Jeanine M. Pennington and Susan M. Rosenberg from Baylor College of Medicine in Houston have developed a fluorescence-activated cell sorting (FACS) assay that detects DNA damage. They used the assay to study breaks that occur naturally in both strands during replication. These breaks can occur when the replication machinery encounters obstacles and no longer can continue the process. Rejoining the broken strands sometimes results in chromosomal rearrangements likely to cause cancer and other genetic diseases, although sometimes these rearrangements can result in beneficial genetic changes that promote evolution.Previous studies have suggested that the breaks happen frequently. Although these studies assessed break frequency indirectly by counting viable cells or lost chromosomes, cells can die or lose chromosomes for other reasons. The studies also employed Escherichia coli mutants lacking repair proteins, but the loss of a repair protein has been shown to affect cellular processes other than repair. To assess the number of breaks more directly, the Baylor researchers tied GFP expression to DNA damage and measured the fluorescence by FACS. They used E. coli as a model organism and tied GFP expression to a promoter involved in the SOS response, a process specific to bacteria that induces the expression of survival genes when DNA damage occurs. However, the assay probably can be modified for use with any organism, Rosenberg said. To do so, GFP expression must be tied to a promoter activated during the DNA damage response of each organism, and good controls must be selected for comparison, she said.The researchers chose a GFP variant that exhibits stable fluorescence for a long time because the stability increased the sensitivity of the assay. For their results to have stastistical significance, they needed to analyze ~100,000 cells, and FACS flow cytometry can analyze that many cells in a single minute, saving time. The researchers used a flow cytometer equipped with a Beckman Coulter forward fluorescence detector. Pennington said that the detector contains a photomultiplier tube that amplifies the forward light-scatter signal, enabling efficient detection of small bacterial cells as well as cell sorting with few errors. In fact, the flow cytometer detected GFP fluorescence with nearly 100 percent accuracy, she said.As controls, the researchers used several E. coli mutants, perhaps the most important of which lacked the enzymes RecA or RecB. RecA is necessary for the SOS response to occur, so the presence of the RecA mutants indicated that the response had happened. An SOS response to double-strand breaks cannot occur without RecB, so the RecB mutants enabled the researchers to differentiate between double-strand breaks and other types of damage.Using FACS, the researchers discovered that 0.9 percent of wild-type bacterial cells had DNA damage. Control RecB mutants indicated that double-strand breaks accounted for 62 percent of this damage (Figure 1).Figure 1. Researchers developed a flow cytometry assay that directly detects DNA damage by measuring GFP expression. Naturally growing wild-type bacteria experienced DNA damage 0.9 percent of the time. RecB mutants indicated how much DNA damage resulted in breaks of both strands, and RecA mutants are a negative control. Images reprinted with permission of Nature Genetics.Sending an SOSThen the researchers experimentally determined how sensitively their assay detected double-strand breaks. To do this, they exposed a set of bacteria to a restriction enzyme that induces double-strand breaks, so that they knew that all the bacteria had a single break per chromosome, and they used E. coli that carried a genetic element that each bacterium could use to repair itself. Because all the bacteria had a break, all of them should have fluoresced. However, only 27 percent of the cells fluoresced, meaning that 27 percent of cells with a double-strand break induced the SOS response, making the assay 27 percent sensitive. Pennington said that this percentage is sufficient to calculate the number of double-strand breaks.Finally, the researchers estimated that about one-half of the fluorescent cells arose from new DNA damage per generation, based on their flow cytometry results. To recap, the experimental data showed that only 0.9 percent of the bacterial cells fluoresced, that 62 percent of the sample cells had double-strand breaks, that the assay can detect 27 percent of these breaks, and that one-half of these cells reproduced per generation. Mathematically, this translates to 1/2 (0.9 percent × 62 percent/27 percent), or 1 percent, meaning that only 1 percent of cells per generation have double-strand breaks. This estimate is 20 to 100 times lower than previous ones, and it signifies that double-strand breaks are much rarer than previously predicted.If we can assume that previous estimates of the number of chromosomal rearrangements that occur as a result of double-strand breaks are correct, then 1 to 10 percent of these breaks would result in rearrangements, Rosenberg said. That equals a 100 times higher risk of rearrangements than previously predicted. Thus, if the results can be generalized to other organisms, they suggest that double-strand breaks are much more likely to lead to chromosomal rearrangements typical of cancer. This means that cells strongly avoid double-strand breaks in the first place, rather than relying on repair mechanisms.Two-thirds of the bacteria that had breaks did not grow or reproduce, which supports the avoidance hypothesis. This senescence-like state is reminiscent of human cells, which also can stop growing or reproducing to avoid cancer, although the physiological mechanism is different. Pennington said that, although a senescence-like state in response to DNA damage has never before been observed in bacteria, it is a documented response to aging and nutrient depletion. She and Rosenberg believe that the senescence-like state may function to prevent less evolutionarily fit cells from proliferating so that hardier cells can reproduce. The researchers reported their findings in the June 2007 issue of Nature Genetics.Figure 2. As shown in this graph, GFP expression was induced by UV light, a known agent of DNA damage. LexA(Def) is a positive control.Rosenberg said that her laboratory plans to use the assay to examine specific genetic changes that lead to particular diseases as well as to observe specific responses to DNA damage caused by external agents such as oxidation. Her group already has exposed the cells to UV light, and that result has shown that the test can detect external agents of genetic change (Figure 2).