IRREVERSIBLE ENZYME INHIBITORS. CLXXII Journal of Medicinal Chemistry, 1870, Vol. 13, No. 3 467

with three IO-ml portions of 0.5 N NaOH, once with 10 ml of brine, arid then dried (MgS04). Spin evaporation in vacuo left a syrup which showed no uv change from pH 1 to 13.

described above) in 25 ml of absolute EtOH arid refluxed with stirring for 4 hr. The cooled solution was acidified to pH 5

with AcOH, then spin evaporated in vacuo. The residue was triturated with several milliliters of i-PrOH, cooled, the product collected and then washed with i-PrOH; yield, 0.680 g (787,),

This syrup was treated with 3.5 mmol of NaOEt (formed as mp 146149". Recrystallization from i-PrOH-EtOH gave clones of white pins, mp 153-1514'. Anal. (CMHI~NZO~) C , H, N.

Irreversible Enzyme Inhibitors. CLXXII.'v2 Proteolytic Enzymes. XVI.3 Covalent Bonding of the Sulfonyl Fluoride Group to Serine Outside

the Active Site of a-Chymotrypsin by exo-Type Active-Site Directed Irreversible Inhibitors

R. c .4RDINAVD4 AXD B. R. BAKER

Dcpartinent of Chemistry, University of California at Santa Barbara, Santa Barbara, California 93106

Received December 5, 1969

Three active-site directed irreversible inhibitors with a teiminal SOIF group (1-3) of a-chymotrypsin have been shown to covalently link a serine OH group by two independent methods. The sulfonate group on a- chymotrypsin in each case was displaced by mercaptoethylamine and resultant 8-aminoethyl->cysteine identi- fied after hydrolysis; the second method converted each sulfonate of serine into pyruvic acid which was identified enzymatically and by conversion into its 2,4-dinitrophenylhydrazorie. Reaction of 2 with a-chymotrypsin was unequivocally shown not to occur on the active-site serine-195, but a serine outside the site tentatively identified as serine-223; these results establish the hypothesis that properly designed active-site directed irreversible inhibitors can covalently link an amino acid residue outside the active site by the e m mechanism.

Active-site directed irreversible inhibitors of enzymes operate by two 5teps.j A reversible complex is first formed between t,he enzyme and inhibitor; if a proper leaving group on the inhibitor is juxtaposed to a nu- cleophilic group on the enzyme, then a rapid and selective neighboring group reaction occurs with forma- tion of a covalent, bond that inactivates the enzyme.j There are two classes of active-site directed irreversible inhibtors, the endo type that forms a covalent bond inside the active-site and the ex0 type that forms a covalent bond outside the active-site.j

The endo type of irreversible inhibit'or is of int'erest in protein structure sbudies for "labeling" amino acids inside the active-site; a now classical example is l-chloro-4-phenyl-3-tosylamido-2-butanone (TPCIQ6 which specifically forms a covalent bond with histidine- 57 in the active site of a-~hymotrypsin.~

The ex0 type of irreversible inhibitor has a consider- ably wider utility in drug design than the endo type.8 The best leaving group yet foundg for the exo type of covalent bond is the F of the SOzF m0iet.y; such a moiety has the electrophilic capacity to react rapidly with a serine OH of an enzyme,I0 but direct chemical proof of covalent bond formation with a serine out'side the active site had yet to be achieved. The SOzF

(1) This work was supported in part by Grant CA-08695 from the Na- tional Cancer Institute, U. S. Public Health Service.

(2) For the previous paper in this series see B. R. Baker, and J. L. Kelley, J. M e d . Chem., 13, 461 (1970).

(3) For the previous paper on proteolytic enzymes see B. R. Baker and .\.I. Cory, ih id . , 12, 1053 (1969).

(4) Visiting research professor from the Department of Biology, Atomic Energy Commission, France, 1968-1969.

( 5 ) B. R. Baker, "Design of Active-SiteDirected Irreversible Enzyme Inhibitors," John Wiley & Sons, New York, N. Y., 1967, pp 17-21.

(6) G. Scboellmann and E. Shaw, Biochemistry, 2, 252 (1963). (7) E. B. Ong, E. Shaw, and G. Schoellmann, J. Biol. Chem., 240, 694

(1965). (8) B. R. Baker, Biochern. Pharmacol., 11, 1155 (1962). (9) B. R. Baker and G. J. Lourens, J . hfed. Chem., 10, 1113 (19671, paper

(10) For discussion of tile Chemistry of the SOzF group, see ref Y. CV of this series.

moiety properly positioned on an appropriate revers- ible inhibitor could inactivate a variety of enzymes, presumably by the ex0 type of active-site directed irreversible inhibition. Examples are dihydrofolic re- ductaselg xanthine oxidase," guanine deaminase,12 tryp- sin,I3 a -~hymot ryps in , '~ -~~ the C' la component of com- plement, l7 and cytosine nucleoside deaminase.l8 With dihydrofolic reductase, SOzF-type inhibitors have been found that can inactivate an L1210 mouse leukemia enzyme with no appreciable inactivation of the enzyme in normal liver, spleen, and intestine of the mouse. l9

Three successive questions can be asked about an active-site directed irreversible enzyme inhibitor of the SOzF type that presumably operates by the ex0 mechanism. (1) Does the SOzF form a covalent bond with a serine? (2) If the enzyme has a serine in the active-site, as in the case of a-chymotrypsin, has the covalent bond formed with the active-site serine (endo) or has the covalent bond been formed with a serine outside the active site (ezo)? (3) If the covalently

(11) (a) l3. R. Baker and W. F. Wood. J. M e d . Chem., 11, 660 (1968), paper CXXIII of this series: (b) B. R. Baker and J. A. Kosma. i h i d . . 11, 652 (1968), paper CXXIV of this series: (e) B. R. Baker and J. A . Kosma, ibid., 11, 656 (1968). paper CXXV of this series: (d) B. R. Baker and W. F Wood, ibid. , 12, 211 (19691, paper CSLVI of this series.

(12) B. R . Baker and W. F. Wood, ihid. , 12, 214 (1969), paper CXLVII of this series.

(13) (a) B. R. Baker and E. H. Erickson, ibid.. 11, 245 (1968), paper CSV of this series: (b) B. R. Baker and E. H. Erickson, ih id . , 12, 112 (19691, paper CXLIV of this series.

(14) B. R. Baker and J. A. Hurlbut, ih id . , 11, 233 (1968), paper CXIII of this series.

(15) B. R. Baker and J. A. Hurlbut, ihid., 12, 118 (1969), paper CXLV of this series.

(16) B. R. Baker and J. A. Hurlbut, i h i d . , 12, 221 (1969), paper CL of this series.

(17) B. R. Baker and J. A. Hurlbut, i h i d . , 12, 902 (1969), paper CLXI of this series.

(18) B. R. Baker and J. L. Kelley, i h id . , 12, 1046 (1969), paper CLXIII of this series.

(19) (a) B. R. Baker, G. J. Lourens, R. B. Meyer, Jr., and N. bf. J. Vermeulen, i b i d . , 12, 67 (196Y), paper CXXXIII of this series; (b) for a recent review see B . R. Baker, Accounts Chem. Res. , 2, 120 (1069).

IRREVERSIBLE ENZYME INHIBITOR^. CLXXII Journal of Medicinal Chemistry, 1970, Vol. 18, No. 3 469

dissolved in 2 ml of H2O and lyophilized with NaOH in the re- ceiver; this operatmion was repeated once more. This treatmerit cleaves the protein chain a t the carboxyl side of methionine and the amino side of S-arninoeth3.l-t-L'J.steine.

The dry residue was dissolved in 2 ml of H20, transferred to a hydrolysis bulh, theu evaporated in uacuo, and dried over PIO: in uucuo for 48 hr. The residue was dissolved iu 0.50 ml of aiihy- drous N2H4, the bulb was sealed under vacuum, then placed in a 80" oven for 24 hr. The residue was dissolved in 2 ml of H2O and transferred to a 3 X 0.9 cm column of Biorex-70 ( H + form). The amino acid hydrazides were absorbed and the amino acids were eluted with 15 ml of HgO.30 The eluate was lyophilized, then the residue was uis- solved in 2 ml of pH 2.2 citrate buffer and the ratio of amino acids determined on an amino acid analyzer. Hydrazinolysis converts all amide linkages t,o hydrazides leaving t,he carboxy terminal amino acid underivatized.al

The aolutioii was evaporated in uucuo.

Results and Discussion

a-Chymotrj.psiri was inactivated 97-100% wlien treated with any one of the inhibitors, 1-5. Activity mas not restored when the inactivated chymotrypsin was exhaustively dialyzed, thus showing that covalent bond formation had occurred; an a-chymotrypsin control under the same conditions still maintained most of its activity.

That covalent bond formation with 1-4 occurred with a serine residue as established by two methods. The first method was that of Gold22 used for a-chymo- trypsin inactivated with a-tolysulfonyl fluoride (5). The sulfonate esters (7) formed between 1 and 4 and a-chymotrypsin (6) were displaced to 8 with mercapto- ethylamine and the S-aminoethyl-L-cysteine (9) re-

CH2OR

T-~BLE I

MODIFIED CY-CHYMOTRYPINS YIELD O F S-AMINOETHYL-L-CYSTEINE FROM

Mo1/14 mol Mo1/2 mol Inhibitor oi Lys of His

4 0.66 0.65 4 0.61 0.69 1 O..i5 0.57 1 0.53 0.53 2 0.37 0.41 3 0.43 0.47

tacked the active-site serine-195, based on analogy with diisopropyl phosphorofluoridate (DFP) and the fact that the enzyme was inactivated; although 5 was shown to attack a serine,22 i t was not proven to be serine- 10.5. 1,ater it was shown by X-ray studies that 4 did indeed form a covalent bond with serine-195. 34

The second method for establishing that serine was attacked by 1-4 was conversion of the modified serine residue 7 into pyruvic acid (11) by the method of Koshland et al. ;23,26 this reaction proceeds via an an- hydrochymotrypsin 10 formed by treatment of 7 with HO-.

Pyruvic acid was determined enzymatically with lactic dehydrogenase. A standard curve was estab- lished by acid hydrolysis of 25 mg of unmodified chymo- trypsin containing varying amounts of added pyruvic acid; the recovery of pyruvic acid was 80-86x. The results of analysis for pyruvic acid of the serine modified a-chymotrypsins are presented in Table 11; no a-

C H ~ S C H ~ C H ~ N H Z ---f I Peptidel-NHCHCOXH-Peptide2

1 Peptidel-XHCHCOIVH-Peptide2

6 , R = H 8 7. R = SOzR'

I

.1 -OH J HaO+

CHz C H ~ S C H ~ C H ~ X H Z il I

Pep tidel-XHCCONH-PeptideZ 10

CCHa d CzHjCOCOzH CHaCOCO2H I /

Peptidei-NHCCONH-Peptide2 11

sultirig after hydrolysis was identified arid quantified with an amino acid analyzer. Results are shown in Table I.

The yields of S-aminoethyl-L-cysteine were deter- mined by peak areas compared to either the peak areas of 2 histidines or 14 lysines present in a-chymo- trypsin;32 the yields were 0.4-0.7 mol/mol of a-chymo- trypsin for 1-4, showing in each case a serine residue had been covalently linked as p r e d i ~ t e d . ~ . ' ~

It was assumed by earlier workers that p-tolylsulfonyl fluoride ( 5 ) 3 3 and a-tolylsulfonyl fluoride (4)20,z2 at-

(30) P. De La Llosa, C. Tertrin, and M. Jutise, Ezperientia, 20, 204 (1964).

( 3 1 ) H. Fraenkel-Conrat and C. M. Tsung, Methods Enzgmol. , 11, 151- 155 (1967).

(32) B. S. Hartley, Nature, 201, 1284 (1964); B. S. Hartley and D. 1,.

(33) D. H. Strumeyer, \V. N. White, and D. E. Koshland, Jr., Proe. "Vat. Kauffman, Biochem. J. 101, 229 (1966).

d c u d . S c i . C. S., SO, 931 (1963).

12 13

ketobutyrate (13) could be detected, which would be formed via 12 if 1-4 had attacked a threonine residue in a-chymotrypsin.

Unmodified a-chymotrypsin gave about 0.6 mol of pyruvic acid (Table 11) ; thus the net yield of pyruvic acid from a-chymotrypsin modified by 1-4 was 0.4- 0.S mol/mol of a-chymotrypsin. Similarly, the spec- trophotometric method used for the 2,4-dinitrophenyl- hydrazone of pyruvic acid showed a net yield of 0.4- 0.S mol/mol of a-chymotrypsin, although the method does not differentiate pyruvate from a-ketobutyrate; that the latter was absent ( < l O ~ o ) was shown by tlc.

a-Chymotrypsin modified with either a-tolylsulfonyl fluoride (5) or 2 were selected for further study to determine which serine of the 26 present in a-chymo- trypsin was covalently linked. The reaction sequence

(34) D. M. BIov, Biochern. J . . 112, 261 (1969).

470 Journal of Xedzcinal ('hcrriistr,y, 1970, Vo/, 13, N o . .3

Pyruvic acid (40 pg) added I)ef(~re :wid liyti~iilyiis. 'J U- Ket,obutyric acid (40 pg) added before acid hydrolysis. The spectrophotometric method does Iwt distitiguish hetwveru a-keto- butyrate and pyruvate.

for selective cleavagt, of tlic peptide :imino group of S-amirioetliyl-L-c~steiiie (8) procetded through 14-16 to generate u new c:trboxxrl tcrmiiiu-, (17). Tlir li -Y HC1 at 2.5" required for the \tepa 15-17 : d ~ o givc- qonie rxtraneous cleavagc of tlir protein. p:wticularl\ at gly cinr. :L rid glut a mat e re-i due i ; t 111 - ' * hac kgrou rid Iioiye'' in:tbt% iiiterprtTutioii niow difficult :tiid w d t s I w rx:ict. Thc rtwlth from thcx chromatogram coming off t l ic umiiio ncid :iii:dj w r could ht> quaii- titated h j cornpariwii with tlicl p : i k : t r w of Jmi-13, 1 I i c cnrboxj 1 termiiius oi tlir. ch t i i r : the p d a rw of lJeu-l:3 \$:is :trhitrarily stit a t 1 .OO : t i id otlit,l' ('- tcrrniiial :Lniiiio acids arc~ relativt. t o 1 mol of l t w k i t x mol of a-chymotrypsin, a ' ~ recorded iii T:ihle 111.

Tlir moyt e a d y understundable form for iiittqjreta- tiori of the results i-, compariwii of tlicl ratio- iii tlir two right hand columns of Table 111. It is c l e w that a-tolylsulfonyl fluoride (5) modified a-chymo- trypsin gcrierates X\p a? :i i i m - c:irhox\ 1 tcrmiriui in the sequence 14-17, thi5 rr-ult could ariye if Ser-19.5 liad tieen cov:tlt~ritIy lirihed by 5, +iiicc. Ser-1% i- the onlj ieriiie linked on it, :mino iide to A l b y .

The results with a-chymotrj p i l i modified ~ I J 2 clearly bhon- the generation of much l e s carbox) 1 ter- minal aqpartic acid than does a-rhymotrypin modified h> a-to13 ld fo i iy l fluoridc (5). Therefor(> it i* clear that 2 h:~* iiot reactctl n i t h the :ictivc-.ite SCT-195; thi- result coi~firms the hypoth t h t ari actwe-szte r l i - i eetetl u i l i i b i f o / su(h C I S 2 o p e ) atr s by the C.I o MY iiaiiis71r, / l i d is, 2jo i )ns a (orwle i i t Ii)tXac/c'oittoiclc t h e a t f u c a r f c