High activity of human butyrylcholinesterase at low pH in the presence of excess butyrylthiocholine Patrick Masson 1 , Florian Nachon 1,2 , Cynthia F. Bartels 2 , Marie-Therese Froment 1 , Fabien Ribes 1 , Cedric Matthews 1 and Oksana Lockridge 2 1 Centre de Recherches du Service de Sante ´ des Arme ´ es, Unite ´ d’Enzymologie, La Tronche, France; 2 Eppley Institute, University of Nebraska Medical Center, Omaha, Nebraska, USA Butyrylcholinesterase is a serine esterase, closely related to acetylcholinesterase. Both enzymes employ a catalytic triad mechanism for catalysis, similar to that used by serine pro- teases such as a-chymotrypsin. Enzymes of this type are generally considered to be inactive at pH values below 5, because the histidine member of the catalytic triad becomes protonated. We have found that butyrylcholinesterase retains activity at pH £ 5, under conditions of excess substrate activation. This low-pH activity appears with wild- type butyrylcholinesterase as well as with all mutants we examined: A328G, A328I, A328F, A328Y, A328W, E197Q, L286W, V288W and Y332A (residue A328 is at the bottom of the active-site gorge, near the p-cation-binding site; E197 is next to the active-site serine S198; L286 and V288 form the acyl-binding pocket; and Y332 is a component of the peripheral anionic site). For example, the k cat value at pH 5.0 for activity in the presence of excess substrate was 32 900 ± 4400 min )1 for wild-type, 55 200 ± 1600 min )1 for A328F, and 28 700 ± 700 min )1 for A328W. This activity is titratable, with pK a values of 6.0–6.6, suggesting that the catalytic histidine is protonated at pH 5. The existence of activity when the catalytic histidine is protonated indicates that the catalytic-triad mechanism of butyrylcho- linesterase does not operate for catalysis at low pH. The mechanism explaining the catalytic behaviour of butyryl- cholinesterase at low pH in the presence of excess substrate remains to be elucidated. Keywords: butyrylcholinesterase; excess substrate activation; mutant enzyme; pH dependence; steady-state kinetics. Human butyrylcholinesterase (EC 3.1.1.8; BuChE) is a serine esterase, which is present in vertebrates. It is routinely isolated from plasma [1] where it is considered to be of pharmacological and toxicological importance because it hydrolyzes numerous ester-containing drugs [2] and scav- enges toxic esters, such as organophosphates [3]. Its primary amino-acid sequence is 54% identical with that of Torpedo californica acetylcholinesterase (EC 3.1.1.7; AChE) [4]. A 3D model for human BuChE has been built [5] from the known co-ordinates for the 3D structure of T. californica AChE [6]. This model agrees with the general features of the recently determined X-ray structure of human BuChE [7,8]. In particular, most of the essential features of the catalytic site (i.e. a catalytic triad of Ser-His-Glu, an oxyanion hole, a p-cation-binding site, and an acyl-binding pocket) are the same in AChE and BuChE (Fig. 1). The acyl-binding pocket, which is responsible for the difference in substrate specificity between the two enzymes, is larger in BuChE [5,8–10]. The active site for both enzymes is located at the bottom of a 20-A ˚ deep gorge. An aspartate residue [D70(72)] is located at the mouth of the gorge. [Italicized numbers in parentheses (N ) after amino-acid numbers refer to residue numbering in T. californica AChE. In human BuChE, the corresponding residue is N)2.] This aspartate, part of the peripheral anionic site, contributes to the affinity of positively charged substrates for the active site, and is a major factor in the binding of excess substrate to these enzymes [11,12]. Neither AChE nor BuChE follows Micha- elis–Menten kinetics with positively charged substrates. Under standard conditions, i.e. at neutral pH and 25 °C, AChE has been shown to be inhibited by excess substrate, whereas BuChE is activated [13]. However, we recently reported that AChE may display substrate activation at low pH [14]. The complete mechanism by which activation or inhibition of cholinesterases by excess substrate occurs is still controversial, but it is now accepted that binding of a second molecule of substrate on the peripheral anionic site (PAS) induces a conformational change that triggers the process. For some time, we have been interested in the molecular basis of substrate activation in wild-type and mutants of human BuChE [11,15,16], as well as substrate activation in wild-type and mutants of human and Bungarus fasciatus AChE at low pH [14]. The term substrate activation describes the situation in which excess substrate causes an increase in the turnover number (k cat )ofanenzyme.For wild-type BuChE reacting with butyrylthiocholine (BTC), Correspondence to P. Masson, Centre de Recherches du Service de Sante ´ des Arme ´ es, Unite ´ d’Enzymologie, B.P. 87, 38702 La Tronche Cedex, France. Fax: + 33 4 76 63 69 63, Tel.: + 33 4 76 63 69 59; E-mail: pymasson@compuserve.com Abbreviations: AChE, acetylcholinesterase; BuChE, butyrylcholine- sterase; BTC, butyrylthiocholine; DTNB, 5,5¢-dithiobis- 2-nitrobenzoic acid; PAS, peripheral anionic site. Enzymes: butyrylcholinesterase (EC 3.1.1.8; BuChE); acetylcholin- esterase (EC 3.1.1.7; AChE). (Received 5 August 2002, revised 29 October 2002, accepted 25 November 2002) Eur. J. Biochem. 270, 315–324 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03388.x the turnover number is determined by BTC concentrations in the 10–100 micromolar range. Millimolar levels of BTC cause the activity to rise above this turnover number, eventually reaching a new, excess-substrate-defined turn- over number, or bk cat (with b > 1). With wild-type human BuChE, the turnover number for BTC increases 2.5–3-fold (b ¼ 2.5–3) in the presence of excess BTC (at pH 7.0). Mutations at the 328(330) position (A328 is at the bottom of theactivesitegorge,nearthep-cation site) have been reported to cause a marked decrease in this substrate activation, i.e. A328F is activated only 20% (b ¼ 1.2), at pH 7.0 [17], whereas A328Y is inhibited by 20% (b ¼ 0.8), at pH 8.0 [9]. To clarify the cause of these differences in substrate activation between the wild-type and A328 mutants, we examined the pH dependence of BTC turnover. Although we found the pK a values for k cat and bk cat of the mutants to be slightly decreased relative to wild-type, these decreases could not explain the observed differences in substrate activation. However, in the course of our studies, we observed that the bk cat values did not fall to zero as the pH was lowered from 8.5 to 5.0, for either wild-type or mutant BuChE. Rather they approached a substantial, nonzero limit. For example, the limiting value for bk cat at low pH was 32 900 ± 4400 min )1 for wild-type, 55 200 ± 1600 min )1 for A328F, and 28 700 ± 700 min )1 for A328W. Similar observations were made for wild-type human and B. fasciatus AChEs and their mutants modified on the equivalent residues [14]. Consistent with previous reports, we found that k cat did approach zero as the pH decreased from 8.5 to 5.0. The pK a values that we found for both k cat and bk cat are consistent with titration of the catalytic histidine, H438(440). The persistence of activity in the presence of the protonated form of the catalytic histidine is inconsistent with the generally accepted mechanism for hydrolysis by cholinesterases [18,19]. This mechanism utilizes the catalytic histidine as an acceptor for a proton from the catalytic serine, therefore protonation of the histidine would be expected to block catalysis. Rather, the appearance of activity under conditions in which the catalytic histidine is protonated indicates a change in the mechanism of BuChE and AChE. Work is in progress to probe the mechanism that could explain these observations. Materials and Methods Chemicals Butyrylthiocholine iodide (BTC) and 5,5¢-dithiobis-2-nitro- benzoic acid (DTNB) were purchased from Sigma Chemical Co., St Louis, MO, USA. Chlorpyrifos-oxon was from Chem Services Inc., West Chester, PA, USA (catalog number MET-674B). All other chemicals, including buffer components, were of biochemical grade. Mutagenesis and expression of recombinant BuChE Mutagenesis and expression were performed as described previously [17]. Briefly, mutations in human BuChE were created, then amplified by PCR using Pfu polymerase. Fragments containing the mutation were cloned into the plasmid pGS. The plasmid was transfected into CHO-KI cells by calcium phosphate coprecipitation. Stable cell lines were selected in methionine sulfoximine. Expressed BuChE was secreted from these cells and collected into serum-free medium. Purification of BuChE Mutant forms of BuChE were purified from culture medium as previously described [1,17]. Briefly, the culture medium was passed over a procainamide–Sepharose affinity column which retained the BuChE, which was then selectively eluted with 0.2 M procainamide hydrochloride. Further purification was obtained by ion-exchange chro- matography on DE52 (Whatman, Clifton, NJ, USA) using an NaCl gradient for elution. The resulting enzyme was typically 70–95% pure. Wild-type BuChE used in this work was purified from human plasma using the same combina- tion of affinity and DE52 chromatography. The concentration of each BuChE mutant was deter- mined by titration with chlorpyrifos-oxon as proposed by Amitai et al. [20]. Chlorpyrifos-oxon concentration was standardized against wild-type BuChE of known concen- tration. Enzyme assay Initial rate of turnover of BTC was measured by the method of Ellman et al.[21]in0.1 M sodium phosphate buffer, pH variable from 5.0 to 8.5 and, in 0.1 M sodium acetate buffer, pH ranging from 4.0 to 5.25. The ionic strength of phosphate buffers varied from 0.1 to 0.29, and that of acetate buffers varied from 0.014 to 0.075. Such changes in ionic strength are known to have no effect on k cat of BuChE-catalyzed hydrolysis of cationic substrates [11,13,22]. Buffers contained 0.33 m M DTNB and 0.01– 50 m M BTC, at 25 °C. Product formation was followed by Fig. 1. Side view of the active-site gorge of acylated (butyrylated) human BuChE. The arrow indicates the entrance of the gorge. D70 and Y332 are the peripheral anionic site residues. The active site is at the bottom of the gorge: the substrate binding subsite is W84 and A328; the acyl- binding pocket is formed from L286 and V288; the catalytic triad is S198, H438 and E325. Residue 197 next to the catalytic serine is involved in stabilization of transition states. 316 P. Masson et al.(Eur. J. Biochem. 270) Ó FEBS 2003 the change in A 420 . Observed rates were corrected for spontaneous hydrolysis of BTC and for spontaneous reduction of DTNB from blank samples. The observed rate, in terms of mol productÆmin )1 ÆL )1 , was obtained by dividing the DA 420 Æmin )1 by the absorption coefficient for the 5-thio-2-nitrobenzoic acid (TNB), which is the product of the reaction of thiocholine with DTNB. This calculation was complicated by the fact that TNB has a pK a of 4.53 [23], which gives rise to a decrease in absorption coefficient at pH values below 7.0. Consequently, we measured the absorp- tion coefficients at 420 nm for TNB, at pH values between 5.0 and 8.5. Representative absorption coefficients are: 8590 M )1 Æcm )1 at pH 5.0; 11 100 M )1 Æcm )1 at pH 5.5; 12 500 M )1 Æcm )1 at pH 6.0; 13 200 M )1 Æcm )1 at pH 7.0; and 13 300 M )1 Æcm )1 at pH 8.0. Data analysis Steady-state turnover of BTC with wild-type BuChE exhibits the phenomenon of excess substrate activation. This is illustrated in Scheme 1. This scheme is also suitable for excess substrate inhibition. This scheme is described by Eqn (1): k app ¼ k cat þ bÃk cat þ½S K ss 1 þ K m ½S  1 þ ½S K ss  ð1Þ where k app is the apparent rate, in terms of mol productÆ(mol BuChE) )1 Æmin )1 , [S] is the concentration of BTC. k cat is the turnover number (min )1 )when [S] << K ss , K m is the Michaelis–Menten constant, bk cat is the turnover number (min )1 ) when [S] >> K ss ,andK ss is the dissociation constant for excess BTC [10,24]. The parameter b reflects the efficiency of product formation from the ternary complex (SES). When b >1, there is substrate activation. When b <1, there is substrate inhibition. When b ¼ 1, the enzyme follows Michaelis– Menten kinetics. The k cat , K m , K ss and b values were obtained by nonlinear fitting of the apparent rate vs. BTC concentration data to Eqn (1), using SigmaPlot v4.16 (Jandel Scientific, San Rafael, CA, USA). The value for bk cat was obtained by multiplying k cat by b. Results pH dependence of turnover The values of k cat , bk cat , K m and K ss were obtained by fitting the apparent rate (k app ) vs. BTC concentration data to eqn (1), at each pH value (data not shown). To within the limits of experimental error, K m was independent of pH for all enzymes. This is consistent with earlier reports on the pH dependence of K m for human BuChE [25,26]. The value of K ss increased as the pH was lowered from 8.5 to 5.2, for all enzymes. This indicates that the binding of BTC to the excess-substrate activation site is becoming weaker as the pH is lowered. The change in K ss varied from fivefold to 20-fold, depending on the enzyme. At pH 8.5, K ss had essentially stopped changing with pH, having reached a limiting value for high pH. As the pH was lowered, the value of K ss became progressively larger; however, by pH 5.2 a clear inflection point had not yet developed. Therefore, pK a values for K ss could not be determined; only an upper limit of 5.0 could be estimated. Such a low pK a is consistent with the involvement of an acidic amino acid in the binding of excess BTC. Excess-substrate activation for human BuChE has been attributed to binding of positively charged substrates, such as BTC, to D70 in the peripheral anionic site [11,27]. The pH dependence of K ss is consistent with protonation of D70. The pH dependence data for the D70G mutant supports this statement. Indeed, although the D70G mutant shows a slight activation by excess substrate (b ¼ 1.2 ± 0.2) at BTC concentrations higher than 2 m M [11,15,27], its high K ss value (> 1 m M [27]) does not significantly change with pH (not shown). However, D70 is not the only residue involved in substrate activation: the D70N mutant was shown to be strongly activated by excess substrate [27] and even the D70G mutant shows substrate activation similar to that of wild-type enzyme in the presence of high concentrations of sugars or polyols [28]. No further discussion on the pH dependence of K m or K ss will be presented, so that we can focus attention on the pH dependence of k cat and bk cat . Figure 2 shows the pH dependence of k cat and bk cat for wild-type BuChE and the mutants A328G, A328I, A328F, A328Y and A328W, using BTC as substrate. Simple inspection of Fig. 2 reveals that both k cat and bk cat exhibit well-defined pH titration profiles, with the minimum rates occurring at low pH. The values for k cat approach zero by pH 5.0. However, the bk cat values clearly approach a nonzero limiting activity at low pH. Limiting activity is the plateau value in a titration curve. All of the titrations extend over a range of at least 3 pH units, indicating that they are more than 90% complete by pH 5. Therefore, the nonzero limiting rates for bk cat at low pH cannot be attributed to incomplete titration. The data in Fig. 2 were in turn fitted to an expression for asinglepK a (see the legend to Table 1 for details). The fitting results are tabulated in Table 1. Wild-type BuChE and all of the A328 mutants show limiting rates for k cat at low pH (k H )thatare10%orlessof their limiting rates at high pH (k A ). Therefore, the limiting values of k cat at low pH may be considered to be effectively zero. This is despite the fact that most of these k H values are statistically greater than zero. The pK a values of 6.5–7.0 for these enzymes are all consistent with the titration of a histidine. These results for k cat are consistent with the what is commonly found for wild-type BuChE and wild-type AChE (see [32] for a review). In these reports, the titrating histidine has consistently been taken to be the catalytic histidine, i.e. H438(440) in BuChE. The bk cat values also gave well-behaved titrations for wild-type BuChE and all of the A328 mutants, except A328W. The pK a values of 6.0–6.5 are consistent with titration of a histidine. By analogy with k cat ,thetitrating Scheme 1. Steady state turnover of BTC. Ó FEBS 2003 High activity of BuChE at low pH (Eur. J. Biochem. 270) 317 histidine is probably the catalytic histidine, H438. It is noteworthy that the limiting values of bk cat at low pH are decidedly greater than zero (e.g. k H ¼ 32 900 ± 4400 min )1 for wild-type BuChE or k H ¼ 28 700 ± 700 min )1 for A328W). Thus, in the presence of excess BTC, BuChE is active even though H438 is protonated. The existence of substantial activity for BuChE when the catalytic histidine is protonated is an unprecedented observation, which has significant implications for the mechanism. We would like to emphasize that the activity that we measure for bk cat does not approach zero at low pH. For a titration that ends at zero activity for the fully protonated histidine, theory predicts that at 1 pH unit below the pK a , only 10% of the histidine is unprotonated and that therefore only 10% of the activity will remain. Our results in Fig. 2 show that for A328F 71% remains, for A328I 32% remains, for wild-type 40% remains, and for A328G 42% remains. In all four cases, the remaining activity at 1 pH unit below the pK a is much higher than the theoretical prediction. As the major point of this paper rests on the observation that there is significant turnover of BTC by BuChE at low pH, it is important to control for artifactual sources of turnover. The first point to be made is that, to our knowledge, BTC is not hydrolyzed, to any significant extent, by any enzyme other than BuChE. AChE will hydrolyze BTC slowly, but our recombinant enzymes were collected into serum-free medium, which contains no AChE. As we are seeing high bk cat values for BTC turnover at low pH, this activity is probably not due to a contaminating enzyme. Secondly, all of the rate data have been corrected for spontaneous hydrolysis of BTC and chemical reduction of DTNB. Thirdly, to control for unexpected contaminations in the expressed enzymes, which might have arisen from the cell culture, we examined wild-type BuChE that had been purified to homogeneity (tetrameric G 4 form) from human plasma. As can be seen in Figs 2 and 3, wild-type BuChE also had substantial activity at low pH, in the presence of excess substrate. It is unlikely that both naturally occurring and cultured BuChE would show the same contaminations; therefore, artifactual hydrolysis from contamination is unlikely. Further titration of wild-type BuChE The titration of wild-type BuChE was extended from pH 8.5 to 4.0. At pH 4.0, the k cat activity was effectively zero, and bk cat activity was approaching zero (Fig. 3). The titration of k cat was monophasic, with a pK a of 6.7 ± 0.09. However, the titration of bk cat was very broad, extending for more than 4.5 pH units. The profile was clearly biphasic, with pK a values of 4.63 ± 0.24 and 6.68 ± 0.20 and a rate for the singly protonated species of 41 400 ± 4500 min )1 . Complete elimination of the bk cat activity required proto- nation of two amino acids. This biphasic titration accen- tuates the fact that the intermediate, singly protonated species is active. That is to say, complete protonation of an amino acid with a pK a of 6.68 results in an enzyme that still hydrolyzes BTC at a rate of 41 400 min )1 .Aswehave suggested, the most likely candidate for this group is the catalytic histidine, H438. The pK a of 4.63 suggests the involvement of an acidic residue in the activity at low pH. Low-pH activity and residue E197 Selwood et al. [34] reported two pK a values for the pH dependence of the reaction of Electrophorus electricus and T. californica AChEs with BTC. The higher pK a (6.3 and 6.1, respectively) was attributed to the titration of the catalytic histidine, H440 (corresponding to H438 in human BuChE). Protonation of this residue in electric eel AChE left a preparation that retained % 30% of the activity found at high pH. The residual activity could be abolished by further titration, yielding a pK a of 4.7. Similarly, the pH profile for T. californica AChE yielded a pK a of 5.0. This pK a was attributed to residue E199, an active-site residue corres- ponding to E197 in human BuChE. They proposed that titration of H440 resulted in a change in mechanism Ôfrom triad catalysis to one that likely involves general base catalysis by E199 of direct water attack on the scissyl carbonylÕ. The similarity between their results and ours is evident and suggests that the low-pH activity of BuChE, in the presence of excess BTC, may be due to general base catalysis by E197. Owing to this similarity, it became necessary to test the involvement of residue E197(199) in the activity of BuChE at low pH. To accomplish this, we determined the pH Fig. 2. pH dependence for the turnover number (k cat ) and the excess- substrate-activated turnover number (bk cat ) of wild-type human BuChE and various 328-position mutants. Each panel represents a different mutant form of BuChE, as indicated. In each panel, the solid circles indicate the measured k cat values, the solid squares indicate the measured bk cat values, and the lines are the result of fitting the meas- ured rates to an expression for a single pK a (see the legend to Table 1 for details). The values for k cat and bk cat ,ateachpH,weretakenfrom fitting of k app vs. BTC concentration data for each mutant (data not shown). 318 P. Masson et al.(Eur. J. Biochem. 270) Ó FEBS 2003 dependence of the activity of the BuChE mutant E197Q. We reasoned that glutamine at position 197 was equiv- alent to protonation of E197. If general base catalysis by E197 was responsible for the residual activity seen with wild-type BuChE, in the presence of excess BTC at low pH, then E197Q would not be able to support that activity. The excess-substrate activity of E197Q (bk cat ) should then titrate to zero with a pK a % 6.5 (correspond- ing to the catalytic histidine, H438). Figure 4 shows that this expectation was not realized. The bk cat rates of E197Q approach a substantial, limiting rate at low pH. The change in bk cat between high pH and low pH is not large, but the trend is clear, and it yields a pK a of 6.17 ± 0.56. Thus, the suggestion that E197 is responsible for the low- pH activity of wild-type BuChE in the presence of excess BTC is not supported. Moreover, it should be noted that the pK a of mutant E1997Q for k cat is shifted by % 1 pH unit below that of k cat of wild-type. Such a shift supports the assumption that the observed pK a is related to His438 because it is consistent with the fact that the electrostatic stabilizing effect of the E197 side chain on the protonated form of H438 is abolished in the E197Q mutant. Role of position 328 in the excess-substrate effect Mutations at position 328 seem to modulate the behaviour of BuChE, rather than to introduce qualitatively new behaviour. The most obvious indication of this modulation appears at high pH where the limiting value of bk cat tends to approach the limiting value of k cat as the size of the residue at position 328 increases (Fig. 2). For example, the differ- ence between bk cat and k cat at high pH is 68 400 min )1 for wild-type BuChE (A328), 19 700 min )1 for A328F, and 13 000 min )1 for A328W. Thus, the larger residues seem to interfere with the ability of excess substrate to increase the activity at high pH. However, the aliphatic or aromatic character of side chains has to be considered too. At low pH, the size of the residue in position 328 has no consistent effect on bk cat .Thevaluesforbk cat are generally between 20 000 and 40 000 min )1 (Fig. 2). The net effect is that the value of bk cat at high pH becomes closer to its value at low pH as the size of the residue at position 328 gets larger (Fig. 2). For example, the difference between bk cat at high pH and bk cat at low pH is 65 600 min )1 for wild-type (A238), 22 700 min )1 for A328F, and 0 for A328W. From the effect that the size of the 328 residue has on bk cat , it is tempting to suggest that the 328 position (which is part of the substrate binding site) plays a special role in the excess substrate effect. However, mutations at other loca- tionsintheactivesitealsoperturbthepHdependenceof bk cat . E197Q (part of the esteratic site) shows a pH dependence for bk cat that is similar to that for A328F (Fig. 4). V288W (in the acyl-binding pocket) and Y332A (in the PAS) show pH dependencies more like A328I, i.e. the difference between bk cat and k cat at high pH is smaller than for wild-type, and the difference between bk cat at high pH and bk cat at low pH is relatively small (Fig. 5). On the other hand, L286W (also in the acyl-binding pocket) is similar to wild-type BuChE, i.e. the difference between bk cat and k cat at high pH is large relative to the difference at low pH, and the difference between bk cat at high pH and bk cat at low pH is relatively large (Fig. 5). In all of these enzymes, there is a substantial activity for bk cat at low pH, strengthening the Table 1. pH dependence of k cat and bk cat for BuChE mutants. Values for the parameters were determined by fitting the data from Figs 1, 3 and 4 to the expression: k ¼ k H þ k A à 10 ðpHÀpK a Þ 1 þ 10 pHÀpK a which is an algebraic rearrangement of the more common expression for the dependence of rate on pH involving a single pK a [29]: pH ¼ pK a À log k À k A k H À k  The term k stands for the observed rate, k A stands for the limiting rate at high pH, and k H stands for the limiting rate at low pH. This is a general expression for a pH titration, which does not exclude the possibility of a nonzero limiting rate at either pH extreme. The rearrangement was required in order to obtain the dependent variable (k) in terms of the independent variable (pH). Fitting was performed using SIGMAPLOT v.4.16. k cat is the turnover number in the absence of excess substrate. bk cat is the turnover number in the presence of excess substrate. Residue volumes were taken from Zamyatnin [30]. Hydrophobicity ratings were taken from Karplus’ ÔpureÕ hydrophobicity scale [31]. NA, not applicable. There is no change in bk cat with pH for A328W. BuChE k cat bk cat Residue volume (A ˚ 3 ) Hydrophobicity k H (min )1 ) k A (min )1 )pK a k H (min )1 ) k H (min )1 )pK a A328G 2000 ± 1100 26 100 ± 1000 6.78 ± 0.11 26 800 ± 1700 70 400 ± 800 6.23 ± 0.07 60.1 1.18 Wild-type 2800 ± 1100 30 100 ± 1200 6.83 ± 0.10 32 900 ± 4400 98 500 ± 3600 6.56 ± 0.15 88.6 2.15 A328I 1230 ± 1100 27 900 ± 1340 6.79 ± 0.11 13 600 ± 3640 51 100 ± 1480 6.02 ± 0.15 166.7 3.88 A328F 6600 ± 1300 58 200 ± 1800 6.58 ± 0.06 55 200 ± 1600 77 900 ± 2200 6.57 ± 0.18 189.9 3.46 A328Y 8200 ± 3000 73 300 ± 3100 7.03 ± 0.11 33 400 ± 9400 71 900 ± 2600 6.12 ± 0.32 193.6 2.81 A328W 3200 ± 1900 41 000 ± 1700 6.59 ± 0.13 28 700 ± 700 28 700 ± 700 NA 227.8 4.11 E197Q 3200 ± 740 12 000 ± 280 5.85 ± 0.12 12 800 ± 1200 16 700 ± 730 6.17 ± 0.56 – – L286W 4300 ± 2700 16 200 ± 2100 6.26 ± 0.45 17 400 ± 4100 93 000 ± 2700 6.17 ± 0.10 – – V288W 6300 ± 1100 52 200 ± 1800 6.70 ± 0.07 62 900 ± 4200 89 300 ± 4200 6.08 ± 0.29 – – Y332A 2900 ± 1400 48 400 ± 1400 6.51 ± 0.07 27 500 ± 3400 64 700 ± 2600 6.29 ± 0.19 – – Ó FEBS 2003 High activity of BuChE at low pH (Eur. J. Biochem. 270) 319 argument that this high activity at low pH in the presence of excess substrate is a common feature of BTC hydrolysis by BuChE. Mutations in the acyl-binding pocket of mouse AChE (F297A and F297I) or in the hydrogen-bonding network (E450Q) have also been shown to alter the excess-substrate effect. In these cases, excess-substrate inhibition was switched into excess-substrate activation [10,35,36]. More- over, mutations in the p-cation-binding site of human and snake AChE have been found to cause activation by excess acetylthiocholine at low pH [14]. Taken together, these observations suggest that any change to the structure of the active site may alter the excess-substrate effect. Thus the response of cholinesterases to excess-substrate binding appears to involve the intricate interplay of a variety of residues in the active site. pH dependence of A328W The A328W mutant has a remarkable pH vs. activity profile (Fig. 2, bottom right panel). At low pH, bk cat is larger than k cat , but at high pH bk cat is smaller than k cat .Thatistosay, A328W goes from substrate activation, at low pH, to substrate inhibition, at high pH. A similar observation was made by Kalow [25] using benzoylcholine as substrate for wild-type human BuChE and by us with benzoylthiocholine as substrate on the same enzyme. In particular, the pH- dependence study of benzoylthiocholine hydrolysis by wild- type BuChE showed a progressive shift from activation by excess substrate (b > 1) at low pH to inhibition by excess substrate (b < 1) at pH > 7.1 (unpublished). There is no reason to believe that excess substrate binds to a different site at high pH than it does at low pH. Therefore the switch from substrate activation to substrate inhibition most probably reflects a pH-dependent difference in the response of the protein to excess-substrate binding. That is to say, the structure of the BuChE active site changes in response to excess-substrate binding, and this change is different at high pH than it is at low pH. Fig. 4. pH dependence for the turnover number (k cat ) in the absence of excess substrate and for the turnover number (bk cat ) in the presence of excess substrate, of the human BuChE mutant E197Q. The solid circles indicate the measured k cat values, the solid squares indicate the measured bk cat values, and the lines are the result of fitting the meas- ured rates to an expression for a single pK a (see the legend to Table 1 for details). The values for k cat and bk cat ,ateachpH,weretakenfrom fittings of k app vs. BTC concentration data (data not shown). Fig. 5. pH dependence for the turnover number (k cat ) in the absence of excess substrate and for the turnover number (bk cat ) in the presence of excess substrate, of various BuChE mutants. Each panel represents a different mutant form of BuChE, as indicated. In each panel, the solid circles indicate the measured k cat values, the solid squares indicate the measured bk cat values, and the lines are the result of fitting the meas- ured rates to an expression for a single pK a (see the legend to Table 1 for details). The values for k cat and bk cat ,ateachpH,weretakenfrom fittings of k app vs. BTC concentration data (see Materials and Meth- ods) for each mutant (data not shown). Fig. 3. pH dependence for the turnover number (k cat ) in the absence of excess substrate and for the turnover number (bk cat ) in the presence of excess substrate, of wild-type human BuChE over the pH range 4–8.5. From pH 5 to 8.5 the assays were performed in 0.1 M sodium phos- phate buffers. From pH 4 to 4.75. the assays were performed in 0.1 M sodium acetate buffers. The solid circles indicate the measured k cat values, the solid squares indicate the measured bk cat values, and the lines are the result of fittings. The k cat rates were fitted to an expression for a single pK a (see the legend to Table 1 for details). The bk cat rates werefittedtoanexpressionfortwopK a values [33]. k ¼ k A à K 2 à K 4 þ k H ýHÃK 4 þ k H2 ýH 2 K 2 à K 4 þ½HÃK 4 þ½H 2 The term k is the observed rate, k A is the limiting rate at high pH, k H is the rate for singly protonated form, k H2 is the limiting rate at low pH (zerointhiscase),K 2 is the dissociation constant for the first proto- nation, K 4 is the dissociation constant for the second protonation, and [H] is the hydrogen ion concentration. The values for k cat and bk cat ,at each pH, were taken from fittings of k app vs. BTC concentration data (data not shown). 320 P. Masson et al.(Eur. J. Biochem. 270) Ó FEBS 2003 Dependence of p K a on the size/hydrophobicity or aromaticity of residue 328 The original motivation for these studies was the hypo- thesis that the residue in position 328 significantly perturbed the pK a values for k cat and bk cat . This, however, is not the case. There is only a slight dependence of pK a on the size of the residue at position 328 (Fig. 6). It is, however, noteworthy, that with any given mutant the pK a for bk cat is generally lower than the pK a for k cat (Table 1). This difference indicates that excess-substrate binding to the PAS site of BuChE affects the environment of H438. It could be due to either the presumed conformational change induced on binding [11,12,15,37,38] or simply the presence of an additional positive charge close to the active site (in the form of BTC or any positively charged substrate) making protonation of the catalytic histidine more difficult. Plots of pK a values against the nonpolar surface area of the residue [31] or against any of a variety of residue hydrophobicity scales, e.g. Chothia’s residue accessible surface area scale [39] or Nozaki and Tanford’s water/ organic solvent partition scale [40], were similar to those in Fig. 6 (data not shown). It is not surprising that the correlation of pK a values with residue size is similar to the correlation of pK a values with residue hydrophobicity, as Chothia [41] has pointed out that hydrophobicity is directly related to the accessible surface area of the residue, i.e. size. In view of this, we believe that it is not possible to conclude whether the variations in pK a of bk cat of the A328 mutants are due to a steric or a hydropho- bic effect. Moreover, results with the bulkiest residue (mutant A328W) do not fit the pattern, suggesting that the tryptophan ring may affect the H438 pK a through p-cation interactions. Discussion The central observation in this paper is that BuChE retains significant hydrolytic activity after protonation of what appears to be the catalytic histidine, H438. This occurs under the influence of binding of excess substrate to the PAS, i.e. for bk cat , but not at lower substrate concentrations, i.e. for k cat in Scheme 1. This creates a major mechanistic puzzle. How can BuChE manage to turnover at a rate of % 20 000–50 000 min )1 when the catalytic histidine appears to be protonated? Let us review the accepted general mechanism of serine hydrolase catalysis. The mechanism of serine esterases is generally considered to be analogous to that of the serine proteases [18]. It goes through four steps [18,42,43] represented in Scheme 2 (human BuChE amino-acid numbering). First, the carbonyl carbon of the substrate undergoes a nucleophile attack by the Oc of the catalytic serine, while the proton is shuttled to the catalytic histidine. This results in the formation of a tetrahedral transition- state intermediate, the negative charge on the former carbonyl oxygen being stabilized by interactions with the dipoles of the oxyanion hole. Secondly, the alcohol product is released, picking up a proton from the catalytic histidine. This results in the formation of a transient acyl-enzyme adduct. The alcohol product is exchanged with a molecule of water. Thirdly, the acyl-enzyme undergoes a nucleophile Scheme 2. Proton shuttle mechanism. Fig. 6. Dependence of pK a for k cat (filled circles) and bk cat (open squares) on the volume of the residue at position 328. The pK a values were taken from fitting the data of Fig. 1 to an expression for a single pK a (see Table 1). The residue volume was taken from Zamyatnin [30] (see Table 1). The letters are the single letter codes for the amino acids at the 328 position. They are provided to help the reader to associate the data with the mutant. The lines are presented to emphasize the trend in these data. They have no analytical significance. Ó FEBS 2003 High activity of BuChE at low pH (Eur. J. Biochem. 270) 321 attack by this molecule and the dissociated proton is transferred to the catalytic histidine. This results in the formation of a second tetrahedral transition-state interme- diate, the negative charge on the former carbonyl oxygen being again stabilized by the oxyanion hole. Fourthly, the catalytic serine is released, picking up a proton from the catalytic histidine. This results in regeneration of the starting enzyme. There are currently two proposals for the driving force behind catalysis: the low-barrier hydrogen-bond model [44,45] and the electrostatic stabilization of the transition state model [46,47]. According to the low-barrier energy model, substrate binding drives a conformational change to form a Michaelis complex in which steric compression is intro- duced between the histidine and carboxylate (aspartate in chymotrypsin, glutamate in cholinesterases) of the cata- lytic triad. Compression of the His-Asp/Glu diad causes the basicity of the histidine to increase, so that it is able to accept/abstract the Oc proton from the catalytic serine. Transferring a proton from the catalytic serine to the catalytic histidine relieves the steric compression by forming a short, strong hydrogen bond between the protonated histidine and the carboxylate. In this way, the ability of the histidine to accept the proton from the serine Oc is greatly increased. The presence of a short, strong hydrogen bond was shown by NMR studies for both human AChE and human BuChE complexes with com- pounds mimicking the transition state [48,49]. In the electrostatic transition-state stabilization model, the dipoles of the oxyanion hole are considered to be optimally prealigned to strongly polarize the carbonyl of the substrate. The carbonyl carbon becomes very electrophilic and as the bond between the serine c-oxygen and the carbonyl carbon forms, the pK a of the Oc proton falls to a point where it can be released to the catalytic histidine, which is positioned to accept it. Thus, no free energy is spent on re-orienting the dipoles of the protein in the transition state. This leads to a large decrease in transition-state energy for the enzyme reaction compared with the chemical hydrolysis reaction in water, and thereby, a large increase in the rate for catalysis. It is noteworthy that the ability of the oxyanion hole to stabilize the tetrahedral transition state is well illustrated in the recently solved X-ray structure of human BuChE [8]. In this structure, the enzyme is not free, but the acidic product of substrate hydrolysis (butyrate) is still loosely bound to the catalytic serine (bond length is 2.16 A ˚ ). The carbonyl carbon of the butyrate adopts a partial tetrahedral character. Such a distortion results from the strong polarization of the C–O bond by the dipoles of the oxyanion hole in conjunction with the influence of a close nucleophile like the Oc of the catalytic serine. The same type of adduct was observed previously for Strepto- myces griseus protease A [50]. Thus, the butyrate–BuChE, quasi-tetrahedral complex is more stable than the free or the nonhydrated acyl-enzyme. We have found that the protonated form of BuChE, in the presence of excess substrate, i.e. bk cat , is active at low pH (see Fig. 2). It is assumed that the change in bk cat as a function of pH reflects protonation of the catalytic histidine. This observation generates a problem for any mechanism in which the catalytic histidine is the proton acceptor for the catalytic serine, because the likelihood of a protonated histidine accepting an additional proton is very low. Thus, at low pH and in the presence of excess substrate, any model for catalysis by human BuChE based on the histidine being the proton acceptor becomes untenable. The protonated form of the catalytic histidine may accept a proton in the transition state, therefore serving as a general base catalyst, only if a concerted proton transfer to the leaving group of the substrate occurs. Proton transfer The requirement of an acceptor for the Oc proton from the catalytic serine is a critical component of the mechanism for the serine proteases/esterases. Based on the 3D structures of a-chymotrypsin [44], AChE [6], and human BuChE [8], the most logical recipient for the serine proton is the catalytic histidine. Most models use the histidine to shuttle protons: first, between the serine and the leaving group; and then between the attacking water and the serine. In its role as a proton shuttle, the catalytic histidine formally accepts a proton from a donor and then delivers it to the acceptor (Scheme 3). On the other hand, our data at low pH indicate that catalysis by BuChE and AChE [14], in the presence of excess substrate, can proceed readily even when the catalytic histidine is protonated. Formal transfer of an additional proton to the cationic form of the histidine is not reasonable. This is the dilemma. One solution would be a concerted proton transfer, in the transition state, via the catalytic histidine. The concerted proton transfer would be between the serine and the leaving group and later between the water and the serine. Cyclic transition-state structures have been proposed for nonenzymatic acid catalysis of ester hydrolysis ([51] and references therein). Such a concerted transition state would not conflict with electrostatic stabilization of the transition state. Finally, a concerted proton transfer mechanism could also explain catalysis by AChE at low pH [14]. How- ever, further studies are needed to test this hypothetical mechanism. Thus, we suggest that BuChE and AChE may use two different mechanisms for transferring protons. At high pH, where the catalytic histidine is unprotonated, both choli- nesterases use the traditional proton shuttle mechanism (Scheme 3) both for k cat and bk cat . At low pH (where the catalytic histidine is protonated), and in the presence of excess substrate, the binding of which induces a conform- ational change, cholinesterases use another mechanism which remains to be elucidated. Scheme 3. Proton shuttle transition state. 322 P. Masson et al.(Eur. J. Biochem. 270) Ó FEBS 2003 Conclusion We have found that, as for AChE [14,34], the turnover of human BuChE reaches a substantial, nonzero limiting rate at low pH, in the presence of excess positively charged substrate. This observation suggests that catalysis at low pH, in the presence of excess substrate, does not involve the classical acid-base triad-mediated mechanism. However, involvement of general base catalysis by a carboxylate, i.e. E197(199), was disproved. The observations of high activity from BuChE and AChE at low pH is a new and important finding which requires further investigation to dissect the molecular mechanisms of hydrolysis of substrates by cholinesterases under extremes of pH and substrate concentration. Acknowledgements We are grateful to Lawrence M. Schopfer for fruitful discussions and constant support during this work. This work was supported by US Medical Research and Materiel Command grants DAMD 35-1905- 2010-00 (to O.L.), DGA/DSP/STTC 99CO-029/PEA and DGA/ ODCA Washington 00-2-032-0-00 (to P.M. and O.L.) and by a National Cancer Institute grant, P30CA36727, to the Eppley Institute. The opinions and assertions contained herein should not be construed as the official views of the US Army or the Department of Defense. References 1. Lockridge, O. (1990) Genetic variants of human serum choli- nesterase influence metabolism of the muscle relaxant succinyl- choline. Pharmacol. Ther. 47, 35–60. 2. Lockridge, O. (1992) Genetic variants of human serum butyryl- cholinesterase influence the metabolism of the muscle relaxant succinylcholine. In Pharmacogenetics of Drug Metabolism (Kalow, W., ed.), pp. 15–50. Pergamon Press, New York. 3. Raveh, L., Grunwald, J., Marcus, D., Papier, Y., Choen, E. & Ashani, Y. (1993) Human butyrylcholinesterase as a general prophylactic antidote for nerve agent toxicity. In vitro and in vivo quantitative characterization. Biochem. Pharmacol. 45, 2465–2474. 4. Lockridge, O., Bartels, C.F., Vaughan, T.A., Wong, C.K., Nor- ton, S.E. & Johnson, L.L. (1987) Complete amino acid sequence of human serum cholinesterase. J. Biol. Chem. 262, 549–557. 5. Harel, M., Sussman, J.L., Krejci, E., Bon, S., Chanal, P., Massoulie, J. & Silman. I. (1992) Conversion of acetylcholinester- ase to butyrylcholinesterase: modeling and mutagenesis. Proc. Natl. Acad. Sci. USA 89, 10827–10831. 6. Sussman, J.L., Harel, M., Frolow, F., Oefner, C., Goldman, A., Toker, L. & Silman, I. (1991) Atomic structure of acetyl- cholinesterase from Torpedo californica: a prototypic acetylcho- line-binding protein. Science 253, 872–879. 7. Nachon, F., Nicolet, Y., Viguie ´ , N., Masson, P., Fontecilla- Camps, J.C. & Lockridge, O. (2002) Engineering of a monomeric and low-glycosylated form of human butyrylcholinesterase. Eur. J. Biochem. 269, 630–637. 8. Nicolet,Y.,Nachon,F.,Masson,P.,Lockridge,O.&Fontecilla- Camps, J.C. (2003) Crystal structure of recombinant human butyrylcholinesterase: new insights into the catalytic mechanisms of cholinesterases. Proceedings of the XIth Cholinergic Mechanisms Symposium (Fisher, A. & Soreq, H., eds). Martin Dunitz Ltd, London (in press). 9. Saxena, A., Redman, A.M.G., Jiang, X., Lockridge, O. & Doctor, B.P. (1997) Differences in active site gorge dimensions of choli- nesterases revealed by binding of inhibitors to human butyryl- cholinesterase. Biochemistry 36, 14642–14651. 10. Radic, Z., Pickering, N.A., Vellom, D.C., Camp, S. & Taylor, P. (1993) Three distinct domains in the cholinesterase molecule confer selectivity for acetyl- and butyrylcholinesterase inhibitors. Biochemistry 32, 12074–12084. 11. Masson, P., Froment, M T., Bartels, C.F. & Lockridge, O. (1996) Asp70 in the peripheral anionic site of human butyrycholinester- ase. Eur. J. Biochem. 235, 36–48. 12. Mallender, W.D., Szegletes, T. & Rosenberry, T.L. (2000) Acetyl- thiocholine binds to Asp74 at the peripheral site of human acetylcholinesterase as the first step in the catalytic pathway. Biochemistry 39, 7753–7763. 13. Tougu, V. (2001) Acetylcholinesterase: mechanism of catalysis and inhibition. Curr.Med.Chem.1, 155–170. 14. Masson, P., Schopfer, L.M., Bartels, C.F., Froment, M T., Ribes, F., Nachon, F. & Lockridge, O. (2002) Substrate activation in acetylcholinesterase induced by low pH or mutation in the p-cation subsite. Biochim. Biophys. Acta 1594, 313–324. 15. Masson, P., Xie, W., Froment, M T., Levitsky, V., Fortier, P L., Albaret, C. & Lockridge, O. (1999) Interaction between the peripheral site residues of human butyrylcholinesterase, D70 and Y332, in binding and hydrolysis of substrates. Biochim. Biophys. Acta 1433, 281–293. 16. Masson, P., Xie, W., Froment, M T. & Lockridge, O. (2001) Effects of mutations of active site residues and amino acids interacting with the W loop on substrate activation of butyryl- cholinesterase. Biochim. Biophys. Acta 1544, 166–176. 17. Xie, W., Varkey-Altamirano, C., Bartels, C.F., Speirs, R.J., Cashman, J.R. & Lockridge, O. (1999) An improved cocaine hydrolase: the A328Y mutant of human butyrylcholinesterase is 4-fold more efficient. Mol. Pharmacol. 55, 83–91. 18. Quinn, D.M. (1987) Acetylcholinesterase: enzyme structure, reac- tion dynamics, and virtual transition state. Chem. Rev. 87, 955–979. 19. Harel,M.,Quinn,D.M.,Nair,H.K.,Silman,I.&Sussman,J.L. (1996) The X-ray structure of a transition state analog complex reveals the molecular origins of the catalytic power and substrate specificity of acetylcholinesterase. J. Am. Chem. Soc. 118, 2340– 2346. 20. Amitai, G., Moorad, D., Adani, R. & Doctor, B.P. (1998) Inhibition of acetylcholinesterase and butyrylcholinesterase by chlorpyrifos-oxon. Biochem. Pharmacol. 56, 293–299. 21. Ellman, G.L., Courtney, K.D., Andres, V. & Featherstone, R.M. (1961) A new and rapid colorimetric determination of acetyl- cholinesterase activity. Biochem. Pharmacol. 7, 88–95. 22. Tougu, V. & Kesvatera, T. (1996) Role of ionic interactions in cholinesterase catalysis. Biochim. Biophys. Acta 1298, 12–30. 23. Riddles, P.W., Blakeley, R.L. & Zerner, B. (1979) Ellman’s reagent: 5,5¢-dithiobis (2-nitrobenzoic acid), a reexamination. Anal. Biochem. 94, 75–81. 24. Webb, J.L. (1963) Enzyme and Metabolic Inhibitors,Vol.1,p.36. Academic Press, New York. 25. Kalow, W. (1964) The influence of pH on the hydrolysis of benzoylcholine by pseudocholinesterase of human plasma. Can. J. Physiol. 42, 161–168. 26. Lockridge, O., Blong, R.M., Masson, P., Froment, M T., Millard, C.B. & Broomfield, C.A. (1997) A single amino acid substitution, Gly117His, confers phosphotriesterase (organopho- sphorus acid anhydride hydrolase) activity on human butyryl- cholinesterase. Biochemistry 36, 786–795. 27. Masson, P., Legrand, P., Bartels, C.F., Froment, M T., Schopfer, L.M. & Lockridge, O. (1997) Role of aspartate 70 and tryptophan 82 in binding of succinyldithiocholine to human butyryl- cholinesterase. Biochemistry 36, 2266–2277. 28. Levitsky, V., Xie, W., Froment, M T., Lockridge, O. & Masson, P. (1999) Polyol-induced activation by excess substrate of Ó FEBS 2003 High activity of BuChE at low pH (Eur. J. Biochem. 270) 323 the D70G butyrylcholinesterase mutant. Biochim. Biophys. Acta 1429, 422–430. 29. Irvin, J.L. & Irvin, E.M. (1947) Spectrophotometric and potentiometric evaluation of apparent acid dissociation exponents of various 4-aminoquinolines. J. Am. Chem. Soc. 69, 1091–1099. 30. Zamyatnin, A.A. (1972) Protein volume in solution. Prog. Biophys. Mol. Biol. 24, 109–123. 31. Karplus, P.A. (1997) Hydrophobicity regained. Protein Sci. 6, 1302–1307. 32. Main, A.R. (1979) Mode of action of anticholinesterases. Phar- macol. Ther. 6, 579–628. 33. Fersht, A. (1999) Structure and Mechanism in Protein Science, pp. 172–173. W.H. Freeman, San Francisco. 34. Selwood,T.,Feaster,S.R.,States,M.J.,Pryor,A.N.&Quinn, D.M. (1993) Parallel mechanisms in acetylcholinesterase- catalyzed hydrolysis of choline esters. J. Am. Chem. Soc. 115, 10477–10482. 35. Hosea, N.A., Berman, H.A. & Taylor, P. (1995) Specificity and orientation of trigonal carboxyl esters and tetrahedral alkyl- phosphonyl esters in cholinesterase. Biochemistry 34, 11528– 11536. 36. Hosea, N.A., Radic, Z., Tsigelny, I., Berman, H.A., Quinn, D.M. & Taylor, P. (1996) Aspartate 74 as a primary determinant in acetylcholinesterase governing specificity to cationic organophos- phates. Biochemistry 35, 10995–11004. 37. Radic, Z., Reiner, E. & Taylor, P. (1991) Role of the peripheral anionic site on acetylcholinesterase: inhibition by substrates and coumarin derivatives. Mol. Pharmacol. 39, 98–104. 38. Shafferman, A., Belan, B., Ordentlich, A., Kronman, C., Grosfeld, H.,Leitner,M.,Flashner,Y.,Cohen,S.,Barak,D.&Ariel,N. (1992) Substrate inhibition of acetylcholinesterase: residues affecting signal transduction from the surface to the catalytic center. EMBO J. 11, 3561–3568. 39. Chothia, C. (1976) The nature of the accessible and buried surfaces in proteins. J. Mol. Biol. 105, 1–14. 40. Nozaki, Y. & Tanford, C. (1971) The solubility of amino acids and two glycine peptides in aqueous ethanol and dioxane solutions. J. Biol. Chem. 246, 2211–2217. 41. Chothia, C. (1974) Hydrophobic bonding and accessible surface area in proteins. Nature (London) 248, 338–339. 42. Kraut, J. (1977) Serine proteases: structure and mechanism of catalysis. Annu. Rev. Biochem. 46, 331–358. 43. Blow, D.M., Birktoft, J.J. & Hartley, B.S. (1969) Role of a buried acid group in the mechanism of action of chymotrypsin. Nature (London) 221, 337–340. 44. Cleland, W.W., Frey, P.A. & Gerlt, J.A. (1998) The low barrier hydrogen bond in enzymatic catalysis. J. Biol. Chem. 273, 25529– 25532. 45. Cassidy, C.S., Lin, J. & Frey, P.A. (1997) A new concept for the mechanism of action of chymotrypsin: the role of the low barrier hydrogen bond. Biochemistry 36, 4576–4584. 46. Warshel, A. (1998) Electrostatic origin of the catalytic power of enzymes and the role of preorganized active sites. J. Biol. Chem. 273, 27035–27038. 47. Fuxreiter, M. & Warshel, A. (1998) Origin of the catalytic power of acetylcholinesterase: computer simulation studies. J. Am. Chem. Soc. 120, 183–194. 48. Viragh, C., Harris, T.K., Reddy, P.M., Massiah, M.A., Mildvan, A.S. & Kovach, I.M. (2000) NMR evidence for a short, strong, hydrogen bond at the active site of a cholinesterase. Biochemistry 39, 16200–16205. 49. Massiah, M.A., Viragh, C., Reddy, P.M., Kovach, I.M., Johnson, J., Rosenberry, T.L. & Mildvan, A.S. (2001) Short, strong hydrogen bonds at the active site of human acetylcholinesterase: proton NMR studies. Biochemistry 40, 5682–5690. 50. James, M.N., Sielecki, A.R., Brayer, G.D., Delbaere, L.T. & Bauer, C.A. (1980) Structures of product and inhibitor complexes of Streptomyces griseus protease A at 1.8 A ˚ resolution. A model for serine protease catalysis. J. Mol. Biol. 144, 43–88. 51. Bender, M.L. (1960) Mechanisms of catalysis of nucleophilic reactions of carboxylic acid derivatives. Quart. Rev. 60, 53–113. 324 P. Masson et al.(Eur. J. Biochem. 270) Ó FEBS 2003 . both for k cat and bk cat . At low pH (where the catalytic histidine is protonated), and in the presence of excess substrate, the binding of which induces. K 4 þ½HÃK 4 þ½H 2 The term k is the observed rate, k A is the limiting rate at high pH, k H is the rate for singly protonated form, k H2 is the limiting rate at low pH (zerointhiscase),K 2 is