.Jnnunry 1967
IRREVERSIBLE
EKZYME
IXHIRITORS.r , x x I I I
59
Irreversible Enzyme Inhibitors. LXXIII.'j2 Inhibitors of Guanine Deaminase. I. Mode of Binding of Guanine B. R. BAKER Department of Chemistry, University of California, Santa Barbara, California 93106, and / h e Department of Medicinal Chemistry, State University of New York at Buffalo, Riifalo, S e w York Received August 26, 1966 Investigation of guanine and 17 related compounds as substrates or inhibitors of guanine deaminase has led to a suggested mode of binding of guanine. It is highly probable that the 1- and 9-hydrogens of giianine are complexed to the enzyme as electron acceptors and the 6-oxo and 7-nitrogen are complexed as electron donors. I t appears that the x cloud, as well as the 2-NHz group of guanine, does not complex to the enzyme. It is possible that the 3-nitrogen of guanine as an electron donor is complexed to the enzyme.
of dihydrofolic reductase.8 In contrast, active-sitedirected irreversible enzyme i n h i b i t o r ~ can ~ ~ ~show ~ tissue specificity; for example, little selectivity with reversible inhibitors of lactic dehydrogenase from skeletal muscle or heart could be shown. but the active-sitedirected irreversible inhibitors could be constructed that did show selective inactivation between these two iso0 0 zymes." An active-site-directed irreversible enzyme inhibitor operates by first forming a reversible complex, then a N slower formation of a covalent linkage between the H H H enzyme and the inhibitor in a facile neighboring group I I1 reaction which leads to inactivation. In order to design an active-site-directed irreversible inhibitor, a de0 S 0 finitive modus operandi has been gradually developedg lo which leads to a greater probability of success; if a covalent forming group is placed on the potential irreversible inhibitor in any position that will interfere I11 Iv V with reversible complex formation, then the facile neighboring group reaction is automatically negated.g*10 tially purified enzyme from rabbit liver is commercially With an uninvestigated enzyme, such as guanine deavailable; it is readily assayed by the rate of decrease aminase, the following 711 odus operandi should be emin optical density a t 245 mp, where guanine has an abployed.1° (a) The groups on the substrate or an insorption maximum and xanthine has a m i n i m ~ m . ~ hibitor necessary for binding should be determined ; Neither guanosine nor 5'-guanylic acid are substrate^.^ a binding group can be eliminated if the binding by a Some years ago it was p r o p ~ s e dtha.t ~ ~ , the ~ selective second group can be increased sufficiently. (b) An area action of thioguanine (IV) on cert,ain t'umors could be on the inhibitor should be found that is not in contact due to the lack of guanine deaminase in these suscepwith the enzyme when the inhibitor forms a coniplextible cell lines; thus, when thioguanine could not be deso-called bulk tolerance. (c) Groups of varying lengths toxified to thioxanthine4 within a cell, cell deat,h reterminating in a covalent-forming moiety such as brosulted. If guanine deaminase could be inhibited selecmoacetamido should be placed on the noncontact area tively in a t,umor cell line with minimum blockade of of the inhibitor, then investigated for inactivation. =1 this enzyme in normal tissues, then such an inhibit'or study of the first phase with guanine deaminase is the could be used in conjunction with thioguanine with cell subject of this paper; in the following paper'z a study lines otherwise less affected by thioguanine. Ordinary of the second phase is presented. reversible inhibitors rarely show tissue or species speciThe inhibitory properties6 of 5-aminoimidazole-4ficity unless some nonfunctional part of t,he enzyme is carboxamide (V) mere checked; T' was complexed to employed for binding, such as the hydrophobic region the enzyme about half as well as the substrate, guanine (1) This work was generously supported by Grants CA-05867 and C.4(I) (see Tables I and 11). This result indicated that 08695 from the National Cancer Institute, U. S.Public Health Service. the 2-NHzC moiety was not necessary for binding. (2) For the previous paper of this series see B. R . Baker, T.J. Schwan, and B.-T. Ha, J. Pharm. Sci., in press. The series on Nonclassical AntiThat the 7-nitrogen of guanine was complexed as an metabolites and the series on Analogs of Tetrahydrofolic Acid have been electron donor to the enzyme was indicated by the lack combined into one series since they have a common objective, namely, the of inhibition of 3-aminopyrazole-karboxamide (VI) ; design of active-site-directed irreversible inhibitors. A collected list of Guanine deaminase (guanase) is a catabolic enzyme that converts guanine (I) to xanthine (11); 3 the enzyme reaction appears t o be irreversible. 8-hzaguanine (111),3thioguanine (IT'),4 and 1-methylguaninej are known substrates. The most potent known inhibitor is 3-nminoimidazole-4-carboxan~ide (Y) .6 The par-
referenceswillbe sent on request. (3) A. Roush and E. R . Norris, Arch. Biochem., 29, 124 (1950). (4)(a) E. C. Moore and G . A. LePage, Cancer Res., 18, 1075 (1958); (b) A. C. Sartorelli, G. A. LePage, and E. C. Moore, ibid., 18, 1232 (1958). ( 5 ) G . H. Hitchings and E. A . Falco, Proc. N a t l . Acad. Sci. L'. S., 30, 291
(1944).
(6) H. G. Mandel, J. L. Way, and P. K . Smith, Biochim. B i o p h y s . Acta, 23,402 (195i). (7) R. R . Baker, Cancer Chemotherapy Rept., 4, 1 (1959); paper I of this
aeries.
(8) B . R . Baker and B.-T. Ho, J . Pharm Scz., 66, 470 (1966). (9) B. R . Baker, zbzd., 67, 347 (1961), a re1 ieir (10) B . R Baker, "Design of Actir e-Site-Directed Irreversible Enzyme Inhibitors. The Organic Chemistry of the Enzsmic i c t w e Site," John Wiley and Sons, Ino., New York, N . Y , in press. (11) B . R. Baker and R . P . Patel. J . Pharm. Scz , 53, 714 (1961); paper XV of this series. (12) B. R . Baker and D . V. Santi, .I. ,'tfllrd. Chem , 10, 62 (1967); paper LXXIV of this series.
.
II I
\
\.I11
,-,II
2,-1ii
0'
s
SI
250
-It1
I)
,January 1967
IRREVERSIBLE ENZYME INHIBITORS. LXXIII
sirice the pyrazole ring is a weaker base than imidazole, it is possible that this wealteniIig of the base strength c~)ultlclecwrlw hiriding if the T cloud of the ring system were a donor. That the latter interpretation is unlikely is indicated by the following. (a) 7-.\Lethylguariiiie (XI) iq not a n inhibitor. If the K cloud of the imidazole moiety were a donor, then XIshould bind to the erizynie at least as well as 9-methyl gunnine (X). (ti) If the lone-electron pair of the 7-nitrogen of guanine iwre a donor, then the more basic this nitrogen is, the better the complexing should be and tiice cersa. S o t e that S-hydroxyguanine (XIII) actually has an acidic S H group a t the 7 position and S-bromoguanine (XII) has a base-weakened i-nitrogen; both compounds are poor inhibitors. Although 2,6-diaminopurine (YII) ar;d 2-amino-6-methylthioguanine (YIII) have a 7-nitroqcn that is more basic than that of guanine. IT1 ? i d 1711 are riot inhibitors; in these two cases the SHC=O grouping a t 1,6, which are probable binding points. has been changed, also indicating that the 1.6 grouping roiltributes more to binding than the 7-ni t rogen. That the SHC=O moiety a t 1,6 complexes to the enzyme with the hydrogen as a n acceptor and the oxygen as a donor is indicated by the following. (a) Removal of the SHC=O moiety, as in 2,B-diaminopurine (T'II) and 2-amino-6-methylthiopurine (T'III) results in no detectable binding. (b) Arethylation of the 1-nitrogen of guanine gives a compound (IX) that is still a substrate, but requires a much higher concentration than guanine for saturation of the enzyme, thus indicating that IX does not complex a ~well ; as guanine due to loss of the I-hydrogen in
IX. (c) Conversion of the 6-oxo group of guanine (I) to 6-thione (IY) still allows substrate properties (Table 11), as previously noted.4 That such a structural change leads to a nine-fold loss in binding can be seen by comparing the 9-phenyl derivatives, XY and XT'I; this result indicates that the 6-oxo group, as a donor, complexes with the enzyme, sirice the 6-thione is a poorer donor. If only the &oxo group of the KHC=O moiety were complexed, then an even bigger diff ererice in binding might have been anticipated; however, sirice the 1-hydrogen is apparently binding as an acceptor, it ~vouldbe more acidic. and a better acceptor with a 6thiorie group, thus partially compensating for the decarease in binding of the 6-thione group. That the 3-nitrogen of guanine (I) may complex as a donor to the enzyme is indicated by the lack of binding by xanthine (11) which has an acidic 3-SH group; this iriterpretation is highly equivocal sirice the decrease in basicity of the 7-nitrogen of xanthine (11) compared to guanine (I) (wild also account for this result. If the 5amino group of the imidazole-4-carboxamide (Y)were removed and tested as an inhibitor, lesser binding could also be due to a decreased basicity of the nitrogen corresponding to the 7-nitrogen of guanine; thus, it might be difficult t o determine with any greater degree of certainty that the 3-nitrogen of guanine is complexed to the enzyme. T h a t the 9-hydrogen of guanine is complexed as an accaeptor to enzyme is indicated by the 21-fold loss in binding when the 9-hydrogen is replaced by methyl, as
61
in 9-methylguanine (X); removal of i-methylat#ion, as in XI, gives an even greater loss in binding. These results might also be interpreted as due to a lack of bulk tolerance for either group within the enzyme-inhibitor complex, but such an interpretation is less likely. Replacement of the 9-hydrogen by phenyl (XY) gives an excellent inhibitor; that this result is most, probably due to a hydrophobic interactioii with the enzyme, combined with some additional forces, is discussed in the accompanying paper.'? Replacement of the S-CH of guanine by nitrogen, as in 8-azaguanine (111) (Table 11).gives a sevenfold loss in binding as noted by their relative K , values. This result is substantiated by comparison of their 9-(pchlorophenyl) derivatives, the guanine derivative (XYII) being a fivefold better inhibitor than the corresponding 5-azaguanine derivative (XYIII). The poorer binding by the 5-azaguanines is more apt to be due to the iveaker basicity of the 7-nitrogen, since L'triazole is a weaker base t'hari imidazole, than due to binding by the 5-CH group. The contribution to binding to guanine deaminase by groups on guanine (I) can be summarized in several classes: (a) those groups that most probably complex t'o the enzyme include t'he 1- and %hydrogens, the 6oxo, arid the 7-nitrogen-the first two as electron acceptors arid the last two as donors; (b) t,hose groups that' most probably do not, complex to t8heenzyme include the cloud system, the 2-SH2C moiety and the 8-CH group; those groups that might be complexed to the enzyme such as the 3-nitrogen as a donor where insufficient' evidence is available to make the probability higher. ((8)
Experimental Section Guanine deaminase (guanase) was a rabbit liver preparation purchased from Sigma Chemical Co. as a 1-mg/ml suspension: at' this concentration it was reputed t o deaminate 0.1 pmole of guanine/min. T h e suspension was stable over 4 months a t 2-5" ; for assay, 50 pl of bulk enzyme was diluted with 1.95 ml of 0.05 .V Tris buffer (pH 7.4), which could be kept at room temperature for a day's run. Guanine (15 mg) was dissolved in 1.00 ml of 1 .V KOH, then diluted to 100 ml Kith water. For assay, 1 ml was diluted with 14 ml of water t o give a 66.7 p M solution. T h e assay was performed as follows. I n a 1-ml cuvette was placed 0.70 ml of 0.05 JI Trih buffer (pH 7.4) and 200 p l of 66.7 pLlf guanine. T h e reaction was started b y addition of 100 pl of diliited enzyme solntion: the decrease in optical density a t 245 m p was recorded continuoiisly with a Gilford 2000 spectrophotometer, being about 0.005 optical density imit/min. The final concentration of guanineuTaq 13.3 p M . " Inhibitors B-ere diwolved in i a l 0.05 Jf Tris buffer IDH 7.4) if soluble, i b ) at 100 mlM in 1 -1-KOH. then 100-fold di1i;tion with Tris buffer, or (c) if base insolitble, in DMSO. In the latter case the assay was nin in l0C; DAISO t,he rate of reaction being identical with or without the presence of DAISO. T h e concentration for SOTc inhibition was determined by testing a series of inhibitor concentrations giving 30-i0% inhibition: when V o / V ~ was plotted against the inhibitor concentration, [I], where Vo = velocity without inhibitor and VI = velocity with inhibitor, the 50% inhibition concentration wax obtained where V o / T r=~ 2.13
Acknowledgment.-The author wishes to t'hank Professor Roland I
