In light of mounting concerns regarding the potential toxicity of some nanomaterials, scientists have designed a rapid screening tool to help predict which ones are likely to be harmful. Hundreds of nanotechnology-based products are already on the market--in everything from sunscreens and cosmetics to paints and car bumpers--and many more are in the pipeline. However, studies assessing the safety of nanomaterials are limited. As a result, scientists and policy makers have been calling for more systematic reviews of the risks that these nanoscale materials might pose to human health.

Given the large diversity of engineered nanomaterials, which can vary in their chemical makeup, size, shape, and coating, assessing their toxicity has been a challenge. Studies in animals are expensive and time consuming, and although testing nanomaterials in cell cultures can yield useful information, different cell types can respond differently to the same nanomaterial.

"Nanomaterials are really complex, and if you just carry out one or two tests, you're going to miss something," says Andrew Maynard, chief science advisor to the Project on Emerging Nanotechnologies, at the Woodrow Wilson International Center for Scholars, in Washington, DC. What's more, results from experiments in cells often don't match those from animal studies.

To address these challenges, Stanley Shaw, a chemical biologist at Massachusetts General Hospital Center for Systems Biology, and his colleagues at the Broad Institute of Harvard and MIT designed a high-throughput screening method. Inspired by cancer studies in which scientists classify different types of cancer based on patterns of gene expression, Shaw and his colleagues sought to develop a tool that could screen large numbers of different nanomaterials and classify them based on their toxicity.

As a proof of concept, the researchers tested 50 different nanoparticles--mainly particles used for medical imaging. These included mostly iron-based particles, as well as several types of quantum dots. The particles also had various chemical coatings.

The researchers tested each of the nanoparticles in four different types of cells--immune cells from mice, two types of human blood-vessel cells, and human liver cells--and at four different dosages. To create the different combinations, a robotic system similar to that used for drug screening placed the nanoparticles inside tiny wells on a plate containing hundreds of separate wells. Each well contained one cell type. The screening system then detected changes in the cells' metabolism in response to the nanomaterial. Computer software analyzed the data, looking for relationships between the different particles.