Figure 2

Identification of hit genes from primary screens with the 27 K Pathway Decipher shRNA library. (a) Scatter plot of gene targets statistically over-represented in anti-FAS Ab- and/or TNF-treated cells (three replicates each; y-axis) relative to median control value (DR-untreated replicates, n=3; x-axis) at the end of the culture period from primary screens. Data is visualized in the scatter plot on a log₂ scale. The middle diagonal line is the identity line, or the x=y line. The other two lines delineate genes with at least a twofold change in the intensity value in one of the data sets. Enriched genes (>2-fold, P<0.05) are indicated in red, and those that were enriched in both FAS and TNF screens are highlighted in black. To detect hits in the primary screens, P-values for each shRNA were calculated using a modified Rank-product test and by comparing the phenotypes produced by shRNAs targeting each gene with the phenotypes produced by negative-control shRNAs. (b) Pseudo-color representation of screen results. Apoptosis-related genes were used to evaluate assay performance. Internal library shRNA controls targeting genes known to be involved in DR-mediated apoptosis (positive controls) or the luciferase gene (negative controls) were used for data normalization. FAS, FADD, CASP8 and BID shRNAs showed the expected inhibition of FASL-induced apoptosis. (c) Venn diagram summarizing the number of genes statistically over-represented in FASL- and TNF-induced apoptosis shRNA screens and their relation to each other as well as known drug targets. (d) Graphical depiction of enriched pathway modules obtained by edge-betweeness network clustering for the altered genes from FAS and TNF shRNA screens FI networks. Each node is a pathway and nodes are linked by directed edges representing parent-to-child relationships. Node size is proportional to the number of proteins annotated within that term. (e) In vitro validation of genes identified as novel candidate apoptosis modulators using transduction of individual shRNAs into HeLa cells followed by FASL or TNF treatment. For validation experiments, we generated 200 unique lentiviral constructs containing 200 different shRNAS corresponding to 100 candidate genes selected in the primary screen (two shRNAs per gene) (Supplementary Figure 2). These were tested as an arrayed shRNA sub-library; representative shRNA hits are shown on the graph. shRNAs were confirmed as hits if they increased survival of FASL- or TNF-treated cells by at least twofold relative to controls expressing shRNA targeting luciferase (LUC). shRNAs against canonical members of the FAS and TNF pathways isolated in the screen served as positive controls in this assay. The bars indicate relative survival of cells (%) versus untreated control cells