Analysis of the substrate specificity of the Staphylococcus aureus sortase transpeptidase SrtA

The Staphylococcus aureus sortase transpeptidase SrtA isoform is responsible for the covalent attachment of virulence and colonization-associated proteins to the bacterial peptidoglycan. SrtA utilizes two substrates, undecaprenol-pyrophosphoryl-MurNAc(GlcNAc)-Ala-D-isoGlu-Lys(epsilon-Gly(5))-D-Ala-D-Ala (branched Lipid II) and secreted proteins containing a highly conserved C-terminal LPXTG sequence. SrtA simultaneously cleaves the Thr-Gly bond of the LPXTG-containing protein and forms a new amide bond with the nucleophilic amino group of the Gly(5) portion of branched Lipid II, anchoring the protein to this key intermediate that is subsequently polymerized into peptidoglycan. Here we describe the development of a general in vitro method for elucidating the substrate specificity of sortase enzymes. In addition, using immunofluorescence, cell adhesion assays, and transmission electron microscopy, we establish links between in vitro substrate specificity and in vivo function of the S. aureus sortase isoforms. Results from these studies provide strong supporting evidence of a primary role of the SrtA isoform in S. aureus adhesion and host colonization, illustrate a lack of specificity cross talk between SrtA and SrtB isoforms, and highlight the potential of SrtA as a target for the development of antivirulence chemotherapeutics against Gram-positive bacterial pathogens.     

Staphylococcus aureus sortase transpeptidase SrtA: insight into the kinetic mechanism and evidence for a reverse protonation catalytic mechanism

The Staphylococcus aureus transpeptidase SrtA catalyzes the covalent attachment of LPXTG-containing virulence and colonization-associated proteins to cell-wall peptidoglycan in Gram-positive bacteria. Recent structural characterizations of staphylococcal SrtA, and related transpeptidases SrtB from S. aureus and Bacillus anthracis, provide many details regarding the active site environment, yet raise questions with regard to the nature of catalysis and active site cysteine thiol activation. Here we re-evaluate the kinetic mechanism of SrtA and shed light on aspects of its catalytic mechanism. Using steady-state, pre-steady-state, bisubstrate kinetic studies, and high-resolution electrospray mass spectrometry, revised steady-state kinetic parameters and a ping-pong hydrolytic shunt kinetic mechanism were determined for recombinant SrtA. The pH dependencies of kinetic parameters k(cat)/K(m) and k(cat) for the substrate Abz-LPETG-Dap(Dnp)-NH(2) were bell-shaped with pK(a) values of 6.3 +/- 0.2 and 9.4 +/- 0.2 for k(cat) and 6.2 +/- 0.2 and 9.4 +/- 0.2 for k(cat)/K(m). Solvent isotope effect (SIE) measurements revealed inverse behavior, with a (D)2(O)k(cat) of 0.89 +/- 0.01 and a (D)2(O)(k(cat)/K(m)) of 0.57 +/- 0.03 reflecting an equilibrium SIE. In addition, SIE measurements strongly implicated Cys184 participation in the isotope-sensitive rate-determining chemical step when considered in conjunction with an inverse linear proton inventory for k(cat). Last, the pH dependence of SrtA inactivation by iodoacetamide revealed a single ionization for inactivation. These studies collectively provide compelling evidence for a reverse protonation mechanism where a small fraction (ca. 0.06%) of SrtA is competent for catalysis at physiological pH, yet is highly active with an estimated k(cat)/K(m) of >10(5) M(-)(1) s(-)(1).     

Characterization of new fluorogenic substrates for the rapid and sensitive assay of cathepsin E and cathepsin D.

Cathepsin E and cathepsin D are two major intracellular aspartic proteinases implicated in the physiological and pathological degradation of intra- and extracellular proteins. In this study, we designed and constructed highly sensitive synthetic decapeptide substrates for assays of cathepsins E and D based on the known sequence specificities of their cleavage sites. These substrates contain a highly fluorescent (7-methoxycoumarin-4-yl)acetyl (MOCAc) moiety and a quenching 2,4-dinitrophenyl (Dnp) group. When the Phe-Phe bond is cleaved, the fluorescence at an excitation wavelength of 328 nm and emission wavelength of 393 increases due to diminished quenching resulting from the separation of the fluorescent and quenching moieties. The first substrate, MOCAc-Gly-Lys-Pro-Ile-Leu-Phe-Phe-Arg-Le u-Lys(Dnp)gamma-NH2, in which the Lys-Pro combination at positions P5 and P4 was designed for specific interaction with cathepsin E, is hydrolyzed equally well by cathepsins E and D (kcat/Km = 10.9 microM(-1) x s(-1) for cathepsin E and 15.6 microM(-1) x s(-1) for cathepsin D). A very acidic pH optimum o was obtained for both enzymes. The second substrate, MOCAc-Gly-Lys-Pro-Ile-Ile-Phe-Phe-Arg-Le u-Lys(Dnp)gamma-NH2, in which the isoleucine residue at position P2 was meant to increase the specificity for cathepsin E, is also hydrolyzed equally by both enzymes (kcat/Km = 12.2 microM(-1) x s(-1) for cathepsin E and 16.3 microM(-1) x s(-1) for cathepsin D). The kcat/Km values for both substrates are greater than those for the best substrates for cathepsins E and D described so far. Unfortunately, each substrate shows little discrimination between cathepsin E and cathepsin D, suggesting that amino acids at positions far from the cleavage site are important for discrimination between the two enzymes. However, in combination with aspartic proteinase inhibitors, such as pepstatin A and Ascaris pepsin inhibitor, these substrates enable a rapid and sensitive determination of the precise levels of cathepsins E and D in crude cell extracts of various tissues and cells. Thus these substrates represent a potentially valuable tool for routine assays and for mechanistic studies on cathepsins E and D.     

Isopeptide ligation catalyzed by quintessential sortase A: mechanistic cues from cyclic and branched oligomers of indolicidin

The housekeeping transpeptidase sortase A (SrtA) from Staphyloccocus aureus catalyzes the covalent anchoring of surface proteins to the cell wall by linking the threonyl carboxylate of the LPXTG recognition motif to the amino group of the pentaglycine cross-bridge of the peptidoglycan. SrtA-catalyzed ligation of an LPXTG containing polypeptide with an aminoglycine-terminated moiety occurs efficiently in vitro and has inspired the use of this enzyme as a synthetic tool in biological chemistry. Here we demonstrate the propensity of SrtA to catalyze "isopeptide" ligation. Using model peptide sequences, we show that SrtA can transfer LPXTG peptide substrates to the ε-amine of specific Lys residues and form cyclized and/or a gamut of branched oligomers. Our results provide insights about principles governing isopeptide ligation reactions catalyzed by SrtA and suggest that although cyclization is guided by distance relationship between Lys (ε-amine) and Thr (α-carboxyl) residues, facile branched oligomerization requires the presence of a stable and long-lived acyl-enzyme intermediate.     

Evaluation of an imaging platform during the development of a FRET protease assay

Synthetic peptide substrates labeled with a fluorescent donor and quenching moiety flanking an enzyme cleavage site provide a reliable method for monitoring enzyme activity. The dye pair Mca/Dnp has been widely used for this purpose, but poor solubility characteristics, combined with fluorescence emission in the region of the spectrum associated with interference from biologicals and library compounds, can limit the usefulness of Mca/Dnp substrates in a high-throughput screening (HTS) environment. Peptide Mca-Arg-Pro-Lys-Pro-Val-Glu-Nva-Trp-Arg-Lys(Dnp)-NH(2) is a matrix-metalloproteinase 3 (MMP-3) enzyme substrate that the authors have labeled with a CyDye pair, Cy3/Cy5Q. The Mca/Dnp- and CyDye-labeled substrates were compared during the development of an MMP-3 inhibitor assay. The results obtained showed that although the peptide substrates behaved similarly throughout the development of the MMP-3 assay, during a test screen of 934 compounds randomly selected from a collection of more than 70000 compounds, the CyDye substrate was considerably more reliable. Screen Z factor values of 0.84 and 0.15 were obtained using the CyDye and Mca/Dnp peptides respectively, and the authors found that although < 1% of the test compounds were auto-fluorescent at Cy3 wavelengths, > 10% could not be screened using the Mca/Dnp substrate because of compound auto-fluorescence and interference. During this study, the authors used a PMT-based fluorescence plate reader and at the same time evaluated a charged couple device (CCD)-based imaging platform specifically optimized for use with CyDye reagents. The imaging platform gave improved read accuracy and faster plate processing times compared with the PMT reader. Overall, the results presented here highlight the potential benefit of employing the red-shifted CyDye reagents and imaging technology during the development and execution of HTS protease screens.     

Inhibition of the Staphylococcus aureus sortase transpeptidase SrtA by phosphinic peptidomimetics

During pathogenesis, Gram-positive bacteria utilize surface protein virulence factors such as the MSCRAMMs (microbial surface components recognizing adhesive matrix molecules) to aid the initiation and propagation of infection through adherence to host endothelial tissue and immune system evasion. These virulence-associated proteins generally contain a C-terminal LPXTG motif that becomes covalently anchored to the peptidoglycan biosynthesis intermediate lipid II. In Staphylococcus aureus, deletion of the sortase isoform SrtA results in marked reduction in virulence and infection potential, making it an important antivirulence target. Here we describe the chemical synthesis and kinetic characterization of a nonhydrolyzable phosphinic peptidomimetic inhibitor of SrtA derived from the LPXTG substrate sequence.     

Use of a fluorescence plate reader for measuring kinetic parameters with inner filter effect correction

A general method is presented here for the determination of the Km, kcat, and kcat/Km of fluorescence resonance energy transfer (FRET) substrates using a fluorescence plate reader. A simple empirical method for correcting for the inner filter effect is shown to enable accurate and undistorted measurements of these very important kinetic parameters. Inner filter effect corrected rates of hydrolysis of a FRET peptide substrate by hepatitis C virus (HCV) NS3 protease at various substrate concentrations enabled measurement of a Km value of 4.4 +/- 0.3 microM and kcat/Km value of 96,500 +/- 5800 M-1 s-1. These values are very close to the HPLC-determined Km value of 4.6 +/- 0.7 microM and kcat/Km value of 92,600 +/- 14,000 M-1 s-1. We demonstrate that the inner filter effect correction of microtiter plate reader velocities enables rapid measurement of Ki and Ki' values and kinetic inhibition mechanisms for HCV NS3 protease inhibitors.     

Kinetics of dipeptidyl peptidase IV proteolysis of growth hormone-releasing factor and analogs.

The kinetics and selectivity of proteolysis of synthetic human growth hormone-releasing factor and analogs by purified human placental dipeptidyl peptidase IV (DPP IV) were studied by HPLC. The initial rates of Ala2-Asp3 cleavage (pH 7.8, 37 degrees C, So = 0.15 mM) were all approx. 5 mumol min-1 mg-1 for the parent hormone, GRF(1-44)-NH2, and the fragments, GRF(1-29)-NH2 and GRF(1-20)-NH2. Lower activities observed for GRF(1-11)-OH, GRF(1-3)-OH, and cyclic lactam analogs indicate S1'-Sn' binding. Assays of [Trp6]-GRF(1-29)-NH2 versus [D-Trp6]-GFR(1-29)-NH2 indicate an S4' binding cavity. Peptides with D-configuration at P2, P1 or P1' and desNH2Tyr1 and N-MeTyr1 analogs of GRF were not cleaved. Catalytic parameters for the P1-substituted analogs [X2,Ala15]-GRF(1-29)-NH2 were found to vary with X as follows, Km: Abu less than Ala less than Pro less than Val less than Ser less than Gly much less than Leu; kcat: Pro greater than Ala greater than Abu greater than Ser greater than Gly much greater than Leu greater than Val; kcat/Km: Abu greater than Pro greater than Ala much greater than Ser greater than Gly = Val much greater than Leu. Km is at a minimum and kcat/Km at a maximum, for a hydrophobic P1 side-chain of about 0.25 nm in length, i.e., the ethyl side-chain of alpha-aminobutyric acid (Abu) is very close to optimal. These results further define the S1 selectivity of DPP IV and may be useful in the design of DPP IV resistant GRF analogs that can be produced by recombinant DNA methods and the design of DPP IV inhibitors.     

Rapid optimization of enzyme substrates using defined substrate mixtures.

A strategy is described for the rapid optimization of kcat/Km for protease substrates. Selected positions of a given peptide substrate sequence are varied through synthesis with mixtures of amino acids. Incubation of the resulting peptide mixture with the enzyme of interest and analysis by high pressure liquid chromatography provides a direct measure of analogs with enhanced kcat/Km. High performance liquid chromatography/continuous flow fast atom bombardment mass spectrometry is used to assign structure to each peak in the chromatogram. As an example of the utility and efficiency of "substrate mapping" we describe optimization of the collagenase substrate Dnp-Pro-Leu-Gly-Leu-Trp-Ala-D-Arg-NH2 (where Dnp is dinitrophenyl) at the P'1 and P'2 positions. Six different mixtures were prepared for evaluation, representing the synthesis of 128 different synthetic substrates. "Substrate mapping" has led to Dnp-Pro-Leu-Gly-Cys(Me)-His-Ala-D-Arg-NH2, a substrate that possesses a 10-fold better kcat/Km than Dnp-Pro-Leu-Gly-Leu-Trp-Ala-D-Arg-NH2.     

Staphylococcus aureus Sortase A-Mediated Incorporation of Peptides: Effect of Peptide Modification on Incorporation

The endogenous Staphylococcus aureus sortase A (SrtA) transpeptidase covalently anchors cell wall-anchored (CWA) proteins equipped with a specific recognition motif (LPXTG) into the peptidoglycan layer of the staphylococcal cell wall. Previous in situ experiments have shown that SrtA is also able to incorporate exogenous, fluorescently labelled, synthetic substrates equipped with the LPXTG motif (K(FITC)LPETG-amide) into the bacterial cell wall, albeit at high concentrations of 500 μM to 1 mM. In the present study, we have evaluated the effect of substrate modification on the incorporation efficiency. This revealed that (i) by elongation of LPETG-amide with a sequence of positively charged amino acids, derived from the C-terminal domain of physiological SrtA substrates, the incorporation efficiency was increased by 20-fold at 10 μM, 100 μM and 250 μM; (ii) Substituting aspartic acid (E) for methionine increased the incorporation of the resulting K(FITC)LPMTG-amide approximately three times at all concentrations tested; (iii) conjugation of the lipid II binding antibiotic vancomycin to K(FITC)LPMTG-amide resulted in the same incorporation levels as K(FITC)LPETG-amide, but much more efficient at an impressive 500-fold lower substrate concentration. These newly developed synthetic substrates can potentially find broad applications in for example the in situ imaging of bacteria; the incorporation of antibody recruiting moieties; the targeted delivery and covalent incorporation of antimicrobial compounds into the bacterial cell wall.     

Mutagenesis studies of substrate recognition and catalysis in the sortase A transpeptidase from Staphylococcus aureus

The Staphylococcus aureus transpeptidase sortase A (SrtA) is responsible for anchoring a range of virulence- and colonization-associated proteins to the cell wall. SrtA recognizes substrates that contain a C-terminal LPXTG motif. This sequence is cleaved following the threonine, and an amide bond is formed between the threonine and the pentaglycine cross-bridge of branched lipid II. Previous studies have implicated the beta6/beta7 loop region of SrtA in LPXTG recognition but have not systematically characterized this domain. To better understand the individual roles of the residues within this loop, we performed alanine-scanning mutagenesis. Val-168 and Leu-169 were found to be important for substrate recognition, and Glu-171 was also found to be important, consistent with its hypothesized role as a Ca(2+)-binding residue. Gly-167 and Asp-170 were dispensable for catalysis, as was Gln-172. The role of Arg-197 in SrtA has been the subject of much debate. To explore its role in catalysis, we used native chemical ligation to generate semi-synthetic SrtA in which we replaced Arg-197 with citrulline, a non-ionizable analog. This change resulted in a decrease of <3-fold in k(cat)/K(m), indicating that Arg-197 utilizes a hydrogen bond, rather than an electrostatic interaction. Our results are consistent with a model for LPXTG recognition wherein the Leu-Pro sequence is recognized primarily by hydrophobic contacts with SrtA Val-168 and Leu-169, as well as a hydrogen bond from Arg-197. This model contradicts the previously proposed mechanism of binding predicted by the x-ray crystal structure of SrtA.     

An improved cathepsin-D substrate and assay procedure

Ten analogs of the peptide A-Phe(NO2)-Phe-Val-Leu-B were synthesized and tested as substrates for cathepsin D and pepsin. The best substrate found for cathepsin D, Phe-Ala-Ala-Phe(NO2)-Phe-Val-Leu-OM4P (kcat = 2.9 s-1; Km = 7.1 microM), has the largest kcat/Km value (408 mM-1 s-1) reported to date for this enzyme. The effect of peptide structure on solubility and kinetic parameters is discussed. The peptide provides a useful new substrate for continuous assay of cathepsin D.     

Specificity of prolyl endopeptidase

A series of tetrapeptides, Cbz(Bz)-Gly-X-Leu-Gly, were synthesized and the kinetic parameters, kcat and kcat/Km, determined for their hydrolyses by prolyl endopeptidase from Flavobacterium. The peptides with X = N-Me-Ala, Sar and Ala as well as the standard substrate (X = Pro) were found to be good substrates, while those with X = alpha-aminobutyryl, Hyp, Ser and Gly were poor substrates, and those with X = pipecolyl, alpha-aminoisobutyryl, N-Me-Val, N-Me-Leu, Hyp(O-Bzl) and Ser(O-Bzl) were not cleaved at all. These results suggest that the specificity-determining site or S1 subsite of the enzyme is designed to fit exactly the proline residue of the substrate with allowance for the residues carrying substituents at the N and/or C alpha which must not exceed the size of the pyrrolidine ring of proline.     

Engineering the substrate specificity of Staphylococcus aureus Sortase A. The beta6/beta7 loop from SrtB confers NPQTN recognition to SrtA

The Staphylococcus aureus transpeptidase Sortase A (SrtA) anchors virulence and colonization-associated surface proteins to the cell wall. SrtA selectively recognizes a C-terminal LPXTG motif, whereas the related transpeptidase Sortase B (SrtB) recognizes a C-terminal NPQTN motif. In both enzymes, cleavage occurs after the conserved threonine, followed by amide bond formation between threonine and the pentaglycine cross-bridge of cell wall peptidoglycan. Genetic and biochemical studies strongly suggest that SrtA and SrtB exhibit exquisite specificity for their recognition motifs. To better understand the origins of substrate specificity within these two isoforms, we used sequence and structural analysis to predict residues and domains likely to be involved in conferring substrate specificity. Mutational analyses and domain swapping experiments were conducted to test their function in substrate recognition and specificity. Marked changes in the specificity profile of SrtA were obtained by replacing the beta6/beta7 loop in SrtA with the corresponding domain from SrtB. The chimeric beta6/beta7 loop swap enzyme (SrtLS) conferred the ability to acylate NPQTN-containing substrates, with a k(cat)/K(m)(app) of 0.0062 +/- 0.003 m(-1) s(-1). This enzyme was unable to perform the transpeptidation stage of the reaction, suggesting that additional domains are required for transpeptidation to occur. The overall catalytic specificity profile (k(cat)/K(m)(app)(NPQTN)/k(cat)/K(m)(app)(LPETG)) of SrtLS was altered 700,000-fold from SrtA. These results indicate that the beta6/beta7 loop is an important site for substrate recognition in sortases.     

Kinetic study of carboxypeptidase P-catalyzed reaction. Pressure and temperature dependence of kinetic parameters

Detailed kinetic analyses of carboxypeptidase P-catalyzed reactions were carried out spectrophotometrically using 3-(2-furyl)acryloyl-acylated peptide substrates. The maximum kcat/Km was observed at around pH 3.5 for the synthetic peptide substrates. The kcat/Km value decreased with increasing pH, with an apparent pKa value of 4.43. However, the maximum kcat was observed at neutral pH (pH congruent to 6) and the pKa was 4.49. These apparently different pH profiles for kcat/Km and kcat of this enzyme were due to the decreasing Km value in the acid pH region. The pressure and temperature dependences of these kinetic parameters were also measured. N-Benzoylglycyl-L-phenyllactate (Bz-Gly-OPhLac) gave dependences similar to those of the peptide substrate, suggesting that there is no distinct difference in the catalytic mechanism between the peptide and the ester hydrolyses.     

Characterization of SARS main protease and inhibitor assay using a fluorogenic substrate

SARS main protease is essential for life cycle of SARS coronavirus and may be a key target for developing anti-SARS drugs. Recently, the enzyme expressed in Escherichia coli was characterized using a HPLC assay to monitor the formation of products from 11 peptide substrates covering the cleavage sites found in the SARS viral genome. This protease easily dissociated into inactive monomer and the deduced Kd of the dimer was 100 microM. In order to detect enzyme activity, the assay needed to be performed at micromolar enzyme concentration. This makes finding the tight inhibitor (nanomolar range IC50) impossible. In this study, we prepared a peptide with fluorescence quenching pair (Dabcyl and Edans) at both ends of a peptide substrate and used this fluorogenic peptide substrate to characterize SARS main protease and screen inhibitors. The fluorogenic peptide gave extremely sensitive signal upon cleavage catalyzed by the protease. Using this substrate, the protease exhibits a significantly higher activity (kcat = 1.9 s(-1) and Km = 17 microM) compared to the previously reported parameters. Under our assay condition, the enzyme stays as an active dimer without dissociating into monomer and reveals a small Kd value (15 nM). This enzyme in conjunction with fluorogenic peptide substrate provides us a suitable tool for identifying potent inhibitors of SARS protease.     

Fluorescent oligopeptide substrates for kinetic characterization of the specificity of Astacus protease

The design of fluorescent N-dansylated oligopeptides based on the tubulin cleavage pattern by Astacus protease yields substrates that are turned over up to 10(5) times faster than those presently available. On the basis of this study, an optimal substrate for Astacus protease contains seven or more amino acids and minimally requires at least five amino acids. Direct examination of the formation and breakdown of the ES complex shows its formation occurs within milliseconds at 25 degrees C. The best heptapeptide substrate, Dns-Pro-Lys-Arg-Ala-Pro-Trp-Val, is cleaved only between the Arg-Ala (P1-P1') bond with kinetic parameters kcat = 380 s-1 and Km = 3.7 x 10(-4) M. The presence of Lys or Arg in the P1 and P2 positions yields high-turnover substrates. In the P3 position, the enzyme prefers Pro greater than Val greater than Leu greater than Ala greater than Gly, following the same order of preference seen in the tubulin cleavage pattern. Substitution of Leu for Ala in P1' and of Ser for Pro in P2' decreases activity by 10(5)- and 10(2)-fold, respectively. In position P3', substitution of Trp for Leu leaves the activity unaltered. However, introduction of the Trp fluorophore greatly enhances the sensitivity of the assay due to a 10-fold increase in indole fluorescence for cleavage of any peptide bond between the tryptophan and the dansyl group. Such an energy-transfer-based assay should have widespread use for detection of neutral proteases.(ABSTRACT TRUNCATED AT 250 WORDS)     

Human cathepsin H: deletion of the mini-chain switches substrate specificity from aminopeptidase to endopeptidase

The mini-chain of human cathepsin H has been identified as the major structural element determining the protease's substrate specificity. A genetically engineered mutant of human cathepsin H lacking the mini-chain, des[Glu(-18)-Thr(-11)]-cathepsin H, exhibits endopeptidase activity towards the synthetic substrate Z-Phe-Arg-NH-Mec (kcat = 0.4 s(-1), Km = 92 microM, kcat/Km = 4348 M(-1) s(-1)) which is not cleaved by r-wt cathepsin H. However, the mutant enzyme shows only minimal aminopeptidase activity for H-Arg-NH-Mec (kcat = 0.8 s(-1), Km = 3.6 mM, kcat/Km = 222 M(-1) s(-1)) which is one of the best known substrates for native human cathepsin H (kcat = 2.5 s(-1), Km = 150 microM, kcat/Km = 16666 M(-1) s(-1)). Inhibition studies with chicken egg white cystatin and E-64 suggest that the mini-chain normally restricts access of inhibitors to the active site. The kinetic data on substrates hydrolysis and enzyme inhibition point out the role of the mini-chain as a structural framework for transition state stabilization of free alpha-amino groups of substrates and as a structural barrier for endopeptidase-like substrate cleavage.     

Mutation of asparagine 229 to aspartate in thymidylate synthase converts the enzyme to a deoxycytidylate methylase.

The conserved Asn 229 of thymidylate synthase (TS) forms a cyclic hydrogen bond network with the 3-NH and 4-O of the nucleotide substrate dUMP. The Asn 229 to Asp mutant of Lactobacillus casei thymidylate synthase (TS N229D) has been prepared, purified, and investigated. Steady-state kinetic parameters of TS N229D show 3.5- and 10-fold increases in the Km values of CH2H4folate and dUMP, respectively, and a 1000-fold decrease in kcat. Most important, the Asp 229 mutation changes the substrate specificity of TS to an enzyme which recognizes and methylates dCMP in preference to dUMP. With TS N229D the Km for dCMP is bout 3-fold higher than for dUMP, and the Km for CH2H4folate is increased about 5-fold; however, the kcat for dCMP methylation is 120-fold higher than that for dUMP methylation. Specificity for dCMP versus dUMP, as measured by kcat/Km, changes from negligible with wild-type TS to about a 40-fold increase with TS N229D. TS N229D reacts with CH2H4folate and FdUMP or FdCMP to form ternary complexes which are analogous to the TS-FdUMP-CH2H4folate complex. From what is known of the mechanism and structure of TS, the dramatic change in substrate specificity of TS N229D is proposed to involve a hydrogen bond network between Asp 229 and the 3-N and 4-NH2 of the cytosine heterocycle, causing protonation of the 3-N and stabilization of a reactive imino tautomer. A similar mechanism is proposed for related enzymes which catalyze one-carbon transfers to cytosine heterocycles.