Review
農薬および工業化学品のカエルにおける濃縮性と代謝分解
2014 年 39 巻 2 号 p. 55-68
詳細
2014 年 39 巻 2 号 p. 55-68
During the past twenty years, much attention has been given by many researchers to the worldwide decline of amphibians. Possible causes of the decline have been investigated; including the possibility of infectious diseases, habitat destruction by land use changes, global climatic changes causing increases in temperature and UV-B radiation, exposure to contaminants including pesticides, and other causes.1–3) No single factor can be attributed to explain the decline, and it is likely that complex interactions between these factors account for the phenomenon. Among these possible causes, the exposure of amphibians to pesticides and industrial chemicals has been targeted, since some chemicals may alter either their normal development by toxic effects, interactions with the thyroid axis and interference with retinoid signaling pathways, or endocrine disruption causing sexual differentiation.4) The toxicity of pesticides to amphibians has recently been reviewed in relation to their habitat and pesticide use, but with many of the existing studies conducted on an acute basis.5) In relation to a possible endocrine disruption, the U.S. Environmental Protection Agency has started to utilize the amphibian metamorphosis assay as a screening tool to scrutinize substances interfering with the normal function of the hypothalamic-pituitary-thyroid axis.6) In the European Union, there are currently no specific data requirements for amphibians for the registration of pesticides; however, available information concerning toxicity profiles, mode of action, and potential effects, including endocrine-disrupting properties, is requested for risk assessment purposes.7)
Although mortality, clinical signs and histopathological examination are very important for investigating the potentially toxic effects of pesticides on amphibians, their body burden and distribution together with information on metabolites, are required to understand the relationship between exposure and toxic effects. Additionally amphibians present a unique situation from the point of view of pesticide exposure, in view of the change of habitat from an aquatic to a terrestrial environment through metamorphosis and also absorption through their permeable skin.4,5,8) Furthermore, the bioconcentration potential of pesticides in amphibians should be also examined from the viewpoint of possible impacts via the food web on biota in higher trophic levels. From the above consideration, this review deals with the bioconcentration and metabolism of pesticides and industrial chemicals known to exist as pollutants in the aquatic environment by focusing on frogs and their tadpoles, since these profiles have been more widely investigated in the frog than in other amphibian species. The bioconcentration profiles are first addressed with some theoretical approaches on uptake and depuration necessary to understand the mechanism. Second, the metabolic profiles for several pesticides and other chemicals are summarized with the review of relevant enzymes involved in their metabolism. Based on these reviews, areas for further investigation are proposed in order to gain a better understanding of the toxic effects of pesticides on the frog.
Pesticides are considered to be taken up into the frog by passive diffusion through the gills, skin, or gastrointestinal tract via the dietary route, as shown in Fig. 1. The routes via gills and skin are involved in the uptake by tadpoles, while the skin would be a main route for the adult frog. After uptake pesticides are either partitioned to a lipid phase or metabolized by various enzymes in each organ and finally depurated as intact molecules or as metabolites. These processes in aquatic species such as fish are usually examined first by exposing a test species to an aqueous solution of pesticide under static, semi-static or flow-through conditions, depending on its stability and uptake rate, followed by transferring to clean water.9) The pesticide concentration in the frog exposed similarly to that above can be expressed by Eq. 1 under the assumption of first-order kinetics, and the flow-through condition is preferable in providing for a more precise estimation of the rate constants.9,10)
 | (1) |
Cf and Cw are the concentrations of pesticide in the frog (mg/kg wet weight) and water (mg/L), and kU and kD are first-order rate constants for uptake to the frog (L/kg/day) and depuration from frog (/day), respectively. The kU term for tadpole represents the uptake via gills and skin in total. The kD term can be divided into the rate constant corresponding to growth, metabolism, and fecal excretion. However, it is usually treated as an apparent depuration process in total and estimated from the decline curve of Cf after the transfer of biota to clean water (Cw=0). In the case of 2,4-D ester susceptible to enzymatic hydrolysis, the contribution of metabolism was concretely included in the kD term through kinetic analysis.11) Since the change in body weight cannot be neglected in a long-term study, the growth dilution factor was considered for the uptake of bisphenol A in the tadpole12) and similarly for the depuration of polychlorinated biphenyls and polycyclic aromatic hydrocarbons from the adult frog.13) There is no kinetic information available for the present on the uptake and depuration of pesticides via the dietary route in frogs.
The similar initial uptake rates (2.6 to 3.0/day) from the exposure water were reported for fipronil (log Kow=2.7; Kow, n-octanol/water partition coefficient) and benzo[a]pyrene (BaP; log Kow=6.0) by using adult Pelophylax kl. esculentus, implying no correlation of kU with log Kow.14) More polar phenol tended to be eliminated faster after its intraperitoneal injection into adult Rana temporaria, but with no clear relationship between the initial depuration rate and log Kow (−0.2 to 3.3).15) In contrast, the kD values of various polychlorinated biphenyls and polycyclic aromatic hydrocarbons injected into two frog species decreased in proportion to their log Kow (4 to 7).13) Gobas et al.16) utilized the diffusion rate of a hydrophobic chemical through membrane-diffusion layer barriers to express the kU and kD terms in fish. Their theory suggested the increase of log kU in proportion to log Kow and the constant kD at log Kow<3 to 4, while the constant kU and the proportional decrease of log kD at log Kow>3 to 4. Although the absence of exposure via gills in the adult frog may result in some differences from the uptake process in fish, the observed profiles in kU and kD may be well accounted for by this theory. Additionally, the temperature effect on the uptake and depuration processes was examined for bisphenol A by using tadpole R. temporaria.12) Both rate constants decreased by a factor of 3 to 8 when the exposure temperature was lowered from 19 to 7°C. By considering a growth factor in the study at 19°C, the kD value decreased remarkably with a slight increase of kU, indicating the contribution of a growth dilution. The kD value in this study gradually increased with an exposure concentration, and a similar trend was reported for embryo Rana pipiens exposed to fluoranthene.17)
In the early developmental stages before hatching, frog embryos are embedded in egg jelly which aids the anchoring of egg masses to aquatic vegetation, protects from mechanical disturbance and UV-B radiation, and prevents polyspermy.11,18,19) This jelly coat is a gelatinous matrix insoluble in water and secreted from glandular cells lining the oviduct of the frog. The jelly coat of Xenopus laevis consists of three layers, J1 to J3, from the inside, with its composition of proteins (37 to 48%) and carbohydrates (63 to 52%) such as hexosamine, galactose, and fucose.18) Nine macromolecules were separated by electrophoresis, with six of them being O-glucosylated proteins. NMR and MS analyses showed that oligosaccharides released from the jelly coats of various frogs are species specific, with a remarkable heterogeneity in their structure.19,20) Similarly to the toxicity assessment on fish in an early stage, the exposure of egg masses including embryos to some pesticides has been investigated for a few frog species. When the spawn of Bombina species at Gosner (G)21) stage 12 was exposed to [14C]isoproturon (log Kow=2.9) and [14C]cypermethrin (log Kow=6), more radioactivity by a factor of 3 to 5 was distributed in embryos as compared with that in jelly coats.22,23) When the spawn of R. temporaria within 3 days before hatching was exposed to DDT (log Kow=6.9) at 0.5 ppm for a day, no pesticide residue was detected in tadpoles hatching 21 days later, while the exposure of spawn 5 days before hatching resulted in significant residues (19.4 ppm) in the tadpoles with toxic signs and poor development.24) This may imply the ability of a well-developed jelly coat to act as a barrier to pesticide exposure, but the possibility of less metabolic detoxification at an early developmental stage cannot be ruled out. When the spawn of Rana sylvatica was exposed to DDT at 0.025 ppm, almost 10 times higher DDT residues were observed at 9 to 21°C in embryos (0.14 to 0.19 ppm) than those in the jelly coat without any morphological alterations.25) Furthermore, the exposure of jelly-free embryos after hatching resulted in more accumulation of DDT (0.8 to 1.6 ppm). Recently, by using jellied and de-jellied embryos of X. laevis at G stages 8 and 9, the processes of uptake, metabolism and depuration have been kinetically analyzed for [14C]2,4-D butoxyethyl ester (log Kow=4.1).11) Approximately double radioactivity was localized in embryos compared with the jelly coat, and the absence of a jelly coat clearly increased the uptake rate per embryo by a factor of 2 to 3. The metabolic rate was insensitive to the presence of a jelly coat but with its effect inconclusive for the depuration rate. These results on several pesticides (log Kow, 2.9 to 6.9) imply that the jelly coat cannot fully protect embryos from a pesticide exposure due to its hydrophilic nature, but it does play some role in reducing the exposure of embryos to pesticides.
The skin absorption of pesticides should be evaluated as one of the exposure routes for both tadpole and adult frog. Its contribution in the tadpole cannot be distinguished from the absorption via gills in a water-exposure study, while it has been examined for the adult frog using different methods. The simplest method is the direct application of a pesticide solution to frog skin, followed by analysis of remaining residues on the skin at appropriate intervals to estimate an absorption rate. When a small aliquot of acetone solution was applied to the dorsal integument of adult Bufo woodhousii, 90–100% of DDT, dimethoate and famphur were rapidly incorporated within 18 min into tissues beneath the integuments, followed by much slower absorption.26) Among the several animal species tested, the frog skin showed the highest absorption of these pesticides. A similar experiment with five pesticides using adult R. pipiens showed the most rapid absorption for carbaryl and the slowest for dieldrin with small portions being detected in the blood and liver (<5% of the applied dose after 8 hr).27) In these two studies, the estimated absorption rates showed no clear relationship with the oil/water partition coefficient of the respective pesticides. The absorption profiles of five organochlorine pesticides (log Kow, 3.8 to 6.9) in the ventral skin of adult R. pipiens showed saturation curves, when the frogs were immersed in the corresponding aqueous solution for 8 hr.28) By fitting an exponential curve, S·(1−exp(−k·t)), to the reported residues in the skin when exposed to each pesticide solution (C, mg/L), the absorption rate (k, /hr) and saturated amount (S, ng/cm2) were newly estimated. Very similar absorption rates (k=0.24 to 0.33) were obtained for these organochlorines in accordance with the Gobas theory,16) and a linear correlation was observed between log (S/C) and log Kow (slope, 0.61; r2=0.91). As another approach, the diffusion cell apparatus, consisting of two chambers with skin clamped between them, is frequently utilized to evaluate absorption and penetration of pesticides. The donor chamber is filled with an aqueous solution of pesticide, and its residue either trapped in the clamped skin or penetrating through the skin to the receptor is monitored at appropriate intervals. Among seven animal species, the fastest skin absorption of three drugs was observed in adult Rana catesbeiana by using this method,29) similarly to the deposit study.26) By applying this method to the ventral skin of adult Rana esculenta, the linear correlation between log Pe (permeation coefficients) and log Kow was confirmed for five chemicals.30) Therefore, the observed discrepancy in the correlation of absorption parameters with log Kow may originate from the different application method of pesticides, i.e., solid deposit or aqueous solution.
The different cutaneous absorption of malathion was demonstrated for Bufo marinus and R. catesbeiana by the diffusion cell technique.31) Insignificant differences in malathion permeability among ventral and dorsal skins of B. marinus and the ventral one of R. catesbeiana were accounted for by similarities in the thickness of stratum externum (1 to 2 cell layers) and density of glandular tissues. In contrast, a 2 to 3 cell layer thickness in the dorsal skin of R. catesbeiana with more glandular tissues resulted in a 25 to 30% reduction in permeability. These authors have further developed a unique method using the harvested limbs of R. catesbeiana, perfused through the common iliac artery with amphibian Ringer’s solution.32) The analysis of perfusate after application of malathion to the dorsal thigh skin showed significantly less total absorption of malathion than that by the diffusion cell method, but with similar residues in the skin. The skin absorption route is important for frogs living in a terrestrial habitat, since they are known to take up water and minerals through the highly vascularized pelvic patch in their ventral skin.6) Recently, the rapid uptake of pesticide with water through the ventral pelvic patch has been demonstrated for adult Bufo americanus, dehydrated overnight, followed by exposure to a small volume of aqueous solution of atrazine.33)
1.2. Bioconcentration factorA bioconcentration factor (BCF) is generally defined by Eq. 2, using the steady-state concentrations of pesticide in the frog (Cfs) and in water (Cws).9) BCF is equal to kU/kD, defined in Eq. 1, when both growth of frog and metabolism therein can be neglected.
 | (2) |
Consistent BCF values were given by the two definitions as shown for adult X. laevis by the flow-through exposure to atrazine.34) Compared with acute toxicity studies, less information on bioconcentration is available for the tadpole and the adult frog. The BCF values of some pesticides and industrial chemicals have been calculated according to Eq. 2 in water-exposure studies, but the experimental design is not harmonized. The steady state, defined by “concentration in fish (Cfs) within ±20% of each other in three successive analyses of Cfs made on samples taken at intervals of at least two days” in the Organisation for Economic Co-operation and Development (OECD) test guideline 305 for fish bioconcentration,9) is assumed for Eq. 2, but it has not been strictly applied to most of the reported studies.
Lipid content in the tadpole and the frog is considered to be the most important determinant for bioconcentration and generally ranges from 1 to 5% on a wet-weight whole-body basis,13,35–38) similarly to other aquatic organisms.10) There are some differences in its content between the sex and age of the frog with a seasonal variation.35,36,39) The liver and fat body of the frog have higher lipid contents than other organs, and therefore, more pesticide residues were detected therein.40,41) The distribution of pesticide and its metabolites in each tissue can be visualized and quantified by using a radio-labeled pesticide with autoradiography. In the adult R. temporaria force-fed with [14C]DDT, the amount of radioactive residues was proportional to the lipid content of each organ with the higher residues in fat body, gall bladder, liver and kidney, and for females in the ovary.42) A similar dependence of 14C distribution in each organ was observed for X. laevis exposed to [14C]atrazine.34) The bioconcentration of triclosan (an antibacterial and antifungal agent) was examined under the same experimental conditions for tadpoles of X. laevis, B. woodhousii and Rana sphenocephala, and the species dependence in BCF (2 to 540) was clearly observed, together with the effect of the developmental stage in X. laevis.43) A similar species dependence was reported for the bioconcentration of dieldrin in X. laevis, R. catesbeiana and R. pipiens, but to a lesser extent.35) The temperature effect on BCF was found to be unclear in studies on bisphenol A and DDT using tadpoles of R. temporaria and R. sylvatica, respectively.12,41) Since the BCF values were estimated on the 14C basis, further studies determining pesticide residues in biota are needed to clarify the temperature effect.
The traditional lipid/water partition model for bioconcentration introduced a linear relationship between log BCF and log Kow for non-polar pesticides. Mackay44) has demonstrated this linearity by introducing the thermodynamical concept of fugacity. This relationship is well known for fish and many other aquatic species,10) but it has not been clearly demonstrated for the frog. The averaged BCF values of some chemicals including pesticides, obtained by exposure of each frog species (tadpole and adult) to aqueous solutions, are plotted in a logarithm scale vs. log Kow, as shown in Fig. 2. A slightly lower bioconcentration of hydrophobic chemicals is observed for tadpoles (marked as filled circles) as compared with the corresponding fish BCF values cited in the U.S. EPA ECOTOX database [www.epa.gov/ecotox; log BCF]: triclosan (5 in Fig. 2), 3.6 in Danio rerio; fluoranthene (6), 4.0 to 4.2 in Pimephales promelas; dieldrin (7), 3.4 to 3.8 in Gambusia affinis; and DDT (8), 4.47 in P. promelas. Among the chemicals at log Kow>5, cypermethrin (9) and deltamethrin (10) showed much lower BCF values in tadpoles than other hydrophobic ones (DDT, dieldrin, and fluoranthene). The hydrophobic methoxychlor (12) and BaP (13) also showed lower bioconcentrations when adult frogs were exposed. These deviations may be accounted for by either much more rapid metabolism of pyrethroids or much lower skin absorption of hydrophobic aromatics in adults compared with uptake through the skin and gills in tadpoles. When these data in adult frogs and for two pyrethroids were excluded, the linear regression analysis (SigmaPlot ver. 10.0, Systat Software, Inc.) gave a poor correlation between log BCF (tadpole) and log Kow: slope, 0.468; r2=0.58 (n=14). In contrast, the non-linear regression using a quadratic function gave a much better correlation between them (r2=0.93, Fig. 2) with a maximum BCF at log Kow of 6.0. This quadratic equation predicts the BCF value of pyriproxyfen (log Kow=5.37) to be 2.58 in a logarithmic unit, which is in good agreement with the experimental result (2.7 to 2.8) obtained for tadpole X. laevis under flow-through conditions.48) A similar quadratic relationship was reported for the constituents of Arochlor 1254 in tadpole X. laevis but with much lower coefficients of correlation.49) The parabolic nature in a log BCF vs. log Kow plot with a maximum at log Kow around 6 is well known for the bioconcentration of chemicals in fish50) and other aquatic species,10) partly due to a larger molecular dimension causing less diffusion through membranes.
In contrast to bioconcentration, bioaccumulation through the dietary route has been experimentally examined in the few studies. No accumulation was observed when mealworms treated with methoxychlor were fed to adult B. americanus for 6 days, while its flow-through water exposure in the same period gave the BCF value of 2.8.47) In a similar dietary exposure study using mealworms, no bioaccumulation of pentachlorophenol was observed for adult X. laevis.51) Similarly to bioconcentration, theoretical approaches using models have been applied to bioaccumulation. The detailed kinetic analysis was conducted for fish by directly considering dietary uptake, growth dilution, egestion, and metabolism,52) but such an approach has never been taken for frogs. However, the bioaccumulation factor can be also estimated from the equation “α·F/kD” by using absorption efficiency (α) and feeding rate (F, in g food/g of frog/day).53) However, much less information on α and F in frogs5) with a large variance in these values, as compared with fish and birds, presently makes it difficult to correctly estimate the bioaccumulation factor. Recently, the fugacity concept was used to examine the relationship between the metamorphosis of Rana clamitans tadpoles and the elimination of orally dosed polychlorinated biphenyls.54) The elimination tendency represented by fugacity was greatly controlled not only by the change of a lipid content in the tadpole but also by the hydrophobicity and metabolic susceptibility of each biphenyl. The elimination of polychlorinated biphenyls after dietary exposure was also examined through hibernation of the same species frog.55) A negative correlation between Kow and the elimination rate was observed. The reduced lipid content during hibernation increased the fugacity of more hydrophobic biphenyls, causing concern about their possible contamination in offspring.
It should also be noted that the metabolic transformation of a chemical reduces its bioconcentration and bioaccumulation, though the behavior of a persistent metabolite in biota should be further assessed. The water exposure of the R. temporaria tadpole with hind limbs to DDT caused the production of DDE.24) Although no useful information on DDE is available for frogs, the comparative assessment of DDT with its potential metabolites DDD and DDE (log Kow=6.0–6.9) on bioconcentration was conducted for two amphipods in the water-exposure study.56) Very similar profiles in uptake and depuration resulted in almost the same BCF values for three organochlorines. Since DDE was nearly non-toxic in the embryonic development of X. Laevis but DDD was more lethal and teratogenic than DDT,57) the bioconcentration profiles of these metabolites should be examined similarly to DDT from a toxicological viewpoint by considering the developmental stage of the frog. The toxicological concern about an environmentally more persistent metabolite was reported for fipronil.14) More residues of the sulfone metabolite than of fipronil were detected in the P. kl. esculentus frog in the water-exposure study. Although its behavior in the frog was not examined, a slower depuration of the sulfone, by a factor of 3 to 4, than of fipronil (log Kow=4.0 and 3.7, respectively) was reported in the dietary exposure of rainbow trout to fipronil.53) The slower depuration is likely to originate from its less metabolic transformation than the parent. This information on the metabolites of DDT and fipronil strongly suggests the necessity of more investigation on the bioconcentration and bioaccumulation profiles of hydrophobic metabolites resistant to metabolism.
Although bio-activation sometimes causes an increased toxicity as known for organophosphorus (OP) pesticides, polar metabolites are usually less toxic than the parent compound, more easily depurated, and thus less likely to be bioconcentrated. In contrast to fish and other aquatic organisms,10) limited information on the metabolism of pesticides is available for the frog. As is common with other species, the metabolic profiles in the frog can be classified to phase-I and -II reactions. Phase I means the primary transformation of a pesticide via enzymatic oxidation and hydrolysis, catalyzed by mixed-function oxidases (mfo) including cytochrome P450 enzymes (CYP) and esterases, respectively. The phase-II reactions are the conjugation with glutathione (GSH), glucuronic acid (Glu), and sulfate (Sulf), at the reactive functional group in a substrate such as OH and COOH, introduced to the pesticide molecule via phase-I transformation. Acetylation and methylation were rarely reported for metabolism in the frog.
2.2. Relevant enzymes2.2.1. OxidasesCYP in mfo, mainly distributed in the microsomal fraction of tissues, catalyzes the insertion of the oxygen atom into a substrate from molecular oxygen with two electrons supplied from nicotinamide adenine dinucleotide phosphate (NADPH) or nicotinamide adenine dinucleotide (NADH).10) The total CYP content in the hepatic microsomal fraction of the frog generally ranges from 0.1 to 0.8 nmol/mg microsomal protein, which is much lower than that in the rat, mouse, and fish.58) Higher CYP contents were reported in warmer seasons for the hepatic microsomes of X. laevis and R. temporaria, but no clear correlation with the enzyme activity was observed.59,60) Its content also increased with development from tadpole to adult by a factor of 2 to 4 in X. laevis61) and R. pipiens.62) The optimal pH and temperature for CYP activity are generally 7.4 to 7.5 and around 25°C, but a higher temperature up to 37°C was known for aminopyrine N-demethylase (AD) in R. pipiens.63)
The hydroxylation of biphenyl and coumarin was reported for adult R. temporaria by using the liver post-mitochondrial supernatant, and either the requirement of NADPH as a co-factor or the reduced activity by a typical CYP inhibitor, SKF-525A, strongly suggested the involvement of CYP.64) The hepatic microsomal CYP content in adult X. laevis was determined by a CO-difference spectrum after reduction by dithionite, and it increased by injection of a CYP inducer β-naphthoflavone (BNF) with the enhanced activity of aldrin epoxidase (AE).61) The enzyme activity of AD, BaP hydroxylase (BH), and p-nitrophenetole O-deethylase (NPD) was detected in the hepatic microsomes of R. pipiens and reduced by the addition of SKF-525A.63) For these enzymes, NADH was a poor electron source as compared with NADPH, but it showed a synergism. The intraperitoneal injection of 3-methylcholanthrene (3MC) into adult R. pipiens increased not only the CYP content in hepatic microsomes but also the activity of BH and NPD, while no effect of phenobarbital (PB) was observed.65) The microsomal CYP content in the liver of X. laevis increased by a factor of 2 through injection of 3MC, but it was much less than those of R. catesbeiana, Rana nigromaculata and Bufo bufo japonicus.66) The authors found that the enzyme activity in the NADPH-supported oxidation of aniline, aminopyrine, 7-ethoxycoumarin and BaP varied among these four species. Furthermore, the spectrophotometric titration of the hepatic microsomes of R. catesbeiana with cyanide showed a titration curve consisting of at least three phases, implying the presence of multiple CYP isoforms.
The presence of various CYP families in the frog has been clarified by using gene and genome sequencing (www.xenbase.org/common). Among them, CYP families 1 to 4 have been extensively studied, not only by the classical biochemical analysis of enzyme activity in relation to substrate specificity, inhibition and induction, but also by the comparison of amino acid sequences and immunochemical and genetic analyses.58) Among them, the CYP1 family has been rather more investigated by using X. laevis. The single band corresponding to a molecular mass of 55 to 66 kDa was detected in the hepatic microsomes by using SDS-PAGE and immunoblot with antisera raised against rat CYP1A1, and the band intensity was strengthened by the intraperitoneal injection of BaP.67) In contrast, no corresponding band was detected in the embryo 8 days post-fertilization, but the enzyme was induced by 1-day exposure to BaP solution. Fujita et al.68) have demonstrated induction of the hepatic enzyme (ca. 54 kDa), detected by the same antisera as above, in the adult females by injection of 3MC. The authors constructed the liver cDNA library for screening using a fragment of rat CYP1A2 cDNA under low stringency conditions. The two largest isolated clones, being 2,618 and 2,905 bp in length, resembled each other with >90% identities in DNA and their coding amino acid sequences. From the poor identity (55 to 63%) in amino acid sequences toward mammalian CYP1A1, 1A2 and fish CYP1A, these enzymes were newly named as CYP1A6 and 1A7. 7-Ethoxyresorufin O-deethylase (EROD) activity in the hepatic microsomes of R. pipiens more than doubled following injection of BNF, and such an induction by PB was observed for AE, EROD, and 7-pentoxyresorufin O-depentylase.62) These results showed the presence of CYP1A and 2B in this frog. The induction of hepatic CYP1A (53.5 kDa) in R. pipiens by 3,3′,4,4′,5-pentachlorobiphenyl (PCB126) was confirmed by using monoclonal antibody 1-12-3 in the Western blot analysis.69) Furthermore, immunohistochemical analysis using this antibody showed that the induced CYP1A was localized in the liver, kidney, stomach, and intestines. The induction of CYP1A by various (polychlorinated) polycyclic aromatic hydrocarbons was also reported for adult R. esculenta.70) Recently, novel CYP1A cDNA fragments (122 bp in length) were isolated from the livers of six frog species together with X. laevis by using a reverse transcription polymerase chain reaction (RT-PCR), and a 75 to 98% identity was found for their cDNA nucleotide sequences.71)
In the liver of adult R. esculenta complex , the induction of 7-methoxyresorufin O-demethylase and EROD activity by BNF together with immunoblot and RT-PCR analyses showed the presence of not only CYP1A but also other CYP families.72) The Western blot analysis using antisera raised against rat CYP2 (2B, 2C11, 2E1), rabbit CYP2A 10/11 and trout CYP3A27 indicated the possible involvement of CYP2 and 3 in its liver, the latter of which was supported by the hydroxylation of testosterone at the 6β-position. By applying successive chromatography and chromato-focusing to a liver microsomal fraction of female X. laevis, Saito et al.73) have purified the enzyme classified to CYP2 by Western blot analysis. This enzyme was found to be closest to rat CYP2B1 by comparing the amino acid sequences of lysylendopeptidase-digested peptides isolated by HPLC with those of the several rat CYPs, while its substrate specificity resembled that of mammalian CYP2E1. This enzyme was later named CYP2Q through cDNA nucleotide sequence analysis.74) The microsomes of Xenopus A6 kidney epithelial cells showed corticosterone-6β-hydroxylase activity, which was inhibited by the anti-rat CYP3A1 or 3A6 antibody.75) SDS-PAGE and immunoblot analyses showed that this enzyme, having a molecular weight of ca. 52 kDa, can be classified to the CYP3A family. In addition, adult R. catesbeiana was found to show the laurate hydroxylase activity characteristic to mammalian CYP4A.60)
2.2.2. EsterasesAcetylcholinesterases (AchE), butylcholinesterases (BchE) and carboxylesterases (CaE) are B-type esterases, typically inhibited by OP chemicals. Cholinesterases have been examined in relation to the toxic effects of pesticides at neuromuscular junctions. CaEs usually distributed in all tissues are known to play a role in detoxifying pesticides having an ester linkage.10) The activity of these esterases has been examined for tadpoles of X. laevis,76) Rhinella arenarum,77,78) and Scinax fuscovarius.79) The relative activity of these enzymes cannot be simply compared between these tadpoles due to the different fractions of tissue homogenates together with a substrate specificity in CaE, but the BchE activity seems to be lower than the other two. By using the post-mitochondrial supernatant of tissue homogenates from tadpoles of eleven species, Lajmanovich et al.80) have reported large variations in the activity of three esterases. The authors also found the significant substrate specificity in CaE, indicating the possible presence of multiple isozymes. The presence of CaE isoforms with the optimal pH around 8 was shown in S. fuscovarius not only by no-saturation kinetics in the Lineweaver-Burk plot of the enzyme activity, but also by the presence of at least five esterasic bands in the zymogram after staining with α- and β-naphthyl acetate.79)
Phenyl methyl sulfonyl fluoride (PMSF) generally inhibits most esterases, but tetramonoisopropyl pyrophosphorotetramide (iso-OMPA) and 2-(O-cresyl)-4H-1,3,2-benzoxaphorin-2-oxide (CBDP) specifically inhibit BchE and CaE, respectively.10) PMSF specifically reduced the CaE activity in S. fuscovarius, while iso-OMPA caused no effects on AchE and CaE.79) When CBDP was subcutaneously applied to adult R. pipiens, the trans-isomers of resmethrin and permethrin exhibited acute toxicity higher (by a factor of 5 to >12) than the untreated controls but with insignificant effects on cis-isomers, indicating that CaE more favorably metabolized trans-isomers of these pyrethroids.81) The hepatic microsomal esterases in R. temporaria similarly hydrolyzed the trans-isomer of cypermethrin more rapidly than the cis-isomer.82) Furthermore, the effects of pesticide exposure on esterases in tissue homogenates have been reported. OP pesticides and carbaryl were found to significantly reduce the B-esterase activity, in line with their mode of action.77–79,83) When tadpole S. fuscovarius was exposed to diazinon dissolved in water for up to 7 days, the AchE activity significantly decreased with its concentration increasing, but an insignificant effect on CaE was observed.79) This difference may be accounted for by the presence of multiple isoforms of CaE.
2.2.3. Glutathione-S-transferases (GST)The transfer of tripeptide GSH to an electrophilic substrate is catalyzed by GST, mainly distributed in a tissue cytosolic fraction, and the resulting conjugate is successively transformed to a mercapturic acid conjugate.10) 1-Chloro-2,4-dinitrobenzene (CDNB) is the most common substrate for GST isozymes, consisting of two subunits, each having a molecular mass of 23 to 28 kDa.10) Di Ilio et al.84) first purified GST from the cytosolic fraction of de-jellied embryo B. bufo by GSH-Sepharose affinity chromatography. Each of six enzymatically active fractions isolated by chromato-focusing showed a 23-kDa single band in SDS-PAGE and exhibited the higher activity characteristic to π-class GST toward ethacrynic acid (ET) and 4-nitroquinoline-1-oxide. The Western blot analysis and the sequences of N-terminal 36 amino acids classified all of these fractions to be π-class GST. The authors further showed that both the most abundant fraction in the purified cytosolic GST from the liver of adult B. bufo (bbGSTP2-2) and its embryonic GST (bbGSTP1-1) are homodimers (ca. 50 kDa) consisting of two subunits (23 kDa).85) Interestingly, adult and embryonic GSTs showed significant differences in substrate specificity; much higher activity for cumene hydroperoxide but much less for ET and trans-non-2-enal in the adult bbGSTP2-2 than in the embryo. Furthermore, the Western blot analysis of bbGSTP2-2 showed no cross-reaction with antisera raised against embryonic bbGSTP1-1 and mammalian GST (α-, μ- and π-classes). The amino acid sequence of bbGSTP2-2 showed the 50 to 60% identity with both mammalian π-class GST and bbGSTP1-1, but much less (<36%) with other class GSTs.86) Eight amino acids contributing to bind GSH in π-class GST were strictly conserved in bbGSTP2-2, but the change of Tyr-109 with Phe-109 resulted in much less activity toward ET.87) The authors have further investigated GST isoforms in each organ of adult B. bufo by amino acid sequence analyses. bbGSTP2-2 was major (65 to 71%) in liver and kidney, while bbGSTP1-1 and other GSTs (α-, μ- and σ-classes) predominated in other organs, and bbGSTP1-1 was >90% in the ovary. Through these investigations, the switch of GST from bbGSTP1-1 to bbGSTP2-2 with development was presumed to relate to the change of habitat from an aquatic environment to a terrestrial one.
A similar approach to isolate and characterize GST was applied to the embryo and adult liver of X. laevis, and its specific activity was found to be higher in the cytosolic fraction of the liver rather than in the embryo.88) Three common bands detected in SDS-PAGE were also confirmed by HPLC analysis of purified proteins, and the molecular mass of each subunit was determined by LC-ESI-MS to be 22, 24, and 25 kDa. These GSTs were confirmed to be μ-, α-, and σ-classes by Western blot analysis. Only quantitative differences in the GST isoforms, where σ- and μ-classes were predominant in the embryo and adult liver, respectively, were presumed to relate to the situation where there was no change in habitat with development, i.e., continuously in an aquatic environment, as distinct from other frogs such as B. bufo. In the Western blot analysis, the cytosolic GST in the liver of R. esculenta complex cross-reacted with the antisera raised against the σ-class GST of X. laevis and α-class one of rat, while that in nasal mucosa only reacted with the antisera against B. bufo bbGSTP1-1.72) The fractions in both organs did not react with any antisera raised against mammalian μ- and π-classes and fish θ-class, indicating the presence of different isoforms.
The GST activity varies not only with developmental stage but also with exposure to pesticides. Much more cytosolic GST activity toward CDNB was detected than with the microsomal one in tadpole Bombina variegata, and the former activity showed a maximum at G stages 24 and 25, where the development of operculum and the disappearance of external gills were observed.22) The authors also reported that the cytosolic GST activity at G stage 25 significantly increased by exposure to isoproturon as either its concentration or exposure period increased. A similar increase of GST activity in the same species was induced by cypermethrin.23) About a 4-fold increase of GST activity was clearly observed in the liver of adult B. regularis by 28-day exposure to endosulfan, and such an effect was also caused in the brain by exposure to diazinon, which may originate from the detoxification processes of these pesticides.89) Azinphos-methyl and carbaryl increased the GST activity in tadpole R. arenarum by 30 to 40%, but GSH reductase and peroxidase were not affected.76) The induction of the cytosolic GST activity was also reported by exposure of P. kl. esculentus to BaP, especially in its liver.14) In contrast, the cytosolic GST in tadpole X. laevis was inhibited by exposure to λ-cyhalothrin and deltamethrin under static conditions, which was explained by their possible association with the active site of GST.76)
2.2.4. Glucuronosyl transferases (UGT)Glucuronidation is one of the major phase-II conjugation pathways known for aquatic organisms such as fish. UGT, distributed in a microsomal fraction, catalyzes the transfer of Glu from the corresponding uridine diphosphate derivative (UDPGA) to O, N, or S atoms of a substrate in the presence of Mg2+.10) UGT activity toward 2-aminophenol was reported for the liver homogenates of adult R. pipiens, but it was absent in adult X. laevis and larvae R. catesbeiana.90) More Glu conjugate of 2-nitrophenol was formed by incubation with kidney homogenates of R. esculenta in the presence of UDPGA than the liver ones.91) The kinetic analysis of UGT activity toward pyrene hydroxylated at the 1-position (1-OH-Pyr) showed that the Michaelis–Menten constants (Vmax and Km) for adult X. laevis were much smaller than those for the amphibian Japanese newt (Cynops pyrrhogaster).92) In regard to UGT activity toward 3-trifluoromethyl-4-nitrophenol in the hepatic microsomal fraction of R. catesbeiana, almost the same Vmax/Km values between tadpoles at G stages 33 to 38 and adult indicated a similar enzyme efficiency, while the different Vmax and Km values implied the involvement of more than one isoform.93) In the tadpoles of wild-type X. laevis at Nieukoop and Faber (NF)94) stage 45, the presence of mRNA encoding UGT1A1 and UGT1A6 was confirmed by quantitative polymerase chain reactions (qPCR), with the level increasing by exposure to triiodothyronine (T3).95)
2.2.5. Sulfotransferases (SULT)Sulf conjugation, catalyzed by SULT, is the other important phase-II transformation. The sulfate group is transferred from phosphoadenosyl phosphosulfate (PAPS) to O or N atoms of a substrate.10) SULT, distributed mainly in a cytosolic fraction, is a homodimer, with its subunit molecular weight of 31 to 35 kDa, and a few isoforms are known to conjugate different chemical classes. Not only the bile alcohols such as β-cyprinol and scymnol but also simple primary alkyl alcohols were effectively sulfated by the liver cytosol fraction of X. laevis at pH 8 in the presence of PAPS.96) The requirement of Mg2+ and the effect of thiols on the enzyme activity are different between bile and alkyl alcohols, indicating the presence of isoforms. SULT in the liver cytosol of adult X. laevis showed about a 10-fold higher substrate specificity on hydroxylated pyrene at the 1-position than that of C. pyrrhogaster.92) Rahman and Yamauchi97) have recently characterized SULTs catalyzing the sulfation of thyroid hormones in the liver cytosol of R. catesbeiana. The enzyme activity was highest in the liver at pre-metamorphic Taylor and Kollros (TK)98) stage 10. As with the development of the tadpole, enzyme activity gradually decreased, but with a slight increase at stage 20 (metamorphic climax), and finally increased again in the adult. The Vmax/Km values in the tadpole at TK stage 10 showed much higher substrate specificity of T3 than thyroxine (T4). The substrate specificity and inhibitory effects by various phenols showed the presence of isoforms, including one similar to mammalian SULT1. In relation to the metabolism of tetrabromobisphenol A in the tadpole of wild-type X. laevis at NF stage 45, the presence of mRNA encoding SULT1A1 was confirmed by qPCR with its level not modified by exposure to this chemical.95)
2.2.6. Other enzymesReductive metabolism of nitrophenols in R. temporaria and X. laevis was rationally presumed by the detection of their acetylamino derivatives,99,100) but the information on corresponding reductases is very limited. No nitro reduction of 4-nitrobenzoic acid was confirmed by using liver homogenates of R. pipiens, R. catesbeiana and B. marinus, while these species could reduce the azo moiety of the antibiotic neoprontosil in the presence of co-factors such as NADPH under N2.101) The presence of acetyl coenzyme A-arylamine-N-acetyltransferases in some frog species was demonstrated by Ho et al.,102) who have investigated their kinetics in the acetylation of 2-aminofluorene and 4-aminobenzoic acid by using the tissue cytosolic fraction of adult Rana tigrina. The highest activity for the former substrate in the liver with a significantly tissue-dependent Km and similar activity among tissues for the latter suggested the presence of at least two isoforms. Furthermore, the wide distribution of the enzyme-specific activity among 100 individuals may imply an enzyme polymorphism. In addition, N- and O-methylation are well-known transformation reactions for small molecules to proteins. S-Adenosyl-L-methionine (SAM)-dependent methylation has been thoroughly investigated in many species for biosynthesis of phosphochloline, and the presence of phosphoethanolamine methyltransferase was reported for X. laevis.103) The soluble fraction in parotid gland homogenates of B. marinus catalyzed many kinds of N- and O-methylation of small organic molecules in the presence of SAM.104) By using the homogenates of the liver, kidney and intestine from B. vulgaris formosus, tryptamine was N-methylated in the absence of any additional co-factor.105) These observations suggested the possible involvement of methyltransferases in the metabolism of pesticides in the frog.
2.3. Metabolism of chemicals including pesticidesMetabolism studies have been conducted on adults and tadpoles of various frog species by intraperitoneal injection of a chemical or exposure to its aqueous solution. Tissue homogenates, microsomal or cytosolic fractions were conveniently utilized to examine basic metabolic profiles. Information on the metabolism of pesticides in the frog is not available from the limited bioaccumulation studies through the dietary route.42,106) Metabolism studies in the tadpole and adult frog where there has been a definite identification of metabolites are listed in Table 1. No frog-specific transformation reaction is known at present, as compared with the metabolism in other aquatic species.10) Enzymatic hydrolysis is a typical reaction for OP and carboxylic esters, but very little information on metabolite identification is available. 2,4-D butoxyethyl ester was quickly hydrolyzed to 2,4-D by jellied embryos of X. laevis.11) trans-Isomers of chrysanthemate-type pyrethroids are more susceptible to hydrolysis catalyzed by hepatic esterases of R. pipiens, and the presence of an α-cyano group in cypermethrin reduced the hydrolysis rate.81)
| Chemical | Species (Developmental stagea) | Appl.b | Fractionc | Profilesd | Reference |
|---|---|---|---|---|---|
| Chlorobenzenes | Rana pipiens (adult) | ip | ex | HD, O1 | 112 |
| Phenol | Rana temporaria (adult) | ip | ex | G, S, O1 | 99, 107 |
| Xenopus laevis (adult) | ip | ex | S, O1 | 99 | |
| 3-Nitrophenol | Rana temporaria (adult) | ip | ex | A, G, O1, R, S | 100 |
| 4-Nitrophenol | Rana temporaria (adult) | ip | ex | A, G, O1, R, S | 99 |
| Xenopus laevis (adult) | ip | ex | A, O1, R, S | 99 | |
| 2-Methylphenol | Rana temporaria (adult) | ip | ex | G, O1, O2, S | 99 |
| Xenopus laevis (adult) | ip | ex | O1, O2, S | 99 | |
| Biphenyl | Rana temporaria (adult) | ivl | — | O1 | 64 |
| 4,4′-Dichlorobiphenyl | Rana aesculenta (adult) | ip | ex | HD, O1 | 111 |
| Chlorobiphenyls | Rana pipiens (adult) | ip | ex | HD, M, O1 | 112 |
| Bisphenol A | Xenopus laevis (NF 45) | w | ex | G, S | 114 |
| Tetrabromobisphenol-A | Xenopus laevis (NF 45) | w | wh | G, S | 95 |
| Benzo[a]pyrene | Pelophylax kl. esculentus (adult) | w | i, l | O1 | 14 |
| Xenopus laevis (adult) | w | ex | O1, S | 92 | |
| Rana tagoi (adult) | w | ex | G, O1, S | 92 | |
| DDT | Rana temporaria (TK 4–14) | w | wh | DC | 24 |
| Bufo bufo (TK 6–16) | w | wh | DC | 24 | |
| Rana japonica japonica (adult) | os | f, l | DC, O2 | 120 | |
| Chlordene | Xenopus laevis (adult) | ivl | — | O2, O4 | 61 |
| 2,4-D BEEe | Xenopus laevis (G 8–9) | w | wh | E | 11 |
| trans-Permethrin | Rana pipiens (adult) | ivl | — | E | 81 |
| Cypermethrin | Rana temporaria (adult) | ivl | — | E, O1, O2, G | 82 |
| Parathion | Rana catesbeiana (tadpole) | w | w | O6 | 46 |
| Isoproturon | Bombina bombina (G 25) | w | wh | O2, O3 | 22 |
| Bombina variegata (G 25) | w | wh | O2, O3 | 22 | |
| Atrazine | Xenopus laevis (NF 66) | w | wh | O3 | 34 |
| Fipronil | Pelophylax kl. esculentus (adult) | w | i, k, l | O5, R | 14 |
| Phenytoin | Rana pipiens (adult) | ip | ex | G, O1, M | 113 |
| Aldosteron | Rana catesbeiana (adult) | ip | ex | G, R | 118 |
| Norepinephrine | Rana pipiens (adult) | ip | ex | DA, M, O2, R | 122 |
| Triiodothyronine (T3) | Rana catesbeiana (TK 5–7) | ip | wh | G, S | 119 |
| Tryptamine | Bufo vulgaris formosus (adult) | ivl | — | DA, M | 105 |
a G, Gosner;21) NF, Nieukoop & Faber;87) TK, Taylor & Kollros.91) b Method of application. ip, intraperitoneal injection; os, oral adminitration; w, exposed to water treated with a chemical; ivl, in-vitro metabolism using liver homogenates. c For analysis of metabolites. ex, water medium including excreta.; f, fat; i, intestine; k, kidney; l, liver; wh, whole body. d Metabolic profiles. A, acetylation; DA, deamination; DC, dechlorination; E, ester cleavage; G, glucuronic acid conjugation; HD, hydrolytic dechlorination; M, methylation; O1, ring oxidation; O2, alkyl oxidation; O3, O(N)-dealkylation; O4, epoxidation; O5, S-oxidation; O6, oxidative desulfuration; R, reduction; S, sulfate conjugation. e Butoxyethyl ester.
Enzymatic oxidation has been most frequently reported for many chemicals, with the reaction profiles classified to hydroxylation of the aromatic ring and alkyl group, dealkylation, epoxidation, and sulfur oxidation. The ring oxidation to form the corresponding quinol, catechol, and resorcinol was reported for phenol,99,107) 3-nitrophenol,100) 4-nitrophenol, and 2-methylphenol,99) all of which were intraperitoneally injected into adult R. temporaria or X. laevis. Since the oxo-iron (IV) porphyrin π-cation radical is considered to be a reactive species in catalytic oxidation by CYP and has an electrophilic nature,108) the regio-selectivity in ring hydroxylation can be qualitatively explained by the electron density in the highest occupied molecular orbital (HOMO) of a chemical; the higher electron density is located, the more hydroxylation proceeds.109) The semi-empirical molecular orbital (MO) calculations were conducted for each chemical through a full optimization of its molecular geometry by the MOPAC 2002 program in WinMOPAC 3.9 (Fujitsu Ltd.)110) with the following calculation keywords: PM3, PRECISE, EF, VECTORS and ALLVEC. The aromatic carbon of phenol with a higher localized HOMO electron density was more susceptible to hydroxylation: carbon position, HOMO electron density (total % of hydroxylated product and its Sulf conjugate);107) o, 0.22 and 0.28 (1 to 9%); m, 0.06 and 0.13 (0.3 to 3%); p, 0.56 (1 to 25%). A similar tendency was observed for the other phenol derivatives, but less formation of the corresponding metabolites made the correlation unclear.
Hydroxylated biphenyl at the 4-position (Fig. 3) was produced about eight fold more by incubation with the hepatic post-mitochondrial supernatant of R. temporaria than that at the 2-position,64) which was qualitatively accounted for by the higher HOMO electron density at the 4-position (0.31) than at the 2-position (0.16). The 4(4′)-Cl carbons of 4,4′-dichlorobiphenyl are considered to be a primary attacking site, judging from the higher HOMO electron density (0.23) than those at the 2(2′)- and 3(3′)-positions (0.06 to 0.11). The formation of a 4-OH derivative by its intraperitoneal injection into R. esculenta coincided with this estimation, and the detection of 3-OH and 3-Cl–4-OH derivatives strongly suggested the presence of an arene oxide intermediate.111) A similar regio-selectivity was reported for many polychlorinated aromatics in R. pipiens.112) By using 4-chloro-4′-deuteriobiphenyl as a substrate, the authors found about 80% retention of deuterium in the 4′-OH derivative produced and revealed that the ring hydroxylation proceeded via NIH shift. One phenyl group of the drug, phenytoin (Fig. 3), was predominantly hydroxylated at the p-position in R. pipiens.113) The HOMO electron density is localized at the p-position of one phenyl ring (0.51) with lower density at the o/m-positions (0.04 to 0.20) and the other ring (0.002 to 0.08), which accounts for this regio-selectivity. Furthermore, the selective oxidation at the 4′-position of the 3-phenoxybenzyl moiety was reported through incubation of cypermethrin with the hepatic post-mitochondrial fraction (R. temporaria),82) which was correctly estimated by the higher HOMO electron density.109)
The static exposure of adult P. kl. esculentus to BaP in water produced a 3-OH derivative in the liver and intestine with a few unknown metabolites.14) In the case of static exposure of many frog species to pyrene (Fig. 3), the formation of 1-OH-Pyr was mainly observed, followed by further ring hydroxylation but with no information on its position.92) The MO calculation showed the localization of HOMO not only at the 3-position of BaP (Fig. 3; density, 0.16) but also comparably at the 12- and 3- to 6-positions (0.17 and 0.13 to 0.33). The same HOMO electron density at the 1- and 3-positions (0.24) was estimated for pyrene, together with slightly lower density at the 4/5-positions (0.17). The comparison of hydroxylated positions in these primary metabolites with HOMO distribution showed that a steric constraint on larger molecules at the catalytic site of CYP also controlled the regio-selectivity in oxidation. Since the HOMO of 1-OH-Pyr is most localized at the 6- and 8-positions (0.22), followed by other secondary carbons (0.13 to 0.18), these are the candidate sites for the several conjugates of the dihydroxylated pyrene.
By exposing tadpoles of two toad species at G stage 25 to isoproturon in water, methyl or methine carbon of the isopropyl group was hydroxylated with a species-dependent regio-selectivity.22) Hydroxylation at the gem-methyl group of the cyclopropyl ring was reported for cypermethrin in adult R. temporaria.82) The application of the semi-empirical MO method is not effective for alkyl oxidation, since the π–π interaction between a ring carbon and the active oxygen species in CYP is not the primary reaction mechanism.109) N-Demethylation of isoproturon was considered to proceed via successive oxidation of the methyl group by oxidases, and a similar mechanism was most likely to be involved in the formation of N-dealkylated atrazine metabolites by larvae X. laevis at NF stage 66.34) The successive oxidation of an aryl methyl group finally to COOH was reported in the intraperitoneal injection of 2-methylphenol into two frog species.99) Epoxidation of cyclodiene insecticides, aldrin and chlordane, was observed by using the hepatic microsomal fraction of adult X. laevis, together with hydroxylation at the 1-position of the latter.61) The indirect evidence on the oxidative desulfuration catalyzed by CYP was given by the reduced toxicity of chlorpyrifos on tadpole X. laevis in the presence of piperonyl butoxide which inhibited the CYP-catalyzed formation of its toxic oxon derivative.115) In the flow-through exposure of parathion, the detection of paraoxon in the exposure water but not in tadpole R. catesbeiana indicated its rapid depuration after oxidative metabolism.46)
Any GSH conjugate has not yet been detected as in other aquatic species,10) although the GST activity is known to be enhanced by exposure to chemicals. The phase-II reactions to form Glu and Sulf conjugates are reported in many metabolism studies. The formation of these conjugates was confirmed by identification of the corresponding free metabolites released by treatment with glucuronidases and sulfatases. The difficulty in examining a Glu conjugation in-vivo studies, due to the co-existence of β-glucuronidase, is known.116,117) Hepatic bilirubin UGT activity with an optimal pH value of 6.85 was confirmed in R. catesbeiana at all developmental stages up to adult, and it was about one-fifth of that in the mouse liver enzyme. However, no Glu conjugate was detected in plasma and bile via intraperitoneal injection of bilirubin. The co-injection of saccharo-1,4-lactone, a β-glucuronidase inhibitor, led to the formation of a Glu conjugate, indicating that the action of β-glucuronidase masked the formation of the conjugate. Phenols are susceptible to Glu and Sulf conjugations with their relative preference depending on ring substituents, based on the metabolism studies using R. temporaria (% Glu, % Sulf): phenol (18%, 25%), 4-nitrophenol (13%, 48%), 2-methylphenol (8%, 33%),99) and 3-nitrophenol (57%, 26%).100) Furthermore, the species difference in the relevant enzyme activity may result in the preferential formation of these conjugates. Glu conjugates were detected for the metabolism of phenytoin in R. pipiens,113) cypermethrin in liver homogenates of R. temporaria,82) and aldosterone in R. catesbeiana,118) while Sulf ones were predominant for T3 in R. catesbeiana,119) pyrene,92) tetrabromobisphenol A95) and phenols99) in X. laevis.
In contrast to the oxidation and conjugation reactions as above, other metabolic pathways have scarcely been reported. The reductive pathways were indirectly speculated from the formation of N-acetyl products of 4-nitrophenol99) and a 3β-OH-5β-tetrahydro derivative of aldosterone.118) Reductive dechlorination and dehydrochlorination were known for DDT in a few species.24,120) Both oxidative and reductive metabolism were observed in female P. kl. esculentus exposed to an aqueous solution of fipronil.14) In each organ, the main metabolite was sulfone, with sulfide as a minor component only detected in the blood, gall bladder and skin. Since the desulfonyl derivative is specifically formed by photolysis,121) its detection not only in all organs but also in the exposure water may indicate its photolytic formation in water followed by uptake into the frog. Methylation was reported for a phenolic OH by the intraperitoneal injection of 4-chlorobiphenyl112) and norepinephrine122) into R. pipiens and for an amino nitrogen by incubation of tryptamine with tissue homogenates of Bufo vulgaris.105) For the latter two chemicals, the deamination was also detected possibly via oxidative processes.105,122)
Several scenarios for pesticide exposure can be supposed for the frog, depending on its developmental stage from embryo in spawn to adult through metamorphosis, as well as its habitat. Different from fish, the spawn and tadpoles of frogs are mostly found in stagnant and shallow water bodies and are more susceptible to pesticide exposure. The hydrophilic jelly coat surrounding embryos, consisting of proteins and carbohydrates, does not seem to fully protect them from pesticide exposure, however the relevant investigations are limited to a few pesticides. Pesticide incorporated by diffusion is expected to be released by the transfer of contaminated spawn to clean water. In order to clarify the toxic effects of pesticides on the embryo, the exposure of spawn should be kinetically examined for more residues of pesticides. The bioaccumulation of pesticides is governed by the balance between uptake, metabolism and depuration, similarly to fish. The more hydrophobic a pesticide is, the higher bioconcentration is observed with more distribution in the organs having a higher lipid content. Either the linear relationship between log BCF and log Kow or parabolic profiles at a wider log Kow range are observed. The maintenance of an exposure concentration at a constant level, lower than the toxic level, is essential to correctly estimate an uptake rate and BCF. The content, composition and tissue distribution of lipids may vary with developmental stage, sex, species, and season, and they are an important affecting factor in BCF. Therefore, the measured lipid content in tadpole and adult frogs should be used for the normalization of BCF. Furthermore, the longer exposure period covering the several developmental stages in the tadpole results in significant changes in body weight, and hence a growth dilution effect should be taken into account. The temperature effect on bioconcentration is another issue to be investigated due to its relation to growth and enzyme activity.
The absorption of pesticides via skin should be better clarified in relation to the adult frog. Since the new technique using excised frog limbs is too cumbersome to be routinely conducted, the basic profiles on absorption of each pesticide should be defined first by their direct application to the skin or the flow-through two-compartment diffusion cell method in relation to the physico-chemical properties of pesticides and skin structures of each species. In contrast, a similar approach cannot be applied to the tadpole due to the technical difficulty of pesticide application outside water, together with yhe much smaller body size. Therefore, the contribution of skin absorption should be estimated in total with the uptake through gills, similarly as for fish. Furthermore, adult frogs except X. laevis mainly live in a terrestrial environment. One of the possible routes of exposure is skin absorption by dipping their partial or whole body in water containing a pesticide.
Frogs occasionally absorb water through the ventral pelvic patch, however such a route of pesticide exposure has rarely been examined to date. The contribution of this route should be further examined with the mechanism, especially for pesticides likely to contaminate stagnant pools as one of the habitats of frogs. Information on a dietary uptake of pesticides in the frog is very limited, and only a few studies have been conducted to examine the bioaccumulation and toxic effects caused by residues in prey. From the viewpoint of biomagnification via the food web, the type of feeding study in accordance with the OECD guideline 305 is suggested to quantify the body burden of persistent and hydrophobic pesticides.
Metabolism studies have been conducted by either in-vitro methods using tissue homogenates and centrifuged fractions or the in-vivo intraperitoneal injection method. Although these methods are convenient for surveying the potential metabolic pathways in the frog, the water exposure using a radio-labeled pesticide at an environmentally relevant concentration would be preferable as a realistic scenario. Furthermore, the distribution of radioactive residues in each organ, together with the analysis of its components, gives valuable information to examine the relationship of toxic signs with residues of pesticides and their metabolites. Similarly to bioaccumulation, the metabolic activity is considered to vary with developmental stage, sex, and species. These effects require further clarification.
Various immunochemical analyses showed the involvement of enzyme isoforms in metabolic transformations with many similarities to other species. Most studies on relevant enzymes are focused on oxidases (CYP) and GST from the viewpoints of detoxification and oxidative stress, followed by esterases in relation to neurotoxic actions of OP and carbamate pesticides. Less information is available for the SULT and UGT that play great roles in detoxification of pesticides as the phase-II reactions. The contribution of these enzymes is likely to be species-dependent, but it has yet to be properly investigated. Although the existence and some activities are reported, very limited information is available for reductases and methyl- and acyl-transferases in relation to the metabolism of xenobiotics. Not only their enzymology but also their importance in pesticide metabolism in the frog should be further investigated. The clarified amino acid sequences of some CYP and GST greatly helped to elucidate their substrate specificity, and further enzymological investigation with species differences is needed for the other enzymes. The identification of relevant metabolites catalyzed by CaE and GST is considered to be insufficient, and neither GSH conjugates nor succeeding metabolites such as N-acetyl-cysteine have been confirmed. The formation of the Glu and Sulf conjugates was indirectly confirmed by identifying phase-I metabolites released via treatment of sulfatases and glucuronidases in a few studies. Therefore, more intense application of LC-MS with various ionization methods and interfaces is highly desired in order to provide direct evidence of free and conjugated metabolites.
