Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Emamectin benzoate (MK-0244) is the benzoate salt of the 4"-deoxy-4"-epimethylamino derivative of abamectin, a macrocyclic lactone produced by the soil actinomycete Streptomyces avermitilis. It is currently being developed as a wide-spectrum Lepidopteran larvicide for use on leafy and fruiting vegetables, on cole crops, and on cotton. Emamectin benzoate is effective at extremely low use rates (<0.015 lb of active ingredient/acre).

Emamectin benzoate is specified as a mixture of two homologous compounds, i.e. a minimum of >90% MAB1a² and a maximum of <10% MAB1b; these differ only by a methylene group on the C25 side chain (fig. 1). Because the structural difference between the two homologues is small and because comparative metabolism studies of MAB1a and MAB1b using rat liver slices have demonstrated homologous metabolism (Mushtaq M, unpublished results), it is likely that the metabolism of MAB1b would also be homologous to that of MAB1a in chickens.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Structure of emamectin benzoate (MK-0244) and positions of the radiolabels.

In previous emamectin benzoate metabolism studies in rats and goats, >99% of the administered dose was recovered in the excreta, with correspondingly low tissue residue levels (Mushtaq et al., 1996, 1997). In those studies, the parent compound was the major residue (85-95% of TRR) in tissues and feces, and the only identified metabolite was the N-demethylated derivative of the parent, AB1a (5-15% of TRR). The purpose of this study was to determine the egg and tissue distribution, excretion, and metabolism of emamectin benzoate and its metabolites in laying hens after oral administration of the major homologue, MAB1a. Such results can be used to evaluate the potential for human exposure to emamectin benzoate residues from the consumption of eggs and tissue from chickens whose diet contained treated crops. The present results indicate that the metabolism of emamectin benzoate in chickens is remarkably different from that in mammals.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Test and Reference Chemicals. Ethanolic solutions of 5-[³H]MAB1a benzoate (16.80 mCi/mg, 16.94 mCi/mmol) and 25-[¹⁴C]MAB1a benzoate (20.86 µCi/mg, 20.69 µCi/mmol) were supplied by the Labeled Compound Synthesis Group of Merck Research Laboratories. The radiochemical purity of both chemicals was >97%. The ethanolic solutions of [³H]MAB1a benzoate and [¹⁴C]MAB1a benzoate were combined to yield 8 ml of a 15.2 mg/ml [³H/¹⁴C]MAB1a benzoate test solution. Unlabeled emamectin benzoate (MK-0244) and AB1 (including both the AB1a and AB1b homologues) were used as reference standards and were supplied by Merck Research Laboratories.

Preparation of the Dosing Capsules. Gelatin capsules (size 3 white ELC/ELC; Shionogi Corp., Whitsett, NC) were half-filled with cellulose powder, and a 100-µl aliquot of either ethanol (control) or the ethanolic test chemical solution (treated) was added to each capsule. The capsules were kept at room temperature in a hood, to allow the ethanol to evaporate, and were then completely filled with cellulose powder and sealed. Approximately 100 mg of cellulose was added to each capsule. The sealed capsules were then placed inside a larger capsule (size 2 clear AA/AA) for administration.

Handling and Dosing of Chickens. Twenty top-quality, laying, Leghorn hens (Gallus domesticus) were procured from Avian Services (Frenchtown, NJ). They were each approximately 61 weeks of age, weighed between 1 and 2 kg, and were uniquely identified by wing band. Chickens were individually housed in galvanized metal cages (multi-stacking poultry laying cages) with ad libitum access to food (Purina Layena) and water. Daily food consumption was monitored by feed weight for each chicken. The chickens were allowed to acclimate to the metabolism cages, diet, and handling procedures for 14 days before the start of dosing and were maintained in a temperature-controlled room at 72°F, with 16 hr of light provided daily with fluorescent lights. After approximately 1 week of acclimation, healthy and actively laying chickens were randomly assigned to test (10 chickens) and control (5 chickens) groups. Chickens were dosed once daily (at approximately 1 mg/kg of body weight) for 7 consecutive days with capsules and were euthanized by carbon dioxide inhalation approximately 20 hr after administration of the last dose. The daily dose corresponded to approximately 10 ppm of emamectin benzoate in the diet, as calculated from food consumption.

Sample Collection. Excreta were collected and composited once daily from all treated chickens. Eggs were collected twice daily from the treated chickens and were separated into whites and yolks. Excreta and eggs were collected from the control chickens before euthanasia only. Liver, kidney, heart, gizzard, GIT, ovaries (immature eggs), muscle (thigh and breast), fat (abdominal and muscle fat, with adhering skin), and carcass (minus head, feet, wings, and feathers) specimens were collected from each chicken. The interiors of the gizzards were flushed with water, and the contents were combined with the GIT specimen. After removal of the chickens, the cages housing the treated chickens were washed with soap and water and an aliquot of the wash was analyzed for radioactivity. All specimens were frozen after collection and were kept frozen except when aliquots were removed for analysis.

Sample Homogenization and TRR Determination. Immediately after collection, the composite excreta samples were homogenized with water (1:1, w/v) in a Waring blender. GIT were composited into three samples (two treated and one control, each containing the GIT from five chickens) and likewise homogenized with water (1:1, w/v) in a Waring blender. In addition, water (1:1-3, w/v) was added to individual liver, kidney, heart, gizzard, egg yolk, and ovary specimens, and these were then homogenized with a Brinkman Polytron homogenizer (Brinkman Instruments, Westbury, NY). Egg white specimens were homogenized using a Brinkman Polytron homogenizer without the addition of water. Muscle and carcass specimens were chopped into small pieces (while still partially frozen) using a meat cleaver and were then passed three times through a Fleetwood meat grinder (W. W. Lowenstein, Inc., Newark, NJ). Fat specimens were homogenized by thorough chopping and mixing with a meat cleaver. Aliquots of tissue and excreta homogenates (0.2-1 g) were then assayed in replicate (three or more determinations) for TRR by radiocombustion analysis using a Packard model 307 sample oxidizer, followed by LSC. The TRR values for egg whites, egg yolks, and ovaries were determined by direct LSC of triplicate aliquots (0.2-0.4 g) of each specimen. The accuracy of direct LSC of these specimens was demonstrated by a comparison of results obtained by radiocombustion analysis followed by LSC with those obtained by direct LSC of composite samples (data not shown).

Sample Extraction. Tissue homogenates (liver, muscle, fat, and ovaries) were composited by combining equal amounts of each individual homogenized specimen. Aliquots of egg yolk homogenates from all eggs collected in the study were similarly composited and were additionally composited according to treatment day, as for excreta. Aliquots of composited tissue, excreta, and egg yolk homogenates (1-2 g) were extracted either with acetone/EtOAc (1:1, v/v) or with acetone/5% aqueous sodium chloride (1:1, v/v) (fat only), the extracts were partitioned with EtOAc, and the organic layers from this partitioning were fractionated by cation exchange SPE, as described in detail elsewhere (Mushtaq et al., 1996, 1997). The EtOAc/NH₃ residue fractions from the SPE fractionation contained 80-95% of the total radioactivity in the original samples (data not shown). Initial attempts were made to analyze the EtOAc/NH₃ SPE fractions by reverse-phase HPLC using a C₁₈ column. However, for all samples except excreta, recoveries were generally unacceptably low, ranging from 30 to 90% (data not shown). Column recoveries were improved to >95% using normal-phase HPLC (method 1; table 1). Therefore, suitable aliquots from the EtOAc/NH₃ SPE fractions were mixed with unlabeled AB1 and MK-0244 standards, dried under N₂, and reconstituted in approximately 100 µl of methanol. Triplicate 10-µl aliquots were counted directly for recovery calculations, and the remaining solution (~70 µl) was analyzed by normal-phase HPLC (method 1).

Metabolite Isolation. Approximately 50 g of composite excreta homogenate was extracted with acetone/EtOAc, and the organic extract was fractionated by cation exchange SPE. After concentration, the EtOAc/NH₃ SPE fraction was fractionated by semipreparative HPLC (method 2; table 1), and selected fractions were pooled into four crude residue fractions. The crude 24-OH-MAB1a and 24-OH-AB1a fractions were combined and refractionated by HPLC method 4. The collected fractions containing the 24-OH-MAB1a and 24-OH-AB1a metabolites from this second HPLC fractionation were then individually purified by HPLC method 1. The two purified residues were then analyzed by NMR spectrometry and/or MS. Similarly, the crude MAB1a and AB1a fractions from semipreparative HPLC were subjected to additional HPLC separation. Nonlabeled MK-0244 and AB1 standards were added as appropriate, and the crude MAB1a and AB1a residue fractions were refractionated by HPLC method 3. The individual residues coeluting with the respective added standards were then analyzed by HPLC method 1. The identities of MAB1a and AB1a were thus confirmed by cochromatography with reference standards, using two HPLC methods with different selectivities.

Lipase Hydrolysis. To each of two 20-ml glass vials was added an aliquot of the EtOAc/NH₃ SPE fraction from the liver with the highest TRR level, containing approximately 0.7 µg of parent equivalent. The aliquots were concentrated to dryness under N₂ and resuspended in 5 ml of phosphate buffer (pH 7.1) containing 0.5 ml of Triton X-100 (Sigma Chemical Co.). Lipase (from Chromobacterium viscosum, 0.5 mg, 1790 units of activity; Sigma) was added to one of the vials, with the second serving as a control. Both vials were incubated for approximately 16 hr at 37°C. After incubation, the contents of both vials were extracted three times, by volume, with methylene chloride. The methylene chloride extracts were then analyzed by HPLC method 1 (table 1).

Lyophilization. Aliquots (~2 g) of composite egg white homogenates from each treatment day were lyophilized in 20-ml glass scintillation vials. To verify the operation of the equipment, an aliquot of a nontreated egg white homogenate was spiked with ³H₂O standard and lyophilized. After lyophilization, the trapped water and the lyophilized egg white sample were analyzed by LSC. Lyophilized egg white residues were resuspended in approximately 1 ml of water before LSC.

HPLC Instrumentation. HPLC was performed using two separate systems. The first system consisted of an IBM 350-P100 personal computer loaded with Thermo Separations UV3000 detector system software version 3.0, a Spectra-Physics SP8700 gradient controller, a Spectra-Physics UV3000 scanning detector, and either a Pharmacia Frac-100 fraction collector or an ISCO Foxy fraction collector. The second system consisted of a Hewlett-Packard Vectra 486/66XM computer loaded with ^3DChem Station software, a Spectra-Physics model 8700 gradient controller, a Hewlett-Packard 1040A diode-array detector, and an ISCO Foxy 200 fraction collector. For both systems, a Rheodyne 7125 injector was used with either a Brownlee reverse-phase or Brownlee normal-phase guard column, as appropriate. Absorbance was continuously monitored at 245 nm and other wavelengths as appropriate. One-minute (1-3-ml) fractions were collected, and total radioactivity in each fraction or an aliquot thereof was determined by LSC. Summaries of the HPLC methods used in this study are provided in table 1.

MS Analysis. MS analysis of isolated metabolites was performed with a Finnigan LCQ quadrupole ion-trap mass spectrometer, using either APCI or electrospray ionization, in positive-ion mode. Nitrogen/helium was the carrier gas. An MK-0244 standard was used to calibrate the instrument. Because both modes of ionization are soft, typically only the protonated parent peaks (M+H) were observed.

NMR Analysis. Proton NMR analysis was performed at 400 MHz using a Varian Unity-400 spectrometer. Isolated metabolites were dissolved in CDCl₃, and analysis was conducted at room temperature using a 1-sec acquisition time and a 45° flip angle.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Excretion of Radioactive Residues. The majority of the total administered dose (72.0/66.9% of ³H/¹⁴C) and approximately 92% of the total recovered radioactivity (both labels) were accounted for in the excreta, GIT, and cage washes (table 2). An average of approximately 13 ± 1.6% of the total recovered radioactivity was recovered in the excreta for each treatment day.

Tissue Distribution of Radioactive Residues. The means of the individual tissue ³H/¹⁴C TRR levels are presented in table 3. The liver and the ovaries had the highest TRR levels (3.149/3.088 and 2.677/2.170 ppm of ³H/¹⁴C, respectively), and TRR levels in all other tissues were <1 ppm. Tissue TRR levels were in the following order: liver > ovaries > kidney  abdominal fat > muscle fat/skin  heart > gizzard > thigh muscle > breast muscle. Tissue TRR levels in one chicken were generally several times higher than the average residue levels in the other nine chickens (data not shown). This disparity was most evident in the liver and in the ovaries, where the TRR levels for this chicken were 12.726/12.204 and 14.279/11.619 ppm of ³H/¹⁴C, respectively, whereas the mean TRR levels in these two tissues for the other nine chickens were 2.084/2.075 and 1.388/1.120 ppm of ³H/¹⁴C, respectively. Average TRR levels in egg yolks generally increased with treatment day to about 3 ppm by day 7, except for a decrease from day 5 to day 6 (fig. 2). This decrease corresponded to and likely resulted from the lack of egg production by the chicken with the highest residue levels on treatment day 6. TRR levels in egg white were approximately 100 times less than those in the yolk. Although ¹⁴C TRR levels in egg whites remained fairly constant at approximately 4 ppb after 3 days, ³H TRR levels consistently increased with time, rising from a mean of 2 ppb in individual specimens collected after application of the first dose to a mean of 21 ppb in individual specimens collected after application of the last dose (fig. 2). Although lyophilization demonstrated that this disparity between the two labels in egg whites was accounted for by the accumulation of ³H₂O (data not shown), the total residues in egg whites were <0.1% of the total recovered radioactivity (table 2). Because no differences were observed in ³H and ¹⁴C TRR levels and ³H and ¹⁴C radioprofiles of extractable residues for any other specimens examined, no other specimen types were lyophilized. Because the H₂O seen in egg whites represented such a minor fraction of the total recovered radioactivity and because there was no evidence for accumulation of ³H₂O in other specimen types, it can be concluded that the ³H label was stable in this study. Approximately 6% of the total recovered radioactivity was associated with tissues and carcass, and approximately 1.5% was associated with egg yolk (table 2).

View larger version (31K): [in this window] [in a new window]   Fig. 2.   TRR in egg whites and egg yolks vs. treatment day. The ppm values represent the average TRR levels in egg white and egg yolk specimens for each treatment day.

Extraction and Metabolic Profiles in Tissues, Excreta, and Egg Yolks. The extractable residues in the tissues, excreta, and egg yolks ranged from 80 to 95% in each type of sample, with the remaining radioactivity being distributed approximately evenly between the other two or three fractions that were generated (data not shown). The metabolic profiles of composited excreta samples from days 1, 3, 6, and 7 were essentially identical (table 4). The major residues in excreta were parent MAB1a (~52%) and 24-OH-MAB1a (~33%) (table 4). Additionally, minor amounts (1-5%) of the N-demethylated metabolites of parent and 24-OH-MAB1a (AB1a and 24-OH-AB1a, respectively) were identified.

View larger version (37K): [in this window] [in a new window]   Fig. 3.   HPLC (method 1) analysis of residues from the liver with the highest TRR level, before (A) and after (B) lipase treatment, and from the liver with the lowest TRR level, without lipase treatment (C). Note that 14C values are offset for clarity.

Isolation and Identification of Residues from Excreta. The procedures used to isolate MAB1a, AB1a, 24-OH-MAB1a, and 24-OH-AB1a from a composite excreta sample and to subsequently identify these residues are summarized in fig. 4. The radiochromatogram from the HPLC (method 2) fractionation of the EtOAc/NH₃ SPE fraction of composite excreta is presented in fig. 5. MAB1a and the AB1a metabolite were identified by coelution with analytical standards using two HPLC methods, with different selectivities. The 24-OH-MAB1a and 24-OH-AB1a metabolites were identified by NMR and/or MS analyses. The APCI MS analysis of 24-OH-MAB1a resulted in an M+H peak of m/z 901.5, which is 16 units greater than that of MAB1a (m/z 886) and is consistent with monohydroxylation of MAB1a (table 5). NMR analysis of the metabolite confirmed its identification as the 24-hydroxymethyl analogue of MAB1a. Key observations were the absence of one of the two methyl doublets near 0.9 ppm and the presence of two new protons, at 3.54 and 3.73 ppm (fig. 6). These findings clearly pointed to hydroxylation of either the 24a- or 26a-methyl group. Irradiation of H24 did not affect the methyl doublet at 0.95 ppm, thus establishing that the derivatized methyl is in the 24a-position. The APCI MS analysis of the 24-OH-AB1a residue isolated from excreta resulted in a M+H peak at m/z 888.2 (table 5), which is 2 mass units more than that of MAB1a and is consistent with monohydroxylation (+16 units) and N-demethylation (14 units). Confirmatory NMR analysis was not performed for this isolated metabolite.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Summary of procedures used to isolate residues from composite excreta samples. ESI, electrospray ionization.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   HPLC (method 2) fractionation of a composite excreta sample. Note that 14C values are offset for clarity.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   1H NMR analysis of MAB1a (A) and 24-OH-MAB1a (B).

Isolation and Identification of Fatty Acid Conjugates from Liver. As for excreta (fig. 4), a 40-g aliquot of the liver with the highest TRR level was extracted with solvent and fractionated by SPE. The resulting EtOAc/NH₃ SPE fraction was subsequently fractionated using HPLC method 1 (table 1). The unresolved fatty acid conjugates of 24-OH-MAB1a and of 24-OH-AB1a (fig. 3A) were separated into seven metabolite peaks each by HPLC method 5 (table 1 and fig. 7). Although the fatty acid conjugates could not be eluted from C₁₈ columns, as previously noted, there was complete recovery using a C₈ column, as in HPLC method 5. Note that the myristic and linolenic acid conjugates eluted as a single peak with this method (fig. 7, peak A). The purified fatty acid conjugates were subsequently identified by MS analysis, and the molecular ion data from APCI and/or electrospray ionization MS analysis of the 14 residue peaks are summarized in table 5. NMR analysis of the 24-OH-MAB1a oleate residue (data not shown) strongly supported its identification as a 24-OH-MAB1a fatty acid conjugate. Key observations were the presence of characteristic signals associated with a long-chain fatty acid, the absence of the 24a-methyl group, and the presence of signals associated with a CH₂O grouping in the 24a-position. Attachment of the oleic acid was indicated by perturbation of nearby signals, notably H24, H22, H23, and 24a-CH₂. The 0.3-0.4-ppm downfield displacement of the 24a-methylene group, relative to 24a-CH₂OH, typically results from esterification. However, the presence of some hydrocarbon contamination in the crucial region, together with contributions from the macrocycle, precluded conclusive identification of the oleic acid portion of the molecule.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   HPLC (method 5) separation of 24-OH-MAB1a fatty acid conjugates for identification. The eight fatty acid conjugates of 24-OH-MAB1a purified by method 1 (fig. 3A) were resolved into seven residue peaks by HPLC method 5 (table 1). The fatty acids conjugated to 24-OH-MAB1a were myristic acid and linolenic acid (peak A), palmitoleic acid (peak B), linoleic acid (peak C), palmitic acid (peak D), oleic acid (peak E), stearic acid (peak F), and gadoleic acid (peak G). Fatty acid conjugates were subsequently identified by MS analysis. Note that 14C values are offset for clarity.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The in vivo metabolism of MAB1a benzoate in chickens is summarized in fig. 8. In previous studies in rats and lactating goats, MAB1a was not extensively metabolized; the N-demethylated derivative of MAB1a (AB1a) was the only reported metabolite, ranging from 5 to 15% in tissues, milk (goat only), and feces (Mushtaq et al., 1996, 1997). However, in chickens MAB1a benzoate was extensively metabolized, with the major metabolite being the 24-hydroxymethyl derivative of MAB1a (24-OH-MAB1a). This metabolite accounted for approximately 33% of excreta TRR (table 4). Smaller amounts of AB1a (2-7% of TRR in excreta, tissues, and egg yolks) and the N-demethylated derivative of 24-OH-MAB1a (24-OH-AB1a, ~1% of TRR in excreta only) were also identified in chickens. Hydroxylation of the 24-methyl group was reported for the structurally similar abamectin in rats and lactating goats (Maynard et al., 1989, 1990) and for ivermectin in steers, sheep, and rats (Chiu et al., 1987). However, excreta and/or tissue levels of 24-hydroxymethylavermectin were only 3-12% of the TRR, and liver levels of 24-hydroxymethyl ivermectin were only 5-17% of the TRR.

View larger version (26K): [in this window] [in a new window]   Fig. 8.   Summary of the in vivo metabolism of MAB1a benzoate in chickens.

Zeng et al. (1996) have demonstrated that in rat liver microsomes cytochrome P450 1A1 plays a major role in hydroxylation of the C24-methyl group of abamectin but not ivermectin. Preliminary results also indicate that hydroxylation of the C24-methyl group of emamectin benzoate occurs in rat liver microsomes and that cytochrome P450 1A1 is involved (Mushtaq M, personal communication). If the same isozyme is responsible for this transformation in chickens, then this would indicate that the activity of hepatic cytochrome P450 1A1 in chickens, as measured by hydroxylation of the C24-methyl group of emamectin benzoate, is greater than in rats. Previous studies have demonstrated that cytochrome P450 1A1 activity, as measured by dealkylation of ethoxyresorufin, is higher in avian species than in rats (Amsallem-Holtzman et al., 1997; Walker and Ronis, 1989).

In tissues and egg yolk, nearly all of the 24-OH-MAB1a was present as fatty acid conjugates, ranging from 7 to 70% of TRR, whereas unconjugated 24-OH-MAB1a ranged from 1 to 8% of TRR (table 4). In addition, whereas unconjugated 24-OH-AB1a was not apparent in tissue and egg yolk, fatty acid conjugates of 24-OH-AB1a were found in both specimen types (1-7.5% of TRR). Similar, although much less extensive, fatty acid conjugation of the 24-hydroxymethyl metabolite of ivermectin in the fat tissue of cattle, sheep, and rats was reported by Chiu et al. (1988). This result is consistent with adipose tissue being the primary site of lipogenesis in cattle and sheep and one of the primary sites in rats (Ballard et al., 1969; Hood et al., 1972; Ingle et al., 1972a,b; Pullen et al., 1990). In contrast, because lipogenesis occurs primarily in the liver of chickens (Pullen et al., 1990; O'Hea and Leveille, 1969), conjugation of 24-OH-MAB1a most likely also occurred there.

In laying chickens, the lipids of egg yolk are synthesized in the liver and are subsequently transported in the blood to the ovaries (McLelland, 1991; Schneider, 1995). More than 30% of egg yolk weight consists of lipids, and up to 2 g of lipid per day are deposited in growing oocytes (Schneider, 1991, 1995). Therefore, the activity of enzymes, such as esterases, that are associated with fatty acid synthesis would be expected to be high in chicken livers. Because fatty acid conjugation of xenobiotics is thought to be associated with esterases (Ansari et al., 1995), a high level of activity coupled with the extensive hydroxylation of MAB1a that most likely also occurs in the liver would account for the high fatty acid conjugate residue levels observed in liver, ovaries, and egg yolk. A total of eight fatty acid conjugates of 24-OH-MAB1a and eight fatty acid conjugates of 24-OH-AB1a were identified by MS analysis. The relative proportions of the individual fatty acid conjugates, which were isolated from an individual liver specimen, generally matched the relative proportions of fatty acids in chicken liver, as reported by the United States Department of Agriculture (1996) (table 6). This suggests that the observed conjugation of 24-OH-MAB1a and 24-OH-AB1a is fairly nonspecific for fatty acids and possibly for the 4"-substituent of avermectin. In addition, the observed general increase in egg yolk residue levels over the 7-day treatment period perhaps corresponded to the egg development cycle in chickens, which is also approximately 7 days (Schneider, 1995).

In previous metabolism studies, emamectin benzoate was excreted mostly (>98%) in the feces of rats and goats (Mushtaq et al., 1996, 1997), as would be expected for a high-molecular weight compound. Similarly, in chickens, slightly more than 92% of total recovered radioactivity was accounted for in excreta and GIT. However, tissue TRR levels were higher in chickens (~4% of the total recovered radioactivity) (table 2) than in rats or goats (<2% and <1%, respectively) (Mushtaq et al., 1996, 1997). Additionally, ~3.5% of the total recovered radioactivity was accounted for in ovary and egg yolk specimens. These higher TRR levels apparently resulted from the distribution of fatty acid conjugates along with fat. Levels of these conjugates were highest in egg yolks, ovaries, liver, and fat. Moreover, in other tissues residue levels correlated with fat levels. For example, residue levels (table 3) in heart (~9% fat) were greater than in thigh muscle (~4% fat), and thigh muscle levels were greater than levels in breast muscle (~1.5% fat) (United States Department of Agriculture, 1996). Based on these findings, the avian metabolism of MAB1a differs significantly from the mammalian metabolism.

No fatty acid conjugates were found in chicken excreta. This indicates that, after the fatty acid conjugates were formed, either they were not excreted or esterases in the GIT or in gut bacteria cleaved the ester bond of excreted conjugates. Because a high rate of liver lipogenesis is necessary to sustain egg laying and most of the fatty acid conjugates were found in the liver, ovaries, and egg yolks (table 4), it seems likely that chickens would not excrete significant amounts of lipid conjugates. Moreover, fatty acid conjugates of the nonsteroidal anti-inflammatory drug etofenamate were isolated from the feces and urine of dogs (Dell et al., 1982), indicating at least some stability of this type of conjugate to gut bacteria.

Although the in vivo formation of fatty acid conjugates of xenobiotic alcohols is apparently uncommon, there are other examples in the literature. Dell et al. (1982) reported the conjugation of etofenamate to oleic, palmitic, linoleic, stearic, palmitoleic, myristic, and lauric acid in dogs. Leighty et al. (1976, 1980) reported the conjugation of ⁹-tetrahydrocannabinol, ⁸-tetrahydrocannabinol, and 2,2-bis(p-chlorophenyl)-1,1,1-trichloroethane (DDT) to several fatty acids in rats. Pryde and Hänni (1983) reported the conjugation of 2-naphthylmethylcyclopropanecarboxylate to C₁₆, C₁₈, C₂₀, and C₂₂ saturated fatty acids in apples. There have also been several reports of the in vitro conjugation of fatty acids to such xenobiotic alcohols as plaunotol (an antiulcer isoprenoid), 2,2-bis(p-chlorophenyl)ethane (DDOH, a hydroxylated metabolite of DDT), phencyclidine, and codeine (Ikeda, 1988; Leighty et al., 1980; Leighty and Fentiman, 1980, 1983). Additionally, because the fatty acid conjugates reported in this article could not be eluted from a C₁₈ column, additional examples may go unobserved if the extent of conjugation is minor and residue analysis is performed using a C₁₈ column.

In this study, chickens were dosed at a greatly exaggerated level (10 ppm dietary equivalents), relative to that expected from consumption of feed derived from treated crops. This is because emamectin benzoate is applied to crops at very low use rates (<0.015 lb of active ingredient/acre), with residues reported to be <10 ppb by 3 days after application (Prabhu et al., 1991). Because the majority of the residues (>92%) were excreted by the chickens, the potential for human exposure to emamectin benzoate residues from consumption of chicken would be extremely low.