The effect of Arg 209 to Lys mutation in mouse thymidylate synthase

Mouse thymidylate synthase R209K (a mutation corresponding to R218K in Lactobacillus casei), overexpressed in thymidylate synthase-deficient Escherichia coli strain, was poorly soluble and with only feeble enzyme activity. The mutated protein, incubated with FdUMP and N 5,10 -methylenetetrahydrofolate, did not form a complex stable under conditions of SDS/polyacrylamide gel electrophoresis. The reaction catalyzed by the R209K enzyme (studied in a crude extract), compared to that catalyzed by purified wild-type recombinant mouse thymidylate synthase, showed the Km value for dUMP 571-fold higher and Vmax value over 50-fold (assuming that the mutated enzyme constituted 20% of total crude extract protein) lower. Thus the ratios kcat, R209K/kcat, ‘wild’ and (kcat, R209K/Km, R209K dUMP )/( kcat, ‘wild’/Km, ‘wild’ dUMP ) were 0.019 and 0.000032, respectively, documenting that mouse thymidylate synthase R209, similar to the corresponding L. casei R218, is essential for both dUMP binding and enzyme reaction.

As the sole de novo source of thymidylate synthesis, it is an attractive target in chemotherapy (Ealick & Armstrong, 1993).
The enzyme's active center requires the phosphate moiety of dUMP to be dianionic for optimum binding (Go³os et al., 2001).Crystallographic studies of thymidylate synthasebound dUMP have shown the dianionic phos-phate to be ligated by a quartet of highly conserved arginines (Arg23, Arg178¢, Arg179¢ and Arg218 in the Lactobacillus casei enzyme) and one serine (Ser219).This pattern appears to be common for the bacterial (Fines-Moore et al., 1993;Stroud & Finer-Moore, 1993;Fauman et al., 1994;reviewed in Carreras & Santi, 1995), fungal (Anderson et al., 2000) and mammalian (Sotelo-Mundo et al., 1999) synthases.The arginines appear to bind cooperatively to the phosphate oxygens, as their energetic contributions indicate ideal complementation of the charge and geometry of the phosphate moiety (Morse et al., 2000).Kawase et al. (2000) replaced each of the four arginine residues in the L. casei enzyme by mutagenesis, proving their differing contributions to binding and catalysis.Interestingly, only one of the four phosphate-binding arginine residues (Arg218) appeared to be indispensable for catalysis (Kawase et al., 2000) and its guanidinium group was suggested to lower, via ion-pairing, the pK a of the catalytic thiol of Cys198, enhancing its nucleophilicity (Hardy et al., 1987).So far no data on the effects of the corresponding replacements in the enzyme of higher organisms have been reported.An unexpected byproduct of our recent experiment, aimed at overexpression of mouse thymidylate synthase and involving PCR-assisted replacement of certain bases in the coding sequence, was the enzyme containing an R209K mutation (R218K in L. casei), resulting probably from a Taq polymerase error.The present paper presents the properties of the recombinant mutated and wild-type mouse thymidylate synthases.

MATERIALS AND METHODS
pBluescript SK -vector and Escherichia coli XL-1 Blue strain were from Stratagene, TX61 -, a derivative of the E. coli BL21(DE3) strain with the endogenous thymidylate synthase gene inactivated by transposon-me-diated insertion, was a gift from Dr. W.S. Dallas (Glaxo Wellcome, U.K.).
Maintenance of L1210 leukemia cells and RNA isolation.The cells were maintained as previously described (Rode et al., 1984).Transplantations were done once every 7 days by intraperitoneal inoculation of 10 4 cells in 0.5 ml of ascites fluid diluted with 0.9% NaCl.The cells were harvested 7 days after inoculation by centrifuging the ascites fluid at 300 ´g for 10 min.The pellet was washed twice with 0.9% NaCl and immediately used for total RNA isolation with TRIzol reagent (Gibco) according to the manufacturer's protocol.
RT-PCR and cDNA cloning.General methods for DNA manipulations were as described (Sambrook et al., 1989).Reverse transcription was performed on total RNA with Moloney Murine Leukemia Virus Reverse Transcriptase (Promega) and oligo(dT) primer with an additional round of reverse transcription reaction -after the first round the sample was denatured for 3 min at 95°C, cooled on ice and, after addition of reverse transcriptase (50 units), incubated for 1 h at 42°C.In the PCR procedure the following primers were used: forward (primer I) 5¢-GGGA-ATTCATATGCTGGTGGTTGGCTCCGAG-3¢, complementing the 21 nucleotide region upstream of the start codon, and the reverse (primer II) 5¢-AAAAAGCTTTTAAACAGCCA-TTTCCATTTTAAT-3¢, complementing the last 24 bases of the coding region.EcoRI and NdeI (in primer I) or HindIII (in primer II) restriction sites were included to permit cloning of the amplified fragment.This fragment, being thymidylate synthase coding region, was ligated into the EcoRI/HindIII sites of the pBluescript SK -vector and the ligation mixture was used to transform E. coli XL-1 cells.One-thousand bp inserts in the isolated recombinant plasmids (pBluescript/TS) were sequenced using USB Sequenase Version 2 DNA sequencing kit.The coding region was restricted from pBluescript as an NdeI/ HindIII fragment and ligated into the pPIGDM4+stop vector, either straight away or after prior PCR with primers III (5¢-GGG-AATTCATATGCTTGTTGTTGGTTCCGG-3¢) and II (in order to substitute T for G or C in the 2nd, 3rd and 5th codons).The pPIGDM4+ stop expression vector is a derivative of the pPIGDM1 plasmid (Mikiewicz et al., 1997), containing ampicillin resistance gene (inserted into the putative rom gene), T7 phage promoter, transcription terminator and rare arginine tRNA genes (tRNA Arg(AGG/AGA ).DH5aF¢ cells were transformed, recombinant plasmid isolated and the construction was confirmed.A pPIGDM4+stop/TS plasmid pool was used to transform the thymidylate synthase-deficient E. coli strain TX61 -.
Induction of thymidylate synthase synthesis.Overnight cultures of untransformed TX61 -cells or TX61 -cells transformed with the TS/pPIGDM4+stop plasmid wild-type or mutated were used for inoculation of superbroth medium (Pedersen-Lane et al., 1997) with 30 mg/ml kanamycin, 50 mg/ml ampicilin and 50 mg/ml thymidine.Cells were shaken at 270 r.p.m. at 37°C until the culture attained an A 600 of 0.6.IPTG (1 mM) was then added and shaking was continued for another 4 h, at which time the cells were centrifuged at 10 000 ´g for 15 min at 4°C.The cell pellets were frozen at -20°C until use.
Cell extracts.The cell pellets were suspended in lysis buffer containing 50 mM Tris/HCl, pH 7.5, 0.1% Triton X-100, 0.1 mM EDTA and 10 mM 2-mercaptoethanol and sonicated.The resulting homogenate was centrifuged for 20 min at 20 000 ´g at 4°C and the pellet discarded.The supernatant, further referred to as crude extract, was used for thymidylate synthase activity, FdUMP binding and protein assays.
Recombinant mouse thymidylate synthase purification.A previously described method (Cieoela et al., 1995) was modified as follows.Purification was done at 2-4°C and all buffers contained 20 mM 2-mercaptoethanol.Cell pellet, corresponding to 1 l of culture, was suspended in 100 ml of 50 mM so-dium/potassium phosphate buffer, pH 7.5, containing 0.1 M KCl, immersed in an ice-ethanol bath and sonicated (Branson Sonifier 250) for ten pulses of 60 s each, separated with 15 s pauses.After removal of cell debris by centrifugation for 20 min at 20 000 ´g, 2% streptomycin sulfate was added to the supernatant, the suspension stirred for 20 min and the precipitated nucleic acid removed by 20 min centrifugation at 20 000 ´g. Solid ammonium sulfate was added to the stirred supernatant to 30% saturation and, after additional 20 min stirring, the resulting mixture centrifuged for 20 min at 20 000 ´g.The ammonium sulfate content of the supernatant was raised to 80% saturation and, after stirring for 20 min, the resulting precipitate was spun down.It was dissolved in 10 mM sodium/potassium phosphate buffer, pH 7.5, dialyzed overnight against two 2-liter changes of the same buffer and loaded onto a DE-52 column (2.5 ´7 cm) equilibrated with the above buffer.The column was washed with 150 ml of the same buffer, followed by 250 ml of 25 mM sodium/potassium phosphate buffer pH 7.5, and the enzyme eluted with 50 mM sodium/potassium phosphate buffer, pH 7.5.Fractions containing thymidylate synthase activity were pooled and precipitated with solid ammonium sulfate at 80% saturation.Following centrifugation the pellet was dissolved in 10 mM sodium/potassium phosphate buffer, pH 7.5, containing 1 M ammonium sulfate, and loaded onto a phenyl-Sepharose column (1.5 ´7 cm) equilibrated in the same buffer.The column was washed with 200 ml of the same buffer and the enzyme eluted with 10 mM sodium/potassium phosphate buffer, pH 7.5, containing 0.8 M ammonium sulfate.The most active and purest fractions (assessed electrophoretically) were pooled and concentrated in an Amicon Column Eluate Concentrator apparatus.
Thymidylate synthase.This was determined by monitoring the release of tritium from [5-3 H]dUMP (Rode et al., 1984).A unit of activity was defined as the amount of enzyme catalyzing formation of 1 mmole of dTMP (or release of 1 mequivalent of 3 H + ) per minute at 37°C.Protein was determined as described by Spector (1978) with bovine serum albumin as a standard.
Kinetic studies.Quantitative analyses of thymidylate synthase interaction with dUMP were performed as reported earlier (Jastreboff et al., 1983).Results are presented as means ± S.E.M., or means ± % difference between the mean and each of two results, followed by the number of experiments (n) in parentheses.
Electrophoretic analysis.In FdUMP binding experiments a crude extract of recombinant thymidylate synthase ('wild' or mutated) was incubated at 37°C for 30 min with 2.5 mM N 5,10 -methylenetetrahydrofolate and 1 mM [ 3 H]FdUMP (18 Ci/mmol).Polyacrylamide (12%) gel electrophoresis in the presence of sodium dodecyl sulfate (Bollag et al., 1991) was employed to test enzyme homogeneity and to determine molecular mass of its monomer.To identify thymidylate synthase on the gel, based on the enzyme's ability to form a ternary complex with 5-fluoro-dUMP and N 5,10 -methylenetetrahydrofolate, a sample of the crude extract of the recombinant enzyme ('wild' or mutated) for electrophoresis was prepared as described elsewhere (Rode et al., 1979).
Western blotting.Standard method was applied (Bollag et al., 1991) with an anti-rat thymidylate synthase monoclonal antibody that recognized both the recombinant and L1210 mouse thymidylate synthases (Go³os et al., 2002) and goat anti-mouse IgG (H+L)-HRP conjugate (BioRad) as a second antibody.

RESULTS
RT-PCR was performed on total RNA isolated from mouse leukemia L1210 cells, with the use of primers designed based on the known mouse thymidylate gene sequence (Deng et al., 1986), and the resulting entire thymidylate synthase coding region was cloned into the pBluescript vector.A direct attempt to express the mouse enzyme in E. coli by subcloning of the NdeI/HindIII fragment, restricted from pBluescript/TS, into the pPIGDM4+stop expression vector was unsuccessful.In order to optimize the coding region sequence, following the approach of Pedersen-Lane et al. (1997), substitutions were made of G or C in the 2nd, 3rd and 5th codons with T (without changing the encoded amino-acid sequence).Unexpectedly, the procedure resulted in two products: a plasmid with the expected sequence and another one having an additional substitution of A for middle G in the AGG codon encoding Arg209 (amino-acid numbers starting from the first Met in the mouse enzyme), apparently introduced incidentally by Taq polymerase during the subcloning procedure.The latter substitution resulted in Lys209 present in the expressed protein.Both mouse thymidylate synthase forms ('wild' and R209K) were successfully overexpressed in the thymidylate synthase-deficient E. coli TX61 -cells, constituting about 20% of total protein (Fig. 1A).
Crude extracts of bacteria expressing the R209K enzyme showed a specific activity of 0.016 ± 0.009 munit/mg protein (n = 4), almost 4000-fold lower than that found with bacteria expressing 'wild' mouse thymidylate synthase (61.8 ± 28.8 munit/mg protein; n = 3).The enzyme activity was undetectable in extracts of untransformed TX61 -cells.Anti-rat thymidylate synthase monoclonal antibodies, recognizing also mouse recombinant enzyme (Go³os et al., 2002), used in immunoblot analysis following electrophoretic separation under denaturing conditions, allowed us to identify the overexpressed protein, present in extracts of bacteria harboring the mutated plasmid, as thymidylate synthase (Fig. 2, cf.Fig. 1A).Interestingly, with both the 'wild' and mutated enzymes the antibodies showed heterogeneity of the enzyme subunits.
While [ 3 H]FdUMP, binding in the presence of N 5,10 -methylenetetrahydrofolate to 'wild' thymidylate synthase, remained bound under the denaturing condition of SDS polyacrylamide electrophoresis, it failed to form so stable a complex with the R209K enzyme (Fig. 1B).
Table 1 compares the K m and V max values describing the reaction catalyzed by the crude extract R209K enzyme, and purified 'wild' recombinant, L1210 cell and mouse thymus thymidylate synthases.The affinity of the R209K enzyme for dUMP, reflected by the K m value, was 571-fold lower than that of the 'wild' recombinant thymidylate synthase, the latter showing 2.6-3.1-foldhigher affinity than the two enzyme forms purified from the natural sources.The R209K enzyme was assessed to constitute about 20% of total crude extract protein.Assuming the latter, the V max value, corresponding to 1 mg of pure R209K thymidylate synthase, would amount to 0.024 unit/mg protein and would be over 50-fold lower than that estimated for 'wild' recombinant thymidylate synthase (k cat, R209K /k cat, 'wild' = 0.019; the k cat /K m dUMP ratio over 30 000 higher for the 'wild' then R209K en-zyme).In accord, the resulting low catalytic potency of the mouse R209K thymidylate synthase was not sufficient to support growth  Anti-rat TS monoclonal antibody was used.MWM, molecular mass markers (in kDa).

MWM TSm TSw pTSw
of E. coli TX61 -cells in the Luria-Bertani medium in the absence of exogenous thymidine (not shown).

DISCUSSION
Construction of a system capable of overexpressing mouse thymidylate synthase in E. coli required optimization of the coding sequence.In accord, the poor expression of the enzyme obtained with the original coding region of the mouse gene, was improved by 17-fold (up to 5-10% of total cell protein) when the second codon was changed from CUG to CUU (Zhang et al., 1989).Also Pedersen-Lane et al. (1997) substituted G or C in the 2nd, 3rd and 5th codons with T (without changing the encoded amino-acid sequence) in human thymidylate synthase gene and obtained thymidylate synthase expression in E. coli to the extent of 25-30% of total protein.The latter approach, followed in the present studies, allowed successful expression of mouse thymidylate synthase in E. coli.
The equivalent of mouse R209 in L. casei thymidylate synthase is R218, suggested, in addition to direct contributions to dUMP phosphate binding, to coordinate the active site in the structure necessary to bind the nucleotide (Kawase et al., 2000).In accord with the results concerning mutagenetic replacement of Arg218 in the L. casei enzyme (Kawase et al., 2000), the corresponding Arg209 in the mouse enzyme appears to be of great importance for catalysis and nucleotide binding (Table 1, Fig. 1b).In fact, the consequence of the R209K mutation of the mouse enzyme, reflected by a decrease of the affinity for the substrate, appears to be even deeper than that described for the corresponding mutation (R218K) of the bacterial enzyme (Kawase et al., 2000).
The heterogeneity of enzyme subunits indicated by immunoblot analysis of both the 'wild' and mutated enzymes present in crude extracts, as well as of the purified 'wild' enzyme (Fig. 2), points to a possibility that the subunits underwent different posttranslational modifications (cf.Go³os et al., 2002).The same is suggested by the K m and V max values, reflecting higher affinity for dUMP and catalytic potency of highly purified 'wild' mouse recombinant enzyme, as compared with several highly purified thymidylate synthases originating from natural sources, including normal mouse thymus (Table 1).The sequence of the enzyme from the latter source may certainly be expected to be the same as the sequence of the recombinant thymidylate synthase.It should be mentioned that recombinant rat and human thymidylate synthases were also found to show catalytic potencies higher than those of the corresponding enzyme species (with each enzyme pair, either rat or human, characterized by the same sequence) isolated from natural sources, and with the rat enzyme the difference was correlated with the presence of an N-acetylated methionine at the NH 2 end (Cieoela et al., 1995).