Migration patterns of individual cells showing accelerated migration determined by time-lapse imaging.

From Nature Cell Biol. 10 August 2008; doi:10.1038/ncb1762.

siRNA technology enables the genetic screening of mammalian cells. However, the most common strategy used by researchers involves minimally validating the vast majority of primary hits and only selectively analysing a few strong hits. In Nature Cell Biology, Joan Brugge, Kaylene Simpson and colleagues now report the use of a high-throughput siRNA screening approach to generate a highly validated dataset of genes that regulate epithelial cell migration.

The screen was based on a classical scratch-wound assay using a mammary epithelial cell line (MCF-10A). Three siRNA libraries were used — a phosphatase library, a kinase library, and a custom migration and adhesion related (MAR) library — which together account for 1,081 human genes. The MAR library was designed to target gene families that encode proteins either known or predicted to affect migration or adhesion.

In a primary screen, siRNAs were classified into hit categories based on migration-specific phenotypes — accelerated migration, impaired migration and low levels of Alamar Blue (the Alamar Blue reagent measures a cell's total reducing activity and is used to assay for non-specific effects such as proliferation, changes in metabolism or cell detachment). The hits were then extensively validated by re-testing the effects of individual siRNAs, analysing potential off-target effects and correlating the knockdown efficiency with the observed phenotypes. All positive hits were then re-evaluated in a different cell line. These screens yielded 66 high-confidence genes that either accelerated or impaired cell migration, which were then studied in greater detail.

The authors used static images and time-lapse video imaging to analyse changes in cell migration during wound healing. This included any changes in cell morphology, lamellae and lamellipodia protrusion dynamics, speed, direction and the regulation of cell adhesions. The phenotypes were classified into groups according to these parameters and the migration patterns of individual cells from each group were then further characterized. The authors observed that knocking down genes that increased cell migration significantly altered cell-cell adhesion.

Interestingly, 42 of the 66 high-confidence genes had not been previously associated with motility or adhesion. Bioinformatics analysis aimed at identifying pathways linking these genes highlighted three main signalling nodes — -catenin, 1-integrin and actin — and expansion of these nodes identified extensive networks.

This screen provides a resource of high-confidence data that greatly aids the understanding of the mechanisms of cell migration regulation. The fully interactive database can be found in the CMC Activity Center pages (http://www.cellmigration.org/resource/discovery/brugge/simpson2008_rnai.cgi).

Author: Kim Baumann