10

Radiation Genetics

Earl L. Green and Thomas H. Roderick

The laboratory mouse has been and probably will continue to be the mammal of choice for experiments on the genetic effects of radiation. This is so because the existing inbred strains, the stocks bearing named mutations, and the already known genetic linkage groups and chromosome maps provide a desirable degree of experimental sophistication. In addition, the large numbers of mice required for mutation studies may be reared in relatively small space with relatively little expense. More importantly, as mammals, mice may help to bridge the gap between Drosophila and other small organisms, on the one hand, and human beings and other large organisms, on the other, in attempts to predict the effects of high-energy radiation on genetic constitutions not amenable to direct experimental analysis.

The genetic effects of radiation are classifiable into three categories: the induction of mutations, the breakage of chromosomes, and the effects of these changes on populations. We use the term "mutation" to mean any heritable change in the genetic constitution detectable by breeding methods. The breakage of chromosomes by radiation and the related effects are discussed in Chapter 7. The other two aspects are discussed in this chapter.

Little and Bagg ( 1923) pioneered in attempting to induce heritable changes in the germ plasm of the mouse. Although it is believed that the variations they observed in the descendants of X-rayed mice may not have been induced by X-rays, still this study called attention to the need for such investigations and to the use of laboratory animals for the purpose. Within a few years their work was followed by studies by Hertwig and Snell and later by larger studies by Charles, Russell, Carter, Searle, Lyon, Lüning, and others. Falconer, Wright, Haldane, and Woolf, among others, have contributed important concepts of theory and method.

The most extensive investigations of radiation-induced mutations and mutation rates in mice have been carried out at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, and at the MRC Radiobiological Research Unit at Edinburgh, Scotland, and Harwell, England. W.L. Russell at Oak Ridge and T.C. Carter at Edinburgh each started large-scale studies of mice exposed to radiation in 1947 with related but not identical major objectives and with practically the same genetic techniques. A few years later, the mouse stocks at Edinburgh were transferred to Harwell and in 1953 were supplemented by sublines from Oak Ridge to make the studies even more comparable. This section will be based mostly on these two studies.

Russell ( 1954) has thoroughly reviewed the studies up to 1952. Since then several reviews of the rapidly expanding literature on radiation mutagenesis of mice have appeared, including a later paper by Russell ( 1963), opening the way to the preceding papers. We will refer to a few specific experiments in this section. References to studies not in the list of Literature Cited may be found in either the 1954 or 1963 papers of Russell.

Types of mutation

Mutations may be classified with respect to type of cell (somatic vs. germinal), chromosomal location (autosomal vs. sex-linked), nature of effect (dominant vs. recessive, visible vs. lethal), and size of effect (major vs. minor, oligogenic vs. polygenic). The studies reviewed in this section deal predominantly with recessive autosomal visible mutations in spermatogonia because these are easier to detect than others in large routine breeding experiments.

Cell stages

Male mice, exposed to 300 R or more of high-energy radiation, are usually fertile for about 3 weeks and then enter a sterile period for about 3 weeks and then enter a sterile period with lasts 4 weeks or more depending on the dose. Doses less than 300 R may depress fertility without a clear-cut sterile period. After the sterile period the males remain fertile throughout the rest of their lives.

Matings in the presterile period make use of germ cells which were in postmeiotic stages (sperm, spermatids, spermatocytes) at the time of irradiation. Matings after the sterile period make use of germ cells descended from irradiated spermatogonia. As found by Hertwig, the types of genetic changes induced in the gonial stages are characteristically different from those in the postgonial stages. Irradiated sperm tend to transmit dominant mutations, whereas irradiated gonia tend to transmit recessive mutations. By selecting the interval of time between irradiation and copulation one may thus determine whether irradiation postmeiotic or premeiotic cells produce the sperm.

Young adult female mice exposed to 50 R or more of high-energy radiation may produce one, sometimes two, litters and thereafter may be permanently sterile. Their ovaries contain a quite uniform population of oocytes, nearly all in a resting stage prior to the first meiotic metaphase and a few about to be extruded as fertilizable ova without yet having completed the second meiotic division. As a consequence, it is not possible to effect a clear-cut separation of germ cell stages in irradiated adult females. Females must be irradiated as fetuses to expose oogonia.

The type of cell — somatic or germinal — in which a mutation occurs affects its detectability and, more importantly, one's judgment about the seriousness of the consequences. In this section we are concerned chiefly with the problem of detecting and measuring the frequency of mutations in germ cells exposed to high-energy radiation.

Breeding tests

The objective of breeding tests of irradiated and control mice is to analyze the genetic constitution of single germ cells by examining the zygotes they produce or of somatic cells by examining the cell clusters they produce. This may require one, two, or three generations of breeding depending on the nature of the genetic change under investigation.

There are four breeding systems each with a large number of variants useful in radiation genetic experiments with mice. We propose that the systems be called: parental outcross or P-cross, filial outcross or F₁-cross, outcross-backcross or cross-backcross, and outcross-intercross or cross-intercross. These are operational terms referring to what one does with the mice and not to what type of genetic change he expects to find. Each of the systems may be used to detect several types of mutations.

P-cross method. Each parental mouse, irradiated or control, to be tested for mutations is crossed to an unrelated mouse and the progeny are scrutinized for phenotypic deviations or presumptive mutations ( Figure 10-1A). Each viable deviant should be tested by further matings to establish whether a mutation has occurred. The P-cross method may be used to search for all types of dominant mutations — autosomal or sex-linked, visible or lethal — and, with appropriate choices of the mates in the cross, for specified types of recessive mutations as well. The rationale may best be explained by considering a few types of mutations. We shall denote alleles at a given locus as dominant D, recessive r, or wild type +, and newly induced mutations as D* or r*.

If the mating type is +/+ x +/+ (autosomal) or +/+ x +/Y (sex-linked), the F₁ progeny will be +/+ or +/+ and +/Y, unless there is a new mutation. Either sex may be irradiated. Dominant visible mutations are indicated by the occurrence of some progeny with the D-phenotype. Since each F₁ mouse tests one treated or control gamete in combination with an untreated gamete, the relative frequency of the D-phenotype in the treated group, corrected for the relative frequency of the D-phenotype in the control group, is an estimate of the radiation-induced mutation rate at this one locus. If n₁ and n₂ are the numbers of F₁ mice tested in the control and irradiated groups, respectively, and if v₁ and v₂ are the numbers of mice with the D-phenotype in the respective groups, then p₂ = v₂/n₂ is the relative frequency or fraction of mice with spontaneous (D and radiation-induced (D*) mutations in the irradiated group and p₁ = v₁/n₁ is the fraction with spontaneous mutations (D) in the control group. The ratio (p₂ - p₁)/(1 - p₁) is an estimate of the radiation-induced rate of mutation at the D locus. If p₁ is very small, the difference (p₂ - p₁) is a sufficiently accurate estimate of the induced mutation rate. If q₁ = 1 - p₁ and q₂ = 1 - p₁ and q₂ = 1 - p₂, the induced rate is also estimated by 1 - (q₂/q₁).

It may not often be practical to confine one's attention to a single locus. Therefore the progeny are examined for any departure from the expected type and, either supposing the departures to be due to dominant mutations or testing them by breeding to establish them as such, the resulting frequency relative to the total examined gives an estimate of the rate of mutation to dominant visible alleles in the genome as a whole. Computing the ratio given above gives the radiation-induced mutation rate per gamete.

The irradiated and control mice should be mated with unrelated or distantly related mice to minimize the possibility of mistaking a recessive allele, present in the stock in a low frequency, for an induced dominant mutation. That is, if the mating is in fact r/+ x r/+, the r/r progeny may be erroneously scored as D*/+.

If females are irradiated, recessive sex-linked mutations may appear in their male progeny. The mating +/+ x +/Y may produce r*/Y sons.

One variant of the P-cross, the "specific locus" method, has been extensively used for detecting recessive mutations. If the irradiated and control mice are dominant homozygotes for one or more loci, say +/+, +/+, ..., and if they are mated with recessive homozygotes a/a, b, ..., mutations from + to a, + to b, ..., will be detected by the appearance of the recessive phenotypes among the progeny. The number of proved mutations recovered may be used directly to estimate the mutation rate at each locus tested, the treated rate being adjusted for the control rate to give an induced rate. Obviously the rates from several loci may be averaged.

The P-cross method may be used to detect dominant autosomal lethals by a significant reduction in litter size, particularly if fathers are irradiated. If mothers are irradiated, litter size cannot be used to measure the incidence of dominant lethals, as Russell and Russell ( 1956) showed. This must be done by using the ratio of the number of living embryos (n - v) to the number of corpora lutea (n). The induced rate is then estimated as 1 - (q₂/q₁).

If males are irradiated, dominant sex-linked lethals may be detected by a deficiency of daughters. If females are irradiated, recessive sex-linked lethals may be detected by a deficiency of sons. Both litter size and sex ratio are so highly variable that changes in their magnitude are difficult to detect without enormous samples, except in cases of dominant semisterility (due to reciprocal translocations) which can be detected by the half-size of a few litters.

The P-cross method or one of its variants has been widely used. In addition to its being used in the early studies by Hertwig and by Snell on the effects of X-rays and by Snell and Aebersold on the effects of neutrons, it has been used by most later investigators as well. Russell and Carter each devised a multiple recessive tester stock ( a/ a b/ b c^ch p/ c^ch p d se/ de se s/ s) to mate with irradiated mice normally carrying wild-type alleles at all seven loci. By far the largest number of mice in radiation genetic experiments have been bred under this system. Lyon ( 1960) used a balanced lethal line of genotype T tf/ t⁶ + to detect mutations from T to t^m which restore the tail length to normal. Russell and Major ( 1957) adapted the specific locus technique to the detection of somatic mutations.

F₁-cross method. This method requires two generations and may be used to detect sex-linked lethal mutations ( Figure 10-1B). It makes use of dominant sex-linked markers such as Bn, Mo^br, and Ta. Tabby males, irradiated or control, are mated to females heterozygous for bent-tail and brindle in coupling:

F₁ daughters, heterozygous for Ta, Bn, and Mo^br and possibly carrying an induced sex-linked mutation in the chromosome marked by Ta, are backcrossed to wild-type males:

The absence of tabby males (+ Ta +/Y) among the backcross progeny of any one family of adequate size indicates that a lethal mutation is located on the Ta chromosome between the Bn and Mo^br loci. Auerbach et al. ( 1962), in proposing and first using this method with mice, pointed out the difficulties of using these specific markers. As more sex-linked alleles are discovered in mice, new improved stocks may be synthesized for detecting sex-linked lethal mutations.

Cross-backcross method. This method, requiring three generations, is useful for detecting recessive visible and lethal mutations. Each exposed or control mouse in the zeroth generation (G0) is mated to a stock mouse with wild-type or specific mutant alleles. If a recessive mutation has occurred in some of the gametes, the progeny, each representing one gamete from an irradiated or control parent, will be heterozygous for an induced or natural mutation (r*?+ or r/+). These mutations are to be uncovered by two more generations of breeding: an outcross and a backcross. Accordingly, the G1 mice are outcrossed to unrelated wild-type mice to produce prospective heterozygous carriers in G2. The G2 mice are, in turn, backcrossed to their parents to produce G3. Homozygotes (r*r* or r/r) in G3 can be recovered if + mutated to r* or r, if the new phenotype is viable, and if a sufficient number of progeny have been reared to reduce the chance of nonrecovery to a low level. I r*/r* or r/r is lethal, its existence must be inferred from a reduction in litter size in G3 ( Figure 10-1C).

Using this method, Carter and Phillips ( 1952) found two new recessive visible mutations in the progeny of nonexposed mice, and Carter ( 1957a) found evidence for the induction of recessive lethal mutations, manifested by reductions in litter sizes.

The strength of this method may be markedly increased by the use of single markers of selected linkage groups as Snell ( 1935) proposed and, still further, by the use of two or more linked markers as Carter ( 1957b) proposed. Each marker having normal viability in homozygotes should appear in one-half of the progeny in G3. The absence of a marker class is evidence for a nearby induced recessive lethal mutation. The statistical problems of estimating mutation rates from the collected data may be quite formidable. They arise from the uncertainty of detecting all recessive mutations, present or induced in the G0 parents, through their failure to appear in litters of finite sizes three generations later. Falconer ( 1949) has dealt with this problem by introducing the idea of reducing the number of incompletely tested gametes to the equivalent smaller number of completely tested gametes.

Cross-intercross method. This method, the first used for detecting recessive mutations in mice, also requires three generations. It is like the preceding method except that the outcross progeny in G2 are intercrossed instead of backcrossed, but its efficiency is less ( Figure 10-1D). For this method also Falconer ( 1949) solved the problem of scoring the results to estimate the equivalent number of fully tested gametes.

Haldane ( 1956, 1957) proposed a breeding design using the cross-intercross method. Males or females or both of a stock of mice homozygous for several known recessives are exposed to radiation for one or more generations, ending with a generation (G1) not itself exposed. Recessive lethal mutations may have accumulated in these mice. The objective of further breeding is to detect the presence of such lethals by observing deficiencies in the recessive marker classes of an F₂ (=G3) generation, indicating that a lethal mutation has been induced at or near each known locus whose recessive class is deficient. Accordingly the G1 mice are outcrossed and their progeny are intercrossed to produce a multiply-segregating F₂ generation. The method has been used by Carter ( 1959) and by Sugahara and his colleagues ( Sugahara, 1964 and earlier).

A related method is adapted to estimating recessive mutation rates within inbred lines being propagated for other purposes. Following a cross and one generation of intercrossing, brother-sister matings are continued indefinitely. The articles by Snell ( 1945) and Woolf ( 1954) set forth the theory of this method.

Estimates of mutation rates

Charles ( 1950) was the first to estimate the rate of induction of mutations by radiation, using fairly large samples of mice. His estimate of 7 x 10^-5 per roentgen per gamete was for all types of mutations, chiefly dominants, induced in sperm under repeated small doses of penetrating radiation. The very large scale studies at Oak Ridge and Harwell in the ensuing years have established the importance of specifying the sex, cell stage, test system, radiation quality, dose rate, and other variables in estimating mutation rates.

The results of numerous experiments are summarized in Tables 10-1, 10-2, 10-3, 10-4, 10-5, 10-6, 10-7, 10-8, and 10-9. The data for male cells exposed to single doses of radiation at high and low rates are given first. Data on female cells are in the last two tables. The breeding methods and other distinctive features of each experiment are also given in the tables. This arrangement does not reveal the sequence of completion of the various experiments and so masks the excitement of the successive stages of discovery. This side of the story may be found in Russell's ( 1963) review paper.

The mutation rates given in Tables 10-1, 10-2, 10-3, 10-4, 10-5, 10-6, 10-7, 10-8, and 10-9 are given in two ways. First, if mutations of a designated type were detectable at all loci, the rate per gamete is given for the radiation does employed. For example, in row 3 of Table 10-2, five dominant visible mutations in 31,253 gametes gives an overall rate of 16.0 x 10^-5 per gamete exposed to 200 R of acute X-rays. Dividing by 200 gives 80.0 x 10^-8/R/gamete. This is called an overall dominant visible rate because all loci were presumably at risk and dominant visible mutations at any loci were detectable. In this example the rate also includes any spontaneous or natural dominant visible mutations and so must be adjusted to give an induced rate. Second, if mutations only at certain specific loci were counted, the rate per locus is given for the radiation dose employed. For example, in row 4 of Table 10-2, nine recessive visible and lethal mutations in 31,253 gametes tested gives a rate of 28.8 x 10^-5 for seven loci exposed to 200 R. Dividing by seven gives 4.1 X 10^-5/locus. A further division by 200 gives 20.6 x 10^-8/R/locus. If the spontaneous rate is independently estimated as 6.2 x 10^-5 for seven loci, the adjusted or induced rate comes out at 16.2 x 10^-8/R/locus. Leaving the rates on a per gamete or per locus basis, Table 10-4. In the text a few rates are given on a per roentgen basis.

The average mutation rate of seven specific loci to recessive visible and lethal alleles in postmeiotic male germ cells irradiated with 300 R of X-rays was estimated as 13.6 x 10^-5 per locus or 45.3 x 10^-8 per roentgen per locus ( Table 10-1, row 1). No extensive data are available for sex-linked lethals ( Table 10-1, row 2). Irradiated fetal spermatogonia gave a rate of 80.0 x 10^-5/R/gamete for dominant visible mutations at all loci ( Table 10-2, row 3) and a rate of 20.5 x 10^-8/R/locus for recessive visibles and lethals at seven specific loci, roughly one-half that of the postmeiotic rate ( Table 10-2, row 4). An earlier small-scale experiment had given a much smaller rate ( Table 10-2, row 5). A variety of experiments with all spermatogonial stages exposed to low dose rates and relatively low total doses of γ-rays have yielded too few data to provide good estimates of mutation rates ( Table 10-3 rows 6 to 10).

In the first large experiment, 48,007 progeny of males exposed to 600 R of acute X-rays and bred after the sterile period gave an induced mutation rate of 25.0 x 10^-8/R/locus for recessive visibles and lethals at seven specific loci. By 1962, with a much larger sample, the improved estimate was very nearly the same: 22.1 x 10^-8/R/locus ( Table 10-4, row 20, 600R). This is about the same as the rate for fetal spermatogonia and about one-half the rate for postmeiotic stages. The spermatogonial mutation rates at 0, 300, and 600 R were satisfactorily fitted by a straight line, but the point for 1,000 R fell significantly below the line ( Table 10-4, row 20). There is no obvious explanation of this result. It can be accounted for by supposing that spermatogonia are differentially susceptible to the action of large acute doses of radiation so that the observed drop in mutation rate at 1,000 R is possibly due to cell selection. Estimates of other rates for acute exposures of spermatogonia are also given in Table 10-4, rows 11 to 18 and 21.

While the results of exposures to acute doses were being accumulated, the results of exposure of males to chronic radiation suggested that radiation at low dose rates (Table 10-5, rows 22 to 24). Further, mutation rates in oocytes also turned out to depend upon dose rate. Since nearly all oocytes in young adult females, unlike spermatogonia in young adult males, appear to be in a single premeiotic stage, the idea of cell selection as an explanation of the dose-rate effect was less satisfactory than a hypothesis of repair of a premutational state induced by radiation at low dose rates.

To investigate further the low mutation rate at 1,000 R, Russell ( 1963) exposed samples of mice to total doses of 1,000 R of acute radiation in various fractions ( Table 10-6, row 25). The mutation rates obtained with some of the fractions fell near the line extrapolated from the 0, 300, and 600 R points. Although the explanation is not clear, it appeared that some factor other than, or in addition to, cell selection must be operative. Finally, two doses of 500 R of acute radiation given one day apart yielded the highest mutation rate so far found in spermatogonia. This result suggests that the first dose imposes a synchronization and heightened mutability upon the cells and that the second dose thus yields more mutations than otherwise expected. Obviously this phenomenon requires further study by variation of the total doses, the sizes and numbers of the fractional doses, the intervals between the fractions, and the cell stage and type irradiated.

Snell and Aebersold ( 1937) first attempted to induce mutations in mice with neutrons and got evidence that neutrons are five to six times as effective as X-rays in inducing dominant lethals in spermatozoa. Snell ( 1939) used the cross-backcross method of searching for recessive visible mutations but found none. Later studies by Batchelor et al. ( 1964), Searle and Phillips ( 1964) and Russell and Kelly ( 1964a, 1964b) have conformed the evidence that neutrons are five to six times more effective than X-rays for inducing mutations in premeiotic as well as postmeiotic stages. Batchelor et al. ( 1964) found a mutation rate of 25.0 x 10^-5/locus or about 110 x 10^-8/rad/locus for spermatogonia exposed to 215 rads of neutrons with 100 R of γ-contamination at a rate of 0.002 rads/min or less ( Table 10-7, row 26). These authors suggested that high linear-energy-transfer radiation (neutrons) may have a dose-rate effect opposite in direction from that of low linear-energy-transfer radiation (X-rays). Russell and Kelly ( 1964b) did not find such an effect, however.

Investigations of mutation rates in female germ cells have been less extensive. The rates for recessive visibles and lethals at seven specific loci in acutely irradiated fetal oogonia were about 10 to 14 x 10^-8/R/locus, slightly more than one-half the spermatogonial rate ( Table 10-8, rows 28 and 29). The mutation rates in oogonia also turned out to be dependent upon the dose rate. When 500 R of X rays were given at a rate of 0.009 R/min, the specific locus rate was about 1.9 x 10^-8/R/locus, whereas at 0.8 R/min it was 12 x 10^-8/R/locus, or about six times higher ( Table 10-9, row 31).

The estimation of somatic mutation rates presents special technical problems, chief of which is that breeding tests of the presumed mutant cells are ruled out. Russell and Major ( 1957) used the specific locus method (P-cross) to estimate the average mutation rate of four loci. C57BL females, mated with NB males, were irradiated on the 10.25 day of pregnancy. The zygotes should be a/ a b/+ c^ch p/+ + d se/+ + and the offspring should be self-colored black. Spots of color other than black anywhere in the coat may be taken to indicate a mutation at the b, c, d, or p locus. This method gave a somatic rate of 70 x 10^-8/R/locus for the same four loci exposed to the same amount of acute X-irradiation.

The preceding discussion has referred only to estimates of rates of mutation to recessive visible and lethal mutations in spermatogonia, postspermatogonial cells, and oogonia, obtained in experiments using the specific locus method. The chief value of the specific locus method lies in its usefulness in comparing mutation rates in different cell stages under exposures to various types of radiation given at various rates and at various total doses. The method has relatively little value in predicting overall mutation rates, nor is any other method fully satisfactory for doing so. Estimates of mutation rates to dominant visibles and dominant lethals, as steps toward estimating overall mutation rates, are given for a variety of types of cells and of radiation in Table 10-2 (row 3), Table 10-4 (rows 11 to 13) and Table 10-8 (row 23). The estimates work out to 17.3 x 10^-8/R/gamete for dominant visible mutations in spermatogonia exposed to acute X-rays and 16.7 x 10^-8/R/gamete in oogonia. In contrast Sugahara ( 1964) found a recessive lethal mutation rate of about 1.8 x 10^-4/R/gamete, both using chromic γ-irradiation. It is apparent that many more studies need to be done before we shall have a clear understanding of the mutability of mouse genes exposed to radiation.

Mutations at specific loci

There are several notable features of mutations induced at specific loci which are briefly mentioned here. Further information will be found in papers by Carter et al. ( 1958), Russell and Russell ( 1959b), Phillips ( 1961), L.B. Russell ( 1962), and W.L. Russell ( 1965).

First, in spermatogonia the induced mutation rates show a 27-fold variation among the seven loci studied by the specific locus method. In post-spermatogonial stages, however, the rates are much more uniform ( Table 10-10). Second, mutations recovered from irradiated spermatogonia are mostly point mutations, whereas those from later stages contain a high proportion of chromosomal deficiencies. Specifically, only one deficiency at the closely linked d and se loci was found in the irradiated spermatogonia, whereas five were found in later spermatogenic stages and six in oocytes. Third, about 75 per cent of the induced mutations are lethal when homozygous, lethals have occurred at all loci, and the time of death is frequently after birth. Fourth, in heterozygotes the induced mutations, although recessive by the usual criteria of their effects on pigmentation, frequently affect body size and viability.

One question of considerable interest is not yet answerable. How do radiation-induced mutations in general compare, in the spectrum of phenotypic effects, with natural mutations? An answer to this question will require the same careful search for and analysis of spontaneous mutations that has been applied to radiation-induced mutations ( Green et al., 1965). The fact that lethal mutations at the seven loci used in the specific locus studies were not known at the time the radiation studies were started may mean that natural mutations are less likely to be lethal when homozygous or it may mean that the existing mutations are the viable survivors out of a very much larger group containing many lethals. Until better information is available, we must withhold judgment about the comparability or noncomparability of induced and natural mutations in mice.

EFFECTS IN POPULATIONS

The occurrence of radiation-induced mutations and chromosome aberrations implies that populations exposed to radiation will suffer deleterious effects. The magnitude of the effect following specified radiation exposures can only be predicted after we have more knowledge about the viability, fertility, and fecundity on nonirradiated descendants of irradiated ancestors. Several studies in which populations of mice or selected members thereof have been irradiated for one or more generations will be summarized in this section. For an introduction to the literature on other laboratory populations of mammals see Roderick ( 1964).

To understand the effects of ancestral irradiation on the genetic constitution of populations, it is necessary to control or deliberately vary the important variables: dose level and dose rate of irradiation, stage of germ cells irradiated, the sex irradiated, size of the breeding population, and genetic background of the population. The mouse is the only mammal for which a wide variety of genetically independent inbred strains and mutant-bearing stocks are available. The foundation populations of long-term experiments can be of wide variety, either hybrid or inbred, and each variety can be almost perfectly replicated genetically at another time or place.

The pioneer work of Snell and Hertwig (see review by Russell, 1954) showed that the reduced number of animals born from irradiated postmeiotic germ cells of the sire was primarily due to dominant mutations of the chromosomal type. From these and subsequent studies it is apparent that dominant lethals and serious postnatal deleterious effects are easily detected among the zygotes formed from irradiated sperm and oocytes, but are comparatively very uncommon in the zygotes formed from irradiated spermatogonia. In female mammals premeiotic germ cells are not present after birth. In population studies, therefore, it is important to know the extent of the male's contribution to successive generations by irradiated sperm and by irradiated spermatogonia.

The following discussion of experiments is arranged according to the germinal stage irradiated.

Irradiation of postgonial stages

In a study Kalmus et al. ( 1952) paired mice and allowed each pair to produce two litters. Then they X-rayed the gonads of males with 150 to 300 R and reconstituted the pairs. They found a reduction in litter size in the subsequent third litter compared to the second preirradiation litter. Also the sex ratio changed in the direction of expectation (fewer females among the offspring of irradiated males) in the third litter. The results are not clearly the consequence of irradiation because parity has effects on litter size and sex ratio.

Spalding et al. ( 1964 and earlier) have reported an irradiation experiment of many generations. All the animals originated from a single pair of inbred RF mice. Two sublines were begun from the pair, one of which received no irradiation and functioned as a control line. The males of the other line were given 200 rads of whole-body X-rays at 24 to 28 days of age and mated to unexposed sisters in each generation. Throughout the experiment mice were conceived as early as 25 days and as late as 120 days after irradiation. The earliest animals conceived were probably used for the parents of following generations, since the sublines average about four generations each year. Therefore, the postspermatogonial germinal stages at the time of irradiation probably constitute most of the sires' genetic contribution to subsequent generations. Differences between the lines were observed in the sixth generation. The irradiated line was found to be significantly lower than the control in weaning weight, litter size at birth and weaning, and length of survival under protracted γ-ray and fractionated X-ray exposures. The irradiated line also exhibited a greater number of litters eaten by their mothers, but no consistent differences were observed in sex ratio, weaned-to-born ratio, or lifespan.

In their 1964 study they also showed values of some of these characteristics for a third line, an irradiated subline derived from the irradiated line but propagated in parallel for a few generations without irradiation. In those traits showing significant differences between irradiated and control lines, the values for the irradiated subline were not intermediate; sometimes the values fell beyond those of the control and sometimes beyond those of the irradiated line. One would expect the irradiated subline to be intermediate or at least consistently nearer one line or the other if the differences between the irradiated and control lines were due solely to the amount of ancestral irradiation.

Russell ( 1957) investigated the lifespan of offspring of (101 x C3H)F₁ males exposed to neutrons from a nuclear detonation. The range of exposure was 0 to 186 rep, and the offspring were conceived in the presterile period. A general life shortening of 0.61 days for each rep was calculated from the data over the entire range of doses. There was no evidence of nonlinearity of this mortality over the dose range.

In a similar study Spalding ( 1964) investigated the lifespan of male RF mice exposed to both fission neutrons and γ-rays and the lifespan of their first and second generation offspring. He exposed five groups of males to various doses of whole-body fission neutrons ranging from 32 to 117 rads and another group to whole-body Co⁶⁰ γ-rays over a range of 60 to 300 rads. A third major group, a control population of males, received no irradiation. To obtain offspring from both pre- and postspermatogonial stages during the time of irradiation, he mated the males immediately after irradiation and also 16 weeks after irradiation to two different samples of RF females. He was unable to demonstrate any shortening of the life or any effect on the sex ratio.

Proshina ( 1961) studied the fitness of nonpedigreed white mice over five generations after a series of graded doses, 50 to 300 R of X-rays, were given to 400 males of the initial generation. We do not know the time between irradiation and mating of the animals nor the germinal stage irradiated. She found more deaths in the first month of life in the descendants of the irradiated males than in the control group. This effect was apparent in all five generations. The weights of 1-month-old mice were also less than in the controls in all generations, and developmental anomalies were noted in the descendants of the irradiated males in the first two generations.

From the summary of the experiment of Lindop and Rotblat ( 1963), it is not clear what stages of spermatogenesis at the time of irradiation contributed to subsequent generations. Offspring from male mice irradiated with a whole-body dose of 350 R of X-rays at 4 weeks of age were compared with offspring of nonirradiated males. They found no effect of three generations of such parental exposure on lifespan or radiation sensitivity.

Continuous irradiation exposure

In some studies, populations of mice were either left continuously in a radiation field or males were periodically removed for irradiation and then replaced with their nonirradiated mates. Fertilizing gametes, therefore, could come from all possible stages of gametogenesis at the time of irradiation.

The early studies of Charles et al. (reported in 1960 and 1961) were on males of strain DBA exposed to radiation of the gonads and mated to nonirradiated females of strain C57BL. The gonadal doses were 0.1, 0.5, 1, and 10 R/day of X-rays given in a period of 1 to 1.5 hours each day except Sunday. The males remained with the females throughout the period except for the short time each day necessary for the irradiation. Therefore, part and probably most of the fertilizing sperm were irradiated in postspermatogonial stages. Charles et al. found increased juvenile mortality in the offspring of the males receiving 10 R/day or those receiving a total dose of 100 R or more. The effect was significantly expressed in larger litters but only suggestive in smaller litters. These investigators thoroughly autopsied all of the male offspring of the irradiated and control sires. Examining mice in such detail in a radiation-genetics experiment has never been attempted elsewhere. They observed a small but significant increase in rare external and internal morphological abnormalities among the offspring of the irradiated males. The female offspring were mated to an albino tester stock for the analysis of partial sterility. There were a greater number of partially sterile females whose fathers had been irradiated. About one-fifth of these females were tested for another generation and proved to be semisterile; that is, their partial sterility could be explained as a consequence of a reciprocal translocation induced by irradiation ( Chapter 7).

Charles et al. were unable to find any differences between the groups in sex ratio at weaning. In the male offspring of the irradiated and control sires they found no differences in body weight, tail length, body length, strength, and number of Peyer's patches on the small intestines. In the female offspring they found no difference in age at time of their fourth litters nor in total offspring born in their first four litters.

In a long-term investigation Gowen and Stadler ( 1964 and earlier) studied the fitness of 10 inbred strains exposed to continuous Co⁶⁰ γ-rays over many generations. They exposed mice to doses ranging from 0.06 to 0.10 R/hour given during 22-hour days. Throughout the study the strains were propagated by brother-sister matings. Mice receiving 3.0 R or more per day were sterile. Five strains did not survive five generations, and the other five survived 12 generations. At least one strain reached the 24th generation. Gowen and Stadler believed that irradiation probably was not a significant factor in extinguishing the five lines because these strains were less vigorous at the outset of the experiment. The surviving strains appeared healthy and did not appear to differ from their respective nonirradiated controls in first-litter productivity or in sex ratio. Lifespan data were collected on all the surviving strains over all the generations and analyzed through the sixth generation. Although there was a significant decline in lifespan with succeeding generations, the authors said the decline could not be explained as a result of the increased number of generations of ancestral irradiation. The average correlation of lifespan and ancestral irradiation within strains and within generations was -0.17. These correlations differed between strains but not between generations.

In another long-term experiment Touchberry and Verley ( 1964) studied the effects of whole-body X-rays given for successive generations to several hybrid populations derived from B6DF₁/J and several inbred populations derived from C57BL/6J. Males of the initial populations were divided into five groups and given 0, 50, 100, 150, and 200 R, respectively. In the next generation, matings were made at random between the irradiated groups within the inbred or within the hybrid populations. Mice were between 60 and 85 days old at the time of irradiation and were mated 14 days after irradiation. They were permitted to raise two litters, from which parents of the subsequent generation were chosen randomly.

Data on six generations of the hybrid populations and four generations of the inbred populations were used for the analysis. In both the inbred and hybrid populations they found a decrease in litter size with an increase in parental irradiation. Surprisingly they also found an increase in 32-day body weight and in growth rate from 2 to 38 days with an increase in parental irradiation. Touchberry and Verley believed the decrease in litter size to be evidence of induced dominant mutations and the increase in 32-day body weight and growth rate to be evidence of heterotic effects of the induced mutations. They pointed out that maternal irradiation produced both genetic and somatic effects, whereas paternal irradiation produced genetic effects only. The supposition of a heterotic effect seems less probable than the possibility that 32-day weight and growth rate are negatively correlated with and partially dependent upon litter size.

Sugahara ( 1964) and his colleagues raised and bred two populations of a tester stock of mice for three generations in a field of Co⁶⁰ γ-rays. In the first experiment the mice were given about 5 to 8 R/22-hour day and in the second experiment 0.43 R/22-hour day. In a third experiment they gave acute doses of 600 R of X-rays to the gonads of males of another population of the same stock. In a fourth experiment they irradiated the adult males of one generation with 597.3 R of Co⁶⁰ γ-rays given at a dose rate of 9.3 R/22-hour day. Nonirradiated but genetically similar mice were raised and maintained in a similar manner to serve as control populations. The fourth generation in the first two experiments and the second generation in the third and fourth experiments, all nonirradiated, were crossed to strain CBA, according to Haldane's ( 1956, 1957) plan. From the F₁ population, siblings were intercrossed to produce an F₂ generation. In addition to the data on suspected lethals induced in these populations (mentioned in the first part of this chapter), they collected data on mean litter size, incidence of sterile pairs, sex ratio, and juvenile death (death before weaning). Only for juvenile death was there a clear effect of ancestral irradiation in the second, third, and fourth experiments. This effect was found in the F₁ and F₂ generations following the outcross.

To determine whether irradiation would produce subline differentiation Searle ( 1964a) studied the effect of continuous Co⁶⁰ γ-rays on 28 pairs of sublines of the inbred strain C3H/HeH. One subline of each pair was exposed to 1R/night with a total dose of about 80 R/generation, whereas the other subline was set aside as a nonirradiated control. All sublines were propagated by brother-sister matings. Data on fitness characteristics were analyzed through 24 generations. Searle found a significant reduction in litter size in the irradiated sublines particularly apparent with greater parity, but he attributed the effect to damage of irradiated somatic tissue and not to mutations in irradiated ancestral germ plasm. No differences were found between the irradiated and control lines in complete sterility. Searle studied the genetic divergence of two pairs of sublines by comparing the lines with respect to 23 quasi-continuous characteristics and four metrical characters of the skeletal system. The unexpected result was that the rate of divergence was somewhat greater for the nonirradiated lines than for the irradiated lines.

Irradiation of spermatogonia

Dominant mutations or chromosomal aberrations can be induced by irradiating spermatogonia, and occasionally such changes do reach zygotes and are passed on to subsequent generations ( Griffen, 1964; Lyon et al., 1964). However, they are far fewer than can be found among zygotes descended from irradiated postspermatogonial stages. There is apparently no period after irradiation when the germ line can be assumed to be free from any kind of genetic damage. Irradiation will apparently have some effect on the fitness of populations, no matter what stage of gametogenesis is irradiated. The problem still remains, however, to show clearly in what ways population fitness is damaged by irradiating spermatogonia. The effect of spermatogonial irradiation on population fitness is important, too, because it is this type of irradiation damage which may be most important to a human population.

Ehling and Randolph ( 1962) reported an experiment in which strain 101 males were exposed to different sources and conditions of radiation and then crossed to C3H females so that both postspermatogonial and spermatogonial stages could be studied independently with respect to their effects on skeletal abnormalities in the F₁ animals. The progeny were conceived during the presterile period after their fathers were irradiated with 600 rads of X-rays or 80 to 200 rads of neutrons with varying intensities. Males mated after the poststerile period had received 600 rads. The litters from the exposed sires were killed for examination between 26 and 28 days of age. Skeletons of 618 offspring from nonirradiated sires and 515 offspring of irradiated sires were examined. The authors classified the abnormalities as either of two types depending upon their frequency: class I abnormalities were those occurring only once in all the animals studied, and class II abnormalities were those occurring more than once. No differences between irradiated and control groups were found in frequencies of class II abnormalities, nor in class I abnormalities when progeny were derived from irradiated spermatogonial cells. A significant difference, however, was discovered between the irradiated and control groups in class I abnormalities when the sires' contribution was derived from irradiated postgonial cells. The authors concluded that class I abnormalities were probably due to genetic effects and class I to nongenetic effects.

In studies of general fitness Kohn ( 1960) irradiated CAF₁ males with X-rays over a range of 0 to 720 rads. During the poststerile period surviving males were mated to nonirradiated BALB/cCrgl females and the offspring were studied for the characteristics: males per litter, females per litter, progeny per litter, sex ratio at birth, and sex ratio at weaning. Regression lines were fitted to these characteristics plotted against the sires' irradiation doses, but the slopes were not significant. Thus no effect of irradiation was demonstrated. Samples of the progeny were also studied for lifespan and tolerance to 500 rads of X-rays per week, but no significant effect of paternal irradiation was observed ( Kohn et al., 1965).

However, Russell and Russell ( 1959a) reported a reduction of mean litter size at 3 weeks of age among offspring derived from spermatogonia given 300 R of X-rays in a single generation. The lower mean litter size at 3 weeks was consistently found in all 10 groups of animals studied over different periods of time.

In further studies of fertility Carter and Lyon ( 1961) gave 600 R of whole-body X-irradiation to (101 x C3H)F₁ males. The males were mated 6 weeks after irradiation to females of the same F₁. The females were allowed to bear four litters and were then dissected during their fifth pregnancies for counts of corpora lutea, moles, and live and dead fetuses. The males from the liveborn litters were crossbred to detect a possible reduction of their fertility, and their mates were dissected during their pregnancies for similar counts. The same males were then mated to females of strain 101 to produce offspring to be backcrossed to the males to get an indication of the induction of recessive lethals and visibles. There was no evidence of reduced fertility in the F₁ males. There was also no difference found in litter size at birth when the F₁ males were crossed to strain 101 females. Nor was there any difference in the counts of corpora lutea, number of implantations, moles, or live or dead embryos. But from backcross matings the granddaughters of the irradiated males had a significant reduction of corpora lutea per pregnancy over that in the control granddaughters, but no differences were found in number born per litter. The authors concluded that there is a compensating mechanism tending to equalize the litter sizes despite differences in the number of eggs ovulated.

Later experiments at Harwell have considerably clarified the mechanisms involved in reduced fertility following spermatogonial irradiation. Lyon et al. ( 1964) irradiated the males of a cross of strains C3H/HeH and 101/H with an acute gonadal dose of 1,200 R of X-rays in two fractions 8 weeks apart. Twelve weeks after the latter dose, the F₁ males were outcrossed to CBA/H females. Controls were propagated and manipulated in the same manner but with no irradiation. The sizes of the litters of the irradiated sires were reduced 15 per cent at birth. The decrease in litter size was mostly due to increased embryonic death just after implantation. Part of this lethality was due to semisterility caused by the induction of reciprocal translocations at a rate of 3.3 per cent per gamete. Subsequent crosses also revealed a higher incidence of recessive lethal mutations induced in the irradiated group. No differences were found in sex ratios.

In the same experiment Searle ( 1964b) outcrossed the F₁ daughters and recorded the sizes of their first four litters. Their fifth litters were observed at the 14-day fetal stage. The mean litter size was 5.5 per cent lower in the daughters of the irradiated males. This reduction in fertility was almost wholly attributable to semisterility, present in 6.7 per cent of the daughters of the irradiated males. Searle estimated the rate of induction of translocations in spermatogonia to be 6.4 x 10^-5/R/genome. This rate is about one-sixth that of spermatozoa; he suggested that spermatogonial stages are less sensitive to irradiation.

In a related study, Philips and Searle ( 1964) looked for a dose-rate effect by spreading the 1,200 R over 12 weeks at 100 R/week. The lethal effect was reduced to a 2 per cent decrease in litter size, and the rate of induction of translocations to about 1 per cent. Neither of these effects was significant when compared with the control values, but the fact that they were far less than the effects observed after an acute dose of 1,200 R is evidence for a dose-rate effect.

In long-term studies Lüning and Sheridan ( 1964 and earlier) irradiated the spermatogonia of males of a randomly mated population from strain CBA with 276 R of X-rays each generation and compared their descendants with those of a parallel nonirradiated control population. The radiation was delivered to the gonads of the males with the rest of the body shielded. In studying the reproductive performance of the fourth to sixth generations under irradiation, they found a consistent reduction of litter size in the irradiated group independent of the number of litters the females produced. The females of the irradiated group had a slightly greater frequency of losses of complete litters. Lüning and Sheridan also found a decrease in the number of corpora lutea in the first litters of the fourth generation in the irradiated group. These findings suggested that damage to this population is through a lowered maternal fitness. There was no change in the sex ratio because of irradiation. Further studies of offspring of sib and nonsib matings of these lines indicated that the major dominant effects of irradiation were on the immediate offspring of the irradiated males and not on future generations. In other words, there was no evidence that dominant detrimental effects accumulated with increasing generations of irradiation. However recessive lethals apparently did accumulate with successive generations of irradiation.

In other long-term studies Green and his colleague ( Green, et al., 1964; Roderick and Schlager, 1965, personal communication) propagated both genetically heterogeneous and genetically homogeneous populations of mice under different levels of irradiation for several generations. The genetically heterogeneous populations were all descended from a 4-way cross of genetically independent inbred strains. The genetically homogeneous populations were all descended from a single pair of mice from strain C57BL/10Gn. The two major populations were subdivided at the outset of the experiment into several different subpopulations and propagated in parallel with different amounts of inbreeding in order to study the possible effect of inbreeding in combination with the effect of irradiation. All groups were further subdivided so that different levels of irradiation could be delivered to populations of each breeding size. Irradiation doses were 0, 50, and 100 R per generation. In the data collected from the mice in the first and second litters used to propagate the main lines, no differences were found in age of dam at parturition, number born, number weaned, sex ratio, weaned-to-born ratio, or frequency of phenotypic deviations attributable to differences in irradiation dose or breeding size of the population.

Samples of nonirradiated males from the 11th generation were mated to siblings as well as to females from strain C57BL/6J for a study of prenatal mortality. The females were dissected before parturition, and data were taken on number of corpora lutea, moles, dead fetuses, and living fetuses. No differences in prenatal mortality between the populations were found which could be attributed to ancestral radiation or inbreeding level.

Every three generations some nonirradiated animals of each dose and breeding size group were set aside to determine their lifespan reproductive performance. In these studies a significant reduction in number of litters was found in the genetically homogeneous populations after nine generations, and this difference could be attributed to the ancestral radiation exposure. The mean litter sizes, however, did not differ. These findings suggested that the animals with ancestral irradiation perform just as well through most of their lives, but that they deteriorate sooner physiologically. The authors point out the value of studies of lifetime performance, because these differences would not have been observed on data of first and second litters only. The effect was not observed in the heterogeneous populations after nine generations, and no effects could be attributed to the inbreeding levels of the subpopulations.

Green and Les ( 1964) tried to bring about genetic extinction in a small hybrid mouse population by irradiating the spermatogonia (body shielded) with a dose of 900 R of X-rays in each generation. After nine generations they found no differences in fertility, number of litters, number of offspring born and weaned, and average litter size between the irradiated and nonirradiated control populations. It is not clear why this study and that of Kohn ( 1960) failed to show the reduced litter size reported in other studies with comparable radiation doses.

Conclusions

Table 10-11 gives an abridged summary of the methods and results of the papers discussed in this section of the chapter. Characteristics such as viability and fertility are the principal components of population fitness. It appears that reduction in fitness of mouse populations as measured in these characteristics can be shown to be a result of irradiation of doses as low as 100 to 300 R per generation. Furthermore, the data suggest that decreased fitness, like chromosome abnormalities, can be more easily brought about and the damage more easily passed on to subsequent generation as a result of postspermatogonial irradiation than of spermatogonial irradiation. Genetically heterogeneous populations of mice may be more resistant than inbred populations to radiation-induced damage to fitness. Selection probably eliminates many mutations in each generation and thus reduces the damage to population fitness. There may be mechanisms of physiological compensation, such as that reported by Carter and Lyon ( 1961), ameliorating the outward phenotypic manifestations of a fitness decrement. Certain fitness traits seem to express the genetic damage from irradiation better than others in different experimental designs and under different conditions. Sex ratios apparently do not change in mouse populations under the irradiation and conditions studied, even though it is the only characteristic which showed an apparent change after irradiation of a human population ( Neel, 1963).

Several experiments, chiefly at the Oak Ridge National Laboratory and at the MRC Radiobiological Research Unit at Harwell, have yielded estimates of induced mutation rates in various types of germinal cells of mice exposed to X-rays, γ-rays, and neutrons. These experiments have revealed that mutation rates depend on many factors including: sex and cell stage irradiated, type of mutation scored, genotype or strain, and kind, amount, and dose of radiation.

A number of other experiments have been initiated by the desire to predict the effect of irradiating one or more generations on the fitness of a population. Measures of fitness include fertility, numbers and sizes of litters, reproductive life, and lifespan. No clear generalization emerges from the experiments so far reported, except that irradiation of postgonial stages appears to be more detrimental to fitness than irradiation of gonial stages. This does not necessarily mean that gonial stages are less sensitive to radiation damage; possibly the damage is compensated by selection in favor of nonmutated cells.

Auerbach, C., D.S. Falconer, and J.H. Isaacson, 1962. Test for sex-linked lethals in irradiated mice. Genet. Res. 3: 444-447.
See also MGI.

Batchelor, A.L., R.J.S. Phillips, and A.G. Searle. 1964. High effectiveness of chronic neutron exposures for the induction of specific locus mutations in mice. Nature 201: 207-208.
See also PubMed.

Carter, T.C. 1957a. Recessive lethal mutation induced in the mouse by chronic γ-irradiation. Proc. Roy. Soc. B 147: 402-411.
See also PubMed.

Carter, T.C. 1957b. The use of linked marker genes for detecting recessive autosomal lethals in the mouse. J. Genet. 55: 585-595.

Carter, T.C. 1958. Radiation-induced gene mutation in adult female and foetal male mice. Brit J. Radiol. 31: 407-411.
See also PubMed.

Carter, T.C. 1959. A pilot experiment with mice, using Haldane's method for detecting induced autosomal recessive lethal genes. J. Genet. 56: 353-362.

Carter, T.C. 1960. Mutation induced in germ cells of the foetal female mouse. Genet. Res. 1: 59-61.

Carter, T.C., and M.F. Lyon. 1961. An attempt to estimate the induction by X-rays of recessive lethal and visible mutations in mice. Genet. Res. 2: 296-305.

Carter, T.C., M.F. Lyon, and R.J.S. Phillips. 1956. Induction of mutations in mice by chronic gamma irradiation; interim report. Brit J. Radiol. 29: 106-108.
See also PubMed.

Carter, T.C., M.F. Lyon, and R.J.S. Phillips. 1958. Genetic hazard of ionizing radiation. Nature 182: 409.
See also PubMed.

Carter, T.C., and R.J.S. Phillips. 1952. An experimental attempt to investigate the induction of visible mutations in mice by chronic gamma irradiation, p. 73-81. In A. Haddow [ed.] Biological Hazards of Atomic Energy. Clarendon Press, Oxford.

Charles, D.R. 1950. Radiation-induced mutations in mammals. Radiology 55: 579-581.
See also PubMed.

Charles, D.R., J.A. Tihen, E.M. Otis, and A.B. Grobman. 1960. Genetic Effects of Chronic X-irradiation Exposure in Mice. Univ. of Rochester, Atomic Energy Project Rep., UR-565. 354 p.
See also PubMed.

Charles, D.R., J.A. Tihen, E.M. Otis, and A.B. Grobman. 1961. Genetic effects of chronic X-irradiation exposure in mice. Genetics 46: 5-8.
See also PubMed.

Ehling, U.H., and M.L. Randolph. 1962. Skeletal abnormalities in the F₁ generation of mice exposed to ionizing radiations. Genetics 47: 1543-1555.
See also PubMed.

Falconer, D.S. 1949. The estimation of mutation rates from incompletely tested gametes, and the detection of mutations in mammals. J. Genet. 49: 226-234.

Gowen, J.W., and J. Stadler. 1964. Lifespans of mice as affected by continuing irradiation from Cobalt-60 accumulated ancestrally and under direct irradiation. Genetics 50: 1115-1142.
See also MGI.

Green, E.L., and E.P. Les. 1964. Persistence of populations of mice exposed to substerilizing doses of spermatogonial X-irradiation. Genetics 50: 497-507.
See also PubMed.

Green, E.L., T.H. Roderick, and G. Schlager. 1964. Embryonic mortality in mouse populations after eleven generations of spermatogonial irradiation. Genetics 50: 1053-1063.
See also PubMed.

Green, E.L., G. Schlager, and M.M. Dickie. 1965. Natural mutation rates in the house mouse: plan of study and preliminary estimates. Mutation Res. 2: 457-465.
See also PubMed.

Griffen, A.B. 1964. The occurrence of chromosomal aberrations in pre-spermatocytic cells of irradiated male mice. II. Cytological studies of sterile and semi-sterile F₁ individuals, p. 175-188. In W.D. Carlson and F.X. Gassner [ed.] Proc. Int. Symp. Effects of Ionizing Radiation on the Reproductive System. Pergamon Press, New York.

Haldane, J.B.S. 1956. The detection of autosomal lethals in mice induced by mutagenic agents. J. Genet. 54: 327-342.

Haldane, J.B.S. 1957. Methods for the detection and enumeration of mutations produced by irradiation in mice, p. 35-39. Proc. Int. Genet. Symp. 1956. Science Council of Japan, Ueno Park, Tokyo.

Kalmus, H., J.D. Metrakos, and M. Silverberg. 1952. Sex ratio of offspring from irradiated male mice. Science 116: 274-275.
See also PubMed.

Kohn, H.I. 1960. The effect of paternal X-ray exposure on the secondary sex ratio in mice (F₁ generation). Genetics 45: 771-778.
See also PubMed.

Kohn, H.I., M.L. Epling, P.H. Guttman, and D.W. Bailey. 1965. Effect of paternal (spermatogonial) X-ray exposure in the mouse: Lifespan, X-ray tolerance, and tumor incidence of the progeny. Radiat. Res. 25: 423-434.
See also PubMed.

Lindop, P.J., and J. Rotblat. 1963. Study of the offspring of irradiated mice. Summarized by E.C. Amoroso, p. 408-429. In M. Ebert and A. Howard [ed.] Radiation Effects in Physics, Chemistry and Biology. North-Holland, Amsterdam.

Little, C.C., and H.J. Bagg. 1923. The occurrence of two heritable types of abnormality among the descendants of X-rayed mice. Amer. J. Roentgenol. Radiat. Therap. 10: 975-989.
See also MGI.

Lüning, K.G., and W. Sheridan. 1964. Dominant effects on productivity in offspring of irradiated mouse populations. Genetics 50: 1043-1052.

Lyon, M.F. 1960. Effect of X-rays on the mutation of t-alleles in the mouse. Heredity 14: 247-252.

Lyon, M.F., R.J.S. Phillips, and A.G. Searle. 1964. The overall rates of dominant and recessive lethal and visible mutation induced by spermatogonial X-irradiation of mice. Genet. Res. 5: 448-467.
See also MGI.

Neel, J.V. 1963. Changing Perspectives on the Genetic Effects of Radiation. Charles C. Thomas, Springfield, Ill. 97 p.

Phillips, R.J.S. 1961. A companion of mutation induced by acute X and chronic gamma irradiation in mice. Brit. J. Radiol. 34: 261-264.

Phillips, R.J.S., and A.G. Searle. 1964. The effect of dose-rate on the yield of translocations and dominant lethals following spermatogonial irradiation of mice. Genet. Res. 5: 468-472.

Proshina, A.D. 1961. Comparative characteristics of the fertility of five generations of mice from irradiated males. [Transl. from Russian] Med. Radiol. 6: 41-48.
See also PubMed.

Roderick, T.H. [ed.] 1964. Proceedings of the symposium: The effects of radiation on the hereditary fitness of mammalian populations. Genetics 50: 1019-1217.

Russell, L.B. 1962. Chromosome aberrations in experimental mammals. Prog. in Med. Genetics 2: 230-294.

Russell, L.B., and M.H. Major. 1957. Radiation-induced presumed somatic mutations in the house mouse. Genetics 42: 161-175.
See also PubMed.

Russell, L.B., and W.L. Russell. 1956. The sensitivity of different stages in oogenesis to the radiation induction of dominant lethals and other changes in the mouse, p. 187-192. In J.S. Mitchell, B.E. Holmes, and C.L. Smith [ed.] Progress in Radiobiology. Proc. 4th Int. Conf. Radiobiol., Cambridge, August, 1955. Oliver and Boyd, London.

Russell, W.L. 1951. X-ray-induced mutations in mice. Cold Spring Harbor Symp. Quant. Biol. 16: 327-336.
See also PubMed.

Russell, W.L. 1954. Genetic effects of radiation in mammals, p. 825-859. In A. Hollender [ed.] Radiation biology. Vol. 1, Part 2. McGraw-Hill, New York.

Russell, W.L. 1956. Radiation-induced mutations in the mouse, p. 25. US AEC Report ORNL-2155, Oak Ridge National Laboratory.

Russell, W.L. 1957. Shortening of life in the offspring of male mice exposed to neutron radiation from an atomic bomb. Proc. Nat. Acad. Sci. 43: 324-329.
See also PubMed.

Russell, W.L. 1962. An augmenting effect of dose fractionation on radiation-induced mutation rate in mice. Proc. Nat. Acad. Sci. 48: 1724-1724.
See also PubMed.

Russell, W.L. 1963. The effect of radiation dose rate and fractionation on mutation in mice, p. 205-217. In F. Sobels [ed.] Repair from Genetic Radiation Damage. Pergamon Press, New York.

Russell, W.L. 1965. Evidence from mice concerning the nature of the mutation process, p. 257-264. In S.J. Geerts [ed.] Proc. XI Int. Congr. Genet. (The Hague, 1963) Vol. 2. Pergamon Press, New York.
See also MGI.

Russell, W.L., J. W. Bangham, and J.S. Gower. 1958. Comparison between mutations induced in spermatogonial and postspermatogonial stages in the mouse. Proc. X Int. Congr. Genet. 2: 245-246. (Abstr.)

Russell, W.L., and E.M. Kelly. 1964a. Relative biological effectiveness of neutron to X-ray induction of gene mutations in mouse postspermatogonial stages. Semiannual Prog. Rep., Biol. Div., Oak Ridge Nat. Lab. 3601: 83-85.

Russell, W.L., and E.M. Kelly. 1964b. Neutron-induced mutation in mouse spermatogonia. Lack of effect of dose rate. Relative biological effectiveness of neutrons. Semiannual Prog. Rep., Biol. Div., Oak Ridge Nat. Lab. 3700: 83-85.

Russell, W.L., and L.B. Russell. 1959a. Radiation-induced genetic damage in mice, p. 179-188. In J.G. Bugher, J. Coursaget, and J.F. Loutit [ed.] Prog. in Nuclear Energy, Ser. 6, Biol. Sci. Pergamon Press, New York.

Russell, W.L., and L.B. Russell. 1959b. The genetic and phenotypic characteristics of radiation-induced mutations in mice. Radiat. Res., Suppl. 1: 296-305.

Russell, W.L., L.B. Russell, and E.M. Kelly. 1960. Dependence of mutation rate on radiation intensity, p. 311-320. In: Immediate and Low Level Effects of Ionizing Radiations: Conference held in Venice, June, 1959. Int. J. Radiat. Biol. Spec. Suppl.

Searle, A.G. 1964a. Effects of low-level irradiation on fitness and skeletal variation in an inbred mouse strain. Genetics 50: 1159-1178.
See also PubMed.

Searle, A.G. 1964b. Genetic effects of spermatogonial X-irradiation on productivity of F₁ female mice. Mutation Res. 1: 99-108.
See also PubMed.

Searle, A.G., and R.J.S. Phillips, 1964. Genetic effects of neutron irradiation in mice, p. 361-370. In Biological Effects of Neutron and Proton Irradiation. Vol 1. Int. Atomic Energy Agency, Vienna.

Snell, G.D. 1935. The induction by X-rays of hereditary changes in mice. Genetics 20: 545-567.
See also PubMed.

Snell, G.D. 1939. The induction by irradiation with neutrons of hereditary changes in mice. Proc. Nat. Acad. Sci. 25: 11-14.
See also PubMed.

Snell, G.D. 1945. The detection of mutations. Relative efficiency of various systems of brother-sister inbreeding in mice. J. Hered. 36: 275-278.

Snell, G.D., and P.C. Aebersold. 1937. The production of sterility in male mice by irradiating with neutrons. Proc. Nat. Acad. Sci. 23: 374-378.
See also PubMed.

Spalding, J.F. 1964. Longevity of first and second generation offspring from male mice exposed to fission neutrons and gamma rays, p. 147-152. In W.D. Carlson and F.X. Gassner [ed.] Proc. Int. Symp. Effects of Ionizing Radiation on the Reproductive System. Pergamon Press, New York.

Spalding, J.F., M. Brooks, and P. McWilliams. 1964. Observations on lifespan, radioresistance, and productivity in offspring from 5 to 25 generations of X-irradiated male mice. Genetics 50: 1179-1186.
See also MGI.

Sugahara, T. 1964. Genetic effects of chronic irradiation given to mice through three successive generations. Genetics 50: 1143-1158.
See also PubMed.

Touchberry, R.W., and F.A. Verley. 1964. Some effects of an inbred and a hybrid population of mice. Genetics 50: 1187-1209.

Woolf, B. 1954. Estimation of mutation rates. I. Visible recessive characters detected in inbred lines maintained by single sib-matings. J. Genet. 52: 332-353.