Aberrant methylation in preneoplastic gallbladder lesions

Chronic cholecystitis is an inflammatory disease associated with the presence of gallstones, which predetermines the appearance of the first morphological changes described in the metaplasia-dysplasia-carcinoma cancer sequence. House et al.[28] evaluated the hypermethylated state of six tumor-associated genes in a normal gallbladder, chronic cholecystitis and adenocarcinomas (samples fixed in formalin and embedded in paraffin) in a series of Chilean patients. In patients with chronic cholecystitis, they reported 28% methylation in some of the genes (APC p16 and hMLH1), whereas only one case (5%) presented multigene methylation [28]. Likewise, García et al.[29] determined the gene methylation pattern in preneoplastic and neoplastic gallbladder lesions, finding that DAPK-1 DLC-1 TIMP-3 and RARβ-2 presented a progressive increase in their state of methylation from chronic cholecystitis to advanced carcinomas [29], reporting at the same time an aberrant methylation pattern of the gene for E-cadherin (CDH1) with a progressive increase in the methylation from chronic cholecystitis without metaplasia to advanced carcinoma (53% to 65.2%) [29]. Recently Shin et al.[30] reported that genes TWIST1 HOXA1 SFRP1 PENK GRIN2B CDH13 NEUROG1 TMEFF2 TIMP3 MINT2 CCND2 RASSF1A RUNX3 DLC1 CRABP1 GATA3 MT1G SEZ6L SOCS3 THBS1 and BCL2 were significantly methylated in extrahepatic cholangiocarcinoma (EHC) tissue samples, at higher levels than in cholecystitis and cholangitis tissues. This analysis was able to define a five-gene panel (CCND2 CDH13 GRIN2B RUNX3 and TWIST1) in bile fluid samples capable of detecting EHC at a sensitivity of 83%, which was far higher than that of bile cytology (46%, P < 0.05) [30].

Aberrant methylation in gallbladder cancer

Recently it has been reported that PSCA is downregulated in non-neoplastic gallbladder lesions such as cholesterolosis, cholecystolithiasis and cholecystitis (9 out of 17; 53%). However, it was disclosed that the expression was decreased in more than 90% of cancers (40 out of 44) [53]. A DNA methylation assay revealed that the methylation level of the gene enhancer region was comparatively low in the cell lines of GBC with a relatively higher PSCA expression, suggesting that the methylation level of the region is related to the level of PSCA expression [53]. However, in the four GBC samples, no correlation was detected between the PSCA expression level and the DNA methylation level of the PSCA enhancer [53]. In others tumors, such as prostate and pancreatic cancers and gliomas, PSCA was recently reported to be upregulated [54], which suggests that PSCA has a different function dependent on the type of cancer. However, its function in normal and malignant epithelial cells is unknown.

CDH1

The E-cadherin gene (CDH1), located on chromosome 16q22.1, is one of the most important tumor suppressor genes [55, 56]. CDH1 encodes for a transmembrane glycoprotein of 120 kDa that intervenes in cell adhesion mediated by calcium. It is a component of the E-cadherin/beta-catenin complex, which is important for cellular polarity, normal tissue morphology and cellular differentiation [56]. It also integrates adherens junctions (originally known as belt desmosomes) that form a continuous belt around the cell to bind the epithelial cells to each other and to maintain the integrity of the stratified epithelium [56]. CDH1 belongs to a family of genes directly related to the processes of tumor invasion and cytoskeleton destabilization. CDH1 expression, it has been reported in less differentiated tumors, that generally have an unfavorable prognosis [57, 58]. The most aggressive carcinomas generally show losses of epithelial cell cohesion, and this is often associated with a reduction in E-cadherin expression [59]. Some mutations of this gene have been detected in gastric, breast and endometrial cancer [60, 61]. The methylation rate of different tumor suppressor genes in cancers of the digestive tract (stomach, colon, pancreas and gallbladder) has been studied, and show a high methylation frequency of CDH1 (77.8%) [46]. CDH1 has been described to be hypermethylated in GBC (Table 1), with frequencies that range from 11.1% to 65.2% [29, 32, 39, 43, 46, 47]. Likewise, the methylation of this gene in advanced stage III and IV of GBC was evaluated, demonstrating approximately 60% methylation [46]. Previously, we reported that reduced expression of E-cadherin in resected gallbladder cancer tissues was significantly correlated with poor prognosis [62]. Loss of E-cadherin expression, followed by expression of the mts1 gene, may be an important event for increasing cell proliferation, motility and invasion activity in the progression of gallbladder cancer [63].

In relation to the degree of tumor differentiation, it was observed that CDH1 presented a high methylation frequency in poorly differentiated tumors and in cases with three or more positive nodes [46]. A significant decrease in CDH1 expression has been recorded as the lesion progresses, which may be due to methylation [39]. As far as patient survival is concerned, this gene does not exhibit significant differences in survival between the methylated and non-methylated cases [29, 39].

Therefore, there is a significant correlation between the methylation of CDH1 and the metastatic phenotype in GBC [32], which has already been described in previous studies on breast cancer [64].

Hypermethylation of CDH1 has been reported in two cell lines of biliary tract cancer (SNU-478 and SNU-1079), with silenced mRNA expression. After treatment with a demethylating agent (5-aza-2′-deoxycytidine), CDH1 was successfully re-expressed [65].

CDKN2A-p16

CDKN2A-p16 is a tumor suppressor gene that encodes the protein p16, which is an inhibitor of cyclin-dependent kinase, involved in cell cycle regulation at Checkpoint G1. The loss of p16 expression is usually connected to homozygote deletion, loss of heterozygosity, mutations and methylation. A lower expression of p16 has been reported to increase the activity of D-type cyclin-dependent kinase activity, which translates into aberrant phosphorylation of the retinoblastoma gene, producing uncontrolled cell proliferation [31, 46]. Loss of heterozygosity and homozygote deletion are two different pathways of p16 inactivation and have been shown to be combined with hypermethylation of the promoter in GBC [50]. In hepatocellular carcinoma and intrahepatic cholangiocarcinoma, p16 is frequently inactivated by methylation of the promoter and rarely by deletion or mutation [31]. Inactivation of p16 through methylation of the promoter region has been frequently identified in breast, prostate, head and neck, liver, lung, brain, colon and esophageal cancers and cell lines of bladder cancer [66–69]. This is a tendency that is also observed in GBC with a loss of expression of up to 62.5% [48, 50]. Takahashi et al.[32] reported that a considerable number of genes (8 out of 24 genes) are frequently methylated in GBC in comparison with chronic cholecystitis, varying in methylation frequency between 4% and 80%, with p16 having the lowest rate [32].

In 2006, Roa et al. published the methylation status of five genes in advanced carcinoma of the gallbladder and their correlation with the immunohistochemical expression, demonstrating that CDH1, APC, FHIT and CDKN2A might be important in the carcinogenesis of gallbladder [39]. Hypermethylation of CDKN2A is also one of the main mechanisms that induce the loss of p16 expression [49]. Furthermore, Ueki et al. studied different alterations of p16 in 68 tumors in Chinese patients, finding that only 14.8% presented aberrant hypermethylation [31]. Other investigations have identified a methylation percentage of 72.5% at different stages of progression; nevertheless, a significant relationship to the loss of expression of this protein was not established [50]. When the methylation state of this gene was evaluated in samples from the US and Chile, methylation frequencies of 56% were recorded, with similar methylation patterns in both populations [28].

In the analysis of p16 and p14, it was noted that they presented a high methylation rate (between 40% and 60%) for carcinoma of the biliary duct, GBC and for primary sclerosing cholangitis. In normal tissue and choledocolithiasis, either no or an extremely low methylation rate was observed in the two genes [48].

The absence of alterations in p16 (methylation, mutation, loss of heterozygosity in chromosome 9p) in cases of GBC means a better overall survival rate, and it is thus considered a significant prognostic factor [31]. A similar relationship has been seen in patients with stage IA non-small cell lung cancer, where the hypermethylation of p16 was related to poorer survival [70]. This was not the case for other cancers (for example, of the biliary duct and the ampulla of vater), where the alteration had no influence on the prognosis [31].

Reprimo

This is a candidate tumor suppressor gene regulating p53, which is commonly altered in numerous human cancers. Reprimo is a highly glycosylated protein located in the cytoplasm that induces cell cycle arrest at the G2 phase, inhibiting the activity of Cdc2 and cyclin B1 [71, 72]. The transcriptional repression of Reprimo by methylation was initially confirmed together with other genes in pancreatic cancer [72]. In addition, it was found to be hypermethylated in 16 different types of tumors, with a high percentage in gastric cancer (79%), gallbladder cancer (62%), lymphoma (57%), colorectal cancer (56%), esophageal adenocarcinoma (40%), breast cancer (37%) and leukemia (31%) [51]. As a result, it has been described as a potential biomarker for the early detection of gastric cancer [73].

Hypermethylation of this gene is infrequent in normal tissue; however, methylation percentages of 32% and 27% have been described in the gastric epithelium and colorectal polyps, respectively [51]. A low methylation pattern in close to 4% of cases has been described in chronic cholecystitis samples [51].

UCHL1

UCHL1 (also known as PGP9.5) is the only gene with a potential oncogenic role that has found to be hypomethylated in the promoter region in GBC [52]. Is located on chromosome 4p14 and was identified originally as a member of a gene family whose products hydrolyze small C-terminal adducts of ubiquitin (Ub) to generate the ubiquitin monomer [74]. The product of the gene is a peptide responsible for eliminating Ub from proteins that have it, and to thereby avoid its degradation by the proteasome. Proteins degraded by this mechanism actively participate in cell cycle control, for example, p53 and a variety of cyclins [75]. Ub, a protein consisting of 76 amino acid residues, is present in all eukaryotic cells tested. It plays a role in the degradation of abnormal and short-lived proteins by the ATP- and Ub-dependent proteolytic systems. UCHL1 dysfunction in neurons is known to be involved in familial Parkinson’s disease [76]. In vivo analysis of UCHL1-deficient mice suggested that it functions as a regulator of apoptosis in neurons [77] and also in germinal cells during spermatogenesis [78].

The accumulation of Ub has been documented in several types of primary cancers [75]. During the past few years, several works have reported the link between UCHL1 expression and tumor progression, and that it may be useful as a potential marker of several human cancers, such as non-small cell lung cancer [79, 80], invasive colorectal cancer [81], pancreatic cancer [82], squamous cell esophageal carcinoma [83] and neuroblastoma [84]. In GBC, a progressive decrease in the methylation of this gene has been observed, with 84.6% in normal epithelium, 37.5% in adenoma and 27.2% in carcinoma. These results suggest that hypomethylation of the PGP9.5 promoter is a reliable marker in GBC and that DNA hypomethylation might play a significant role in the re-expression of the gene in GBC [52]. In addition, hypomethylation of UCHL1 had previously been found in its promoter region in colorectal cancer, and lymph node metastasis was significantly associated with a lower frequency of methylation [85], poorer survival and a high incidence of recurrence [86].