On the Active Site of &hymotrypsin

THE Joum.u. OF I3romorca~ CHEM~~Y Vol. 244, No. 10,Iasueof %fay 25, pp. 2M3-2674, 1969

Printed in U.S.A.

On the Active Site of &hymotrypsin

ARSOLUTli: COSE’IGUKATIOSS ASD KINETICS OF HYDKOLI-SIS OF CYCLIZED ASD T\‘OKCI-CLIZED SUBSTRATES*

(Received for publication, December 30, 1968)

SAUL G. &HEX., ALEKSASDER NILOVANOVI&S RICHARD 31. SCHULTZ,! AND SASDRA I'. WEISSTEIN

Fwm ihe Department of Chemistry, BTandeis University, Waltham, Massachusetts 0215/,

SUMMARY

The absolute steric course, stereoselectivity, and kinetics of hydrolysis of cyclized and noncyclized ester substrates by a-chymotrypsin are compared. The cyclized substrates are of the form

0 II

D(-)-I-A,-X-Y- = -NHC-; D(-)-I-B, -X-Y- = 0 II

-OC-, S(-)-I-C, -X-Y- = -CH=CH-; D(-)- 0 II

I-D, -x-Y- = -0-; S( +)-I-F, -X-Y- = -CHtC-; R(+)-I-G, -X-Y- = -CHCHz--. The noncyclized substrates are analogous a-substituted @phenylpropionates, CJ-I&H~CH(x)COrR, II-A, x = -NHCOCH,; II-B, x = OCOCHI; II-C, x =-CH,CO,R; V-A, x = Cl; V-B, x = OH; V-C, X = CHr. The absolute configurations of I-C, I-F, and I-G are established in this work; absolute con- figurations of the other compounds were previously known or recently established.

The active site of a-chymotrypsin is described in terms of loci complementary to a conformation of L( +)-H-A, in which the aryl and ester groups are transoid, the /3-a@ and a-hydrogen are directed into the enzyme, and the a-aceta- mido and hydrolyzing ester groups are directed along the exterior of the enzyme. Conditions for hydrolysis of L and D enantiomers by a-chymotrypsin are described. Associa- tion is generally dominated by the /3-a@ group at its locus, ar. Hydrolysis requires the ester group to be at the hy- drolytic site R. Compounds H-A, II-B, and II-C hydrolyze in the L sense as the large a-substituents occupy the a- acetamido site, am, and may not fit into the a-hydrogen

*This work was supported by Grant GM94584 from the Nn- tional Institute of General Medical Sciences.

1 Present address, Department of Chemistry, University of Bel- grade, Yugoslavia.

$ United States Public Health Service Predoctoral Fellow.

locus, h. Both enantiomers of V-A and V-B are hydrolyzed as the small polar substituents, Cl and HO, may be directed in the L sense toward am and in the D sense toward h. Only the L enantiomer of V-C is hydrolyzed, and the small hydro- carbon group, CHs, may not fit at h. All of the cyclized sub- strates, IA-IG, with five- and six-membered second rings, are hydrolyzed preferentially in the D sense, but with widely varying reactivity and stereoselectivity. In every case the more effective D sense orientation directs the -X-Y- grouping toward h-ur, leaving the remainder of the carbon chain in the same conformation and orientation as that of L-II-A during its enzymic hydrolysis. In D( -)-I-A, D( -)- I-B, S( -)-I-C, and D( -)-I-D, the groupings --X-Y- fit well into ar-h and very high reactivity and high stereoselec- tivity result. The a-heteroatoms of I-A and I-B and I-D, and the planar v-electron structure of I-C permit fit in w-h, as do the a-heteroatom substituents in V-A and V-B. In I-F and I-G, the a-CH2, like the a-CH3 in V-C, does not permit the -X-Y- grouping to fit easily into bar, and low re- activity and low stereoselectivity result, still with D prefer- ence. During hydrolysis of these cyclized substrates in the D sense, the am site is unoccupied. When the second ring in a cyclized substrate is eight-membered, hydrolysis in the L sense may be preferred.

There has been interest. in the hydrolysis by cu-chymotrypsin of cyclizcd substrates since the discovery of the first of these compounds, l-keto-3-carbomethoxytetrahydroisoquinoline, I-A (1, 2). The D( -) enantiomer of this compound was hydrolyzed very rapidly by the enzyme, kcat/K, = 4.3 x 10’ ~i SW-1 (2), essentially as effectively ils t.he natural substrate, L(+)-methyl N - acetyl - p - phenylalaninate, C6H&H&H(NHCOCHJCOn- CHJ, II-A (3). Compound I-A may be considered to be an analogue of Compound II-A, in which the cyclized structure restricts severely the number of possible conformat.ions’ as compared with the many which are accessible to the noncyclized substrate. Because of the high reactivity of Compound I-A, and because no substrate was known at the time which showed high reactivity in the absence of an a-acylamido group, the properties of Compound I-A were thought to give direct informa- tion as to the mode of association of nat.ural substrates and the

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Issue of May 25, 1969 S. G. Cohen, A. klilovanocii, R. Af. Schultz, and S. Y. Weinstein 2665

geometry of the artive site. It was inferred (l-4) (a) that the reacting conformation of 1,(+)-11-A is described by the rigidly constrained stereochemistry of the cyclized compound D( -)-I- :\, (b) that the topography of the act.ive site would be comple- mcntary to that of II-A, and (c) that the amide linkage in D( -)-I-h is important for ib reactivity and occupies the same position in the active site as doti the amide group in a noncyclized subst.rate, such as I>(+)-II-A. However, impressed by the fact that the 11 enantiomer of I-A is hydrolyzed rapidly, and not the L, we came to a different view as to the inferences that. should be drawn about its mode of association and about the propert.ies of the active site.

I-A, X = NH, Y = C=O

y\

cc<

I-B, X = 0, Y = C=O

0 XH I-C, X-Y = CH=CH I-D, X-Y = 0

CO& I-E, X = CH?, Y = C=O, It = CHJ I-F, X = CHZ, Y - C=O, R = C?H; I-G, X-Y = CH&Hz

It seemed to us t.hat the active site of a-chymotrypsin might comprise parts complementary to the parts of a small molecule subst.ratc such as L(+)-II-A, in a particular conformation in which the P-aryl group and hydrolyzing ester groups are transoid (Fig. I) (5, 6).

(a) The UT site’ is a cavity or fold into the enzyme in which the P-aryl substituent provides the primary associative effect. Association at this site leads (7, 8) to the activating conforma- tional change of the enzyme (9). (6) The h site is one of re- stricted volume into which the (Y-H or other small substituent may fit. It is essentially continuous with the UT site, the /3-aryl group and a-11 of L(S)-II-A being cisoid in the reacting con- formation. (c) The urn site provides a hydrogen-bonding asso- ciation with the a-acylamido group of the substrate, which, when combined with the association at UT, rest.ricts rotation about C,--C,q and CcCphenyl of the bound substrate (7, 8). (d) The hydrolytic site, n, contains the serine hydroxyl (10) and imidazole groups (11). In an effective nssociution the restric- tions imposed by t.hc binding at ar and am and t.he fit at h deliver the group to be hydrolyzed to t.he site n. The a-acylamido and ester groups of a small molecule substrate such ‘as L( +)-II-A arc models for the large YHrterminal and COOH-terminal chains of a natural polypeptide substrate of this endopeptidase, and the complementary sites am and n are at or near the enzyme surface

(798). Hydrolysis in the L sense proceeds via Ltn association like that

indicated in Fig. I. Such stereospecificity is observed with Compound II-A (12) and with oxygen and methylene analogues, ethyl cr-acetoxy-/?-phenylpropionate (11-D) (13) and diethyl cr-bcnzylsuccinat (II-C) (14), because the unreactive enantio- mers may not fit effectively into the active site. The association for all of t.hese compounds is dominated by the /?-aryl group at ur, and for hydrolysis to occur the ester group must be at 72. One of each pair of enantiomers would fit as in Fig. 1, while the other enantiomers would have to reverse the positions of the a-13 and the cu-substituent., placing the large cY-acylamido (II-A), cY-acyloxy (II-B), or a-carbethoxymethylenc (II-C) group at /L. These groups cannot be accommodated at h and high stereo- specificity is observed. The hydrolyses show widely varying

‘The abbreviations used are: ar, B-aryl site; am, cr-acetnmido site; h, cu-hydrogen site; n, hydrolytic site.

4 R ‘“\\, J am -

PIG. 1. Association of L enantiomers of esters of a-substituted B-phrnylpropionic acids with a-chymotrypsin: Compound II-A, S = P;‘H, It = CHB, L(+)-methyl .V-acetyl-fl-phenylnlaninatc; Compound II-B, X = 0, H = -CII,, L(-)-e&y1 Lcetoxy-6: phenylpropionate; Compound II-C, 9 = CHQ, It = OC2H6, It(+)-diethyl a-benzylsuccinnte.

/ I

I I I I

\ \ g ---rC\

FIN;. 2. Association of D enantiomers of esters of a-substituted propionic acids with a-chymotrypsin: Compound III-A, X = 0, It = CH3, u(+)-ethyl a-ucetoxypropionate; Compound III-B, X = 0, Ar = CsIIb, D(-)-ethyl a-benzoyloxypropionate; Com- pound III-C, X = NH, Ar = &Ha, D(-)-methyl cu-bcnzamido- propionatc; Compound III-D, X = NH, Ar = CsHS, I)(+)- methyl a-picolinylamidopropionate.

reactivity depending upon the capacity of the cY-substituents to bind at am and restrict the conformat.ions of the substrate.

Hydrolysis in the D sense may occur under several sets of circumstances. (a) If a substrate lacks a j3-aryl substituent, binding at ar may be provided by the a-substituent, as indicated in Fig. 2. This leads to hydrolysis of esters of D enantiomers of est.ers of cr-acetoxy-, cr-bcnzoyloxy-, a-benzamido-, and (Y- picolinylamidopropionic acids, compounds III-A, 13, C, I) (13, 15, 16). To the extent that the cr-substituents of L enantio- mers of these compounds bind at am, these enantiomers also hydrolyze, leading to diminished stereoselectivity. Rinding of the Lu-subsbituents of the L enantiomers at ar leads to incffecti~e a5socintion, since this does not place the group to be hydrolyzed at site n. Inversion of stereospecificity, preferred hydrolysis in the D sense occurs with Compounds III-A, IS, and 1). (b) D

Enantiomers may hydrolyze when the (Y- and P-substituents arc similar and may each associate at am and ur, and this is seen in the low stereoselectivity in hydrolysis of esters D- and L-(Y- acetoxysuccinic acid, IV (6). (c) n Enantiomers of substrates which have a /?-aryl substit.uent may hydrolyze if the cr-substit- uent is such that it may fit into h (Fig. 3).

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2666 On the Active Site of ad’hymotrypsin Vol. 244, No. 10

FIG. 3. Association of u cnantiomers of esters of a-substihlted fi-phenylpropionic acids with cu-chymotrypsin: Compormd V-A, X = Cl, effective association; Compound V-13, X = OH, elrective association; Cornpour~d \‘-C, X = CIIS, Ihssociation fails.

&---2

- FIG. 4. Association of LI enantiomers of cyclized substrates

with WY-chymotrypsin: Compound I-A, X = NH, Y = C-=0, I)( -)-l-keto-3-cart~omett~oxytetrahydroisoquinoline; Compollnd I-B, X = 0, I’ = C=O, n(-)-methyl 3,4-dihydroisocoumarin- 3-carhoxylate; Compound I-C, X-Y = CH=CH, S(-)-methyl 1,2-dihydro-2-naphthoate; Compound I-D, X-Y = 0, D(--)- methyl hydrocoumsrilatc:; Compound I-F, X = CIIy, Y = C-O, S(+)+cart)ethoxy-a-tetralone; Compound I-(;, X-Y = C!II,CH2, R(+)-methyl 1,2,3,4-tet.r2th3dro-2-naphthoatc.

This is observed in the hydrolysis of both enantiomers of esters of a-chloro- (li) and cr-hydroxy-B-ghellyll,rol,iollic acid (V-A, R) (13, 17). IIowevrr, the a-methyl compound (V-C) was hy- drolyzed with L specificity, indicating that the a-CH3 group may not fit at h, and t,hat there may be either a sharp space restriction or a polar requirement for fit at h (14).

It appeared to us most likely (t&8) that the cyclized substrate I)( -)-I-.1 associates with the enzyme with its aryl group at UT and the ester group at n, placing the fused amide group at h- ar (Fig. 4).

The second ring would provide the restriction of rotations about Co--C8 and CrCphsnyl, otherwise provided by binding at the two sites, am and ur, iu L( +)-II-A, place the ester group at n, and lead to high reactivity (7, 8). A favorable orientation, such as that indicated in Fig. 4, may not be achieved by L(S)-I-A,

when the active site is constructed in a way complementary to the indicated conformation of L(+)-11-d (Fig. 1) (7, 8). This indicated that Compound D( -)-I-A made no use of the am site, and the amide group in I-A was not required for high reactivity in such cyclized substrates. This idea w&s confirmed in our

laboratory in the high reactivity of I)( -)-methyl 3,4-dihydroiso- coumarin-3-carbosylate (I-B) (7,8), and a study of other cyclizcd substrates was initiated. At about the same time similar ideas were developed aud confirmed by the high reactivity of the S( -)- p-nitrophenyl ester of 1,2-dihydro-2-naphthoic acid (I-C), R = p-NO&,Hs (18). Soon D( -)-methyl hydrocoumarilste (I-D) was also reported to be rapidly hydrolyzed by a-chymotrypsin (19). Thus, reactive cyclized substrates either do not use specifically, or do not have, amide groups, and would not be directly informative as to the placing of the acetumido group of L( +)-II-A in the active site. However, in our view, they do utilize the primary binding site, which we now refer to as ur-h. Information about the properties of this important part of the active site might bc inferred from coneideration of the widely varying reactivit.ies of poc;lsible cyclized substrates, and of their absolute configurations and stereospecificities, and by comparison with structurally related noncyclized conforma- tionally free compound?. Further aspects of this study are now described.

I<XPERIMENTAL I’IlOC!~:I~URE

a-Benzylsuccinic Acid (Compound II-L))

This compound (20) was prepared as described previously (14) by condensation of diethyl a-benzylmalonate with cr-bromo- acetate to triethyl I-phenylpropane-2,2,3-tricarbosylate, saponi- fication of the tricster, and decarbosylation, 53% yield over-all, m.p. l&161”, literature (20) m.p. 160-161”.

&ethyl a-~enz~ykuzcinak (Compound II-C)

This was prq)ared from cY-benzylsuccinic acid, boiled in ethanol in a stream of IICI for 1.5 hours, 62y0 yield, b.1). 129” (0.7 to 0.8 mm), literature (21) b.p. 169-170” (5 mm).

S-Carboxy-cY-t&alone (Compound VI)

Benzylsuccinic acid (1.01 g, 4.8 mmolcs) was allowed to stand in 10 ml of concentrated ILSOa overnight. The solution was poured onto ice and the precipitate WI.S collected and wmhed wit.h water, Compound VI, 0.70 g (3.7 mmoles) 77yc yield, n1.p.

149 -149.5” from Walter, literature (22) rixp. 149”.

CdLoOa Calculated: C 69.46, H 5.30 Found2: c 69.45, II 5.11

S-Carbomethoxy-cY-telralonc! (Compound I-E)

.4 solution of 0.40 g (2.11 mmolw) of the acid, VI, in 20 1111 of methanol was saturated with dry HCl and allowed to stand over- night. It was poured into water, clxtracted with ethrr, and washed with NaHC03 and water, leading to Compound I-E, 0.27 g (1.34 mmoles) 63% yield, m.p. 30-31”, literature (23) 32-33”.

Calculated: C 70.57, I1 5.!12 I”ollrld : C 70.60, H 5.85

IIydrolysis of S-CarbollletAocy-a-ttralonc (Compound I-B) by a-Chymotryps-in

A suspension of 0.0799 g (0.391 mmolc) of Compound I-E in 20 ml of H20 containing 0.0207 g of a-chymotrypsin, in a pl I-

* Analyses by Schwarzkopf Microunalytictrl Laboratory.

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Issue of May 25, 1969 S. G. Cohen, A . JZilocanooiC, 12.

stat at. pII 7.2, 27”, consumed 2.79 ml of 0.1 IS IKaOH in 2.5 hours, 71 yi rcxtinn, the reaction then l~rocrcding very slowly. ‘I’ht: rnisturt WIS cstr:wt.ed wit,11 ether, lwding to 0.0205 g (0.10 nrn~ole), 877; recovery of unhydrolyxed I-E, with infrnred spectrum iu chloroform identical with t.hat of t.hc starting ma- terial, aOl,., -0.51”, c, 2.05 in mtthanol, [a]:’ -25”. The :~queous solution \vitS acidified alld lynphilizcd, a11t1 the residue wns estrnctcd with ether, leading to ol)tically active keto acid VI, 0.0541 g (0.284 mmole, -1OOL;;. yield), with iiifrarcd spectrum iu chloroform identical with that of the inactive acid, III.~.

139-143”, c&I,> +0.28”, f, 1.88 ii1 nwth:inol, [a],?” +15’.

Calculated: C G!).4B, II 5.30 Found : C 694. H 5.45

S-Carbelhox!l-cr-telralonc (Compound I-F)

This WI.‘; prepared iu 89(il yield from l.iO g (8.63 m111o1c~) of 3-c:lrbos~-Lu-tetrnloiic folloniug ii reported lnwwdurc, n1.p. 38” from cthru~nl, litrrature (24) [email protected]”.

ZZUdrotysis of ,%Carbethoxylelratone (Compound OL-I-1“) by cY-ChymolrypsirL

A suy)enGon of 0.836 g (3.83 mmnlcs) of C’om~~nund T-F iu 400 ml of Hz0 coiilniuiug O.Oi55 g of c+cl~~mntry~~siii iii n 1jtr- stat at pH 7.2, 27”, consumed l.i24 ml of 1 s SaOlI in 1 hours, 45.0’:; rcactiou. The reaction slnwd dowt~ but did not stol). The misturc was cstrwted with etlirr, lwdiiig to 0.396 g (1.82 mmolcs), SS(;l, rccnvrry of uuhydrolyzed o~~tically active I-F, with infrared spectrum in chloroform itlrnt~kd with that of IJL I-F, (~,l,~ - 0.50”, c, 2.81 in mcthannl [a]:’ -18”. The :~queous solution was acidified and lyophiliztd, and the rwidur was estractcd with ether, leading to 0.311 g (1.6~ mmnles), 957; yield of partially active 3-c:Irbos~-cY-tetrtllolle, (+)-VI, with infrared spectrum iu c+lornform identical wit.11 that of I)L-VI,

n1.p. 139-140”, a,bs +0.55O, c, 1.65 ill ruethannl, [a]:’ +33”.

GlI~L003

C&!111ut.ed: c (it).‘l(i, II 5.30 Found : C 09.30, H 5.42

Methyl 1,~,3,~-Tetrahydro-2-)7npilthoate (Compound Z-G)

Sodium (10 g) WRY added, over :I period of 45 mill, to :1 SOIII- t.ion of 4 g (0.0232 mole, Alathrsou, m.1). 183-186”) of 2-naphthnic acid in 80 ml of boiling isoxmyl :~lrohnl, and boiling was cou- tiuuctl for 1 hour. The misturr was trrnted with wxtrr :Illd ether, nud the aqueous solution was ncaidifkd and c~st.r:wted with ether, leading to 1.7 g (9.7 mmole~), 42(>; yirld of I ,2,3,4- tetrahydro-2-naphthoic acid, Compound VI I, n1.p. 9(t!)2”, litera- ture (25) 111.p. 94”. .A solution of 1.45 g (8.2 mrnole~) of the acid in 25 ml of mrth:mnl was saturated wit.11 IICl and boiled ulldcl reflux for 3 hours. The mixture was diluted with water, ex- tracted with ethrr, and washrd with sodium bicarbnnutc, leading to methyl 1 ,2,3,4-trtmh~dro-2-Il~ll)hthn:lt~, Compound I-G, 1.20 g (6.3 mmoles), 77yG yield, b.1,. 129” (5 to 6 mnl), literature (26) b.p. 102-105” (1 em).

llydrolysis oj Nefhyl 1,2 ,S ,/t-Tetrahy(lro-8-naivi,fhoate (Compound I-G) by cu-Ch!yntoirypsin.

A susprilhioll of 0.730 g (3.85 mmole~) of (‘oiu~~oiu~d 1-G in 245 ml of wter w:w treated with 0.105 g of a-chymotr~l~~in in 5 ml of wit.er :it 111 r i.2 in il 111 r-st.:lt. The react ioii wis lnwcced- iug very slowly nftrr 48 hnurd, 1 .M ml of I s NnOII having been coilwmed, -197; rwctiou. l’hr re:wt.ion mixture WIS rrtrarted with ether, leading to 0.316 g (1.66 mmoles), 855; wrovrry of ol)tic:dly twtivc uuhydrolprd (-)-I-G, a,,~,~ - 1 .W, c, 3.8 ill (Ill(‘ln, [a]Z’.’ -29.2”. The iufrnred *prctrun~ was idriitiral with that of the T)I. starting mrlteri:ll. The aquoori+ wlulion leas brought to J)H 2 with lIC1 :u~tl l~ophilizcd. The twitlw was estrwted with dry rthrr Irnding to opt iwIly nctivc~ tetrahydro-2- i~:iphtlinic arid, (+)-VI 1, 0.305 g (I .74 rum&~), 9Xc; yield, 111.1). 97-98”, litcraturr (27) m.1). 99”, (Y~,I,~ +0.61”, c, 1.6 in C’H(&, [a]:‘.’ +38”, liter:lture (27) 51.8”. The infmrrd ,+pec- trurn in CJHCln W:IS ideuti~;d with that of thr initial DI, :Icitl.

Calculated: C 74.97, II C,.M Found : C 74.75, H 70)

.I[et/?,tJl 1 ,2-Z>ih!Jdro-2-?laph(t~ofl~e (Compound I-C)

.I solution of 2-n:ll)hthok wid (20 g, 0.1 I6 molr) in x~ueou~ ~~ota~siuni c:irboiiate was treated with I80 g of 5’;‘; wtlium :~rnal- gam at 0” for 5 hours :IS dtwritwd in :I publishrtl procedure (28). ‘I’hc~ solution was acidified, lwdiug to :I rnisturc of the 1 ,2- :md 1 ,4-tlihydro- acids. ‘I’hk WIS diw~lwd iu l~ota~sium bicorbnu:~te a11t1 fr:wtioncdly acidified with 2 s 110 until thr melting point of the lwccil)itak iviis below 162”, that. of the 1 ,-I-dihydro- acid (29). The remaiudrr WIS precipituted with arid, leading to 1 ,2- tiih~dro-2-I1:ll)hthoic wid, C’ompnrmd VI II, 4.0 ,g 0.023 mole), 2Ooi, yield, m.p. 102- 103” from witrr, literaturr (28) m.1). 103”. l)ittzometh:uw, prepared from 9 g of .\‘..l-‘-tlilrlcth~l-.~ ,S’- di~~itro~otcr~~htll~~l~rnidc (30), was distillrrl into :I solution of 3.60 g (0.0207 molt) of 1 ,2-dihydrn-2-naphthoir wid in rthel and allowed to stxud ovrrnight.. Thr aolutinu w:is roriwntrated :iritl distillctl lrading to mrthyl I .2-dill?-dro-2-nnl,hthoatc, (‘ompou~~tl T-V, 3.20 g (0.0180 mole), 877; yield, b.1’. 1-13-143” (9 mm).

c,211*202

Calclllatcd: C 76.57, H (iA;< I“ound: C i43.3, II 6.25

Ziydrotysis o-f nL-.Uethyt I ,B-ZXhydro-%naph!hoate (Compound I-C) by a-Chymolrypsin

.I suq)ensinu of 1.026 g (5.45 inmoles) of Compound 1-C iii 395 ml of water, trrnted with 0.091 g of a-chymotrypsiu ill 5 1111 of water at 131-f 7.2 in a 131 I-stat., consumed 2.18 ml of 1 K N:IOH iu 3 linurs, 4O’y, reaction. Addition~il portions of rnzyrne, 0.066 g and 0.095 g, added during t.he nest 18 hours led to consumption of only 0.16 ml more NaOH, total rractinu, 43:;. The misture WIS cstracted with ether leading to 0.536 g (2.85 mmolcs), 897; recovery, of uuhydrolyzcd (+)-I-(‘, \vith infrared spectrum identical with that of the starting material, cyO~,~ +3.10”, c, 2.75 in met.hannl, [a]z6.’ +150”, 1r1.1). 35. 36”. The aqueous nolutinn wls acidified and lyophilized :mtl the rrsidw tvas rstrwted with ether, leading to 0.353 g (2.03 ~n~~~oles), Sic/;, yield of (-)-1 ,2-dihydro-2-naphthnic arid, Compound (-)- VIII, r11.1). 101.-102” literature (28) lOlo, a,,hh -3.80”, c, 1.38 in

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2668 On the Act,itle Site of a-Chymotrypsin Vol. 244, Ko. 10

cIIcl~ [aIt -276”, literature (18) -283”. The infrared spcc- trum was identical with that of the DL compound.

Assignment of d bsolute Conjgurations

S( -)-a-Benzylswcinic Acid (??I)--1)iethyl cr-benzylsuccinate, Compound II-C (0.98 g, 3.7 mmolcs), suspended in 20 ml of water containing 0.11 g of cY-chymotrypsin in a pH-stat at 1~11 i.2, consumed 1.9 mmolea of NaOH in 18 hours and the reaction stopped, equivalent to 100% hydrolysis of one cnantiomer (14). The mixture was extracted with ether, and the extract was con- centrated leading to unhydrolyzed (-)-diester II-C, 0.43 g, 88% recovery, with infrared spectrum in CHC13 ident,ical with that of DL-11-c. A portion (0.091 g, 0.34 mmole) was hydrolyzed with I ml of 1 K EaOH and 1 ml of ethanol at room temperature for 24 hours, leading to S( -)-a-benzylsuccinic acid, 0.061 g (0.29 mmolr, 85%:. yield) n1.p. 159161”, U&s -0.40”, c, 1.5 in ethyl acetate, [a]z6 -26”, literature m.p. 164.5’, [cy]t5 -27”, c, 2 in ethyl acetate (31).

S( +)-S-Carboxy-c~-t&atone (Compound 1’1)-:1 solution of 0.100 g (0.48 mmole) of S( -)-a-benzylsuccinic acid, above, in I .O ml of concentrakd ITZSO~ W:LS allowed to stand at room tempera- ture for 2.5 horns. The solution was poured onto ice and the precipitate was collected and washed with water, leading to S( +)-3-carboxy-a-tetralone, Compound VI, 0.030 g (0.16 mmole) 33% yield, melting in part at, 139” and completely at 149”, with infrared spectrum in chloroform identical with those of inactive VI aid of (+)-VI isolated from enz.vmic hydrolysis of DI,-I-E,

a0bs +0.30’, c, 1.19 in methanol, [(Y]:’ + 25”.

CLlK003 Calculated: C 69.46, II 5.30 Found: C (39.50, H 5.45

R( +)-Tetrahydro-2-naphthok Acid (Compound VII)----Corn- pound S( +)-VI, 0.140 g (0.76 mmole), prepared above by cycli- zation of S( -)-benzylsuccinic acid, 0.6 ml of t.oluene, 0.6 1111 of acetic acid, 0.6 ml of water, and 1.5 ml of concentrnkd hydrochlo- ric acid were added to 0.6 g of amalgamated mossy zinc and the mixture was boiled for 5 hours (32). Three 1.5~ml portions of concentrated hydrochloric acid were added at I .5-hour intervals. The mixture was cooled, diluted with water, a~~1 extracted with ether, leading to R( +)-tekahydro-2-naphthoic acid, Compound WI, 0.081 g (0.46 mmole) 61% yield, 111.p. 93-9-P, a,~,, +0.56”, c, 1.5 in CHCl,, [a]:’ +35”. The infrared spectrum in chloroform was identical with that of the acid ohtained hy the enzymic hy- drolysis of DL-I-C.

CnH,zOz

Calculated: C 74.97, H 6.86 Found : C 74.77, H 6.83

S( -)-Jlethyl 1 ,%Dihydro-2-naphthoab (Compound Z-C)- S( -)-I ,2-l)ihydro-2naphthoic acid (0.40 g, 2.5 mmoles, [cr]:” -276”) obtained by the enzymic hydrolysis of nn-methyl 1 ,2- dihydro 2-naphthoate, was dissolved in anhydrous ether and treated with excess diazomethane overnight. The solution was concentrated u11d the residue was dissolved in CHCl,. This solution was n-ashcd with water, 5% NaHCOI, and water, dried, filtered, diluted with hesanr, filtered, and concentrated, leading

TABLE I Hydrolysis oj cyclized substrates by a-chymotrypsin

pH 7.8, 0.1 h’ N:iCl, 2-S’=.

CO&Ha --

(‘oqound

. Substrate

I Sol\vlt I [El x

--~ 1.

- .- -.

I - 10’ x N

c--o TT*O” 0.11 D(-)-I-A NII D(-)-I-A SII c=o 1)(-)-I-A XII c=o n(-)-I-B 0 c-0 ~(-)-I-13 0 C=O 1)(-)-I-B 0 c-o DJ.-I-B 0 c=o 1.(+)-I-B 0 c=o s(-)-I-C CH CH I,J.-I-F , Cl12 : c-o DI.-I-G CH? / CHt

- ..- ~- . .--.-. -

1O7o CHaOIP 10% CII#.xc H*0d 10% CHzOHe

0.31 0.10 0.09 0.39

107, CII&N’ 0.11 56 2.5 Hs00 0.09 25 1.3 H?O* 17.0 1.8 28 10% CIJ$JH’ 4.9 0.2 0.23

set-’ ?ndl

26 0.64 12 3.7 20 0.98 42 0.70 27 : 12

kc.JG

Y-1 XL-1

4.1 x IW 3.3 x 103 2.0 x 101 8.0 x 10’ 2.4 X 18 2.2 x 101 1.7 x 10’ 64

9.1 x 102 lWc C2H50Hj 18.0 0.2 / 23 ; 8.8 10% CHsOH 37. : I i -lk

D Eight runs, 0.5 to 3.0 X 10e3 M substrate, 2.5 X lWa M potas- sium phosphate.

* Ten runs, 06 to 2.8 X 10mZ .M substrate, 1.25 X 10ea M potas- sium phosphate.

c Seven runs, 0.6 to 2.4 X IO-* M substrate, 1.25 X 10m3 M potas- sium phosphate.

d Reference 8, nine runs, 0.25 to 1.7 X lo+ 11 substrate, 2.5 X 1OmJ M potassium phosphate.

e I:ipht runs, 0.5 to 2.3 X 10m3 M substrat,e, 1.25 X 10Aa M potas- sium phosphate.

1 Nine runs, 0.5 to 2.1 X 1P M substmt.e, 1.25 X lo-* M potas- sium phosphate.

* Six runs, 0.2 to 1.5 X 10-J M substrate, 5 X 10e3 Y potassium phosphate.

* Reference 8, five runs, 0.5 to 1.3 X lOma M substrate. i Thirteen runs, 0.15 to 1.5 X IO-* M substrate. j Ethyl ester, six runs, 1 to 7 X lO+ H substrate. *Three runs, 1 to 2 X 10e3 M substrate, average value

of V/(E)@).

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Issuct of Rlay 25, 1969 S. G. Cohen, A. MilovanoviC, R. Ai. Schultz, and S. Y. Weinstein 2669

TABLE II Hydrolysis of noncyclized subslrales by a-chymotrypsin

PII 7.8, 0.1 s NaCl, 2.5". CeHa-CII~-CIICO&~Ha

I A

_.-_ - .-

Substrate _._. .~

Compound A

_ ._

1,(+)-11-A CH&ONH L(+)-II-A CH&ONH L(+)-II-A CII,CONH DL-11-c c211s0*cc11* DL-V-C CHa

II-E II

Solvent El

H20a lOcr, CH1011” 10% CH&Nc 10% C2H5011d 10% C,HsOIl’ 100/b CsHsOH’

10” x Y

0.10 0.18 0.10

37. 7'. :,5.

a Methyl ester (8) seven runs, 0.3 to 1.6 X lo+ Y substrate, 2.5 X lo-* M potassium phosphate.

D Methyl ester, 11 runs, 0.6 to 2.6 X 10m3 Y substrate, 1.25 X 10-a M potassium phosphate.

c Methyl ester, nine runs, 0.5 to 2.4 X 10e3 M substrate, 1.25 X 10ms M potassium phosphate.

to 0.28 g of a mixed liquid-solid, 6Oc/l, yield of the methyl ester I-C, &l>fi -1.69”, c, 1.09 in methanol, [(Y]:’ - 169”. The nuclear magnetic resonance spectrum in CDC13 showed 4 aromatic hydrogens at 2.7 T, two olefinic doublets centered at. 3.4 and 3.9 r corresponding to 2 vinyl hydrogens, a methyl peak at 6.3 7, and a multiplet at 6.3 to 7.0 r representing 3 methylene and methylene protons. Except for absorption due to methyl the spectrum c’or- responded closely to that of 1,2-dihydro-2-naphthoic acid. Im- purity was observed at 6.0 and 8.7 r, corresponding to 0.5 proton, 7% of the total hydrogen.

In the kinetic studies t.he enzymic hydrolysis was carried to completion in six of the kinetic runs. In each case 0.63 eq of alkali was consumed rapidly, indicating presence of 637, of S( -)-cqter, the remainder being largely the less rrsct.ive lt( +)- ester.

Kinetic Studies

These were carried out as described previously (8, 13). Init.ial zero order rates of enzymic hydrolysis were determined as a function of concentration of substrate. Plots of l/I against l/S were made and k,,, and K, were calculated. Details are given in Tables I and II and their footnotes.

IWWLTS

Compounds I-E and I-F, 3-carbomethoxy- and 3-carbethoxy- cr-t&alone, t.he cyclized methylene ketone analogues of the previ- ously studied highly react.ive cyclized amide and ester Substrates I-A and I-B, were prepared by acid-catalyzed cyclization of cY-benzylsuccinic acid, followed by estcrification (Equations 1 and

2). ,A suspension of DL-I-E was hydrolyzed (Equation 3) by 4 X

10-j M a-chymotrypsin, considerably more slowly than Com- pounds I-A and I-B had been. The reaction proceeded beyond hydrolysis of one enantiomer and the mixture was worked up when it had slowed down markedly, at 7152 over-all reaction. Optically active I-E was recovered in high yield [(Y]:’ -25”. Be- cause of the high extent of reaction, this may approximate the

Lit

xc- 7nJf

87. 1.4 55. 6.4 65. 2.7

0.02 2.7

0.09 6.0

kodKm

Jr’ SEC-’

6.2 X 10’ 8.6 x 103 2.4 X 10' 8. 3.

15.

J Five runs, 0.40 to 1.4 X 10ea M substrate. e Six runs, 0.6 t.o 1.5 X 1W3 M substrate, plot extrapolated to

origin. j Six runs, 0.G to 2.2 X IO+ M substrate.

CJ~sCH~~HC0.H

CH,CO,H DL SC-)

H,SO.! Ila ‘lb i 1 I

(1)

01. %+)-VI

ROHl (‘2) 1 0

I-E, R = CH,

I-F, 11 = CzHj

I)L-I-K, I-F

I a- Chymotrypsin (3)

R(-)-I-E, I-F s(+)-VI [a]: -25" [a]:* +37"

true specific rotation of I-E. The hydrolysis product (+)-VI was also obtained in high yield and had [cy]is +15”. Correction for 71,s hydrolysis indicat.es a possible minimum value of +37” for the specific rotation of the acid VI.

A suspension of DL-I-F was hydrolyzed (Equation 3) by 7.5

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

X IO 6 .\I a-Chytttotryl)s;itt :tt it ])rnl)nrt ionalcly lo\~t’r rate tltatt that of 1-E above. The reactiott rlowd down md the misture \vas work(ul up after -15’;s o vcr-all hydrolysis. The u11hJdrolym?rl

ester I-F was recoww1, [a]:’ -1i.i” , :md the hydrolysis product (+)-I’1 was isolwtrrl, with [cy]:’ +33.5”. 130th were: isolated in high yield. (‘orrwtiotls based OII :t minimum \-alum: of 37” for tltc arid rind OII the 4.5,;. estrnt of hydrolyris indicate a specific rotation of 24” for tltcl ethyl ester I-F, similar to that inferred above for the tttethyl ester I-E.

Compound I-G, methyl 1 ,2,3,4-tctralt?-dro-2-tt:~l)htlto:lt~, the saturatctl hydrocarbntt member of this set. of cyclized cstct suhstratcb2 was prepared by wdium-alcohol reduotiott of 2- ttaphthoic acid, followtd by crterification. A suspeusioti of DL I-G was hydrolyzed quite slowly by 1.7 x lo-” Y cr-chgmotryp- sin, :ipparentiy only :hoitt 5 7;. as rapidly as DL-1-F. Agaiti op- tically active products were obtained (Equation 4). The rcac- lion mistuw was worked up after 49’7;, hydrolysis, leadittg, itt high yield, to uttltydrolyzed wter (-)-I-G, [(r]i6 -29”, uttd to hydrolysis product (+)-I ,2,3,4-tc(rah?.dlonal,hthoic. acid (VII), [a]y + 38”. The slwcific rotatiott of this acid is re- ported (27) to he 51.8”.

SC-)-I--G R(+)-\.II

Methyl 1 ,2-dihydro-2-ttaphthnate I-C, the LP.~ utvaturated hydrocurbot~ memhcr of this set attd t.lte methyl ester corrwpnnd- ing tn the previously reported (IX) highly reactive p-rtitt-ophcnyl ester, was prcparrd by (a) sodium amalgam reductintt of 2- naphthoic acid, (b) scparatiott of the 1 ,4-tlihydro- acid from the 1 ,2-dilt~dro-2-naph thoic, Con~pnuttd VII I, attd (c:) esterificat iott of the 1 ,2-dihydro- acid. A suslwnsiott of UL-I-C WLS hydro- lyzed (Equnt.iott 5) by 9 x 1OP AI a-chytnnt.rypsitt, ctplwetttl~ somewhat more rapidly than I-I:, in a rcactinn which appeared to stop brforc complete hydrolysis of ottc rnantiomcr, 43’;;., reac- tion. The reaction led to (+)-I-c’, [,]g’.’ +150”, and (-)-acid VIII, [(Y]y -276”, which cotttpares favorably with the reported (18) value, 282”, indicating high stereospecificity.

Dr.---I--C

CO&II3

(5)

R(+j-I--C S(-)-VI II

The absolute steric course of thaw hydrolyses n-as shown as follows. (a) cr-Renzylsuccinic acid, Compound II-D, was con- verted to diethyl a-henzylsuccinatc, II-C. (b) Compound II-C was hydrolyzed stereospecific:tlly by cu-chynotrypsitt, w ittdi- cated in Fig. lC, leading to K(+)-B-Ethel-cu-bettz~lsu~~itt~~t~ and to utlhydrolyzed S( -)-diethyl cY-brttz?.lsUCrittaIte, S( -)-II-C

(14). (c) Thr S( -)-diestw was hgdrolyzed by alkali to S( -)- cY-bettzylsuccitric acid, [a]E6 -26”, which was wetttially optirall~ l)ure (31). (d) The S( -)-cr-bt~ttz~lsitrcitti~ acid wits cyclizcd iti sulfuric acid to S( +)-3-c;lrbos2-(Y-tetr~tlotte, (‘otttlwuttd VI (IGlual iott 1). Thus tltr absolute cottfiguratiou of thr prndwt s formed by the rtlzymic hydrolysis of I)I,-1-E a11r1 DL-I-F is a~- rigtted, acid S( +)-VI and rsters R( -)-I-E aud I-P, and thrsc are indicatrd itt the rtructut-al formulas of Squat ion 3. (e) S( +)-3-(‘arbosy-cu-tetralnttr (VI) w:\s reduced by zitic and hydrochloric wid to H( +)-1 ,2,3,4-tetrah~dro-2-tt:tl)litlioi~ acid, establishitg the cottfiguratiotts of the Iwntluct> fortnrd by the ertzymic hydrolysis of nr,-I-G, arid R( +)-VII, aud e,stcr S( -)- T-G, and thwr at-r indicated itt I<yrt:tt.inn 4. (f) ‘I%> (+)-dihydro acid (VII I) had ~WJI hydrogrnoted (18) to the ( -)-trtrahydro acid (S( -)-VII), establislting the wttfiguratiotts of the lxoducts forttwd by the ettzymir hydrolysis of DL-T-C, acid S( -)-VI II, and R( +)-I-C:, as ittdiwted itt Kquatiott 5. This is itt agrctetrtcttt with the Iwevious (18) tentative ;trsigttrnrttt of tltcl :tl~solutc steric course’ of the hydrolysis of the (-)-P-ttitro1)ltt’tt!.1 cstct- cnrrrs~~ottdittp to 1-c’.

‘I’hr kinrtirs of hydrolysis of nr.-3-c:trhethos~-~-lc~trn1otte, I-F, hy cY-c~h?-motr~l)sitl in water-ethattol (9: l), 0.1 5 Sac’l, l)H 7.8, autl of S( -)-tttethyl 1 ,‘2-dilt~dro-2-tt:ll)htlto:tt(~, I-(‘, itt \~:IIor-tltcth:trtol (9: 1) was studied attd cntnpart~d with the hydrolysis, uttdtlr similar cnuditiotis, of home twttcyclizrd am- loguw: dicthyl cu.l,c~tlz~l?;ttccitt:~tt~ (IT-C), etltyl cr-tttelhyl-P- I)hett!ll)rol)iott:~t~~, and ethyl 8-l)ltett~l1)ro1)iorlnte. The wzytttic ltydroly& of nL-methyl I ,2,3,4-tc:trah?dro-2-nal,llttltoate was stereoselec~tive but. very slow, Tv/(I:‘) (S) -1 >I-’ see-’ iti vxtrr-

tncthattnl (9: 1) and a det.ailed kinetic attalysis was not carried out. ‘I’hc kittet.icr; of hydrolysis of the wry rwrtive cyclizctd substrates, u( -)-methyl 3,4-dilt?-droisocounl~~ritt-3-~~trbns~l:tte, 1-B (7, 8), and D( -)-I-keto-3-curbotttethos~tet rahydrnisn- quittolitte (I-.4) (2), previously studied in water, was rwsttminrd in ~~-ater-mrtlt:~~~nl (9: 1) for direct comparison with Cnnil~outtds I-P aud I-C, and also in ~~-oter-:lc,etottitrilr (!a: 1). The tt:~tur:tl substrate L( +)-methyl .~-:tc!et~l-P-l)hett~li~latlinate was also rc- csamitwd itt water-methanol and in \~~tter-ac!etnttiIrilt,. Some data are suntmarizcd in Tables I at1t1 II.

Thrsc substrates may be thought of as cu-suhstitutcd @- pli(~tt~l1)rol)iott~~t~s, aud the cyclizcd substratrs haw the (Y- substituettt cyclized ittto the phettyl ring. In all of thaw c’n~es, in which t.his second rittg is sis-tttctnhrrrd, the cyrlized substrate is ltydrolyzrd by cY-chytnotrypritt preferetttially itt the I) renrr, in t.he opposite absolute steric sense from that in which the anal- ogous noncyclized substrates arc hydrnlyzrd by thr ettzyte. For l-keto-3-c:irhnnictlio~~tetrah~droisoquitio~ittc (I-A) and for methyl 3,4-dih~droisocoun~aritt-3-~~tr~~o~~l~~te (I-13) thi.c h:ts bren shown by direct stereospecific sytthcsis of the more rapidly hy- drnlyzcd enantiomers from t)( +)-phcttylalanina (1, 2, 8) and n( +)-P-l)ltettvll:tctic acid (8), rrspect.ively. For the cstera of 3-carbosywtetralotte, I-E and I-F, t.his was show by formation of that enant iomer, which was more rapidly hydrolyzed 1)~ a!-chymotrypnitt, via cyclizatiott of that mantiontcr of c+brttzy1- succinic acid, II-J), which ww nbtaincd from the want inmer of diethyl a-bettz2lsuc:cirtate which was not hydrolyzed by cu-chymo- trypsin. This rttantiomer had been shown (14) to hc the S( -)- diester, and the enzymically more reactive forms of the tetmlnne derivatives are thus S(S) (Equation 3). (+)-I ,2,3,4-T&m-

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Issue of May 25, 1969 S. G. Cohen, A. MilocanoviC, R. iIf. Schultz, and S. Y. Weinstein 2671

hydra-2-naphthoic acid (VII) formed in the enzymir hydrolysis of methyl 1,2,3,4-tetrahgdro-2-naphthonte (I-G) was also formed by reduction of the product of enzynic hydrolysis of 3- cnrbethosv-cY-tt:tralone, S( +)-3-carbosy-cu-tetmlone, e&blish- ing that it.5 configuration is R(+) and that the hydrolyses of I-E and I-G proceeded in the same abso1ut.r sense (Equations 3 and 4). The relative wnfigurstions of the 1 ,2,3,4tetrahydro- and the I ,2, -dih~dro-2-nnphthoic acids had hccn previousl) correlat,ed by reduction of the dihydro- acid (18). The present work wtablished that the (-)-dihydro- acid, VIII, which is rc- lated to t.he (+)-tctmhydro- acid, VII, ia formed in the enzymic hydrolysis of the 8( -)-dihgdro- ester I-C, and thus hydrolysis proceeds in the same absolute sense as the other cyclized sub- strates (I+mation 5).

We propose then that the more reactive enantiomers of all of these cxclized substrates, Compounds l-A, I-R, I-c’, I-E, I-F, I-G, associated with the enzyme during hydrolysis as indicated iii Fig. 4. Iii t.his mode of association, the --S---I’-- grou1)ing fiti into h-w; the liartin structure IX, kq)t. rigid by thr. sccontl ring,

has the s:ur~e conformation as t.hnt of the natural noncyclized Substrate 11-A during its enzymic hydrolysis (Fig. 1). This places the phengl group at ur, and the ester group at the hydro- lytic site n. So use is made of the n?n site. Ilgdrolysis in t.he opposite, or L, sense of the less rapidly attacked cylizcd cnant.io- mers occurs via an association in mliiA the transposition of the -X-Y- and -CH,- groups may be indicated as in Fig. 5. In this figure the a-amide, cr.1 I, and @-CH2 are placed essent.inlly as in the binding of noncyclized natural r,substratw (Fig. I), hut the ring closure draws the aryl group from t.he ar site and de- creases reactivity. The most reactive orientation of the I. enantiomer need not be this clear-cut, and, in the absence of an amide group to associate at anz, the aryl group and the &CHz may be displaced toward nr (8).

The efficiency of hydrolysis in the D sense of the rwctire enantiomers reflects the case of fit of the ---X--I’- grouping into w-h and indicates some propertirs of this part of the site. When

0 ‘I ;7

--)i’.-I’- is amide, -XII-C-, D( -)-I-A, or ester, q-C--, I)( -)-I-R, the efficicnc~- of hydrolysis is high, the values of Iz,,,J K, in water being for I)( -)-I-A, 4.1 x 10’ Y-’ w-i, and for D( -)-I-B, 6.0 x 10’ Y-I WC-~ (Table I), similar lo that of the natural noncyclized substrate L(+)-II-h, 6.2 x 10’ h1-I WC-I (Table II). When -S-Y is vinyl, HC-CH, S(-)-I-(‘, the efhcieyv of hydrolysis is also high. This compound was not soluble enough in water and was st.udicd in 10% mrthanol, ~c.tlKn ~900. Comparison with the preceding compounds under t.hc same conditions (Table I) indicated that the value of k rat/K,n for 1-c: was about one-third that of the cyclized sub- strates, I-h and I-13, and one-tent.h that of the derivative of 1)t1(:llyl:ll;lllill1:, II-.-\. The high reactivity of I-C‘ nppcared to he due largely to a favorable K, which compensated for an unfsvor- able kc,,,. We were unable t.o obtain this compound in a high slate of purity. The true reactivity may be somewhat greater

than indicated and we may not comment in detail on this com- pound.

0

II I lowever, when -X-Y- is --Ct Ix-C--, in DL-S-carbethosy-

cu-tetralone, I-F, the reactivity is greatly decreased, kcut/Km = 9 M+ SW-* in 1072 ethanol, less than 1% as reactive as I-A, I-R, I-C, and II-X. When - -X--- l’- is -CH*--CH2-, in I-G, the reactivity is reduced further by an order of magnitude, Y/(E) (S) -1 Jr-l see-*. The arnidr, e&r, and vinyl groups of I-.A\, I-R, and 1-C fit well into the ar-h slit, either because of their size or their relative planarity or because of a polar interaction with an electrophile in the enzyme at that posit.ion. h methylene substitucnt on the carbon (Y t.o t.he wter in I-F prevents good fit at. h, and a second n~etl~ylene in I-G prevents it still more, either because of size or nonpolarity. These results with the cyclized substrates are consistent with the behavior of related noncyclized compounds. Enzymic hydrolysis of the D enantiomers of wters of ar-chtoro- (17) and Lu-tl~dros~-P-phell~l1)ropionic acids (13, 17) indicates that a small polar a-substituent may fit at h (Fig: 3); the singly bonded heteroutonx, nitrogen and oxygen, in I-A and 1-R do so similarly. The faihwe of the D ennnt.iomcr of ethyl a-methyl-~-phellyl1~ro1~iollate Lo be hydrolyzed by cr-chymo- trypsin (14) indicated that, the methyl group may not fit at h and t.his is consistent with the very low reactivity of I-F and I-G. This very marked difference between the effect, of a methylene group and of a heteroutom attached to the cr-carbon is also SWII in the compounds with five-membered fused hydroaromatic rings. Methyl illdane-2-carboxll~ltc is not a good substrate while the oxygen analoguc, met.hyl I,-hydrocorlrnarilate, I-L), is a good substrate for wchymotrypsin (19). This co~npound, too, associates as in Fig. 4, with the, het~roatom in the ar-h site.

The mode of association of Fig. 4 and the greater reactivity in the D sense result bwause the -.li-Y--- grouping fits into W-IL better than does the -CIIn- of the hydroaromutic ring. When the --X---Y- grouping fits in very well, great reactivit.y results and is accompanied by high stereospecificity, ahhough not as high as in the noncyclized substrates (t(), where the large wsubst,itu- ents simply cannot fit into h. The stereospecificity of the reactive substrates may be indicated by the ratio of k,,,/K, values for the I) and I, enantiomers (8). These ratios arc D-I-A:I,-I->~ = 4100, D-I-13 : ~-1-13 = 900. IVhrn the --X-Y- grouping dota not fit well, as in I-F and I-G, react.ivity dccre:hses, and the dif- ference in t.he ease of placing the ---y-I’- and --Cl Ir- groups in ar-h also dccreasw, leading to diminished stcreospccificity.

FIG. 5. hsoriation of I, en.zrltiomers of cyclizc!d slhstrntc!s

with n-ctl~rrlotr~psiil. S--I- as in Fig. 1.

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2672 On the Active Site of a-Chymotrypsin Vol. 244, 30. 10

03 Q

/+ NY.. / -.

\ N’ 4 -. ,I / ,, ,y

‘.

I//

“X,

,//I P 4’ I

I

B H

‘I \ ,’ / // /’ p

l!!C02R

\ \/Q,/' '2

a

a

H \ 4

FIG. 6. Combined perspective scheme of u-1, L-II, and L-X. Broken line vorlion is directed toward or-h. P. nhenyl rinr re- sponsible fo; primary binding at ar in the three system-s. Q,-sec- ond phenyl ring in I.-X, absent in D-1 and I.-II. A--B, NH-CO in L-II and L-X, in am site; A-13 absent in I. X, hydrogen in L-II and L-X. Xv.. Y, NH-CO, O-CO, and CH=CH, in n-I-A, n-I-B, and D-I-C, respectively. There is no X-Y bond in L-X; Y = C in L-X; Y is absent in L-II.

Comparison of the specific rotation, +33.5”, of the acid S(+)- VI obtained in enzymic hydrolysis of the DMetralone derivative, I-F, with the estimated minimum value for the rotation, 37”, indicates that S(+)-T-F hydrolyzed only about 20 times as rapidly as R( -)-I-F from the RS material under our experimen- tal conditions. Similarly, the specific rotation, +38”, of the acid R(+)-VII obtained in the hydrolysis of the DGtet.ralin derivative, I-G, compared with the reported (27) rotation, 51.8”, indicates that R( +)-I-G was hydrolyzed only about 6 times as rapidly ss S( -)-I-G from the RS mixture in our experiment. The slightly more effective association directs the -CIIz---CH- group rather than the -CH,-- group toward ur-h, although the latter is smaller and might fit better. The more effective asso- ciation more nearly retains the proper location and conformation of the partial structure IX. The other orientation may in effect intercalate three carbon atoms between the phenyl and the ester group and lead to less efficient hydrolysis, much as glutarate de- rivatives are less reactive than succinates (33).

Some brief comments may be made about the data in Tables I and II. The cyclized and noncyclized amide subst.rates D( -)- I-A and L(+)-11-A have similar over-all reactivit.y; the cyclized compound has a somewhat more favorable K,, less favorable It cat. Both are affected similarly by change in solvent; the re- activity falls by an order of magnitude in 10% methanol as com- pared with water, with somewhat more of the decrease due to change in K,, less due to decrease in 12,,,. Acetonitrile, lo%, has only a minor effect in both cases, largely on K,,,. The cyclized ester substrate, D( -)-I-B, shows a 25-fold decrease in reactivity in 10% methanol, largely due to less favorable K,. Again acetonitrile IpAds to a small decrease in reactivity, due to its effect on K,.

The a-hydrocarbon substituents, methyl in ethyl a-methyl &phenylpropionatc (V-C) and carbethoxymethylene in diethyl a-benzylsuccinate (II-C), do not fit into IL and they occupy am; they lead to high L specificity, but with very low reactivity. They occupy am but do not bind there by hydrogen bonding, as the a-acylamido group does in Compound II-A. The cu-alkyl substituents do not restrict the substrate, and the reactivity is

low and comparable to that of the unsubstituted parent con-

pound, ethyl @phenylpropionate (II-E). The kinetic data for II-C and V-C are derived from study of the DL materials. The correct values of k,,,/K, for t.he L enantiomers, which are under- going hydrolysis, would be something more than twice as great as those listed.

Some inferences may be drawn from t.he proposed n~odcs of association of cyclized substrates which suggest further esperi- men&. Cyclized D substrates, associating with the enzyme as in Fig. 4, leave the am site unoccupied, and t,his site should he able to accommodate a fourth substituent instead of hydrogen on the a-carbon. Replacement by a methyl group of the ar-hydro- gen of noncyclized natural substrates, a-N-ttc:ylamino-P-aryl- propionate esters, decreases substrate activity by a factor of IO5 (34). On the other hand, we have found that when the cr-hydro- gen of the cyclizcd substrate I-B is replaced by a methyl group the compound remains an active substrate,3 supporting association as described in Fig. 4.

A cyclized La-amide substrate, associating with the enzyme as in Fig. 5 has decreased reactivity, as utilization of the am siie diminishes association at ar, and this may decrease the activating reorganization of the fine structure of enzyme (8, 9). If the a-amino group should be cyclized into the P-aryl group by a larger number of atoms than in the six-membered ring, then this larger grouping might not fit into wh, preventing hydrolysis of the D enantiomer. The larger ring might, however, allow the cu-acylamino group to associate at. urn without withdrawing t.he p-aryl group from t.he ar site, allowing hydrolysis of the L enantio- mer. This appears to be the case with a recent.ly reported (35) 2,2’-bridged biphenyl analogue of methyl benzoylphenylalani- nate, X, in which the second ring is eight-membered and leads to

c-l 0 /-d--Lo IfiT i

- --%OCH 2 3 X

high reactivity in the L sense. We find unconvincing the infer- ence of these authors (35) that the high reactivity of this com- pound strongly supports the hypothesis that the carbalkoxyl group in the reacting conformation of D( -)-I-A is axial. Models indicate that this molecule also may have i& carbalkoxyl group in an equatorial conformat.ion, and have the appropriate part, IX, in an orientation quite similar to those of L(+)-II-A and D( -)- I-A, as indicated in Fig. 6.

High reactivity of this bridged hiphengl system may indicate that the acylamino group in a noncyclized, substrate, II-A, is itself directed toward the /3-aryl group. The bridged biphenyl cu-acylamido substrate has a partially constrained conformation; it occupies the am site and may require the hydrogen-bonding properties of the site for further restraint and for its reactiviby. It is not clear whether its ester analogue might also be reactive. Also, replacement by a methyl group of the a-hydrogen in this kind of cyclized substrate, which utilizes the urn site, might lead to a great decrease in reactivity, similar to that observed in non- cyclized substrates.

3 1,. W. Lo and S. G. Cohen, unpublished results.

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Issue of May 25, 1969 S. G. Cohen, A. JlilocanoviC, R. 41. Schultz, and S. Y. Weinstein 2673

On Ihe rlctiue &U-The inferences which we draw about the topography of the active site may be considered in detail and recapitulated. The parts of the site, ar, h, am, and 12, are comple- mentary to the parts of the natural small molecule model sub- strate L( +)-II-A, in the indicated conformation (Fig. 1).

Part UT is a fold into the enzyme at which primary binding occurs. This is best provided by a @-aryl substituent, which may be mononuclear, as in derivatives of phenylnlanine and tyrosine, or binuclear, as in tryptophan. This indicates that aromatic and planar two-ring nuclei may slide completely into t.he ar fold. The CH2 group of dihydro- or tetrahydro- binuclenr aromatic systems, D( -)-I-A, D( -)-I-B, s( -)-I-C, acting like the &CH2 group of L( +)-II-A, allows only one ring to fit into ar and the hydrolyzing carbalkoxyl group to find its hydrolytic site at n. In the corresponding aromatic or dchydro systems, esters 3-carboxyisocarbostyril (1, 2), 3-carboxyisocoumarin$ and 2- naphthoic acid (18) show essentially no react.ivity toward LY- chymotrypsin. In each case the binuclear moiety presumably may fit fully into ar, and the ester group is not placed at 72. It. is not clear whether larger aromatic ring systems may fit into ur.

The p-C& group is at the entrance to the ar fold. Its L sense hydrogen, H, (Fig. l), is directed into a restricted domain of the enzyme and an L-P-substituent prevents reaction. The D sense hydrogen of the &carbon is directed toward the enzyme surface and a D-@substituent allows enzymic reaction, leading to D-/I?- specificity (36, 37).

The a-carbon is placed so as to deliver the group to be hydro- lyzed to the imidazole-serine locus, 7~. In L(+)-II-A the (Y- carbon is located by the binding of the &phcnyl group at ar nrd of the L-cr-acetamide group at am. These forces rest.rict rota- tions about C,--C,q and C&--Cphenyl and fix the a-carbon in position. Of the other ccY-substituents examined, the hydroxyl group provides moderately high reactivity (13), k&K, -1500 ~-1 set-1 in 20% ethanol, presumably by donating a hydrogen bond and leading to some restriction. The acetoxyl group leads to low reactivity, k&Km -26 in 207, ethanol. cr-hlkyl sub- stituents, methyl and carbalkoxymethylene, also lead to low re- activity (Table II). They may fit at am but do not bind there and do not restrict the substrate. The a-acetoxy and cy-alkyl substituents lead to L specificity, as does the cu-acetamido group, because they are too large to fit at the restricted site h, as would be required for the D enuntiomers.

In cyclized substrates, restriction of the conformation of the substrate is provided by the second ring (7, 8). A conformation and a complementary relation to the active site very similar to that of L( +)-II-A in Fig. 1 are provided when the cgclization is completed from the I) configuration of t.he a-carbon (Fig. 4). Binding at am is not required for high reactivity and any activat- ing reorganization of the enzyme that may occur would result from interaction at ar. Cyclizing groups X--Y, which can fit well into ar-h, lead to high reactivity. These groups are ester, amide, and vinyl, indicating that the ar-h site is restricted and may have electrophilic properties.

. Cgchzatlon via an Q(-

aliphatic group leads to low reactivity. The amide of a large cyclizing group, like the four-membered one in the eight-mem- bered ring of the L-bridged-biphenyl derivative (35), X, may associate at am and bc fused into the P-phenyl group without drawing it from the ar site.

‘It. hi. Schultz and 6. G. Cohen, unpnblished reslllts.

The cr-acetamido group and the hydrolyzing group of L(+)- II-A represent the peptide chains of a polymeric substrate for the endopeptidase and lie at the surface of the enzyme, as do their complementary sites am and n. The acetamido group, like the a-hydrogen, is cisoid or gauche to t.he /3-aryl group. The aryl group and the a-hydrogen are directed to the fold of the enzyme. The hydrolyzing carbalkoxyl group appears to be transoid to the aryl group, and directed away from it. Evidence for this was seen in the reactivity of diethglfumarate and the inertness of diethgl maleatc (5). This may also support the view that the cstcr group in cyclized substrates is equatorial, although the axial conformation has its supporters.5 The angular and linear dimensions of our preferred conformat.ion have been reported re- cently (8).

1.

2.

3. 4. 5.

6.

7.

8.

9.

10.

11.

12. 13.

14.

15.

HEIN, G. E., MCGRIFF, It. B., MD XIEYANN, C., J. Amer. Chem. Sot., 82, 1830 (1960).

HEIN, G. E., .~XD NIEM,\NN, C., J. Amer. Chem. Sot., 84,4487, 4495 (1962).

*JONES, J. B., .~NI) NIEMIASN, C., Biochemistry, 2, 498 (1963). HARTLEY. 1). S.. Brookhaven Swrw. Biol.. 16.85 (1962). COHEN, d. G., KLEE, I,. II., AN; WEIN&TE;N, s‘. Y.; J. Amer.

Chem. Sot., 88, 5302 (1966). COHES, S. G., NEUF~IRTH, Z., AND WEISSTEIN, S. Y., J. Amer.

Chem. Sot., 38, 5306 (1966). COHEN, S. G., AND SCHULTZ, It. M., Proc. Nal. Acad. Sci.

U. S. A., 67, 243 (1967). COHEN, S. G., ‘\ND SCH~IY~Z, H. M., J. Biol. Chem., 243, 2607

(1968). WOOTTON, J. F., .~NI) HESS, G. P., J. Amer. Chem. Sot., 84, 440

(1962). SCHAFFER, N. K., MAY, S. C., JR., ASD SUUUEIZSON, W. H.,

J. Biol. Chem., 202,67 (1953); J. Biol. Chem., 208,201 (1954). KOSHL~IND, I). E., JR., STRUMEYER, D. H., AND RAY, W. J., JR.,

Brookhaven Symp. Biol., 16. 101 (1962). NEURATH, II., AND SCH~VERT, G., Chem. Rev., 46,69 (1950). COAES, S. G., AND WEIXSTEIN, 8. Y., J. Amer. Chem. Sot., 88,

5326 (1964). COHEX, S. G., AND MILOV~~Z~OVI~, A., J. Amer. Chem. Sot., 80,

3495 (1968). COHES, S. G., CROSSLEY, J., KHEDOUHI, E., ANO ZAND, H., J.

Amer. Chn. Sot., 84.4163 (1962); COIIEN, S. G., CROSSLEY, J.. KIIBDOURI. E.. Zas~. I~.. ASD KLEE. 1,. H.. J. Amer.

16.

17.

18.

19. 20. 21.

I I I I

Chem. Sot., 85, lG8.5 (1963). RAPP, J. H., AND XIEMANN, C., J. Amer. Chem. Sot., 85, 1896

(1963). SNOKE, J. E., AND NE~R~TII, II., Arch. Biochem. Biophys., 21,

351 (1949). SILVEH, 11. S., AND SOSE, T., J. -4mer. Chem. Sot., 89, 457

(1967). L.\wsos, W. B., .I. Biol. Chem., 242, 3397 (1967). WEIZY.\S, A., J. Org. Chem., 8, 285 (1943). hfa~su~a, S., YAMAUCHI, T., ,~SD TORI, K., J. Chem. Sot.

Jap. Znd. Cha. Sect., 68, 60 (1955); Chem. Abstr., 50, 3308 (1956).

22.

23.

24.

Awwoo~, A. S.. STEVENSON. A., .~NI) TIIORPE, J. F., J. Chem. Sot., 1?64 (19i3).

H~RII, Z., MOMOSE, T., AND TAMURA, Y., Chem. Pharm. Bull. (Japa7c), 10, 946 (1962); Chem. Abstr., 69, 3850 (1963).

Vas TAMELES, E. E., MCNAI~Y, J., ASD LOIMTZ, F. A., J. Amer. Chem. SOL, 79, 1235 (1957).

25. SERGIEVSKAYA. S. I.. AND VOLYXSKII. N. P.. Zh. Obshch.

REFERENCES

Khim., 22, 3il (195i). 26. ~EWHAS, M. S., ASD M~SGHAM, J. R., J. Amer. Chem. Sot.,

71, 3342 (1949).

6 Discussion of the axial-equatorial question and references are given in Reference 8.

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

2674 On the Active Site of a-Chymotrypsin Vol. 244, No. 10

27. PICKARD, R. H., AND YATES, J., J. Chem. Sm., 89, 1101 (1906). 33. COHEN, S. G., AND CROSSLEY, J., J. Amer. Chem. SOL, 86, 2% PICKARD, R. H., AND YATES, J., J. Chem. Sot., 96, 1011 (1909). 4999 (1964). 29. DEIUCK, C. G., AND KAMM, O., J. Amer. Chem. Sot., 36, 388 34. ALMOND, H. R., JR., MANNIHG, D. T., AND NIEMANN, C., Bio-

(1917). chemistry, 1. 243 (1962).

30. MOORE, J. A., AND REED, D. E., Org. Syn., 41, 16 (1961). 35. BELLEAU, S., ASII CIIEVALIER, R., J. Amer. Chem. Sot., 60,

31. FREI)G‘\, A., Ark. Mineral Geol., 26, No. 11 (1948); Chem. 6804 (1968).

Absl~., 43, 1747 (1949). 36. COIIEN, S. G., AND WEINSTEIN, S. Y., J. Amer. Chem. Sot., 86,

32. ;\lAR’rIS, E. L., Organic reactions, Vol. I, John Wiley and Sons, 725 (1964).

37. COHEN, S. G., SCHULTZ, K. M., AXI) WEISS’I’EIN, S. Y., .J. Amer. Inc., New York, 1942, p. 155. Chem. Sot., 88, 5315 (1966).

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from

WeinsteinSaul G. Cohen, Aleksander Milovanovic, Richard M. Schultz and Sandra Y.

SUBSTRATESAND KINETICS OF HYDROLYSIS OF CYCLIZED AND NONCYCLIZED

-Chymotrypsin: ABSOLUTE CONFIGURATIONSαOn the Active Site of

1969, 244:2664-2674.J. Biol. Chem. 

  http://www.jbc.org/content/244/10/2664Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/244/10/2664.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 7, 2018

http://ww

w.jbc.org/

Dow

nloaded from