Mutant manifestation: the time factor in somatic mutagenesis

manifestation n. the act, process, or instance of making evident (Webster's Dictionary, 1961) Mutagenesis in live mammals is much more complex than in bacteria, fungi or cultured mammalian cells. In microorganisms, metabolism is rapid, replication of the genome may take less than 1 h, and mutation fixation is necessarily quick. In microbes, because of their rapid cell division, pre-existing non-mutant gene products are quickly diluted or replaced, making the mutant phenotype evident within a few hours. Even in cultured cells, mutant frequencies are typically maximal within a few days. In these cases, the time at which samples are taken after treatment is readily standardized for a particular locus. In vivo, however, most of these processes happen much more slowly and other processes are also involved, so the maximal mutant frequency may be observed as long as 3 months after exposure in sperm (Douglas et al., 1995) and possibly longer in other tissues. It can be as short as 1 week in other tissues (Tao et al., 1993), so tissues differ widely in this respect (Hoorn et al., 1993; Tao et al., 1993; Douglas, 1995; Heddle and Swiger, 1996). Hence, one of the most important variables in experiments on mutagenesis in vivo is the time between treatment and assay, yet there has been no generally accepted name for this variable. The names that have been used include fixation time, expression time and sampling time, but each has an implication that is not justified. For this reason the term mutation manifestation is suggested as a suitable term to represent the interval between treatment and measurement. In general, the manifestation time should be the time required for the mutant frequency in the target cells to reach a maximum. The processes involved in mutagenesis in vivo can be conveniently grouped into three phases, as illustrated in Figure 1. The first phase includes the physiological processes leading to DNA damage and the repair or fixation of the damage to create a mutation. The second phase is expression of the mutant phenotype, which involves transcription, dilution of the pre-existing non-mutant RNA and protein and the subsequent physiological change in the mutant cell. The third phase is cellular expansion of the mutant clone in the tissue. These phases have different durations. The length of the first phase (largely pharmacokinetics) varies with the chemical, but is ordinarily quite short, perhaps 1 or 2 days. The duration of the DNA repair processes has not been measured, except in occasional circumstances, and may vary from tissue to tissue and from chemical to chemical, not to mention as a function of dose. Nevertheless, it is likely that the first phase, here called by a name used for the last step, fixation, will occur within a few days in most tissues for most chemicals. The duration of the second phase, phenotypic expression, will vary with the locus as some proteins turn over rapidly and others slowly, but should be the same for all mutagens. For transgenic assays involving shuttle vectors, including the commonly used Big Blue™ Mouse and Muta™Mouse assays, phenotypic expression occurs in vitro and does not affect the timing of appearance of mutants. For loci that are expressed in vivo, such as Dlb-1 and hprt, however, this is a factor. The third phase is the longest phase in most tissues and deserves special mention. Probably the most important variable in the appearance of somatic mutations in vivo is tissue turnover (Heddle et al., 1995). For somatic mutation of most tissues, it is the stem cells that are most important (Cairns, 1975; Heddle et al., 1996). As Heinrich Malling says, `We are not interested in mutation in dead cells!' (c.f. Burkhart and Malling, 1994). However, most of the cells in any tissue are reproductively dead. The cells that are more at risk for oncogenic mutation are the proliferating cells, which includes the stem cells. In some of the important tissues, such as liver, cell turnover is slow and the important cells may be a minority population (see Sell, 1990 for a review). Even in rapidly dividing tissues, such as the bone marrow and the intestinal epithelium, stem cells are <1% of the cells present. Analysis of the tissue after mutagenesis but before replacement by stem cells is the equivalent of assaying dead cells in many cases. The stem cells may be at greater or lesser risk of mutation than most of the cells of the tissue. Since tissue turnover times vary enormously, this aspect of the manifestation time is specific to each tissue. It may be that by killing some of the somatic cells and stimulating the stem cells to replace them, the tissue turnover time and thus the time required for the appearance of mutants is reduced at high doses. This is not to say that mutations in non-stem cells are necessarily unimportant. Certainly for germ cells, which are not reproductively dead, mutation at any stage is of equal importance. This may also be true for some tissues but, unlike germ cells, once the cell has been shed from the tissue, any mutations it had are no longer of any importance. Nevertheless, if a mutation should arise in a non-stem cell, it could be of potential importance so long as that cell survives. In addition, non-dividing cells are at lower risk of mutation, as is clear from the long slow rise in mutant frequency in the liver and lower maximum as compared with rapidly dividing tissues (Douglas et al., 1996). For quantitative determination of mutant frequencies and a proper understanding of somatic mutation in vivo, the time at which the measurement of mutation frequency is made is crucial. This time has been called the fixation time, but this is understood by those who have used it to be a misnomer because it is more than that, including expression and all of the other processes described above. Similarly, `expression time', which has also been used, is unsuitable because tissue turnover and other processes are important factors in the timing of the appearance of mutants. `Sampling time' has fewer implications (although it might be taken to mean the time required to take a sample for analysis) but carries less meaning: it is possible to sample both before and after the mutant frequency has reached the maximum in a tissue. For these reasons, the term `manifestation time' seems valuable. The term would include the various biological events that intervene between treatment and measurement, as illustrated in Figure 1. In practice it would be estimated by measuring the minimum time required for the mutant frequency to reach its plateau value after an acute treatment, as illustrated in Figure 2. This figure assumes that the mutations are neutral, i.e. that there is no selection against them (Heddle et al., 1995; Cosentino and Heddle, 1996). As illustrated, there is both a minimum manifestation time, before which no induced mutations are recovered, and an optimal manifestation time, after which the maximum mutant frequency is always obtained. The existence of a minimum manifestation time, before which no induced mutations are detected, has been clearly demonstrated for sperm cell mutagenesis by Douglas et al. (1995) in the Muta™Mouse assay and for intestinal mutagenesis in the Dlb-1 assay (Cosentino and Heddle, 1996). For other loci that are expressed, such as TK, there may be selective factors that alter the mutant frequency during or after manifestation, which requires a test at each such locus. Nevertheless, to the extent that tissue turnover is the largest component of the manifestation time, this time will be largely a characteristic of the tissue and will not vary much with the mutagen used nor the treatment protocol. This provides a simple and testable model for mutagenesis in vivo. It may be that, sometimes, the biological situation will prove more complex than envisioned here. Jargon must evolve as the understanding of the situation it describes improves. However, currently, use of the term manifestation time would simplify description of mutagenesis in vivo, help to avoid misconceptions and improve communication within the field. Illustration of the stages involved in mutagenesis in vivo. The durations of these phases may be very unequal, with the last phase being much the longest. In transgenic assays involving shuttle vectors, like the Big Blue™ Mouse assay and the Muta™Mouse assay, phenotypic expression occurs in vitro and is very rapid. For endogenous genes, mutants would begin to appear during the expression period and would reach a stable plateau at the end of the tissue turnover time. Illustration of the theoretical change in mutant frequency with time for a neutral gene, such as the lac1 or lacZ transgenes (redrawn from Heddle et al., 1995). A minimum manifestation time is shown before which no induced mutations are detectable (Phase I), a period of time during which tissue turnover is occurring and the mutant frequency is rising (Phase II) and a period of time when the mutant frequency is stable (Phase III). Thus the minimum time between treatment and sampling is the time to reach the beginning of Phase III, which would be the characteristic manifestation time of the tissue. More complex patterns are possible, as discussed in the text. Research in my laboratory is supported by the Cancer Research Society Inc., the National Cancer Institute of Canada, with funds from the Canadian Cancer Society, and by the Natural Sciences and Engineering Research Council. I am grateful to Lidia Cosentino and Roy R.Swiger for helpful discussions and criticism.