The Selby-Russell Dispute Regarding the Nonreporting of Critical Data in the Mega-Mouse Experiments of Drs William and Liane Russell That Spanned Many Decades: What Happened, Current Status, and Some Ramifications.

The Russells began their studies of the hereditary effects of radiation in the late 1940s, and their experiments contributed much to what is known about the induction of gene mutations in mice. I had a close association with them for about 26 years, and they relied on me considerably for database management and statistical support. In 1994, I was shocked to discover that, in experiments on males, they had failed to report numerous spontaneous mutations that arose during the perigametic interval and were detected as clusters of mutations. I realized that their nondisclosure of this information meant that the decades-long application of their data to estimate hereditary risks of radiation to humans using the doubling-dose approach had resulted in a several-fold overestimation of risk. I accordingly reported the situation to funding agencies. The resulting complicated situation is referred to here as the Selby-Russell Dispute. Highlights of the resulting investigation, as well as what occurred afterward, are described, and reasons will be provided to show why, in my opinion, the hereditary risk from radiation in humans was likely overestimated by at least 10-fold because the Russells decided not to report critical information from their massive experiments.

Major impacts of gonadal mosaicism on hereditary risk estimation, origin of hereditary diseases, and evolution.

The specific-locus test in mice is by far the most extensively applied method for precisely defining gene mutation frequencies in mammals. Computer simulations of control experiments involving 57.4 million offspring, based on vast amounts of historical data, show that because of gonadal mosaicism, the total frequency of spontaneous mutations per generation is much higher than has been thought. The estimated combined spontaneous mutation frequency for both sexes for the seven genes tested in specific-locus experiments is 39.6 x 10(-5) mutation/gamete. Division of this frequency by the combined induced mutation frequencies in parents of both results in an estimate of the doubling-dose (DD) of from 5.4 to 7.7 Gy. For decades, the DD has been thought to be about 1 Gy. As the DD increases, estimates of hereditary risk that are based upon it decrease. Thus, one important ramification of this new understanding is that estimates of the hereditary risk to humans from radiation, commonly made by the doubling-dose (DD) approach, are probably at least five times too high. It also appears that gonadal mosaicism is likely to play a much more important role both in evolution and the origin of hereditary diseases than has been appreciated in the past.

Discovery of numerous clusters of spontaneous mutations in the specific-locus test in mice necessitates major increases in estimates of doubling doses.

Both the precision with which mutations can be quickly identified and the extensive application of the method make results from specific-locus experiments in mice especially important for estimating the doubling dose, which is the radiation exposure that induces a mutation frequency equal to the total spontaneous mutation frequency per generation. Because of gonadal mosaicism and the mechanism by which it occurs, the frequency with which new spontaneous mutations occur per generation is much higher than has been thought. While it will be some time before many of the newly-apparent uncertainties related to understanding this phenomenon can be resolved, consideration of what is known suggests that it would already be reasonable to raise the doubling dose from 1 to 5 Gy for low-dose-rate exposures to X and gamma radiation. Doing so would reduce risk estimates made by the doubling-dose method fivefold. Because the doubling dose for chemical mutagens is also calculated by division of the total spontaneous mutation frequency per generation by the induced mutation frequency per unit of chemical exposure, hereditary risks for chemicals have also been considerably overestimated if they are based on specific-locus data.

The thick tail mutation contains anomalies of the axial skeleton.

Analysis of the skeletal effects of thick tail (Tht), a radiation-induced mutation, has revealed numerous anomalies in the axial skeleton. The affected regions include the atlantal-occipital region as well as the lumbar (Lu) and caudal (Ca) vertebrae in which the ossified adult structures are either missing or reduced in size. Skeletons of juvenile Tht heterozygotes exhibit a malformed occipital bone, atlas, smaller Ca vertebrae, and delayed ossification of the affected adult structures. The diminished amount of cartilage and bone suggests that the Tht gene may be functioning during the formation of these tissues.

Mouse clavicular development: analysis of wild-type and cleidocranial dysplasia mutant mice.

Cleidocranial dysplasia (CCD) is an autosomal dominant disease characterized by hypoplasia or aplasia of clavicles, open fontanelles, and other skeletal anomalies. A mouse mutant, shown by clinical and radiographic analysis to be strikingly similar to the human disorder and designated Ccd, was used as a model for the human disorder. Since malformation of the clavicle is the hallmark of CCD, we studied clavicular development in wild-type and Ccd mice. Histology and in situ hybridization experiments were performed to compare the temporal and spatial expression of several genes in wild-type and Ccd mutant mouse embryos. Bone and cartilage specific markers--type I, II, and X collagens, Sox9, aggrecan, and osteopontin were used as probes. The analyses covered the development of the clavicle from the initial mesenchymal condensation at embryonic day 13 (E13) to the late mineralization stage at embryonic day 15.5. At day 13.5, cells in the center of the condensation differentiate into characteristic precursor cells that were not observed in other bone anlagen. In the medial part of the anlage these cells express markers of the early cartilage lineage (type II collagen and Sox9), whereas cells of the lateral part express markers of the osteoblast lineage (type I collagen). With further development the medial cells differentiate into chondrocytes and start to express chondrocyte-specific markers such as aggrecan. Cells of the lateral part differentiate into osteoblasts as indicated by the production of bone matrix and the expression of osteopontin. At day 14.5 a regular growth plate has developed between the two parts where type X collagen expression can be demonstrated in hypertrophic chondrocytes. The data indicate that the medial part of the clavicle develops by endochondral bone formation while the lateral part ossifies as a membranous bone. The clavicle of Ccd mice showed a smaller band of mesenchymal cell condensation than in wild-type mice. Cells of the condensation failed to express type I and type II collagen at E13.5. In the lateral part of the clavicle type I collagen expression was not detected until E14.5 and osteopontin expression only appeared at E15.5. At E15.5, a small ossification center appears in the lateral part which is, in contrast to the wild-type clavicular bone, solid and without primary spongiosa as well as bone marrow. In the medial portion, type II collagen expression and endochondral ossification never occurs in Ccd mice; this portion of the clavicle is therefore missing in Ccd.

Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development.

We have generated Cbfa1-deficient mice. Homozygous mutants die of respiratory failure shortly after birth. Analysis of their skeletons revealed an absence of osteoblasts and bone. Heterozygous mice showed specific skeletal abnormalities that are characteristic of the human heritable skeletal disorder, cleidocranial dysplasia (CCD). These defects are also observed in a mouse Ccd mutant for this disease. The Cbfa1 gene was shown to be deleted in the Ccd mutation. Analysis of embryonic Cbfa1 expression using a lacZ reporter gene revealed strong expression at sites of bone formation prior to the earliest stages of ossification. Thus, the Cbfa1 gene is essential for osteoblast differentiation and bone formation, and the Cbfa1 heterozygous mouse is a paradigm for a human skeletal disorder.

Cyclophosphamide: review of its mutagenicity for an assessment of potential germ cell risks.

Cyclophosphamide (CP) is used to treat a wide range of neoplastic diseases as well as some non-malignant ones such as rheumatoid arthritis. It is also used as an immunosuppressive agent prior to organ transplantation. CP is, however, a known carcinogen in humans and produces secondary tumors. There is little absorption either orally or intravenously and 10% of the drug is excreted unchanged. CP is activated by hepatic mixed function oxidases and metabolites are delivered to neoplastic cells via the bloodstream. Phosphoramide mustard is thought to be the major anti-neoplastic metabolite of CP while acrolein, which is highly toxic and is produced in equimolar amounts, is thought to be responsible for most of the toxic side effects. DNA adducts have been formed after CP treatment in a variety of in vitro systems as well as in rats and mice using 3H-labeled CP. 32P-postlabeling techniques have also been used in mice. However, monitoring of adducts in humans has not yet been carried out. CP has also been shown to induce unscheduled DNA synthesis in a human cell line. CP has produced mutations in base-pair substituting strains of Salmonella tryphimurium in the presence of metabolic activation, but it has been shown to be negative in the E. coli chromotest. It has also been shown to be positive in Saccharomyces cerevisiae in D7 strain for many endpoints but negative in D62.M for aneuploidy/malsegregation. It has produced positive responses in Drosophila melanogaster for various endpoints and in Anopheles stephensi. In somatic cells, CP has been shown to produce gene mutations, chromosome aberrations, micronuclei and sister chromatid exchanges in a variety of cultured cells in the presence of metabolic activation as well as sister chromatid exchanges without metabolic activation. It has also produced chromosome damage and micronuclei in rats, mice and Chinese hamsters, and gene mutations in the mouse spot test and in the transgenic lacZ construct of Muta Mouse. Increases in chromosome damage and gene mutations have been found in the peripheral blood lymphocytes of nurses, pharmacists and female workers occupationally exposured to CP during its production or distribution. Chromosome aberrations, sister chromatid exchanges and gene mutations have been observed in somatic cells of patients treated therapeutically with CP. In general, there is a maximum dose and an optimum time for the detection of genetic effects because the toxicity associated with high doses of CP will affect cell division. In germ cells, CP has been shown to induce genetic damage in mice, rats and hamsters although the vast majority of such studies have used male mice.(ABSTRACT TRUNCATED AT 400 WORDS)

Molecular and phenotypic characterization of a new mouse insertional mutation that causes a defect in the distal vertebrae of the spine.

We have identified and characterized the phenotype of a new insertional mutation in one line of transgenic mice. Mice carrying this mutation, which we have designated TgN(Imusd)370Rpw, display undulations of the vertebrae giving rise to a novel kinky-tail phenotype. Molecular characterization of the insertion site indicates that the transgene integration has occurred without any substantial alterations in the structure of the host sequences. Using probes that flank the insertion site, we have mapped the mutation to chromosome 5 near the semidominant mutation, thick tail (Tht). Thick tail does not complement the TgN(Imusd)370Rpw mutation; compound mutants containing one copy of each mutation display a more severe phenotype than either mutation individually.

Synergistic interactions between two skeletal mutations in mice: individual and combined effects of the semidominants cleidocranial dysplasia (Ccd) and short digits (Dsh).

Heterozygotes for cleidocranial dysplasia (Ccd) and short digits (Dsh) were crossed to test whether synergistic interactions occur between different dominant mutations whose individual pleiotropic phenotypic effects exhibit a common feature. These unlinked mutations are homozygous lethal, and they are congenic on the C57BL/10 background. Each mutation caused more than 10 different anomalies and showed variable expressivity. Each mutation produced several malformations that were present in every heterozygote. Seven different synergistic interactions were found, including one that yielded an entirely new abnormality not predicted from any abnormalities found in either of the single heterozygotes. Although synergistic interactions between dominant mutations have not, to our knowledge, been described in humans, these findings in mice increase the probability that they occur in humans. Under certain circumstances in human populations, the segregation of mutations causing synergistic interactions of the type demonstrated might be confused with recessive inheritance. It will be important to learn whether synergistic interactions can occur between other mutations. If they can, it will probably become important to take synergistic interactions into account when estimating the genetic hazards to humans from mutagens. Three antagonistic interactions were also found.

Lifespan and autopsy findings in the first-generation offspring of X-irradiated male mice.

Male mice of the C3Hf strain were exposed to 600 R of acute X-rays and, along with unirradiated control males, mated with 101-strain females. The offspring of the treated males were all conceived more than 7 weeks after irradiation, thereby ensuring that they were derived from germ cells exposed as stem-cell spermatogonia. After weaning, the offspring were caged individually and allowed to live their normal lifespan. Tumors and other major pathological disorders were recorded at a careful post-mortem examination. The lesions encountered were typical of those characteristically seen in aging (101 x C3Hf)F1 mice. The results showed no significant differences in lifespan between experimentals and controls. This held true when allowance was made for littermate correlations and for other factors that might contribute to differences among litters. Likewise, there were no significant differences between experimentals and controls in the frequency, severity, or age distribution of neoplasms and other diseases.

Specific-locus experiments show that female mice exposed near the time of birth to low-LET ionizing radiation exhibit both a low mutational response and a dose-rate effect.

Female mice were exposed to 300 R of 73-93 R/min X-radiation either as fetuses at 18.5 d post conception (p.c.) or within 9 h after birth. Combining the similar results from these two groups yielded a specific-locus mutation frequency of 9.4 X 10(-8) mutation/locus/R, which is statistically significantly higher than the historical-control mutation frequency, but much lower than the rate obtained by irradiating mature and maturing oocytes in adults. Other females, exposed at 18.5 days p.c. to 300 R of 0.79 R/min gamma-radiation, yielded a mutation frequency that was statistically significantly lower than the frequency at high dose rates. The low-dose-rate group also had markedly higher fertility. It appears that the dose-rate effect for mutations induced near the time of birth may be more pronounced than that reported for mature and maturing oocytes of adults. A hypothesis sometimes advanced to explain low mutation frequencies recovered from cell populations that experience considerable radiation-induced cell killing is that there is selection against mutant cells. The reason for the relatively low mutational response following acute irradiation in our experiments is unknown; however, the finding of a dose-rate effect in these oocytes in the presence of only minor radiation-induced cell killing (as judged from fertility) makes it seem unlikely that selection was responsible for the low mutational response following acute exposure. Had selection been an important factor, the mutation frequency should have increased when oocyte killing was markedly reduced.

Molecular and genetic characterization of a radiation-induced structural rearrangement in mouse chromosome 2 causing mutations at the limb deformity and agouti loci.

Molecular characterization of mutations in the mouse, particularly those involving agent-induced major structural alterations, is proving to be useful for correlating the structure and expression of individual genes with their function in the whole organism. Here we present the characterization of a radiation-induced mutation that simultaneously generated distinct alleles of both the limb deformity (ld) and agouti (a) loci, two developmentally important regions of chromosome 2 normally separated by 20 centimorgans. Cytogenetic analysis revealed that an interstitial segment of chromosome 17 (17B- 17C; or, possibly, 17A2-17B) had been translocated into the distal end of chromosome 2, resulting in a smaller-than-normal chromosome 17 (designated 17del) and a larger form of chromosome 2 (designated 2(17). Additionally, a large interstitial segment of the 2(17) chromosome, immediately adjacent and proximal to the insertion site, did not match bands 2E4-2H1 at corresponding positions on a normal chromosome 2. Molecular analysis detected a DNA rearrangement in which a portion of the ld locus was joined to sequences normally tightly linked to the a locus. This result, along with the genetic and cytogenetic data, suggests that the alleles of ld and a in this radiation-induced mutation, designated ldIn2 and ajIn2, were associated with DNA breaks caused by an inversion of an interstitial segment in the 2(17) chromosome.

The 1986 and 1988 UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation) reports: findings and implications.

The United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) has published a substantive series of reports concerning sources, effects, and risks of ionizing radiation. This article summarizes the highlights and conclusions from the most recent 1986 and 1988 reports. The present annual per person effective dose equivalent for the world's population is about 3 mSv. The majority of this (2.4 mSv) comes from natural background, and 0.4 to 1 mSv is from medical exposures. Other sources contribute less than 0.02 mSv annually. The worldwide collective effective dose equivalent annually is between 13 and 16 million person-Sv. The Committee assessed the collective effective dose equivalent to the population of the northern hemisphere from the reactor accident at Chernobyl and concluded that this is about 600,000 person-Sv. The Committee also reviewed risk estimates for radiation carcinogenesis which included the new Japanese dosimetry at Hiroshima and Nagasaki. These data indicate that risk coefficient estimates for high doses and high dose rate low-LET radiation in the Japanese population are approximately 3-10% Sv-1, depending on the projection model utilized. The Committee also indicated that, in calculation of such risks at low doses and low dose rates, a risk-reduction factor in the range of 2-10 may be considered.

A strategy for fine-structure functional analysis of a 6- to 11-centimorgan region of mouse chromosome 7 by high-efficiency mutagenesis.

A refined functional map of a 6- to 11-centimorgan region surrounding the albino (c) locus in mouse chromosome 7 is being generated by N-ethyl-N-nitrosourea (EtNU) "saturation" mutagenesis of stem-cell spermatogonia. In the first phase of an experiment that will eventually test at least 3000 gametes, we screened 972 mutagenized gametes for the induction of both lethal and visible mutations with a two-cross breeding protocol. Thirteen mutations mapping within the limits of a segment corresponding to the cytologically visible Df(c Mod-2 sh-1)26DVT deletion were recovered. They represented three phenotypic groups: prenatal lethality (six mutations); a fitness/runting syndrome (three mutations, provisionally designated as fit variants); and a neurological/balance-defect abnormality (four mutations). Complementation analysis provided evidence for a true repeat mutation at the sh-1 (shaker-1) locus (for the neurological mutations) and another at the here defined fit-1 (fitness-1) locus. In addition, four complementation groups were defined by induced lethal mutations; the two other lethal mutations were each part of a cluster. The recovery of the repeat mutations suggests that the EtNU-induced mutation rate, estimated from specific-locus tests, should make it possible to achieve saturation mutagenesis of a chromosomal region. This experiment is providing basic logistical and statistical information on which to base strategies for expanding the functional map of larger segments of the mouse genome by experimental mutagenesis. It is also yielding additional mutations useful in dissecting the functional and molecular complexity of this segment of chromosome 7.

Animal model: skeletal anomalies in mice with cleidocranial dysplasia.

Cleidocranial dysplasia in mice, a radiation-induced skeletal mutation, showed striking homology with cleidocranial dysplasia in humans. Genetic studies indicated that the condition in mice is inherited as an autosomal dominant trait with variable expressivity and almost complete penetrance. The homozygous condition was lethal in utero. Radiographic and alcian blue/alizarin red S-stained whole-skeletal preparation studies were used to determine the extent, pattern, incidence, and distribution of skeletal abnormalities in heterozygous mice. Cleidocranial dysplasia in mice was characterized by variable clavicular hypoplasia, delayed closure of cranial fontanelles and sutures, and variable hypoplasia of pelvic bones, in particular ischiopubic rami. The gene symbol Ccd is proposed for the cleidocranial dysplasia mutation in mice and humans.

A rapid method for preparing high quality alizarin stained skeletons of adult mice.

A simple three-day technique is described for preparing completely cleared and high quality alizarin stained total skeletons of adult mice. Unfixed specimens are partially macerated during staining. Older specimens are heated for 15 min in 1% KOH. A heated solution of benzyl and ethyl alcohol, glycerin, and water is used for final clearing and hardening. This procedure requires about 10 min work per specimen and greatly simplifies preparation of stained and cleared skeletons of adult mice. Another technique, giving slightly better preparations, but requiring 11-14 days, is also described.

First-generation litter-size reduction following irradiation of spermatogonial stem cells in mice and its use in risk estimation.

Litter-size reduction (LSR) is a useful measure of part of the overall F1 radiation-induced damage. Extensive LSR data were obtained as a by-product of specific-locus experiments. Fourteen such experiments involving 158,490 F1 litters have been analyzed for the extent of LSR induced by x- or gamma-irradiation of spermatogonia. Litter sizes were compared between experimental and control groups at about 3 weeks after birth. In order to reduce variability, comparisons were made only with concurrent controls and between groups of litters having mothers of approximately the same age. At the high dose rate of 90 R/min, the LSRs showed a humped dose-response curve. There was a pronounced dose-rate effect, the mutational responses being much less at dose rates of 0.009 R/min and 0.001 R/min. It is estimated that if men were exposed to 1 R of radiation delivered at low linear energy transfer (low LET) and low dose rate, the number of deaths caused by induced dominant mutations among their children before late childhood would be about 19 per million live-born. This can be added to the earlier estimate of an approximately equal number of viable disorders in all body systems as based on dominant skeletal mutations. This gives a total estimate of induced dominant damage, but much of this addition represents death in very early embryonic life that would not be recognized in humans. The LSR data also permit the conclusion that only an extremely small proportion of serious radiation-induced genetic disorders among live-born humans would be expected to result from segmental aneuploidy.