DNA Adduct Formation by 2,6-Dimethyl-, 3,5 ... - ACS Publications

Jul 8, 2006 - Paul L. Skipper,*,† Laura J. Trudel,† Thomas W. Kensler,‡ John D. ... for carcinogenicity of monocylic aromatic amines is limited ...
0 downloads 0 Views 59KB Size
1086

Chem. Res. Toxicol. 2006, 19, 1086-1090

DNA Adduct Formation by 2,6-Dimethyl-, 3,5-Dimethyl-, and 3-Ethylaniline in Vivo in Mice Paul L. Skipper,*,† Laura J. Trudel,† Thomas W. Kensler,‡ John D. Groopman,‡ Patricia A. Egner,‡ Rosa G. Liberman,† Gerald N. Wogan,† and Steven R. Tannenbaum† Biological Engineering DiVision, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, and Department of EnVironmental Health Sciences, Bloomberg School of Public Health, Johns Hopkins UniVersity, Baltimore, Maryland 21205 ReceiVed April 18, 2006

Aromatic amines such as 2-naphthylamine and 4-aminobiphenyl are established human bladder carcinogens. Experimental evidence for carcinogenicity of monocylic aromatic amines is limited mostly to other organs, but a recent epidemiologic study of bladder cancer found that 2,6-dimethyl- (2,6-DMA), 3,5-dimethyl- (3,5-DMA), and 3-ethylaniline (3-EA) may play a significant role in the etiology of this disease in man. The present work was undertaken to test whether a genotoxic mechanism can account for the presumptive activity of 2,6-DMA, 3,5-DMA, and 3-EA by quantifying the binding of these compounds to DNA in vivo. Each of these three [14C]alkylanilines was administered at approximately 100 µg/kg to C57BL/6 mice, which were subsequently sacrificed 2, 4, 8, 16, and 24 h post-dosing. Bladder, colon, kidney, liver, lung, and pancreas were harvested from each animal, and DNA was isolated from each tissue. Adduct levels were determined by quantifying bound isotope using accelerator mass spectrometry. Adducts were detectable in the bladder and liver DNA samples from every animal at every time point at levels that ranged from 3 per 109 to 1.5 per 107 nucleotides. Adduct levels were highest in animals given 3,5-DMA and lowest in those given 3-EA. Levels in both bladder and liver declined by severalfold over the course of the experiment. Adducts were detected less frequently in the other four tissues. Taken together, the results strongly suggest that these three alkylanilines are metabolized in vivo to electrophilic intermediates that covalently bind to DNA and that adducts are formed in the DNA of bladder, which is a putative target organ for these alkylanilines. Introduction Aromatic amines constitute one of the most extensively studied classes of chemical carcinogens as detailed in several recent reviews (1-3). Some of these compounds exhibit associations with human cancer, specifically cancer of the urinary bladder, that are among the strongest observed for chemical exposures. The exposure-disease linkage is nearly indisputable for 4-aminobiphenyl, 2-naphthylamine, and benzidine. Assays with experimental animals have also identified other amines that are strongly carcinogenic, including acetylaminofluorene, 4-aminostilbene, and several aminoazodyes. A common feature of these compounds is that they are composed of more than one aromatic ring system. Evidence for carcinogenicity of monocyclic aromatic amines is relatively modest. Experimental animal models generally demonstrate only weak carcinogenic activity for most compounds. Occupational studies have implicated o-toluidine (4) in an exceptional case of reasonably well-defined exposure. More recently, a nonoccupational epidemiologic study in which exposure to ethyl- and dimethylanilines was determined through hemoglobin adduct assays found that three of these isomeric compounds were significantly and independently associated with bladder cancer incidence in Los Angeles County (5). The three are 2,6-dimethylaniline (2,6-DMA),1 3,5-dimethylaniline (3,5* To whom correspondence should be addressed at 77 Massachusetts Avenue, Building 56, Room 753, Cambridge, MA 02139. E-mail: skipper@ mit.edu. † Massachusetts Institute of Technology. ‡ Johns Hopkins University.

DMA), and 3-ethylaniline (3-EA) (Scheme 1). Little is known regarding the genotoxicity of 3,5-DMA or 3-EA, nor has their DNA binding previously been studied in vivo. 2,6-DMA is a rodent nasal cavity carcinogen (6), categorized by IARC as possibly carcinogenic to humans (group 2B), and it has been shown to form DNA adducts in rats (7). We are unaware of any reports regarding carcinogenicity tests of 3,5-DMA or 3-EA. No further direct chemical evidence has been reported regarding the formation of covalent adducts between monocyclic aromatic amines and DNA in vivo. In principle, there is nothing to preclude that adducts would be formed; many synthetic O-acetyl-N-hydroxylamines, including the one made from 2,6DMA, react with deoxyguanosine to form N-(deoxyguanosin8-yl) adducts (8). Adducts of other nucleobases have also been described (9). Since the O-acetyl-N-hydroxylamine is a potential metabolic product in vivo, it is possible, at least, that DNA adducts could be formed as a result of exposure to the amine. Other potential metabolites such as the O-sulfate of the N-hydroxylamine might also reasonably be expected to form adducts. Evidence that monocyclic amines of varied structure form DNA adducts in vivo would contribute greatly to the biologic plausibility of the epidemiologic findings cited above. The present limited range of such evidence partly reflects the analytical challenge of detecting adducts of unknown structure 1 Abbreviations: AMS, accelerator mass spectrometry; CBI, carcinogen binding index; 2,6-DMA, 2,6-dimethylaniline; 3,5-DMA, 3,5-dimethylaniline; 3-EA, 3-ethylaniline; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)1-butanone; LSC, liquid scintillation counting.

10.1021/tx060082q CCC: $33.50 © 2006 American Chemical Society Published on Web 07/08/2006

DNA Adduction by Three Alkylanilines in ViVo

Chem. Res. Toxicol., Vol. 19, No. 8, 2006 1087

Scheme 1. Structures of the Three Alkylanilines

at levels of 1 per 107-109 nucleotides. Further difficulty arises in the use of small animals, such as mice, and when the experimental design includes examining organs, such as the bladder, from which relatively little DNA can be obtained. The analytic demands can be met through the use of accelerator mass spectrometry (AMS) to quantify isotope-labeled products following administration of labeled precursors, as described in a recent review (10). With AMS now available to us, we undertook the study reported here to ascertain whether 2,6DMA, 3,5-DMA, and 3-EA produce DNA adducts in vivo when administered to mice and, if so, to obtain a preliminary characterization of adduct yields as well as stability.

Figure 1. Plasma clearance of 14C. Isotope concentrations are dosenormalized to permit direct comparison between compounds. Each point is the average of values from two animals except for the 2,6-DMA 8 h point, which corresponds to a single animal, and the error bars represent the individual values.

Materials and Methods Caution: Alkylanilines are toxic and mutagenic and may haVe carcinogenic actiVity. These compounds should be handled using appropriate precautions. Chemicals and Reagents. [Ring-14C(U)]-3-EA, 3,5-DMA, and 2,6-DMA with specific activities, respectively, of 53, 50, and 70 mCi/mmol were obtained from American Radiolabeled Chemicals (St. Louis, MO). All other chemical used were reagent grade unless otherwise noted. Animals and Dosing. Male 7-9 weeks old C57BL/6 mice were obtained from Harlan (Indianapolis, IN) and were acclimated for 1 week on AIN-76A purified diet without ethoxyquin prior to use. All experiments were approved by The Johns Hopkins University Animal Care and Use Committee. [14C]Alkylanilines were dissolved in ethanol and administered ip in 200 µL of 0.9% normal saline. Final ethanol concentration was 5%. The actual doses given, per animal, were 3-EA, 1.48 µCi (27.9 nmol, 3.34 µg); 3,5-DMA, 0.96 µCi (19.2 nmol, 2.32 µg); 2,6-DMA, 1.10 µCi (15.9 nmol, 1.92 µg). Animal weights ranged from 22 to 24 g. On a body weight basis, the doses averaged 145 µg/kg (3-EA), 100 µg/kg (3,5-DMA), and 83 µg/kg (2,6-DMA). A total of 39 mice, divided into 3 cohorts of 13 for each alkylaniline, was used in this study. Two from each cohort were sacrificed at each of 4 times, 2, 4, 8, and 16 h postdosing. The remaining 5 from each cohort were housed in metabolic cages in order to collect urine and were sacrificed at 24 h. Tissues were harvested from only 2 of each group of 5. Animals were sacrificed by CO2 asphyxiation and subsequently exsanguinated via cardiac puncture. Bladder, colon, kidney, liver, lung, and pancreas were harvested from each animal. Immediately after harvest, 25 mg of each tissue was removed for DNA isolation and snap-frozen in liquid nitrogen. DNA. DNA was isolated from tissue and initially purified using Qiagen DNeasy Kit (Qiagen, CA). Tissues were minced with a razor blade, placed in Eppendorf tubes, and processed as recommended, including the optional RNase step. DNA was eluted in the final step with pure water instead of buffer. DNA was repurified using a slightly modified protocol provided by Qiagen technical service. Briefly, 200 µL of DNA in water plus 20 µL of AW1 buffer and 500 µL of AW2 buffer were vortexed 10 s and processed as above. Samples were eluted in 50 µL of water. Repurifications were conducted after several weeks, during which time the samples were stored frozen at -20 °C. DNA concentrations were determined by measuring absorbance at 260 nm. Accelerator Mass Spectrometry. AMS analyses were conducted at the MIT BEAMS Lab using procedures described elsewhere in detail (11). In this study, DNA was isolated in the form of unbuffered solution in H2O and was analyzed by AMS without further processing. Aliquots (1.50 µL) were absorbed into pellets

of packed CuO powder that had previously been exposed for 30 min to an atmosphere of O2 at 720 °C. After drying briefly in a vacuum oven, the sample-loaded CuO pellets in their holders were transferred to a laser-induced combustion interface for subsequent AMS analysis. The current interface design has a capacity of 15 samples. Each run of 15 samples was organized as follows: the first sample was a quantitation standard (0.0030 dpm/µL) and was followed by a blank (1 µg/µL albumin), 5 samples, a second quantitation standard, a second blank, another 5 samples, and a third standard. The 10 samples typically comprised all those associated with one organ and one compound. Quantitation was performed by integrating peaks produced in the continuous trace of 14C detector count rate versus time generated during operation of the combustion interface, which produces and delivers the CO2 of combustion to the AMS ion source, and taking the product of the sample/standard peak area ratio multiplied by the standard concentration as the concentration of the sample. The purpose of the blank was to test for “memory” in the ion source cathode, which occurs to a variable extent; detection of 14C when analyzing a blank is not an indication that the blank contains detectable 14C. In our experience, whenever a cathode exhibits “memory” as determined by a response to a blank, then a subsequent test sample analyzes high by a comparable amount. Thus, whenever the albumin blank produced a peak capable of being integrated, the area of the peak was subtracted from the area of the following 5 sample peaks. Initially isolated DNA samples were analyzed once; repurified DNA samples were analyzed twice in order to improve precision, as these samples had lower levels of isotope.

Results Clearance. Pooled urine from 5 animals from each cohort was collected over 24 h. Total isotope excreted in urine, determined by liquid scintillation counting (LSC), amounted to 45% of dose for 3-EA, 45% of 3,5-DMA, and 25% of 2,6DMA. These results may underestimate the actual totals because of the difficulty of quantitatively collecting, in the available metabolic cages, the small volumes of urine produced by the mice, but they are useful for comparing the different alkylanilines. Total plasma isotope concentrations were determined by LSC with the results shown in Figure 1. Since the study protocol was not specifically designed to provide a complete description of plasma clearance, there were no samples taken at early time points. As a result, the data obtained do not characterize the absorption phase for any of the three compounds. They also do

1088 Chem. Res. Toxicol., Vol. 19, No. 8, 2006

Skipper et al.

Figure 2. DNA adduct levels in bladder and liver at different times following administration of 2,6-DMA, 3,5-DMA, and 3-EA. Each point is the average of values from two animals except for the 2,6-DMA 8 h time point. Horizontal dashed lines at 300 and 30 amol/µg correspond to levels of one adduct per 107 and 108 nucleotides, respectively.

not appear to characterize any of the maximum plasma concentration values (Cmax), but they do depict clearance beginning some time after the attainment of Cmax. Clearance curves differed depending on the identity of the alkylaniline, as is readily apparent from the Figure. Clearance of 2,6-DMA was apparently biphasic: in the loglinear plot shown, the curve exhibits a large change in slope between the early and late time points. Clearance of 3,5-DMA is less obviously biphasic, but also is not straight-lined in a loglinear plot. Clearance of 3-EA, in contrast, is well-described with a single first-order rate constant which, as determined by linear regression analysis, corresponds to a half-life of 33 h (R2 ) 0.967; P ) 0.003). In all three cases, later clearance is slow relative to rates that would be expected for clearance of metabolites. The representative 33 h half-life observed in the case of 3-EA corresponds more closely to clearance of protein-bound isotope, considering that the half-life for serum albumin in the mouse is estimated to be 35-39 h (12). It may be inferred from the clearance curves that metabolism and elimination of metabolites is largely complete after no more than 8 h, or less in the case of 3-EA. In keeping with this inference, we note that total plasma isotope at 8 h represents only 0.05-0.3% of the dose, approximately, which is entirely consistent with the isotope being mostly protein-bound. DNA Adducts. Adduct levels were calculated from the ratio of 14C concentration determined by AMS, and DNA concentration determined by A260. To facilitate comparisons between compounds, measured values were normalized to a doses20 nmol/animalswhich is intermediate between the lowest and highest doses actually given. All DNA samples were first analyzed as isolated from tissues. For reasons discussed below, all samples were then repurified and analyzed again. The entire data set is given in the Supporting Information in both tabular and graphic form. Shown here in Figure 2 are the data for the final bladder and liver DNA samples, all of which had quantifiable adduct levels. Adduct levels in other tissues were less frequently detected and are not shown here because the relatively fewer number of data points from the repurified samples makes their interpretation difficult. Figure 2 reveals both the clearance of adducts with time over the course of the experiment and the relative potency of the three alkylanilines. 3,5-DMA was clearly the most effective of the three alkylanilines in forming adducts in both tissues. 3-EA

Table 1. Kinetic Parameters for Clearance of Bladder and Liver DNA Adducts calculated adduct levelaat t1/2 (h)

P

b

2,6-DMA 3,5-DMA 3-EA

8 15 23

Bladder .004 .03 .28

2,6-DMA 3,5-DMA 3-EA

10 21 12

Liver .04 .17 .008

t)0

t ) 12

64 145 25

24 82 17

184 411 55

77 275 27

a Calculated from slope and intercept obtained by linear regression analysis of ln(adduct level) vs time. Units are amol/µg DNA/20 nmol dose. b P value for slope.

Table 2. Relative Adduct Levels in Isolated and Repurified DNA adduct level ratioa tissue

2,6-DMA

3,5-DMA

3-EA

bladder colon kidney liver lung

9.5 3.8 16 4.9 4.2

23 2.5 8.0 4.6 3.4

24 6.6 11 5.4 -

a Values are the median ratio of the adduct level in the DNA as isolated to the level in the repurified sample.

was the least effective, though the difference between it and 2,6-DMA in bladder was small. More quantitative comparisons are available in Table 1, which gives the parameters derived by linear regression analysis of the data using a semi-log model. On the basis of the zero-time intercepts, 3,5-DMA formed 2-7 times more adducts in liver and 2-6 times more in bladder than 2,6-DMA and 3-EA. The same comparisons using values calculated for 12 h are 4-10 and 3-5. Half-lives ranged from 8 to 23 h. At 12 h, adduct levels in bladder ranged from 6 per 109 to 3 per 108 nucleotides and in liver from 1 per 108 to 9 per 108. Repurification of DNA resulted in a substantial decrease in apparent adduct levels, and the reductions were both tissue- and time-related. Differences between median adduct level declines in all the various tissues are compared numerically by the data given in Table 2. The specific differences at each sample time can be found in the Supporting Information and are summarized here as follows. Adduct levels at 2 and at 4 h typically, but not

DNA Adduction by Three Alkylanilines in ViVo

always, declined more than at 8, 16, or 24 h. While repurification led to an order-of-magnitude decrease in adduct levels, with few exceptions, it had minimal effect on the shape of the adduct level versus time curves, mainly just decreasing the early slope. In particular, if adduct levels peaked later than 2 h in the plot of once-purified adduct levels versus time, this feature was preserved in the plot of repurified adduct levels versus time. There are several other points to note regarding the effect of DNA repurification on apparent adduct levels. One is the relative persistance of liver DNA adducts of 3,5-DMA. Not the most abundant in the initially isolated DNA, these adducts become 5-10-fold higher than adducts of 3,5-DMA in other tissues after DNA repurification. A similar persistencesbut relative to adducts of other compoundssis observed for adducts of 2,6DMA in pancreatic DNA. Whereas pancreatic DNA adducts of 3,5-DMA were undetectable after 4 h and adducts of 3-EA not at all after repurification, adducts of 2,6-DMA remained detectable at all time points. A last observation is that bladder DNA adducts of all three alkylanilines declined more than adducts in any other tissue, and more so at 2 and 4 h than at later times, effectively inverting their relative abundance as a result of DNA repurification.

Discussion This study was designed principally to address the question of whether any of the several selected monocyclic aromatic amines are capable of forming significant levels of DNA adducts in vivo in the bladder or any of several other organs. To our knowledge, the existing literature contains only a single report of a comparable experiment (7). That publication indicates that 2,6-DMA, one of the alkylanilines used in the present study, does bind to DNA of liver and nasal cavity epithelium of rats after induction by pretreatment with the amine. A second objective of the present study, in the event that detectable adduct levels were induced, was to develop a preliminary overview of binding index, organ specificity, and adduct clearance to be used in subsequent study designs. All the DNA samples acquired in this study were repurified for a second round of adduct quantitation after results were obtained from the initially isolated samples. The second quantitative analysis was prompted primarily by the observation of some initially precipitous declines in adduct levels, which might have reflected contamination of the DNA since high levels of isotope in various form would be expected to be present in tissues during the first hours of the experiment. A second consideration was the high levels in bladder DNA relative to DNA from other tissues, since 25-45% of the isotope was collected in urine, whereas other tissues, except perhaps liver, would be less exposed. Repurification led to, at first glance, a roughly 10-fold decrease in adduct level. As noted earlier, the extent of decrease was not uniform across the different time points or tissues. Apparent adduct levels declined considerably more in bladder and kidney than in other tissues as judged by comparisons of the median decline for all samples taken from one organ. This tissue dependence is most noticeable in the case of bladder and liver adducts of 3,5-DMA. Adduct levels at 2 and at 4 h typically, but not always, declined more than at 8, 16, or 24 h. The greater decline in adduct levels in bladder and kidney relative to other organs, and between early time samples relative to late, is consistent with our first thought that the DNA, as originally isolated, was not entirely free of contaminating isotope. However, the fact that repurification had little impact on the overall form of the adduct level versus time curves, except

Chem. Res. Toxicol., Vol. 19, No. 8, 2006 1089

to reduce the early slope, suggests that contamination was unimportant because it is more likely to occur randomly than systematically unless it were to occur as a result of material proportionally co-purifying. To address the matter of copurifying isotope, we took a subset of 12 isolated and repurified DNA samples from five different tissues of animals given 3,5DMA and subjected them to a third purification using traditional chloroform/phenol extraction methods, including proteinase K and RNase treatment. This procedure was adopted for its greater capacity relative to the Qiagen methods to remove traces of protein, including protein that might contain bound isotope and would thus contribute to the apparent DNA isotope level. The effect was to reduce specific isotope levels by severalfold, but leave every DNA sample with isotope still present well above background, which strongly suggests the presence of true DNA adducts in the samples. An alternative explanation for the reduction of bound isotope upon repurification of DNA samples is that the isotope detected is indeed present in the form of adducts, but that a substantial proportion of the adducts are unstable to the conditions of DNA purification. It is not unexpected that unstable adducts might be formed. N-7 alkylguanine adducts, for example, readily undergo depurination. Ultimately stable adducts that form through initially reversible reactions such as carbonyl condensation might also be unstable at an early stage of formation. At this point, though, nothing is known about the structures of the adducts quantitated in this study, nor even about the metabolites from which they arise, so the discussion regarding stability remains speculative. It should also be noted that adduct instability and incomplete purification are not mutually exclusive possible explanations for reduced adduct levels in repurified DNA; it may be that both are involved. The results presented here enable a preliminary assessment of the carcinogenic potential of 2,6-DMA, 3,5-DMA, and 3-EA based on the carcinogen binding index (CBI) described by Lutz (13). CBI is a measure of the amount of adduct formed from a given dose; it is calculated by dividing the amount of adduct per unit of DNA by the dose per unit of body mass and multiplying by a factor to adjust the values so that they fall within a convenient range. CBIs can be calculated for the alkylanilines at any of the time points used in this study. The most conservative approach uses the data from DNA that underwent repurification after isolation and from animals that were sacrificed at 16 and 24 h. The resulting CBIs for bladder and liver, respectively, are 2,6-DMA, 5 and 14; 3,5-DMA, 21 and 84; and 3-EA, 5 and 7. According to Lutz, a CBI of 1-10 is characteristic of weak carcinogens, while moderate carcinogens are characterized by values on the order of 100. If the CBI is less than 0.1, then the compound is considered unlikely to induce carcinogenesis by itself in a long-term bioassay. According to this analysis, 3,5-DMA would be ranked a moderate carcinogen, the other two alkylanilines would be weak carcinogens. These assessments may actually underestimate the carcinogenic potencies for at least two reasons. First, the lowest adduct levels measured in this study were used to calculate the CBIs. Lowest levels were used because there is uncertainty regarding whether the DNA, as isolated, was truly free of contaminating isotope and because it is questionable that the CBI values calculated using rapidly cleared DNA adduct levels, that is, those present at 2-4 h post-dosing but not at 16-24 h, are reasonably comparable to previously reported CBI values. The second reason for considering the carcinogenicity assessments as likely underestimates highlights an unusual and important, though not unique, feature of this study, which is

1090 Chem. Res. Toxicol., Vol. 19, No. 8, 2006

the low dosessapproximately 100 µg/kgsthat were used. One hundred micrograms per kilogram (µg/kg) is greater than normal daily exposure to ambient levels of alkylanilines,2 but it is within the range of occupational exposures (15). The doses used in this study are much lower than the doses used in most studies of DNA adduct formation and are more likely to be in the linear portion of the DNA adduct dose-response relationship. A study of the potent carcinogen NNK (16), published after the CBI was proposed as a measure of carcinogenic potential, provides an interesting and informative comparison. In the NNK study, doses of 12-20 000 µg/kg were used, and it was reported that the dose-response decreased at the higher doses. The following CBIs are calculated for the lowest doses from the reported data: pyridyloxobutylation in liver DNA, 16; in lung, 53; N7guanine methylation in liver, 680; in lung 730. Corresponding CBIs at the highest doses are up to 9 times lower, as expected for nonlinear dose-response. The CBI values that emerge from the present study are quite comparable to those derived from the data reported for pyridyloxobutylation by NNK, even under the conservative criteria used. The CBI for N7-guanine methylation is an order of magnitude higher but of questionable relevance since N7methylguanine is generally considered to be of minor significance for carcinogenicity. If our criteria are actually too conservative, then the CBI for not only 3,5-DMA but also 2,6DMA and 3EA approach or exceed the CBI calculated for NNK pyridyloxobutylation. As noted earlier, adduct formation by 2,6-DMA has previously been investigated (7). In the rat, a species different from that used in the present study, 2,6-DMA formed DNA adducts in liver and ethmoid turbinate after pretreatment for 9 days with a daily dose of unlabeled amine equal to 25% of the LD50. The CBI for liver was then very similar to the CBI we observed for liver and was considerably higher in ethmoid turbinate. Without pretreatment, adduct formation was observed only in liver near the limit of detection. In the present study, pretreatment was not required. A possible explanation for the different outcomes of the two studies, besides the species difference, lies in the doses used. Short et al. (7) administered 2,6-DMA at about 4.3 mg/kg, which differs by a factor of nearly 60 from our dose of about 75 µg/kg. Quite possibly, at the higher dose, a much smaller fraction of the amine is metabolized to DNA-binding metabolite(s) in the uniduced animal. Notably, the effect of induction may be reversed: Short et al. found that pretreatment with acetylaminofluorene sharply reduced the CBI of acetylaminofluorene (7). A fuller understanding of the enzymology of 2,6-DMA metabolism is required to determine whether the proposed explanation is correct. In conclusion, the present study indicates that all three of the alkylanilines investigated appear to have significant carcinogenic potential, based on their ability to bind to DNA in vivo. The evidence also suggests that 3,5-DMA has greater carcinogenic potential than the other two. Whether DNA binding is mediated by N-hydroxylamine metabolites, as has been exten2 Based on the hemoglobin adduct data in ref 5 and the dose-response data for adduct formation given in ref 14. Environmental source data are too limited for conventional exposure assessment.

Skipper et al.

sively documented for multicyclic aromatic amines, remains undetermined but will be the subject of future studies. Acknowledgment. This work was supported by grants PO1ES006052 and P30-ES002109 from the National Institute for Environmental Health Sciences. Supporting Information Available: Table listing all the data and plots of all adduct levels as a function of time. This material is free of charge via the Internet at http://pubs.acs.org.

References (1) Woo, Y. T., and Lai, D. Y. (2001) Aromatic amino and nitro-amino compounds and their halogenated derivatives. In Patty’s Toxicology (Bingham, E., Cohrssen, B., and Powell, C. H., Eds.) pp 969-1105, Wiley, New York. (2) Talaska, G. (2003) Aromatic amines and human urinary bladder cancer: exposure sources and epidemiology. J. EnViron. Sci. Health, Part C: EnViron. Carcinog. Ecotoxicol. ReV. 21, 29-43. (3) Vineis, P., and Pirastu, R. (1997) Aromatic amines and cancer. Cancer Causes Control 8, 346-355. (4) Markowitz, S. B., and Levin, K. (2004) Continued epidemic of bladder cancer in workers exposed to orthno-toluidine in a chemical factory. J. Occup. EnViron. Med. 46, 154-160. (5) Gan, J., Skipper, P. L., Gago-Dominguez, M., Arakawa, K., Ross, R. K., Yu, M. C., and Tannenbaum, S. R. (2004) Alkylaniline-hemoglobin adducts and risk of non-smoking-related bladder cancer. J. Natl. Cancer Inst. 96, 1425-1431. (6) U. S. National Toxicology Program (1990) Toxicology and Carcinogenesis Studies of 2,6-Xylidine (2,6-Dimethylaniline) (CAS No. 8762-7) in Charles RiVer CD Rats (Feed Studies). Technical Report Series No. 278, NTP, Research Triangle Park, NC. (7) Short, C. R., Joseph, M., and Hardy, M. L. (1989) Covalent binding of [14C]-2,6-dimethylaniline to DNA of rat liver and ethmoid turbinate. J. Toxicol. EnViron. Health 27, 85-94. (8) Marques, M. M., Mourato, L. L. G., Amorim, M. T., Santos, M. A., Melchior, Jr., W. B., and Beland, F. A. (1997) Effect of substitution site upon the oxidation potentials of alkylanilines, the mutagenicities of N-hydroxyalkylanilines, and the conformations of alkylaniline-DNA adducts. Chem. Res. Toxicol. 10, 1266-1274. (9) Gonc¸ alves, L. L., Beland, F. A., and Marques, M. M. (2001) Synthesis, characterization, and comparative 32P-postlabeling efficiencies of 2,6dimethylaniline-DNA adducts. Chem. Res. Toxicol. 14, 165-174. (10) Brown, K., Tompkins, E. M., and White, I. N. H. (2006) Applications of accelerator mass spectrometry for pharmacological and toxicological research. Mass Spectrom. ReV. 25, 127-145. (11) Liberman, R. G., Tannenbaum, S. R., Hughey, B. J., Shefer, R. E., Klinkowstein, R. E., Prakash, C., Harriman, S. P., and Skipper, P. L. (2004) An interface for direct analysis of 14C in nonvolatile samples by accelerator mass spectrometry. Anal. Chem. 76, 328-334. (12) Chaudhury, C., Mehnaz, S., Robinson, J. M., Hayton, W. L., Pearl, D. K., Roopenian, D. C., and Anderson, C. L. (2003) The major histocompatibility complex-related Fc receptor for IgG (FcRn) binds albumin and prolongs its lifespan. J. Exp. Med. 197, 315-322. (13) Lutz, W. K. (1979) In vivo covalent binding of organic chemicals as a quantitative indicator in the process of chemical carcinogenesis. Mutat. Res. 65, 289-356. (14) Sabbioni, G. (1992) Hemoglobin binding of monocyclic aromatic amines: molecular dosimetry and quantitative structure activity relationships for the N-oxidation. Chem.sBiol. Interact. 81, 91-117. (15) National Institute for Occupational Safety and Health (1991) Interim report No. 2, HETA 88-159, Goodyear Tire and Rubber Co. Niagara Falls, N. Y., NIOSH, Cincinnati, OH. (16) Murphy, S. E., Palomino, A., Hecht, S. S., and Hoffmann, D. (1990) Dose-response study of DNA and hemoglobin adduct formation by 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone in F344 rats. Cancer Res. 50, 5446-5452.

TX060082Q