Cadmium Alters the Biotransformation of Carcinogenic Aromatic Amines by Arylamine N-Acetyltransferase Xenobiotic-Metabolizing Enzymes: Molecular, Cellular, and in Vivo Studies

Background Cadmium (Cd) is a carcinogenic heavy metal of environmental concern. Exposure to both Cd and carcinogenic organic compounds, such as polycyclic aromatic hydrocarbons or aromatic amines (AAs), is a common environmental problem. Human arylamine N-acetyltransferases (NATs) are xenobiotic-metabolizing enzymes that play a key role in the biotransformation of AA carcinogens. Changes in NAT activity have long been associated with variations in susceptibility to different cancers in relation with exposure to certain AAs. Objective We explored the possible interactions between Cd and the NAT-dependent biotransformation of carcinogenic AAs. Methods We exposed purified enzymes, lung epithelial cells, and mouse models to Cd and subsequently analyzed NAT-dependent metabolism of AAs. Results We found that Cd, at biologically relevant concentrations, impairs the NAT-dependent acetylation of carcinogenic AAs such as 2-aminofluorene (2-AF) in lung epithelial cells. NAT activity was strongly impaired in the tissues of mice exposed to Cd. Accordingly, mice exposed to Cd and 2-AF displayed altered in vivo toxicokinetics with a significant decrease (~ 50%) in acetylated 2-AF in plasma. We found that human NAT1 was rapidly and irreversibly inhibited by Cd [median inhibitory concentration (IC50) ≈ 55 nM; rate inhibition constant (kinact) = 5 × 104 M−1 · sec−1], with results of acetyl coenzyme A (acetyl-CoA) protection assays indicating that Cd-mediated inhibition was due to the reaction of metal with the active-site cysteine residue of the enzyme. We found similar results for human NAT2, although this isoform was less sensitive to inactivation (IC50 ≈ 1 μM; kinact = 1 × 104 M−1 · sec−1). Conclusions Our data suggest that Cd can alter the metabolism of carcinogenic AAs through the impairment of the NAT-dependent pathway, which may have important toxicological consequences.


Research
Cadmium (Cd), a ubiquitous toxic element and widespread pollutant, is introduced to the environment mainly through anthropo genic activities, such as lead mining, fos sil fuel combustion, and the manufacturing of Cdcontaining products. Another major source for human exposure to Cd is cigarette smoke (Martelli et al. 2006). The chief route of Cd exposure is via the respiratory system (Potts et al. 2001).
Because of its stability in the environment and long retention time in the human body (halflife, ~ 20 years), Cd can accumulate and cause a variety of adverse effects (Joseph 2009;Waalkes 2003). The target organs for Cd toxicity include the liver, kidney, lung, testis, prostate, and bladder. However, pro longed human exposure to Cd results mainly in diseases affecting lungs and kidneys [International Agency for Research on Cancer (IARC) 1993]. Cd has been classified as a class 1 human carcino gen by IARC (1993). Most of our current knowledge regarding the mechanisms of Cd carcino genesis is derived from experi ments conducted with cell cul tures and animal models. These studies have shown that the mechanisms underlying Cd carcinogenesis are multi factorial (Huff et al. 2007). Among these mechanisms, accumula tion of DNA damage due to inhibition of DNA repair enzymes by Cd is considered as one of the major underlying processes (Jin et al. 2003;McNeill et al. 2004;Potts et al. 2001). Simultaneous and/or sequential expo sure to Cd has been suggested to contribute to the carcinogenic potential of other geno toxic chemicals commonly found in the environ ment and in the workplace, such as aromatic carcinogens (Prakash et al. 2000;Rivedal and Sanner 1981;Williams et al. 1984).
Aromatic amines (AAs) represent one of the most important classes of occupational or environmental pollutants (Kim and Guengerich 2005). AAs account for 12% of the chemicals known or strongly suspected to be carcinogenic in humans (National Toxicology Program 2005). AAs are by products of chemical manufacturing (e.g., pesticides, dyestuff, rubbers), gasoline com bustion, or pyrolysis reactions (Palmiotto et al. 2001). Carcinogenic AAs such as 4aminobi phenyl (4ABP) and βnaphthylamine are also present in cigarette smoke (Hein 1988;Hein et al. 2000).
Human arylamine Nacetyltransferases (NATs) are xeno bioticmetabolizing enzymes (XMEs) that play a major role in the bio transformation of AA carcinogens (Dupret and RodriguesLima 2005;Hein 1988). The inter relationship between variable NAT activities and the toxicity or carcino genicity of environ mental AAs has been reported (Badawi et al. 1995;Gemignani et al. 2007;Hein 1988;Minchin et al. 1993). NATdependent acetyla tion of AAs leads mainly to xeno biotic detoxi cation but also to bio activation. Detoxification into innocuous compounds is mainly medi ated by NATdependent Nacetylation, whereas Oacetylation of chemi cals previously hydroxy lated by cytochrome P450 1A1 (CYP1A1) or CYP1A2 promotes their metabolic activation into DNAbinding electrophiles (Hein 1988). Several studies have shown associations between NAT1 and/or NAT2 activities and increased risk of cancer, particularly in relation with expo sures to AAs (Hein 1988;Hein et al. 2000).
Exposure to both Cd and AAs occurs commonly, such as through cigarette smoke (Joseph 2009;Stavrides 2006). Although Cd has been shown to modify the expres sion and/or activity of CYP1A1 (Elbekai and ElKadi 2007;Vakharia et al. 2001), the effect of this metal on the metabolism of aro matic carcinogens remains poorly defined. In this study, we found molecular, cellular, and in vivo evidence that Cd, at biologically rele vant concentrations, can alter the biotrans formation of AA carcinogens through the impairment of the NATdependent acetyla tion pathway. This process may represent an additional mechanism contributing to Cd carcinogenesis. volume 118 | number 12 | December 2010 • Environmental Health Perspectives
Recombinant enzyme production and purification. Hexahistidine (6xHis)tagged human NAT1 was produced and purified from Escherichia coli strain BL21 (DE3) as previously described (Dairou et al. 2003). We used the same approach to prepare human NAT2, except that induction with iso propyl βd1thiogalactopyranoside (500 μM) was performed at 30°C for 8 hr followed by a 12hr incubation at 4°C.
Effects of Cd on recombinant human NAT enzymes. We tested the effect of Cd on recombinant human NAT1 and NAT2 by incubating purified enzymes (0.3 μM final concentration) with increasing concentra tions of CdCl 2 in 25 mM TrisHCl (pH 7.5) for 10 min at 37°C. Recombinant NAT1 or NAT2 enzymatic activities were determined spectrophotometrically using pnitro phenyl acetate (PNPA) as the acetyl donor and 2AF as arylamine substrate, as reported previously (Dairou et al. 2003;Mushtaq et al. 2002). In all reaction mixtures, the final concentrations of NAT1 and NAT2 were 15 nM and 30 nM, respectively. For the controls, we omitted the enzyme, 2aminofluorene (2AF), or PNPA. All enzyme reactions were performed in tripli cate, in conditions in which the initial reaction rates were linear.
We assessed the reversibility of the reac tion of Cd with NAT1 and NAT2 by incu bating recombinant enzymes with Cd (final concentrations, 0.3 μM for NAT1 and 2 μM for NAT2) for 10 min at 37°C. Mixtures were then dialyzed overnight at 4°C against 25 mM TrisHCl, pH 7.5. For controls, enzymes not treated with Cd were dialyzed overnight. After dialysis, residual enzyme activities were assayed.
We tested whether reducing agents [reduced glutathione (GSH) and dithio threitol (DTT)] or the chelating agent EDTA was able to restore the activity of Cdinhibited NAT1 and NAT2. To this end, recombinant enzymes were pre incubated with Cd (final concentra tions, 0.3 μM in experiments with NAT1 and 2 μM in those with NAT2) for 10 min at 37°C. Mixtures were then incubated with different concentrations of DTT or GSH (up to 10 mM final concentration) or EDTA (up to 5 mM final concentration) for 10 min at 37°C. Residual enzyme activities were then assessed. Control assays were carried out as described above in the absence of Cd but with GSH, DTT, or EDTA. We also tested whether these reducing or chelating agents were able to protect NAT1 and NAT2 enzymes from the inhibitory effects of Cd by carrying out Cd treatments (final concentration, 0.3 μM in experiments with NAT1 and 2 μM in those with NAT2) in the presence of high concentra tions of DTT, GSH, or EDTA (up to 10 mM, which corresponds to > 33,000 times the Cd concentration) and then determining residual enzyme activities.
We investigated the extent to which acetylCoA and CoA protected NAT1 and NAT2 from Cddependent inhibition. To this end, recombinant enzymes were preincubated with different concentrations of acetylCoA or CoA in 25 mM TrisHCl, pH 7.5, for 5 min at 37°C. Mixtures were then incu bated with Cd (final concentration, 0.3 μM for NAT1 experiments and 2 μM for NAT2 experiments) for 10 min at 37°C (final con centration of acetylCoA or CoA, 0-3 mM). Samples were then assayed. Control assays carried out in the absence of Cd treatment gave 100% enzyme activity.
Kinetic analysis: determination of the second-order rate inhibition constant (k inact ). NAT1 or NAT2 was incubated under second order conditions as described previously by CornishBowden (2001). Briefly, recombinant enzymes (0.3 μM final concentration) were incubated with Cd (0.3 μM final concentra tion) in 25 mM TrisHCl, pH 7.5, at 37°C. Every 2 min, aliquots of the reaction mixture were taken and quenched by dilution with buf fer containing 1 mM EDTA, and the residual enzyme activity was assayed as described below. The values of k inact were obtained by fitting residual enzyme activity to the equation where E is the enzyme concentration, E 0 is the initial enzyme concentration, and t is time. We used KaleidaGraph, version 3.5 (Abelbeck/Synergy, Reading, PA, USA) for mathematical analyses of the data.
Cell culture, exposure to Cd, and wholecell extracts. The murine mtCC12 Clara lung epithelial cell line (Magdaleno et al. 1997) was provided by J.M. Sallenave (Institut Pasteur, Paris, France) and grown in Dulbecco's modi fied Eagle's medium (DMEM) supplemented with 20% (vol/vol) fetal bovine serum. Cells were cultured as mono layers in 35 or 100mm Petri dishes at 37°C. This lung epithelial cell line is known to express only functional Nat2 (the murine ortholog of human NAT1) (Dairou et al. 2009).
Cell monolayers (~ 80% confluence) were washed with phosphatebuffered saline (PBS) and exposed to different concentrations of CdCl 2 in 10 mL PBS or DMEM for 2 hr at 37°C in a cell incubator. Control cells were incubated with PBS or DMEM only. After incubations, mono layers were washed with PBS and scraped into 0.5-1 mL lysis buf fer (25 mM TrisHCl, pH 7.5; 0.1% Triton X100) and protease inhibitors. Extracts were sonicated and centrifuged for 15 min at 13,000 × g. Supernatants (wholecell extracts) were removed, and their protein concentra tion was determined using Bradford reagent with bovine serum albumin as a standard. All cell extracts were adjusted to the same pro tein concentration by adding lysis buffer and were used for enzyme assays. We determined NAT activity in cells or mouse tissue extracts by measuring the formation of Nacetylated  metabolites using reversephase high performance liquid chromatography (HPLC) as described previously (Grant et al. 1991). All assays were performed in triplicate under initial reaction rate conditions. Enzyme activi ties were normalized according to the protein concentration of cellular extracts determined using the BioRad protein assay kit (BioRad, Hercules, CA, USA). We analyzed CdCl 2 cytotoxicity using a 3(4,5dimethyl thiazol2yl)2,5diphenyl tetrazolium bromide (MTT)based assay (Mosmann 1983) with concentrations of CdCl 2 ranging from 0 to 100 μM. In the conditions used above, CdCl 2 cytotoxicity was < 5% for concentrations up to 50 μM.

Acetylation of AA carcinogens by intact cultured cells in the presence or absence of Cd.
Acetylation of 2AF and 4ABP by endoge nous Nat2 in growing cells was measured by reversephase HPLC as described previously (Wu et al. 2000). Cells were incubated with Cd at different concentrations (up to 50 μM) for 2 hr. After treatment, cells were grown in fresh culture medium containing 750 μM 2AF or 4ABP. Controls were incubated in the same conditions but with cell monolayers not exposed to Cd. Reactions were found to be linear with time.
In a second set of experiments, Clara cells were coexposed to Cd at different concentra tions (up to 50 μM final) and to 750 μM 2AF (or 4ABP) in culture medium. At dif ferent time points, aliquots were analyzed as described above.
Mouse Cd exposure and plasma pharmacokinetics. All procedures involving animals were carried out in accordance with the French Agriculture Ministry's internal guidelines for animal handling. The number of mice and suffering were minimized whenever possible.
We investigated the ability of Cd to inhibit NAT functions in vivo by treating 12weekold female C57BL/6J mice with a sub lethal dose of CdCl 2 (2 mg/kg in PBS), as reported previously (Martin et al. 2007). Two hours after intra peritoneal (IP) injection, mice (n = 7) were sacrificed by cervical dis location, and endogenous NAT activity was measured in protein lysates of tissues known to be targeted by Cd, such as blood, liver, kid ney, and lung. Tissue extracts were prepared as described previously (Smelt et al. 2000).
In a second set of experiments, we ana lyzed the in vivo pharmaco kinetics of acety lated 2AF in mice (n = 7) with or without CdCl 2 treatment (2 mg/kg in PBS). Two hours after Cd treatment, 2AF (50 mg/kg) dissolved in dimethyl sulfoxide was adminis tered by IP injection. Animals were then anesthetized with 10 mg/kg ketamine and 1 mg/kg xylazine, and blood samples were drawn from retroorbital venous plexus at five sequential time points (30 min, 1 hr, 2 hr, 3 hr, and 6 hr). Samples were diluted 1:50 in HPLC mobile phase and analyzed for par ent and acetylated metabolites by HPLC as described above. Area under the curve (AUC) was determined by the trapezoidal rule using Microsoft Excel 2007 (Microsoft Corporation, Paris, France).
Statistical analysis. Data are presented as mean ± SD of three independent experiments performed in triplicate, unless otherwise stated. Oneway analysis of variance was performed, followed by Student's ttest between two groups using StatView 5.0 (SAS Institute Inc., Cary, NC, USA).

Cd impairs recombinant human NAT1 and NAT2 activity.
To test whether Cd can inhibit human NAT1 and NAT2 activity, we assessed its effect on purified recombinant enzymes. As shown in Figure 1, Cd demon strated a dosedependent inhibitory effect on NAT1 activity. Full inhibition of NAT1 was obtained with Cd concentrations as low as 0.3 μM. The median inhibitory concentration (IC 50 ) for Cd was approxi mately 0.055 μM (Figure 1, inset).
We also observed dosedependent inhibition of the human NAT2 isoform by Cd (data not shown) but with a higher IC 50 (~ 1 μM). These data indicate that both human NAT isoforms are readily inhibited in vitro by low, biologically relevant concentrations of Cd.
Inhibition of human NAT enzymes by Cd is rapid and irreversible. We tested whether the inhibition of NAT1 by Cd could be reversed by physiological (GSH) and non physiological (DTT) reducing agents that are known to react with Cd. DTT and GSH (1-10 mM final concentrations) did not significantly reverse Cddependent inhibition of NAT1 (Figure 2A). We observed a modest reactivation effect (~ 20% of control activity) with high concentrations of DTT or GSH (10 mM final concentration). Similar results were obtained for human NAT2 (data not shown).
We also tested whether EDTA, a known Cdchelating agent, was able to reverse Cddependent inhibition of NAT1 and NAT2. Incubation of inhibited enzymes with differ ent concentrations of EDTA for 30 min at 37°C did not reverse the inhibitory effect of Cd ( Figure 2A). To further analyze the irreversible reaction of Cd with human NAT enzymes, we carried out dialysis experiments. In agree ment with the results reported above, dialysis of Cdinhibited NAT1 and NAT2 enzymes did not allow any significant recovery of enzymatic activity (data not shown). Taken together, these results indicate that Cddependent inhibition of NAT enzymes is irreversible.
In a second set of experiments, we ana lyzed the ability of DTT and GSH to pre vent Cddependent inhibition of NAT1. Incubation of the enzyme with Cd in the pres ence of these compounds at high concentra tions (> 33,000 times the Cd concentration) protected NAT1 only partially against inhi bition ( Figure 2B). At high final concentra tions of 10 mM GSH or DTT, we observed approximately 60% residual NAT1 activity. Similar results were obtained for NAT2 (data not shown). These data indicate that reducing agents, even at high concentrations, provide only partial protection against Cddependent inhibition of NAT enzymes, suggesting that Cd reacts more quickly with NAT enzymes than with GSH or DTT. Conversely, EDTA (2 mM final concentration) provided almost full protection (~ 85%), indicating the depen dence of inhibition on the presence of free metal ions in solution ( Figure 2B).
To further characterize the reaction of Cd with NAT enzymes, we performed kinetic analy ses. Pseudofirstorder conditions (i.e., implying Cd concentrations were well above enzyme concentration) could not be used because the enzymes were almost instantly inhibited by Cd under these conditions. Therefore, we used secondorder conditions (see "Materials and Methods") to determine the secondorder rate constant of inhibition (k inact ). The k inact constant for Cddependent inhibition of NAT1 was 5.2 × 10 4 M -1 • s -1 (Figure 3, inset). The k inact for NAT2 was slightly lower (1 × 10 4 M -1 • s -1 ). These results confirm the high reactivity of Cd toward NAT1 and NAT2 enzymes in vitro.

Cd-dependent inhibition is due to interaction with the active-site cysteine residue of NAT enzymes.
Alteration of cellular func tions by binding to certain thiol groups of bio molecules is the most commonly invoked pathway for Cd toxicity (Joseph 2009). To investigate whether the Cddependent inhibi tion of these XMEs could be due to direct reaction of Cd with the activesite cysteine residue of NAT enzymes, we carried out sub strate protection assays using acetylCoA and CoA as reported previously (Liu et al. 2008). This protection assay relies on the specific acetylation of the NAT activesite cysteine residue by acetylCoA, which protects this residue from further chemical reaction (Liu et al. 2008). Conversely, because CoA is unable to acetylate the activesite cysteine residue, this amino acid is thus suscepti ble to chemical reaction in the presence of CoA. AcetylCoA afforded significant dose dependent protection against Cddependent NAT1 inhibition (up to 70% residual NAT1 activity; Figure 4). In contrast, CoA did not provide any significant protection against Cddependent inhibition. We observed similar results with the NAT2 isoform (data not shown). These data suggest that in vitro, Cd irreversibly inhibits NAT1 and NAT2 enzymes through the direct interaction with their activesite cysteine residues.

Impairment of the endogenous NATdependent biotransformation pathway in lung epithelial Clara cells by Cd.
We tested the effect of Cd on endogenous NAT activ ity by measuring acetylated metabolites of 2AF and 4ABP in the culture medium of mtCC12 Clara cells not exposed or previ ously exposed to different concentrations of Cd. Clara cells are known to play a major role in lung xeno biotic metabolism and are the progenitor cells for broncho genic carcinomas (Oreffo et al. 1990). Recently, Clara cells were shown to biotransform AA chemicals through the Nat2dependent pathway (Dairou et al. 2009). We exposed mtCC12 Clara cells [which express only Nat2, the murine coun terpart of human NAT1 (Kawamura et al. 2008)] to Cd for 2 hr and then to 2AF or 4ABP (in fresh medium with no Cd). We observed that the amount of acetylated 2AF and 4ABP in cell culture medium decreased in a dosedependent manner with an IC 50 value for CD around 17 μM ( Figure 5A). We observed similar results in cells co exposed to Cd and AAs (data not shown). Moreover, these results are in agreement with enzyme assays carried out with extracts of treated mtCC12 cells ( Figure 5B). Overall, these data indicate that exposure to Cd alters the acetylation of AAs in lung epithelial Clara cells through impairment of the endogenous NATdependent pathway.

Effect of Cd on the in vivo NAT-dependent biotransformation of 2-AF.
In C57BL/6J mice (n = 7) treated with a sub lethal dose of CdCl 2 (2 mg/kg) as described previously (Martin et al. 2007), we found that Cd to significantly decreased NAT activity in the lungs, kidneys, liver, and blood of treated mice (up to 52% inhibition in the lung; Figure 6A). These data were further confirmed by plasma toxico kinetics experiments with 2AF, which showed that Cdtreated mice exhibited altered levels of Nacetylated 2AF with an approximately 50% decrease in the AUC (mean ± SD, 1.43 ± 0.15 mmolmin/L for control mice and 0.81 ± 0.12 mmolmin/L for treated mice; Figure 6B). Taken together, these data indicate that Cd alters the NATdependent acetylation of 2AF in vivo. Figure 3. Determination of k inact for inhibition of NAT1 by Cd under second-order conditions. After equimolar concentrations of NAT1 and CdCl 2 were incubated, residual NAT1 activity was assayed at 2-min intervals. Inset: k inact obtained by fitting the data to the equation 1/E = 1/E 0 + k inact × t and taking k inact from the slope of 1/E versus time.

Discussion
The cellular effects of the toxic metal Cd are manifold. In particular, exposure to Cd is asso ciated with cancers of the prostate, bladder, kidney, and lung (Huff et al. 2007;Waalkes 2003). Because Cd interacts with cellular components in many ways, no factor fully accounts for its spectrum of toxic and carci nogenic effects (Joseph 2009). Several stud ies have demonstrated that Cd is a complex carcinogen, and the mechanisms underlying Cd carcino genesis are multifactorial (reviewed by Joseph 2009). Cd interacts synergistically with DNAdamaging agents, such as aromatic carcinogens, which may enhance their muta genic potential and result in biologically rele vant genotoxic effects (Godschalk et al. 2005;Prakash et al. 2000;Rivedal and Sanner 1981;Williams et al. 1984). Although Cd has been suggested to modify the expression of certain XMEs such as CYP1A1 (Elbekai and ElKadi 2007;Maier et al. 2000), the potential of Cd to alter the biotransformation of aromatic carcino gens remains poorly defined. NATs are XMEs that play a major role in the biotransformation of AA carcinogens, and changes in the N and/or Oacetylation of these chemicals have been linked to carcino genesis (Hein 1988). We report here that biologically relevant levels of Cd (≤ 50 μM) (Apostolova et al. 2006) alter the biotransfor mation of carcinogenic AAs through impair ment of NAT enzyme functions. We found that Cd can irreversibly inhibit NAT1 and NAT2 acetylation activities in vitro with IC 50 values as low as 75 nM and 1 μM, respec tively. Kinetic analysis of these Cddependent inhibitions gave secondorder k inact values of 5 × 10 4 M -1 • s -1 and 1 × 10 4 M -1 • s -1 for human NAT1 and NAT2, respectively. Studies on DNA repair enzymes known to be impaired by Cd, such as Ogg1 and poly nucleotide kinase, have reported IC 50 values ranging from 5 to 100 μM (Bravard et al. 2006;McNeill et al. 2004;Whiteside et al. 2010;Zharkov and Rosenquist 2002) and k inact values around 5 M -1 • s -1 (Zharkov and Rosenquist 2002). The data we obtained for NAT enzymes indicate that these XMEs are extremely sensitive to Cd exposure, which leads to their rapid functional impairment. Moreover, our results suggest that in vitro, NAT1 is more susceptible to Cddependent inhibition than is NAT2. A similar trend with isoformselective inactivation of human NAT enzymes by 4nitrosobiphenyl and 2nitroso fluorene was recently reported, with NAT1 appearing to be more sensitive to these com pounds (Liu et al. 2009).
Cd has been reported to act as either a reversible or an irreversible inhibitor of cer tain enzymes, depending on the nature of the enzyme-Cd interaction. For instance, Whiteside et al. (2010) have shown that different DNA repair enzymes exhibit differ ent behaviors in this regard. In the present study, we found that the in vitro inhibition of NAT1 and NAT2 by Cd was irreversible because extensive dialysis did not restore enzyme activities. Moreover, the reduc ing agents GSH or DTT and the chelating agent EDTA did not restore the activity of NAT1 and NAT2, further supporting the irreversible nature of the Cddependent inhi bition of these XMEs. We also found that the presence of high concentrations of GSH or DTT (10 mM final) afforded only partial protection (~ 60% residual activity) against Cddependent inhibition (with Cd at a molar concentration > 33,000 times lower than that of GSH or DTT). These data suggest that Cd reacts more rapidly with NAT1 (and NAT2) than with GSH or DTT. Accordingly, the kinetics of the Cd reaction with GSH is at least two orders of magnitude lower (k assoc < 10 2 M -1 • s -1 ) than the inactiva tion rates found for NAT1 and NAT2 (k inact > 10 4 M -1 • s -1 ). Conversely, we found that EDTA afforded significant protection against Cddependent inhibition of human NAT enzymes. These results are similar to data obtained with the DNA repair enzyme Ogg1 and indicate the dependence of inhibition on the presence of free metal ions in solution (Zharkov and Rosenquist 2002). AcetylCoA protection assays indicated that Cddependent irreversible inhibition of human NAT1 and NAT2 was due to the reaction of the metal with the reactive catalytic cysteine residue (Ragunathan et al. 2008), which is in agree ment with the fact that Cd exhibits high affin ity for certain reactive thiols (Bravard et al. 2006). A similar mechanism of inhibition has been reported for other enzymes inhibited by Cd, such as Ogg1 and nicotinamide adenine dinucleotide phosphate-dependent isocitrate dehydrogenase (Kil et al. 2006).
The amount of Cd absorbed in the body after exposure is principally due to inhala tion of Cdcontaminated smoke and parti cles and results mainly in diseases affecting kidneys and lungs (Joseph 2009). Cd has been shown to accumulate in tissues at up to tensofmicromolar concentrations (Jin et al. 2003). Furthermore, occupational expo sures in industrial society have been reported to produce renal cortical Cd concentrations around 300 μM (Apostolova et al. 2006). We exposed murine Clara cells, which are known to express functional Nat2, the murine Figure 5. Inhibition of endogenous Nat2 activity and AA acetylation in Clara cells by Cd. Cells in Petri dishes were exposed to different concentrations of CdCl 2 for 2 hr and then grown in fresh culture medium in the presence of 750 μM 2-AF or 4-ABP. (A) The amount of acetylated 2-AF or acetylated 4-ABP quantitated (in triplicate) in culture medium by HPLC. (B) Cells were washed, and a whole-cell extract was made; Nat2 activity was measured by HPLC in normalized extracts using 2-AF as substrate. Data are mean ± SD. *p < 0.05.  Figure 6. Impairment of Nat2 activity in tissues of mice exposed to Cd (A) and pharmacokinetics of acetylated 2-AF (B). CdCl 2 (2 mg/kg) was injected IP. Adult C57BL/6J mice (n = 7) were injected IP with 2 mg/kg CdCl 2 . (A) Two hours after injection, mice were sacrificed, and Nat2 activity was assessed in lysates from the liver, kidney, lung, and blood. (B) Two hours after CdCl 2 injection, 2-AF (50 mg/kg) dissolved in dimethyl sulfoxide was administered by IP injection, and acetylated 2-AF was measured in blood at different time points. Acetylated 2-AF AUC values: untreated, 1.43 ± 0.15 mmol-min/L; Cd treated, 0.81 ± 0.12 mmol-min/L. Data are mean ± SD.   (Potts et al. 2001). Our results are also in agreement with previous data showing that the acetylation of 3chloroaniline in isolated rat hepatocytes could be suppressed by Cd concentrations close to 25 μM (Alary et al. 1989). Cd significantly inhibited (by 36-52%) endogenous NAT activity in lung, liver, kid ney, and blood, tissues known to accumulate Cd (Huff et al. 2007;Joseph 2009), from mice exposed to Cd IP. These data indicate that NATdependent biotransformation of AAs may be altered by Cd in several tissues expressing these XMEs. Toxicokinetics stud ies provided further evidence that in vivo Cd alters the NATdependent biotransformation of carcinogenic AAs. Mice exposed to Cd exhibited altered biotransformation of 2AF, as shown by the 50% decrease in acetylated 2AF AUC.
Increasing evidence suggests that geno toxicity induced by Cd also depends on the synergic interactions of Cd with genotoxic chemicals (Joseph 2009). Cd inhibits several enzymes involved in DNA repair, and this has been identified as a major mechanism underlying the carcinogenic potential of Cd (Joseph 2009;Zharkov and Rosenquist 2002). Furthermore, synergic interactions between Cd and carcinogenic aromatic chemi cals that can lead to aromatic DNA adducts have been reported (Godschalk et al. 2005;Prakash et al. 2000;Rivedal and Sanner 1981;Williams et al. 1984). Moreover, studies have shown that Cd can modify the expression and/or activity of certain XMEs, such as CYP1A1, leading to altered metabolism of polycyclic aromatic hydrocarbons such as benzo[a]pyrene (Elbekai and ElKadi 2007;Maier et al. 2000;Vakharia et al. 2001). Interestingly, Cd has been proposed to enhance mutagenicity of benzo[a]pyrene metabolites (Prakash et al. 2000).
Cd toxicity is caused by both acute and chronic exposure. Although the toxicity of Cd is considered mostly chronic, several studies using acute conditions have yielded a substantial amount of information pertinent to Cd toxic ity, including mechanistic information (Bravard et al. 2006;McNeill et al. 2004). Moreover, acute inhalation of high levels of Cd in humans may result in longlasting impairment of lung functions (Agency for Toxic Substances and Disease Registry 1997). Here, we showed that acute exposure to Cd alters the biotransforma tion of AAs in several tissues, including lung. Further studies are needed to assess whether chronic exposure to Cd leads to similar effects. Humans are exposed to AA carcinogens mainly through cigarette smoke or occupa tional/industrial pollutants. Several studies have suggested that NATdependent acetyla tion is a susceptibility factor for cancers associ ated with AA exposures (Hein 1988), possibly through formation of aromatic DNA adducts (Badawi et al. 1995). Interestingly, simulta neous and/or sequential exposure to Cd and AAs occurs commonly, such as in cigarette smoke (Hein et al. 2000;Stavrides 2006). Our results provide evidence that Cd alters AA carcinogen metabolism by interfering with the NATdependent acetylation pathway. This may represent an additional mechanism con tributing to Cd carcinogenesis.