Low Doses of Radiation are Protective in vitro and in vivo : Evolutionary Origins

Plants

Cortes et al. (1990a) examined plant cells for evidence of an adaptive response to radiation. When Allium cepa root-tip cells were exposed to low doses of X-rays or to incorporated tritium and subsequently given a ‘challenge’ dose of X-rays, the X-ray-induced chromosomal damage was reduced, indicating an adaptive response capable of repairing DNA double-strand breaks exists in plants.

Insects

Adaption to radiation has been shown in insects. Experiments in lepidopteran cells (Koval, 1988) showed that the survival of cells plated immediately following the second of two equivalent doses separated by several hours was greater than the survival of cells plated immediately following a single dose equal to the sum of the split doses. The survival was approximately five times greater than that expected after the second dose, suggesting that the first split dose stimulated a repair system not present in unirradiated cells. In other experiments, exposure of flour beetles to gamma-rays enhanced longevity (Ducoff 1975) and induced resistance to oxygen toxicity (Lee and Ducoff 1984). The magnitude of the effect was dose-dependent, and irradiation under anoxic conditions reduced the development of oxygen resistance to the same degree that it reduced acute radiation lethality, a dose dependence and radiobiological oxygen effect similar to the observations in the simple eukaryote yeast (Mitchel and Morrison 1982a, 1984b, 1987). In Drosophila, induction of the adaptive response to radiation in mature oocytes was at a maximum after an adapting dose of 0.2 mGy (Schappi-Buchi 1994), suggesting that the cells respond to first track of radiation, in a manner similar to prokaryotes (Ewing 1995).

Fish

Kurihara et al. (1992) examined the response of fish cells to low doses of X-rays. They showed that in cells cultured from two species, either goldfish or mudminnows, low doses of radiation reduced both chromosomal aberrations and the frequency of micronuclei resulting from a subsequent exposure to a high dose of radiation, indicating that the low dose had increased the ability of the fish cells to repair DNA double strand breaks.

Mammalian cells

DNA repair and mutagenesis. Exposure of human lymphocytes to betaradiation by culture in the presence of tritiated thymidine decreased the yield of chromosomal aberrations from a subsequent X-ray exposure (Olivieri and Wolff 1984). A similar observation was made using gamma-rays and lymphoblastoid cells (Rigaud 1993). The protection was not due to a diffusible factor and was consistent with a chromosomal repair mechanism (Wiencke 1986). Doses as low as 1 cGy of X-rays protected against DNA double strand breaks induced by either radiation or chemicals (Wolff 1988), a finding that is consistent with observations in both bacteria and yeast. Other oxidizing agents such as hydrogen peroxide also induced the protective effect in human lymphocytes (Cortes 1990b), as well as in bovine or rabbit lymphocytes (Flores 1996). This adaptive process required protein synthesis (Youngblom 1989), a characteristic also conserved from prokaryotes.

Low doses and low dose-rate exposures have also been shown to increase the ability of human skin fibroblasts to repair radiation-generated breaks in chromosomes, as shown by an increased rate of repair (Azzam et al. 1994a). An increase in the rate of an enzymatic process implies either elevated enzyme levels or increased access to the substrate. The enhanced repair occurred after 1 mGy, which represents, on average, a single track per cell (Broome et al. 2002). Higher doses, representing multiple tracks/cell, produced the same result as one track/cell when those tracks from the higher doses were spaced out in time. This ability of human cells to respond to the first track of radiation also appears to have been conserved from prokaryotes (Ewing 1995) and insect cells (Schappi-Buchi 1994).

In mammals, the induced protection is not unique to human cells. Cultured Chinese hamster V79 cells exposed to chronic beta-rays from incorporated tritiated thymidine also showed reduced sister-chromatid exchange and micronucleus formation when subsequently exposed to a high dose of gamma-rays (Ikushima 1987). When exposed to a low dose of gamma-rays 4h prior to a challenge dose, the rate of DNA damage repair in the adapted cells was higher than that in non-adapted cells, and the residual damage was less in adapted cells than in non-adapted cells. In these cells, there was no indication of any difference in the initial yields of DNA double-strand breaks induced by challenging doses from non-adapted cells and from adapted cells (Ikushima et al. 1996), indicating that the predominant feature of the induced resistance was enhanced DNA repair capacity, not a reduction in the initial damage by, for instance, elevated radical scavenging.

The enhanced repair capacity in adapted cells also protected against mutation in human-hamster hybrid cells, where a low dose followed by a later challenge dose significantly reduced the yield of mutants compared to the challenge dose alone. As observed for enhanced resistance to killing and enhanced DNA repair, incubation with a protein synthesis inhibitor largely negated the decrease in mutant yield (Ueno et al. 1996). The reductions in chromosomal aberrations and mutation yield observed in the various mammalian cells used for these low dose studies implies that the net error rate for the overall DNA repair must have been reduced in adapted cells, further implying that the induced repair process is relatively error-free. The reduction in mutations in radiation adapted mammalian cells was also seen in a single cell eukaryote, indicating evolutionary conservation of this feature of the stress response.

Exposure to low doses has also been shown to induce an adaptive response in cells taken from wild white-tailed deer. Exposure to doses as low as 1 mGy increased the ability of skin fibroblasts to repair chromosomal breaks from a second high dose (Ulsh et al. 2004), a response that replicates that observed in human skin fibroblasts exposed to the same low dose (Broome et al. 2002), and that was also seen in prokaryotes (Ewing 1995).

When lower eukaryotic cells were exposed to a heat shock, which induces heat-shock proteins, the cells became radioresistant (Mitchel and Morrison 1982a) and the extent of the radioresistance was regulated by a heat shock protein (Boreham and Mitchel 1994). This association of heat-shock protein induction with induction of radioresistance has apparently also been conserved in mammalian cells. When RIF cells, which did not show an adaptive response to radiation, were transfected with inducible HSP70 or HSP25, the adaptive response was restored and both radioresistance, measured by a clonogenic assay, and a reduction of apoptosis were detected in these cells (Kang et al. 2002, Lee et al. 2002). In human lymphocytes, either a mild heat stress or a low dose of ionizing radiation induced an adaptive response that protected against cytogenetic damage from a subsequent high radiation exposure (Cai and Jiang 1995). When the two adaptive treatments were combined, no additive effects on the magnitude of the adaptation induced were observed, suggesting that low-dose radiation and hyperthermia may share one mechanism of induction of adaptation to cytogenetic damage. This same observation has been reported in the single cell eukaryote S. cerevisiae (Mitchel and Morrison 1982a, 1984a), again indicating that adaption to radiation is part of a general stress response conserved from lower to higher eukaryotes.

Although the studies noted above clearly show that low doses of radiation or other oxidizing stress induce DNA repair processes that increase the ability of the cell to repair broken chromosomes, and reduce mutation and chromosomal aberrations, these endpoints are not the direct measures of risk, such as cancer, usually considered when assessing radiation risk. However, the impact of these evolutionarily conserved, inducible protective effects on the risk of malignant transformation has been tested in both rodent and human cells.

Malignant transformation. The essential features of the oxidative stress-inducible DNA-repair processes appear to have been tightly conserved during the evolution of prokaryotes to mammalian cells, including human cells, a strong argument that this process is highly important to life. However, this does not necessarily indicate a reduction in the risks normally associated with and measured after a radiation exposure. The most important of those risks is generally considered to be cancer. One characteristic of cancer cells is genomic instability. Suzuki et al. (1998) showed that low doses reduced the level of genomic instability induced in human cells by a high dose. Azzam et al. (1994b) measured both the ability of cells to repair chromosome breaks (DNA double-strand breaks) and the frequency of malignant transformation in rodent cells exposed to a high dose and dose rate of gamma radiation, with and without a prior low-dose and low-dose-rate exposure. The prior low dose, given 24h before the high dose, reduced both the frequency of unrepaired chromosomes and the frequency of malignant transformation induced by the high dose, indicating that the enhanced DNA repair was relatively error-free, and that the low-dose exposure was actually reducing cancer risk. Evidence presented by Broome et al. (1999) indicated that the application of the induced repair in human cells was biased toward some chromosomes and away from others suggesting that the decrease in risk may depend on or reflect this bias.

The same group extended those experiments by testing the effects of low gamma-radiation doses alone (without a second high dose) on spontaneous malignant transformation in rodent cells (Azzam et al. 1996). Doses between 1 and 100 mGy, given at low dose rate, reduced the risk of spontaneous malignant transformation by about 3–4 fold, and all doses were equally effective. This same concept was tested in human cells by Redpath et al. (1998, 2001) who reached a similar conclusion. These results demonstrate the tight evolutionary conservation, from prokaryotes (Ewing 1995) to complex invertebrates (Schappi-Buchi 1994) to human cells (Broome et al. 1996), of the ability of very low doses of radiation to induce DNA double strand break repair in cells, and the importance of that conserved response for the reduction of risk in mammalian cells, including human cells.

The lowest dose tested by Azzam et al. (1996) and Redpath et al. (2001), 1 mGy of gamma-radiation, represents an average of 1 ionization track per mammalian cell nucleus. That data shows that mammalian cells respond to the first track of radiation, and that the response reduces risk. Since radiation tracks are random, at 1 mGy most cells receive one track but some receive none or more than one. However, all respond to the same extent as they did when they certainly received one or more tracks at the higher (10–100 mGy) doses. This maximum risk-reduction effect at 1 mGy parallels the results on induced DNA double strand break repair (Broome et al. 1996, Ulsh 2004). This is evidence, therefore, that not all cells are actually required to be exposed (hit) by radiation in order to enhance their defences and reduce their risk. Such distributed effects are known as bystander effects and result from inter-cell signalling.

Control of adaptive response. These observations in mammalian cells allow an interesting comparison with the induced DNA repair process in lower eukaryotes. Although the data noted above strongly indicate that the fundamental process has been tightly conserved, the results also indicate that the strategy for control of the process has evolutionarily diverged. Although all prokaryotic and eukaryotic cells appear to respond to the first track of radiation, in lower eukaryotes the extent of induction of the DNA repair process by radiation, and therefore the net effect of the process, is clearly dose-dependent. The extent of induction increased with dose (to a maximum induction) (Chankova and Bryant 2002, Le and Ducoff 1984, Mitchel and Morrison 1984b, 1987) and therefore the degree of protection against, for example, cell killing by a second dose or mutation by a chemical mutagen, was also dose dependent. In contrast, the response of mammalian cells to the first track of radiation appears to be amplified, via signalling to bystander cells, to produce a maximum response in some volume of tissue (Broome et al. 2002, Ulsh et al. 2004). This maximum response to the first track, rather than a graded response with increasing dose, could reflect the necessity to provide maximum protection in the more radiation-sensitive mammalian cells.

The cellular data strongly suggest that the biological processes that respond to radiation are inducible and protective, and that these processes have been tightly conserved throughout evolution. The data also suggest that these processes are likely to impact on risk in whole organisms, and in ways that would preclude a linear increase in risk with increasing dose.

Chromosomal changes

In vivo exposure of higher eukaryotes, including humans, can induce protective effects in cells that are similar to those seen using cells in culture. Exposure of rabbits to low doses prior to a challenge dose in vivo reduced the frequency of chromatid aberrations seen in bone marrow (Cai and Liu 1990). Liver cells taken from frogs living in radioactively contaminated ponds (1 mGy/y) were resistant to chromosomal breakage from high doses (preliminary data of Stuart and Morrison, described in Mitchel 2004). This result in frogs was also observed in lymphocytes, taken from humans living in a high background area (260 mSv/y), that showed a reduced level of chromosomal aberrations after a high challenge dose in vitro, compared to people living in a nearby area without the high background (Ghiassi-nejad et al. 2002). In similar studies of lymphocytes taken from occupationally exposed nuclear workers (Barquinero et al 1995, Thierens et al. 2002) or medical radiation workers (Gourabi and Mozdarani 1998), the occupational exposures were shown to increase the capacity of lymphocytes to repair broken chromosomes (DNA double strand breaks) resulting from an in vitro high dose exposure. These data show that environmental and occupational exposures induce the same protective response in mammalian and amphibian cells that has been seen in prokaryotic and eukaryotic cells grown in culture, and support the idea that this response is relevant to real-life situations in mammals and non-mammals.

Cancer Risk

The studies noted above indicate that low doses induce an evolutionarily conserved DNA repair system that reduces the risk of spontaneous and radiation induced chromosomal aberrations, mutations and malignant transformation. Several studies show that low doses also reduce the risk of cancer in animals.

Exposure of mice with a high spontaneous risk of lymphoma to 15 or 5 cGy two or three times per week respectively, for about 29 weeks significantly reduced the lymphoma incidence and increased the lifespan of the mice (Ishii 1996). A single exposure of mice to 100 mGy at low dose rate increased the latency of myeloid leukemia in mice subsequently exposed to a high dose (Mitchel et al. 1999). These data suggest that a major influence of a single low dose may be a reduction in the rate of progression of genomic instability (the process which controls tumor latency) and multiple exposures may extend tumor latency beyond normal lifespan. Low adapting doses have been shown to reduce genomic instability induced by high doses in human cells in culture (Suzuki et al. 1998).

Radiation protection standards and practices applied to humans must also consider the possibility that some individuals may be more radiation-sensitive and cancer-prone than others, for genetic reasons. This raises the possibility that low doses may produce different, and potentially more harmful effects in such individuals, than those seen in genetically normal individuals.

Mice that are heterozygous for the Trp53 gene (Trp53 +/-) are compromised in their ability to repair DNA damage and in their ability to initiate cell death in improperly repaired cells, two of the processes most important for controlling cancer risk.

Consequently, such mice are cancer-prone (Kemp 1994). A dose of 10 mGy, given at low dose rate the day before a 4 Gy exposure delayed the onset of lymphomas in Trp53 +/- mice, but did not change their frequency (Mitchel et al. 2004) paralleling the effect for myeloid leukemia seen in genetically normal (Trp53 +/+) mice (Mitchel et al. 1999). Increasing the low adapting dose to 100 mGy caused this protective effect to disappear in the Trp53 +/- mice. While not increasing harm, 100 mGy apparently represented an upper threshold for doses that are protective against radiation-induced lymphomas in the cancer-prone mice. In the normal mice, protection was still seen at 100 mGy indicating a higher upper threshold for protective effects in the Trp53 +/+ mice.

Experiments testing the in vivo effect of low doses on cancer risk produced by high dose exposure are important for improving our understanding of the influence of induced protective processes on the dominant biological outcome of such exposures. However, to estimate risk for radiation protection purposes it is more important to understand the influence of low doses alone on spontaneous cancer risk. In cancer prone Trp53 heterozygous mice exposed to low doses, the tumor types and frequencies did not change, indicating that the tumors appearing after irradiation were the same spontaneous tumors seen in the absence of radiation. However, a single exposure of either 10 or 100 mGy, given at low dose rate to young mice, increased lymphoma latency and restored a portion of the lifespan lost due to the disease in the unexposed, cancer-prone mice (Mitchel et al 2003). In the same mice, the single exposure to 10 mGy delayed the appearance of spontaneous spinal osteosarcomas, but increasing the dose to 100 mGy resulted in a general acceleration of their appearance. Since, in the same animals, the upper dose threshold for protection against lymphomas exceeded 100 mGy, the dose threshold, where protective effects give way to detrimental effects, must be tissue-type specific. The data show, therefore, that protective effects against cancer show thresholds, a result also observed for malignant transformation in human cells (Redpath 2001). The observations indicate that the main effect of a single low dose exposure in animals was to increase tumor latency, suggesting that the rate of progression of genomic instability is slowed after an adapting dose. This observation is consistent with an evolutionarily conserved induction of the capacity for error-free DNA double-strand-break repair, increasing the probability of correct chromosomal rejoining in genomically unstable cells.

Teratogenic risk

Fetal cells are thought to be especially susceptible to radiation exposure, and the teratogenic effects of radiation are another measure of radiation risk in humans. Temple and Schleifer (1995) exposed chicken embryos in vivo to low doses 24h prior to a high dose. They report enhanced regeneration of DNA structure in brain and liver in the embryos exposed to the prior low dose, indicating that adaption to radiation exists in embryonic birds.

Wang et al. (1999) showed a correlation between radiation-induced apoptosis and teratogenesis in mice. That group exposed mouse fetuses on embryonic day 11(during the period of organ development) to a priming dose of 0.3 Gy, which significantly suppressed prenatal death and malformation induced by a challenging dose of radiation on embryonic day 12 (Wang et al. 1998). In a similar experiment by Mitchel et al. (2002), the degree of protection afforded by the low dose was shown to be proportional to Trp53 gene status, with maximum protection in Trp53 normal fetal mice, reduced but significant protection in Trp53 heterozygous mice and sensitization in Trp53 null mice. These observations show that adaption to radiation exists in fetal mammals as well as birds, and that these responses reduce teratogenic risk. The results showing reduced protection against teratogenesis in Trp53 heterozygous fetal mice, as compared to Trp53 normal mice, parallel the observations for reduced protection against cancer in Trp53 heterozygous mice. Since apoptosis in mammalian cells is closely linked to a lack of, or incomplete DNA repair, these results are consistent with protective effects that result from the induction of error-free DNA repair by low doses of radiation in fetal mice, with the consequences of that induction under the control of Trp53 gene activity.

Heritable mutation risk

Radiation protection assumptions also consider the risk that radiation will create heritable mutations. Boreham et al. (2004) have examined the risk of heritable mutations from radiation exposure in mice. In their experiments, male mice were exposed to 1 Gy of gamma radiation, with or without an exposure to 100 mGy 24h prior to the 1 Gy. The exposed mice were held for 10 weeks to remove all radiation-exposed mature sperm and then bred to unexposed females. The offspring were examined for mutations received from the exposed male parent germline stem cells. They observed that the 1 Gy exposure approximately doubled the spontaneous rate of mutation in the offspring, but that the prior exposure to 100 mGy reduced the effect of the 1 Gy such that the resulting mutation rate was not different from the spontaneous rate in the absence of any exposure. This observed reduction in heritable mutations is similar to the observations seen throughout simple and lower eukaryotes, and in other mammalian cells, and is consistent with the induction of a DNA repair system by low doses in the sperm stem cells, correctly repairing subsequent radiation damage, and thereby reducing the frequency of heritable mutations in the offspring of the exposed male mice.