Roles of Asp75, Asp78, and Glu83 of GTP-dependent Phosphoenolpyruvate Carboxykinase from Mycobacterium smegmatis*

The roles of Asp75, Asp78, and Glu83 of the 75DPSDVARVE83 element of Mycobacterium smegmatis GTP-dependent phosphoenolpyruvate (PEP) carboxykinase (GTP-PEPCK) were investigated. Asp78 and Glu83 are fully conserved in GTP-PEP-CKs. The human PEPCK crystal structure suggests that Asp78 influences Tyr220; Tyr220 helps to position bound PEP, and Glu83 interacts with Arg81. Experimental data on other PEPCKs indicate that Arg81 binds PEP, and the phosphate of PEP interacts with Mn2+ of metal site 1 for catalysis. We found that D78A and E83A replacements severely reduced activity. E83A substitution raised the apparent Km value for Mn2+ 170-fold. In contrast, Asp75 is highly but not fully conserved; natural substitutions are Ala, Asn, Gln, or Ser. Such substitutions, when engineered, in M. smegmatis enzyme caused the following. 1) For oxaloacetate synthesis, Vmax decreased 1.4-4-fold. Km values for PEP and Mn2+ increased 3-9- and 1.2-10-fold, respectively. Km values for GDP and bicarbonate changed little. 2) For PEP formation, Vmax increased 1.5-2.7-fold. Km values for oxaloacetate increased 2-2.8-fold. The substitutions did not change the secondary structure of protein significantly. The kinetic effects are rationalized as follows. In E83A the loss of Glu83-Arg81 interaction affected Arg81-PEP association. D78A change altered the Tyr220-PEP interaction. These events perturbed PEP-Mn2+ interaction and consequently affected catalysis severely. In contrast, substitutions at Asp75, a site far from bound PEP, brought subtle effects, lowering oxaloacetate formation rate but enhancing PEP formation rate. It is likely that Asp75 substitutions affected PEP-Mn2+ interaction by changing the positions of Asp78, Arg81, and Glu83, which translated to differential effects on two directions.


EXPERIMENTAL PROCEDURES
Strains and Growth Conditions-E. coli DH5␣ (36) was used as a cloning host, and E. coli C41(DE3) (37) was used for the overexpression of PEPCK. These strains were grown in Luria-Bertani (LB) medium. The E. coli transformants were selected on plates or grown in liquid medium containing 100 g of ampicillin/ml. DNA Techniques-Standard techniques were used for the manipulation of the DNA (38). Qiagen QIAprep spin miniprep kit (Qiagen Inc., Valencia, CA) was used for the purification of plasmids. The sequence for the mutagenized DNA was verified by determining the sequence for both strands. Site-directed mutagenesis experiments were performed by using the QuikChange kit from Stratagene (La Jolla, CA) and a pair of mutagenic oligonucleotides. The oligonucleotides were purchased from Integrated DNA Technologies (Coralville, IA).
Site-directed Mutagenesis of M. smegmatis PEPCK-The desired alterations were introduced via two sets of experiments. In the first set, Asp 75 was the target and the plasmid pBE129-19b, an expression vector for M. smegmatis PEPCK (2), was used as the template. pBE129-19b was constructed by cloning the pck coding sequence as an NdeI-BamHI fragment into pET19b (2). The primer pairs used in this first set of experiments (Table 2) were designed to cause the following two changes in the pck coding sequence: 1) replacement of the codon representing Asp 75 with that for Ala, Asn, Gln, or Ser; 2) introduction of a PmlI site at nucleotide position 239 -244 without causing any change in the amino acid sequence. pBE129-19b does not contain a PmlI site. The newly introduced PmlI site provided a rapid screening method for plasmids containing the designed mutations. Upon digestion with BamHI and PmlI, a mutant plasmid produced two fragments of 1.5 and 6 kb, whereas this treatment generated only a 7.5-kb fragment from pBE129-19b. From these experiments the plasmids pBE129-19bD75A, pBE129-19bD75N, pBE129-19bD75Q, and pBE129-19bD75S were obtained. In the second set of mutagenesis experiments, either Asp 78 or Glu 83 of the PEPCK protein was targeted. For this work the template was pBE129-19bD75A, which carried a PmlI site and encoded an Ala residue at the 75th amino acid sequence position of the enzyme ( Table  2). Each primer pair used in the second round restored the wild-type sequences that were altered in the first round and replaced the codon for either Asp 78 or Glu 83 with that for Ala (Table 2). A loss of the PmlI site was taken as a preliminary indication for the success of a mutagenesis experiment. The plasmids pBE129-19bD78A and pBE129-19bE83A were generated from this round. For each of six mutant clones, the DNA sequence for both strands of the entire pck gene was determined. The data showed that in every case the mutant gene carried only the desired alterations, and the rest of the DNA sequence remained unchanged.
Expression and Purification of Recombinant Enzymes-From the above-mentioned constructs, the wild-type and variant M.   (2,37). The recombinant proteins were purified as described previously (2), except the steps beyond nickel-nitrilotriacetic acid chromatography were eliminated. As judged from SDS-PAGE analysis with Coomassie Blue staining (39), the protein preparations obtained from this modified method were homogeneous.
Enzyme and Protein Assays and Analysis of the Enzyme Kinetics Data-Protein was assayed according to Bradford (40) using the dye reagent from Bio-Rad. PEPCK activity was determined as described previously (2), but with modifications as indicated in the tables and figures, and the following are some of the details. The OAA forming or anaplerotic activity was measured by coupling it with the malate dehydrogenase (MDH) reaction and monitoring NADH oxidation spectrophotometrically at 340 nm (11). The assays for the PEP forming or gluconeogenic activity were conducted in a mixture where OAA generated in situ from L-malate by MDH served as the substrate (41,42). As OAA was consumed by PEPCK, MDH produced more OAA from L-malate to maintain equilibrium (41,42). The consequent formation of NADH was monitored spectrophotometrically. The standard assay mixture for measuring the OAA-forming activity contained 100 mM HEPES-NaOH buffer, pH 7.2, 200 mM KHCO 3 , 10 mM PEP, 2 mM GDP, 2 mM MgCl 2 , 0.2 mM MnCl 2 , 37 mM dithiothreitol, 0.25 mM NADH, and 2 units of MDH in a total volume of 1 ml. For the PEP-forming reaction, the standard assay mixture contained 100 mM HEPES-NaOH buffer, pH 7.2, 3 mM L-malate, 0.2 mM GTP, 3 mM NAD ϩ , 2 mM MgCl 2 , 0.2 mM MnCl 2 , 37 mM dithiothreitol, and 6 units of MDH in a total volume of 1 ml. The L-malate consistently provided an initial OAA concentration of 14 M. The assay temperature for both directions was 37°C. All initial rate data were analyzed according to Cleland (43) using the KinDist, a PC graphics program obtained from Prof. Bryce V. Plapp, University of Iowa (Iowa City). Each data set fit to the standard Henri-Michaelis-Menten equation well. The k cat values were calculated from V max data, considering that the subunit molecular mass of the monomeric enzyme is 71.2 kDa as calculated from the nucleic acid sequence-derived amino acid sequence (2).
CD Spectroscopy-The CD spectra of purified PEPCK proteins were collected between 190 and 260 nm at 0.5-nm intervals by using a Jasco J710 spectropolarimeter (Easton, MD) and a 0.1-cm cylindrical quartz cuvette. In each case, a solution containing 2.8 M PEPCK protein (calculated based on a subunit molecular mass of 71.2 kDa, which was derived as indicated in the preceding paragraph (2)) and 10 mM potassium phosphate buffer, pH 7, was analyzed, and the data from nine acquisitions were averaged. Each averaged spectrum was corrected for the contribution by the buffer. The measured ellipticity data ( values in millidegrees) were converted into mean residue ellipticity or [] MRE values (degrees square centimeters/ dmol) according to the following relationship: [] MRE ϭ /10nCl, where l is the cuvette path length in centimeters; C is the molar concentration of monomeric PEPCK subunits, and n is the number of residues per PEPCK subunit (644 amino acids, including the His tag).
Amino Acid Sequence Alignment and Analysis of Crystal Structure-A multiple alignment of the primary amino acid sequences for GTP-dependent PEPCKs (   Wild-type sequence of the relevant region of M. smegmatis PEPCK with the areas targeted for mutagenesis are underlined is as follows: 5Ј-CTTTCCGATCCGTC-CGACGTCGCGCGTGTCGAATCCCGCACCTTC-3Ј. c First round of mutagenesis is as follows: underlined bases show a PmlI restriction site that was introduced by changing one base (G to A, shown in boldface) without altering the amino acid sequence; white letters on a black background represent the changes providing desired alterations in the amino acid sequence. d Second round of mutagenesis is as follows: white letters on a gray background indicate restoration of the wild-type sequence; white letters on a black background represent the changes providing desired alterations in the amino acid sequence.

Alanine-scanning
mM. The D75A enzyme was active, and the values of the corresponding specific OAA-forming and PEP-forming activities at a Mn 2ϩ concentration of 0.2 mM were about 25 and 126% of that of the wild-type or Asp 75 enzyme, respectively (Table 3). A more detailed study on the kinetic properties of the D75A variant is presented below. The D78A and E83A variants exhibited very low activity in either the OAA-or the PEP-forming direc-tion (Table 3). With Mn 2ϩ at 0.2 mM, the specific OAA-forming activities of D78A and E83A were 0.3 and 0.9% that of the wild-type enzyme, respectively (Table 3). These values increased to 0.5 and 3.4%, respectively, when the Mn 2ϩ concentration in the assay was raised to 2 mM ( Table 3). The D78A and E83A enzymes were impaired in the PEP forming activity as well, and the corresponding specific activities at a Mn 2ϩ concentration of 0.2 mM were 0.4 and 2.3% that of the wild type, respectively (Table 3). An increase in Mn 2ϩ concentration beyond 0.2 mM did not improve the PEP forming activities of D78A and E83A PEPCK (data not shown). Substrate Kinetics for the OAA and PEP Forming Activities of E83A Variant of M. smegmatis PEPCK-Assays for OAA forming activity were run at approximately saturating conditions for the substrates that were held at fixed concentrations. Assay mixtures were standard except for Mn 2ϩ , which was provided at a concentration of 2.0 mM. For the PEP-forming assays, two sets of conditions were used. Because the assay used did not allow for establishing a saturating concentration for OAA, the GTP kinetics were performed at a subsaturating OAA level of 14 M. The initial velocity data at varying concentrations of OAA (1-14 M) were collected at a GTP concentration of 0.2 mM. Each initial rate data set fit the standard Henri-Michaelis-Menten kinetics equation (Fig. 2), and the values of the kinetic constants obtained from the fits are shown in Table 4. The apparent K m values for PEP and GDP of E83A were not significantly different from that of Glu 83 enzyme. A similar conclusion could also be made for the K m value for bicarbonate (Table 4), if one considers the fact that the measured values had large errors because of low activities. The apparent V max FIGURE 2. Substrate kinetics of wild-type (F) and E83A (छ) M. smegmatis PEPCK for the OAA and PEP synthesis activities. The left axis represents data for the wild-type enzyme (F), and the right axis is for E83A (छ). For each study, concentration of the respective substrate was varied. The concentrations for other compounds were standard, except that for the OAA-forming assays the concentrations of PEP and MnCl 2 were 4 and 2 mM, respectively. A, Mn 2ϩ kinetics; A (inset), expanded version for wild type; B, PEP kinetics; C, GDP kinetics; D, bicarbonate kinetics; E, OAA kinetics. The line drawn through the points for each data set is the best fit to the hyperbola v ϭ V max ϫ S/(K m ϩ S).  (Table 4). When the MgCl 2 concentration was raised from 2 to 5 mM, the apparent K m of E83A for Mn 2ϩ dropped by 17-fold to 33 Ϯ 3 M. In contrast, there was no change in the apparent K m value of wild-type PEPCK for this rise in Mg 2ϩ concentration. An increase in the MgCl 2 level from 5 to 7.5 mM did not change the apparent K m value for Mn 2ϩ of either E83A or wild-type PEPCK significantly. Also, a change in the Mg 2ϩ concentration from 2 to 5 or 7.5 mM did not alter the V max values for E83A or Glu 83 enzyme. Similarly, the apparent K m value for GDP was determined at fixed MgCl 2 concentrations of 5 and 7.5 mM and compared with that obtained at 2 mM MgCl 2 . For these studies, the MnCl 2 concentration was held constant at a saturating level of 0.8 mM. The apparent K m value of E83A for GDP at 5 mM MgCl 2 (82 Ϯ 6 M) was 1.5-fold lower than at 2 mM MgCl 2 (121 Ϯ 12 M). The corresponding drop for the wild-type enzyme was 1.4-fold (from 69 Ϯ 9 to 49 Ϯ 5 M). An increase in the MgCl 2 concentration from 5 to 7.5 mM raised the apparent K m value for GDP of the Glu 83 and E83A enzymes 1.1-and 1.08-fold (to 55 Ϯ 4 and 89 Ϯ 8 M), respectively.
Mn 2ϩ Requirements of M. smegmatis PEPCK Asp 75 Variants-The assays were conducted in the OAA-forming direction. The concentration of Mn 2ϩ was varied although that of PEP, GDP, and added bicarbonate was held constant at saturating levels of 10, 2, and 200 mM, respectively. The apparent values of K m , V max , and k cat /K m for Mn 2ϩ derived from the Henri-Michaelis-Menten fits (Fig. 3A) are shown in Table 5. The apparent K m value for Mn 2ϩ of the Asp 75 or wild-type PEPCK was 2.4 Ϯ 0.3 M, and a 3.5-fold higher value was observed for the D75N and D75S variants. It increased 9-fold when Asp 75 was replaced with Ala, but a substitution with Gln did not cause a significant change. The engineered alterations reduced the catalytic efficiency of the enzyme significantly (Table 5).
Substrate Kinetics for the OAA Forming Activity of M. smegmatis PEPCK Asp 75 Variants-The replacement of Asp 75 with Ala, Asn, Gln, or Ser significantly reduced the apparent maximum specific activity or V max of M. smegmatis PEPCK in the anaplerotic or OAA-forming direction (Fig. 3, B-D, and Table 5). In studies where the concentrations of Mn 2ϩ , GDP, and added bicarbonate were held constant at the saturating concentrations of 0.2, 2, and 200 mM, respectively, and the PEP concentration was varied in the 0.025-10 mM range, the apparent V max values for the altered enzymes were found to be 1.4 -4-fold lower than that of the wild type ( Table 5). The order of V max values was Asp 75 Ͼ D75N Ͼ D75S Ͼ D75Q Ͼ D75A ( Table 5). The apparent K m values for PEP of the D75A, D75N, and D75S variants were 3-fold higher than that of the wild-type enzyme ( Table 5). For D75Q, the apparent K m values for PEP was 9-fold higher than that of Asp 75 . Consequently, the catalytic efficiencies for D75A, D75N, D75Q, and D75S were 13-, 4-, 28-, and 5-fold lower than that of Asp 75 enzyme, respectively (Table 5). In contrast, the apparent K m values for GDP of the variant enzymes were similar to that of the wild type (Table 5). A similar observation was made for bicarbonate (Table 5), except the apparent K m value for bicarbonate with D75N was 2-fold higher than that of Asp 75 .  1-14 M) and a GTP concentration of 0.2 mM showed that the apparent V max values for the altered enzymes were 1.5-2.7-fold higher than that of the wild-type enzyme ( Fig. 3E and Table 6). The order for the apparent V max values was Asp 75 Ͻ D75Q Ͻ D75N Ͻ D75A Ͻ D75S. However, the substitutions increased the apparent K m value of OAA by about 2-2.8-fold and the order was Asp 75 Ͻ D75N Ͻ D75Q ϭ D75A ϭ D75S. The corresponding changes in the k cat /K m value fell in the range of almost none to 1.8-fold.
CD Spectra of Wild-type and Variant PEPCKs-The CD spectrum of each protein showed minima at 208 and 220 nm (Fig. 4).

DISCUSSION
We have investigated the functions of two Asp residues and one Glu residue of a well conserved sequence element of a GTP-PEPCK (Table 1), which have been called the second Mn 2ϩ -binding site for the enzyme (23). The Mn 2ϩ -binding role for these residues was proposed based on the data from binding studies (23). It was thought that the Mn 2ϩ bound to this site facilitates phosphoryl transfer between PEP and the nucleotide substrate (23). However, the GTP-PEPCK crystallographic data show that the aforementioned residues are situated near the PEP-binding pocket and do not serve as ligands for the enzyme-bound Mn 2ϩ (21,24). Therefore, it is necessary to re-evaluate the roles of these residues. We had reported previously that these Asp and Glu residues show different degrees of conservation in the GTP-PEPCK amino acid sequences (2). Consequently, they all might not play the same role in a PEPCK reaction. To investigate this possibility, we performed a more extensive sequence comparison (Table 1), site-directed mutagenesis ( Table 2) and kinetic studies (Figs. 1 and 2 and Tables 3-6), and we analyzed the results in light of the x-ray crystallographic structure of human cytosolic GTP-PEPCK (21). We used the M. smegmatis GTP-PEPCK (2) for our sitedirected mutagenesis and kinetic studies. The three-dimensional structural information for M. smegmatis PEPCK is not available. The percentage identity and percentage strong similarity between the primary structures of the mycobacterial and human PEPCKs are 50 and 19%, respectively (2). The identity values are much higher for the PEP-binding pocket. Almost every residue of this site of the human enzyme is conserved in the M. smegmatis PEPCK. Therefore, inferences on the structural aspects of the bacterial enzyme could be made from the structure data for the human enzyme. Our results show that two of the above-mentioned residues (an Asp and one Glu) are important for catalysis. Interestingly, the other residue (an Asp) influences the activity of the enzyme in a subtle manner. Our analysis suggested that through long range structural interactions this Asp residue influences the optimal position of one or  Table 5. E, data from the PEP-forming direction assays with standard assay mixture described in the legend of Table 6. For each study, concentration of the respective substrate was varied. The concentrations for other compounds were standard. A, Mn 2ϩ kinetics; B, PEP kinetics; C, GDP kinetics; D, bicarbonate kinetics; E, OAA kinetics. The line drawn through the points for each data set is the best fit to the hyperbola v ϭ V max ϫ S/(K m ϩ S).

Asp 75 , Asp 78 , and Glu 83 of M. smegmatis PEPCK
more catalytically active or important residues. Interestingly, a replacement of this Asp with Ala, Asn, Gln, or Ser changed the V max value for two directions of the PEPCK reaction differently. We discuss these findings below.
Secondary Structure Changes in M. smegmatis PEPCK Variants-Far-UV CD spectra showed that the engineered substitutions did not alter the secondary structure of the protein significantly, although some small changes occurred because of D75N and D75Q substitutions (Fig. 4). A CDSSTR analysis (47)(48)(49) of the CD data identified minor changes in all variants except D75A and E83A (Table 7).
Asp 78 and Glu 83 of M. smegmatis PEPCK-These residues are fully conserved in the GTP-PEPCKs (Table 1). Therefore, they are expected to play important roles in catalysis. Our experimental data supported this conclusion. A replacement of either Asp 78 or Glu 83 with Ala severely diminished activity of the enzyme in both anaplerotic and gluconeogenic directions ( Table 3). The x-ray crystallographic data for the human cytosolic PEPCK (21) helped to rationalize our kinetic data. Asp 78 of M. smegmatis PEPCK is equivalent to Asp 84 of the human enzyme (2,21). The Asp 84 ␥-COO Ϫ interacts directly with the backbone amino group of Tyr 235 , for they are only 2.65 Å apart from each other. Tyr 235 (underlined in sequence below) is one of the fully conserved active site residues of GTP-dependent PEPCKs (21). It belongs to the highly conserved 230 SF(Y)S-GYGGNLGKKC 243 element (2, 21) (Fig. 1, A and B). The binding of PEP to the enzyme causes displacement of the aromatic group of Tyr 235 from its original location by about 2.1 Å (Fig.   FIGURE 4. Far-UV CD spectra for the wild-type and variants of M. smegmatis PEPCK. A, spectra for wild-type (F), D75A (E), D75S (ࡗ), D78A (OE), and E83A (छ). B, spectra for wild type (F), D75N (■), and D75Q (‚). In each case, a solution of protein (0.2 mg/ml) in 10 mM potassium phosphate, pH 7.0, was analyzed. Each spectra is an average of data from nine individual measurements. 3 indicates changes in the spectra because of amino acid substitutions.  1A) (21). After this movement, the aromatic side chain of tyrosine and the carboxylate group of PEP is separated by about 3.5 Å (21). As a result, an energetically favorable, edge-on interaction between the carboxylate group of PEP and the aromatic side chain of Tyr 235 is established (21). We hypothesize that for this role of Tyr 235 and the above-described interactions between Tyr 235 and Asp 84 , the latter residue influences catalysis. The activity of the D78A enzyme was too low to allow meaningful kinetic studies. In contrast, E83A PEPCK was stimulated by Mn 2ϩ . As a result, we were able to perform a detailed kinetic study with this enzyme (Fig. 2). The results were analyzed with respect to the following structure information. Glu 83 of the M. smegmatis enzyme is equivalent to Glu 89 of the human PEPCK (2,21). We found that in the PEP-free state, Glu 89 appears to bind Arg 87 , which is equivalent to Arg 81 of the M. smegmatis enzyme (2,21). In the PEP-bound state, the positively charged side chain of Arg 87 interacts with the negatively charged phosphate group of PEP (21) (Fig. 1, A and B). From the structure data (21) we found that Glu 89 is displaced by over 2 Å when PEP binds to the enzyme (Fig. 1A). We hypothesize that in the absence of PEP, Glu 89 stabilizes Arg 87 through charge interactions and maintains the latter ready for establishing a similar interaction with the phosphate group of PEP. Arg 87 is fully conserved in all GTP-PEPCKs (Table 1). Therefore, Arg 87 is a catalytically important residue in the human PEPCK. The equivalent residue in an ATP-PEPCK (Arg 70 in the S. cerevisiae PEP carboxykinase) is necessary for the complete reaction (50). Our data on the K m values provided insight on how the role of this Arg residue is linked to the observed substantial loss of enzymatic activity caused by the E83A replacement in M. smegmatis PEPCK. The E83A residue change did not affect the apparent K m values for PEP, GDP, and HCO 3 Ϫ significantly, but increased the apparent K m value for Mn 2ϩ by more than 2 orders of magnitude (Table 4). Therefore, we conclude that in the E83A enzyme, Arg 81 remained efficient in binding PEP, but the environment of Mn 2ϩ bound at the metal-binding site 1 (shown in purple in Fig. 1) changed. We hypothesize that a loss of the favorable charge-charge interaction between Glu 83 and Arg 81 because of the E83A substitution caused a perturbation in the Arg 81 -PEP interaction, which in turn changed the location of bound PEP with respect to site 1 Mn 2ϩ (Fig. 1A) (21). The result was a substantial rise in the K m value for Mn 2ϩ and a severe loss of the enzymatic activity. We examined whether the observed increase in the K m value for Mn 2ϩ was a mere reflection of a rise in the requirement of the variants for Mg 2ϩ . When the Mg 2ϩ concentration was raised to 5 mM, the apparent K m value for Mn 2ϩ of the E83A enzyme dropped substantially, but still was 1 order of magnitude higher than that of the wild-type enzyme. The apparent K m value for Mn 2ϩ of the wild-type enzyme did not change because of such a change in the Mg 2ϩ concentration. The K m value for GDP of both the Glu 83 and E83A enzymes decreased following an elevation of the Mg 2ϩ concentration, but these effects were small in magnitude. These observations can be explained as follows. At an elevated Mg 2ϩ concentration, more nucleotide was complexed with this cation, and less Mn 2ϩ was combined with the nucleotide. As a result, more free Mn 2ϩ was available to bind at site 1, and the enzyme was saturated at a lower concentration of total Mn 2ϩ . However, this change could not restore the wild-type interactions between PEP and Mn 2ϩ in the variant enzyme. This conclusion is supported by the fact that a rise in the Mg 2ϩ level did not improve the activity of the enzyme significantly. In summary, our results suggest a critical role for PEP and site 1 Mn 2ϩ interaction in the OAA synthesis. However, the available data do not provide any clue to exactly how a change in the location of PEP affected the rate-determining step for the OAA formation activity of PEPCK. In fact, the identity of the rate-determining step for this direction of the PEPCK reaction remains unknown. For the PEP synthesis activity of chicken liver mitochondria PEPCK, the catalytic cleavage of the ␥-phosphate bond of GTP has been argued to be the rate-limiting step (51).
Asp 75 of M. smegmatis PEPCK-As elaborated in the Introduction, this residue is not invariant (Table 1) (2). On the other hand, it is highly conserved and belongs to a sequence element that bears catalytically critical or important residues (Table 1) (Fig. 1A) (21). Therefore, we hypothesized that Asp 75 of M. smegmatis PEPCK is not critical for catalysis. However, it could influence the PEPCK reaction through indirect interactions with the catalytically active residues. Evolutionarily, such a role makes this residue an avenue for diversifying the nature of PEPCK activity through mutations. To explore this hypothesized role of Asp 75 of M. smegmatis PEPCK, we have investigated the kinetic properties of variant enzymes generated via substitutions at this position. We found that unlike the D78A and E83A variants, which were substantially less active in both directions than the wild type, D75A retained significant activity in the OAA formation direction and was more active than the wild type in the PEP formation direction (Table 3). Then a kinetic study with the variant enzymes that represented all substitutions for Asp 75 (Ala, Asn, Gln, and Ser) that are found in nature (Table 1) provided the following details. The engineered changes caused a moderate decrease in the specific activity and efficiency of the enzyme in the anaplerotic or OAA-forming direction (Fig. 3, A-D, and Table 5) and improved the gluconeogenic activity of the enzyme (Fig. 3E and Table 6). The apparent K m values for Mn 2ϩ , PEP, and OAA were elevated, whereas that for GDP remained unchanged (Table 5). Therefore, the substitutions altered the interactions between enzyme, PEP, and Mn 2ϩ . The three-dimensional structure of the human enzyme (Fig. 1, A and B) provides an explanation for these results. Asp 81 of the human enzyme, which is equivalent to Asp 75 of M. smegmatis PEPCK, is situated near the PEP-binding pocket, but unlike Asp 84 and Glu 89 does not interact directly with an active site residue or a substrate (Table 1 and Fig. 1, A Data were obtained by processing the data shown in Fig. 4 using the CDSSTR program (47,48) at the Dichroweb website (49).  (21). As discussed above, Asp 84 , Arg 87 , and Glu 89 of the human enzyme, three of the fully conserved residues of PEPCK, influence the PEP-Mn 2ϩ interaction. In addition, the ␣-carbonyl group of Val 85 , another fully conserved residue, interacts with the aromatic ring of Tyr 235 (Fig. 1B) and therefore indirectly determines the characteristics of the Tyr 235 -PEP and PEP-Mn 2ϩ interactions. A replacement of Asp 81 in human PEPCK is likely to change the positions of Asp 84 , Val 85 , Arg 87 , and Glu 89 (Fig. 1, A and B), which in turn could influence the placement of PEP, the PEP-Mn 2ϩ interaction, and consequently the catalysis. Such indirect effects are expected to bring about subtle changes in catalysis, which is consistent with our observations with the Asp 75 variants of M. smegmatis PEPCK (Tables 5 and 6). Examples of an effect on the active site over long distances through multiple interactions have been presented previously (52)(53)(54)(55). We are currently investigating exactly how the structural perturbations caused by substitutions at Asp 75 position translated into two opposite effects on catalysis in two directions without violating the Haldane's equilibrium relationship (56) of the reaction. Nevertheless, our data show that it is possible to influence the activity of PEPCK by manipulating a residue distal to the active site. Asp 81 of human PEPCK (Asp 75 of M. smegmatis PEPCK) is not located at the active site but near the surface of the enzyme (21). Because such an approach involving a surface residue has the potential of reducing the activity of the enzyme without directly interfering with catalysis, it could be used to design a compound that can prevent PEPCK from participating in the overproduction of glucose in the liver of a person with type 2 diabetes.