The Signal Transduction Laboratory is interested in the mechanisms by which regulatory signals affecting the growth and survival of cells are transmitted from receptors at the cell surface to target enzyme systems within the cell. We study both the normal function of these signalling pathways and also the defects in their regulation found in cancerous cells.
Investigation of mechanisms of transformation by RAS oncogenes
Much of the work in the laboratory has focused on the RAS family of oncogenes and the signalling pathways that they control. RAS genes are activated by point mutation in some 20% of human tumours and are known to play a key role in the establishment of the transformed phenotype. While the early signalling pathways activated by RAS are now well characterised and the transcriptional programmes they induce have been documented in detail using microarray technology (e.g. Schulze et al., 2001; Genes & Development; 15: 981 and Schulze et al., 2004; Mol. Biol. Cell 15: 3450), it remains a major challenge to understand later events in oncogene-induced signalling and, in particular, what proteins are important in the establishment and maintenance of the transformed phenotype and may therefore act as potential therapeutic targets for cancer treatment. In order to investigate novel aspects of these pathways in cancer cells, especially those with activated RAS oncogenes, we have employed a functional genomics approach using post-transcriptional gene silencing by genome-scale libraries of RNA interference agents.
Two approaches to screening have been applied. In one, genes corresponding to a large fraction of the genome are systematically silenced one by one, allowing identification of genes required for a particular aspect of the transformed phenotype. An example of a successful high throughput screen from the lab (Swanton et al., 2007; Cancer Cell 11: 498; Swanton et al., 2009; PNAS 106, 8671) studied the effect of knock down of expression of different genes on the sensitivity of cancer cells with activated RAS oncogenes to four common chemotherapeutic agents. This has lead to the identification of proteins that might be potential therapeutic targets for overcoming resistance of tumours to existing drugs. In particular, targeting a ceramide transport protein, COL4A3BP/CERT, results in multidrug sensitization in several tumour lines, with less effect on untransformed cells, suggesting that it could be a promising chemosensitization target. We have also used a similar genome-wide approach to identify genetic determinants of sensitivity and resistance of tumour cells to targeted therapies, including the EGFR receptor tyrosine kinase inhibitor erlotinib, which is a licensed treatment for lung cancer, and the mTOR inhibitors everolimus, a rapamycin analogue that is a licensed treatment for renal cancer, and PP242, an experimental inhibitor of mTOR kinase activity (Treins et al., Oncogene 2009, in press; Cully et al., Mol. Cell. Biol. 2009. in press).
In addition, a whole genome-scale screen has been carried out searching for synthetic lethal interactions between gene silencing and activation of the RAS oncogene. This compared a colon cancer cell line containing an activated KRAS allele with a normal derivative in which this has been deleted by homologous recombination, with the top fifty hits then being tested on a panel of thirty or so cancer cell lines, half of which were mutant and half wild type for KRAS. This has uncovered proteins whose therapeutic targeting might be expected to provide differential toxicity towards KRAS mutant tumour cells, with several being broadly applicable to cells derived from lung and colon cancers. The relationship between synthetic lethality with RAS oncogene activation and oncogene addiction to activated RAS is also being studied with hits from this screen. A drug library screen carried out at the same time identified drugs that are selectively toxic for KRAS mutant cells, some of which target proteins identified in the RNA interference screen. Potential RAS selective therapeutic targets being investigated from this work include the matrix metalloproteinase MMP7, proteasome components and topoisomerases.
In a second approach to genome scale screening, viral RNA interference vectors are used to silence many genes at the same time in a mixed pool of cells, with cells only surviving that acquire the desired phenotype as a result of knock down of a specific gene. These cells, along with the RNAi sequence they carry, are then identified as they emerge at the end of the screen. An example of this approach was published previously (Nicke et al., 2005; Molecular Cell 20: 673) in which we identified MINK, a MAP4 kinase acting in the p38 stress activated protein kinase pathway, as a protein required for RAS oncogene induced senescence in ovarian epithelial cells. A novel adaptation of this methodology allows the identification of shRNA sequences lost from a population of cells under selective pressure by the use of high throughput sequencing of barcodes in the shRNA library to give a digital readout of library sequence representation. This methodology is being used currently in the lab to address the mechanisms used by tumour cells to grow in the absence of adhesion to extracellular matrix.
The role of phosphatidylinositol 3-kinase in RAS-driven oncogenesis
RAS proteins signal through direct interaction with a number of effector enzymes, including type I phosphatidylinositol (PI) 3-kinases. Although the ability of RAS to control PI 3-kinase has been well established in manipulated cell culture models, evidence for a role of the interaction of endogenous RAS with PI 3-kinase in normal and malignant cell growth in vivo has been lacking. We have generated mice with mutations in the Pi3kca gene encoding the PI 3- kinase catalytic p110¿ isoform that block its ability to interact with RAS (Gupta et al., 2007; Cell; 129: 957). Cells from these mice show proliferative defects and selective disruption of signaling from certain growth factors to PI 3-kinase. The mice also display defective development of the lymphatic vasculature due to reduced signalling from VEGF-C to PI 3-kinase. Most importantly, the mice are highly resistant to endogenous KRAS oncogene induced lung tumourigenesis and HRAS oncogene induced skin carcinogenesis. The interaction of RAS with p110α¿is thus required in vivo for certain normal growth factor signaling and for RAS-driven tumour formation. The demonstration of the importance of the RAS/PI 3-kinase interaction in tumourigenesis raises the prospect that agents that disrupt this interaction might have particular value in cancer therapy.
This work is being further pursued by the generation of mice with inducible expression of the inactivating mutation in the RAS binding domain of p110α¿so that the requirement of this interaction for tumour maintenance, rather than simply tumour initiation and development, can be assessed. In addition, the effect of this mutation in PI 3-kinase on tumourigenesis driven by other oncogenes acting upstream of RAS, such as EGF receptor, is also being studied. Furthermore, we have also created a mouse with an inactivating mutation in the RAS binding domain of p110α the other ubiquitously expressed PI 3-kinase catalytic subunit isoform, and will test the effects of this mutation on tumour initiation and maintenance.
Figure 1. Genes whose knock down is selectively harmful to RAS mutant relative to RAS wild type cells. siRNAs targeting the top 52 genes idenitified as selectively toxic to HCT-116 colon cancer cells relative to a KRAS deleted derivative were tested on a panel of 14 KRAS mutant and 14 KRAS wild-type cell lines from various cancer types. Unpaired t-test analysis of viability data is shown, identifying genes whose knockdown has a differential effect in KRAS mutant (blue) versus KRAS wild-type (red) cells across the whole panel. Significant differences (p>0.05) are found for a total of ten genes, including KRAS.
For a list of refereed research papers, see Publications (in navigation on left).
