36 | MARCH 2022 | VOLUME 22 www.nature.com/nrc REV I EWS 0123456789( ) ; : This Review, updated from a previous Review1, addresses chemical mechanisms and prevention of frequently fatal human cancers induced by two worldwide addictions — smokeless tobacco use and cigarette smoking. Although there is general awareness that tobacco products cause cancer, there are nevertheless more than one billion users of these products worldwide. We discuss chemical mechanisms by which smokeless tobacco and cigarette smoking cause cancer, focusing on carcinogenic constituents, their DNA adducts and mutational consequences, biomarkers of these processes and regulatory approaches to prevention, including reduction of tobacco-specific nitrosamines in smokeless tobacco and nicotine in cigarettes. These mechanisms proceed in a well-established sequence fromrepeated carcinogen exposure due to nicotine addiction to metabolically activated carcinogens to the formation of DNA adducts and consequent critical mutations in growth control genes resulting in cancer1,2 (Fig. 1). We also describe how we can use biomarkers not only to understand the carcinogenic process but also to examine their association with cancer risk. Finally, we discuss selected means to prevent and reduce the harm caused by these addictive and deadly products. Massive scope of the problem Smokeless tobacco The worldwide prevalence and distribution of smokeless tobacco use is perhaps underappreciated. In 121 countries worldwide, there are more than 350 million smokeless tobacco users, 67% of whom are men3. Nearly 95% live in developing countries, 82.7% in the World Health Organization (WHO) South-East Asia Region, where smokeless tobacco use is sometimes the predominant form of tobacco use, exceeding cigarette smoking3. More than 90% of smokeless tobacco users live in 11 countries: India (237.4 million users), Bangladesh (30.9 million), Myanmar (12.6 million), Pakistan (10.1 million), the USA (9.6 million), China (4.1 million), Indonesia (3.2 million), Nepal (2.7 million), Madagascar (2.6 million), and Germany and Uzbekistan (2.4 million each)3. Norway, Sweden, Yemen and parts of Africa are also relatively high prevalence areas4. In India, the prevalence of current smokeless tobacco use among adults was 25.9% overall, 32.9% among men and 18.4% among women3. The International Agency for Research on Cancer (IARC) concluded that smokeless tobacco causes cancers of the oral cavity, oesophagus and pancreas in humans5; the US National Cancer Institute and the American Cancer Society concur with this evaluation6,7. The evidence supporting these conclusions, from studies in laboratory animals and epidemiologic studies, has been reviewed5. A systematic review and meta-analysis of epidemiologic studies from South Asia demonstrated a combined odds ratio of 4.7 for oral cancer in users of these products8. A recent review further summarizes this convincing body of research9. Consistent with the high DNA adducts Compounds formed by the reaction of DNA bases with certain electrophilic intermediates generated during metabolism, or with inherently reactive substances. Odds ratio A statistic that explains the association between two events. Smokeless tobacco and cigarette smoking: chemical mechanisms and cancer prevention Stephen S. Hecht ✉ and Dorothy K. Hatsukami Abstract | Tobacco products present a deadly combination of nicotine addiction and carcinogen exposure resulting inmillions of cancer deaths per year worldwide. A plethora of smokeless tobacco products lead to unacceptable exposure tomultiple carcinogens, including the tobacco-specific nitrosamine N′-nitrosonornicotine, a likely cause of the commonly occurring oral cavity cancers observed particularly in South-East Asian countries. Cigarettes continue to deliver a large number of carcinogens, including tobacco-specific nitrosamines, polycyclic aromatic hydrocarbons and volatile organic compounds. Themultiple carcinogens in cigarette smoke are responsible for the complexmutations observed in critical cancer genes. The exposure of smokeless tobacco users and smokers to carcinogens and toxicants can nowbemonitored by urinary andDNA adduct biomarkers that may be able to identify those individuals at highest risk of cancer so that effective cancer prevention interventions can be initiated. Regulation of the levels of carcinogens, toxicants and nicotine in tobacco products and evidence-based tobacco control efforts are now recognized as established pathways to preventing tobacco related cancer. Masonic Cancer Center, University of Minnesota, Minneapolis, MN, USA. ✉e-mail: hecht002@umn.edu https://doi.org/10.1038/ s41568-021-00423-4 Nature reviews | CanCer Rev i ews volume 22 | march 2022 | 143
VOLUME 22 | MARCH 2022 | 37 NATURE REVIEWS | CANCER REV I EWS 0123456789( ) ; : prevalence of smokeless tobacco use, the incidence of oral cancer in South Asia is one of the highest in the world, and it has been estimated that approximately 50% of oral cancers in India are due to smokeless tobacco use, amounting to about 35,000 cancers per year10. Worldwide, attributable disability-adjusted life years for upper aerodigestive tract cancer from smokeless tobacco use amounted to an estimated 1.5 million11 to 2 million12, mostly from the South-East Asia region. Extraordinarily diverse types of smokeless tobacco products with differing carcinogenic properties are consumed worldwide. Monographs published by the US Centers for Disease Control and Prevention (CDC) and the IARC describe some of the many types of these products4,13. Broadly, they can be classified as pre-made or custom-made products. Pre-made products can be further subdivided into two types. The first are made in a manufacturing environment and commercially distributed in sealed packaging, commonly found in the USA and other Western countries as well as India and parts of Asia. The second are those handmade in cottage industries and non-traditional environments and having inconsistent types of packaging; these are more common in, for example, the South-East Asia region. Custom-made products are produced individually for immediate consumption according to a customer’s preference. Smokeless tobacco products have been divided into four categories4. Category 1 products containmainly tobacco. Category 2 products have tobacco and alkaline modifiers, which raise the pH, facilitating absorption of nicotine. Category 3 products contain tobacco, alkaline modifiers and areca nut, as in the well-known betel quids consumed in South-East Asia. Category 4 products contain tobacco and additional substances, such as stimulants, flavouring agents or spices. Examples of products in each category are presented in a CDC monograph4. Cigarette smoking Despite the harmful effects of cigarette smoking being nearly universally known, worldwide there were 933 million smokers in 2015 (reF.14). More than 80% of the world’s smokers live in low-income and middle-income countries14. It is estimated that about two-thirds of lung cancer deaths worldwide are due to smoking15. There were about 2.2 million new cases of lung cancer and 1.8 million deaths from lung cancer in 2020, accounting for 11.4% of new cancer diagnoses and 18.0% of cancer-related deaths15. Lung cancer was the leading cause of cancer-related death worldwide in men and the second leading cause (after breast cancer) in women. It was the leading cause of cancer-related death in men in 93 countries and in women in 25 countries. Convincingly, international variations in lung cancer rates and trends mainly reflect tobacco use15. Cigarette smoking is also a cause of cancers of the oral cavity, pharynx, larynx, oesophagus, nasal cavity, pancreas, bladder, stomach, liver, kidney, ureter, cervix, colorectum and ovary (mucinous), as well as of myeloid leukaemia2,5. The devastating effects of cigarette smoking result from the combination of rapid delivery of nicotine to the brain, leading to its highly reinforcing effects and high potential for addiction16, along withmultiple carcinogens in the smoke that accompany each dose of nicotine. Collectively, these highly addictive tobacco products continue to cause a worldwide epidemic of morbidity and death from cancer. We also note that all smokeless and combustible tobacco products, including those not discussed here (various types of smokeless tobacco, cigars, cigarillos and so on), are established causes of cancer, although the risk may differ depending on the product type (combustible or non-combustible) and within categories of tobacco products, particularly with smokeless tobacco. Whether smokeless or combustible, the problem is the tobacco. While not the focus of this Review, bidi smoking and waterpipe use are additional popular international tobacco use practices. Bidis are small hand-rolled tobacco products commonly smoked in South-East Asia, where it is likely that there are more than 50 million users17. There are millions of waterpipe users worldwide. In waterpipe smoking, the tobacco is heated and the smoke is drawn through water, whichmay contain flavouring additives18. Each of these habits entails exposure to tobacco smoke toxicants and carcinogens. Carcinogens, DNA adducts and mutations Tobacco-specific nitrosamines All commercial tobacco products, including smokeless tobacco, cigarette smoke and cigar smoke contain tobacco-specific nitrosamines, which form during the curing and processing of tobacco19–22. Seven tobacco-specific nitrosamines have been identified in unburned, cured tobacco, but three of these — N′- nitrosonornicotine (NNN), 4-(methylnitrosamino)- 1-(3-pyridyl)-1-butanone (NNK) and itsmajormetabolite 4-(methylnitrosamino)-1-(3- pyridyl)-1- butanol (NNAL) — have received by far the most attention23,24 (Fig. 2). The major source of tobacco- specif ic nitrosamines is cured and processed tobacco, where they are formed by the reaction of tobacco alkaloids with nitrite; there is also evidence for their endogenous formation, particularly by nitrosation of nornicotine25,26. NNN and NNK are the most carcinogenic of the tobacco-specific nitrosamines; their mean levels in US Nicotine addiction and continuous exposure to carcinogens Smokeless tobacco use Cigarette smoking Carcinogens Excretion Normal DNA Apoptosis DNA adducts Mutations in multiple genes Cancer Metabolic activation Persistence leading to miscoding Metabolic detoxification Repair Urinary metabolite biomarkers DNA adduct biomarkers Mutational signatures and biological changes Fig. 1 | Overall established scheme relating smokeless tobacco use and cigarette smoking, as driven by nicotine addiction, to cancer. Long-termuse of tobacco products results in continuous exposure to carcinogens, the formation of DNA adducts andmultiplemutations in critical cancer control genes1,2. Biomarkers andmutational signatures can elucidate each step. Areca nut The seed of the areca palm, which grows in South-east Asia; it is a common constituent of a quid known as paan, which is chewed by millions of people in the region. www.nature.com/nrc Rev i ews 144 | march 2022 | volume 22
38 | MARCH 2022 | VOLUME 22 www.nature.com/nrc REV I EWS 0123456789( ) ; : brands of cigarette tobaccowere 1,901ng g−1 and 523ng g−1 tobacco, respectively, while the levels in cigarette mainstream smoke (the smoke taken in by the smoker as estimated by machine smoking) were 189 ng and 122ng per cigarette, respectively20. The relative levels of NNN and NNK in cigarette smoke are dependent on the type of tobacco used27. NNAL, which is also highly carcinogenic, is present in the urine of all people who use tobacco products, and is also commonly detected in the urine of non-smokers exposed to second-hand tobacco smoke28. NNN and NNK, which always occur together, are considered carcinogenic to humans by the IARC13. NNN, NNK and NNAL are typically strong nitrosamine carcinogens, inducing tumours in laboratory animals through a genotoxic mechanism after metabolic activation in specific tissues. Tumour induction is generally independent of the route of administration and occurs in a dose-responsive manner, including at low doses relevant to the levels of exposure of people who use tobacco products13. Some particularly relevant carcinogenicity data for these compounds are presented in the following sections. Tobacco-specific nitrosamines and other carcinogens in smokeless tobacco users. The concentrations of NNN and other tobacco-specific nitrosamines in some smokeless tobacco products widely consumed in South-East Asia are remarkably high9,29,30. In khaini products consumed in India, Stanfill et al. reported NNN concentrations of 37.7–48.1 µg g−1 wet weight of tobacco, and Stepanov et al. reported levels of 16.8–29.4µgg−1 (reFS21,31). Nasrin et al found NNN concentrations as high as 59 µg g−1 in zarda products and 25 µg g−1 in gul, a powdered tobacco snuff product, both consumed in Bangladesh32. These high levels of NNN (and other tobacco-specific nitrosamines) are in contrast to average amounts of NNN found in 56 northern European snus products, 0.6µgg−1 (reFS33), while the amounts in brands sold in the USA are typically 1–4 µg g−1 (reFS34,35). The levels of other tobacco- specific nitrosamines were similarly lower in northern European and US brands than in the products from South-East Asia. Thus, the concentrations of NNN and other tobacco-specific nitrosamines in smokeless tobacco continue to be a problem in US brands, although they are generally considerably less than those found in South-East Asia. Because of the ability of NNN to induce tumours of the oral mucosa (described further later), this suggests an association with the high incidence of oral cavity cancer in South-East Asia. Considerable evidence indicates that NNN in particular is likely to play a critical role as a cause of cancer in users of smokeless tobacco products, which all have something in common: they contain both nicotine and NNN. Nicotine is the addictive constituent of all tobacco products, including smokeless tobacco36,37. NNN, the main component of smokeless tobacco that induces cancer of the oral cavity and oesophagus, is closely related structurally to nicotine38,39, the only structural difference being that nicotine has an N-methyl group, while NNN has an N-nitroso group (Fig. 3). This seemingly minor difference causes a major difference in biological activity. Nicotine binds to nicotinic acetylcholine receptors, causing addiction; NNN binds far less readily to these receptors, and its concentration in tobacco is more than 1,000 times less than that of nicotine40. Nicotine is metabolized by cytochrome P450 2A6 (CYP2A6) at its 5′ position to α-hydroxynicotine, which is in equilibrium with the nicotinium ion (Fig. 3); these are further oxidized by aldehyde oxidase and CYP2A6 to cotinine, the major metabolite of nicotine41,42. The nicotinium ion is not reactive with DNA43. NNN is similarly metabolized at its 5′ position by CYP2A6 and CYP2A13, resulting in the formation of 5′-hydroxyNNN44. This highly reactive intermediate undergoes ring opening to a diazohydroxide, which readily reacts with DNA, forming adducts that can cause miscoding when DNA is replicated, thereby leading to permanent mutations and eventually cancer45. The DNA adducts formed from diazohydroxide intermediates produced in the metabolism of NNN have been extensively characterized in rats treated with NNN; these adducts result from both 2′-hydroxylation and 5′-hydroxylation of NNN45,46. Nicotine and NNN have chiral carbons at their 2′ position, as indicated in Fig. 3. Nicotine in tobacco is more than 99% (S)-nicotine, while (S)-NNN is the predominant enantiomer of NNN, constituting 57–83% of NNN in smokeless tobacco products47,48. A carcinogenicity study of the NNN enantiomers in F-344 rats demonstrated that (S)-NNN, administered in the drinking water at a dose of 14 ppm for 17 months, is a powerful oral cavity and oesophageal carcinogen, causing a total of 89 benign and malignant oral cavity tumours, including tumours of the tongue, oral mucosa, soft palate, epiglottis and pharynx, in a group of 20 rats and 122 oesophageal tumours in these same rats, while (R)-NNN was less active but enhanced the carcinogenicity of (S)-NNN39. These results clearly demonstrate the powerful oral carcinogenicity of (S)-NNN in rats, while there is no conclusive evidence that (S)-nicotine is carcinogenic49. The total dose of (S)-NNN used in that study was less than that calculated for 30 years of human use of a smokeless tobacco product containing 3 µg NNN per gram of tobacco39. (S)-NNN preferentially formed pyridyloxobutyl DNA adducts in the oral mucosa and oesophagus of F-344 rats treated identically to those in the carcinogenicity study46. No other known smokeless tobacco constituent induces tumours in the oral mucosa5,13. Taken together with the high levels of NNN in multiple types of smokeless tobacco products consumed in South-East Asia as noted earlier herein, the collective NNK NNAL NNN N N OH O N N N O O N N N N O Fig. 2 | Structures of NNN, NNK and NNAL. N′-Nitrosonornicotine (NNN) and 4-(methylnitrosamino)- 1-(3-pyridyl)-1-butanone (NNK) are carcinogenic tobacco- specific nitrosamines found in all tobacco products. 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) is the major metabolite of NNK and, likeNNK, a potent pulmonary carcinogen. NNAL is found in the urine of all tobacco users. Nature reviews | CanCer Rev i ews volume 22 | march 2022 | 145
VOLUME 22 | MARCH 2022 | 39 NATURE REVIEWS | CANCER REV I EWS 0123456789( ) ; : data strongly suggest that NNN is a cause of oral and oesophageal cancer in smokeless tobacco users. But NNN is not the only carcinogen in smokeless tobacco products. As summarized previously and in recent publications, smokeless tobacco products also contain other carcinogens, including multiple nitrosamines, certain polycyclic aromatic hydrocarbons (PAHs), formaldehyde, acrolein and metals such as cadmium, but none of these carcinogens induces oral tumours in laboratory animals and is consistently found in these products at levels as high as those of NNN5,30,50,51. Thus, we conclude that the path to prevention is clear: find ways to prevent the uptake and promote cessation of use of these products in addition to drastically reducing or eliminating NNN from smokeless tobacco products. Approaches to decreasing NNN in smokeless tobacco include plant breeding, agronomics, and tobacco processing and storage35. Tobacco-specific nitrosamines in cigarette smokers. While the effects of the carcinogenic properties of cigarette smoke on the lung and oral cavity (and other tissues) of smokers are undoubtedly collectively due to exposure to multiple carcinogens, co-carcinogens, inflammatory agents and oxidants, there is little doubt that NNK and NNN are very important in this process on the basis of their levels in smoke and carcinogenic activities. All of the properties noted earlier herein for NNN are also relevant to cigarette smokers and to cancers of the oral cavity and oesophagus in smokers. NNK and NNAL, which are potent lung carcinogens, are particularly relevant to cigarette smoking and lung cancer. As reviewed previously, the lung is the major target organ of NNK in the rat, hamster, mouse, ferret and mink52. NNK readily induces adenocarcinoma of the lung, the major cancer type seen in smokers, independently of the route of administration and at low doses in these laboratory animals. Thus, lung tumours are induced in F-344 rats when NNK is given in the drinking water, by subcutaneous injection, gavage, oral swabbing or even intravesicular administration. Lung tumours are always induced preferentially over tumours in other tissues; this is due to efficient metabolic activation of NNK in the lung by cytochromes P450, resulting in the formation of methyl and pyridyloxobutyl DNA adducts in the lung and consequent mutation of oncogenes. Other effects of NNK, such as activation of the α7 nicotinic acetylcholine receptor–ERK–contactin 1 pathway and regulation of gene expression through DNA methyltransferase 1-mediated epigenetic modifications, have also been described53–55. Extensive dose– response data are available for induction of lung tumours by NNK in rats; the lowest total dose in these studies, only 1.8mg kg−1, administered by subcutaneous injection, induced a significant number of lung tumours; this dose was not dissimilar from the lifetime dose of NNK in a smoker52,56. One carcinogenicity study conducted more recently deserves mention57. In that study, NNK and enantiomers of NNAL were administered in the drinking water to F-344 rats at a dose of 5 ppm for 20 months, resulting in a total dose of 130mgkg−1 body weight. All three compounds as well as racemic NNAL (10 ppm) induced a high incidence of lung tumours, both adenoma and carcinoma; all treated animals had lung lesions, while no lung lesions were observed in the controls. These results again illustrated the powerful pulmonary carcinogenicity of NNK and NNAL, further supporting their role as causes of lung cancer in users of tobacco products. Multiple studies have examined the mutational properties of DNA adducts induced by NNK and NNAL; these include not only pyridyloxobutyl and pyridylhydroxybutyl DNA adducts but also methyl DNA adducts, aldehyde DNA adducts and adducts that generate apurinic sites. As previously noted, the adducts that contribute to the genotoxic effects of NNK and NNAL depend on the context, such as the relative amounts of each DNA alkylating pathway in a given cell, the activity of repair enzymes and the gene targeted for mutation54,58. Thus, the relationship between tobacco- specific nitrosamine DNA adducts and cancer in people who use tobacco products is complex and requires further study. Continuing interest in tobacco-specific nitrosamines. It has been more than 40 years since we first identified tobacco-specific nitrosamines as products of the reaction of nicotine with nitrite and as constituents of tobacco59–61. There has been sustained research interest in the presence of nitrosamines in tobacco products and the carcinogenic properties of these compounds62–64. (S)-Nicotine N N CH3 5' 2' CYP2A6 (S)-NNN CYP2A6 and CYP2A13 N N N 5' 2' O N N CH3 N N CH3 OH N N N O OH DNA adducts Mutations Cancer (S)-Cotinine Excretion in urine (S)-Nicotinium ion (S)-5′-Hydroxynicotine 5′-HydroxyNNN Diazohydroxide intermediate N N CH3 O N N N OH O Fig. 3 | Metabolism of (S)-nicotine and (S)-NNN by 5′-hydroxylation. Metabolismof (S)-nicotine andmetabolismof (S)- N′-nitrosonornicotine ((S)-NNN) by this pathway are similar, yet lead to drastically different results. The red atoms in the structures are the different functional groups that explain the results. For NNN, after 5′-hydroxylation by cytochrome P450 2A6 (CYP2A6) or CYP2A13, the properties of the N–N=Ogroup in 5′-hydroxyNNN cause immediate transformation to the DNA-damaging diazohydroxide intermediate illustrated, leading to DNA adducts, which can result inmutations leading to cancer. In contrast, the highlighted N–CH3 group of nicotine does not have this effect, but rather is further oxidized to the innocuous compound (S)-cotinine, which is stable and is excreted in urine. Brackets indicate unstable intermediates. Agronomics The branch of economics dealing with the distribution, management and productivity of land. Apurinic sites Sites in DNA lacking the usual guanine or adenine bases. The sites can be formed when certain relatively unstable DNA adducts, such as 7-methyldeoxyguanosine, lose their purine base (in this case 7-methylguanine) due to spontaneous decomposition or hydrolysis. www.nature.com/nrc Rev i ews 146 | march 2022 | volume 22
40 | MARCH 2022 | VOLUME 22 www.nature.com/nrc REV I EWS 0123456789( ) ; : Even the tobacco industry, after its usual initial attempts to discredit this research, joined academic and government scientists to investigate methods to monitor and decrease the occurrence of nitrosamines in both tobacco and smoke from cigarettes, and to determine their mechanisms of action. Three recently published studies are briefly summarized here. Investigators from the US Food and Drug Administration (FDA) conducted the first comprehensive inhalation study of NNK, comparing exposure to NNK and its metabolism to NNAL in rats treated with NNK by three different routes: intraperitoneal injection, oral gavage or nose-only inhalation for 1 h using single doses ranging from 5×10−5 to 50mgkg−1. NNK was rapidly absorbed and metabolized extensively to NNAL after NNK administration by all routes, but NNK metabolism to NNAL appeared to be more efficient via inhalation than by the other routes. However, NNK significantly increased DNA damage in multiple tissues via all three routes63. A second study quantified nicotine and tobacco-specific N-nitrosamine levels in 64 snus products made by ten manufacturers in the USA and northern Europe. The level of NNN plus NNK was higher in snus sold in the USA (1,360ng g−1, standard error 207ngg−1) than that of snus sold in northern Europe (836ngg−1, standard error 132ngg−1), likely reflecting the more stringent manufacturing processes used in northern Europe33. Another study tested the effects of co-administration of formaldehyde, acetaldehyde or CO2 by nose-only inhalation on lung tumour development in A/J mice treated with NNK by intraperitoneal injection. The mice treated with aldehydes had more adenomas with dysplasia or progression than those given only NNK, while CO2 treatment increased the number of NNK-induced lung adenomas64. These results demonstrate that gas-phase constituents of tobacco smoke can enhance the lung carcinogenicity of NNK. While the mechanisms of enhancement are not clear and require further study, the presence of considerable amounts of these compounds in tobacco smoke (CO2 is 12.5% of smoke) suggest that the effects of cigarette smoke inhalation may be far greater than estimated by consideration of only individual constituent concentrations. Carcinogens formed by combustion While carcinogenic tobacco-specific nitrosamines are present in both smokeless tobacco and cigarette smoke, combustion products are generally far more prevalent in cigarette smoke because they are often volatile and are generated by burning tobacco at temperatures as high as 910–920 °C. Important among these are the PAHs. Multiple well-established PAH carcinogens are found in tobacco smoke, including benzo[a]pyrene (BaP) (TAble 1), benz[a]anthracene, 5-methylchrysene, dibenz[a,h] anthracene, dibenz[a,l]pyrene, indeno[1,2,3- cd]pyrene and benzofluoranthenes65,66. BaP is considered carcinogenic to humans by the IARC67. Additionally, more than 500 other PAHs have been identified or partially identified in cigarette smoke68. Other combustion products found in cigarette smoke include aldehydes such as formaldehyde, acetaldehyde and acrolein; heterocycles such as furan; volatile nitrosamines such as dimethylnitrosamine; aromatic amines such as 4-aminobiphenyl and 2-naphthylamine; heterocyclic aromatic amines such as 2-amino-1- methyl-6-phenylimidazo[4,5- b]pyridine; and volatile hydrocarbons such as 1,3-butadiene and benzene69. Thus, most carcinogens in cigarette smoke are combustion products. PAHs are perhaps the most extensively studied of all tobacco smoke carcinogens, and considerable data, reviewed previously, indicate that they are among the most important causes of cancer in cigarette smokers1,5,16,67. DNA adducts fromcigarette smoking DNA adduct formation is the critical step in mutagenesis induced by tobacco (Fig. 1). There is no doubt that DNA adducts resulting from exposure to the complex mixture of genotoxic compounds in tobacco products are central in the carcinogenic process70–72. When DNA repair systems in the cell are inefficient or error-prone in their response to DNA adducts, the adducts can persist, causing mutations when adducted bases are misread and the wrong base is inserted by DNA polymerases. The result is a permanent mutation, which may occur in critical genes involved in growth control, ultimately leading to cancer. Thus, DNA adducts form the requisite link between exposure to carcinogens or their metabolically formed electrophiles and mutations in critical genes, although the relationship between mutations and specific DNA adducts is complex73. Multiple studies have reported DNA adduct levels in cells and tissues of smokers comparedwith non-smokers, andmost of these studies demonstrate significantly higher levels of DNA adducts in smokers, but some of the methods used, such as immunoassays and 32P postlabelling, are only semiquantitative or did not identify, or misidentified, the relevant carcinogens. Difficulties in obtaining adequate amounts of relevant tissue DNA and questions of the timing of adduct formation and persistence are additional challenges in most DNA adduct studies, which continue to be relatively small. Comprehensive reviews of these studies are available74–77. We focus here on selected recent studies that used validated mass spectrometric methods for DNA adduct analysis; this topic has also been reviewed recently55. Tobacco-specific DNA adducts. Metabolism by α-hydroxylation at the 2′ position of NNN or the methyl group of NNK produces a reactive intermediate that pyridyloxobutylates DNA. When this DNA is treated with acid, 4-hydroxy-1-(3-pyridyl)-1-butanone (HPB) (TAble 1) is released and can be quantified by mass spectrometry. A highly sensitive high-resolution mass spectrometric method for analysis of HPB released from DNA was applied to oral cell samples obtained from65 current daily cigarette smokers, 30 with head and neck squamous cell carcinoma (HNSCC) and 35 cancer- free controls78. Significantly higher levels of HPB were released from the DNA of smokers with HNSCC than from the DNA of smokers who were cancer-free. These intriguing results require confirmation in larger studies and could lead to a test for early detection and/or prevention of these disfiguring and often fatal cancers. Nature reviews | CanCer Rev i ews volume 22 | march 2022 | 147
VOLUME 22 | MARCH 2022 | 41 NATURE REVIEWS | CANCER REV I EWS 0123456789( ) ; : BaP DNA adducts. The major DNA adduct of BaP is BaP diol epoxide N2-deoxyguanosine (BPDE- N2-dG) (TAble 1), but reliable detection and quantitation of this intact adduct in human tissues was lacking until recently. We developed a liquid chromatography–nanoelectrospray–tandem high-resolution mass spectrometry method for its analysis in human lung tissue79. The detection limit was one adduct per 1011 nucleotides (corresponding to about 1amol of the adduct). Application of this method to human lung tissue DNA from 29 individuals resulted in an average level of 3.1 adducts per 1011 nucleotides in smokers, three times higher than in non-smokers. While this study identified the presence of the intensively studied BPDE- N2-dG as the intact adduct in pulmonary DNA of smokers, further studies with larger sample sizes are required. In addition, as mentioned earlier, BaP is only one of multiple carcinogenic PAHs in cigarette smoke, so quantitation of total PAH DNA adducts in DNA from smokers is an important goal. 1,3-Butadiene DNA adducts. 1,3-Butadiene is a simple volatile hydrocarbon, is considered carcinogenic to humans by the IARC and is one of the most abundant carcinogens in cigarette smoke. A metabolite of 1,3-butadiene, 3.4-epoxy-1-butene (EB), reacts with DNA to form adducts, among which is a guanine N7 adduct named ‘EB-GII’ (TAble 1). This adduct is removed fromDNA either by spontaneous depurination or repair and is excreted in urine. It has been quantified in cigarette smokers as a biomarker of 1,3-butadiene exposure and metabolism; its levels decrease upon smoking cessation80. Acrolein DNA adducts. Acrolein is an abundant and highly toxic compound in cigarette smoke; it is considered probably carcinogenic to humans by the IARC. Table 1 | Structures of some compounds mentioned in the text Compound Structure Function Benzo[a]pyrene (BaP) Highly carcinogenic polycyclic aromatic hydrocarbon found in all combustion products, including tobacco smoke 4-Hydroxy-1-(3-pyridyl)- 1-butanone (HPB) N OH O Tobacco-specific compound that can be released fromDNA or haemoglobin of people who use tobacco products and has been used as amonitor of tobacco-specific DNA damage Benzo[a]pyrene diol epoxide N2-deoxyguanosine (BPDE- N2-dGuo) N N N N N dR O H OH OH OH H Major DNA adduct formed by metabolic activation of benzo[a]pyrene via its 7,8-diol 9,10-epoxide Guanine N7 adduct of 3,4-epoxy-1-butene (EB-GII) NH2 N N N N O H HO DNA adduct formed from the carcinogen 1,3-butadiene (8R/S)-3-(2′-deoxyribos-1′-yl)-5,6,7, 8-tetrahydro-8-hydroxypyrimido [1,2- a]purine-10(3H)-one (γ-OH-Acr-dGuo) N N N N N O OH H dR Major DNA adduct formed from acrolein (Z)-7-[1R,2R,3R,5S)-3,5-Dihydroxy2-[(E,3S)-3-hydroxyoct-1-enyl] cyclopentyl]hept-5-enoic acid (8- iso-PGF2α) CO2H OH OH OH Biomarker of oxidative damage found in the urine and blood of all humans Phenanthrene tetraol (PheT) HO OH HO HO Metabolite of the polycyclic aromatic hydrocarbon phenanthrene formed by the diol epoxidemetabolismpathway andexcreted in theurine. It has beenused as abiomarker ofmetabolic activationof polycyclic aromatic hydrocarbons www.nature.com/nrc Rev i ews 148 | march 2022 | volume 22
42 | MARCH 2022 | VOLUME 22 www.nature.com/nrc REV I EWS 0123456789( ) ; : Multiple studies show that acrolein reacts with DNA to form exocyclic 1,N2-deoxyguanosine adducts, which were detected in various human tissues81,82. Oral cells are a target for DNA adduct formation in cigarette smokers as the oral cavity is the first site of exposure77. In a recent study, we found average levels of the exocyclic adduct (8R/S)-3-(2′-deoxyribos-1′-yl)-5,6,7,8-tetrahydro- 8- hydroxypyr imido[1,2- a]pur ine-10(3H)- one (γ-OH-Acr-dGuo) (TAble 1) were 27 times higher in oral cells of smokers than of non-smokers83. However, no difference was seen in the levels of these adducts in lung tissue or leukocytes from smokers versus non- smokers84,85, and further studies are required to investigate these differences. Mutational signatures andmutated genes Most studies on mutational signatures potentially traceable to specific DNA adducts and carcinogen exposures have been performed using samples from smokers. In one study, somatic mutations were measured and compared in more than 5,000 tumour samples from cigarette smokers and non-smokers86. Five mutational signatures were increased in smokers compared with non-smokers. Important among these was the mutational signature assigned to exposure to BaP (and likely other PAHs), called ‘SBS4’ in the Catalogue of Somatic Mutations In Cancer (COSMIC) database, which was observed particularly in lung and larynx cancers. These results are consistent with the massive amount of data implicating PAHs as critically important carcinogenic agents in tobacco smoke, as noted earlier herein. Overall, however, the generation of mutations from cigarette smoking as measured in that study was complex and not fully understood, likely reflecting the complexity of cigarette smoke and the contribution and interactions of multiple genotoxic and carcinogenic agents, and it is still not clear which other constituents contribute to the observed changes86. A further study sequenced whole genomes derived from normal bronchial epithelial cells in smokers and non-smokers87. Smoking added thousands of mutations per cell, including driver mutations, and the signature assigned to BaP exposure. Significant within-person variation in mutational burden was observed, also reflecting the complexity of carcinogen exposure from tobacco smoke87. One recent study examined exome sequences and copy number profiles in 660 lung adenocarcinoma and 484 lung squamous cell carcinoma (SCC) tumour and normal pairs; 38 genes were significantly mutated in adenocarcinoma and 20 were significantly mutated in SCC88. Multiple mutated genes were discovered, with distinct mutational patterns in adenocarcinoma and SCC. Among the multiple significantly mutated genes in adenocarcinoma were KRAS, KEAP1, EGFR, STK11, SMARCA4, RBM10, TP53, NF1 and RB1, while in SCC they included TP53, RB1, CDKN2A, NFE2L2, PTEN and MLL2 (also known as KMT2D). These results are completely consistent with the many carcinogens in tobacco smoke, most of which form multiple different types of DNA adducts. Fewer studies have characterized changes in samples originating from smokeless tobacco users. While a comprehensive genomic characterization of HNSCCs has been published, it is not specifically focused on smokeless tobacco users89. The India Project Team of the International Cancer Genome Consortium reported on the mutational landscape of gingivobuccal oral SCC90, which is prevalent in regions such as India where tobacco chewing is common. Significantly and frequently altered genes specific to this cancer type included USP9X, MLL4, ARID2, UNC13C and TRPM3, while others were shared with HNSCCs, including TP53, FAT1, CASP8, HRAS and NOTCH1. A high proportion of G–T transversions was observed. Additional smokeless tobacco-associated genetic alterations were identified in cell lines established from gingivobuccal oral SCC; these included PCLO, FAT3 and SYNE2 mutations as well as oncogenic PIK3CA mutations90. A further study of patients with human papillomavirus-negative early-stage tongue cancer who were habitual chewers of betel nuts, areca nuts, lime or tobacco revealed common G–T transversions and mutations in TP53, NOTCH1, CDKN2A, HRAS, USP6, PIK3CA, CASP8, FAT1, APC and JAK1 among other genes91. Some of the same mutational spectra as well as other mutational spectra were identified by exome sequencing of oral SCC in users of shammah, an Arabian smokeless tobacco preparation92. The relationship of these changes to DNA adduct formation by 5′-hydroxylation of NNN, the major metabolism pathway observed in human enzyme systems, is unclear45. DNA adducts formed by this pathway have been characterized (Fig. 3), but their mutational consequences are unknown45,93,94. Urinary metabolite biomarkers Biomarkers of exposure can potentially identify those cigarette smokers at highest risk of cancer. These individuals can be targeted for intensive smoking cessation interventions and lung cancer screening. The quantitation of urinary metabolites of cigarette smoke constituents as biomarkers of exposure, and in some cases biomarkers of cancer risk, has matured in recent years mainly due to advances in high-throughput liquid chromatography–mass spectrometry techniques, allowing relatively large studies with thousands of urine samples to be performed. These are significant advances permitting tobacco exposure science to proceed from the earlier approaches, which depended mainly on machine measurement of smoke constituents. We present some recent studies as examples of the application of several urinary biomarkers: total nicotine equivalents (TNEs), NNAL, phenanthrene metabolites and (Z)-7-[1R,2R,3R,5S)-3,5-dihydroxy-2- [(E,3S)-3-hydroxyoct-1-enyl]cyclopentyl]hept-5-enoic acid (8- iso-PGF2α), which is a biomarker of oxidative damage (TAble 1). Total nicotine equivalents Urinary levels of nicotine and several of its metabolites constitute the TNE biomarker. Various combinations have been used, but one recent study is illustrative95. The combination of the molar quantities of nicotine, cotinine, 3′-hydroxycotinine and their glucuronide conjugates plus nicotine N-oxide, cotinine N-oxide, nornicotine and norcotinine in urine accounts for more than 90% Nature reviews | CanCer Rev i ews volume 22 | march 2022 | 149
VOLUME 22 | MARCH 2022 | 43 NATURE REVIEWS | CANCER REV I EWS 0123456789( ) ; : of the nicotine dose and is therefore not substantially affected by metabolic differences between individuals. Comparison of this panel with the total urinary molar levels of nicotine, cotinine and 3′-hydroxycotinine and their glucuronides demonstrated a strong correlation; thus omission of the other technically more demanding minor urinary metabolite assays did not affect the overall results. TNEs were quantified in smokers from five ethnic groups with differing risks of lung cancer in the Multiethnic Cohort Study: African Americans, Native Hawaiians, white Americans, Latino Americans and Japanese Americans96. The highest levels of TNEs were found in African Americans, with intermediate levels in white Americans, and the lowest levels were found in Japanese Americans, consistent with their lung cancer risk, but the amounts in Native Hawaiians and Latino Americans did not clearly relate to lung cancer risk. Further studies demonstrated that low-activity forms of CYP2A6, the principal nicotine-metabolizing enzyme, accounted for the lower levels of TNEs in Japanese Americans because more unchanged nicotine remained in the body, thus alleviating the need for more intense smoking97. The reduced risk of lung cancer in smokers with lower CYP2A6 activity was explained by lower consumption of cigarettes, less intense smoking and reduced CYP2A6-catalysed activation of NNK98, while greater CYP2A6 activity causes smokers to smoke more and therefore have a higher risk of lung cancer99. NNAL plus its glucuronides (total NNAL) NNAL, a major and highly carcinogenic metabolite of NNK, is found in free form and as its glucuronide in the urine of all smokeless tobacco users and cigarette smokers. Total NNAL levels in the Population Assessment of Tobacco and Health (PATH) study, a US-representative longitudinal cohort study of approximately 46,000 adults and youths, aged 12 years or older, were highest in the urine of approximately 11,000 everyday users of smokeless tobacco (993ngg−1 creatinine), followed by cigarette smokers (285 ng g−1 creatinine) and e-cigarette users (6.3 ng g−1 creatinine); these amounts compared with 1.0 ng g−1 creatinine in non-users, which is consistent with the tobacco specificity of NNK and its relatively high levels in smokeless tobacco100. Similar results were obtained in the US National Health and Nutrition Examination Survey (NHANES), which was based on a different nationally representative sample of the US population. The urinary level of total NNAL in smokeless tobacco users averaged 583ngg−1 creatinine, while in exclusive cigarette smokers it was 218ngg−1 creatinine101. The levels of total NNAL in urine often correlate with TNEs100. As in the analysis of TNEs in the Multiethnic Cohort Study, urinary levels of total NNAL in 2,252 smokers were consistent with lung cancer risk in African Americans (highest), white Americans (intermediate) and Japanese Americans (lowest), but not in Native Hawaiians and Latino Americans102. Polycyclic aromatic hydrocarbons Urinary phenanthrene metabolites, 1-hydroxypyrene and related PAHs were monitored in the NHANES as biomarkers of PAH exposure. Consistently, the levels of these metabolites are significantly higher in smokers than in non-smokers, and cigarette smoking remains one of themajor sources of human exposure to PAHs103. Urinary phenanthrene tetraol (PheT; TAble 1) is an excellent biomarker of PAH exposure and of metabolic activation of PAHs by the well-established diol epoxide pathway that results in DNA adduct formation by multiple PAH carcinogens, including BaP104. The metabolism of phenanthrene to PheTmimics that of BaP to its ultimate carcinogen, but phenanthrene metabolites are more readily measured in urine because their concentrations are about 1,000 times higher than those of BaP metabolites105,106. We took advantage of this strategy to determine whether the activation of aryl hydrocarbon receptor (AHR) and the resulting induction of cytochromes P450 by cigarette smoking, which has been known for more than 50 years107, resulted in the increased metabolic activation or detoxification of this representative PAH, as measured by the amounts of deuterated PheT and phenanthrene phenols in the urine of 170 smokers and 57 non-smokers given defined doses of deuterated phenanthrene, thus avoiding possible complications due to exposure to phenanthrene in the environment108. This clearly demonstrated that cigarette smoking increases the metabolic activation of phenanthrene via the diol epoxide pathway, consistent with the conclusion that cigarette smoking enhances the metabolic activation and carcinogenicity of PAHs via activation of AHR and induction of CYP1A1, CYP1A2 and CYP1B1. Urinary biomarkers and cancer risk Collaborative studies with Jian-Min Yuan, who leads the Shanghai Cohort Study and the Singapore Chinese Health Study, both of which are prospective molecular epidemiology studies, have resulted in important findings relating urinary biomarkers and CYP2A6 genetic polymorphisms to lung cancer risk in cigarette smokers98,109,110. The Shanghai Cohort Study collected a single urine sample and followed up more than 18,000 men since 1986–1989, and the Singapore Chinese Health Study obtained a single urine sample frommore than 63,000 Chinese men and women followed up since 1993–1998. These single-void urine samples were stored frozen until biomarker analysis in the laboratory, beginning in 2007. The relationships of the urinary biomarkers to lung cancer risk were determined in participants, all cigarette smokers. The results were adjusted for smoking intensity and duration. In the Shanghai Cohort Study, the smoking-adjusted odds ratio of lung cancer for the highest quartile versus the lowest quartile of TNEs was 4.71, while for total NNAL it was 3.15, both of which are significant. Similar results were obtained in the Singapore Chinese Health Study. In both studies, lower nicotine metabolism (as determined by CYP2A6 phenotype (3′-hydroxycotinine to total cotinine ratio) and CYP2A6 genotype) was also significantly associated with a reduced risk of lung cancer98,109. Further studies demonstrated significant relationships between urinary PheT and lung cancer risk and between urinary NNN and oesophageal cancer risk110; however, urinary mercapturic acid metabolites of volatile organic compounds, such as acrolein, benzene and 1,3-butadiene, were not Creatinine A waste product from normal metabolism that is commonly used as a denominator in biomarker studies. www.nature.com/nrc Rev i ews 150 | march 2022 | volume 22
44 | MARCH 2022 | VOLUME 22 www.nature.com/nrc REV I EWS 0123456789( ) ; : related to lung cancer risk in smokers, after correction for urinary cotinine and years of smoking111. Thus, selected urinary biomarkers and, importantly, CYP2A6 genotype, as shown by the Transdisciplinary Research in Cancer of the Lung (TRICL) consortium99, have the potential to identify smokers at high risk of cancer, a critical element in cancer prevention. While these biomarkers relate directly to carcinogen exposure, 8- iso-PGF2α is an accepted urinary biomarker of oxidative damage, likely involved in co-carcinogenesis and tumour promotion by cigarette smoke. 8- iso-PGF2α was also significantly associated with lung cancer risk among smokers and former smokers, but not non-smokers, in the Shanghai Cohort Study112. Overall, these relationships of urinary biomarkers of carcinogen exposure and oxidative damage recapitulate in smokers the known effects of cigarette smoke constituents observed in experimental studies with laboratory animals39,52,113,114. In summary, these biomarker studies in cigarette smokers are consistent with the hypothesis that NNK and PAHs are causative agents for lung cancer, that NNN is a causative agent for oesophageal cancer and that oxidative damage as represented by 8-iso-PGF2α is important in human lung carcinogenesis. The studies also demonstrate that TNEs and CYP2A6 polymorphisms are potentially important biomarkers of cancer risk. We envision that a combination of biomarkers could be incorporated into a predictive algorithm for cancer risk to improve lung cancer screening and hopefully prevent fatal lung cancers. Regulatory approaches We recognize that there are multiple critically important approaches to prevention of cancer due to tobacco use, including public health strategies (for example, taxation), recognition and treatment of preneoplastic lesions of the oral cavity and low-dose CT lung cancer screening115–117. Each of these could be the basis of a separate and extensive review. In this Review, we focus on examples of the approaches that we have explored in our interdisciplinary research, involving mainly regulatory strategies related to reducing harmful constituents in tobacco products (Articles 9 and 10 of the WHO Framework Convention on Tobacco Control)118. Regulating smokeless tobacco One approach to reducing cancer risk is to establish standards for the levels of harmful constituents in smokeless tobacco119. This approach has been implemented by Swedish Match, a tobacco company that manufactures snus, and is referred to as the ‘GOTHIATEK standard’120. This company regulates its smokeless tobacco products tomeet certain standards such as 0.95µgg−1 of NNN plus NNK, which is a likely reason that there is a lower risk of oral and oesophageal cancer from smokeless tobacco use in Sweden compared with South-East Asia121,122. However, few smokeless tobacco manufacturers have established limits on harmful constituents in their products. In 2009, the WHO Study Group on Tobacco Product Regulations recommended, under Article 9 of the WHO Framework Convention on Tobacco Control, that industry-manufactured smokeless tobacco products should not exceed 2μgg−1 dry tobacco weight for NNN plus NNK and 5ngg−1 for BaP, and that the levels of arsenic, cadmium and lead in tobacco should be monitored by regulatory authorities123. In 2017, the FDA issued an advance notice of proposed rulemaking (ANPRM) proposing a limit of NNN of 1.0 μg g−1 dry tobacco weight124. The FDA deemed that there was insufficient evidence linking NNK to cancer in smokeless tobacco users and thereby focused the product standards on NNN (although reduction inNNN levels is likely to also lead to a reduction in NNK levels). Additionally, because NNN levels could increase over time in smokeless tobacco products, the rule would also require an expiration date on each batch of smokeless tobacco products and the manufacturers would have to show that the NNN levels are sustained below the allowable limit until the expiration date. Furthermore, instructions for storage would also be provided on the label if storage conditions, which are known to influence NNN levels124, impact NNN levels in a given product. It is important to recognize that implementation of product standards does not indicate that the product is safe, and tobacco companies should not be allowed to directly or indirectly promote or market their products in this manner, or declare that the product has been approved by the government123. The NNN product standard is expected to lead to a reduced incidence of oral cancer124, but unfortunately few countries have imposed any standards. Furthermore, a product standard may be very difficult to implement in some countries, such as those in South-East Asia, where the rate of smokeless tobacco use is high and there are a plethora of different smokeless tobacco products, some which are handmade by consumers or street vendors. Potentially, educating consumers and vendors about the differing amounts of carcinogens in smokeless tobacco and factors that might contribute to the formation of carcinogens such as tobacco type, processing and storage might be a first step, as reviewed by the WHO125,126. Other means to reduce the prevalence of smokeless tobacco use include restricting sales to minors, requiring prominent health warning labels, restricting or banning advertising, promotion or sponsorship of smokeless tobacco products, taxation and pricing policies, providing public education and promoting or provision of evidence-based smokeless tobacco use cessation treatments (for extensive reviews, see reFS4,125). Regulating cigarettes Reducing nicotine levels in cigarettes to minimally addictive levels would likely lead to a reduction in the prevalence of smoking by significantly dampening the trajectory towards cigarette dependence among naive tobacco users and facilitating the cessation of cigarette use among smokers127,128. A substantial body of scientific literature demonstrates that if the nicotine level in cigarettes is reduced by about 95%, then reductions in the number of cigarettes smoked, cigarette dependence and exposure to carcinogens and toxicants129–131 are observed along with an increase in quit attempts131 or smoking cessation129. Reductions in smoking behaviours and dependence have been observed among young adults132, smokers of low socio-economic status133, smokers in Advance notice of proposed rulemaking (ANPrM). A document that an agency such as the US Food and Drug Administration may choose to issue before it is ready to issue a notice of proposed rulemaking. it is the first public step in the notice and comment rulemaking process, and the comments received could affect the final rule making process. Nature reviews | CanCer Rev i ews volume 22 | march 2022 | 151
RkJQdWJsaXNoZXIy MTYzOTI3MA==