The Role of Adaptive Responses Following Exposure to Ionizing Radiation

Ludwig E. Feinendegen, M.D.

National Institutes of Health

Department of Nuclear Medicine

Building 10, Room 1C401

10 Center Drive - MCS1180

Bethesda, MD 20892-1180

Tel: (301) 496-6455

Fax: (301) 496-0114



The definition of dose:

AR's express temporarily stimulated physiological protections of cells against concentration changes of potentially toxic agents. In toxicology as in life sciences, the amount of an agent given to, taken up by, or effective in, a biological system is simply termed dose. This is not so in radiation sciences, radiation biology, and radiation protection. Regarding ionizing radiation, "absorbed dose" is the conventional unit; it expresses energy absorbed per unit mass, i.e., dose is expressed as agent concentration. The SI unit of absorbed dose D is the gray (Gy) where 1 Gy = 1 J/kg (ICRU 1980). The difference between "dose" and "absorbed dose" is often not properly seen. This is not surprising, since the word dose is used in radiation sciences assuming that everyone knows what it means. The situation becomes even more obscure when biological effects in individual cells are related to low and very low values of absorbed dose in tissues. This needs clarification before radiation-induced AR's are addressed.

Tissue dose and cell-dose:

Ionizing radiation is absorbed in matter by the deposition of discrete amounts of energy along the track of charged particles throughout the exposed matter; the lower the dose the more spaced apart are these energy deposition events, also called hits, and vice versa. Depending on the type and energy of the radiation, these hits have their characteristic spectrum. The amount of energy absorbed per hit in a micromass of tissue is formally denoted specific energy z1 (ICRU 1983). Because cells are the elemental units of life, and the average mass of a typical mammalian cell is about 1 ng, it is justified to choose this mass to express the specific energy here as microdose or hit size (Feinendegen et al. 1994). The cell-dose is a multiple of hit sizes. The mean hit size, z1, is calculated from the hit frequency. The spectrum of hit sizes, and thus the mean hit size, is constant for a given type of radiation. For example, the z1 from 100 keV X-rays is about 1 mGy. For densely ionizing radiation z1 is larger, for example by a factor of 300 for alpha-particles.

The ratio of the number of hits of all sizes, NH, to the number of exposed micromasses, Ne, multiplied by the mean hit size, z1, is equal to D:

D = z1 ·NH/N (1)

Clearly, D does not directly reveal the value of cell-dose, {z1 NH}. For a given D, z1 and NH are reciprocally related to each other. Since z1 remains constant for a given type of radiation, only NH/Ne increases in proportion to D. At low values of D, single cells experience only single hits or none at all, i.e., the ratio NH/Ne is much lower than 1. With increasing values of D, all cells eventually experience multiple hits, i.e., the ratio NH/Ne becomes increasingly larger than 1. For a given radiation quality with it's known z1 , NH can be a more meaningful quantity than just energy in tissue of any mass (Feinendegen et al. 1994, 1995). The components of equation (1) are measured by microdosimetry techniques (ICRU 1983).

The meaning of dose rate:

The above approach to radiation dosimetry also explains the meaning of dose rate, absorbed dose per unit time, D/t. It has long been known that the response of irradiated tissues depends on the rate at which the dose is absorbed. Repair of damaged tissue takes time. Indeed, dose rate in the conventional term D/t can be converted to the average time interval between two consecutive hits per exposed cell. According to equation (1), during exposure to a defined radiation field the dose rate is (Feinendegen et al. 1985, 1994):

D/t = z1 ·(NH/Ne) ·(1/t)

or D/t = z1/(t· Ne/NH). (2)

The denominator (t · Ne/NH) is the mean time interval between two consecutive hits; it allows the hit cell to acutely respond and may be long enough for a second hit not to interfere. The meaning of dose rate is crucial for assessing an AR. For instance, a 100 keV X-ray background radiation of 1 mGy per year causes about 1 hit per cell per year, and 365 mGy per year cause about 1 hit per cell every day (Feinendegen, 1991, 1995). It is likely that an AR following the first hit will be altered by the second hit at the higher dose rate.

AR's to radiation:

Various types of radiation-induced AR's have been reported and reviewed (Sugahara et al. 1992; UNSCEAR 1994; Academie des Sciences 1997; Feinendegen et al. 1999). Several AR's are readily seen after very low cell-doses, even below 0.01 Gy. Whether these low-dose specific AR's are mainly triggered by hit cells or also by hits in cell free matrix is an open debate. Hit cells possibly affect also non-hit cells. These AR's include temporary activation of prevention, repair, and elimination of damage to cellular DNA. For instance, damage prevention by way of detoxification of reactive oxygen species, ROS, is induced in mouse bone marrow cells for up to about 10 hours after irradiation; radiation-induced repair of DNA damage in human lymphocytes may last for several days; radiation-induced immunogenic elimination of damaged cells in rodents appears to be activated for several weeks. Programmed cell death, apoptosis, is not easily seen after very low cell-doses; it is a short term reaction of irradiated cells with more severe DNA damage (Ohyama et al. 1998). Despite recent progress in unraveling the molecular biological and genetic complexities of some AR's to very low cell-doses, their immediate consequences to cancer risk are still elusive. Much literature, however, points to fixed DNA damage to contribute to cancer induction. It appears, therefore, justified to discuss these AR's in the context of radiation-induced cancer.

Dose-response of AR's:

The AR's, so far examined, are specific for low-doses; they decline with doses above about 0.1 - 0.2 Gy and eventually disappear with doses exceeding about 0.5 Gy (UNSCEAR 1994; Feinendegen et al. 1995, 1996, 1999). The lowest measured dose to induce an AR is about the average hit size from X- or rays (Feinendegen et al. 1982). Since mean hit sizes from densely ionizing radiation, such as alpha-particles, already have values of about 0.3 Gy, it is not surprising that they have not yet been described to induce low-dose specific AR's. They may, nevertheless, be predicted to occur by way of so-called clastogenic factors that are released by the irradiated cells and may affect neighboring non-irradiated cells. In fact, recent studies show both DNA damage induction and various gene activation in non-hit cells in response to clastogenic factors released by alpha-irradiated cells (Azzam et al. 1998).

Low-dose specific AR's appear physiological:

Since irradiation always causes a spectrum of hit sizes that is typical of the radiation quality, the mean hit size is used here for assessing the relation between cell-dose and AR's. Taking a single hit from 100 keV X-rays, the hit size is just about 1 mGy. This amounts to 6.24 keV/ng, or about 2 · 102 ionizations per cell average. Of those, some 80 % are ionized water molecules with the concomitant production of ROS. With a background radiation of 1 mGy assumed to be solely 100 keV X-rays, one cell average in exposed tissues would be hit once a year, or one out of 365 cells hit per day. It is, thus, most unlikely that natural radiation has triggered the evolutionary acquisition of any type of cellular protective response, especially not the low-dose specific AR's. Since AR's are observed after low doses of X-irradiation, the responding cell obviously recognizes a radiation-induced variation of an agent that is natural to the cell.

Effects of radiation and metabolic ROS:

Some low-dose specific AR's appear to be directly or indirectly linked to natural, metabolically produced ROS (Feinendegen et al. 1987). It has been estimated that the oxidative metabolism of an average normal mammalian cell produces about 109 ROS per day (Beckmanm et al. 1998). These are taken to lead to about 105 DNA alterations in terms of oxidative adducts per cell per day; yet, only about 0.1 fixed DNA damage is likely produced per cell per day (Pollycove et al. 1998). Against this level of metabolically induced DNA damage, the DNA damage from background radiation is comparatively very small: only about 1 in 365 cells is hit per day from a background X-ray dose of 1 mGy per year and suffers in all about 200 ionizations, mainly ROS, per hit. The rates of ROS production per cell average per day from metabolism and background radiation, then, yield a quotient of more than 109. The radiation-induced DNA alterations are approximately 105 times more effective than those from metabolically produced ROS in causing, for example, DNA double strand breaks; nevertheless, the totality of metabolic induced DNA damage in the total exposed cells still far outnumbers the corresponding damage from background radiation by a factor of about 106 (Pollycove et al. 1998). The corresponding ratio per hit cell, however, is in the order of 103 - 104; this is similar to the ratio of 103 for leukemia induction calculated from epidemiological data to occur from spontaneous sources and a single hit from X- or rays, per hemopoietic stem cell in vivo (Feinendegen et al. 1991, 1995).

Low-dose specific AR's appear to operate on ROS effects:

Normal cells control DNA damage by ROS detoxification, DNA damage repair and removal (Alberts et al. 1989; Hanawalt 1994; Friedberg et al. 1995; Feinendegen et al. 1995; Pollycove et al. 1998). This physiological DNA damage control system focuses on protecting genetic integrity of the involved cells and, thus, on organism survival that is constantly threatened by potentially toxic agents, such as metabolically produced cellular ROS. The huge ROS production rate normally fluctuates with cellular oxygen consumption, for example, in response to demand by work loads; this fluctuation triggers biochemical feed-back controls affecting the DNA damage control system. Radiation also boosts the cellular ROS concentration, and, thus, may induce the corresponding feedback controls leading to AR's. Whatever the mechanisms involved, low-dose specific AR's apparently operate on all components controlling DNA damage from metabolically produced ROS. The special role of metabolically produced ROS in the DNA damage control system suggests that they are physiological signal substances essential to cellular survival.

The dual low-dose effect; DNA damage reduction may outweigh induction:

In view of the above named relatively large contribution to DNA damage from constantly produced metabolic ROS and the very low hit incidence and oncogenic effect from background radiation, radiation-induced suppression of DNA damage from normal metabolism appears to outweigh radiation-induced DNA damage at low cell-doses (Feinendegen et al. 1995, 1996; Pollycove et al. 1998). Moreover, the rarity of a cellular hit from background radiation predicts that radiation-induced AR's do not contribute much to biological system stability. However, larger variations in background radiation may have important consequences. For instance: let background radiation be increased tenfold from 1 to 10 mGy per year, or from about 0.003 to about 0.03 tissue cells hit per day; assume the ensuing AR's to cause a 25% enhanced effectiveness of each principal component of the DNA damage-control system in the hit cells for 10 days; then, the daily produced DNA alterations from metabolic ROS in the total exposed cells would be reduced stepwise by 7.5 %, involving radical detoxification, DNA repair and damage removal. This would result in a 20 % fall in the production rate of fixed DNA damage, from 0.1 to 0.08 DNA mutation per average cell per day (Pollycove et al. 1998). On the other hand, the radiation-induced DNA damage per average cell per day is calculated to likely increase only by an infinitesimally small amount: the fraction of daily hit cells rises from 0.003 to 0.03; each cell experiences on average 1 mGy; this increases the radiation-induced fixed DNA damage from about 10-7 to 10-6 per average cell per day (Pollycove et al. 1998).


Low-dose specific AR's are different from high-dose cell response:

Obviously, mammalian cells command various types of low-dose specific AR's with different duration. It is unknown to what degree these AR's interact; however, they appear to express various lines of defense and protection against accumulation of DNA damage in tissues. The low-dose specific AR's not detectable after high cell-doses appear to belong to a DNA damage-control system which is more sensitive to agent interventions than those biochemical networks that are conventionally measured after cell-doses higher than about 0.2 Gy; they can induce besides DNA repair genetic instability and apoptosis. The latter responses may be regarded as more robust, belonging to a higher order cascade of reactions to acute DNA damage than the more subtle AR's exclusively seen at low cell-doses.

Maximal low-dose specific AR's are seen after single doses of 0.1 - 0.2 Gy X- or rays. The dose of 0.1 Gy of 100 keV X-rays causes on average about 100 hits in each ng, i.e., each cell, in the exposed tissue. With about 200 ionizations and ROS per hit, a total of 2 104 ionizations and ROS occur per cell within the short time of exposure, for instance, a minute. This, then, adds to the 109 ROS produced per average cell per day, or about 7 105 ROS per average cell per minute. Considering the roughly 105 times greater effectiveness of radiation-induced ionizations and ROS than metabolic ROS in causing cell and DNA damage, the sudden radiation-induced burst of some 2 104 radiation-induced ROS per cell may well be close to the tolerance of the physiological DNA damage control system. At higher doses, i.e. higher concentrations of radiation-induced ionizations and ROS, the control system obviously falters and eventually succumbs fully so that DNA damage will prevail (Feinendegen et al. 1995, 1996).

Genetic instability:

Another aspect addresses the question of radiation-induced genetic instability; it is seen after relatively high doses (Morgan et al. 1996; Meydan et al. 1998) and may be reduced by low-dose specific AR's (Suzuki et al. 1998). Genetic instability increases the rate of somatic mutations in the descendent cell population and contributes to carcinogenesis. On the other hand, the increased incidence of chromosomal aberrations in circulating lymphocytes in irradiated human beings is well known to decline with given half times for various aberration types. Such mutated cells in an adult organism may be lost by physiological differentiation or become more susceptible to being scavenged, for example, by apoptosis or a competent immune system. It appears reasonable to consider genetic instability also in terms of protection by augmenting the removal of damaged cells.

No evidence speaks in favor of a supralinear oncogenic cellular transformation induced by low doses of loosely ionizing radiation. If it occurs it may be so small as to escape detection. Indeed, the calculated probability of cancer induction per hit, as referred to above, is so small that a hit-induced enhancement at low doses would be drowned by the protective effect of low-dose specific AR's.

Programmed cell death, apoptosis:

Ionizing radiation also may induce removal of damaged cells from tissues through programmed cell death, apoptosis; in various radiosensitive cell lines, its incidence rises linearly with dose above about 0.1 Gy, with a slope of around 5 ¥ 10-3 per mGy (Ohyama et al. 1998). This value is high in comparison with the probability of cancer induction per mGy per cell in vivo, namely about 10-14, and may, therefore, be a major mechanism of protection against radiation-induced cancer. After X-irradiation with 0.5 Gy, mouse spleen cells have a maximum incidence of apoptosis at 4 hours with a return to normal values within about 20 hours thereafter (Fujita et al. 1998). Different cell lines which were all radioresistant at high doses, show an increased lethality with a maximum at about 0.2 - 0.3 Gy; this appears to be programmed cell death at low doses (Joiner et al. 1996). These data emphasize that apoptosis is a physiological response to DNA damage accumulation from whatever cause, in terms of what has been called "altruistic cell suicide" (Kondo 1993). Indeed, apoptosis can be considered a major protective response not restricted to low cell-doses and of a last resort preventing damaged cells from accumulating in tissues.


It is generally too early to answer this question. More research is needed. Current data, however, support the potential applicability of low-dose specific AR's for medical benefit. Envisaged benefits may include low-dose induced protection of normal tissue against higher doses in the course of radiotherapy, or against cell and DNA damage in certain chronic infectious diseases not directly involving an increase in cellular ROS. Especially, the stimulation of the cellular immune competence appears an interesting approach to treat all such conditions where an enhanced immune competence may be beneficial. Clinical studies for cancer therapy have been reported in Japan (Sakamoto et al. 1997).

Quite controversial is the clinical use of radon therapy. Despite it's densely ionizing alpha-radiation, radon has been popular for decades with beneficial subjective reports from patients suffering from painful chronic infectious and degenerative diseases. Some of these subjective reports could be objectively registered medically (Pratzel et al. 1997). A biological explanation is not yet available. Moreover, epidemiological studies indicate no significantly increased cancer incidences from radon exposure to low doses (Lubin et al. 1997). In fact, a large population cohort study has consistently shown a significant decrease in lung cancer in those counties of the USA where the natural radon concentrations in dwellings is slightly elevated (Cohen 1995). Also these results still await a biological explanation.

The above two types of potential benefits from low-dose radon exposure may eventually be explainable by the effectiveness of clastogenic factors that are released from cells with large hit sizes, such as from alpha-particles of radon. Non-hit cells in the neighborhood of alpha-irradiated cells have been demonstrated to respond with an increase in chromosomal aberrations; various genes were also activated, one of which relates to programmed cell death (Azzam et al. 1998). It needs to be seen to what degree "by-stander effects" from clastogenic factors may initiate AR's, for instance, involving the immune system (Soto 1997). The chemical nature of the clastogenic factors has not yet been identified. It would indeed be strange, if cells would respond to a toxic agent only by suffering damage without the activation of ubiquitous AR's in a dose dependent manner. AR's to clastogenic factors should be investigated.


Single low-dose irradiation:

Risk assessment after single irradiation assumes a linear relationship between absorbed dose D and the probability of detriment R, such as cancer, in the exposed tissues, with a being the constant of proportionality:

R = · D. (3)

Substituting for R the ratio Nq /Ne, with Nq being the number of induced cancers and Ne the number of exposed cells, and for D {z1 NH/Ne} using equation (1) and multiplying each side of the equation by Ne:

Nq = · z1 ·NH. (4)

This equation transforms the conventionally used dose-risk function into a hit-number-effectiveness function at the cellular level (Bond et al. 1995).

Because z1 in equation (4) is a constant for a given radiation quality, the term a alone expresses the biological response of the irradiated system over a certain range of NH.

On the cell level is a composite of the various probabilities that express the principal dual radiation effect of low doses, discussed above, i.e., the damaging and the protective effect in the exposed tissue. The carcinogenic radiation effect in vivo is assigned a constant probability pind per average hit, with a value of about 10-14 for X-or irradiation; it includes the constant probability of any protection per hit; the life time probability of "spontaneous" carcinogenesis in vivo, pspo, mainly from metabolic ROS, has a value of about 10-11 (Feinendegen et al. 1991, 1995). The effect of the low-dose specific AR's per average hit on reducing DNA damage is assumed to correspondingly apply to reducing the cancer incidence. The degree of this reduction is here expressed by the cumulative protection probability pprot per average hit, mainly against pspo . As explained above, low-dose specific AR's disappear with increasing D, or, more precisely, with the increasing NH in the exposed tissue. Moreover, AR's are only effective over a fraction of the time a cancer develops after an oncogenic cell transformation step. Lastly, AR's may also be triggered rather than suppressed by intercellular communication, otherwise they would not be observed so easily. Even if most of the confounding factors are only partly quantified for a given cell system, the final outcome of protection can be measured. Considering the dependence of AR's on D and additional needed amendments, the value of pprot is here given as {pprot(D)A}. Then, the radiation-induced Nq may be expressed by (Feinendegen et al. 1995, 1996):

Nq = [pind - {pprot(D)A} pspo ] ·NH. (5)

Substituting equation (5) for (4):

{ · z1 } = [pind - {pprot(D)A} pspo ],

or: = [pind - {pprot(D)A} pspo ] / z1 (6)

The value of can not be constant at low doses when they stimulate AR's to protect against DNA damage and cancer from metabolic ROS, and when this effect declines against pind with increasing D and fully disappears with cell-doses above about 0.5 Gy. According to this dose dependent change of AR's, approaches constancy with doses above about 0.1 - 0.2 Gy, as was shown epidemiologically and experimentally. With values of pind and pspo as given above, and a value of 10-3 for {pprot(D)A}, Nq in equation (5) would be zero, i.e., cancer induction would not be observed. For {pprot(D)A} larger than 10-3, a net reduction in cancer incidence would result. The present model on the basis of observed AR's argues against the linear-no-threshold hypothesis and, in fact, portrays what is commonly called a hormetic effect. This approach, although incomplete, offers a conceptual framework for investigating the probability of cancer from low doses, where epidemiological analyses are severely limited by the need for large populations.

Protracted low-dose irradiation:

The above example deals with the life long risk from a single low-dose irradiation. With repeated or protracted irradiations, the approach to risk assessment on the basis of the dual low-dose effect is principally similar. The constraint here is the mean time interval between two consecutive hits per cell. It may be long enough for AR's not to be altered by the second hit, as explained in the section" the meaning of dose rate" under Question 1. It is not known whether AR's can be amplified by a second hit shortly after the first one. An interval of 10 days appears sufficiently long to avoid interference from a second hit. Also, assuming a constant relationship between fixed DNA damage and cancer induction, the p-values and Nq in equation (5) may be expressed in terms of fixed DNA damage produced or reduced per time in the entirety of exposed cells from endogenous sources and radiation. This approach is applied to estimate the effect of increase of background X-irradiation from 1 to 10 mGy per year, in the section "The dual low-dose effect; DNA damage reduction may outweigh induction" under Question 1 (Pollycove et al. 1998). The result is a 20% reduction of fixed DNA damage from about 0.1 to about 0.08 per average cell per day, a hormetic effect. In fact, this estimate is supported by reports of decreased age-adjusted cancer mortality rates and decreased mortality rates observed in populations living in high background areas, by various studies on medical populations and on cultured cells (UNSCEAR 1994; Pollycove 1994, 1998; Azzam et al. 1996).


The above discourse shows the need for more research. Radiation is a weak carcinogen, and low-dose specific AR's to loosely ionizing radiation exist; they temporarily protect cells and apparently reduce the production rate of fixed DNA damage mainly from metabolic ROS; their impact on tissue function and late effects such as carcinogenesis, even if very suggestive, is still elusive. Projected studies should include low doses also from densely ionizing radiations; the resulting large hit sizes to relatively few single cells produce clastogenic factors; these may not only cause DNA damage, but also AR's in non-hit cells.

Low-dose specific AR's produce the dual effect of low-dose irradiation, in terms of reducing and inducing fixed DNA damage and, presumably, detrimental late tissue effects, including carcinogenesis. The answers to the emerging scientific questions are expected to influence risk assessment and radiation protection. The expected socio-economic and political consequences surely warrant concerted efforts.


Academie des Sciences, Institut de France Problems associated with the effects of low doses of ionizing radiations. Rapport de l'Academie des Sciences, No.38; Lavoisier, TecDoc, Paris, London, New York, 1995

Alberts B., Bray D., Lewis J., Raff M., Roberts K., Watson J.D. Eds. Molecular Biology of the Cell Garland Pub., New York, New York, 1989

Azzam E.I., de Toledo S.M., Raaphorst G.P., Mitchel R.E.J. Low-dose ionizing radiation decreases the frequency of neoplastic transformation to a level below the spontaneous rate in C3H 10T1/2 cells. Radiation Res. 146: 369-373 (1996)

Azzam E.L., de Toledo S.M., Gooding T., Little J.B. Intercellular communication is involved in the bystander regulation of gene expression in human cells exposed to very low fluence of alpha particles. Radiation Res. 150: 497-504 (1998)

Beckman K.D., Ames B.N. The free radical theory of aging matures.
Physiol. Rev. 78: 547-581 (1998)

Bond V.P., Benary V., Sondhaus C.A., Feinendegen L.E. The meaning of linear dose-response relations, made evident by use of absorbed dose to the cell. Health Phys. 68: 786-792 (1995)

Cohen B.L. Test of the linear no-threshold theory of radiation carcinogenesis for inhaled radon decay products. Health Phys. 68: 157-174 (1995)

Feinendegen L.E., Muehlensiepen H., Porschen W., Booz J. Acute non-stochastic effect of very low-dose whole-body exposure, a thymidine equivalent serum factor. Intern.J.Radiat.Biol. 41: 139-150 (1982)

Feinendegen LK.E., Booz J., Bond V.P., Sondhaus C.A. Microdosimetric approach to the analysis of cell responses at low dose and low dose rate. Rad.Prot.Dosim. 13: 299-306 (1985)

Feinendegen L.E., Muehlensiepen H., Bond V.P., Sondhaus C.A. Intracellular stimulation of biochemical control mechanisms by low-dose low-LET irradiation. Health Phys. 52: 663-669 (1987)

Feinendegen L.E. Radiation risk of tissue late effect, a net onsequence of probabilities of various cellular responses. Europ.J.Nucl.Med. 18: 740-751 (1991)

Feinendegen L.E., Bond V.P., Booz J. The quantification of physical events within tissue at low levels of exposure to ionizing radiation. ICRU-News 2: 9-13 (1994)

Feinendegen L.E., Loken M., Booz J., Muehlensiepen H., Sondhaus C.A., Bond V.P. Cellular mechanisms of protection and repair induced by radiation exposure and their consequences for cell system responses. Stem Cells 13(1): 7-20 (1995)

Feinendegen L.E., Bond, V.P., Sondhaus C.A., Muehlensiepen H. Radiation effects induced by low doses in complex tissue and their relation to cellular adaptive responses. Mutation Res. 358: 199-205 (1996)

Feinendegen L.E., Bond V.P., Sondhaus C.A., Altman K.I. Cellular signal adaptation with damage control at low doses versus the predominance of DNA damage at high doses. Compt. Rend. (in press) 1999

Friedberg E.C., Walker G.C., Siede W. DNA Repair and Mutagenesis. ASM Press, Washington D.C., 1995

Fujita K, Ohtomi M., Ohyama H., Yamada T. Biphasic induction of apoptosis in the spleen after fractionated exposure of mice to very low doses of ionizing radiation. In: Apoptosis, its Roles and Mechanisms. Ed.: Yamada T., Hashimoto Y. Business Center for Academic Societies Japan , Tokyo, 1998, 201-218.

Hanawalt P.C. DNA repair comes of age. Mutation Res. 336(2) 101-113 (1995)

ICRP (International Commission on Radiological Protection). Recommendations of the International Commission on Radiological Protection. ICRP Publication 60. Annals of the ICRP 21 (1-3), Pergamon Press Oxford, 1991

ICRU (International Commission on Radiation Units and Measurements). Radiation Quantities and Units. ICRU, Bethesda, MD, USA, Report 33 (1980)

ICRU (International Commission on Radiation Units and Measurements). Microdosimetry. ICRU, Bethesda, MD, USA, Report 36 (1983)

Joiner M.C., Lambin P., Malaise E.P., Robson T., Arrand J.E., Skov K.A., Marples B. Hypersensitivity to very low single radiation doses: Its relationship to the adaptive response and induced radioresistance. Mutation Res. 358: 171-183 (1996)

Kondo S. Health Effects of Low Level Radiation. Kinki Univ.Press, Osaka, Japan; Medical Physics Publishing, Madison, WI, 1993.

Lubin J.H., Boice J.D. Lung cancer risk from residential radon: Meta-analysis of eight epidemiological studies. J.Natl.Cancer Inst. 89: 49-57 (1997)

Meydan D., Hellgren D., Lambert B. Variations in the frequency of the T-cell receptor beta/gamma-interlocus recombination in long-term cultures of non-irradiated and X- and gamma-irradiated human lymphocytes. Intern.J.Radiat.Biol. 74: 697-703 (1998)

Morgan W.F., Day J.P., Kaplan M.I., McGhee E.M., Limoli C.L. Genomic instability induced by ionizing radiation. Rad.Res. 146: 247-258 (1996)

Ohyama H., Yamada T. Radiation-induced apoptosis: a review. In: Apoptosis, its Roles and Mechanisms. Ed.: Yamada T., Hashimoto Y. Business Center for Academic Societies Japan , Tokyo, 1998, 141- 186.

Pollycove M. Positive health effects of low level radiation in human populations. In: Biological Effects of Low Level Exposures: Dose-Response Relationships. Ed. E.J. Calabrese. Lewis Pub. Inc., Chelsea, Michigan, 1994, 171-187

Pollycove M. Nonlinearity of radiation health effects. Env.Health Perspect. 106: 363-368 (1998)

Pollycove M., Feinendegen L.E. The relationship between spontaneous and background radiation induced DNA damage in preparation, 1998

Pratzel H.G., Legler B., Heisig S., Klein G., Franke Th., Aurand K. Efficacy of radon baths on rheumatic pain in patients with degenerative spinal of joint disorders. In: Radon in der Kurortmedizin. Ed . Pratzl, H.G., Deetjen P. I.S.M.H. Verlag, Geretsried, 1997, 114 - 143

Soto J. Effects of radon on the immune system. In: Radon in der Kurortmedizin. Ed . Pratzl, H.G., Deetjen P. I.S.M.H. Verlag, Geretsried, 1997, 103 - 113

Suzuki K., Kodama S. Watanabe M. Suppressive effect of low-dose preirradiation on genetic instability induced by X rays in normal human embryonic cells. Rad.Res. 150 : 656-662 (1998)

Sakamoto K., Myojin M., Hosoi Y., Ogawa Y., Nemoto K., Takai Y., Kakuto Y., Yamada S., Watabe N. Fundamental and clinical studies on cancer control with total or upper half body irradiation. J. Jpn. Soc. Ther. Radiol. Oncol. 9: 161-175 (1997)

UNSCEAR. Sources and Effects of Ionizing Radiation. United Nations, New York, N.Y., 1994; Annex B, Adaptive Responses to Radiation in Cells and Organisms.