Targeting Translation with Small Molecule Inhibitors. A major checkpoint controlling translation in eukaryotes involves mRNA selection for ribosome binding by eukaryotic initiation factor (eIF)4F (Fig. 1) - a node that is usurped in many human cancers. Although increasing eIF4F levels only modestly augments global translation, the expression of a limited set of mRNAs is dramatically altered. Indeed, analysis of transcripts that are differentially translated due to altered expression of eIF4E suggests that eIF4F can simultaneously exert effects on many different physiological pathways, including cell death and cell cycle programs. Currently, we are actively probing the ribosome recruitment phase of translation initiation as an emerging chemotherapeutic target.

Our lab has designed 3 high throughput screens (HTS) which have been used to prosecute over 750,000 compounds in search of chemical inhibitors that specifically target the eIF4F complex. As a consequence of this effort, we have identified and characterized several novel translation initiation and elongation inhibitors. We have shown that several of these have promising in vivo activity as anti-neoplastic and chemosensitization agents (Fig. 1) and current efforts are aimed at better understanding the mechanism of action of these unique compounds.


Synthetic Lethal RNAi Screens. It is not difficult to find small organic molecules that can kill cancer cells. Indeed, 0.1-1% of compounds in typical collections will kill at concentrations used in high-throughput assays. The challenge is to find molecules that will kill transformed cells while sparing normal ones. One approach to doing this lies in identifying drugs that exert selective toxicity towards transformed cells since many oncogenic mutations will alter the requirement (qualitatively or quantitatively) for particular biochemical activities. This concept is not new and has been well described from studies of synthetic lethal interactions in model organisms (yeast and Drosophila). Two genes are synthetic lethal if mutation in either gene alone is compatible with viability but where simultaneous mutation of both is lethal. Although synthetic lethal interactions are often viewed in the context of gene loss, they can also occur when one or both genes have acquired gain-of-function mutations. Exploiting this concept to treat cancer cells is therefore very attractive since it provides a conceptual framework for developing drugs with higher selectivity towards transformed cells. Synthetic lethal interactions are thought to be common in biology [perhaps on the order of 10 interactions per gene] but are difficult to predict given our limited understanding of metazoan pathway networks. To best explore synthetic lethal interactions in cancer, unbiased genetic screening approaches are required. One such approach has been to use chemical compound libraries to identify drug leads that preferentially kill cells with cancer-predisposing mutations relative to isogenic cells lacking these mutations. Another approach is to undertake RNA Interference (RNAi) and CRISPR/Cas9-based genetic screens. Our laboratory has pursued several synthetic lethal screens using custom designed, focused shRNA libraries that target the translatome. We are currently unravelling the underlying biology of several “hits” from these powerful genetic screens.


Tractable Mouse Cancer Models. Our laboratory uses a unique tractable mouse model, called the Eu-Myc mouse, to explore oncogene cooperation, synthetic genetic relationships, and drug action ex vivo and in vivo. In this model, c-Myc oncogene expression is ectopically driven in pre-B cells by the Ig heavy chain enhancer (Eu) leading to lymphoma development in mice 4-6 months after birth. This latency allows one to undertake genetic screens for novel oncogenes and tumor suppressor genes that co-operate with Myc to accelerate tumorigenesis in vivo. There are several attractive features of this model that make it well suited for undertaking genetic screens. (I) Eu-Myc lymphomas recapitulate typical genetic and pathological features of human non-Hodgkin’s lymphomas; (II) tumor burden is easily monitored by lymph-node palpation or blood smears; (III) lymphomas are detectable long before the animal dies; (IV) large numbers of pure tumor cells can be isolated from enlarged lymph nodes for ex vivo studies; (V) genetically engineered tumor cells with defined mutations can be generated; and (VI) therapy is performed in immunocompetent mice to study drug performance and target behavior in vivo.

            Recently we have modeled drug response in the mouse by engineering tissue-specific inducible RNAi (Fig. 2). We extended current inducible RNAi technology and generated novel models that exhibit inducible and tissue-restricted suppression of eIF4E. These mice show reversible inhibition of eIF4E at the organismal level and are being used to model drug responses in vivo. This enabling technology allows us to assess both the consequences of eIF4E suppression on in vivo tumor initiation and maintenance as well as uncover potential side-effects of systemic eIF4E suppression. 



Developing Tools for Precise Genome Engineering. Genome engineering using custom-built DNA targeting domains to recruit nuclease activity to specific chromosomal sites has tremendous potential for basic research and clinical applications. The current most powerfull approach for undertaking this is CRISPR (Clustered Regularly InterSpaced Short Palindromic Repeats), a technology that has its roots in a bacterial RNA-mediated adaptive defense system. In geneeral, as RNA guide strand of ~20 nucleotides (called the “trigger”) is used to direct the Cas9 endonuclease to a specific DNA target sequence to create a double stranded cut (Fig. 3). The use of base complementarity to target the Cas9 nuclease (via the sgRNA) to a specific genomic address indicates that virtually any desired site in the genome can be easily targeted. We have repurposed this technology for genetic screening purposes and are applying it to probe and dissect translational control processes in normal and tumor cells. 

Figure 3