- Email: [email protected]
Site-directed mutagenesis of a potential catalytic and formyl phosphate binding site and substrate inhibition of N10-formyltetrahydrofolate synthetase
ABB Archives of Biochemistry and Biophysics 408 (2002) 137–143 www.academicpress.com
Site-directed mutagenesis of a potential catalytic and formyl phosphate binding site and substrate inhibition of N 10-formyltetrahydrofolate synthetase Adam B. Leaphart,a H. Trent Spencer,b and Charles R. Lovella,* a
Department of Biological Sciences, University of South Carolina, 715 Sumter Street, Columbia, SC 29208, USA b Department of Pediatrics, Emory University School of Medicine, Atlanta, GA 30322, USA Received 19 August 2002, and in revised form 13 September 2002
Abstract Structural studies of N 10 -formyltetrahydrofolate synthetase (FTHFS) have indicated the involvement of Arg 97 in the binding of the formyl phosphate intermediate. Two site-directed mutants were constructed to test this hypothesis: R97S (Ser substitution) and R97E (Glu substitution). The kcat of R97S was approximately 60% that of the wild-type enzyme and had Km for ATP and formate twofold higher than those of wild type. R97E was completely inactive and had a Km for ATP nearly six times that of wild type. Substrate inhibition by tetrahydrofolate was shown to occur in wild-type and R97S enzymes using both steady-state and transientstate kinetic approaches. These results lend greater insight into the mechanistic function of FTHFS by confirming the interaction of both ATP and formate with Arg 97 and introducing the aspect of substrate inhibition by tetrahydrofolate with regard to substrate binding and dissociation. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: N 10 -Formyltetrahydrofolate synthetase; Formyl phosphate intermediate; Ligand binding; Substrate inhibition
Tetrahydrofolate (H4 folate)1 and its pteroylpolyglutamate derivatives constitute the biologically active forms of folic acid. These compounds play a fundamental role in metabolism by acting as specialized cosubstrates for a wide range of enzymatic reactions involving C1 units. Folate is primarily responsible for bringing formate into the C1 pool. This is accomplished through the ATP-dependent formylation of H4 folate catalyzed by N 10 -formyltetrahydrofolate synthetase (FTHFS) (EC 6.3.4.3). The product, N 10 -formyltetrahydrofolate, is then either directly utilized in the biosynthesis of fMet-tRNAfMet and purines or dehydrated by 5,10-methenyltetrahydrofolate cyclohydrolase and reduced by 5,10-methylenetetrahydrofolate dehydrogenase for use in biosynthesis of amino acids and pyrimidines [1] or acetate in acetogenic bacteria [2]. In *
Corresponding author. Fax: 1-803-777-4002. E-mail address: [email protected] (C.R. Lovell). 1 Abbreviations used: H4 folate, tetrahydrofolate; FTHFS, N 10 -formyltetrahydrofolate synthetase.
eukaryotes, the FTHFS, cyclohydrolase, and dehydrogenase activities are found in one trifunctional enzyme, C1 -tetrahydrofolate synthase [3,4]. FTHFS is ubiquitous in Eucarya and Bacteria. By far the highest levels have been seen in the acetogenic bacteria that synthesize acetyl-CoA via the Wood-Ljungdahl (acetyl-CoA) pathway of autotrophic C1 fixation [2,5–7] and in the purine-fermenting bacteria that synthesize acetate via the glycine synthase/glycine reductase pathway [5,8–10]. FTHFS enzymes from both acetogenic and purinolytic bacterial sources [11,12] have very similar physical, chemical, kinetic, and genetic characteristics [13,14]. All bacterial enzymes studied to date are homotetrameric, with a Mr ¼ 240,000, and have four identical substrate binding sites (reviewed in [1,15]). Previous studies utilizing steady-state kinetics and isotope exchange reactions [13,16–18] have suggested that catalysis by FTHFS proceeds via a random sequential mechanism. Results of Himes and Rabinowitz [16] were consistent with the participation of a nondissociable, tightly bound intermediate. Buttlaire et al.
0003-9861/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 3 - 9 8 6 1 ( 0 2 ) 0 0 5 5 2 - 0
138
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
[19,20] showed that FTHFS could catalyze the phosphorylation of ADP in the presence of carbamoyl phosphate and H4 folate. Carbamoyl phosphate is a potent inhibitor of the forward reaction catalyzed by FTHFS. Additionally, it was shown that formate and phosphate could inhibit ATP synthesis in the reverse reaction. Given that various ligase enzymes have been shown to chemically activate substrates via nucleotide triphosphate dependent phosphorylation, the inference could be made that formyl phosphate could be the enzyme-bound intermediate of the naturally catalyzed reaction. More recently, Smithers et al. [21] showed that chemically synthesized formyl phosphate could support both the forward and the reverse reactions by formylating H4 folate and phosphorylating ADP, respectively. It was ultimately demonstrated by Mejillano and coworkers [22] that formyl phosphate could act as a kinetically competent intermediate of the reaction and that FTHFS could form this putative intermediate through a designated ‘‘formate kinase’’ activity. Based on data obtained from the three-dimensional structure of FTHFS [23], it has been hypothesized that Arg 97 is involved in binding the formyl phosphate intermediate. In this study, we have constructed sitedirected mutants of the Arg 97 residue and examined them kinetically. In the process of this analysis, incidental observation of substrate inhibition by H4 folate was observed.
Materials and methods Materials A crystalline, racemic mix of [6R,S]-H4 folate was purchased from Sigma (St. Louis, MO, USA) and purified using DEAE cellulose (Sigma) chromatography with a solution of 50 mM Tris, pH 7.0, and 40 mM 2mercaptoethanol as the eluent. Purity of [6S]-H4 folate was determined by calculating the ratio of absorbance at 290 versus 245 nm. A ratio of 3.5 or greater indicated a pure fraction. Additionally, an enzymatically prepared stock of [6S]-H4 folate was used for comparison purposes. HPLC was performed on both H4 folate stock solutions to confirm purity using the method of McMartin et al. [24]. Site-directed mutagenesis The gene encoding Moorella thermoacetica FTHFS [14] was subcloned into the phagemid pAlter-1 (Promega, Madison, WI, USA) as described previously [25,26]. Site-directed mutations of the FTHFS gene were introduced at the Arg 97 position using the Altered Sites II in vitro mutagenesis system (Promega). Mutagenic oligonucleotides coding for Ser and Glu in place of Arg
97 were synthesized by Cyber Syn (Aston, PA, USA). The resulting mutants were designated R97S and R97E, respectively. The mutations were confirmed by two rounds of bidirectional sequencing. Purification of wild-type and mutant FTHFS Wild-type FTHFS, R97S, and R97E were purified using a modification of the procedure described by Staben et al. [27] given in Radfar et al. [26]. Activity was assayed by spectrophotometric analysis of product formation [28] for wild-type and R97S FTHFS. R97E was inactive and was purified in parallel to wild type and R97S. Presence and purity of R97E, as well as wild type and R97S, were confirmed using native and SDS–PAGE [29]. The molarity of concentrated, pure enzyme was determined by use of the Lowry et al. [30] and Bio-Rad (Hercules, CA, USA) assays. The results were compared for accuracy against the molar extinction coefficient at 280 nm of 1:80 105 M1 cm1 . Kinetic characterization of wild-type FTHFS—steady-state determinations
and
R97S
kcat , Km , and Vmax were determined for H4 folate, ATP, and formate. Substrate concentrations ranged from 100 to 1000 lM for H4 folate, 30–5000 lM for ATP, and 500–10,000 lM for formate. Reactions were performed at 27 °C and product formation was quantified spectrophotometrically using the assay of Shoaf et al. [28]. Reaction velocities were plotted against substrate concentration and the values for Km and kcat determined. Experiments were conducted in triplicate. Steady-state analysis of H4 folate employed nonlinear least-squares fitting of the data. Kinetic characterization of wild-type FTHFS—transient-state determinations
and
R97S
Values for the Kd and kchem of H4 folate were determined for wild type and R97S using an RQF-3 quenchflow instrument (KinTek, Austin, TX, USA). The chemical quench-flow apparatus allows the mixing of two reactants for a controlled length of time after which a quenching agent is mixed with the reactants to denature the enzyme and liberate any enzyme-bound intermediates or products. The three syringes of the quench-flow apparatus were filled with: (1) reaction components including enzyme, but excluding H4 folate; (2) 1.12 M HCl, as the chemical quencher; and (3) H4 folate diluted in potassium maleate buffer, pH 7.0 [28], with concentrations of 0.4, 0.8, 1.6, 3.2, and 6.4 M [6S]-H4 folate. Equal volumes of the contents of syringes 1 and 3 were combined and allowed to react for a standardized length of time. After this time, a volume of HCl equal to the total volume of reactants was added to
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
139
stop the reaction. Reaction products were collected in duplicate and then quantified spectrophotometrically. After collection, data were fitted by nonlinear leastsquares analysis using Kaleidagraph version 3.0.9 (Synergy Software, Reading, PA, USA). Equilibrium dialysis of wild-type and R97E FTHFS Purified enzyme and 1:85 105 Bq of [2,8-3 H]ATP (ICN, Irvine, CA, USA) were combined with 100 mM potassium maleate buffer, pH 8.0, 1 mM NH4 Cl, 5 mM MgCl2 , and 4.6 mM ATP and brought to a volume of 500 ll. This mixture was dialyzed against 230 ml of a solution of 100 mM potassium maleate buffer, pH 8.0, 1 mM NH4 Cl, and 5 mM MgCl2 for 2.5 h using 0.5– 3.0 ml capacity 10,000 MWCO Slide-a-Lyzer dialysis cassettes (Pierce Biochemical, Rockford, IL, USA). Upon completion of dialysis, the contents of the dialysis cassettes were withdrawn and the volume was measured [31,32]. The contents were added to Ultima Gold scintillation cocktail (Packard Bioscience, Boston, MA, USA) and radiolabel was quantified using a Packard Bioscience Model 1500 liquid scintillation counter. Additionally, a 500-ll aliquot of dialysis buffer was removed before and after dialysis and radiolabel quantified in the same manner. The duration of dialysis was established by dialyzing 1:85 105 Bq of [2,83 H]ATP under the same conditions listed above, excluding enzyme. Aliquots of the contents of the dialysis chamber were taken at half-hour intervals and radioactivity was quantified. The dialysis reached equilibrium after 2 h. Additionally, the membrane was removed from this dialysis cassette and radiolabel quantified to establish that binding to the membrane was negligible. Two additional experiments were run combining pure enzyme, ATP, formate, and all reaction buffer components or pure enzyme, ATP, H4 folate, and all reaction buffer components. Each was processed identical to the experiments containing only ATP. All experiments were run simultaneously, in quadruplicate.
Fig. 1. Effect of concentration of tetrahydrofolate (H4 folate) on the rate of N 10 -formyltetrahydrofolate production by wild-type and R97S N 10 -formyltetrahydrofolate synthetase. The dashed line indicates data fitted to the simple Michaelis–Menten equation. The solid line indicates data fitted to Eq. (1).
the experiments were repeated to match the temperature used in subsequent single turnover experiments. To verify that the inhibition shown in these experiments was not an artifact resulting from some competitive inhibitor occurring as an impurity in the H4 folate stocks, HPLC analysis of the H4 folate stock solutions was performed. Both H4 folate stock solutions, whether prepared enzymatically or chemically, proved to have only negligible contamination by other folate species (data not shown). Substrate inhibition prevents the accurate determination of Km or kcat using steady-state kinetic techniques. That is, two values can be obtained for each of these constants, one that occurs with the reaction path prevailing at low substrate concentrations (Km and kcat ) and one that occurs with the reaction path prevailing at 0 0 higher substrate concentrations (Km and kcat ). The data in Fig. 1 were fitted to the following equation for substrate inhibition,
Results
v=½E ¼ m1 S þ m2 S2 =1 þ m3 S þ m4 S2 ;
Substrate inhibition of H4 folate and single turnover experiments
where S equals the substrate concentration, v equals reaction velocity, ½E equals enzyme concentration, and m1 , m2 , m3 , and m4 are constants that are functions of the rate constants in the reaction pathway [33]. As previously derived, the value of m1 equals kcat =Km and the 0 value of m2 =m4 equals kcat . For the wild-type enzyme these values were 0.00511 and 0.00206, respectively, and for the R97S mutant they were 0.00182 and 0.00124, respectively. To more accurately determine kinetic constants governing ligand binding, single turnover experiments were conducted using quench-flow techniques. Transient-state kinetic experiments were performed
When steady-state reaction velocity was determined as a function of H4 folate concentration, both wild type and the R97S mutant showed noticeable substrate inhibition. The data fit poorly to simple Michaelis– Menten kinetics (Fig. 1), but fit well to an equation incorporating substrate inhibition (Eq. (1)). Inhibition was seen in initial experiments performed at 50 °C (data not shown), the reported optimum temperature for M. thermoacetica FTHFS [14], and confirmed at 25 °C when
ð1Þ
140
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
using high enzyme concentrations that allowed accurate quantification of the concentration of enzyme–substrate species. The burst phase in transient-state kinetics is representative of the formation of products, formyltetrahydrofolate and ADP þ Pi , during the first round of catalysis and prior to their dissociation. After the burst phase, a linear increase in product formation is measurable until substrates are depleted. In order to evaluate the burst phase preceding steady state, kburst , the first-order rate constant governing this phase was calculated using the equation Aðm0 Þ ¼ m1 þ m2 ð1 e ðm3 m0 ÞÞ þ m4 m0 ;
ð2Þ
where m0 is time, Aðm0 Þ is product (absorbance at 350 nm) at time m0 , m1 is product (absorbance at 350 nm) at time ¼ 0, m2 is the amplitude of the burst phase, m3 is kburst , and m4 is the linear rate of increase in product (absorbance at 350 nm) during steady state [34]. The amplitude, or m2 , is directly proportional to the concentration of enzyme under the assumption that substrate binding is not limiting and release of product is slower than formation of the enzyme–product complex [35]. This held true with minor variation for each concentration of H4 folate used. Fig. 2 is a representative plot of the data sets collected at various concentrations of H4 folate showing absorbance of product at 350 nm versus time of reaction and represents the least-squares fit of the data to Eq. (2). The values obtained for kburst were used to determine the Kd for dissociation of H4 folate from the enzyme– substrate complex as well as the value of kchem , which is the rate constant for the chemical transformation step. The calculated values for kburst were plotted against H4 folate concentration and the data fitted to the following equation [33]:
Fig. 3. Effect of tetrahydrofolate (H4 folate) concentration on the conversion of enzyme-bound substrates to enzyme-bound products by R97S and wild-type N 10 -formyltetrahydrofolate synthetase.
kburst ¼ ðkchem ½H4 folateÞ=Kd þ ½H4 folate:
ð3Þ
This equation was used under the assumption that kon and koff for H4 folate are fast compared to kchem of the reaction and that the dissociation of ATP and formate from the enzyme–substrate complex is not appreciable. Fig. 3 shows the hyperbolic dependence of kburst on H4 folate as described by the fit of the data to Eq. (3). From these fits the values of Kd and kchem were determined for both wild-type ð219 50 lM H4 folate and 2:96
0:158 s1 , respectively) and R97S FTHFS (285 63 lM H4 folate and 1:98 0:115 s1 , respectively). In addition, these values were compared to those from the steadystate data using the portion of the data occurring before the start of the inhibition [35,36]. Eadie–Hofstee and Lineweaver–Burk plots yielded Km and kcat values comparable to the values seen for the Kd and kcat for the transient-state data (see Table 1). However, the kcat for wild type and R97S determined from these studies is somewhat lower than the kchem seen from the transientstate data. Steady-state analysis for substrates ATP and formate
Fig. 2. Time course of product formation when wild-type N 10 -formyltetrahydrofolate synthetase was mixed with 200 lM tetrahydrofolate. The enzyme concentration was 1.25 lM.
Steady-state analysis of R97S and wild-type FTHFS was completed for both ATP and formate. Inhibition kinetics were not observed with either substrate, and Km and kcat values were determined using Michaelis– Menten kinetics. Studies were also attempted with R97E, but the mutant enzyme was found to be inactive, i.e., no measurable product was generated at 27 or 50 °C. Comparing wild type and R97S, a drastic difference was observed for the Km and kcat for both ATP and formate, although only minor differences were observed for the Km for H4 folate (Table 1). For R97S, the Km of
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
141
Table 1 Steady-state Km and kcat values for wild-type and R97S N 10 -formyltetrahydrofolate synthetase Kd a H4 folated
Km b H4 folate L–Bf
Km H4 folate E–Hg
Km ATP
Km formate
Wild type R97S
219 50 285 63
233 10 296 12
248 25 260 29
87 4 156 27
459 73 951 177
Wild type R97S
kcat c H4 folatee 1:12 0:124 0:52 0:036
kcat H4 folate L–Bf 1:41 0:068 0:603 0:036
kcat H4 folate E–Hg 1:436 0:073 0:566 0:031
kcat ATP 0:996 0:007 0:391 0:015
kcat formate 1:125 0:034 0:398 0:020
a
Kd in units of lM of substrate. Km in units of lM of substrate. c kcat in units of ðs1 Þ. d Single turnover experimental values. e kcat value calculated using m1 from Eq. (1). f Lineweaver–Burk fit. g Eadie–Hofstee fit. b
ATP and formate are roughly twofold that of the wildtype enzyme. Additionally, the wild-type FTHFS kcat is roughly threefold that of R97S. The values established here for the kcat of both mutant and wild-type FTHFS are lower than those previously reported due to incubation of assays at 27 °C rather than 50 °C, which was required for kburst measurements. Equilibrium dialysis In order to examine the nature of the inactivity seen with R97E, stopped-flow experiments were initiated with wild-type and mutant FTHFS. However, no quenching of enzyme fluorescence could be detected, likely due to the high value of koff . Therefore, equilibrium dialysis experiments were performed using wildtype and R97E FTHFS in order to examine the behavior of ligand binding. The Kd for ATP was determined using the equation Kd ¼ ð½ET ð½LT ½Lf ÞÞ½Lf =ð½LT ½Lf Þ;
phosphate formation compared to the overall rate constant for the reaction is examined, the participation of this intermediate in the normal reaction pathway of FTHFS is not certain [22]. However, both the presence of a formate kinase activity of FTHFS and participation of formyl phosphate as a kinetically competent, enzymebound intermediate in the reaction have all but indicated such. In this set of experiments, we sought to establish the participation of a specific residue, Arg 97, in binding the formyl phosphate intermediate. As shown in Fig. 4, structural data have indicated that the side chain of Arg 97 has the appropriate spacing and tangency to the triphosphate moiety to accommodate such an interaction [23]. Kounga et al. [37], using site-directed mutagenesis, have shown the Lys 71 residue in Clostridium
ð4Þ
where ½ET is the total concentration of enzyme active sites, ½LT is the total ligand concentration, bound and free, and ½Lf is the free ligand concentration. This equation was used under the assumption that saturation of a single class of active site occurs, as is the case for FTHFS [31,32]. The calculated values of Kd for ATP using wild-type and R97E FTHFS were 75 and 421 lM, respectively, corresponding to the value obtained for the Km of ATP for wild-type FTHFS (Table 1). Combinations of substrates (i.e., ATP and formate or ATP and H4 folate) yielded only a slight, almost equivalent increase in the amount of ATP bound.
Discussion The participation of formyl phosphate in the normal reaction pathway of FTHFS has not been conclusively shown to occur. When the rate constant for formyl
Fig. 4. Hypothetical interaction of formyl phosphate in the active site of N 10 -formyltetrahydrofolate synthetase. ATP is shown in its interaction with Lys 74.
142
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
cylindrosporum FTHFS (Lys 74 in the M. thermoacetica enzyme, Fig. 4) to be involved in catalysis. This finding, in concert with active site modeling showing a presumptive stabilizing effect on the triphosphate moiety of ATP by Lys 74, strengthens the hypothesis that Arg 97 is an important residue interacting with the formyl phosphate intermediate. Mutagenesis has helped to establish the involvement of this residue in interactions with both ATP and formate. Kinetic analysis of R97S, a neutrally charged substitution with a significantly shorter side chain, showed an approximate 60% decrease in kcat and a twofold increase in Km for both ATP and formate with respect to wild type. Since no repulsion of the phosphate or formate could occur due to the absence of charge of the side chain, catalysis could still occur. However, catalysis proceeded at approximately one-third that of wild type, due primarily to the decreased stabilization or incorrect orientation of ATP and formate in the active site, as shown by the reduced affinity of the R97S mutant FTHFS for both. R97E FTHFS, an opposite charge substitution with approximately equivalent side chain length, was completely inactive. Equilibrium dialysis experiments showed there to be an almost sixfold increase in Kd for ATP with R97E compared to the Km value observed in wild-type enzyme. Clearly, a repulsion of both phosphate and formate could occur as a result of this substitution. Repulsion of either substrate, individually, could also play a role in the inactivity of this mutant. However, several pieces of evidence taken together seem to refute an effect based solely on ATP. The position of bound sulfate ion in crystallized enzyme [23] is from the Arg 97 side chain and it partially overlaps 2.6 A the triphosphate moiety. The absence of the positively charged side chain of arginine in R97S produced a significant disruption of catalytic activity as well as a twofold lowered affinity for both ATP and formate. Substitution of a negatively charged glutamic acid residue in place of the positively charged arginine residue produced a sixfold increase in the Kd of ATP versus wildtype enzyme in equilibrium dialysis experiments. Comparing the Kd obtained from these experiments for R97E to the Km obtained for ATP from R97S in steady-state experiments, there is a threefold increase in Km versus R97S, just by restoring the length of the side chain and changing the net charge to negative. The Kd for formate reasonably could not be much out of proportion for R97E. Considering all of this, it is proposed that Arg 97 is neither solely a formate-binding nor an ATP-binding residue, but is one of the key residues involved with enzyme-bound formyl phosphate in the active site. Although substrate inhibition of H4 folate was not initially a focus of this study, its well-established presence could have some bearing on the actual reaction mechanism for FTHFS. Both an ordered [16] and a random [13,17] sequential mechanism have been pro-
posed for FTHFS thus far. Substrate inhibition can occur due to either of two circumstances associated with increased levels of substrate concentration, nonproductive substrate binding and abortive complex formation. The binding site for substrate is more often than not composed of several subsites that each recognize a specific structural feature of the substrate molecule. Nonproductive binding involves the binding of more than one substrate to different recognition subsites in the active site, thus producing a catalytically inactive complex. Abortive complex formation occurs in a multisubstrate enzyme system in which one of the substrates resembles the product. Depending on the order of product release during catalysis, enzyme with simultaneously bound product and substrate may occur. If product is released from the enzyme–product–substrate complex more slowly than from the enzyme–product complex, then the overall reaction rate is slowed. However, substrate has to have greater affinity for the enzyme–product complex versus enzyme alone. FTHFS has been shown to undergo a conformational change upon binding of H4 folate [38,39] that could potentially accommodate the presence of another H4 folate molecule in the active site, thus promoting nonproductive binding. However, the steric hindrances involved in allowing two H4 folate molecules into the active site would more than likely prevent such an occurrence from happening. Based on product inhibition studies [13,17], the observed slow dissociation of ADP [22], and the relative similarity of H4 folate to product, it is proposed that abortive complex formation occurs. Orders of substrate arrival and binding thus far have not been shown to definitively be ordered or random, but it would seem that the dissociation of product does in fact occur in a defined order. Complementing the results of Mejillano et al. [22] showing that formyl phosphate is synthesized and bound by FTHFS, this study has established that Arg 97 is one of the amino acid residues that is responsible for binding the intermediate. The presence of all three substrates has been shown to be obligatory and a sequential mechanism has been suggested [13,17,18,22]. This mechanism occurs through nucleophilic attack by the carbonyl of formate on the cphosphate of ATP to produce formyl phosphate. Nucleophilic attack by N 10 of H4 folate yields N 10 -formyltetrahydrofolate. The discovery of inhibition by H4 folate may suggest an ordered release of product, which could affect the randomness of substrate arrival in the active site. With the knowledge garnered by this study, greater insight can be lent to establishing and understanding the catalytic mechanism of FTHFS.
Acknowledgments This research was supported by National Science Foundation Grant MCB-9873606.
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
References [1] R.E. Mackenzie, in: R.L. Blakely, S.J. Benkovic (Eds.), Folates and Pterins, vol. 1, Wiley-Interscience, New York, 1984, pp. 255–306. [2] L.G. Ljungdahl, Annu. Rev. Microbiol. 40 (1986) 415–450. [3] D.W. Hum, A.W. Bell, R. Rozen, R.E. MacKenzie, J. Biol. Chem. 263 (1988) 15946–15950. [4] T.R. Whitehead, J.C. Rabinowitz, J. Bacteriol. 170 (1988) 3255– 3261. [5] G. Fuchs, FEMS Microbiol. Rev. 39 (1986) 181–213. [6] A.B. Leaphart, C.R. Lovell, Appl. Environ. Microbiol. 67 (2001) 1392–1395. [7] H.G. Wood, L.G. Ljungdahl, in: L.L. Barton, J. Shively (Eds.), Variations in Autotrophic Life, Academic Press, San Diego, CA, 1991, pp. 201–250. [8] L.G. Ljungdahl, in: R.L. Blakely, S.J. Benkovic (Eds.), Folates and Pterins, vol. 1, Wiley-Interscience, New York, 1984, pp. 555–579. [9] H. Schiefer-Ulrich, R. Wagner, P. Durre, J.R. Andreesen, Arch. Microbiol. 138 (1984) 345–353. [10] L.J. Waber, H.G. Wood, J. Bacteriol. 140 (1979) 468–478. [11] L.G. Ljungdahl, J.M. Brewer, S.H. Neece, T. Fairwell, J. Biol. Chem. 245 (1970) 4791–4797. [12] J.C. Rabinowitz, W.E. Pricer Jr., J. Biol. Chem. 237 (1962) 2898–2925. [13] B.K. Joyce, R.H. Himes, J. Biol. Chem. 241 (1966) 5725–5731. [14] W.E. OÕBrien, J.M. Brewer, L.G. Ljungdahl, in: H. Zuber (Ed.), Enzymes and Proteins from Thermophilic Microorganisms. Experentia Suppl, Birkhauser-Verlag, Basel and Stuttgart, 1976, pp. 249–262. [15] R.H. Himes, J.A.K. Harmony, CRC Crit. Rev. Biochem. 1 (1973) 501–535. [16] R.H. Himes, J.C. Rabinowitz, J. Biol. Chem. 237 (1962) 2903–2914. [17] B.K. Joyce, R.H. Himes, J. Biol. Chem. 241 (1966) 5716–5724. [18] J.J. McGuire, J.C. Rabinowitz, J. Biol. Chem. 253 (1978) 1079–1085. [19] D.H. Buttlaire, C.A. Balfe, M.F. Wendland, R.H. Himes, Biochim. Biophys. Acta 567 (1979) 453–463.
143
[20] D.H. Buttlaire, R.H. Himes, G.H. Reed, J. Biol. Chem. 251 (1976) 4159–4161. [21] G.W. Smithers, H. Jahansouz, J.L. Kofron, R.H. Himes, G.H. Reed, Biochemistry 26 (1987) 3943–3948. [22] M.R. Mejillano, H. Jahansouz, T.O. Matsunaga, G.L. Kenyon, R.H. Himes, Biochemistry 28 (1989) 5136–5145. [23] R. Radfar, R. Shin, G.M. Sheldrick, W. Minor, C.R. Lovell, J.D. Odom, R.B. Dunlap, L. Lebioda,Biochemistry 39 (2000) 3920–3926. [24] K.E. McMartin, V. Virayotha, T.R. Tephly, Arch. Biochem. Biophys. 209 (1981) 127–136. [25] C.R. Lovell, A. Przybyla, L.G. Ljungdahl, Arch. Microbiol. 149 (1988) 280–285. [26] R. Radfar, A.B. Leaphart, J.M. Brewer, W. Minor, J.D. Odom, R.B. Dunlap, C.R. Lovell, L. Lebioda, Biochemistry 39 (2000) 14481–14486. [27] C. Staben, T.R. Whitehead, J.C. Rabinowitz, Anal. Biochem. 162 (1987) 257–264. [28] W.T. Shoaf, S.H. Neece, L.G. Ljungdahl, Biochim. Biophys. Acta 334 (1974) 448–458. [29] U.K. Laemmli, Nature 227 (1970) 680–685. [30] O.H. Lowry, N.J. Rosebrough, A.L. Farr, R.J. Randall, J. Biol. Chem. 193 (1951) 265–275. [31] J.E. Bell, E.T. Bell, Proteins and Enzymes, Prentice-Hall Inc., Englewood Cliffs, NJ, 1988. [32] D. Freifelder, Physical Biochemistry: Applications to Biochemistry and Molecular Biology, W.H. Freeman and Company, San Francisco, CA, 1982. [33] T. Nakano, H.T. Spencer, J.R. Appleman, R.L. Blakley, Biochemistry 33 (1994) 9945–9952. [34] H.T. Spencer, J.E. Villafranca, J.R. Appleman, Biochemistry 36 (1997) 4212–4222. [35] A. Cornish-Bowden, Fundamentals of Enzyme Kinetics, Portland Press Ltd., London, 1995. [36] M. Dixon, E.C. Webb, Enzymes, Academic Press Inc., New York, 1964. [37] K. Kounga, S. Song, G.C. Haslam, R.H. Himes, Arch. Biochem. Biophys. 1296 (1996) 112–120. [38] D.H. Buttlaire, G.H. Reed, R.H. Himes, J. Biol. Chem. 250 (1975) 254–260. [39] R.H. Himes, M. Cohn, J. Biol. Chem. 242 (1967) 3628–3635.
Site-directed mutagenesis of a potential catalytic and formyl phosphate binding site and substrate inhibition of N 10-formyltetrahydrofolate synthetase Adam B. Leaphart,a H. Trent Spencer,b and Charles R. Lovella,* a
Department of Biological Sciences, University of South Carolina, 715 Sumter Street, Columbia, SC 29208, USA b Department of Pediatrics, Emory University School of Medicine, Atlanta, GA 30322, USA Received 19 August 2002, and in revised form 13 September 2002
Abstract Structural studies of N 10 -formyltetrahydrofolate synthetase (FTHFS) have indicated the involvement of Arg 97 in the binding of the formyl phosphate intermediate. Two site-directed mutants were constructed to test this hypothesis: R97S (Ser substitution) and R97E (Glu substitution). The kcat of R97S was approximately 60% that of the wild-type enzyme and had Km for ATP and formate twofold higher than those of wild type. R97E was completely inactive and had a Km for ATP nearly six times that of wild type. Substrate inhibition by tetrahydrofolate was shown to occur in wild-type and R97S enzymes using both steady-state and transientstate kinetic approaches. These results lend greater insight into the mechanistic function of FTHFS by confirming the interaction of both ATP and formate with Arg 97 and introducing the aspect of substrate inhibition by tetrahydrofolate with regard to substrate binding and dissociation. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: N 10 -Formyltetrahydrofolate synthetase; Formyl phosphate intermediate; Ligand binding; Substrate inhibition
Tetrahydrofolate (H4 folate)1 and its pteroylpolyglutamate derivatives constitute the biologically active forms of folic acid. These compounds play a fundamental role in metabolism by acting as specialized cosubstrates for a wide range of enzymatic reactions involving C1 units. Folate is primarily responsible for bringing formate into the C1 pool. This is accomplished through the ATP-dependent formylation of H4 folate catalyzed by N 10 -formyltetrahydrofolate synthetase (FTHFS) (EC 6.3.4.3). The product, N 10 -formyltetrahydrofolate, is then either directly utilized in the biosynthesis of fMet-tRNAfMet and purines or dehydrated by 5,10-methenyltetrahydrofolate cyclohydrolase and reduced by 5,10-methylenetetrahydrofolate dehydrogenase for use in biosynthesis of amino acids and pyrimidines [1] or acetate in acetogenic bacteria [2]. In *
Corresponding author. Fax: 1-803-777-4002. E-mail address: [email protected] (C.R. Lovell). 1 Abbreviations used: H4 folate, tetrahydrofolate; FTHFS, N 10 -formyltetrahydrofolate synthetase.
eukaryotes, the FTHFS, cyclohydrolase, and dehydrogenase activities are found in one trifunctional enzyme, C1 -tetrahydrofolate synthase [3,4]. FTHFS is ubiquitous in Eucarya and Bacteria. By far the highest levels have been seen in the acetogenic bacteria that synthesize acetyl-CoA via the Wood-Ljungdahl (acetyl-CoA) pathway of autotrophic C1 fixation [2,5–7] and in the purine-fermenting bacteria that synthesize acetate via the glycine synthase/glycine reductase pathway [5,8–10]. FTHFS enzymes from both acetogenic and purinolytic bacterial sources [11,12] have very similar physical, chemical, kinetic, and genetic characteristics [13,14]. All bacterial enzymes studied to date are homotetrameric, with a Mr ¼ 240,000, and have four identical substrate binding sites (reviewed in [1,15]). Previous studies utilizing steady-state kinetics and isotope exchange reactions [13,16–18] have suggested that catalysis by FTHFS proceeds via a random sequential mechanism. Results of Himes and Rabinowitz [16] were consistent with the participation of a nondissociable, tightly bound intermediate. Buttlaire et al.
0003-9861/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 3 - 9 8 6 1 ( 0 2 ) 0 0 5 5 2 - 0
138
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
[19,20] showed that FTHFS could catalyze the phosphorylation of ADP in the presence of carbamoyl phosphate and H4 folate. Carbamoyl phosphate is a potent inhibitor of the forward reaction catalyzed by FTHFS. Additionally, it was shown that formate and phosphate could inhibit ATP synthesis in the reverse reaction. Given that various ligase enzymes have been shown to chemically activate substrates via nucleotide triphosphate dependent phosphorylation, the inference could be made that formyl phosphate could be the enzyme-bound intermediate of the naturally catalyzed reaction. More recently, Smithers et al. [21] showed that chemically synthesized formyl phosphate could support both the forward and the reverse reactions by formylating H4 folate and phosphorylating ADP, respectively. It was ultimately demonstrated by Mejillano and coworkers [22] that formyl phosphate could act as a kinetically competent intermediate of the reaction and that FTHFS could form this putative intermediate through a designated ‘‘formate kinase’’ activity. Based on data obtained from the three-dimensional structure of FTHFS [23], it has been hypothesized that Arg 97 is involved in binding the formyl phosphate intermediate. In this study, we have constructed sitedirected mutants of the Arg 97 residue and examined them kinetically. In the process of this analysis, incidental observation of substrate inhibition by H4 folate was observed.
Materials and methods Materials A crystalline, racemic mix of [6R,S]-H4 folate was purchased from Sigma (St. Louis, MO, USA) and purified using DEAE cellulose (Sigma) chromatography with a solution of 50 mM Tris, pH 7.0, and 40 mM 2mercaptoethanol as the eluent. Purity of [6S]-H4 folate was determined by calculating the ratio of absorbance at 290 versus 245 nm. A ratio of 3.5 or greater indicated a pure fraction. Additionally, an enzymatically prepared stock of [6S]-H4 folate was used for comparison purposes. HPLC was performed on both H4 folate stock solutions to confirm purity using the method of McMartin et al. [24]. Site-directed mutagenesis The gene encoding Moorella thermoacetica FTHFS [14] was subcloned into the phagemid pAlter-1 (Promega, Madison, WI, USA) as described previously [25,26]. Site-directed mutations of the FTHFS gene were introduced at the Arg 97 position using the Altered Sites II in vitro mutagenesis system (Promega). Mutagenic oligonucleotides coding for Ser and Glu in place of Arg
97 were synthesized by Cyber Syn (Aston, PA, USA). The resulting mutants were designated R97S and R97E, respectively. The mutations were confirmed by two rounds of bidirectional sequencing. Purification of wild-type and mutant FTHFS Wild-type FTHFS, R97S, and R97E were purified using a modification of the procedure described by Staben et al. [27] given in Radfar et al. [26]. Activity was assayed by spectrophotometric analysis of product formation [28] for wild-type and R97S FTHFS. R97E was inactive and was purified in parallel to wild type and R97S. Presence and purity of R97E, as well as wild type and R97S, were confirmed using native and SDS–PAGE [29]. The molarity of concentrated, pure enzyme was determined by use of the Lowry et al. [30] and Bio-Rad (Hercules, CA, USA) assays. The results were compared for accuracy against the molar extinction coefficient at 280 nm of 1:80 105 M1 cm1 . Kinetic characterization of wild-type FTHFS—steady-state determinations
and
R97S
kcat , Km , and Vmax were determined for H4 folate, ATP, and formate. Substrate concentrations ranged from 100 to 1000 lM for H4 folate, 30–5000 lM for ATP, and 500–10,000 lM for formate. Reactions were performed at 27 °C and product formation was quantified spectrophotometrically using the assay of Shoaf et al. [28]. Reaction velocities were plotted against substrate concentration and the values for Km and kcat determined. Experiments were conducted in triplicate. Steady-state analysis of H4 folate employed nonlinear least-squares fitting of the data. Kinetic characterization of wild-type FTHFS—transient-state determinations
and
R97S
Values for the Kd and kchem of H4 folate were determined for wild type and R97S using an RQF-3 quenchflow instrument (KinTek, Austin, TX, USA). The chemical quench-flow apparatus allows the mixing of two reactants for a controlled length of time after which a quenching agent is mixed with the reactants to denature the enzyme and liberate any enzyme-bound intermediates or products. The three syringes of the quench-flow apparatus were filled with: (1) reaction components including enzyme, but excluding H4 folate; (2) 1.12 M HCl, as the chemical quencher; and (3) H4 folate diluted in potassium maleate buffer, pH 7.0 [28], with concentrations of 0.4, 0.8, 1.6, 3.2, and 6.4 M [6S]-H4 folate. Equal volumes of the contents of syringes 1 and 3 were combined and allowed to react for a standardized length of time. After this time, a volume of HCl equal to the total volume of reactants was added to
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
139
stop the reaction. Reaction products were collected in duplicate and then quantified spectrophotometrically. After collection, data were fitted by nonlinear leastsquares analysis using Kaleidagraph version 3.0.9 (Synergy Software, Reading, PA, USA). Equilibrium dialysis of wild-type and R97E FTHFS Purified enzyme and 1:85 105 Bq of [2,8-3 H]ATP (ICN, Irvine, CA, USA) were combined with 100 mM potassium maleate buffer, pH 8.0, 1 mM NH4 Cl, 5 mM MgCl2 , and 4.6 mM ATP and brought to a volume of 500 ll. This mixture was dialyzed against 230 ml of a solution of 100 mM potassium maleate buffer, pH 8.0, 1 mM NH4 Cl, and 5 mM MgCl2 for 2.5 h using 0.5– 3.0 ml capacity 10,000 MWCO Slide-a-Lyzer dialysis cassettes (Pierce Biochemical, Rockford, IL, USA). Upon completion of dialysis, the contents of the dialysis cassettes were withdrawn and the volume was measured [31,32]. The contents were added to Ultima Gold scintillation cocktail (Packard Bioscience, Boston, MA, USA) and radiolabel was quantified using a Packard Bioscience Model 1500 liquid scintillation counter. Additionally, a 500-ll aliquot of dialysis buffer was removed before and after dialysis and radiolabel quantified in the same manner. The duration of dialysis was established by dialyzing 1:85 105 Bq of [2,83 H]ATP under the same conditions listed above, excluding enzyme. Aliquots of the contents of the dialysis chamber were taken at half-hour intervals and radioactivity was quantified. The dialysis reached equilibrium after 2 h. Additionally, the membrane was removed from this dialysis cassette and radiolabel quantified to establish that binding to the membrane was negligible. Two additional experiments were run combining pure enzyme, ATP, formate, and all reaction buffer components or pure enzyme, ATP, H4 folate, and all reaction buffer components. Each was processed identical to the experiments containing only ATP. All experiments were run simultaneously, in quadruplicate.
Fig. 1. Effect of concentration of tetrahydrofolate (H4 folate) on the rate of N 10 -formyltetrahydrofolate production by wild-type and R97S N 10 -formyltetrahydrofolate synthetase. The dashed line indicates data fitted to the simple Michaelis–Menten equation. The solid line indicates data fitted to Eq. (1).
the experiments were repeated to match the temperature used in subsequent single turnover experiments. To verify that the inhibition shown in these experiments was not an artifact resulting from some competitive inhibitor occurring as an impurity in the H4 folate stocks, HPLC analysis of the H4 folate stock solutions was performed. Both H4 folate stock solutions, whether prepared enzymatically or chemically, proved to have only negligible contamination by other folate species (data not shown). Substrate inhibition prevents the accurate determination of Km or kcat using steady-state kinetic techniques. That is, two values can be obtained for each of these constants, one that occurs with the reaction path prevailing at low substrate concentrations (Km and kcat ) and one that occurs with the reaction path prevailing at 0 0 higher substrate concentrations (Km and kcat ). The data in Fig. 1 were fitted to the following equation for substrate inhibition,
Results
v=½E ¼ m1 S þ m2 S2 =1 þ m3 S þ m4 S2 ;
Substrate inhibition of H4 folate and single turnover experiments
where S equals the substrate concentration, v equals reaction velocity, ½E equals enzyme concentration, and m1 , m2 , m3 , and m4 are constants that are functions of the rate constants in the reaction pathway [33]. As previously derived, the value of m1 equals kcat =Km and the 0 value of m2 =m4 equals kcat . For the wild-type enzyme these values were 0.00511 and 0.00206, respectively, and for the R97S mutant they were 0.00182 and 0.00124, respectively. To more accurately determine kinetic constants governing ligand binding, single turnover experiments were conducted using quench-flow techniques. Transient-state kinetic experiments were performed
When steady-state reaction velocity was determined as a function of H4 folate concentration, both wild type and the R97S mutant showed noticeable substrate inhibition. The data fit poorly to simple Michaelis– Menten kinetics (Fig. 1), but fit well to an equation incorporating substrate inhibition (Eq. (1)). Inhibition was seen in initial experiments performed at 50 °C (data not shown), the reported optimum temperature for M. thermoacetica FTHFS [14], and confirmed at 25 °C when
ð1Þ
140
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
using high enzyme concentrations that allowed accurate quantification of the concentration of enzyme–substrate species. The burst phase in transient-state kinetics is representative of the formation of products, formyltetrahydrofolate and ADP þ Pi , during the first round of catalysis and prior to their dissociation. After the burst phase, a linear increase in product formation is measurable until substrates are depleted. In order to evaluate the burst phase preceding steady state, kburst , the first-order rate constant governing this phase was calculated using the equation Aðm0 Þ ¼ m1 þ m2 ð1 e ðm3 m0 ÞÞ þ m4 m0 ;
ð2Þ
where m0 is time, Aðm0 Þ is product (absorbance at 350 nm) at time m0 , m1 is product (absorbance at 350 nm) at time ¼ 0, m2 is the amplitude of the burst phase, m3 is kburst , and m4 is the linear rate of increase in product (absorbance at 350 nm) during steady state [34]. The amplitude, or m2 , is directly proportional to the concentration of enzyme under the assumption that substrate binding is not limiting and release of product is slower than formation of the enzyme–product complex [35]. This held true with minor variation for each concentration of H4 folate used. Fig. 2 is a representative plot of the data sets collected at various concentrations of H4 folate showing absorbance of product at 350 nm versus time of reaction and represents the least-squares fit of the data to Eq. (2). The values obtained for kburst were used to determine the Kd for dissociation of H4 folate from the enzyme– substrate complex as well as the value of kchem , which is the rate constant for the chemical transformation step. The calculated values for kburst were plotted against H4 folate concentration and the data fitted to the following equation [33]:
Fig. 3. Effect of tetrahydrofolate (H4 folate) concentration on the conversion of enzyme-bound substrates to enzyme-bound products by R97S and wild-type N 10 -formyltetrahydrofolate synthetase.
kburst ¼ ðkchem ½H4 folateÞ=Kd þ ½H4 folate:
ð3Þ
This equation was used under the assumption that kon and koff for H4 folate are fast compared to kchem of the reaction and that the dissociation of ATP and formate from the enzyme–substrate complex is not appreciable. Fig. 3 shows the hyperbolic dependence of kburst on H4 folate as described by the fit of the data to Eq. (3). From these fits the values of Kd and kchem were determined for both wild-type ð219 50 lM H4 folate and 2:96
0:158 s1 , respectively) and R97S FTHFS (285 63 lM H4 folate and 1:98 0:115 s1 , respectively). In addition, these values were compared to those from the steadystate data using the portion of the data occurring before the start of the inhibition [35,36]. Eadie–Hofstee and Lineweaver–Burk plots yielded Km and kcat values comparable to the values seen for the Kd and kcat for the transient-state data (see Table 1). However, the kcat for wild type and R97S determined from these studies is somewhat lower than the kchem seen from the transientstate data. Steady-state analysis for substrates ATP and formate
Fig. 2. Time course of product formation when wild-type N 10 -formyltetrahydrofolate synthetase was mixed with 200 lM tetrahydrofolate. The enzyme concentration was 1.25 lM.
Steady-state analysis of R97S and wild-type FTHFS was completed for both ATP and formate. Inhibition kinetics were not observed with either substrate, and Km and kcat values were determined using Michaelis– Menten kinetics. Studies were also attempted with R97E, but the mutant enzyme was found to be inactive, i.e., no measurable product was generated at 27 or 50 °C. Comparing wild type and R97S, a drastic difference was observed for the Km and kcat for both ATP and formate, although only minor differences were observed for the Km for H4 folate (Table 1). For R97S, the Km of
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
141
Table 1 Steady-state Km and kcat values for wild-type and R97S N 10 -formyltetrahydrofolate synthetase Kd a H4 folated
Km b H4 folate L–Bf
Km H4 folate E–Hg
Km ATP
Km formate
Wild type R97S
219 50 285 63
233 10 296 12
248 25 260 29
87 4 156 27
459 73 951 177
Wild type R97S
kcat c H4 folatee 1:12 0:124 0:52 0:036
kcat H4 folate L–Bf 1:41 0:068 0:603 0:036
kcat H4 folate E–Hg 1:436 0:073 0:566 0:031
kcat ATP 0:996 0:007 0:391 0:015
kcat formate 1:125 0:034 0:398 0:020
a
Kd in units of lM of substrate. Km in units of lM of substrate. c kcat in units of ðs1 Þ. d Single turnover experimental values. e kcat value calculated using m1 from Eq. (1). f Lineweaver–Burk fit. g Eadie–Hofstee fit. b
ATP and formate are roughly twofold that of the wildtype enzyme. Additionally, the wild-type FTHFS kcat is roughly threefold that of R97S. The values established here for the kcat of both mutant and wild-type FTHFS are lower than those previously reported due to incubation of assays at 27 °C rather than 50 °C, which was required for kburst measurements. Equilibrium dialysis In order to examine the nature of the inactivity seen with R97E, stopped-flow experiments were initiated with wild-type and mutant FTHFS. However, no quenching of enzyme fluorescence could be detected, likely due to the high value of koff . Therefore, equilibrium dialysis experiments were performed using wildtype and R97E FTHFS in order to examine the behavior of ligand binding. The Kd for ATP was determined using the equation Kd ¼ ð½ET ð½LT ½Lf ÞÞ½Lf =ð½LT ½Lf Þ;
phosphate formation compared to the overall rate constant for the reaction is examined, the participation of this intermediate in the normal reaction pathway of FTHFS is not certain [22]. However, both the presence of a formate kinase activity of FTHFS and participation of formyl phosphate as a kinetically competent, enzymebound intermediate in the reaction have all but indicated such. In this set of experiments, we sought to establish the participation of a specific residue, Arg 97, in binding the formyl phosphate intermediate. As shown in Fig. 4, structural data have indicated that the side chain of Arg 97 has the appropriate spacing and tangency to the triphosphate moiety to accommodate such an interaction [23]. Kounga et al. [37], using site-directed mutagenesis, have shown the Lys 71 residue in Clostridium
ð4Þ
where ½ET is the total concentration of enzyme active sites, ½LT is the total ligand concentration, bound and free, and ½Lf is the free ligand concentration. This equation was used under the assumption that saturation of a single class of active site occurs, as is the case for FTHFS [31,32]. The calculated values of Kd for ATP using wild-type and R97E FTHFS were 75 and 421 lM, respectively, corresponding to the value obtained for the Km of ATP for wild-type FTHFS (Table 1). Combinations of substrates (i.e., ATP and formate or ATP and H4 folate) yielded only a slight, almost equivalent increase in the amount of ATP bound.
Discussion The participation of formyl phosphate in the normal reaction pathway of FTHFS has not been conclusively shown to occur. When the rate constant for formyl
Fig. 4. Hypothetical interaction of formyl phosphate in the active site of N 10 -formyltetrahydrofolate synthetase. ATP is shown in its interaction with Lys 74.
142
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
cylindrosporum FTHFS (Lys 74 in the M. thermoacetica enzyme, Fig. 4) to be involved in catalysis. This finding, in concert with active site modeling showing a presumptive stabilizing effect on the triphosphate moiety of ATP by Lys 74, strengthens the hypothesis that Arg 97 is an important residue interacting with the formyl phosphate intermediate. Mutagenesis has helped to establish the involvement of this residue in interactions with both ATP and formate. Kinetic analysis of R97S, a neutrally charged substitution with a significantly shorter side chain, showed an approximate 60% decrease in kcat and a twofold increase in Km for both ATP and formate with respect to wild type. Since no repulsion of the phosphate or formate could occur due to the absence of charge of the side chain, catalysis could still occur. However, catalysis proceeded at approximately one-third that of wild type, due primarily to the decreased stabilization or incorrect orientation of ATP and formate in the active site, as shown by the reduced affinity of the R97S mutant FTHFS for both. R97E FTHFS, an opposite charge substitution with approximately equivalent side chain length, was completely inactive. Equilibrium dialysis experiments showed there to be an almost sixfold increase in Kd for ATP with R97E compared to the Km value observed in wild-type enzyme. Clearly, a repulsion of both phosphate and formate could occur as a result of this substitution. Repulsion of either substrate, individually, could also play a role in the inactivity of this mutant. However, several pieces of evidence taken together seem to refute an effect based solely on ATP. The position of bound sulfate ion in crystallized enzyme [23] is from the Arg 97 side chain and it partially overlaps 2.6 A the triphosphate moiety. The absence of the positively charged side chain of arginine in R97S produced a significant disruption of catalytic activity as well as a twofold lowered affinity for both ATP and formate. Substitution of a negatively charged glutamic acid residue in place of the positively charged arginine residue produced a sixfold increase in the Kd of ATP versus wildtype enzyme in equilibrium dialysis experiments. Comparing the Kd obtained from these experiments for R97E to the Km obtained for ATP from R97S in steady-state experiments, there is a threefold increase in Km versus R97S, just by restoring the length of the side chain and changing the net charge to negative. The Kd for formate reasonably could not be much out of proportion for R97E. Considering all of this, it is proposed that Arg 97 is neither solely a formate-binding nor an ATP-binding residue, but is one of the key residues involved with enzyme-bound formyl phosphate in the active site. Although substrate inhibition of H4 folate was not initially a focus of this study, its well-established presence could have some bearing on the actual reaction mechanism for FTHFS. Both an ordered [16] and a random [13,17] sequential mechanism have been pro-
posed for FTHFS thus far. Substrate inhibition can occur due to either of two circumstances associated with increased levels of substrate concentration, nonproductive substrate binding and abortive complex formation. The binding site for substrate is more often than not composed of several subsites that each recognize a specific structural feature of the substrate molecule. Nonproductive binding involves the binding of more than one substrate to different recognition subsites in the active site, thus producing a catalytically inactive complex. Abortive complex formation occurs in a multisubstrate enzyme system in which one of the substrates resembles the product. Depending on the order of product release during catalysis, enzyme with simultaneously bound product and substrate may occur. If product is released from the enzyme–product–substrate complex more slowly than from the enzyme–product complex, then the overall reaction rate is slowed. However, substrate has to have greater affinity for the enzyme–product complex versus enzyme alone. FTHFS has been shown to undergo a conformational change upon binding of H4 folate [38,39] that could potentially accommodate the presence of another H4 folate molecule in the active site, thus promoting nonproductive binding. However, the steric hindrances involved in allowing two H4 folate molecules into the active site would more than likely prevent such an occurrence from happening. Based on product inhibition studies [13,17], the observed slow dissociation of ADP [22], and the relative similarity of H4 folate to product, it is proposed that abortive complex formation occurs. Orders of substrate arrival and binding thus far have not been shown to definitively be ordered or random, but it would seem that the dissociation of product does in fact occur in a defined order. Complementing the results of Mejillano et al. [22] showing that formyl phosphate is synthesized and bound by FTHFS, this study has established that Arg 97 is one of the amino acid residues that is responsible for binding the intermediate. The presence of all three substrates has been shown to be obligatory and a sequential mechanism has been suggested [13,17,18,22]. This mechanism occurs through nucleophilic attack by the carbonyl of formate on the cphosphate of ATP to produce formyl phosphate. Nucleophilic attack by N 10 of H4 folate yields N 10 -formyltetrahydrofolate. The discovery of inhibition by H4 folate may suggest an ordered release of product, which could affect the randomness of substrate arrival in the active site. With the knowledge garnered by this study, greater insight can be lent to establishing and understanding the catalytic mechanism of FTHFS.
Acknowledgments This research was supported by National Science Foundation Grant MCB-9873606.
A.B. Leaphart et al. / Archives of Biochemistry and Biophysics 408 (2002) 137–143
References [1] R.E. Mackenzie, in: R.L. Blakely, S.J. Benkovic (Eds.), Folates and Pterins, vol. 1, Wiley-Interscience, New York, 1984, pp. 255–306. [2] L.G. Ljungdahl, Annu. Rev. Microbiol. 40 (1986) 415–450. [3] D.W. Hum, A.W. Bell, R. Rozen, R.E. MacKenzie, J. Biol. Chem. 263 (1988) 15946–15950. [4] T.R. Whitehead, J.C. Rabinowitz, J. Bacteriol. 170 (1988) 3255– 3261. [5] G. Fuchs, FEMS Microbiol. Rev. 39 (1986) 181–213. [6] A.B. Leaphart, C.R. Lovell, Appl. Environ. Microbiol. 67 (2001) 1392–1395. [7] H.G. Wood, L.G. Ljungdahl, in: L.L. Barton, J. Shively (Eds.), Variations in Autotrophic Life, Academic Press, San Diego, CA, 1991, pp. 201–250. [8] L.G. Ljungdahl, in: R.L. Blakely, S.J. Benkovic (Eds.), Folates and Pterins, vol. 1, Wiley-Interscience, New York, 1984, pp. 555–579. [9] H. Schiefer-Ulrich, R. Wagner, P. Durre, J.R. Andreesen, Arch. Microbiol. 138 (1984) 345–353. [10] L.J. Waber, H.G. Wood, J. Bacteriol. 140 (1979) 468–478. [11] L.G. Ljungdahl, J.M. Brewer, S.H. Neece, T. Fairwell, J. Biol. Chem. 245 (1970) 4791–4797. [12] J.C. Rabinowitz, W.E. Pricer Jr., J. Biol. Chem. 237 (1962) 2898–2925. [13] B.K. Joyce, R.H. Himes, J. Biol. Chem. 241 (1966) 5725–5731. [14] W.E. OÕBrien, J.M. Brewer, L.G. Ljungdahl, in: H. Zuber (Ed.), Enzymes and Proteins from Thermophilic Microorganisms. Experentia Suppl, Birkhauser-Verlag, Basel and Stuttgart, 1976, pp. 249–262. [15] R.H. Himes, J.A.K. Harmony, CRC Crit. Rev. Biochem. 1 (1973) 501–535. [16] R.H. Himes, J.C. Rabinowitz, J. Biol. Chem. 237 (1962) 2903–2914. [17] B.K. Joyce, R.H. Himes, J. Biol. Chem. 241 (1966) 5716–5724. [18] J.J. McGuire, J.C. Rabinowitz, J. Biol. Chem. 253 (1978) 1079–1085. [19] D.H. Buttlaire, C.A. Balfe, M.F. Wendland, R.H. Himes, Biochim. Biophys. Acta 567 (1979) 453–463.
143
[20] D.H. Buttlaire, R.H. Himes, G.H. Reed, J. Biol. Chem. 251 (1976) 4159–4161. [21] G.W. Smithers, H. Jahansouz, J.L. Kofron, R.H. Himes, G.H. Reed, Biochemistry 26 (1987) 3943–3948. [22] M.R. Mejillano, H. Jahansouz, T.O. Matsunaga, G.L. Kenyon, R.H. Himes, Biochemistry 28 (1989) 5136–5145. [23] R. Radfar, R. Shin, G.M. Sheldrick, W. Minor, C.R. Lovell, J.D. Odom, R.B. Dunlap, L. Lebioda,Biochemistry 39 (2000) 3920–3926. [24] K.E. McMartin, V. Virayotha, T.R. Tephly, Arch. Biochem. Biophys. 209 (1981) 127–136. [25] C.R. Lovell, A. Przybyla, L.G. Ljungdahl, Arch. Microbiol. 149 (1988) 280–285. [26] R. Radfar, A.B. Leaphart, J.M. Brewer, W. Minor, J.D. Odom, R.B. Dunlap, C.R. Lovell, L. Lebioda, Biochemistry 39 (2000) 14481–14486. [27] C. Staben, T.R. Whitehead, J.C. Rabinowitz, Anal. Biochem. 162 (1987) 257–264. [28] W.T. Shoaf, S.H. Neece, L.G. Ljungdahl, Biochim. Biophys. Acta 334 (1974) 448–458. [29] U.K. Laemmli, Nature 227 (1970) 680–685. [30] O.H. Lowry, N.J. Rosebrough, A.L. Farr, R.J. Randall, J. Biol. Chem. 193 (1951) 265–275. [31] J.E. Bell, E.T. Bell, Proteins and Enzymes, Prentice-Hall Inc., Englewood Cliffs, NJ, 1988. [32] D. Freifelder, Physical Biochemistry: Applications to Biochemistry and Molecular Biology, W.H. Freeman and Company, San Francisco, CA, 1982. [33] T. Nakano, H.T. Spencer, J.R. Appleman, R.L. Blakley, Biochemistry 33 (1994) 9945–9952. [34] H.T. Spencer, J.E. Villafranca, J.R. Appleman, Biochemistry 36 (1997) 4212–4222. [35] A. Cornish-Bowden, Fundamentals of Enzyme Kinetics, Portland Press Ltd., London, 1995. [36] M. Dixon, E.C. Webb, Enzymes, Academic Press Inc., New York, 1964. [37] K. Kounga, S. Song, G.C. Haslam, R.H. Himes, Arch. Biochem. Biophys. 1296 (1996) 112–120. [38] D.H. Buttlaire, G.H. Reed, R.H. Himes, J. Biol. Chem. 250 (1975) 254–260. [39] R.H. Himes, M. Cohn, J. Biol. Chem. 242 (1967) 3628–3635.