Genomic shRNA screen (University)
The aim of this project was to define genes specifically essential for cancer cells, the hypothesis being that corresponding proteins are candidate drug targets. The genomic screen, the main part of the project at the University side, could be executed remarkably well. A robotic system was purchased and installed. Preliminary quality control experiments had to be performed which documented that the approach was feasible. Then, a commercial genome-wide lentiviral shRNA was tested consisting of >80'000 single viruses, each gene was targeted by at least 5 library shRNAs, and over 13'000 genes were interrogated. Our specific assays allowed to state whether a gene is essential for tumor cells. We could define a first group of  800 essential genes, where downregulation inhbited cell proliferation >85%. This gene set identifyed was analyzed by bioinformatics (gene enrichment analysis) using commercial  Ingenuity software, and genes enriched for specific categries and pathways with very high statistical significance emerged. The screening hits were also analyzed in silico using commercial Ingenuity software (gene enrichment analysis). Importantly, we saw that our screening system had enriched for genes involved in signaling, cancer, inflammation, and mitochondrial functions. Also, the analysis indicated that many known targets of established anticancer drugs, such als rapamycin, sorafenib and others were enriched. This gives us confidence that some of the identified targets are similarly valuable candidates for drug development.

In follow up experiments concentrating on 250 genes from these 800 genes using other mTOR-addicted cells from our isogenic cell panel, we could narrow the list to a core set of 54 genes essential for mTOR addicted cells. Interestingly, the core set of genes contains the TOR kinase gene itself (so to speak as inbuilt positive control),  other enzymes, G proteins, many signaling elements and genes involved in metabolism.These genes represent potential drug targets.  A fraction of the corresponding gene products, mainly enzymes,  are "druggable".

Validation of hits in human cancer. Genes identified as "essential" in the murine model system were then tested to see whether they are also essential in human cancer lines from lung, breast, melanomas and other tumors. We used a cell panel of 14 lines, 6 of them from one type of a non-treatable tumor (melanoma). As control, we used normal fibroblasts. For this part of the project, we purchased human orthologs of the murine genes as siRNA and used siRNA transfection technology. We identified a sizable group of genes which were essential in more than one tumor, but non-essential in normal cells. Of particular interest are 6 genes which were essential in 2 or 3 out of six melanoma lines, but not in normal fibroblasts  We predict therefore that potential inhibitors targeting the corresponding proteins should be selective for certain cancer cells.