386
Chem. Res. Toxicol. 1989,2, 386-391
Preparation, Analysis, and Anticholinesterase Properties of 0 ,&Dimethyl Phosphorothioate Isomerides Charles M. Thompson,* Jeffrey A. Frick, Barbara C. Natke, and Linda K. Hansen Department of Chemistry, Loyola University of Chicago, Chicago, Illinois 60626 Received March 30, 1989
Previous studies have shown that impurities in commercial organophosphorus insecticides induce a variety of toxicological manifestations. Few studies have contrasted common impurity types and their comparative chemical and bioch2mical properties. In this study, five 0,O-dimethyl phosphorothioate compounds were converted to their corresponding 0,s-dimethyl phosphorothioates (isomerides) by a stepwise dealkylation-alkylation process (yields 58-76% ). The 0,S-isomerides and parent material were characterized by reverse-phase high-performance liquid chromatography (RPHPLC) and phosphorus (31P)nuclear magnetic resonance (NMR) spectroscopy. Methanol-water mixtures were found to adequately separate isomeride from parent structure with the isomeride eluting first. In general, the 0,s-isomerides were found to be shifted about 40 ppm upfield relative to the 0,O material. Isomerides were also determined to be significantly more potent as anticholinesterases (rat brain), with lzi values approximating 1000-fold those of the parent material.
Introduction Organophosphorus (OP) insecticides comprise a major portion of all currently applied insecticides and, generally, are in continued use as safe compounds for the control of agricultural pests. Most commercial OP insecticides are phosphorothioates (e.g., la-c) or phosphorodithioates (e.g., ld,e), the most notable dithioate being malathion (la) (see Figure 1). The toxicity of OP esters to insects, mammals, and other organisms can be mostly attributed to the inhibition of acetylcholinesterase (AChE) (I), the enzyme responsible for the hydrolysis of the neurotransmitter acetylcholine (ACh). Once inhibited, the phosphorylated AChE complex is stable and unable to participate in the hydrolysis of ACh leading to repetitive firing of the neural signal, possibly resulting in respiratory collapse and death. In addition to this major biochemical interaction (AChE inhibition), there is growing concern centered on the varied toxicological effects induced in warm-blooded mammals by OP compounds. For example, they are known to be mutagenic, teratogenic, and delayed neurotoxic (2). In many instances, these latter manifestations of exposure to OP’s are the result of “impurities” in the commercial sample rather than a direct consequence of the parent material. For example, specific delayed toxic effects due to trialkylphosphorothioate impurities in malathion have been examined (3-6). Although the exact origin of the impurities is unknown, there are ample opportunities for them to form during the preparation, storage, and environmental lifetime. The “isomerization” of alkoxy-thiophosphoryl t o alkylthio-phosphoryl linkage is one such example of an impurity-forming reaction (eq 1;see Figure 1). This particular rearrangement may occur thermally (7,8), may occur photochemically (7,9),or may be induced chemically (10). This chemical alteration was brought to public attention when 7500 Pakistan spraymen were exposed to 50% malathion; over 2800 became poisoned and 5 died (11). The
* Author to whom correspondence should be addressed. 0893-228x/89/2702-0386$01.50/0
cause of the poisoning was directly related to the corresponding S-methyl isomer, isomidathion (2d) (12,13).This atomic rearrangement apparently exacerbates the toxicity and may be explained, in part, by consideration of the phosphorylating species. First, conversion of the P=S to P=O results in enhanced electrophilicity at the phosphorus and, hence, increased reactivity toward nucleophiles (14). Second, the alkylthio group is a somewhat better leaving group than the corresponding methoxy (15). Third, the alkylthio group is susceptible to oxidation (16, 17), which may result in the formation of an even more labile leaving group (the sulfoxide). Lastly, alkylthio phosphorothioate compounds may be prone to “aging” (18, 19). Despite this strong evidence, there have been few systematic studies (7, 8 ) directed toward examining the “isomerization”products of phosphorothioates and related materials. The present study was undertaken to characterize and briefly investigate the relative AChE inhibitory potency of 0,O-dimethyl phosphorothioate isomerides.
Materials and Methods Chemicals. Acetylcholine iodide (ATCh-I), 5,5‘-dithiobis(2nitrobenzoic acid) (DTNB), and electric eel acetylcholinesterase (EEAChE; type VI-A) were obtained from Sigma Chemical Co. (St. Louis) and used directly without further purification. Azinphos and leptophos were purchased from Chem Services (Westchester, PA) and used without further purification. Parathion and fenitrothion were prepared from 0,O-dimethyl phosphorochloridothioateby standard procedures (20).Malathion and isomalathion were prepared by literature methods (3). Potassium ethyl xanthate was freshly prepared (21)prior to use. Dimethyl sulfate and iodomethane were purchased from Aldrich Chemical Co. (Milwaukee). Methamidophos was available from a previous study. Elemental analyses were performed by Midwest Microlab Ltd., Indianapolis, IN. HPLC solvents were obtained from EM Science and were filtered (Millipore; 0.2 pm)under vacuum prior to use. Analytical thin-layer chromatography (TLC) was conducted on E. Merck aluminum-backed, silica TLC plates with fluorescent indicator. The plates were visualized with DBQ (5% 2,4-dibromoquinone 4-chloroimine in ether) stain, UV light, or both. Flash chroma-
0 1989 American Chemical Society
Chem. Res. Toxicol., Vol. 2, No. 6, 1989 387
0,O-Dimethyl Phosphorothioate Isomerides
CH3C-
c1
SI P-X I
heat, light or
b
1
CH3S-P-X
I
chemically induced
OCH,
OCH,
la-e -
2a-e -
(1)
0
Figure 1. 0,O-Dimethyl phosphorothioates and their S-methyl isomerides. tography (22) was performed with Kieselgel60,0.04-0.06 mesh (Merck). Nuclear Magnetic Resonance. Proton ('H), carbon (13C), and phosphorus (31P)nuclear magnetic resonance (NMR) spectra were recorded on a Varian VXR 300-MHz instrument in deuterated chloroform (CDC13). Pertinent proton frequencies are tabulated in the following order: chemical shift (ppm in a), multiplicity (s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet), coupling constant (J in hertz), and number of hydrogens. Proton frequencies are relative to tetramethylsilane (TMS), and carbon frequencies are relative to the chloroform triplet (77.06 ppm) as internal standards. Phosphorus chemical shift data are relative to phosphoric acid (H3P04)in CDC13 as an external standard. Synthesis of Isomerides. General Procedure for the Preparation of 0,s-Dimethyl Phosphorothioates. To a 100-mL flask fitted with a reflux condenser, N2 inlet/outlet, and stir bar were added the requisite 0,O-dimethyl phosphorothioate (4 mmol) and 40 mL of anhydrous methanol. Potassium ethyl xanthate (5 mmol) was added and the solution refluxed for 4 h or until all the starting material appears consumed by TLC. The heat was removed and the solution allowed to come to room temperature. Dimethyl sulfate (5 mmol) was added and the heating resumed for 1h. The solution was cooled once more and rotary evaporated to a semisolid. The residue was taken up in a minumum amount of chloroform and flash chromatographed with ether/petroleum ether mixtures depending on the respective TLC properties (vide infra). Yields averaged about 70%. 0-(4-Nitrophenyl) 0-methyl S-methyl phosphorothioate (S-methyl methyl parathion) (23) (2a): yields 82%; R = 0.2 (1:lether/petroleum ether); 'H NMR 6 2.3 (d, J = 16,3 3.9 (d, J = 13, 3 H), 7.38 (d, J = 9.2, 2 H), 8.2 (d, J = 9, 2 H); 13C NMR 6 12.5, 54.4, 115.5, 121.0, 125.6, 125.9. 0-(3-Methyl-4-nitrophenyl) 0-methyl S-methyl phosphorothioate (S-methyl fenitrothion) (23) (2b): yields 76%; Rf = 0.07 (1:lether/petroleum ether); 'H NMR 6 2.2 (d, J = 16, 3 H), 2.5 (s, 3 H), 3.8 (d, J = 13, 3 H), 7.1 (d/s, 2 H), 7.0 (d, J = 10, 1 H); 13C NMR 6 12.4, 20.4, 54.2, 118.4, 123.8, 126.6, 136.2, 145.8, 152.9. 0-(2,4,5-Trichlorophenyl) 0-methyl S-methyl phosphorothioate (S-methyl fenchlorphos) (2c): yields 65%; Rf = 0.29 (1:l ether/petroleum ether); 'H NMR 6 2.3 (d, J = 16.1, 3 H), 3.9 (d, J = 12.8, 3 H), 7.5 (s, 1 H), 7.6 (s, 1 H); 13C NMR 6 12.7, 54.7, 123.2, 124.7, 129.6, 131.1, 131.6, 145.1; MS, m / z (M+)calcd, 321.8968; obsd, 321.8962 (-1.9 ppm). S -[(3,4-Dihydro-4-oxo-1,2,3-benzotriazin-3-yl)methyl] 0-methyl S-methyl phosphorothioate (S-methyl azinphos) (2e): yields 58%; Rf = 0.16 (100% diethyl ether); 'H NMR 6 2.3 (d, J = 17, 3 H), 3.7 (d, J = 13.0, 3 H), 5.8 (d, J = 16, 2 H), 7.8 (t,J = 7.5, 1 H), 7.95 (t, J = 8, 1 H), 8.1 (d, J = 7, 1 H), 8.3 (d, J = 7, 1 H); 13C NMR 6 13.2, 49.9, 53.8, 119.4, 125.1, 128.8, 132.7, 135.2, 143.9, 154.7. Anal. Calcd: C, 37.85; H, 3.81; N, 13.24. Found: C, 38.09; H, 3.94; N, 13.36.
k),
0,s-Dimethyl phenylphosphonothioate (4): yields 40%. In addition to the general procedure noted above, the following experiment was also conducted. Leptophos (4 mmol) was dissolved in 30 mL of dry methanol under nitrogen a t room temperature. Dimethyl sulfate (5 mmol) ww added and the reaction brought to reflux for 2 h. The mixture was cooled, the solvent removed by rotary evaporation, and the oil concentrated in vacuo overnight. The residue was flash chromatographed with 100% ethyl ether as the eluent. 'H NMR 6 2.25 (d, J = 1 8 , 3 H), 3.8 (d, J = 14, 3 H), 7.3-8.1 (m, 5 H). Chromatography. High-performance liquid chromatography (HPLC) was accomplished with a Beckman Model 334 HPLC system equipped with a system controller, variable-wavelength detector, and integrator-recorder. All separations and purity determinations were conducted on a 30 cm X 4.6 mm h g k 10-pm Spherisorb S10 ODS Workhorse using methanol-water eluent mixtures at a flow rate of 1.0 mL/min. Injection volumes were 3-10 pL. Detection was conducted at 270 nm for all compounds except malathion and isomalathion which were detected a t 240 nm. Triplicate samples were determined for each compound with a minimum reproducibility of f O . l min. Column equilibration with 100% methanol followed every sixth sample injection. Animals and Organ Procurement. The rat (Sprague-Dawley) starter colony was purchased from Charles River breeding labs (Indianapolis, IN) and was further bred on site. The rats were given free accesa to water and chow and were environmentally controlled for temperature, humidity, and light. Male rats with a body weight range of 150-200 g were sacrificed via decapitation and the brains excised, dried by paper towel, and weighed (range 1.01-1.92 g). The brains were then placed in pH = 7.6, 0.1 M phosphate buffer and chilled to 0 "C if stored. Cholinesterase Determinations. Rat brain acetylcholinesterase activity determinations were accomplished by our previously reported method (24). The procedure, with some minor changes, is presented here for clarity. Freshly excised brains (stored no more than 6 days) were homogenized in phosphate buffer (0.1 M, pH = 7.6) to give a final volume of 20 mL. This stock solution was kept at 0 "C up to 7 days. For analysis, a 1:4 dilution of stock solution to buffer was employed (approximating a hydrolysis rate of 0.04-0.07 A unit/min). This dilution was vortexed gently, and 1.09 mL of this sample was added to a test tube and placed in a 25 "C Forma Scientific constant-temperature shaker bath. A modified Ellman method (25) determined enzyme activity as follows. To each of six cuvettes was added 2.5 mL of DTNB solution (3.33 X M DTNB, 5.9 X M sodium bicarbonate; phosphate buffer) and 0.020 mL ATCh-I solution (7.5 X 10' M ATCh-I; phosphate buffer) and these cuvettes were placed in a Beckman DU-40 spectrophotometer equipped with a kinetic Soft-Pac module. From the test tube of brain homogenate 0.1 mL was added to cuvette 1 to serve as control. To the remaining 0.990 mL in the test tube 0.01 mL of the inhibitor (varied concentrations) was added and the solution gently vortexed. At 3,6,9,12, and 15 min, 0.1 mL of the homogenate-inhibitor solution was added to cuvettes 2, 3, 4, 5, and 6, respectively. The rate of hydrolysis of acetylthiocholine was monitored at 412 nm a t 60-5 intervals for 30 min from the addition of enzyme. The bimolecular inhibition constants (hi) were determined in triplicate by plotting the slopes against time, and the resulting slope was analyzed by linear regression. The enzyme dissociation constant (KD) and phosphorylation constant (k,) for S-methyl methyl parathion were determined by monitoring the loss of AChE activity over time at a variety of inhibitor concentrations (19). The inverse slopes of these rates were plotted against inverse concentration and evaluated as previously described (26, 27). Hydrolysis of S-Methyl Methyl Parathion. To three 10-mL vials, fitted with stir bars and septa, were added 4.9 mL of pH = 7.6 phosphate buffer, methanol, and acetone (control), respectively. Each of the vials was then charged with 100 p L of 1X M solution of 2a to give a final concentration of 2 X 10" M. The solutions were stirred at room temperature for 48 h while monitoring for the decrease in the 2a absorption and increase in p-nitrophenol absorption by the HPLC method previously described (70:30 methanol-water). The loss of 2a versus time was evaluated for reaction order, half-life, and rate by traditional
388 Chem. Res. Toxicol., Vol. 2, No. 6, 1989
Thompson et al.
Table I. HPLC and Phosphorus NMR Data
retention time, min no.
compound methyl parathion S-methi1 methyl parathion fenitrothion S-methyl fenitrothion fenchlorphos S-methyl fenchlorphos malathion S-methyl malathion azinphos S-methyl azinphos leptophos fonofos
la 2a lb 2b IC
2c Id 2d
le 2e 3 4 a
C H 3 0 H / H 2 0 (70:30) C H 3 0 H / H 2 0 (6535) 7.9 7.3 9.5 6.7 27.8 10.8 8.5 5.9 8.4 5.8
11.2 9.9 14.4 9.0 53.4 15.8 12.1 7.4 12.0 7.1
b b
b b
chemical shift, b 66.0 27.76 65.84 27.41 66.67 28.12 95.94 58.4 /57.05 96.23 57.45 88.73 107.81
Diastereomers. *Not determined.
methods. The slope was determined by linear regression analysis.
Results and Discussion We examined the isomerization, spectral and chromatographic properties, and biochemical interaction with AChE of five 0,O-dimethyl phosphorothioates la-e and their respective isomerides 2a-e (Figure 1). These five were chosen to represent thioate and dithioate compound classes of the symmetrical dialkyl type. The conversion of commercial material to 0,s-dimethyl phosphorothioates is of major concern since enhanced and, perhaps, different mechanisms of toxicity may be elicited by these compounds. The isomerides 2a-e were prepared by a one-pot dealkylation-realkylation of the corresponding 0,O-dimethyl phosphorothioate (eq 2). Potassium ethyl xanthate (PEX)
I
I
Y'r'OCH
CH,OH,
n
la-e -
CH3S-P-X
I
(2)
OCH, 2a-e -
was found to be superior to other demethylating agents (n-BUS-Na+,BBr3, etc.). Further, the use of PEX permitted direct realkylation without intermediate isolation. Of the methylating agents examined (CH31, CH3S03CH3, CF3S03CH3,etc.), dimethyl sulfate afforded the most reliable conversion while minimizing side reactions. All isomerides described were prepared by this optimized procedure. In related studies, we also noted that leptophos (3) is converted to 0,s-dimethyl phenylphosphonothioate (4) (eq 3) upon exposure to the aforementioned conditions. 1. EtOC(S)S-K+ 2. (CH,O),SO,
Br
-3 O ! - S C HI , OCH,
-4
(3)
This product presumably results from methanol-induced loss of the halogenated phenoxy moiety to give 1 (R = Ph), which then undergoes conversion to 4 in the usual manner (28). This suggestion is supported by the fact that leptophos converts into 4 by treatment with dimethyl sulfate in refluxing methanol (approximately 40% in 2 h). With this exception, this procedure appears suitable for the general conversion of 0-methyl thiophosphoryl to S-methyl thiophosphoryl compounds. A notable feature of this process is the avoidance of aqueous workup, which minimizes hydrolysis (7) and worker exposure. Immediately following purification, the isomerides were stored at -20 "C to suppress decomposition. Decomposition was an advance concern for 2a, which has been reported to exhibit questionable hydrolytic stability (7,29). The purity and structural integrity of the material were further confirmed by HPLC (30) and nuclear magnetic resonance (NMR) spectroscopy. The eluotropic properties of both parent and isomeride are listed in Table I. Two solvent systems were examined by using a reverse-phase column to ensure purity. As expected, all isomerides eluted correspondingly faster than the parent material, testimony to the more polar nature of the phosphoryl linkage as compared to thiophosphoryl. Of particular importance is the relative elution properties of the isomerides, which distinguishes the contribution of the leaving group, X. Thus, the rather lipophilic 2,4,5-trichlorophenyl lc/2c moiety significantly retarded the elution as compared to the 4-nitrophenyl derivatives or the dithioates (isomalathion and azinphos). This finding would suggest that isofenchlorphos ( 2 4 may be metabolized differently and at different rates owing to higher lipid solubility. The S-methyl and 0-methyl proton NMR absorbances of the isomerides appeared as doublets (JH+)at 6 2.2-2.3 and 3.7-3.9, respectively. The decoupled carbon NMR shows the S-methyl and 0-methyl peaks ranging from 6 12.4 to 13.2 and from 6 53.8 to 54.7, respectively. 31PNMR also proved useful in analyzing the isomerides (Table I). All isomerides were shifted approximately 40 ppm upfield relative to parent material. This finding is consistent with a previous report (31),and it is noteworthy that the (S-methyl) isomerides are readily differentiated from phosphonothioates [e.g., leptophos (3)] and phosphonodithioates [e.g., fonophos (4)]. Because of the isomeride sensitivity to acids and bases, we chose to employ an external standard (H3P04in deuterated chloroform) in our experiments. The chemical shifts reported are correspondingly affected. In addition to providing specific shift data, 31PNMR served to identify any phosphorus-containing impurity that may have resulted from the isomerization reaction sequence. In particular, we were concerned whether trialkyl
Chem. Res. Toxicol., Vol. 2, No. 6, 1989 389
0,O-Dimethyl Phosphorothioate Isomerides
no. la 2a lb 2b IC 2c Id 2d le 2e 3 4 5
Table 11. Bimolecular Inhibition Constants (k# ki X lo-', M-' compound min-' methyl parathion 0.691 (k4.3) S-methyl methyl parathion 714 fenitrothion
