Before any drug can enter clinical trials in humans, the US Food and Drug Administration (FDA) requires that developers check to see if it is genotoxic - if it causes DNA in the nucleus of a cell to break.

"Genotoxicity, as a side effect, is of particular concern because it can lead to mutations, and, through these mutations, can yield heritable defects and cancer," Joe Shuga, who designed the test, told DrugResearcher.com. Although most commonly associated with radiotherapy and chemotherapy, these side effects can also be caused by some external factors, such as ionising radiation.

Typically, genotoxicty levels are tested by injecting the potential drug into live mice and then, after 24 to 48 hours, looking for damaged red blood cells formed in the bone marrow of the mice. However, at the Whitehead Institute for Biomedical Research and Massachusetts Institute of Technology, US, scientists have now developed a cell culture alternative to traditional methods.

Rather than relying on live mice, Shuga developed an assay based on cells extracted from mouse bone marrow that go on to produce red blood cells. Since hundreds or even thousands of tests could be performed on cells extracted from a single animal, the assay could dramatically reduce the numbers of animals needed.

Not only that, but the increased throughput of the test would make it cheaper to run. As a result, drug developers could afford to use it earlier in the development process, saving time.

Shuga explained that there are other in vitro screens but they are sometimes conducted in Salmonella or Chinese hamster ovary (CHO) cells, whereas this test is "closer to normal mammalian physiology".

In fact, the test could also be extended to human tissue, according to Professor Harvey Lodish, who supervised the research, along with Professors Linda Griffith and Leona Samson.

"This research is the first stage in a new type of clinical drug toxicity test," said Lodish.

"And although we haven't done it, you may be able to extend the technique to humans. Humans are the gold standard in that one wants an assay that directly predicts toxicity in humans, not animals, and you could obtain human bone marrow that's left over from medical procedures."

"If you could change the micronucleus assay to have a human cell readout, that would be pretty amazing," agreed Samson. Such a test could offer a new way to examine how different individuals respond to chemotherapeutic agents, she explained.

Shuga first worked along with postdoctoral researcher Jing Zhang to adapt techniques from an established cell-culture system based on mouse foetal liver cells to create a new system based on adult red blood cell precursors from mouse bone marrow.

If a drug causes DNA to break, the resultant fragments are not incorporated into the nucleus of newly formed cells. Instead they form 'micronuclei' and if the body cannot repair the damage, the malformed cells either die or survive with mutations that can cause serious side effects, such as leukaemia and anaemia.

Shuga also pointed out that cheaper results don't necessarily mean poorer quality results. The new test was validated against three DNA alkylating agents that all produce a positive result in live mice, and the compounds all produced a positive result in the new test too. He did say that the new test had higher levels of baseline genotoxicity - probably because of elevated levels of red blood cell growth in the cells, compared to in live mice.

If mutant mice that are deficient in certain DNA-repair systems are tested in vivo, they are seen to be more sensitive to the toxic agents than normal mice. Shuga then validated his test further by showing that these results could be repeated in vitro by incorporating bone marrow cells derived from mutant mice into the new test. The results of the test are published in the 14 May edition of the Proceedings of the National Academy of Science journal.

There is a downside though: being based only on cells, the test is isolated from other bodily functions. Some compounds only become genotoxic after they are metabolised by the liver, according to Shuga.

"These agents would not be detected in the system as we have described it. Fortunately, it is well established that liver microsomes can be added to in vitro genotoxicity screens in order to restore some measure of metabolic activation," he said.