66 Chem. Res. Toxicol. 1992,5, 66-71
(35) Di Netta, J., and Egerton, J. R. (1967) Compositions and methods for treating helminthiasis, US. Patent 3,325,356.
(36) Di Netta, J., and Egerton, J. R. (1969) Compositions and methods for treating helminthiasis comprising combinations of organo-phophates and certain dibenzocycloheptenes, U.S. Patent 3,484,520.
(37) Egerton, J. R., and Di Netta, J. (1970) Anthelminthic compo- sitions and method of using same, U.S. Patent 3,524,000.
(38) Little, P. J., and Ryan, A. J. (1982) Inhibitors of hepatic mix- ed-function oxidases. 4. Effecta of benzimidazole and related compounds on aryl hydrocarbon hydroxylase activity from phe- nobarbitone and 3-methylcholanthrene induced rats. J. Med. Chem. 25,622426.
(39) Hansch, C., and Leo, A. J. (1979) Substituent Constants for Correlation Analysis in Chemistry and Biology, Wiley-Intersci- ence, New York.
(40) March, J. (1980) Advanced Organic Chemistry: Reactions, Mechanisms, and Structure, pp 89-92, McGraw-Hill Kogakusha La., Tokyo.
(41) Ortiz de Montellano, P. R., Kunze, K. L., Yost, G. S., and Mico, B. A. (1979) Self-catalyzed destruction of cytochrome P-450 Covalent binding of ethynyl sterols to prosthetic heme. h o c . Natl. Acad. Sci. U.S.A. 76, 746-749.
(42) Ortiz de Montellano, P. R., and Kunze, K. L. (1980) Inactiva- tion of hepatic cytochrome P-450 by allenic substrates. Biochem. Biophys. Res. Commun. 94,443-449.
(43) Guengerich, F. P. (1990) Mechanism-based inactivation of hu- man liver microsomal cytochrome P-450 IIIA4 by gestodene. Chem. Res. Toxicol. 3, 363-371.
(44) Tephly, T. R., Black, K. A., Green, M. D., Coffman, B. L., Dannan, G. A., and Guengerich, F. P. (1986) Effect of the suicide
substrate 3,5-diethoxycarbonyl-2,6-dimethyl-4-ethyl-l,4-dihydro- pyridine on the metabolism of xenobiotics and on cytochrome P-450 apoproteins. Mol. Pharmacol. 29, 81-87.
(45) B&ker, R. H., and Guengerich, F. P. (1986) Oxidation of 4aryl- and 4-alkyl-substituted 2,6-dimethyl-3,5-bis(alkoxycarbonyl)-1,4 dihydropyridines by human liver microsomes and immunochem- ical evidence for the involvement of a form of cytochrome P-450. J. Med. Chem. 29, 1596-1603.
(46) Halpert, J., Hammond, D., and Neal, R. A. (1980) Inactivation of purified rat liver cytochrome P-450 during the metabolism of parathion (diethyl p-nitrophenyl phosphorothionate). J. Biol. Chem. 255,1080-1089.
(47) Guengerich, F. P. (1986) Covalent binding to apoprotein is a major fate of heme in a variety of reactions in which cytochrome P-450 is destroyed. Biochem. Biophys. Res. Commun. 138,
(48) Buening, M. K., and Franklin, M. R. (1976) SKF 525-A inhib- ition, induction, and 452-nm complex formation. Druz Metab.
193-198.
Dispos. 4, 244-255. (49) Pershing, L. K., and Franklin, M. R. (1982) Cytochrome P-450
metabolic-intermediate complex formation and induction bv ma-
-
crolide antibiotics: A ned class of agents. Xenobiotiia 12, 687-699.
(50) Reidy, G. F., Mehta, I., and Murray, M. (1989) Inhibition of oxidative drug metabolism by orphenadrine: In vitro and in vivo evidence for isozyme-specific complexation of cytochrome P-450 and inhibition kinetics. Mol. Pharmacol. 35, 136-743.
(51) Murray, M., Zaluzny, L., and Farrell, G. C. (1986) Selective reactivation of steroid hydroxylases following dissociation of the isosafrole metabolite complex with rat hepatic cytochrome P-450. Arch. Biochem. Biophys. 251,471-478.
Quantitative Structure-Activity Analysis of Acetylcholinesterase Inhibition by Oxono and Thiono Analogues of Organophosphorus
Compounds
Donald M. Maxwell* and Karen M. Brecht United States Army Medical Research, Institute of Chemical Defense, Aberdeen Proving
Ground, Maryland 21010-5425 Received June 27, 1991
A comparison of the bimolecular rate constants (hi) for inhibition of electric eel acetyl- cholinesterase (AChE) by the oxono (i.e., P=O) and thiono (i.e., P=S) analogues of parathion, methylparathion, leptophos, fonofos, sarin, and soman revealed that the oxono/thiono ratios of ki values varied from 14 for soman to 1240 for parathion. Analysis of the relative importance of the dissociation equilibrium constant and the phosphorylation rate constant in producing this variation in ki values indicated that the oxono analogues had phosphorylation rate constant values that varied in a narrow range from 8- to 14-fold greater than their thiono counterparts, while the oxono/thiono ratios for dissociation constants varied widely from 1 for soman to 82 for fonofos. The lower affmities of thiono analogues for AChE probably resulted from differences in the hydrophobic binding of oxono and thiono analogues to the active site of AChE, inasmuch as the hydrophobicities (i.e., octanol/water partition coefficients) of thiono organophosphorus compounds were much greater than the hydrophobicities of their oxono analogues. Quantitative structure-activity analysis indicated that the hydrophobic effects of oxono and thiono moieties correlated with log ki for AChE inhibition to a greater extent (r2 = 0.79) than their electronic effects (r2 I 0.48). These observations suggest that the differences in hydrophobicity of oxono and thiono analogues of organophosphorus compounds may be as important as their electronic differences in determining their effectiveness as AChE inhibitors.
Introduction The low reactivity of thiono organophosphorus compounds for mammalian AChE provides a safety factor for agri-
The poor reactivity of phosphothioates and phospho- cultural of pesticides, such as or nothioaks for acetylchoheskrase (AChE)’ in while the rapid metabolic oxidation of these thiono (i-e., to their oxono analogues has been well documented (1-3).
Abbreviations: AChE, acetylcholinesterase; Pr’, isopropyl; Pin, pi- * Author to whom correspondence should be addressed. nacolyl.
This article not subject to U.S. Copyright. Published 1992 by the American Chemical Society
AChE Inhibition by Oxono and Thiono Analogues
Table I. Structures of Oxono and Thiono Analogues of Organophosphorus Compounds
0,s II I A2
A,-PP-X
oxono/thiono analogues A, A2 X" paraoxon/ parathion Et0 Et0 OPh-4-NO2 methylparaoxon/methylparathion Me0 Me0 OPh-4-N02 fonofos oxon/fonofos Et Et0 SPh leptophos oxon/leptophos Ph Me0 OPh-4-Br-2,5-C12 sarin/ thiosarin Me Pr'O F soman/ thiosoman Me Pin0 F
a Leaving group.
P=S) pesticides to their oxono (i.e., P=O) analogues in insects, but not mammals, produces a beneficial selective toxicity. Although oxono organophosphorus inhibitors of AChE are consistently better inhibitors than their thiono analogues, the ratios of their reactivities with AChE vary considerably. The bimolecular rate constant for inhibition of AChE by paraoxon was reported to be 10000-fold greater than that of its thiono analogue ( l ) , while the rate constant for inhibition of AChE by soman was only 3-fold greater than the rate constant for its thiono analogue (2).
The usual explanation for the greater inhibition of AChE by oxono organophosphorus inhibitors is that oxygen is more electronegative than sulfur, which results in a cor- respondingly greater reduction in the electron density around phosphorus (4). The greater reduction of the electron density around phosphorus produced by oxygen vs sulfur enhances the electrophilicity of phosphorus and thereby increases its reactivity toward nucleophiles, such as water or the active site serine of AChE. This expla- nation appears to adequately explain the 10-fold greater hydrolysis rate of paraoxon vs parathion and methyl- paraoxon vs methylparathion (5), but it was inadequate to explain the 10000-fold difference in AChE reactivities of paraoxon vs parathion (1,6) or to account for the tre- mendous variation in the effect of oxono vs thiono ana- logues for AChE inhibition by alkyl methylphosphono- fluoridates (2).
To address these unresolved questions concerning the differences in AChE inhibition by oxono and thiono or- ganophosphorus compounds, we performed a quantitative structure-activity analysis of the oxono and thiono ana- logues of a heterogeneous group of organophosphate and organophosphonate inhibitors (see Table I). The in vitro activity of these inhibitors was evaluated by measurement of their bimolecular rate constants for AChE inhibition, as well as their dissociation and phosphorylation constants, which had not been previously measured for any thiono organophosphorus compounds. A variety of electronic, hydrophobic, and steric properties for the substituents of these organophosphorus compounds were used as inde- pendent variables for quantitative structure-activity analysis.
Materials and Methods Chemicals. Parathion, paraoxon, methylparathion, methyl-
paraoxon, leptophos, leptophos oxon, fonofos, and fonofos oxon were obtained from the Pesticides and Chemicals Repository, U.S. Environmental Protection Agency (Research Triangle Park, NC). Soman, thiosoman, sarin, and thiosarin were obtained from the Chemical Research, Development and Engineering Center (Ab- erdeen Proving Ground, MD). The purity of all organophosphorus compounds was >98.9% as determined by 31P nuclear magnetic resonance spectroscopy. Caution: All of the aboue-mentioned organophosphorus compounds are hazardous and should be
Chem. Res. Toxicol., Vol. 5, No. 1, 1992 61
handled carefully (7). Acetylthiocholine iodide, 5,5'-dithiobis- (Znitrobenzoic acid), and electric eel AChE (type V-S) were purchased from Sigma Chemical Co. (St. Louis, MO).
Determination of Kinetic Constants for AChE Inhibition. The bimolecular rate constants (ki), dissociation equilibrium constant (&), and phosphorylation rate constants (k ) for in- hibition of AChE by organophosphorus compounds, wiere ki = k /Kd, were determined by the method of Wang and Murphy (8). dectric eel AChE dissolved in 0.1 M phosphate buffer (pH 8.0) containing 0.1% bovine serum albumin was incubated a t 5 OC with a small volume of organophosphorus inhibitor dissolved in ethanol. The concentration of AChE used for inhibition studies was 0.3 pM except for organophosphorus compounds with ki > 10' M-l min-l, where a 10-fold reduction in enzyme concentration was necessary. In order to satisfy the experimental requirements for pseudo-firsborder inhibition kinetics, inhibitor concentrations were a t least 10-fold greater than enzyme concentrations. For inhibitors with adequate aqueous solubility, k, and Kd were de- termined from first-order inhibition kinetics using inhibitor concentrations from <Kd to ZKd. The time course for inhibition of the AChE activity was determined from samples of the incu- bation mixture that were diluted 100-fold when aesayed in 2 mM substrate. Residual AChE activity was assayed by the method of Ellman et al. (9) using acetylthiocholine as substrate. The rates of inhibition for AChE were determined with 4-6 concentrations of each organophosphorus compound.
Quantitative Structure-Activity Analysis. The struc- ture-activity relationships between the electronic, hydrophobic, and steric properties of organophosphorus compounds and their inhibition of AChE were examined by stepwise multiple regrewion using RS/I software purchased from Bolt Beranek and Newman, Inc. (Cambridge, MA). Regression analysis was performed on the combined data set of inhibition constants for both oxono and thiono analogues. The octanol/water partition coefficients (P) of oxono and thiono analogues were taken from the compilation shown in Table V, while the hydrophobicities (T) of Substituents were taken from Hansch and Leo (IO). The molecular volumes (MV) that were used as steric parameters for substituents were calculated using an algorithm developed by Hopfiger (11). The ionization constants (pK,) of the acid formed by the hydrolysis of organophosphorus compounds were taken from Ooms (12) and Dean (13). The properties of the organophosphorus compounds that were evaluated for their effect on AChE inhibition were the pKa of the leaving group (substituent X), the hydrophobicities (al and a,) of the small (substituent A,) and large (substituent A,) groups that were attached to the phosphorylating moiety, the difference in hydrophobicities (al - T,) between substituents AI and A,, the differences in hydrophobicities (al - "01s and u2 - sop) between substituents Al or A, and P = O or P=S, the mo- lecular volumes (MV1 and MV2) of substituents A, and A,, a quantal electronic indicator variable (Dl) that took the value 0 for thiono analogues and 1 for oxono analogues, and a quantal electronic indicator variable (D,) that took the value 0 for orga- nophosphates and 1 for organophosphonates. The degree of correlation between structural properties and AChE inhibition was expressed as P', the fraction of total variation explained by each structure-activity regression equation, where r was the correlation coefficient. The r2 values for regression equations containing different numbers of independent variables were corrected for the number of variables by the procedure of Won- nacott and Wonnacott (14).
Results The ki values describing the inhibition of AChE by ox-
on0 and thiono analogues of organophosphorus compounds varied from 55 M-' min-' for methylparathion to 1.3 X lo7 M-' min-' for soman (Table 11). The oxono analogues invariably had larger ki values for AChE inhibition than their thiono counterparts, but the oxono/ thiono ratios of ki values varied considerably from 1240 for paraoxon/ parathion to 14 for soman/thiosoman.
To evaluate the importance of phosphorylation and affinity in producing the wide variation in ki values be- tween inhibitors, the k , and Kd values were determined
68 Chem. Res. Toxicol., Vol. 5, No. 1, 1992 Maxwell and Brecht
Table 11. Kinetic Constants for Inhibition of Eel Acetylcholinesterase" phosphorylation dissociation bimolecular
rate constant O P , ki, M-' min-' ratio paraoxon 32.7 (5.5) 3.5 x 10-4 (0.7) 9.3 x 104 1240
methylparaoxon 22.7 (7.1) 9.1 6.5 X lo-' (1.4) 70 3.5 x 104 640
fonofos oxon 27.4 (4.3) 8.2 2.3 X lo4 (0.3) 82 1.2 x 105 680
leptophos oxon 28.3 (4.5) 1.3 X (0.4) 2.1 x 105 960
sarin 30.2 (9.1) 13.5 7.0 X 10" (1.4) 3.8 4.3 x 106 51
soman 45.1 (9.8) 12.9 3.5 X 10" (1.5) 1.1 1.3 x 107 14
constant 0 ! S b inhibitor k,, min-' ratio Kd, M ratio O P , rate constant
7.5 x 10' parathion -C -C
methylparathion 2.5 (0.9) 4.6 X (0.7) 5.5 x 10'
fonofos 3.3 (1.4) 1.8 x 10-2 (0.2) 1.8 x 102
-C 2.2 x 102 leptophos -C
thiosarin 2.2 (0.8) 2.6 X (0.4) 8.5 x 104
thiosoman 3.5 (0.3) 3.8 X 10" (0.9) 9.2 x 105
a Inhibition constants determined at 5 "C in pH 8.0 phosphate buffer. Standard deviations are in parentheses. Oxono/thiono ratio of kinetic constants for analogue pairs. Inadequate inhibitor solubility in water to perform measurement.
Table 111. Quantitative Structure- Activity Relationships for AChE Inhibition by Oxono and Thiono Analogues of
Organophosphorus Compounds independent explained variation (r2)
variables log k; log k, log (I/&) 0.36 0.18 0.05 0.01 0.33 0.02 0.01 0.01 0.34 0.40 0.79 0.50 0.23 0.01 0.12 0.01 0.48 0.87 0.17 0.01 0.44 0.01
0.12 0.15 0.54 0.01 0.24 0.67 0.39 0.21 0.11 0.29 0.60
for all inhibitors except parathion and leptophos where the aqueous solubilities were too low to achieve inhibitor concentrations LKd, which is required for determination of these kinetic constants (Table 11). For the oxono ana- logues the k, values varied only 2-fold from 22.7 min-' for methylparaoxon to 45.1 min-' for soman, while the Kd values varied 186-fold from 6.5 X 10" M for methylpara- oxon to 3.5 X lo* M for soman. Similarly, the thiono analogues had a 2-fold range of k, values (2.2 min-' for thiosarin to 3.5 min-' for thiosoman) and a 12000-fold range of Kd values (4.6 X loT2 M for methylparathion to 3.8 X lo4 M for thiosoman).
The relative importance of phosphorylation and affinity in producing the differences in ki values between oxono and thiono analogues was examined by comparing the oxono/thiono ratios of k, and Kd values (Table 11). The oxono analogues had k, values that were consistently 8- to 14-fold greater than those of their thiono analogues. However, the Kd values of oxono analogues were from 1- to 82-fold greater than those of their thiono counterparts. Therefore, the greater ki values for AChE inhibition by the oxono analogues of fonofos and methylparathion in com- parison to the ki values of their thiono analogues were primarily due to the 70- to 82-fold differences in Kd and not to the 8- to 9-fold differences in k,. In contrast, the 51-fold difference in ki values between sarin and thiosarin was more the result of their 14-fold difference in k, than their 4-fold difference in Kd, and the 14-fold difference in the ki values for soman vs thiosoman was entirely due to their oxonof thiono difference in k,.
Quantitative structureactivity regression analysis (see Table 111) indicated that the independent variables that were most closely correlated with the variation of log ki were the pKa of the leaving group (r2 = 0.44), the oxo- nolthiono indicator variable D1 (r2 = 0.48), and the hy-
Table IV. Best Structure-Activity Regression Equations for AChE Inhibition Constants
kinetic constant regression equations ?
ki log ki = 1.36(*2 - ~ 0 , s ) + 5.89 0.79" O M b
Kd log (I/&) = 0.86(~2 - Tois) + 4.62 0.67"
k, log k, = 28.2D1 + 2.88 0.87"
log ki = 2.880, - O.83pKa + 7.57
log (I/Kd) = 0.58(~2 - ~ 0 1 s ) - 0.38pKa + 6.45 0.82b9C
log k, 29.301 + 4.32~2 - 5.00 0.94b*c log k, 23.701 + 3.51(~2 - T O / S ) + 9.28 0.94b*c
" Best one-variable regression equation for kinetic constant. * Best multiple-variable regression equation for kinetic constant. Significant difference between one-variable and multiple-variable
regression equations for kinetic constant.
drophobicity difference r2 - r o / s (r2 = 0.79). The best multiple-variable regression equation for explaining the variation in log ki was achieved with two variables, pKa and D1, which resulted in r2 = 0.85 (Table IV). Including additional independent variables in the regreasion equation did not increase ? values. When the 1.2 values for the best multiple-variable equation (i.e., pKa and Dl) and the best single-variable equation (i.e., r2 - were corrected for the number of variables in each equation, the regression equations were not significantly different. Therefore, the variation in log ki was explained equally well by either the combined electronic effects of the P=S or P=O moiety and the pK, of the leaving group, or by the hydrophobic effect resulting from the difference in the hydrophobicities of the largest alkoxy substituent and the P=S or P=O moiety.
The single variable that best explained the experimental variation in the affinity [Le., log (l/Kd)] of the organo- phosphorus compounds was r2 - ao/s (r2 = 0.67). The explained variation in log (l/Kd) could be significantly increased to ita maximal value (r2 = 0.82) by including a second variable, pKa, in the multiple regression equation.
The independent variable that was most closely corre- lated with differences in k was the indicator variable D, (r2 = 0.87). The explainecfvariation was significantly in- creased to ? = 0.94 when either of the hydrophobic terms r2 or 7r2 - ro was included in multiple regression equa- tions. Including additional independent variables in the regression equations did not improve the explained vari- ation.
From the graphs of log ki and log (I/&) vs r2 - AO/S, the single variable that best explained their experimental variation, it was concluded that the oxono and thiono analogues could be treated as a single data set rather than as different subpopulations of data, inasmuch as their
AChE Inhibition by Oxono and Thiono Analogues
respective groups of data overlap and both groups are equally distributed around the regression line of the com- bined data set (Figures 1 and 2). Comparison of these graphs for log ki and log (l/Kd) against 7r2 - also illustrated that log ki had a greater dependence on a2 - sols than did log (l/Kd), which can be quantitively estimated by comparison of the slopes of their respective regression equations [1.36 for log ki; 0.86 for log (l/Kd)] shown in Table IV. A graph of log k, vs its best independent var- iable, D1, was not included because D1 is a quantal indi- cator variable with only two possible values rather than a continuously variable parameter, such as A.
Discussion Our analysis of the factors that influence the differences
in reactivity of oxono and thiono analogues of organo- phosphorus inhibitors of AChE indicated that the hydro- phobic effect of sulfur vs oxygen may be as important as its electronic effect on phosphorus electrophilicity. Re- gression analysis suggested that the experimental variation in ki was explained equally well either by a single hydro- phobic variable, ?r2 - A ~ / ~ , or by a combination of two electronic variables, D1 and pKa (Table IV).
The electronic effect on ki was the result of a minor effect on Kd, where pKa explained 15% of the experimental variation in Kd, and a major electronic effect on k,, where D1 explained 87% of the experimental variation in k, (Table IV). The magnitude of the electronic effect of sulfur vs oxygen on the chemical reactivity of phosphorus, which can be estimated from the 10-fold reduction in al- kaline hydrolysis rates of thiono organophosphorus com- pounds vs their oxono counterparts (5), was similar to the 8- to 16fold reductions in phosphorylation rate constants for AChE inhibition that we observed for thiono vs oxono analogues (Table II). This relatively small electronic effect of sulfur vs oxygen substitution on k, correlated with previous observations of the magnitude of electronic effects on phosphorus by other substituents, such as leaving groups, where 1OOO-fold changes in the lability of the bond between phosphorus and the leaving groups of organo- phosphorus compounds resulted in 2-fold changes in k for AChE inhibition (15). These electronic effects on pbos- phorylation constants may be subject to the same elec- tronic limitations that have produced a plateau effect on the increase in ki values for AChE inhibition as pKa values for leaving groups of organophosphorus inhibitors are re- duced below 7.0 (16).
A large effect of sulfur vs oxygen substitution was evi- dent in comparisons of the hydrophobicity of thiono analogues vs oxono analogues. Regression analysis sug- gested that 79% of the experimental variation in ki could be explained by 7r2 - The major hydrophobic effect on ki was exerted through effects on Kd where 1r2 - a0 explained 67% of the experimental variation in Kd, whiie a minor hydrophobic effect was evident with k, where 7r2 or 7r2 - AO/S explained only 7% of the variation in k, (Table IV).
Oxono analogues had 1- to 82-fold greater Kd values for AChE than their thiono counterparts, which was probably due to the hydrophobic effects of P=S vs P=O moieties, inasmuch as the affinity of organophosphorus compounds to AChE is determined to a great extent by their binding to various hydrophobic binding sites in the active site of the enzyme (1 7,18). Alkyl groups are so hydrophobic that they bind tightly to these hydrophobic sites. However, alkoxy groups, particularly those with small numbers of carbons (Le., methoxy, ethoxy), are much less hydrophobic (IO) and presumably bind poorly to these hydrophobic
Chem. Res. Toxicol., Vol. 5, No. 1, 1992 69
binding sites. The P=O moiety is even less hydrophobic than small alkoxy groups (10). However, the P=S moiety is more hydrophobic, on the basis of our compilation of the logarithms of the octanol/water partition coefficients (log P) that have been published for pairs of thiono/oxono analogues of organophosphorus compounds (Table V). Regardless of the wide variation of the compiled log P values, which ranged from 2.70 to 5.88 for thiono analogues, the differences in log P between thiono and oxono ana- logues were essentially a constant value (1.65 f 0.18). This indicated that organophosphorus compounds containing P=S moieties partition 60-fold (Le., more into oc- tanol than their P=O analogues. A A value of 1.65 for P=S is roughly equal to the A value for ethyl substituents ( A = 1.55). Therefore, the P=S moiety of thiono analogues could compete for AChE hydrophobic binding sites with small alkoxy substituents that the P 4 moiety of oxono analogues would not be hydrophobic enough to displace.
Those organophosphorus compounds with methoxy or ethoxy substituents (see Table I) exhibited greater oxo- no/ thiono differences in affinity than organophosphorus compounds with larger alkoxy substituents (Table 11). A possible explanation for this variation in oxono/ thiono differences in affinity may be provided by previous ob- servations of AChE inhibition by chiral organophosphorus compounds. The variation in ki values for inhibition of AChE by P(-) vs P(+) stereoisomers of organophosphorus compounds was primarily the result of differences in af- finity of these stereoisomers for AChE and not of differ- ences in phosphorylation (19), which was similar to our observations of variation in AChE inhibition by oxono and thiono analogues. In addition, the ratio of ki values for P(-) vs P(+) stereoisomers was approximately equal to the largest oxono/thiono ratios for ki values that we observed. These similarities between chiral effects and oxono/ thiono effects in the magnitude of the differences in ki values and the importance of affmity on AChE inhibition suggest that the affinity differences between oxono and thiono ana- logues are due to the differences in the preferred orien- tation (i.e., stereospecificity) of oxono vs thiono analogues for hydrophobic binding to AChE. Therefore, oxono and thiono analogues with small alkoxy Substituents may have a different binding orientation at the active site of AChE, while oxono and thiono analogues with large alkoxy sub- stituents have the same binding orientation.
Quantitative structure-activity analysis indicated that the hydrophobicity difference between alkoxy groups and P=S or P=O moieties explained 79% of the variation in ki values of oxono and thiono analogues. The residual unexplained variation in ki was presumably the result of other interactions such as steric effects or hydrogen bonding. Steric variables such as log MV for substituents AI and A2 did not increase the explained variation in re- gression equations for ki, k,, or Kd. Although some authors have suggested that the difference in the size of sulfur vs oxygen may produce a steric effect on AChE inhibition (ZO), the difference in the van der Waals radius of sulfur (1.8 A) vs oxygen (1.5 A) is considerably less than the differences in the sizes of the Al and A2 substituents. A more likely explanation for the residual unexplained ex- perimental variation is the difference in the hydrogen bonding of P=S vs P=O. Hydrogen bonding was not examined in this report because of the unavailability of quantitative estimates of this parameter. However, pre- vious studies of oxono and thiono analogues of substrates with proteases suggested that variation in the hydrogen bonding of the sulfur and oxygen moiety in an oxyanion hole of the enzymatic active site may be an important
70 Chem. Res. Toricol., Vol. 5, No. 1, 1992
10-
9
8
7
2- 6
2 5- 0
4
3
Maxwell and Brecht
desired, then the other substituents on phosphorus should be alkyl or large alkoxy (Lpropoxy) groups to avoid dis- placement of these substituents from hydrophobic binding sites by sulfur. If both the electronic and hydrophobic effects of sulfur are desired, the inclusion of at least one small alkoxy group (lethoxy) will allow the expression of the hydrophobic effect of sulfur.
--
-- -- --
-- --
/ e: d 2L 1
n2-no/s
Figure 1. Correlation between ki and hydrophobicity for oxono (0) and thiono (0) analogues.
10
8 g l 7
7J Y \ 7
- 4 -3 -2 - 1 0 1 2
=2-no/s
Figure 2. Correlation between Kd and hydrophobicity for oxono (0) and thiono (0) analogues.
Table V. Octanol/Water Partition Coefficients of Oxono and Thiono Analogues of Organophosphorus Compounds
difference log P in log P
thiono oxono (thiono - oxono analogue analogue analogue oxono)
azinophos ethyloxon" 3.40 1.63 1.77 azinophos methyloxona 2.69 0.78 1.91 diazinon oxon" 3.81 2.97 1.74 fonofos ox0n4 3.89 2.11 1.78 leptophos oxon" 5.88 4.58 1.30 paraoxona 3.77 1.98 1.79 methylparaoxonb*c 2.99 1.21 1.78 3-ethylmethylparaoxone 3.74 2.20 1.54 3-isopropylmethylparaoxonb~c 4.05 2.57 1.48 3,5-dimethylmethylparaoxone 3.88 2.26 1.62 fenitroxon' 3.30 1.69 1.61 chloroxon' 3.45 1.83 1.62 dicapthoxon' 3.44 1.84 1.60 0-cyclohexyl 2.70 1.49 1.21
methylphosphonofluoridatec 1.65 ( O . W d
From ref 23. "From ref 10. From ref 24. Mean value. Standard deviation in parentheses.
mode of enzyme-substrate interaction (20-22). The significance of our observations is that they suggest
that other factors besides the electronic effects of sulfur must be considered when designing thiono organo- phosphorus compounds as molecular probes (i.e., inhibitors of enzyme active sites, tetrahedral haptens for catalytic antibodies, ligands for receptors). If only the electronic effects of sulfur on the organophosphorus compounds are
Registry No. AChE, 9000-81-1; paraoxon, 311-455; parathion, 56-38-2; methylparaoxon, 950-35-6; methylparathion, 298-00-0; fonofos oxon, 944-21-8; fonofos, 944-22-9; leptophos oxon, 25006-32-0; leptophos, 21609-90-5; sarin, 107-44-8; thiosarin, 4241-37-6; soman, 96-64-0; thiosoman, 97931-17-4.
References (1) Aldridge, W. N., and Davison, A. N. (1952) The inhibition of
erythrocyte cholinesterase by triesters of phosphoric acid. Bio- chem. J. 52,663-671.
(2) Boter, H. L., and Ooms, A. J. J. (1966) Organophosphorus com- pounds. 11. Synthesis and cholinesterase inhibition of a series of alkyl and cycloalkyl methylphosphonofluoridothionates. Recl. Trav. Chim. PQYS-BQS 85, 21-30.
(3) Thompson, C. M., Frick, J. A., Natke, B. C., and Hansen, L. K. (1989) Preparation, analysis, and anticholinesterase properties of 0,O-dimethyl phosphorothioate isomerides. Chem. Res. Toxicol.
(4) Kier, L. B. (1971) Molecular Orbital Theory in Drug Research, pp 116-119, Academic Press, New York.
(5) Cox, J. R., and Ramsay, 0. B. (1964) Mechanisms of nucleophilic substitution in phosphate esters. Chem. Rev. 64, 317-352.
(6) Heath, D. F. (1961) Organophosphorus Poisons, pp 150-202, Pergamon Press, New York.
(7) Taylor, P. (1985) Anticholinesterase agents. In The Pharma- cological Basis of Therapeutics (Gilman, A. G., Goodman, L. S., Rall, T. W., and Murad, F., Eds.) 7th ed., pp 110-129, Macmillan, New York.
(8) Wang, C., and Murphy, S. D. (1982) Kinetic analysis of species differences in acetylcholinesterase sensitivity to organophosphorus insecticides. Toxicol. Appl. Pharmaco!. 66, 409-419.
(9) Ellman, G. L., Courtney, K. D., Andrea, V., and Featherstone, R. M. (1961) A new and rapid colorimetric determination of ace- tylcholinesterase activity. Biochem. Pharmacol. 7,88-95.
(10) Hansch, C., and Leo, A. J. (1979) Substituent Constants for Correlation Analysis in Chemistry and Biology, John Wiley, New York.
(11) Hopfinger, A. J. (1980) A QSAR investigation of dihydrofolate reductase inhibition by Baker triazines based upon molecular shape analysis. J. Am. Chem. SOC. 102, 7196-7206.
(12) Ooms, A. J. J. (1961) The reactivity of organic phosphor com- binations in regards to a number of esterases, Ph.D. Thesis, University of Leiden, The Netherlands.
(13) Dean, J. A. (1985) Lange's Handbook of Chemistry, Chapter 5, pp 15-60, McGraw-Hill, New York.
(14) Wonnacott, T. H., and Wonnacott, R. J. (1981) Regression: A Second Course in Statistics, pp 180-182, Wiley, New York.
(15) Forsberg, A., and Puu, G. (1984) Kinetics for the inhibition of acetylcholinesterase from the electric eel by some organo- phosphates and carbamates. Eur. J . Biochem. 140, 153-156.
(16) Ashani, Y., Wins, P., and Wilson, I. B. (1972) The inhibition of cholinesterase by diethylphosphorochloridate. Biochim. Bio-
(17) Kabachnik, M. I., Brestkin, A. P., Godovikov, N. N., Michehon, M. J., Rozengart, E. V., and Rozengart, V. I. (1970) Hydrophobic areas on the active surface of cholinesterases. Pharmacol. Rev. 22, 355-388.
(18) Jarv, J., Aaviksaar, A., Godovikov, N., and Lobanov, D. (1977) The arrangement of substrate and organophosphorus-inhibitor leaving groups in the acetylcholinesterase active site. Biochem.
(19) Berman, H. A,, and Leonard, K. (1989) Chiral reactions of acetylcholinesterase probed with enantiomeric methyl- phosphonothioates. J . Biol. Chem. 264, 3942-3950.
(20) Asboth, B., and Polgar, L. (1983) Transition-state stabilization at the oxyanion binding sites of serine and thiol proteases: Hy- drolysis of thiono and oxygen esters. Biochemistry 22,117-122.
(21) Campbell, P., Nashed, N. T., Lapinskas, B. A., and Gurrieri, J. (1983) Thionesters as a probe for electrophilic catalysis in the
2,386-391.
phys. Acta 284, 427-434.
J. 167,823-825.
Chem. Res. Toxicol. 1992,5, 71-76 71
serine protease mechanism. J. Biol. Chem. 258, 59-66. (22) Fersht, A. (1985) Enzyme Structure and Mechanism, 2nd ed.,
pp 405-413, W. H. Freeman, New York. (23) Bowman, B. T., and Sans, W. W. (1983) Determination of
Octanol-water partitioning coefficients (KJ of 61 organo- phosphorus and carbamate insecticides and their relationship to
respective water solubility (S) values. J. Enuiron. Sci. Health B18,667-683.
(24) Mundy, R. L., Bowman, M. C., Farmer, J. H., and Haley, T. J. (1978) Quantitative Structure activity study of a series of sub- stituted 0,O-dimethyl 0-(p-nitropheny1)phosphorothioates and 0-analogs. Arch. Toxicol. 41, 111-123.
Human Serum Albumin-Benzo[ a Ipyrene anti-Diol Epoxide Adduct Structure Elucidation by Fluorescence Line Narrowing
Spectroscopy
Billy W. Day,+Jp§ Mark M. Doxtader," Robert H. Rich,$ Paul L. Skipper,' Kuldip Singh,II Ramachandra R. Dasari,ll and Steven R. Tannenbaum*yt*$
Department of Chemistry, Division of Toxicology, and George R. Harrison Spectroscopy Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139
Received September 16, 1991
Cryogenic (4-10 K) laser-induced vibrationless ground state and vibronic excited state fluorescence emission spectra of the adducts resulting from reaction in vitro of human serum albumin and the carcinogen (f)-r-7,t-8-dihydroxy-c-9,c-lO-epoxy-7,8,9,lO-tetr~ydrobenzo[a]- pyrene were recorded in order to determine the structures formed. Comparison of these fluorescence line-narrowed (FLN) spectra to those obtained from BaP-7,8,9,10-tetrahydrotetrols, synthetic N-t-BOC-alaninate ester, and N- and N"-histidine amine anti-BuPDE adducts revealed that a mixture of adduct types are formed with the protein. Extensive dialysis of the adducted protein simplified the FLN spectrum, causing it to become nearly identical to the FLN spectrum obtained from the stable peptide adduct. Comparison of the FLN spectra of the synthetic histidine adducts to those obtained from peptide adducts isolated from enzymic digestion of the adducted protein indicated that only one of the imidazole nitrogens is the nucleophile which forms a stable adduct with anti-BaPDE. The FLN studies confirm that "-histidine adducts are formed between human serum albumin and the C-10 position of anti-BaPDE.
Introduction Fluorescence line narrowing (FLN)' spectroscopy ori-
ginally arose from the desire to obtain narrow-band vi- brational excitation and emission spectra from molecular species which do not easily dinsolve in the paraffii solvents so successfully implemented as Shpol'skii matrices (1,2). An entirely different class of compounds from those studied by Shpol'skii spectroscopy is amenable to sensitive and selective analyses using FLN. Compounds in biolog- ical studies are usually only sparingly soluble in the hy- drophobic alkanes and require polar solvents which adopt a much less defined orientational structure in the solid state than do the alkanes. At cryogenic temperatures the chromophores & p e d in the heterogeneous solvents used for FLN reside in a number of energetically inequivalent microenvironments or "sites". Because of the relative orientation of the solute within the solvent, each slightly different microenvironment imparts a specific energy to
*Author 6 whom correspondence should be addressed at Room 56-309, Department of Chemistry, Division of Toxicology, Massa- chusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, MA 02139.
'Department of Chemistry. * Division of Toxicology. 8 Present address: Departments of Pharmaceutical Sciences and
Environmental & Occupational Health, University of Pittsburgh, Pittsburgh, PA 15261.
11 George R. Harrison Spectroscopy Laboratory.
the chromophore. A change of substituents on the same chromophore imparts additional differences in microen- vironments. The result is a collection of chromophores arising from structurally similar molecules which may undergo the same nominal transition, but only at specific excitation energies corresponding to their various rela- tionships to the host matrix. Narrow vibrational spectra are revealed in heterogeneous solvents only if narrow-band excitation, such as that originating from a laser, is used as a photon source. In addition, the temperatures ordi- narily required to achieve narrow lines in paraffi matrices are not adequately low to reduce phonon contributions from the heterogeneous glass matrices used in FLN spectroscopy. While the ultimate solvent system for FLN has not been rigorously defined with respect to molecular dimensions, it is often possible to find suitable solvents which minimize electron-phonon coupling.
FLN spectroscopy has been successfully applied to a variety of problems, including the analysis of coal tars (3) and petroleum distillates ( 4 5 ) and, because of its appli- cability to aqueous systems, the analysis of polar com- pounds in waste water (6). FLN has also been used to study fast kinetic processes (7,8) and the vibronic effects
Abbreviations: BaP, benzo[a]pyrene; anti-BaPDE, r-7,t-8-di- hydroxy-c-9,c-l0-epoxy-7,8,9,lO-tetrahydrobenzo[a]pyrene; BaP tetra- hydrotetrols, r-7,t-8,t-9,t-10- and r-7,t-8,t-9,c-lO-tetrahydroxy-7,8,9,10- tetrahydrobenzo[a]pyrenes; t-BOC, tert-butyloxycarbonyl; FLN, fluorescence line narrowing.
0893-228x/92/2705-0071$03.00/0 0 1992 American Chemical Society
