RNA interference (RNAi) is proving to be a powerful tool in cancer research for determining the potential function of any gene whose sequence is known. We have used an RNAi approach for genes involved in cancer cell proliferation, with the ultimate goal of generating functional blocking therapeutic antibodies. Our screening process was designed to minimize the frequency of false positives due to off-target effects and includes: (i) using multiple siRNAs to each target gene, (ii) monitoring the efficacy of each siRNA by measuring the extent of mRNA reduction by Taqman and/or protein downmodulation by flow cytometry, and, (iii) when possible, including in our functional assays, a number of cell lines which express, as well as cell lines that do not express the target gene. These critical steps allow us to detect off-target effects in two different ways: (i) siRNAs that affect proliferation in cell lines that do not express the target gene or (ii) siRNAs that downmodulate protein and/or mRNA levels equivalently but do not have the same effect on cell proliferation. siRNAs which are identified as having off-target effects are eliminated from further evaluation. Using these criteria, we have found that the rate of off-target effects is much higher than has been previously reported. 47% (150/317) of the siRNAs tested have measurable off-target effects. Using multiple siRNAs to the same target gene greatly reduces the impact of this problem. For target genes where we have tested 3 different siRNAs, 87% (54/62) have at least one siRNA that does not produce detectable off-target effects. The use of multiple siRNAs per target gene has also proven useful to ensure that at least one siRNA is effective (at least 50% reduction in protein and/or mRNA levels). Overall, 70% (100/143) of siRNAs tested were effective at reducing protein levels, and 76% (104/137) of siRNAs reduced mRNA levels. However, when we assessed the effectiveness of 3 siRNAs per target gene, 86% (19/22) and 94% (30/32) of those target genes had at least one effective siRNA, as tested by protein and mRNA levels, respectively. Thus, including multiple controls in the RNAi screening process greatly reduces the rate of both false positives and false negatives. Using this strategy, to date, we have screened 97 target genes and have identified 7 functional genes in our assay that previously had not been associated with cancer cell proliferation.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]