J BIOCHEM TOXICOLOGY Volume 7, Number 4, 1992
Pharmacokinetics of Propetamphos Following Intravenous Administration in the F344 Rat Kelly Dix,'I2 Leo T . Burka,' and Walter C. Dauterman2 'National Institute of Environmental Health Sciences, Research Triangle Park, N C 27709; and 'Department of Toxicology, North Carolina State University, Raleigh, NC 27695
ABSTRACT: Propetamphos [(E)-1-methylethyl 3[[(ethylamino)methoxyphosphinothioylloxy1-2-b~-
tenoatel, the active ingredient in Safrotin,@ is an organophosphate developed by Sandoz, Ltd.@ (Switzerland) as an insecticide (1).Although metabolism of propetamphos has been previously investigated (2,3), there is no pharmacokinetic data available in the literature. The current studies were undertaken to investigate the pharmacokinetics of propetamphos following intravenous administration in male and female Fischer 344 (F344) rats. Rats were dosed via an indwelling jugular cannula at a dose of 12 mg/kg (one-tenth the oral LD-50). Blood samples were withdrawn via the cannula at predetermined timepoints to quantitate plasma concentrations of propetamphos over time. Propetamphos is highly bound to plasma proteins (free fraction = 0.06). Free propetamphos concentration in plasma vs. time data were analyzed by noncompartmental methods. The terminal elimination rate constant, A, was significantly different for males versus females (0.015 min-' for males and 0.037 min-' for females, p = 0.001). Plasma was cleared of unbound propetamphos at rates of 0.559 f 0.069 and 0.828 f 0.181 L/min/kg for males and females (mean 2 standard error). Mean residence times (MRTs) for propetamphos in the body for males and females were 28.3 f 5.7 and 14.4 +: 3.5 min, and the volume of distribution at steady state (Vss) was 14.7 2 2.6 and 12.3 2 4.5 L/kg. The differences in these parameters, clearance (CI), MRT, and Vss, were not statistically significant at the p < 0.05 level for males versus females, but MRT was nearly significantly different ( p = 0.08). Because of the rapid elimination of propetamphos from plasma following intravenous administration, it is unlikely that propetamphos would bioaccumulate in environmentally exposed animals. Although the pharmacokinetic parameters were not statistically different for males and females in these studies, there was a clear clinical difference in their sus-
ceptibility to propetamphos toxicity. Female rats presented with overt signs of organophosphate intoxication, whereas males were only slightly effected. The observed gender-related clinical difference in susceptibility to toxicity suggests that there may be a difference in the extent of elimination due to activation versus detoxication of propetamphos in males and females. Another possible explanation for the clinical difference in propetamphos toxicity is that inhibition of acetylcholinesterase by the activated, oxygenated form of propetamphos (propetamphos oxon) may be greater in females than in males. KEYWORDS: Noncompartmental Pharmacokinetics, Propetamphos, Rat.
INTRODUCTION
Received June 8, 1992. Address correspondence to Kelly Dix, National Institute of Environmental Health Sciences, Research Triangle Park, NC 27709.
Propetamphos, PRO (Figure l), is a vinyI organothiophosphate insecticide developed by Sandoz, Ltd., which is effective against a wide range of household pests, including cockroaches, flies, spiders, and mosquitoes. It is also marketed as a sheep dip (EctomortB) to protect against sheep scab, blowflies, ticks, and lice. The insecticide is used in Australia, France, Ireland, and the United Kingdom. Although PRO is not commercially available in the United States, it is used by licensed pest control personnel in this country. The E isomer, shown in Figure 1, is the active ingredient in Safrotin, and its reported acute oral LD50 in the rat is 119 mg/kg (1).The Z isomer has approximately the same insecticidal activity as the E isomer, but it is much more toxic to mammalian species (acute oral LD-50 in the rat is 8 mg/kg) (1). Pesticide formulations containing PRO consist only of the E isomer. The metabolism of PRO in three strains of housefly has been characterized by Wells et al. (2) and in the rat by Winkler and Patel (3). Propetamphos oxon (Figure 1) is the suspected toxic, anticholinesterase metabolite of propetam-
0 1992 VCH Publishers, Inc.
0887-2082/92/$3.50
+
.25
199
K. DIX ET AL.
200
Volume 7, Number 4, 1992
S
P-o\ CH,CH2-
matographic methods) and a new technique which couples in vivo microdialysis in the rat jugular vein with on-line tandem mass spectrometry (5,6).
I
H3C -O\ N/
I
H
/c= CH,
c
/H
O \C* H \O-CH
/CH3 \ CH,
PRO PETAM PHOS
0
I
H&-O\ CH3CHz-
P-o\ N/
I
H
/c=c CH,
/H
C‘ OR
\O-CH
/CH3 \
CH,
PROPETAMPHOS
OXON
FIGURE 1. Chemical structures of propetamphos (E isomer) and its suspected toxic metabolite, propetamphos oxon. Asterisks on carbons in propetamphos indicate positions of the radiolabel.
phos. Winkler and Pate1 were not able to detect propetamphos oxon in rats dosed orally with propetamphos, but this does not mean the oxon was not formed in vim. The major route of PRO metabolism in the rat is hydrolysis of the enol phosphate bond. In the housefly, hydrolysis also occurs, but there is a second major metabolic pathway that involves direct glutathione conjugation. Propetamphos oxon was detected in homogenates from the CSMA (Chemical Specialties Manufacturers Association) strain of housefly (susceptible to most insecticides) dosed with propetamphos but not in resistant strains. In vitro metabolism in mouse liver, housefly, and cockroach preparations has been investigated by Wells et al. (4). In those studies, propetamphos oxon was detected in micro soma1 preparations from houseflies and mouse livers in the presence of NADPH (reduced nicotine adenine dinucleotide phosphate). However, levels of the oxon were quite low in mouse liver preparations. The lack of pharmacokinetic data in the literature prompted the current studies. In conjunction with these studies, the pharmacokinetics of structurally similar vinyl organophosphates are being investigated by conventional methods (jugular vein cannulation followed by dose administration, blood withdrawal, and analysis by chro-
MATERIALS AND METHODS Animals Six male (177 to 321 g) and five female (163 to 204 g) Fischer 344 (F344) rats were obtained from Charles River Breeding Laboratory (Raleigh, NC) at least 1 week prior to treatment. Animals were housed in animal quarters with a 12 h light/dark cycle and received NIH-31 Rodent Chow and water ad libitum. The animal quarters were maintained at 10% relative humidity. 21-22°C and 50
*
Chemicals Unlabeled and I4C-labeled propetamphos (labeled in the carbonyl and vinyl groups, as shown in Figure 1, specific activity = 19.8 pCi/mg) were received as gifts from Sandoz, Ltd. High performance liquid chromatography (HPLC) grade water and acetonitrile were purchased from J.T. Baker, Inc. (Phillipsburg, NJ), trifluoroacetic acid was purchased from Fluka (Ronkonkoma, NY), and heparin sodium was purchased from Eastman Kodak Company (Rochester, NY).
Chemical Treatment Propetamphos was purified by HPLC (CIScolumn; isocratic, 60% water, 40% acetonitrile; 1 mL/ min; UV = 220 nm). 14C-propetamphos and unlabeled propetamphos or unlabeled propetamphos alone was mixed with a 1:1 ethanol :emulphor SOlution. Water was added so that the final dosing solution contained 12 mg propetamphos per milliliter of 8 :l :l water :ethanol: emulphor. On the day prior to chemical treatment, the right jugular vein of each animal was cannulated using the method described by Harms and Ojeda (7). Following surgery, animals were housed in individual metabolism cages for ease in handling. Animals were allowed free access to food and water for the remainder of the experiment. On the day of treatment, animals were dosed intravenously with propetamphos (12 mg/kg) via the indwelling jugular cannula. Blood samples were withdrawn at 2, 4, 6, 8, 10, 15, 20, 30, and 45 min and at 1, 1.25, 1.5, 1.75, 2, 2.5, and 3 h. The blood volume removed from the animal at each sample withdrawal (approximately 100 pL) was replaced as heparinized
Volume 7, Number 4, 1992
saline (10 units heparin per milliliter physiological saline).
Preliminary Studies In preliminary studies, blood was also withdrawn from the indwelling jugular cannula at 4, 6, 8, 10, 12, and 24 h. The purpose of these studies was to determine how long radioactivity remained in blood. This information allowed us to determine the duration of sampling in subsequent pharmacokinetic studies in which unlabeled PRO was administered.
Plasma Protein Binding Blood was obtained from donor male F344 rats by heart puncture, and the heparinized blood was centrifuged to separate plasma from red blood cells. Five PRO solutions were prepared in 8 : l : l water: ethanol :emulphor at concentrations of 10, 5, 2.5, 1.25, and 0.625 mg/mL. Approximately 1 pL of '4C-propetamphos (negligible mass) was added to each dilution, and triplicates of each dilution were counted in a Beckman LS 9800 liquid scintillation counter (Beckman Instruments, Inc., Fullerton, CA). Twenty microliters from the appropriate PRO dilution was added to five l-mL plasma aliquots so that final propetamphos concentrations were 196, 98, 49, 25, and 12 pg/mL. These concentrations are total PRO concentrations in plasma (i.e., bound plus free). Triplicate aliquots of each plasma sample were counted. Two 400-pL aliquots of each plasma sample were pipetted into separate Centricon-10 microconcentrators (Amicon, Beverly, MA). The Centricon-10 tubes were then centrifuged in a Sorvall RC-2B centrifuge (Sorvall, Inc., Newton, CT) in a fixed angle (28") SM-24 rotor at 4400 X g for approximately 5 min. Triplicate 10-pL aliquots of filtrate were counted, and the concentration of propetamphos in the filtrate was assumed to be the free propetamphos concentration in plasma. Free fraction (fu) was determined as free propetamphos concentration in plasma divided by total propetamphos concentration in plasma. Propetamphos appeared to be stable in plasma over time (data not shown).
PROPETAMPHOS PHARMACOKINETICS
201
g in a tabletop microfuge (Eppendorf Centrifuge Model 5415 C, Eppendorf-Netheler-Hinz, Hamburg, Germany) for 5 min to obtain plasma. Fifty microliters of plasma were pipetted into a new microfuge tube, and plasma proteins were precipitated with either cold acetonitrile alone, 500 or 200 pL, or cold acetonitrile containing benzophenone as an internal standard (51 mg/mL), 200 pL. Samples were vortexed then centrifuged for 10 min at 16000 x g to precipitate plasma proteins and form a pellet. Supernatants from radiolabeled samples were counted in a Beckman LS 9800 liquid scintillation counter. The protein precipitation procedure resulted in 90-95% recovery. Because of the high efficiency of the precipitation procedure, propetamphos concentrations in supernatants were not corrected for efficiency. Samples were stored on ice until they were transferred to a refrigerator. Total propetamphos concentration in plasma was determined by HPLC analysis of supernatant and quantitated using a standard curve. The HPLC system consisted of a Waters Associates (Milford, MA) chromatography pump (Model 6000a Solvent Delivery System), U6K injector, UV spectrophotometer (Lambda-Max, Model 481), and data module (Model 730). A Microsorb C,, column (Rainin Instrument Co., Woburn, MA) was used with an isocratic solvent system (45:55, A = 0.1% trifluoroacetic acid, B = acetonitrile; flow rate = 1.5 mL/ min) and ultraviolet (UV) detection at 210 nm. Free propetamphos concentration in plasma was calculated as total propetamphos concentration multiplied by fu.
Pharmacokinetic Analysis Free propetamphos concentration in plasma vs. time data were analyzed by noncompartmental methods (8). Terminal elimination rate constants, A, were determined graphically on the natural logarithm of free concentration versus time curve. Area under the curve (AUC) and area under the first moment curve (AUMC) were calculated by the trapezoidal method with extrapolation to t = to. Clearance (Cl) and mean residence time (MRT) were calculated as C1 = dose/AUC and MRT = AUMC/ AUC. Volume of distribution at steady state (Vss) was calculated as Vss = C1. MRT.
Sample Preparation and Analysis Blood samples were withdrawn from each animal at specified timepoints via the indwelling cannula, and then transferred to a polypropylene microfuge tube and stored on ice or immediately centrifuged. Samples were centrifuged at 16000 X
Statistical Analysis Mean values of pharmacokinetic parameters (A, AUC, C1, MRT, and Vss) for males and females were compared using Student's t-test and considered significantly different if p < 0.05.
202
K. DIX ET AL.
Volume 7, Number 4, 1992
TABLE 1. Free Fraction of Propetamphos in Plasma" Concentration ( p g / m L ) 196 98 49 25 12
10
CI
i
A
E"
Free Fraction
Y
male 1
w male.!
male2
male:
a
0.08 0.06 0.06 ND' 0.09
-n5m
C
0 ~
"Plasma samples containing '*C-propetamphos were centrifuged in Centricon-10 tubes at 4400 x for 5 min, and free fraction in plasma was calculated as the ratio of C-propetamphos concentrations in the filtrate vs. the sample reservoir. "No radioactivity detected in filtrate.
6
c
8
1
0 U P al
2
c
-001 -01 0
30
60
120 1 5 0 180 210 240
time (min)
RESULTS Results from plasma protein binding studies are shown in Table 1. Propetamphos is highly bound to plasma proteins over a total plasma concentration range of 12-196 pg/mL (fu = 0.06). Free concentrations of propetamphos in plasma vs. time curves following intravenous administration of propetamphos (12 mg/kg) for males and females are shown in Figure 2. Propetamphos was detected and quantitated by HPLC analysis with UV detection at 210 nm. In preliminary studies, 14Cpropetamphos was administered intravenously to determine how long radioactivity remained in plasma. The duration of sampling in subsequent pharmacokinetic experiments, which used unlabeled propetamphos, was based on radioactivity data from earlier studies. Pharmacokinetic parameters for propetamphos following intravenous (iv) administration are shown in Table 2. Propetamphos was rapidly cleared from plasma following iv administration (Cl = 0.559 and 0.828 L/min/kg for males and females, respectively). In males, propetamphos had a MRT of 28.3 min, and for females, MRT = 14.4 min. Mean AUC values for males and females were 23.5 and 18.2 min-mg/L, respectively. The Vss for males was 14.7 L/kg, and for females, Vss = 12.3 L/kg. There was no statistically significant difference between males and females for any of these pharmacokinetic parameters (AUC, CI, MRT, V s s ) at the p < 0.05 level. However, the terminal elimination rate constants were significantly different (A = 0.015 and 0.037 min-' for males and females, respectively; p = 0.001) and MRTs were nearly significantly different ( p = 0.08). Clinically, females appeared to be much more susceptible to the organophosphate toxicity than did males. Lacrimation, salivation, and trembling were observed in females. Males exhibited some signs of organophosphate toxicity but to a much lesser extent, and
90
-
i
10
B
E"
0
Y
female 1 female2
0 0
female3 female4
m
-a5m C
0
c
8 0 U
n a,
t
c
1
.001 .O1 0
30
60
90
1 2 0 150 180 210 240
time (min)
FIGURE 2. Free concentration of propetamphos in plasma vs. time for (A) male and (B) female F344 rats. Propetamphos was administered intravenously (12 mg/kg) via an indwelling jugular vein cannula and quantitated by HPLC analysis of the parent compound, as described in the text.
TABLE 2. Noncompartmental Pharmacokinetic Parameters" for Male and Female F344 Rats Following Intravenous Administration of Propetamphos (12 mg/kg) Males (n = 6 )
Parameters
Frrnales ( n = 5)
~~~~~
A (min-')
AUC (min.mg/L) C1 (L/min/kg) MRT (min) vss ( L / W
*
0.015 0.002' 23.5 t 3.2 0.559 & 0.069 28.3 2 5.7 14.7 t 2.6
0.037 ? 0.004h 18.2 * 4.5 0 828 2 0.181 14.4 3.5 12.3 t 4.5
*
'mean 5 standard error h p = 0.001. Parameters were compared for males and females using Student's ttest and considered significantly different if p < 0.05. The value of A was significantly different ( p = 0.001) and the MRT was nearly significantly different ( p = 0.08) for males and females.
Volume 7, Number 4, 1992
they generally appeared to recover within 15 min of dosing.
DISCUSSION Plasma PRO concentration vs. time data were analyzed by noncompartmental methods in these studies. Noncompartmental analysis was used for several reasons. First, the necessary data for a physiologically based pharmacokinetic model were lacking (e.g., blood/tissue partition coefficients, metabolic rate constants, velocities, etc.), so the development of such a model was not attempted. Second, although noncompartmental analysis is not as widely used as classical compartmental models, it is gaining favor with pharmacokineticists and classical compartmental models are being used less frequently. Initially, PRO data were fit to compartmental models (both one- and two-compartment open models were attempted) using the computer program NONLIN84 (9). However, the resulting errors in parameter estimates (10-50%) were not acceptable. In addition, NONLIN84 is sensitive to variability in concentration data in the terminal elimination phase of the decay curve. For one of the data sets, NONLIN84 converged to negative numbers for the terminal elimination rate constant and elimination half-life. Another reason for using noncompartmental analysis is that when using compartmental models, it is often the case that all animals in a study do not conform to the same model (10). Thus, the pharmacokinetic parameters obtained are not shared by all animals in the study. For example, animals which are best described by a two-compartment model have parameters describing a distribution phase. But parameters describing a distribution phase are not obtained from animal data that are fit to a one-compartment model, because no such phase exists in a one-compartment model. It is difficult to make statistical comparisons when some of the parameters are not shared by all the data sets. In noncompartmental analysis, the same parameters are estimated for all animals in the study and statistical comparisons can easily be made. Except for the terminal elimination rate constants, PRO pharmacokinetic parameters in male and female rats following iv administration of 12 mg/kg are not signhcantly different. In both sexes, a short MRT and rapid plasma C1 were observed for PRO. This supports the rapid elimination of PRO-derived radioactivity reported by Winkler and Patel (3). Those investigators have shown that the majority of PRO-derived radioactivity is elimi-
PROPETAMPHOS PHARMACOKINETICS
203
nated as 14C0, following a single oral dose in female rats. As the dose increased (0.6, 6, and 16 mg/kg), the percent of radioactivity eliminated as exhaled carbon dioxide decreased (80, 60, and 40%, respectively) with corresponding increases in radioactivity excreted in urine (12, 20, and 38%).This suggests that the rapid oxidative detoxication pathways leading to C 0 2 formation observed at lower doses are saturated at higher doses. The urinary metabolites were primarily volatile hydrolysis products containing acetone and acetoacetate moieties resulting from cleavage of the P-0-vinyl bond. Desmethyl and desisopropyl propetamphos represented a minor fraction of administered radioactivity, and no PRO oxon or N-desethyl propetamphos was found. Total excretion of PRO-derived radioactivity was the same over the dose range studied, but there were differences in the ratio of 0-dealkylation (desmethyl and desisopropyl PRO) vs. hydrolysis products (carbon dioxide and other volatiles). Winkler and Patel (3) also showed that following multiple oral doses of PRO (6.4 mg/kg), male and female rats excreted 60 and 40% of administered radioactivity as exhaled carbon dioxide, respectively, and females excreted more of the dose in urine than did males (44 versus 33%). Total elimination of PRO-derived radioactivity, however, was approximately the same for males and females at this dose. The iv dose administered in the current studies was 12 mg/kg. Only parent PRO was measured, and, although they are fairly low, the resulting plasma levels of PRO are likely to be high enough to saturate detoxication pathways leading to carbon dioxide formation. A marked gender-related clinical difference in susceptibility to PRO toxicity was observed at this iv dose. Females presented with overt signs of organophosphate intoxication, whereas males were effected to a much lesser extent. This is also consistent with the Winkler and Patel study. The iv PRO dose of 12 mg/kg was selected for two reasons. First, it was well below the reported oral LD-50 and presumably would not be a lethal dose. The 12 mg/kg dose did result in observed clinical toxicity. Second, very little radiolabeled PRO was available and the HPLC assay was developed for unlabeled PRO with UV detection. The sensitivity of the assay was such that if a lower iv dose were administered, PRO would probably not be consistently detected in plasma/acetonitrile supernatants. If a higher dose were administered, toxicity would interfere with the pharmacokinetic study. Therefore, studying PRO pharmacokinetics over a range of doses was not practical. Similarly,
204
K. DIX ETAL.
oral administration of PRO at the dose used in iv studies would likely result in very low plasma levels (i.e., undetectable by UV methods) of the parent compound. Because of the rapid rate of elimination in both sexes, it is unlikely that propetamphos would bioaccumulate in environmentally exposed animals unless repeated exposures occurred at very short time intervals. However, such repeated exposures would probably result in death because of the neurotoxic nature of organophosphate action, namely, cholinesterase inhibition.
ACKNOWLEDGMENT Portions of this work were supported by PHS grants ES-07046 and ES-00044. The authors would like to thank Adrian Phillips for her excellent technical assistance.
REFERENCES 1. J. P. Leber (1972). New class of vinyl thionophosphate insecticides. In Insecticides, Proceedings of the
2nd International IUPAC Congress of Pesticide Chemistry, A. S . Tahori, ed., vol. 1, pp. 381-401, Gordon and Breach, New York, NY.
Volume 7,Number 4, 1992
2. D. S. Wells, N. Motoyama, and W. C. Dauterman (1986a). The in vivo metabolism of propetamphos by insecticide-resistant and susceptible strains of the housefly, Musca domestica. Pestic. Sci., 17, 631-640. 3. V. W. Winkler and J. R. Patel (1982). Metabolism of propetamphos in the rat after single and multiple dosing. In The Fifth international Congress of Pesticide Chemistry (IUPAC), Abstracts VE-20, Human HealthEnvironment-Pesticides, August 29-September 4, 1982, Kyoto, Japan. 4. D. S. Wells, L. M. Afifi, N. Motoyama, and W. C. Dauterman (1986b). In vitro metabolism of propetamphos by housefly, cockroach, and mouse liver preparations. J. Agric. Food Chem., 34, 79-86. 5. L. J. Deterding, K. Dix, L. T. Burka, and K. B. Tomer (1992). On-line coupling of in vivo microdialysis with tandem mass spectrometry. Anal. Chem., 64(21), 2636-2641. 6 . K. Dix, L. J. Deterding, L. T. Burka, and K. B. Tomer (in press). Tris(2-chloroethyl) phosphate pharmacokinetics in the Fischer 344 rat: a comparison of conventional methods and in vivo microdialysis coupled with tandem mass spectrometry. 7. P. G. Harms and S. R. Ojeda (1974). A rapid and simple procedure for chronic cannulation of the rat jugular vein. J. A p p l . Phys., 36(3), 391-392. 8. M. Gibaldi and D. Perrier (1982). Pharmacokinetics, 2nd ed., Marcel Dekker, Inc., New York, NY. 9. Statistical Consultants, Inc. (1986). PCNONLIN and NONLIN84: software for the statistical analysis of nonlinear models. A m . Statistician, 40(1). 10. W. R. Gillespie (1991). Noncompartmental versus compartmental modelling in clinical pharmacokinetics. Clin. Pharrnacokinetics, 20(4), 253-262.
Pharmacokinetics of Propetamphos Following Intravenous Administration in the F344 Rat Kelly Dix,'I2 Leo T . Burka,' and Walter C. Dauterman2 'National Institute of Environmental Health Sciences, Research Triangle Park, N C 27709; and 'Department of Toxicology, North Carolina State University, Raleigh, NC 27695
ABSTRACT: Propetamphos [(E)-1-methylethyl 3[[(ethylamino)methoxyphosphinothioylloxy1-2-b~-
tenoatel, the active ingredient in Safrotin,@ is an organophosphate developed by Sandoz, Ltd.@ (Switzerland) as an insecticide (1).Although metabolism of propetamphos has been previously investigated (2,3), there is no pharmacokinetic data available in the literature. The current studies were undertaken to investigate the pharmacokinetics of propetamphos following intravenous administration in male and female Fischer 344 (F344) rats. Rats were dosed via an indwelling jugular cannula at a dose of 12 mg/kg (one-tenth the oral LD-50). Blood samples were withdrawn via the cannula at predetermined timepoints to quantitate plasma concentrations of propetamphos over time. Propetamphos is highly bound to plasma proteins (free fraction = 0.06). Free propetamphos concentration in plasma vs. time data were analyzed by noncompartmental methods. The terminal elimination rate constant, A, was significantly different for males versus females (0.015 min-' for males and 0.037 min-' for females, p = 0.001). Plasma was cleared of unbound propetamphos at rates of 0.559 f 0.069 and 0.828 f 0.181 L/min/kg for males and females (mean 2 standard error). Mean residence times (MRTs) for propetamphos in the body for males and females were 28.3 f 5.7 and 14.4 +: 3.5 min, and the volume of distribution at steady state (Vss) was 14.7 2 2.6 and 12.3 2 4.5 L/kg. The differences in these parameters, clearance (CI), MRT, and Vss, were not statistically significant at the p < 0.05 level for males versus females, but MRT was nearly significantly different ( p = 0.08). Because of the rapid elimination of propetamphos from plasma following intravenous administration, it is unlikely that propetamphos would bioaccumulate in environmentally exposed animals. Although the pharmacokinetic parameters were not statistically different for males and females in these studies, there was a clear clinical difference in their sus-
ceptibility to propetamphos toxicity. Female rats presented with overt signs of organophosphate intoxication, whereas males were only slightly effected. The observed gender-related clinical difference in susceptibility to toxicity suggests that there may be a difference in the extent of elimination due to activation versus detoxication of propetamphos in males and females. Another possible explanation for the clinical difference in propetamphos toxicity is that inhibition of acetylcholinesterase by the activated, oxygenated form of propetamphos (propetamphos oxon) may be greater in females than in males. KEYWORDS: Noncompartmental Pharmacokinetics, Propetamphos, Rat.
INTRODUCTION
Received June 8, 1992. Address correspondence to Kelly Dix, National Institute of Environmental Health Sciences, Research Triangle Park, NC 27709.
Propetamphos, PRO (Figure l), is a vinyI organothiophosphate insecticide developed by Sandoz, Ltd., which is effective against a wide range of household pests, including cockroaches, flies, spiders, and mosquitoes. It is also marketed as a sheep dip (EctomortB) to protect against sheep scab, blowflies, ticks, and lice. The insecticide is used in Australia, France, Ireland, and the United Kingdom. Although PRO is not commercially available in the United States, it is used by licensed pest control personnel in this country. The E isomer, shown in Figure 1, is the active ingredient in Safrotin, and its reported acute oral LD50 in the rat is 119 mg/kg (1).The Z isomer has approximately the same insecticidal activity as the E isomer, but it is much more toxic to mammalian species (acute oral LD-50 in the rat is 8 mg/kg) (1). Pesticide formulations containing PRO consist only of the E isomer. The metabolism of PRO in three strains of housefly has been characterized by Wells et al. (2) and in the rat by Winkler and Patel (3). Propetamphos oxon (Figure 1) is the suspected toxic, anticholinesterase metabolite of propetam-
0 1992 VCH Publishers, Inc.
0887-2082/92/$3.50
+
.25
199
K. DIX ET AL.
200
Volume 7, Number 4, 1992
S
P-o\ CH,CH2-
matographic methods) and a new technique which couples in vivo microdialysis in the rat jugular vein with on-line tandem mass spectrometry (5,6).
I
H3C -O\ N/
I
H
/c= CH,
c
/H
O \C* H \O-CH
/CH3 \ CH,
PRO PETAM PHOS
0
I
H&-O\ CH3CHz-
P-o\ N/
I
H
/c=c CH,
/H
C‘ OR
\O-CH
/CH3 \
CH,
PROPETAMPHOS
OXON
FIGURE 1. Chemical structures of propetamphos (E isomer) and its suspected toxic metabolite, propetamphos oxon. Asterisks on carbons in propetamphos indicate positions of the radiolabel.
phos. Winkler and Pate1 were not able to detect propetamphos oxon in rats dosed orally with propetamphos, but this does not mean the oxon was not formed in vim. The major route of PRO metabolism in the rat is hydrolysis of the enol phosphate bond. In the housefly, hydrolysis also occurs, but there is a second major metabolic pathway that involves direct glutathione conjugation. Propetamphos oxon was detected in homogenates from the CSMA (Chemical Specialties Manufacturers Association) strain of housefly (susceptible to most insecticides) dosed with propetamphos but not in resistant strains. In vitro metabolism in mouse liver, housefly, and cockroach preparations has been investigated by Wells et al. (4). In those studies, propetamphos oxon was detected in micro soma1 preparations from houseflies and mouse livers in the presence of NADPH (reduced nicotine adenine dinucleotide phosphate). However, levels of the oxon were quite low in mouse liver preparations. The lack of pharmacokinetic data in the literature prompted the current studies. In conjunction with these studies, the pharmacokinetics of structurally similar vinyl organophosphates are being investigated by conventional methods (jugular vein cannulation followed by dose administration, blood withdrawal, and analysis by chro-
MATERIALS AND METHODS Animals Six male (177 to 321 g) and five female (163 to 204 g) Fischer 344 (F344) rats were obtained from Charles River Breeding Laboratory (Raleigh, NC) at least 1 week prior to treatment. Animals were housed in animal quarters with a 12 h light/dark cycle and received NIH-31 Rodent Chow and water ad libitum. The animal quarters were maintained at 10% relative humidity. 21-22°C and 50
*
Chemicals Unlabeled and I4C-labeled propetamphos (labeled in the carbonyl and vinyl groups, as shown in Figure 1, specific activity = 19.8 pCi/mg) were received as gifts from Sandoz, Ltd. High performance liquid chromatography (HPLC) grade water and acetonitrile were purchased from J.T. Baker, Inc. (Phillipsburg, NJ), trifluoroacetic acid was purchased from Fluka (Ronkonkoma, NY), and heparin sodium was purchased from Eastman Kodak Company (Rochester, NY).
Chemical Treatment Propetamphos was purified by HPLC (CIScolumn; isocratic, 60% water, 40% acetonitrile; 1 mL/ min; UV = 220 nm). 14C-propetamphos and unlabeled propetamphos or unlabeled propetamphos alone was mixed with a 1:1 ethanol :emulphor SOlution. Water was added so that the final dosing solution contained 12 mg propetamphos per milliliter of 8 :l :l water :ethanol: emulphor. On the day prior to chemical treatment, the right jugular vein of each animal was cannulated using the method described by Harms and Ojeda (7). Following surgery, animals were housed in individual metabolism cages for ease in handling. Animals were allowed free access to food and water for the remainder of the experiment. On the day of treatment, animals were dosed intravenously with propetamphos (12 mg/kg) via the indwelling jugular cannula. Blood samples were withdrawn at 2, 4, 6, 8, 10, 15, 20, 30, and 45 min and at 1, 1.25, 1.5, 1.75, 2, 2.5, and 3 h. The blood volume removed from the animal at each sample withdrawal (approximately 100 pL) was replaced as heparinized
Volume 7, Number 4, 1992
saline (10 units heparin per milliliter physiological saline).
Preliminary Studies In preliminary studies, blood was also withdrawn from the indwelling jugular cannula at 4, 6, 8, 10, 12, and 24 h. The purpose of these studies was to determine how long radioactivity remained in blood. This information allowed us to determine the duration of sampling in subsequent pharmacokinetic studies in which unlabeled PRO was administered.
Plasma Protein Binding Blood was obtained from donor male F344 rats by heart puncture, and the heparinized blood was centrifuged to separate plasma from red blood cells. Five PRO solutions were prepared in 8 : l : l water: ethanol :emulphor at concentrations of 10, 5, 2.5, 1.25, and 0.625 mg/mL. Approximately 1 pL of '4C-propetamphos (negligible mass) was added to each dilution, and triplicates of each dilution were counted in a Beckman LS 9800 liquid scintillation counter (Beckman Instruments, Inc., Fullerton, CA). Twenty microliters from the appropriate PRO dilution was added to five l-mL plasma aliquots so that final propetamphos concentrations were 196, 98, 49, 25, and 12 pg/mL. These concentrations are total PRO concentrations in plasma (i.e., bound plus free). Triplicate aliquots of each plasma sample were counted. Two 400-pL aliquots of each plasma sample were pipetted into separate Centricon-10 microconcentrators (Amicon, Beverly, MA). The Centricon-10 tubes were then centrifuged in a Sorvall RC-2B centrifuge (Sorvall, Inc., Newton, CT) in a fixed angle (28") SM-24 rotor at 4400 X g for approximately 5 min. Triplicate 10-pL aliquots of filtrate were counted, and the concentration of propetamphos in the filtrate was assumed to be the free propetamphos concentration in plasma. Free fraction (fu) was determined as free propetamphos concentration in plasma divided by total propetamphos concentration in plasma. Propetamphos appeared to be stable in plasma over time (data not shown).
PROPETAMPHOS PHARMACOKINETICS
201
g in a tabletop microfuge (Eppendorf Centrifuge Model 5415 C, Eppendorf-Netheler-Hinz, Hamburg, Germany) for 5 min to obtain plasma. Fifty microliters of plasma were pipetted into a new microfuge tube, and plasma proteins were precipitated with either cold acetonitrile alone, 500 or 200 pL, or cold acetonitrile containing benzophenone as an internal standard (51 mg/mL), 200 pL. Samples were vortexed then centrifuged for 10 min at 16000 x g to precipitate plasma proteins and form a pellet. Supernatants from radiolabeled samples were counted in a Beckman LS 9800 liquid scintillation counter. The protein precipitation procedure resulted in 90-95% recovery. Because of the high efficiency of the precipitation procedure, propetamphos concentrations in supernatants were not corrected for efficiency. Samples were stored on ice until they were transferred to a refrigerator. Total propetamphos concentration in plasma was determined by HPLC analysis of supernatant and quantitated using a standard curve. The HPLC system consisted of a Waters Associates (Milford, MA) chromatography pump (Model 6000a Solvent Delivery System), U6K injector, UV spectrophotometer (Lambda-Max, Model 481), and data module (Model 730). A Microsorb C,, column (Rainin Instrument Co., Woburn, MA) was used with an isocratic solvent system (45:55, A = 0.1% trifluoroacetic acid, B = acetonitrile; flow rate = 1.5 mL/ min) and ultraviolet (UV) detection at 210 nm. Free propetamphos concentration in plasma was calculated as total propetamphos concentration multiplied by fu.
Pharmacokinetic Analysis Free propetamphos concentration in plasma vs. time data were analyzed by noncompartmental methods (8). Terminal elimination rate constants, A, were determined graphically on the natural logarithm of free concentration versus time curve. Area under the curve (AUC) and area under the first moment curve (AUMC) were calculated by the trapezoidal method with extrapolation to t = to. Clearance (Cl) and mean residence time (MRT) were calculated as C1 = dose/AUC and MRT = AUMC/ AUC. Volume of distribution at steady state (Vss) was calculated as Vss = C1. MRT.
Sample Preparation and Analysis Blood samples were withdrawn from each animal at specified timepoints via the indwelling cannula, and then transferred to a polypropylene microfuge tube and stored on ice or immediately centrifuged. Samples were centrifuged at 16000 X
Statistical Analysis Mean values of pharmacokinetic parameters (A, AUC, C1, MRT, and Vss) for males and females were compared using Student's t-test and considered significantly different if p < 0.05.
202
K. DIX ET AL.
Volume 7, Number 4, 1992
TABLE 1. Free Fraction of Propetamphos in Plasma" Concentration ( p g / m L ) 196 98 49 25 12
10
CI
i
A
E"
Free Fraction
Y
male 1
w male.!
male2
male:
a
0.08 0.06 0.06 ND' 0.09
-n5m
C
0 ~
"Plasma samples containing '*C-propetamphos were centrifuged in Centricon-10 tubes at 4400 x for 5 min, and free fraction in plasma was calculated as the ratio of C-propetamphos concentrations in the filtrate vs. the sample reservoir. "No radioactivity detected in filtrate.
6
c
8
1
0 U P al
2
c
-001 -01 0
30
60
120 1 5 0 180 210 240
time (min)
RESULTS Results from plasma protein binding studies are shown in Table 1. Propetamphos is highly bound to plasma proteins over a total plasma concentration range of 12-196 pg/mL (fu = 0.06). Free concentrations of propetamphos in plasma vs. time curves following intravenous administration of propetamphos (12 mg/kg) for males and females are shown in Figure 2. Propetamphos was detected and quantitated by HPLC analysis with UV detection at 210 nm. In preliminary studies, 14Cpropetamphos was administered intravenously to determine how long radioactivity remained in plasma. The duration of sampling in subsequent pharmacokinetic experiments, which used unlabeled propetamphos, was based on radioactivity data from earlier studies. Pharmacokinetic parameters for propetamphos following intravenous (iv) administration are shown in Table 2. Propetamphos was rapidly cleared from plasma following iv administration (Cl = 0.559 and 0.828 L/min/kg for males and females, respectively). In males, propetamphos had a MRT of 28.3 min, and for females, MRT = 14.4 min. Mean AUC values for males and females were 23.5 and 18.2 min-mg/L, respectively. The Vss for males was 14.7 L/kg, and for females, Vss = 12.3 L/kg. There was no statistically significant difference between males and females for any of these pharmacokinetic parameters (AUC, CI, MRT, V s s ) at the p < 0.05 level. However, the terminal elimination rate constants were significantly different (A = 0.015 and 0.037 min-' for males and females, respectively; p = 0.001) and MRTs were nearly significantly different ( p = 0.08). Clinically, females appeared to be much more susceptible to the organophosphate toxicity than did males. Lacrimation, salivation, and trembling were observed in females. Males exhibited some signs of organophosphate toxicity but to a much lesser extent, and
90
-
i
10
B
E"
0
Y
female 1 female2
0 0
female3 female4
m
-a5m C
0
c
8 0 U
n a,
t
c
1
.001 .O1 0
30
60
90
1 2 0 150 180 210 240
time (min)
FIGURE 2. Free concentration of propetamphos in plasma vs. time for (A) male and (B) female F344 rats. Propetamphos was administered intravenously (12 mg/kg) via an indwelling jugular vein cannula and quantitated by HPLC analysis of the parent compound, as described in the text.
TABLE 2. Noncompartmental Pharmacokinetic Parameters" for Male and Female F344 Rats Following Intravenous Administration of Propetamphos (12 mg/kg) Males (n = 6 )
Parameters
Frrnales ( n = 5)
~~~~~
A (min-')
AUC (min.mg/L) C1 (L/min/kg) MRT (min) vss ( L / W
*
0.015 0.002' 23.5 t 3.2 0.559 & 0.069 28.3 2 5.7 14.7 t 2.6
0.037 ? 0.004h 18.2 * 4.5 0 828 2 0.181 14.4 3.5 12.3 t 4.5
*
'mean 5 standard error h p = 0.001. Parameters were compared for males and females using Student's ttest and considered significantly different if p < 0.05. The value of A was significantly different ( p = 0.001) and the MRT was nearly significantly different ( p = 0.08) for males and females.
Volume 7, Number 4, 1992
they generally appeared to recover within 15 min of dosing.
DISCUSSION Plasma PRO concentration vs. time data were analyzed by noncompartmental methods in these studies. Noncompartmental analysis was used for several reasons. First, the necessary data for a physiologically based pharmacokinetic model were lacking (e.g., blood/tissue partition coefficients, metabolic rate constants, velocities, etc.), so the development of such a model was not attempted. Second, although noncompartmental analysis is not as widely used as classical compartmental models, it is gaining favor with pharmacokineticists and classical compartmental models are being used less frequently. Initially, PRO data were fit to compartmental models (both one- and two-compartment open models were attempted) using the computer program NONLIN84 (9). However, the resulting errors in parameter estimates (10-50%) were not acceptable. In addition, NONLIN84 is sensitive to variability in concentration data in the terminal elimination phase of the decay curve. For one of the data sets, NONLIN84 converged to negative numbers for the terminal elimination rate constant and elimination half-life. Another reason for using noncompartmental analysis is that when using compartmental models, it is often the case that all animals in a study do not conform to the same model (10). Thus, the pharmacokinetic parameters obtained are not shared by all animals in the study. For example, animals which are best described by a two-compartment model have parameters describing a distribution phase. But parameters describing a distribution phase are not obtained from animal data that are fit to a one-compartment model, because no such phase exists in a one-compartment model. It is difficult to make statistical comparisons when some of the parameters are not shared by all the data sets. In noncompartmental analysis, the same parameters are estimated for all animals in the study and statistical comparisons can easily be made. Except for the terminal elimination rate constants, PRO pharmacokinetic parameters in male and female rats following iv administration of 12 mg/kg are not signhcantly different. In both sexes, a short MRT and rapid plasma C1 were observed for PRO. This supports the rapid elimination of PRO-derived radioactivity reported by Winkler and Patel (3). Those investigators have shown that the majority of PRO-derived radioactivity is elimi-
PROPETAMPHOS PHARMACOKINETICS
203
nated as 14C0, following a single oral dose in female rats. As the dose increased (0.6, 6, and 16 mg/kg), the percent of radioactivity eliminated as exhaled carbon dioxide decreased (80, 60, and 40%, respectively) with corresponding increases in radioactivity excreted in urine (12, 20, and 38%).This suggests that the rapid oxidative detoxication pathways leading to C 0 2 formation observed at lower doses are saturated at higher doses. The urinary metabolites were primarily volatile hydrolysis products containing acetone and acetoacetate moieties resulting from cleavage of the P-0-vinyl bond. Desmethyl and desisopropyl propetamphos represented a minor fraction of administered radioactivity, and no PRO oxon or N-desethyl propetamphos was found. Total excretion of PRO-derived radioactivity was the same over the dose range studied, but there were differences in the ratio of 0-dealkylation (desmethyl and desisopropyl PRO) vs. hydrolysis products (carbon dioxide and other volatiles). Winkler and Patel (3) also showed that following multiple oral doses of PRO (6.4 mg/kg), male and female rats excreted 60 and 40% of administered radioactivity as exhaled carbon dioxide, respectively, and females excreted more of the dose in urine than did males (44 versus 33%). Total elimination of PRO-derived radioactivity, however, was approximately the same for males and females at this dose. The iv dose administered in the current studies was 12 mg/kg. Only parent PRO was measured, and, although they are fairly low, the resulting plasma levels of PRO are likely to be high enough to saturate detoxication pathways leading to carbon dioxide formation. A marked gender-related clinical difference in susceptibility to PRO toxicity was observed at this iv dose. Females presented with overt signs of organophosphate intoxication, whereas males were effected to a much lesser extent. This is also consistent with the Winkler and Patel study. The iv PRO dose of 12 mg/kg was selected for two reasons. First, it was well below the reported oral LD-50 and presumably would not be a lethal dose. The 12 mg/kg dose did result in observed clinical toxicity. Second, very little radiolabeled PRO was available and the HPLC assay was developed for unlabeled PRO with UV detection. The sensitivity of the assay was such that if a lower iv dose were administered, PRO would probably not be consistently detected in plasma/acetonitrile supernatants. If a higher dose were administered, toxicity would interfere with the pharmacokinetic study. Therefore, studying PRO pharmacokinetics over a range of doses was not practical. Similarly,
204
K. DIX ETAL.
oral administration of PRO at the dose used in iv studies would likely result in very low plasma levels (i.e., undetectable by UV methods) of the parent compound. Because of the rapid rate of elimination in both sexes, it is unlikely that propetamphos would bioaccumulate in environmentally exposed animals unless repeated exposures occurred at very short time intervals. However, such repeated exposures would probably result in death because of the neurotoxic nature of organophosphate action, namely, cholinesterase inhibition.
ACKNOWLEDGMENT Portions of this work were supported by PHS grants ES-07046 and ES-00044. The authors would like to thank Adrian Phillips for her excellent technical assistance.
REFERENCES 1. J. P. Leber (1972). New class of vinyl thionophosphate insecticides. In Insecticides, Proceedings of the
2nd International IUPAC Congress of Pesticide Chemistry, A. S . Tahori, ed., vol. 1, pp. 381-401, Gordon and Breach, New York, NY.
Volume 7,Number 4, 1992
2. D. S. Wells, N. Motoyama, and W. C. Dauterman (1986a). The in vivo metabolism of propetamphos by insecticide-resistant and susceptible strains of the housefly, Musca domestica. Pestic. Sci., 17, 631-640. 3. V. W. Winkler and J. R. Patel (1982). Metabolism of propetamphos in the rat after single and multiple dosing. In The Fifth international Congress of Pesticide Chemistry (IUPAC), Abstracts VE-20, Human HealthEnvironment-Pesticides, August 29-September 4, 1982, Kyoto, Japan. 4. D. S. Wells, L. M. Afifi, N. Motoyama, and W. C. Dauterman (1986b). In vitro metabolism of propetamphos by housefly, cockroach, and mouse liver preparations. J. Agric. Food Chem., 34, 79-86. 5. L. J. Deterding, K. Dix, L. T. Burka, and K. B. Tomer (1992). On-line coupling of in vivo microdialysis with tandem mass spectrometry. Anal. Chem., 64(21), 2636-2641. 6 . K. Dix, L. J. Deterding, L. T. Burka, and K. B. Tomer (in press). Tris(2-chloroethyl) phosphate pharmacokinetics in the Fischer 344 rat: a comparison of conventional methods and in vivo microdialysis coupled with tandem mass spectrometry. 7. P. G. Harms and S. R. Ojeda (1974). A rapid and simple procedure for chronic cannulation of the rat jugular vein. J. A p p l . Phys., 36(3), 391-392. 8. M. Gibaldi and D. Perrier (1982). Pharmacokinetics, 2nd ed., Marcel Dekker, Inc., New York, NY. 9. Statistical Consultants, Inc. (1986). PCNONLIN and NONLIN84: software for the statistical analysis of nonlinear models. A m . Statistician, 40(1). 10. W. R. Gillespie (1991). Noncompartmental versus compartmental modelling in clinical pharmacokinetics. Clin. Pharrnacokinetics, 20(4), 253-262.
