0090-9556/79/0701 -0007$02.

00/O
DRUG METABOLISM AND DISPOSITION Vol. 7. No. I
Copyright 1979 by The American Society for Pharmacology and Experimental Therapeutics Printedin U.S.A.

PHARMACOKINETICS OF a- AND fl-ISOMERS OF RACEMIC ENDOSULFAN FOLLOWING

INTRAVENOUS ADMINISTRATION IN RABBITS

P. K. GUPTA1 AND M. EHRNEBO

Industrial Toxicology Research Centre, Lucknow (P.K.G.) and Karolinska Pharmacy (ME.)

(Received April 17, 1978, and in revised form September 20, 1978)

ABSTRACT:

The pharmacokinetics of the a- and fl-Isomers of endosulfan in sequently, the plasma clearance was considerably lower for a-en-
rabbits was investigated following intravenous injection. Endosulfan dosulfan (2.70 1 .33 mI/hr/kg) than for fl-endosulfan (70.1
(2 mg/kg) was given as a mixture of the two isomers in the ratio 70: 18.6 mI/hr/kg). The a-isomer had a higher fraction unchanged In
30. The plasma concentration-time data for a-endosulfan was best the urine (37% of dose) than did the fl-isomer (11%) and a slightly
fitted to the three-compartment model with a terminal slope half-life higher fraction unchanged In feces (2.7% and 0.4%, respectively)

Downloaded from dmd.aspetjournals.org at ASPET Journals on September 12, 2017

of 235 168 (SD; N = 6) hr. The data for f1-endosulfan was at 5 days. Thus, the two isomers of endosulfan showed pronounced
adequately described by the two-compartment model, the corre- differences in their pharmacokinetic profile. These dissimilarities
sponding half-life was 5.97 2.41 hr. The total distribution volumes may partly explain reported differences in toxicity between the a-
during the terminal slopes were, however, similar for a- and /1- and fl-isomer.
endosulfan (675 246 and 565 126 mI/kg, respectively). Con-

The insecticide endosulfan has been widely used for the control lected from the cannulated femoral vein and subsequent samples were
of crop pests in agriculture (1, 2). Earlier we have reported data drawn directly from the vein into heparinized tubes at various times. The
on its acute and subacute toxicity in rats, mice, and rabbits (3-5), blood collected at zero time served as control. Urine and fecal samples
were collected up to 5 days. Attempts were made to analyze the samples
its regional distribution in the brain (6), effects on hepatic micro-
on the same day. The samples postponed to the next day were deep-frozen.
somal enzymes (7-9), and its teratogenic and embryotoxic effects
Analysis of Samples. Urine (1 ml) or plasma (0.5 ml) samples were
in rats (10). Following po administration of 14C-labeled endosulfan
extracted for 10 mm with 25 ml of hexane/acetone (9: 1 , v/v). For fecal
in mice, the highest relative amounts of radioactivity were found
samples, 10% homogenates were prepared in hexane/acetone, blended at
in feces, visceral fat, urine, and liver; both a- and fl-isomers could high speed for 10 mm. The emulsion formation was avoided by the
be detected in feces (1 1). Following a single oral dose (0.3 mg/kg) addition of Na2SO4 by the method of Mills ( 13) and Johnson ( 14). The
of 4C-labeled endosulfan to sheep, the compound was completely homogenates were filtered. The residue was washed twice with hexane/
eliminated in 22 days. About 50% of the 4C was excreted in the acetone and the extract partitioned. The hexane extracts were cleaned up
feces, 41% in the urine, and 1% in the milk (12). No insecticide by passage through a Florisil column with 1 g of Na2SO4 on top. The
was recovered in the milk of cows that had been fed for 21 column was rinsed three times with 10 ml of hexane each time. The
consecutive days on silage containing 0.41-2.35 ppm endosulfan volume of the extract was reduced to 10 ml. The concentrated extract was
analyzed by gas chromatography (electron-capture detector) in a I .8-rn x
(12).
6.5-mm column packed with 1.5% OV-l7 + 1.95% OV-210. Other condi-
In order to utilize the results of toxicological studies and also
tions were: column, 180#{176}C;
detector and injector, 200#{176}C;high-purity N2
because of its contamination of human and animal systems either
carrier gas, 60 mI/mm. The identification of the peaks was based on the
directly or indirectly as environmental pollutants, it was consid- retention times and the absence of these peaks in samples collected at zero
ered desirable to elucidate the pharmacokinetics of the compound time. Recoveries of a- and fl-isomers of endosulfan, under in vitro condi-
following administration to laboratory animals. tions, ranged from 95 to 98%. The sensitivity and specificity for a- and
$-endosulfan have been demonstrated earlier (15).
Materials and Methods Pharmacokinetics. Graphic analysis of the logarithm of the plasma
Animals and Treatment. Six Industrial Toxicology Research Centre concentrations vs. time after iv administration revealed three exponential
colony-bred, female, albino rabbits weighing 1.7-2.0 kg were used. The segments for a-endosulfan and two for $-endosulfan. The time-course of
animals were preconditioned on the day of experiment, and placed in a-endosulfan concentrations in plasma (Ce) could therefore be described
individual metabolic chambers designed for separate collection of urine by the following equation:
and feces. They had free access to food and water.
Technical grade endosulfan containing a- and fl-isomers in the ratio of (p = Ae + Be1 + Ce, (1)
7:3 was obtained from the National Chemical Laboratory, Pune, India.
where A, B, and C are the intercepts with the ordinate and a, fi, and y are
Very fine powder of endosulfan (2 mg/kg) suspended in peanut oil was
the hybrid disposition rate constants. The time-course of fl-endosulfan
injected in the ear vein of the rabbits.
concentrations could accordingly be described as
For blood collection, the rabbits were anesthetized with ether, and the
femoral vein was cannulated. The initial seven blood samples were col- C = Ae + Be. (2)

1 Present address: Industrial Toxicology Research Centre. P.8. No. 80, Luck- The plasma concentration time data for a- and f-endosulfan were fitted
now 226001, India. to Eqs. 1 and 2, respectively, by use of the nonlinear least-squares
regression program BMDP3R ( 16). The plasma concentrations were
Send reprint requests to: Dr. Mats Ehrnebo, Karolinska Pharmacy, Karolinska
weighted according to the reciprocal of concentration. The data for a-
Hospital. S-1O4 01 Stockholm 60, Sweden.
endosulfan were also fitted to the two-compartment model (Eq. 2) and a
7
 8 GUPTA AND EHRNEBO

comparison was made with the data from the three-compartment model (Eq. 2) models. As shown in table 1, the three compartment model
(Eq. 1) by use of the F-ratio criteria suggested by Boxenbaum et al. (17). gave in all cases a smaller WSSR, indicating a better fit to Eq. 1
The microscopic rate constants km to k31 (fig. I) were calculated from the as compared to Eq. 2. The improvement in WSSR with the F-
hybrid constants of Eqs. 1 and 2 (18, 19).
ratio test (17) was significant at the p < 0.05 level in four cases out
The apparent distribution volume of the central compartment (com-
of six (table 1). The improvement in curve fitting is also evident
partment I in fig. I) per kg body weight (Vc/B.W.) was calculated for a-
endosulfan as in fig. 2, in which both curves are drawn and compared with the
experimental data points for rabbit 4. Thus, the three-compart-
Vc/B.W. = D/C#{176}= D/(A + B + C), (3) ment model was chosen in the following pharmacokinetic analysis.
and for fi-endosulfan as As seen in fig. 2, the number of data points at the terminal slope
were somewhat insufficient to give an accurate estimation of y,
Vc/B.W. = D/Cp#{176}
= D/(A + B), (4) this is further evident in table 1, where the values for -y showed
where D is the dose per kg body weight of a-endosulfan (1.4 mg/kg) and great interindividual differences (0.00472 0.00311 SD; N = 6)1.
fi-endosulfan (0.6 mg/kg), respectively, the sum being the total dose of However, the two-compartment model was unable to describe this
racemic endosulfan administered iv (2 mg/kg). The total apparent volume terminal phase, giving values of the slope (fi) that overestimated
of distribution at pseudodistribution equilibrium, i.e., during the y-phase the coefficient of the terminal slope (table 1).
for a-endosulfan and during the fl-phase for fl-endosulfan, was calculated The plasma levels for fJ-endosulfan were adequately described
as by the two-compartment model, as seen in fig. 2 and evident from
the small WSSR shown in table 2.

Downloaded from dmd.aspetjournals.org at ASPET Journals on September 12, 2017

(Vd)/B.W. = D/(y.AUCo..) (5)
The microscopic rate constants for a- and /J-endosulfan are
and shown in table 3. It can be seen that for a-endosulfan the rate
(Vd)fl/B.W. =
(6) constant for return from compartment 3 (k31) was of the same
magnitude as the rate constant of elimination (k10). For fi-endo-
sulfan, k21 had a value similar to k10.
k2
The half-life of the terminal slope (t112) was much longer for
a-endosulfan [235 168 hr SD; N = 6) than for fi-endosulfan
(5.97 2.41 hr) as seen in table 3. The slow elimination of a-

20
-J

10

Two-compartment body model LI

z 5
0
(-3

k12
U,

-J
2
0

0 10 20 30 qO 50 60 70 80 90 100110 120
HOURS

-I
E
k0

Three-compartment body model

LI
FIG. 1 . Compartment models used to describe rite thsposion of a- and /3- z
endosulfan.

respectively. The area under the plasma concentration-time curve from

time zero to infinity (AUCo) was calculated as (A/a) + (B/fl) + (C/y) 0.
for a-endosulfan and as (A/a) + (B/fl) for /I-endosulfan.
The total plasma clearance per kg body weight (Cl/B.W.) was deter-
mined according to Eq. 7:
0.
12 lq 16 ii
Clp/B.W. = V.k19. (7) HOURS

The half-life during the y- and fl-phase was calculated as O.693/y and FIG. 2. Plasma concentrations of a-endosulfan (0), /3-endosulfan (0), and
0.693/fl, respectively. total endosulfan (y) in rabbit 4 following administration of racemic endo-
The derivations of Eqs. 1-7 can be found elsewhere (18). sulfan (2 mg/kg) by intravenous injection.

The predicted curve for a-endosulfan according to the three-compart-

Results
ment (--) and two-compartment ( ) models (top), and that for /1-
For a-endosulfan, the curve-fitting procedure was made accord- endosulfan according to the two-compartment model (--) (bottom) are
ing to both the three-compartment (Eq. 1) and two-compartment shown.
 PHARMACOKINET1CS OF ENDOSULFAN ISOMERS 9

endosulfan compared to /3-endosulfan is further demonstrated by (sum ofall compartments) during the terminal slope was, however,
the value for the total plasma clearance, which is only 2.7 ml/hr/ similar for both isomers.
kg for a-endosulfan compared to 70 mI/hr/kg for /3-endosulfan Higherpercentages of the dose were excreted unchanged in the
(mean values). As seen in table 3, the apparent volume of distri- urine for a-endosulfan (37%) than f.3-endosul.fan (1 1%) during 0-
bution of the central compartment was twice as large for /3- 5 days (table 4). It should be pointed out that the urine collections
endosulfan as for a-endosulfan. The apparent total distribution of a-endosulfan were incomplete, because of the long half-life of

TABLE 1
Dalafor a-endosulfan obtainedfollowing a single intravenous injection ofracemic endosulfan (2 mg/kg) in rabbits according to the three- or two-
compartment body model
Critical
No.ofCorn-
Rabbit No. A a B fi C -y WSSR Nb F-Ratio Value
pa men
(p = 0.05)
g/ml hr pg/mI h1 g/m1 hr

1 3 18.7 3.53 10.7 0.177 2.49 0.00266 0.0977 10 23.9 6.94

2 15.3 0.320 3.01 0.00484 1.26 10
2 3 11.4 1.25 2.90 0.137 1.57 0.00165 0.107 9 4.81 9.55
2 1 1.9 0.776 2.34 0.0063 1 0.449 9
3 3 15.6 2.02 2.47 0.114 1.25 0.00155 0.0404 10 31.6 6.94

Downloaded from dmd.aspetjournals.org at ASPET Journals on September 12, 2017

2 14.9 1.41 2.22 0.00966 0.678 10
4 3 13.3 1.87 4.45 0.219 2.75 0.00762 0.0487 10 19.5 6.94
2 13.8 0.918 3.57 0.0117 0.523 10
5 3 15.2 0.893 2.97 0.0417 1.63 0.00658 0.284 9 0.365 9.55
2 15.4 0.767 3.82 0.0141 0.353 9
6 3 16.6 2.45 7.86 0.151 2.47 0.00828 0.145 9 13.0 9.55
2 14.2 0.447 3.59 0.0125 1.40 9

Mean 3 15.1 2.00 5.23 0.440 2.03 0.00472

2 14.3 0.773 3.09 0.00985
SD 3 2.5 0.93 0.68 0.060 0.62 0.00311
2 1.3 0.384 3.33 0.00364
a WSSR = weighted sum of squared residuals (reciprocal of concentration).
h Number ofexperimental data points.
C Three-compartment model shows significant improvement compared to two-compartment model.

TABLE 2
Datafor fi-endosulfan obtainedfollowing a single intravenous injection ofracemic endosulfan (2 mg/kg) in rabbits according to the two-compartment body
model
Rabbit No. A a B /3 WSSR N

g/m1 hr g/ml hr
1 2.04 1.12 0.945 0.119 0.0437 7
2 2.90 2.63 1.26 0.253 0.0120 5
3 2.60 2.07 0.801 0. 1 19 0.0384 6
4 7.11 4.87 0.989 0.153 0.0205 7
5 3.99 1.34 0.582 0.0986 0.167 8
6 4.47 2.18 0.854 0.0701 0.377 9

Mean 3.85 2.20 0.905 0.135

SD 1.83 0.99 0.225 0.064
a WSSR = weighted sum of squared residuals (reciprocal of concentration).
I, Number of experimental data points.

TABLE 3

Microscopic rate constants, half-lives ofterminal slope (t,1), distribution volumes [V, and (Vd)pJ and total body clearance (Clp)for a- and fi-endosulfan of
racemic endosulfan (2 mg/kg) administered iv to rabbits
Data rep resent means SD.

Isomer k:,1 k2 k, k2 km ti,2 VJB.W. (V4/B.W Cl,./B.W.

hr hr hr hr hr hr mi/kg mi/kg mi/hr/kg
a 0.04.09 0.738 0.0565 1.06 0.262 235 65.9 675 2.70
0.0200 0.483 0.0126 0.47 0. 109 168 15.8 246 1.33
/3 0.564 0.555 1.39 5.97 140 565 70.1
0.255 0.263 0.99 2.41 45 126 18.6
a For y for the three-compartment model (a-endosulfan), or for /3 for the two-compartment model (J3-endosulfan).
b Total distribution volume during the y-phase for a-endosulfan and during the fl-phase for fl-endosulfan.
10 GUPTA AND EHRNEBO

TABLE 4
Urinary andfecal excretion (0-5 days) of a- and fi-endosulfan administered intravenously as racemic endosulfan (2 mg/kg) to rabbits
Data represent means SD. X, Cumulative amount excreted in 5 days;f, percent of dose excreted in 5 days.
a-Endosulfan $-Endosulfan Total Endosulfan
(1.4 mg/kg) (0.6 mg/kg)
Excreta (2.0 mg/kg)
x I X f X f
mg % mg % mg %
Urine 0.903 36.7 0.111 10.5 1.01 28.8
0.346 15.0 0.041 4.1 0.39 11.7
Feces 0.0683 2.7 0.00394 0.37 0.0709 2.0
0.0173 0.6 0.00140 0.10 0.0194 0.5

this isomer. If calculated with reference to total endosulfan, 29% anticoagulant, both in pharmacokinetics and pharmacodynamic
of the dose had been eliminated unchanged in urine up to 5 days. effect (23, 24). Recently, differential disposition of dextrococaine
Very small amounts of endosulfan were excreted unchanged in and cocaine was reported (25). Endosulfan has been shown in this
feces: only 2.7% of a-endosulfan and 0.4% of the /3-isomer (table study to be another important example of how different isomers
4). of the same compound behave essentially like two different com-
Discussion pounds in the body.

Downloaded from dmd.aspetjournals.org at ASPET Journals on September 12, 2017

The total levels of endosutfan showed very slow decreases in Acknowledgements. The first author is grateful to Dr. C. R.
plasma with time. This was mainly due to the slow elimination of Krishnamurty, Acting Director, Industrial Toxicology Research
the a-isomer, inasmuch as the /3-isomer of racemic endosulfan had Centre, Lucknow for his valuable suggestions. Thanks are also
a much more rapid elimination. The three-compartment model due to Mrs. Kajoli Kanwar and Messrs. A.K. Khanna and G.S.
seemed to describe better the pharmacokinetics of a-endosulfan. Tandon for their technical assistance. Free supply of endosulfan
Use of the three compartment model for a-endosulfan gave some from the Director, National Chemical Laboratory, Pune, is thank-
interesting information that would have been lost in a simpler fully acknowledged.
compartmental treatment: as seen in table 3 the rate constant (k31) References
for return from compartment 3 to the central compartment was of 1. H. Martin, Insecticide and Fungicide Handbook for Crop Protec-
the same magnitude as that for elimination (k10). The third corn- tion, 4th ed., p. 44. Blackwell Scientific Publications, Oxford, 1964.
partrnent in this case represents a so-called deep compartment; 2. E. J. Miller, Residue Rev. II, 100 (1955).
it will act as a sink for a-endosulfan in the body. Thus, both k10 3. P. K. Gupta, Bull. Environ. Contam. ToxicoL 15, 708 (1976).
and k31 are responsible for the slow overall elimination of a- 4. P. K. Gupta and V. C. Satya, Bull. Environ. Contam. ToxicoL 15, 513
endosulfan from the rabbit. Moreover, it was impossible to get a (1975).

good fit to the terminal slope of a-endosulfan by use of the two- 5. P. K. Gupta and V. C. Satya, Bull. Environ. Contam. ToxicoL 18, 378
(1977).
compartment model; the ti,2 of /3 in table 1 is 81 37 hr (mean
6. P. K. Gupta, Toxicology 9, 371 (1978).
SD, N = 6), which is an underestimation of the frnal slope
7. P. K. Gupta and R. C. Gupta, J. Pharm. Pharmacol. 29, 245 (1977).
according to the three-compartment model (t112 of -y 235 168
8. P. K. Gupta and R. C. Gupta, Toxicology 7, 283 (1977).
hr). 9. D. K. Agarwal, P. K. Seth, and P. K. Gupta, J. Environ. Sci. Health,
There were marked differences in the pharmocokinetics be- Part C 13, 49 (1978).
tween the two isomers. For example, the total plasma clearance 10. P. K. Gupta, V. C. Satya, and D. K. Saxena, Acta Pharmacol. ToxicoL
was about 26 times greater for /3-endosulfan than for a-endosulfan. 42, 150 (1978).
The only similarity was seen in the distribution volume during the 1 1. P. Deema, E. Thompson, and G. W. Ware, J. Econ. EntomoL 59, 546
disposition phase, which was about the same for the isomers. The (1966).
slower plasma clearance of the a-isomer is in accordance with our 12. H. Maier-Bode, Residue Rev. 22, 1 (1968).
earlier observation that the concentration of a-endosulfan in 13. P. A. Mills, J. Assoc. Off AnaL Chem. 42, 734 (1959).
14. L. Johnson, J. Assoc. Off AnaL Chem. 45, 363 (1962).
plasma and in the brain tissue of rats given 5 or 10 mg of
15. A.S.Y. Chau, J. Assoc. Off AnaL Chem. 55, 1232 (1972).
endosulfan per kg daily for 15 days was manyfold higher than
16. W. J. Dixon, BMDP-Biomedical Computer Programs, p. 541. Urn-
that of /3-endosulfan (6).
versity of California Press, Berkeley, 1975.
It is interesting to note that a-endosulfan seems to have the 17. H. G. Boxenbaum, S. Riegelman, and R. M. Elashoff, I. Pharmaco-
higher toxicity of the two isomers in white rats; LDo of the a- kinet. Biopharm. 2, 123 (1974).
isomer was 76 mg/kg and that of the /3-isomer was 240 mg/kg 18. M. Gibaldi and D. Perrier, Pharmacokinetics, p. 48. Marcel Dekker,
(12). The higher toxicity may be due partly to a slower elimination New York, 1975.
of a-endosulfan in rats. However, it is not certain that the differ- 19. T. L. Loo, B. B. Tanner, G. E. Householder, and B. J. Shephard, J.
ences in disposition are in the same direction in rats as in rabbits. 57, 2126 (1968).
Pharm. Sci.

Considerable species differences have been reported in this respect 20. D. D. Breimer and J. M. van Rossum, Eur. J. PharmacoL 26, 321
(1974).
for another substance, hexobarbital (20, 21).
21. D. D. Breimer and J. M. van Rossum, J. Pharm. PharmacoL 25, 762
It has earlier been shown that isomers may show pronounced
(1973).
differences in pharmacokinetics, metabolism, and excretion. For
22. K. H. Palmer, M. S. Fowler, and M. E. Wall, J. PharmacoL Exp. Ther.
example (-)-hexobarbital has a much longer half-life than does 175, 38 (1970).
(+)-hexobarbital in rats (20). The isomers of another barbiturate, 23. A. Yacobi and G. Levy, J. Pharmacokinet. Biopharm. 2, 239 (1974).
pentobarbital, have been demonstrated to yield different fractions 24. A. Breckenridge, M. Orme, H. Weaseling, R. J. Lewis, and R. Gibbons,
of the hydroxy metabolite in vivo (22). Marked differences have Clin. PharmacoL Ther. 15, 424 (1974).
also been demonstrated for the enantiomers of warfarin, an oral 25. A. L. Misra and R. B. Pontani, Drug Metab. Dispos. 5, 556 (1977).