1. Volume 12, Issue 3, Pages 627-637 (September 2003)
Identification of Modulators of TRAIL-Induced Apoptosis via RNAi-Based Phenotypic Screening 
Pedro Aza-Blanc, Christopher L. Cooper, Klaus Wagner, Serge Batalov, Quinn L. Deveraux, Michael P. Cooke 
Molecular Cell 
Volume 12, Issue 3, Pages (September 2003)
DOI: /S (03)

2. Figure 1 Evaluation of siRNA Activity
(A) Evaluation of siRNA activity using quantitative PCR. Five siRNAs were tested for each of ten genes in HeLa cells. Depicted is the percentage inhibition of target RNA levels for each siRNA.
(B) Histogram of siRNA activity levels. Data from the above analysis are compiled and displayed with the number of oligonucleotides in each bin displayed. The red bars highlight oligos (31/50) that inhibited target mRNA by >70%.
(C) Correlation of GC content and position with siRNA activity. GC content in a three base window was calculated for each siRNA oligonucleotide and correlated with siRNA activity data. Plotted is the correlation coefficient and 95% confidence intervals of GC content with activity for each three base window from the 5′→3′ end of the siRNA. Note the correlation of activity with high GC content in the first four bases and low GC content in the last four bases. Significant correlations are marked. **, p < 0.01; *, p < 0.05.
Molecular Cell  , DOI: ( /S (03) )

3. Figure 2 siRNA-Based Screen for Trail Sensitivity
(A) Screening strategy. siRNA duplexes were spotted onto 384-well plates in duplicate and HeLa cells were reverse transfected onto the wells. Cells were incubated for 48 hr to allow target decay and treated with or without TRAIL. Viability was measured 20 hr after TRAIL treatment using Alamar Blue reagent. A sensitivity ratio was determined for each siRNA for comparison with a total of 60 values obtained with negative control siRNAs targeting luciferase.
(B) Viability distributions from Alamar Blue readings after transfecting the siRNA collection in the absence (blue line) or presence (red line) of TRAIL compared to negative control siRNAs targeting luciferase (hatched lines). Data correspond to average values obtained from two screens done in parallel.
(C) Histogram showing distribution sensitivity ratios derived from two parallel experiments across the siRNA collection and compared to negative controls siRNA targeting luciferase (siRNA collection, blue line; controls, blue hatched line).
Molecular Cell  , DOI: ( /S (03) )

4. Figure 3
Identification of TRAIL Enhancers and Inhibitors by siRNA Screening
(A) Pseudo-color representation of screening results. Sensitivity ratios (SR) obtained in screens 1 and 2 for each siRNA in the collection were ordered according to the average of the sensitivity ratios obtained in both screens. siRNAs which desensitize cells to TRAIL lie in the upper part (red). siRNAs that sensitized cells to TRAIL (green) are in the lower portion. On the left, apoptosis-related genes that were used to evaluate the screening performance. CASP8, p53, BID, TRAIL receptor DR4, APAF1, and CASP3 siRNAs showed the expected inhibition of TRAIL-induced apoptosis. FADD siRNA showed moderate inhibition. DR5 and CASP9 siRNAs showed no activity.
(B and C) List of the 20 top inhibitor and enhancer siRNAs. p value was determined by t test comparing values obtained for each siRNA in the two screens (four data points) with the control siRNA population (60 wells/screen). (B) Inhibitor siRNAs, targeting genes required for effective TRAIL-induced apoptosis. (C) Enhancer siRNAs, targeting genes that prevent TRAIL-induced apoptosis.
Molecular Cell  , DOI: ( /S (03) )

5. Figure 4 Confirmation of Selected Genes Detected in the Screen
Two additional siRNAs with independent sequence (siRNA 1 and siRNA 2) were designed for each target and used to confirm that the screen results are due to target inhibition. siRNA control used was siGL2. Experiments were performed in normal serum conditions.
(A) Effect of selected inhibitor siRNAs on TRAIL-dependent caspase-3/7 activation. Blue columns, no TRAIL treatment; red columns, treatment with 1 μg/ml TRAIL. Selected targets were GSK3α, the uncharacterized FLJ32312, and the signal recognition particle component SRP72. GSK3β siRNAs were also included. Values are normalized to caspase-3/7 activity detected for control (siGL2) in the presence of TRAIL (= 3). GSK3β siRNAs behaved as control siRNA and did not prevent caspase-3/7 activation.
(B) An antibody that detects both GSK3α and GSK3β was used to detect GSK3α and GSK3β levels in extracts from cells transfected with GSK3α and GSK3β siRNAs 1 and 2 (see Experimental Procedures). An extract from cells transfected with the GSK3α siRNA present in the screen (GSK3α S) was also included. All of them were efficient inhibitors of their respective targets.
(C) Effect of selected enhancer siRNAs on caspase activation by TRAIL. Selected targets were the semaphorin receptor PLXNB1, JNK inhibitory kinase (JIK), and the uncharacterized gene FLJ PAK1 was also included in the study as a kinase with known antiapoptotic activity. Effects on caspase activity in the absence of TRAIL (blue columns) or under 100 ng/ml TRAIL treatment (red columns) are normalized as in (B). In (A) and (C), asterisks indicate values that were significantly different from the control siGL2 (**, p < 0.01; *, p < 0.05).
Molecular Cell  , DOI: ( /S (03) )

6. Figure 5 Biochemical Mapping of Inhibitory siRNAs
(A) siRNAs against SRP72, GSK3α, and DOBI/FLJ32312 were transfected in parallel with a negative control (sigl2) and two positive controls (siCASP8 and siBID). Forty-eight hours later cells were treated with or without TRAIL (1 μg/ml), and Western analysis was performed with antibodies to detect caspase-8, BID, caspase-9, and caspase-3 cleavage. SRP72 is required for caspase-8 activation by TRAIL signaling. GSK3α showed a similar though weaker activity and might be acting at other levels. Inhibition of DOBI blocked caspase-9 activation, but it did not prevent Bid or caspase-3 cleavage.
(B) DOBI is required for the release of cytochrome c from the mitochondria. HeLa cells were transfected with siRNA-2 against DOBI or control siGL3. Forty-eight hours later, cells were treated in the presence or absence of TRAIL (1 μg/ml) followed by separation of soluble cytoplasmic and heavy membrane fractions. Samples were analyzed by Western blotting using an antibody against cytochrome c (upper panel) or β-actin (lower panel) as loading control.
(C) Schematic representation of the pathway and positions at which interaction of these genes is detected.
Molecular Cell  , DOI: ( /S (03) )

7. Figure 6
Correlation among Known Hits Detected in the Screen and Their Functional Relations
Well-established interactions among top enhancers and inhibitors identified in the screen highlight certain pathways as determinants of cell susceptibility to TRAIL-induced apoptosis. TRAIL-enhancer genes (corresponding to inhibitor siRNAs) are shown in red; TRAIL-inhibitor genes (corresponding to enhancer siRNAs) are shown in green. See text for details.
Molecular Cell  , DOI: ( /S (03) )