INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough,margins, substandard and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600 Order Number 9411984
Quantitative structure activity relationships to predict the fate and effects of selected organophosphorus and carbamate insecticides in aquatic systems
Kallander, David Brian, Ph.D.
The Ohio State University, 1993
UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106 QUANTITATIVE STRUCTURE ACTIVITY RELATIONSHIPS TO PREDICT THE FATE AND EFFECTS OF SELECTED ORGANOPHOSPHORUS AND CARBAMATE INSECTICIDES IN AQUATIC SYSTEMS
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University
By
David B. Kallander, B.S., M.S.
*****
The Ohio State University 1993
Dissertation Committee: Approved By
Dr. Susan W. Fisher
Dr. Richard W. Hall Sc. Advisor Dr. David J. Horn Department of Entomology ACKNOWLEDGEMENTS
To Dr. Susan Fisher for her advice, guidance, and
putting up with me for so long. A special thanks also to
my other dissertation committee members Dr. David Horn and
Dr. Rich Hall for their help and support over the years.
To my co-workers Kathy Bruner, Denise Boulet, Mark
Atanasoff, Steve Chordas, Kristen Manter, Lisa Jackson,
Xiaosong Zhang, Jane Zumwalt, and Rick Stock. I wish you the best of luck. I'll miss all of you.
To Mike and Liz Lydy for advice, suggestions, and years of friendship. I would also like to thank Dr. Bill
Collins for the use of his midge cultures and lab space.
Finally, I would like to thank my parents for their years of unwavering support.
ii VITA
November 5, 1963 ...... Born in Hamilton, Ohio
1986 ...... B.S., Miami University Oxford, Ohio
1989 ...... M.S., The Ohio State University, Columbus, Ohio
PUBLICATIONS
Horn, D.J., D.K. Pearl, R. Bartoszynski, and D.B. Kallander 1990. "Refinement of a stochastic model of spider mite predator-prey interactions" in Modern Acarology, F. Dusbabek and V. Bukva, Eds. Volume I, 1991.
FIELD OF STUDY
Major Field: Entomology
iii TABLE OF CONTENTS
ACKNOWLEDGEMENTS ...... ii
VITA ...... iii
LIST OF TABLES ...... vi
LIST OF FIGURES ...... viii
INTRODUCTION ...... 1
CHAPTER PAGE
I. QUANTITATIVE STRUCTURE ACTIVITY RELATIONSHIPS FOR PREDICTING THE IN VIVO AND IN VITRO INHIBITION OF ACETYLCHOLINESTERASE IN THE MIDGE, CHIRONOMUS RIPARIUS ...... 17
Introduction...... 17 Materials and Methods ...... 20 R e s u l t s ...... 27 D i s c u s s i o n ...... 49
II. SYMPTOMATIC AND ENZYMATIC RECOVERY FROM EXPOSURE TO ORGANOPHOSPHORUS AND CARBAMATE COMPOUNDS IN THE MIDGE, CHIRONOMUS RIPARIUS: TEMPERATURE AND EPISODIC EXPOSURE ...... 64
Introduction...... 64 Materials and Methods ...... 67 R e s u l t s ...... 76 D i s c u s s i o n ...... 95
iv III. QUANTITATIVE STRUCTURE ACTIVITY RELATIONSHIPS FOR PREDICTING THE PERSISTENCE OF ORGANOPHOSPHORUS AND CARBAMATE INSECTICIDES IN WATER, WITH AND WITHOUT THE MIDGE, CHIRONOMUS RIPARIUS ...... 104
Introduction ...... 104 Materials and Methods ...... 107 R e s u l t s ...... 117 D i s c u s s i o n ...... 122
CONCLUSIONS ...... 135
LIST OF REFERENCES ...... 138
V LIST OF TABLES
TABLE PAGE
1. Mean in vivo I50 Values of Organophosphorus and Carbamate Insecticides in the Midge, Chironomus riparius at 10, 20 and 30 "C (± S . E . ) ...... 28
2. Mean in vivo and in vitro I50 Values of Organophosphorus and Carbamate Insecticides in the Midge, Chironomus riparius at 20 °C (+ S . E . ) ...... 30
3. Molecular Descriptors used in Regression Analysis with in vivo and in vitro I50 v a l u e s ...... 3 3
4. Linear Solvation Energy Relationships for Insecticides used in Regression Analyses .... 36
5. Molecular Connectivity Indices used in Regression Analysis ...... 38
6. Summary of Best Regression Analyses of In vivo I50 values for LSER, and MC Models...... 40
7. Summary of regression analyses of in vitro I50 values for LSER and MC models...... 42
8. Exposure level (M) of one organophosphorus and four carbamate insecticides at 10, 20 and 30 °C for the midge, Chironomus riparius, producing 50% inhibition of AChE . . . 72
9. Mean AChE activity (in nmoles substrate hydrolyzed/min/mg protein) for Chironomus riparius at 3, 6, and 12 h following exposure to one of five OP or carbamate insecticides at 10, 20, and 30 ° C ...... 87
vi 1 0 . Mean adjusted percent effect for one Organophosphorus and four carbamate insecticides at 20 °C to the midge, Chironomus riparius following two 1 h exposures separated by 0, 2,6, 12, and 24 h in undosed w a t e r ...... 90
11. Dose levels of 3 OP and 2 carbamate compounds and sampling dates for determination of abiotic halflives at 10, 20 and 30 ° C ...... 110
12. Mean abiotic halflives of Organophosphorus and Carbamate Insecticides in sterilized deionized water at 10, 20 and 30 °C (± S . E . ) ...... 116
13. Mean halflives of Organophosphorus and Carbamate Insecticides in water with 40 midges (Chironomus riparius) at 10, 20 and 30 °C (± S.E.) ...... 118
14. Summary of Best Regression Analyses of hydrolysis and biological halflife values for LSER, and MC Mo d e l s ...... 12 0
vii LIST OF FIGURES
FIGURE
1. Some transport and transformation of organic chemicals in aquatic systems......
2. Regression of in vivo Iso's (20 °C) for 7 OP and 5 carbamate insecticides against corresponding 24 h EC50's......
3. Regression of in vitro I50's for 2 OP and 5 carbamate insecticides against corresponding 24 h EC jo's (20 °C)......
4. Regression of in vivo I50's for 5 carbamate and 2 OP insecticides against corresponding in vitro I50's (20 °C)......
5. Reactivation of midge AChE at 10, 20 and 30 °C following exposure to aldicarb. Data points are mean /xmoles substrate hydrolyzed/min/mg protein (+ standard error). Solid lines are mean control activities at respective temperatures ......
6. Reactivation of midge AChE at 10, 20 and 30 °C following exposure to propoxur. Data points are mean mmoles substrate hydrolyzed/min/mg protein (+ standard error). Solid lines are mean control activities at respective temperatures ......
7. Reactivation of midge AChE at 10, 20 and 30 °C following exposure to carbofuran. Data points are mean ^moles substrate hydrolyzed/min/mg protein (+ standard error). Solid lines are mean control activities at respective temperatures ......
viii 8. Reactivation of midge AChE at 10, 20 and 30 °C following exposure to carbaryl. Data points are mean /nmoles substrate hydrolyzed/min/mg protein (+ standard error). Solid lines are mean control activities at respective temperatures...... 83
9. Reactivation of midge AChE at 10, 20 and 30 °C following exposure to parathion. Data points are mean /moles substrate hydrolyzed/min/mg protein (+ standard error). Solid lines are mean control activities at respective temperatures ...... 85
10. AChE activity (+ S.E.) in 20 fourth instar midges following two 1 h pulsed exposure to carbaryl (T = 20 °C, n = 3). Midges were pulsed at time 0 and 24 h (exposure time not included on graph). Solid line is mean control activity...... 93 INTRODUCTION
More chemicals are being used today than ever before
(Jorgensen 1990). In the United States, novel chemicals are being evaluated for possible use at a rate of 1000 -
1600 per year (Nirmalakhandan and Speece 1988). In
Europe, a report by the European Inventory cited over
100,000 chemicals in use (Hart 1991). Since nearly all of the chemicals in use are likely to find their way into terrestrial and aquatic environments, regulating authorities such as the US Environmental Protection Agency require that these compounds be assessed as to their possible harm to living organisms. However, regulating authorities have been faced with two significant obstacles: 1) problems related to chemicals in the environment are highly complex, with many variables such as temperature, pH, and sunlight having an impact on a chemical's fate and effects (Jorgensen 1990) and 2) there is a serious lack of biological data on chemicals in use
(Hart 1991). Both of these problems involve the lack of adequate information concerning the fate and effects of chemicals in the environment. Since the number of compounds in use and the cost of evaluating the hazards of these compounds is ever increasing, researchers have turned to the development of predictive models to estimate the required information. The development of accurate predictive models,, called quantitative structure activity relationships (QSAR), would allow industry and regulating authorities to quickly and cheaply determine the possible hazards of a chemical without exhaustive and costly laboratory research. However, to develop accurate predictive models a good understanding of the various factors that will affect a chemical in the environment is necessary.
Any compound entering an aquatic system is subject to many different processes that could affect availability to aquatic organisms, convert it from one form to another or degrade the chemical (Haque et al. 1980). What transport or transformation processes a chemical will likely undergo depends on how the chemical partitions between the different compartments (water, sediment, biota) in an aquatic system (Fig. 1).
Water
Hydrolysis is the splitting of an organic compound with water, hydroxide and/or hydrogen ions (Jorgensen
1990). It has long been known to be an important factor in the degradation of many organic compounds. Mabey and
Mill (1978) reported the abiotic halflives of a number of Figure 1. Some transport and transformation of organic chemicals in aquatic systems.
3 4
WATER hydrolysis photolysis volatilization BIOTA sorption to DOC uptake metabolism CHEMICAL elimination / storage t J SEDIMENT sorption desorption degradation
Figure 1. organic chemicals which ranged from seconds to thousands of years. Eichelberger and Lichtenberg (1971) in a study conducted with 28 pesticides in river water, reported that all but one organophosphorus (OP) insecticide had undergone some transformation in eight weeks. Seven carbamates were found to be significantly altered in one week and completely degraded after 8 weeks. The remaining organochlorine compounds were unaltered after 8 weeks.
Chemicals in water can be transformed by hydrolysis, and can also be altered by sunlight. The photolysis of chemicals in water is reported to be affected by sunlight intensity, latitude, time of day and depth of water (Zepp and Cline 1977). Wolfe et al. (1978) reported that the halflife of carbaryl in distilled water exposed to sunlight was 6.6 days, while the halflife from abiotic hydrolysis was 15 days.
Not only can pollutants be hydrolyzed in the water column, but they are also capable of sorbing to dissolved organic carbon (DOC) as well as suspended particulates.
Nebeker et al. (1989) reported that sorption of xenobiotics to dissolved organic matter in the interstitial water of sediment can decrease the bioavailability of such compounds. Lydy et al. (1990a) reported that when several neutral lipophilic compounds were added to beakers containing water and sediment, up to 60% of the aqueous fraction of the chemicals were found to bind to DOC and suspended particulates. The reduction of
availability due to sorption on DOC, particulates and
sediment was cited as the cause of subsequent decrease in
toxicity of these compounds to midges added to the beakers
of water and sediment.
Water-borne xenobiotics leave aquatic environments by volatilization, thereby decreasing their concentration in water. The rate of volatilization is dependant on the vapor pressure and solubility of the compound in water
(Haque et al. 1980). Dilling et al. (1975) reported that
five organochlorine compounds (methylene chloride,
chloroform, 1,1,1-trichloroethane, trichloroethylene, and tetrachloroethylene) placed in water at 25 °C and stirred
continuously, evaporated to 10% of the original
concentration in less than 90 min. Henry's Law Constant
(HLC), which relates the concentration of a gas dissolved
in a liquid to the partial pressure of the gas in
solution, is often used to describe the volatility of a
compound. Mackay and Leionen (1975) demonstrated that HLC
could be estimated by the ratio of the partial pressure of
a compound divided by it's solubility in water.
Sediment
Sediment has long been known to affect many chemicals which enter aquatic communities by changing their bioavailability. Researchers have determined that
numerous nonpolar organic compounds adsorb to the organic
carbon in sediment, reducing their availability and
therefore toxicity to aquatic organisms. Pittinger et al.
(1989) found that the toxicity of four surfactants was
lower to the midge, Chironomus riparius if the route of
exposure was through contaminated sediment. Nebeker et
al. (1989) found that higher quantities of total organic
carbon in sediment caused a reduction in the toxicity of
DDT to the freshwater amphipod, Hyalella azteca. Although
the carbon content of sediment will affect availability of many compounds, other factors such as structural
characteristics of the chemical, as well as the pH and
temperature of the medium will also be important in
determination of the availability of a chemical in an
aquatic environment (Haque et al. 1980).
Biota
Unquestionably the most important compartment which a
compound can partition into are the different organisms
which make up an aquatic community. The degree of harm
any compound will inflict on the inhabitants of our
streams, rivers, lakes and oceans will depend on a
chemical's ability to enter an organism, negotiate the
various detoxification mechanisms possessed by the organism, reach the target site and successfully interact with the appropriate receptor.
For example, for a compound to reach the target site within the body of an aquatic invertebrate (such as the midge, C. riparius) the compound must pass through the insect's cuticle which is covered by a thin wax layer.
Three factors have been cited as important for this: 1) lipid solubility of the compound, 2) affinity of the compound for other cuticular components such as protein and chitin, and 3) solubility of the compound in the insect's hemolymph (Matsumura 1985). According to Olson and O'Brien (1963), the lipophilicity of the compound is very important since this determines the rate and amount of compound that penetrates the wax layer covering the cuticle. As the compound passes through the cuticle, there is the opportunity for it to be stored, limiting the amount entering the organism. Once the compound has dissolved into the hemolymph it would be transported throughout the insect's body. However, at this point the compound would be subject to attack by the insect's mixed function oxidase system, which could detoxify or possibly activate the compound. Further, the insecticide may be stored in various organs or tissues (in vertebrates, adipose tissue has long been known to be a storage site) making the insecticide bio-unavailable. Once the compound has reached the target site, it must then react successfully with the target site for intoxication to occur. Even so, a compound that has bound to the appropriate target site it may be removed from the active site by hydrolases present in the insect, or simply fall off due to weak interactions between the compound and target site.
Overall it is clear that any model which would describe the effects of a compound on aquatic biota must account for many different chemical and physical processes that occur within living organisms.
Predictors of fate and effects
The parameter describing the hydrophobicity of a compound, log K^, has been used more often in QSAR studies than any other descriptor (Nirmalakhandan and
Speece 1988). It has been used to describe both the effects on biota and fate of xenobiotics in aquatic systems. Researchers have shown that a linear relationship exists between the sorption of neutral organic compounds to sediment. Karickhoff et al. (1979) demonstrated that sorption of several hydrophobic pollutants was linearly related to log K,^. Landrum et al. 1987 also demonstrated that sorption to sediment will increase with increasing log Kow for nonionic compounds. Other researchers have demonstrated a linear relationship between such important parameters as the bioconcentration factor and log K^, Neely et al. (1974) demonstrated a linear relationship between log and the log of the bioconcentration factor of seven neutral organic compounds in trout muscle. Kenega and Goring (1980) found a strong relationship between Kw and bioaccumulation of hydrophobic compounds in Daphnia magna. Not only has the fate of many chemicals been found to be predicted by
K^, but K,^ has been shown to correlate with the toxicity of many compounds under site specific conditions. Konneman
(1981) and Veith (1983) reported that a good correlation existed between log K^, values and toxicity of various industrial chemicals to fish. Zaroogian et al. (1985) also demonstrated that log K^'s for several lipophilic compounds correlated fairly well with toxicity, although correlations with pesticides did not result in significant relationships.
The steric parameter molecular volume (MV), has been used in a number of studies to describe the toxicity of neutral organic compounds which exert their toxicity through narcosis. McGowan and Mellors (1986a) found that the toxicities of nonpolar organic compounds to juvenile sheepshead minnows (Cyprinodon variecatus) as well as fathead minnows (Primephales promelis) could be predicted 1 1 to within an order of magnitude by MV. McGowan and
Mellors (1986b) assert that although toxicity has been described by such parameters as log K^, water solubility, vapor pressure, and chemical potential, all these properties are related in that they are dependant on characteristic volume and temperature, and therefore claimed MV should be able to describe toxicity for a wide variety of compounds.
The solubility of a compound has been known to be an important factor in the fate and toxicity of chemicals
(Lydy 1990). Linear solvation is a method which can be used to quantify solubilities based on energy relationships involved in the solution process. According to Kamlet et al. (1986), solubilization consists of three energy dependant steps:
1) A cavity is formed in the solvent within which a
solute particle can enter;
2) One molecule of solute will enter the cavity;
3) Attractive forces form between solute and solvent.
Solvatochromic parameters were developed which represent linear combinations of energy contributions in each step.
The result of this was a term which described many of the solubility and solvent-dependant properties of a compound.
One or more values for these solvatochromic parameters are
known for hundreds of compounds and because of this large
data base, it has become possible to correlate many
physicochemical properties and reactivity parameters
including toxicity of aliphatics and aromatic compounds
with biological endpoints in a variety of organisms. For
example, Kamlet et al. (1987) found a linear relationship
(r = 0.983) between the toxicities of several organic
nonelectrolytes to the golden orfe fish (Leuciscus idus melanotus) and LSER parameters 7r*, /3, and a. Kamlet et
al. (1986) found a strong linear relationship (r2 = 0.987)
between EC50's (where the effect was inhibition of
bioluminescence) of neutral organic compounds to the
bacterium Photobacterium phosphoreum and LSER parameters
Vj/100, 7r*, (8, and a.
Of the topographical indices developed, molecular
connectivity (MC) has been used often over the past decade
to describe such properties as toxicity, molar refraction,
polarizability, water solubility, molar volume, and heat
of vaporization (Kier and Hall 1986). MC is a procedure
for quantifying molecular structure based on fragments or
subgraphs weighted by the degree of bonding between atoms
(Hall and Kier 1989). Quantifying molecular structure is
done simply by counting structural features weighted by adjacent atoms where delta values are assigned to all
atoms in a molecule except hydrogen (Hall et al. 1989).
The summation of all delta values in a molecule results in
a first order index. Higher order indexes, based on more
complex substructures, are possible and are used for multiple regression analyses with physical and biological
processes (Hall et al. 1989). Leegwater (1989) used MC
indices to predict acute toxicity of a variety of pollutants including benzene and toluene to the guppy.
Results showed that MC indices were equivalent and
sometimes superior to predictions from log Kow values.
Shigoeka et al. (1988) compared the predictability of
several quantitative structure-activity relationship
parameters to toxicity of chlorophenols to green algae,
Selenastrum capricornutum. Log Kow produced the best
correlation, however MC indices also proved to be
satisfactory.
Modeling organophosphorus and carbamate fate and effects
Although numerous models have been developed around
the steric, electronic and topographical aspects of molecules, most of the studies have related these
parameters to biological activity for neutral organic
compounds whose mode of action is nonspecific narcosis.
Compounds which exert their toxic effect through narcosis are believed to dissolve into cell membranes and disrupt vital processes, such as the transmission of nervous system impulses (Franks and Lieb 1985). The toxic action of these compounds appears to be less related to particular functional groups and specific interactions with receptor sites than simple partitioning into lipid membranes. Therefore, parameters such as and MV which are useful in describing effects related to partitioning have been used quite successfully in the past. However, few studies have been conducted which attempt to predict the toxicity and persistence of organophosphorus (OP) and carbamate insecticides. These compounds have a specific mode of action— they inhibit the nervous system enzyme acetylcholinesterase (AChE) (Casida 1964). Any model which would predict the toxicity and fate of these compounds would have to account for specific interaction between these compounds and biological or chemical receptor sites. Attempts at predicting OP and carbamate toxicity using parameters which work well for neutral organic compounds, such as log Kow, have proven problematic (see Lydy et al. 1990b, Zaroogian et al. 1985, de Bruijn and Hermens 1991a). Further, Fisher et al.
(1993) asserted that it is impossible to know what aspect of the intoxication process is being modeled even if predictors are adequately correlating with OP and 15
carbamate toxicity. Clearly more research needs to be conducted to be determine how the fate and effects of OP and carbamate compounds are to be predicted.
Objectives
The purpose of this dissertation was to determine, and if possible predict, some of the biological and physical processes which are important in the intoxication of the aquatic invertebrate the midge, Chironomus riparius by selected OP and carbamate insecticides. In Chapter 1, the ability of 3 unidimensional and 2 multidimensional models were assessed as to whether they could predict the movement of OP and carbamate insecticides into the midge and interact with the target enzyme, AChE at varying temperatures.
The purpose of chapter 2 was to determine if symptomatic and enzymatic recovery occurred in the midge following exposure to selected OP and carbamate compounds.
Recovery was measured at 3 temperatures to determine if temperature during recovery could affect the post-exposure toxicity of the insecticides.
In Chapter 3, of the same 5 physical and structural parameters were assessed as to whether they could predict the persistence of OP and carbamate insecticides in water and their susceptibility to in vivo metabolism, by the midge. Three temperatures were used during persistence studies to assess whether the QSAR models could predict persistence at varying temperatures. CHAPTER I
QUANTITATIVE STRUCTURE ACTIVITY RELATIONSHIPS FOR
PREDICTING THE IN VIVO AND IN VITRO INHIBITION OF
ACETYLCHOLINESTERASE IN THE MIDGE,
ChlRONOMUS RIPARIUS vv
INTRODUCTION
Many models developed to predict toxicity based on quantitative structure activity relationships (QSAR) have focused on the lipophilicity of a compound, as expressed by Kow, the n-octanol/water partition coefficient. Log has been used to predict the fate (Halfon and Brueggemann
1990) and toxicity (Konneman 1981, Veith et al. 1983,
Hermens et al. 1984, Hermens 1986, Lipnick 1989, Kaiser and Esterby 1991, Newsome et al. 1991) of certain hydrophobic organic compounds. The mode of action for these compounds is narcosis, in which toxicant molecules penetrate cell membranes and exert their toxic effect by disrupting nerve cell membranes either by altering the lipid component of cell membranes or by binding to
17 18
specific receptor sites on protein molecules embedded in
cell membranes (Seeman 1972, Franks and Lieb 1985) .
Recently, efforts have been made to predict the toxicity of compounds with specific modes of action (eg
La11 1990, Schuurmann 1990). Organophosphorus (OP) and
carbamate compounds are toxic because they inhibit the
enzyme acetylcholinesterase (AChE) (Casida 1964). OP's
and carbamates posses functional groups which impart a high level of reactivity to the compounds. These
functional groups subjects OP's and carbamates to
oxidation, reduction, hydrolysis and synthesis reactions
which could affect their toxicity and bioavailablility
(Matsumura 1985). OP's and carbamates can also be metabolically activated (Matsumura 1985) or detoxified
(Aldridge 1953, Plapp and Casida 1958) further
complicating prediction.
While models predicting toxicity of hydrophobic
compounds center on the importance of partitioning through
biological membranes, predicting OP and carbamate toxicity
requires describing not only penetration into the organism
but interactions between cholinergic inhibitor and target
site as well. The three dimensional structure of the
compound, which would be important in determining the fit
into the receptor site, is likely to be of critical
importance to toxicity and therefore must be described by 19 a predictive model.
Although modeling the toxicity of various OP and carbamate compounds has had some success, limitations are still evident. In particular, Fisher et al. (1993) point out that it is impossible to know which of the numerous reactions in the intoxication process is being modeled, or if all reactions are being modeled together. Modeling the inhibition of AChE, using both in vivo and in vitro exposure, should provide valuable insight into determining what processes are being described by available models.
In this study we examined the ability of three physical and structural parameters, Kow, Molecular volume
(MV) and Henry's Law constant (HLC) and two multidimensional models, molecular connectivity (MC) indices and linear solvation energy relationships (LSER), to explain the variation, through regression analysis, in the in vivo and in vitro inhibition of AChE by twelve OP and carbamate compounds. In vivo inhibition was also measured under varying temperatures to determine the ability of the models studied to explain variation in the data under changing environmental conditions. In vivo and in vitro I50's were determined for each compound and regressed against the physical parameters selected to determine their predictive value. Finally, in vivo and in vitro Iso values were regressed against published toxicity 20
data.
MATERIALS AND METHODS
Organisms
The midge, Chironomus riparius, used throughout this
study, was chosen because it is easy to culture, has a
cosmopolitan distribution, and is important in aquatic
food chains. Midge cultures were maintained at room temperature (23 + 3 °C) according to the methods of
Estenik and Collins (1978). Fourth instar larvae were used in all tests.
Compounds
For measurement of in vivo inhibition studies, seven
OP (dichlorvos, ethion, leptophos, malathion, methyl parathion, parathion, and phosdrin) and five carbamate
compounds (aldicarb, carbaryl, carbofuran, mexacarbate,
and propoxur) of varying inhibitory capabilities were
studied. For in vitro studies, the same compounds were
used as well as the oxidative metabolites (aldicarb
sulfoxide, aldicarb sulfone, methyl paraoxon and paraoxon). All compounds were of reagent grade ( >96% purity) and were purchased from Chem Services Co (West
Chester, PA). 21
Physical and Structural Parameters
K^, and HLC values were taken from Suntio et al.
(1988) and MV values from Hansch and Leo (1979). Values for MC and LSER were obtained from Fisher et al. (1993) or were calculated using the MOLCONN computer program.
Values for LSER variables (Vj/100, it*, a, and /3) used in this study were either taken from Fisher et al. (1993) or were calculated by methods described in Hickey and
Passino-Reader (1991). V;/100, the intrinsic molecular volume (adjusted by a factor of 100 to make it comparable to the size of the other three parameters), represents the energy required to open a cavity in a solvent. IT, represents dipole interactions between solute and solvent and is a measure of a molecule's ability to stabilize a neighboring charge or dipole. Alpha (a) and beta (j8) represent the ability of a molecule to donate a hydrogen, or accept a hydrogen, respectively.
The four levels of MC indices used in this study were
°X, lXr 2X> an<* 3X* °X represents the volume of a molecule. xX, the simple connectivity index, is the sum of non hydrogen atoms adjacent to each other connected by a single bond. 2x, and 3x represent the sum of nonhydrogen atoms connected by 2 and 3 bonds respectively, x parameters were also included which account for multiple bonds and the presence of ionized groups, designated °xv - 22
3XV. The cluster type subclass, both valence and non valence, for the third level of connectivity, designated
3cX and 3cxv, was also used. MC indices were developed to describe relationships within molecules and how this relates to the more complex three dimensional aspects.
Enzyme preparation
Groups of midges were homogenized using a motor driven Potter-Elvehjem tissue homogenizer. Each group of twenty midges was homogenized in 1 ml 0.1 M tris-HCL buffer (pH 8.0) for 20 s at 3-5 °C. Homogenates were poured into 1.5 ml microcentrifuge tubes and centrifuged for 10 min at 8500 rpm in a Dupont Sorvall RC-5B refrigerated centrifuge. Enzyme preparations for in vivo
I50 measurements were kept on ice throughout the homogenization and assay procedures. Enzyme for in vitro
I50 measurements was kept at 20 °C during the assay procedures. Only supernatant from the centrifuged samples was used for enzyme activity measurements and total protein content analysis.
Enyzme assay methods
AChE activity was measured colorimetrically (Ellman et al. 1961). The standard reaction mixture contained 2.6 ml 0.01 M tris-HCL buffer (pH 8.0), 1 ml 0.58 M dithio- 23 bis-nitrobenzoic acid (DTNB), 0.4 ml 3.84 x 10"4 M acetylthiocholine iodide (ATC), and 0.05 ml enzyme preparation. The blank contained buffer, DTNB, and deionized water in place of ATC. The reaction mixture was poured into a cuvette and change in absorbency read at 412 nm in a Varian DMS100 UV/VIS spectrophotometer. Change in absorbency was calculated by the DMS100 and reported every
45 s. Four readings were combined and the average change in absorbency/min calculated. Specific activity of each sample was calculated from equation (1) reported in
Fairbrother et al. (1991):
(1) Specific activity = (A x jumoles) / (min x 13.6 x mg
protein)
Where specific activity is in units/min/mg protein, A is the change in absorbance, 13.6 is the thionitrobenzoic acid extinction coefficient and mg protein is the total protein present in the enzyme mixture.
Protein Measurement
The filter paper dye-binding method was used to determine total protein content of midge samples (Minamide and Bamburg 1990). A grid of 1.5 x 1.5 cm squares was drawn on a sheet of clean Whatman No. 1 filter paper with a No. 2 pencil. Each enzyme preparation (4 jtil) was spotted onto a square and allowed to dry. Protein standard solutions of bovine serum albumin (BSA) from 0 to
30 nq were applied in triplicate amounts. The filter paper was rinsed with absolute methanol for 10-20 s and allowed to dry. The filter paper was gently agitated in a shallow glass tray filled with approximately 200 ml
Coomasie brilliant blue G (in 7% acetic acid) for 30 min at room temperature. The paper was then removed and destained in 7% acetic acid with gentle agitation for 3 h.
After destaining, the squares were cut out and each placed in a 1.5 ml microcentrifuge tube containing 1.5 ml extraction buffer (66% methanol, 33% water, 1% ammonium hydroxide). Tubes were mixed vigorously (vortex mixer) for several min, allowed to stand for 5 min, then mixed again. The absorbance was measured at 610 nm. A standard curve was constructed from the BSA standards.
Enzyme protein was determined by comparison to the standard curve.
Dosing Procedure for In Vivo AChE Inhibition Measurements
Groups of 20 fourth instar midges were exposed in vivo to a given insecticide in 1 1 beakers filled with 500 ml soft standard reference water (USEPA 1975) adjusted to pH 7.0 for 24 hr. All beakers were covered with aluminum foil to prevent excessive evaporation of water and were
placed in a Forma Scientific (#3740) environmental chamber
at the appropriate temperature. Midges were acclimated
for 6 h prior to dosing. A photoperiod of 14/10 h 1/d was
maintained through all experiments. Five concentrations
of each insecticide were prepared in a small volume of
acetone. Quantities of acetone carrier did not exceed 0.5 ml/500 ml water. Controls received 0.5 ml acetone only.
All treatment and controls were prepared in groups of
three replicates. Concentrations of each insecticide were
selected both above and below the 24 h EC50 for a particular chemical, determined in preliminary tests.
Three different exposure temperatures were used: 10, 20,
and 30 °C. After the 24 h exposure period, each beaker was emptied through a fine mesh sieve and the midges
collected and wrapped in aluminum foil for freezing. All midges were kept at -20 °C until assayed. All midges were
assayed within 72 hours of dosing. In vivo I50 values were
determined for each compound for the three temperatures by
regressing molar concentration of insecticide (based on
exposure water concentration) against percent inhibition
and obtaining the concentration necessary to inhibit 50%
of AChE from the regression line. Percent inhibition was
normalized by arcsine square root transformation (Sokal
and Rolf 1988). 26
Dosing Procedure for In Vitro AChE Inhibition Measurement
Dosing was performed on 1 ml supernatant obtained after groups of 20 midges were homogenized and centrifuged as described previously. After centrifuging, supernatants were poured into clean 1.5 ml microcentrifuge tubes and the temperature allowed to equilibrate at 20 °C for 30 min.
Five concentrations of each insecticide were prepared with a small volume of acetone. Quantities of carrier did not exceed 1 ixl/m 1 supernatant. Controls received 1 fil acetone only. All treatment and controls were prepared in groups of three replicates. Concentrations of each insecticide were selected above and below the in vivo I50 value for a particular compound. The exposure period was for 30 min, which was found to be sufficient for enzyme inhibition to level off. After dosing, enzyme preparations were assayed immediately and in vitro Iso values determined as described previously.
Data Analysis
I50 data were analyzed by an analysis of variance
(ANOVA) to determine significant differences between treatment means. Mean separation was achieved by using
Tukey's Studentized Mean test. In vivo Iso data for all three temperatures were regressed against each of the molecular descriptors (SAS 1982), with the following 27 exception. Since only 20 °C HLC values were obtained, regressions were only conducted between HLC and in vivo and in vitro Iso's using 20 °C data. In vitro I50 data for those compounds which inhibited AChE equal to or below in vivo IJ0 values were also regressed against each molecular descriptor. The thiono form OP's, aldicarb and aldicarb sulfone were not used in regression analysis since results showed they did not contribute significantly to toxicity.
For multidimensional descriptors (MC indices and LSER) I50 values were regressed against all possible combinations of up to three descriptors. Regressions were sorted by dosing procedure, temperature and class of compound and the best regressions selected. Regressions were only considered significant if P values were 0.05 or below.
Finally, in vivo Iso's at 20 °C and in vitro I50's were regressed against each other and 24 h EC^'s from Fisher et al. (1993).
RESULTS
In Vivo I so Data
The in vivo Ij0 values of the 12 insecticides tested were significantly affected by temperature (Table 1). In general, an increase in temperature resulted in a decrease in Iso, indicating that less compound was required to cause inhibition as temperature increased. However, variation Table 1. Mean in vivo I50 Values of Organophosphorous and Carbamate Insecticides in the Midge, Chironomus rioarius at 10, 20 and 30 °c (± S.E.)• © H o u o O O Compound N) 30 °C
Aldicarb 9.64 X l 0-8» 3.10 X l0-8a 7.92 X 10'8a (6.75 X 1 0 ’8) (1.14 X 10'8) (2.04 X 10'8)
Carbary1 7.33 X 10'7a 4.22 X 10‘7a 3.16 X l 0 -7a (3.53 X I O ’7) (1.60 X 10'7) (1.18 X 10'7)
Carbofuran 1.28 X 1 0 ’7a 8.96 X IQ-Sa 6. 06 X 10-81) (5.24 X 109) (2.29 X 10'8) (9.18 X 10'9)
Dichlorvos 1.83 X 1 0 ‘7a 2.25 X 10'8b 1.07 X 10-8c (3.18 X 10'7) (2.50 X 109) (7.33 X IO-10)
Ethion 3.86 X l 0-6a 2.89 X IO-78 8.19 X 1 0 ‘8c (4.91 X I O 8) (3.33 X 10'9) (1.51 X I O ’8)
Leptophos 4.01 X 10*7a 3.52 X 10'8a 1.36 X 10'8a (2.54 X 1 0 ‘7) (2.98 X 10-9) (2.31 X 10'9)
Malathion 1.65 x l 0-8a 4.76 X 10’9b 1.65 X 10"9b (7.04 X io-10) (1.02 X 109) (1.36 X IO-10) Table 1 (continued)
Methyl 2.48 X 10'8a 5.23 X 10'9b 2.51 X 10'9b Parathion (5.66 X IO'9) (2.01 X 10'9) (1.70 X I O ’10)
Mexacarbate 8.56 X 1 0 7a 1.77 X 10'7b 1.69 X 10'8b (2.03 X 1 0 ’7) (5.53 X 10*8) (7.97 X 1 0 ’8)
Parathion 6.82 X 10'8a 9.18 X 10 % 3.94 X 1 0 % (3.86 X 1 0 ’9) (1.32 X 10'9) (1.56 X 1 0 ’9)
Phosdrin 2.61 X 1 0 8a 2.51 X 10'8b 7.85 X 10-9b (6.75 X 10'9) (1.33 X I O 9) (8.49 X I O ’10)
Propoxur 2.75 X 10-7a 2.61 X 10'7a 2.83 X 1 0 8b (4.36 X IO'8) (3.55 X 10*8) (1.67 X IO'9) aI50 values for a given compound followed by the same letter are not significantly different based on Tukey's Studentized Mean test. Table 2. Mean in vivo and in vitro 1^ Values of Organophosphorous and Carbamate Insecticides in the Midge, Chironomus riparius at 20 °C (+ S.E.).
in vivo in vitro in vitro for metabolites Compound Io,a fM> Iov. fM^ sulfoxide sulfone______-oxon
Aldicarb 3.10 x IO-8* 2.79 x 10-* 9.68 x 10-®* 3.10 x lO^ --- (1.14 x 10-8) (1.87 x IO’7) (9.90 x 1 0 ‘9) (6.38 x 10'7)
Carbaryl 4.22 x IO'7* 3.45 x IO'71 ------(1.60 x IO'7) (4.30 x 10-*)
Carbofuran 8.96 x IO-81 1.28 x 10'8b ------(2.29 x IO4*) (1.51 x IO'9)
Dichlorvos 2.25 x IO'8* 7.06 x 10'7b ------(2.50 x 10‘9) (1.20 x IO'7)
Ethion 2.89 x IO’7* >5.20 x lO^ ------(3.33 x I O ’9) ( )
Leptophos 3.52 x 10-®“ >7.28 x 10'7b ------(2.98 x IO9) ( )
Malathion 4.76 x IO'9" >4.33 x 1 0“* ------(7.04 x I O ’10) ( )
Methyl Parathion 5.23 x IO’91 3.53 x IO'51’ 3.25 x IO"8* (2.01 x IO'9) (2.99 x I O ’5) (2.57 x I O ’9)
Mexacarbate 1.77 x 10'7‘ 2.59 x IO'7* ------(3.53 x 10-*) (1.10 x 10-8) Table 2 (continued)
Parathion 9.18 x I O ’91 8.98 x 10-» 1.99 x 10‘7“ (2.30 x IO9) (3.85 x 10-6) (2.31 x IO'9)
Phosdrin 2.51 x IO-8* 2.62 x 1 0 -7b ------(1.33 x I O ’9) (7.47 x IO"8)
Propoxur 2.61 x IO'7* 1.50 x 1 0 ‘7b ---- (3.55 x 10-8) (2.34 x IO-8)
“Ijo values followed by the same letter are not significantly different based on Tukey's Studentized Mean test. 32 was seen among compounds with respect to the influence of temperature on I50. The Iso's of three compounds, aldicarb, carbaryl, and leptophos showed no significant change with changing temperature. I50 values for the remaining nine compounds decreased significantly at either 20 or 30 °C.
In Vitro %so Data
The in vitro I 50's of thiono form OP's were significantly higher than corresponding in vivo I50 values
(Table 2). The water solubility was exceeded for three compounds, ethion, leptophos and malathion before the I50 was reached. Carbamates carbofuran and propoxur had in vitro I50 values significantly lower than in vivo values, whereas no significant differences were found between in vivo and in vitro I50's for carbaryl and mexacarbate. The in vitro I50 for aldicarb was significantly higher than the in vivo I50.
The metabolites aldicarb sulfoxide, methyl paraoxon and paraoxon were not significantly different than the in vivo Iso for the respective parent compound (Table 3). The in vitro I50 for aldicarb sulfone was significantly higher than the in vivo I50 for aldicarb and the in vitro IJ0 for aldicarb sulfoxide but not significantly different than the in vitro Iso for aldicarb. Table 3. Molecular Descriptors used in Regression Analysis with in vivo and in vitro I50 values.
Compound Molecular Volume8 Henry's Law Constant8 Log (cm3/mol) (Pa m3/mol)
Aldicarb 224.3 0.00032 1.10 Aldicarb Sulfoxide NDb ND ND
Carbaryl 218.7 0.0013 2.81
Carbofuran 240.8 0.00051 2.32
Dichlorvos 167.5 0.190 1.40
Ethion 350.2 0.032 5.07
Leptophos 317.8 0.250 6.31
Malathion 319.1 0.0023 2.36
Methyl Parathion 207.5 0.021 3.32
Methyl Paraoxon ND ND 1.41
Parathion 251.9 0.012 3.80
Paraoxon ND ND 2.11 Table 3 (continued)
Phosdrin ND ND 1.11
Propoxur 244.5 0.130 1.50
"Data from Suntio et. al. (1988). bData from Hansch and Leo (1979). cNo data available.
U) 35
Regression Analyses with MV, HLC and Log Kolv
The one-dimensional molecular descriptors used in this study are given in Table 3. Regression analysis was performed between these values and in vitro I50 data and in vivo I50 data to determine if a linear relationship exists between these variables. For MV and Kw , regressions were conducted between each descriptor and in vivo and in vitro
I50 values for each temperature.
The results of the regression analysis showed no linear relationship between any of the one dimensional descriptors and either in vivo or in vitro I50. Molecular volume gave the best results for in vivo l50 data but described only 33% of the variability for the 10 °C values. No consistent temperature effects were observed for any of the three models. In vivo and in vitro data were separated according to class of compound and again regression analysis performed, but no statistically significant relationships were observed.
Regression Analysis with LSER and MC Indices
Physical and structural parameters for LSER and MC indices (Tables 4 and 5) yielded significant correlations
(Tables 6 and 7) . When LSER parameters n*, a, and j8 were Table 4. Linear Solvation Energy Relationships for Insecticides used in Regression Analyses".
Compound Vj/100 7r* a _ A _ Aldicarb 1.202 1.14 0.30 1.50 - sulfoxide 1.235 1.78 0.30 2 .bo
Carbaryl 1.183 1.17 0.30 0.90
Carbofuran 1.303 1.45 0.30 1.15
Dichlorvos 0.861 0.85 0.10 0.77
Ethion 1.800 1.14 0.10 2.00
Leptophos 1.643 1.43 0.00 1.01
Malathion 1.421 1.65 0.00 1.75
Methyl Parathion 1.214 1.30 0.16 1.15
Methyl Paraoxon 1.142 1.35 0.16 0.95
Mexacarbate 1.379 1.16 0.30 1.22
w G\ Table 4 (continued)
Parathion 1.410 1.30 0.16 1.16
Paraoxon 1.338 1.33 0.16 0.96
Phosdrin 1.016 1.30 0.22 1.25
Propoxur 1.245 1.35 0.30 1.10
“Data from Fisher et al. (1993) or calculated using guidelines in Hickey and Passino-Reader (1991). Table 5. Molecular Connectivity Indices used in Regression Analysis*.
Compound °X *X 2X 3X 3cx °XV *XV 2XV 3x v 3cXv
Aldicarb 5.4 9.7 5.1 3.4 1.7 8.6 4.7 4.3 2.5 1.7
Aldicarb sulfoxide 10.5 6.0 5.5 3.8 1.6 9.0 5.2 4.9 3.0 1.9
Carbaryl 10.6 7.3 6.1 5.1 0.7 8.4 4.8 3.2 2.3 0.3
Carbofuran 12.1 7.3 6.5 3.9 1.7 9.8 5.2 4.2 2.2 1.0
Dichlorvos 8.0 4.6 3.6 2.8 0.9 7.2 4.1 3.2 2.0 0.3
Ethion 14.8 8.9 7.5 4.9 1.8 16.8 4.5 15.9 13.6 4.4
Leptophos 14.7 9.5 8.7 7.3 1.8 15.1 9.7 8.9 7.2 1.9
Malathion 14.9 8.9 7.4 5.5 1.6 13.9 9.8 9.7 7.8 2.5
Methyl Parathion 12.2 7.5 6.7 5.3 1.6 10.4 6.7 5.8 4.2 1.0
Methyl Paraoxon 12.2 7.5 6.7 5.3 1.6 9.5 5.6 4.5 3.0 0.5
Mexacarbate 12.3 7.5 6.9 4.9 1.4 10.4 5.1 3.9 7.3 0.7 Table 5 (continued)
Parathion 13 .9 8.3 8.0 6.2 1.9 12.4 7.5 7.7 6.4 2.2
Paraoxon 13.6 8.5 7.5 5.3 1.6 11.0 6.8 4.7 3.4 0.5
Phosdrin 10.3 6.1 4.8 3.4 0.7 11.3 8.1 5.7 3.6 0.5
Propoxur 11.3 7.1 6.2 4.2 1.0 9.1 4.8 3.3 1.6 0.4
“Data from Fisher et al. (1993) or calculated according to the methods of Kier and Hall (1986). Table 6. Summary of Best Regression Analyses of In vivo I50 values for LSER, and MC Models.
Compound class Treatment Descriptor Variables r2
OP's and Carbamates 10 °C LSER 7r% P, a 0.62 MC °x, lx, V 0.75 20 °C LSER NSRa MC °X/ lx / 3x 0.71 30 °C LSERNSR MC NSR
OP's Only 10 °c LSER n\ P , a 0.99 MC °x / V 0.94 20 °C LSER n , P, a 0.97 0<< 1 vv MC Xr X 0.92 30 °C LSER NSR MC °x / V 0.93 Table 6 (continued)
Carbamates Only 10 °C LSER NSR MCNSR
20 °C LSER 0 0.79 MC 3x, V 0.98 30 °C LSERNSR MC 2 X, 3 X 0.99
“No significant relationships 42
Table 7. Summary of regression analyses of in vitro I50 values for LSER and MC models.
Compound Class Descriptor Variables_____ ri_
OP's and Carbamates LSER j3 0.88
MC ix, V 0.91
Carbamates LSER |3 0.88
MC 3x, 2XV 0.99 regressed against I50 values at 10 °C, for all compounds a
linear relationship was evident (adjusted r2 = 0.62; p <
0.04; df = 11; F = 4.41; I50 = -4.37 x lO^ff*) + 1.65 x 10'
6 (j8) - 3.38 X 10'6(a) + 4.77 x 10-6) (Table 6). However, at
2 0 and 30 °c no significant linear relationships was found with.LSER. When the data were separated by compound class correlations improved significantly. When n*, a, and /3 parameters were regressed against the OP's alone, adjusted the best at 20 °C.
Regressions of in vivo l5Q's with MC indices produced the best results overall (Table 6) . For 10 °C, the best relationship for combined OP and carbamate data combined was produced when °x, *x, and 3xv were used in combination
(r2 = 0.75; p < 0.008; df = 11; F = 8.18; I50 = -1.79 x 10'
7(°x) + 2.12 x 10'7(1x) + 2.15 X 10-8(3xv) + 5-44 X 10'7) . This combination was not, however, the best relationship for 20
0C, where °x, lx, and 3x produced the best relationship (r2
= 0.71; p < 0.01; df = 11; F = 6.73; I50 = -2.00 x 10'7(°x)
+ 5.40 X 10’7(1x) - 2.14 X 107(3x) + 2.82 X 10'7) . No relationships were discerned between independent and dependant variables for OP and carbamate data at 30 °c.
When data were separated by class of compound, correlation coefficients for significant models increased to nearly
1.00. When OP's were regressed against I50's at 10 °C, °x and *xv produced a significant correlation (r2 = 0.94; df = 44
11; I50 = -5.81 X 10-7(°X) + 6.66 X 10-7(‘xv) + 2.60 X 10'6) .
At 20 °C °x and V (r2 = 0.92; df = 6; I50 = -4.36 x 10'8(°x)
+ 4.84 x 10'8(1xv) + 2.15 x 10'7) and at 30 °C °x and V (r2 =
0.93; df = 6; I50 = -1.26 x lO'Vx) + 1.36 x 10'8(1xv) + 6.66
x 10'8) produced the best relationships. Regressions with
carbamates alone produced significant correlations at 20
(r2 = 0.98; df = 11; I50 = 6.79 x 10'8(3x) - 2.52 x 10'7(2xv) +
8.36 x 107) and 30 °C (r2 = 0.99; df = 4; I50 = -2.74 x 10'
7(2X) + 2.98 x 107(3x) + 4.70 X 10'7 ).
Regressions of in vitro I50 data and LSER and MC also yielded higher correlation coefficients than regressions using unidimensional parameters (Table 7). Regression of
LSER parameter /3 against in vitro I50 data gave an adjusted r2 of 0.88 for either OP's and carbamates combined (p <
0.002; df = 6; i50 = 2.47 x 10*6(/3) - 2.38 x 10’6) or carbamates alone (p < 0.02; df = 4; I50 = 2.59 x 10'6(j8) -
2.59 x 10'6) . Regression of *x and *xv gave an adjusted r2 of 0.91 (p < 0.008; df = 6; F = 21.16; I50 = -1.57 x 10'
6(’x) + 9.07 x 10-7(‘xv) + 7.18 x 10-6) for OP's and carbamates in combination. Regression of 3x and 2xv gave an adjusted r2 of 0.99 against carbamates alone (p <
0.005; df = 4; F = 204.1; IJ0 = -4.08 x 106(3x) + 3.18 x 10'
6(2XV) “ 6.28 x IQ’6) . Figure 2. Regression of in vivo Iso's (20 °C) for 7 OP and
5 carbamate insecticides against corresponding 24 h EC50's.
45 EC50 (M) 0E-7 E .0 3 6.0E—7 - ■ 0.0 0.0 5 = .8-(C0 1.57E-8 + 2.68E-9(EC50) = l50 .E- 20 7 .E- 4. 7 50 -7 5.0E -7 E .0 4 -7 3.0E -7 2.0E -7 1.0E In iue 2. Figure VIVO 5 (M) I50 46 Figure 3. Regression of in vitro I50's for 2 OP and 5 carbamate insecticides against corresponding 24 h EC50's
(20 °C) .
47 EC50 (M) .E-7 4.0E .E-7 6.0E 0.0 0.0 l 50 24e9ES) 9.82e-8 + 2.41e-9(ECS0) = 1.0E-7 n ir I0 (M) I50 vitro In iue 3. Figure .E73.E-7 .0E 3 2.0E-7 .E-7 4.0E 48 49
Regression of I50's with toxicity
There was a strong relationship between the in vivo
I50's measured in this study and EC50 values of these 12 insecticides to the midge, Chironomus riparius (from
Fisher et al. 1993) (r2 = 0.82; p < 0.0001; df = 11; F =
47.01) (Fig. 2). There was a linear relationship between in vitro Iso's and the corresponding ECS0's (r2 = 0.59;p <
0.04; df = 6; F = 7.15) (Fig. 3). There was also a significant relationship between in vivo and in vitro Iso's
(r2 = 0.52; p < 0.0000001; df = 6 ; F = 5.36) (Fig. 4).
DISCUSSION
Midge AChE was exposed to cholinergic insecticides both directly and indirectly. Dosing a tissue homogenate
(in vitro exposure) allowed the insecticide direct access to the target enzyme without having to negotiate the various biological impediments present in a living midge such as the cuticle, and the many degredative enzymes midges possess. Only volatilization from the test system, sorption to the sides of the container or binding of the toxicant to other molecules (e.g., proteins) would affect the concentration at the receptor site. The factors which would be most important for inhibition should be the steric and electronic characteristics of the toxicant and how these affected interaction with the target site. Figure 4. Regression of in vivo ISQ's for 2 OP and 5 carbamate insecticides against corresponding in vitro
Iso's (20 °C) .
50 ,n v'tro 50 i cm; .E-7+ 1.0E 0E- + -7 E .0 3 .E-7 4.0E 0.0 0.0 in vivo in .E-7 1.5E I qq = 0.93(in vitro vitro 0.93(in = n io 50 (M) 0 l5 vivo In Figure 4. Figure 3.0E-7 3.0E-7 I^ q ) - 2.80E—9 - ) .E-7 4.5E 6.0E—7 51 Indirect exposure of AChE was accomplished by exposing living midges, through contaminated water.
Numerous physical and biological processes would affect the ultimate concentration at the AChE receptor site, and resulting inhibition. These processes can be placed into three general categories: 1) factors affecting exposure;
2) toxicokinetics of the insecticide and; 3) toxicodynamics (Ariens 1980, McCarty 1989). Exposure involves the factors which affect the availability of the toxicant to the organism, such as temperature, route of exposure, stability of the compound in the test system
(abiotic and biotic hydrolysis), and volatility of the toxicant. Toxicokinetics involves the uptake, metabolism, distribution and elimination of the toxicant by the exposed organism. Toxicodynamics involves the interaction of the toxicant with the target site. It is the complex interaction of all these factors which ultimately determines the amount of toxicant at the receptor site and any model developed to predict the toxicity of cholinergic compounds must account for these processes.
Regressions with MV, HLC, and Log K^
The one dimensional models were not able to explain the variation in the in vitro I50 data or the in vivo I50 data at any of the three temperatures. This agrees with 53
the results of Fisher et al. (1993), but contrasts with
the results of other studies.
The toxicity of nonionic organic compounds to fish
was successfully predicted by McGowan and Mellors (1986)
using molecular volume. However, they stated that
compounds with selective toxicity would not be accurately predicted by this method. Another study found a strongly negative correlation (r = -0.89) between LC50 of a wide range of organic compounds to fish and parachor (molar volume) (Hodson et al. 1988). Again the compounds tested
exhibit a non-specific mode of action. Molecular volume
appears to predict toxicity best for narcotic compounds where the size, not the structure, of the molecule is
important (de Bruijn and Hermens 1991a).
HLC values for dichlorvos, leptophos and propoxur were sufficiently high enough that volatilization from the test system could occur and thus decrease exposure. This would not have been a factor for in vitro studies since dosing was performed in epindorf tubes which were sealed until assayed. However, In vivo studies were conducted in beakers where compound could escape, yet cholinesterase
inhibition did occur. Therefore insufficient compound
left the system to affect the results.
The results showed that was insufficient to explain the variation in cholinesterase inhibition for the compounds tested. This agrees with the results of Fisher et al. (1993) who found no relationship between toxicity of these same 7 OP and 5 carbamate compounds to the midge,
Chironomus riparius, and Log Kw . This also agrees with results of de Bruijn and Hermens (1991a) who found no relationship between the toxicity of 2 0 OP compounds to fish and Log K^. The relatively low success (r2 = 0.53) in modeling OP toxicity was attributed to important biotransformation reactions such as deactivation of the
OP's through enzymatic hydrolysis as well as different rates of metabolic conversion of thio to oxon analogues
(de Bruijn and Hermens 1991a). Both studies also attributed the poor correlation of and toxicity with the specific mode of action of these compounds. has been traditionally used to predict the toxicity of compounds with a general narcosis-type toxicity (eg
Konneman 1981, Veith et al. 1983, Hermens et al. 1984) .
However, the results of this study and the others mentioned are in contrast with the results of Vighi et al.
(1991) . They found that (combined with Kw2) was the most important parameter (r2 = 0.56) in describing the toxicity of 22 OP compounds to Daphnia. When Kow was combined with other parameters to describe the volumetric and electronic characteristics as well as the electronegativity of the leaving group on the OP compound, the r2 = 0.89. Although Kw was found to correlate with toxicity to Dapnia it had little direct relationship (r2 =
0.22) with toxicity of 15 OP compounds to honeybees (Aphis mellifera) where uptake was from contaminated honey (Vighi
et al. 1991). However, when Kw was combined in multiple regressions with the other parameters mentioned before gave r2 = 0.91.
The discrepancy between the results of this study and
Vighi et al. (1991) may be due to several factors. The compounds used in this study and Fisher et al. (1993) were both carbamates and OP's. These two classes are structurally diverse and the descriptive abilities of many one dimensional parameters are severely limited when compounds of different classes are compared together
(Eriksson et al. 1991). The predictive ability of Kow in
Vighi et al. (1992) was diminished when exposure was through ingestion of contaminated honey. After ingestion, the compounds would have entered the honey stomach then the digestive system of the bee probably increasing the likelihood that the compounds would be metabolically altered. In fact, Vighi et al. (1991) reported that better results were obtained for correlations of toxicity to honeybees using values for the oxon metabolites of the compounds tested. Further, Fisher et al. (1993) suggested that the predictive ability of breaks down when attempting to
account for internal processes which are not dependant on
lipid partitioning. This is supported by de Bruijn and
Hermens (1991b) who found a definite correlation between uptake and elimination rate constants for 12 OP compounds
in fish and Log K,^. However, they found that compounds which were metabolized during the experimental period were not eliminated at a rate predictable by K,^. Apparently, the midges in this study and that of Fisher et al. (1993) metabolically altered the compounds rendering unsuitable for predicting the toxicity or AChE inhibition.
Regressions with multidimensional descriptors
LSER uses the energy requirements in the solubilization process to relate the properties of a compound to molecular structure (Hickey and Passino-Reader
1991). LSER has been used to predict the toxicity to fish
(Kamlet et al. 1987) and even bacteria (Kamlet et al.
1986) .
When OP's and carbamates were analyzed together, LSER was only marginally successful at describing the variation of in vivo I50 data over the three temperatures. LSER was best able to describe the combined I50 data at 10 °C. When regressions were performed with OP's and carbamates combined tt*, /?, and a surfaced as the most important 57 parameters describing changes in toxicity. However, LSER was not able to describe significant variation when the temperature was raised to either 20 or 30 °C. This combination of n*, j8, and a produced an even higher correlation when OP's were analyzed alone and worked at both 10 and 20°C indicating that LSER can be used across different temperatures.
The fact that LSER had difficulty describing the combined data at higher temperatures, producing no significant linear relationships at 30°C, was unexpected.
The solubility of many compounds in water is known to increase with increasing temperature (Stryer 1989).
Therefore it is reasonable that LSER would better describe the variation in toxicity data at higher temperatures, if the availability of insecticide is critical to intoxication. It may be that the compounds were differentially affected by the increase in temperature thus making it more difficult for LSER to describe the data, although further research would be necessary to determine if this is true. Also, higher temperatures would likely have increased the rate of uptake due to increased respiration by the midges. This would likely cause more compound reaching the target site than would be expected by passive diffusion across the insect's cuticle.
It is likely that LSER, which describes phenomena by 58
solute-solvent interactions, could not account for this
factor.
Although 7r*, 0, and a proved best for regressions which included OP compounds, 0 was the only parameter which was used consistently. It was also the only LSER parameter which described AChE inhibition by carbamates.
The predominance of j8 was also reported in Fisher et al.
(1993) , in which /3 was consistently correlated with
changes in ECS0 of pesticides to midges. They suggested that (5 might reflect the susceptibility to nucleophilic attack leading to phosphorylation or carbamylation of AChE
in the midge. This is supported by the regressions of
LSER parameters with in vitro I50 data (Table 7) and most of the in vivo I50 data (Table 6) . The fact that alone can describe changes in the in vitro I50 data, in which all of the organismal physiological impediments to
intoxication have been removed, suggested that j(3 is characterizing interaction with AChE. However, regressions of in vitro I50 data with ECS0's from Fisher et al. (1993) did not result in a highly significant correlation, as it did when in vivo I50's were used in regression with toxicity. If /3 is describing interaction between toxicant and target site and this is of key importance to both inhibition and toxicity, then there should have been a close correlation between in vitro Ij/s and toxicity. Therefore it is likely that /3 is not characterizing the same phenomena in both studies. It is possible that in regressions with toxicity and in vivo
Iso's |3 is describing interactions between the insecticide and surrounding media which allow it to remain in solution. This would be important for penetration of the cuticle and solubility in the midge's hemolymph which would be necessary for the insecticide to reach the target site. Regressions with in vivo I50 data required two other parameters, n* and a, which also describe interactions between solute and solvent which keeps the solute in solution (Hickey and Passino-Reader 1991).
Except for the predominance of j8, the parameters which best described AChE inhibition at 20 °C were not the same combination of parameters which described toxicity at
20 °C in Fisher et al. (1993). It is likely that the difference is due to the fact that Fisher et al. (1993) used an alteration in the behavior of the midge as their biological endpoint. The later is more complex and involves many additional interactions other than inhibition of AChE, which are not characterized by 0.
MC indices, like LSER, relate the physical properties of a compound to it's molecular structure (Kier and Hall
1986). It encodes information not only about molecular size, but also skeletal branching, unsaturation and 60 heteroatom content (Hall and Kier 1984). MC indices have been used in modeling not only toxicity of hydrophobic organics (Hall and Kier 1984, Protic and Sabljic 1989), but also bioaccumulation and aerobic biodegradation
(Sabljic 1991).
MC indices, in contrast with LSER, gave significant correlation coefficients for all in vivo and in vitro I50 values. For in vivo I50 data, parameters describing size, such as °x and ‘x as well as higher order indices describing three dimensional structure such as 3xv surfaced as significant most often. This suggests that MC is describing the interactions between the insecticide and target site. This is supported by the regressions with in vitro I50 data, in which parameters describing size and three dimensional structure gave high correlation coefficients.
MC indices also showed the ability to describe AChE inhibition across temperatures, in some situations with the same group of parameters. For regressions with OP's alone the same parameters explained over 92% of the variation at all three temperatures. Evidently the shift between in vivo l50's over the three temperatures for OP's was small enough that MC indices could still describe a large percentage of the variation. The same parameters were also used for regressions with OP's and carbamates 61
combined between 10 and 30, but not between 10 and 2 0 or
20 and 30 °C. We expected that the parameters used to
describe AChE inhibition would be the same between 10 and
20 or 20 and 30 °C but different between 10 and 30 °C since
the results showed a significant decrease for in vivo I50
values between these two temperatures.
Although MC and LSER did give strong correlations with the inhibition data obtained in this study, there are
several problems. Whenever regression analysis is conducted there is always a possibility of chance correlations between the observations and independent variables. It has been suggested that the ratio of
observations to independent variables be greater than 10
(Hall and Kier 1984, Protic and Sabljic 1989). Since the ratio of observations to variables did not exceed 10 in
every instance, the results of these regressions must be considered preliminary.
Regressions of in vivo and in vitro Iso's with toxicity
There was a strong correlation between in vivo I50's and toxicity as defined by the ECS0 Fisher et al. 1993) .
This was expected since AChE inhibition is the mode of action for these compounds and both toxicity and the in vivo Ijq's were based on exposure water concentration.
Thus the concentration in the water which inhibits 50% of 62
the target enzyme should have a strong relationship
between the amount of compound necessary to cause 50%
effect.
There was, however, only a weak relationship between
in vitro l50's and toxicity (Fig. 3) . There was also only
a weak relationship between in vivo and in vitro Iso's
(Fig. 3) . The lack of linearity between the in vitro I50
and ECJ0 is due in part to the reduction in metabolism which occurs in the in vitro preparations. Once the toxicant enters the organism, it is subject to a variety of activation and/or detoxification reactions which are mediated primarily by multifunction oxidases (Matsumura
1985) . The increase in the in vitro I50 values of several organophosphorus compounds (dichlorvos, ethion, leptophos, malathion, parathion, and phosdrin) as well as for aldicarb, the one carbamate which is oxidatively activated, clearly suggests that the lack of such oxidation in vitro, reduces correlation with E C 50. This further substantiated by the finding that in vitro I50's for the oxon forms of methyl parathion and parathion, and the sulfoxide form of aldicarb were not significantly different than the in vivo IJ0 values for the parent forms of the compounds (Table 2). Thus the strong correlation between in vivo I50 and ECJ0 and the weak correlation between in vivo and in vitro I50 must reflect the 63
requirement for metabolic oxidation.
Conclusions
Overall, this study shows that the inhibition of AChE by cholinergic insecticides could be described by multidimensional models. Although the descriptive parameters varied with changing temperature, MC descriptors displayed some flexibility in their predictive abilities. The greatest limitation in their use lies in the inability to determine which parameters are important until after regression analysis has been performed on the test results. Further, related phenomena such as AChE inhibition and toxicity may require dissimilar predictive parameters. MC indices showed the most promise in their ability to explain > 95% of the variability under most of the conditions tested. Further exploration of the multidimensional parameters and their relationship to prediction of pesticide fate and effects appears warranted. Refinement of these models may lead to an enhanced ability to predict the biological activities of structurally diverse chemicals with specific modes of action under a variety of environmental conditions. CHAPTER II
SYMPTOMATIC AND ENZYMATIC RECOVERY FROM EXPOSURE TO
ORGANOPHOSPHORUS AND CARBAMATE INSECTICIDES IN THE
MIDGE, CHIRONOMUS RIPARIUS: TEMPERATURE AND
EPISODIC EXPOSURE
INTRODUCTION
Toxicity bioassays have focused primarily on
assessing the effects of continuous exposure of xenobiotics on aquatic organisms (Pasco and Shazili 1986,
Jarvinen et al. 1988). However, exposure of aquatic communities to pollutants is often periodic rather than continuous (Siem et al. 1984, Poirier and Surgeoner 1988,
Baughman et al. 1989, Parsons and Surgeoner 1991a).
Studies have shown that the impact of periodic exposure of aquatic organisms to pollutants can be quite different from continuous exposure. For example, Siem et al. (1984) found that steelhead trout (Salmo gairdneri) exposed to intermittent doses of copper suffered higher mortality than when exposed to continuous doses of equal
64 concentration. Daphnia pulex suffered a higher mortality rate when exposed to pulsed doses of copper than when exposed to copper continuously (Ingersoll and Winner
1982). Mortality of steelhead and cutthroat trout (Salmo clarki) was also higher when the fish were given multiple pulses of ammonia than when given continuous exposure, although sublethal exposure allowed the fish to tolerate higher doses given subsequently (Thurston et al. 1981).
However, Eaton et al. (1985) found no significant difference between the impact of chlorpyrifos on aquatic invertebrate communities receiving 24 h pulses every 14 days and communities receiving continuous exposure.
In spite of this, some researchers have argued that episodic pollution should be less toxic to aquatic organisms if sufficient time is allowed between exposures for recovery to take place (Mancini 1983, Wang and Hanson
1985). Breck (1988) suggested that the toxicity of individual pulses might be cumulative if the compounds in question were either eliminated slowly or the damage done not readily reversible. However, Parsons and Surgeoner
(1991a) found no significant difference in the L C 50 values to mosquito larvae (Aedes aegypti) between two 1 h pulse and 2 h continuous exposure to carbary1, carbofuran, or fenitrothion. Comparisons of multiple-pulse and continuous exposure also yielded no differences in 66 toxicity, even when the larvae were given up to 24 h to recover between exposures. They suggested that either there was little recovery from the effect of poisoning, or that 24 h between pulses was insufficient time for full recovery to take place.
Since acetylcholinesterase (AChE) inhibition is the mode of action for carbamate and organophosphorus (OP) compounds (Casida 1964), measuring recovery rates of inhibited enzyme would provide useful insight into assessing the toxicity of periodic exposure to these compounds. Further, since environmental factors, such as temperature, have been known to increase the toxicity of many cholinergic compounds (Rattner et al. 1987, Chapman et al. 1982, Lydy et al. 1990a) evaluating temperature's effects on recovery would be beneficial. If temperature affects the reactivation rate of AChE, then the amount of recovery between pulsed exposure to insecticides would also likely be affected by temperature.
Further, recovery of AChE following exposure is likely to limit it's usefulness as an indicator of exposure to OP and carbamate insecticides. The monitoring
AChE levels in aquatic and terrestrial organisms has often been proposed as a method for determining when exposure to cholinergic compounds has taken place (Coppage and
Matthews 1975, Verma et al. 1979). Depression of AChE activity levels in animals suspected to be poisoned to
levels 20% below control activity is assumed to indicate
exposure to cholinergic insecticides (Zinkl et al. 1980,
Fleming and Grue 1980). Rapid recovery of AChE activity at certain temperatures would make it unlikely that animals would be diagnosed as exposed even if they had been.
The purpose of this study was to observe and measure the recovery of the midge, Chironomus riparius, after being exposed to OP and carbamate compounds and relate this to the devlopment of QSAR. This was accomplished by:
1) measuring the in vivo recovery of acetylcholinesterase
(AChE) under varying temperatures; and 2) measuring the acute toxicity of one OP and four carbamate insecticides following continuous and pulsed exposure, where pulses were separated by up to 24 h in clean (undosed) water.
MATERIALS AND METHODS
Organisms
The midge, chironomus riparius, was used throughout this study. The midge was selected because it is easy to culture, and is important in aquatic food chains. Midge cultures were maintained at room temperature (23 + 3 °C) according to the methods of Estenik and Collins (1978). 68
Fourth instar larvae were used in all tests.
Compounds
For measurement of acute toxicity, two OP's
(parathion and malathion) and four carbamate insecticides
(aldicarb, carbaryl, carbofuran, and propoxur) of varying
inhibitory potential were examined. AChE reactivation studies were conducted with the four carbamates and one
OP (parathion). Pulsed exposure studies were conducted using the same compounds as for the toxicity measurements.
All compounds were of reagent grade ( >96% purity) and were purchased from Chem Services Co.
Enzyme preparation
Groups of twenty midges were homogenized using a
Potter-Elvehjem tissue homogenizer. Groups of midges were homogenized in 1 mL of 0.1 M tris-HCL buffer (pH 8.0) for
15 s at 3-5 °C. Each homogenate was then poured into a
1.5 mL microcentrifuge tube and centrifuged for 10 min at
8500 rpm in a Dupont Sorvall RC-5B refrigerated centrifuge. All enzyme preparations were kept on ice during the homogenization and assay procedures. Only the supernatant from the centrifuged samples was used for enzyme activity measurements and total protein content analysis. 69
Enzyme assay methods
AChE activity was measured colorimetrically using the
method described in Ellman et al. (1961). The standard
reaction mixture consisted of 2.6 mL 0.01 M tris-HCL
buffer (pH 8.0), 1 mL 0.58 M dithio-bis-nitrobenzoic acid
(DTNB), 0.4 mL 3.84 x 10"4 M acetylthiocholine iodide
(ATC), and 0.05 mL enzyme preparation. The blank
contained buffer, DTNB, and deionized water instead of
ATC. The reaction mixture was poured into a cuvette and
change in absorbency read at 412 nm in a Varian DMS100
UV/VIS spectrophotometer. Change in absorbency was
calculated by the DMS100 and reported every 45 s. Four
readings were combined and the average change in
absorbency/min calculated. The specific activity for each
sample was calculated based on the equation (2) reported
in Fairbrother et al. (1991):
(2) Specific activity = (A x nmoles) / (min x 0.0136 x c)
Where specific activity is in nmoles substrate hydrolyzed/min/mg protein, A is the change in absorbance,
0.0136 is the thionitrobenzoic acid extinction
coefficient, and C is the amount of total protein present
in the enzyme mixture (mg). 70
Protein Measurement
The filter paper dye-binding method described by
Minamide and Bamburg (1990) was used to determine the
total protein. Each enzyme preparation (4 /u 1) was spotted
onto a sheet of clean Whatman No. 1 filter on which a grid
of 1.5 x 1.5 cm squares was drawn with a No. 2 pencil.
Protein standard solutions of bovine serum albumin (BSA)
from 0 to 30 jug were applied in triplicate amounts. The
filter paper was rinsed with absolute methanol for 10-20
seconds, allowed to dry, and the filter paper gently agitated for 30 min in a shallow glass tray filled with approximately 200 mL Coomasie brilliant blue G (in 7% acetic acid). The paper was removed and destained in 7%
acetic acid with gentle agitation for 3 h. The squares were then cut out and each placed in a 1.5 mL microcentrifuge tubes containing 1.5 mL extraction buffer
(66% methanol, 33% water, 1% ammonium hydroxide). Tubes were mixed vigorously (vortex mixer) for several minutes, allowed to stand for 5 min, then mixed again. The absorbance was measured at 610 nm. A standard curve was constructed from the BSA standards and enzyme protein determined by comparison to the standard curve. 71
Dosing Procedure for AChE Reactivation Measurements
Groups of 20 fourth instar midges were exposed for 24 h to a given insecticide in 1 1 beakers filled with 500 mL soft standard reference water (USEPA 1975) adjusted to pH
7.0. Beakers were covered with aluminum foil to prevent excessive evaporation of water and placed in a Forma
Scientific (#3740) environmental chamber at the appropriate temperature. Midges were acclimated for at least 12 h prior to dosing. A photoperiod of 14/10 h was used throughout the experiments. One concentration of each insecticide was prepared in a small volume of acetone. Quantities of acetone carrier did not exceed 0.5 ml/500 ml water. Midges were dosed with 0.5 mL of a concentration which would produce 50% inhibition of AChE
(Table 8). Controls received 0.5 mL acetone alone. All treatment and controls were prepared in groups of three replicates. Three different exposure temperatures were used: 10, 20, and 30 °C. After the 24 h exposure period was over, each beaker was emptied through a fine mesh sieve, and the midges rinsed with deionized water. A time
0 sample was taken (the midges collected and wrapped in aluminum foil for freezing), then all other midges placed
in beakers of clean, soft standard reference water (pH 7), which had been previously acclimated to the proper temperature, and all beakers returned to the environmental 72
Table 8. Exposure level (M) of one organophosphorus and four carbamate insecticides at 10, 20 and 30 °C for the midge, Chironomus riparius, producing50% inhibition of AChE.
Compound 10 °C 20 °C 30 °c
Aldicarb 9.64 x 10'8 3.10 x 10'8 7.92 x 1 0 ‘8
Carbaryl 7.33 x 10"7 4.22 x 10'7 3.16 x 10'7
Carbofuran 1.28 x 10'7 8.96 x 1 0 ’8 6.06 x 10*8
Parathion 6.82 x 1 0 ‘8 9.18 x 109 3.94 x 10'9
Propoxur 2.75 X 10'7 2.61 X 10'7 2.83 x 10'8 chamber. At 3, 6, 12 and 24 h three replicate beakers were removed and the midges wrapped in foil and placed in a freezer. Midges were removed and frozen at -20 °C until assayed. Enzyme assays were performed within 72 hours of dosing. Specific activities were determined for each compound and the values graphed against time. Mean specific activities at 3, 6, and 12 h for each compound were normalized by log transformation and subjected to analysis of variance
(ANOVA) to determine significant differences between mean activities over the three temperatures. Mean separation was achieved using Tukey's Studentized Mean test. Mean values were considered significantly different if p <
0.05.
Dosing Procedures for Pulsed Exposure Study
A single concentration of each insecticide was used in the pulsed exposure study. The concentration selected was one which was found to cause effect in approximately
20% of the midges after a 1 h exposure period (EC20) . This was determined by exposing groups of twenty midges in beakers of 500 mL of pH 7 soft standard reference water at
20 °c. Beakers were covered with aluminum foil and placed in an environmental chamber to allow midges overnight acclimation. After acclimation, 0.5 ml of a graduated series of different insecticide concentrations was added to each beaker where three replicates were made for each
concentration. Three replicate control beakers received
0.5 mL of acetone. After 1 h of exposure midges were counted as affected if they failed to execute three
figure-eight motions after pinching or prodding. This standard has been used in many studies as a measure of the well-being of Chironomus spp. (Fisher et al. 1993, Detra
1982). The toxicity data were analyzed by Finney's probit analysis (Finney 1971) and EC20 determined for each compound (aldicarb 26 ppb; carbaryl 40 ppb; carbofuran 2.5 ppb; parathion 55 ppb; propoxur 25 ppb) . The EC20 was selected as the exposure concentration since this would produce a low level effect after 1 h and would not cause
100% effect after 2 h continuous exposure for most of the compounds used in this study. A higher exposure concentration would minimize the possible difference between pulsed and continuous effect.
For the pulsed exposure study, groups of twenty midges were placed in beakers of soft standard reference water at pH 7.0 and 20 °C. Beakers were covered with aluminum foil and placed in an environmental chambers to allow the midges overnight acclimation. Each insecticide was prepared in a small volume of acetone. The effect used was the failure to execute three figure eight motions if pinched or prodded with forceps. Five replicates were used for all treatments and controls were in groups of five replicates. For each compound, four groups of five replicates were placed into treated water for one hour then transferred to clean water. Groups of five replicates were dosed for a second hour after a fixed recovery period. Recovery periods were 2, 6, 12 and 24 h.
After the second dosing, five replicate treatment and controls were removed from the environmental chamber and the number of affected control and treatment midges assessed. Studies were re-initiated if more than 20% of the control midges were affected. For each compound, one group of five replicates was dosed for a full 2 h then the number of affected midges determined to evaluate effects of continuous exposure. All effect data were adjusted for control effect using Abbott's equation (Amdur et al.
1991), and mean adjusted percent effect calculated for each group of five replicates.
In order to help confirm that enzymatic recovery did take place during the recovery period between exposures, a second study was conducted with carbaryl where AChE activity was measured after each exposure period and during recovery. Groups of twenty midges were placed in beakers filled with 500 mL pH 7.0 standard reference water at 20 °C. Three replicates were exposed for 1 h to a 76
single concentration of carbaryl. After transfer to clean
water groups, three replicates were removed at 0, 3, 6 and
12 hrs. Midges were collected and frozen at -20 °C for
analysis. At 24 h, the remaining beakers were again dosed
for 1 h, and the midges were then transferred to clean
water. Thereafter, three replicate beakers were removed
at 0, 3, 6, 12 and 24 hrs. Midge AChE activity was
measured and AChE activity graphed against time.
Pulsed exposure data were subjected to ANOVA after
being adjusted by arcsine square root to transform the
percentage data to normal distribution. Mean separation
was achieved by using Tukey's Studentized Mean test.
Means were considered significantly different if P < 0.05.
RESULTS
AChE Reactivation Data
For all temperatures studied, the AChE activity of
the midges exposed for 24 h to the four carbamate
compounds recovered to activity levels at or near control
levels within 24 h after being placed in undosed water
(Figs. 5-9). Recovery times ranged from a high of more
than 15 h for aldicarb at 10 °C to less than 3 h for
propoxur at 30 °C. However, the mean recovery times of
only three out of four compounds (aldicarb, carbofuran and Figure 5. Reactivation of midge AChE at 10, 20 and 30 following exposure to aldicarb. Data points are mean
^moles substrate hydrolyzed/min/mg protein (+ standard error). Solid lines are mean control activities at respective temperatures. RATE(nmolee/mln/mg protein) 0.0 2.0 1.0 1.0 3.0
6 2 8 4 30 24 18 12 6 0 i i a I i -i- A □ A 1 J 1 I I Figure 5. Figure 0 A TIME(h) B I I ' ■ 30 °C 30 ■ ' a 20 °C 20 a 10°CA 10 °C10 30 °C 30 20 °C 20 78
Figure 6. Reactivation of midge AChE at 10, 20 and 30 °C
following exposure to propoxur. Data points are mean
/moles substrate hydrolyzed/min/mg protein (± standard
error). Solid lines are mean control activities at respective temperatures.
79 RATE (nmolee/mln/mg protein) 12 18 18 12 Figure 6. Figure TIME(h) □ 20 °C 20 □ ■ 30 °C 30 ■ 0 °C 20 10 °C 10 30 °C 30 80 Figure 7. Reactivation of midge AChE at 10, 2 0 and 30 °C
following exposure to carbofuran. Data points are mean
/moles substrate hydrolyzed/min/mg protein (+ standard error). Solid lines are mean control activities at respective temperatures.
81 RATE (nmolea/mln/mg protein)
01 o« o
Ol - • ►
*1 TIME (h) H* M *
(D
00
to * I • O H
■ □ ► y to -4 O o~ o O o o O a o
-» to o< o o o o o o o 00 o o t\J Figure 8. Reactivation of midge AChE at 10, 20 and 30 °C following exposure to carbaryl. Data points are mean jumoles substrate hydrolyzed/min/mg protein (± standard error). Solid lines are mean control activities at respective temperatures.
83 RATE(nmoles/mln/mg protaln) ^ 6 . 0 2 18 12 Figure 8. Figure TIME(h) 3 °C 30 ■ °C 20 □ 0 °C 30 0 °C 20 10 °C °C 10 84 Figure 9. Reactivation of midge AChE at 10, 20 and 30 °c
following exposure to parathion. Data points are mean mmoles substrate hydrolyzed/min/mg protein (+ standard error). Solid line is the mean control activity.
85 RATE (nmolaa/mln/mg protein) 0.0 0.5 1.0 2.0 0 6 iue 9. Figure TIME(h) 12 18 24 □ 20 °C 20 □ 30 20 °C 20 30, °C 30, °C 10 °C 10 86 Table 9. Mean AChE activity (in nmoles substrate hydrolyzed/min/mg protein) for Chironomus riparius at 3, 6, and 12 h following exposure to one of five OP or carbamate insecticides at 10, 20, and 30 °C.
Time Acivity* (+ S.E.b) Q
Compound fhl 10 °C to o o 30 °C Aldicarb 3 0.263* 0.257* 1.469b (0.020) (0.007) (0.324) 6 0.283* 0.447* 2.423b (0.040) (0.027) (0.292) 12 0.266* 0.702b 2.884° (0.026) (0.075) (0.520) Control 0.355 0.666 2.330 (0.024) (0.152) (0.070) Carbaryl 3 0.496* 0.603* 0.739* (0.199) (0.037) (0.081) 6 0.647* 0.699* 0.902* (0.046) (0.068) (0.091) 12 0.712* 0.693* 0.955* (0.081) (0.081) (0.035) Control 0.730 0.697 0.992 (0.027) (0.054) (0.068)
Carbofuran 3 0.560* 0.618* 1.390b (0.074) (0.036) (0.155) 6 0.431* 0.594* 1.711b (0.051) (0.022) (0.037) 12 0.624* 0.694“ 1.9 9 lb (0.093) (0.008) (0.209) Control 0.523 0.726 2.730 (0.032) (0.025) (0.304) Table 9 (continued)
Parathion 3 0.556* 0.243b 0.315*b (0.111) (0.010) (0.038) 6 0.528* 0.258b 0.300*b (0.118) (0.002) (0.014) 12 0.480* 0.497* 0.407* (0.076) (0.094) (0.026) Control 1.283 0.575 0.703 (0.144) (0.005) (0.126)
Propoxur 3 0.323* 0.694b 1.802® (0.029) (0.046) (0.083) 6 0.470* 0.732* 1.326b (0.070) (0.147) (0.098) 12 0.524* 0. 626*b 1.189b (0.125) (0.024) (0.079) Control 0.569 0.617 1.676 (0.049) (0.011) (0.084)
•Recovery time values followed by the same letter are not significantly different based on Tukey's Studentized Mean test. bStandard error propoxur) were significantly affected by changing temperature (Table 9). The results showed that increasing temperature caused a increase in the mean specific activities for each compound after 3, 6, or 12 h. There was a significant difference between the mean activities after 12 h at all three temperatures for aldicarb (Table
9). Carbofuran and propoxur showed significant differences between mean activities for 20 and 30 °C as well as 10 and 30 °C but not between 10 and 20 °C. For midges exposed to aldicarb, AChE activity reached it's highest level at 12 h when exposed at 30 °C and 24 h when exposed at 10 or 20 °C (Fig. 5) . Midges dosed with propoxur at 30 °C, however, showed a somewhat erratic reactivation curve with the highest AChE activity at 3 h
followed by a rapid decline at 6 and 12 h then a sharp
increase at 24 h (Fig. 6). Peak AChE values for midges dosed with carbofuran occurred at 24 h for 20 and 30 °C but generated a near horizontal line at 10 °C
(Fig. 7). Mean AChE specific activities for midges dosed with carbaryl did not significantly change over the three temperatures examined (Table 9) .
Although the midges were dosed at concentrations which should have produced the same initial inhibition, midges dosed at 30 °C showed higher initial activity than midges dosed at either 10 or 20 °C, except with parathion. Table 10. Mean adjusted percent effect for one Organophosphorus and four carbamate insecticides at 2 0 °C to the midge, Chironomus riparius following two 1 h exposures separated by 0, 2, 6, 12, and 24 h in undosed water.
Percent of Midges Effecteda,b
Compound continuous 2-h 6-h 12-h 24-h 2-h
Aldicarb 81“ 84a 12b 27b 27b (7) (14) (5) (6) (7) Carbaryl 68a 56a 12b 14b 27b (7) (6) (5) (3) (5)
Carbofuran 58a 33b 10c 8C 10c (3) (4) (4) (4) (4)
Malathion 100a 100a 99a 99a 97a (1) (0) (1) (1) (0) Parathion 93a 90a 48b 46b 58a (3) (6) (10) (18) (15) Table 10 (continued)
Propoxur 59a 38b 10c 12c 12c (4) (5) (3) (4) (7)
“Values are mean of three replicates (numbers in parentheses are standard error) Values followed same letter are not significantly different based on Tukey's Studentized Mean test. 92
Initial AChE activity levels of midges dosed at 10 and 20
°C were nearly identical for all compounds (Figs. 5-9).
Midges exposed to parathion showed no significant reactivation over the 24 h recovery period except at 10 °C
(Fig. 9) . AChE activity at 10 °c was consistently at or above activity at either 20 or 3 0 °C.
Pulsed Exposure Data
Midges exposed for two i h periods with four carbamate compounds showed significantly fewer symptoms of intoxication than when exposed 2 h continuously if recovery in clean water was provided for 6 or more hours
(Table 10). Carbofuran and propoxur showed significantly less toxic effect than with continuous exposure if provided only 2 h in clean water between exposures.
However, two 1 h exposures to parathion or malathion were not significantly more toxic than a 2 h continuous exposure (Table 10).
AChE activity in the midges exposed to carbaryl showed a significant increase (Pr>F = 0.035, df = 2) after the midges were placed in clean water (Fig. 10). Recovery of the enzyme to control levels occurred within 6 hours after the first pulse. After the second pulse activity dropped to a level which was not significantly different from the activity after the first pulse (p < 0.027, df = Figure 10. AChE activity (± S.E.) in 20 fourth instar midges following two 1 h pulsed exposure to carbaryl (T =
20 °C, n = 3). Midges were pulsed at time 0 and 24 h
(exposure time not included on graph). Solid line is mean control activity.
93 Pulse 1 t Pulse 2
----- 1---- H- _t------1------1------— f-----1------1----- 0 6 12 18 24 30 36 42 48 54 TIME (h)
Figure 10. 95
2). Recovery after the second pulse also occurred within
6 hours (Fig. 10).
DISCUSSION
Recovery of midges from exposure to OP and carbamate insecticides was measured by two methods. First, reactivation of midge AChE was measured directly following exposure to a particular toxicant (Figs. 5-9). This made it possible to determine the amount of time necessary before the toxicant ceased to act on the target site.
Since the amount of the toxicant at the target site and the interaction between toxicant and target site is the closest measurement of the actual toxicity of a compound
(McCarty 1989), measuring AChE reactivation (disappearance of the toxin from the target site) provides a good estimate of how much time is necessary after exposure before recovery will take place. Second, symptomatic recovery of the midges was measured after first being exposed to a toxin, and then being placed in clean water.
This was performed to determine if: 1) symptomatic recovery from exposure occurred and 2) if reactivation of
AChE activity and symptomatic recovery from intoxication coincided. 96
The effect of temperature on AChE reactivation
The results showed that midge AChE inhibited by carbamate insecticides reactivated within 24 h even at temperatures as low as 10 °C. This rapid rate of AChE reactivation agrees with the findings of Reiner (1971), who determined that 50% of AChE inhibited by a variety of carbamates recovered to control levels 2 to 240 min after exposure. However, the results of this study did not clearly show that temperature affected reactivation rate.
Aldicarb, propoxur, and carbofuran showed an increase in
AChE activity with an increase in temperature. This is in agreement with the conclusions of Green et al. (1988) who found that the freshwater isopod Asellus aquaticus could recover from brief periods of exposure to phenol if placed in clean water. Rate of recovery was found to be dependant on temperature during exposure and recovery periods as well as exposure time, and exposure concentration. They found that an increase in temperature resulted in a concurrent increase in recovery rate. These results also agree with the findings of Honkakoski et al.
(1988). They determined that reactivation of rat and mouse blood and lung AChE inhibited with di
isoprop lyphospho-fluordate (DFP) was significantly affected by temperature, where increasing temperature caused an increase in reactivation rate. 97
However, activity for AChE inhibited with carbary1 did not appear to be significantly affected by temperature. This was unexpected since AChE inhibited by the other three carbamates tested did demonstrate a change in AChE activity as a function of temperature. Webb
(1963) asserted that the hydrolysis of a phosphorylated or carbamylated enzyme is dependant on temperature. However, he stated that the reversal of inhibition would require an energy of activation (14.4 kcal/mole for phosphorylated
AChE). A rise in temperature might therefore increase the rate of reactivation if the amount of free energy exceeded the energy of activation, but below this level there could be an increase in enzyme inhibition rather than reactivation (Webb 1963). It is therefore possible that
AChE inhibited by carbaryl needed a greater increase in temperature if the reactivation rate were to be substantially altered.
AChE inhibited by parathion showed no clear reactivation at 20 or 30 °C, even after 24 h in clean water. This was not surprising since OP compounds have been characterized as irreversible inhibitors of AChE.
However, some studies have reported that spontaneous dephosphorylation of AChE does occur. Recovery of AChE activity following exposure to fenitrothion in the freshwater field crab, Oziotelphusa senex was reported by Bhagyalakshmi and Ramamurthi (1980). They determined that
there was a significant increase in AChE activity within
72 h after exposed crabs were placed in clean water.
Similar results were found by Bhagyalakshmi et al. (1984), where AChE in the thoracic ganglion of the field crab recovered to control levels within 10 days after exposure to 1 to 2 ppm sumithion. They cited dephosphorylation, resynthesis of new enzyme, biodegradation and rapid excretion of the toxin as reasons for recovery. Recovery of AChE inhibited by OP compounds has also been reported
in fish (Coppage et al. 1975, Morgan et al. 1990), the housefly, Musca domestica (Ahmad 1970) and the penaeid prawn, Metapenaeus monoceros (Reddy et al. 1987). The discrepancies can be explained by differences in recovery times. All the above studies found recovery after 72 h to
15 days. Midges used in this study were only allowed to recover for 24 h, which was evidently too short to see any reactivation from parathion inhibition if it were to occur. Also, the time required to recover will be dependant on the initial inhibition level, where a longer recovery interval will be necessary for organisms that have had a greater reduction in AChE (Weiss 1961, Morgan et al. 1990). Although AChE activity at 24 h at 10 °C was higher than the previous activity, it is impossible to tell if this was actual recovery or if the data point were 99 just an outlier. Certainly, no recovery in AChE activity occurred at either 20 or 30 °C.
The results of this study support the limitations reported by Greig-Smith (1991) concerning the monitoring of AChE levels in wildlife. Since AChE inhibited by carbamates reactivates rapidly following intoxication, an animal exposed to a carbamate would show normal AChE activity shortly after exposure, leading anyone monitoring
AChE to the erroneous conclusion that no exposure took place. Greig-Smith (1991) correlated cholinesterase measurements obtained from 13 years of investigation into suspected wildlife poisoning incidents in England and
Wales with the presence and amount of pesticide residues in animals7 gut contents. It was found that there was a close correlation between AChE activity and OP residues but no relationship between AChE and carbamate residues.
The conclusion reached was that the animals tested had been dead long enough for AChE inhibited by carbamates to have spontaneously reactivated. Therefore animals suspected of being exposed to carbamates should be collected live for analysis. Our results indicate that these animals should also be stored at low temperatures until assayed to slow reactivation of inhibited AChE. 100
Toxicity of pulsed and continuous exposure to insecticides
The results clearly showed that the toxicity of
carbamate compounds for two 1 h pulses is significantly
lower than 2 h continuous exposure if 6 h in clean water
is provided between pulses. For exposure to aldicarb and carbaryl, less than 6 h in clean water between exposures produced no significant differences between the effect caused by the two 1 h pulses and to the 2 h continuous exposure. The need for 6 h in clean water corresponds with the reactivation curves produced for these compounds.
After 6 h in clean water, the AChE in the midges would have recovered to somewhere near control levels. A second exposure would have produced an inhibition only slightly above the initial inhibition level.
The lower toxicity of the two 1 h pulses agrees with
Mancini (1983) as well as Wang and Hanson (1985) who concluded that episodic exposure to pollutants is less toxic if recovery in clean water is allowed between exposures. Since the effects of the insecticides in this study were not determined based on the number of midges surviving to adult, a direct comparison between our results and those of Parsons and Surgeoner (1991a) is not possible. They found that two 1 h exposures to carbaryl or carbofuran were equally toxic as 2 h continuous exposure to mosquito larvae. They concluded that either 24 h recovery between exposures (the maximum time allowed
in clean water before the second pulse) was too little time for recovery of the mosquito AChE or that damage had been done other than inhibition of AChE. Based on our results, it is likely that the AChE activity in the mosquito larvae did recover within the 24 h period in clean water. Our results showed that no cumulative increase in AChE inhibition occurred in midges dosed with carbaryl when two 1 h pulses were separated by 24 h in clean water (Fig. 10). The last data point did show a significant drop in AChE activity in comparison to the control level; however this was likely due to a drop in total amount of cholinesterase due to death from stress and starvation (many controls were also affected by this point). Although some studies have concluded that OP and carbamate insecticides have target sites other than AChE
(see Marquis 1986), it is impossible to know based on the results of our study to determine if the OP's or carbamates acted on anything other than AChE.
The results of our study also showed that the OP compounds parathion and malathion were equally toxic when exposed either for 1 h pulses or for 2 h continuously.
This is in agreement with the lack of recovery found in
AChE activity for parathion at 20 °C following exposure.
Since no recovery in AChE activity took place after the 102
first 1 h pulse, during the period while the midges were
in clean water, the AChE activity would have been further
depressed after the second pulse. Thus the toxic effect
of the two pulses was not significantly different than the
effect of continuous exposure on the midge. This is also
in agreement with the results of Parsons and Surgeoner
(1991b) who found no symptomatic recovery of mosquito
larvae exposed to fenitrothion.
The fact that enzymatic recovery leads to decreased toxicity for organisms receiving pulsed exposure to an
insecticide is likely to be of importance in the devlopment of QSAR. Currently there have been no attempts to model episodic exposure, with all predictors being developed for continuous exposure data. Development of
QSAR for episodic exposure to insecticides, and other
compounds as well, would allow regulating authorities to recommend effluent release rates for industry or field application rates for farmers which would minimize the
impact of these compounds on indigenous biota.
Overall, this study shows that midge AChE recovers
from exposure to carbamate insecticides within 24 h.
Temperature during the recovery period can have a marked affect on the rate of recovery but this may be compound
specific. Midge AChE inhibited by the OP compound parathion does not recover within 24 h even at temperatures as high as 30 °C. The ability of the midges to recover from exposure to carbamates makes episodic exposure to these compounds less toxic to the midge than continuous exposure, so long as the midges are allowed to recover in clean water between doses. Because temperature affects the rate of recovery, and recovery is of key importance in determining the effects of episodic exposure, it is probable that temperature during recovery also affects the toxicity of episodic exposure of aquatic organisms to carbamate insecticides. However, since temperature has been found to affect the toxicity of OP's and carbamates to midges during continuous exposure studies (Lydy et al. 1990a), this might mitigate the effects of temperature on the toxicity of these compounds during episodic exposure. For example, lower temperatures results in slower recovery, but a carbamate insecticide would be less toxic at lower temperatures as well. CHAPTER III
QUANTITATIVE STRUCTURE ACTIVITY RELATIONSHIPS TO PREDICT
THE PERSISTENCE OF ORGANOPHOSPHORUS AND CARBAMATE
INSECTICIDES IN AQUATIC SYSTEMS WITH AND WITHOUT
THE MIDGE, CHIRONOMUS RIPARIUS
INTRODUCTION
With the environmental problems caused by the use of organochlorine insecticides, emphasis has shifted to the use of carbamate and organophosphorus (OP) compounds which tend to be less persistent in aquatic systems (Aly and El-
Dib 1972) . However, many researchers have reported incidents in which OP and carbamate compounds have been found in surface and ground water samples, challenging the belief that they are ephemeral in the environment (Liu et al. 1981, Rothschild et al. 1982, Zaki et al. 1982,
Crathorne et al. 1984, Marshall 1985). A contributing factor to the appearance of these insecticides in aquatic systems is their enhanced stability under certain conditions. Temperature and pH were reported by McEwen and Stephenson (1979) to be the most important
104 105
environmental factors in determining the aquatic fate of
OP's and carbamates. Generally, lower water temperature
(Aly and El-Dib 1971, Badawy and El-Dib 1984) and pH
(Chapman and Cole 1982, Given and Dierberg 1985, Lohner
and Fisher 1990, Lydy et al. 1990a) tend to increase the
stability and persistence of OP and carbamate
insecticides.
Although environmental parameters such as pH and
temperature affect the persistence of insecticides in
aquatic systems, according to Matsumura (1985) degradation
by living organisms such as microorganisms, animals, plants is fundamental to determining the amount of
insecticides in the environment. For example, Lohner and
Fisher (1990) found that the midge, Chironomus riparius was able to metabolize 12% of the available carbaryl over
a 24 h exposure period at 20 °C. They found that the rate
of metabolism was temperature dependent, where increasing
temperature resulted in an increase in the amount of
carbaryl metabolized. Further, Lydy et al. (1990a) found
that the midge was able to metabolize 69% of parathion
over 24 h at 20 °C. They also found that an increase in
temperature tended to result in an increase in the amount
of parathion metabolized.
Although there is a need for determining the
persistence of xenobiotics in the environment, few attempts have been made to predict degradation rates in aquatic systems through the use of QSAR. Drossman et al.
(1988) found that the base-promoted hydrolysis of carbamates could be modeled using Hammett and Taft parameters a , a*, and Es. They found that correlation with carbamate hydrolysis rates and the Hammett and Taft parameters gave adjusted r2 values as high as 0.97.
Models based on linear free energy relationships have been developed to predict the hydrolysis rate constants for carbamate (Wolfe et al. 1978) and OP (Wolfe 1980) insecticides in water. Correlations between the linear free energy relationships parameters and OP hydrolysis rate constants gave adjusted r2 values as high as 0.99.
The purpose of this study was to examine the ability of three physical and structural parameters, Kow/
Molecular volume (MV) and Henry's Law constant (HLC) and two multidimensional models MC indices and LSER, to explain the variation, through regression analysis, in the halflives of three OP and two carbamate compounds in water at varying temperatures. Halflives were determined for each compound in water with and without the midge,
Chironomus riparius, and were regressed against the physical parameters selected to determine their predictive value. Finally, halflives were regressed against published toxicity data to determine if halflives related 107
to the toxicity of the insecticides.
MATERIALS AND METHODS
Test organisms
The midge, Chironomus riparius, used throughout this
study in tests which required midges, was chosen because
it is easy to culture, has a cosmopolitan distribution,
and is important in aquatic food chains. Midge cultures were maintained at room temperature (23 ± 3 °C) according
to the methods of Estenik and Collins (1978). Fourth
instar larvae were used in all tests.
Compounds
For measurement of hydrolysis halflives, three OP
(malathion, methyl parathion, and parathion) and two
carbamate compounds (aldicarb, carbaryl) were studied.
The same five compounds were used in the metabolism study.
All compounds were of reagent grade ( aldicarb, 79%,
specific activity = 1.05 x 108 dpm/mg; carbaryl, 99%, 9.39
x 107 dpm/mg; malathion, 99%, 4.17 x 107 dpm/mg; parathion,
99%, 7.63 x 107 dpm/mg; methyl parathion, 97%, 5.99 x 107
dpm/mg) and were purchased from Sigma Chemical Co.
Physical and Structural Parameters
Molecular connectivity and LSER values were obtained
from Fisher et al. (1993) or were calculated using the MOLCON computer program. Values for LSER parameters
(Vj/100, 7r*, a, and j8) used in this study were taken either
from Fisher et al. (1993) or were calculated by methods
described in Hickey and Passino-Reader (1991). The
intrinsic molecular volume, Vj/100 (adjusted by a factor
of 100 to make it comparable in size to the other three
parameters), represents the energy required to open a
cavity in a solvent. IT, represents dipole interactions
between solute and solvent. It is a measure of a molecule's ability to stabilize a nearby charge or dipole.
Alpha and beta, both hydrogen bonding terms, represent the
ability of a molecule to donate a hydrogen, or accept a hydrogen, respectively.
The levels of MC indices used in this study were °x,
lXt 2Xr an(* 3x* °X describes the volume of a molecule. *x
is the sum of non-hydrogen atoms adjacent to each other
connected by a single bond. 2x, and 3x represent the sum
of nonhydrogen atoms connected by 2 and 3 bonds respectively, x parameters were also included which
account for multiple bonds and the presence of ionized groups, designated °xv “ 3XV* The cluster type subclass, both valence and non-valence, for the third level of connectivity, designated 3cx and 3cxv/ was also used. MC
indices were developed to describe relationships within molecules and how this relates to the more complex three 109
dimensional aspects of molecular structure. K,^ and HLC
values were taken from Suntio et al. (1988) and MV values
from Hansch and Leo (1979).
Measurement of abiotic halflives in water
Hydrolysis halflives for 14C-radioactive samples of
the three OP and two carbamate were measured as a function
of temperature. A total of three temperatures were used:
10, 20 and 30 °C. Two liter beakers containing
approximately 1.2 L of deionized water buffered to pH 7.0
(using potassium phosphate and sodium hydroxide) was
autoclaved for 25 min at 120 °C in an American Sterilizer
Co. autoclave. 14C-labeled insecticides dissolved in
acetone (0.5 mL) were added to the sterilized water, and
stirred vigorously. Initial water samples were taken to determine average l4C-insecticide concentration (Table 11) .
The dosed water was divided into aliquots of 100 mL and poured into brown glass bottles. Each bottle was covered with sterilized aluminum foil, capped, and placed in an
environmental chamber at the appropriate temperature.
After varying time intervals, three replicate samples were removed and extracted with the appropriate solvent system
(aldicarb, methylene chloride; carbaryl, chloroform/methanol, 49:1; malathion, benzene/acetic acid,
4:1; methyl parathion and parathion, ether/hexane, 4:1) 110
Table 11. Dose levels of 3 OP and 2 carbamate compounds and sampling dates for determination of abiotic halflives at 10, 20 and 30 °C.
Temperature Dose sampling dates Compound f°C> fmq/L) (dj______
Aldicarb 10 14, 151, 199, 294 20 0.011 14, 44, 150, 294 30 21, 35, 45, 62
Carbaryl 10 14, 44, ill, 159 20 0.013 1, 4, 7, 10 30 1, 4, 7, 10
Malathion 10 14, 44, ill, 159 20 0.029 1, 4, 7, 10 30 1, 4, 7, 10
Methyl 10 14, 151, 199, 294 Parathion 20 0.020 14, 43, 158, 199 30 21, 35, 45, 62
Parathion 10 14, 151, 199, 294 20 0.016 14, 43, 158, 191 30 21, 35, 45, 62 Ill
which was determined to extract >80% of the parent
compound from the water. The water was extracted
according to the procedures of Fisher (1985). In a 500 mL
separatory funnel, 100 mL of water were extracted three
times with 50 mL aliquots of the appropriate solvent.
Extracts were combined and dried over anhydrous MgSo4 and
filtered through Whatman #1 filter paper. The MgS04 was
washed three times with approximately 10 mL aliquots of
solvent; washes and filtrates were combined and the volume
measured. One mL samples of each extract were removed for
scintillation counting and the remaining extract
evaporated to a small volume for thin layer chromatography
(TLC).
Fifty ul of each extract and standards were applied
to silica gel TLC plates. The plates were developed in
the appropriate solvent system and the spots visualized
under 254 nm UV light parent compounds and metabolites.
The plates were cut into sections and each section placed
in a scintillation vial containing 5 mL 14C-cocktail. The
scintillation vials were counted after 24 h in a Beckman
LS6000IC scintillation counter (5 min/vial), and the data
were used to determine total amount of parent compound and metabolites. The amount of parent compound and metabolites was adjusted for the initial impurity of each
compound. 112
Hydrolysis halflives were calculated as a function of temperature for each compound by plotting log percent parent compound against time. Halflives were determined from the slope of the line using equation (3):
(3) T]/2 = In 2/slope
Significant differences between halflives as function of temperature were determined by analysis of variance
(ANOVA). Mean separation was achieved using Tukey's
Studentized Mean test.
Persistence in water with midges
The persistence of insecticides in water with midges, at varying temperatures, was measured for the same 2 carbamate and 3 OP compounds. Groups of 40 fourth instar midges were exposed for a 32 h period to a given insecticide in 1 1 beakers filled with 500 ml soft standard reference water adjusted to pH 7.0. All beakers were covered with aluminum foil to prevent excessive evaporation of water and were placed in a Forma Scientific
(#3740) environmental chamber at the appropriate temperature. Midges were acclimated for at least 12 h prior to dosing, and a photoperiod of 14/10 h was maintained through all experiments. Each beaker then received 0.5 mL of a given concentration of insecticide
dissolved in acetone; controls received 0.5 mL acetone
only. All treatments and controls were prepared in groups
of three replicates. A sublethal concentration was used
(20% the reported 24 h EC50 at 20 °C: aldicarb, 1.98 /xg/L;
carbaryl, 22 /xg/L; malathion 0.41 /xg/L; methyl parathion,
0.32 /xg/L; parathion, 1.38 /xg/L) for each compound at each
temperature. The dose level of one-fifth the EC50
concentration was selected since this was a sublethal
exposure level and still provided enough compound in the water for accurate measurements of parent compound and metabolite concentrations. Three different exposure
temperatures were used: 10, 20, and 3 0 °C. After dosing,
the beakers were returned to the environmental chamber and
at 4, 8, 24 and 32 h after exposure, three replicate
beakers were removed, and emptied through a fine mesh
sieve and the midges collected and wrapped in aluminum
foil for freezing. All midges were kept frozen at 0 °C
until assayed.
The water extractions were performed as described for
the abiotic halflife measurements, except that 500 mL of water were extracted with three aliquots of 100 mL
solvent. Midges were extracted by homogenizing groups of
40 midges in 3 mL acetonitrile with a Potter-Elvehjem
tissue homogenizer. Homogenates were centrifuged in 15 mL conical centrifuge tubes at high speed in a Fisher
clinical centrifuge for 3 min. The supernatant was
decanted, collected, and the pellet resuspended in 3 mL
acetonitrile. The homogenate was again centrifuged and
the procedure repeated for two more times. Combined
supernatants were dried over anhydrous MgS04 then
filtered. The MgS04 was rinsed 3 times with acetonitrile,
and the rinses combined with the filtered extracts. The
combined volume was recorded and a 1 mL sample taken from
each extract for scintillation counting. The remaining
extract was evaporated to a small volume and reconstituted to 2 mL with acetone. Total amount of parent compound and metabolites were determined for each extract by TLC.
Midge pellets were dried, weighed and placed in
scintillation vials with 0.5 mL tissue solubilizer. After
24 h, each sample was neutralized using 0.5 mL acetic acid
and 5 mL scintillation cocktail added for scintillation
counting. Total amount of parent compound and metabolites was determined for each replicate at each time interval.
The amount of parent compound and metabolites was adjusted
for the initial impurity of the compound. For carbaryl and malathion, disappearance of parent compound due to hydrolysis was subtracted from each replicate at each time
interval. The amount of hydrolysis at each time interval was extrapolated from the hydrolysis regression lines 115
previously determined. No adjustment was made for the
remaining three compounds since the amount of hydrolysis
over 32 h was negligible. The data were graphed as log
percent parent compound against time and the halflife for
each compound was determined using equation (1).
Significant differences between halflives, as a function
of temperature, were determined by ANOVA, where mean
separation was performed using Tukey's Studentized Mean
test.
Data Analysis
Abiotic hydrolysis and metabolic halflife data were
analyzed by ANOVA to determine significant differences between treatment means. Mean separation was achieved by using Tukey's Studentized Mean test. Halflife data for all three temperatures were regressed against each of the molecular descriptors (SAS 1982). For multidimensional descriptors (MC indices and LSER) halflife values were regressed against all possible combinations of up to three descriptors. For the multidimensional descriptors only the regressions which gave the highest correlation coefficients were reported. Regressions were considered significant if P values were 0.05 or below. Finally, abiotic and metabolic halflives were regressed against the 116
Table 12. Mean abiotic halflives of Organophosphorous and Carbamate Insecticides in sterilized deionized water at 10, 20 and 30 °C (± S.E.) .
Half life8 (d) Compound 10 °c 20 °c 30 °c Aldicarb 4 2 2 . 58 3 07 . 3" 70.5" (210.6) (48.7) (9.8)
Carbary 1 23.9" 9.5" 11.1" (7.6) (0.3) (0.9)
Malathion 80.9" 7.7b 7.2b (9.0) (0.2) (0.9)
Methyl Parathion 340.0" 123 .6b 40.5= (8.0) (4.7) (2.3)
Parathion 797.4" 239.9b 70.2° (44.6) (2.8) (1.7)
"Halflife values for a given compound followed by the same letter are not significantly different based on Tukey's Studentized Mean test. 117
24 h EC50's from Fisher et al. (1993) for all compounds used to determine if hydrolysis and metabolism correlates with toxicity.
RESULTS
Abiotic hydrolysis data
The abiotic halflives of the 5 insecticides tested were significantly affected by temperature (Table 12). In general, an increase in temperature resulted in a decrease in halflife, indicating that hydrolysis rate increases with increasing temperature. There was, however, significant variation seen among compounds with respect to the influence of temperature on halflife.
The aquatic halflives of aldicarb and carbaryl were not significantly affected by temperature over the range tested. There was also no significant change in the halflife of malathion between 20 and 30 °C, although there was a significant change between 10 and 20 °C (Table 12) .
The halflife values for the remaining two compounds decreased significantly between 10 and 20 as well as 20 and 30 °c.
Biological halflife data
The biological halflives of the five insecticides were considerably lower than in water only tests (Table 118
Table 13. Mean halflives of Organophosphorous and Carbamate Insecticides in water with 40 midges (Chironomus riparius) at 10, 20 and 30 °C (+ S.E.).
Half life* (d) Compound 10 °C 20 °C 30 °C
Aldicarb 1.1“ 2.3a 1.2* (0.1) (1.0) (0.4)
Carbary 1 3.1“ 3.4“ 1.4“ (0.9) (0.8) (0.2)
Malathion 24.3* 4.4“ 3.0“ (0.9) (1.3) (0.2)
Methyl Parathion 3.1* 1.8“ 2.8* (0.9) • (4.0) (1.0)
Parathion 2.4* 2.6“ 1.3* (0.7) (0.3) (0.5)
“Halflife values for a given compound followed by the same letter are not significantly different based on Tukey's Studentized Mean test. 119
13). However, these halflives were not significantly
affected by temperature (Table 13). This indicates that the ability of the midges to degrade the insecticides tested was not significantly altered by changing temperature.
Regression Analyses with MV, HLC and Log Kow
The unidimensional molecular descriptors used in this
study are given in Table 3. Regression analysis was performed between these values and abiotic and biological halflife data at each temperature to determine if a linear relationship exists between these variables.
The results of the regression analysis showed no
linear relationship between any of the one dimensional descriptors and the abiotic halflife data. However, when
the halflives of the insecticides in water with midges at
20 °C were regressed against MV, a significant linear
relationship was evident (r2 = 0.78; p < 0.04; df = 4; F =
10.93; t1/2 = 0.475(MV) - 48.4). However, there was no
linear relationship between MV and halflives at either 10
or 3 0 °C, and no linear relationship between K^, or HLC and
any of the halflives with midges. No consistent
temperature effects were observed for any of the three models. 120
Table 14. Summary of Best Regression Analyses of hydrolysis and biological halflife values for LSER, and MC Models.
Halflife Treatment Descriptor Variables r2
Hydrolysis 10 °C LSER NSR" MC 3cx 0.93
20 °C LSER NSR MC 3cx, V 0.95
30 °C LSER NSR MC 3cx, V 0.95
Biological 10 °C LSER n* 0.94 MC 3xv 0.83
20 °C LSER NSR MC °x, V 0.99
30 °C LSER Vj/100, a 0.99 MC NSR
“No significant relationships 121
Regression Analysis with LSER and MC Indices
Physical and structural parameters for LSER and MC indices (Tables 4 and 5) yielded some significant correlations with the halflife data. When the LSER parameter n* was regressed against halflives with midges, values at 10 °C, a linear relationship was evident (r2 =
0.94; p < 0.007; F = 45.05; df = 4; t1/2 = 1095.9 (7T*) -
1244.6) (Table 14). At 30 °C, there was a linear relationship between LSER parameters VJ 100, a, and halflives with midges (r2 = 0.97; p < 0.03; df = 4; F =
38.09; t,/2 = -179.4(Vj/100) - 275.6(a) + 328). However, there was no significant relationship between any of the
LSER parameters and abiotic halflives, or biological halflives at 20 °C.
Regressions of biological and abiotic halflives with
MC indices produced the best results overall (Table 14).
For regressions against abiotic halflives at 10 °C, the best relationship was produced using 3cx (r2 = 0.99; p <
0.003; df = 4; F = 284.04; l/t1/2 = -0.045(3cx) + 0.075)
(Table 14) . The best relationship for 20 °C was found when 3cx and 3xv were used in combination (r2 = 0.95; p <
0.05; df = 4; F = 17.44; l/t,/2 = -0.20(3cx) + 0.03 (V) +
0.20) . However, at 30 °C 3cx and 2xv gave the best correlation (r2 = 0.95; p < 0.05; df = 4; F = 17.94; l/t1/2
= -0.17(3cx ) + 0.02(3x v) + 0.15). 122
Regression of 3xv against biological halflives produced a linear relationship at 10 °C (r2 = 0.83; df = 4; ti/2)05 = 3.05(3x v) - 0.79) (Table 14). At 20 °C regression of °x and 2xv gave the best results (r2 = 0.99; df = 4; t1/2)05 = 2.78 (2xv) “ 1.88 (°x) + 24.45). However there was no significant relationship between any of the MC parameters and biological halflife data at 30 °C.
Regression of halflives with toxicity
There was no linear relationship between biological
(r2 = 0.01; p < 0.83; df = 4; F = 0.06) halflives measured in this study and ECS0 values of these 5 insecticides to the midge, Chironomus riparius (from Fisher et al. 1993).
There was also no relationship between the abiotic half lives measured and EC50 values (r2 = 0.23; p < 0.41; df
= 4; F = 0.92).
DISCUSSION
The persistence of the insecticides were measured under two different situations. Halflives were measured in containers of water alone and in beakers of water with midges. When the insecticides were placed in water without any living organisms, the compounds were degraded by simple hydrolysis (Chapman and Cole 1982) . However, when midges were added, the insecticides would have 123 entered the midge by such mechanisms as cuticular absorption or possibly ingestion of contaminated water.
The insecticides would then have been transported to various tissues within the midge and there likely biotransformed to metabolites structurally different from the parent compound. For example, the OP's would have likely undergone a desulfuration reaction catalyzed by the midges mixed function oxidase system (MFO) to the more toxic oxon metabolites (Davison 1955, Neil 1967,
Nakatsugawa and Dahm 1967). Aldicarb likely underwent a sulfoxidation reaction to the more toxic sulfoxide and less toxic sulfone metabolites (Andrawes et al. 1967,
Oonithan and Casida 1966, Shrivastava et al. 1971). The metabolites produced by the midge would either be excreted, react with a target site (Acetylcholinesterase) or might be redistributed to other tissues and again undergo biotransformation, reaction with target sites, or excretion. The factors which would be most important in determining the halflives of the insecticides should be the steric and electronic characteristics of the toxicant molecules and how these affect the hydrolysis of the insecticides or interaction with the various degradative enzymes in the midge. Any model designed to predict the halflife of insecticides in systems with or without living organisms must be able to account for both the 124 susceptibility to abiotic hydrolysis by water and degradation by MFO's, hydrolases, etc. within an organism.
The abiotic halflives in water obtained in this study were, in general, close to the values determined in other studies. For example, the abiotic halflife of methyl parathion at 20 °C was 123.6 d (Table 12) which was close to the value of 13 6 d reported in Badaway and El-Dib
(1984). The halflife of carbaryl was determined to be 9.5 d (Table 12) which also closely matched the 10.5 d reported in Aly and El-Dib (1971). However, when temperature was changed, there was a subsequent change in abiotic halflife.
In general, the abiotic halflives of the insecticides used decreased with increasing temperature (Table 12).
This is in agreement with Badaway and El-Dib (1984) who found that the rate of methyl parathion hydrolysis nearly doubled with each 10 °C rise in temperature. Mabey and
Mill (1978) also reported that increasing temperature should cause a decrease in abiotic halflife of many organic compounds in water. However, the halflives of aldicarb and carbaryl were not significantly affected by temperature (Table 12). There was wide variation in the halflives of aldicarb at each temperature, especially at
10 °C which would account for the lack of significance between halflives even though there were large differences 125 between the mean values. There was also wide variation in the abiotic halflife of carbaryl at 10 °C.
The results showed that midges decreased the halflives of all five compounds substantially (Table 13).
This is in agreement with the results of Lydy et al.
(1990a) who found that 59% of parathion was metabolized by midges over a 24 h period. This is also in agreement with
Lohner and Fisher (1990) who found that 12% of carbaryl was metabolized in 24 h. However, temperature did not significantly affect halflives in water with midges for any of the five compounds tested. This agrees with the results of Lohner and Fisher (1990) where carbaryl degradation by midges did not significantly change between
10 and 30 °C. However, the results of this study do not agree with the results of Lydy et al. (1990a), where the amount of parathion in the water after 24 h significantly decreased when water temperature was increased from 10 to
20 °C. These results also disagree with the conclusions of Cairns et al. (1975) who concluded that increasing temperature would increase detoxification of xenobiotics.
One possible reason for the discrepancy between the our results and those of Lydy et al. (1990a) is that 40 midges were used per replicate in this study while 20 midges were used in Lydy et a. (1990a). The higher number of midges and extremes in temperature could have resulted in 126
decreased levels of 02 and/or elevated levels of stress
which might have decreased the midges' ability to
metabolize the insecticides, although it is impossible to
tell without more information.
Regressions with MV, HLC, and Log
None of the unidimensional descriptors was able to
describe the variation in the abiotic haIflives of the
pesticides tested. Only MV at 20 °C could describe the
variation in halflife data when midges were present in the
water. Measurement of metabolites for halflives with and without midges was done by mass balancing, such that
compound which left the system through evaporation or
sorption to glassware was not measured. Thus it is
plausible that HLC would not correlate with halflives
since HLC is a measure of volatility, and any compound
volatilizing from the system was not included in total metabolite analysis.
Log Kow has been used to measure the fate of many
neutral, lipophilic compounds. The sorption of
xenobiotics to sediment was accurately described by Kow
(Adams et al. 1985). The uptake (kj and elimination (k2) rate constants of 12 OP compounds in guppies was also
found to correlate closely with (de Bruijn and Hermens
1991) . However, in both of the above studies Kow was probably describing the lipophilic partitioning of a compound between different media. Bonding of a lipophilic compound to sediment as well as passage of an OP into a guppy are likely to involve different processes from simple abiotic hydrolysis where water is the only medium present. However, even when midges were present, was not able to describe the variation in halflife data.
Lipophilicity was probably important in the penetration of the insecticides into the midge, but once there the insecticides would have been subjected to numerous alterations in chemical structure, the results of which were too complex for K^, to predict. Further, most of the insecticides studied by de Bruijn and Hermens (1991) were reportedly not metabolized during the test period. The uptake and elimination constants for those compounds which were metabolized were not predictable by Kow. Unlike Kow,
MV was able to describe most of the variation for halflives measured when midges were present, albeit at only one temperature (20 °C) . MV has been most successfully used in correlations with the toxicity of neutral, organic compounds whose mode of action is narcosis (McGowan and Mellors 1986a, McGowan and Mellors
1986b). It is likely that the volume of a molecule is important in determining the rate of penetration, and this may be in part what molecular volume is describing in this case. However, the results of Fisher et al. (1993) did not find any correlation between MV and the toxicity of these same five compounds to the midge. They believed that MV was not able to describe the variety of mechanisms involved in the intoxication process. Since the metabolism of an insecticide also involves many different processes, it seems strange that the insecticide halflives could be predicted by MV but not toxicity. The reason for the discrepancy probably lies with the different endpoints
MV was attempting to measure. The toxicity of OP and carbamate insecticides is the result of inhibition of acetylcholinesterase (AChE) (Casida 1964). Here, MV would have to be able to describe not only the penetration of the pesticide into the midge, but also the three dimensional fit of the insecticide into the receptor site on the AChE molecule (Fisher et al. 1993). With metabolism, the fit of the molecule into a receptor site is probably less important. It has long been known that
MFO systems are generalists in nature, able to handle a wide variety of substrates (Wipke et al. 1983, Low and
Castagnoli 1980). Therefore, MV would likely not have describe the fit of a toxin molecule into a receptor site since a wide variety of sites are available for the metabolism of xenobiotics. 129
Regressions with multidimensional descriptors
LSER uses the energy requirements to surround a
solute molecule with a solvent molecule and the energy
necessary to maintain the solute molecule within the
solvent to relate the properties of a compound to
molecular structure (Hickey and Passino-Reader 1991).
Therefore, properties which depend on solute-solvent
interactions should be predictable using LSER energy
terms. LSER has been used successfully to predict aqueous
solubility (Kamlet et al. 1986, Kamlet et al. 1987), K^
values (Kamlet et al. 1988), as well as toxicity to a
variety of organisms (Kamlet et al. 1986, Kamlet et al.
1987) .
The fact that LSER was unable to describe any of the
variation in the abiotic halflife data leads us to
conclude that hydrolysis of these compounds is not related
to their solubility in water or partitioning behavior
since LSER attempts to describe biological phenomena based
on these characteristics. However, when midges were added
to the water, LSER was able to explain a high percentage
of the variation in halflife data. At 10 °C n* explained
92% of the variation in the data, while V;/100 and a
explained 99% of the variation at 30 °C. However, LSER was not able to explain any of the variation in the data
at 20 °C. It is difficult to discern why there is a strong relationship between tt* and half lives at 10 °C. it* is a measure of a compound's ability to stabilize a neighboring charge or dipole (Kamlet et al. 1986). It is possible that tt* is describing the attachment of oxygen to the sulphur group on an OP, by cytochrome P-450, to form the dipole sulfine, the intermediate form between the thio and oxon form of an OP compound (Kulkarni and Hodgson 1980).
Although, without further study it is impossible to determine if n* is describing the activating or detoxifying portion of metabolism, or possibly all reactions together. Although regressions with n* produced the best results at 10 °C, at 3 0 °C the best relationship was between half lives with midges and Vj/100 and a.
Vj/100 is the intrinsic molecular volume, which describes the energy required to separate the solvent and produce a suitably sized cavity for the insecticide (Hickey and
Passino-Reader 1991). The parameter a describes the acidity of the compound (ability of the compound to donate a hydrogen) (Kamlet et al. 1986) . Vj/100 is likely describing the importance of the volume of the insecticide
in penetrating the insect cuticle and partitioning into the various tissues within the midge, while a is possibly describing the likelihood of the compound to be oxidized during the degradation process. Overall, LSER was only 131
slightly successful in describing halflives, but was able
to account for some temperature shift, albeit with
different parameters.
MC in contrast to LSER gave significant correlations with both biological and abiotic halflives at all
temperatures tested. For regressions with abiotic
halflives, parameters such as lx and 2\ surfaced as the most important parameters. This would suggest that MC is
describing the importance of simple molecular structure in
the interactions between the insecticides and water. For
regressions with halflives when midges were present, *x
and 3x as well as higher order indices like 3cx describing
three dimensional structure were important. This is
likely showing not only the importance of interactions
between water and insecticide but also the importance of
the interaction between the three dimensional structure of
the insecticide and the receptor sites on the degradative
enzymes present in the midge.
MC indices were not only able to describe haIflives
both with and without midges, they were also able to
describe a large percentage of the variation in the data
at different temperatures. For abiotic half lives, 3cx
surfaced at all three temperatures. The results indicate
that as clusters of molecules connected by three bonds
increases the halflives also tend to increase (Table 14). 132
For biological halflives at 10 °C, only 3xv was needed to adequately describe the variation in the data. This indicates that insecticides containing three and higher order paths result in a small increase in the biological half life. At 20 °C, the presence of second order paths causes a small increase in halflife but increasing size causes a small decrease in biological halflife.
Although MC and LSER gave strong correlations with the halflife data obtained in this study, there are several problems. There always exists, when regression analysis is conducted, the possibility of chance correlations between the observations and independent variables. It has been suggested that the ratio of observations to independent variables be 10 or more (Hall and Kier 1984,
Protic and Sabljic 1989). Since the ratio of observations to variables did not exceed 10, we must consider the results of these regressions to be preliminary.
Regressions of halflives with toxicity
The lack of correlation between abiotic halflives and
ECS0 values reported in Fisher et al. (1993) is not surprising. The halflives of three out of the five compounds used in this study were greater than 10o d at 20
°C. Since the ECj0' values were determined over a 24 h period, only a tiny fraction of the insecticides would have hydrolyzed during that period, making it unlikely that this would have any effect on toxicity. The lack of correlation between halflives when midges were present
(adjusted for abiotic hydrolysis) and toxicity was unexpected since the toxicity of a compound to an organism has been reported as being related to the ability of the organism to metabolize/detoxify the compound (Brealey et al. 1980, Kulkarni and Hodgson 1980, Cheminitius et al.
1983). However, the correlation of the biological halflives with toxicity significantly improved if the halflives of carbaryl and malathion were left unadjusted for abiotic hydrolysis. This would indicate that for compounds whose hydrolysis halflives are very short, the rapid breakdown of the insecticide by hydrolysis may be as important to determining the toxicity as the ability of the midge to detoxify the compound.
Conclusions
Overall, this study shows that aquatic halflives of
OP and carbamate insecticides both with and without living organisms could be described by one unidimensional and two multidimensional models. Although the descriptive parameters varied with changing temperature, MC descriptors showed close correlation with halflives at all three temperatures studied. MC indices showed the most 134 promise in their ability to explain > 96% of the variability. Further exploration of the multidimensional parameters and their relationship to prediction of pesticide fate appears warranted. CONCLUSIONS
The first parameters used to predict a chemical's fate and effects tended to center on basic properties of a molecule, such as water solubility or lipophilicity.
Numerous studies demonstrated that these descriptors could predict the toxicity, bioconcentration, sorption to sediment and other properties of certain neutral, lipophilic compounds. However, when studies were conducted using compounds which were more water soluble or had a specific mode of action, the ability of these one dimensional descriptors to predict the desired endpoints broke down.
The results of this work corroborate the findings of these earlier studies. None of the one dimensional parameters used, HLC, MV or K,^ could predict either the potency at the receptor site or persistence of OP and carbamate insecticides. Apparently the many biochemical and/or biological processes that were involved with intoxication, hydrolysis and metabolism were too complex for these parameters to describe.
135 136
However, both of the multidimensional models showed
considerable success describing the fate and effects of OF
and carbamate insecticides even at different temperatures.
Overall, MC showed the highest correlations and the most versatility across temperatures. This indicates the
importance of molecular structure in the interaction between toxicant and receptor site or degredative enzyme as well as chemical reactions between toxicant and water.
However, the predictive ability of both LSER and MC
indices was diminished when regressions were performed with OP's and carbamates in combination.
The fact that regressions of Iso's and toxicity in resulted in significant correlations emphasizes the
connection between toxicity and AChE inhibition. It also
suggests that in studies where MC and LSER were found to describe a high percentage of OP and carbamate toxicity data (e.g. Fisher et al. 1993), the models were actually describing the inhibition of AChE by the toxicant. The
fact that no relationship was found between either abiotic or biological halflives and toxicity suggests that models describing toxicity are not characterizing these properties.
The results of this work also showed that recovery from exposure occurs quite rapidly if the intoxicated organism was returned to undosed water. The time required to recover was affected by temperature, with increasing temperature resulting in a more rapid recovery. This recovery resulted in a decrease in toxicity if the organism was exposed episodically, with at least 6 h between exposures. The fact that recovery during toxicant free periods potentially results in lower toxicity to an exposed organism could be very important for hazard assessment and the development of QSAR's. Most toxicity measurements, and therefore predictive models developed from these measurements, are based on continuous exposure studies. If recovery results in a lower toxicity of a compound to an organism, then predictive models should reflect this. LITERATURE CITED
Adams, W.J., R.A. Kimerle and R.G. Mosher 1985. Aquatic safety assessments of chemicals sorbed to sediments. In R.D. Cardwell, R. Purdy and R.C. Bahner eds., Aquatic Toxicology and Hazard Assessment: Seventh Symposium. ASTM STP 854, American Society for Testing and Materials. Philadelphia, PA. pp. 429-453.
Ahmad, S. 1970. The recovery of esterases in organophosphate treated housefly (Musca domestica L.). Comp. Biochem. Physiol. 33: 579-586.
Aldridge, W.N. 1953. Serum esterases. 2. An enzyme hydrolyzing diethyl p-nitrophenyl phosphate (E600) and its identity with A-esterase of mammalian sera. Biochem. J. 53: 117-124.
Alexander, M. 1980. Biodegradation of toxic chemicals in water and soil. In: Dynamics, Exposure and Hazard Assessment of Toxic Chemicals, R. Haque, ed. Ann Arbor Science, Pub., Ann Arbor, MI. pp. 179-190.
Aly, O.M. and M.A. El-Dib 1971. Studies on the persistence of some carbamate insecticides in the aquatic environment— I: Hydrolysis of sevin, baygon, pyrolan and dimilin in waters. Water Res. 5: 1191- 1205.
Aly, O.M. and M.A. El-Dib 1972. Studies on the persistence of some carbamate insecticides in the aquatic environment. Adv. Chem. Ser. Ill: 210-243.
Andrawes, N.R., H.W. Dorough, D.A. Linquist 1967. Degradation and elimination of temik in rats. J. Econ. Ent. 60: 979-987.
Ariens, E.J. 1980. Design of safer chemicals. In: E.J. Ariens (ed.), Drug Design, Vol. 9. Academic Press, New York, N.Y. pp. 1-46.
138 139
Badaway, M.I. and M. A. El-Dib 1984. Persistence and fate of methyl parathion in sea water. Bull. Environ. Contain. Toxic. 33: 40-49.
Baughman, D.S., D.W. Moore, and G.I. Scott 1989. A comparison and evaluation of field and laboratory toxicity tests with fenvalerate on an estuarine crustacean. Environ. Toxicol. Chem. 8: 417-429.
Bhagyalakshmi, A. and R. Ramamurthi 1980. Recovery of acetylcholinesterase activity from fenitrothion induced inhibition in the freshwater field crab (Oziotelphus senex senex). Bull. Environ. Contam. Toxicol. 24: 866-869.
Bhagyalakshmi, A., P.S. Reddy, and R. Ramamurthi 1984. The in vivo inhibition and recovery of acetylcholinesterase in the thoracic ganglionic mass of the freshwater rice field crab (Oziotelphus senex senex) during and after exposure to sumithion. Tox. Let. 21: 135-139.
Brady, J.E. and J.R. Holum 1981. Fundamentals of Chemistry. John Wiley and Sons, New York. 797 pp.
Brealey, C.J., C.H. Walker and B.C. Baldwin 1980. A- Esterase activities in relation to the differential toxicity of pirimiphos-methyl to birds and mammals. Pestic. Sci. 11: 546-554.
Breck, J.E. 1988. Relationships among models for acute toxic effects: Applications to fluctuating concentrations. Environ. Toxicol. Chem. 7: 775-778.
Borman, S. 1990. New QSAR techniques eyed for environmental assessments. Chem. Eng. News 68: 20-23.
Cairns, J. Jr., A.G. Heath and B.C. Parker 1975. The effects of temperature upon the toxicity of chemicals to aquatic organisms. Hydrobiologia 47: 135-171.
Casarett, L.J. 1991. Casarett & Doull's Toxicology: The Basic Science of Poisons, M.O. Amdur, J. Doull, C.D. Klassen (Eds.). Fourth Edition. Pergamon Press, New York, N.Y. 1033 pp.
Casida, J.E. 1964. Esterase inhibitors as pesticides. Science 146: 1011-1017. 140
Chapman, R.A., and C.M. Cole 1982. Observations on the influence of water and soil pH on the persistence of insecticides. J. Environ. Sci. Health B17: 487-504.
Chapman, P.M., M.A. Farrell and R.O. Brinkhurst 1982. Relative tolerances of selected aquatic oligochaetes to combinations of pollutants and environmental factors. Aquat. Tox. 2: 47-67.
Cheminitius, J., H. Losch, K. Losch and R. Zech 1983. Organophosphate detoxicating hydrolases in different vertebrate species. Comp. Biochem. Physiol. 76C: 85- 93.
Coppage, D.L., E. Matthews, G.H. Cook and J. Knight 1975. Brain acetylcholinesterase inhibition in fish as a diagnosis of environmental poisoning of malathion, O, O-dimethyl-S-(1,2-di-carbethoxyethyl) phosphorodithioate. Pestic. Biochem. Physiol. 5: 536- 542.
Crathorne, B., M. Fielding, C.P. Steel, and C.D. Watts 1984. Organic compounds in water: Analysis using coupled-column high-performance liquid chromatography and soft-ionization mass spectrometry. Environ. Sci. Technol. 18: 797-802.
Davison, A.N. 1955. The conversion of schradan (OMPA) and parathion into inhibitors of cholinesterase by mammalian liver. J. Biochem. 61: 203-209. de Bruijn, J. and J. Hermens 1991a. Qualitative and quantitative modelling of toxic effects of organophosphorus compounds to fish. Sci. Total Environ. 109/110: 441-445. de Bruijn, J. and J. Hermens 1991b. Uptake and elimination kinetics organophosphorus pesticides in the guppy (Poecilia reticulata): correlation with the octanol/water partition coefficient. Environ. Toxicol. Chem. 10: 791-805.
Detra, R.L. 1982. Organophosphate toxicity of larval Chironomus riparius Meigen. (Diptera: Chironomidae). Ph.D. Dissertation. The Ohio State University, Department of Entomology, Columbus, Ohio. 141
Dilling, W.L., N.B. Tefertiller and G.J. Kallos 1975. Evaporation rates and reactivities of methylene chloride, chloroform, 1,1,1-trichloroethane, trichloroethylene, tetrachloroethylene, and other chlorinated compounds in dilute aqueous solutions. Environ. Sci. Technol. 9: 833-838.
Drossman, H . , H. Johnson, T. Mill 1988. Structure activity relationships for environmental processes 1: Hydrolysis of esters and carbamates. Chemosphere 8: 1509-1530.
Eaton, J . , J. Arthur, R. Hermanutz, R. Kiefer, L. Mueller, R. Anderson, R. Erickson, B. Nordling, J. Rogers and H.Pritchard 1985. Biological effects of continuous and intermittent dosing of outdoor experimental streams with chlorpyrifos. In: R.C. Bahner and D.J. Hansen, eds., Aquatic Toxicology and Hazard Assessment: Eighth Symposium. STP 891. American Society for Testing and Materials, Philadelphia, PA, pp. 85-118.
Eichelberger, J.W. and J.J. Lichtenberg 1971. Persistence of pesticides in river water. Environ. Sci. Technol. 5: 541-544.
Ellman, G.L., K.D. Courtney, V. Andres and R.M. Featherstone 1961. A new and rapid colorimetric determination of acetylcholinesterase activity. Biochem. Pharm. 7: 88-95.
Eriksson, L., J. Jonson, S. Hellberg, F. Lindgren, M. Sjostrom and S. Wold 1991. A strategy for ranking environmentally occurring chemicals: Part V: The development of two genotoxicity QSARs for halogenated aliphatics. Environ. Toxicol. Chem. 10: 585-596.
Estenik, J.F., and W.J. Collins 1978. In vivo and in vitro studies of mixed function oxidase in an aquatic insect, Chironomus riparius. In: Pesticide and Xenobiotic Metabolism in Aquatic Organisms. ACS Symposium. Series 99. Washington D.C. pp 349-370.
Fairbrother, A., B.T. Marden, J.K. Bennett, and M.J. Hooper 1991. Methods used in the determination of cholinesterase activity. In: Cholinesterase Inhibiting Insecticides: Their Impact on Wildlife and the Environment. P. Mineau, Ed. Elsevier Pub. pp 35-71. 142
Finney, D.J. 1971. Probit Analysis. Cambridge University Press, London. 333 pp.
Fisher, S.W., M.J. Lydy, J. Barger, and P.F. Landrum 1993. Quantitative structure activity relationships for predicting the toxicity of pesticides in aquatic systems with sediment. Environ. Toxicol. Chem. (in press).
Fisher, S.W. and R.W. Wadleigh 1985. Effects of temperature on the acute toxicity of 14c-lindane in the midge, Chironomus rinarius (Meigen). J. Econ. Entomol. 78: 1222-1226.
Franks, N.P. and W.R. Lieb 1985. Mapping of general anaesthetic target sites provides a molecular basis for cutoff effects. Nature 316: 349-351.
Fukuto, T.R. 1969. Physico-organic chemical approach to the mode of action of organophosphorus insecticides. Residue. Rev. 25: 327-332.
Given, C.J. and F.E. Dierberg 1985. Effect of pH on the rate of aldicarb hydrolysis. Bull. Environ. Contam. Toxic. 34: 627-633.
Green, D.W.J., K.A. Williams, and D.R.L. Hughes 1988. Toxicity of phenol to Asellus aquaticus (L.)— effects of temperature and episodic exposure. Water Res. 22: 225-231.
Halfon, E. and R. Brueggemann. 1990. Simulation of disulfoton fate in the River Rhine with the toxfate model. Sci. Total Environ. 97/98: 385-394.
Hall, L.H., and L.B. Kier 1984. Molecular connectivity of phenols and their toxicity to fish. Bull. Environ. Contam. Toxicol. 32: 354-362.
Hall, L.H. and L.B. Kier 1989. Estimation of environmental and toxicological properties: Approach and Methodology. Environ. Toxicol. Chem. 8: 19-24.
Hall, L.H., E.L. Maynard and L.B. Kier 1989. QSAR investigation of benzene toxicity to fathead minnow using molecular connectivity. Environ. Toxicol. Chem. 8: 783-788. 143
Hansch, C. and A. Leo 1979. Substituent Constants for Correlation Analysis in Chemistry and Biology. John Wiley and Sons, New York. 339pp
Haque, R . , J. Falco, S. Cohen and C. Riordan 1980. Role of transport and fate studies in the exposure, assessment and screening of toxic chemicals. In: Dynamics, Exposure and Hazard Assessment of Toxic Chemicals, R. Haque, ed. Ann Arbor Science, Ann Arbor, MI. pp. 47- 67.
Hart, J.W. 1991. The use of data estimation methods by regulatory authorities. Sci. Total Environ. 109/110: 629-633.
Hermens, J., H. Canton, P. Jessen and R. De Jong 1984. Quantitative structure-activity relationships and toxicity studies of mixtures of chemicals with anaesthetic potency: acute lethal and sublethal toxicity to Daphnia magna. Aquat. Tox. 5:143-154.
Hermens, J. 1986. Quantitative structure-activity relationships in aquatic toxicology. Pestic. Sci. 17: 287-296.
Hickey, J.P. and D.R. Passino-Reader 1991. Linear solvation energy relationships: "Rules of thumb" for estimation of variables values. Environ. Sci. Technol. 25: 1753-1760.
Hodson, P.V., D.G. Dixon, and K.L.E. Kaiser 1988. Estimating the acute toxicity of waterborne chemicals in trout from measurements of median lethal dose and the octanol-water partition coefficient. Environ. Toxicol. Chem. 7: 443-454.
Honkakoski, P., R. Ryhanen, M. Harri, P. Ylitalo and 0. Hanninen 1988. Spontaneous recovery of cholinesterases after organophosphate intoxication: Effect of environmental temperature. Bull. Environ. Contam. Toxicol. 40: 358-364.
Ingersoll, C.G. and R.W. Winner 1982. Effect on Daphnia pulex (De Geer) of daily pulse exposures to copper or cadmium. Environ. Toxicol. Chem. 1: 321-327.
Jarvinen, A.W., D.K. Tanner and E.R. Kline 1988. Toxicity of chlorpyrifos, endrin, or fenvalerate to fathead minnows following episodic or continuous exposure. Ecotox. Environ. Saf. 15: 78-95. 144
Jorgensen, S.E. 1990. Needs of models in ecotoxicology. In: Developments in Environmental Modelling, 16, Modelling in Ecotoxicology, S.E. Jorgensen (pd.). Elsevier, New York. pp. 11-13.
Kaiser, L.E. and S.R. Esterby 1991. Regression analysis of the acute toxicity of 267 chemicals to six species of biota and the octanol/water partition coefficient. Sci. Total Environ. 109/110: 499-514.
Kamlet, M.J., R.M. Doherty, J.M. Abboud, M.H. Abraham and R.W. Taft 1986. Solubility: a new look. Chemtech 202: 566-576.
Kamlet, M.J., R.M. Doherty, G.D. Veith, R.W. Taft, and M.A. Abraham 1986. Solubility properties and biological media. 7. An analysis of toxicant properties that influence inhibition of bioluminescence in Photobacterium phosphoreum (the microtox test). Environ. Sci. Technol. 20: 690-695.
Kamlet, M.J., R.W. Doherty, R.W. Taft, M.A. Abraham, G.D. Veith, and D.J. Abraham 1987. Solubility properties in polymers and biological media. 8. An analysis of the factors that influence toxicities of organic nonelectrolytes to the golden orfe fish (Leuciscus idus melanotus). Environ. Sci. Technol. 21: 149-155.
Kamlet, M.J., R.M. Doherty, P.W. Carr, D. Mackay, M.H. Abraham, and R.W. Taft 1988. Linear Solvation Energy Relationships. 44. Parameter estimation rules that allow accurate prediction of octanol/water partition coefficients and other solubility and toxicity properties of polychlorinated biphenyls and polycyclic aromatic hydrocarbons. Environ. Sci. Technol. 22: 503-509.
Kenega, E.E. and C.A.I. Goring 1980. Relationship between water solubility, soil sorption, octanol-water partitioning and concentration of chemicals in biota. In: Aquatic Toxicology, ASTM STP 707. J.G. Eaton, P.R. Parrish, and X.C. Hendricks, eds. pp. 78-115.
Kier, L.B. and L.H. Hall 1986. Molecular Connectivity in Structure-Activity Analysis. Research Studies Press. 547 pp. 145
Kirsch, J.F., W. Clewell and A. Simon 1968. Multiple structure-reactivity correlations. The alkaline hydrolyses of acyl- and aryl-substituted phenyl benzoates. J. Org. Chem. 33: 127-132.
Klein, W. 1989. Mobility of environmental chemicals, including abiotic degradation. In: Ecotoxicology and Climate, P. Bourdeau, J.A. Haines, W. Klein, C.R. Krishna Murti, eds. John Wiley & Sons, New York. pp. 65-78.
Konneman, H. 1981. Quantitative structure-activity relationships in fish toxicity studies. Toxicology 19: 209-221.
Kulkarni, A.P. and E.H. Hodgson 1980. Metabolism of insecticides by mixed function oxidase systems. Pharmac. Ther. 8: 379-475.
Lall, R.S. 1990. Structure activity relationship of Organophosphorus insecticides. Asian J. Chem. 2: 37- 42.
Landrum, P.F., S.R. Nihart, B.J. Eadie and L.R. Herche 1987. Reduction in bioavailability of organic contaminants in the amphipod Pontoporeia hoyi by dissolved organic matter of sediment interstitial waters. Environ. Toxicol. Chem. 6: 11-20.
Leegwater, D.C. 1989. QSAR-analysis of acute toxicity of industrial pollutants to the guppy using molecular connectivity indices. Aquat. Toxicol. 15: 157-168.
Lipnick, R. 1989. Narcosis, electrophile and proelectrophile toxicity mechanisms: Application of SAR and QSAR. Environ. Toxicol. Chem. 8: 1-12.
Liu, D., K. Thompson and M.W. Strachan 1981. Biodegradation of carbaryl in simulated aquatic environment. Bull. Environ. Contam. Toxicol. 27: 412- 417.
Lohner, T.W. and S. Warwick Fisher 1990. Effects of pH and temperature on the acute toxicity and uptake of carbaryl in the midge, Chironomus riparius. Aquat. Toxicol. 16: 335-354. 146
Low, L.K. and N. Castagnoli Jr. 1980. Drug biotransformation. In M.E. Wolf (ed.) Burger's Medical Chemistry. Part I. The Basis of Medicinal Chemistry. Wiley, New York. pp. 107-226.
Lydy, M.J. 1990. The usefulness of partition coefficients in predicting the fate and effects of neutral lipophilic and moderately water soluble metabolizable compounds in the midge, Chironomus riparius. PhD Dissertation. The Ohio State University, Dept, of Zoology, Columbus, Ohio.
*Lydy, M.J., T.W. Lohner, and S.W. Fisher 1990a. Influence of pH, temperature and sediment type on the toxicity, accumulation and degradation of parathion in aquatic systems. Aquat. Toxicol. 17: 27-44.
Lydy, M.J., K.A. Bruner, D.M. Fry and S.W. Fisher 1990b. Effects of sediment and route of exposure on the toxicity and accumulation of neutral lipophilic and moderately water soluble metabolizable compounds in the midge, Chironomus riparius. Aquatic Toxicology and Risk Assessment: Thirteenth Volume, ASTM STP 1096, W.G. Landis and W.H. van der Schalie, Eds., American Society for Testing and Materials, Philadelphia, pp. 140-164.
Mabey, W. and T. Mill 1978. Critical review of hydrolysis of organic compounds in water under environmental conditions. J. Phys. Chem. Ref. Data 7: 383-415.
Mackay, D. and P.J. Leinonen 1975. Rate of evaporation of low-solubility contaminants from water bodies to atmosphere. Environ. Sci. Technol. 9: 1178-1180.
Mackay, D. and W.Y. Shiu 1981. Critical review of Henry's Law Constant for chemicals of environmental interest. J. Phys. Chem. Ref. Data 10: 1175-1199.
March, R.B., T.R. Fukuto, R.L. Metcalf, and M.G. Moxon 1956. Fate of P32 labeled malathion in the laying hen, white mouse, and American cockroach. J. Econ. Entomol. 49: 185-195.
Mancini, J.L. 1983. A method for calculating effects on aquatic organisms of time varying concentrations. Water Res. 17: 1355-1362. 147
Marquis, J.K. 1986. Contemporary Issues in Pesticide Toxicology and Pharmacology, F. Homburger, ed. Karger, New York. 108 pp..
Marshall, E. 1985. The rise and decline of temik. Science 229: 412-417.
Matsumura, F. 1985. Toxicology of Insecticides. 2nd ed. Plenum Press, N.Y. 598 pp.
McCarty, L.S. 1989. A kinetics-based analysis of quantitative structure-activity relationships in aquatic toxicity and bioconcentration bioassays with organic chemicals. Ph.D. Dissertation. The University of Waterloo, Department of Biology, Waterloo, Ontario, Canada.
McEwen, F.L. and G.R. Stephenson 1979. The use and significance of pesticides in the environment. John Wiley and Sons, New York, 538 pp.
McGowan, J.C. 1952. The physical toxicity of chemicals: III. A systematic treatment of physical toxicity in aqueous solutions. J. Appl. Chem. Lond. 4: 41-47.
McGowan, J.C. and A. Mellors 1986a. Molecular volumes and the toxicities of chemicals to fish. Bull. Environ. Contam. Toxicol. 36: 881-887.
McGowan, J.C. and A. Mellors 1986b. Molecular Volumes in chemistry and Biology. Applications Including Partitioning and Toxicity. John Wiley and Sons, New York. 259 pp.
Metcalf, R.L. and T.R. Fukuto 1962. m- Sulfurpentafluorophenyl diethylphosphate and m- sulfurpentafluorophenyl N-methylcarbamate as insecticides and anticholinesterases. J. Econ. Entomol. 55: 340-341.
Minamide, L.S. and J.R. Bamburg 1990. A filter paper dye- binding assay for quantitative determination of protein without interference from reducing agents or detergents. Analyt. Biochem. 190: 66-70.
Morgan, M.J., L.L. Fancey, and J.W. Kiceniuk 1990. Response and recovery of brain acetylcholinesterase activity in atlantic salmon (Salmom salar) exposed to fenitrothion. Can. J. Fish. Aquat. Sci. 47: 1652- 1654. 148
Motoc, I. Steric and other structural parameters of QSAR 1980. In: Steric Fit in Quantitative Structure- activity Relations, A.T. Balaban, A. Chiriac, I. Motoc and Z. Simon, eds. Springer-Verlag, Berlin.
Natsugawa, T. and P.A. Dahm 1967. Microsomal metabolism of parathion. Biochem. Pharmac. 16: 25-38.
Neil, R.A. 1967. Studies on the metabolism of diethyl 4- nitrophenyl phosphorothionate (parathion) in vitro. J. Biochem. 103: 183-191.
Nebeker, A.V., G.S. Schuytema, W.L. Griffis, J.A. Barbitta and L.A. Carey 1989. Effect of sediment organic carbon on survival of Hyalella azteca exposed to DDT and endrin. Environ. Toxicol. Chem.
Neely, W.B., D.R. Branson and G.E. Blau 1974. Partition coefficients to measure bioconcentration potential of organic chemicals in fish. Environ. Sci. Technol. 8: 1113-1115.
Newsome, L.D., D.E. Johnson, R.L. Lipnick, S.J. Broderius and C.L. Russom 1991. A QSAR study of the toxicity of amines to the fathead minnow. Sci. Total Environ. 109/110: 537-551.
Nirmalakhandan, N. and R. E. Speece 1988. structure- activity relationships: Quantitative techniques for predicting behavior of chemicals in the ecosystem. Environ. Sci. Technol. 22: 606-615.
Olson, W.P. and R.D. O'Brien 1963. The relation between physical properties and the penetration of solutes into the cockroach cuticle. Pestic. Biochem. Physiol. 3:384.
Oonithan, E.S. and J.E. Casida 1966. Metabolites of methyl- and dimethylcarbamate insecticide chemicals as formed by rat liver microsomes. Bull. Environ. Contam. Toxic. 1: 59-69.
Ormerod, W.E. 1953. Hydrolysis of benzoylcholine derivatives by cholinesterase in serum. Biochem. J. 54: 701-709.
Parsons, J.T. and G.A. Surgeoner 1991a. Effect of exposure time on the acute toxicities of permethrin, fenitrothion, carbaryl and carbofuran to mosquito larvae. Environ. Toxicol. Chem. 10: 1219-1227. 149
Parsons, J.T. and G.A. Surgeoner 1991b. Acute toxicities of permethrin, fenitrothion, carbaryl and carbofuran to mosquito larvae during single- or multiple-pulse exposures. Environ. Toxicol. Chem. 10: 1229-1233.
Pasco, D. and N.A.M. Shazili 1986. Episodic pollution— a comparison of brief and continuous exposure of rainbow trout to cadmium. Ecotox. Environ. Saf. 12: 189-198.
Pittinger, C.A., D.M. Woltering and J.A. Masters 1989. Bioavailability of sediment-sorbed and aqueous surfactants to Chironomus riparius (Midge). Environ. Toxicol. Chem. 8: 1023-1033.
Plapp, F.W., and Casida, J.E. 1958. Hydrolysis of alkyl- phosphate bond in certain dialkyl aryl phosphorothioate insecticides by rats, cockroaches, and alkali. J. Econ. Entomol. 51: 800-803.
Poirier, D.G. and G.A. Surgeoner 1988. Evaluation of a field bioassay technique to predict the impact of aerial applications of forestry insecticides on stream invertebrates. Can. Entomol. 120: 627-637.
Protic, M. and A. Sabljic 1989. Quantitative structure- activity relationships of acute toxicity of commercial chemicals on fathead minnows: effect of molecular size. Aquatic Toxicology 14: 47-64.
Randic, M. 1975. On characterization of molecular branching. J. Am. Chem. Soc. 97: 6609-6615.
Rattner, B.A., J.M. Becker and T. Nakatsugawa 1987. Enhancement of parathion toxicity to quail by heat and cold exposure. Pestic. Biochem. Physiol. 27: 330- 339.
Reddy, M.S., S.G. Lazam, and K.V.R. Rao 1987. Inhibition and recovery of acetylcholinesterase activity in the nervous tissue of prawn, Metapenaeus monoceros (Fabricius) exposed to phosphamidon and methylparathion. Curr. Sci. 56: 764-767.
Reiner, E. 1971. Spontaneous reactivation of phosphorylated and carbamylated cholinesterases. Bull. World Health Org. 44: 109-112. 150
Rothschild, E.R., R.J. Manser and M.P. Anderson 1982. Investigation of aldicarb in ground water in selected areas of the central sand plain of Wisconsin. Ground Water 10: 437-445.
Sabljic, A. 1991. Chemical topology and ecotoxicology. Sci. Total Environ. 109/110: 197-220.
SAS. 1982. User's Guide, SAS Institute, Cary, N.C.
Schuurmann, G. 1990. QSAR analysis of the acute fish toxicity of organic phosphorothionates using theoretically derived molecular descriptors. Environ. Toxicol. Chem. 9: 417-428.
Shrivastava, S.P., G.P. Georghiou, T.R. Fukuto 1971. Metabolism of N-methyl carbamate insecticides by mosquito larval enzyme system requiring NADPH2. Entomologia Exp. Appl. 14: 338-348.
Seeman, P. 1972. The membrane actions of anaesthetics and tranquilizers. Pharmacol. Rev. 24: 583-655.
Shigeoka, T., Y. Sato, Y. Takeda 1988. Acute toxicity of chlorophenols to green algae, Selenastrum capricornutum and Chlorella vulgaris, and quantitative structure-activity relationships. Environ. Toxicol. Chem. 7: 847-854.
Siem, W.K., L.R. Curtis, S.W. Glenn and G.A. Chapman 1984. Growth and survival of developing steelhead trout (Salmo gairdneri) continuously or intermittently exposed to copper. Can. J. Fish. Aquat. Sci. 41: 433- 438.
Suntio, L.R., W.Y. Shiu, D. Mackay, J.N. Seiber and D. Glotfelty 1988. Critical review of Henry's Law Constants for pesticides. Rev. Environ. Toxicol. Chem. 103: 1-59.
Thurston, R.V., C .Chakoumakos and R.C. Russo 1981. Effect of fluctuating exposures on the acute toxicity of ammonia to rainbow trout (Salmo gairdneri) and cutthroat trout (S. Clarki). Water Res. 15: 911-917.
USEPA 1975. Methods for acute toxicity tests with fish, macroinvertebrates and amphibians. EPA-6601/3-75-009, National Technical Information Service, Springfield, VA. 151
Veith, G.D., D.J. Call, and L.T. Brooke 1983. Structure- toxicity relationships for the fathead minnow, Pimephales promelas: Narcotic industrial chemicals. Can. J. Fish. Aquat. Sci. 40: 743-748.
Verma, S.R., A.K. Tyagi, M.C. Bhatnagar and R.C. Dalela 1979. Organophosphate poisoning of some fresh water teleosts-acetylcholinesterase inhibition. Bull. Environ. Contam. Toxicol. 21: 502-506.
Vighi, M . , M.M. Garlanda and D. Calamari. 1991. QSARs for toxicity of organophosphorus pesticides to Daphnia and honeybees. Sci. Total Environ. 109/110: 605-622.
Wang, M.P. and S.A. Hanson 1985. The acute toxicity of chlorine on freshwater organisms: Time-concentration relationships of constant and intermittent exposures. In R.C. Bahner and D.J. Hansen, eds., Aquatic Toxicology and Hazard Assessment: Eighth Symposium. STP 891. American Society for Testing and Materials, Philadelphia, PA, pp. 213-232.
Webb, J.L. 1963. Enzyme and Metabolic Inhibitors. Vol. I, General Principles of Inhibition. Academic Press, New York, London. 949 pp.
Weiss, C.M. 1961. Physiological effect of organic phosphorus insecticides on several species of fish. Trans. Am. Fish. Soc. 90: 143-152.
Wipke, W.T., G.I. Ouchi and J.T. Chou 1983. Computer- assisted prediction of metabolism. In L. Goldberg ed. Structure-activity Correlation as a Predictive Tool in Toxicology. Fundamentals, methods and applications. McGraw-Hill International Book Co. New York, N.Y. pp. 151-169.
Wolfe, N.L. 1980. Organophosphate and organophosphorothionate esters: Application of linear free energy relationships to estimate hydrolysis rate constants for use in environmental fate assessment. Chemosphere 9: 571-579.
Wolfe, N.L., R.G. Zepp, and D.F. Paris 1978. Carbaryl, propham and chloroproham: A comparison of the rates of hydrolysis and photolysis with the rate of biolysis. Water Res. 12: 565-571. 152
Wolfe, N.L., R.G. Zepp, and D.F. Paris 1978. Use of structure-reactivity relationships to estimate hydrolytic persistence of carbamate pesticides. Water Res. 12! 561—563.
Zaki, M.H., E. Moran and D. Harris 1982. Pesticides in ground water. The aldicarb story in Suffolk County. N.Y. Am. J. Pub. Health 72: 1391-1395.
Zaroogian, G., J.F. Heltshe and M. Johnson 1985. Estimation of toxicity to marine species with structure-activity models developed to estimate toxicity to freshwater fish. Aquat. Toxicol. 5: 251- 270.
Zepp, R.G. and D.M. Cline 1977. Rates of direct photolysis in aquatic environment. Environ. Sci. Technol. 11: 359-366.