Colorectal cancer is the second leading cause of cancer related dead in the western world and represents a serious health concern. To put in perspective the magnitude of the problem posed by colorectal cancer it is important to highlight that approximately one in 17 EU citizens will develop malignant tumors in their colon or rectum in the course of their life. Patients diagnosed with early stage (I and II) tumors have good prognosis (5-year survival greater than 80%). However, the majority of patients have advanced disease (stage III or IV) at the moment of their initial diagnosis and the 5-year survival rates for these patients ranges from 40% to less than 5%. There is, therefore, great need to improve the treatment of these patients.

We use high throughput techniques to find new markers that when used alone or in combination with other makers, they can be used to discriminate between patients that have high and low probability of recurrence after treatment. We then follow up these experiments using in vitro and in vivo experiments to investigate the functional relevance of these new markers for colorectal cancer initiation and progression.

EPH receptors are the largest family of receptor tyrosine kinases (RTKs), proteins that play a crucial role in many biological processes such as embryonic development, cell proliferation and differentiation. The first member of the EPH family was identified and cloned in 1987 by Hirai et al. from an Erytropoietin Producing Hepatocellular carcinoma cell line (EPH). To date, 16 receptors (14 found in mammals) and 9 ligands (8 in mammals) have been described. EPH receptors and ephrins (ligands) are implicated in a great variety of processes such as regulation of cell proliferation, migration, cell attachment and shape, axon guidance and synaptic plasticity. EPH receptors play important roles in tumorigenesis and metastasis and high levels of EPH have been related to angiogenesis in many tumor types including breast and lung.

Although overexpression of EPHB2 is observed in some tumor types, in gastrointestinal cancers, low levels of EPHB2 expression have been reported and found to be significantly associated with advance disease stage and poor survival (Lugli et al., 2005). In colorectal carcinoma (CRC), a progressive reduction in EPHB2 levels has been reported in the progression from normal epithelial cells to benign adenomas and to low and high stage tumors as well as lymph node and liver metastases, demonstrating a clear tendency to decreasing EPHB2 levels as CRC progress towards a more aggressive and metastatic phenotype. The loss of EPHB2 expression was also significantly associated with poor tumor differentiation and shorter patient survival. We have made some contributions towards increasing our understanding of the mechanisms responsible of this EPHB2 down-regulation in CRC (Alazzouzi et al 2005 Cancer Res 65:10170; Davalos et al 2007 Oncogene 26:308).

Similarly, EPHB4 expression in cancer is up- or down-regulated depending on the tumor type.  A drastic increase of EPHB4 protein has been observed in endometrial hyperplasias and carcinomas, suggesting EPHB4 as an early indicator of malignant development.  Moreover, EPHB4 overexpression has been associated with high histological grade and certain clinical stages in endometrial cancer.  An increase in breast carcinoma has been also reported and high levels of EPHB4 correlated with histological grade and stage.  In addition, strategies to block EPHB4 expression, both using siRNA and antisense led to dose-dependent reduction in cell survival and increased apoptosis in breast.  In the normal colonic mucosa, we revealed a gradient of EPHB4 expression from the lower crypt to the colonic flat mucosa, and a substantial variability of EPHB4 expression in colorectal tumors from complete lack of immunoreactivity to very high levels of expression (Davalos et al 2006 Cancer Res 66:8943). Furthermore, our group has shown that low EPHB4 tumor levels identify a subset of colorectal cancer patients with poor prognosis and high risk of recurrence, and demonstrated that promoter hypermethylation was a common mechanism associated to the loss of EPHB4 expression. Moreover, reintroduction of EPHB4 into EPHB4-deficient tumor cells significantly reduced their long-term clonogenic potential, which taken together contributed to establish EPHB4 as a new putative tumor suppressor gene, and a useful prognostic marker in colorectal cancer (Davalos et al 2006 Cancer Res 66:8943).

A significant proportion of patients with colorectal cancer are hereditary. Hereditary Nonpolyposis Colorectal Cancer (HNPCC) and Familial Adenomatous Polyposis (FAP) are the two most common forms of hereditary predisposition to colorectal cancer. FAP is characterized by the presence of hundreds to thousands of adenomatous polyps in the colon and rectum of affected individuals and would lead to colorectal cancer in virtually all the patients if left untreated. An attenuated form of FAP (AFAP) is characterized by a reduced number of polyps compared to classical FAP and a later age of onset. Together, FAP and AFAP affect up to 1 in 5,000 individuals. Inherited mutation or deletion of one allele of the adenomatous polyposis coli (APC) gene is responsible for 80% of FAP or AFAP cases. A small proportion of the FAP/AFAP individuals that do not have germline mutations in APC carry homozygous mutations in the gene MutY homolog (MUTYH). However, the underlying genetic cause of FAP/AFAP is not known for approximately 20% of the affected families. Precise identification of the genetic cause of this condition has a profound impact on the management of FAP/AFAP family members, and there is therefore an acute need to identify new genetic abnormalities that could be responsible for a significant number of these adenomatous polyposis syndromes in APC/MUTYH mutation negative families. We are actively investigating new genetic and epigenetic causes of colorectal cancer predisposition, both to FAP and HNPCC.

RhoA is a member of the small GTPase family that regulates cytoskeletal remodeling, protein and lipid trafficking, transcriptional activation and cell growth. We have recently demonstrated that patients whose tumors have low levels of RhoA have significantly worse prognosis than patients with high RhoA tumor levels (Arango et al., 2005).

The membrane-bound small GTPase RAS was one of the first oncogenes to be identified (Sukumar et al., 1983). Activating mutations in KRAS have been found in more than one third of the human tumors of the colon and rectum (Oliveira et al., 2004), highlighting the importance of this protein in promoting tumor initiation and progression. The role of RhoA and other members of the small GTPase super-family on colorectal carcinogenesis has not been as extensively studied. In contrast to the high frequency of KRAS mutations in colorectal tumors, no mutations have been identified in RHOA in this tumor type (Arango et al., 2005; Rihet et al., 2001). However, RhoA signaling regulates an important signal transduction pathway linking plasma membrane receptors to the assembly of focal adhesions and actin stress fibers. This signaling cascade can regulate cell morphology, attachment to the substrate and motility, and this regulation is cell type dependent (Van Aelst & D'Souza-Schorey, 1997). In addition, RhoA signaling has been shown to regulate the expression and activity of multiple key members of the cell cycle machinery. Thus, inhibition of RhoA activity can lead to either cell cycle arrest or increased growth, depending on the cellular contest (Bellovin et al., 2005; Bellovin et al., 2006; Pille et al., 2005).

We are studying the molecular mechanism underlying our previous observation showing that low RhoA levels are associated with poor prognosis of colorectal cancer patients. For this purpose, we are using in vitro isogenic systems as well as animal studies and analysis of achieved materials from human tumor samples.