The Adaptive Response to Ionizing Radiation: Low Dose Effects Unpredictable from High Dose Experiments

Odile Rigaud, Ph.D.

Institut Curie-Recherche,



26 rue d'Ulm,

75248 Paris cedex 05, France

Tel: 33-1-42-34-67-08

Fax: 33-1-46-33-30-16


Despite considerable worldwide effort, uncertainties underlying the biological effects of exposure to low doses of genotoxic agents still exist. Thus, in general, the protection guidelines recommend the extrapolation of data from high dose experiments to estimate the low dose effects. The adaptive response (AR) is one example of protective effects induced by low doses of DNA damaging agents. Although the existence of the AR is clearly demonstrated at the cellular level, its practical interest for risk assessment is still questionable.

The adaptive response occurs in response to low level exposure to various DNA damaging agents including alkylating compounds, oxidative radicals and ionizing radiation. Some evidence for cross adaptive responses by some chemicals and radiation has been reported. Comments on the following issues have been mainly illustrated by experiments dealing with the adaptive response to ionizing radiation.

1. How does the dose that induces the adaptive response (AR) relate to human and environmental (ecological) exposures?

In the case of ionizing radiation (IR), both the dose and the dose rate that induce AR are above the level of exposure to terrestrial ambiant radiation , i. e. few millisieverts per year (one mSv is the absorbed dose corresponding to a delivered dose of one milligray, mGy).

Indeed, the adaptive response is induced by a single exposure to a low dose usually within the dose range of 0.1 to 10 cGy. Chronic exposure at a very low dose rate (mGy. h-1), though not as well documented, can result in a protective effect against subsequent damage induced by a high dose. A relatively narrow window of doses is effective if delivered at an appropriate dose rate depending on the total dose applied. This means that the cells must receive a certain amount of signals within a given interval of time for the AR to be expressed1.

The level of spontaneous DNA damage mainly due to physiological metabolism has been estimated to be 6-1000 damage events per cell per hour. Exposure to low LET radiation yields approximatly 100 DNA lesions per cell and per cGy2. The prime importance of the dose-rate is illustrated by the experiments described by Shadley and Wiencke1 showing that a dose-rate < 5cGy/min was required for inducing adaptation by 1 cGy. That is to say, within the 12 sec. of exposure at this dose-rate, irradiation produces approximatly 100 lesions since 25 DNA damage events are estimated to arise spontaneously. Therefore, under such conditions, the level of damage occuring within the cell is five fold higher than that spontaneously arising. Cells would likely react to such changes in the level of damage by activating some molecular processes that ultimately lead to a better tolerance to a subsequent high dose.

The influence of background radiation on a genotoxic stress response can be illustrated by the experiments carried out on yeast cells maintained in the presence of radiation at either 5 mSv per day (ambient radiation) or at 0.6 mSv per day (conditions for shielded cultures)3. When exposed to methylmethane sulfonate (MMS), the frequency of induced reciprocal recombination frequency was higher in cells maintained at the lowest background. These data indicate that the environmental background radiation would allow the maintenance of an appropriate antioxidant defenses or a pool of repair proteins for preventing or repairing DNA damage as it occurs. These data are reminiscent of those showing hormetic effects on growth of procaryotic cells depending on the level of background radiation4.

Lymphocytes from individuals occupationally exposed to low doses (maximum of 28mSv per year) were found to be more sensitive to the clastogenic effect of a challenge in vitro irradiation with 2 Gy, indicating in vivo induction of an AR. Furthermore, this study indicates higher basal levels of chromosome aberrations in exposed individuals indicating that the previous exposure would not be beneficial per se5. An AR was also reported to occur in lymphocytes from children exposed to the fallout of the Chernobyl accident and was ascribed to the persistent internal contamination to 137Cs rather than to the initial acute absorbed dose of high doses of IR6. Clearly, these studies on the human population show how changes in human and environmental exposures are able to induce an AR.

2. What are the potential up and down sides of having one's AR induced?

The adaptive response to IR has been extensively demonstrated at the cellular level. It can be considered beneficial for cells with respect to its contribution to the maintenance of genomic integrity, as a result of less damage (chromosomal breaks, gene mutations) and cell killing by a subsequent treatment. Reproducible data have pointed out the existence of AR for somatic and germinal cells following in vivo exposure to a previous low dose.

In this context, several observations suggest that the induction of the AR might be restricted to specific experimental systems or biological end-points rather than a general phenomenon. Indeed, the AR in terms of chromosomal damage has been shown to be variable from one donor to another with some individuals being unresponsive and a few exhibiting a synergistic response to the two doses7. The same holds true for different cell lines or mouse strains and for other biological end-points, for instance adaptation expressed as cell survival8. The reasons for this individual variability still remains unknown. One possibilitity is that it is related to the intrinsic radiosensitivity of the individual; owing to the conflicting results as to whether the AR is inducible in cells from patients with radiosensitive syndromes such as ataxia telangiectasia, experimental data are at present inconclusive9, 10. It might also be possible that the dose range required for the AR to be induced depends on the cell sensitivity. It has also to be mentioned that a high constitutive level of certain proteins involved in cell cycle and damage repair are thought to be responsible for the development of survival adaptation of some human cells8.

Moreover, according to the experimental systems, the adaptive response in terms of cytogenetic, mutagenic and lethal effects is not always concomitantly triggered by low dose pre-exposure11. It can be hypothesized that the protective mechanism on the different biological end-points analyzed in the same cell line could arise from activation observed for different pathways of signalling or processing of DNA lesions.

Clearly, elucidating the molecular mechanism underlying the AR is necessary. In the case of adaptation to IR, the activation of DNA repair (double strand-breaks and base excision repair systems) as well as the contribution of increased activities of antioxidant enzymes have been reported to explain the protective effect observed in adapted cells12-14. Changes in gene expression are involved as suggested by the observation that de novo synthesis of messengers and transcripts within the first hours after the adaptive dose are required for the AR to be fully developped. Identifying the genes for which changes in expression levels are associated with the development of the response would give indications as to whether the AR is a general phenomenon that when induced might be beneficial for human health.

Far less has been reported regarding the AR on survival and cancer induction in whole animals (see below issue 5). This is the most important area for future research before stating that having ones AR induced constitutes a benefit for the whole organism.

3. Can the induction of AR be manipulated for medical and other benefits?

Our previously reported data showing a decreased mutation frequency with unaltered cell lethality in adapted cells led us to suggest that an adaptive protocol might be beneficial in radiotherapy treatment; it was expected that the low dose preexposure would minimize the mutagenic effect of the radiation treatment whereas the same cytotoxic efficiency should be obtained11. Later on, data from Boothman's laboratory showed that amongst eight cell lines tested (4 normal, 3 neoplastic and 1 cancer-prone cell lines) the induction of an AR in terms of survival seems to be restricted to some specific tumoral cells8. Clearly, further knowledge is needed before defining the potential implications of AR for medical or other benefits.

Another example of induced radioresistance by low doses illustrated by the multiphasic shape of single dose survival curve in the low dose range, shares some homology with the AR15. The survival curve has been characterized by an hypersensitive region at doses <0.5Gy (HRS for hyper radiosensitivity) followed by an induced radioresistance as doses increase up to 1 Gy (phase IRR for induced radioresistance). The assumption that the AR and the HRS/IRR phenomenon could arise from similar mechanisms comes from data showing that the low dose hypersensitivity disappeared if cells have been pre-exposed to a low adaptive dose of 0.2 Gy prior to the challenge dose. This hypothesis remains to be proven. The extent of the HRS/IRR phenomenon has been related to the intrinsic radioresitance of tumoral cells, that is to say the more radioresitant tumor cells exhibited higher levels of HRS and IRR. The authors proposed that a possible therapeutic gain by lowering dose per fraction in order to avoid induced radioresistance for the class of tumors clinically identified to be radioresistant15.

From a general point of view, developping new therapeutic means by manipulating the AR requires further knowledge of the consequences of low dose exposure in tumoral and normal experimental cell systems. Most knowledge regarding the signalling transduction pathways and the resulting alterations in gene expression, apoptotic processes and interaction with the immune system comes from studies using high doses and it is not known whether the activated molecular processes are similar after low-level doses.

4. How does the AR relate to the concept of hormesis?

The term hormesis derived from the greek word «hormôn» - to excite- was used to describe the stimulatory effects on growth following single exposures to low doses of genotoxic agents which are toxic at high doses. The AR refers to the process whereby cells exposed to small doses became less susceptible to the genotoxic effects of a subsequent high dose. This response takes some hours to be induced. The AR differs from the the hormetic effect since it modifies the response to subsequent high doses by decreasing the slope of the dose-effect curve. The AR can be related to the concept of hormesis as it can be expected to result from stimulation of some cellular defenses by low doses of an agent which otherwise is toxic. It would not indicate that the agent is beneficial per se. The mechanisms are likely stimulated depending on the dose level. The counterpart of these so-called beneficial or hormetic effects is that unexpectedly harmful effects might also occur at doses unable to trigger the defense systems. Evidence for such effects is clearly indicated by the HRS region of the single dose response curve evoked before, i. e. a sensitivity per unit dose higher than expected from the extrapolation of high dose data15. Numerous examples have also been reviewed elsewhere16. One must thus keep in mind that consequences of low-level exposure might refer to both harmful and beneficial effects.

5. Should knowledge of the AR affect current risk assessment methods for carcinogens?

Due to the low statistical reliability of experimental and epidemiological data already published, it cannot be concluded whether or not a threshold dose below which no increase in the frequency of cancer is observed. Thus, the low dose effects are currently estimated by extrapolation of high dose data. The assumption of a linear dose effect relationship for the low dose range is questionable after considering the current data concerning carcinogenesis by chemicals as well as radiation17. The existence of low dose effects such as the AR and the HRS/IRR phenomenon evoked before, show evidence for responses unpredictable from high dose experiments. In other words, both infra and supra-linear dose-response relationships might be expected, as has been previously reviewed16. This also raises the question as to whether or not a threshold dose would exist for cells to react to changes of damage following low-level exposure by inducing some protective mechanisms.

Mutations as well as chromosomal damage are likely involved in the developpment of cancer. Numerous data show evidence for a protective effect against the induction of gene mutations in various cell lines including human cells adapted by low-level radiations18. The frequency of mutations, especially deletion is reduced by more than one half in radioadapted cells. The non essential gene, hypoxanthine phosphoribosyl transferase gene (HPRT) is often the target gene choosen in these mutagenesis studies because mutants deficient in HPRT activity can be easily selected by acquired resistance to 6-thioguanine. This seems to be beneficial for the cells. However, it is of prime importance to determine if a low dose preexposure would similarly affect the fate of mutagenic lesions in essential genes, such as oncogenes or tumor suppressor genes that play a key role in the control of genomic stability. This requires the experimental development of a means to analyze and quantify such rare mutational events in cell populations without a selectable phenotype. Until such experiments are performed, it is premature to evoke beneficial effects of AR, in terms of cancer risk.

The incidence of the AR on the estimation of cancer risk related to previous exposure to carcinogens is at present poorly documented. In the case of ionizing radiation, recent data indicate that an adaptive dose would induce a decrease in the frequency of neoplastic transformation spontaneously arising as well as that induced by a subsequent high dose (18 and references therein). Bhattarcharjee described recently that preirradiating Swiss mice with five repeated exposures to small doses of 1 cGy per day appeared to reduce from 46% to 16% the incidence of thymic lymphoma by a challenge high dose of 2 Gy19. It is of prime interest to perform further experiments before AR should be taken into account for risk evaluation.


There is no doubt that the AR is an important phenomenon at the cellular level. A better understanding of the associated molecular mechanisms will explain the individual variability and determine whether it is a general response. The current knowledge of the AR at the level of whole organism needs to be improved, especially for its possible relevance to cancer induction, a major concern in protection of human health.

I wish to thank Dr E. Moustacchi for helpful discussion and critical reading of the manuscript. Research was supported by research contracts from «Association pour la Recherche contre le Cancer» and «Ligue Nationale contre le Cancer».


(1) Shadley J-D, Wiencke J-K. Induction of the adaptive response by X-rays is dependent on radiation intensity. International Journal of Radiation Biology 1989; 56: 107-118.

(2) Ward J-F. DNA damage producedproduced by ionizing radiation/ identities, mechanisms of formation and repairability. Progress in nucleic acids and molecular biology 1988; 35: 93-125.

(3) Satta L et al. Biological response of the yeast Saccharomyces cerevisia to chemical radiomimetic agents: influence of low background environment. in Low dose irradiation and biological defense mechanisms. T. Sugahara, L.A. Sagan, T. Aoyama, (edit) Elsevier Science: Amsterdam, 1992; pp.383-384.

(4) Planel H et al. Influence on cell proliferation of background radiations or exposure to very low chronic gamma radiation. Health Physics 1987; 52: 571-578.

(5) Barquinero J-F et al. Occupational exposure to radiation induces an adaptive response in human lymphocytes. Intenational Journal of Radiation Biology 1995; 67: 187-191.

(6) Tedeschi B et al. Do human lymphocytes exposed to the fallout of the Chernobyl accident exhibit an adaptive response? III.Challenge with bleomycin in lymphocytes from children hit by the initial acute dose of ionizing radiation. Mutation Research 1996; 354: 77-80.

(7) Bosi A, Olivieri G. Variability of the adaptive response to ionizing radiations in humans. Mutation Research 1989; 211: 13-17.

(8) Boothman D-A, Meyers M, Odergaard E, Wang M. Altered G1 checkpoint control determines adaptive survival response to ionizing radiation, Mutation Research 1996; 358: 143-154

(9) Nemethova G, Kalina I, Racekova N. The adaptive response of peripheral blood lymphocytes to low doses of mutagenic agents in patients with ataxia telangiectasia. Mutation Research 1995; 348: 101-104.

(10) Seong J, Suh C-O, Kim G-E. Adaptive response to ionizing radiation induced by low doses of gamma-rays in human cell lines. International Journal of Radiation. Oncology Biology Physics 1995; 33: 869-874.

(11) Rigaud O, Papadopoulo D and Moustacchi E. Decreased deletion-type mutation in radiodapted human cells. Radiation Research 1993; 133: 94-101.

(12) Zhou P-K et al. Adaptive response of mutagenesis and DNA double-strand break repair in mouse cells induced by low dose of g-ray. in Low dose irradiation and biological defense mechanisms,T. Sugahara, L.A. Sagan, T. Aoyama, (edit) Elsevier Science:Amsterdam 1992; pp271-274.

(13) Le X-C et al. Inducible repair of thymine glycol detected by an ultrasensitive assay for DNA damage. Science 1998; 280: 1066-1069.

(14) Bravard A, Luccioni C, Moustacchi E, Rigaud O. Contribution of antioxidant enzymes to the adaptive response to ionizing radiation of human lymphoblasts, International Journal of Radiation Biology (in press)

(15) Joiner M-C etal. Hypersensitivity to very-low single radiation doses: its relationship to radioadaptive response and induced resistance. Mutation Research 1996; 358: 171-183

(16) Oftedal P. Biological low-dose radiation effects. Mutation Research 1991; 258: 191-205.

(17) Abelson P-H Risk assessment of low-level exposures. Science 1994; 265: 1507 and 266: 1141-1145

(18) Rigaud O, Moustacchi E. Radioadaptation for gene mutation and the possible molecular mechanisms of the adaptive response. Mutation Research 1996; 358: 127-134.

(18) Redpath J-L, Antoniono R-J. Induction of an adaptive response against spontaneous neoplastic transformation in vitro by low-dose gamme radiation. Radiation Research 1998; 149, 517-520.

(19) Bhattacharjee D. Role of radioadaptation on radiation-induced thymic lymphoma. Mutation Research 1996; 358: 223-230.