Measurement and characterization of micronuclei in exfoliated human cells by fluorescence in situ hybridization with a centromeric probe.

The micronucleus (MN) assay in human exfoliated cells has been widely used to detect the genotoxic effects of environmental mutagens, infectious agents and hereditary diseases. Substantial variability characterizes the MN frequencies reported by different research groups. One reason for this may be the restricted resolution power of the Feulgen-Fast-Green staining that is routinely used. Here we describe a new version of the MN assay that employs fluorescent propidium iodide staining along with fluorescence in situ hybridization (FISH) with a centromeric probe. Buccal and urothelial cells were collected from 5 healthy unexposed female volunteers and 55,000 cells analyzed for MN frequency and abnormal nuclear events. The Feulgen-Fast-Green and the new fluorescent staining produced very similar results. The frequency of MN in buccal cells was 0.145 +/- 0.118% and in urothelial cells 0.083 +/- 0.074%. No correlation was found between the frequencies of MN in the two types of exfoliated cells. FISH with a centromeric probe allowed MN containing whole chromosomes with a centromere to be differentiated from those containing only acentric fragments. The former appear as a result of chromosome lagging in mitosis, while those without a centromere are due to chromosome breakage. In urothelial cells 43% of MN were centromere-negative and in buccal cells -44%. Fluorescent staining provided more accurate scoring of degenerative cells than standard Feulgen-Fast-Green staining. The combined frequency of pycnotic cells, "broken eggs" and cells with fragmented nuclei did not exceed 2%, while that of karyorrhexis and karyolysis together was as high as 21%. Significant interindividual variability was found in the frequency of cells with karyolysis and karyorrhexis. Thus, the new version of micronucleus assay allows for MN to be scored more precisely, the mechanism of MN formation to be determined and abnormal nuclear events to be readily identified in exfoliated human cells. It is therefore ideal for studying genotoxicity in human populations using exfoliated cells from the mouth, bladder and nose.