The data on the protein analysis of β-catenin overexpression with AXIN2 and BTRCP low mRNA/protein expression were cross-referenced to investigate the correlations between these four proteins. An association between the overexpression of β-catenin protein with low expression of either AXIN2 or BTRCP mRNA/protein was found (Table 7). Using a definition of alteration as low expression in either one of the upstream effectors in β-catenin degradation pathway, we found that among the tumors with overexpression of β-catenin, 68%
(23/34) showed a low expression of either AXIN2 or BTRCP (P = 0.032) (Table 7). The correlation appeared to be also significant especially when mRNA and protein of the AXIN2 and BTRCP genes were analyzed (P = 0.004 ~ 0.013) (Table 7).
DISCUSSION
Previous studies have shown the overexpression of β-catenin protein in 17-63% of lung cancer (88, 92, 102) with only 0-5% of patients showing mutations in β-catenin gene (92, 104, 156). It is possible that β-catenin protein accumulation may be a result of alternative mechanisms other than mutation leading to protein stabilization in lung cancer. In an effort to better understand the molecular pathogenesis of Wnt pathway in NSCLC, we collected 78 primary NSCLC tumors and carried out a comprehensive molecular analysis including mRNA/protein expression and DNA methylation/deletion of four genes (β-catenin, AXIN2, BTRCP, and ICAT) encoding proteins known to function in the Wnt signaling pathway. We found that 55% of NSCLC tumors showed β-catenin overexpression and that the β-catenin deregulation was most often attributable to inactivation of AXIN2 or BTRCP. In addition, β-catenin overexpression and low expression of AXIN2/BTRCP were frequent in squamous cell carcinomas. To our knowledge, our study is the first to identify decreased mRNA/protein expression and promoter hypermethylation of AXIN2, BTRCP, and ICAT genes in lung cancer.
Our data also provide new evidence for diverse mechanism of β-catenin deregulation in the NSCLC tumorigenesis.
Low levels of mRNA and protein AXIN2 expressions occurred in 35-44% of primary NSCLC tumors and they correlated with elevated expression of β-catenin. A high concordance was observed between alterations in protein and mRNA expression and promoter hypermethylation of the AXIN2 gene. The data suggest that promoter
hypermethylation is the predominant mechanism involved in the deregulation of the AXIN2 gene. Note that deletion of two microsatellite markers locating in chromosome 17q24.3 (where AXIN2 is located) was found in 32% of NSCLC tumors. This chromosomal region has also been shown to be frequent LOH in breast cancer, neuroblastoma, heptocarcinoma, and oesophageal squamous carcinoma (111).
A number of studies have shown that AXIN2 is critical for mediating the down-regulation of β-catenin (97). The intracellular concentrations of AXIN2 are much lower than those of other degradation complex components in the β-catenin and it is upregulated in response to increased β-catenin concentrations, suggesting that AXIN2 is the limiting factor in β-catenin degradation pathway. Although high levels of AXIN2 mRNA expression have been detected in human colon, ovarian and liver cancers (97, 144, 157). However, mutations in the AXIN2 gene rarely occur in human cancers. For example, mutations in AXIN2 were found in approximately 20% of colorectal tumors (137). A fraction of hepatocellular carcinomas (~10%) with wild-type β catenin have mutations in AXIN1 and AXIN2 (158). In addition, no significant overexpression of AXIN2 mRNA was observed in breast, lung, and uterus tumors (143, 144). Lusting et al. (2002) previously showed a decrease of AXIN2 mRNA expression in uterine and lung carcinomas.
Their findings were consistent with our data that low AXIN2 mRNA expression occurred in lung cancer. Our observation of loss expression of AXIN2 supports to the premise that AXIN2 is a tumor suppressor in NSCLC carcinogenesis and shows for the first time a frequent AXIN2 5’CpG hypermethylation in lung cancer and an inverse correlation of
AXIN2 and β-catenin expression in NSCLC tumorigenesis. To our knowledge, the present study is the first report to show the expression of AXIN2 at both the RNA and protein levels in lung cancer.
In addition, a high frequency of low AXIN2 mRNA or protein can be detected in early-stage and well and moderately differentiated lung tumors, indicating that loss of AXIN2 might be an early event during lung cancer initiation. The expression analysis of the AXIN2 gene may be examined in precancerous samples such as dysplasic and metaplasic lesions from patients. Note that decreased mRNA and protein expression of AXIN2 was lowered significantly at early stage, but not at stage III and IV. A relationship might exist between the expression of AXIN2 and poor survival, and only cancer patients without AXIN2 alteration could be alive with the advanced stages of the disease. How the expression of AXIN2 influences the survival of lung cancer patients is a question that warrants further analysis.
As an important regulatory molecule, BTRCP is the degradation of major regulatory proteins of the Wnt/β-catenin signaling pathway.
Consequently, loss of BTRCP function results in the stabilization of β-catenin (159). We found a decreased mRNA and protein expression of BTRCP1 in 29-32% of NSCNC tumors. Consistent with our studies, He et al. (2005) indicated that loss of BTRCP protein is also found in lung cancer cell lines (152). In addition, prior studies showed that mutations in BTRCP gene are rarely, suggested additional silencing mechanisms for the BTRCP gene such as hypermethylation (148-150, 152). In our data, a highly significant correlation was observed between aberrant
mRNA/protein expression and hypermethylation of BTRCP, suggestion that the promoter hypermethylation is the predominant mechanism of BTRCP deregulation. Our data also provide the first clinical evidence that BTRCP mRNA is inversely correlated with β-catenin expression in supporting its role as a tumor suppressor and a regulator of Wnt signaling.
Presence of multiple splicing variants of the BTRCP gene had only been reported in Xenopus laevis (160), brain tissue (150), and melanomas (149). Their data indicated that BTRCP transcript variant 1 and transcript variant 2 showed approximately equal expression in normal and tumor tissues. To our knowledge, the present study is the first report to show the expression of splicing variants of BTRCP in lung tissue. Our sequencing data demonstrated an mRNA splicing variant that had a deletion of 108 bp at the exon 2 region. However, we thought that this splicing variant might not play an etiological role in tumorigenesis. This is because the exon 2 is not locating in functional F-box and WD-40 domains of BTRCP protein (119). In addition, the 5’ splicing variant were also expressed in the normal lung tissue. The question whether this splicing variant can be encoded to a protein and the relationship with this BTRCP variant to its β-catenin binding properties will be matter of further investigation.
No ICAT studies have been reported previously in lung cancer. The present study showed that low levels of mRNA and protein expression of the ICAT gene occurred in 29-35% of primary NSCLC tumors and the decreased expression correlated with the promoter hypermethylation of the gene suggesting that ICAT alteration plays a role in a subset of NSCLC
tumors. However, ICAT deregulation did not significantly associated with aberrant β-catenin expression. It is likely that ICAT might affect other cellular functions in lung cancer. In this regard, ICAT has been shown to inhibit both cadherin and TCF binding to β-catenin in vitro (130).
Whether ICAT deregulation could promote NSCLC tumorigenesis, perhaps by inhibiting cadherin function, needs to be further examined.
Note that 10-12 patients with positive protein expression showed promoter methylation. This could be explained by the presence of several distinct tumor subpopulations, one of which has methylation and does not express the protein, the others having no methylation and expressing the protein. In addition, methylation at only one allele could explain the discordant result for these patients. For tumors with no demonstrable protein staining and no evidence of promoter methylation, some had LOH and others may had mutations. In our data, correlations were found between LOH of microsatellite markers in AXIN2 and BTRCP genes suggesting synergistic interactions between these loci. Alternatively, these correlations may be the consequence of sequential molecular genetic changes that occur in the development of lung cancer.
In conclusion, our findings corroborate that low mRNA and protein expression in three tumor suppressor-like genes, AXIN2, BTRCP, and ICAT in primary NSCLC tumors and that promoter hypermethylation is the predominant mechanism in deregulation of these genes. Overall, 67% (52/78) of NSCLC tumors had alteration in at least one of these three genes and the decreased AXIN2 and BTRCP expression significantly contributed to overexpression of β-catenin. Our data
provide compelling evidence that alteration in Wnt pathway is involved in NSCLC tumorigenesis. We proposed that the β-catenin deregulation in lung cancer was mostly attributed to the alteration in its cytoplasmic degradation pathway involving the AXIN2 and BTRCP proteins (Fig. 11), but not through its nuclear inhibition signaling mediated by ICAT.
Alterations in other components of Wnt signaling, including TCF, AXIN1, APC, GSK, DVL, and PTEN in lung cancer are under the investigation to more clearly dissect the aetiology of this disease. The search for antagonists or agonist might lead to the discovery of compounds that have potential for the treatment of lung cancer. In addition, other related signaling pathways such as SAPK/JNK and TGFβ are worth examining.