Studying the function of the alternative spliced isoform Rac1b in colorectal cancer
Gudiño, Maria Victoria
Colorectal cancer (CRC) is a leading cause of cancer-related deaths, being the secondary most lethal cancer type in UK and the third worldwide. In the recent years, detection methods and therapies have improved overall survival and life quality of patients diagnosed with CRC but it is still a major health problem. There are a few signalling pathways whose alteration are hallmarks for CRC such as the WNT signalling pathway, RAS, TGFβ and the TP53 gene related signalling. For instance, the Apc gene, the core negative regulator of the WNT signalling pathway, is mutated in over 85% of total sporadic CRC cases and it is considered an initiating event for tumour formation. Targeting any of these pathways is the paradigm for conventional anticancer treatments. However, off-target effects or the development of resistance urge for additional treatments. Using multidrug therapies to tackle resilient cells or drugs affecting downstream pathway activation could be promising strategies to overcome treatment failure. Many molecular processes are altered during cancer evolution in order to favour tumour survival. One that has caught the attention of cancer researchers over the past decades is alternative splicing. Alternative-produced isoforms of proteins have been observed overexpressed in a wide spectrum of tumour types compared to non-tumoral healthy tissues and oncogenic properties have been attributed to many of them. RAC1B is an example of an alternative spliced isoform, which was firstly discovered in colorectal cancer cells. Since then, it has been associated with tumour-related functions. Its canonical protein product is RAC1, a small Rho GTPase protein which cycles from active to inactive states and drives cell migration and invadopodia formation, among other functions. Interestingly, Rac1 signalling, but not its expression, has been detected overactivated in tumours, suggesting an oncogenic role for those proteins that promote its GTP-bound state. RAC1B is the result of the insertion of 19 amino acids within the switch II domain, which is responsible for the conformational change during its cycling from GTP to GDP. Consequently, RAC1B remains in a GTP bound state and it is considered a constitutive active version of RAC1. Since it was first discovered in breast and colon cancer cells, it has been assigned a tumour-related role. However, little is known about its downstream activated pathways and its function in both homeostasis and cancer. The goal of the present thesis was to describe RAC1B role in the intestine, either within a neoplastic context or during normal conditions. Besides, I aimed to characterise RAC1B relevance in tumour formation and to elucidate its potential as a therapeutic target. Initially, RAC1B tumour expression was studied through analysis of The Cancer Genome Atlas (TCGA) dataset. Interestingly, it was detected highly expressed in many cancer types such as lung, pancreas, thyroid and gastric-related cancers. Moreover, patients with high levels of RAC1B have a worse overall survival and disease-free survival. CRC can be divided into four different subtypes, the CMS subtypes. Curiously, even though RAC1B correlated with the canonical/WNT-related subtype, which has one of the best prognosis (CMS2), patients within this group with high RAC1B expression had a survival rate comparable to those with the most malignant subtype, CMS4. Therefore, these data indicate that RAC1B expression correlates with a worse overall prognosis for CRC patients. I then studied both its tumour promoter role and its requirement to drive tumorigenesis by modulating its expression in tamoxifen-inducible mouse models, which emulated different stages across tumour evolution. Other than survival and tumorigenesis data, intestinal tissue was used for histologic and molecular analysis, as well as for extraction of intestinal stem cells to culture in a 3D organoid system for following functional analysis. A benign adenocarcinoma model driven by heterozygous loss of Apc (APC) and an invasive tumour model with Apc and TP53 co-deletion (APC P53) showed the most interesting results and provided most information about Rac1b-related signalling pathways. On one side, Rac1b depletion in the APC model prolonged mice survival by decreasing tumorigenesis. Phenotypically, these tumours were less proliferative and RNAseq analysis demonstrated a strong WNT-driven phenotype, since most of its target genes were depleted in the co-deleted group. Moreover, a protein interactome analysis by BioID showed a large interactive network, with EGFR and ErbB2 receptors and other RTK and proteins involved in the EGFR signalling pathway. Functional analysis using a colorectal 2D cell line validated its interaction and presented a decreased EGFR activation when Rac1b was deleted. Interestingly, APC-derived organoids with Rac1b deletion were more sensitive to the treatment with an EGFR inhibitor drug than organoids with wild type Rac1b expression, indicating a synergistic effect. Moreover, tumours from the aging cohort also presented a downregulation in two transcriptional targets of the ERK pathway, suggesting that depletion of Rac1b was modulating both WNT and EGFR signalling pathways to diminish tumorigenesis. On the other side, the opposite effect was observed when Rac1b was overexpressed. The APC P53 model was used in order to model tumour invasion, although these mice presented a rapid tumour initiation phenotype rather than long-developed invasive tumours. RAC1B overexpression drove an increased tumorigenesis in the small intestine which led to a significant decrease in mice survival. Promotion of tumorigenesis was also observed when RAC1B was overexpressed in a pre-tumorigenic APC P53 model, which developed significantly more micro lesions or early adenomas compared to the control mice. Intriguingly, these microadenomas presented a marked decrease in the percentage of apoptosis. Transcriptome and functional analysis on the organoids derived from these mice revealed a decreased TGFβ-induced apoptosis, potentially through downregulation of its pro apoptotic target gene Bim and its signal transductor SMAD4. Histoscore quantification of endpoint aging tumours stained for SMAD4 antibody demonstrated a significant reduction in immune positivity in the group with RAC1B overexpression, pointing to SMAD4 as a potential Rac1b downstream mediator to regulate cell death and TGFβ tumour suppressor role. Altogether, this thesis provides evidence for a tumour promoter role of Rac1b. Given that its expression positively correlates with a worse patient prognosis, therapeutic targeting of RAC1B in combination with other anticancer treatments could efficiently amplify their effect. Nonetheless, cellular context and tumour stage might vary its downstream function, highlighting the need of studying the tumour mutational landscape prior therapy.