[16]
SERINE PROTEASES WITH HALOMETHYL KETONES
197
[16] Reaction of Serine Proteases with Halomethyl Ketones By
JAMES C. POWERS
The use of the chloromethyl ketones Tos-PheCH.~C1 ( T P C K ) and Tos-LysCH2C1 ( T L C K ) as specific reagents for chymotrypsin and trypsin, respectively, is one of the classic demonstrations of the value of affinity labels for enzyme studies? Subsequently, halomethyl ketones were used for characterization of functional groups and sequences in active sites, as probes of structure by providing labeled proteins that contain spectroscopic handles, in crystallographic studies of substrate binding sites, and in the study of the biological function of proteases. Today, halomethyl ketones are probably the most widely studied class of affinity label and have been discussed in numerous reviews. 2-6 The reaction of a serine protease with most substrate-related haloketones probably first involves the formation of an enzyme inhibitor complex (Scheme l) in which the inhibitor is recognized by specific interactions between the side chain of the P1 amino acid residue and the $1 or primary substrate binding site of the enzyme (nomenclature of Schecter and B u r g e r ) . Irreversible inhibition takes place within this complex by covalent bond formation between the active-site histidine residue and a methylene group of the inhibitor. In addition, crystallographic studies in a few cases indicate that the active-site serine oxygen has added to the carbonyl group of the inhibitor to give a tetrahedral
The nomenclature used in this chapter for amino acid and peptide derivatives Conforms to the 1971 Recommendations of the IUPAC-IUB Commission on Biochemical Nomenclature, Biochemistry 11, 1726 (1972). Thus the chloromethyl ketone derived from tosyl-L-phenylalanine will be abbreviated as Tos-PheCH2C1 instead of the more commonly used TPCK. The author recommends the use of systematic abbreviations such as Tos-PheCH2C1 since they define the compound concisely, show the relationship between several related halomethyl ketones, and lead to less confusion for those not familiar with shorter abbreviations. See this volume [10]. 3B. R. Baker, "Design of Active-Site-Directed Irreversible Enzyme Inhibitors." Wiley, New York, 1967. 4E. Shaw, in "The Enzymes" (P. Boyer, Ed.,), 3rd ed., Vol. 1, pp. 91-146. Academic Press, New York, 1970; Physiol. Rev. 50, 244 (1970). 5S. J. Singer, Adv. Protein Chem. 22, 1 (1967). J. C. Powers, in "Chemistry and Biochemistry of Amino Acids, Peptides, and Proteins" (B. Weinstein, ed.), Vol. 4, in press. Dekker, New York, 1977. 7I. Schechter and A. Berger, Biochem. Biophys. Res. Commun. 27, 157 (1967).
198
[15]
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
Enzyme Sn- 52
51
Set
I
His . . . .
Cr
RCO- NH-CH-- ~-- CH2--X -
O Inhibitor (X= CI, Br, I )
Sn'S2
$1
Set
His
,I [I ,Io', RCO--NH --CH - - C - - C H 2 I
O-
SCHEMB 1. Reaction of a halomethyl ketone with the active site of a serine protease.
hemiketal,s It is likely that this is a general structural feature of serine proteases that are inhibited by haloketones. In designing a haloinethyl ketone affinity label for a new serine protease, some knowledge of the substrate specificity is required in order to choose the proper amino acid residue to place at the P1 position of the inhibitor, i.e., the position that interacts with the enzymes $1 subsite. The remainder of the inhibitor structure should optimally be free of nonpeptide functional groups. Tos-PheCH.~C1, for example, will not inhibit subtilisin BPN' owing to the sulfonamide group, which would be forced into excessively close contact with the enzyme in any transition state leading to alkylation of the active-site histidine, whereas other chloromethyl ketones (RCO-PheCH~C1) are excellent inhibitors2 -11 Even though the synthesis is more difficult, design of an inhibitor with an extended chain should be considered. In several crystallographic studies, interactions between the extended substrate binding site ( $ 2 - - S , subsites) of the enzyme and an extended peptide chain of the inhibitor has been observed. 8,12,13 For enzymes such as elastase, inhibitors without extended chains react slowly or not at all. ~4,~5 There are now numerous s T. Poulos, R. A. Alden, S. T. Freer, J. J. Birktoft, and J. Kraut, J. Biol. Chem. 251, 1097 (1976). J. D. Robertus, R. A. Alden, J. J. Birktoft, d. Kraut, J. C. Powers, and P. E. Wilcox, Biochemistry 11, 2439 (1972) ~ K. Morihara and T. Oka, Arch Biochem. Biophys. 138, 526 (1970). ~1j. C. Powers, M. O. Lively III, and J. T. Tippett, Biochim. Biophys. Acta. in press (1977). ~2D. M. Segal, J. C. Powers, G. H. Cohen, D. R. Davies, and P. E. Wilcox, Biochemistry 10, 3728 (1971). ~3j. D. Robertus, J. Kraut, R. A. Alden, and J. J. Birktoft, Biochemistry I1, 4293 (1972). ~4L. Visser, D. S. Sigman, and E. R. Blout, Biochemistry 10, 735 (1971). ~ A. Thomson and I. S. Denniss, Eur. J. Biochem. 38, 1 (1973).
[16]
S E R I N E P R O T E A S E S W I T H HALOMETHYL K E T O N E S
199
examples reported in which extending the chain of the inhibitor has accelerated the rate of inhibition. 1°'15-1s Increased reactivity is also obtained by proper choice of the amino acid residues at the various subsites. 11,18 In addition, selectivity within a group of related enzymes can be obtained by use of subsite inactions. 19,-'° The reactivity of haloketones is in the order I > Br > C1. -~ However, increasing the reactivity of the haloketone in this way is dangerous because of the possibility of accelerating competing side reactions. For example, haloketones can react with nucleophilic side chains such as thiols 22 and thioethers. Therefore it is probably more advantageous to increase the reactivity of chloromethyl ketone by altering the peptide structure than by changing the leaving group. Synthesis Halomethyl ketone derivatives of blocked amino acids are readily prepared by the reaction of mineral acids (hydrochloric and hydrobromic) with the corresponding diazomethyl ketone (Scheme 2). Iodomethyl Z-AA-OH
l, Z - A A C H N 2
HX jZ_AACH2X HBr
Boc-AA-OH --e B o c - A A C H N 2
HX
NH3-~HCOCH2X R
SCHEME2. Synthesis of halomethyl ketones. ketones are prepared by reaction of a bromo- or chloroketone with N a I since reaction of H I with a diazomethyl ketone yields the methyl ketone. Usually the diazomethyl ketone is not isolated, but is generated in situ from the corresponding blocked amino acid. A number of blocking groups have been used including tosyl (Tos), benzyloxycarbonyl (Z), and t-butyloxycarbonyl (Boc); only the latter two should be used if the ~°J. C. Powers and P. M. Tuhy, Biochemistry 12, 4767 (1963). ~7R. C. Thompson and E. R. Blout, Biochemistry 12, 44 (1973). '~ K. Kurachi, J. C. Powers, and P. E. Wilcox, Biochemistry 12, 771 (1973). is j. R. Coggins, W. Kray, and E. Shaw, Biochem. J. 138, 579 (1974). :° C. Kettner and E. Shaw, Fed. Proc., Fed. Am. Soc. Exp. Biol. 35, 1464 (1976). :1 E. Shaw and J. Ruscica, Arch. Biochem. Biophys. 145, 484 (1971). "~T. Rossman, C. Norris, and W. Troll, J. Biol. Chem. 249, 3412 (1974).
200
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
[15]
goal is a peptide halomethyl ketone. Simple acyl groups should be avoided, since reaction with diazomethane will probably yield an oxazolone. 2~,~4 The diazomethyl ketone is prepared by reaction of diazomethane with the appropriate acid activated by means of dicyclohexylcarbodiimide (DCCI), by the mixed anhydride method ~5 or via the acid chloride. The mixed anhydride procedure is probably the most convenient, avoids oxazolone formation, and allows the use of acid-labile protecting groups. Extreme caution must be exercised in using diazomethane since it is both toxic and explosive. ~6 Unblocked amino acid chloromethyl ketones are prepared by reaction of benzyloxyearbonyl blocked derivatives with HBr in HOAc, ~2,27 trifluoroacetic acid, 21or by hydrogenation.TM The hydrogenolysis procedure can lead to difficulty in some cases since the chlorine is also susceptible to hydrogenolysis.2s-3° A very useful method for the synthesis of chloromethyl ketone derivatives of amino acids utilizes the ready removal of the t-butyloxycarbonyl (Boc) group with hydrochloric acid (Scheme 1). In one step the chloromethyl ketone moiety is generated from the diazomethyl ketone and the protecting group is removed. 1~,8~ Synthesis of peptide chloromethyl ketones can be accomplished simply bY coupling of an appropriate peptide or amino acid with an unblocked amino acid chloromethyl ketone) °,12,27 A few dipeptides have been converted directly to the chloromethyl ketone using the mixed anhydride and CH2N2 followed by HC122 Various synthetic problems are encountered in the preparation of chloromethyl ketone derivatives of basic amino acids. The side chain usually must be blocked during synthesis, and difficulties are often encountered during removal of the blocking group. For example, deblocking of Tos-Lys(Z)CH2C1 to Tos-LysCH2C1 by several methods 2sin gives a 2sp. Karrer and R. Widmer, Helv. Chim. Acta. S, 203 (1925). 54p. Karrer and G. Bussman,Helv. Chim. Acta. 24, 645 (1941). B. Penke, J. Czombos, L. Balaspiri, J. Petres, and K. Kovacs, Helv. Chim. Acta. 53, 1057 (1970). ~J. A. Moore and D. E. Reed, Org. Syn. 41, 16 (1961). 2~j. C. Powers and P. E. Wilcox,J. Am. Chem. Soc. 92, 1782 (1970). 2sE. Shaw, M. Mares-Guia, and W. Cohen, Biochemistry 4, 2219 (1965). L. Y. Frolova, G. K. Kovaleva, M. B. Agalarova, and L. L. Kisselev, FEBS Left. 34, 213 (1973). R. C. Thompson, Biochemistry 13, 5495 (1974). 3,p. L. Birch, H. A. El-Obeid, and M. Akhtar, Arch. Biochem. Biophys. 148, 447 (1972). asM. Tejima, M. Takeuchi, E. Ichishima, and S. Kobayashi, Agr. Biol. Chem. 39, 1423 (1975). 2sF. Sebestyen and J. Samu, Chem. Ind. (London), p. 1568 (1970).
[16]
SERINE PROTEASES WITH HALOMETHYL KETONES
201
low yield of product. Use of trifluoroacetic acid "~ or H F "~ was eventually found to give a good conversion to product. A number of peptide chloromethyl ketones with terminal LysCH.,C1 residues }lave been prepared via Lys(Z)CH2C1. L9 The chain was extended by coupling with Boc-blocked amino acids or peptides. Deblocking of the Boc group was accomplished with 5 M HC1 in ethanol. Use of CF:~C()~H in the case of dipeptides results in loss of 'the chloromethyl ketone group probably due to a cyclization reaction. In an analogous reaction Tos-OrnCH~Cl slows cyclizes by reaction of the chloromethyl ketone moiety with the side-chain amino group?' Tos-LysCH.,C1, on tile other hand, is completely stable. Arginine chloromethyl ketones, although of considerable interest, h'~ve proved to be difficult to synthesize. Z(NO=.~-ArgCH=,C1 was obtained as a contaminant (1.5-2%) of a cyclized product that was prepared by reaction of Z (NQ)-Arg-C1 with CHIN., followed by HC1. :~4Tos-ArgCH~C1 has been successfully synthesized by using the nitro blocking group for the arginine side chain. ~'~ Peptide chloromethyl ketones with terminal Arg residues have recently been reported. ~ The synthesis of several chloromethyl ketones have been reported in other volumes of this series. These include Tos-PheCH~C1, Z-PheCH._,CI, Tos-LysCH~C1, PheCH~C1, and Z-Ala-Gly-PheCH~C12 ~,:~: Halomethyl or diazomethyl ketones derivatives of most amino acids (Ala, Arg, Asp, Cys, Glu, Gln, Gly, Ile, Leu, Lys, Phe, Pro, Thr, Trp, Tyr, Val, and several unusual amino acids) has been described in the literature. A recent comprehensive review lists all halomethyl and diazomethyl ketones reported in the literature through September 1975. Over a hundred derivatives of amino acids and approxim'~tely 60 peptide derivatives are listed. 6 Two representative procedures are presented in detail. The first, preparation of Z-Gly-Leu-PheCH~C1, involves deblocking of Z-PheCH~C1 and subsequent coupling of the deblocked chloromethyl ketone with Z-Gly-Leu-OH. 18 This compound is an excellent inhibitor of chymotrypsin 's and cathepsin G. 3s The second procedure, that for Ac-Ala-Ala-ProValCH~C1, illustrates the synthesis of a peptide chloromethyl ketone from Boc-Val-0H with a minimum of isolation and purification along the way3~; this compound is an excellent inhibitor of porcine pancreatic and human leukocyte elastase29 3~E. Shaw and G. Glover, Arch. Biochem. Biophy~. 139, 298 (1970). '~ K. Inouye, A. Sasaki, and N. Yoshida, Bull. Chem. Soc. Jp~. 47, 202 (1974). E. Shaw, this series, Vol. 11, p. 677 (1967). 37E. Shaw, this series, Vol. 25, p. 655 (1972). 38j . C. Powers and B. F. Gupton, unpublished observations. 39j . C. Powers and R. J. Whitley, unpublished observations.
202
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
[16]
N-Benzyloxycarbonylglycyl-L-leucyl-L-phenylalanine Chloro~ethyl Ketone (Z-Gly-Leu-PheCH2Cl)W To a saturated solution of HBr in acetic acid are added 3.5 g (10.4 mmoles) of Z-PheCH~C1. After 20 min, ether is added and the resulting crystals are filtered. Recrystallization from ethyl acetate-methanol will allow an approximately 70% yield of L-phenylalanine chloromethyl ketone hydrobromide (m.p. 180°-181 °, dec). It is not necessary to recrystallize the product; simply washing with dry ether provides material suitable for subsequent coupling. The hydrobromide must be kept anhydrous. In the present procedure, the intermediate was not recrystallized. The mixed anhydride prepared from 3.2 g (10 mmoles) of Z-Gly-LeuOH, 1.1 ml (10 mmoles) of N-methylmorpholine and 1.34 ml (10 mmoles) of isobutyl chloroformate in 60 ml of anhydrous tetrahydrofuran at --10 ° to --15 ° is allowed to react with L-phenylalanine chloromethyl ketone hydrobromide and 1.3 ml of N-methylmorpholine for 1.25 hr while the reaction mixture is allowed to warm to room temperature under anhydrous conditions. Work-up of the reaction mixture involves evaporation of tetrahydrofuran, extraction of the residue into ethyl acetate, washing with citric acid and sodium bicarbonate solutions, drying over magnesium sulfate, and evaporation. The product is recrystallized from ethyl acetatecyclohexane to yield 3.41 g (68%) of product with m.p. 140.5°-143 °.
N-A cetyl-L-alanyl-L-alanyl-L-prolyl-L-valine C hloromethyl Ketone (Ac-Ala-Ala-Pro-ValCH2C1). 4° The diazomethyl ketone Boc-ValCHN2 is prepared from Boc-Val-OH (10 g, 45 mmoles) by the mixed anhydride method (N-methylmorpholine, 45 mmoles; isobutyl chloroformate, 45 mmoles) in tetrahydrofuran at --10 ° to --15 °. The mixed anhydride solution is filtered under anhydrous conditions to remove N-methylmorpholine hydrochloride and is then added to excess CHIN2 in ether36 After standing overnight at room temperature, anhydrous HC1 is bubbled through the solution of Boc-ValCHN_o at 5 ° for 15 min as the solution turns from yellow to colorless. The product (HC1.ValCH.~CI) is isolated by evaporation of the solvents and dried under reduced pressure to give 4.5 g (55%) of an extremely hydroscopic white solid. Ac-Ala-Ala-Pro-OH (1.5 g, 5 mmoles) is dissolved in 50 ml of anhydrous tetrahydrofuran and stirred at --15 ° while N-methylmorpholine (0.55 ml, 5 mmoles) and isobutyl chloroformate (0.65 ml, 5 mmoles) are *°Anhydrous conditions must be maintained during the synthesis of diazomethyl ketones, their reaction with HCI to form chloromethyl ketones, and coupling reactions. Solvents and most reagents can be purchased in anhydrous form. Prior to use, all apparatus is dried in an oven or flamed while passing a stream of dry nitrogen through glassware. Mixed anhydride solutions are filtered through a jacketed sintered glass filter, using dry nitrogen to exclude moisture.
[16]
SERINE PROTEASES WITH HA.LOMETttYL KETONES
203
added. After 10 min, the N-methylmorpholine salts were filtered and HC1 • ValCH~C1 (0.93 g, 5 mmoles) in 5 ml of D M F and N-methylmorpholine (0.55 ml, 5 mmoles) are added. Stirring is continued for 4 hr as the mixture warmed to 25 ° . Thereafter, the solvent is removed under reduced pressure and the resulting yellow oil is dissolved in methanol and chromatographed on Merck silica gel G (0.063-0.2 mm). The product, eluted with 8% methanol in chloroform, is recrystallized from methanol to give 1.05 g (49% yield, m.p. 195°-196 °, dec.). Kinetics of the Reaction
The irreversible reaction of a ehloromethyl inhibitor with a protease may be represented by the overall reaction sequence in Eq. (1), where E . I represents a noncovalently bound complex of the enzyme with the inhibitor and E - I is the final product with the inhibitGr irreversibly bound to the enzyme via a covalent linkage. k3
E + I ~- E.I--, E - I
(1)
Kx = [E][I] [E. II
(2)
If the inhibitor concentration is sufficiently greater than the total enzyme concentration ( [ I ] / [ E ] > 10-20), it can be shown that the decrease in E + E . I concentration in the inhibition mixture follows pseudo-firstorder kinetics at any fixed value of I. The pseudo-first-order rate constant can easily be measured by periodically drawing an aliquot from the inhibition mixture and assaying for residual enzyme activity. Dilution of the inhibitor concentration in the assay mixture will likely result in the absence of E . I in the assay mixture, so that the enzyme concentration obtained will represent the total enzyme ( E ~ E . I ) present in the aliquot. A semilog plot of enzyme activity or concentration vs time will then give ko,~ directly. It can be shown that kob~ is related to K~ by Eq. (2) and k~ by Eq. (3).1s,41 1/kobs = (gi/k3[I]) -t- 1/k3
(3)
If data are available over a sufficient range of inhibitor concentrations, both K, (dissociation constant of the E . I complex) and ka (the limiting rate of inactivation) may be evaluated by the use of a double-reciprocal plot (1/koh.~ vs 1/[I]) and Eq. (3). When the K~ value is much greater than the chosen inhibitor concentrations, Eq. (3) reduces to Eq. (4) and kobs/[I] = k ~ / g l 41R. Kitz and I. B. Wilson, J. Biol. Chem. 277, 3245 (1970).
(4)
204
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
[15]
predicts that a reciprocal plot of 1/kobs vs 1/[I] will pass through the origin and that K~ cannot be evaluated. From a practical standpoint, this situation occurs often, since K~ values are in the millimolar range whereas inhibitor concentrations are frequently limited to lower values owing either to low solubility of the inhibitor or its high rate of reaction, is It should be pointed out that in a situation wherein kob~ is measured at one inhibitor concentration or where kobJ[I] does not vary in the range of inhibitor concentrations utilized, then it is not possible to demonstrate the existence of an enzyme-inhibitor complex. The most appropriate parameter for comparison of the reactivity of various inhibitors is the inhibition parameter k3/Ki. In the absence of sufficient data for determining k3/K~, kobJ[I] may be used. The differences in the magnitude of the numbers for related inhibitors reflect mostly the effect of structural changes on the binding of the inhibitor to the enzyme (K~) and on the rate of reaction within the bound complex (k3). It will be apparent that values of kob~/[I] are subject to distortion from nonlinear concentration effects when the inhibitor concentration is close to the K~ value of the inhibitor. Since most halomethyl ketones are only slightly soluble in water, the kinetics of inhibition are usually measured in a mixed solvent. The inhibitors are dissolved in such solvents as methanol or dimethoxyethane and then diluted with buffer. If the solution becomes cloudy or the haloketone precipitates at this point, either the concentration of organic solvent must be increased or the concentration of inhibitor decreased. Solutions of chloromethyl ketones in organic solvents are fairly stable, of the order of a week or more, but aqueous solution slowly lose inhibitory activity owing to hydrolysis. 16,1s A solution of Ac-Gly-PheCH~Cl was hydrolyzed 30% after standing for 20 hr in a buffer solution at pH 5.80. The rates of inhibition with halomethyl ketone are quite variable and depend on the pH, the enzyme, and the structure of the inhibitor. The fastest reactions have half-lives of less than 1 min whereas poor inhibitors may take many hours. The optimum in most cases is at pH 7 to 8, but rates are often measured at lower values to slow the reaction with reactive inhibitors to a more manageable range. Identification o] the Site o] Reaction. The alkylation of the active site histidine residue by a substrate related haloketone has been used often to locate a specific histidine in the primary sequency of a serine protease. However, haloketones are general alkylating agents, and if the reagent is too far removed in structure from a substrate, it is likely to alkylate nucleophilic sites other than the catalytic histidine. 42 In one 42K. J. Stevenson and L. B. Smillie, Can. J. Biochem. 46, 1357 (1968).
[1~]
SERINE PROTEASES WITH HALOMETHYL KETONES
205
case, alkylation of the catalytic serine residue of trypsin was observed when an inhibitor was synthesized with a decreased distance between the chloromethyl ketone group and a charged side chain relative to normal substrate or inhibitor. 43 However, this must be regarded as an exception, since it is likely that formation of a tetrahedral intermediate 8 by reaction of the catalytic serine with the carbonyl group of the inhibitor is a requirement for inhibition. The site of reaction can be identified by isolation of a peptide fragment containing the modified amino acid residue. This process is quite efficient when a diagonal electrophoresis technique is utilized. This involves electrophoresis of a peptide mixture on some support, treatment with H C Q H , and then electrophoresis in a direction perpendicular to the first. Oxidation of histidine alkylated by a halomethyl ketone yields earboxymethyl histidine. Disulfides are also oxidized to cysteic acid residues. All peptides that contained cystine residues or amino acid residues alkylated by the haloketone will lie off the diagonal after the second electrophoresis and can easily be isolated. ~-~G A modification of this technique allows separation of only the peptide containing the alkylated amino acid residue. ~7 The alkylated protein is first reduced, then eyanoethylated, and the eyanoethylated cysteine residues are oxidized to the sulfone. This change is not sufficiently large to remove such peptides from the diagonal, and only the peptide with the alkylated amino acid residue is found off the diagonal.
Applications Representative examples of halomethyl ketones which have been used as affinity labels for serine proteases are listed in the table. Several thiol proteases (papain, clostripain, cathepsin B, ~ and cocoonase) are also listed since the mechanism of the inhibition is probably similar for the two classes of enzymes. The inhibitors chosen for the table were either the most readily available or among the more reactive for a particular enzyme. For more information the reader is referred to a recent comprehensive review of all the enzymes that have been studied with halomethyl ketone inhibitors2 Most of the potential applications of affinity labeling have been realized by the use of halomethyl ketones. They have been used to label '~ D. D. Schroedcr and E. Shaw, Arch. Biochem. Biophys. 142, 340 (1971). ~ K. J. Stevenson and L. B. Smillie. J. Mol. Biol. 12, 937 (1965). ~5L. B. Smillie and B. S. Hartley, Biochem. J. 101, 232 (1966). 46K. J. Stevenson and L. B. Smillie, Can. J. Biochem. 46, 1357 (1968). 47K. J. Stevenson, Biochem. J. 139, 215 (1974).
206
[15]
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
INHIBITION OF PROTEOLYTIC ENZYMES WITH HALOKETONE DERIVATIVES OF AMINO ACIDS AND PEPTIDES
Enzyme Chymotrypsin Cathepsin G Subtilisin
Streptomyces griseus protease B Elastase
Trypsin
Acrosin Thrombin
Plasmin Kallikrein Carboxypeptidase Y Carboxypeptidase C Acid carboxypeptidases Papain Clostripain Cathepsin B x Cocoonase
Inhibitor
References a
TosPheCH,CI Z-Gly-Leu-PheCH2C1 Z-Gly-Leu-PheCH2Cl Z-PheCH~Br Z-Ala-Gly-PheCH~C1 Ac-Phe-Gly-Ala-LeuCH2C1 Phe-Ala-LysCH,C1 Boc-Gly-Leu-PheCH..C1 Ac-AlaCH~C1 Ac-Ala-Ala-AlaCH:C1 Ac-Ala-Ala-Pro-AlaCH2C1 Tos-LysCH~CI Tos-ArgCH~Cl Lys-Ala-LysCH2C1 Tes-LysCH2CI Tos-LysCH2C1 Phe-Ala-LysCH~C1 Glu-Ala-LysCH~C1 Tos-LysCH~C1 Z-LysCH2C1 Ala-Phe-LysCH2C1 Z-PheCH2C1 Tos-PheCH2C1 Tos-PheCH,C1 Tos-PheCH2Cl Z-Gly-Leu-PheCH2C1 Tos-LysCH,C1 Tos-PheCH2C] Ac-Ala-Ala-Ala-AlaCH~C1 Tos-LysCH..C1
5 18 38 b 10 11 19 c 15 16, d 16, 17, e,f 34, g g 19 h-j k 19, l 19, l m, n 19, o 19 p, q r s, t u, v w, x y z aa
a Numbers refer to text footnotes. b F. S. Markland, E. Shaw, and E. L. Smith, Proc. Natl. Acad. Sci. U.S.A. 61, 1440 (1968). c A. Gertler, FEBS Lett. 43, 81 (1974). d j. Travis and R. C. Roberts, Biochemistry 8, 2884 (1969). J. C. Powers and P. M. Tuhy, J. Am. Chem. Soc., 94, 6544 (1972). I p. IV[. Tuhy and J. C. Powers , FEBS Lett. 50, 359 (1975). N. Yoshida, A. Sasaki, and K. Inouye, Biochem. Biophys. Acta 321, 615 (1973) i, L. Zaneveld, K. Polakoski, and W. Williams, Biol. Repro& 6, 30 (1972). i L. Zaneveld, B. Dragoje, and G. Schumacher, Science 177, 702 (1972). i K. L. Plaskoski and R. A. McRorie, J. Biol. Chem. 248, 8183 (1973). k T. M. Chulkova and V. N. Orekhovick, Biokhimiya 33, 1222 (1968). ~E. Shaw, in "Proteinase Inhibitors" (H. Fritz, H. Tschesche, L. J. Green, and E. Truscheit, eds.), pp. 531-540. Springer-Verlag, Berlin and New York, 1974.
[16]
SERINE PROTEASES WITH HALOMETHYL KETONES
207
F. F. Buck, B. Hummel, and E. DeRenzo, J. Biol. Chem. 243, 3648 (1968). W. R. Groskopf, B. Hsieh, L. Summaria, and K. Robbins, J. Biol. Chem. 244, 359 (1969). ° C. Sampaio, S. Wong, and E. Shaw, Arch. Biochem. Biophys. 165, 133 (1974). R. Hayashi, Y. Bai, and T. Hata, J. Biochem. (Tokyo) 76, 1355 (1974). q R. Hayashi, Y. Bai, and T. Hata, J. Biol. Chem. 250, 5221 (1975). r R. W. Kuhn, K. A. Walsh, and H. Neurath, Biochemistry 13, 3871 (1974). E. Ichishima, S. Sonoki, K. Hirai, Y. Torii, and S. Yokoyama, J. Biochem. (Tokyo) 72, 1045 (1972). T. Nakadai, S. Nasuno, and N. Iguchi, Attic. Biol. Chem. 37, 1237 (1973). S. S. Husain and G. Lowe, Chem. Commun. 1965, 345 (1965). " M. L. Bender and L. J. Brubacker, J. Ant. Chem. Soc. 88, 5880 (1966). S. Roffman, M. Levy, and W. Troll, Fed. Proc., Fed. Am. Soc. Exp. Biol. 32, 466 (1973). S. L. Roffman, Diss. Abstr. Int. 35B, 3218 (1975). W. H. Porter, L. W. Cunningham, and W. M. Mitchell, J. Biol.Chem. 246, 7675 (1971). M. C. Burleigh, A. J. Barrett, and G. S. Lazarus, Biochem. J. 137, 387 (1974). ~ F. Kafatos, J. H. Law, and A. M. Tartakoff, J. Biol. Chem. 242, 1488 (1967). and locate the active site histidine in numerous serine proteases including chymotrypsin, trypsin, subtilisin, thrombin, and elastase. H e a v y atomcontaining halomethyl ketones have been synthesized for possible use in crystallographic phasing during the determination of the three-dimensional structure of proteins. ~s Indeed, the determination of the structures of chymotrypsin and subtilisin t h a t had been treated with halomethyl ketones led to the discovery of the extended substrate binding site in these enzymes2,9,1~ Numerous kinetic studies with chloromethyl ketones have shown the existence of extended binding sites in other proteases or have sought to determine the nature of the individual subsites.~°, l~,15--~° H a l o m e t h y l ketones have been used to place either spectroscropic, 49 fluorescent, ~° or spin-labeled 51 reporter groups in the active sites of serine proteases. A chloromethyl ketone has been used in a nuclear magnetic resonance study of catalytic residues of a serine protease. ~2 More recently, halomethyl ketones have been extensively used in studies of role of proteases in such diverse biological processes as fertilization, cell growth, protein synthesis, virus maturation, and diseases such as e m p h y s e m a and cancer. 6 However, in m a n y cases appropriate controls were not carried out and it is impossible to discern whether the effect of the haloketone is A. Tulinsky, R. L. Vandlen, C. N. Morimoto, N. V. Mani, and L. H. Wright, Biochemistry 4185 (1973).
,9 D. S. Sigman and E. R. Blout, J. Am. Chem. Soc. 89, 1747 (1967). 50G. Schoellmann, Int. J. Peptide Protein Res. 4, 221 (1972). 51D. J. Kosman, J. Mol. Biol. 67, 247 (1972). 55G. Robillard and R. G. Shulman, J. Mol. Biol. 86, 519 (1974).
208
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
[17]
due to specific inhibition of a serine protease or to other alkylation reactions. In the future, there is an excellent possibility that some of these affinity labels may be used medicinally in the control of both normal and abnormal physiological processes involving serine proteases. [17] R e a c t i o n o f S e r i n e P r o t e a s e s w i t h A z a - A m i n o A c i d a n d Aza-Peptide Derivatives
By
JAMES C. POWERS and B. FRANK GUPTON
Aza-amino acid residues are analogs of amino acids in which the a-CH has been replaced by a nitrogen atom. 1 This substitution has a profound effect on the reactivity of aza-peptides, i.e., those containing ~NH~CH--CO--
I R
~NH--N~CO
I
R
aza-amino acid residues. In particular, it has been shown that azapeptide p-nitrophenyl esters are inhibitors and active site titrants of several serine proteases. 2-4 Inhibition of these enzymes is believed to arise from the acylation of the active-site serine residue yielding an acylated enzyme (Fig. 1), which is substantially less reactive toward deacylation than a normal acylated enzyme owing to the influence of the adjacent nitrogen atom. At present, there is no evidence that the activesite serine residue is actually the site of reaction. However, the close structural resemblance of aza-peptides to normal peptides substrates of serine proteases points strongly to the serine residue as the one being labeled. The only requirement for design of an aza-peptide reagent for a new serine protease is knowledge of the enzyme's substrate specificity so that an appropriate side chain may be placed on the aza-amino acid residue. The employment of p-nitrophenol as a product of the acylation step is advantageous since it is a good leaving group, and the release of p.nitroThe use of standard IUPAC nomenclature (2-substituted carbazic acid derivatives) for the naming of aza-amino acid residues is particularly disadvantageous since the names are cumbersome and give no indication of the structural relationship with amino acids. For this reason, the method of naming these analogs in this text will be to precede the name of the corresponding amino acid with "aza." For example, 2-benzylcarbazoie acid p-nitrophenyl ester will be referred to as aza-phenylalanyl p-nitrophenyl ester. Likewise, the method of abbreviating these analogs will be to prefix the abbreviation of the corresponding amino acid with an "A." For example, the abbreviation of aza-phenylalanine will be Aphe. A. N. Kurtz and C. Niemann, J. Am. Chem. Soe. 83, 1879 (1961). s D. T. Emore and J. J. Smyth, Biochem. J. 107, 103 (1968). J. C. Powers and D. L. Carroll, Bioehem. Biophys. Res. Commun. 67, 639 (1975).
SERINE PROTEASES WITH HALOMETHYL KETONES
197
[16] Reaction of Serine Proteases with Halomethyl Ketones By
JAMES C. POWERS
The use of the chloromethyl ketones Tos-PheCH.~C1 ( T P C K ) and Tos-LysCH2C1 ( T L C K ) as specific reagents for chymotrypsin and trypsin, respectively, is one of the classic demonstrations of the value of affinity labels for enzyme studies? Subsequently, halomethyl ketones were used for characterization of functional groups and sequences in active sites, as probes of structure by providing labeled proteins that contain spectroscopic handles, in crystallographic studies of substrate binding sites, and in the study of the biological function of proteases. Today, halomethyl ketones are probably the most widely studied class of affinity label and have been discussed in numerous reviews. 2-6 The reaction of a serine protease with most substrate-related haloketones probably first involves the formation of an enzyme inhibitor complex (Scheme l) in which the inhibitor is recognized by specific interactions between the side chain of the P1 amino acid residue and the $1 or primary substrate binding site of the enzyme (nomenclature of Schecter and B u r g e r ) . Irreversible inhibition takes place within this complex by covalent bond formation between the active-site histidine residue and a methylene group of the inhibitor. In addition, crystallographic studies in a few cases indicate that the active-site serine oxygen has added to the carbonyl group of the inhibitor to give a tetrahedral
The nomenclature used in this chapter for amino acid and peptide derivatives Conforms to the 1971 Recommendations of the IUPAC-IUB Commission on Biochemical Nomenclature, Biochemistry 11, 1726 (1972). Thus the chloromethyl ketone derived from tosyl-L-phenylalanine will be abbreviated as Tos-PheCH2C1 instead of the more commonly used TPCK. The author recommends the use of systematic abbreviations such as Tos-PheCH2C1 since they define the compound concisely, show the relationship between several related halomethyl ketones, and lead to less confusion for those not familiar with shorter abbreviations. See this volume [10]. 3B. R. Baker, "Design of Active-Site-Directed Irreversible Enzyme Inhibitors." Wiley, New York, 1967. 4E. Shaw, in "The Enzymes" (P. Boyer, Ed.,), 3rd ed., Vol. 1, pp. 91-146. Academic Press, New York, 1970; Physiol. Rev. 50, 244 (1970). 5S. J. Singer, Adv. Protein Chem. 22, 1 (1967). J. C. Powers, in "Chemistry and Biochemistry of Amino Acids, Peptides, and Proteins" (B. Weinstein, ed.), Vol. 4, in press. Dekker, New York, 1977. 7I. Schechter and A. Berger, Biochem. Biophys. Res. Commun. 27, 157 (1967).
198
[15]
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
Enzyme Sn- 52
51
Set
I
His . . . .
Cr
RCO- NH-CH-- ~-- CH2--X -
O Inhibitor (X= CI, Br, I )
Sn'S2
$1
Set
His
,I [I ,Io', RCO--NH --CH - - C - - C H 2 I
O-
SCHEMB 1. Reaction of a halomethyl ketone with the active site of a serine protease.
hemiketal,s It is likely that this is a general structural feature of serine proteases that are inhibited by haloketones. In designing a haloinethyl ketone affinity label for a new serine protease, some knowledge of the substrate specificity is required in order to choose the proper amino acid residue to place at the P1 position of the inhibitor, i.e., the position that interacts with the enzymes $1 subsite. The remainder of the inhibitor structure should optimally be free of nonpeptide functional groups. Tos-PheCH.~C1, for example, will not inhibit subtilisin BPN' owing to the sulfonamide group, which would be forced into excessively close contact with the enzyme in any transition state leading to alkylation of the active-site histidine, whereas other chloromethyl ketones (RCO-PheCH~C1) are excellent inhibitors2 -11 Even though the synthesis is more difficult, design of an inhibitor with an extended chain should be considered. In several crystallographic studies, interactions between the extended substrate binding site ( $ 2 - - S , subsites) of the enzyme and an extended peptide chain of the inhibitor has been observed. 8,12,13 For enzymes such as elastase, inhibitors without extended chains react slowly or not at all. ~4,~5 There are now numerous s T. Poulos, R. A. Alden, S. T. Freer, J. J. Birktoft, and J. Kraut, J. Biol. Chem. 251, 1097 (1976). J. D. Robertus, R. A. Alden, J. J. Birktoft, d. Kraut, J. C. Powers, and P. E. Wilcox, Biochemistry 11, 2439 (1972) ~ K. Morihara and T. Oka, Arch Biochem. Biophys. 138, 526 (1970). ~1j. C. Powers, M. O. Lively III, and J. T. Tippett, Biochim. Biophys. Acta. in press (1977). ~2D. M. Segal, J. C. Powers, G. H. Cohen, D. R. Davies, and P. E. Wilcox, Biochemistry 10, 3728 (1971). ~3j. D. Robertus, J. Kraut, R. A. Alden, and J. J. Birktoft, Biochemistry I1, 4293 (1972). ~4L. Visser, D. S. Sigman, and E. R. Blout, Biochemistry 10, 735 (1971). ~ A. Thomson and I. S. Denniss, Eur. J. Biochem. 38, 1 (1973).
[16]
S E R I N E P R O T E A S E S W I T H HALOMETHYL K E T O N E S
199
examples reported in which extending the chain of the inhibitor has accelerated the rate of inhibition. 1°'15-1s Increased reactivity is also obtained by proper choice of the amino acid residues at the various subsites. 11,18 In addition, selectivity within a group of related enzymes can be obtained by use of subsite inactions. 19,-'° The reactivity of haloketones is in the order I > Br > C1. -~ However, increasing the reactivity of the haloketone in this way is dangerous because of the possibility of accelerating competing side reactions. For example, haloketones can react with nucleophilic side chains such as thiols 22 and thioethers. Therefore it is probably more advantageous to increase the reactivity of chloromethyl ketone by altering the peptide structure than by changing the leaving group. Synthesis Halomethyl ketone derivatives of blocked amino acids are readily prepared by the reaction of mineral acids (hydrochloric and hydrobromic) with the corresponding diazomethyl ketone (Scheme 2). Iodomethyl Z-AA-OH
l, Z - A A C H N 2
HX jZ_AACH2X HBr
Boc-AA-OH --e B o c - A A C H N 2
HX
NH3-~HCOCH2X R
SCHEME2. Synthesis of halomethyl ketones. ketones are prepared by reaction of a bromo- or chloroketone with N a I since reaction of H I with a diazomethyl ketone yields the methyl ketone. Usually the diazomethyl ketone is not isolated, but is generated in situ from the corresponding blocked amino acid. A number of blocking groups have been used including tosyl (Tos), benzyloxycarbonyl (Z), and t-butyloxycarbonyl (Boc); only the latter two should be used if the ~°J. C. Powers and P. M. Tuhy, Biochemistry 12, 4767 (1963). ~7R. C. Thompson and E. R. Blout, Biochemistry 12, 44 (1973). '~ K. Kurachi, J. C. Powers, and P. E. Wilcox, Biochemistry 12, 771 (1973). is j. R. Coggins, W. Kray, and E. Shaw, Biochem. J. 138, 579 (1974). :° C. Kettner and E. Shaw, Fed. Proc., Fed. Am. Soc. Exp. Biol. 35, 1464 (1976). :1 E. Shaw and J. Ruscica, Arch. Biochem. Biophys. 145, 484 (1971). "~T. Rossman, C. Norris, and W. Troll, J. Biol. Chem. 249, 3412 (1974).
200
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
[15]
goal is a peptide halomethyl ketone. Simple acyl groups should be avoided, since reaction with diazomethane will probably yield an oxazolone. 2~,~4 The diazomethyl ketone is prepared by reaction of diazomethane with the appropriate acid activated by means of dicyclohexylcarbodiimide (DCCI), by the mixed anhydride method ~5 or via the acid chloride. The mixed anhydride procedure is probably the most convenient, avoids oxazolone formation, and allows the use of acid-labile protecting groups. Extreme caution must be exercised in using diazomethane since it is both toxic and explosive. ~6 Unblocked amino acid chloromethyl ketones are prepared by reaction of benzyloxyearbonyl blocked derivatives with HBr in HOAc, ~2,27 trifluoroacetic acid, 21or by hydrogenation.TM The hydrogenolysis procedure can lead to difficulty in some cases since the chlorine is also susceptible to hydrogenolysis.2s-3° A very useful method for the synthesis of chloromethyl ketone derivatives of amino acids utilizes the ready removal of the t-butyloxycarbonyl (Boc) group with hydrochloric acid (Scheme 1). In one step the chloromethyl ketone moiety is generated from the diazomethyl ketone and the protecting group is removed. 1~,8~ Synthesis of peptide chloromethyl ketones can be accomplished simply bY coupling of an appropriate peptide or amino acid with an unblocked amino acid chloromethyl ketone) °,12,27 A few dipeptides have been converted directly to the chloromethyl ketone using the mixed anhydride and CH2N2 followed by HC122 Various synthetic problems are encountered in the preparation of chloromethyl ketone derivatives of basic amino acids. The side chain usually must be blocked during synthesis, and difficulties are often encountered during removal of the blocking group. For example, deblocking of Tos-Lys(Z)CH2C1 to Tos-LysCH2C1 by several methods 2sin gives a 2sp. Karrer and R. Widmer, Helv. Chim. Acta. S, 203 (1925). 54p. Karrer and G. Bussman,Helv. Chim. Acta. 24, 645 (1941). B. Penke, J. Czombos, L. Balaspiri, J. Petres, and K. Kovacs, Helv. Chim. Acta. 53, 1057 (1970). ~J. A. Moore and D. E. Reed, Org. Syn. 41, 16 (1961). 2~j. C. Powers and P. E. Wilcox,J. Am. Chem. Soc. 92, 1782 (1970). 2sE. Shaw, M. Mares-Guia, and W. Cohen, Biochemistry 4, 2219 (1965). L. Y. Frolova, G. K. Kovaleva, M. B. Agalarova, and L. L. Kisselev, FEBS Left. 34, 213 (1973). R. C. Thompson, Biochemistry 13, 5495 (1974). 3,p. L. Birch, H. A. El-Obeid, and M. Akhtar, Arch. Biochem. Biophys. 148, 447 (1972). asM. Tejima, M. Takeuchi, E. Ichishima, and S. Kobayashi, Agr. Biol. Chem. 39, 1423 (1975). 2sF. Sebestyen and J. Samu, Chem. Ind. (London), p. 1568 (1970).
[16]
SERINE PROTEASES WITH HALOMETHYL KETONES
201
low yield of product. Use of trifluoroacetic acid "~ or H F "~ was eventually found to give a good conversion to product. A number of peptide chloromethyl ketones with terminal LysCH.,C1 residues }lave been prepared via Lys(Z)CH2C1. L9 The chain was extended by coupling with Boc-blocked amino acids or peptides. Deblocking of the Boc group was accomplished with 5 M HC1 in ethanol. Use of CF:~C()~H in the case of dipeptides results in loss of 'the chloromethyl ketone group probably due to a cyclization reaction. In an analogous reaction Tos-OrnCH~Cl slows cyclizes by reaction of the chloromethyl ketone moiety with the side-chain amino group?' Tos-LysCH.,C1, on tile other hand, is completely stable. Arginine chloromethyl ketones, although of considerable interest, h'~ve proved to be difficult to synthesize. Z(NO=.~-ArgCH=,C1 was obtained as a contaminant (1.5-2%) of a cyclized product that was prepared by reaction of Z (NQ)-Arg-C1 with CHIN., followed by HC1. :~4Tos-ArgCH~C1 has been successfully synthesized by using the nitro blocking group for the arginine side chain. ~'~ Peptide chloromethyl ketones with terminal Arg residues have recently been reported. ~ The synthesis of several chloromethyl ketones have been reported in other volumes of this series. These include Tos-PheCH~C1, Z-PheCH._,CI, Tos-LysCH~C1, PheCH~C1, and Z-Ala-Gly-PheCH~C12 ~,:~: Halomethyl or diazomethyl ketones derivatives of most amino acids (Ala, Arg, Asp, Cys, Glu, Gln, Gly, Ile, Leu, Lys, Phe, Pro, Thr, Trp, Tyr, Val, and several unusual amino acids) has been described in the literature. A recent comprehensive review lists all halomethyl and diazomethyl ketones reported in the literature through September 1975. Over a hundred derivatives of amino acids and approxim'~tely 60 peptide derivatives are listed. 6 Two representative procedures are presented in detail. The first, preparation of Z-Gly-Leu-PheCH~C1, involves deblocking of Z-PheCH~C1 and subsequent coupling of the deblocked chloromethyl ketone with Z-Gly-Leu-OH. 18 This compound is an excellent inhibitor of chymotrypsin 's and cathepsin G. 3s The second procedure, that for Ac-Ala-Ala-ProValCH~C1, illustrates the synthesis of a peptide chloromethyl ketone from Boc-Val-0H with a minimum of isolation and purification along the way3~; this compound is an excellent inhibitor of porcine pancreatic and human leukocyte elastase29 3~E. Shaw and G. Glover, Arch. Biochem. Biophy~. 139, 298 (1970). '~ K. Inouye, A. Sasaki, and N. Yoshida, Bull. Chem. Soc. Jp~. 47, 202 (1974). E. Shaw, this series, Vol. 11, p. 677 (1967). 37E. Shaw, this series, Vol. 25, p. 655 (1972). 38j . C. Powers and B. F. Gupton, unpublished observations. 39j . C. Powers and R. J. Whitley, unpublished observations.
202
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
[16]
N-Benzyloxycarbonylglycyl-L-leucyl-L-phenylalanine Chloro~ethyl Ketone (Z-Gly-Leu-PheCH2Cl)W To a saturated solution of HBr in acetic acid are added 3.5 g (10.4 mmoles) of Z-PheCH~C1. After 20 min, ether is added and the resulting crystals are filtered. Recrystallization from ethyl acetate-methanol will allow an approximately 70% yield of L-phenylalanine chloromethyl ketone hydrobromide (m.p. 180°-181 °, dec). It is not necessary to recrystallize the product; simply washing with dry ether provides material suitable for subsequent coupling. The hydrobromide must be kept anhydrous. In the present procedure, the intermediate was not recrystallized. The mixed anhydride prepared from 3.2 g (10 mmoles) of Z-Gly-LeuOH, 1.1 ml (10 mmoles) of N-methylmorpholine and 1.34 ml (10 mmoles) of isobutyl chloroformate in 60 ml of anhydrous tetrahydrofuran at --10 ° to --15 ° is allowed to react with L-phenylalanine chloromethyl ketone hydrobromide and 1.3 ml of N-methylmorpholine for 1.25 hr while the reaction mixture is allowed to warm to room temperature under anhydrous conditions. Work-up of the reaction mixture involves evaporation of tetrahydrofuran, extraction of the residue into ethyl acetate, washing with citric acid and sodium bicarbonate solutions, drying over magnesium sulfate, and evaporation. The product is recrystallized from ethyl acetatecyclohexane to yield 3.41 g (68%) of product with m.p. 140.5°-143 °.
N-A cetyl-L-alanyl-L-alanyl-L-prolyl-L-valine C hloromethyl Ketone (Ac-Ala-Ala-Pro-ValCH2C1). 4° The diazomethyl ketone Boc-ValCHN2 is prepared from Boc-Val-OH (10 g, 45 mmoles) by the mixed anhydride method (N-methylmorpholine, 45 mmoles; isobutyl chloroformate, 45 mmoles) in tetrahydrofuran at --10 ° to --15 °. The mixed anhydride solution is filtered under anhydrous conditions to remove N-methylmorpholine hydrochloride and is then added to excess CHIN2 in ether36 After standing overnight at room temperature, anhydrous HC1 is bubbled through the solution of Boc-ValCHN_o at 5 ° for 15 min as the solution turns from yellow to colorless. The product (HC1.ValCH.~CI) is isolated by evaporation of the solvents and dried under reduced pressure to give 4.5 g (55%) of an extremely hydroscopic white solid. Ac-Ala-Ala-Pro-OH (1.5 g, 5 mmoles) is dissolved in 50 ml of anhydrous tetrahydrofuran and stirred at --15 ° while N-methylmorpholine (0.55 ml, 5 mmoles) and isobutyl chloroformate (0.65 ml, 5 mmoles) are *°Anhydrous conditions must be maintained during the synthesis of diazomethyl ketones, their reaction with HCI to form chloromethyl ketones, and coupling reactions. Solvents and most reagents can be purchased in anhydrous form. Prior to use, all apparatus is dried in an oven or flamed while passing a stream of dry nitrogen through glassware. Mixed anhydride solutions are filtered through a jacketed sintered glass filter, using dry nitrogen to exclude moisture.
[16]
SERINE PROTEASES WITH HA.LOMETttYL KETONES
203
added. After 10 min, the N-methylmorpholine salts were filtered and HC1 • ValCH~C1 (0.93 g, 5 mmoles) in 5 ml of D M F and N-methylmorpholine (0.55 ml, 5 mmoles) are added. Stirring is continued for 4 hr as the mixture warmed to 25 ° . Thereafter, the solvent is removed under reduced pressure and the resulting yellow oil is dissolved in methanol and chromatographed on Merck silica gel G (0.063-0.2 mm). The product, eluted with 8% methanol in chloroform, is recrystallized from methanol to give 1.05 g (49% yield, m.p. 195°-196 °, dec.). Kinetics of the Reaction
The irreversible reaction of a ehloromethyl inhibitor with a protease may be represented by the overall reaction sequence in Eq. (1), where E . I represents a noncovalently bound complex of the enzyme with the inhibitor and E - I is the final product with the inhibitGr irreversibly bound to the enzyme via a covalent linkage. k3
E + I ~- E.I--, E - I
(1)
Kx = [E][I] [E. II
(2)
If the inhibitor concentration is sufficiently greater than the total enzyme concentration ( [ I ] / [ E ] > 10-20), it can be shown that the decrease in E + E . I concentration in the inhibition mixture follows pseudo-firstorder kinetics at any fixed value of I. The pseudo-first-order rate constant can easily be measured by periodically drawing an aliquot from the inhibition mixture and assaying for residual enzyme activity. Dilution of the inhibitor concentration in the assay mixture will likely result in the absence of E . I in the assay mixture, so that the enzyme concentration obtained will represent the total enzyme ( E ~ E . I ) present in the aliquot. A semilog plot of enzyme activity or concentration vs time will then give ko,~ directly. It can be shown that kob~ is related to K~ by Eq. (2) and k~ by Eq. (3).1s,41 1/kobs = (gi/k3[I]) -t- 1/k3
(3)
If data are available over a sufficient range of inhibitor concentrations, both K, (dissociation constant of the E . I complex) and ka (the limiting rate of inactivation) may be evaluated by the use of a double-reciprocal plot (1/koh.~ vs 1/[I]) and Eq. (3). When the K~ value is much greater than the chosen inhibitor concentrations, Eq. (3) reduces to Eq. (4) and kobs/[I] = k ~ / g l 41R. Kitz and I. B. Wilson, J. Biol. Chem. 277, 3245 (1970).
(4)
204
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
[15]
predicts that a reciprocal plot of 1/kobs vs 1/[I] will pass through the origin and that K~ cannot be evaluated. From a practical standpoint, this situation occurs often, since K~ values are in the millimolar range whereas inhibitor concentrations are frequently limited to lower values owing either to low solubility of the inhibitor or its high rate of reaction, is It should be pointed out that in a situation wherein kob~ is measured at one inhibitor concentration or where kobJ[I] does not vary in the range of inhibitor concentrations utilized, then it is not possible to demonstrate the existence of an enzyme-inhibitor complex. The most appropriate parameter for comparison of the reactivity of various inhibitors is the inhibition parameter k3/Ki. In the absence of sufficient data for determining k3/K~, kobJ[I] may be used. The differences in the magnitude of the numbers for related inhibitors reflect mostly the effect of structural changes on the binding of the inhibitor to the enzyme (K~) and on the rate of reaction within the bound complex (k3). It will be apparent that values of kob~/[I] are subject to distortion from nonlinear concentration effects when the inhibitor concentration is close to the K~ value of the inhibitor. Since most halomethyl ketones are only slightly soluble in water, the kinetics of inhibition are usually measured in a mixed solvent. The inhibitors are dissolved in such solvents as methanol or dimethoxyethane and then diluted with buffer. If the solution becomes cloudy or the haloketone precipitates at this point, either the concentration of organic solvent must be increased or the concentration of inhibitor decreased. Solutions of chloromethyl ketones in organic solvents are fairly stable, of the order of a week or more, but aqueous solution slowly lose inhibitory activity owing to hydrolysis. 16,1s A solution of Ac-Gly-PheCH~Cl was hydrolyzed 30% after standing for 20 hr in a buffer solution at pH 5.80. The rates of inhibition with halomethyl ketone are quite variable and depend on the pH, the enzyme, and the structure of the inhibitor. The fastest reactions have half-lives of less than 1 min whereas poor inhibitors may take many hours. The optimum in most cases is at pH 7 to 8, but rates are often measured at lower values to slow the reaction with reactive inhibitors to a more manageable range. Identification o] the Site o] Reaction. The alkylation of the active site histidine residue by a substrate related haloketone has been used often to locate a specific histidine in the primary sequency of a serine protease. However, haloketones are general alkylating agents, and if the reagent is too far removed in structure from a substrate, it is likely to alkylate nucleophilic sites other than the catalytic histidine. 42 In one 42K. J. Stevenson and L. B. Smillie, Can. J. Biochem. 46, 1357 (1968).
[1~]
SERINE PROTEASES WITH HALOMETHYL KETONES
205
case, alkylation of the catalytic serine residue of trypsin was observed when an inhibitor was synthesized with a decreased distance between the chloromethyl ketone group and a charged side chain relative to normal substrate or inhibitor. 43 However, this must be regarded as an exception, since it is likely that formation of a tetrahedral intermediate 8 by reaction of the catalytic serine with the carbonyl group of the inhibitor is a requirement for inhibition. The site of reaction can be identified by isolation of a peptide fragment containing the modified amino acid residue. This process is quite efficient when a diagonal electrophoresis technique is utilized. This involves electrophoresis of a peptide mixture on some support, treatment with H C Q H , and then electrophoresis in a direction perpendicular to the first. Oxidation of histidine alkylated by a halomethyl ketone yields earboxymethyl histidine. Disulfides are also oxidized to cysteic acid residues. All peptides that contained cystine residues or amino acid residues alkylated by the haloketone will lie off the diagonal after the second electrophoresis and can easily be isolated. ~-~G A modification of this technique allows separation of only the peptide containing the alkylated amino acid residue. ~7 The alkylated protein is first reduced, then eyanoethylated, and the eyanoethylated cysteine residues are oxidized to the sulfone. This change is not sufficiently large to remove such peptides from the diagonal, and only the peptide with the alkylated amino acid residue is found off the diagonal.
Applications Representative examples of halomethyl ketones which have been used as affinity labels for serine proteases are listed in the table. Several thiol proteases (papain, clostripain, cathepsin B, ~ and cocoonase) are also listed since the mechanism of the inhibition is probably similar for the two classes of enzymes. The inhibitors chosen for the table were either the most readily available or among the more reactive for a particular enzyme. For more information the reader is referred to a recent comprehensive review of all the enzymes that have been studied with halomethyl ketone inhibitors2 Most of the potential applications of affinity labeling have been realized by the use of halomethyl ketones. They have been used to label '~ D. D. Schroedcr and E. Shaw, Arch. Biochem. Biophys. 142, 340 (1971). ~ K. J. Stevenson and L. B. Smillie. J. Mol. Biol. 12, 937 (1965). ~5L. B. Smillie and B. S. Hartley, Biochem. J. 101, 232 (1966). 46K. J. Stevenson and L. B. Smillie, Can. J. Biochem. 46, 1357 (1968). 47K. J. Stevenson, Biochem. J. 139, 215 (1974).
206
[15]
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
INHIBITION OF PROTEOLYTIC ENZYMES WITH HALOKETONE DERIVATIVES OF AMINO ACIDS AND PEPTIDES
Enzyme Chymotrypsin Cathepsin G Subtilisin
Streptomyces griseus protease B Elastase
Trypsin
Acrosin Thrombin
Plasmin Kallikrein Carboxypeptidase Y Carboxypeptidase C Acid carboxypeptidases Papain Clostripain Cathepsin B x Cocoonase
Inhibitor
References a
TosPheCH,CI Z-Gly-Leu-PheCH2C1 Z-Gly-Leu-PheCH2Cl Z-PheCH~Br Z-Ala-Gly-PheCH~C1 Ac-Phe-Gly-Ala-LeuCH2C1 Phe-Ala-LysCH,C1 Boc-Gly-Leu-PheCH..C1 Ac-AlaCH~C1 Ac-Ala-Ala-AlaCH:C1 Ac-Ala-Ala-Pro-AlaCH2C1 Tos-LysCH~CI Tos-ArgCH~Cl Lys-Ala-LysCH2C1 Tes-LysCH2CI Tos-LysCH2C1 Phe-Ala-LysCH~C1 Glu-Ala-LysCH~C1 Tos-LysCH~C1 Z-LysCH2C1 Ala-Phe-LysCH2C1 Z-PheCH2C1 Tos-PheCH2C1 Tos-PheCH,C1 Tos-PheCH2Cl Z-Gly-Leu-PheCH2C1 Tos-LysCH,C1 Tos-PheCH2C] Ac-Ala-Ala-Ala-AlaCH~C1 Tos-LysCH..C1
5 18 38 b 10 11 19 c 15 16, d 16, 17, e,f 34, g g 19 h-j k 19, l 19, l m, n 19, o 19 p, q r s, t u, v w, x y z aa
a Numbers refer to text footnotes. b F. S. Markland, E. Shaw, and E. L. Smith, Proc. Natl. Acad. Sci. U.S.A. 61, 1440 (1968). c A. Gertler, FEBS Lett. 43, 81 (1974). d j. Travis and R. C. Roberts, Biochemistry 8, 2884 (1969). J. C. Powers and P. M. Tuhy, J. Am. Chem. Soc., 94, 6544 (1972). I p. IV[. Tuhy and J. C. Powers , FEBS Lett. 50, 359 (1975). N. Yoshida, A. Sasaki, and K. Inouye, Biochem. Biophys. Acta 321, 615 (1973) i, L. Zaneveld, K. Polakoski, and W. Williams, Biol. Repro& 6, 30 (1972). i L. Zaneveld, B. Dragoje, and G. Schumacher, Science 177, 702 (1972). i K. L. Plaskoski and R. A. McRorie, J. Biol. Chem. 248, 8183 (1973). k T. M. Chulkova and V. N. Orekhovick, Biokhimiya 33, 1222 (1968). ~E. Shaw, in "Proteinase Inhibitors" (H. Fritz, H. Tschesche, L. J. Green, and E. Truscheit, eds.), pp. 531-540. Springer-Verlag, Berlin and New York, 1974.
[16]
SERINE PROTEASES WITH HALOMETHYL KETONES
207
F. F. Buck, B. Hummel, and E. DeRenzo, J. Biol. Chem. 243, 3648 (1968). W. R. Groskopf, B. Hsieh, L. Summaria, and K. Robbins, J. Biol. Chem. 244, 359 (1969). ° C. Sampaio, S. Wong, and E. Shaw, Arch. Biochem. Biophys. 165, 133 (1974). R. Hayashi, Y. Bai, and T. Hata, J. Biochem. (Tokyo) 76, 1355 (1974). q R. Hayashi, Y. Bai, and T. Hata, J. Biol. Chem. 250, 5221 (1975). r R. W. Kuhn, K. A. Walsh, and H. Neurath, Biochemistry 13, 3871 (1974). E. Ichishima, S. Sonoki, K. Hirai, Y. Torii, and S. Yokoyama, J. Biochem. (Tokyo) 72, 1045 (1972). T. Nakadai, S. Nasuno, and N. Iguchi, Attic. Biol. Chem. 37, 1237 (1973). S. S. Husain and G. Lowe, Chem. Commun. 1965, 345 (1965). " M. L. Bender and L. J. Brubacker, J. Ant. Chem. Soc. 88, 5880 (1966). S. Roffman, M. Levy, and W. Troll, Fed. Proc., Fed. Am. Soc. Exp. Biol. 32, 466 (1973). S. L. Roffman, Diss. Abstr. Int. 35B, 3218 (1975). W. H. Porter, L. W. Cunningham, and W. M. Mitchell, J. Biol.Chem. 246, 7675 (1971). M. C. Burleigh, A. J. Barrett, and G. S. Lazarus, Biochem. J. 137, 387 (1974). ~ F. Kafatos, J. H. Law, and A. M. Tartakoff, J. Biol. Chem. 242, 1488 (1967). and locate the active site histidine in numerous serine proteases including chymotrypsin, trypsin, subtilisin, thrombin, and elastase. H e a v y atomcontaining halomethyl ketones have been synthesized for possible use in crystallographic phasing during the determination of the three-dimensional structure of proteins. ~s Indeed, the determination of the structures of chymotrypsin and subtilisin t h a t had been treated with halomethyl ketones led to the discovery of the extended substrate binding site in these enzymes2,9,1~ Numerous kinetic studies with chloromethyl ketones have shown the existence of extended binding sites in other proteases or have sought to determine the nature of the individual subsites.~°, l~,15--~° H a l o m e t h y l ketones have been used to place either spectroscropic, 49 fluorescent, ~° or spin-labeled 51 reporter groups in the active sites of serine proteases. A chloromethyl ketone has been used in a nuclear magnetic resonance study of catalytic residues of a serine protease. ~2 More recently, halomethyl ketones have been extensively used in studies of role of proteases in such diverse biological processes as fertilization, cell growth, protein synthesis, virus maturation, and diseases such as e m p h y s e m a and cancer. 6 However, in m a n y cases appropriate controls were not carried out and it is impossible to discern whether the effect of the haloketone is A. Tulinsky, R. L. Vandlen, C. N. Morimoto, N. V. Mani, and L. H. Wright, Biochemistry 4185 (1973).
,9 D. S. Sigman and E. R. Blout, J. Am. Chem. Soc. 89, 1747 (1967). 50G. Schoellmann, Int. J. Peptide Protein Res. 4, 221 (1972). 51D. J. Kosman, J. Mol. Biol. 67, 247 (1972). 55G. Robillard and R. G. Shulman, J. Mol. Biol. 86, 519 (1974).
208
ENZYMES, ANTIBODIES, AND OTHER PROTEINS
[17]
due to specific inhibition of a serine protease or to other alkylation reactions. In the future, there is an excellent possibility that some of these affinity labels may be used medicinally in the control of both normal and abnormal physiological processes involving serine proteases. [17] R e a c t i o n o f S e r i n e P r o t e a s e s w i t h A z a - A m i n o A c i d a n d Aza-Peptide Derivatives
By
JAMES C. POWERS and B. FRANK GUPTON
Aza-amino acid residues are analogs of amino acids in which the a-CH has been replaced by a nitrogen atom. 1 This substitution has a profound effect on the reactivity of aza-peptides, i.e., those containing ~NH~CH--CO--
I R
~NH--N~CO
I
R
aza-amino acid residues. In particular, it has been shown that azapeptide p-nitrophenyl esters are inhibitors and active site titrants of several serine proteases. 2-4 Inhibition of these enzymes is believed to arise from the acylation of the active-site serine residue yielding an acylated enzyme (Fig. 1), which is substantially less reactive toward deacylation than a normal acylated enzyme owing to the influence of the adjacent nitrogen atom. At present, there is no evidence that the activesite serine residue is actually the site of reaction. However, the close structural resemblance of aza-peptides to normal peptides substrates of serine proteases points strongly to the serine residue as the one being labeled. The only requirement for design of an aza-peptide reagent for a new serine protease is knowledge of the enzyme's substrate specificity so that an appropriate side chain may be placed on the aza-amino acid residue. The employment of p-nitrophenol as a product of the acylation step is advantageous since it is a good leaving group, and the release of p.nitroThe use of standard IUPAC nomenclature (2-substituted carbazic acid derivatives) for the naming of aza-amino acid residues is particularly disadvantageous since the names are cumbersome and give no indication of the structural relationship with amino acids. For this reason, the method of naming these analogs in this text will be to precede the name of the corresponding amino acid with "aza." For example, 2-benzylcarbazoie acid p-nitrophenyl ester will be referred to as aza-phenylalanyl p-nitrophenyl ester. Likewise, the method of abbreviating these analogs will be to prefix the abbreviation of the corresponding amino acid with an "A." For example, the abbreviation of aza-phenylalanine will be Aphe. A. N. Kurtz and C. Niemann, J. Am. Chem. Soe. 83, 1879 (1961). s D. T. Emore and J. J. Smyth, Biochem. J. 107, 103 (1968). J. C. Powers and D. L. Carroll, Bioehem. Biophys. Res. Commun. 67, 639 (1975).
