Somatic cells as indicators of germinal mutations in the mouse.

In attempts to find a prescreen for mutagens that may induce heritable mutations in mammals, an in vivo somatic mutation test has been developed in the mouse that uses a localized gene product (hair pigment), is relatively fast and cheap, and gives results that have some predictive value for point mutation induction in spermatogonia. Embryos heterozygous at specific coat color loci are exposed to the presumptive mutagen, and 3 weeks later the fur is observed for spots of altered color. It is possible to distinguish spots resulting from expression of the recessive (RS's) from spots having various other causes. In tests with seven compounds, mutation rates per locus and unit dose have been calculated on the assumption that 175 cells are at risk per 10¼-day embryo (a number derived from distribution of spot proportions). These rates are found to be roughly parallel to, but uniformly higher than spermatogonial point-mutation rates for the same seven compounds. The higher somatic rates are presumably due to the fact that RS's can result from several genetic mechanisms besides point mutations. The spot test, which has not to date given any false negatives, may thus be considered a useful in vivo prescreen for heritable germinal mutations in mammals.

by Liane B. Russell* In attempts to find a prescreen for mutagens that may induce heritable mutations in mammals, an in vivo somatic mutation test has been developed in the mouse that uses a localized gene product (hair pigment), is relatively fast and cheap, and gives results that have some predictive value for point mutation induction in spermatogonia. Embryos heterozygous at specific coat color loci are exposed to the presumptive mutagen, and 3 weeks later the fur is observed for spots of altered color. It is possible to distinguish spots resulting from expression of the recessive (RS's) from spots having various other causes.
In tests with seven compounds, mutation rates per locus and unit dose have been calculated on the assumption that 175 cells are at risk per 101/4-day embryo (a number derived from distribution of spot proportions). These rates are found to be roughly parallel to, but uniformly higher than spermatogonial point-mutation rates for the same seven compounds. The higher somatic rates are presumably due to the fact that RS's can result from several genetic mechanisms besides point mutations. The spot test, which has not to date given any false negatives, may thus be considered a useful in vivo prescreen for heritable germinal mutations in mammals.
Despite the recent proliferation in development of submammalian and in vitro test systems for mutagenic studies, there is general agreement that in vivo studies on heritable genetic effects in mammals must, at some stage, enter into the risk evaluation for compounds that show a potential for affecting human populations. Since such testing can, however, be relatively timeand space-consuming, it must be reserved for only selected possible mutagens.
For this reason, some effort has been devoted to the perfection of an in vivo mammalian somaticmutation test that might be capable of acting as a prescreen for germinal mutations of various kinds. This test was first developed by us in an x-ray experiment about 20 years ago and, at that time, explored for its various-complexities (1). The usefulness of the method in chemical mutagenesis has been tested more recently (2)(3)(4)(5)(6)(7)(8)(9). We have elsewhere presented some of our recent results in some detail (3), have discussed some of the more basic aspects of the method (9), and have compared the interpretations various investigators have made of their data (9). For the purposes of this symposium, some of *Biology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830. the findings will be summarized, and the relation between somatic and germinal mutation rates explored.
In a germinal mutation experiment, each individual examined for presence or absence of a whole-body mutation scores one cell that was at risk to the mutagen. In an in vivo somatic mutation experiment, an individual is examined, instead, for clones of mutant cells (Fig. 1). While each clone again represents just one cell at risk, the whole irndividual represents a population of cells at risk-an obvious advantage in numbers. It should be noted, however, that the altered clone (consisting, as it does, of somatic cells) cannot ordinarily be involved in breeding tests, so that the nature of the genetic event that produced it cannot be genetically analyzed.
The method as originally developed (1) meets the following prerequisites of a successful somatic mutation prescreen: it uses a localized gene product (namely hair pigment) so that clones of exceptional cells can be directly identified; it involves an in vivo mammalian situation; and it is relatively fast and cheap. In addition, as more recently demonstrated, the results have predictive value for germinal point mutations (3,9).
The in vivo somatic mutation test consists of treating embryos (i.e., populations of cells) that are heterozygous at specific coat color loci, allowing animals to be born and to grow a coat, and examining this for mosaic patches, indicative of clones of mutant cells. From the time of mating the animals to the final result, this process takes only 5 weeks.
Recessive mutations at the heterozygous loci do not, however, constitute the only cause of spots. This feature of the results, if properly used, can lend additional resolving power to the method. It can, however, also lead to misinterpretation in inexperienced hands. Our interpretation of the various causes of spots was derived from a comparison of results from heterozygous embryos with those from homozygous wild-type embryos (having mothers of the same strain): types of spots found exclusively in the former population are presumably due to an "'uncovering" of the recessive, while types of spots found in both populations presumably have other causes (1). The various causes of spots are summarized in Table 1.
In order to determine whether there was any relation between the frequency of recessive spots (RS) and the frequency of germinal point mutations, we carried out in vivo somatic-mutation tests with compounds for which specific-locus mutation-rate data are available for spermatogonia. [Spermatogonia are the male cell stage of importance for risk evaluations (1O).] These results, an extension of some published earlier (3), are summarized in Table  2. White near-midline ventral spots (WMVS), which are scored independently of RS's, are found with elevated frequencies in some of the treatment groups, generally those yielding also appreciable frequencies of malformed animals (observed by external examination at birth). Both WMVS's and teratogenic effects have been ascribed to cell killing (3).
Doses that have been used in spermatogonial specific-locus experiments are generally near the highest compatible with continuous fertility of the males. In the spot test, however, the maximum tolerated doses are considerably lower, since the method requires the treatment of embryos during one of their most vulnerable stages, day 10¼/4 (1). In general, we have found it possible to use arrays of exposures that straddle a dose one-third of the one used in spermatogonia. In such dose ranges, one proportion of total fur occupied, and assuming that The mutation rates per locus per unit dose for (I Abbreviations: EMS, ethyl methanesulfonate; MMS, methyl both sets of data are listed in Table 3. The induced methanesulfonate, TEM, triethylenemelamine; MC, mitomycin spermatogonial rate is shown on line with the C: HC, hycanthone; BP, benzo[a]pyrene; DEN, diethylnitsomatic rate per unit dose is, in each case, considrosamine. b See Table 1, last column, for explanation of symbols.
erably higher than the spermatogonial rate (and somatic rate per unit dose is, in each case considerably higher than the spermatogonial rate (and Table 3. Comparison of mutation rates at specific loci induced by higher even than the upper 95% confidence limit in seven agents in spermatogonia or somatic cells. cases where the spermatogonial rate was zero).
The relatively higher RS   Weighted average these positives are probably the result of genetic d Where no mutations were observed, the upper 95% confi-changes, other than point mutations, that represent dence limits of zero-frequency is shown in parentheses. certain types of hereditary risks. The spot test has I not, to date, given any false negatives. Because of this, and because of the finding of a rough parallelism (i.e., a relatively higher frequency of RS's in the case of compounds that induce spermatogonial mutations), the spot test can be a useful in vivo prescreen for heritable germinal mutations in mammals.