Relationships between the structure of organophosphorus compounds and their activity as acetylcholinesterase inhibitors*

Bull. Org. mond. Sante 1971, 44, 31-42 Bull. Wid Hith Org. Relationships Between the Structure of Organophosphorus Compounds and their Activity as Acetylcholinesterase Inhibitors* T. R. FUKUTO 1 This paper discusses the relation between chemical structure and inactivation of the enzyme acetylcholinesterase (AChE) by organophosphorus esters, in terms of reactivity and steric effects. The discussion is centered in organophosphorus esters of the type (R) (R')P(O)X, whereX is a readily displaceable group and R and R' are various com- binations of alkyl, alkoxy, alkylthio, and amido moieties. Specific examples illustrating the effect of AChE inhibition on the selective toxicity of organophosphorus esters for insects and mammals are also presented. The inactivation of acetylcholinesterase (AChE) by organophosphorus esters has been demonstrated to be the result of an actual chemical reaction between the enzyme and the phosphorus compound (O'Brien, 1960; Heath, 1961; Fukuto, 1957). The overall inhibition process, leading ultimately to the formation of a covalently bonded phosphoryl- ated enzyme, may be depicted by the equations in Fig. 1-i.e., the ester and enzyme first combine to form a complex and this is followed by phosphorylation. The model for the active site of AChE proposed by Krupka (1964) is used here for the purpose of illustration. In this scheme, B is a basic group (histidine imidazole nitrogen), OH is a serine hydro- xyl, HA is an acidic group (tyrosine hydroxyl), and S is the anionic site. The anionic site, whose normal function is to attract the trimethylammonium moiety of the natural substrate acetylcholine (ACh) to the active site, may be visualized as a flexible pouch in which resides a negative charge. In certain cases a group in the organophosphorus ester- e.g., R in the 3-phenyl position-may interact with the anionic site and either aid or hinder the inhibition process. By using different kinetic methods, the bimolecular inhibition constant ke, the equilibrium or affinity constant for enzyme- inhibitor-complex formation Ka, and the phosphoryl- * This study was supported in part by US Public Health Service Research Grant FD 00239 from the Food & Drug Administration, Rockville, Md., USA, and by the Rockefeller Foundation, New York, USA. 1 Professor of Entomology and Professor of Chemistry, University of California, Riverside, Calif., USA. ation constant kp may be determined (Aldridge & Davison, 1952; Main, 1964). According to this mechanism of inhibition, phosphorylation of AChE takes place by serine hydroxyl attack on the phos- phorus atom, the reaction being catalysed by the basic and acidic moieties in the active site. The phosphorylated enzyme thus obtained is unable to catalyse the hydrolysis of ACh. The relation between the chemical structure of organophosphorus esters and the inactivation of AChE has been studied extensively. These studies have shown that the anticholinesterase activity of organophosphorus esters depends largely on the reactivity of the ester. For example, in pioneering work by Aldridge & Davison (1952) the inhibition of erythrocyte AChE by paraoxon and related substituted-phenyl diethyl phosphates was shown to take place in a bimolecular manner and a direct relationship was established between the inhibition constant ke and rates of solvolysis of these esters in phosphate buffer. Subsequently, it was demon- strated from a study of a larger series of substituted- phenyl diethyl phosphates that the inhibition of fly-head AChE by these compounds was related to the effect imposed by the substituent on the lability of the P-O-phenyl bond as estimated by Hammett's a constants, shifts in P-O-phenyl stretching fre- quencies, and hydrolysis rates (Fukuto & Metcalf, 1956). Fig. 2 shows the correlation between Ham- mett's cr constant, a parameter that provides an estimate of the electron-withdrawing or -donating properties of the substituent, and the logarithm 2614 31- T. R. FUKUTO Fig. 1 Inhibition of acetylcholinesterase by an organophosphorus ester* 0 \ ~~~~(R)N/ / P 0 NO2 -B + (RO) -B. N N<R /\ NO, /B A A 0 (OR), 0 (OR)2 P + HO/NO2 -B JBHOSN A A- R See text for explanation. of the I50 value (molar concentration required to Fig. 2 produce 50% inhibition of a given amount of enzyme Relation between log 150 for fly-head AChE and in 15 minutes). The linear correlation in Fig. 2 Hammett's sigma constant for substituted-phenyl clearly establishes a direct relationship between diethyl phosphates the reactivity of the organophosphorus ester and its anticholinesterase activity. Although reactivity is by far the most important factor that determines the anticholinesterase activity p-SOtCH3 of an organophosphorus ester, steric properties m8IM p-NO2 of the molecule often may significantly affect in- hibition rates. This is demonstrated in part by z mNO2 the unusually high activity of the m-dimethylamino c 7 p-CHO and m-tert-butylphenyl diethyl phosphate esters j p-CN p-SOCHincluded in Fig. 2. The effectiveness of these com- O m-NMe2 03 pounds as inhibitors is approximately 1000 times m-u the level that would be predicted from their reac- w 6 0 p-COON tivity parameter (a), and enhancement in inhibition has been attributed to interaction of the dimethyl- / amino and tert-butyl moieties with the anionic / X'5 site of the enzyme. Molecular models have shown 5 that the distance between the central atom in the Z p-SCHN /0 p-Clsubstituent and the electrophilic phosphorus atom '°r is very close to the distance between the quaternary 4 p-t-bu nitrogen and the carbonyl carbon in acetylcholine g -OCH3 (Fig. 3), and interaction with the anionic site probably ' / occurs by hydrophobic bonding (Hansch & Deutsch, p-OCH H 1966). 3 0-C30 There are numerous other examples of the in- fluence of steric effects on the rate of reaction be- / tween an organophosphorus ester and AChE. Fig. 4 -02 0 02AMMT 056 05 01O2NSAN 32 STRUCTURE-ACTIVITY RELATIONSHIPS FOR AChE INHIBITION BY OP COMPOUNDS Fig. 3 Structures of m-tert-butylphenyl O,O-diethyl phosphate (left) and acetylcholine (right) HiCHoP(C2H5)2 Hfe CH3 0 ,CH20CCC3 H3C CH3 shows the relation between log ke (bimolecular inhi- bition constant) for housefly-head AChE and log khyd (solvolysis constant) for a series of ethyl p-nitro- phenyl alkylphosphonates (Fukuto & Metcalf, 1959a). The plot shows that although there is a general trend -i.e., ke increases with an increase in khyd-the rela- tionship is not linear. Closer examination of the data shows that there is approximately a 20-fold drop in ke without a commensurate drop in khaYd when the alkyl group is increased in size from ethyl to n-hexyl, sugges- ting that the larger alkyl groups prevent the phospho- nyl moiety from making proper contact with the este- ratic site. That steric effects play a strong role in esta- blishing the anticholinesterase activity of these esters is supported by results from multiple regression analysis by Hansch & Deutsch (1966), who found Fig. 4 Relation between log ke for housefly-head AChE and log khyd for ethyl p-nitrophenyl alkylphosphonates * * 'CH 0 ntC-3I7 i-C5H1j S 0 n-C*H9 I-C6H13 4,4 -dimethyoentyI I &-C6$13 i-C4H9 S i-03H7 0 (C-H2)3P 9.0 1.5 2.0 LOG (Khd.X 105) * Reproduced, with permission, from Fukuto & Metcalf (1 959a). 3.0 2.51. * CH3 le2.C 10 w 1.5 0 0 -I.- 0 pherel 5- 1.0p. cyclohexy 0.5k 0.5 2.5 3.01 33 I a I I I I T. R. FUKUTO an improved correlation between ke and Taft's steric substituent constant Es (Taft, 1956), as follows: log ke = 3.738 E±+7.539 (1) Eq. (1) produced a correlation coefficient (r) of 0.901. The inclusion of other free-energy parameters in the regression equation had no effect on the correlation coefficient, indicating that variation in anticholinesterase activity is attributable mainly to steric effects. Perhaps a more striking example that illustrates the importance of steric effects in AChE inhibition is seen in data obtained from a series of methyl 2,4,5-trichlorophenyl N-alkyl phosphoramidates (Fu- kuto et al., 1963). Fig. 5 shows log ke plotted against Taft's polar substituent constant or*, a parameter that provides an estimate of the effect of the polar properties of substituents on the reactivity of the phosphorus ester (Taft, 1956). The figure shows that a linear relationship (solid line) is obtained except for the isopropyl and tert-butyl derivatives, Fig. 5 Relation between log ke for fly-head AChE and Taft's a* values for methyl 2,4 5-trichlorophenyl N-alkylphosphoramidates 8.0 I 7.0 N l 6.0 1- 5.0 F 4.0 3.0 0.4 -N(CI 1 V32 I ICH.TNH-0/-3 I t-C4H9N H- 0.6 ~ ~0. . . 0 0 0 34 -NH2 0.6 0.8 1.0 LIr * 1.2 1.4 I. 6 STRUCTURE-ACTIVITY RELATIONSHIPS FOR AChE INHIBITION BY OP COMPOUNDS 35 which were substantially less effective as anti- cholinesterases than was predicted from or* values. Further analysis of the data by Hansch & Deutsch (1966) resulted in Eq. (2), which gave considerable improvement to the correlation, with r = 0.939. log ke = 2.359Es-3.913 a*+4.948 (2) Fig. 6 shows a plot of the experimental values of log ke against the values of log ke calculated from Fig. 6 Relation between log ke for fly-head AChE and log ke calculated from (2.359 Es- 3.913 a*) for methyl 2,4,5-trichlorophenyl N-alkylphosphoramidates 7.0- H2 NHCH3 6.0 NHPr * NHEt NHBu Me2N * o 5.0o NH-i-Pr 4.0 NJi-t-Bu 3.0 I I I I I -2.0 -1.0 0 1.0 2.0 2.359 Es - 3.913 cr Eq. (2). Eq. (2), which includes Taft's steric substi- tuent constant Es, leads to an excellent linear re- lationship and provides strong support for the view that the bulky isopropyl and tert-butyl moieties decrease inhibition rates by steric interference. Another interesting example that shows the effect of steric factors on AChE inhibition is found in the data given in Table 1. Measurements of the inhibition of housefly-head AChE by a series of cyclic esters (p-nitrophenoxy-1,3,2-dioxaphosphori- nane 2-oxides) show that the AChE-inhibiting effectiveness of these compounds is approximately 1/100 000 times that of paraoxon (diethylp-nitrophe- nyl phosphate), in spite of their similar reactivity to hydroxide ion. Furthermore, the 2,4-dinitrophenyl es- ter (lV),' which underwent hydrolysis at twice the rate 1 Roman numerals indicate compounds identified by the same numerals in the tables and figures. Table 1 Second-order alkaline hydrolysis constants (koH) and anticholinesterase activity (Iso) for 6-membered-ring cyclic phosphate esters * No. Structure rn(M-1 f (M) 0\ 0/ \ < 1.56 > 1.3 x 10-3/ /\0 NO, o 0 11 t0/ o 0.69 >1.3 x 10-3 0/O _/ NO, O 0 lil >CO/ \o_No2 1.49 >1.3 x 10-3 0 / \~ NO, IV Co/ \oaNO2 3.10 1.3 x 10-3 V'0 / \ NO, Va* C2H50)2 11/0, 0.94 2.6 x 10-8 Data of Fukuto & Metcalf (1 965). a Compound V is paraoxon. of the corresponding mononitro compound (I), show- ed little increase in anticholinesterase activity. Since organophosphorus esters of the same order of susceptibility to hydrolysis as paraoxon are gene- rally effective inhibitors of AChE, the virtual ab- sence of anticholinesterase activity of the cyclic esters is unexpected and their inactivity is attri- buted to steric effects. Stereospecificity in AChE inhibition has also been demonstrated with cis and trans isomers of organophosphorus esters. Mevinphos consists prin- cipally of two geometric isomers, the cis-crotonate (VI) and the trans-crotonate (VII) forms shown in Fig. 7 (Fukuto et al., 1961). Compound VI showed approximately 10 times the effectiveness of com- Fig. 7 Mevinphos: cis-crotonate and trans-crotonate forms 0 0II pi(CH0)2 PO ,H (CH3O)2PQ COOCH3 H5C ICOOCH3 C H cis-crotonate trons-crotonate T. R. FUKUTO pound VII as an inhibitor of housefly-head AChE (Table 2). Table 2 Bimolecular rate constants (ke) for the inhibition of fly-head AChE by the cis-crotonate and trans-crotonate isomers of mevinphos ke (M-' min-') Temperature (°C) cis-crotonate trans-crotonate 29.0 33.0 37.5 3.9 x 104 1.0 x 105 2.8 x 105 8.7 x 103 1.4 x 104 2.2 x 104 The activation energy of the reaction of com- pound VI with fly-head AChE (41.6 kcal) is much greater than that for compound VII (19.9 kcal), indicating that the phosphoryl ester linkage in the latter is intrinsically more reactive. The higher rate of inhibition by compound VI than by com- pound VII must then be attributed to a sterically more favourable transition state, as reflected in entropy of activation (z1S+) values of 90.3 e.u. for compound VI and 6.6. e.u. for compound VII. These results suggest that the carbomethoxy group tranis to the dimethyl phosphoryl moiety is less likely to interfere sterically with the reaction be- tween the nucleophilic centre in the esteratic site and the phosphorus atom. Recently, the inhibition of bovine erythrocyte AChE by compounds VI and VII has been examined by the kinetic procedure developed by Main & Iverson (1966) for determining the affinity constant for complex formation, Ka, and the phosphorylation constant, kp (Chiu & Dauterman, 1969). The results, given in Table 3, are somewhat in disagree- ment with the preceding conclusions regarding Table 3 Ka and kp values for inhibition of bovine erythrocyte AChE by the cis- and trans-crotonate isomers of mevinphos at 50C Isomer Ka (M) kp (min-') ke (M-1 min-') cis-crotonate 1.7 x 10-3 trans-crotonate 3.2 x 10-3 59.0 5.6 3.5 x 104 1.7 x 103 steric effects. The data in Table 3 show that the approximately 10-fold greater phosphorylation rate (kp) for the cis-crotonate isomer is mainly responsible for the difference in overall inhibition rates (ke). In contrast, the Ka values for the two isomers are similar, the value for the trans-crotonate being less than twice that of the cis-crotonate. Since kp, the phosphorylation constant, is associated with reacti- vity, the overall conclusion to be drawn is that the anticholinesterase activity of these isomers depends more on reactivity than on steric effects. The kinetic approach developed and refined by Main (1964) and by Main & Iverson (1966) for determining Ka and kp has been extremely useful in separating steric effects from reactivity effects in AChE inhibition by organophosphorus esters. Particularly relevant in this connexion is work carried out by Hollingworth et al. (1967) on the selective toxicity of fenitrothion [O,O-dimethyl 0-(4-nitro-m- tolyl) phosphorothioate] and some of its analogues. Fenitrothion (IX, Table 4) is a wide-spectrum insecti- cide whose activity is comparable with that of para- thion-methyl (VIII), but its toxicity for mammals is only 0.02-0.05 times that of the latter (see Table 4). Since the toxic action of P= S phosphorus esters occurs through their corresponding P=0 esters after metabolic activation, the anticholinesterase activities of the latter compounds were determined. Values for ke in Table 4 show that introduction of alkyl groups in the 3-position of the phenyl ring renders the compound more inhibitory to housefly- head AChE, with inhibition increasing in the order H <CH3 <iso-C3H7, while the opposite effect is observed with bovine erythrocyte (mammalian) AChE. A similar decrease in anticholinesterase activity against mouse-brain AChE was observed with substitution in the 3-position, although the data are not included here. Furthermore, there was signi- ficant correlation between toxicity for the mouse and inhibition of bovine erythrocyte AChE (also mouse- brain AChE) for compounds XI, XII, and XIII. Analysis of the inhibition of fly-head and bovine erythrocyte AChE by the kinetic procedure developed by Main shows that (with increasing substitution) change in Ka is primarily responsible for change in ke. There is only a slight decrease in kp for both enzymes in proceeding from compound XI to compound XIII and this may be attributed to the electron-donating tendency of the alkyl group and to the inhibition of resonance of the nitro group by steric interference. A comparison of the val- ues for kp with the rate constants for solvolysis 36 STRUCTURE-ACTIVITY RELATIONSHIPS FOR AChE INHIBITION BY OP COMPOUNDS Table 4 Toxicity and anticholinesterase data for fenitrothion and related esters GENERAL FORMULA x 11 02N / OP(OCH3)2 R Substituents LDso (mg/kg) for: Compound w ke x 1O-5 Ka x 105 kp kbNo. R | X housefly whmouse (M-1 min-') (M) (min-') (min-') ouse VIlI H S 1.2 23 - - - IX CH3 S 3.1 1 250 _ - _ _ X iso-C3H 7 S 6.3 880 - _ _ - Housefly-head AChE Xi H 0 2.5 21 2.9 3.7 10.6 5.25 Xii CH3 0 4.3 120 7.6 1.1 8.3 3.50 Xlii iSo-C31H-7 0 6.5 > 500 22.6 0.33 7.5 2.89 Bovine erythrocyte AChE Xi H 0 - - 5.2 1.3 6.6 - XII CH3 0 - - 0.73 6.7 5.0 - XIII iso-C3H7 0 - . 0.22 15.8 3.5 - (kb) gives a good linear relationship, indicating that the factors that control phosphorylation and hydrolysis rates are similar. Since K. measures the tendency for the enzyme-inhibitor complex to dissociate-i.e., it is a binding constant-it should be dependent largely on the structural and steric features of the molecule. The large change in Ka for the three compounds suggests that enzyme- inhibitor-complex formation with housefly AChE is aided by interaction of the 3-alkyl substituent with the anionic site but is hindered in the case of bovine erythrocyte AChE. Similar studies have been carried out to explain the more than 200-fold greater toxicity of the diiso- propyl homologue of parathion (O,O-diisopropyl O-p-nitrophenylphosphorothioate)forhousefliesthan forhoney-bees (LD50 = 4.7 ,tg/g for the fly and > 1000 ,ug/g for the bee) (Camp et al., 1969). The values for ke, Ka, and kp for the diisopropyl homologue of paraoxon, the anticholinesterase intermediate, against housefly and honey-bee AChE are given in Table 5. The value of ke for housefly AChE is approximately 37 times that for honey-bee AChE, Table 5 Anticholinesterase activity of the diisopropyl homologue of paraoxon AChE ke (M-' min-') f K. (M) T k_p (min-') honey-bee-head 2.4 x 103 9.1 x 104 2.2 housefly-head 8.8 x 104 9.5 x 1Q0-5 8.3 which is consistent with the 18-fold greater tole- rance that honey-bees show to the diisopropyl homologue of paraoxon. Examination of the values of Ka and kp shows that the great difference in anticholinesterase activity is caused by the Ka for bee AChE being 9.6 times that for fly AChE and by the kp for fly AChE being 3.9 times that for bee AChE. According to these constants, steric effects, as estimated by Ka, are significantly more important than reactivity in determining the sus- ceptibility of these two enzymes to inhibition by the diisopropyl homologue of paraoxon. 37 T. R. FUKUTO Phoxim is another highly effective insecticide that has been studied because of its unusual safety to mammals (Vinopal & Fukuto, 1971). Toxicity and anticholinesterase data are summarized in Table 6. All of the compounds were quite toxic to houseflies and, except for the phosphonate XXI, were safe to the white mouse. Although concurrent studies on the metabolism of phoxim (XIV) have shown that selectivity may also be accounted for by diffe- rences in rates of phoxim detoxification in the housefly and mouse, the data in Table 6 show that housefly-head AChE is significantly more susceptible to inhibition than bovine erythrocyte AChE (see the last column in Table 6), suggesting that differences in sensitivity to inhibition between insect and mam- malian AChE also contribute to the selective action of the compounds. Because of the favourable toxicological properties of phoxim, a series of diethyl phosphate esters of ring-substituted acetophenone oximes were examined for insecticidal and anticholinesterase activity (Fukuto et al., 1969). Although most of the com- pounds studied were ineffective as insecticides, they were generally strong inhibitors of housefly-head AChE. Multiple regression analysis of the inhi- bition data (I50) with different free-energy para- meters for the p-substituted acetophenone 0- (diethylphosphoryl) oximes (XXII, Fig. 9) produced the following equation that gave best fit to the data: log 1/I50 = 7.19- 2.37F- 2.47R (3) where F and R are the field and resonance con- stants of Swain & Lupton (1968). Eq. (3) gave a correlation coefficient r of 0.89 and the correlation is shown graphically in Fig. 8. The plot shows that anticholinesterase activity is directly related to the electron-donating tendency of the substituent on the phenyl ring, a relationship opposite to that dis- cussed earlier for substituted-phenyl diethyl phos- phates. These results were unexpected and difficult to rationalize in terms of a mechanism involving nucleophilic attack by the serine hydroxyl in the esteratic site on the phosphorus atom, a process that should be aided by electron-withdrawing sub- stituents. Subsequently, it was shown that substituted acetophenone 0-(diethylphosphoryl) oximes that contain electron-donating substituents readily under- go the Beckmann rearrangement to the corres- ponding enol phosphate according to the equation ble 6 Toxicity and anticholinesterase data for phoxim and analogues GENERAL FORMULA R'/ CN R/ \ON = C Substituents LDso (mg/kg) for ke (M-1 min-') No.pun fl-ed bovine ratoeCompound 1 R R' X housefly mouse ACflyhEd erythrocyte ratio AChE AChE XIV C2HsO C2HsO S 2.1 >2 000 - XV CH30 CH30 S 3.3 >2 000 - - - XVI iso-C3H7O iso-C3H70 S 16.0 >1 500 - - _ XVII C2H50 C2Hs S 5.8 > 500 _ - - XVIII C2H50 C2H50 0 3.4 1 000 8.4 x 107 3.1 x 105 270 XIX CH30 CH30 0 5.0 >1 000 1.3 x 108 1.7 x 105 730 XX iso-C3H7O iso-C3H70 0 5.0 1 250 8.1 x 107 1.1 x 105 756 XXI C2HSO C2Hs Q 5.9 70 5.9 x 107 1.2 x 106 49 a i.e., ke (housefly-head AChE)/ke (bovine erythrocyte AChE). 38 STRUCTURE-ACTIVITY RELATIONSHIPS FOR AChE INHIBITION BY OP COMPOUNDS Fig. 8 Relation between log observed 1/150 and log 1/150 calculated * for p-substituted acetophenone oxime diethyl phosphates 9.0 CH/O *CH30 H Fe Cl *Br /*CF3, CN I I I I I I I I I I 4.0 5.0 6.0 7.0 8.0 9.0 LOG /I (observed) * Calculated from the equation log 1/150 = 7.19-2.37 F - 2.478 R; see text. given in Fig. 9. Oxime phosphates in which substi- tuent X was electron-attracting did not undergo this rearrangement. Since the enol phosphate (XXIII) is a highly reactive ester and would be expected to react rapidly with AChE, it is probable that the high anticholinesterase activity ofcompounds of type XXII, where X is electron-donating, is caused by the rearranged product. Fig. 9 Beckmann rearrangement of substituted acetophenone 0-(diethylphosphoryl)oxime (XXII) to enol phosphate (XXIII) 0 0 xIc=N0P(oc2H82 X4 N=COP(0C2H5)2 CH3 CH3 XXII XXIII Recently, a number of O-alkyl S-alkyl phosphor- amidothioates have been shown to possess out- standing insecticidal activity (Quistad et al., 1970). Toxicological data for some of these esters are given in Table 7. Examination of the data shows that the compounds are, on the whole, not very Table 7 Toxicological data for O-alkyl S-alkyl phosphoramidothioates and related esters GENERAL FORMULA R\ //R RV \N/ R \R4 Substituents LDso for Compound ke the kb No. R1 R2 R3 R4 (M-1 min-') housefly (min-1)No. ~ ~ ~ ~II(,4g/g) XXIV CH3S CH30 H H 9.2 x 102 1.3 3.2 x 10-2 XXV C2H5S CH30 H H 9.9 x 102 1.9 1.6 x 10-2 XXVI C2HsS C2H50 H H 1.5 x 102 3.1 1.1 x 10-2 XXVII C3H7S C3H70 H H 5.0 x 103 24.0 9.7 x 10-3 XXVIII CH3S CH30 CH3 H <10 115.0 1.8 x 10-2 XXIX CH3S CH30 CH3 CH3 <10 >500 0 XXX C2H5S C2Hs H H 2.1 x 104 0.64 5.2 x 10-2 XXXI C2H5S C2Hs CH3 CH3 <10 60.0 0 XXXII CH3S CH3S H H 1.3 x 104 0.75 - Xxxiii C2H5S C2H5S H H 1.7 x 104 32.0 - 39 2 8.00 -3 o -) 070o 0 6.0 5.0 I I T. R. FUKUTO effective inhibitors of fly-head AChE. The strongest anticholinesterases of the series are compounds XXX, a phosphonate ester, and XXXIII, a dithioate, with bimolecular inhibition constants (ke) of 2.1 x 104 and 1.7 x 104 M-1 min-'), respectively. Thus, compared to paraoxon (ke = 2.7x 107) these compounds are approximately 1/1000 as effective in inhibiting fly-head AChE, although both are significantly more toxic to the housefly. Minor modi- fication of the structure, particularly with respect to groups on the nitrogen atom, had profound effect on anticholinesterase activity (compare compounds XXIV, XXVIII, and XXIX). The substantial de- crease in activity upon insertion of a single methyl group is difficult to rationalize in terms of the reac- tivity of the molecule, since compound XXVIII is only slightly less susceptible to alkaline hydrolysis than compound XXIV (compare the values for kb, the pseudo first-order hydrolysis constants in pH 11.5 phosphate buffer for P-S bond cleavage). On the other hand, the N,N-dimethyl derivative XXIX was completely resistant to alkaline hydrolysis and its poor anticholinesterase activity may be ex- plained on the basis of the stable character of this compound. The phosphoramidothioates are unusual in other respects. In aqueous potassium hydroxide 0-methyl S-methyl phosphoramidothioate (XXIV, Fig. 10) is Fig. 10 P-OCH3 and P-SCH3 bond cleavage in hydrolysis of 0-methyl S-methyl phosphoramidothioate CH30, ,,0 CH3S, , PI ± KOH PfPCH3S~ NH2 -o' NH2 CH30, ,, + H-O .N -0 'NH2 hydrolysed by predominantly P-OCH3 bond (82%) cleavage to produce XXIV a, while in less polar orga- nic solvent-water mixtures P-SCH3 bond cleavage predominates to produce XXIVb (Fahmy, unpublish- ed data). Based on the acidities of the conjugate acids of the three types ofgroup bonded to the central phosphorus atom, it was anticipated that the thio- methylate moiety would be displaced under alkaline conditions. The fact that the methoxide rather than the thiomethylate anion is displaced in an aqueous environment raises the important question concerning the nature of the leaving group when phosphoramidothioates react with AChE. The unsubstituted 0-alkyl S-alkyl phosphoramidothio- ates and N-monomethyl derivatives were hydrolysed more quickly than anticipated. For example, at pH 11.5 compound XXIV (kb = 3.2x 10-2 min -1) hydrolysed at twice the rate of methyl paraoxon (k5 = 1.6 x 10-2 min-') under the same conditions. Thus, it appears that phosphoramidothioates, al- though having approximately the same reactivity as methyl paraoxon in terms of hydrolysis rates, are substantially less effective as anticholinesterase agents. Because of their unusual toxicological and chemical properties, phosphoramidothioates deserve further study. Finally, the effect of chirality in the phosphorus molecule on AChE inhibition deserves comment. There have been a number of reports describing differences in cholinesterase inhibition by enantio- morphs of organophosphorus esters containing an asymmetric phosphorus atom. Table 8 gives toxicological data for the (+)- and (-)-isomers of O-ethyl S-2-(ethylthio)ethyl ethylphosphonothioate (Fukuto & Metcalf, 1959b). Table 8 Anticholinesterase and toxicity data for enantiomers of S-2-(ethylthio)ethyl ethylphosphonothioate * ke for housefly- LDso (gg/g) for LC5o for Isomer head AChE Culex pipiens (M-1 min-') housefly honey larvae (ppm) (+) 6.7 x 104 30.0 6.6 0.80 (-) 7.6 x 105 5.0 0.68 0.08 (±t) 4.4 x 1 05 9.5 - 0.15 X Data of Fukuto E Metcalf (1959). The ke values given in Table 8 show that the rate of reaction of the (-)-isomer with fly-head AChE is 11 times that of the (+)-isomer and 1.8 times that of the (+)-mixture. These values are reflected in the toxicity data, which show that the toxicity of the (-)-isomer for the housefly, honey bee, and mosquito larva is 6, 9.7, and 10 times that of the (+)-isomer, respectively. More recently, the toxicological properties of the four isomers of O-sec-butyl S-2-(ethylthio) ethyl ethylphosphonothioate (Table 9), a compound containing two chiral centres, have been determined (Wustner & Fukuto, unpublished data, 1971). The asymmetric centres are indicated in Fig. 11. 40 STRUCTURE-ACTIVITY RELATIONSHIPS FOR AChE INHIBITION BY OP COMPOUNDS 41 Table 9 Anticholinesterase and toxicity data for the isomers of O-sec-butyl S-2-(ethylthio)ethyl ethylphosphonothioate ke (M-1 min-') for the folowing ChE: LDso for the LCso for No. designation bovine housefly Culex pipiensde erythrocyte horse serum fly head (4g/g) larvae (ppm) XXXIV + - 2.49 x 103 5.12 x 1O3 1.72 x 106 6.6 0.25 XXXV - - 2.10 x 104 1.72 x 103 1.95 x 106 6.9 0.15 XXXVI + + 5.05 x 10 1.16 x 103 1.36 x 103 >500 >1 XXXVII - + 1.28 x 102 1.43 x 103 2.80 x 103 500 >500 XXXVIII ± ± 1.55 x 103 1.16 x 103 1.48 x 106 10.8 0.65 Fig. 11 Chiral centres in O-sec-butyl S-2- (ethylthio)ethyl ethylphosphonothioate H3C,* ,CHO% OC2H5 'Pi C2H5 SCH2-CH2SC2H5 Several points of interest emerge from the data in Table 9. Most striking is the almost complete absence of stereospecificity in the inhibition of horse serum cholinesterase (butyrylcholinesterase). This finding is in agreement with observations by others who found that horse serum cholinesterase was equally sensitive to the enantiomers of sarin t (Boter & van Dijk, 1969) and S-alkyl p-nitrophenyl t Names against which this symbol appears are identifiedin the glossary on pages 445-446 methylphosphonothioates (Ooms & Boter, 1965). On the other hand, bovine erythrocyte AChE and fly-head AChE exhibited marked stereospeci- ficity with compounds XXXIV and XXXV, being substantially more effective as anticholinester- ases. Overall, compound XXXV, with both carbon and phosphorus atoms designated as (-), was the strongest inhibitor of AChE. The effect of chirality in the sec-butyl moiety on AChE was small but significant-e.g., the effectiveness of compound XXXV is approximately 10 times that of compound XXXIV. The results also show that toxicity to houseflies and to Culex pipiens fatigans larvae is consistent with anticholinesterase activity, and the two enantiomers containing (-) phosphorus showed the greatest toxicity. Overall, these results show that the toxic properties of organophosphorus esters are strongly affected by chirality in the molecule. REFERENCES Aldridge, W. N. & Davison, A. N. (1952) Biochem. J., 51, 62-70 Boter, H. L. & Dijk, C., van (1969) Biochem. Pharmacol., 18, 2403-2407 Camp, H. B., Fukuto, T. R. & Metcalf, R. L. (1969) J. agric. Food Chem., 17, 243-248 Chiu, Y. & Dauterman, W. C. (1969) Biochem. Pharmacol. 18, 359-364 Fukuto, T. R. (1957) Advances in pest control research, New York, Interscience, vol. 1, pp. 147-191 Fukuto, T. R., Hornig, E. O., Metcalf, R. L. & Winton, M. Y. (1961) J. org. Chem., 26, 4620-4623 Fukuto, T. R. & Metcalf, R. L. (1956) J. agric. Food Chem., 4, 930-935 Fukuto, T. R. & Metcalf, R. L. (1959a) J. Amer. chem. Soc., 81, 372-377 Fukuto, T. R. & Metcalf, R. L. (1959b) J. econ. Ent., 52, 739-740 Fukuto, T. R. & Metcalf, R. L. (1965) J. med. Chem., 8, 759-762 Fukuto, T. R., Metcalf, R. L., Jones, R. L. & Myers, R. 0. (1969) J. Agric. Food Chem., 17, 923-930 Fukuto, T. R., Metcalf, R. L., Winton, M. Y. & March, R. B. (1963) J. econ. Ent., 56, 808-810 42 T. R. FUKUTO Hansch, C. & Deutsch, E. A. (1966) Biochem. Biophys. Acta (Amst.), 126, 117-128 Heath, D. F. (1961) Organophosphorus poisons, London, Pergamon Hollingworth, R. M., Fukuto, T. R. & Metcalf, R. L. (1967) J. agric. Food Chem., 15, 235-241 Krupka, R. M. (1964) Canad. J. Biochem., 42, 677-693 Main, A. R. (1964) Science, 144, 992-993 Main, A. R. & Iverson, F. (1966) Biochem. J., 100, 525-531 O'Brien, R. D. (1960) Toxic phosphorus esters, New York, Academic Press Ooms, A. J. J. & Boter, H. L. (1965) Biochem. Pharmacol., 14, 1839-1845 Quistad, G. B., Fukuto, T. R. & Metcalf, R. L. (1970) J. agric. Food Chem. 18, 189-194 Swain, C. G. & Lupton, E. C. (1968) J. Amer. chem. Soc., 90, 4328 Taft, R. W., Jr (1956) In: Steric effects in organic chemistry, New York, Wiley, pp. 556-675 Vinopal, J. H. & Fukuto, T. R. (1971) Pestic. Biochem. Physiol., 1, 44-60 DISCUSSION HOLAN: Were cyclic phosphorinanes that were not anti- cholinesterases active insecticidally? FUKuTrO: No. AUGUSTINSSON: Were the specific activities of bovine erythrocyte acetylcholinesterase and the fly-head acetyl- cholinesterase comparable? Were both enzymes in solu- tion? FUKUTO: Both enzymes were soluble. In general, we have attempted to balance the specific activities of fly acetylcholinesterase and bovine erythrocyte acetyl- cholinesterase in the determination of inhibition con- stants. ALDRIDGE: In the O-alkyl S-alkyl phosphoramidothio- ates, is there any relation between inhibitory power and the ratio of P-S and P-O cleavage with different alkyl groups under chemical hydrolysis? FUKUTO: Apart from the O,S-dimethyl compounds, we have not examined phosphoramidothioates in detail. However, we have examined the effect of increasing methyl-group substitution on the nitrogen: when one methyl group is placed on the nitrogen, there is increased P-S cleavage, and with two methyl groups, hydrolysis occurs entirely by P-S cleavage. NEAL: Have you compared the energy of activation of hydrolysis for a series of dialkyl aryl phosphates with the Hammett sigma constants? FUKUTO: No. WEIDEN: Have you made a Main kinetic analysis of cholinesterase inhibition by the optically active phos- phonates ? FUKUTO: Such a study in now in progress. NEAL: When we see a linear free-energy relationship between sigma constants and the reactions of a series of phosphorus compounds, does this reflect a change in the energy of activation of a reaction or rather a change in a kinetic parameter? FUKUTO: The correlations that we have made between Hammett's sigma constants and anticholinesterase acti- vity have been with overall rates of inhibition. This would suggest that we are seeing a change in a kinetic parameter rather than a change in activation energy. Unfortunately, we have not determined Ka values for a large series of phosphates, and therefore a definite answer cannot be given. However, some limited studies show a direct correlation between kp values and sigma constants. In the series of compounds that we have studied, the substituent on the phenyl ring was in the para position, and presumably the Ka values for these compounds are similar. Hence, one would expect a correlation between sigma constants and a kinetic parameter-i.e., ki.