chymotrypsin mechnism

7/28/2019 chymotrypsin mechnism

1/8

Vol. 9, 1976 Structure and Mechanism of Chymotrypsin 14 5Structure and Mechanismof Chymotrypsin

David M . BlowMRC Laboratory of Molecular Biology, Hil ls Road, Cambridge, United Ki ngdom

Received December 27,1974

Our understanding of enzyme catalysis has ad-vanced steadily since x-ray diffraction results gavethe first detailed information about the architectureof an enzyme molecule 10 years ago. The mechanismof chymotrypsin is probably understood in more de-tail than any other enzyme at the present time. Asenzymes go, it is a very simple one, and it has provedaccessible to study by awide range of different tech-niques. Although it is 7 years since the first molecularmodel of chymotrypsin was built,l it is only recentlythat I have begun to believe that the major effects arequalitatively understood. A quantitatiye assessment,which will require a quantum mechanical descriptionof the active site and a detailed thermodynamic anal-ysis of each step of the reaction, is still probably sev-eral years away.Chymotrypsin catalyzes the hydrolysis of peptidebonds of protein foods in the mammalian gut. I t is se-creted in the pancreas as chymotrypsinogen, a single-chain protein of 245 amino acids,2 and is activated tochymotrypsin by the hydrolysis of a single peptidebond, catalyzed by tryp~i n.~Amino acid sequence analysis has shown that en-zymes which have a close homology to chymotrypsinare widespread. In addition to chymotrypsinB, tryp-sin, and elastase, all secreted in the pancreas, thereare several similar enzymes in the blood-clotting sys-tem, including thrombin and blood-clotting factor X ,and in the complement-fixing ~ystem.~-~trypsin-like enzyme occurs in the Pacific dogfish and the seaanemone, and the enzyme cocoonase used by thepupa of the silkworm to digest its cocoon is closely re-lated.7 Proteolytic enzymes in bacteria mostly showno sign of genetic relationship to chymotrypsin. Sub-tilisin, found inB.subtilis andB.amyloliquefaciens,has a completely unrelated structure,8 though it usesan almost identical catalytic me~hanism.~hymo-trypsin-like enzymes have, however, been found insome species of soil bacteria, including Myxobacter495. In these enzymes, large tracts of the amino acidsequence, including the catalytically important resi-dues, show extensive homology with chymotrypsin,and there can be little doubt that the basic moleculararchitecture is similar.1 Possibly these enzymes havedeveloped through the protozoa from a primitive an-cestral gene, but it seems more likely that some kindof genetic transfer has occurred from higher organ-isms.Chymotrypsin is a monomeric enzyme which ex-hibits no allosteric effects. The conformational

David M. Blow studied natural sciences at Corpus Christi College, Cam-bridge, and studied for his Ph.D. under M. F. Perutz in the Medical ResearchCoun cils Unit for the study of molecular biologic al systems in the CavendishLaboratory, Cambridge. After 2 years in the United States, he returned toCambridg e and has been on the staff of the Laboratory of Molec ular Biologysince it was established in 1962.

SchemeIk, k,k-l Michaelis

complexEXH2OH +RCOX ECH2OH:RCOX *

kECHzOCORA CHzOH +RCOOHacyl enzyme HZ0

+H x

changes which occur in the process of normal proteol-ysis seem to be very small. One of the substrates inproteolysis is water, which has to be present in all ex-periments on the enzyme. These features have great-ly simplified study of the catalytic mechanism.The Acyl-enzyme Mechanism

Chymotrypsin has a uniquely reactive seTine, Ser-195, which may be acylated by a variety of agents.When p-nitrophenyl acetate is used, a rapid burstof 1mol of p-nitrophenollmol of enzyme is produced,followed by aslower steady release of nitrophenol asthe acetyl-chymotrypsin is hydrolyzed.ll Both stepsare slowed down at low pH, and at pH 4 acetyl-chy-motrypsin is sufficiently stable to be isolated.12 Sincethe pKa of the reaction is about 6.7, it was early sug-gested that histidine was involvedinthe reaction.13Many substrates give similar kinetic behavior, andall esters of the same acyl group show the samesteady-state rate of hydrolysis. This suggested thatthe same acyl-enzyme intermediate was formed inevery case, and that hydrolysis of the common acyl-enzymes was the rate-limiting step. Amides, on theother hand, are hydrolyzed at slower rates, which de-pend on the amide group inv01ved.l~This is consis-tent with Scheme I, in which CHzOH represents the(1) B. W. M atthews, P. B. Sigler, R. Henderson, and D. M . Blow, Nature(2) B. S.Hartley, Nature (L ondon),201,1284-1287 (1964).(3) P.Desnuelle, Enzymes,1stEd., 4, Chapter 5(1960).(4) B. S.Hartley, Symp. SOC. en. Microbiol., 24,151-182 (1974).(5) E. W. Davie and E. P.K irby, Curr. Top. Cell.Regul., 7,51-86 (1973).(6) T. Barkas, G. K . Scott, and J . E. Fothergill, Biochem. SOC. rans., 1,1219-1220 (1973).(7) H. Neurath, R. A. Bradshaw, and R. A rnon in Structure-FunctionRelationships of Proteolytic Enzymes, P. Desnuelle, H. Neurath, and M .Ottesen, M unksgaard, Copenhagen, 1970,pp 113-134.(8) F. S. M arkland and E. L . Smith, J . Biol. Chem., 242, 5198-5211(1967).(9) J . D. Robertus, R. A. Alden, J . J . Birktoft, J . K raut, J . C. Powers, andP.E.Wilcox, Biochemistry, 11,2439-2449 (1972).(10) M. 0. . Olson, N. Nagabhushan, M . Dzwinich, L . B. Smillie, and D.R. Whitaker, Nature (London), 228, 438-442 (1970);L . T. J . Delbaere, W.L. B. Hutcheon, M. N. G. J ames, and W. E. Thiessen, Nature (Londonj,257,758-763 (1975).

(London),214,652-656 (1967).

(11) B. S.Hart,ley and B. A . K ilby, Biochem. J . ,50,672-678 (1952).(12) A. K . Balls and J .H. Wood, J . Biol.Chem.,219,245-256 (1956).(13) L. Cunningham,Compr. Biochem., 16,85-188 (1965).(14) M. L . Bender and F. J . K ezdy, J . Am. Chem. Soc., 86, 3704-3714(1964).


2/8

146 Blow Accounts of Chemical ResearchScheme I 1ECH,OH + RCOX *

LECH,O x \tetrahedral

intermediate

H R ,o-/OH\CH,O

ECH,OH t RCOOHproduct 2

EC H OH RCOXMichaelis complex

HX product 1tECH,O C0.R

acyl enzymeECH O.CO.R

+HO H

ECH,OH:RCOOHproduct complex

Ser-195 side-chain of the enzymeE, and RCOX is anamide or ester substrate.Herekllk-1 gives the binding constant K s for for-mation of the Michaelis complex; k2, the acylationrate constant, is rate limiting for amide hydrolysis;and k3, the deacylation rate constant, is rate limitingfor ester hydrolysis.Even quite recently, experimental data have beenpublished which seemed to contradict the acyl-en-zyme me~hanism,l ~-~ut further analysis has showna particular disturbing effect in each case18-20 andthe acyl-enzyme mechanism may be taken as estab-lished.The Tetrahedral Intermediate

Formation and hydrolysis of an ester bond, as indi-cated by steps 2 and 3 of Scheme I , are supposed toproceed by a mechanism in which the trigonally coor-dinated C is attacked by the approach of a fourth li-gand, approximately perpendicular to the plane ofc~ordi nation.~l -~~his approach provides a fourth li-gand to C, which moves slightly out of the plane ofthe three original ligands to form a tetrahedrallycoordinated form. The tetrahedral form may collapseto a more stable state, either by repelling the attack-ing ligand and returning to its original state or by re-pelling another ligand and forming a new three-coor-dinated state, by moving into the plane which thenew ligands define (Scheme11;Figure 1).Due to the relative instability of the tetrahedralform, it is not present in sufficient abundance to havea detectable effect on the kinetics of a normal hydro-

(15) R. M . Epand, Biochem. Biophys.Res. Commun., 37,313-318 (1969).(16) S.E. Bresler, G. P. Y easov, and V. M. K rutyakov, Mol. Biol. (Mos-(17) G. P. Hess, J . McConn, E. K u, and G. M cConkey, Philos. Tmns.R.(18) I. V. Berezin, N. F. K azanskaya, A. A. K lyasov, and V. K . Svedas,(19) A . R. Fersht and Y . Requena, .Mol. Biol., 60,279-290 (1971).(20) J .Fastrez and A. R. Fersht, Biochemistry, 12,2025-2034 (1973).(21) L. P. Hammett, Physical Organic Chemistry, M cCraw-Hil l, New(22) M. L. Bender, J .Am. Chem Soc., 73,1626-1629 (1951).(23) H. B. Burgi, J . D. Dunitz, and E. Shefter, J . Am. Chem. Soc., 95,5065-5067 (1973).The initial approach makes an angle of 109 to the C-0direction.

cow ) , 3,15-28 (1969).Soc. London,Ser. B, 257,89-104 (1970).Eur. J .Biochem., 38,529-536 (1973).

Y ork, N.Y.,1940,pp 355-357.

/Figure 1. T he carbon atom C moves between one planar trigonalstate and another by passing through the center of the tetrahe-dron.

0-6, 91 195np 2.SOHL

K, 6.7

195

HO

Figure 2. T he charge relay system, indicating the transition be-tween the state characterizing the active enzyme and the ( inactive)form which exists below pH 6.7, according to Hunkapil ler et aL31lytic reaction. Until recently, only indirect evidence22for the existence of such a form had been obtained,but two crystallographic studies of complexes oftrypsin with protein trypsin inhibitors have shownthat these are stabilized in the tetrahedral form.24i25The Charge-Relay System

The first interpretable electron-density map ofchymotrypsin1 showed, with the aid of the aminoacid sequence, that the side chain of His-57 was lyingclose to Ser-195. This gave satisfying agreement withthe supposed involvement of histidine from the pH-activity curve. It was only when a more accurate elec-tron density map was obtained, after further crystal-lographic work, and a correction to the amino acid se-quence, that a far more interesting situation was re-vealed26 (Figure2) .Ser-195 lies in a shallow depression on the enzymesurface, with its 0 7 at hydrogen-bond distance fromone of the ring nitrogens (NE2) f His-57. The imidaz-ole ring of His-57 has its edge at the surface of themolecule, so that NC2s accessible from the solvent;the other nitrogen (N61) is buried. Immediately be-hind N61, and completely buried from contact withthe solvent, is the side chain of Asp-102, with its car-boxyl group at hydrogen-bonding distance fromN6 .Two water molecules are buried closeto the carboxylgroup, but not hydrogen bonded to it;27otherwise theenvironment of the carboxyl group is completely hy-drophobic.The unique juxtaposition of a buried acid groupwith the unusually polarizable system of the imidaz-ole ring, which in its uncharged form can carry a pro-ton on either of the two ring nitrogens, seemed atfirst to be the key to the activity of the trypsin familyof enzymes. This view was strengthened when itturned up again and again, not only in the crystal

(24) R. H uber, D. K ukla,U . ode, P.Schwager, K . Bartels, tJ . Deisenhof-er. and W teigemann, J Mol. Biol., 89, 73-101 (1974). Refined, unpub-lished coordinates show a tetrahedral conformation around the carbonylcarbon, but the distance to the serine01 is 2.6 A.jo(25) R. M. Sweet, H. T. Wright, J . J anin, C . H. Chothia, and D. M. Blow,Biochemistry, 13,4212-4228 (1974).(26) D. M. Blow, J. J. Birktoft, and B. S.Hartley, Nuture (London), 221,337-340 (19691.(27) J . J . Birktoft and D. M. Blow, J .Mol . Bio!., 68,187-240 (1972).


3/8

Vol. 9, 1976 Structure and Mechanismof Chymotr.ypsin 147structures of elastasez8 and of trypsin,29 but also inthe structure of s~bti l i si n.~~We proposedz6 that the function of the buried acidgroup was to polarize the imidazole ring, since theburied negative charge would induce a positivecharge adjacent to it. The exposed nitrogen, Nc2,would carry an excess negative charge, strengtheningthe hydrogen bond between Ser-195 and His-57. Thisgave a possibility of proton transfer along hydrogenbonds, which could possibly allow the hydroxyl pro-ton to be transferred to His-57 from Ser-195. A de-protonated Or would be highly reactive to attack thescissile peptide bond. There were, however, obviousdifficulties about a scheme which involved protontransfer. The pKa of -CH20H is about 15,sothat theamount of proton transfer from serine to histidine atpH 7 is likely to be very small. Also, the system clear-ly has a pKa at about 6.5 to 7, which is absolutelynormal for a histidine, while this particular histidineis in an extremely abnormal environment. In our1969 paper,26 therefore, the issue of proton transferwas carefully left open.

These problems seem to have been resolved by theexperiments of Hunkapiller et al.31 on the a-lyticprotease of Myxobacter495. Although little is knownof the structure of this protein at present, it is as-sumed that the homologies of active-site sequencesimply that there is a similar Asp-His-Ser system inthe active site. The important feature of the enzymeis that it contains only one histidine residue, so thatthe 13C magnetic resonance of the one Ca2histidineatom can be clearly identified. The pH-dependentchanges of chemical shift and coupling constant ofthis resonance were carefully analyzed, and they indi-cate that the pKa of the histidine residueisabout 4.2.The change at pH 6.5 to 7 is ascribed to protonationof the carboxylate of Asp-102 by these workers (al-though others have criticized this interpretati~n~~).This means that the buried environment of Asp-102 perturbs its pKa strongly, making it much moredifficult to ionize than a normal carboxyl group. His-57 is strongly perturbed by the adjacent protonatedcarboxyl group, and also becomes much more diffi-cult to ionize than normal. In terms of the protontransfer system, the new result is that below pH 7 theproton between His-57 and Asp-102 is localized onAsp-102 (Figure 2). Above pH 7 it is transferred toN*l, but the histidine remains electrically neutral bylosing the proton on Ns2.The transition at pH 7 thusreverses the polarization of the imidazole ring. Ndlshows the anomaly that it is deprotonated below pH6.7, but protonated above pH 7.In another series of NMR measurements, the pro-ton resonance due to the proton between Asp-102and His-57 of chymotrypsin has been observed. Thisshows a transition near pH 7, corresponding to thetransfer of the proton.33

(28) D. M . Shotton and H. C. Watson, Nature (London), 225, 811-816(29) R. M. Stroud, L. M. K ay, and R. E. Dickerson, J . Mol. B ol. , 83,(30) C. S. Wright, R. A . A lden, and J . K raut, Nature (London), 221,(31) M. Hunkapiller, S.H. Smallcombe, D. R. Whitaker, and J . H. Rich-(32) G. Robil lard and R. G. Shulman, J .Mol. Biol., 86,519-540 (1974).(33) G. Robillard and R. G. Shulman, J . Mol. B ol. ,71,507-511 (1972).

(1970).185-208 (1974).235-242 (1969).ards, J . Biol. Chem., 248,8306-8308 (1973).

In the active form of the enzyme, the polarizationof the imidazole ring above pH 7 creates an environ-ment which can readily accommodate the proton ofSer-195 when this attacks the scissile bond. The re-peated occurrence of the charge-relay system in neu-tral proteases leaves little room for doubt that thisconstitutes the proton-transfer system required toimplement Scheme 11. But the discovery that thecharge-relay system exists in identical form in the"inactive" precursor chym~trypsi nogen~~?~~howedthat this was not enough to make an effective en-zyme. Trypsinogen and chymotrypsinogen do exhibitenzymatic activity, but at such a low levelthat of trypsin) that it can only be detected underspecial condition^^^^^^The Substrate Binding Site

Close to the active serine of chymotrypsin is adeeply invaginated pocket, large enough to accommo-date an indole or a toluene molecule. When an aro-matic group occupies this pocket, it is sandwichedbetween peptide bonds 190-191 and 191-192 on oneside and 215-216 on the other. A hydrophobic groupso accommodated can make numerous hydrophobiccontacts, especially involving Ser-190, Cys-191-220,Val-213, Trp-215, and T~r-2283~Figure 3). Thispocket does not exist in chymotrypsinogen, but isformed by the rearrangement of residues 191-194which occurs on a~ti vati on.~~The narrowness of the pocket defines the plane ofthe aromatic side chain of a Tyr, Phe, or T rp residue,including the CP atom. The shape of the mouth of thepocket leaves little freedom for the C" position. Thewhole conformation of the amino acid side chain inthe primary specificity pocket is thus rather closelydefined. The shape of the pocket corresponds exactlywith the range of amino acids which define the pri-mary specificity of the enzyme.The substrate binding pocket does not restrict theconformational angleXI which defines rotation aboutthe C"-CP bond and which thus controls the positionof adjacent atoms of the polypeptide chain (Figure4). Possible values for X I are partly restricted by theshape of the molecular surface outside the substratebinding pocket. A conveniently placed hydrogen-bond acceptor, the CO group of Ser-214, close to themouth of the pocket, further controls XI. When theNH group of an amino acid in the primary bindingsite is directed toward this acceptor, its carbonyl car-bon is brought close to the side chain of Ser-195.

If the above interactions are made, the orientationof the scissile bond depends on the conformationalangle$, which defines rotation about the Ca-C bond(Figure4). Once again, the adjacent surface of the en-zyme restricts the possibilities, and permits tworanges for this angle, about 180' apart. One of thesebrings the carbonyl oxygen of the scissile bond closeto two hydrogen-bond donors which are directedaway from the enzyme surface. These are the peptideNH groups of Gly-193 and Ser-195.37f38The alterna-(34) S.T. Freer, J .K raut, J . D. Robertus, H. ?.Wright, and N. H. Xuong,(35) J .K ay and B. K assell , J . Biol. C hem.,246,6661-6665 (1971).(36) P. H. M organ, N. C. Robinson, K. A. Walsh, and H. Neurath, Proc.(37) T. A. Steitz, R. Henderson, and D. M. Blow, J .Mol. B i d , 46, 337-

Biochemistry,9,1997-2009 (1970).

Natl. Acad. Sei. U.S.A.,69,3312-3316 (1972).348(1969).


4/8

148 Blow Accounts of Chemical ResearchI I

0Qe oOS

Figure 3. Complex of chymotrypsin with formyl-L -tryptophan, asscissile bond would be adjacent to the carboxylate oxygen S.scissilebond

Figure4. T he conformational angles X I , @, and $which determinethe orientation of the scissile bond when an aromatic side chain islocated in the primary substrate binding pocket.tive range of $, which brings the N of the scissilebond into this position, is unfavorable for hydrogenbonding.) If the carbonyl oxygen accepts these hy-drogen bonds, the NH of the scissile bond is broughtclose to His-57 and Ser-195. The scissile peptidebond lies in a plane almost perpendicular to theplane of the charge relay system (defined by the car-boxyl oxygen, the center of the imidazole ring, and0 7 (195)).

These interactions are sufficient to define a confor-(38) R. Henderson, J Mol B d , 4,341-354 (1970).

~ ~determined crystallographically by Steitz et al.37 I n a substrate, the

mation for a small molecular substrate, for exampleN-acetyl-L-tryptophan amide (Figure4).For trypsinsubstrates, an exactly similar mode of binding exists,in which a lysine or arginine side chain is stabilizedby an acid group (Asp-189) which is at the bottom ofthe substrate binding pocket in tr y p~i n . ~~2~~muchsmaller pocket exists in elastase, due to the presenceof two bulky side chains (Val-215 and Thr-226)which replace glycine in trypsin and chymotrypsin.28These will fill up most of the pocket, which can ac-cept the methyl side chain of an alanine residue, butlittle more. This small binding pocket does not com-pletely define the direction of the C'(-CP bond, nordoes it give enough binding energy by comparisonwith other modes of binding. Secondary substrate in-teractions are far more important for elastase, whichfunctions most efficiently with favorable polypeptidesub~trates. ~~-~lAn extremely similar binding site exists for subtil-isin, except that the primary binding pocket is lesswell defined, being more clearly described as a shal-low gro~ve.~The highly developed primary binding site made itrelatively easy to study the binding of substrates to(39) D. Atlas, S.Levit, I. Schechter, and A . Berger, F E E S Lett., 11,281-(40) A. Gertler and T.Hofmann, Can. . Biochem., 48,384-386 (1970).(41) R. C. Thompson and E. E. Blout, Proc. Natl. Acad. Sei . U.S.A., 47,

283 (1970).1341-1355 (1970).


5/8

Vol. 9, 1976 Structure and Mechanismof Chymotrypsin 149the enzyme on one side of the scissile bond. Thebinding of the other part of the substrate,X, has tobe studied by less direct means. Scheme I shows thecourse of a reaction in which the solvent containsonly water. If the solution is enriched in XH, theacyl-enzyme can break down in one of two ways:

ECH,OH + RCOOHx3ECH,OCOR ECH,OH + RCOX

and the relative efficiencies of various XH as de-acylating agents can be compared.42 Glycinamide andalaninamide were shown to be 10-100 times more po-tent deacylating agents than water, while hydrazine,a very reactive nucleophile in nonenzymatic reac-tions, was only as effective as water. This indicated apreferred mode of binding of a peptide group X tothe enzyme, believed to involve hydrophobic contactswith the cystine bridge42-58. There was no evidence,however, of strong binding of alaninamide to theacyl-enzyme. The binding energy was thus responsi-ble for accelerating the reaction, rather than assistingwith binding the product XH.Two crystallographic studies of complexes of tryp-sin .with protein trypsin i nhi bi t~rs~~p~~onfirmedthese ideas in a striking way. The inhibitors bindvery much like a peptide substrate, but the deacyla-tion step is prevented (or diminished) by the exclu-sion of water from the active site and by the high sta-bility of the complex. Both the studied inhibitors, al-though completely different molecules, make the an-ticipated interactions on both sides of the scissilebond. But in order to make all these interactions si-multaneously, the atomic arrangement at the activesite is forced into a very crowded situation. Bothstudies showed that this crowding results in stabiliza-tion of the tetrahedral intermediate (Scheme 11), nwhich four atoms are all within covalent bond dis-tance of the carbonyl carbon.24 n this conformation,the carbonyl oxygen is optimally aligned toward thetwo hydrogen-bond donors NH-193 and NH-195.The partial negative charge which it probably carrieswill strengthen these bonds and help in delocalizingthe buried negative charge.I t has been found that many inhibitors bind aboutequally strongly to anhydrotrypsin, a form of trypsinin which Ser-195 is converted to dehydr~al ani ne.~~>~~In this case no chemical bond can be made betweenenzyme and inhibitor; but the removal of the serineOr relieves the overcrowding at the active site, and inother respects the close fit between the surfaces ofenzyme and inhibitor is not di ~turbed.~~rystallo-graphic analysis confirms this.45 The chemicalchanges at the active site, stabilizing the tetrahedralintermediate, involve little net energy change, sincethe affinities of inhibitors for trypsin and anhydro-

(42) A. R. Fersht, D. M . Blow, and J . Fastrez, Biochemistry, 12, 2035-(43) H. Ako, R. J . Foster, and C. A. Ryan, Biochemistry, 13, 132-139(44) J .-P. V incent, M . Peron-Renner, J . Pudles, and M . L azdunski, Bio-(45) R. Huber, W. Bode, D. K ukla, U. K ohl, and C. A. Ryan, Riophys.

2041 (1973).(1974).chemistry, 13,4205-4211 (1974).Struct.Mech., 1,l-13 (1975).

trypsin are ~i mi l ar . ~~? ~~he energy gained from thetwo polarized hydrogen bonds from NH-193 andNH-195, and from the delocalization of the buriednegative charge, evidently balances the energy of for-mation of the strained bond between enzyme and in-hibitor (see below).25Transition-State T h e ~r y ~~- ~~

In each step of a chemical reaction, reactants passfrom one relatively stable state to another through astate of higher energy. The state of highest energythrough which the reactants must pass, to get fromone stable state to another, is the transition state.The additional free energy needed to reach the tran-sition state is the activation energy of the reaction.Part of the activation energy is enthalpic: for ex-ample, the work which needs to be done to bring twoatoms close enough that a covalent bond begins to beformed between them. Another part is entropic, andexpresses the additional order which needs to be im-posed on the system to reach the transition state: forexample, the chance that two atoms would approachsufficiently close to react. Obviously the entropiccomponent for a unimolecular reaction of two atomswhich can collide by bond rotation is less than it isfor a bimolecular reaction of two atoms floating free-ly in a box.The rate of a chemical reaction depends on the ac-tivation energy, according to the Arrhenius equation.Although an absolute rate can be calculated in someidealized cases, these are rare. We do not knowenough about the modes of vibration of the reactants,nor about the range of possible configurations whichwill still allow the reaction to occur. In enzyme reac-tions the correct orientation of the substrate cangreatly reduce the entropy of a~ti vati on.~~hecharge distribution at the active site and the geome-try of the binding site reduce the enthalpy of thetransition state.The appearance of a tetrahedral adduct in the twocrystalline trypsin-inhibitor compl e~es~~2~~ivesclear evidence about the catalytic mechanism of theenzyme. The tetrahedral form, which in normal hy-drolytic reactions exists too briefly to be observed,has been stabilized to exist permanently in crystals.Hubers results24 show that the stabilized form hasan abnormal bond length-a form which might be as-sumed to be even less stable. The observed conforma-tion gives us some idea of how the reduction in theactivation energy has been achieved, as detailed atthe end of the last section.The energy profile as the tetrahedral complex de-velops must be rather flat. Probably the Or-C bondcan only shorten to a normal value of 1.5A in the tet-rahedral intermediate by straining adjacent bond an-gles. In the trypsin-inhibitor complexes, where nu-merous secondary interactions are optimal, the com-plex is actually stabilized by 17 kcal with its abnor-mally long Or-C bond. Small substrates of chymo-trypsin, which make a few favorable secondary inter-actions, need 10-15 kcal of activation energy for acyl-

(46) M. I. Page and W. P. J encks, Proc. Natl. Acad. Sci. U.S.A., 68,1678-1683 m7i).~ ~~.~~., .(47) R. Wolfenden, Acc. C hem Res., 5,lO-18 (1972).(48) G. E. L ienhard,Science, 180,149-154 (1973).(49) A. R. Fersht, Proc. Roy. SOC. ondon, Ser.R, 187,397-407 (1974).


6/8

150 Blow Accounts of Chemical Research

QAla 16'

L ys 15'i'tb

> rg65'

Asp 194

(e)

Arg65'6Figure5 . T he relative positions of key amino acids at the active site of (a) a complex of bovine chymotrypsin with the substrate analogueformyl-L -tryptophan,': (b) the complex of bovine trypsin with bovine pancreatic trypsin i nhibi tor,24 c) the complex of pork trypsin withsoybean trypsin inhibitor;25 (d-f) show the same atoms from a viewpoint 90 away. T he ori entation and structures have been arranged togive, as nearly as possible, an exactly comparable view i n each case. These diagrams are derived from three completely independent crys-tallographic structure determinations of three dif ferent pairs of molecules. T he largest difference is in the position of the histidi ne ring,confi rming the idea that this side chain is able to move after the substrate is bound. W ith this exception, all homologous atoms are simi lar-ly placed within less than 1& I thank Dr. R. Huber for providing the unpublished coordinates used to generate (b) and (e).ation,2Q nd the tetrahedral complex has only a tran-sient existence. These differences of stabilization en-contact areas in the complexes which the enzyme

Proposed Catalysis in SubtilisinSubtilisin has a similar charge-relay system, nd

around the substrate.51 I t has been pro osed that theergy can be reasonably explained by the different there is a similar arrangement of reacting groupsforms with an inhibitor or a small s~bs trate.~~ subtilisin substrate moves by about 1K n the transi-(51) J D Robertus, R A A lden, J J Birktoft, J Kraut, J C Powers,

(50) C H Chothiaand J Janin J Mol Bioi 100, 197-212 (1976) andP E Wdcox, Blochemutry 11,2439-2449 (1972)


7/8

Structure and Mechanism of Chymotrypsin 151ol. 9,1976( a ) G y193n G y193

#%-' His 57

G y 193 G y 193

Figure 6. A stereochemically realistic series of drawings,54 ndicating the proposed steps in formation of the acyl-enzyme, following thescheme proposed pr ev i ~usl y .~* * ~~onds between atoms in the substrate are indicated in black. (a) Rotation of Or (Ser-195) to attack thecarbonyl carbon of the substrate, which moves into a tetrahedral conformation. Thi s must be accompanied by some displacement of His-57, since otherwi se Or-195 would pass too close to C"-57.55 (b) The tetrahedral intermediate, as observed in the inhibitor complexes. (c)T he hydrogen bond f rom the displaced H is-57 is transferred from 07-195 to the leaving group X. After the proton has been transferredfrom His-57 toX, the leaving group is expelled. (d) The tetrahedral carbon moves into the plane defini ng the ester group of the acyl en-zyme. The leaving group is replaced by a water molecule hydrogen bonded to H is-57, which can move in to attack the ester group. Furthersteps of the process would be indicated by (c) , (b), (a), where this water molecule (or an OH group) takes the place of the leaving group.

tion from the Michaelis complex to the acyl-enzyme.In this process, the distance from the substrate NHto CO-125 (the analogue of chymotrypsin's CO; seeFigure 2) shortens from about 4 A to a hydrogen-bond distance. It has been suggested that the tighterbinding of the substrate which results from thismovement providesadriving force for catalysis.52 Byanalogy, it is argued that a similar process might op-erate in trypsin and chymotrypsin.Such a movement appears unlikely in chymotryp-sin, since the substrate-binding pocket greatly re-(52) J . D.Robertus, J . K raut, R. A. Alden, and J . J . Birktoft, Biochemis-try, 11,4293-4303 (1972).

stricts possible movements of the substrate. More-over, the chemical changes in the various steps of thereaction provide no obvious mechanism which wouldexplain a change in the length of the hydrogen bondThe structural results for the two trypsin-inhibitorcomplexes confirm that the substrate movement pro-posed for subtilisin does not occur in trypsin. In Fig-ure 5 we show comparable views of the complex ofchymotrypsin with the pseudosubstrate formyl-L-tryptophan (which we take as a model of a Michaelis

complex)37 and of the two trypsin inhibitor com-p l e ~e s . ~~~~~heCaatom in the mouth of the primary

toCO-214.


8/8

152 Blow Accounts of Chemical Researchbinding site is in the same position (relative to meanenzyme coordinates) within experimental error. TheNH-CO-214 hydrogen bond is shorter in both inhib-itor complexes (3.324and 3.2h;25) han in the pseudo-substrate complex (at east 3.5&,but this differenceis probably at the level of experimental error.Stereochemistryof an Enzyme-CatalyzedReaction38*53*54

When the scissile bond lies in the orientation dic-tated by the primary binding site and by the hydro-gen bonds made from NH-193 and NH-195 to thecarbonyl oxygen 0, the carbonyl carbon C is nearC@-195, nd the amide nitrogen N is close to NC2-57and Or-195. The leaving group X cannot be broughtinto the position observed in the trypsin-inhibitorcomplexes, because these contacts would become tooshort.Development of the Tetrahedral Intermediate(Figure 6a).In the first step of the reaction, the ser-ine 07 rotates about its C-Cp bond. This rotationmust involve a small displacement of His-57, sinceotherwise 07 would pass too close to CC157.55 When07 begins to come in contact with C, i t is movingroughly perpendicular to the plane defined by thethree ligands of C. On closer approach to C, bondingwill develop, and as a consequence C will be dis-placed toward Or, out of this plane and into a tetra-hedrally coordinated conformation.In this conformation, observed in the trypsin in-hibitor complexes, the NC2-0distance is still suitablefor a hydrogen bond (Figure 6b). I t is now possiblefor the full leaving group interactions to be made, be-cause of the closer approach of C to Ca-195 whichfollows from the C-Or-C@ bonds. The relaxation ofthe planarity of the peptide group which results fromformation of the additional bond may assist in devel-oping the leaving group interactions.At this stage we may assume the proton originatingfrom the serine OrH to be localized on Ne2, eavingHis-57 and Asp-102 electrically neutral, and with anegative charge developing on0,as in Scheme11Collapse of the Tetrahedral Intermediate tothe Acyl Enzyme (Figure6c).The structures of theinhibitor complexes show that NC2s too far from Nfor a hydrogen bond, and probably His-57 mustswing back slightly toward the substrate, away fromOr. During the process the hydrogen bond is trans-ferred from07 to N. The proton can now be trans-ferred onto N, initiating the collapse of the tetrahe-

(53) R. Henderson, C. S. Wright, G. P. Hess, and D. M. Blow, ColdSpring Harbor Symp.Quant.Eiol., 36,63-70 (1971).(54) D. M. Blow and J . M. Smith,Philos. Trans.R. SOC. ondon, Ser. E ,272,87-97 (1975).(55) Huber et al.24 have suggested that, in the active enzyme at neutralpH , 07 is in a different orientation from that observed in crystals, for whichthis movement of His-57 would not be needed,

dral intermediate toward the acyl-enzyme. Protona-tion of N implies the repulsion of N fromC, leavingC three-coordinated, so that it will be pulled a fewtenths of an angstrom into the plane of Or, 0,andC. This plane also includes Cp-195, as is usuallyfound in esters.The repulsion of N from C must move it farenough to bring it to a nonbonded distance from C,while accommodating the proton accepted from NC2.The movement of N is sufficient to dislodge the in-teractions made by the leaving group X, facilitatingdiffusion of the leaving group into the solvent. Assoon as possible, N will pick up a further protonfrom the solvent, becoming NH3+ The Asp-His sys-tem again polarizes His-57 because N61 s permanent-ly protonated (above pH 7) and NC2 s now depro-tonated. This polarization causes a water molecule tobe strongly hydrogen bonded to NC2t site W (Figure6d).Deacylation. This water molecule also forms a hy-drogen bond to CO-40, and in this position its lone-pair electrons point toward C, from a directionroughly perpendicular to the plane of the esterThe attack of these electrons on C causes itto revert to four-coordination, with development of anegative charge on0and movement of C back to itstetrahedral position. The tetrahedral conformationcan collapse by expulsion of Or. The proton is trans-ferred from the water molecule first to NC2,hen toOr, as His-57 and07 move back to their starting po-sitions. C and its ligands now form a carboxylategroup, and the remaining product of hydrolysisleaves the enzyme in this form.Conclusion

The activity of chymotrypsin can be considered toderive from several different effects: (i)chemical ap-paratus (the charge relay system) which, by polariz-ing hydrogen bonds facilitates the proton transfersteps which occur in different steps of the reaction;(ii) a binding site, which immobilizes the scissilebond of a substrate at the active site; (iii) a preciseorientation between the two, which permits each re-action step to involve only a small rotation about asingle bond, and which aligns the attacking groupsnear to optimal directions; (iv) lowering the actiua-ti on energy of the reaction by features which reducethe energy of the enzyme-substrate complex, at thetransition state.All of these effects are, however, different aspectsof the structure of the active site, and cannot reallybe separated from one another. Together they in-crease the rate of the hydrolytic reaction and causethe enzyme to function as a catalyst.The rather simple nature of the reaction, and thesmall number of atoms which need to move, allow usto define the process in detail.

chymotrypsin mechnism

Documents