Ionizing radiation and genetic risks. VII. The concept of mutation component and its use in risk estimation for Mendelian diseases

Consideration of hereditary effects in the radiological protection system: evolution and current status

PURPOSE

The purpose of this paper is to provide an overview of the methodology used to estimate radiation genetic risks and quantify the risk of hereditary effects as outlined in the ICRP Publication 103. It aims to highlight the historical background and development of the doubling dose method for estimating radiation-related genetic risks and its continued use in radiological protection frameworks.

RESULTS

This article emphasizes the complexity associated with quantifying the risk of hereditary effects caused by radiation exposure and highlights the need for further clarification and explanation of the calculation method. As scientific knowledge in radiation sciences and human genetics continues to advance in relation to a number of factors including stability of disease frequency, selection pressures, and epigenetic changes, the characterization and quantification of genetic effects still remains a major issue for the radiological protection system of the International Commission on Radiological Protection.

CONCLUSION

Further research and advancements in this field are crucial for enhancing our understanding and addressing the complexities involved in assessing and managing the risks associated with hereditary effects of radiation.

Genome-based, mechanism-driven computational modeling of risks of ionizing radiation: The next frontier in genetic risk estimation?

Research activity in the field of estimation of genetic risks of ionizing radiation to human populations started in the late 1940s and now appears to be passing through a plateau phase. This paper provides a background to the concepts, findings and methods of risk estimation that guided the field through the period of its growth to the beginning of the 21st century. It draws attention to several key facts: (a) thus far, genetic risk estimates have been made indirectly using mutation data collected in mouse radiation studies; (b) important uncertainties and unsolved problems remain, one notable example being that we still do not know the sensitivity of human female germ cells to radiation-induced mutations; and (c) the concept that dominated the field thus far, namely, that radiation exposures to germ cells can result in single gene diseases in the descendants of those exposed has been replaced by the concept that radiation exposure can cause DNA deletions, often involving more than one gene. Genetic risk estimation now encompasses work devoted to studies on DNA deletions induced in human germ cells, their expected frequencies, and phenotypes and associated clinical consequences in the progeny. We argue that the time is ripe to embark on a human genome-based, mechanism-driven, computational modeling of genetic risks of ionizing radiation, and we present a provisional framework for catalyzing research in the field in the 21st century.

Estimation of the hereditary risks of exposure to ionizing radiation: history, current status, and emerging perspectives

This paper provides a brief overview of the advances in the field of the estimation of the genetic risks of exposure of human populations to ionizing radiation from the early 1950's to the present and of the developments that are anticipated in the coming years. The latter are based on the view that the insights gained from human genetics, especially human molecular genetics, will be increasingly applied to address problems in risk estimation. Owing to the paucity of human data on radiation-induced mutations, mouse data on radiation-induced mutations are used to predict the risk of genetic diseases in humans using the doubling dose method. With this method, the risk per unit dose is expressed as a product of three quantities, i.e., P x 1/DD x MC where P is the baseline frequency of genetic diseases, 1/DD (the relative mutation risk per unit dose; DD refers to the doubling dose, i.e., the radiation dose required to produce as many mutations as those that occur spontaneously in a generation) and MC is the disease class-specific mutation component (a measure of the relative increase in disease frequency per unit relative increase in mutation rate). The five important changes that are now introduced in genetic risk estimation include (1) an upward revision of the baseline frequency of Mendelian diseases to 2.4% (from 1.25% used until the early 1990's); (2) a reversion to the conceptual basis for DD calculations used in the 1972 BEIR report of the U.S. National Academy of Sciences, namely, the use of human data on spontaneous mutation rates and mouse data on induced mutation rates (instead of the use of mouse data for both these rates as has been the case from mid-1970's until the early 1990's); (3) the fuller development and use of the MC concept for predicting the responsiveness of Mendelian and multifactorial diseases to increases in mutation rate; (4) the introduction of a new disease-class-specific quantity called the "potential recoverability correction factor" or PRCF in the risk equation to bridge the gap between the rates of induced mutations in mice and the risk of inducible genetic diseases in humans; and (5) the introduction of the concept that multisystem developmental abnormalities are likely to be among the principal phenotypes of radiation induced genetic damage in humans. All these advances now permit, for the first time in 40 y, the estimation of risks for all classes of genetic diseases. For a population exposed to low-LET, chronic or low-dose irradiation, the risks predicted for the first generation progeny are the following (all estimates are per million live born progeny per gray of parental irradiation): autosomal dominant and x-linked diseases, approximately 750 to 1,500 cases; autosomal recessive, nearly zero; chronic multifactorial diseases, approximately 250 to 1,200 cases; and congenital abnormalities, approximately 2000 cases. The total risk per gray is of the order of approximately 3,000 to 4,700 cases, which represent approximately 0.4 to 0.6% of the baseline frequency of these diseases (738,000 per million) in the population. The advances anticipated in the coming years are likely to permit the estimation of genetic risks of radiation with greater precision than is now possible.

Ionizing radiation and genetic risks. XIII. Summary and synthesis of papers VI to XII and estimates of genetic risks in the year 2000

This paper recapitulates the advances in the field of genetic risk estimation that have occurred during the past decade and using them as a basis, presents revised estimates of genetic risks of exposure to radiation. The advances include: (i) an upward revision of the estimates of incidence for Mendelian diseases (2.4% now versus 1.25% in 1993); (ii) the introduction of a conceptual change for calculating doubling doses; (iii) the elaboration of methods to estimate the mutation component (i.e. the relative increase in disease frequency per unit relative increase in mutation rate) and the use of the estimates obtained through these methods for assessing the impact of induced mutations on the incidence of Mendelian and chronic multifactorial diseases; (iv) the introduction of an additional factor called the "potential recoverability correction factor" in the risk equation to bridge the gap between radiation-induced mutations that have been recovered in mice and the risk of radiation-inducible genetic disease in human live births and (v) the introduction of the concept that the adverse effects of radiation-induced genetic damage are likely to be manifest predominantly as multi-system developmental abnormalities in the progeny. For all classes of genetic disease (except congenital abnormalities), the estimates of risk have been obtained using a doubling dose of 1 Gy. For a population exposed to low LET, chronic/ low dose irradiation, the current estimates for the first generation progeny are the following (all estimates per million live born progeny per Gy of parental irradiation): autosomal dominant and X-linked diseases, approximately 750-1500 cases; autosomal recessive, nearly zero and chronic multifactorial diseases, approximately 250-1200 cases. For congenital abnormalities, the estimate is approximately 2000 cases and is based on mouse data on developmental abnormalities. The total risk per Gy is of the order of approximately 3000-4700 cases which represent approximately 0.4-0.6% of the baseline frequency of these diseases (738,000 per million) in the population.

Ionizing radiation and genetic risks. XII. The concept of "potential recoverability correction factor" (PRCF) and its use for predicting the risk of radiation-inducible genetic disease in human live births

Genetic risks of radiation exposure of humans are generally expressed as expected increases in the frequencies of genetic diseases over those that occur naturally in the population as a result of spontaneous mutations. Since human data on radiation-induced germ cell mutations and genetic diseases remain scanty, the rates derived from the induced frequencies of mutations in mouse genes are used for this purpose. Such an extrapolation from mouse data to the risk of genetic diseases will be valid only if the average rates of inducible mutations in human genes of interest and the average rates of induced mutations in mice are similar. Advances in knowledge of human genetic diseases and in molecular studies of radiation-induced mutations in experimental systems now question the validity of the above extrapolation. In fact, they (i) support the view that only in a limited number of genes in the human genome, induced mutations may be compatible with viability and hence recoverable in live births and (ii) suggest that the average rate of induced mutations in human genes of interest from the disease point of view will be lower than that assumed from mouse results. Since, at present, there is no alternative to the use of mouse data on induced mutation rates, there is a need to bridge the gap between these and the risk of potentially inducible genetic diseases in human live births. In this paper, we advance the concept of what we refer to here as "the potential recoverability correction factor" (PRCF) to bridge the above gap in risk estimation and present a method to estimate PRCF. In developing the concept of PRCF, we first used the available information on radiation-induced mutations recovered in experimental studies to define some criteria for assessing potential recoverability of induced mutations and then applied these to human genes on a gene-by-gene basis. The analysis permitted us to estimate unweighted PRCFs (i.e. the fraction of genes among the total studied that might contribute to recoverable induced mutations) and weighted PRCFs (i.e. PRCFs weighted by the incidences of the respective diseases). The estimates are: 0.15 (weighted) to 0.30 (unweighted) for autosomal dominant and X-linked diseases and 0.02 (weighted) to 0.09 (unweighted) for chronic multifactorial diseases. The PRCF calculations are unnecessary for autosomal recessive diseases since the risks projected for the first few generations even without using PRCFs are already very small. For congenital abnormalities, PRCFs cannot be reliably estimated. With the incorporation of PRCF into the equation used for predicting risk, the risk per unit dose becomes the product of four quantities (risk per unit dose=Px(1/DD)xMCxPRCF) where P is the baseline frequency of the genetic disease, 1/DD is the relative mutation risk per unit dose, MC is the mutation component and PRCF is the disease-class-specific potential recoverability correction factor instead of the first three (as has been the case thus far). Since PRCF is a fraction, it is obvious that the estimate of risk obtained with the revised risk equation will be smaller than previously calculated values.

Ionizing radiation and genetic risks. XI. The doubling dose estimates from the mid-1950s to the present and the conceptual change to the use of human data on spontaneous mutation rates and mouse data on induced mutation rates for doubling dose calculations

This paper provides an overview of the concept of doubling dose, changes in the database employed for calculating it over the past 30 years and recent advances in this area. The doubling dose is estimated as a ratio of the average rates of spontaneous and induced mutations in a defined set of genes. The reciprocal of the doubling dose is the relative mutation risk per unit dose and is one of the quantities used in estimating genetic risks of radiation exposures. Most of the doubling dose estimates used thus far have been based on mouse data on spontaneous and induced rates of mutations. Initially restricted to mutations in defined genes (with particular focus on the seven genes at which induced recessive mutations were studied in different laboratories), the doubling dose concept was subsequently expanded to include other endpoints of genetic damage. At least during the past 20 years, the magnitude of the doubling dose has remained unchanged at approximately 1 Gy for chronic low LET radiation exposures. One of the assumptions underlying the use of the doubling dose based on mouse data for predicting genetic risks in humans, namely, that the spontaneous rates of mutations in mouse and human genes are similar, is incorrect; this is because of the fact that, unlike in the mouse, the mutation rate in humans differs between the two sexes (being higher in males than in females) and increases with paternal age. Further, an additional source of uncertainty in spontaneous mutation rate estimates in mice has been uncovered. This is related to the non-inclusion of mutations which arise as germinal mosaics and which result in clusters of identical mutations in the following generation. In view of these reasons, it is suggested that a prudent way forward is to revert to the use of human data on spontaneous mutation rates and mouse data on induced mutation rates for doubling dose calculations as was first done in the 1972 BEIR report of the US National Academy of Sciences. The advantages of this procedure are the following: (i) estimates of spontaneous mutation rates in humans, which are usually presented as sex-averaged rates, automatically include sex differences and paternal age-effects; (ii) since human geneticists count all mutations that arise anew irrespective of whether they are part of a cluster or not, had clusters occurred, they would have been included in mutation rate calculations and (iii) one stays close to the aim of risk estimation, namely, estimation of the risk of genetic diseases in humans. On the basis of detailed analyses of the pertinent data, it is now estimated that the average spontaneous mutation rate of human genes (n=135 genes) is: (2.95+/-0.64)x10(-6) per gene and the average induced mutation rate of mouse genes (n=34) is: (0.36+/-0.10)x10(-5) per gene per Gy for chronic low LET radiation. The resultant doubling dose is (0.82+/-0.29) Gy. The standard error of the doubling dose estimate incorporates sampling variability across loci for estimates of spontaneous and induced mutation rates as well as variability in induced mutation rates in individual mouse experiments on radiation-induced mutations. We suggest the use of a rounded doubling dose value of 1 Gy for estimating genetic risks of radiation. Although this value is the same as that used previously, its conceptual basis is different and the present estimate is based on more extensive data than has so far been the case.

Ionizing radiation and genetic risks. X. The potential "disease phenotypes" of radiation-induced genetic damage in humans: perspectives from human molecular biology and radiation genetics

Estimates of genetic risks of radiation exposure of humans are traditionally expressed as expected increases in the frequencies of genetic diseases (single-gene, chromosomal and multifactorial) over and above those of naturally-occurring ones in the population. An important assumption in expressing risks in this manner is that gonadal radiation exposures can cause an increase in the frequency of mutations and that this would result in an increase in the frequency of genetic diseases under study. However, despite compelling evidence for radiation-induced mutations in experimental systems, no increases in the frequencies of genetic diseases of concern or other adverse effects (i.e., those which are not formally classified as genetic diseases), have been found in human studies involving parents who have sustained radiation exposures. The known differences between spontaneous mutations that underlie naturally-occurring single-gene diseases and radiation-induced mutations studied in experimental systems now permit us to address and resolve these issues to some extent. The fact that spontaneous mutations (among which are point mutations and DNA deletions generally restricted to the gene) originate through a number of different mechanisms and that the latter are intimately related to the DNA organization of the genes, are now well-documented. Further, spontaneous mutations include those that cause diseases through loss of function as well as gain of function of genes. In contrast, most radiation-induced mutations studied in experimental systems (although identified through the phenotypes of the marker genes) are predominantly multigene deletions which cause loss of function; the recoverability of an induced deletion in a livebirth seems dependent on whether the gene and the genomic region in which it is located can tolerate heterozygosity for the deletion and yet be compatible with viability. In retrospect, the successful mutation test systems (such as the mouse specific locus test) used in radiation studies have involved genes which are non-essential for survival and are also located in genomic regions, likewise non-essential for survival. In contrast, most of the human genes at which induced mutations have been looked for, do not seem to have these attributes. The inference therefore is that the failure to find induced germline mutations in humans is not due to the resistance of human genes to induced mutations but due to the structural and functional constraints associated with their recoverability in livebirths. Since the risk of inducible genetic diseases in humans is estimated using rates of "recovered" mutations in mice, there is a need to introduce appropriate correction factors to bridge the gap between these rates and the rates at which mutations causing diseases are potentially recoverable in humans. Since the whole genome is the "target" for radiation-induced genetic damage, the failure to find increases in the frequencies of specific single-gene diseases of societal concern does not imply that there are no genetic risks of radiation exposures: the problem lies in delineating the phenotypes of recoverable genetic damage that are recognizable in livebirths. Data from studies of naturally-occurring microdeletion syndromes in humans and those from mouse radiation studies are instructive in this regard. They (i) support the view that growth retardation, mental retardation and multisystem developmental abnormalities are likely to be among the quantitatively more important adverse effects of radiation-induced genetic damage than mutations in a few selected genes and (ii) underscore the need to expand the focus in risk estimation from known genetic diseases (as has been the case thus far) to include these induced adverse developmental effects although most of these are not formally classified as "genetic diseases". (ABSTRACT TRUNCATED)

Ionizing radiation and genetic risks IX. Estimates of the frequencies of mendelian diseases and spontaneous mutation rates in human populations: a 1998 perspective

This paper is focused on baseline frequencies of mendelian diseases and the conceptual basis for calculating doubling doses both of which are relevant for the doubling dose method of estimating genetic risks of exposure of human populations to ionizing radiation. With this method, the risk per unit dose is obtained as a product of three quantities, namely, the baseline frequency of the disease class under consideration, the relative mutation risk (which is the reciprocal of the doubling dose, which in turn, is calculated as a ratio of spontaneous and induction rates of mutations) and mutation component, i.e., the responsiveness of the disease class to an increase in mutation rate. The estimates of baseline frequencies of mendelian diseases that are currently used in risk estimation date back to the late 1970s. Advances in human genetics during the past two decades now permit an upward revision of these estimates. The revised estimates are 150 per 10(4) livebirths for autosomal dominants (from the earlier estimate of 95 per 10(4)), 75 per 10(4) livebirths for autosomal recessives (from 25 per 10(4)) and to 15 per 10(4) livebirths for X-linked diseases (from 5 per 10(4)). The revised total frequency of mendelian diseases is thus 240 per 10(4) livebirths and is about twice the earlier figure of 125 per 10(4) livebirths. All these estimates, however, pertain primarily to Western European and Western European-derived populations. The fact that in several population isolates or ethnic groups, some of these diseases (especially the autosomal recessives) are more common as a result of founder effects and/or genetic drift is well known and many more recent examples have come to light. These data are reviewed and illustrated with data from studies of the Ashkenazi Jewish, Finnish, French Canadian, Afrikaner and some other populations to highlight the need for caution in extrapolating radiation risks between populations. The doubling dose of 1 Gy that has been used for the past 20 years for risk estimation is based on mouse data for both spontaneous and induction rates of mutations. In extrapolating the mouse-data-based doubling dose to humans, it is assumed that the spontaneous rates in mice and humans are similar. This assumption is incorrect because of the fact that in humans, for several well-studied mendelian diseases, the mutation rate differs between the two sexes and it increases with paternal age. In estimates of spontaneous mutation rates in humans (which represent averages over both sexes), however, paternal age effects are automatically incorporated. In the mouse, these effects are expected to be much less (if they exist at all), but the problem has not been specifically addressed. The complexities and uncertainties associated with assessing the potential impact of spontaneous mutations which arise as germinal mosaics (and which can result in clusters of mutations in the following generation) on mutation rate estimates (in the mouse) and on mutation rate estimates and disease frequencies (in humans) are discussed. In view of (i) the lack of comparability of spontaneous mutation rates in mice and humans and (ii) the fact that these estimates for human genes already include both paternal age effects and correction for clusters (if they had occurred), it is suggested that a prudent procedure now is to base doubling dose calculations on spontaneous mutation rates of human genes (and induction rates of mouse genes, in the absence of a better alternative). This concept, however, is not new and was used by the US National Academy's Committee on the Biological Effects of Ionizing Radiation in its 1972 report.