Introduction

Epigenetic modifications of DNA or associated proteins cause changes in gene expression without altering DNA sequence. This includes various pathways and mechanisms that maintain alternate states of chromatin structure by affecting protein composition and transcriptional activity through DNA methylation, histone modifications, nucleosome positioning, and non-coding RNAs [1]. Collectively, these processes are critical players that regulate accessibility of the transcription machinery to gene promoters [2]. Epigenetic changes that modify the contact between transcription machinery and gene promoters can lead to aberrant activation or inhibition of numerous signaling cascades resulting in different diseases including cancer [3]. Therefore, understanding epigenetic alterations involved in cancer development, progression, therapeutic response, and resistance hold great significance. Phytochemicals are plant-derived substances that can affect disease development and have been used for the prevention of chronic and degenerative diseases [4]. Flavonoids represent a class of low molecular weight polyphenolic phytochemicals found universally in plants [5]. Based on their chemical structure, they are categorized as flavonols, flavones, flavonones, isoflavones, catechins, anthocyanidins, and chalcones (see Fig. 1). They are characterized by a flavan nucleus and C6-C8-C6 carbon skeleton [5]. The antioxidant, anti-inflammatory, anti-thrombogenic, anti-angiogenic, and anti-cancer activities and their ubiquitous presence in fruits, vegetables, beverages, etc. underscore the importance of flavonoids in the human diet and diseases such as cancer [6]. Cancers of the genitourinary organ system include prostate, testicular, ovarian, cervical, urinary bladder, and kidney. Genitourinary cancers contribute to more than a fourth of the total cancer incidence in the USA and approximately 15 % of cancer-related deaths. An estimated 480,000 people are expected to be diagnosed with these collective cancers, and more than 89,000 deaths are expected to occur in 2014 [7]. Prostate cancer is the second most commonly diagnosed cancer in men and is generally associated with aging men. Testicular germ cell tumors on the other hand are a common solid malignancy of young male adults that are either seminomas or non-seminomas and occur with almost equal frequency. Ovarian cancer is the most deadly form of gynecological cancer in women. Urinary bladder cancer has the infamous distinction of being within the top 10 most common cancers and the most common urinary system cancer that affects both men and women. Kidney cancer is the second most common urinary system cancer affecting both men and women. In this chapter, we discuss DNA methylation changes in different genitourinary cancers (summarized in Table 1), their ability to modulate expression of genes involved in cancer development and progression, and the role of flavonoids in regulating epigenetic modifications in these malignancies.

Fig. 1
figure 1

Chemical structure of the flavan nucleus (adapted from www.mdpi.com)

Full size image
Table 1 List of epigenetically silenced genes in GU malignancies with the target organs
Full size table

DNA Hypermethylation

Archetypal CpG islands are devoid of methylation to allow transcription of required genes in the presence of essential cofactors and appropriate chromatin structure. DNA methylation at palindromic CpG sites by DNA methyltransferases (DNMTs) generates a pattern of methylated CpG dinucleotides at the 5′ end of the promoter of human genes. Methylation is a mechanism to control embryonic development, silence the X chromosome, prevent unwanted transcription, etc. DNA methylation can also occur in the “shore regions” of CpG islands, which have a comparatively lower CpG density [8, 9]. DNA promoter hypermethylation-mediated transcriptional repression of genes prevents the binding of transcriptional activators and helps in the recruitment of methylcytosine-binding proteins (MBPs). These proteins help load DNMTs and histone deacetylases (HDACs) participating in chromatin remodeling and transcriptional silencing [10, 11]. DNMT1 shows a preference for hemimethylated DNA and is considered to be a maintenance enzyme that keeps the methylation pattern intact after DNA replication [12, 13]. DNMT3a and DNMT3b on the other hand do not distinguish between methylated substrates and are involved in de novo DNA methylation [13]. Loss of genome-wide DNA methylation and lethality of DNMT-knockout mice demonstrates the importance of these enzymes in mammalian development [14, 15]. DNA methylation is a critical process during germ cell development that relies heavily on DNMTs. Among them, the de novo DNMTs mainly exert their function during prenatal germ cell development while the maintenance DNMTs become critical in proliferating spermatogonia shortly after birth in the male [16]. Changes in DNA methylation profile and heritable decrease in DNA-5-MeC due to reduced DNA methylation fidelity maintenance was initially considered to be solely a hypomethylation process that resulted in overexpression of oncogenes [17]. However, the theory of demethylation of oncogenes leading to their activation has been replaced by the growing popularity of hypermethylation of tumor suppressor genes. These CpG islands become hypermethylated in malignant cells, thus inactivating certain tumor suppressor genes through a progressive process of numerous “waves” of dysregulated methylation unlike a gene mutation. There are two hypotheses by which hypermethylation takes place. The first is that methylation spreads from normal methylation centers to CpG islands devoid of methylation, and the second involves “seeding” of methylation that is already present and certain single-CpG dinucleotides become methylated inducing more cooperative methylation in the surrounding to finally lead to hypermethylation [18]. As discussed below, DNA methylation plays an important role in genitourinary cancers through the modulation of numerous genes that play critical roles in cancer cell biology [1]. Genes involved in the regulation of cellular processes such as hormone response, cell cycle progression, DNA damage and repair, signal transduction, and tumor invasion and architecture have deregulated hypermethylation providing the needed advantage to the sustenance of cancer cells.

Tumor Suppressors

The tumor suppressor phosphatase and tensin homolog (PTEN) is commonly silenced by promoter methylation in many of the genitourinary cancers. PTEN hypermethylation is an early event seen in patients with recurrent or fatal cervical cancer [19]. Several other potential tumor suppressors are atypically methylated in cervical cancer. Loss of Ras association domain family 1 isoform A (RASSF1A) leads to tumor progression, suggesting a tumor-suppressive role for this protein. In cervical cancer cells, hypermethylation of RASSF1A is a mechanism through which cervical cancer cells extinguish death receptor-mediated cell death [20, 21]. Aberrant DNA methylation of the POU2F3 promoter, which is a transcription factor with putative tumor-suppressive function involved in cell type-specific differentiation, is common in cervical cancer [22]. Testisin, a putative tumor suppressor and testicular protease involved in sperm cell maturation, and the CDK inhibitor p16INK4a are hypermethylated in testicular cancer [23•, 24]. Methylation of the p16INK4a promoter and subsequent inactivation are involved in the initiation of bladder cancer [25]. Hypermethylation of the other gene product of CDKN2A, p14, in normal bladder samples after resection has been found to be a predictor of bladder cancer recurrence [26]. The most important genetic event in clear cell renal cell carcinoma (ccRCC) is the hypermethylation-mediated inactivation of the von Hippel-Lindau (VHL) tumor suppressor gene, which stabilizes hypoxia-inducible transcription factors HIF-1 and HIF-2 and the induction of a multitude of hypoxia-inducible genes [27, 28]. RASSF1 is frequently methylated in sporadic renal cell carcinoma (RCC) (either biallelically or as a second hit following 3p deletion) [29]. RASSF1A methylation was detected in normal kidney tissues adjacent to the tumor but not in distant normal tissues, indicating that the TSG methylation is part of a “field effect” and an early event in cancer development causing epigenetic alterations in a large number of cells which is then accompanied by additional genetic and epigenetic changes [30]. Another tumor suppressor gene, TU3A (located at 3p21.1), is also found to be methylated in 42 % of ccRCC and 25 % of papillary renal cell carcinoma (pRCC) [31]. The FHIT gene that encodes a diadenosine triphosphate hydrolase is an important player in purine metabolism, and its promoter methylation has been seen in both ccRCC and pRCC [32]. Methylation of retinoic acid receptor beta (RAR beta), which regulates cell proliferation and differentiation, is also reported in less than 20 % of RCC cases [32–34].

Steroid Receptors

Steroid receptors are critically involved in the regulation proliferation, differentiation, and development of prostate cells with the androgen receptor (AR) being a key player in prostate cancer. Interestingly late-stage prostate cancer tissues have methylated AR promoter, implicating methylation of the AR promoter as a late event in prostate cancer [35]. Estrogens, which are important in female sexual development, also affect benign prostatic hyperplasia and the development of prostate cancer with estrogen receptor (ER) expression being inversely correlated with histologic grade or pathologic stage of prostate cancer. The expression of ER-α isoforms was found to be dependent on the methylation status of their respective promoters with ER-α-C being unmethylated whereas ER-α-A and ER-α-B being methylated in prostate cancer tissue samples [35]. Methylation of progesterone receptor (PR) is found in some prostate cancer cell lines and is believed to also be a late-stage event in prostate cancer [35]. Estrogen and progesterone receptor genes are also methylated in cervical cancer [19]. Retinoic acid receptor beta2 (RARbeta2) is methylated in invasive cervical cancer, and methylation levels were found to correlate with the pathological state of prostate cancer [32].

Cell Cycle Regulation and Apoptosis

D-type cyclins play a very important part in cell cycle regulation, and high cyclin D2 methylation levels have been associated with invasive prostatic carcinoma and clinically aggressive prostate cancer [36]. The human papillomavirus (HPV), a cervical cancer causative agent, can exist either in the integrated or episomal state in the cancer cells. The integrated form of HPV has been reported to be associated with methylation of the cell cycle regulatory protein cyclin A1 (CCNA1), which is inversely correlated with p53 mutation status in head and neck cancers [37]. Transcriptional silencing of apoptosis-related genes such as death-associated protein kinase (DAPK) due to aberrant promoter methylation leads to the impairment of the apoptotic machinery and disrupts cell growth and death homeostasis in prostate and cervical cancer cells [38–41].

DNA Repair

The glutathione S-transferases (GSTs) consist of a family of enzymes that convert electrophilic and hydrophobic compounds into more soluble and easily excreted forms by catalyzing their conjugation to glutathione. They participate in the detoxification of several potentially carcinogenic compounds, thereby contributing to cancer prevention. Glutathione S-transferase P1 (GSTP1) was one of the earliest genes identified to be hypermethylated in prostate cancer [42]. Subsequently, others such as glutathione S-transferase M1 (GSTM1), glutathione peroxidase 3 (GPX3), and O-6-methylguanine DNA methyltransferases (MGMT) have been identified as having hypermethylated promoters in prostate cancer [43–46]. MGMT encodes an enzyme involved in DNA repair and removes alkylated guanine in DNA through stoichiometric transfer of the alkyl group at the O-6 position to a cysteine residue. Methylation of the MGMT promoter leads to its inactivation and the development of non-seminomatous testicular cancer and is also seen in cervical cancer [20, 47]. Given the importance of oxidative stress in cancer, loss of function of gene products involved in neutralizing oxidative DNA damage is of great significance. Promoter hypermethylation of BRCA1 observed in ovarian cancer may be an alternative mechanism in silencing the tumor-suppressive function of this gene in ovarian cancer [48].

Migration, Invasion, and Metastasis

Laminins are components of the extracellular matrix (ECM) that maintain the architecture of the basal lamina surrounding the epithelial cells and modulate cell migration and invasion, whereas the E-cadherin-catenin complex plays a key part in epithelial cell-cell adhesion and tissue architecture. Methylation of LAMA3 and LAMB3 has been found to be associated with poor grade and stage in bladder cancer, while the methylation of LAMC2 has been linked to a drop in survival time in patients [49]. Aberrant methylation of the LN5 and E-cadherin has been found to correlate with clinicopathological features and invasion in prostate cancer [50, 51]. Other genes implicated in tumor invasion such as familial adenomatous polyposis (APC), caveolin 1 (CAV 1), cadherin 1CD44, cluster differentiation antigen (CDH1), α-3 laminin (LAMA3), γ2 laminin (LAMC2), and TIMP3 have also been found to undergo hypermethylation in prostate cancer [38, 44, 45, 51]. In bladder cancer, downregulation of the CDH1 gene through promoter hypermethylation is frequently correlated with staging and grading of bladder cancer, but the relation with methylation patterns is somewhat inconsistent with other studies showing high to low level of methylation at the promoter of CDH1 in bladder cancer [45, 52, 53]. Promoter methylation of E-cadherin is also seen in cervical cancer [54]. Other cell adhesion genes including OPCML, ICAM-1, and CDH1 are hypermethylated in ovarian cancer [55, 56].

Miscellaneous

Genes involved in normal development such as the homeobox genes (HOXA9 and HOXB5) are variably methylated in testicular cancer [57]. Global methylation at long interspersed nuclear element-1 (LINE-1) is inherited in familial testicular cancer kindred [58]. Interestingly, hypermethylation-mediated silencing of the small regulatory PIWI-interacting RNAs (piRNAs) including PIWIL1, PIWIL2, PIWIL4, and TDRD1, which regulate male germ line development, has been observed in primary seminoma and non-seminoma testicular cancer [23]. Several genes involved in signal transduction including Dickkopf (DKK3), endothelin receptor type B (EDNRB), RASSF1A, runt-related transcription factor 3 (RUNX3), and secreted frizzled-related protein 1 (SFRP1) have been found to be hypermethylated in prostate cancer [38, 43, 44, 46].

Hypomethylation

Loss of DNA methylation was the first reported epigenetic change in human cancer [59]. Approximately half of the genome comprises of highly repeated DNA sequences, which contribute to the global DNA hypomethylation commonly observed in cancers [60–62]. The most commonly studied DNA hypomethylated repeats in cancer include tandem centromeric satellite a, juxta-centromeric satellite 2, interspersed Alu, and LINE-1 repeats [61, 62, 63••, 64, 65]. The association between loss of methylation in mice lacking DNMT or fed methyl-deficient diet and cancer development provides evidence for the involvement of DNA hypomethylation in cancer development [66, 67]. In general, unlike hypermethylation, hypomethylation of transcriptional regulatory sites in cancer appears to occur infrequently [68]. Hypomethylation of gene promoters in cancer is associated with decreased overall genomic methylation or satellite DNA that correlates with increased transcription [69–72]. Evidence for the contribution of gene region hypomethylation to oncogenesis suggests a greater involvement in activating genes involved with tumor invasion and metastasis as well as drug resistance [73, 74].

DNA Hypomethylation in Genitourinary Cancers

DNA hypomethylation appears to be a late event in prostate cancer contrary to other cancers. Although some LINE-1 promoter hypomethylation is seen in primary prostate cancer, the extent of hypomethylation is much greater when there is lymph node involvement [75]. Hypomethylation of wingless-related MMTV integration site 5A (WNT5A), S100 calcium-binding protein P (S100P), and cysteine-rich protein 1 (CRIP1) has been observed in prostate cancer [76]. Hypomethylation and overexpression of the protease urokinase are associated with prostate cancer progression [77]. Testicular germ cell seminomas display a large amount of genomic hypomethylation, although this may be linked to the cell type from which this tumor originates [78]. Hypomethylation has been found to cause re-expression of previously hypermethylated genes, such as the HRAS, BCL2, ABCB1, S100A4, SNCG, and CCND2 [39], and numerous cancer-/testis-associated genes [40] in testicular cancer. Global DNA hypomethylation has been observed in cervical cancer, and increased DNA hypomethylation is involved in progression and appears to correlate with grade [79]. Hypomethylation of satellite DNA has been observed in ovarian cancer and found to increase with stage and grade of ovarian cancer. Further, hypomethylation of satellite 2 DNA correlates with poor prognosis [80]. Genomic DNA hypomethylation is considered to be a biomarker for bladder cancer with increased LINE-1 hypomethylation and greater number of hypomethylated loci in cells derived from blood and urine of bladder cancer patients [81]. Heparanase, an enzyme that acts both on the cell surface and within the extracellular matrix to degrade polymeric heparan sulfate molecules into shorter-chain-length oligosaccharides, is also regulated by hypomethylation in bladder cancer [82].

Flavonoids as Modulators of DNA Methylation

Fruits, vegetables, and other plant products used in beverages are a rich source of polyphenols with flavonoids accounting for a large proportion of the polyphenolic compounds. Almost all categories of flavonoids have cancer prevention properties due either to their antioxidant or antiinflammatory properties. However, a comprehensive analysis of the effect of this class of polyphenolic compounds on DNA methylation-mediated effects in cancer still remains to be carried out. Table 2 shows the flavonoids that have been tested in GU malignancies.

Table 2 List of commonly tested flavonoids in GU malignancies
Full size table

Flavonols include quercetin, myricetin, catechins, etc. Quercetin is present in red onions, buckwheat, red grapes, apple skin, green tea, etc. It also forms the structural backbone of the flavonones hesperitin. Despite its ability to demethylate the p16INK4a gene promoter in colon cancer cells, its ability to modulate DNA methylation in genitourinary cancers remains to be vigorously tested. Quercetin was shown to decrease bladder cancer cell growth and to induce apoptosis by decreasing the DNA methylation of the estrogen receptor (ER-β), p16INK4a, and RASSF1A [83]. Myricetin, which is found in many grapes, berries, fruits, vegetables, and herbs, can also decrease DNA methylation by inhibiting SssI DNMT [84].

The green tea polyphenol (−)-epigallocatechin-3-O-gallate (EGCG) is the most studied catechin. It inhibits tumorigenesis partially by affecting DNA methylation through inhibition of DNMTs [84]. EGCG can re-express many transcriptionally silenced genes through inhibition of DNMT1 enzymatic activity in prostate cancer cell lines [85–89]. EGCG was shown to decrease growth and induce apoptosis in renal cell carcinoma by re-expressing tissue factor pathway inhibitor-2 (TFPI-2), a member of the Kunitz-type serine proteinase inhibitor family, and decreasing its promoter hypermethylation [90]. EGCG also decreases the promoter methylation of other genes such as hTERT and CDX2, thereby helping in tumor suppression [91, 92].

Isoflavones present in legumes, pomegranate seeds, etc. are another important group of dietary flavonoids, which includes the soybean isoflavone genistein. Genistein has been most widely studied for its effect on DNA methylation in cancer including genitourinary cancers. Genistein has also been shown to exert its anticancer effect through modulation of DNA methylation [93, 94]. Genistein can change DNA methylation at gene promoters in the prostate of C57BL/6J mice [95]. Genistein was also found to decrease DNMT activity resulting in transcriptional activation of genes such as p16INK4a, RAR beta, MGMT, PTEN, and CYLD in prostate cancer and RAR beta 2 in cervical cancer [93, 96]. Genistein-mediated modulation of methylation of pro-tumorigenic miRNA-1260b and its targets sFRP1 and Smad4 inhibits prostate cancer cell proliferation, invasion, and TCF reporter activity [97•]. Genistein upregulates the GTP-binding RAS-like gene (ARHI), which is an imprinted tumor suppressor gene that is downregulated in prostate cancer [98].

Other common dietary phenolic compounds found in many fruits and vegetables including hesperetin, naringin, apigenin, and luteolin can also modulate DNA methylation. Apigenin, a flavone found in parsley, artichoke, basil, celery, and other plants, has been shown to inhibit the hypermethylation of various tumor suppressor genes. It has also been shown to reduce MssI enzyme-mediated hypermethylation in prostate epithelial cells [99]. Although these compounds are not as efficient as EGCG in directly inhibiting DNMT activity, they can indirectly regulate DNMT activity by regulating the ratio of S-adenosyl methionine (SAM) and S-adenosyl homocysteine (SAH) during their metabolic methylation by catechol-O-methyltransferase (COMT) [84, 100]. Curcumin was found to bring about global hypomethylation in the MV4–11 leukemia cell line by inhibiting DNMT [101]. It was also found to negatively regulate DNMT1 in ovarian cancer and melanoma [102, 103]. Curcumin was also observed to decrease CpG promoter methylation of Neurog1, a highly hypermethylated marker in prostate cancer [104], and also the hypermethylation at the promoter region of the tumor suppressor retinoic acid receptor 2 gene in cervical cancer [105]. It was also found to be involved in the re-expression of Nrf2, a critical regulator of the antioxidant response, by reducing promoter hypermethylation in transgenic adenocarcinoma of the mouse prostate (TRAMP) prostate cancer cells [106]. Curcumin also reduces hypermethylation of the Fanconi anemia (FANCF) promoter, thereby regulating growth and proliferation in cervical cancer [107].

Future Directions

Clearly, dietary components including flavonoids affect DNA methylation patterns. Most reports in the literature have used in vitro model systems to test the demethylation effects of bioactive dietary compounds to restore functions associated with tumor suppression. It remains unclear how demethylation of methylated cytosines in the gene promoter regions would affect hypomethylation given the suggestion that there is crosstalk between demethylation and de novo methylation pathways during tumorigenesis that can make one pathway dependent on the other. Future work is needed to systematically analyze the influence of these bioactive components on the distribution of DNA methylation and demethylation patterns in cancer and normal tissues in order to predict the usefulness of these agents as clinical epigenetic modulators. Further, more attention should be exercised with regard to dietary consumption of bioactive foods in clinical trials when DNMT inhibitors are being tested. Future studies should include a more thorough investigation of specific flavonoids to determine their global epigenetic effects and how they influence outcomes in normal and cancer states. Differential effects on each flavonoid on specific genes and their functional outcomes need further elucidation. Obtaining a comprehensive understanding of the capabilities of flavonoids can result in their development as novel, natural epigenetic modulators for cancer prevention.