part 1 . PART 1 concordance between cancer in humans and in experimental CHAPTER 2 animals chapter 2. Aromatic amines and aristolochic acids Frederick A. Beland and M. Matilde Marques Carcinogenicity in humans uation for these dyes was raised to can be attributed to 4-aminobiphe- Group 1 based on this mechanistic nyl, o-toluidine, 2-naphthylamine, or Exposure to 4-aminobiphenyl, o-tolu- information, although at present the other aromatic amines. Hair dyes idine, 2-naphthylamine, and benzi- corresponding epidemiological data are an additional source of expo- dine (Fig. 2.1) has been consisten- are considered to provide inade- sure to 4-aminobiphenyl and o-tolu- tly associated with the induction quate evidence for the carcinogeni- idine (IARC, 2010, 2012a; Lizier and of cancer of the urinary bladder in city of these dyes in humans (IARC, Boldrin Zanoni, 2012). humans. This association is based 2010, 2012a). Exposure to phenacetin (Fig. 2.1), upon occupational exposures, pri- Cigarette smoke contains 4-amino- through its use as an analgesic, marily of workers in the rubber and biphenyl, o-toluidine, and 2-naphthyl- causes cancer of the kidney and dye industries (IARC, 2010, 2012a). amine, and tobacco smoking caus- ureter in humans (IARC, 2012c). Similarly, occupational exposure to es cancer of the bladder in humans Chlornaphazine (Fig. 2.1), a chemo- 4,4′-methylenebis(2-chloroaniline) (IARC, 1986, 2004, 2010, 2012a, b). therapeutic agent that has been used (MOCA; Fig. 2.1), a curing agent for The contribution of 4-aminobiphenyl, for the treatment of Hodgkin lympho- polyurethane pre-polymers, causes o-toluidine, and 2-naphthylamine ma and for the control of polycythae- cancer of the bladder in humans, al- to the induction of smoking-related mia vera, causes cancer of the though the epidemiological data are cancer of the bladder is confounded bladder in humans, presum ably due not as strong as those for the other by the presence of numerous other to metabolism to 2-naph thylamine agents (IARC, 2010, 2012a). Certain carcinogens, including carcinogenic (IARC, 2012c). An additional source azo dyes that are used in commercial aromatic amines, in tobacco smoke. of human exposure to o-toluidine products, for example, Direct Black Cigarette smoking also causes oth- is from the anaesthetic prilocaine 38, Direct Blue 6, and Direct Brown er cancers (e.g. cancer of the lung, (Fig. 2.1) (IARC, 2010, 2012a). 95 (Fig. 2.2), are known to undergo oral cavity, and pancreas, and pos- Exposure to herbal remedies pre- azo reduction in vivo to yield the car- sibly breast cancer), but at present pared from plant species of the ge- cinogen benzidine. The overall eval- it is unclear whether these cancers nus Aristolochia has been causally Part 1 • Chapter 2. Aromatic amines and aristolochic acids 23 associated with the induction of Fig. 2.1. Structures of IARC Group 1 aromatic amines, drugs that are urothelial cancer in humans (IARC, metabolized to Group 1 aromatic amines, and aristolochic acids. MOCA, 2002, 2012c). The induction of 4,4′-methylenebis(2-chloroaniline). urothelial tumours has been attributed to the presence of aristolochic acid I and aristolochic acid II (Fig. 2.1). Metabolism and DNA adduct formation 4-Aminobiphenyl, o-toluidine, 2-naph- thylamine, and MOCA are metabolized to electrophilic N-hydroxyarylamines by hepatic cytochrome P450 en- zymes (IARC, 2010, 2012a, c). The N-hydroxyarylamines undergo ac- id-catalysed reactions with DNA to form a variety of DNA adducts. With the exception of MOCA, C8- substituted deoxyguanosine adducts are typically the major products, along with smaller amounts of N2- substituted deoxyguanosine and N6-substituted deoxyadenosine ad- ducts (Fig. 2.3); with MOCA, only C8-substituted deoxyadenosine ad- ducts have been detected (IARC, 2010, 2012a). These DNA adducts can also be formed from reactive esters of N-hydroxyarylamines (e.g. N-sulfoxyarylamines and N-aceto xy- arylamines). Benzidine, which has two amino groups, also forms a C8- substituted deoxyguanosine adduct via a pathway involving an initial ronides can undergo acid-cata- adduct detected in each instance N-acetylation followed by N-hydroxyl- lysed hydrolysis to release the was a C8-substituted deoxyguano- ation of the remaining amino function N-hydroxyarylamines, which can sine adduct. The importance of uri- (Fig. 2.3). enter the bladder epithelium and re- nary acidity for the hydrolysis of the The carcinogenic activity of ar- act with DNA either directly or after N-hydroxyarylamine N-glucuronides omatic amines in the bladder in esterification. DNA adducts derived and perhaps for the reaction of the humans has been attributed to an from 4-aminobiphenyl, o-toluidine, N-hydroxyarylamines with urothe- initial N-hydroxylation, catalysed benzidine, and MOCA have been lial DNA has been demonstrated by hepatic cytochrome P450 en- detected in bladder tissue or exfoli- by the positive correlation between zymes, followed by transport of ated bladder cells from exposed in- urinary acidity and the levels of ben- the N-hydroxyarylamines to the dividuals (IARC, 2010, 2012a; Böhm zidine DNA adducts in exfoliated bladder as either aglycones or et al., 2011; Lee et al., 2014). With bladder cells from exposed workers N-glucuronide conjugates (Bois et the exception of MOCA, which forms (Rothman et al., 1997). Additional al., 1995). In the bladder lumen, only C8-substituted deoxyadenosine support for this mechanism comes the N-hydroxyarylamine N-glucu- adducts, the major – if not the only – from the observation of a positive 24 correlation between urinary acidity Fig. 2.2. Structures of benzidine-derived azo dyes. and the incidence of bladder cancer in smokers (Alguacil et al., 2011). The major metabolic activation PART 1 pathway for aristolochic acid I and CHAPTER 2 aristolochic acid II involves nitro reduction, followed by cyclization to give N-hydroxyaristolactams, which – in contrast to other N-hydroxyarylamides – do not ap- pear to require additional activa- tion to react with DNA (Stiborová et al., 2011, 2013). Nonetheless, N-hydroxyaristolactams have been shown to serve as substrates for human sulfotransferases, particu- larly sulfotransferase family cy- tosolic 1B member 1 (SULT1B1), forming highly reactive N-sulfoxy derivatives (Sidorenko et al., 2014). The major adducts resulting from the N-hydroxyaristolactams are N2- substituted deoxyguanosines and N6- substituted deoxyadenosines (Fig. 2.4) (IARC, 2002, 2012c). DNA adducts derived from aris- Alterations in the TP53 higher-grade tumours (grades 2 or 3, tolochic acids have been detected tumour suppressor gene in i.e. moderately or poorly differentiat- in renal tissue from patients who humans ed) and only at G:C base pairs. had been exposed to aristolochic Mutant p53 protein has also been acid-containing herbal products and Mutations in the TP53 tumour sup- detected in workers exposed oc- from individuals who had consumed pressor gene have been found in cupationally to benzidine (Xiang et wheat grains contaminated with approximately 50% of all bladder al., 2007). The occurrence and the Aristolochia (IARC, 2002, 2012c; cancers in humans (Petitjean et al., amount of mutant protein were pos- Chen et al., 2012; Jelaković et al., 2007), with G:C base substitution itively correlated with the level of 2012; Schmeiser et al., 2012, 2014; mutations occurring to a greater benzidine exposure and the extent Yun et al., 2013, 2014). Typically, extent than A:T base substitution of neoplastic changes in exfoliated the major lesion detected is an N6- mutations. urothelial cells. deoxyadenosine adduct derived TP53 gene mutations have been Urothelial tumours arising from from aristolochic acid I, accompa- detected in bladder cancer patients exposure to aristolochic acids have nied by smaller amounts of a similar exposed occupationally to 4-amino- been consistently shown to carry mu- adduct derived from aristolochic acid biphenyl, 2-naphthylamine, and/or tations in the TP53 tumour suppres- II and an N2-deoxyguanosine adduct benzidine (Sørlie et al., 1998). The sor gene, of which the most common derived from aristolochic acid I. mutations occurred exclusively in mutation is an A → T transversion Part 1 • Chapter 2. Aromatic amines and aristolochic acids 25 Fig. 2.3. Structures of representative DNA adducts obtained from Group 1 Support for this mechanism comes aromatic amines. dR, deoxyribose. from the observation that the DNA lesions detected in the bladders of dogs treated with 4-aminobiphenyl or MOCA appear to be C8-substituted deoxyguanosine and deoxyaden- osine adducts that are identical to the DNA adducts detected in blad- der tissues or exfoliated bladder cells from humans exposed to these carcinogens (IARC, 2010, 2012a). 2-Naphthylamine DNA adducts de- tected in the bladders of dogs ex- posed to 2-naphthylamine are en- tirely consistent with a mechanism involving the formation of N-hydroxy- 2-naphthylamine (IARC, 2010). In contrast to what is observed in humans, benzidine in not a bladder carcinogen in dogs. This lack of car- cinogenicity has been attributed to the inability of dogs to N-acetylate aromatic amines (IARC, 2010). With most aromatic amines, N-acetylation is considered to be a detoxifica- tion event; however, with benzi- dine, N-acetylation appears to be required to give N-acetylbenzidine, which undergoes a subsequent mutation (reviewed in IARC, 2002, Tumour sites and the N-hydroxylation of the second amino 2012c; Hollstein et al., 2013; see also mechanism of tumour function. This metabolic pathway oc- induction in experimental Schmeiser et al., 2012; Chen et al., curs in humans but not in dogs. animals 2013; Hoang et al., 2013; Poon et 4-Aminobiphenyl, o-toluidine, al., 2013; Aydin et al., 2014). These 2-naph thylamine, benzidine, ben- 4-Aminobiphenyl, 2-naphthylamine, mutations have been demonstrated zidine-based dyes, and MOCA in- and MOCA are bladder carcinogens in tumour tissue from patients who duce hepatocellular tumours in mice in dogs (IARC, 2010, 2012a). Bladder had consumed herbal preparations (IARC, 2010, 2012a).