymes [8–10]. 5-FormylTHF is the most chemically stable form of reduced folate, and it is administered clinically to patients to elevate intracellular folate levels [11]. Because of its stability, it has been suggested that 5formylTHF may function as the primary storage form of folate within the cell [12]. Studies have shown that dormant cells such as Neurospora crassa spores and
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
soybeans have high levels of 5-formylTHF [11,13]. The enzyme 5,10-methenyltetrahydrofolate synthetase (MTHFS; EC 6.3.3.2), which is the only enzyme that utilizes 5-formylTHF as a substrate, catalyzes the following irreversible reaction: 5-formylTHF C Mg–ATP!5, 10-methenylTHF C Mg–ADP C Pi MTHFS activity is found in many organisms such as bacteria, plants, and mammals [14], suggesting that it plays an important role in regulating folate metabolism. Recently, we demonstrated that elevations in MTHFS expression increase rates of folate turnover or degradation [15]. MCF-7 and SH-SY5Y neuroblastoma cells expressing the human MTHFS cDNA exhibit severe folate deWciency, increased rates of folate turnover, and elevated concentrations of folate degradation products in culture medium relative to control cells [15]. These data suggest that MTHFS may have a second activity that catalyzes the degradation of folate cofactors. To determine if MTHFS catalyzes folate degradation in vitro, milligram quantities of stable MTHFS protein are required. EVorts to express mammalian MTHFS proteins in bacteria have not been successful because of low yields and lack of protein stability [16]. In this study, we report the cloning, expression, and puriWcation of recombinant Mus musculus MTHFS. The expression and puriWcation conditions were optimized to yield milligram quantities of pure and functional protein that remained stable after puriWcation. This is the only mammalian form of MTHFS that has been successfully expressed in Escherichia coli and puriWed in milligram quantities, allowing for characterization of the recombinant protein including the determination of its threedimensional structure by NMR.
Materials and methods Materials Restriction enzymes were purchased from New England Biolabs. Mes, Hepes, Tris, and ATP were purchased from Sigma. (6S)-5-FormylTHF and (6R,S)-5formylTHF were a generous gift from Eprova AG. [15N]ammonium chloride and [13C]D-glucose were purchased from Isotech. All other materials were of high quality and obtained from various commercial vendors. cDNA cloning of MTHFS from M. musculus and construction of the MTHFS expression vector The murine MTHFS cDNA was generated by RTPCR using the high Wdelity Pfu polymerase (Stratagene). RNA was isolated from mouse liver (129/Sv strain) using the Total RNA Isolation Kit (Purgene) and con-
277
verted to cDNA using a polyT primer. The cDNA was used as the template to amplify the 30 end of the mouse MTHFS cDNA by PCR using degenerate primers based on human MTHFS sequence 50 GTCCCGCGTACT GAGCCAGAAG 30 (forward primer) and oligo(dT) (reverse primer). The PCR conditions used were 95 °C for 45 s, 42 °C for 1 min, and 72 °C for 1 min, for a total of 30 cycles. The PCR products were sequenced from two independent PCRs and found to be identical. This partial mouse MTHFS sequence was used to determine the sequence of the entire MTHFS open reading frame by 50 RACE following the manufacturer’s instructions (Gibco). Two primers were derived based on the 30 end of the partial mouse MTHFS cDNA sequence, the reverse primer (50 GAGCAGTAAGTGTAGCCGTCG TG 30) and the nested reverse primer (50 ATAAGGC CTCCTCCCGAACAT 30). Once the 50 sequence was determined, the mouse MTHFS open reading frame was ampliWed using the forward primer 50 GCGCCATATG GCAGACAGGGATGCCAC 30 (NdeI site in italics) and the reverse primer 50 GCGCAAGCTTTCAGAAG TCAGGAATTCCCC 30 (HindIII site in italics). Three independent PCR products were sequenced and found to be identical. The products were cloned into the pCR2.1 vector (Invitrogen), creating clone pCRMTHFS, and transformed into E. coli TOP10 cells (Invitrogen). Individual colonies were selected for DNA sequencing analyses to verify both the integrity of the NdeI and HindIII cloning sites, and the MTHFS cDNA sequence. The pCR-MTHFS vector was digested with NdeI and HindIII and ligated into the corresponding sites of the pET28a (+) expression vector (Novagen) inframe with an N-terminal hexa-histidine tag. DNA sequence analysis and amino acid alignment Alignment of the mouse, human, rabbit, Arabidopsis, Caenorhabditis elegans, and Saccharomyces cerevisiae MTHFS amino acid sequences was performed using the Pole Bio-Informatique Lyonnais Network Protein Sequence Analysis server (available at http://npsapbil. ibcp.fr/) using the Clustal W algorithm [17]. DNA sequencing of all constructs was performed at Cornell University BioResource Center. Expression of recombinant mouse MTHFS in E. coli cells Mouse MTHFS cDNA in the pET28a vector was used to transform E. coli BL21 Star (DE3) cells (Invitrogen). A single colony was used to inoculate 5 mL LB containing 50 g/mL kanamycin and the culture was agitated at 225 rpm overnight at 37 °C. The cells were pelleted by centrifugation, then resuspended in LB media, and used to inoculate a 1 L of SuperBroth medium (Bio101). The culture was incubated at 37 °C until OD550 nm D 0.8, then protein expression was induced
278
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
using 1.5mM isopropyl--D-thiogalactopyranoside (IPTG), and the incubation temperature was lowered to 15 °C for 20 h. Cells were harvested by centrifugation at 7000 rpm at 4 °C and cell pellets were stored at ¡80 °C. Expression of 15N and 13C labeled mouse MTHFS protein using M9 minimal media A conXuent 5 mL culture of E. coli BL21 Star cells expressing mouse MTHFS cDNA was used to inoculate a 50 mL starter culture containing LB supplemented with 50 g/mL kanamycin. The 50 mL culture was grown at 37 °C for 3 h, then pelleted by centrifugation, and resuspended in 1 mL of 2£ M9 media. Resuspended cells were used to inoculate 750 mL of M9 media containing 2£ M9 salts, 8 g D-glucose (or [13C]D-glucose), 2 g [15N]ammonium chloride, 5 mM magnesium sulfate, and 50 g/mL kanamycin. Cells were grown at 37 °C to OD600 nm D 1.0, then protein expression was induced using 1.5 mM IPTG, and the incubation temperature was lowered to 15 °C for 20 h. Cells were harvested by centrifugation at 7000 rpm at 4 °C and cell pellets were stored at ¡80 °C. Cell lysis and puriWcation of mouse MTHFS Frozen cell pellets were resuspended on ice in 10 mM Tris, pH 8.0, 100 mM NaCl, and 100 M PMSF, and the cells were lysed by four passages through a French press. Cell debris was pelleted by centrifugation at 15,000 rpm for 15 min at 4 °C. The supernatant was added to a column containing Talon metal aYnity resin (Clontech) equilibrated with wash buVer (10 mM Tris, pH 7.0, 100 mM NaCl, and 12.5 mM imidazole). The column was washed with 5 column volumes of wash buVer. The protein was eluted using 10 mM Tris, pH 7.0, 100 mM NaCl, and 125 mM imidazole. The absorbance at 280 nm was determined for all 10 fractions collected, and fractions containing the eluted protein (fractions 1–8) were pooled together and concentrated to 2–3 mL using a Centricon-8 (Millipore). Cleavage of the fusion protein was initiated by addition of 5 U of biotinylated thrombin (Novagen) to the eluted protein fractions, followed by incubation for 2 h at room temperature. Thrombin was removed using streptavidin–agarose beads and spin columns (Thrombin Cleavage Capture kit, Novagen). The puriWed protein was dialyzed overnight at 4 °C against 2 L of 50 mM Tris, pH 8.0. For 15N and 13C labeled MTHFS, the dialysis buVer was Sorenson’s Phosphate (67 mM), pH 6.0. The purity of the recombinant MTHFS protein was determined from an SDS– PAGE gel and analyzed using ChemiImager 4400 from Alpha Innotech (San Leandro, CA). The concentrated puriWed protein was aliquoted and stored at ¡80 °C. Protein concentrations were determined by the Lowry method.
MTHFS activity assay MTHFS activity was determined by monitoring the increase in absorbance at 355 nm, which is due to the formation of the product 5,10-methenylTHF (360 D 25,100 M¡1 cm¡1). For determination of speciWc activity values during the puriWcation, protein samples were added to a cuvette containing 100 mM Mes, pH 6.0, 100 M (6S)-5-formylTHF, and 1 mM Mg–ATP. Kinetic parameters of both substrates for MTHFS were obtained using saturating concentrations of one substrate with a range of concentrations of the second substrate in 100 mM Mes buVer, pH 6.0. Km values were derived from Lineweaver–Burk plots, and activity assays were performed at 28 °C with readings taken in quadruplicate.
Results and discussion Cloning of the mouse MTHFS open reading frame and sequence comparisons Total RNA from a 129/Sv mouse liver was isolated and converted to cDNA. The 30 end of the mouse MTHFS cDNA sequence was ampliWed by PCR using a high Wdelity thermostable polymerase and degenerate primers based on the human MTHFS cDNA sequence [16]. The 30 sequence derived from the PCR product was used to obtain the murine MTHFS cDNA by 50 RACE. PCR ampliWcation of the MTHFS open reading frame yielded a 612 nucleotide DNA sequence coding for the mouse form of MTHFS, which encoded a 203 amino acid protein with a predicted molecular mass of 23,210 Da. After verifying the nucleotide sequence of three independently derived clones, the mouse MTHFS cDNA was cloned in-frame with an N-terminal hexahistidine tag into a pET28 expression vector. This MTHFS cDNA is nearly identical to a C57/Bl/6 cDNA sequence obtained from embryo tissue [GenBank Accession No. NM026829 (nucleotide) and GenBank Accession No. Q9D110 (protein)], but the two sequences diVer by one nucleotide (T298C) that results in an amino acid substitution (S98P). The human MTHFS protein sequence has a proline residue at position 98 [16], indicating that either the database sequence for C57/Bl/6 mouse MTHFS (GenBank Accession No. NM026829) is incorrect or that the mouse MTHFS protein is polymorphic. To determine the localization of the mouse MTHFS protein and putative substrate binding motifs, the MTHFS amino acid sequence alignment comparisons and cellular localization predictions were obtained using several software programs. Mouse MTHFS was predicted to be cytoplasmic determined by the PSORTII program (http://psort.nibb.ac.jp/). Analysis of MTHFS
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
amino acid sequence from S. cerevisiae using the same program also predicted a cytoplasmic protein [14]. Human and rabbit liver MTHFS activity is found primarily in the cytoplasm [18,19], but one study detected about 15% of total human MTHFS activity in mitochondria from human liver [20]. Mitochondria isolated from rabbit liver do not contain MTHFS activity [19]. Fig. 1 contains the amino acid sequence alignment, using the Cluster W algorithm, for the mouse, human, rabbit, S. cerevisiae, C. elegans, and Arabidopsis forms of MTHFS. The protein sequence of mouse MTHFS is
279
84% identical with the human sequence and 76% identical with the rabbit protein. The mouse MTHFS sequence is 36, 28, and 26% identical compared to the C. elegans, S. cerevisiae, and Arabidopsis sequences, respectively. There is 8.8% sequence identity among all six protein sequences of MTHFS (Fig. 1). Carbodiimideactivated 5-formylTHF(Glu)1 and 5-formylTHF(Glu)5 covalently attach to lysine-68 of rabbit liver MTHFS, indicating that this residue is involved in a salt bridge with the -carboxyl group of the Wrst glutamate residue of the folate substrate [19]. A lysine residue at position
Fig. 1. Alignment of MTHFS protein sequences from mouse, human (GenBank Accession No. AAC41945), rabbit (GenBank Accession No. A53688), S. cerevisiae (GenBank Accession No. P40099), C. elegans (GenBank Accession No. 492692), and Arabidopsis (GenBank Accession No. 568284). The symbol (*) beneath the amino acid residue indicates identity, the symbol (:) indicates a conservative substitution, and the symbol (.) indicates a non-conservative substitution. The sequences were aligned by the Pole Bio-Informatique Lyonnais Network Protein Sequence Analysis server (available at http://npsa-pbil.ibcp.fr/) using the Clustal W algorithm [17]. The mouse, human, and rabbit proteins are 70% identical and have 16% conservative substitutions. All six sequences are 8.8% identical.
280
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
68 is present in MTHFS sequences from mouse, human, rabbit, C.elegans, and S.cerevisiae, and MTHFS from Arabidopsis contains an arginine residue at this position, which could also interact with the glutamate moiety (Fig. 1). The peptide region spanning residues 60–72 of mouse, human, and rabbit liver MTHFS contains alternating Arg and Lys residues that may interact with the negatively charged polyglutamate portion of 5-formylTHF(Glu)n [19]. The human MTHFS enzyme contains a putative folate binding signature sequence (S159-L-L-PX-X-X-W171) [16,21], and these residues are nearly identical in the mouse and rabbit MTHFS proteins (Fig. 1). Analysis of the mouse MTHFS amino acid sequence using the PROSITE database failed to identify any binding motifs, however the MOTIF sequence analysis tool searching the BLOCKS database (http://motif. genome.ad.jp) revealed a dihydropteroate synthase motif in region 186–207, which may correspond to the pterin binding site of the folate substrate of mouse MTHFS. MTHFS is known to bind Mg–ATP [13,19,22,23], and there is a probable ATP-binding site motif at positions 209–217 that is similar to the ATP/GTP-binding site motif (P-loop; [AG]-x-x-x-x-G-K-[ST]). This region contains four glycine residues that would allow Xexibility for the positively charged lysine and arginine residues to interact with the negatively charged phosphate groups of ATP. The last residue in this putative ATP-binding motif is a glycine residue instead of a serine or threonine, which is similar to the ATP-binding sequence for adenylate kinase (ExPASy documentation). The area spanning residues 209–217 displays high sequence identity among the six MTHFS sequences compared in Fig. 1, suggesting that this area is important for binding the ATP substrate. ConWrmation of these possible 5-formylTHF and Mg–ATP-binding motifs requires the three-dimensional structure of mouse MTHFS. Expression and puriWcation of recombinant mouse MTHFS Previous attempts to express the recombinant human MTHFS in E. coli resulted in very low yields of functional protein that required a multi-step puriWcation procedure (unpublished results and [16]). The purpose of this study was to determine if the murine MTHFS protein exhibited greater stability as a recombinant protein than the human protein, and if milligram amounts of protein could be obtained and puriWed with one column. E. coli BL21 Star cells were transformed with the pET28a vector containing the mouse MTHFS cDNA and transformants were screened for optimal protein expression. Initially, 1 L cultures were grown in SuperBroth media, induced with IPTG, and incubated at 25 °C for 5 h. However, only 1–2 mg of puriWed protein was obtained per liter of culture. By lowering the induction temperature to 15 °C and extending the induction time
Fig. 2. SDS–PAGE analyses of recombinant mouse MTHFS protein during puriWcation. Recombinant mouse MTHFS protein, containing an N-terminal hexa-histidine tag, was expressed in E. coli and puriWed using a metal aYnity column. (A) Cells were lysed using a French press, and the soluble and insoluble fractions (25–50 g) were run on a 12% SDS–PAGE gel and stained with Coomassie brilliant blue. Lane 1, Precision Plus Protein Standards (BioRad); lane 2, the uninduced culture; lane 3, the insoluble pellet after cell lysis; and lane 4, the soluble fraction of the cell lysate (supernatant). (B) SDS–PAGE gel (12%) showing the fractions from a typical puriWcation and stained with Coomassie brilliant blue. Lane 1, Precision Plus Protein Standards (BioRad); lane 2, 50 g of soluble supernatant following cell lysis; lane 3, 50 l sample of the column wash; lane 4, 50 l sample of elution fraction 1; lane 5, 20 g sample of elution fraction 2; lane 6, 20 g sample of pooled fractions of eluted protein treatment with thrombin; and lane 7, 10 g sample of puriWed protein after overnight dialysis.
to 20 h, the yield of puriWed protein increased to 8 mg/L of culture. Analysis of the soluble and insoluble fractions of the cell lysate from E. coli expressing mouse MTHFS indicated that the majority of recombinant MTHFS protein was present as inclusion bodies (Fig. 2A). SDS–PAGE analyses of uninduced and induced cell lysates indicated that the recombinant mouse MTHFS protein was primarily present in the insoluble fraction (Fig. 2A). Refolding attempts using sodium deoxycholate and a
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
Protein Refolding Kit (Novagen) were unsuccessful because the protein rapidly denatured during the Wnal dialysis. However, there were suYcient quantities of protein present in the soluble fraction to allow for protein puriWcation and characterization. Cells were lysed using a French press and the soluble MTHFS protein was puriWed using metal aYnity chromatography with Trisbased buVers. The presence of Hepes in the puriWcation buVers resulted in the rapid denaturation of the protein during the Wnal dialysis step. Recombinant mouse MTHFS was puriWed to greater than 98% purity using a single metal aYnity column. Samples from a typical puriWcation of mouse MTHFS were run on a 12% SDS– PAGE gel and are shown in Fig. 2B. MTHFS activity assays of the total protein, total activity, and speciWc activity values for the initial and Wnal steps in a typical puriWcation are shown in Table 1. The recombinant protein was eluted from the column using 125 mM imidazole, and the second elution fraction contained the majority of MTHFS protein (Fig. 2B). The eluted fractions were pooled, and the hexa-histidine tag was removed with thrombin treatment. Complete removal of the hexa-histidine tag occurred without degradation of the puriWed protein (Fig. 2B, lanes 6–7). Removal of the hexa-histidine tag did not aVect the activity nor stability of recombinant mouse MTHFS because the protein remained stable and active after a 22 h dialysis (Table 1 and Fig. 2B). We cannot account for the increased MTHFS activity in the puriWed protein compared to the cell lysate (Table 1), although inhibitory molecules may have been present in the crude cell lysates. Image densitometry analysis of the SDS–PAGE gel in Fig. 2B indicated that the purity of recombinant mouse MTHFS protein was greater than 98% and other puriWcations yielded protein with similar purity. The puriWed protein was concentrated to 20–30 mg/mL without denaturation and the addition of 0.1% Tween 20 stabilized the protein through multiple rounds of freeze–thawing with no changes in speciWc activity (data not shown). The literature reports that MTHFS proteins from rabbit, S. cerevisiae, and Lactobacillus casei are stabilized by the presence of non-ionic detergents because of the hydrophobicity of the MTHFS protein [14,19,24]. The recombinant MTHFS enzyme is not stable after several rounds of freeze–thawing, but it can be stabilized with
281
the presence of either substrate, (6S)-5-formylTHF or Mg–ATP. Expression and puriWcation of mouse MTHFS using minimal media The recombinant MTHFS protein was expressed in minimal media and puriWed using the protocols described previously to determine if NMR could be used to solve the three-dimensional structure of mouse MTHFS protein. A 750 ml culture of E. coli BL21 cells expressing recombinant mouse MTHFS was grown in M9 minimal medium. The culture reached an OD600 nm of 1.0 in approximately 6.5 h, whereas cultures grown in SuperBroth achieved an OD600 nm of 1.0 in approximately 3 h (data not shown). The Wnal yields of puriWed recombinant MTHFS protein expressed in E. coli grown in diVerent culture media are summarized in Table 2. Typical puriWcations of mouse MTHFS from cultures grown using SuperBroth medium yielded between 6 and 8 mg of puriWed protein per liter of culture. Expression of recombinant mouse MTHFS in M9 minimal media containing [15N]ammonium chloride yielded an average of 2 mg of puriWed protein per liter of culture, and media containing both [15N]ammonium chloride and [13C]Dglucose also yielded 2 mg puriWed protein per liter of culture. Therefore, milligram quantities of 15N and 13C labeled recombinant mouse MTHFS protein can be obtained and used for NMR structural determination. Single and double labeled mouse MTHFS protein was Table 2 PuriWed protein yields of recombinant mouse MTHFS grown in diVerent culture media Culture media
PuriWed protein yield (mg puriWed protein/L culture media)
SuperBroth 2£ M9 2£ M9 with [15N]ammonium chloride 2£ M9 with [15N]ammonium chloride and [13C]glucose
7.0 8.3 1.7 2.2
Escherichia coli expressing mouse MTHFS was cultured in diVerent bacterial growth media, then puriWed as described in the Materials and methods section. Values represent the average protein quantities obtained from three individual preparations.
Table 1 Summary of recombinant mouse MTHFS puriWcation Step
Total protein (mg)
Total activity (U) £ 1000
SpeciWc activity (U/g) £ 10,000
PuriWcation fold factor
Cell lysate (supernatant) PuriWed protein (without His tag)
720 5.5
4.1 5.6
0.13 22.2
1.0 171
The values for total protein, total activity, speciWc activity, and the puriWcation fold factor for a typical puriWcation of mouse MTHFS are shown below. Protein concentrations were determined using the Lowry method. The activity values represent the average of triplicate measurements from a representative preparation. Assays were performed at 24 °C using 100 M (6S)-5-formylTHF, 1 mM Mg–ATP, and 100 mM Mes, pH 6.0. One unit (U) corresponds to 1 mol 5,10-methenylTHF/min.
282
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283
Table 3 Comparison of kinetic parameters for recombinant mouse MTHFS and other forms of MTHFS Species
(6R,S) 5-CHO-THF Km (M)
Mg–ATP Km (M)
Turnover number (min¡1)
Reference
Mouse Human Rabbit
5.0 4.4 8.0 33
769 20 330 43
980 1000 300 —a
Present report [20] [25] [14]
S. cerevisiae L. casei
0.6
1.0
88
[24]
The Km values of the substrates for MTHFS and the turnover numbers for various forms of MTHFS from diVerent species are shown. The mouse, human, rabbit, and L. casei enzymes were assayed using buVers at pH 6.0 and S. cerevisiae MTHFS was assayed at pH 7.0. a Value not reported.
puriWed using the metal aYnity procedure described previously, and either 5-formylTHF or Mg–ATP was added to the puriWed protein to aVect stability. Biochemical characterization of recombinant mouse MTHFS The molecular mass of puriWed recombinant mouse MTHFS containing a hexa-histidine tag is approximately 25 kDa as determined by SDS–PAGE gels (Fig. 2B, lane 5). Analysis by MALDI-TOF mass spectrometry of recombinant mouse MTHFS (with the hexahisitidine tag) indicated that the molecular mass is 25,221 Da (data not shown). The predicted molecular mass of mouse MTHFS without the hexa-histidine tag is 23,210 Da, and this value agrees with values obtained by SDS–PAGE of mouse MTHFS following thrombin treatment (Fig. 2B, lanes 6 and 7). The Michaelis–Menten constants for the MTHFS-catalyzed formation of 5,10-methenylTHF were determined using a spectrophotometer. Both Km and kcat values for recombinant mouse MTHFS were determined using Lineweaver–Burk plots and compared to known values for MTHFS proteins from other species (Table 3). The Km of 5-formylTHF for mouse MTHFS is similar to the Km values for the human, rabbit, and L. casei enzymes, and it is 6-fold lower than the Km for MTHFS from S. cerevisiae (Table 3). The Km of Mg–ATP for mouse MTHFS is between 2.3and 700-fold higher than the values reported for the human, rabbit, S. cerevisiae, and L. casei forms of MTHFS (Table 3). Recombinant mouse MTHFS has a turnover number of 980 min¡1, similar to the human form of the enzyme (Table 3). This is consistent with the high amino acid sequence identity (84%) found between mouse and human MTHFS, indicating that the three-dimensional structures for these two proteins may be similar.
Conclusion Here, we report a method for expression and puriWcation of MTHFS from M. musculus. This is the Wrst example of a mammalian form of MTHFS that has been successfully and stably expressed in E. coli with a hexahistidine tag allowing for rapid puriWcation. Although
the majority of recombinant protein is present as inclusion bodies, it is possible to purify 2–8 mg of MTHFS per liter of culture medium from the soluble fraction of the cell lysate. The recombinant MTHFS protein can be stabilized during multiple freeze–thawing using Tween 20 or the addition of either substrate. The expression and puriWcation procedures presented here yield suYcient amounts of unlabeled and 15N and 13C labeled protein to allow for further biochemical characterization and solution of the three-dimensional structure by NMR, which are currently in progress. The three-dimensional structure of MTHFS is not known, yet its elucidation will be very important in understanding the multiple roles that this enzyme plays in folate metabolism and folate homeostasis. Because of the high sequence similarity between the mammalian forms of MTHFS, the threedimensional structure of mouse MTHFS can be used to model the structure for the human form of this enzyme. Acknowledgment This work was supported by PHS HD35687 to P.J.S. References [1] C. Wagner, Biochemical role of folate in cellular metabolism, in: L.B. Bailey (Ed.), Folate in Health and Disease, Marcel Dekker, New York, 1995, pp. 23–42. [2] B. Shane, Folate chemistry and metabolism, in: L.B. Bailey (Ed.), Folate in Health and Disease, Marcel Dekker, New York, 1995, pp. 1–22. [3] P.J. Stover, C. Garza, Bringing individuality to public health recommendations, J. Nutr. 132 (2002) S2476–S2480. [4] S. Friso, S. Choi, D. Girelli, et al., A common mutation in the 5,10methylenetetrahydrofolate reductase gene aVects genomic DNA methylation through an interaction with folate status, Proc. Natl. Acad. Sci. USA 99 (2002) 5606–5611. [5] A. Fleming, A.J. Copp, Embryonic folate metabolism and mouse neural tube defects, Science (Washington, DC) 280 (1998) 2107– 2109. [6] P.A. Jones, S.B. Baylin, The fundamental role of epigenetic events in cancer, Nat. Rev. Genet. 3 (2002) 415–428. [7] C.K. Barlowe, D.R. Appling, In vitro evidence for the involvement of mitochondrial folate metabolism in the supply of cytoplasmic one-carbon units, BioFactors 1 (1988) 171–176. [8] P. Stover, V. Schirch, 5-Formyltetrahydrofolate polyglutamates are slow tight binding inhibitors of serine hydroxymethyltransferase, J. Biol. Chem. 266 (1991) 1543–1550.
M.C. Anguera et al. / Protein Expression and PuriWcation 35 (2004) 276–283 [9] P. Stover, V. Schirch, The metabolic role of leucovorin, Trends Biochem. Sci. 18 (1993) 102–106. [10] R. Bertrand, J. Jolivet, Methenyltetrahydrofolate synthetase prevents the inhibition of phosphoribosyl 5-aminoimidazole 4-carboxamide ribonucleotide formyltransferase by 5-formyltetrahydrofolate polyglutamates, J. Biol. Chem. 264 (1989) 8843–8846. [11] E. Chu, J.L. Grem, P.G. Johnston, C.J. Allegra, New concepts for the development and use of antifolates, Stem Cells 14 (1996) 41– 46. [12] H. Kruschwitz, D. McDonald, E.A. Cossins, V. Schirch, J. Biol. Chem. 269 (1994) 28757–28763. [13] S. Roje, M.T. Janave, M.J. Ziemak, A.D. Hanson, Cloning and characterization of mitochondrial 5-formyltetrahydrofolate cycloligase from higher plants, J. Biol. Chem. 277 (2002) 42748–42754. [14] W.B. Holmes, D.R. Appling, Cloning and characterization of methenyltetrahydrofolate synthetase from Saccharomyces cerevisiae, J. Biol. Chem. 277 (2002) 20205–20213. [15] M.C. Anguera, J.R. Suh, H. Ghandour, I.M. Nasrallah, J. Selhub, P.J. Stover, Methenyltetrahydrofolate synthetase regulates folate turnover and accumulation, J. Biol. Chem. 278 (2003) 29856– 29862. [16] A. Dayan, R. Bertrand, M. Beauchemin, et al., Cloning and characterization of the human 5,10-methenyltetrahydrofolate synthetase-encoding cDNA, Gene 165 (1995) 307–311. [17] C. Combet, C. Blanchet, C. Geourjon, G. Deléage, G. D. TIBS 25 (2000) 147–150.
283
[18] S. Girgis, J.R. Suh, J. Jolivet, P.J. Stover, 5-Formyltetrahydrofolate regulates homocysteine remethylation in human neuroblastoma, J. Biol. Chem. 272 (1997) 4729–4734. [19] B. Maras, P. Stover, S. Valiante, D. Barra, V. Schirch, Primary structure and tetrahydropteroylglutamate binding site of rabbit liver cytosolic 5,10-methenyltetrahydrofolate synthetase, J. Biol. Chem. 269 (1994) 18429–18433. [20] R. Bertrand, M. Beauchemin, A. Dayan, M. Ouimet, J. Jolivet, IdentiWcation and characterization of human mitochondrial methenyltetrahydrofolate synthetase activity, Biochim. Biophys. Acta 1266 (1995) 245–249. [21] R.J. Cook, R.S. Lloyd, C. Wagner, Isolation and characterization of cDNA clones for rat liver 10- formyltetrahydrofolate dehydrogenase, J. Biol. Chem. 266 (1991) 4965–4973. [22] J. Jolivet, A. Dayan, M. Beauchemin, D. Chahla, A. Mamo, R. Bertrand, Biochemical and molecular studies of human methenyltetrahydrofolate synthetase, Oncologist 1 (1996) 248–254. [23] D.R. Appling, Compartmentation of folate-mediated one-carbon metabolism in eukaryotes, FASEB J. 5 (1991) 2645–2651. [24] C.E. Grimshaw, G.B. Henderson, G.G. Soppe, G. Hansen, E.J. Mathur, F.M. Huennekens, PuriWcation and properties of 5,10methenyltetrahydrofolate synthetase from Lactobacillus casei, J. Biol. Chem. 259 (1984) 2728–2733. [25] S. Hopkins, V. Schirch, 5,10-Methenyltetrahydrofolate synthetase. PuriWcation and properties of the enzyme from rabbit liver, J. Biol. Chem. 259 (1984) 5618–5622.