Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The food supply of this country is extensively regulated at the federal level and considered to be quite safe relative to most other countries. However, the food supply contains several thousand additives in trace amounts. The additives include preservatives and nutritional supplements, as well as veterinary pharmaceuticals and feed additives, in the case of animal products. Eighty percent or more of consumable animal protein in this country comes from animals exposed to "medicated" feeds for part of their lives (Hayes and Campbell, 1986).

Organoarsenicals such as roxarsone (4-hydroxy-3-nitrobenzenearsonic acid)¹ and p-arsanilic acid (4-aminobenzenearsonic acid) can be used to control bacterial and parasitic infections and to promote growth and feed efficiency in animals. These compounds have been used as feed additives in poultry and in swine since the 1960s. The arsenicals nitarsone (4-nitrobenzenearsonic acid) and p-ureidobenzenearsonic acid are used therapeutically in turkeys. All four of these compounds have been approved for use by the U.S. Food and Drug Administration. U.S. Food and Drug Administration regulations require that the arsenical feed additives or medications be terminated at least 5 days before slaughter of the animal, to allow tissue levels of arsenic to drop to allowable levels of 0.5 ppm for fresh, uncooked muscle and 2.0 ppm for fresh uncooked by-products (Calvert, 1975; Calvert and Smith, 1980; Aschbacher and Feil, 1991). The allowable tissue levels listed translate to 250 µg/lb muscle and 1.00 mg/lb liver for benzenearsonate content.

Structural formulae are given for several benzenearsonic acids in Fig. 1. These benzenearsonates are chemically reactive compounds, with their phenolic, aromatic amine, and aromatic nitro groups. Chronic exposure of humans to such compounds in the food supply is a matter of serious concern because structurally similar compounds are metabolized to active electrophiles that react with protein and DNA molecules to form tissue toxins or mutagens (Long and Rickert, 1982; Vineis et al., 1994).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Structural formulae of several benzenearsonic acids.

Older whole animal experiments (Moody and Williams, 1964) revealed excretion of N-acetylated arsanilate, and reduced, N-acetylated roxarsone (4-hydroxy-3-N-acetylbenzenearsonic acid, acetarsone), both more toxic than arsanilate and roxarsone (Donoghue et al., 1994; Merck Index, 1996; Rath et al., 1998). The National Toxicology Program data on carcinogenicity of roxarsone and arsanilate (Abdo et al., 1989; Ashby and Tennant, 1991) raise concerns because of some pancreatic tumors in male rats exposed to roxarsone long-term, and because of positive results from arsanilate and roxarsone in Salmonella mutagenicity tests. Inclusion of benzenearsonates in feed to enhance animal growth rate could pose a significant human risk if the potential for activation by phase I and phase II enzymes is realized. Individuals vary considerably in their capacity to metabolically activate or deactivate compounds, due to genetic differences in levels of basal expression of bioactivation enzymes. There is more individual variation in levels of metabolites generated from xenobiotics because many of these enzymes are inducible (Parkinson, 1996). In spite of these considerations, a Federal Register summary notice dated April 1998 gives approval for continued use of arsanilate (21 CFR 558.62, 1998) and roxarsone (21 CFR 558.530, 1998) as feed additives.

The N-acetyltransferase (NAT) enzyme (EC 2.3.1.5.) represents one of the phase II conjugation enzymes (Evans, 1992) that catalyze N-acetylation of aromatic amine substrates, structurally similar to arsanilate. The enzyme is cytosolic, acetyl coenzyme A (acetylCoA)-dependent and has sensitive sulfhydryl groups. There are substantial species differences in levels of this enzyme. Humans have two distinct but related forms, NAT1 and NAT2, and also exhibit polymorphism, with "slow-acetylators" and "rapid-acetylators" of both forms identifiable in the population (Evans, 1992; Bell et al., 1995).

Numerous whole animal studies have been published on arsanilic acid and similar feed additives, but an in vitro system to measure the biotransformation of these compounds has not yet been reported. Because purified human source NAT enzyme is not available commercially, a commercially available source for pigeon liver was used, and in addition, the supernatant fraction from guinea pig liver homogenate was isolated as a source of mammalian enzyme (Weber and King, 1981). Recombinant human NAT1 and NAT2, kindly supplied by a colleague, were tested as well.

The long-term objectives of this project are to characterize the metabolism and assess the potential toxic effects on humans of the benzenearsonate additives in poultry and swine feed. Such information, currently lacking in the literature, would provide data for consideration of potential metabolic or disease impact in humans on ingestion of these compounds. With an HPLC separation system and the in vitro assay conditions now defined, enzyme preparations from human and other mammalian sources can be tested for NAT activity toward arsanilic acid and related feed additives.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. AcetylCoA, arsanilic acid, roxarsone, acetarsone, p-aminobenzoic acid (PABA), 4-N-acetyl PABA, dithiothreitol (DTT), EDTA, phosphotransacetylase, acetyl phosphate, and pigeon liver arylamine acetyltransferase were obtained from Sigma Chemical Co. (St. Louis, MO). Nitarsone was purchased from Pfaltz and Bauer, Inc. (Waterbury, CT). Bradford reagent was purchased from Bio-Rad Laboratories (Hercules, CA). 4-N-acetylarsanilic acid (NAA) was synthesized by Dr. Cornelia Gillyard and her research group at Spelman College, by reacting acetic anhydride with arsanilic acid, in a modification of a method for acetanilide preparation (Moore et al., 1982). Dr. Eyerce Armstrong and Mr. Michael Halfhill at DuPont Central Research and Development evaluated the recrystallized product both by proton NMR and by ¹³C NMR on a 400-mHz instrument, and concluded that the synthesized product (NAA) is 96% pure. Chemicals were dissolved in line-demineralized water, unless otherwise indicated. Phosphate buffer and sulfuric acid were obtained from Fisher Scientific Co. (Norcross, GA). Water used for the HPLC mobile phase of 5.0 mM sulfuric acid was HPLC grade from Jackson and Burdick (Muskegon, MI).

Supplies. Polysulfone 0.45 µm membrane syringe filters were obtained from Whatman, Inc. (Clifton, NJ).

Equipment. The HPLC system was obtained from ThermoSeparation Products (Riviera Beach, FL) and used a 100-µl injection loop and a variable wavelength UV detector. The LC Talk software was mounted on a Dell 486 P/50 computer with a Hewlett-Packard LaserJet 5L printer.

Spectral Data. Benzenearsonic acids were dried overnight at 110°C, then prepared analytically as duplicate 1.0 mM aqueous solutions. Duplicate dilutions to 0.05 mM were made with the HPLC mobile phase, and the spectra were determined in quartz cells of 1-cm light path, using mobile phase as a blank.

HPLC. The column used was obtained from Interaction Chromatography, Inc. (San Jose, CA), now Transgenic (Omaha, NE) and was an ARH-601 Weak Organic Acids column, 6.5 mm × 100 mm heated at 45°C. An isocratic mobile phase of 5.0 mM sulfuric acid was used at a flow rate of 0.6 ml/min. NAA was detected at 256 nm with a retention time of 33 min. 4-N-acetyl PABA was detected at 269 nm with a retention time of 42 min. NAA and acetylated PABA are well separated from other assay components. The retention times in minutes for the coenzymes/cofactors are EDTA, 1.6; acetylCoA, 1.6, 1.8; and DTT, 10.1.

Guinea Pig Liver NAT Activity. NAT activity was determined in guinea pig liver in the supernatant fraction from homogenate prepared by a literature method (Weber and King, 1981). The livers were provided by Dr. Pamela Gunter-Smith of the Biology Department at Spelman College, and were obtained immediately from animals sacrificed by CO₂ asphyxiation. The enzyme was stored at 80°C. The proportion of NAT1 and NAT2 in the preparation is not known. Protein concentration was determined by the Bradford method (Bradford, 1976) using BSA as a standard.

Recombinant Human NAT1 and NAT2. Escherichia coli lysates containing recombinant human NAT1 and NAT2 were kindly provided by Dr. David W. Hein, Department of Pharmacology and Toxicology, University of Louisville School of Medicine (Louisville, KY). These human enzymes were expressed in pKK223-3 vector and E. coli strain JM1055. Lysates were centrifuged at 15,000g for 20 min. The resulting supernatant was used as a source of NAT activity.

NAT Assay Conditions. The N-acetylation of arsanilic acid was assayed in vitro by modifying a literature method (Ward et al., 1995) for acetylation of PABA. Conditions included final concentrations of 1.0 mM DTT, 1.0 mM EDTA, 0.45 mM acetylCoA, an acetylCoA regenerating system using 0.1 U bacterial phosphotransacetylase with 5.0 mM acetyl phosphate, and 5.0 mM arsanilate substrate, in 25 mM sodium/potassium phosphate buffer, pH 7.4, in a total volume of 0.5 ml. Incubation was at 37°C for 45 to 60 min, with 0.5 to 2.0 mg of NAT protein from a preparation of guinea pig liver. With commercial pigeon liver NAT (arylamine acetyltransferase), 0.5 U was used. Using recombinant human NAT1 and NAT2, incubation was for 30 min with 0.2 to 2 µg, and 30 to 300 µg, respectively.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

UV Absorption Characteristics and HPLC Retention Times. Table 1 gives the low pH  maxima values and millimolar absorptivity values for five benzenearsonic acids. Arsanilate and two potential arsanilate metabolites, one oxidized (nitarsone), the other acetylated (NAA), are included. Values are given as well for roxarsone and its reduced, acetylated metabolite, acetarsone. Table 1 also gives HPLC retention times for the benzenearsonic acids. NAA is well separated from other assay components, (EDTA, acetylCoA/CoA, DTT) although its 33-min retention time is relatively long. The nature of the column material precludes a higher flow rate. Both NAT substrates, arsanilic acid and PABA, are noneluting in this system. However, both ortho amino derivatives do elute from the column.

Pigeon Liver NAT. Pigeon liver NAT activity was assayed by a literature method (Ward et al., 1995) using PABA as substrate. Arsanilic acid was unreactive as a substrate in this system.

Guinea Pig Liver NAT Activity. Using the assay conditions outlined in Materials and Methods for acetylation of arsanilic acid, the production of NAA with this mammalian enzyme source was linear with time over a 1.0-h incubation period, and linear with enzyme concentration at least to 2 mg of enzyme protein per assay. These results are shown in Figs. 2 and 3, respectively. Preliminary experiments indicate a pH optimum of 7.4 and a substrate saturation level of 5 mM for arsanilate acetylation.

View larger version (10K): [in this window] [in a new window]   Fig. 2.   NAA produced from arsanilate versus time. Arsanilate (5 mM) was incubated in phosphate buffer at pH 7.4 with an acetylCoA regenerating system and the guinea pig liver NAT preparation (2.1 mg). Product NAA is plotted as the mean of duplicate reaction mixtures versus time.

View larger version (10K): [in this window] [in a new window]   Fig. 3.   NAA produced from arsanilate versus enzyme concentration. Arsanilate (5 mM) was incubated for 45 min in phosphate buffer at pH 7.4 with an acetylCoA regenerating system and the guinea pig liver NAT preparation (0.5-2.1 mg). Product NAA is plotted as the mean of duplicate reaction mixtures versus mg of enzyme protein.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Representative HPLC chromatograph of NAA produced from arsanilate with and without guinea pig liver NAT preparation. Arsanilate (5 mM) was incubated for 45 min in phosphate buffer at pH 7.4 with an acetylCoA regenerating system and the guinea pig liver NAT preparation (2.1 mg). Top, no enzyme; bottom, plus enzyme.

Recombinant Human NAT 1 and NAT2. The supernatant from bacterial lysates containing recombinant NAT 1 and NAT 2 was incubated with arsanilate and PABA under conditions given in the NAT assay procedure. Figure 5. gives the chromatograph for arsanilate acetylation by recombinant human NAT1, with a nonenzyme control (top) and a NAT1-containing reaction mixture (bottom). Table 2. summarizes the specific activities of the NAT enzyme sources tested with arsanilate as substrate. Activity toward PABA is included for comparison.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Representative HPLC chromatograph of NAA produced from arsanilate with and without recombinant human NAT1 enzyme. Arsanilate (5 mM) was incubated for 30 min in phosphate buffer at pH 7.4 with an acetylCoA regenerating system and supernatant from bacterial lysate containing recombinant human NAT1 (4.4 µg). Top, no enzyme; bottom, plus enzyme.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

UV spectral characteristics for five benzenearsonic acids were determined at low pH (2.2), the pH of the HPLC mobile phase used. These compounds undergo a red shift with increasing pH and thus should be reassessed for wavelength maxima and molar absorptivity when used at other pH values. This allows for choice of the most sensitive/feasible wavelength and concentration for use in the HPLC system.

The HPLC system presented here with the ARH-601 column separates several benzenearsonic acids and their metabolites, making it valuable in general for the study of the benzenearsonate feed additives. With the NAT assay components, coenzymes and cofactors elute early, giving excellent separation from the acetylated products NAA and N-acetyl PABA.

An in vitro assay system was modified from the literature (Ward et al., 1995) and used to follow N-acetylation of arsanilate by NAT activity in the supernatant fraction of guinea pig liver. The NAA enzymatic product identity was confirmed by comparison of HPLC retention time with standard NAA. UV spectral analysis of the collected product and comigration in the HPLC system with standard NAA gave additional proof of product identity.

Arsanilate was also acetylated by recombinant human NAT1, at about 15% the rate at which PABA was acetylated, as shown in Table 2. Both arsanilate and PABA were unreactive as substrates with NAT2. Arsanilate does not serve as a substrate for pigeon liver NAT. This result is in agreement with the work on colostomized roosters reported by Aschbacher and Feil in 1991. Roosters were fed ring-labeled [¹⁴C]arsanilate, and excreted about 90% of the arsanilate dose unchanged. Highest tissue levels of the remaining radioactivity were found in liver, but identification of the specific metabolites was not performed. Avian species, then, appear unable to acetylate arsanilate.

In this study, pigeon liver NAT, the guinea pig liver cytosol preparation, and recombinant human NAT1 all were able to acetylate PABA. The guinea pig preparation and human NAT1 were able to acetylate arsanilate. Human NAT2 was unreactive toward both substrates.

Although arylamine acetylation is considered a detoxification reaction, subsequent N-hydroxylation by cytochrome P-450 or flavin monooxygenase (FMO) enzymes, followed by deacetylation, would produce a reactive nitrene, which can bind to DNA and protein. Likewise, N-hydroxylation of arsanilate, followed by N-acetylation, then deacylation, would produce the reactive nitrenium ion (Parkinson, 1996).

Investigators are actively searching for linkages between arylamine exposure from environmental and dietary sources, and increased risk of cancer due to genetic makeup (Hein, 1988; Kirlin et al., 1991; Bell et al., 1995). NAT1 polymorphism is now recognized in the population and variant alleles continue to be described (Doll et al., 1997; Payton and Sim, 1998). With the HPLC system and the in vitro assay system for acetylation described here, other human NAT1 variants can be tested for acetylation activity toward arsanilate.