Overexpression of CrkL as a novel biomarker for poor prognosis in gastric cancer

1. Introduction

Gastric cancer (GC), although declining in incidence in the last decades, is still the fifth most common malignancy and the third leading cause of cancer-related death worldwide [1]. This disease affects nearly one million people per year, particularly in areas of Eastern Asian, Western Europe, and Southern and Central America [2]. In spite of the routine use of endoscopy combined with biopsy for diagnosis and health screening, due to the relatively asymptomatic nature of this disease in its early stages, the majority of GC patients are still diagnosed at advanced stages which negatively affect patient’s outcomes [3].

Nowadays, radical gastrectomy with D2 lympha- denectomy remains the best treatment in patients with resectable tumors [4], unfortunately, most GC patients are diagnosed in an advanced and unresectable stage; therefore, the best treatment for these patients includes chemotherapy, target therapies, and best supportive care, however, the clinical outcomes of these patients remain unsatisfactory with a 5-year overall survival rate of less than 30% [5].

Since no internationally standardized scheme of treatment could be unequivocally applied to GC patients, the identification of novel prognostic biomarkers which may provide important information on the likely course of disease, the prognosis as well as the suggestions for further treatment is urgently needed with the aim to improve the clinical outcomes of GC patients.

Our previous study showed that the expression of CrkL protein was remarkably up-regulated in GC tumor tissue compare to that in matched adjacent non-tumor tissue in a small set of GC patients. Moreover, a series of function assays in vitro, utilizing small-interfering RNA, were carried out. These assays included proliferation assay, cell cycle analysis, and apoptosis detection. Our finding revealed that knockdown of CrkL in SGC-7901 GC cells induced a suppression of cell proliferation along with a significant arrest of cell cycle in G0/G1 phase. Furthermore, we validated that CrkL was a direct target of miR-126, which functions as a tumor suppressor in GC, and over-expression of miR-126 inhibited the protein expression of CrkL significantly. Therefore, CrkL may contribute to GC progression through regulating proliferation, and miR-126 may function as a tumor suppressor in GC with CrkL as its direct target [6]. However, the prognostic value of CrkL protein in GC patients is yet to be elucidated. The aim of current study is thus to further analyze the relationship between tissue-based CrkL protein level and clinicopathological parameters, and its prognostic value in a larger set of GC patients. We expected that CrkL might become a novel prognostic biomarker of GC in the near future.

2. Materials and methods

3. Results

GC is a leading cause of cancer-related mortality, and a major health problem worldwide [7]. Despite improvements in surgical and adjuvant treatment approaches, the clinical outcome for patients with GC remains worse than that for patients with other common malignancies such as colon, breast, and prostate cancer. The TNM staging (tumor, lymph nodes and metastasis) is the most important and widely used tool to predict prognosis of GC, and the main therapeutic reference and guidelines in current clinical practice as well [8]. However, the fact that prognosis often varies between GC patients even at the same TNM staging manifests clearly that there must be molecular factors responsible for these clinical differences [9].

It is generally believed that GC results from the combined effects of environmental factors and genetic alterations, including tumor suppressor genes, oncogenes, growth factors, and cell adhesion molecules [10]. Although previous studies have reported that multiple aberrantly expressed genes could contribute to the malignant behavior of GC cells, biomarkers that are able to identify tumor progression and predict the aggressive phenotype are still lacking in GC. Thus, identifying novel prognostic biomarkers of GC in addition to TNM staging will be definitely helpful to refine the accuracy of prognostic stratification, and henceforth design more reasonable treatment and surveillance strategy for this devastating illness [11].

The signaling adaptor protein Crk was originally isolated as an oncogene fusion product of an avian sarcoma CT10 retrovirus (v-Crk) [12]. Its mammalian homologs including Crk I, Crk II and CrkL proteins are the first identified adaptor proteins which connect with upstream proteins through their SH2 (Src homology 2) domain and with downstream proteins through their NH2-terminal SH3 (Src homoloyg 3) domain [13]. Crk proteins and their effectors are highly conserved throughout evolution. Several binding partners have been identified for the SH2 and SH3 domains of Crk proteins so far. The SH2 domain binds a specific phosphorylated tyrosine motif found in proteins involved in cell spreading, actin reorganization, and cell migration. These SH2 domain binding partners include p130Cas and paxillin [14, 15], which are effectors downstream of integrin activation, as well as the Gab1 and Gab2, which are effectors downstream from many tyrosine kinases [16]. The NH2-terminal SH3 domain binds proline-rich motifs present within several proteins which include Abl tyrosine kinase, Arg, C3G (a nucleotide exchange factor for Rap1), DOCK180 (an exchange factor for Rac1), etc. [17, 18, 19, 20]. However, the binding proteins for the COOH-terminal SH3 domain are still poorly understood. It’s now clear that Crk proteins play central roles in signaling cascades that regulate a variety of cellular and biological processes such as proliferation, differentiation, migration and apoptosis through i) protein complex formation downstream from receptor activation, ii) regulation of cellular tyrosine kinase activity, and iii) upstream roles in signal initiation [21].

Although early studies focused on Crk proteins in the transformation of fibroblasts [22], the studies of Crk proteins in epithelial cell models has further elucidated the role of Crk proteins in the amplification and/or initiation of oncogenic signals, suggesting that Crk proteins dysregulation can cooperate with a variety of oncogenic signals to promote tumorigenesis. For example, overexpression of Crk II and CrkL in breast cancer epithelial cells enhances epithelial-to-mesenchymal transitions (EMT), characterized by the breakdown of adherens junctions, enhanced cell dispersal, and remodeling of cortical actin [23]. Whereas, siRNA-mediated knockdown of Crk significantly reduced invasion and migration of breast cancer cells [24], implicating Crk proteins as critical signal transducers for cell invasion in breast cancer. Similar conclusions have been demonstrated in lung, oral, head, prostate, and ovarian cancer. For example, Crk II knockdown could down-regulate IGF-IR levels and suppressed PI3K/Akt signaling pathway in prostate cancer cells, and further inhibited their migration and invasion [25].

The role for Crk proteins in promoting gastric tumorigenesis is suggested by multiple studies. The phosphorylated CagA binds Crk proteins (Crk I, Crk II, and CrkL), and that interaction is important for the CagA-mediated host responses during H. pylori infection which may further cause GC [26]. Crk genetic variants can be susceptible genetic factors for the development of GC by controlling signals through the CagA transduction pathways, and an interaction between Crk genetic polymorphism and phytoestrogen biomarkers appear to play a role as risk modifiers in gastric carcinogenesis [27]. Knockdown of CrkL abrogates CCL20-induced Erk1/2 signaling and the progression of EMT which plays an important role in the initiation of pathogenesis of normal epithelial cells, migration, invasion as well as metastasis in GC [28]. Of note, miR-126 which was found to be down-regulated and function as a tumor suppressor in a variety of cancers also affects the proliferation and metastatic potential of GC cells, partly through regulating Crk I, Crk II and CrkL in our previous studies [6, 29].

Pathophysiologically, Crk proteins are overex- pressed in a wide array of human cancers, including but not limited to ovarian cancer, lung cancer, and breast cancer; whereby the level of expression often correlates with the tumor aggressiveness, disease progression, and inversely segregates with prognosis, indicating that Crk proteins may be used as potential prognostic biomarkers. For example, Miller et al. demonstrated that CRK I/II mRNA expression in lung cancer was increased in more advanced, larger and poorly differentiated tumors and was associated with poor survival [30]. Wang et al. showed that overexpression of CrkL correlated with poor prognosis in lung cancer patients [31]. Analogous to overexpression of Crk proteins in aggressive lung cancer, analysis of CrkL in ovarian cancer also revealed strong correlations between CrkL expression and poor prognosis [32]. CrkL overexpression was also found to be a significant, independent poor prognostic indicator for hepatocellular carcinoma. Results from IHC assay indicated that the high expressions of CrkL strongly correlated with reduced disease-free and overall survival of hepatocellular carcinoma patients [33]. Glioblastoma patients with high Crk II and Crk phosphoY251 expression had distinctly lower overall survival rates than those with low Crk II and Crk phosphoY251 expression [34]. In addition, our previous study demonstrated clearly that up-regulation of Crk I/II proteins was associated with poor prognosis of GC patients [35].

Despite growing evidence highlighting its pivotal roles in the progression of GC, no study has systematically explored the prognostic value of CrkL in GC. In the current study, we evaluated the clinical and prognostic significance of CrkL expression in GC using immunohistochemistry. We observed the overexpression of CrkL in GC tumor tissues comparing to that in adjacent non-tumor tissues. We further examined the correlation between CrkL level and the progression and prognosis of GC. Our results demonstrated that the overexpression level CrkL was associated with various clinicopathological factors related with poor prognosis, such as tumor size, local invasion, lymph nodes metastasis and TNM staging, suggesting that CrkL expression may influence biological features of GC and might play a pivotal role in tumorigenesis and development of GC. Moreover, GC patients with higher CrkL level had worse prognosis compared with those with lower level. Cox proportional hazards regression model analysis revealed that high expression level of CrkL was an independent factor predicting poor outcome of GC patients, which was in accordance with the results of lung, ovarian, and liver cancer. In this respect, CrkL might possess the same potential in prognosis prediction of GC as found with the classical prognostic factors such as lymph node metastasis or TNM stage. To the best of our knowledge, this is the first study to investigate the clinical significance of CrkL expression in a large set of GC patients.

In conclusion, our data offer the convincing evidence for the first time that the CrkL expression was markedly and consistently up-regulated in GC patients, and high CrkL level was associated with poor prognosis. CrkL thus has potential clinical utility as a promising biomarker for molecularly monitoring the progression and predicting the clinical outcome of patients with GC. This discovery may help to better identify GC patients with poor prognostic potentials and facilitate personalized treatment of GC patients.