Oncogenetics of Lung Cancer Induced by Environmental Carcinogens

The molecular landscape of non-tobacco-induced primary lung tumors displays specific oncogenetic features. The etiology of these tumors has been largely associated with exposure to well-established environmental lung carcinogens such as radon, arsenic, and asbestos. Environmental carcinogens can induce specific genetic and epigenetic alterations in lung tissue, leading to aberrant function of lung cancer oncogenes and tumor suppressor genes. These molecular events result in the disruption of key cellular mechanisms, such as protection against oxidative stress and DNA damage-repair, which promotes tumor development and progression. This chapter provides a comprehensive discussion of the specific carcinogenic mechanisms associated with exposure to radon, arsenic, and asbestos. It also summarizes the main protein-coding and non-coding genes affected by exposure to these environmental agents, and the underlying molecular mechanisms promoting their deregulation in lung cancer. Finally, the chapter examines the anticipated challenges in personalized intervention strategies in non-tobacco-induced lung cancer.


Introduction
Lung cancer remains the deadliest form of cancer across the globe [1]. While smoking rates have decreased in many areas, it remains to be seen if the incidence and mortality of primary lung cancer will experience a similar shift, particularly in light of the observation that close to 25% of cases arise in individuals who have never smoked [2]. As one of the most environmentally-influenced malignancies, lung tumorigenesis can result from exposure to both physical and chemical carcinogens. Exposure to the mix of compounds present in particulate matter is another well-known factor affecting the development of lung cancer [3]. However, a number of single-agent compounds in the environment have been identified as key lung carcinogens, particularly arsenic, asbestos and radioactive radon ( 222 Rn) gas [4]. These compounds are distributed at varying, potentially-dangerous concentrations in the environment, affecting hundreds of millions of people worldwide.
Exposure to each of arsenic, asbestos, and radon has been shown to induce widespread genetic and epigenetic alterations, which may account for their strong carcinogenicity, independent of smoking status [4]. Interestingly, the molecular aberrations associated with these compounds and the onset of lung cancer in never-smokers follows a mechanism distinct from that of tobacco smoke [5]. While strict guidelines regarding exposure to these compounds have been implemented in some regions, mounting evidence suggests that carcinogenic effects may result from chronic exposure to environmental levels that are well below those currently deemed "safe" [6,7]. Additionally, individual differences may contribute to varying degrees of susceptibility to the carcinogenic effects of these compounds. For instance, women have been shown to have a higher incidence of lung cancer arising in never-smokers. This inequality can potentially be attributed to a historical bias towards women being more present in the home, resulting in increased exposure to high radon concentrations and polyaromatic hydrocarbons from various home combustion sources [8]. As these genetic and epigenetic aberrations might be indicative of specific molecular damage induced by these carcinogens, they may be able to be used to develop personalized approaches for risk assessment, monitoring and subsequent disease treatment. Thus, it is critical to uncover the extent of these events associated with exposure to environmental carcinogens.
Arsenic is a class I International Agency for Research on Cancer (IARC) carcinogen that threatens global health through its persistent accumulation in drinking water sources, leading to the onset of skin and lung cancers, among other diseases [9]. Asbestos fibers are naturally occurring silicate mineral fibers that have long been used in industry as building insulation, and are closely linked with not only the well-known outcome of mesothelioma, but also to 5-7% of all lung cancer cases [10]. Radon gas accounts for between 3 and 14% of all lung tumors in a given country and is the second most-common cause of lung cancer, behind smoking [11]. While the radioactive gas normally diffuses easily in open air, it can build up in indoor environments and is readily dissolved into water, which can lead to malignancies through radioactive decay and alpha particle emission [11]. Moreover, drinking water may be a particularly prevalent source of exposure to environmental carcinogens, as it is a primary route of exposure for both arsenic and radon, emphasizing the need for a focus on water contamination measurement and remediation. As arsenic, asbestos, and radon exert their carcinogenic effects through different exposure routes, they display similar, yet distinct mechanisms of genetic and epigenetic aberration, which may be useful in the identification and treatment of tumors caused by these agents.
In this chapter we highlight the molecular alterations induced by exposure to arsenic, asbestos, and radon in key lung cancer pathways, and finish with a discussion of the potential translational applications of environmentally-induced molecular damage.

Physiological and molecular impact of exposure
Arsenic exposure largely occurs through contaminated drinking-water sources, but this problem extends well beyond known arsenic-endemic areas. In fact, it is estimated that 200 million individuals are exposed worldwide to levels deemed non-toxic by the WHO, but shown to induce molecular damage [12].
The toxic effects of arsenic are prevalent from ingestion to excretion and are largely attributed to its various metabolites (Figure 1). Once ingested, arsenate (As V )-the most common form of the compound in the environment-is taken into cells through membrane transporters, where it is quickly reduced to arsenite (As III ) by oxidoreductases including purine nucleotide phosphorylase (PNP) and glutathione-s-transferase omega (GSTO). As III is the most toxic form of arsenic, largely due to its subsequent methylation by methyltransferase enzymes such as arsenic (+3) methyltransferase (As 3 MT), a process exploited for promoting the excretion of arsenic [13]. However, methyl groups are provided by S-adenosylmethionine (SAM), a key cellular methyl group donor. Methylation of arsenic inside the cell can thus lead to the depletion of the cellular methyl pool through a high demand on SAM, which then promotes global DNA hypomethylation and aberrant histone modification [14][15][16][17]. Disruptions in the cellular methyl pool can lead to major disruptions in gene expression, which is known to contribute to malignant transformation [16].
The genomic instability and global changes in gene expression resulting from the exposure and biotransformation of arsenic is exacerbated by the widespread induction of DNA damage from toxic arsenic byproducts. In fact, arsenic has been demonstrated to cause distinct alterations in chromatin, gene expression (both coding and non-coding), as well as splicing, and transcription initiation [18]. In particular, one of the methylated species of arsenic, monomethylarsonic acid (MMA III ), can interrupt the electron transport chain in mitochondria, liberating electrons and inducing the formation of reactive oxygen species (ROS) [15,19,20]. ROS generated from arsenic exposure result in widespread DNA damage, including single-and double-stranded DNA breaks, DNA base oxidation leading to mutations (largely G>C → T>A transversions), adducts, deletions and even damage to mitochondrial DNA (mtDNA) [20][21][22]. Unsurprisingly, as oxidative stress is a known driver of tumorigenesis in multiple tissues, the DNA damage induced from arsenic exposure is thought to be a main mechanism of its carcinogenicity [23][24][25]. The disruption of the electron transport chain produces ROS such as hydroxyl radicals (OH•), superoxide anion radicals (O 2 •−), and hydrogen peroxide (H 2 O 2 ), which can further damage cells through lipid oxidation, protein oxidation, and reduction of the mitochondrial membrane potential [26]. The subsequent liberation of cytochrome c can activate apoptotic pathways through caspases, leading to an abnormal rate of cell death. However in addition to faulty apoptotic signaling, exposure to arsenic can also lead to further aberrations in DNA-repair pathways. Here, arsenic affects the expression of genes involved in both nucleotide-(NER) and base-excision repair (BER) mechanisms, allowing the cell to continue through the cell cycle despite extensive damage and genomic instability [27][28][29][30]. Thus, arsenic exposure can induce an array of molecular damage across the genome and epigenome, culminating in malignant transformation.

Carcinogenic mechanisms
While it is exposure to the methylated metabolic byproducts that yields the largest toxic effects resulting from exposure to environmental arsenic, it is noteworthy that even at very low doses, arsenic may be able to act as a co-mutagen to other known carcinogens, such as ultraviolet light, X-rays, methyl methane sulfonate, and tobacco smoke [15]. ROS are perhaps more immediately damaging to cells, as they can lead to alterations in a variety of lung cancer-specific pathways. As stated previously, arsenic exposure can interfere with DNA damage repair pathways, which exacerbates the effects of ROS generation. In the NER pathway, arsenic can alter the expression of key damage-repair genes, such as XPC, in a process that may be mediated by the proteasome [31].
Collectively, aberrations in cellular DNA-damage repair pathways may not only highlight mechanisms of arsenic toxicity, but also its co-mutagenic effects. One of the most common pathways affected in lung cancer is the constitutive activation of the epidermal growth factor receptor (EGFR), especially in women and individuals who have never smoked [32]. Both amplification and mutation can lead to EGFR activation, which subsequently stimulates cell proliferation. As III can activate protooncogene c-Src (c-Src) through vicinal sulfhydryl groups, which then promotes phosphorylation events in intracellular EGFR tyrosine residues (Tyr845) [32]. As tyrosine phosphorylation is a key event in EGFR activation, As III thus promotes EGFR constitutive signaling. Alternatively, arsenic exposure may also indirectly affect downstream members of the EGFR pathway, through arsenic-induced oxidative stress and ROS, a common mechanism of environmentally-induced lung carcinogenesis. In a mechanism similar to that of EGFR activation, arsenic has been shown to induce the phosphorylation of several potential substrates of protein kinase B (Akt), a regulator of epithelial-to-mesenchymal transition (EMT) and metastasis, inducing cell migration [33]. Specifically, arsenic may affect c-Jun N-terminal kinase (JNK) activation and subsequent activation of signal transducer and activator of transcription 3 (STAT3), resulting in Akt growth and migration signaling [34]. Similarly, arsenic may increase the enzymatic activity of phosphoinositide 3-kinase (PI3K) and Akt phosphorylation, a key pathway in lung cancer tumorigenesis and progression [35]. The mechanism of PI3K/AKT activation has proven elusive, yet evidence suggests that ROS may play a mediating role, as well as alterations in histone modifications and activation of other related pathways, such as EGFR, mammalian target of rapamycin (mTOR), or polo-like kinase 1 (PLK1) signaling [35,36]. Phenotypically, activation of the PI3K/Akt signaling axis by arsenic can result in a variety of changes, including cellular growth and angiogenesis [37]. There are many other lung cancer-specific pathways that may be altered upon exposure to arsenic and its toxic byproducts, including the nuclear factor (erythroid-derived 2)-like 2/kelch-like ECH-associated protein 1 (NRF2/ KEAP1) pathway, the nuclear factor kappa-light-chain-enhancer of activated B cells pathway (NF-κB), and various epigenetic pathways [35,38]. Further experimental work is required to fully characterize and distinguish the molecular mechanisms of the pathways affected by chronic exposure to arsenic.

Prominent cancer genes affected by arsenic
As evidenced by its genome-wide effects on cellular physiology and molecular pathways, gene expression alterations cause by arsenic exposure can potentiate negative health outcomes. In fact, there are a growing number of genes that have been observed to have abnormal expression resulting from arsenic exposure, in cell lines, mouse, and human samples. Many of these genes have accepted roles in cancer, both as tumor-suppressors and oncogenes. Most notably, the tumor suppressor gene TP53 has been shown to be epigenetically inactivated in arsenic-exposed cell lines [39]. Similarly, other cell line studies have suggested that low concentrations of arsenic may upregulate the known lung oncogene Myc (also related to the cell cycle) through aberrant expression of miRNAs targeting upstream regulators of its transcription [40].
As previously discussed, the frequent disruption of DNA damage repair and stress response pathways is a common feature of arsenic-induced lung tumors. Notably, arsenic has been associated with stimulation of the DNA damage response through the upregulation of critical genes, such as the gene encoding DNA excision repair protein ERCC1 (ERCC1) [41], confirming that DNA damage is prevalent in arsenic-exposed individuals. Alternatively, arsenic may induce repression and decreased activity of main DNA repair enzymes, including poly [ADP-ribose] polymerase 1 (PARP1) inhibition (through ROS) [42], proteasomal degradation of xeroderma pigmentosum, complementation group C (XPC) [31], and widespread hypermethylation of NER genes [43]. Additional lung cancer-related genes affected by arsenic include: EGFR [44], cyclin-dependent kinase inhibitor 1A (CDKN1A) [45], and B-cell lymphoma 2 (BCL2) [46]. Despite the mounting evidence of the toxic effects of arsenic, the concentration and identity of key damage-related arsenic compounds varies widely between studies. While different arsenic-based compounds affect similar pathways, specific physiological responses may vary greatly depending on compound type and dose response, necessitating closer examination of these factors in future studies.
However, it is important to note that variations in these genes may exist within individuals prior to arsenic exposure, and that certain genetic polymorphisms may make some individuals more susceptible to the genotoxic effects of arsenic. For instance, a single nucleotide polymorphism (rs238406; C > A) in ERCC2 (part of the DNA-damage response) leads to the inclusion of an alanine residue in the place of a cysteine in the complete protein, increasing an individual's odds ratio for skin cancer to 2.04 [47]. Additionally, polymorphisms in many of the genes involved in the metabolism and biotransformation of arsenic may result in the production of different metabolic byproducts, conferring differential susceptibility and cancer risk [48]. This is exemplified by the rs1191439 polymorphism of As 3 MT, which is correlated with elevated MMA levels in urine [49]. Thus, the landscape of arsenicinduced carcinogenesis is quite complex, with multiple types and outcomes of the molecular aberrations that can result from chronic exposure. A more comprehensive understanding of the mechanisms at play may result in the identification of the underlying causes of lung cancer in never-smokers, and may help to direct the development of novel treatment strategies for these affected individuals.

Physiological and molecular impact of exposure
Asbestos is a term used to define a group of mineral fibers incorporated in a wide variety of products, including talcum powder, brake pads, and construction materials. While more than 50 countries have banned the use of asbestos-containing materials, more than 2 million metric tonnes are still produced every year, which still poses a great public health risk for asbestos-related diseases [50,51]. There are two main classes of asbestos: chrysotile (spiral-shaped, the most common form) and amphibole (needle-shaped). Other elements such as iron (which can constitute up to 30% of the weight of asbestos fibers) embedded in the surface of fibers can potentiate asbestos-related pathogenic effects [52,53]. Importantly, all identified forms of asbestos have been classified as carcinogens to humans (Group 1) by the IARC [54].
Exposure to asbestos fibers has been strongly linked to the development of malignant mesothelioma, but it is also a known contributor to the development of lung cancer [55][56][57]. Between 5 and 7% of all lung cancer cases worldwide have records of high levels of asbestos, mostly derived from occupational exposure (e.g., mining) [10]. Exposure is usually determined by the presence of pleural plaques (areas of fibrosis associated with past exposure to asbestos), or by detection of asbestos fibers in bronchoalveolar lavage (BAL) and lung tissue [58]. The primary source of asbestos exposure comes from inhaled fibers [54]. However, the mechanism of disruption that occurs as a result of asbestos exposure is determined by the efficiency of fiber clearance from airway cells. Longer fibers are cleared at a slower rate than short fibers, and are associated with higher carcinogenic potential [59]. Similarly, thin fibers (width <0.25 μm) are more carcinogenic than thicker ones [60], likely because they can penetrate deeper in airways. Accumulation of asbestos fibers in the lung leads to fibrosis, inflammation, and carcinogenesis, although specific effects depend on the cumulative dose and the type of fiber inhaled [61,62].
Asbestos-related carcinogenesis is thought to primarily result from the ability of the fibers to induce oxidative stress (Figure 2), although the specific mechanisms are not yet fully understood [63]. Asbestos induces the recruitment of alveolar macrophages, followed by an inflammatory reaction [64][65][66]. Failed phagocytosis of these fibers by macrophages results in the generation of ROS, together with the release of cytokines, chemokines, proteases, and growth factors further amplifying deleterious effects of asbestos [10,56,67]. Additionally, the iron contained in asbestos fibers deposits in the lungs and cycles between the reduced and oxidized forms, potentially inducing further oxidative DNA damage in nearby cells via the Fenton reaction which converts H 2 O 2 into more reactive ROS [10,56,68,69].
In lungs, oxidative stress following asbestos exposure can activate several signaling pathways including mitogen-activated protein kinases (MAPK), NF-κB, and activator protein 1 (AP1). All of these pathways have been linked to increases in early response genes (e.g., JUN and FOS) that govern cell proliferation, apoptosis, and inflammatory signaling [55,56].

Carcinogenic mechanisms
The most frequent asbestos-induced alterations in cancer-related genes have been reported in tumor suppressor genes (TSGs). Activation of p53 and p21 are frequently described, both in animal models and lung cancer patients with asbestosis (reviewed in [63]). This likely represents the initial DNA-damage response following exposure to asbestos-induced oxidative stress. In lung cancer patients, the frequency of TP53 gene mutations is similar between asbestos-exposed and unexposed NSCLC cases; however, a higher frequency of G:C to T:A transversions in the sequence of TP53 is observed in asbestos-exposed cases [70,71]. Contrarily, other tumor suppressor genes such as CDKN2A are inactivated in asbestos-exposed lung cancer cases, mostly via segmental copy-number losses [72]. In murine models, chrysotile fibers are able to induce the activity of the c-Jun and c-Fos oncogenes and inactivate p53 and p16 tumor suppressors, both at the mRNA and protein levels [73].
Additionally, other well-known lung cancer genes and pathways have been shown to display aberrant functions in response to asbestos exposure. Different mechanisms of asbestos-mediated activation of the EGFR pathway have been described. Asbestos-induced chronic inflammation has been associated with activation of the EGFR-related and extracellular signal-regulated kinase (ERK) signaling pathway that promote lung epithelial cell and fibroblast proliferation [55,56,74]. Also, asbestos fibers can induce over-expression of EGFR mRNA and induce protein dimerization, phosphorylation, and subsequent pathway activation by directly interacting with the surface portion of the receptor [63,75,76]. On the other hand, DNA mutations affecting EGFR do not seem to be main mechanisms of asbestosinduced EGFR activation. Asbestos-exposed patients displayed a significantly lower rate of EGFR mutations compared to non-exposed patients [77]. Moreover, it is unclear if there is a causal relationship between the mutations found in EGFR and exposure to asbestos fibers [78,79].
Other genes, such as MAP4K3, CEBPZ, QPCT, FANCG, IGFBPL1, CCL19, MELK, FANCM, and CDKL1 have shown aberrant gene expression in human epithelial bronchial cell lines (Beas-2B), following asbestos exposure [80]. Asbestos inhalation also causes up-regulation of mRNA levels of matrix metalloproteinase family members in rat lungs, suggesting induction of extracellular matrix remodeling [81]. At the epigenetic level, alterations affecting tumor suppressor genes have been observed in lung cancer cases associated with asbestos exposure, including those in the promoter regions of RASSF1A and CDKN2A (p16) [82]. Additionally, a genome-wide DNA methylation study identified differentially methylated CpGs in regions nearby the transcription start site of genes such as NPTN, NRG2, GLT25D2 and TRPC3 to be significantly associated with asbestos exposure [83].
Despite the known genetic and epigenetic abnormalities resulting from asbestos exposure, a relatively small proportion of exposed individuals develop thoracic malignancies (mesothelioma or lung cancer). It has been hypothesized that specific genetic variants may confer increased risk of developing asbestos-related diseases [85]. Thus, recent studies have investigated the association between genomic variants and risk of lung cancer following asbestos exposure. In a genome-wide association study (GWAS) performed in the Texas lung cancer GWAS dataset, the authors did not find statistical evidence for gene-asbestos interaction in the etiology of lung cancer [86]. However, the Fas signaling pathway (regulation of tissue homeostasis in the immune system by inducing apoptosis) was identified as the most significant pathway associated with asbestos exposure in the etiology of lung cancer. Another study identified three single nucleotide polymorphisms (SNPs) in the MIRLET7BHG (MIRLET7B host gene located at 22q13.31) significantly associated with increased lung cancer risk among individuals exposed to asbestos [36].
The identification of risk variants linked with asbestos-related lung cancer is a challenging task. Sample sizes for asbestos-related lung cancer cohorts are particularly limited by the number of cases that can be unequivocally attributed to asbestos exposure despite other well-known factors (e.g., smoking). Thus, focusing on the genes and chromosomal regions found by these preliminary studies might be useful for more targeted strategies aiming to validate these results.

Carcinogenic potential of other fibers
While the oncogenic effects of asbestos have been extensively established, recent evidence indicates that non-asbestos fibers, both natural and synthetic in nature can also cause thoracic cancers. Non-asbestos mineral (natural) fibers include erionite and fluoro-edenite, among others. Erionite is a naturally occurring fibrous mineral that shares some physical properties with asbestos, although it is less widespread. In fact, it has been shown that erionite is a more potent carcinogen in causing malignant mesothelioma [87,88]. Erionite activates the NLR family pyrin domain containing 3 (NLRP3, NALP3) inflammasome, inducing the transcription and production of cytokines critical to cancer initiation [89]. On the other hand, Fluoro-edenite (originating from volcanic activity) can induce ROS that result in DNA damage and increase in lactic dehydrogenase release (a damage and toxicity marker) in human lung adenocarcinoma (A549) and monocyte-macrophage (J774) cell lines [90]. Synthetic graphene-based fibers are widely used in several industries. They have also been explored as a drug delivery system for cancer treatments. Physical similarities to asbestos, particularly its high length-to-width ratio, have raised some concerns about the potential carcinogenicity effects of these fibers [91]. Exposure to carbon nanotubules has been shown to induce oncogenic pathways, such as TGF-β and Akt/GSK-3β, resulting in activation of the SNAIL-1 signaling pathway and epithelial-mesenchymal transition [92]. Additionally, carbon nanotubules can generate ROS, activating MAPKs, AP-1, NF-κB, and Akt in normal and malignant human mesothelial cells [93]. Other genetic alterations, including micronuclei formation, disruption of mitotic spindles, and polyploidy have also been observed in response to carbon nanotubule exposure [94][95][96]. Moreover, it has been shown that exposure to carbon nanotubules can induce specific methylation changes at the promoter regions several genes, including DNMT1, ATM, SKI, and HDAC4, while they seem to have only a marginal effect on miRNA expression [97]. Thus, the oncogenetic factors of natural and synthetic fibers, while similar in morphology, are distinct entities that may collectively culminate in tumor development.

Physiological and molecular impact of exposure
Radon is the second most common cause of lung cancer in many countries; however, the intricacies of its mechanism of action remain underappreciated. The genotoxicity of radon is largely the result of alpha particle emission during its spontaneous decay into short-lived radioactive progeny ( 218 Po and 214 Po) and comparably long-lived radioactive 210 Pb, which also induces cellular damage through alpha decay (Figure 3) [98].
Alpha decay is the emission of a 4 atomic mass unit helium ion (two protons and two neutrons), which can liberate electrons from water molecules and result in the generation of several types of ROS [15]. Much like the mechanisms of arsenic and asbestos toxicity, ROS generated as a consequence of radon exposure can lead to widespread molecular aberrations, especially base oxidation (leading to mismatches and mutagenesis), DNA strand breaks, chromosomal aberrations, and deletions. For example, chromatid deletions in blood lymphocytes may be a result of radon exposure, which may in part explain the associations between radon exposure and blood malignancies [8]. These events may occur at levels well below those currently deemed safe in many countries, exemplified by the observation of chromosomal abnormalities in lymphocytes at very low doses of polonium-214, a radioactive progeny of radon [99].
Beyond the molecular events resulting from ROS generation, alpha radiation from radon exposure can induce bystander responses in cells that have not been directly affected by alpha particles [100]. The bystander effect of radiation exposure can occur through the release of signals from nearby irradiated cells, generating a physiological response in non-irradiated cells, even at relatively low doses of radiation [101]. The effect requires direct contact between adjacent cells, such as through gap junctions, as well as compounds in the surrounding medium, including cytokines [102]. One of these compounds, nitric oxide (NO), has been shown to be an important factor for the cell-killing effects of the bystander response, largely through the direct interaction with and damage of DNA [103]. Moreover, NO byproducts such as dinitrogen trioxide (N 2 O 3 ) can promote nitrosation of other amines, such as those of DNA bases, leading to cross-linking and DNA alkylation [102]. Another compound that may be relevant to the bystander effect of cellular radiation exposure is cyclooxygenase 2 (COX-2), which is related to the NF-κB pathway, an effect that is attenuated upon COX-2 inhibition [103,104]. Finally, this response may be dependent on TP53 status, which will be discussed in Section 4.3.

Carcinogenic mechanisms
Despite differences in the details of exposure, the molecular mechanisms contributing to carcinogenesis in individuals exposed to arsenic, asbestos, and radon converge in that they all produce ROS. Radon has a half-life of 3.8 days, and as previously mentioned, commonly generates alpha particles and polonium decay products, which themselves emit further alpha radiation [105]. Alpha particles have a high linear energy transfer (LET) despite having relatively low penetration capability, meaning that they interact readily with DNA, especially in regions close to their site of exposure, such as the bronchial epithelium [105]. Thus, it is not surprising that lung malignancies are the most common type of radon-induced cancer. High LET radiation is distinct from low LET radiation (such as x-rays or gamma rays) in that it produces a substantially greater proportion of clustered damage, meaning the occurrence of ≥2 lesions of ≥1 different types within 1-2 helical turns of DNA. Clustered DNA damage is typically repaired with slower kinetics and has a greater likelihood of producing sequence alterations, as repair pathways converge and conflict with one another [106][107][108]. The largest radon-induced mechanisms of carcinogenesis include DNA damage, ROS, and alpha particle generation; likewise, pathways associated with these functions are also known to be associated with lung cancer. In fact, patients positive for rearrangements in the gene encoding anaplastic lymphoma kinase (ALK)-an event frequently found to drive lung tumorigenesis-were found to have two-fold increases in residential radon levels than those without these rearrangements [109,110]. While a synergistic effect between radon and smoking has been suggested [11], the G:C to T:A transversions associated with tobacco-related molecular damage are not as commonly observed in individuals exposed to radon, suggesting a unique molecular signature in radon-associated lung tumors [15]. Again, it is important to note that a number of the pathways affected by radon exposure, including gene expression alterations and apoptotic disturbances, may actually be from cells neighboring those that are irradiated [104]. In fact, pro-inflammatory and ROS-generating cytokines such as tumor necrosis factor alpha (TNF-α) may be released upon radiation exposure, which may perpetuate the damage enacted by ROS [111]. Thus, key pathways such as DNA repair, proliferation, and cell death can be altered in cells beyond those that are irradiated [111].

Prominent cancer genes affected by radon
ROS-induced DNA damage is a large factor in radon-induced carcinogenesis, thus, many of the examinations into genes affected by radon are relevant to DNArepair and apoptotic pathways. Naturally, a heavy focus is placed on TP53. Many investigations into TP53 examine whether hotspot mutations in TP53 can act as a molecular signal for radon-induced genotoxicity in at-risk populations. Although TP53 is observed to be altered in high exposure populations, there are limited observations available to suggest a consistent mutational landscape [112]. However, the role of TP53 in the molecular response to radon exposure may be relevant to the bystander effect, wherein TP53 may mediate the inhibition of response signals coming from irradiated cells [103]. Additionally, other key lung cancer-related genes may also be mutated by radon exposure, including EGFR and phosphatase and tensin homolog (PTEN), but the exact mechanisms remain to be characterized [113].
As previously discussed, radon may also exhibit its carcinogenic effects epigenetically, as evidenced by the promoter hypermethylation of the tumor suppressor genes CDKN2A and MGMT. In normal human lung cell lines, miRNAs shown to be primarily involved in cell proliferation, differentiation, and adhesion displayed aberrant expression upon radon exposure [114]. Moreover, the miRNA let-7e-an epigenetic regulator of the RAS oncogene-was found to be upregulated upon low radon exposure [115]. In this study, the upregulation of miRNAs targeting tumor suppressor genes was also noted, including PTEN, which may present an alternative mechanism of radon-induced carcinogenesis.
Finally, a number of studies have examined the effect of genetic polymorphisms of DNA damage repair genes in the outcome of individuals exposed to radon. For instance, individuals with a polymorphism leading to the Asp1104His substitution of DNA repair gene ERCC5 (XpG) displayed a higher frequency of micronuclei in their lymphocytes, representative of elevated cytogenetic damage and decreased radiosensitivity [116]. Alternatively, the absence of GSTM1 and GSTT1, members of the glutathione-s-transferase enzyme family-critical to detoxification and excretionis associated with an increased risk of lung cancer development [117,118]. When radon exposure is considered, individuals with null alleles show a doubly increased odds ratio of lung cancer development [118]. Notably, this enzyme is relevant in the biotransformation and excretion of arsenic, suggesting similar carcinogenic pathways between these two environmental agents.
Taken together, the molecular landscape of radon-induced carcinogenesis is complex and diverse, with effects being observed at the genetic, epigenetic and extracellular level. Future studies may examine the underlying molecular events common to radon-induced lung cancer, to aid in diagnosis and perhaps novel treatment strategies.

Common oncogenic features exhibited by environmental carcinogens
The landscape of the genomic disruptions induced by environmental carcinogens is extensive. It has been demonstrated that these compounds can induce alterations such as chromosomal abnormalities, DNA double-strand breaks, gene expression dysregulation, and epigenetic aberrations. While each agent presents a unique mechanism and clinical challenge, a number of parallels can be seen. The molecular effects of exposure to arsenic, asbestos, and radon converge in that each compound can result in DNA damage induced by ROS and inflammation. As these events occur early during tumor development, the identification of the underlying genomic and epigenomic abnormalities caused by these compounds is extremely relevant in identifying early oncogenic events and individual susceptibility differences.
Although the intricacies of the molecular mechanisms of alteration may differ between the various toxic agents, ROS generation is a common outcome of exposure that can lead to extensive DNA damage and further perturbations in various cellular compartments and processes [119]. As mitochondria are one of the primary sources of ROS, they are also key targets of oxidative toxicity [120]. Arsenic exposure is associated with dysfunction of the mitochondria, through the ability of its metabolites to disrupt the mitochondrial membrane potential and reduce mitochondrial ATP levels, as well as ROS-induced mitochondrial damage [121,122]. Mitochondrial damage induced by arsenic can then lead to numerous alterations in key signaling pathways, such as the decreased expression of apoptotic regulator protein Bcl-2 [122]. Regardless of the molecular mechanism, mitochondrial insult culminates in apoptosis and increased inflammation, in addition to the exacerbation of reactive species generation; events that commonly precede tumorigenesis [121,123].
Another frequently observed early consequence of exposure to environmental carcinogens is an inflammatory response. Indeed, inflammation caused by infiltrating immune cells underlies numerous hallmarks of cancer biology by providing key molecules for tumor survival and growth, as well as the promotion of genomic aberrations, again through the generation of ROS [124]. Asbestos-induced carcinogenesis is thought to rely heavily on the inflammatory response, where the macrophages of the innate immune system attempt to clear the carcinogenic fibers through phagocytosis [125]. However, these fibers are inherently difficult to digest, leading to the eventual death of the macrophage and subsequent release of proinflammatory cytokines, ROS, and other growth factors [126]. Interestingly, many malignancies have noticeable local immune responses prior to tumor development, highlighting the complex and dichotomous role of host immune cells in both proand anti-tumor functions [127]. Thus, exposure to environmental carcinogens threatens the genetic and epigenetic landscape of oncogene expression in the development of malignancies, and subsequently changes cellular and systemic processes.
The intertwined role of genetic and epigenetic aberrations resulting from exposure to these compounds highlights the complexity of environmentally-induced lung cancer. However, the carcinogenic mechanisms associated with exposure to these agents have been mainly identified using a "one-agent-at-a-time" approach. Further, we have yet to understand how these factors interplay with one another in cases of combined exposure and how individual genomes modulate the molecular events that arise following exposure. For example, it is difficult to accurately assess the relative risk of lung cancer in an individual who is exposed to occupational asbestos, arsenic-contaminated water, and high levels of domestic indoor air radon. Whether these factors synergize in terms of their molecular effects is not clearly understood and has critical implications to patient monitoring and disease management.
Recently, the idea of the human exposome has sought to provide a method for analyzing individual risk factors by integrating the effects of factors ranging from DNA-level alterations to geographic location. The human exposome is defined as the sum of every exposure to which an individual is subjected to from conception to death [128]. The exposome is dynamic: the nature, amount, and conditions of exposure change over time. It also includes exposure from internal (e.g., metabolism, endogenous hormones, gut microflora, inflammation, oxidative stress, etc.) and external (e.g., radiation, infectious agents, chemical contaminants and environmental pollutants, among others) sources [129]. The lungs are one of the organs at the highest risk of disease development from environmental exposures as the lung exposome can be comprised of an array of molecules and environmental insults. Arsenic, asbestos, and radon, together with air pollution and tobacco smoke, constitute a fraction of the complex mix of environmental carcinogens posing risks for developing thoracic malignancies in humans. However, understanding the oncogenic events following exposure to these agents may allow for the identification of key intervention points to minimize environmentally-induced lung cancer in at-risk populations.

Translational outlook for environmentally-induced cancer
As the molecular mechanisms of environmentally-induced carcinogenesis continue to emerge, a need to characterize the clinical utility of these findings should be underscored. This need is further emphasized by the complex interplay between the numerous features of the lung exposome. Many of the single cancer-associated genes that are affected by exposure to these environmental agents are promising therapeutic intervention points. For instance, targeted inhibitors of EGFR (e.g., erlotinib, afatinib)-a protein transcribed from a gene commonly up-regulated upon exposure to arsenic-are used in lung cancer treatment to interfere with the aberrant growth pathways activated by the upregulation of this signaling receptor [130]. Additionally, the association between radon exposure and ALK gene rearrangements in lung cancer patients may be amenable to therapy with inhibitors of the ALK protein (e.g., crizotinib, ceritinib) [131]. However, patients that do not present with alterations in genes that are clinically actionable remain extremely difficult to treat beyond standard regimes. Thus, it is critical to analyze the oncogenetic alterations induced by environmental carcinogens, to not only identify the contribution of each of these widely-distributed agents to tumorigenesis, but also to direct the development of novel treatment and risk-management strategies. Concurrent analysis of altered genes, transcripts, and proteins may help to parse out the risk associated with the varying molecular aberrations that have been observed to be induced by these compounds [132]. This approach, while difficult in terms of scale, necessitates the use of geographic, demographic, and exposome level data, which can be scarce in areas where environmental carcinogen levels are especially concerning. Table 1 summarizes the currently available sources of information for carcinogens found in the environment that are associated with lung cancer. Overall, future mitigation of the environmental risk factors that lead to lung cancer will rely on the integration of information from the genomic to epidemiological levels.

Conclusions and future directions
The geographical conditions facilitating human exposure to environmental lung carcinogens such as arsenic, asbestos and radon occur commonly across the globe. While millions of individuals are known to be exposed to potentially damaging doses of these carcinogens, another significant part of the population is unaware of its exposure. Despite the worldwide impact of the public health risk posed by these compounds, the genomic and epigenetic consequences of these exposures are drastically understudied. Barriers such as: (i) availability of individual-level exposure data; (ii) collection of genomic, epigenomic, and transcriptomic readouts following acute and chronic exposure to carcinogens; and (iii) obtaining enough samples to reach statistical power; impose even further challenges to determining the true extent of environmentally-induced health effects.
Understanding these mechanisms could have a significant impact on the establishment of safe exposure limits for each of these agents. For instance, most of the current frameworks used to regulate arsenic exposure in drinking water have been derived from studies performed in specific populations exposed to high levels of arsenic, such as Bangladesh, Chile, and China [9,133,134]. However, an increased risk of arsenic-related health effects (including cancer) has been documented at levels below current safety thresholds that are commonly found in water sources throughout North America and Europe [7]. Thus, characterizing the effects of these agents at the genomic/epigenomic level will not only aid in determining the oncogenes that are perturbed in environmentally-induced lung cancers, but may also uncover early molecular events that can be used as diagnostic and prognostic markers.
The fraction of lung cancer patients who have never smoked or have ceased smoking is likely to increase in the coming years. Exposure to environmental carcinogens, such as arsenic, asbestos, and radon will play a key role in their etiology. Further elucidation of the detailed mechanisms driving environmentally-induced lung tumors will provide the much-needed insight to define specific detection methods and intervention strategies. Collectively, uncovering these carcinogen-specific mechanisms, as well as the affected genes driving malignant transformation, will greatly contribute to the development of personalized approaches to provide better support to lung cancer patients.