Dietary Carcinogens and DNA Adducts in Prostate Cancer

Abstract

Prostate cancer (PC) is the most commonly diagnosed non-cutaneous cancer and the second leading cause of cancer-related to death in men. The major risk factors for PC are age, family history, and African American ethnicity. Epidemiological studies have reported large geographical variations in PC incidence and mortality, and thus lifestyle and dietary factors influence PC risk. High fat diet, dairy products, alcohol and red meats, are considered as risk factors for PC. This book chapter provides a comprehensive, literature-based review on dietary factors and their molecular mechanisms of prostate carcinogenesis. A large portion of our knowledge is based on epidemiological studies where dietary factors such as cancer promoting agents, including high-fat, dairy products, alcohol, and cancer-initiating genotoxicants formed in cooked meats have been evaluated for PC risk. However, the precise mechanisms in the etiology of PC development remain uncertain. Additional animal and human cell-based studies are required to further our understandings of risk factors involved in PC etiology. Specific biomarkers of chemical exposures and DNA damage in the prostate can provide evidence of cancer-causing agents in the prostate. Collectively, these studies can improve public health research, nutritional education and chemoprevention strategies

Introduction

The World Health Organization (WHO) has reported that prostate cancer (PC) is the second most common cancer in men worldwide, with an estimated 1.1 million incident cases and 0.3 million deaths occurring in 2012 [1]. PC is more commonly diagnosed in economically developed countries, which may be attributed to more extensive PC screening programs. The major risk factors identified for PC include increased incidence with age, family history, and ethnicity, with African-American men having a two-time higher risk compared to Caucasians [2]. The occurrence of PC varies widely worldwide. Many studies of migrant populations show a significant increase in the incidence of PC and mortality rates in migrants from regions of the world with a low prevalence of PC, following their relocation to countries with high risk for PC, suggesting that environmental or dietary factors influence the risk factors for PC development [3]. The frequent consumption of high-fat diets from dairy products and red meats and alcohol are implicated as risk factors for PC [4]. However, the precise role of dietary factors and specific chemicals in the diet and mechanisms involved in the development of this malignancy remain unclear.

The diet as a risk factor for human PC

High-fat diet

Dietary fat and several fatty acids are postulated to play a role in PC etiology and tumor progression, although the findings of epidemiologic studies are inconsistent. Some studies found a strong positive association between fat consumption, PC incidence and mortality [5-10], whereas other investigations have not detected a correlation [11-14].

Several studies conducted in vivo in animal models and in vitro have shown a role for a high-fat diet in the development and progression of PC. Tissue culture medium conditioned with adipose tissue obtained from mice fed with high-fat Western-style foods enhanced cell proliferation, migration, and invasion of human prostate cancer cells in vitro [15, 16]. In the transgenic adenocarcinoma mouse prostate (TRAMP) and xenograft models, circulating adipokine and cytokine alterations and other factors induced by a high-fat diet contributed to PC progression [15-18]. Although strong evidence supports the effects of a high-fat diet on PC development and progression, the exact mechanism(s) by which a high-fat diet underlines PC etiology remain uncertain. Several hypotheses proposed include intake of fatty acids, resulting in inflammation, induction of oxidative stress, and cell signaling alteration.

Fatty acids

Fatty acids, such as n-3, and n-6 polyunsaturated fatty acids, and their metabolites are involved in numerous pathways that can affect PC development and progression. For example, n-6 fatty acids linoleic acid and arachidonic acid enhance proliferation of human prostate cell lines [19, 20]. Moreover, n-6 fatty acids are precursors of eicosanoids, which are converted to prostaglandins (PGs). The n-6 fatty acid arachidonic acid is metabolized by the enzyme cyclooxygenase (Cox-1 and Cox-2) to form prostaglandin E2 (PGE2) [21]. PGE2 is a short-lived hormone-like molecule involved in cell proliferation, cell differentiation and inflammation [21]. Notably, the growth stimulation of PC cells, by treatment with arachidonic acid, is correlated to the induction of COX2 expression and an increased synthesis of PGE2 [19, 22]. At the molecular level, PGE2 binds to its EP4 and EP2 receptors, resulting in the subsequent activation of the protein kinase A (PKA) pathway, which leads to expression of early growth-related response genes including c-fos [23]. Arachidonic acid also activates the phosphatidylinositol-4,5-bisphosphate 3-kinase signaling pathway (PI3K/Akt) [20]. The PI3K/Akt cascade is involved in the progression and aggressiveness of PC. In fact, after long-term androgen deprivation therapy, there is constitutive activation of the PI3K/Akt pathway, a mechanism that leads to increased resistance of tumor cells to apoptosis [24, 25]. The activation of the PI3K/Akt pathway by arachidonic acid in human prostate cells also results in the activation of the Nuclear Factor Kappa Beta (NF-κB) cascade [20]. The induction of the NF-κB pathway increases cell resistance to chemotherapy and radiation therapy. Moreover, the activation of NF-κB also stimulates tumor cell growth, the inhibition of apoptosis, and enhances tumor invasion, metastasis, and angiogenesis [26] (Figure 1).

Figure 1:

Effect of Arachidonic acid (AA) and its metabolite prostaglandin E2 (PGE2) on cell signaling in human prostate cells

In contrast to the n-6 polyunsaturated fatty acids, n-3 long-chain fatty acids protect against PC development. For example, a significant decrease in the growth of PC xenografts occurs in nude mice fed with a diet containing high levels of eicosapentaenoic and docosahexaenoic acids [27]. These effects have been supported by in vitro studies where both eicosapentaenoic acid and docosahexaenoic acid inhibit the proliferation of human PC cell lines [19, 28]. Moreover, eicosapentaenoic acid and docosahexaenoic acid prevent the progression of human prostate cells toward an aggressive androgen-independent phenotype. At the molecular level, eicosapentaenoic and docosahexaenoic acid treatments inhibit the PI3K/Akt signaling pathway and decrease expression of the androgen receptor (AR), a master regulator of prostate cell proliferation and PC development [29].

Inflammation

Inflammation often occurs in the prostates of aging men, and plays a critical role in the development of benign prostatic hyperplasia and PC incidence [30-32]. Androgen levels, genetic predisposition, obesity, and a high-fat diet are associated with prostatic hyperplasia and PC [33].

The association between a high-fat diet, the induction of inflammation, and PC markers have been reported in several in vivo studies. Prostatic inflammation correlates with cell proliferation and an increase in prostate gland size of mice consuming a high-fat diet [34, 35]. Consumption of a high-fat diet also elevates ataxin levels in the adipose tissue, leading to a significant increase in the production of lysophosphatidic acid, which can act directly on the prostate and induce hyperplasia and cell proliferation [36].

While inflammation is associated with an enhancement of PC development, there is not a clear understanding of the mechanisms involved in this effect. Several studies have reported the involvement of immune cells and the production of pro-inflammatory cytokines. In clinical studies, patients with benign prostatic hyperplasia contain infiltrates of macrophages, T-lymphocytes, and B-lymphocytes that are chronically activated [37]. These infiltrating cells produce cytokines including IL-2, IFN-γ IL-6, IL-8, IL-17, and TGF-β that maintain a chronic immune response and induce persistent intra-prostatic inflammation and fibromuscular growth by an autocrine or paracrine effect [38, 39]. These pro-inflammatory cytokines can modulate prostate growth in mice fed a high-fat diet [40] and correlate with the production of pro-inflammatory cytokines such as IL-1α, IL-1β, IL-6, IL-17 and TNF-α [41, 42]. These cytokines can induce prostate growth through induction of secondary mediators such as Cox-2. For example, IL-17 serves to stabilize and increase the enzymatic activity of Cox-2 [43]. Of note, the induction of Cox-2 expression in the prostate epithelium is associated with increased cell proliferation and apoptotic resistance [44]. Furthermore, the treatment of human prostate cells in vitro with serum obtained from obese mice containing elevated levels of pro-inflammatory cytokines promotes cell proliferation, invasion, migration and, epithelial-mesenchymal transition [45].

Oxidative stress

A disproportionate generation of reactive oxygen species (ROS) causes tissue injury, DNA damage, and post-translational DNA modifications, which can lead to neoplasia in the human prostate [46, 47]. ROS are generated from the mitochondrial respiratory chain, an uncontrolled arachidonic acid cascade, and NADPH oxidase [33]. The expression of NADPH oxidase subunits such as gp91phox, p47phox, and p22phox is increased in the prostates of mice fed a high-fat diet, [34]. Also of note, human PC cells harbor an increased level of ROS compared to normal prostate cells [48]. The ROS activity correlates with dysregulation of the NADPH oxidase system, which is a critical event for the malignant phenotype of human PC cells [47, 49-52].

At the molecular level, continual oxidative stress leads to the activation of two critical signaling pathways: the signal transducer and activator of transcription (STAT-3) and NF-κB pathways [33]. The activation of both cascades leads to the expression of transcription factors required for regulating genes involved in proliferation, survival, angiogenesis, invasion, and inflammation [53] (Figure 2).

Figure 2:

Effect of pro-inflammatory cytokines such as IL-6 and IL-1 on the activation and the cross talk of STAT-3 and NF-κB pathways

A high-fat diet also leads to increased activation of NF-κB in many organs of mice, including prostate [54]. In humans, there is constitutive activation of NF-κB in prostate adenocarcinoma [55]. Moreover, this constitutive activation is associated with upregulation of pro-survival molecules including Bcl-2, Bcl-XL, and Mcl-1 [56]. Similarly, increased STAT-3 activation and its DNA binding occur in the prostate of mice fed a high-fat diet [57]. In human PC cells, the inhibition of STAT-3 results in the inhibition of cell proliferation and a significant decrease in cell viability [58].

Dairy products

Several epidemiological studies have reported that frequent intake of high-fat dairy products is associated with an increased risk of developing PC [59-61]; however, other studies failed to observe this association [62, 63]. The role of high-dairy fat intake in PC risk is supported by studies conducted in vitro where milk modulated and promoted the proliferation of the human prostate LNCaP and PC-3 cancer cell lines [64, 65]. Saturated fat intake, high-calcium intake, decreased circulating levels of 1,25-dihydroxy-vitamin D (the active form of vitamin D), and increasing levels of insulin-like growth factor-1 (IGF-1) are several potential mechanisms by which milk and dairy product intake may impact the incidence and the progression of PC.

Saturated fat intake

Saturated fat is another likely factor in dairy products that may influence the development and the aggressiveness of PC. A higher intake of low-fat milk is associated with a greater risk of non-aggressive PC, whereas whole-fat milk intake is frequently associated with a higher incidence of aggressive PC phenotypes [63, 66-69].

High-calcium intake

Intake of calcium above the recommended daily doses (~1000 mg/day) is associated with increased risk of developing PC but also with aggressive and highly malignant PC [70-75]. The underlying mechanisms of high calcium intake and the risk of PC are not yet elucidated. Over-activation of the calcium-sensing receptor and calcium-dependent voltage-gated channel expressed in human prostate cells by the high levels of ionized calcium circulating in the bloodstream are two potential mechanisms involved in PC etiology [76-79]. The stimulation of these receptors by extracellular calcium increases PC cell proliferation, apoptosis resistance, and metastatic potential in vitro and in vivo [80-83].

Vitamin D

Several epidemiological studies have reported an association between low levels of vitamin D and higher risk for PC [84-86]. The modulation of vitamin D metabolism and the decrease of 1,25-dihydroxyvitamin D levels are associated with an increased risk of PC [87]. Once ingested, vitamin D is metabolized to its biologically active form 1,25-dihydroxyvitamin D through a two-step oxidation. The first oxidation reaction catalyzed by CYP2R1 occurs in the liver leading to the formation of 25-hydroxyvitamin D, and the second oxidation catalyzed by CYP27B1 occurs in the kidney, producing the 1,25-dihydroxyvitamin D [88]. 1,25-Dihydroxy vitamin D has anti-proliferative effects that are driven through the nuclear Vitamin D receptor (VDR) pathway, leading to the expression of genes involved in cell cycle arrest, cell apoptosis and differentiation [89]. VDR is expressed in both normal and cancer prostate cells [90-92]. In human prostate cells, 1,25-dihydroxyvitamin D produces anti-proliferative effects [92-96], reduces oxidative stress [97], and up-regulates pro-apoptotic genes [98]. More than 2000 genes are modulated by 1,25-dihydroxy vitamin D, including genes encoding for androgen metabolism [99]. More detailed studies are required to elucidate the critical roles of vitamin D in PC development.

IGF-1

The IGF system includes three ligands (insulin, IGF-1, IGF-2), their receptors (insulin receptor (INSR), IGF-1 receptor (IGF-1R), the mannose 6-phosphate receptor (M6P/IGF-2R), and six circulating IGF-binding proteins (IGFBP1–6) [100]. In the human prostate, every element of this system is expressed in normal, hyperplastic, and neoplastic prostate tissues, as well as in primary and prostate cell lines [101-110]. The IGF system plays a critical role in normal gland growth and development of the prostate [104, 108, 111, 112]. A higher serum IGF-1 concentration is correlated with an increased risk of PC [113-119]. The biological functions of IGF-I are mediated primarily through the IGF-IR, a tyrosine kinase transmembrane receptor that binds IGF-I with higher affinity than IGF-II [120]. Interestingly, inhibition of IGF-1R is associated with decreased androgen-dependent and androgen-independent growth in vitro as well as a suppression of in vivo tumor growth and PC cell invasiveness [108, 121-123]. Conversely, activation of IGF-1R by its ligand IGF-1 leads to the activation of several signaling pathways including mitogen-activated protein kinase (MAPK) and PI3K/Akt [124]. The activation of these signaling pathways induces proliferation and migration, and inhibits apoptosis in the PC cell [125-128] (Figure 3).

Figure 3:

Insulin/IGF signaling in prostate carcinogenesis.

The IGFBPs provide an additional, extracellular mechanism to regulate IGF activity. The IGFBPs bind to IGF-1 and IGF-2 with high affinity and thereby diminish their binding to IGF-R, resulting in the inhibition of the IGF signaling pathway [100]. Altered IGFBP plasma levels are found in PC patients, and a decrease in IGFBP-3 is associated with higher risk and progression of PC [113, 114, 129, 130]. IGFBP-3 is a substrate for the serum protease PSA [104]. Therefore, high levels of PSA in PC patients may result in a decrease in circulating levels of IGFBP-3 by proteolytic cleavage, leading to an increase in IGFs including IGF-1, thus facilitating disease progression (Figure 3).

Alcohol

Alcohol consumption accounts for about 5% of all cancer deaths worldwide [131]. In the USA, 92% of adult males self-report a long-term use of alcohol [132], and up to 3.7% of total cancer deaths are linked to alcohol [133]. Elevated alcohol consumption can contribute to a number of malignancies, including cancer of the oral cavity, pharynx, larynx, esophagus, and liver of both sexes and colorectal cancer in women. However, findings from epidemiological studies on the role of alcohol consumption in PC risk are inconsistent. Several studies found that alcohol consumption is a risk factor for PC [134-136], whereas other studies reported a decreased risk of PC [137]. The compilation of meta-analyses are also inconsistent: several reports found no association between alcohol consumption and PC risk [138-140], whereas others reported a significantly increased risk of PC with alcohol [141-144]. One meta-analysis reported an increased risk for PC for men drinking more than 50 g of alcohol per day, with the risk becoming slightly higher for men who consume more than 100 g per day [141]. Three other meta-analyses also reported a significantly increased risk in PC for light and moderate drinking (one to four drinks per day) [142, 143] or the equivalent of up 24 g of alcohol per day [144]. An association between alcohol intake and the degree of aggressiveness of PC was reported in some studies [145-148], but not other studies [135, 149, 150].

Ethanol is the primary form of alcohol in alcoholic beverages. Ethanol is classified as a human carcinogen [151]. The genotoxic effects and carcinogenicity of ethanol are thought to be driven by its major metabolite, acetaldehyde. Acetaldehyde forms DNA adducts in human cells in vitro and in vitro [152]. In humans, levels of acetaldehyde DNA adducts present in lymphocytes are seven times higher in alcohol users compared non-users [153]. In vivo studies demonstrated that ethanol is efficiently bioactivated into acetaldehyde in rat prostate by different enzymatic pathways involving xanthine oxidoreductase and cytochrome P450 2E1 [154, 155]. Moreover, acetaldehyde formation is linked to an increase in prostate epithelial cell death and ultrastructural alterations in epithelial cells including chromatin condensation around the perinuclear membrane and endoplasmic reticulum dilatation, an ultra-structural marker of endoplasmic reticulum stress [156]. The rat prostate lacks alcohol dehydrogenase and aldehyde dehydrogenase activities, resulting in an accumulation of acetaldehyde and thus, an increase in genomic damage in the prostate of rats exposed to ethanol [156]. Chronic ethanol exposure also leads to oxidative stress and a diminution in the antioxidant defense system in the rat ventral prostate [156, 157].

An association of PC risk with cancer aggressiveness was observed with high intake of beer [145, 146, 148]; while modest protective effects were observed for red wine consumption in some but not all studies [158-160]. The protective effect of wine, especially red wine, is likely attributed to its high contents of polyphenols such as flavonoids and resveratrol [161]. Polyphenolic compounds harbor antioxidant and anti-androgenic activities and therefore are thought to act as anti-carcinogens [162]. In vitro, nanomolar concentrations polyphenols inhibit cell growth in a dose and time-dependent manner in both androgen-dependent LNCaP and androgen-independent DU145 and PC3 PC cell lines. Treatment of LNCaP and PC3 cells with flavonoids, including catechin, epicatechin, and quercetin, inhibites cell proliferation, whereas resveratrol is the most potent inhibitor of DU145 cell growth. The proposed mechanism for the antiproliferative effect of polyphenols is through the modulation of NO production [163].

Red and processed meat

Many epidemiological studies have focused on the role of red and processed meats in PC risk. Some meta-analyses report an elevated risk for PC with frequent consumption of meats, whereas other studies failed to find an overall effect on risk [164, 165]. It is hypothesized that DNA damaging agents, including heme iron, N-nitroso compounds (NOCs) formed in processed meats [166], polycyclic aromatic hydrocarbons (PAHs) formed in smoked meats and meats cooked under flame [167], and heterocyclic aromatic amines (HAAs) formed in well-done grilled meats [168, 169], contribute to PC risk. Of note, the risk of PC for African American men is ~2-fold greater than for Caucasians [2]. One paradigm proposed for the increased risk of PC in African American men is based on their preference for frequent consumption of well-done cooked meats containing the HAA 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP). PhIP is a rodent prostate carcinogen and potential human prostate carcinogen [170, 171], and may explain the higher risk of PC for African-American men compared to white men [172].

Heme iron

Once ingested, heme-containing proteins such as myoglobin or hemoglobin are hydrolyzed to peptides, amino acids, and heme iron. Feeding ferriheme to rodents induces cytotoxicity, enhances cell proliferation of colonic mucosa, promotes oxidative stress, and thus, may contribute to colorectal cancer [173]. Heme iron is transported through the bloodstream to all organs of the body and can catalyze oxidative reactions, causing DNA, protein and lipid oxidation in multiple organs including the prostate [174]. The overall level of free radical damage induced by heme-catalyzed oxidation is estimated to be comparable to that resulting from ionizing radiation [174]. Two epidemiological studies have examined the role of heme iron in PC development [175, 176]. One reported a positive association between total heme iron intake and advanced PC risk, whereas another reported no associations of the dietary factors with PC risk irrespective of stage or grade. Thus, additional studies are necessary to evaluate any potential associations between heme iron consumption and PC.

N-Nitroso compounds (NOCs)

Carcinogenic NOCs include two chemical classes, N-nitrosamines and N-nitrosamides, formed by the reaction of nitrosating agents derived from nitrite with amines and amides respectively [177]. Nitrites added to processed meat serve as anti-bacterial agents as well as curing agents, and they produce the characteristic red-pink color of cured meats. However, nitrites also react with amines in processed meats to produce dietary sources of NOCs. Moreover, the consumption of processed meat is a significant dietary source of nitrite, secondary amines, and amides, which can undergo nitrosation to form NOCs within the gastrointestinal tract [177-179]. The ingestion of heme contained in red meat can stimulate the endogenous formation of NOCs in the digestive tract [180-182]. More than 300 NOCs have been detected in 39 different animal species, including six species of nonhuman primates. Of these, 85% of N-nitrosamines and 92% of N-nitrosamides were reported to induce cancer in multiple organs including liver, lung, esophagus, bladder, and pancreas [183]. NOCs or their metabolites alkylate DNA. While N-nitrosamides react spontaneously with DNA, N-nitrosamines require metabolism by cytochromes, such as by cytochrome P450 2E1, which is expressed in the gastrointestinal tract [178]. Among the different types of DNA adducts formed with NOCs, the alkylation of the O⁶-position of guanine is a primary lesion that induces G to A transitions [184-186]. The majority of epidemiological studies have focused on the role of NOCs in gastric, esophageal, and colorectal cancers [187-192]. Two epidemiology studies studied the etiology of NOCs and PC risk: there was no significant association between dietary NOCs and risk of development of PC in either study [176, 192]. Thus, further studies on the role of processed meats and NOCs in PC risk are warranted.

Polycyclic aromatic hydrocarbons (PAHs)

PAHs constitute a broad class of compounds that have two or more fused aromatic rings. PAHs arise by the incomplete combustion or high-temperature pyrolysis of organic materials [193]. PAHs are ubiquitous environmental pollutants that occur as complex mixtures but never as individual components [194]. Several PAHs are classified as human carcinogens by the World Health Organization [151, 193]. Apart from occupational exposure, such as the case of coke-oven workers [195-197], the general population is exposed primarily to PAHs from dietary sources [167] and cigarette smoke [198]. The preparation of meats, mainly by a direct open flame, results in pyrolysis of the fat drippings, leading to the formation of PAHs, which are deposited through the smoke particulates on the surface of the grilled meats [199, 200]. Various PAHs occur in some charcoal-broiled, grilled, and smoked meats [194, 201-203]. The estimates of the daily dietary intake for the general population are imprecise and range widely: levels of total daily PAH intake range from 3.7 μg up to 17 μg [167], The most well-studied PAH is benzo[a]pyrene (B[a]P). B[a]P occurs in some grilled meats at levels up to ~ 4 ng/g [204, 205].

The carcinogenic properties of PAHs are attributed to their ability to form mutation-prone DNA adducts [193]. PAHs undergo metabolism by cytochrome P450 enzymes to form reactive dihydrodiol epoxides, which react with DNA to form covalent adducts, leading to mutations [206]. B[a]P-DNA adducts are formed in human prostate cells in vitro after exposure to B[a]P [207-209], B[a]P treatment also leads to an increase in DNA double-strand breaks when measured by the comet assay [208, 210]. PAH adducts, including B[a]P adducts, are frequently detected in human prostate tissues by immunohistochemistry (IHC)-[211-215] with an antibody, which was raised against B[a]P-modified DNA, but which also cross-reacted with DNA adducts of at least five other PAHs [216]. Levels of PAH-DNA adducts is higher in adjacent human non-tumor prostate tissue compared with prostate tumor tissue, possibly due a higher cell proliferation rate in the tumor [211, 213, 217]. The occurrence of putative PAH-DNA adducts is associated with a higher risk for PC and cancer recurrence after prostatectomy within one to two years after surgery [211]. This risk was prominent in patients younger than 60 years old, patients with advanced-stage disease, and African Americans patients [211]. However, these data should be interpreted with caution since IHC is not a specific method of DNA adduct detection, even for assays performed with monoclonal antibodies, where possible cross-reactivity of the antibodies with other DNA adducts or endogenous cellular components can lead to false positivity. The occurrence of DNA adducts of B[a]P was not confirmed in one cohort of PC patients when analyzed by liquid chromatography/mass spectrometry (LC/MS), a more specific analytical method than IHC [218]. Thus, there is a critical need to characterize DNA adducts on the same specimens by IHC and LC/MS to determine the validity of the analyses.

Heterocyclic aromatic amines (HAAs) and PC

HAA formation and sources of exposure

HAAs are a class of more than 25 genotoxic chemicals known to form in cooked meats, fish, poultry, and tobacco smoke [168, 169, 219]. HAAs are sub-classified into the aminoimidazoarenes (AIAs) and the high-temperature pyrolytic HAAs (Figure 4). AIAs contain the N-methyl-2-aminoimidazole moiety derived from creatinine in muscle tissue. The AIAs form in meats, fish, and poultry cooked at temperatures above 150 °C and arise through the reaction of pyridine or pyrazines, derived from Strecker reactions, and condensation with creatine [220, 221]. Pyrolytic HAAs form by high-temperature pyrolysis ( >250 °C) of protein or amino acids, such as glutamic and tryptophan. Pyrolytic HAAs occur when proteinaceous foods are heated at temperatures generally above 250 °C [168, 222, 223]. Several HAAs are also formed in tobacco smoke [223, 224]. 2-Amino-9H-pyrido[2,3-b]indole (AαC), a pyrolysis product of tryptophan, is the major carcinogenic HAA formed in combusted tobacco and occurs in mainstream tobacco smoke at levels up to 258 ng per cigarette [225-227]. Unexpectedly, PhIP, an AIA containing the N-methyl-2-aminoimidazole moiety of creatine, was detected in tobacco smoke [224]; however, the mechanism of PhIP formation in tobacco smoke has not been determined. The principal sources of exposure to most HAAs occur through the consumption of well-done cooked meats and poultry [168, 228]. HAA formation in cooked meats generally occurs at the low parts-per-billion (ppb) range, but the levels of some HAAs can approach several hundred ppb in well-done cooked meats or poultry [168, 219, 229, 230]. PhIP and 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) are often the most prevalent HAAs formed in cooked meats and poultry [219, 221, 230-233]. The average dietary HAAs intake ranges from less than 2 to up to 25 ng/kg per day [172, 234].

Figure 4:

Chemical structures of prevalent HAAs in cooked meat.

Bioactivation and Formation of DNA Adducts

HAAs undergo extensive metabolism by hepatic cytochrome P450 1A2 (CYP 1A2)-catalyzed N-oxidation of the exocyclic amine groups, leading to the formation of N-hydroxy-HAAs, as reactive intermediates [228, 235, 236]. CYP 1A1 and CYP 1B1 catalyze this reaction in extrahepatic tissues [228, 235]. The N-hydroxy-HAAs are further bioactivated by acetylation or sulfation catalyzed by N-acetyltransferases (NATs) or sulfotransferases (SULTs), respectively, either in the liver or extrahepatic tissues [228]. These unstable esters react with DNA to form DNA adducts [228, 237]. HAA-DNA adducts are mainly formed at C-8 on deoxyguanosine (dG) through the exocyclic amine of the HAAs, to produce dG-C8-HAA adducts as the major DNA adducts [238, 239].

Carcinogenesis of HAAs

HAAs are multisite carcinogens in rodents and induce cancers of the oral cavity, liver, stomach, colon, pancreas, and the mammary gland in females [168, 170]. Notably, PhIP is the only HAA reported to induce PC in rodents [168, 170] Carcinogenesis studies in rodents used chronic doses of HAAs ranging between 0.1 to 64 mg/kg/day to induce tumors [168, 228]. These doses are more a million-fold higher than the daily intake of HAAs. Thus, one might surmise that the levels of human exposure are too low to contribute to human cancers. However, a linear relationship between HAA dose and HAA-DNA adduct formation occurs in rodent tissues for PhIP, MeIQx, and IQ [240-242], signifying mutation-prone DNA adducts of HAAs can still form in tissues at dosing regimens approaching human exposure levels.

Animal toxicity studies may underestimate the carcinogenic potential of HAAs in humans. For example, the levels of HAA-DNA adducts formed in primary human hepatocytes are significantly higher than those formed in primary rat hepatocytes, under the same doses and times of exposure [243]. Human CYP1A2, which is principally expressed in the liver, is the major CYP involved in the metabolism of many HAAs. Human CYP1A2 is catalytically more efficient than the rat homolog in the bioactivation of PhIP and MeIQx, and perhaps other HAAs [244]. Human CYP1A2 and human liver microsomes preferentially bioactivate HAAs through N-oxidation of the exocyclic amine group. In contrast, rat CYP1A2 and rat liver microsomes preferentially catalyze the detoxication of HAAs by oxidation of the heterocyclic rings [245]. The superior activity (lower K_m and higher k_cat) and ability of human CYP1A2 to catalyze N-oxidation can explain the higher levels of HAA adducts formed in human compared to rat hepatocytes. Several HAA-DNA adducts have been detected in human tissues by various techniques, indicating that even at low levels of exposure, HAAs can form DNA adducts in humans [246-256].

PhIP DNA Damage, Mutation, and Carcinogenicity in Prostate Rodent Studies

PhIP is the only HAA studied thus far that targets the prostate as a principal site for DNA adduct formation and carcinogenesis in rodents [168]. PhIP undergoes metabolism to form high levels of DNA adducts in the prostate of Wistar and Fischer 344 rats [170, 257-260] and induces high levels of mutations in the prostate of the Big Blue lacI transgenic rat [260, 261]. PhIP is a prostate carcinogen in the Fischer 344 rat [170]. PhIP also induces prostate tumors in CYP1A-humanized (hCYP1A) mice but not in wild-type mice [262]. This finding reinforces the concept that human CYP1A is superior to the rodent orthologue in the bioactivation of PhIP [244, 263].

Extensive inflammation occurs in the dorsolateral prostate lobe marked by CD45⁺ mononuclear leukocyte and CD8⁺ T lymphocyte infiltration in PhIP-induced tumors in the prostates of hCYP1A mice [262]. This inflammation is associated with atrophic glands, high-grade prostatic intraepithelial neoplasia, and oxidative stress [262, 264]. In contrast, the prostatic intraepithelial neoplasia lesions are significantly less severe and infrequently associated with inflammation and oxidative stress in the ventral prostate glands [262]. These observations are noteworthy because the dorsolateral prostate is homologous to the human peripheral prostate zone, the most common site of PC development in humans [265]. Similarly, PhIP treatment leads to inflammation with a marked increase in mastocyte and macrophage infiltration and glandular atrophy of the prostate of Fischer 344 rats [261, 266]. These pathologies induced by PhIP in rodent models of PC are significant because inflammation, oxidative stress, and glandular atrophy are common features in the pathology of human PC [267].

Several key features of cancer biology often reported in human PC occur in PhIP-induced prostate tumors in rodents. For example, treatment of hCYP1A mice with PhIP results in a time-dependent increase in expression of AR protein in prostate tumor epithelial cells [262]. An up-regulation of AR leading to a higher rate of cell proliferation occurs in human PC [268]. Furthermore, PhIP-induced tumors in h-CYP1A mice display significant decreases in levels of E-Cadherin and p63 expression [262]. E-Cadherin is an epithelial cell adhesion molecule involved in the maintenance of normal cell architecture, while the p63 transcription factor has multiple functions in cancer cell biology. PhIP-treatment in Fisher 344 and Big Blue rats also results in significant increases in the levels of Ki-67, a well-established marker of cell proliferation, in the intraepithelial neoplasia regions of the prostate [261, 266]. Dysregulated expression and distribution of these proteins are hallmarks of epithelial malignancies and serve as major diagnostic criteria for human PC [269-271].

PhIP-induced tumors in the h-CYP1A mice also exhibit increased levels of oxidative stress markers, including 8-oxo-dG and nitrotyrosine, markers of oxidative DNA damage and reactive nitrogen species. PhIP treatment results in the up-regulation of COX-2 expression, a cyclooxygenase that catalyzes the formation of pro-inflammatory prostaglandins and a loss of Nuclear factor (erythroid-derived 2)-like 2 (Nrf2) a key transcription factor responsible of the expression of many cytoprotective proteins [264]. Oxidative stress is a key contributor to the development of PC in humans [46, 272]. The PTEN/PI3K/Akt signaling pathway is another critical feature of cell proliferation, cell cycle progression, and survival. Activation of AKT in response to oxidative stress and the loss of PTEN are critical events in human PC progression [273-275]. PhIP-induced prostate tumors in h-CYP1A mice display a significant decrease in PTEN expression and an elevation of phospho-AKT, leading to cell proliferation [264].

These mechanistic data in rodent models reinforce the biological plausibility that PhIP plays an important role in dietary-linked human PC. However, the biological events observed in rodents occurred at very high doses of PhIP treatment - up to 200 mg/kg. Humans are exposed to one million-fold or lower daily amounts of PhIP [172], and the capacity of such lower concentrations of PhIP to induce similar biological effects has not been investigated. Therefore, the interpretation of the carcinogenic effects of PhIP in PC of rodents and their extrapolation to human PC should be done with caution.

b. Human studies

PhIP is the only HAA reported thus far to form DNA adducts in the prostate of human PC patients [218, 255, 276-278] (Figure 5). This biomarker data provides support for some of the epidemiological studies that have linked the frequent consumption of well-done cooked red meat containing PhIP with increased risk of PC. [279-281] However, other investigations have failed to find an association between cooked red meat and increased PC risk [175, 282, 283]. The concentrations of PhIP and other HAAs can vary by more than 100-fold in cooked meats [169, 219]. There is a critical need to conduct such epidemiological studies with more precise exposure measurements of HAAs.

Figure 5:

Representative chromatograms at the MS² scan stage of human prostate samples that were negative and positive for dG-C8-PhIP.

The frequency of detection and the levels of PhIP-DNA adducts in human prostate range widely between studies, depending on the analytical method of adduct measurement. For example, using a high-resolution LC/MS method, PhIP-DNA adducts were detected in 13 out of 54 PC patients with levels ranging from 2 to 120 adducts per 10⁹ DNA bases [218, 278]. However, PhIP-DNA adducts were detected in a very high percentage of prostate tissues in another cohort, occurring at levels exceeding several adducts per 10⁷ DNA bases, when measured by IHC [255, 276, 277]. These discrepancies in adduct measurements may imply a high level of false positivity obtained by IHC, possibly due to cross-reactivity of the polyclonal antibodies raised against PhIP-modified DNA with other DNA adducts or endogenous cellular components [284].

Cytotoxicity and DNA adduct formation induced by PhIP and other HAAs has been studied in primary and PC cell lines. The parent HAAs, PhIP, MeIQx, IQ and AαC were not toxic at doses up to 10 μM and formed low levels of DNA adducts in LNCaP cells [285]. However, HONH-PhIP, the genotoxic metabolite of PhIP, induced a dose-dependent increase in cytotoxicity, whereas HONH-MeIQx, HONH-IQ, and HONH-AαC were not toxic [285, 286]. Moreover, HONH-PhIP forms DNA adducts at levels that are 20-fold higher than other HONH-HAAs in LNCaP cells [285]. These data suggest that the initial bioactivation step of PhIP to form HONH-PhIP occurs in the liver through CYP 1A2-catalyzed N-oxidation, followed by systemic circulation to reach the prostate, where bioactivation is mediated by Phase II enzymes (Figure 6) [285]. Similar data were reported in primary human prostate epithelial cells, where HONH-PhIP formed 50- to 100-fold higher levels of DNA adducts than IQ, MeIQx, and HONH-MeIQx [207, 287].

Figure 6:

Metabolic activation of PhIP and dG-C8-PhIP formation in human prostate.

HONH-PhIP also induces unscheduled DNA synthesis, and DNA single-strand breaks in primary human prostate epithelial cells at 100-time higher levels than HONH-MeIQx [208, 210]. The higher susceptibility of human prostate cells to the DNA-damaging and genotoxic effect of PhIP compared to other prominent HAAs formed in cooked meats recapitulates the DNA adduct biomarker data in prostate tissues of PC patients and provides support for the possible causal role of PhIP in human PC.

PhIP can also act through non-genotoxic mechanisms via androgenic effects to contribute to the development of PC. PhIP binds to the AR and modulates cell proliferation. Using, in silico analysis, binding of PhIP and HONH-PhIP to the AR was found to be comparable to that of the endogenous AR ligand, dihydrotestosterone (DHT), when based on the predicted free energy of binding [288]. Through computational docking studies, both PhIP and HONH-PhIP displayed similar binding modes to DHT and docked with high affinity in the same cavity of the AR ligand binding domain as DHT [288]. Moreover, treatment of the human prostate epithelial cell line LNCaP with PhIP or HONH-PhIP up-regulated AR and PSA expression [288].

PhIP also induces proliferation of human prostate epithelial cells in an AR-independent manner, through the activation of pro-proliferation cell signaling pathways. Low concentrations of PhIP (10⁻¹²–10⁻⁸ mol/L) increase the proliferation, migration and invasion properties of PC-3, an AR-negative human prostate cell line [289]. Proliferation and migration are mediated through the activation of the ERK signal transduction cascade and a rapid, transient increase in phosphorylation of both MEK1/2 and ERK1/2. Interestingly, mitogenic stimulation with epithelial growth factor (EGF), induces the same pattern of activation [290]. Proliferation, migration, and invasiveness are crucial events in the oncogenic progression of cells [291]. Thus, all these biological phenomena induced by PhIP suggest a carcinogenic potential of PhIP in human prostate. The challenge in risk assessment is to determine if the amounts of PhIP in the diet are sufficient to be a significant risk for PC.

Conclusion

There is growing mechanistic and epidemiological data supporting a role for the diet in the development of PC. Multiple mechanisms and hypotheses have been brought forward for PC risk, ranging from different classes of dietary genotoxicants acting as initiators of PC to different dietary factors involved in tumor promotion. However, the precise roles of specific genotoxicants and nutritional factors in PC remain to be clarified. Prospective epidemiological studies on PC risk with improved assessments of dietary habits, including protective nutrient biomarkers in plasma and urine are needed [292]. The identification of micronutrients that protect against PC [293, 294], and the detection of biomarkers of DNA damage in the prostate, such as DNA adducts, by specific mass spectrometric methods and their linkage to mutations [218, 285], can advance our understanding of the micronutrients and genotoxicants in the diet that impact PC risk.