Prognostic significance of TRAIL death receptors in Middle Eastern colorectal carcinomas and their correlation to oncogenic KRAS alterations

Colorectal Cancer (CRC) is a major cause of mortality and morbidity worldwide. In Saudi Arabia, the incidence of CRC is increasing. According to the latest statistics, CRC is the second most common cancer among Saudi males and the third most common among Saudi females [1]. Currently available therapeutic approaches for CRC are less effective, and thus the prognosis is poor. Despite a growing number of publications about biomarkers that give information on disease outcome, the best prognostic factors are still simple clinical parameters like number of lymph nodal metastasis, presence of distant metastasis, tumour grade and AJCC stage. Prognostic biomarkers might especially be useful for hypothesis testing for their relevance as predictive markers, as targets for therapy and for the selection of patients for adjuvant treatment [2].

Apoptosis, or programmed cell death, is a major control mechanism by which cells die if DNA damage is not repaired [3]. Apoptosis is an essential biochemical pathway for normal tissue homeostasis, cellular differentiation, and development. Derangements of normal apoptotic mechanisms provide a growth advantage to cancer cells [4]. The understanding of apoptosis has provided the basis for novel targeted therapies that can induce death in cancer cells or sensitize them to established cytotoxic agents and radiation therapy [5]. In addition, as apoptosis usually does not elicit host inflammatory or immune response, this type of cell death is the preferred way of cancer cell killing by various treatments. Accordingly, selectively inducing apoptosis in tumour cells is gaining recognition as a promising therapeutic approach for many cancers [6]. Tumour necrosis factor-related apoptosis-inducing ligand (TRAIL or Apo2 ligand) is a member of the tumour necrosis factor (TNF) cytokine family that induces apoptosis upon binding to its death domain containing receptors, TRAIL receptor 1 (death receptor 4, DR4) and TRAIL receptor 2 (death receptor 5, DR5) [7]. The TRAIL receptors, TRAIL-R1 and TRAIL-R2, are highly expressed in many cancer cells including CRC [8‐10]. A further three TRAIL receptors exist, which are unable to induce apoptosis and act as decoys. Decoy receptors 1 (DcR1) and 2 (DcR2), similar to TRAIL-R1 and TRAIL-R2, are expressed on the cell surface. Thus, overexpression of either DcR1 or DcR2 confers protection against TRAIL-induced apoptosis [11, 12]. The fifth TRAIL receptor is osteoprotegerin (OPG), a secreted, low affinity receptor for TRAIL [11, 12]. Binding of TRAIL to TRAIL-R1 and TRAIL-R2 induces trimerization of TRAIL-R1 and TRAIL-R2 [13]. The trimerized TRAIL-R1 and TRAIL-R2 bind to FADD, which recruits caspase 8 and initiates a proteolysis cascade that eventually leads to cell death by apoptosis. Many cancer cells are resistant to death receptor induced apoptosis [4]. The mechanisms of resistance include the presence of decoy receptors for TRAIL [12], the loss of TRAIL receptor expression [14], the overexpression of inhibitory proteins in signal transduction pathways such as FLICE-inhibitory protein [14], and the mutation of TRAIL-R2 gene [15‐18].

Oncogenic mutations such as ras may enhance expression of TRAIL receptors; potentially sensitizing these tumors to TRAIL based therapies [19‐21]. Constitutively activated Ras increases the tumorigenic potential of cells because it causes deregulation of important intracellular signaling pathways [22]. Activated RAS mediates its biological activity through interaction with various downstream effector targets, thus activating pathways like MEK, PI3K, and Rho GTPases [22, 23]. RAS regulates a RAF-MEK-ERK1/2 kinase cascade and this pathway is found to be active in human colon adenocarcinomas cells [24] as well as in human colorectal tumors [25]. Drosopoulos et al. [21] have shown transformation of the colon cell line Caco-2 by ras oncogenes sensitizes these cells to TRAIL induced apoptosis by causing specific MEK-dependent up-regulation of TRAIL-R1 and TRAIL-R2. Nesterov A et al. [20] have demonstrated that normal cells are sensitized to TRAIL when TRAIL-R2 is up regulated by overexpression of c-myc or oncogenic ras mutants. Thus, RAS-MEK-ERK1/2 signaling pathway can sensitize cells to TRAIL-induced apoptosis by up-regulating TRAIL-R1, TRAIL-R2 and TRAIL-based therapeutic strategies using TRAIL agonists could be used in cases of human colon cancers bearing RAS mutations. Therefore, we also sought to explore the potential link between expression of TRAIL and its receptors with KRAS alterations in CRC.

The aims of the present study were: (a) to determine the TRAIL/TRAIL receptor expression pattern in normal and neoplastic colon epithelium; (b) to correlate immunohistochemical expression patterns with KRAS alterations, microsatellite instability and pro apoptotic markers; (c) to correlate immunohistochemical expression patterns with overall survival.

We conducted this study to examine the relations of TRAIL and it receptors: TRAIL-R1 and TRAIL-R2 with clinical, pathologic, molecular characteristics and patient survival in Saudi colorectal cancers. Expression of TRAIL-R1 or TRAIL-R2 was associated with a less aggressive phenotype characterized by an early AJCC stage and well-differentiated tumors. TRAIL-R2 expression was associated with microsatellite stable phenotype and with absence of KRAS mutations. TRAIL-R1 but not TRAIL-R2 was an independent prognostic marker for better survival.

Using immunohistochemistry, we have studied the expression of TRAIL and its receptors in Saudi CRC; incidence of TRAIL R1, TRAIL-R2 and TRAIL expression was 85.5%, 59.4% and 31.5% respectively. In agreement with earlier studies, we have also observed a progressive increase in expression of TRAIL and its receptors: TRAIL-R1 and TRAIL-R2 in colorectal carcinoma and noted a strong association of TRAIL -R1 or TRAIL-R2 expression with differentiation and an early stage. The prognostic implication of TRAIL receptor expression is the subject of intensive investigation as malignant cells are more sensitive to TRAIL- induced apoptosis than their benign counterparts are and this potentially affects the future management of patients [32‐34]. Furthermore, our data indicates that high TRAIL-R1 expression was an independent prognostic marker for better survival in Saudi CRC patients. TRAIL-R2 was also associated significantly with better outcome but failed to remain significant in multivariate analysis. TRAIL-R1 expression was also associated with better outcome in the following subgroups: Stage III and IV (p = 0.0060) and CRC subgroup who received adjuvant therapy(p = 0.0033). To elucidate the role of TRAIL expression further analysis was done in the following subgroup: CRC subgroup with high co-expression of TRAIL and TRAIL-R1 and CRC subgroup with high co-expression of TRAIL and TRAIL-R2. Both these combination groups were not associated with outcome (data not shown). Thus, TRAIL ligand co-expression with TRAIL receptors does not influence the outcome.

These findings are in agreement with earlier studies by Starter et al [26] where TRAIL-R1 expression was associated with a better disease free survival in a cohort of 129 Stage II and III CRC [26]. Granci et al. [28] studied the TRAIL receptors TRAIL-R 1, -2, -3 and -4 expression by immunohistochemistry in metastatic stage IV CRC and found that concomitant low/mediumTRAIL-R1 and high TRAIL-R3 expression in primary CRC is significantly associated with a poor response to 5-FU-based first-line chemotherapy and with a shorter progression-free survival. Surprisingly, high TRAIL R1 was associated with worse disease free survival and overall survival in 376 CRC patients with Stage III [30]. Ullenhag et al. [29] analyzed FLICE inhibitory protein (c-FLIP) and TRAIL receptors(TRAIL-R1 and R-2) in 476 CRC of all Stage groups(I to IV): Overexpression of FLIP_L, (the long form of FLICE inhibitory protein) but not TRAIL-R1 or TRAIL-R2, was an independent prognostic factor for shorter disease free survival. In an attempt to explain these conflicting results of TRAIL and its pro-apoptotic receptors in CRC, we offer the following explanations: a) differences and heterogeneity in samples studied: sample size, ethnic differences, different Stage groups, tumor site- colon or rectal tumors, type of treatment- surgery and/or chemo/radiotherapy; b) differences in scoring system could be another important reason for this difference. The varied effects of TRAIL signaling could be also attributed to the following factors: TRAIL resistance due to presence of decoy receptors [11], number, type and functionality of TRAIL receptors [7, 12] and intracellular anti apoptotic molecules like c-FLIP [35], IAP [36], Mcl-1 [37] and bcl2 [38].

Although TRAIL-R1 lost its statistical significance when included as a prognostic marker in multivariate analysis with p27 and KRAS4A (see Additional File 1 Table S2), this does not argue against the biological role of TRAIL-R1 in CRC as much as it reflects that p27 and KRAS4A are a more powerful predictor of clinical outcome of CRC than TRAIL-R1 expression. We can hypothesize that the TRAIL-R1 functions most effectively in the cells which show co-expression of p27^kip1 in concordance with an earlier study [39]. Despite some studies that show a role of Ras signaling pathway in modulating the TRAIL system, studies on the KRAS isoforms - KRAS4A and KRAS 4B are lacking. Alternate approaches to modulate the expression of KRAS isoforms, a greater understanding of the role(s) that each oncoprotein plays in malignant transformation, including the signal transduction pathways affected, is crucial in the development of therapeutic approaches in cancer treatment, which include the use of drugs that target isoform-specific post-translational modifications [40] and of antisense oligonucleotides to modulate alternative splicing [41].

Oncogenic mutations such as ras may enhance expression of TRAIL receptors; potentially sensitizing these tumors to TRAIL based therapies [19‐21]. TRAIL-based therapeutic strategies using TRAIL agonists could be used in cases of human colon cancers bearing RAS mutations. In a small cohort of 51 CRC, Oikonomou E et al.[42] have reported a much lower incidence of KRASG12/13 mutations(10%) and have concluded that there is clear correlation between these mutations(KRAS and BRAF) and upregulation of TRAIL-R1 and TRAIL-R2. Despite lack of statistical significance they have concluded that CRC with mutations in KRAS or BRAF gene had significantly upregulated both TRAIL death receptors. In our earlier study [43] KRAS gene mutations were seen in 80/285 CRC (28.1%) and were an independent prognostic marker for poor survival. Interestingly we have observed a significantly higher expression of TRAIL-R2 (p = 0.0481) in CRC subgroup lacking KRAS mutations(75.5%) as compared to the CRC subgroup with KRAS mutations(61.1%). In view of the recent findings of KRAS mutations and PIK3CA mutations contributing to resistance to EGFR inhibitors like Cetuximab,[44, 45] a better understanding of the TRAIL system with context to KRAS mutations might be useful. The KRAS gene has two alternative fourth exon variants that result from differential splicing and activating mutations affect both isoforms [46‐48]. Studies in animals indicate that KRAS4A promotes apoptosis while KRAS4B inhibits it, and KRAS4B promotes differentiation [49, 50]. In our study [43], KRAS 4A a pro apoptotic isoform, in particular was found to be an independent prognostic marker for better survival in all CRC patients. Even in the CRC subgroup lacking KRAS mutations KRAS4A was associated with better survival. Furthermore, we have observed a highly significant association of KRAS4A and both the TRAIL receptors: TRAIL-R1(p < 0.0001) and TRAIL-R2(p < 0.0001). Considering the tight linkage between TRAIL-R1 and KRAS4A future studies should be conducted to understand the association between these markers.

In summary, our study shows high TRAIL-R1 expression to be an independent prognostic marker for better survival in colorectal cancer. High TRAIL-R1 or TRAIL-R2 expression was associated with a less aggressive phenotype characterized by early AJCC stage, well-differentiated tumors, microsatellite stable cancers, absence of KRAS mutations and expression of pro apoptotic molecules: KRAS4A, p27^kip1 and cleaved caspase 3. Further work is needed to elucidate the biological significance of high TRAIL-R1 expression and better outcome, and to establish the association between TRAIL-R1 expression and response to therapy that targets this receptor. The biological effects of TRAIL in CRC models, its enhancement of chemosensitivity with standard chemotherapeutic agents and the effect of endogenous TRAIL receptor levels on survival make TRAIL an extremely attractive therapeutic target.