Solid-State 17O NMR of Unstable Acyl-Enzyme Intermediates: ADirect Probe of Hydrogen Bonding Interactions in the Oxyanion Holeof Serine ProteasesAaron W. Tang,† Xianqi Kong,† Victor Terskikh,†,‡ and Gang Wu*,†

†Department of Chemistry, Queen’s University, 90 Bader Lane, Kingston, Ontario K7L 3N6, Canada‡Department of Chemistry, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada

*S Supporting Information

ABSTRACT: We report preparation, trapping, and solid-state17O NMR characterization of three unstable acyl-enzymeintermediates (≈ 26 kDa): p-N,N-dimethylamino-[17O]-benzoyl-chymotrypsin, trans-o-methoxy-[17O]cinnamoyl-chymo-trypsin, and trans-p-methoxy-[17O]cinnamoyl-chymotrypsin. Weshow that both the 17O chemical shifts and nuclear quadrupolarparameters obtained for these acyl-enzyme intermediates in thesolid state are correlated with their deacylation rate constantsmeasured in aqueous solution. With the aid of quantummechanical calculations, the experimental 17O NMR parameterswere interpreted as to reflect the hydrogen bonding interactionsbetween the carbonyl (C17O) functional group of the acyl moiety and the two NH groups from the protein backbone (Ser195and Gly193) in the oxyanion hole, a general feature of all serine proteases. Our results further suggest that the 17O chemical shiftand quadrupole coupling constant display distinctly different sensitivities toward different aspects of hydrogen bonding, such ashydrogen bond distance and direction. This work demonstrates the utility of 17O as a useful nuclear probe in NMR studies ofenzymes.

1. INTRODUCTION

Oxygen is one of the most common elements in organic andbiological molecules, but remains largely inaccessible as anuclear probe in NMR studies of these molecular systems.While the exceedingly low natural abundance (0.037%) of theonly NMR-active oxygen isotope, 17O, poses some challenges,the primary obstacle of 17O NMR studies is the intrinsicquadrupolar nature of the 17O nucleus (spin-5/2). In general,NMR spectra for quadrupolar nuclei have significantly lowerresolution than those from more conventional spin-1/2 probessuch as 1H, 13C, and 15N. However, recent advances haveshown that, with the availability of ultrahigh magnetic fields(e.g., 21 T), 17O NMR is beginning to become applicable instudies of biological macromolecules in both solution and solidstate.1−9

In this work, we explore the utility of solid-state 17O NMR at21.1 T as a new technique for studying unstable intermediatesformed in enzymatic reactions. In particular, we set out toinvestigate whether 17O NMR can be used to probe thehydrogen bonding interactions between the substrate andprotein backbone in the so-called oxyanion hole of a modelenzyme, chymotrypsin (26 kDa); see Scheme 1. Chymotrypsinbelongs to a family of enzymes known as serine proteases,named for a catalytically active nucleophilic serine residue inthe active site. For chymotrypsin, the three catalytic residuesknown as the “catalytic triad”, Ser195, His57, and Asp102, form

a hydrogen-bond network at the active site.10,11 This hydrogen-bond network activates Ser195 for nucleophilic attack on thesubstrate. It is well-established that the serine protease catalyzesthe hydrolysis reaction in two stages: first, the acylation stepwhere the substrate is covalently bonded to Ser195 to form theacyl-enzyme intermediate, breaking the amide or ester bondand releasing part of the substrate with the free amino terminus,and second, the deacylation step where the ester bond in theacyl-enzyme intermediate is hydrolyzed. When the substratecontains a suitably stable leaving group, the formation of theacyl-enzyme intermediate occurs rapidly and the hydrolysis ofthe acyl-enzyme is rate-limiting.12

Received: August 31, 2016Revised: October 11, 2016

Scheme 1. Formation and Hydrogen Bonding Environmentof the Oxyanion Hole in Serine Proteases

Article

pubs.acs.org/JPCB

© XXXX American Chemical Society A DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

In addition to the “catalytic triad”, another common featurein serine proteases is the presence of the oxyanion hole.13 Asseen in Scheme 1, the oxyanion hole, first identified in thepioneering work of Henderson,14 consists of two hydrogenbond donors from the backbone of Ser195 and Gly193, thatstabilize the oxygen on the scissile amide/ester bond in thetetrahedral transition state. The oxyanion hole is important forboth the acylation and deacylation steps. The contribution ofthe oxyanion hole to catalysis is significant, as enzymes with thecatalytic triad disabled via site-directed mutagenesis still retainhydrolytic activity with a reaction rate a thousand-fold higherthan that for the uncatalyzed hydrolysis.15,16

Inspired by previous resonance Raman spectroscopic studiesof serine proteases that showed a correlation between theCO stretching frequency and acyl-enzyme deacylationrate,17−20 we hypothesize that 17O NMR should be a sensitiveprobe to the hydrogen bonding interaction at the carbonyloxygen atom of the substrate which is located at the center ofthe oxyanion hole. Our general strategy consists of twocomponents. First, we chose to use substrates that are known toform acyl-enzyme intermediates with relatively slow deacylationrates. Second, once the acyl-enzyme intermediates are formedin aqueous solution, we attempt to trap them by quickly freeze-drying the solution and then perform solid-state 17O NMRmeasurements.

2. EXPERIMENTAL SECTION2.1. Synthesis of 17O-Labeled N-Acyl Imidazole

Substrates. Three N-acyl-imidazoles (Scheme 2) were

prepared by coupling imidazole with the 17O-labeled carboxylicacids using N,N′-carbonyldiimidazole (CDI). Synthetic detailsfor preparation of 17O-labeled carboxylic acid precursors areprovided in the Supporting Information.p-N,N-Dimethylamino-[17O]benzoylimidazole (DAB-Im).

This compound was synthesized following a method describedpreviously.21 p-N,N-Dimethylamino-[17O2]benzoic acid (100.0mg, 0.6054 mmol) and CDI (148.6 mg, 0.9164 mmol, 1.514mol equiv) were added to a 10 mL round-bottom flask, whichwas then capped with a rubber septum. The headspace waspurged with nitrogen gas with a needle for 1 min, andanhydrous THF (3 mL) was added with a syringe and needle.Some insoluble gray/blue solids remained. The headspace wasagain purged with nitrogen gas for 5 min. The reaction mixturewas stirred for 19 h, after which insoluble solid remained visible.The solvent was removed under reduced pressure until a light-gray oil and solid remained. The residue was transferred to anextraction funnel with CHCl3 (2 × 2 mL). The organic layerwas washed with NaHCO3 solution (2 × 2 mL 0.8% w/v),saturated Na2SO4 solution (3 mL), and then dried (Na2SO4).The organic solvent was removed under reduced pressure untila light gray/purple solid remained. The residue was dissolved inCHCl3 (2 mL), and some activated charcoal was added. Themixture was filtered through a Pasteur pipet with a cotton plug,packed with a layer of Celite. The pipet was rinsed with CHCl3

(2 mL, 1 mL). The eluent was light purple, and the solvent wasremoved under reduced pressure. The off-white residue wasdried under vacuum overnight. The residue was dissolved inacetone (1.5 mL) and precipitated with cold H2O (8 mL),forming a white, flocculent solid. The solid was collected viasuction filtration, and washed with cold H2O (2 × 4 mL) andhexanes (2 × 2 mL) to yield 47.5 mg (36.5%) of final product.1H NMR (400 MHz, acetone-d6) δ 8.09 (s, 1H), 7.77 (dt, J =9.2, 2.2 Hz, 2H), 7.58 (t, J = 1.38 Hz, 1H), 7.10 (d, J = 0.5 Hz,1H), 6.87 (dt, J = 9.1, 2.2 Hz, 2H), 3.14 (s, 6H); 17O NMR(54.1 MHz, acetone-d6) δ 410.4 (br).

trans-o-Methoxy-[17O]cinnamoylimidazole (oMC-Im).trans-o-Methoxy-[17O2]cinnamic acid (200 mg, 1.12 mmol)and CDI (303 mg, 1.87 mmol, 1.67 mol equiv) were added to a10 mL round-bottom flask. Anhydrous THF (4 mL) was addedand the flask was sealed with a rubber septum stopper. Theheadspace of the flask was purged with nitrogen gas introducedwith a needle. The reaction mixture was stirred for 13 h. Thesolvent was evaporated under reduced pressure. Crude productwas dissolved in dichloromethane (5 mL) and toluene (15mL), and washed with NaHCO3 solution (2.5 mL 0.8% w/v),saturated Na2SO4 solution (3 × 3 mL). The organic layer wasdried (Na2SO4), then the solvent was removed to yield a crudesolid product. The residue was recrystallized from acetone (14mL) and cold H2O (50 mL), then washed with cold H2O (3 ×10 mL) to yield 181 mg (70.7%) of final product. 1H NMR(500 MHz, acetone-d6) δ 8.54 (s, 1H), 8.38 (d, J = 15.6 Hz,1H), 7.92 (dd, J = 7.7, 1.6 Hz, 1H), 7.82 (t, J = 1.3 Hz, 1H),7.61 (d, J = 15.6 Hz, 1H), 7.50 (ddd, J = 8.7, 6.9, 1.6 Hz, 1H),7.15 (d, J = 8.5 Hz, 1H), 7.10 (s, 1H), 7.05 (t, J = 7.57 Hz, 1H),3.99 (s, 3H); 17O NMR (67.7 MHz, acetone-d6) δ 384.5 (br).

trans-p-Methoxy-[17O]cinnamoylimidazole (pMC-Im). Thesynthesis of this compound was also described previously.21

trans-p-Methoxy-[17O2]cinnamic acid (200 mg, 1.12 mmol) andCDI (303 mg, 1.87 mmol, 1.67 mol equiv) were added to a 10mL round-bottom flask. Anhydrous THF (4 mL) was addedand the flask was sealed with a rubber septum stopper. Theheadspace of the flask was purged with nitrogen gas introducedwith a needle. The reaction mixture was stirred for 13 h. Thesolvent was evaporated under reduced pressure. Crude productwas dissolved in dichloromethane (5 mL) and toluene (15mL), and washed with NaHCO3 solution (2.5 mL 0.8% w/v),saturated Na2SO4 solution (3 × 3 mL). The organic layer wasdried (Na2SO4), then the solvent was removed to yield a crudesolid product. The residue was recrystallized from acetone (14mL) and cold H2O (50 mL), and then washed with cold H2O(3 × 10 mL) to yield the final product. 1H NMR (300 MHz,CDCl3) δ 8.33 (s, 1H), 8.06 (d, 1H, J = 15.5 Hz), 7.63−7.65 (d+s, 3H), 7.10 (s, 1H), 6.98 (d, 2H, J = 6.6 Hz), 6.95 (d, 1H, J =15.3 Hz), 3.90 (s, 3H); 17O NMR (67.7 MHz, acetone-d6) δ352.9 (br).

2.2. Chymotrypsin Activity Assay. The activity of α-chymotrypsin (from bovine pancreas, 3× crystallization,essentially salt free, purchased from Sigma-Aldrich) wasmeasured with a chymotrypsin activity assay adopted fromthe testing protocol of Sigma-Aldrich,22 where the enzyme-catalyzed hydrolysis of N-benzoyl-L-tyrosine ethyl ester (BTEE)was tracked via UV−vis spectroscopy. The 1.18 mM BTEEsolution was prepared by dissolving 18.5 mg BTEE in 31.7 mLmethanol, and filled to the mark with water in a 50 mLvolumetric flask. A 2 M CaCl2 solution was prepared bydissolving 221.96 mg/mL anhydrous CaCl2. HCl solution (1mM) was prepared by serial dilution from concentrated HCl.

Scheme 2. Molecular Structures of 17O-Labeled N-AcylImidazoles

The Journal of Physical Chemistry B Article

DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

B

The enzyme solution was prepared by dissolving enough α-chymotrypsin (approximately 30−80 μg) in 1 mM HCl toreach approximately 2−5 units/mL (1 unit is defined as 1μmole BTEE consumed per minute at pH 7.80 at 25 °C). Then1.420 mL of Tris buffer (80 mM, pH 7.80), 1.400 mL of BTEE,and 80 μL of CaCl2 solutions were mixed in a quartz cuvetteand equilibrated at 25 °C. 100 μL of the enzyme solution wasadded to the cuvette, which was inverted five times. UV−visspectra were recorded immediately with a JASCO J-815 CDspectrometer using time course measurement at 256 nm over3−5 min (slit width = 2500 nm).2.3. Preparation, ESI-MS Characterization, and Kinetic

Measurement of Acyl-Enzyme Intermediates. p-N,N-Dimethylamino-[17O]benzoyl-chymotrypsin (DAB-CHT). α-Chymotrypsin (50.1 mg) was dissolved in acetate buffer (1mL 0.1 M, pH 4.00) in a 2.5 mL microcentrifuge tube. p-N,N-Dimethylamino-[17O]benzoylimidazole (4.4 mg, 10 mol equiv)was dissolved in acetonitrile (100 μL). The two solutions weremixed and allowed to react for 30 min. Four 1 mL plasticsyringes were plugged with glass wool and packed withSephadex G-25. The reaction mixture was split into four 275 μLfractions and added to the syringes, which were thencentrifuged briefly. The gel filtration process successivelyremoved the excess free substrates, which was confirmed bysolution 17O NMR. The eluent fractions were combined andimmediately frozen in a dry ice/acetone bath, and lyophilizedovernight to yield 39.1 mg (77%) of solid product.trans-o-Methoxy-[17O]cinnamoyl-chymotrypsin (oMC−

CHT). α-Chymotrypsin (25.3 mg) was dissolved in acetatebuffer (1 mL 0.1 M, pH 4.00) in a 2.5 mL microcentrifuge tube.trans-o-Methoxy-[17O]cinnamoylimidazole (3.8 mg) was dis-solved in acetonitrile (165.2 μL). One hundred microliters (10mol equiv) of the above solution was added to the enzyme andallowed to react for 1 min. Following the same procedure ofrunning the Sephadex G-25 column and lyophilization asdescribed earlier, we obtained 20.6 mg (81%) of solid product.trans-p-Methoxy-[17O]cinnamoyl-chymotrypsin (pMC−

CHT). α-Chymotrypsin (25.1 mg) was dissolved in acetatebuffer (1 mL 0.1 M, pH 4.00) in a 2.5 mL microcentrifuge tube.trans-p-Methoxy-[17O]cinnamoylimidazole (2.8 mg, 12 molequiv) was dissolved in acetonitrile (100 μL). The twosolutions were mixed and allowed to react for 1 min. Followingthe same procedure of running the Sephadex G-25 column andlyophilization as described earlier, we obtained 20.1 mg (80%)of solid product.Mass Spectrometry. Formation of covalently bonded acyl-

enzymes was confirmed via electrospray ionization ion-trapmass spectrometry (ESI-MS). Mass spectra were recorded on aThermo Scientific Orbitrap Velos Pro Mass spectrometer withESI in the positive ion mode. Spectral deconvolution wasperformed using Thermo Scientific ProMass, under positive ionmode with an adduct ion mass of 1.0079 Da. The output massrange was set to 20 000 Da to 80 000 Da, the default setting forlarge proteins.Kinetic Measurement of Deacylation. Deacylation rates of

the acyl-enzymes were followed using UV−vis spectrosco-py23,24 on a JASCO J-815 CD spectrometer. The acyl-enzymesolution was prepared by dissolving 1.5 mg lyophilized acyl-chymotrypsin intermediate in 3 mL buffer (80 mM Trizma, 50mM CaCl2, pH 7.80) in a quartz cuvette. UV−vis spectrum wasrecorded over wavelengths of 250−350 nm (slit width = 100μm, data pitch = 0.1 nm, scanning speed =100 nm/min), withsample cell kept at 25.0 °C. The data collection time points

ranged between a period of 120 h for DAB-CHT, 210 min foroMC-CHT, and 10 min for pMC-CHT. The last reading foreach inhibitor is treated as the baseline against which adifference spectrum can be calculated, as enough time haspassed to assume that hydrolysis of the acyl-enzyme complex iscomplete. Spectroscopic data for deacylation are provided asthe Supporting Information.

2.4. Solid-State 17O NMR. Solid-state 17O NMR experi-ments were performed on a Bruker Avance-II 900 (21.1 T)NMR spectrometer. A rotor-synchronized Hahn-echo sequencewas used for MAS experiments to eliminate the acoustic ringingfrom the probe. A 3.2 mm Bruker HX MAS probe was used onwhich the effective 90° pulse for the 17O central transition(CT) was 1.0 μs. A liquid H2O sample was used for both RFpower calibration and 17O chemical shift referencing (δ = 0ppm). All spectral simulations were performed with DMfit.25

2.5. Computational Details. Quantum chemical calcu-lations for NMR parameters were performed with Gaussian0926 on High Performance Computing Virtual Laboratory(HPCVL) servers. The calculations were performed using theBecke-3-Parameter, Lee−Yang−Parr (B3LYP) exchange func-tional, and a 6-311++g(d,p) basis set. The experimental solid-state 17O NMR results obtained for p-methoxy-[17O]cinnamatemethyl ester (pMC-Me) were used to calibrate the 17O nuclearquadrupolar moment (Q = −2.305 fm2) as demonstratedpreviously.27,28 The computed 17O magnetic shielding con-stants (σ) were converted to chemical shifts (δ) using δ = 270.2ppm − σ. The experimental and computed 17O NMRparameters for pMC-Me are provided in SupportingInformation.

3. RESULTS AND DISCUSSION

As acyl-enzyme intermediates are generally unstable in aqueoussolution, our strategy was to quickly freeze-dry the reactionsolution so that the acyl-enzymes can be trapped in the solid

Figure 1. Deconvoluted ESI mass spectra of (a) DAB-CHT, (b) oMC-CHT, and (c) pMC-CHT.

The Journal of Physical Chemistry B Article

DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

C

state. The deacylation of acyl-enzymes is known to occur veryslowly in the solid state.14 To confirm the identity of the solidacyl-enzyme products formed between α-chymotrypsin and thethree N-acyl imidazole substrates shown in Scheme 2, weanalyzed them via electrospray ionization-ion trap massspectrometry (ESI-MS). As seen from Figure 1, the majorpeaks in the ESI mass spectra of DBA-CHT, oMC-CHT, andpMC-CHT appear at 25597.2 ± 1.5, 25608.9 ± 0.3, and25609.7 ± 0.6 Da, respectively. The parent enzyme peak wasfound to be at 25449.2 ± 0.7 Da, which corresponds to the α1-isoform of CHT.29,30 The mass differences between the acyl-enzyme and the free enzyme match those expected for thethree moieties, DAB (147.2Da), oMC (160.2 Da), and pMC(160.2 Da). It is also interesting to note that, in each case, boththe acyl-CHT and free CHT major peaks are flanked bysatellite peaks. The MS satellite peaks having higher masses

than the free CHT and acyl-CHT are clearly due to differentNa+ adduct ions because of the mass separation of a multiple of22 Da. The MS peaks below the base peaks of CHT and acyl-CHT appear at M-18 and can be attributed to the presence of aminor isoform of the enzyme, δ-CHT, which has a theoreticalmolecular mass of 25 430.9 Da.30 The α1- and δ-isoforms ofCHT differ only in that the peptide bond of Thr147 ishydrolyzed in the α1-isoform during activation from chymo-trypsinogen A, resulting in a mass difference of a watermolecule (18 Da). The presence of multiple isoforms incommercial preparations of α-CHT has been previouslydemonstrated.29,30

We should note that the coexistence of α1- and δ-isoforms ofCHT in our samples does not affect the conclusion of thepresent study because the minute difference between the twoisoforms does not markedly alter the overall properties of theprotein, nor does it eliminate the catalytic activity of theenzyme.31,32 More importantly, the goal of the present work isto study the hydrogen bonding interaction in the oxyanion holeof chymotrypsin, and these two isoforms are identical in thisaspect. As also seen from Figure 1, the very weak peaks (>25700 Da) are likely due to some protein impurities, but they arenot identified at this time. On the basis of the peak areas of theacyl-CHT and free CHT signals, we estimated the level ofacylation to be in a range from 79 to 92% for the three acyl-CHT products; see Table 1. These are very satisfactory results.Now that the integrity of the solid acyl-enzyme samples is

established, the solid-state 17O NMR spectra can then be

Table 1. Experimental Deacylation Rate Constants (k3),a Percent Acylation,b and 17O NMR Tensor Parameters Obtained for

the Three Acyl-Enzyme Intermediates

acyl-enzyme k3 (s−1) % acylation δiso (ppm) Ω (ppm)c κd CQ (MHz) ηQ

DAB-CHT (6.4 ± 0.1) × 10−6 92 ± 5 323 ± 5 650 ± 80 0.2 ± 0.2 10 ± 1 0.6 ± 0.2oMC-CHT (2.8 ± 0.1) × 10−4 87 ± 5 319 ± 5 560 ± 80 0.2 ± 0.2 9.5 ± 0.8 0.6 ± 0.2pMC-CHT (8.5 ± 0.3) × 10−3 79 ± 5 288 ± 5 470 ± 80 0.2 ± 0.2 7.0 ± 0.5 0.8 ± 0.2

aMeasured in aqueous solution at pH 7.8 and 25 °C as described in the Experimental Section. bDetermined from ESI-MS data shown in Figure 1.cSpan Ω = δ11 − δ33.

dSkew κ = 3(δ22 − δiso)/Ω.

Figure 2. Experimental (blue trace) and simulated (red trace) 17OMAS NMR spectra of (a) DAB-CHT, (b) oMC-CHT, and (c) pMC-CHT. The weak signal marked with * at 380 ppm is from the ZrO2rotor. For each sample, approximately 20 mg of solid acyl-enzymewere packed into a 3.2 mm ZrO2 rotor. The sample spinningfrequency was 20 kHz and the recycle delay of 30 ms. A total of 2.0 ×106 transients were collected for each spectrum (total experimentaltime ≈ 23 h).

Table 2. Summary of Calculated 17O NMR Parameters forAcyl-Enzyme Analogsa

moleculeδiso

(ppm)Ω

(ppm) κCQ

(MHz) ηQ

DAB-Et 329.1 629.3 0.64 8.44 0.0oMC-Et 330.1 630.6 0.68 8.32 0.0pMC-Et 326.1 627.0 0.67 8.27 0.0(exptl for pMC-Me) (325) (512) (0.75) (8.20) (0.02)aThe experimental 17O NMR data of p-methoxycinnamic acid methylester (pMC-Me) are given in parentheses. The solid-state 17O NMRspectra of pMC-Me are provided in the Supporting Information(Figure S4).

Figure 3. (a) A computational model for Oγ-acetyl-chymotrypsin. TheCα-Cβ-Oγ-C1 dihedral angle (φ) and Cβ-Oγ-C1 bond angle (θ) aredefined. The two hydrogen bonds between backbone amide protonsfrom Ser195 and Gly193 and the carbonyl oxygen atom are shown tohighlight the oxyanion hole. (b) Definition of torsional angles Θ193 andΘ195 for describing the directions of the two hydrogen bonds in theoxyanion hole of acyl-enzymes.

The Journal of Physical Chemistry B Article

DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

D

attributed with certainty to the trapped acyl-enzymeintermediates. Each of the 17O MAS NMR spectra shown inFigure 2 consists of a central signal flanked by two weakspinning sidebands. These 17O NMR signals can be properlymodeled by considering both the second-order quadrupolarand magnetic shielding anisotropy interactions. The 17O NMRparameters obtained from such spectral analyses are summar-ized in Table 1, together with the kinetic data measured for thethree acyl-enzymes. It is interesting to note that the deacylationrate constant (k3) is increased from DAB-CHT to pMC-CHTby 1000-fold. Our k3 values are comparable to the literaturedata reported for the same acyl-enzymes.33,34 We should pointout that, because the experimental values on the 17O chemicalshift anisotropy (i.e., the CS tensor components) are far lessreliable than the isotropic chemical shift values, in thediscussion that follows we will discuss only the isotropicchemical shift (δiso) and quadrupole coupling constant (CQ).As seen from Table 1, the values of δiso and CQ observed for

acyl-enzymes are generally consistent with those previouslyreported for the ester functional group.35,36 Among the threeacyl-enzymes, we found the following trend in the isotropic 17Ochemical shifts (δiso): DAB-CHT > oMC-CHT > pMC-CHT.The same trend also holds for CQ. More interestingly, both ofthe 17O NMR parameters appear to be correlated with thedeacylation rate constants measured in aqueous solution forthese acyl-enzymes, as shown in Table 1. That is, the smallerthe k3, the larger the value of δiso (or CQ). As far as the

17ONMR parameters are concerned, numerous previous studies ofcarbonyl compounds have firmly established that, for theoxygen atom involved in a hydrogen bond, its δiso and CQvalues decrease with the increase in the hydrogen bondingstrength.37−48 Thus, the observed trends in δiso and CQ amongthe three acyl-enzyme intermediates suggest that the hydrogenbonding strength is in the following order: DAB-CHT < oMC-CHT < pMC-CHT. Previous Raman spectroscopic studies ofacyl-chymotrypsin complexes17−20 suggested that a strongerhydrogen-bonding environment in the oxyanion hole providesgreater transition state stabilization, leading to a higherdeacylation rate of the acyl-enzyme. Therefore, our solid-state17O NMR results for DAB-CHT, oMC-CHT, and pMC-CHTare consistent with this view.To further investigate the relationship between 17O NMR

parameters and hydrogen bonding interactions in the oxyanionhole, we performed extensive quantum chemical computations.First, to rule out the possibility that the observed 17O NMRparameters reflect only the electronic effects from different acylmoieties, we computed 17O NMR parameters for three ethyl

ester analogs where the acyl moieties of DAB, oMC, and pMCare maintained. The computational results listed in Table 2show that the three analogs exhibit very similar 17O NMRparameters. For example, the values of δiso and CQ for the threeethyl ester analogs differ by only 4 ppm and 0.2 MHz,respectively. These results strongly suggest that it is necessaryto model the hydrogen bonding interaction in the oxyanionhole for the acyl-enzymes. Before we build a computationalmodel to mimic the hydrogen bonding environment in theoxyanion hole of chymotrypsin, we surveyed the crystallo-graphic data for acyl-enzymes of closely related serine proteasescurrently available in the Research Collaboratory for StructuralBioinformatics Protein Data Bank (RCSB PDB). After a carefulexamination of the structural features, we identified thefollowing two structural parameters as the major variablesthat have the biggest impact on the placement of the carbonyloxygen atom in the oxyanion hole: the Cβ-Oγ-C1 bond angle(θ) and the Cα-Cβ-Oγ-C1 dihedral angle (φ), as shown inFigure 3. The surveyed data set is listed in Table 3. Among thesystems examined, the bond angle θ ranges from 114° to 131°and the dihedral angle φ ranges from 83° to 110°. As a result,the two O···N hydrogen bond distances in the oxyanion holecan vary between 2.45 and 3.46 Å. This is essentially the wholerange of the O···N hydrogen bond. Table 3 also listed thevalues of two additional torsional angles, Θ193 and Θ195, thatdefine the directions of the two hydrogen bonds in theoxyanion hole (vide infra).We chose to use the crystal structure of trans-2,4-

dihydroxycinnamoyl-γ-chymotrypsin (PDB ID: 1K2I)51 toconstruct our acyl-enzyme model because the cinnamoylmoiety in 1K2I is structurally similar to the acyl groups usedin this study. In addition, γ-chymotrypsin, being a complex ofα-chymotrypsin and its autolysis products, shares the identicalprimary sequence with α-chymotrypsin.58 To reduce thecomputational load, only amino acid residues from 192 to196 were used in the quantum chemical calculation to emulatethe environment surrounding the oxyanion hole, as shown inFigure 3. Those residues were chosen to extend one residuebeyond Gly193 and Ser195, whose backbone hydrogen atomsare responsible for the hydrogen bonding in the oxyanion hole.The C-terminus of Gly196 is terminated with N-methylamide,with the amide nitrogen and the methyl carbon occupying thesame position as the backbone N and Cα of Gly197 to mimicneighboring peptide backbones. Likewise, the N-terminus ofMet192 is terminated with an acetyl group, with the carbonatoms occupying the same positions as the backbone carbonylcarbon and Cα of Cys191. The trans-2,4-dihydroxycinnamoyl

Table 3. Crystal Structural Data for Acyl-Enzyme Intermediates of Serine Proteases Reported in the Literature

rO···N, Å (Θ, N···O1C1−Oγ, °)

structure PDB ID X-ray data set resolution (Å) Ser195 Gly193 θ (deg) φ (deg)

A 2GCT49 1.80 3.253 (54) 2.452 (120) 131 87B 1GMC50 2.20 3.046 (46) 2.952 (140) 119 100C 1K2I51 1.80 3.405 (47) 2.990 (114) 120 110Da 1AB952 1.60 3.459 (39) 2.986 (102) 127 101Db 2.887 (49) 2.762 (165) 117 85E 1HAX53 1.60 2.806 (43) 2.729 (174) 124 83F 1GVK54 0.94 2.723 (28) 2.849 (160) 117 89G 2AGE55 1.15 2.855 (36) 3.016 (173) 120 91H 1XVM56 1.10 2.965 (36) 2.836 (148) 121 94I 2AH455 1.13 3.142 (45) 3.454 (104) 118 98J 1GBT57 2.00 3.280 (43) 3.409 (90) 114 102

The Journal of Physical Chemistry B Article

DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

E

acyl group in 1K2I was replaced with an O-acetyl group toreduce the computational cost. Using this model, we examinedhow 17O NMR parameters change as a function of the twoaforementioned structural variables: θ and φ. The bond angle θwas varied systematically from 100° to 130°, and the dihedralangle φ was varied from 80° to 120°, both in 5° increments, toproduce a total of 63 θ−φ combinations. Before calculating the17O NMR parameters for each θ−φ pair, a partial geometryoptimization was performed with the three atoms involved inthe hydrogen bonding in the oxyanion hole (the two amidehydrogen atoms from the protein backbone Ser195 and

Gly193, and the carbonyl oxygen, O1) allowed to move freely,while the remaining atoms were frozen in place. This permitsthe O atom and the hydrogen bonded H atoms to explore localenergy minima without interference from the rest of the model.The calculated 17O CQ and δiso are displayed in Figure 4 ascontours in the θ−φ space.It can be seen immediately from Figure 4a and b that, within

the ranges of θ and φ examined, the 17O NMR parameters ofthe acyl-enzyme model display significant changes: CQ by 1MHz and δiso by 55 ppm. These changes clearly reflect thechanging degree of hydrogen bonding within the oxyanion

Figure 4. Dependences of computed 17O NMR parameters (a: CQ in MHz; b: δiso in ppm) and hydrogen bond distances (c: O···NGly193; d: O···NSer195, Å) on the Cβ-Oγ-C1 bond angle (θ) and the Cα-Cβ-Oγ-C1 dihedral angle (φ). The filled circles (A−J) indicate the positions of the knownacyl-enzymes listed in Table 3. Note that, because the acyl-enzyme shown in Figure 3 does not necessarily produce the same hydrogen bonddistances as the actual crystal structures, filled circles for A−J are not plotted onto (c) and (d).

The Journal of Physical Chemistry B Article

DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

F

hole. In Figure 4a and b, we also indicate where the real crystalstructures of the acyl-enzymes listed in Table 3 would appearon the 2D contour maps. One striking observation is that CQand δiso display different responses to the hydrogen bondingchange. That is, while CQ shows a minimum at (θ = 112°, φ =80°), the minimum δiso value occurs at around (θ = 123°, φ =95°). To better decipher the relationship between the 17ONMR parameters (CQ and δiso) and hydrogen bonding, it is alsoimportant to evaluate how the two H-bond distances in themodel, O···N(Ser195) and O···N(Ser193), would change in thesame θ−φ space. These are shown in Figure 4c and d wheretwo general trends can be seen immediately. (1) Decreasing theCα-Cβ-Oγ-C1 dihedral angle, φ, turns the carbonyl oxygen atomtoward the oxyanion hole, moving closer to both hydrogenbond donors. (2) Decreasing θ from 130° to 100° rotates thecarbonyl oxygen toward N(Ser195), thus shortening the O···N(Ser195) distance, but has little effect on the O···N(Gly193)distance. Now inspection of Figure 4 reveals a remarkablesimilarity between Figure 4a and c. This means that CQ appearsto be determined primarily by the O···N(Gly193) distance. Atthe same time, however, it remains unclear from Figure 4c andd why the landscape of δiso should be different.To further investigate the puzzling results on δiso in Figure

4b, we now consider another important aspect of hydrogenbonding: the hydrogen bond direction. In a recent survey ofenzymes containing an oxyanion hole and small moleculeanalogs, Simon and Goodman59 noted that enzymes oftencontain hydrogen bonds that deviate from an arrangement thatallows for ideal (or maximum) hydrogen bonding interactions.Specifically, they found that, for enzymes, the N···O1C1−Oγdihedral angles, Θ (as defined in Figure 3), are more frequentlyfound to be in a range closer to 90° (where the hydrogen bondis “out-of-the plane”), than being “in-plane” (Θ ≈ 0°), which iscommonly seen in the crystal structures of small molecules.This suboptimal hydrogen bonding arrangement in enzymes istheorized to allow for stabilization of the transition statewithout stabilizing the ground state reactant, which wouldhinder the catalysis provided by the enzyme.57 Now using Θ asa measure of the hydrogen bond direction, we examined theacyl-enzymes with known crystal structures (see Table 3). Firstwe noticed that Θ195 values are narrowly distributed around 40°(meaning the N···O hydrogen bond being in the “out-of-the-plane” mode), but Θ193 values span a much larger range. Theformer feature is clearly due to the fact that the access of theHN(Ser195) to the acyl CO, which is attached to the Oγ ofthe same residue, Ser195, is highly restricted. We furtherdiscovered that the acyl-enzymes close to the δiso minimum(e.g., B, Da, H) have the O···N(Gly193) hydrogen bond largelyin the “out-of-the-plane” mode whereas those close to the CQminimum (e.g., Db, F) are in the “in plane” mode. This analysisthen suggests that, within the narrowly defined acyl-enzymesstudied here, CQ is primarily determined by the O···N(Gly193)hydrogen bond distance and δiso is more sensitive to the O···N(Gly193) hydrogen bond direction. Another interestingobservation from Figure 4 is that the acyl-enzymes withknown crystal structures (labeled A through J) are clusteredaround the δiso minimum. This leads us to hypothesize that,between CQ and δiso, the latter is a more direct measure of thehydrogen bonding environment in the oxyanion hole of serineproteases. As we have established a correlation between δiso anddeacylation kinetics, it is tempting to generalize this idea to linkδiso to the catalytic efficiency of serine proteases. Of course, thetwo aspects of hydrogen bonding, distance and direction, are

often coupled to each other. The complex relationshipsbetween the 17O NMR parameters (CQ and δiso) and hydrogenbonding geometry within the oxyanion hole make quantitativeinterpretation of the experimental 17O NMR data difficult.Nonetheless, the results presented here serve as a promisingstart for further experimental and computational studies.

4. CONCLUSIONS

We have carried out a solid-state 17O NMR study for threeunstable acyl-enzyme intermediates. We found that both the17O isotropic chemical shift and quadrupole coupling constantobtained from solid-state NMR are correlated to thedeacylation rate constant of the acyl-enzymes measured inaqueous solution. This may suggest that the deacylationkinetics of acyl-enzymes is determined mainly by the hydrogenbonding interaction that the carbonyl oxygen atom experiencesin the oxyanion hole. This interpretation is consistent with theconclusions drawn from previous resonance Raman spectro-scopic studies. Our quantum mechanical calculations for anacyl-enzyme model generated further insight into the relation-ship between 17O NMR parameters and the hydrogen bondinginteraction in the oxyanion hole of series proteases. Inparticular, we demonstrated that both hydrogen bond distanceand direction have significant impacts on the 17O chemical shiftand quadrupolar coupling constant. The computationsconfirmed that the variations in 17O NMR parameter observedexperimentally for the three acyl-enzyme intermediates are dueto the different hydrogen bonding environments in theoxyanion hole. The present study demonstrates the utility ofsolid-state 17O NMR in studying a model serine protease. Webelieve that not only can this work be extended to other serineproteases, but solid-state 17O NMR in general can provideunique insight into enzyme kinetics and mechanisms.

■ ASSOCIATED CONTENT

*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.jpcb.6b08798.

Synthesis of 17O-labeled carboxylic acids, kinetic data fordeacylation of acyl-enzymes, experimental and simulatedsolid-state 17O NMR spectra of p-methoxycinnamic acidmethyl ester (full citation of ref 26) (PDF)

■ AUTHOR INFORMATION

Corresponding Author*gang.wu@chem.queensu.ca.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

This work was supported by the Natural Sciences andEngineering Research Council (NSERC) of Canada. Accessto the 900 MHz NMR spectrometer was provided by theNational Ultrahigh Field NMR Facility for Solids (Ottawa,Canada), a national research facility funded by a consortium ofCanadian universities, National Research Council Canada, andBruker BioSpin and managed by the University of Ottawa(http://nmr900.ca). We thank Dr. Jiaxi Wang for assistance inrecording ESI-MS spectra for acyl-enzymes.

The Journal of Physical Chemistry B Article

DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

G

■ REFERENCES(1) Wu, G. Solid-state 17O NMR studies of organic and biologicalmolecules. Prog. Nucl. Magn. Reson. Spectrosc. 2008, 52, 118−169.(2) Wong, A.; Poli, F. Solid-state 17O NMR studies of biomolecules.Annu. Rep. NMR Spectrosc. 2014, 83, 145−220.(3) Gerothanassis, I. P. Oxygen-17 NMR spectroscopy: Basicprinciples and applications. Part I. Prog. Nucl. Magn. Reson. Spectrosc.2010, 56, 95−197.(4) Gerothanassis, I. P. Oxygen-17 NMR spectroscopy: Basicprinciples and applications. Part II. Prog. Nucl. Magn. Reson. Spectrosc.2010, 57, 1−110.(5) Wu, G. Solid-state 17O NMR studies of organic and biologicalmolecules: Recent advances and future directions. Solid State Nucl.Magn. Reson. 2016, 73, 1−14.(6) Zhu, J.; Kwan, I. C. M.; Wu, G. Quadrupole-central-transition

17O NMR spectroscopy of protein-ligand complexes in solution. J. Am.Chem. Soc. 2009, 131, 14206−14207.(7) Zhu, J.; Ye, E.; Terskikh, V.; Wu, G. Solid-state 17O NMRspectroscopy of protein-ligand complexes. Angew. Chem., Int. Ed. 2010,49, 8399−8402.(8) Zhu, J.; Wu, G. Quadrupole central transition 17O NMRspectroscopy of biological macromolecules in aqueous solution. J. Am.Chem. Soc. 2011, 133, 920−932.(9) Young, R. P.; Caulkins, B. G.; Borchardt, D.; Bulloch, D. N.;Larive, C. K.; Dunn, M. F.; Mueller, L. J. Solution-state 17Oquadrupole central-transition NMR spectroscopy in the active site oftryptophan synthase. Angew. Chem., Int. Ed. 2016, 55, 1350−1354.(10) Blow, D. M.; Birktoft, J. J.; Hartley, B. S. Role of a buried acidgroup in the mechanism of action of chymotrypsin. Nature 1969, 221,337−340.(11) Hedstrom, L. Serine protease mechanism and specificity. Chem.Rev. 2002, 102, 4501−4524.(12) Hartley, B. S.; Kilby, B. A. The reaction of p-nitrophenyl esterswith chymotrypsin and insulin. Biochem. J. 1954, 56, 288−297.(13) Kraut, J. Serine proteases: structure and mechanism of catalysis.Annu. Rev. Biochem. 1977, 46, 331−358.(14) Henderson, R. Structure of crystalline α-chymotrypsin: IV. Thestructure of indoleacryloyl-α-chymotrypsin and its relevance to thehydrolytic mechanism of the enzyme. J. Mol. Biol. 1970, 54, 341−354.(15) Corey, D. R.; Craik, C. S. An investigation into the minimumrequirements for peptide hydrolysis by mutation of the catalytic triadof trypsin. J. Am. Chem. Soc. 1992, 114, 1784−1790.(16) Carter, P.; Wells, J. A. Dissecting the catalytic triad of a serineprotease. Nature 1988, 332, 564−568.(17) Tonge, P. J.; Carey, P. R. Direct observation of the titration ofsubstrate carbonyl groups in the active site of alpha-chymotrypsin byresonance Raman spectroscopy. Biochemistry 1989, 28, 6701−6709.(18) Tonge, P. J.; Carey, P. R. Length of the acyl carbonyl bond inacyl-serine proteases correlates with reactivity. Biochemistry 1990, 29,10723−10727.(19) Tonge, P. J.; Carey, P. R. Forces, bond lengths, and reactivity:fundamental insight into the mechanism of enzyme catalysis.Biochemistry 1992, 31, 9122−9125.(20) Tonge, P. J.; Pusztai, M.; White, A. J.; Wharton, C. W.; Carey, P.R. Resonance Raman and Fourier transform infrared spectroscopicstudies of the acyl carbonyl group in [3-(5-methyl-2-thienyl)acryloyl]-chymotrypsin: evidence for artifacts in the spectra obtained by bothtechniques. Biochemistry 1991, 30, 4790−4795.(21) Kong, X.; Tang, A.; Wang, R.; Ye, E.; Terskikh, V.; Wu, G. Arethe amide bonds in N-acyl imidazoles twisted? A combined solid-state17O NMR, crystallographic, and computational study. Can. J. Chem.2015, 93, 451−458.(22) Wirnt, R. Chymotrypsin. In Methods of Enzymatic Analysis, 2nd

ed.; Bergmeyer, H.-U., Ed.; Academic Press: New York, 1974; pp1009−1012.(23) Bender, M. L.; Schonbaum, G. R.; Zerner, B. Spectrophoto-metric investigations of the mechanism of α-chymotrypsin-catalyzed

hydrolyses. Detection of the acyl-enzyme intermediate. J. Am. Chem.Soc. 1962, 84, 2540−2550.(24) Schonbaum, G. R.; Zerner, B.; Bender, M. L. Thespectrophotometric determination of the operational normality of anα-chymotrypsin solution. J. Biol. Chem. 1961, 236, 2930−2935.(25) Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calve, S.;Alonso, B.; Durand, J.-O.; Bujoli, B.; Gan, Z.; Hoatson, G. Modellingone- and two-dimensional solid-state NMR spectra. Magn. Reson.Chem. 2002, 40, 70−76.(26) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,B.; Petersson, G. A., et al. Gaussian 09, Revision B.01; Gaussian, Inc.:Wallingford, CT, 2009.(27) Dong, S.; Ida, R.; Wu, G. A combined experimental andtheoretical 17O NMR study of crystalline urea: An example of largehydrogen-bonding effects. J. Phys. Chem. A 2000, 104, 11194−11202.(28) Wu, G.; Hook, A.; Dong, S.; Yamada, K. A Solid-state NMR andtheoretical study of the 17O electric field gradient and chemicalshielding tensors of the oxonium ion in p-toluenesulfonic acidmonohydrate. J. Phys. Chem. A 2000, 104, 4102−4107.(29) Ashton, D. S.; Beddell, C. R.; Cooper, D. J.; Green, B. N.;Oliver, R. W. A.; Welham, K. J. Some electrospray mass spectrometricevidence for the existence of covalent O-acyl enzyme intermediates.FEBS Lett. 1991, 292, 201−204.(30) Ashton, D. S.; Beddell, C. R.; Cooper, D. J.; Green, B. N.;Oliver, R. W. A.; Welham, K. J. On the purity of 3X-recrystallisedbovine α-chymotrypsin. Biochem. Biophys. Res. Commun. 1993, 192,75−81.(31) Bender, M. L.; Killheffer, J. V.; Cohen, S. Chymotrypsin. Crit.Rev. Biochem. Mol. Biol. 1973, 1, 149−199.(32) Krigbaum, W. R.; Godwin, R. W. Molecular conformation ofchymotrypsinogen and chymotrypsin by low-angle x-ray diffraction.Biochemistry 1968, 7, 3126−31.(33) Whiting, A. K.; Peticolas, W. L. Details of the acyl-enzymeintermediate and the oxyanion hole in serine protease catalysis.Biochemistry 1994, 33, 552−561.(34) Bernhard, S. A.; Hershberger, E.; Keizer, J. The influence of pHon the rate of hydrolysis of acylchymotrypsins. Biochemistry 1966, 5,4120−4126.(35) Hagaman, E. W.; Chen, B.; Jiao, J.; Parsons, W. Solid-state 17ONMR study of benzoic acid adsorption on metal oxide surfaces. SolidState Nucl. Magn. Reson. 2012, 41, 60−67.(36) Kong, X.; Shan, M.; Terskikh, V.; Hung, I.; Gan, Z.; Wu, G.Solid-state 17O NMR of pharmaceutical compounds: Salicylic acid andaspirin. J. Phys. Chem. B 2013, 117, 9643−9654.(37) Butler, L. G.; Brown, T. L. Nuclear quadrupole couplingconstants and hydrogen bonding. Molecular orbital study of oxygen-17and deuterium field gradients in formaldehyde-water hydrogenbonding. J. Am. Chem. Soc. 1981, 103, 6541−6549.(38) Butler, L. G.; Cheng, C. P.; Brown, T. L. Oxygen-17 nuclearquadrupole double resonance. 6. Effects of hydrogen bonding. J. Phys.Chem. 1981, 85, 2738−2740.(39) Gready, J. E. The relationship between nuclear quadrupolecoupling constants and the asymmetry parameter. The interplay oftheory and experiment. J. Am. Chem. Soc. 1981, 103, 3682−3691.(40) Gready, J. E. Theoretical study of the variability of the electricfield gradient tensor of oxygen nuclei in organic molecules. J. Phys.Chem. 1984, 88, 3497−3503.(41) Kuroki, S.; Takahashi, A.; Ando, I.; Shoji, A.; Ozaki, T.Hydrogen-bonding structural study of solid peptides and polypeptidescontaining a glycine residue by 17O NMR spectroscopy. J. Mol. Struct.1994, 323, 197−208.(42) Kuroki, S.; Ando, S.; Ando, I. An MO study of nuclearquadrupolar coupling constant and nuclear shielding of the carbonyloxygen in solid peptides with hydrogen bonds. Chem. Phys. 1995, 195,107−116.(43) Wu, G.; Yamada, K.; Dong, S.; Grondey, H. Intermolecularhydrogen-bonding effects on the amide oxygen electric-field-gradient

The Journal of Physical Chemistry B Article

DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

H

and chemical shielding tensors of benzamide. J. Am. Chem. Soc. 2000,122, 4215−4216.(44) Yamada, K.; Dong, S.; Wu, G. Solid-state 17O NMRinvestigation of the carbonyl oxygen electric-field-gradient tensorand chemical shielding tensor in amides. J. Am. Chem. Soc. 2000, 122,11602−11609.(45) Wu, G.; Dong, S.; Ida, R.; Reen, N. A solid-state 17O nuclearmagnetic resonance study of nucleic acid bases. J. Am. Chem. Soc.2002, 124, 1768−1777.(46) Wu, G.; Yamada, K. Determination of the 17O NMR tensors inpotassium hydrogen dibenzoate: A salt containing a short O···H···Ohydrogen bond. Solid State Nucl. Magn. Reson. 2003, 24, 196−208.(47) Wong, A.; Pike, K. J.; Jenkins, R.; Clarkson, G. J.; Anupold, T.;Howes, A. P.; Crout, D. H. G.; Samoson, A.; Dupree, R.; Smith, M. E.Experimental and theoretical 17O NMR study of the hydrogen-bonding on CO and O-H oxygens in carboxylic solids. J. Phys.Chem. A 2006, 110, 1824−1835.(48) Kwan, I. C. M.; Mo, X.; Wu, G. Probing hydrogen bonding andion-carbonyl interactions by solid-state 17O NMR spectroscopy: G-ribbon and G-quartet. J. Am. Chem. Soc. 2007, 129, 2398−2407.(49) Dixon, M. M.; Brennan, R. G.; Matthews, B. W. Structure of γ-chymotrypsin in the range pH 2.0 to pH 10.5 suggests that γ-chymotrypsin is a covalent acyl enzyme adduct at low pH. Int. J. Biol.Macromol. 1991, 13, 89−96.(50) Yennawar, N. H.; Yennawar, H. P.; Farber, G. K. X-ray crystalstructure of gamma-chymotrypsin in hexane. Biochemistry 1994, 33,7326−7336.(51) Ghani, U.; Ng, K. K.; Atta-ur, R.; Choudhary, M. I.; Ullah, N.;James, M. N. Crystal structure of γ-chymotrypsin in complex with 7-hydroxycoumarin. J. Mol. Biol. 2001, 314, 519−525.(52) Kashima, A.; Inoue, Y.; Sugio, S.; Maeda, I.; Nose, T.;Shimohigashi, Y. X-ray crystal structure of a dipeptide-chymotrypsincomplex in an inhibitory interaction. Eur. J. Biochem. 1998, 255, 12−23.(53) Wilmouth, R. C.; Edman, K.; Neutze, R.; Wright, P. A.; Clifton,I. J.; Schneider, T. R.; Schofield, C. J.; Hajdu, J. X-ray snapshots ofserine protease catalysis reveal a tetrahedral intermediate. Nat. Struct.Biol. 2001, 8, 689−94.(54) Katona, G.; Wilmouth, R. C.; Wright, P. A.; Berglund, G. I.;Hajdu, J.; Neutze, R.; Schofield, C. J. X-ray structure of a serineprotease acyl-enzyme complex at 0.95-A resolution. J. Biol. Chem.2002, 277, 21962−21970.(55) Radisky, E. S.; Lee, J. M.; Lu, C.-J. K.; Koshland, D. E. Insightsinto the serine protease mechanism from atomic resolution structuresof trypsin reaction intermediates. Proc. Natl. Acad. Sci. U. S. A. 2006,103, 6835−6840.(56) Schmidt, A.; Lamzin, V. S. Extraction of functional motion intrypsin crystal structures. Acta Crystallogr., Sect. D: Biol. Crystallogr.2005, 61, 1132−1139.(57) Mangel, W. F.; Singer, P. T.; Cyr, D. M.; Umland, T. C.;Toledo, D. L.; Stroud, R. M.; Pflugrath, J. W.; Sweet, R. M. Structureof an acyl-enzyme intermediate during catalysis: (guanidinobenzoyl)-trypsin. Biochemistry 1990, 29, 8351−8357.(58) Harel, M.; Su, C. T.; Frolow, F.; Silman, I.; Sussman, J. L. γ-Chymotrypsin is a complex of α-chymotrypsin with its own autolysisproducts. Biochemistry 1991, 30, 5217−5225.(59) Simon, L.; Goodman, J. M. Enzyme catalysis by hydrogenbonds: The balance between transition state binding and substratebinding in oxyanion holes. J. Org. Chem. 2010, 75, 1831−1840.

The Journal of Physical Chemistry B Article

DOI: 10.1021/acs.jpcb.6b08798J. Phys. Chem. B XXXX, XXX, XXX−XXX

I