Quantification of thioguanine-resistant lymphocytes from mice irradiated in vivo.

Adult mice were Co-60 gamma irradiated, and 7 months later splenocytes were isolated, cultured in microwells, and the frequency of hprt-deficient mutants was determined by measuring the cloning efficiency in media with 6-thioguanine. The mutant frequency at 2, 4, and 6 Gy was 1.6 x 10(-5), 4.4 x 10(-5), and 12.7 x 10(-5), respectively. The frequency of spontaneous mutants was 2.5 x 10(-6). The effect of metabolic cooperation on the cloning efficiency of thioguanine-resistant T-cells in selective medium was evaluated in co-cultures with wild-type T-cells. We found that the growth of hprt-deficient T-cells is supported in the presence of thioguanine-inactivated wild-type splenocytes up to a cell density of 5 x 10(5) cells per well. When cell density was higher, cell growth was inhibited. Possibilities and limitations of cloned lymphocytes for the analysis of somatic mutations that occur in vivo are discussed.


Introduction
In vitro assays such as the Ames test or the mouse lymphoma test cannot replace in vivo tests as predictors of genotoxicity in mammals. Jones and co-workers (1)(2)(3) have demonstrated that cloning mouse lymphocytes permits analysis of mutations in the hprt gene that have occurred in vivo in T-cell progenitors. Possibilities and limitations of this method were investigated, and preliminary results on the mutant frequency in gammairradiated animals will be given.

Thioguanine-Resistant Mutants and Cultivation of Splenocytes
A resistant clone was isolated from wild-type L5178Y mouse lymphoma cells after treatment with N-ethyl-Nnitrosourea as described by others (4). Resistant splenocytes were prepared from hprt-deficient mice (5).
Splenocytes were cultivated in general as described by Jones et al. (1). In brief, spleen cells are isolated from 3 to 6 adult male C57bl mice and cultured in roundbottomed 96-well microtiter plates in the presence of 60 Gy irradiated feeder cells (104 hprt-L5178Y mouse lymphoma cells and a variable number of splenocytes). Modified RPMI1640 medium in an atmosphere of 5% CO2 was used for culture. The medium was supplemented with 10% fetal calf serum, HEPES (25 mM), mercaptoethanol (5 x 10-5 mM), 0.2 mg/mL sodium pyruvate, recombinant IL-2 (100 U/mL)i penicillin (100 U/mL), streptomycin (100 ,ug/mL), 1 ,ug/mL indomethacin, and 5 jig/mL Con A. Medium of selective plates contained in addition 2.5 ,ug/mL 6-thioguanine (TG). The total volume per well was 0.2 mL. In each well, 100 ,uL medium was replaced twice a week. Fresh medium improves cell growth. Late on day 14 in culture, 0.5 ,uCi [Me-3H]thymidine (3H-TdR), specific activity 20 Ci/ mmole, dissolved in 100 ,uL fresh medium, was added to each well. Cells were harvested 12 to 16 hr later, and the 3H-uptake was measured by liquid scintillation counting.

Detection of Wells with Cell Growth
Cell growth was detected visually as well as by 3H-TdR uptake. When 3H-TdR uptake is used to evaluate the wells, a cutoff point has to be defined to detect wells with proliferating cells. In nonselective plates the 3H-TdR uptake in wells that contained only feeder cells was used to discriminate between wells without cell growth and wells with growth. Wells were considered positive for proliferation when they exceeded by more than 3 SD the mean value of the 3H-TdR incorporation in wells that contained only feeder cells.
The uptake of H-TdR by cells inactivated through irradiation is not the same as in TG-inactivated cells. As a consequence, the cutoff point of nonselective plates cannot be used in selective ones. For these plates the cutoff point was defined as the mean value + 3 SD of at least 100 wells without cell growth in selective medium that were prepared with splenocytes from untreated animals. Since spontaneous TG-resistant mutants are rare, positive wells could be eliminated by inspection of the data. (1) is based on P(0) the proportion of wells in which a colony has not grown (6). Mutant frequency (MF) was calculated using Equation (2). Confidence limits of MF were determined as described by Furth et al. (7). CE = -In P(0) number of cells per well MF = CE in selective medium (2) CE in nonselective medium

Results and Discussion
Cloning of T-lymphocytes in microwells was used to determine thioguanine (TG)-resistant mutants (Fig. 1). TG is taken up into the cell by the purine scavenger pathway using the enzyme hypoxanthine phosphoribosyl transferase, the product of the hprt gene. TG-resistant variants of T-cells can survive in the presence of thioguanine as a consequence of a loss of function of this enzyme.
Cell growth was quantified by measuring the 3H-TdR uptake. An uptake of 3 cpm was found per growing cell. In Figure 1, between 500 and 8 x 104 cpm was measured per well with cell growth. This indicates a great variability in the T-cell colony size. Each well was visually screened for cell growth before harvest. In nonselective plates, 70% of all wells in which a colony had grown, as estimated by 3H-TdR uptake, could be detected by eye. In selective plates this fraction was not more than 27%. This indicates that small colonies are often covered with inactivated cells. In rare cases, cell growth was detected visually even though the 3H-TdR uptake in the well was lower than the cutoff point.
Metabolic cooperation by transfer of purine metabolites through the medium or via cell contact might inhibit the growth of hprt-deficient mutants in medium with thioguanine when wild-type cells are present (8,9). We have studied the role of metabolic cooperation in Tcell culture. Figure 2 demonstrates that wild-type mouse lymphocytes inhibit the growth of thioguanineresistant L5178Y mouse lymphoma cells, whereas the growth of hprt-deficient splenocytes is supported by the presence of thioguanine-inactivated wild-type splenocytes up to a concentration of 5 x 105 cells/well. However, when the cell density is higher than 5 x 105, wildtype splenocytes may inhibit the growth of TG-resistant Selective medium  T-cells. Therefore, most experiments were performed with 5 x 105 cells per well. Figure 3 gives the mutant frequency (MF) of the hprt locus in T-cells from mice irradiated 7 months previously. The increase of MF with dose resembles a linear quadratic dose-effect curve. Recently Jones (10) reported about data on the mutant frequency in T-cells from gamma-irradiated animals. In her experiments (10), the dose-effect curve was strikingly curvilinear at weeks 3 to 5 postirradiation. Thereafter the quadratic component was less, and the frequency of mutants present in spleen cells of mice given high doses declined to one-third the maximum observed frequency. At the maximum, the observed frequencies in 4 Gy irradiated animals, measured 4 or 5 weeks after exposure, were 7.2 x 10-5 or 9.5 x 10-5, respectively. These values are of the same order of magnitude as the results of our investigation. Another relationship between MF in spleen cells and dose was found by Dempsey and Morley (11) 4 weeks after X-ray irradiation. In their experi-ments, X-rays at a dose of 1.5 Gy produced an approximately 20-fold increase in MF. Thereafter the number of induced mutant cells tended to plateau between 1.5 and 4.5 Gy. An increased MF was also found in T-cells from humans exposed to cytostatica (12) and from atomic bomb survivors more than 40 years after irradiation (13).
In high dose irradiation experiments, most of the Tcells and T-cell progenitors will die because they belong to the most radiosensitive cells ofthe body (14). Mutants can only be detected in the fraction of surviving T-cells and their progenitors. Therefore, interpretations of dose-effect curves of MF in T-cells have to consider the hierarchical organization of the T-cell renewing system and its radiosensitivity (Fig. 4). The renewing system comprises the virtually quiescent self-renewing pluripotent stem cell population in the bone marrow and two rapidly dividing non-renewing populations of differentiating cells in bone marrow and thymus. A final selective expansion takes place in spleen and lymph nodes subsequent to contact with antigen. Cells of the stem cell compartment are normally released at slow rates. Hence there is relatively more time for repair of any genetic damage in these cells than in any other part of the renewing system. These cells are relatively radioresistant, whereas all other cells of the renewing system belong to the most radiosensitive cells in mammals. In which compartment mutants are induced is not known. not find an increase of hprt-T-cells in the thymus before 1 week after irradiation and in the spleen not before 3 weeks. This is an argument for the hypothesis that the mutations we measured do not arise in virgin or mature T-cells of primary or secondary lymphoid organs, but in stem cells or pre-T-cells in the bone marrow. The fact that the MF is increased even 40 years after irradiation (13) indicates that mutants are preserved in the stem cell pool.
In most experiments on the mutant frequency in mammalian cells, mutations of the hprt locus have been measured. This gene is located on the X chromosome; the hprt gene is hemizygous. Genetic damage involving this gene and extending into adjacent essential hemizygous genes results in lethality. Therefore, some hprt gene mutations will not give a viable clone, and the mutant frequency will be underestimated. It would thus be more appropriate to use mutational systems employing autosomal markers such as the HLA locus (15).
In summary, the measurement of the mutant frequency in T-cells might be a good system for the biomonitoring of genotoxic agents under in vivo conditions. It would be desirable to screen not only for mutations at the hprt locus but also for autosomal genes. The direct DNA sequence analysis of in vitro amplified hprt cDNA from mutants will lead to a better understanding of the final alterations in DNA by mutagenic events.