Kovember 1967
IRREVERSIBLE ESZPME INHIBITORS.
solution of 0.73 g 6-Benzylaminouracil (16). Method B.-A ( 5 mmoles) of 2418 and 1.07 g (10 mmoles) of benzylamine in 100 ml of water was refluxed for about 18 hr. T h e cooled solution was filtered and the product was washed with water; yield 0.35 g ( 3 2 % ) , mp 313-314" dec. Three recrystallizations from aqueous AcOH gave whit,e cryst,als, mp 316-317" dec. The compound moved as a single spot, on tlc in 1: 3 AcoH-c~Hs. See Table 11 for additional data. 6-Phenylthiouracil (13). Method C.-To a solution of 1.10 g (10 mmoles) of thiophenol and 10 mmoles of S a O H in H20 (50 ml) were added 0.73 g ( 5 mmoles) of 2418 and 50 ml of 2methoxyethaiiol. After being refluxed for 12 hr, the solution was spin evaporated in vacuo. To the residue was added 50 ml of water, then the mixtiire was acidified (AcOH), and agaiii spin evaporated in z'acuo. The residue was heated to boiling with 100 ml of water, then cooled. The product was collected on a filter and washed wit.h hot water: yield 1.00 g (YO%), mp 267270". Recrystallization from E t O H gave white crystals, mp 270-272°, that moved as a single spot on tlc in 1: ,5 AcOH-CsHs. See Table I1 for additional data. Method D used for 9 was the same except the product separated directly on cooling the reaction mixture; in the case of 14, the reaction mixture was merely acidified with HC1 to precipitate t,he product. 6-Benzoyluracil (ll).--A mixture of 300 mg (1.5 mmoles) of 1, 330 mg ( 3 mmoles) of Se02, and 50 ml of AcOH was refluxed
1113
for 2 hr, then filtered to remove Se. The filtrate was spiii evaporated in vacuo. T h e residue was dissolved in 50 ml of water, then the solution was clarified by filtration; the product separated on cooling. Two more recrystallizations from water gave 100 mg (31%,) of light yellow needles: mp 250-252'; , , ,X 277 mp ( p H l), 257 mp ( p H 13). The compound moved as a single spot on tlc in 5 : 1 CsH;,-EtOA4c. Lang1ey2O has recorded mp 2522,53' for this compound prepared by a different route. 6 4 cY-Hydroxybenzyl)uracil (12).-A mixture of 300 mg (1.5 mmoles) of 1, 165 mg (1.5 mmoles) of SeO?, and 50 ml o f AcOH was refluxed 1 hr, then filtered t o remove Se. The residue remaiiiing after spin evaporation of the filtrate in mcuo was dissolved ill 50 ml of water. The hot solution was filtered, theii cooled. Filtration removed 50 mg ( 16Tc) of 11, mp 250-25'2'. The filtrate was concentrated t o aboiit 20 ml, then allowed to stand at 3". The product was collected on a filter: yield 120 mg (38%); mp 224-226'; , , ,A 264 mp ( p H l ) , 237 mp ( p H 13). The compound moved as a single spot on tlc in 5:1 C6H6-EtOA4c. Langley20 recorded mp 224-228' for this compound prepared by an alternate route. 6-(p-Methylbenzyl)uracil (5).-A mixture of 1.1 g (5mmoles) of 25,4119 0.40 g of chloroacetic acid, and 30 ml of water was refluxed for 48 hr with stirring. The cooled mixture was filtered and the product was washed with water. Recrystallization from EtOH gave 0.85 g (84%) of white crystals, mp 289-271'. See Table TI for additional data.
Irreversible Enzyme Inhibitors. CV.''2 Differential Irreversible Inhibition of Vertebrate Dihydrofolic Reductases by Derivatives of 4,6-Diamino-1,Z-dihydro-Z,Z-dimethyl-l -phenyl-s-triazines Substituted with a Terminal Sulfonyl Fluoride3 B. R.BAKERA N D
GERHARUUS
J. L O U R E S S 4
Department of Chemistry, Cnioersity of California at Santa BaTbara, Santa Barbara, Calijornia $4106 Receaved June 27, 1967 Derivatives of 4,6-diamirio-1,2-dihydro-2,2-dimeth~l-s-t~riazir~e bridged from its 1 position to sulfanilyl fluoride with six different bridges have been synthesized; these compounds have been evaluated as reversible and irreversible enzyme inhibitors of the dihydrofolic reductases from Walker 236 rat, tumor, rat liver, L1210/FR8 mouse leukemia, and pigeon liver. For each compound little difference in reversible binding to the four dihydrofolic reductases were seen. I n contrast, dramat,ic differences in irreversible iiihibition were seen. Foiir of the six compounds that irreversibly inhibited pigeon liver dihydrofolic reductase failed to irreversibly iuhibit the dihydrofolic reductases from Walker 256 rat tumor and L1210 mouse leukemia. The two compounds containing a p-benzoyl ( 15d) and a p-phenylpropionyl ( 15f) bridge irreversibly iiihibit,ed the two tumor enzymes and the pigeon liver enzyme. However, 15d inactivated the rat tumor >70 t,imes as fast as the mouse leukemia enzyme. Furthermore, 15f inactivated the rat tumor enzyme eight times as fast as the rat liver enzyme. The dihydro-st.riazine moiety of 15 is believed to complex within the active site of the enzyme, but the sulfonyl fluoride is believed to form a covalent bond outside the site; it is the latter area where evolutionary differences are more apt to have occurred. Thus, the differences in irreversible iiihibit,ion of these enzymes can be accounted for if these compounds are operating by the active-site-directed ezo mechanism of irreversible inhibition, such a mechanism account,iiig for the specificity pattern by t,he bridge principle of specificity.
Once it had been established that the hydrophobic. The discovery6t6 of a potent hydrophobic bonding bonding region was outside the active site, lo near region on dihydrofolic reductase considerably comwhere either the 4 or S position of dihydrofolate (1) plicated the successful design of the first active-siteon the enzyme, two active-site-directed irdirected irreversible inhibitors5l7 for this e ~ i z y r n e . ~ * resides ~ reversible inhibitors soon followed ;8,9 for example, the 5-phenylbutyl group of 2 complexes with the hy(1) This work was penerously supported b y G r a n t C,\-08695 from t h e National Cancer Institute, U. S.Public Health Service. drophobic bonding region, thus allowing the 6-phenethyl (2) For t h e previous paper of this series see 33. R. Baker a n d \V. Rzeszogroup t o project back into the active site.8 tarski, J . M e d . Chem., 10, 1109 (1967). (3) For t h e previous paper o n inhibitors of dihydrofolic reductases see R . R. Baker a n d M. A . Johneon, J . Heterocyclic Chem., i n press. (4) G. J. L. wishes to t h a n k t h e Council for Scientific a n d Industrial Research, Republic of South Africa, for a tuition fellowship. ( 6 ) For a reviea on t h e mode of binding of inhibitors t o diliydrofolic reductase, see J3. R. I3aker. "1)esign of ,~ctive-Site-nirecte[I Irreversilile
Enzyme Inhibitors. T h e Orvanic Chemistry of t h e Enzymic .Ictive-Site," .lolin Wiley and Sons, Inc., New York, N . Y., 1967, Chapter S. 16) 13. R. Raker. H.-T. Ho, a n d D. V. Santi, J . Phnrm. S c i . . 64, 1415 (1965).
( 7 ) l3. R . Baker, ihid., 69, 347 (1961). (8) B. R. liaker and .J. H. Jordaan, i b i d . , 66, 1417 (1966); paper L S V I I of ttii8 series. (91 13. R. Baker a n d 11. S. Stiapiro, ibi,l,, 66, 1422 (1966); paper L S V I I I of this series. (10) B, R. Ijaker, T. J . Scliwan, J. Koootny, anti B.-T. Ho, nbid., 6 6 , 295 (1966).
4a. S ('H 5a. S = 0
phenoxybutyl group of 5.*O I t had previously been demonstrated that when a benzyl group ~ a complexed 5 to the hydrophobic bonding region of dihydrofolic reductase, the more polar phenoxy group gave a 30fold loss in hi~idirig.~ T h e above two studiez were then reconverged onto oiic objectivc diagrammatically indicated in 7; if the proper bridge lengths, Bl, between the two phenyl groupb, and B?, between the outside phenyl arid the leaving group, I,, and if the proper leaving group could he found, then an active-site-directed irreverqible
I NH, 7 . B = bridge L = leaving group
inhibitor should emerge which could utilize the small differences in the hydrophobic bonding regiori of dihydrofolic reductases from different tissues. Initial work was focused o n the bromoacetamido arid chloromethyl ketone groups for the L group of 7, since these two groups have the electrophilic character to be able t o react with any one of seven out of the total of 1.5 different’ proteinic amino acids containing a third functional group. About 20 compounds with these tu-o le:iving groups and varying B1and Bzgroups were synthesized and evaluated; none showed irreversible inhibition of the dihydrofolic reductases from pigeon liver, W:dl I o ~ r r a l ,J . I ’ i i n r , m . Nci., 62, 840 f l U 8 : i ) : ( I , ) 1%. I. V. Sanii, 1’. I. .\IinaiiIa, and \\’, c‘. \\‘erkileiaer, ./. I l r , / . C h e m . , 7, 24 (1964): (e) U. R . Baker and .J. H , .Jordaan, i h i n . , 8 , : 3 5 (196.51; ((1) 13, R . Haker, T. ,J. Sellwan, anTl I cr, J r . , i/jf,/,, 66, >titi ( l ! l G i J ,
1116
9
8
i
16
nif'tu series :1. / I
1).
=I)
I1 =
1
C', 11 = 2
react ivitli titi aldehyde due t o inactivatioit 1)). thc. electroii-n-ithdra\viiig sulfoiiyl fluoride group; this lowered reactivity was previously noted with soin(' sulfariilaniide derivatives.:j4 Inhibitor Evaluation.--.hi :irt ive-site-directed irreversible inhibitor operates by first forming :L reversiblc: romplex with the enzyme. If a iiiic1eol)hilir g r o u ~ ) O I I the cwynie and it leaving g r o u ~ (I,) ) are juxta,po within t h e complex, then :i fast irrev t : t l w place that ir1wtiv:ites t h c ~c~nzj~rn
I~WWEI~SIBLE ENZYME INHIBITOHS. CV
Soveinber 1967
1117
-Irr?verml>le----
- _Reversible--1';nsylne suurcea
Estd K , 106 .\I
x
L1210
3 . .i
0.6
Pigeoii liver
1.3
0.2
kV2.56 Itat liver
0.046 0.060
0.008 0.01
5 .6 .5 . 6 18 6.3 1.3 0 . 1.5 0 . 18 0.30
1,1210
0.30
0 . 0.i
0 . 30 1 .:3
0.31
0.03
W2.56
0.064
0.01
Rat liver
0.028
0.003
L1210 Pigeoii liver
0,078 0.10
0.01 0.02
W256
21 600 144 0,020 0.097
0 . 020
0.06
j1.7f
0.003
0.001
Ll"0
0 . 080
0.01
l'igeoii liver
0.07
0.01
0.34
0.40
Large Large Large
4 100 24 0.003 0.016
0.0060
I'igeoii liver
1.6 0 . 3.i 0.3.7 0.32 0.21 0.14 0.14 0.21 0.21 2.i 2.i 2 .i 0.10 0.N 0.40 0.26 0.26 0.13 0.15 0.0.5 0 .o2 0.0,i 0.02 0.4 0.07 0.07 0.1 0 .os 2 23 2.i
Rat liver
\2'256 L1210 Pigeon liver
SH9
.M
0.2
Pigeon liver
I
F
1.1
FV2.56 L1210 Pigeoii liver JV256 L1210
13e
pJI
F2'2.56
Pigeon liver
Lid
150.
Inhibitor concn,
7
TP" concn,
Yo
p -11
Elb
60 0 60 30 60 60 60 0 60 60 30 0 60 60 0 60 0 60 60 0 60 60 60 60 60 30 0 60 0 60 60 30 60
120 120 60 120
97
97 91 87 42 97 !)i
96 97 88 97
97 97 90
88 20 .->0 97 98 97
97 98
!G
%
t l , ?.
inactiv
mine
tored at 3" and was stable for. months. For assay, it was diliited 1 : % O : 100 pI of this diliite s o l i i l i o t i i i i a total o f 1.00 ml gave ail 01) rhange of : i l ~ o i i t 0.01 0 1 ) iiiiit/miii \Theti ashayed as described below. C. Walker 256 Rat Tumor.-A mixture of 30 g of frozeti tumor42 mid 60 ml of ice-cold buffer A was blended for 2 miti at high speed i i i a prechilled small head of a \Taring bleridor, then centrifuged for 20 miii a t 20,000 rpm i n a no. 40 rotor of a Spiiico T, centrifuge. To the stirred supernatant ( i 3 ml) cooled in an ice bath was added 4.3 ml of 5% aqueous streptomycin. After
ndditioiinl 1 0 miit, the mixture wac c,etitrifuged rpni. TCJthe rtiwed siipei,iistaiit (65 ml) cooled i t i ail ire bath W R A added 17.9 g of enzyme grade ( S H 4 ) r S 0 4 over a period o f 1-2 miti (4,ic: of hatiiratioii). After being stirred 10 miti mcii'e, the mistiii,e was cetitrifiiged fat, 10 rniii at 20,000 rpni. To the stii,i,ed siiperiiataiit 162 ml) cooled i i i an ice bath was added 21 g of (XH4j,S04 (905; of saturatioii). After being stirred for an additional 10 miii, the mixtiire was centrifiiged for 20 min at 20,000 rpm. The siiperiiataiit was rejected aiid the pellets mere dissolved iii briffer A giving a final volume of 40 nil: this soliitioii cwiild be stored at 3" with little or no loss i n activity over 3 months. For assay, the h o l i i t i o t i wah diliited 1:2, theii .io p1 gave an 01) chaiige cif aboiit 0.01 iuiit/miii Liiider the cwndit i o i i h dehcribed below. If an extraiieoiih T P X H oxidase was preseiit, the soliitioii w:is iiiriihated at :37" for I hr, then chilled aiid filtered. D. Rat Liver.-The eiizyme w:is prep:ii'ed as de,scribed for the rat tumor. The f i l i a l volume \KIA 35 nil: this solutioii was *table at 3" over 6 moiiths. For ascay, t h e solution was diluted 1 : s ; ,ii) p1 of t h e diliited soliiticiii gave NII ciptival delisit)- rhaiige of aboiit 0.01 iiiiit/miti wheti amiyed as desci,ihed below. Reversible Inhibition of Dihydrofolic Reductases.-A 3.72 m J l solutioii of T P N I I iii 0.01 .I/ Sa011 was sufficieiitly .-table t o be stored oiily 4 days at 3'. FOI.eiizyme away 0 . 2 0 ml was diluted with 1.04 ml of hiiffer B to give a 600 p.11 soliitioti: this soliitioii shoiild be .tored at 0" and i,eiiewed evei'y 4 hi.. 1)ihydrofolate was the 1.82 m J l siihpetisioti above diliited with biiffer B to give a 120 pJ1 h o l i i t i o i i . This soliitiori shoiilti he stored at 0' and renewed every 4 hr. 1 1 1 a I-nil ciivette were placed 0.85 nil of biift'er B, 5 0 pl of 600 p.lI TPNT-I, atid 50 p1 of eiizyme. After the sy.;tem had balanced, .i0 p1 of 120 p J f dihydi,cifiilate was added and the decrease iti optical deiihity at 340 nip was followed oti B 0-0.1 slide wire range of a Gilfoid 2000 o r C;ti,y 11 rpec.ii~o~hutunieter. Siiticietit enzyme shoiild be iised i o give 0.008-0.0li) 0 1 ) iinit chaiige/miii. .Ally sniall de(-lease 1 o p t i c d tleiisity p h i , t o :idditioti of tlihydrofolate is sutitrar1etl l t ~ i ~the n etizynir ],ate. This w t e wit t i o i i t inhitiitor is termed V,);the ciivatle [*[iii(,~iitt,atiotis were (i pJ1 i n dihydrufolate atid :$O w-11 iri T P N H . Inhibitors were diwilved i i i 1 m.11 HC1: if not solrthle, a 1 1 cc!iinl voliime of 1)3IF w-iih added, theii the a o l i i t i o i i was tiiliitcd with 1 in.l/ IICI. Fi,esh sriliiiioiis wet'e pi,ep:ii,ed daily. To tlcbeitig *tirred
mi
for 10 m i t i at 20,Oi)O
II~KEVERSIBLE ESZIME ISHIBITOI~S. CY
Sovember 1l)tii
no inactivation of the enzyme was converted to the diamino-s-triazine, 15a, which could rapidly inactivate the enzyme; therefore 15a could not have contained trace metals arising from the sulfanilyl fluoride, else 16a should also have inactivated the enzyme. Further verification was obtained by synthesis of the ethanesulforiate salt of 15c, the latter showing the same rate of inactivation as the hydrochloride salt of 15c; since the solubilities of these two salts are quite different, it is unlikely the\- would have an identical quantity of a given trace metal after purification. (e) h third mechanism for enzyme inactivation by irreversible inhibitors ha5 been observed with tryp~ i n ; ~when ~ 8 the ~ ~appropriate inhibitor is reversibly complexed to the enzyme, a conformational change can sometimes occur which exposes an enzymic nucleophilic group to bimolecular attack as shown in eq 4. L-1.. . E + 1-L +L-I.. . .Z-I (4) That this mechanism was not operational was shon-n by incubating the enzyme with a. combination of 19. which does not have the sulfonyl fluoride group, and 16a, which doe< not have the diamino-s-triazine moiety. Sirice no inactivation was been, it is highly improbable that 15b formed a reversible complex that was then attacked bimolecularly b\- a second molecule of 15b. Thus, the active-site-directed mechanism is the onl\possible mechanism yet thought of that is compatible with the data i n Table I. Walker 256 Enzyme.-Of the six different bridges (15) between the dinmino-s-triaziiie binding to the active site and the sulfanilyl fluoride group, only the p-benzoyl (15d) arid p-l)henyll)rol)iotiyl (15f) bridges gave molecules that allowed irreversible inhibition of the dihydrofolic reductase from this rat tumor (Table I). A relatively high concentration of 23 p J 1 of 15d had to be employed due to the relatively poor K , of this compound. That 15d did not inactivate the enzyme by the random bimolecular mechanism of eq 3 was show1 by the lack of inactivation of this tumor enzyme by 25 p-11 16a, which lacks the diaminos-triazine needed for complexing to the active site, but is still capable of bimolecular attack. There are two requirements for an active-site-directed irreversible inhibitor to be effective oii an enzyme system, as caii be seen from eq 1 : (a) the loxer the reversible dissociation constant the less coticentratiori of inhibitor is needed to give an effective amount of the rate determining [ E . . . I ] species, and (b) the faster the rate of covalent bond formation within the complex, the more effective the inhibitor. If both parameters are good, the amount of random reaction with other proteins or isozymes in the host will be decreased. S o l e that the p-benzoyl-bridged compound ( 1 5 ~meets ) only one of these two requirements; 15c caii rapidly inactivate the enzyme in less than 1 min, but a relatively high concentration of 4-2ri X lovfi31 nould be needed to form the required amount of [ E . . . I ] complex. 111 contrast, the p-phenylpropionyl bridged compound (15f) meets both requirements; it can inactivate this rat tumor enzyme i n less than 2 mill :it a n inhibitor concentration of 2-,5 x 10-8 J1. The next requirement for an active-site-directed irreversible inhibitor to be effective on a tumor in ~
(45) l'. Inagami. J . B i d . Chem., 240, PCdl33 ( 1 5 6 5 ) .
1121
vivo is to have sufficient isozyme specificity so that little or no attack occurs on the target enzyme in normal host tissues. The effect of 15f on the dihydrofolic reductase from the liver of rats bearing the Walker 256 tumor n a s investigated. S o t e first that 15f differed in reversible binding to the dihydrofolic reductase from the two tissues about threefold. Secondly, rat liver enzyme was also rapidly inactivated by 15f at 37". When the inactivation rate was sloned at 2s" so that it could be measured, the rat liver enz! me was half-inactivated in S min in the absence of dihydrofolate, but the tumor enzyme was half-inactivated in 2 miri i n the presence of G p J 1 dihydrofolate; since dihydrofolate at this conceiitration ~t-ouldprotect against iriactivatioii of 15f by a factor of t\To, it can be estimated that 15f inactivates the tumor enzyme about eight times as fast as the liver enzj me. This eightfold difference is insufficient for effective chemotherapy, but does indicate that the bridge principle of specificity is partially operational. By further modification of the bridge of 15f, of the hydrophobically bonded 1-phenyl group, of the environment of the leaving group, or of combinations of these, it ihould be theoretically possible to separate these rates a 1000-fold or more. S o t e that a higher degree of specificity is already present with the same tissue from two different species; namely, rat liver arid pigeon liver. At near equal concentration, 15b can inactivate pigeon liver enzyme with a half-life of less than 2 min, but the rat liver enz\me shows no inactivation in 60 min. Sirice 1.iYc irreversible inhibition i< readily detectable, the half-life of inactivation of rat liver is greater than 250 min; therefore, 15b inactivates the pigeon liver isozyme more than 120 times as fast as the rat liver isozyme. L1210 Enzyme.-This isozyme had an irreversible inhibition profile c.oincidentally (?) similar t o the rat tumor isozyme with the seven compounds in Table I ; that is, the L1210 dihydrofolic reductase was inactivated by 15d and 15f, but not by the other analogs of 15 or by 16a lacking a triazine moiety. However, the rates of inactivation by 15d were different; corrected for the same amount of [E. . .I] complex, 15d inactivated the rat tumor enzyme greater than 1.5 times as fast as 1,1210 enzyme. With the same 23 11-11concentration of 15d, the rat tumor enzyme was iriactivated greater than 6h times more rapidly than 1,1210 enzyme, onefourth of this difference being due to the relative abilities of the two isozyme t o complex reversibly with E d . The diamino-s-triazine bridged by p-phenylpropionyl (15f) \vas a good irreversible inhibitor of L1210 enzyme; at a concentration of 4 X lop7 M ,15f inactivated the enzyme with a half-life of less than 1 min. Algainnote that this inactivation is not due to a random bimolecular reaction since 250 X 10-' 11 16a failed to \how irreverqible inhibition of the 1,1210 isozyme. Other Observations.-Little difference i n the rates of iiiactivation could be seen whether 01' not TPSH wai ])resent; this alqiareiit iiidependence of TPSH coilcentration is mnvenieiit for chemotheral)eut ic purposes since higher or lower TPX" ( ~ o t i ~ e i i t r ~ ~i ltlithe ( ~ i target i\ vel1 will riot effect the rate of inactivation. Hon-ever, this independence of TPSH coiicentratioii is rather surprisiiig in view of the diff ererice 111 reversible binding
