Activation of Latent Dihydroorotase from Aquifex aeolicus by Pressure*

Elevated hydrostatic pressure was used to probe conformational changes of Aquifex aeolicus dihydroorotase (DHO), which catalyzes the third step in de novo pyrimidine biosynthesis. The isolated protein, a 45-kDa monomer, lacks catalytic activity but becomes active upon formation of a dodecameric complex with aspartate transcarbamoylase (ATC). X-ray crystallographic studies of the isolated DHO and of the complex showed that association induces several major conformational changes in the DHO structure. In the isolated DHO, a flexible loop occludes the active site blocking the access of substrates. The loop is mostly disordered but is tethered to the active site region by several electrostatic and hydrogen bonds. This loop becomes ordered and is displaced from the active site upon formation of DHO-ATC complex. The application of pressure to the complex causes its time-dependent dissociation and the loss of both DHO and ATC activities. Pressure induced irreversible dissociation of the obligate ATC trimer, and as a consequence the DHO is also inactivated. However, moderate hydrostatic pressure applied to the isolated DHO subunit mimics the complex formation and reversibly activates the isolated subunit in the absence of ATC, suggesting that the loop has been displaced from the active site. This effect of pressure is explained by the negative volume change associated with the disruption of ionic interactions and exposure of ionized amino acids to the solvent (electrostriction). The interpretation that the loop is relocated by pressure was validated by site-directed mutagenesis and by inhibition by small peptides that mimic the loop residues.

present in many amidohydrolases (13)(14)(15)(16). There are three flexible loops, designated A, B, and C that are disordered and invisible in the X-ray structure.
The isolated DHO subunit crystallized in two different space groups. These independent structures have nearly identical folds, except a water molecule was bound to the zinc in one X-ray structure (Protein Data Bank code 1xrf), whereas in the other structure (Protein Data Bank code 1xrt), Cys 181 in loop A was ligated to the active site zinc, and a section of the loop completely occludes the active site, blocking access of the substrates. We then determined the X-ray structure of the A. aeolicus DHO-ATC complex (17), which showed that all of the disordered loops now had a well defined conformation and that an extensive movement of loop A had displaced it from the active site (Protein Data Bank code 3d6n). The experimental use of pressure can provide valuable and specific information about structure-function relationships in biomacromolecules, particularly proteins. This information relates to the volume changes (⌬V) associated with conformational changes, association-dissociation processes, and the variations of solvation that accompanies these phenomena. This approach is particularly useful in the study of enzyme catalytic and regulatory properties (18 -21). In the present study, elevated hydrostatic pressure was employed to obtain information about its quaternary structural organization and interactions between the subunits in the A. aeolicus complex. However, during the course of this investigation, it was observed that moderate hydrostatic pressure could reactivate the isolated DHO in the absence of its partner, ATC. The hypothesis that elevated pressure mimics the conformational changes induced by the binding of ATC and displaces the loop occluding the DHO active site was verified by sitedirected mutagenesis of residues that tether a segment of the loop in position (17) and with peptide inhibitors. The negative ⌬V associated with the disruption of these interactions explains the reactivation of the isolated DHO by pressure. This interpretation is currently under further investigation by high pressure crystallography.

Results
Effect of Pressure on the DHO-ATC Dodecamer-As reported previously (10), the isolated DHO is inactive but is activated when associated with ATC. Elevated pressure typically, but not always, results in the dissociation of oligomers (19 -21). To test the effect of pressure on the oligomeric structure and activities of the DHO-ATC complex, it was subjected to 800 bars for 1 h at 25°C. Samples were taken at 10-min intervals, and both enzymatic activities were measured at atmospheric pressure and 75°C (Fig. 1A). Both activities decreased with time of exposure to elevated pressure as the DHO-ATC complex dissociated.
As first observed for E. coli ATC (22-29), A. aeolicus ATC is a trimer, and each catalytic site is a composite, comprised of residues from adjacent subunits (9,17). Consequently, dissociation of the trimer results in abolition of catalytic activity. Thus, the decrease in activity of ATC accompanying the dissociation of the trimer was anticipated, and because DHO is active only when associated with ATC, its activity might also be expected to decrease in parallel with the inactivation of ATC. However, FIGURE 1. Effect of pressure on the DHO-ATC complex. A, the DHO-ATC complex was subjected to 800 bars for 1 h. Samples of 0.250 ml were taken at the indicated times, and the ATC (E) and DHO (F) activities were assayed as described under "Experimental Procedures." B, the same kind of experiment was performed at 400 bars (ATC E and DHO F) and 2000 bars (ATC Ⅺ and DHO f). After 1 h, the complex was returned for 30 min at atmospheric pressure, and a sample was taken. C, the DHO-ATC complex was incubated at 1500 bars in the presence (ATC Ⅺ and DHO f) or absence (ATC E and DHO F) of 1 mM carbamoyl phosphate and 5 mM of the aspartate analog, succinate for 1 h and then returned to 1 bar for an additional hour. As a control, the ATC (‚) and DHO (OE) activities were measured in an identical sample kept at atmospheric pressure.
the DHO activity decreases more slowly than that of ATC, suggesting that there are other factors in play. The same experiment was performed at 400 and 2000 bars (Fig. 1B). The inactivation of both enzymes was slower at 400 bars and more rapid at 2000 bars as compared with the results obtained at the intermediate pressure. Again, DHO was inactivated more slowly at 400 bars, although this difference disappeared at 2000 bars where both activities were rapidly abolished at a comparable rate. No significant decrease of the activities was observed in the control samples kept at atmospheric pressure. After incubation under pressure, no recovery of the catalytic activities was observed after 30 min at atmospheric pressure. To verify that the pressure-induced inactivation of the DHO and ATC subunits is not reversible, the DHO-ATC complex was exposed to a pressure of 1500 bars for 1 h and then returned to atmospheric pressure. As expected, the rate of ATC inactivation was faster than that observed in Fig. 1A, and the difference in the rate of inactivation of ATC and DHO was less pronounced. When returned to atmospheric pressure, the activities remained virtually constant, indicating that the pressure-induced inactivation was irreversible.
The interpretation that dissociation of the ATC trimer was responsible for the loss of its catalytic activity was supported by subjecting the DHO-ATC complex to a pressure of 1500 bars in the presence of the ATC substrate, 1 mM carbamoyl phosphate, and 5 mM of the aspartate analog, succinate (Fig. 1C). Previous studies showed that these ligands protect the E. coli catalytic trimer from dissociation and inactivation by pressure. 4 Similarly, the inactivation of the A. aeolicus DHO-ATC complex is appreciably slower in the presence of the substrates (Fig. 1C). Moreover, the loss of DHO activity closely paralleled the inactivation of ATC. Again, no recovery was observed at atmospheric pressure.
The Effect of Pressure on the Isolated Subunits-The isolated DHO and ATC subunits were separately subjected to elevated hydrostatic pressure (1500 bars) for 1 h and then brought back to atmospheric pressure for an additional hour. Samples were taken periodically throughout the time course and assayed at 75°C and atmospheric pressure (Fig. 2). Because DHO activity had only been observed in the DHO-ATC complex, a stoichiometric amount of ATC, which had not been subjected to pressure, was added to the assay after the sample had been removed from the chamber. Because adjacent subunits in the ATC trimer contribute active site residues, its dissociation results in abolition of catalytic activity. Previous studies have shown that elevated hydrostatic pressure dissociates the E. coli ATC catalytic subunit with a concomitant inactivation. 4 Similarly, the isolated A. aeolicus ATC subunit was rapidly inactivated at 1500 bars as the trimer dissociates (Fig. 2). There was no recovery of catalytic activity during the 1-h incubation after the sample was returned to atmospheric pressure, confirming that dissociation was irreversible, as observed previously for the E. coli catalytic subunit. In contrast, the monomeric DHO subunit appeared to be quite stable under these conditions; its activity, when subsequently assayed at 1 bar, was reduced by only 10% after 1 h at 1500 bars. This modest loss of activity was not reversed during the 1 h post-pressure incubation at 1 bar.
Taken together, these results suggested that the inactivation of the two activities of the complex resulted from the pressure induced dissociation of ATC and that the integrity of the complex relies on the interactions between the ATC monomers in the catalytic trimer, a conclusion consistent with the dodecameric structural organization of the complex (17).
Influence of Pressure on the Activity of the Isolated DHO-In view of the slower pressure-induced decrease of DHO activity as compared with that of ATC, a direct influence of pressure on the activity of the isolated DHO subunit was suspected. In the experiments described thus far, the assays were carried out after the DHO-ATC complex had been subjected to elevated pressure and then returned to atmospheric pressure. In the following experiments, the substrate, dihydroorotate was present in the pressure chamber to assay the reaction under pressure. At atmospheric pressure, there is a very small residual DHO activity in the absence of ATC (Fig. 3) as reported previously (10). However, at 200 bars, the reaction proceeded at 57% of the rate observed at atmospheric pressure in the presence of the ATC subunit, indicating that there is a pressure induced activation of the DHO subunit. Control experiments verified that dihydroorotate is stable. No trace of carbamoyl aspartate could be detected during a 1-h incubation of dihydroorotate at 200 bars in the absence of the enzyme, whereas at 1 bar, little carbamoyl aspartate was produced by DHO.
Kinetics of the DHO Subunit under Pressure-Dihydroorotate saturation curves of the pressure activated DHO were conducted at increasing hydrostatic pressures ( Fig. 4 and Table 1). The substrate was present in the pressure cell, and the reaction at each concentration was allowed to proceed for 10 min. The k cat increases with increasing pressure to a maximum of 12.4 mol/min/mg corresponding to a kcat of 10.1 s Ϫ1 , close to the 4 G. Hervé and A. Else, manuscript in preparation. FIGURE 2. Effect of pressure on the isolated ATC and DHO subunits. ATC (F) and DHO (E) were incubated individually at 1500 bars for 1 h and then returned to atmospheric pressure for an additional hour. Samples of 0.250 ml were taken at the indicated times, and the activity was measured. For the DHO activity assays, ATC was added to the sample after removal from the high pressure instrument because at atmospheric pressure, the DHO is active only when the ATC is present.
value obtained for the DHO-ATC complex. There was little pressure-dependent difference in the K m values.
Reversibility of the Reactivation of DHO by Pressure-To determine whether the reactivation of DHO by pressure is a reversible process, the isolated enzyme was preincubated for 30 min at 200 bars and then returned to atmospheric pressure. Samples were taken at various times and tested for activity at 75°C in the absence of ATC. No activity could be detected, indicating that the pressure-induced reactivation of DHO is fully reversible and that once returned to atmospheric pressure, DHO reverts to the inactive conformation.
Pressure Dependence of the Reactivation of the Isolated DHO and ⌬V of Reactivation-The kinetics of the reaction catalyzed by the isolated DHO was measured as a function of pressure over one h incubations to determine the pressure-dependent variation of the rate constant (k). Half-maximal reactivation was observed at about 500 bars. The variation of Ln(k) as a function of pressure ( Fig. 5) is linear as expected on theoretical grounds, allowing calculation of the ⌬V associated with reactivation. The slope of the plot gave a negative ⌬V of Ϫ177 Ϯ 4 ml/mol, a value consistent with the exposure to the solvent of a limited number of charged groups. This result suggested that the reactivation of the isolated DHO may result from the dissociation of loop A from the core of the enzyme similar to the displacement that occurs upon binding of the ATC subunit. This interpretation has been further investigated by site-directed mutagenesis and peptide inhibition experiments.
Mutation of Loop A Residues-Most of loop A is disordered. Only residues 179 -187 are visible in the electron density map (Fig. 6). In addition to the interaction between Cys 181 in loop A and the active site zinc ion (2.30 Å), two other residues anchor the loop in position (11). Asp 179 forms a salt bond with Arg 216 (3.15 Å), and Asp 183 is hydrogen-bonded to Asn 95 (2.78 Å) and forms a salt link with His 63 (2.33 Å). Each of these loop A residues was substituted with valine or glycine.
The effect of mutation of these residues on the DHO activity was assayed in the isolated DHO subunit as a function of the dihydroorotate concentration ( Fig. 7 and Table 2). Each of these mutations significantly decreased the DHOase activity of wild type DHO-ATC. Consequently, the activation of the isolated DHO produced by the mutation was expressed relative that observed for the mutant DHO-ATC complex. The kinetic parameters for the isolated DHO mutant and in the DHO-ATC  Top panel, dihydroorotate saturation curves. A reaction mixture contained 40 g of the purified DHO subunit, and the indicated concentration of dihydroorotate was subjected to elevated pressures (P 1-700 bars) for 10 min at 25°C. Carbamoyl aspartate formed was assayed as described under "Experimental Procedures." The data were fit to the Michaelis-Menten equation. Bottom panel, a plot of k cat , calculated from the V max values, against hydrostatic pressure. Thus, the replacement of two of the residues known to dock loop A to the remainder of the protein results in activation of the latent DHO activity of the isolated subunit. These results are consistent with the hypothesis that pressure-induced activation results from a conformational change that moves loop A away from the active site, allowing substrate binding. The failure of the Cys 181 mutant to restore catalytic activity suggests that this interaction does not significantly contribute to tethering loop A to the active site.
Effect of Peptides on the DHO Activity-The inhibition of the DHO activity by a series of small peptides that mimic the ordered region of loop A was assessed. If loop A is responsible for the inactivity of the isolated DHO subunit, then these peptides would also be expected to inhibit the active enzyme. The peptides 180 HCE 182 , 179 DHCEDD 185 , and 179 DHCEDDKLA 187 were incubated at a concentration of 10 M with the fully active   DHO-ATC complex for the indicated times (Fig. 8). 180 HCE 182 , which includes the two residues flanking the cysteine bound to the zinc in the inactive enzyme, had no effect, whereas the other two larger peptides significantly inhibited the DHO activity of the complex: 179 DHCEDD 185 , 50% inhibition; and 179 DHCED-DKLA 187 , 76% inhibition.

Discussion
Elevated hydrostatic pressure was used to investigate the interactions and disassembly of the dodecameric DHO-ATC complex. When the complex was subjected to elevated pressure, in the absence of its substrates, both activities were rapidly lost at a rate that increased with increasing hydrostatic pressure. There was a parallel loss of DHO activity as expected because DHO is active only when associated with the intact ATC trimer (10). The rate of dissociation/inactivation was dependent on the magnitude of the pressure (Fig. 1). When the DHO-ATC complex was inactivated by subjecting it to elevated pressure for 1 h and then returned to atmospheric pressure for 1 h more, the activities were not recovered, indicating that the inactivation was irreversible.
Protection from pressure-induced inactivation of the E. coli trimeric ATC catalytic subunit by substrate analogs was previously observed. 4 The same result has been obtained here in the case of the DHO-ATC complex (Fig. 1C). The loss of ATC activity can be attributed to the dissociation of the trimer because the bound substrate analogs bridge the active site residues on adjacent subunits and stabilize the complex (24 -29). The DHO activity was lost because it is dependent on association with a structurally intact ATC subunit. However, ATC and DHO activities were not simultaneously inactivated. Except at very high pressure (2000 bars), the inactivation of ATC significantly preceded the loss of DHO activity (Fig. 1). The interface between the DHO and the ATC monomers is quite extensive corresponding to 788 Å 2 (17). It is stabilized by nine H-bonds and seven salt bonds, so it is possible that the DHO subunit remains partially active via interactions with the fully or partially dissociated ATC subunit. However, it appeared that pressure had a positive influence on the activity of the isolated DHO.
The effect of pressure on isolated constituent subunits was also tested (Fig. 2). The activity of the ATC subunit decreased exponentially and irreversibly when incubated at 1500 bars (k ϭ 0.018 min Ϫ1 ) because of the dissociation of the obligatory trimeric structure. There was little effect on the activity of the DHO subunit when it was returned to atmospheric pressure and assayed in the presence of the ATC subunit, so pressure had little direct irreversible effect on the DHO structure and activity.
In the second series of experiments, the assays were conducted by introducing the substrate dihydroorotate into the pressure chamber rather than assaying the sample after it was returned to atmospheric pressure. These experiments were designed to directly access the influence of pressure on the catalytic activity of DHO. Unexpectedly, the isolated DHO became fully active in the absence of ATC. Up to this point, the DHO activity had only been observed in the complex, and virtually no activity has been observed in the presence of the same concentration of the substrate at ambient pressure.
We cannot rule out the possibility that reorganization of the other loops, which are disordered at ambient pressure, is also involved in catalysis. Two residues, Asn 278 in loop B and Pro 322 in loop C, are within hydrogen bonding distance of dihydroorotate in the DHO-ATC structure (17). These residues are not involved in the catalytic mechanism but may be significant in substrate binding. Once ordered (17), the remainder of loop is far from the catalytic site and appears to be primarily involved in establishing the interface between the DHO and ATC subunits. However, a more plausible explanation for the restoration of the catalytic activity under pressure is that loop A, which blocks access of the substrates to the active site, is displaced.
We can postulate that there is equilibrium between the inactive and active conformations of the isolated DHO subunit. At ambient pressure the equilibrium lies far toward the inactive conformation with loop A occluding the active site, but the direction of the equilibrium is reversed at elevated pressure.  The interpretation that the two conformations of the protein are accessible at 1 bar is supported by the X-ray structures of the isolated DHO subunit; one structure (IIa) has the loop anchored and occludes the active site, whereas in the other (Ia) the loop is displaced with Cys 181 relocated 8.36 Å from the catalytic zinc ion (11). We have no evidence that the postulated open conformation (Ia) is, in fact, catalytically active, although the X-ray structure indicates that although all of loop A residues (Asp 179 -Asp 210 ) are disordered, the DHO active site residues are appropriately positioned as in the active site of the DHO-ATC complex (Fig. 8). The interatomic distances between key active site residues in the loop-out and the complex structures are within 0.5 Å in the two independent structures ( Table 3). The salt bridge between Asp 179 in loop A and Arg 16 is disrupted in the open structure as it is in the partially active D179V mutant, an observation consistent with the idea that the open structure may be active.
There are four prominent interactions between bound dihydroorotate and the enzyme involving the residues Arg 65 , Asn 95 , Asn 278 , and His 309 (17) in DHO-ATC. With the exception of the Asn 278 , which is part of the disordered loop B, these residues are appropriately placed in the DHO open structure ( Fig. 9 and Table 3), albeit with slightly different orientations, suggest-ing that the loop-out structure of the subunit may be able to bind the substrate.
A major effect of pressure on protein structure results from electrostriction (18,19). Upon disruption of a salt bond, the water molecules surrounding the liberated charge are compressed as a result of the coulombic field of the charged group (electrostriction). This results in a decrease in the volume of the solution, a process that is favored by pressure.
Examination of the structure of the isolated DHO subunit (IIA) showed that loop A is tethered to the active site region by the zinc ligand, Cys 181 , a salt link between Asp 179 in loop A and Arg 283 and a salt link and H-bond between Asp 183 with His 63 and Asn 95 , respectively (Fig. 5). With the exception of Cys 181 , mutation of these residues disrupts these interactions and partially restores the latent activity of the isolated DHO subunit.
The inhibition of DHO activity by a series of peptides with sequences corresponding to the segment of loop A near the active site was also assessed. The smallest peptide containing Cys 181 ( 180 HCE 182 ) does not inhibit, consistent with the minimal effect of the C181G mutant on the activity of the isolated DHO subunit. The peptide 179 DHCEDD 184 , containing tethering residues Asp 179 and Asp 183 identified in the loop-in DHO structure, is a good inhibitor, and a longer peptide from the same region ( 179 DHCEDDKLA 187 ) is more effective. This result suggests that the binding of loop A to the active site region is sufficient to inactivate DHO.
Except for pressure-induced dissociation of obligate oligomers, there are few examples in which moderate hydrostatic pressure alters enzyme activity. One notable exception is the pressure induced inactivation of ␣-chymotrypsin (32) resulting from the disruption of a salt link between Asp 194 and Ile 16 . Similarly, moderately elevated hydrostatic pressure induces the same open conformation in plasminogen as the binding of 6-aminohexanoate (33), a change that promotes the binding of streptokinase and plasminogen activation. Large changes in the rate of hydrolysis of ␤-lactoglobulin by thermolysin and pepsin were observed under elevated hydrostatic pressure (34).
The occlusion of the DHO active site by loop A prevents substrate binding and can largely account for the intrinsic inactivity of the isolated DHO subunit. This mechanism ensures that the DHO is active only when associated in a complex with  1 1 1 1 1 1 1 1 1 1 1 1 1

A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A A p A A A A A A A A A A A A A A A A A A As s s
A A n n n n n n n n n n n n n n n n n n n n n n9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 95 5 5

H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H H His309
Asp305 Asp305

Experimental Procedures
Materials-All chemicals were purchased from Sigma-Aldrich. E. coli strains DH5␣ and BL21 (DE3) were from Invitrogen.
Isolation of the DHO and ATC Subunits and Formation of the DHO-ATC Complex-The genes encoding A. aeolicus DHO (pyrC) and ATC (pyrB) subunits were previously cloned and separately expressed in E. coli (8,9). The genes were identified in the A. aeolicus genome (6), amplified by PCR, and inserted into pRSETC (Invitrogen), an expression vector that incorporates a 3-kDa His tag on the amino end of the recombinant protein. The proteins were expressed individually in BL21 (DE3) and purified by Ni ϩ2 affinity chromatography on a 1-ml Ni ϩ2 Probond column (Invitrogen). The purified subunits were mixed in an equimolar ratio to reconstitute the dodecameric DHO-ATC complex (10,17,36). The complex was isolated by gel filtration chromatography on a Sephacryl S-300 column. The composition of the complex was analyzed by SDS-polyacrylamide gel electrophoresis (37) and ATC and DHO assays. Protein concentrations were determined by the Lowry method (38) using bovine serum albumin as a standard.
Enzyme Assays-The ATC activity was measured by monitoring the formation of carbamoyl aspartate using a colorimetric method previously described (39,40). The reaction contained 2 mM aspartate, 5 mM carbamoyl phosphate, 1-11 g of the purified enzyme, and 50 mM Tris-HCl, pH 8, in a total volume of 0.5 ml. The incubation time was 10 min at 25°C or 1.5 min at 75°C. The reactions were quenched by the addition of an equal volume of 5% acetic acid. The color was developed and quantitated as described previously (10).
DHO activity was measured (10) in the reverse direction, the formation of carbamoyl aspartate from dihydroorotate, because the equilibrium strongly favors dihydroorotate hydrolysis (35) at pH 8.3. The assay mixture consisted of 20 -45 g of DHO in 50 mM Tris acetate, pH 8.3. The reaction was initiated by the addition of dihydroorotate (8 mM or variable) and quenched by the addition of an equal volume of 5% acetic acid after 1.5 min (75°C) or 10 min (25°C). For reactions assayed at 75°C, the assay mixture was equilibrated by preincubation at temperature for 1.5 min prior to initiation. Carbamoyl aspartate formed was determined by the same colorimetric method. The kinetic parameters were obtained by least squares analysis of the dihydroorotate saturation curves using the program KaleidaGraph 4.1 (Synergy Software).
High Pressure Experiments-The protein was subjected to elevated hydrostatic pressure in a reactor that allows removal of the samples of the reaction mixture via a pressure lock so that pressure inside the reaction chamber is not perturbed (30).
For assays at elevated hydrostatic pressure, the reaction mixture consisted of 2.10 ml of 0.5 M Tris HCl, pH 8, 0.40 ml of 2.0 mg/ml DHO (800 g), and 15.1 ml of water. The reaction was initiated by the addition of 3.4 ml of 50 mM dihydroorotate. The reaction mixture, 19 ml was immediately transferred to the pressure chamber with the remaining 2 ml maintained at atmospheric pressure as a control. The chamber was pressurized, and 0.250-ml samples were taken at the indicated times and quenched into 0.250 ml of 5% acetic acid. The carbamoyl aspartate produced was measured as described above. For some experiments, a ISS (Champaign, Il) high pressure system with a 1.0 quartz cell was used.
Site-directed Mutagenesis-Site-directed mutagenesis was carried out by PCR as described previously (31) using Pfu Turbo polymerase from Stratagene (ThermoFisher) and the plasmid encoding the DHO subunit as a template. The residues in loop A suspected of anchoring the loop at the entry to the active site were replaced by valine or glycine using forward and reverse oligonucleotides from Invitrogen (ThermoFisher Scientific).
Author Contributions-G. H., H. G. E., and D. R. E. conducted the pressure studies, analyzed the results, and wrote the paper. R. F. conducted some of the pressure studies. R. F., F. H., and C. P. constructed the mutants and conducted the kinetic studies.