Genetic and epigenetic aspects of bladder cancer

Precise reason why specific individuals get bladder cancer remains unknown. Genetic polymorphism may be involved in tumorigenesis of urinary bladder, including xenobiotic metabolism, DNA repair, cell cycle control mechanisms, and many cytokine production [Kim et al., 2000, 2002, 2005; Wang et al., 2002; Jeong et al., 2004], which probably reflects not only different incidence rates in different countries [Morrison et al., 1984] but also the combined effects of environmental and hereditary factors. Recently, there has been major progress in both genetic polymorphism in relation to bladder cancer, and molecular genetic and epigenetic changes leading to the development of transitional cell carcinoma.

More than 90% of bladder cancers are transitional cell carcinomas and roughly 60% are low-grade superficial transitional cell carcinomas. The majority of these patients develop cancer recurrences after endoscopic resection, 16%–25% with high-grade cancers. Approximately 10% of patients with superficial bladder cancers subsequently develop invasive or metastatic disease. Almost 25% of patients with newly diagnosed bladder cancer have muscle-invasive disease, the vast majority being cancers of high histologic grade. Almost 50% of patients with muscle-invasive bladder cancer already have occult distant metastases [Koss, 1979; Cutler et al., 1982; Greenlee et al., 2001; Messing, 2002].

Transitional cell carcinoma is a mixture of heterogeneous cell populations having different metastatic potentials. Despite the apparent de novo clinical presentation of invasive bladder cancers, cytogenetic, and antigenic evidence supports the hypothesis that transitional cell carcinomas follow the general concept of multi-step carcinogenesis and proceed through two distinct genetic pathways responsible for generating different cancer morphologies. These are the inactivation of cyclin-dependent kinase inhibitors in low-grade transitional cell carcinoma and early p53-mediated abnormalities in high-grade transitional cell carcinoma. The progression of transitional cell carcinoma correlates with genetic instability and accumulation of collaborative genetic lesions [Al-Sukhun and Hussain, 2003]. In addition, the advent of high-throughput is accelerating the identification process of the molecular events characteristic of bladder cancers' phenotype and subsequent behavior. The information provided by these analyses is resulting not only in the identification of novel therapeutic targets for bladder cancer, but also in the development of diagnostic tools.

CpG island methylation is rare in normal cells. It plays a role in X-chromosome inactivation in females and genomic imprinting, and increases with age and in vitro cell culture. Abnormal methylation of CpG islands can efficiently repress transcription of the associated gene in a manner akin to mutations and deletions, and act as one of the ‘hits’ in the Knudsen two-hit hypothesis for tumor generation [Jones and Laird, 1999; Baylin and Herman, 2000]. There are several examples of aberrant CpG island promoter hypermethylation of tumor suppressor genes, genes involved in cell–cell adhesion, and genes involved with DNA repair in relation to bladder cancer [Maruyama et al., 2001; Tada et al., 2002; Dulaimi et al., 2004]. Examples in which one copy of a tumor suppressor gene is either mutated or lost and the wild-type copy is transcriptionally silenced by hypermethylation have been found [Robertson, 2001]. It has recently been recognized that aberrant hypermethylation events can occur early in tumorigenesis and disrupt pathways that may predispose cells to malignant transformation. Aberrant methylation was detectable in pre-neoplastic lesions and the frequency of aberrations increased with disease progression [Belinsky et al., 1998; Nuovo et al., 1999; Wong et al., 1999]. Thus, ample evidence exists to support the notion that DNA hypermethylation events could act as a primary inactivating event contributing directly to tumorigenesis of the urinary bladder. This review focuses on the recent advances of genetic polymorphism, genetic alterations, and hypermethylation events in relation to bladder transitional cell carcinoma.

GENETIC POLYMORPHISM

Genetic variation in the human genome is an emerging resource for studying cancer, a complex set of diseases characterized by both environmental and genetic contributions. It is becoming increasingly apparent that most of population-attributable cancer heritability is related not to the rare deleterious gene defects but to polymorphic variations in the DNA sequence [Ponder, 2001]. This concept has been substantiated in a variety of cancer settings, particularly by the genetic epidemiology of bladder cancer. N-Acetylation of aromatic amine procarcinogens by N-acetyltransferase 2 (NAT2) is generally considered to be a detoxifying mechanism [Cartwright et al., 1982]. Slow and rapid acetylators were shown to have variable susceptibility to bladder cancer. Several comparative studies among different ethnic groups observed less than 20% and more than 55% of slow acetylators among Asians and Whites [Grant et al., 1983; Sinues et al., 1992]. Hein [1988] combined data from 12 studies and noted a statistically significant relative risk for bladder cancer of 1.5 in slow versus rapid acetylators. Among Korean subjects in our study, 7.1% of bladder cancer patients and 11.0% of controls had slow acetylator genotypes [Kim et al., 2000]. Yu et al. [1994] observed that a difference in prevalence of slow acetylator (the high-risk phenotype) among the three ethnic groups (Whites, Blacks, and Asians) closely paralleled their varying incidence of bladder cancer. It might be inferred that the lower frequencies of slow acetylator genotypes in Asians relative to the Westerners could result in less incidence of bladder cancer.

Glutathione (GSH) and GSH-dependent enzymes are involved in the cellular metabolism and detoxification of cytotoxic and carcinogenic products. Accordingly, much interest is concentrated on the significance of polymorphism in the glutathione S-transferase (GST) supergene families, as these enzymes metabolize exogenous and endogenous molecules involved in cell-specific functions, such as proliferation and apoptosis [Nebert, 1994]. The absence of GST-μ (GSTM1) activity is due to homozygous deletion of the gene. A similar polymorphism of the GSTT1 encoding the θ class enzymes were discovered [Pemble et al., 1994]. GSTM1 null genotype has attracted much attention due to its possible association with increased susceptibility to certain malignancies such as lung cancer [Seidegard et al., 1986] and bladder cancer [Bell et al., 1993; Katoh et al., 1998], although the high risk of bladder cancer in individuals with the GSTM1 null genotype is still controversial. Kim et al. [2000] demonstrated an association between the GSTM1 null genotype and increased risk of bladder cancer. On the other hand, few studies showed GSTT1 being a possible risk factor in bladder cancer. Kempkes et al. [1996] reported that the frequency of GSTT1 null genotype was similar in patients with urothelial cancer and controls. Brockmoller et al. [1996] reported that there was no significant difference in the GSTT1 genotype between urothelial cancer cases and controls in all subjects including smokers and non-smokers, although a significantly high frequency of GSTT1 null genotype in the cancer cases was observed in the non-smoker group. To identify the effects of GSTM1-null and GSTT1-null types and smoking status on the bladder cancer incidence with ethnicity as determined by the previous 29 reports, Kim et al. [2002] compared the frequencies of GSTM1-null and GSTT1-null genotypes and estimated smoking prevalence with the age-standardized bladder cancer incidence. In the univariate and multivariate analyses with the ecological data of various countries and ethnic groups, cigarette smoking positively, but the frequency of the GSTT1-null type negatively, correlated with the age-standardized bladder cancer incidence. These results suggest that the GSTT1-null genotype might not be a risk factor but a protective factor of bladder cancer.

Genes encoding a number of cytokines are polymorphic. Tumor necrosis factor-α (TNF-α) is a multifunctional cytokine. Genotype changes at position −308 of the TNF-α promoter are frequently observed in several cancers [Warzocha et al., 1998; Sasaki et al., 2000]. Vascular endothelial growth factor (VEGF) is a major angiogenic factor. VEGF mRNA is markedly up-regulated in human tumors, such as kidney and bladder cancer [Brown et al., 1993]. Several studies reported that the cancer stage and grade were significantly associated with the TNF-α and VEGF genotypes [Jeong et al., 2004; Kim et al., 2005]. These cytokines have angiogenic function, possibly critical for cancer invasion and metastasis. High-grade and stage bladder cancers are more prone to invasion and metastasis, which suggests that the aggressiveness of primary bladder cancer might be associated with the TNF-α and VEGF genotypes.

The mutagenic base, 8-oxoguanine, is removed from damaged DNA by base excision repair. The human 8-oxoguanine DNA glycosylase 1 (hOGG1) gene encodes DNA glycosylase that catalyzes the excision of the mutagenic lesion 8-oxoguanine from oxidatively damaged DNA [Tchou et al., 1991; Lee et al., 1993]. A Ser326Cys polymorphism in hOGG1 has been identified in bladder cancers [Kim et al., 2005]. Distribution of the hOGG1 codon 326 genotypes (Cys326Cys versus Ser326Ser and Ser326Cys) of controls was significantly different from the bladder cancer patients in that Cys326Cys genotype displayed a protective effect against bladder cancer development, compared to Ser326Ser and Ser326Cys genotypes. Currently, there is no convincing explanation that Cys326Cys is a highly protective genotype. These reflect that hOGG1 may not be the only gene associated with oxidative damage. An alternative DNA oxidative damage repair pathway to minimize the effects of 8-oxoguanine in genomes was reported [Klungland et al., 1999].

Numerous single nucleotide polymorphism (SNP) studies have been currently reported in relation to the bladder cancer (Table I). The existence of low-penetrance cancer predisposing polymorphisms is undisputable; unfortunately, very few consistent gene-disease associations have been identified so far. It is hoped, that the ongoing worldwide efforts in obtaining large and informative DNA collections, combined with the rapid development of high-throughput genotyping technologies, will provide useful prognostic markers for clinicians applicable in clinical setting.

GENETIC ALTERATIONS

Cytogenetic studies have identified many structural and numerical chromosomal changes in bladder transitional cell carcinomas. Chromosomal abnormalities, including increased or decreased number of chromosomes (hyperdiploidy and aneuploidy), marker chromosomes, and the chromosomes of abnormal size or configuration, have also been shown to correlate with an increased risk of tumor recurrence and cancer progression [Sandberg et al., 1986; Falor and Ward-Skinner, 1988]. Molecular genetic studies on bladder cancer have demonstrated deletion of several chromosomal arms, including 3p, 6q, 9q, 11p, 17p, and 18q [Sandberg and Berger, 1994]. Loss of 9q were demonstrated in low as well as high-grade bladder cancer suggesting that the loss of 9q may be a primary event in the genesis of bladder cancer [Tsai et al., 1990]. Among the early changes appearing in superficial bladder cancers are deletions of 11p and 8p, and gains of 8q and 1q [Fadl-Elmula et al., 2001]. In contrast, deletions of 17p were not identified in superficial noninvasive papillary lesions, while 60% of invasive cancers exhibited loss of heterozygosity (LOH) of 17p, suggesting involvement in the progression of bladder cancer [Olumi et al., 1990; Presti et al., 1991].

New high-throughput microarray technologies make it possible to gain a comprehensive insight into the molecular basis of human diseases [Bubendorf, 2001]. Search for the genes that are differentially expressed in human tumors has been greatly facilitated by the DNA microarray technology. In this technology, the RNA expression level of hundreds or thousands of genes in a tumor can be surveyed simultaneously. Recently, the tissue microarray technology has also been developed to facilitate large-scale molecular analyses in hundreds of tumors simultaneously [Kononen et al., 1998]. These microarray technologies have widened applications in the urological field. However, only few studies dealt with molecular classification of bladder cancer expression profiling using DNA microarrays. Sanchez-Carbayo et al. [2003] introduced cDNA microarrays containing 17,842 known genes to allow hierarchical clustering of early-stage and invasive transitional carcinomas. It is possible to separate carcinoma in situ from papillary superficial lesions and subgroups within early-stage and invasive cancers displaying different overall survival. The most extensive expression profiling study of bladder cancers reported to date has dealt with the development of a predictive classifier of Ta, T1, and T2 + bladder carcinoma subclasses. Dyrskjot and coworkers [Dyrskjot, 2003; Dyrskjot et al., 2003] reported the identification of clinically relevant subclasses of bladder carcinoma using expression microarray analysis of bladder cancers. Cluster analysis identified three major stages, Ta, T1 and T2-4, with the Ta tumors further classified into subgroups. A 32-gene molecular classifier was able to classify benign and muscle-invasive cancers with close correlation to pathological staging. This also provided new predictive information on disease progression in Ta tumors compared with conventional staging. Furthermore, gene-expression profiles characterizing each stage and subtype identified their biological properties, producing new potential targets for therapy. Dyrskjot et al. [2004] recently reported microarray expression profiling to examine the gene expression patterns in superficial transitional cell carcinoma with surrounding carcinoma in situ, without surrounding carcinoma in situ lesions, and in muscle invasive carcinomas. They identified a few gene clusters that contained genes with similar expression levels in transitional cell carcinoma with surrounding carcinoma in situ and invasive carcinomas. A 16-gene molecular classifier for identification of the carcinoma in situ gene expression signature was constructed and suggested to be useful in the follow-up of bladder cancer patients.

The application of tissue microarrays represents a high-throughput approach for validation of potential novel markers for bladder cancer by immunohistochemistry or in situ hybridization in paraffin blocks [Kononen et al., 1998; Schraml et al., 1999; Richter et al., 2000; Nocito et al., 2001]. Several reports describe the rapid evaluation of targets of interest, such as cytoskeletal actin associated gelsolin or E-cadherin [Rao et al., 2002], Na, K-ATPase or cell-cycle-related markers [Espineda et al., 2003]. The most extensive tissue microarray in bladder cancer, analyzing over 2000 bladder carcinomas, revealed the prognostic utility of cyclin E [Richter et al., 2000]. Focus is intensified in this field to automate the construction of tissue microarrays.

The usefulness of the microarrays has not yet been evaluated by enough clinical trials. Identification of tumor subtypes within the superficial disease and patients more likely to develop positive lymph nodes or distant metastases is critical subclassification issues to be investigated. In the near future, gene profiling will provide an effective means of predicting the response to specific therapeutic regimes based on the molecular signatures of the tumors associated with their chemosensitivity or resistance to anticancer drugs. Moreover, the discovery of molecular pathways altered in cancer progression, as well as the identification of molecule-susceptible targets, would lead to the development of novel alternative therapies. The classical tumor marker concept of an individual biological determinant will be substituted by the use of cluster of genes as predictive classifiers.

DNA replication errors (RER) have been characterized by the addition or deletion of bases within the simple mononucleotide or dinucleotide repeat sequences in the DNA of tumors compared with the matching DNA of normal tissue specimens [Thibodeau et al., 1993]. Microsatellite instability (MSI) is the alteration of the length of simple repetitive sequences (microsatellites) throughout the genome. It is due to dysfunction of the DNA mismatch repair genes (hMSH2, hMLH1, hMSH6, PMS1, PMS2) leading to the accumulation of DNA replication errors. These alterations mainly affect microsatellite sequences in some cases residing in the coding sequence of important growth regulatory genes, hence contributing to the development of cancer [Peltomaki and de la Chapelle, 1997]. That is to say, RER is judged based on the degree of microsatellite instability, which is believed to result in a rapid accumulation of somatic mutations once carcinogenesis is initiated. The status of RER may thus be important for determining both the prognosis and the optimal therapeutic strategy of carcinomas in the future [Lothe et al., 1993]. Microsatellite alterations in bladder cancer have been reported to be associated with invasive phenotype and young age [Sardi et al., 1999; Thykjaer et al., 2001]. Regardless of the underlying mechanism, many groups are testing the possibility of using this microsatellite alteration phenotype for the early diagnosis and follow up of bladder cancer by detecting these shifted DNA bands in exfoliated cell in urine, or bladder washings [Uchida et al., 1996; Larsson et al., 2001; Seripa et al., 2001]. However, microsatellite alterations in urine are indicators not only of malignancy but also of inflammatory conditions [Christensen et al., 2000]. Thus, their role will be additive to the current diagnostic methods.

EPIGENETIC ALTERATIONS

The hereditability of newly acquired traits is a hallmark of clonal expansion in cancers. The inheritance of information on the basis of gene expression levels is known as epigenetics, as opposed to genetics, which refers to information inherited on the basis of gene sequence. Enzymatic methylation of the C-5 position of cytosine residues can affect epigenetic inheritance by altering the expression of genes and by transmission of DNA methylation patterns through cell division. Thus, in addition to its well-known role in deamination mutational hotspots in human DNA, DNA methylation may contribute to gene inactivation in cancer. DNA methylation is a powerful mechanism for the suppression of gene activity. It should be emphasized that this inverse correlation has been demonstrated conclusively only for methylation in the promoter regions and not in transcribed part of genes. The methyl groups do not affect base pairing but can influence protein–DNA interactions by protruding into the major groove [Razin and Riggs, 1980].

Several tumor suppressor genes related to the bladder cancer contain CpG islands in their promoters, prompting many studies to investigate the role of methylation in silencing of these genes (Table II). Hypermethylation in the 5′ promoter region of genes is associated with transcriptional silencing and is an alternative mechanism for down-regulating gene expression rather than gene deletion or mutation. Hypermethylation of the promoter region of the p16INK4A gene in a subset of patients with superficial transitional cell carcinomas has been described and may be an early event in transitional cell carcinoma pathogenesis [Orlow et al., 1999]. Other genes also have been found to undergo aberrant promoter methylation in transitional cell carcinoma but are not found in the normal bladder mucosa associated with these tumors, including DBCCR1 (deleted-in bladder cancer 1 gene), and the 5′ end of transmembrane protein containing epidermal growth factor and follistatin domains [Salem et al., 2000]. Maruyama et al. [2001] determined the methylation status of 10 genes in 98 surgically resected bladder cancers, and calculated the median methylation index (MI), a reflection of the methylated fraction of the genes tested. All of the tumors were transitional cell carcinomas except for one squamous cell carcinoma. Their results indicated that multiple genes are methylated during the process of bladder cancer development. They found frequent methylation of four genes (two members of the cadherin family-CDH1 and CDH13-; a recently identified putative tumor suppressor gene, RASSF1A; and APC). Methylation of these genes and the MI correlated with one or more parameters of worse prognosis, specifically high tumor stage and grade, muscle invasion, nonpapillary growth pattern, and shortened survival. In multivariate analyses, CDH1 methylation-positive status was independently associated with poor survival. Sathyanarayana et al. [2004] analyzed the methylation pattern of LAMA3, LAMB3, and LAMC2, which encoded laminin-5, in 128 bladder cancers and 71 urine samples. Methylation of LAMA3 and LAMB3, and MI were correlated with several parameters of poor prognosis (grade, stage, growth pattern, invasion, and ploidy pattern), whereas methylation of LAMC2 and MI were associated with shortened patient survival. Tada et al. [2002] determined the frequency of aberrant promoter hypermethylation of 7 genes, hMLH1, O6-methylguanine-DNA-methyltransferase (MGMT), p16, Von Hippel-Lindau (VHL), death-associated protein kinase (DAP-kinase), glutathione S-transferase P1 (GST-P1) and E-cadherin in superficial bladder cancers. Simultaneous hypermethylation of three genes or more among the seven genes was a significant concordance between the number of methylated genes and the development of recurrence. In particular, the recurrence rate for 24 months was 88% for hypermethylation of DAPkinase and 28% for nonmethylation of DAP-kinase. The methylation profile may represent a new potential biomarker of risk prediction in transitional cell carcinoma.

SUMMARY

The origin of bladder cancer is not fully explained by a single risk factor. Numerous factors may be involved in carcinogenesis, progression and patient's survival, including genetic polymorphisms, genetic and epigenetic alterations such as chromosomal alterations, and loss of heterozygous (LOH) of specific allele and aberrant DNA methylation, and factors controlling the cell cycle, cell proliferation and apoptosis. In spite of numerous SNP studies in relation to bladder cancer, only a few genetic polymorphisms give us marginal information for patient's prognosis. It is expected that the increasing capacities of available DNA collections, coupled with the rapid development of high-throughput genotyping technologies, will vastly accelerate the research on bladder cancer susceptibility. Though the usefulness of the microarrays has not yet been evaluated by enough clinical trials, this technology will allow a rapid molecular analysis of thousands of tumors within a few hours, with the premise of a better tool for biological classification of bladder cancers. The epigenetic silencing of tumor suppressor genes is interesting from a clinical standpoint because it is possible to reverse epigenetic changes and restore gene function onto a cell. Treatment with DNA methylation inhibitors can restore the activities of dormant genes such as CDKN2A and decrease the growth rate of cancer cells in a heritable fashion. DNA repair capacity can be restored by activation of MLH1. It should therefore be possible to partially reverse the cancer phenotype by the use of methylation inhibitors. Aforementioned genetic and epigenetic signatures will allow a better chance of cure opting for the most appropriate treatment, while maintaining quality of life.