DETECTION OF 6-THIOGUANINE-RESISTANT HUMAN PERIPHERAL-BLOOD LYMPHOCYTES USING 5-BROMODEOXYURIDINE LABELING IN COMBINATION WITH IMMUNOCYTOCHEMICAL STAINING

Measurement of frequencies of 6-thioguanine-resistant (TG(r)) human peripheral lymphocytes may contribute to quantitative genetic risk assessment in occupationally or environmentally exposed human populations. A simple procedure for the detection of TG(r) human peripheral blood lymphocytes was developed in our laboratory, using whole blood culturing and 5-bromodeoxyuridine (BrdU) labeling in combination with immunocytochemical staining. Modifications of the procedure designed to reduce the false positive effects of spontaneously cycling lymphocytes (phenocopies), and to optimize duration of BrdU labeling and the culturing period, were evaluated. A standard procedure was developed which applied 24 h cold storage of the diluted heparinized blood (1:10, v:v in RPMI 1640 medium) at 4-degrees-C to reduce the effect of spontaneously cycling lymphocytes, and whole blood culturing in RPMI 1640 complete medium with stimulation of T lymphocytes using phytohemagglutinin (PHA), selection of TG(r) lymphocytes by adding TG to a final concentration of 2 X 10(-4) M and, after 24 h of incubation, labeling of TG(r) lymphocytes with 2.5 X 10(-5) M BrdU during 16 h. Using this standard procedure, frequencies of TG(r) cells in 45 healthy individuals (aged 21-64) were observed to range from 0.3 to 229.8 x 10(-6), with a mean variant frequency (VF) (+/-SD) of 136 X 10(-6) (+/-35.8). After exclusion of the one extremely high value of 229.8 X 10(-6), mean VF was 8.7 X 10(-6) (+/- 14.1). A significant inverse correlation was found between logVF and the labeling index of control cultures (LI(c)), indicating that cultures with low LI(c) tend to yield higher VF. No effect of donor sex or smoking behavior was observed, but logVF significantly increased with age; a 2.4% increase per year was found. After adjustment for the effect of LI(c) on VF and after exclusion of lymphocyte cultures with LI(c) < 10% which are considered to be less reliable, the positive correlation between age and VF still existed. In addition, an inverse correlation was observed between age and LI(c). Even after adjustment for the effects of LI(c) and age, however, no statistically significant effect of smoking behavior on VF was found.