Using functional genetics to understand breast cancer biology

Abstract

Genetic screens were for long the prerogative of those that studied model organisms. The discovery in 2001 that gene silencing through RNA interference (RNAi) can also be brought about in mammalian cells paved the way for large scale loss-of-function genetic screens in higher organisms. In this article, we describe how functional genetic studies can help us understand the biology of breast cancer, how it can be used to identify novel targets for breast cancer therapy, and how it can help in the identification of those patients that are most likely to respond to a given therapy.

Figure 1.

High-throughput siRNA screening (Adapted from Iorns et al. 2008). (A) MCF7 cells plated in 96-well plates were transfected with siRNAs. Each plate contains 80 experimental siRNAs directed against a different gene (black) and 10 nontargeting siControl siRNAs. Transfected cells were divided and half treated with the drug tamoxifen and half with the drug vehicle. After 7 d cell viability was measured. (B) Reproducibility of the method. Correlation of the effect of each siRNA on viability, in vehicle treated cells, between two duplicate screens. The Spearman correlation coefficient r² was 0.71. (C) Scatter plot of averaged Z scores from tamoxifen resistance screen carried out in duplicate screens. Black diamonds are experimental siRNAs targeting 779 kinase genes and red diamonds are control siRNAs.

Figure 2.

RNAi bar-code screens: Identification of candidate drug response biomarkers. ShRNA bar code loss-of-function genetic screens. Collections of shRNA vectors are expressed polyclonally in drug-sensitive cells and subjected to drug selection. Cells harboring a shRNA vector that confers drug resistance will become enriched in the population, shRNAs that enhance the sensitivity to a cancer drug will become depleted under drug selection compared to a reference population that is not exposed to drug. Each shRNA vector contains a unique identifier sequence (the bar code), which can be recovered by PCR and its abundance quantified on a dedicated DNA micro-array containing the bar code sequences. shRNAs that cause drug resistance are enriched and appear “red” on the micro-array, depleted shRNAs appear “green.”

Figure 3.

Finding druggable genes in cancer relevant pathways. In a first step, all members of a druggable gene family are identified through bioinformatics approaches. Next, multiple siRNA or shRNA vectors are designed for each of the members of the gene family yielding a gene family knockdown library. This library can then be used to screen for their ability to modulate a cancer-relevant pathway, for instance by using a specific reporter gene construct that responds to pathway activity.