**7. Fractionated irradiation and bystander effect**

Experimental data on bystander effect mostly come from single-dose application experiments *in vitro*. However, there is a lack of knowledge, which would have potential clinical implication, e. g. whether bystander effect occurs during fractionated treatment. Mothersill and Seymour (2002a) performed experiments involving repeated treatment of bystander cells with medium collected from irradiated cells as well as involving repeated dose exposure of cells producing bystander signals, as a way of mimicking fractionated exposures. The recovery factor was defined as the surviving fraction of the cells receiving two doses (direct, or ICM) separated by an interval of 2 h divided by the surviving fraction of cells receiving the same dose in one exposure. The authors observed that fractionated bystander treatments removed the effect of dose sparing that is observed after conventional fractionated regime, during which cells can repair DNA damage. Using Me45 human melanoma cell line established at the Center of Oncology in Gliwice (Kramer-Marek et al, 2006) we compared frequency of apoptosis and micronuclei formation in directly irradiated and bystander cells after single doses (1.5 - 6 Gy) and after doses divided into 3 fractions given at consecutive days (3 x 0.5 Gy – 3 x 2 Gy). We used a transwell system of co-incubation which allows co-culturing the irradiated cells growing in wells with non- irradiated cells growing in inserts. This system to some extent resembles situation *in vivo*, due to prolonged contact of non-irradiated and irradiated cells. As a source of X-rays (6 MV) Clinac 600 therapeutic accelerator was used. Non irradiated control cells were-sham exposed. After irradiation, inserts with growing non irradiated cells were placed into wells with irradiated ones and co-incubated. Before irradiation medium in both, wells and inserts, was replaced by fresh aliquots. To observe the response of hit and bystander cells after the set time of incubation (0, 24 and 48 h), we performed microscopic analysis of micronuclei induction and apoptosis. The results obtained show that both single dose irradiation and fractionation of the dose into three fractions effectively induced bystander effect in malignant Me45 melanoma cells. However, fractionated irradiation at low doses (Fig. 2) appears to be much more effective in inducing micronuclei in directly hit and bystander cells, whereas higher apoptosis induction was clearly seen in hit, and especially in bystander cells, at all doses in fractionated system (Fig. 3).

Intercellular Communication in Response to Radiation Induced Stress:

(\* denotes statistical difference from corresponding control, p<0.05, Student's t-test).

deviation and were obtained from three independent experiments

Fig. 2. Yield of micronuclei induction in Me45 melanoma cells irradiated with single or fractionated doses, in comparison with bystander cells. Data show means ± standard

Bystander Effects in Vitro and in Vivo and Their Possible Clinical Implications 347

(\* denotes statistical difference from corresponding control, p<0.05, Student's t-test).

Fig. 2. Yield of micronuclei induction in Me45 melanoma cells irradiated with single or fractionated doses, in comparison with bystander cells. Data show means ± standard deviation and were obtained from three independent experiments

Intercellular Communication in Response to Radiation Induced Stress:

(\* denotes statistical difference from corresponding control, p<0.05, Student's t-test).

deviation and were obtained from three independent experiments

Fig. 3. Yield of apoptosis induction in Me45 melanoma cells irradiated with single or fractionated doses, in comparison with bystander cells. Data show means ± standard

Bystander Effects in Vitro and in Vivo and Their Possible Clinical Implications 349

(\* denotes statistical difference from corresponding control, p<0.05, Student's t-test).

Fig. 3. Yield of apoptosis induction in Me45 melanoma cells irradiated with single or fractionated doses, in comparison with bystander cells. Data show means ± standard deviation and were obtained from three independent experiments

Intercellular Communication in Response to Radiation Induced Stress:

compared to γ-rays.

Bystander Effects in Vitro and in Vivo and Their Possible Clinical Implications 351

transfected with neurotransmiter (NAT) gene that enabled greater MIBG uptake. A similar U-shaped bystander phenomenon was observed for clonogenic cell-survival curve in case of high-LET alpha and Auger-electron emitters. No corresponding plateau in toxicity was observed after exposure of cells to the medium from β-irradiated cells. The reason for such behavior is not clear as yet. However, identification of the pathways involved in this process might pinpoint ways of manipulating the bystander effect for therapeutic purposes, i.e. to gain selective increase in tumor cell killing, accompanied by reduced side effects in normal tissue. Proliferative bystander responses have been also observed *in vitro* after irradiation with β-particles emitted by tritiated thymidine (3HTdR). The rat liver epithelial cells (WB-F344 line) not treated with tritiated thymidine (unlabeled cells), in the presence of radiolabeled cells that received absorbed doses from 0.14 – 1.7 Gy, showed statistically significant increase of cell growth by 9-10% in comparison to control (Gerashchenko and Howell, 2004). The mean energy of β-particles is only 5.7 keV, (a range of ca. 1 μm in water). Thus, the probability that β-particles emitted from radiolabeled cells will target the nucleus of adjacent unlabeled cells in non-confluent co-culture used in the study is very low, because the majority of unlabeled cells were far beyond the range of β-particles emitted from radiolabeled cells. The authors compared 3HTdR (the β-emitter) results with their earlier results obtained for γ–rays and found that a much lower dose of radionuclide (0.14 Gy) induced maximum response of bystander cells, whereas the maximum bystander response to γ-rays was not seen, even up to 1 Gy. According to the authors it is possible that the differences in the bystander dose response between γ-rays and 3HTdR may be related to higher relative biological effectiveness (RBE) that has been observed for 3HTdR, as

The presence of bystander effect initiated by *in vivo* decay of radionuclide was demonstrated by Xue et al. (2002). When human colon LS174T adenocarcinoma cells prelabeled with lethal dose of Auger electron–emitting 5-[125I]iodo-2-deoxyuridine (125IUdR) were subcutaneously co-injected with LS174T unlabeled cells into nude mice, a considerable inhibition of tumor growth was observed. Since the 125I present within the cells is DNA-bound, and 99% of the electrons emitted by the decaying 125I atoms have a subcellular range (<0.5 µm), and since the overall radiation dose deposited by radiolabeled cells in the unlabeled cells within the growing tumor is less than 10 cGy, these authors concluded that the results obtained are a consequence of a bystander effect generated *in vivo* by factor(s) present within and/or released from the 125IUdR-labeled cells. Radionuclides differ in their physical characteristics such as type of decay, the mean energy, the half-life and range of penetration. However, in spite of the identical decay, the Auger electrons for both, 123I (half–life, 13.3 h) and 125I (halflife, 60.5 d) they differ in mean energy which is 1.234 MeV and 179 keV for 123I and 125I respectively (Prise, 2008). The bystander effect induced *in vivo* by co-injection of radiolabeled and unlabeled LS174T cells was totally different (Kishikawa et al., 2006). 125I labeled cells stimulated tumor growth, and inversely, 123I labeled cells inhibited tumor growth after subcutaneous co-injection of cell mixture into nude mice. Similar pattern of response was observed in experiment *in vitro*. These contrasting effects were accompanied by different biochemical events; supernatants from cultures with 125I-labeled cells were positive for tissue inhibitors of metalloproteinases (TIMP1 and TIMP2), and those from cultures with 123I-labeled cells were positive for angiogenin (Kishikawa et al., 2006). These all studies demonstrate the potential of internalized radionuclides to generate bystander effects *in vivo* for therapeutic treatment, however many question remain in regard to bystander signaling evoked by application of different radionuclide, as pointed out in the

Our data indicate that the bystander effect may play some role during fractionated radiotherapy and should be regarded as an important part of ionizing radiation effect on living cells. Although fractionated irradiation was also applied *in vivo* to study the bystander effect on the level of DNA epigenetic changes in the non-exposed spleen of cranial irradiated mice (Ilnytskyy et al., 2009), the fraction doses used were far below those clinically applied. However, the authors observed that acute irradiation induced more pronounced bystander effect than fractionated irradiation.

#### **8. Radionuclide induced bystander effect**

Induction of the bystander effect is prevalent at low radiation doses and low dose rates (Seymour and Mothersill, 2000), the characteristic features of targeted radionuclide treatment of cancer. Thus, one could expect that bystander effects induced by targeted radionuclides could have a strong impact on radiotherapeutic and diagnostic treatment (Prise and O'Sullivan, 2009). An increasing body of data indicates the involvement of bystander phenomenon after radionuclide application under experimental conditions. It can appear as damaging or protective effects in dependence on dose and dose rate. A very low dose of photon radiation (~ 30 keV) emitted by iodine-125 radioisotope (4mGy dose/day to 1,4 mGy/day) during a three month exposure of hybrid HeLa cells with human fibroblasts caused resistance of these cells to neoplastic transformation when they were challenged by subsequent irradiation with 3 Gy of 137Cs gamma rays (Elmore et al., 2008). Lowering of dose rate below 1mGy/day abolished the adaptive answer, suggesting that low dose-rate above a certain threshold is responsible for this type of radio-adaptation. The damaging bystander effect induced by radionuclide is also frequently observed in *in vitro* experiments. Various type of cells may differ however in response to radionuclide induced bystander signals. Chen et al. (2008) using 125I seeds irradiated two lung cancer cell lines that had different sensitivities to HDR gamma-ray irradiation and investigated the bystander effect of DNA DSBs as histone H2AX phosphorylation, and micronuclei formation. They found that the proportion of bystander cells with micronuclei and number of γH2AX foci was higher in radiosensitive NCI-H446 cell line than in more radioresistant A549 cell line. Interesting from clinical point of view was the observation that bystander effect compensated for the nonuniform distribution of radiation dosage in their experimental system. However, radionuclide induced bystander effect depends on the linear energy transfer (LET) of radionuclide emitters, being either damaging, or protective (Boyd et al., 2006, Mairs et al., 2007). Cells exposed to media collected from gamma-irradiated cells exhibited a dose-dependent reduction in survival fraction at low dosage and a plateau in cell-kill at >2 Gy. Cells exposed to media from metaiodobenzylguanidine-treated cells ([131I]MIBG, a low LET β-emitter), demonstrated a dose-response relationship with respect to clonogenic cell death and no annihilation of this effect at high radiopharmaceutical dosage. Contrarily, cells exposed to media from cultures treated with meta-211Atastatobenzylguanidine ([211At]MABG, a high LET α-emitter) exhibited dose-dependent toxicity at low dose, but elimination of cytotoxicity with increasing radiation dose, i.e. Ushaped survival curves (Mairs et al., 2007). Biologically similar analogs of halobenzylguanidines radiolabeled with radionuclides emitting β-particles (131I-MIBG), αparticles (211At-MABG), or Auger electrons emitting 123I-MIBG, were also tested in experiments performed by the same group (Boyd et al., 2006) on a human glioma cell line (UVW) and a cell line derived from human bladder transitional carcinoma (EJ138), both

Our data indicate that the bystander effect may play some role during fractionated radiotherapy and should be regarded as an important part of ionizing radiation effect on living cells. Although fractionated irradiation was also applied *in vivo* to study the bystander effect on the level of DNA epigenetic changes in the non-exposed spleen of cranial irradiated mice (Ilnytskyy et al., 2009), the fraction doses used were far below those clinically applied. However, the authors observed that acute irradiation induced more

Induction of the bystander effect is prevalent at low radiation doses and low dose rates (Seymour and Mothersill, 2000), the characteristic features of targeted radionuclide treatment of cancer. Thus, one could expect that bystander effects induced by targeted radionuclides could have a strong impact on radiotherapeutic and diagnostic treatment (Prise and O'Sullivan, 2009). An increasing body of data indicates the involvement of bystander phenomenon after radionuclide application under experimental conditions. It can appear as damaging or protective effects in dependence on dose and dose rate. A very low dose of photon radiation (~ 30 keV) emitted by iodine-125 radioisotope (4mGy dose/day to 1,4 mGy/day) during a three month exposure of hybrid HeLa cells with human fibroblasts caused resistance of these cells to neoplastic transformation when they were challenged by subsequent irradiation with 3 Gy of 137Cs gamma rays (Elmore et al., 2008). Lowering of dose rate below 1mGy/day abolished the adaptive answer, suggesting that low dose-rate above a certain threshold is responsible for this type of radio-adaptation. The damaging bystander effect induced by radionuclide is also frequently observed in *in vitro* experiments. Various type of cells may differ however in response to radionuclide induced bystander signals. Chen et al. (2008) using 125I seeds irradiated two lung cancer cell lines that had different sensitivities to HDR gamma-ray irradiation and investigated the bystander effect of DNA DSBs as histone H2AX phosphorylation, and micronuclei formation. They found that the proportion of bystander cells with micronuclei and number of γH2AX foci was higher in radiosensitive NCI-H446 cell line than in more radioresistant A549 cell line. Interesting from clinical point of view was the observation that bystander effect compensated for the nonuniform distribution of radiation dosage in their experimental system. However, radionuclide induced bystander effect depends on the linear energy transfer (LET) of radionuclide emitters, being either damaging, or protective (Boyd et al., 2006, Mairs et al., 2007). Cells exposed to media collected from gamma-irradiated cells exhibited a dose-dependent reduction in survival fraction at low dosage and a plateau in cell-kill at >2 Gy. Cells exposed to media from metaiodobenzylguanidine-treated cells ([131I]MIBG, a low LET β-emitter), demonstrated a dose-response relationship with respect to clonogenic cell death and no annihilation of this effect at high radiopharmaceutical dosage. Contrarily, cells exposed to media from cultures treated with meta-211Atastatobenzylguanidine ([211At]MABG, a high LET α-emitter) exhibited dose-dependent toxicity at low dose, but elimination of cytotoxicity with increasing radiation dose, i.e. Ushaped survival curves (Mairs et al., 2007). Biologically similar analogs of halobenzylguanidines radiolabeled with radionuclides emitting β-particles (131I-MIBG), αparticles (211At-MABG), or Auger electrons emitting 123I-MIBG, were also tested in experiments performed by the same group (Boyd et al., 2006) on a human glioma cell line (UVW) and a cell line derived from human bladder transitional carcinoma (EJ138), both

pronounced bystander effect than fractionated irradiation.

**8. Radionuclide induced bystander effect** 

transfected with neurotransmiter (NAT) gene that enabled greater MIBG uptake. A similar U-shaped bystander phenomenon was observed for clonogenic cell-survival curve in case of high-LET alpha and Auger-electron emitters. No corresponding plateau in toxicity was observed after exposure of cells to the medium from β-irradiated cells. The reason for such behavior is not clear as yet. However, identification of the pathways involved in this process might pinpoint ways of manipulating the bystander effect for therapeutic purposes, i.e. to gain selective increase in tumor cell killing, accompanied by reduced side effects in normal tissue. Proliferative bystander responses have been also observed *in vitro* after irradiation with β-particles emitted by tritiated thymidine (3HTdR). The rat liver epithelial cells (WB-F344 line) not treated with tritiated thymidine (unlabeled cells), in the presence of radiolabeled cells that received absorbed doses from 0.14 – 1.7 Gy, showed statistically significant increase of cell growth by 9-10% in comparison to control (Gerashchenko and Howell, 2004). The mean energy of β-particles is only 5.7 keV, (a range of ca. 1 μm in water). Thus, the probability that β-particles emitted from radiolabeled cells will target the nucleus of adjacent unlabeled cells in non-confluent co-culture used in the study is very low, because the majority of unlabeled cells were far beyond the range of β-particles emitted from radiolabeled cells. The authors compared 3HTdR (the β-emitter) results with their earlier results obtained for γ–rays and found that a much lower dose of radionuclide (0.14 Gy) induced maximum response of bystander cells, whereas the maximum bystander response to γ-rays was not seen, even up to 1 Gy. According to the authors it is possible that the differences in the bystander dose response between γ-rays and 3HTdR may be related to higher relative biological effectiveness (RBE) that has been observed for 3HTdR, as

compared to γ-rays. The presence of bystander effect initiated by *in vivo* decay of radionuclide was demonstrated by Xue et al. (2002). When human colon LS174T adenocarcinoma cells prelabeled with lethal dose of Auger electron–emitting 5-[125I]iodo-2-deoxyuridine (125IUdR) were subcutaneously co-injected with LS174T unlabeled cells into nude mice, a considerable inhibition of tumor growth was observed. Since the 125I present within the cells is DNA-bound, and 99% of the electrons emitted by the decaying 125I atoms have a subcellular range (<0.5 µm), and since the overall radiation dose deposited by radiolabeled cells in the unlabeled cells within the growing tumor is less than 10 cGy, these authors concluded that the results obtained are a consequence of a bystander effect generated *in vivo* by factor(s) present within and/or released from the 125IUdR-labeled cells. Radionuclides differ in their physical characteristics such as type of decay, the mean energy, the half-life and range of penetration. However, in spite of the identical decay, the Auger electrons for both, 123I (half–life, 13.3 h) and 125I (halflife, 60.5 d) they differ in mean energy which is 1.234 MeV and 179 keV for 123I and 125I respectively (Prise, 2008). The bystander effect induced *in vivo* by co-injection of radiolabeled and unlabeled LS174T cells was totally different (Kishikawa et al., 2006). 125I labeled cells stimulated tumor growth, and inversely, 123I labeled cells inhibited tumor growth after subcutaneous co-injection of cell mixture into nude mice. Similar pattern of response was observed in experiment *in vitro*. These contrasting effects were accompanied by different biochemical events; supernatants from cultures with 125I-labeled cells were positive for tissue inhibitors of metalloproteinases (TIMP1 and TIMP2), and those from cultures with 123I-labeled cells were positive for angiogenin (Kishikawa et al., 2006). These all studies demonstrate the potential of internalized radionuclides to generate bystander effects *in vivo* for therapeutic treatment, however many question remain in regard to bystander signaling evoked by application of different radionuclide, as pointed out in the

Intercellular Communication in Response to Radiation Induced Stress:

human tissues.

Bystander Effects in Vitro and in Vivo and Their Possible Clinical Implications 353

acetylation, and ubiquitination, referred to epigenetic changes. Pogribny et al. (2004, 2005) investigated the effect of low-dose radiation exposure on the accumulation of DNA lesions and alterations of DNA methylation and histone H4-Lys20 trimethylation in the thymus tissue using an *in vivo* murine model. They found that fractionated whole-body application of 0.5 Gy X-ray leads to decrease in histone methylation and DNA damage accumulation in the thymus gland. The radiation-induced global genome DNA methylation changes were shown to be dose-dependent, sex- and tissue specific and long-persistant. Tissue specificity of bystander responses within the same organism has also been examined by Ilnytskyy et al. (2009). They analyzed changes in global DNA methylation in spleen of mice whole-body or cranial exposed to single 0.5 Gy of X-rays or to the same dose given in five 0.1 Gy fractions. After acute cranial exposure the major changes were observed in the animal spleen such as a significant loss of global DNA methylation 6 hr, 96 hr, and 14 days after irradiation, resembling those induced in whole body irradiated rats. These changes also include DNA binding protein methylation, expression of methylotransferases and the methyl group binding retrotransposomal element LINE-1, and overexpression of micro RNA, miR-194. Therefore, these transcriptionally regulated epigenetic changes seem undoubtedly to be related to the radiation induced bystander effect, although they may be specific to certain tissues, because similar changes were absent in the dermal tissue (Koturbash et al, 2007, Ilnytskyy et al., 2009). *TP53* overexpression, change of proliferation rate measured as Ki67 antigen expression, as well as the increase in the percentage of apoptosis and DNA double strand breaks, the marker of which was the histone H2AX phosphorylation were also observed in bystander spleen of mice exposed to 1 Gy X-rays to their heads. These changes persisted from 24 hours to seven months (Koturbash et al, 2008). All of those experiments indicate that cells and tissues irradiated *in vivo* send signals which are transmitted by paracrine and endocrine systems and are able to induce damage in DNA, apoptosis, clastogenic effects, and epigenetic changes that lead to genetic instability. The consequence of the long-persisting changes may be the late effects including mutation and induction of second primary cancer. In good agreement with data presented above are results of the elegant study on bystander effect in human tissue models, which preserve the threedimensional structure and communication of cells present in tissues *in vivo* (Sedelnikova et al. 2007). The artificial skin which is able to survive 2-3 weeks in culture was irradiated with microbeam helium ions (7 MeV 4He, range in tissue 31 μm). The beam size was restricted to a 1 to 2 nuclei width along the line of irradiation. Bystander effect was studied on histological slides prepared at various time post irradiation (up to seven days). The authors observed increases in bystander cells the double strand breaks formation, followed by increased levels of apoptosis and micronucleus frequency, hypomethylation of nuclear DNA, and by an increased fraction of senescent cells. These findings point out the DNA DSBs induced by bystander signals as precursors of different cellular consequences in

**10. The potential clinical consequences of radiation induced bystander effect**  Although direct extrapolation of data from *in vitro* experiments to *in vivo* radiotherapy is not possible, (three-dimensional structure of tissues), one could assume that the bystander effect implies a risk of post-radiation complications in healthy tissues. It is suggested that genetic instability, which takes the form of delayed reproductive death (DRD), can participate in late side effects in patients treated with radiotherapy, because of damage, increased cell loss

review of Sgouros et al. (2007). E.g. are the protective or damaging effects specific for different types of radionuclides or type of cell? Is the in vivo bystander effect restricted to the damage to DNA by ionization secondary to Auger-elctron cascade or is it also possible when radionuclides deposit their energies within the cell cytoplasm or membrane? Additional studies are required to fully understand the bystander effects in radionuclide therapy.

#### **9.** *In vivo* **bystander effect**

Bystander effect in tissues distant from the radiation field, named "abscopal effect", was observed more than 50 years ago as haematological changes of bone marrow in children, who were given radiotherapy to the spleen in the treatment of leukemia (Parsons et al., 1954). Until recently, the abscopal effect was referred to the distant effects seen after local radiation therapy. Although the abscopal effect is potentially important for tumor control, it is still extremely controversial. However, it inspired *in vitro* and *in vivo* studies. It is believed mediated through cytokines and/or the immune system and results from loss of growth stimulatory or immunosuppressive factors from the tumor (Kaminski et al., 2005). The observation that irradiation of a murine tumor caused growth inhibition of another tumor outside of the radiation field was explained as the effect of immune system activation (Demaria et al., 2004). Interestingly, growth inhibition of tumors remote from the radiation field was tumor-specific. Camphausen et al. (2003) observed an abscopal effect as significant growth delay of distally implanted Lewis lung carcinoma and T241 fibrosarcoma cells in mice when they irradiated the non-tumorbearing legs. Furthermore, the authors compared this effect after fractionated irradiation with five 10-Gy fractions or twelve 2-Gy fraction and found dose dependent inhibition of tumor growth, being greater with higher fraction dose. Persuasive evidence of the bystander effect presence *in vivo* comes from experiments on rats in which the bases of the lungs were exposed to 10 Gy, while the remaining 70% of lungs were protected (Khan et al., 2003). A considerable increase in the DNA damage (micronuclei) was observed in the shielded lung. In addition, various parts of the lungs differed in the micronuclei frequency in response to direct irradiation, or only to bystander signals. The protective effects of two radical scavengers, superoxide dismutase (SOD) and nitro-L-arginine methyl ester (L-NAME), suggest that inflammatory cytokines induced by the irradiation may be involved in the initiation of a reaction generating reactive oxyradicals and nitric oxide that cause indirect DNA damage, both in and out of the radiation field (Khan et al, 2004). The mediators of bystander effect *in vivo* may be macrophages and inflammatory cytokines. Calveley et al. (2005) showed that activation of macrophages and expression of inflammatory cytokines fluctuated in a cyclic pattern in the directly irradiated and bystander regions of the same lung tissues. Cytokines including IL-1a, IL-1 IL-6, TNF-a and TGF-β were expressed to a similar degree in both, radiation targeted and non targeted lung tissues when measured on RNA levels. The results of animal studies involving irradiation of one side of the mouse body with 1 Gy X-ray showed DNA DSBs induction and increase in the levels of Rad51 (DSBs repairing protein) in non exposed skin (completely protected by lead shield). Furthermore, the levels of two methyl-binding proteins known to be involved in transcriptional silencing, MeCP2 and MBD2, were also increased in bystander tissue suggesting that radiation induced bystander effect may be epigenetically regulated. Global DNA hypomethylation is a typical feature of cancer cells. The methylation is one of the many types of histone modification processes which include, phosphorylation,

review of Sgouros et al. (2007). E.g. are the protective or damaging effects specific for different types of radionuclides or type of cell? Is the in vivo bystander effect restricted to the damage to DNA by ionization secondary to Auger-elctron cascade or is it also possible when radionuclides deposit their energies within the cell cytoplasm or membrane? Additional studies are required to fully understand the bystander effects in radionuclide

Bystander effect in tissues distant from the radiation field, named "abscopal effect", was observed more than 50 years ago as haematological changes of bone marrow in children, who were given radiotherapy to the spleen in the treatment of leukemia (Parsons et al., 1954). Until recently, the abscopal effect was referred to the distant effects seen after local radiation therapy. Although the abscopal effect is potentially important for tumor control, it is still extremely controversial. However, it inspired *in vitro* and *in vivo* studies. It is believed mediated through cytokines and/or the immune system and results from loss of growth stimulatory or immunosuppressive factors from the tumor (Kaminski et al., 2005). The observation that irradiation of a murine tumor caused growth inhibition of another tumor outside of the radiation field was explained as the effect of immune system activation (Demaria et al., 2004). Interestingly, growth inhibition of tumors remote from the radiation field was tumor-specific. Camphausen et al. (2003) observed an abscopal effect as significant growth delay of distally implanted Lewis lung carcinoma and T241 fibrosarcoma cells in mice when they irradiated the non-tumorbearing legs. Furthermore, the authors compared this effect after fractionated irradiation with five 10-Gy fractions or twelve 2-Gy fraction and found dose dependent inhibition of tumor growth, being greater with higher fraction dose. Persuasive evidence of the bystander effect presence *in vivo* comes from experiments on rats in which the bases of the lungs were exposed to 10 Gy, while the remaining 70% of lungs were protected (Khan et al., 2003). A considerable increase in the DNA damage (micronuclei) was observed in the shielded lung. In addition, various parts of the lungs differed in the micronuclei frequency in response to direct irradiation, or only to bystander signals. The protective effects of two radical scavengers, superoxide dismutase (SOD) and nitro-L-arginine methyl ester (L-NAME), suggest that inflammatory cytokines induced by the irradiation may be involved in the initiation of a reaction generating reactive oxyradicals and nitric oxide that cause indirect DNA damage, both in and out of the radiation field (Khan et al, 2004). The mediators of bystander effect *in vivo* may be macrophages and inflammatory cytokines. Calveley et al. (2005) showed that activation of macrophages and expression of inflammatory cytokines fluctuated in a cyclic pattern in the directly irradiated and bystander regions of the same lung tissues. Cytokines including IL-1a, IL-1 IL-6, TNF-a and TGF-β were expressed to a similar degree in both, radiation targeted and non targeted lung tissues when measured on RNA levels. The results of animal studies involving irradiation of one side of the mouse body with 1 Gy X-ray showed DNA DSBs induction and increase in the levels of Rad51 (DSBs repairing protein) in non exposed skin (completely protected by lead shield). Furthermore, the levels of two methyl-binding proteins known to be involved in transcriptional silencing, MeCP2 and MBD2, were also increased in bystander tissue suggesting that radiation induced bystander effect may be epigenetically regulated. Global DNA hypomethylation is a typical feature of cancer cells. The methylation is one of the many types of histone modification processes which include, phosphorylation,

therapy.

**9.** *In vivo* **bystander effect** 

acetylation, and ubiquitination, referred to epigenetic changes. Pogribny et al. (2004, 2005) investigated the effect of low-dose radiation exposure on the accumulation of DNA lesions and alterations of DNA methylation and histone H4-Lys20 trimethylation in the thymus tissue using an *in vivo* murine model. They found that fractionated whole-body application of 0.5 Gy X-ray leads to decrease in histone methylation and DNA damage accumulation in the thymus gland. The radiation-induced global genome DNA methylation changes were shown to be dose-dependent, sex- and tissue specific and long-persistant. Tissue specificity of bystander responses within the same organism has also been examined by Ilnytskyy et al. (2009). They analyzed changes in global DNA methylation in spleen of mice whole-body or cranial exposed to single 0.5 Gy of X-rays or to the same dose given in five 0.1 Gy fractions. After acute cranial exposure the major changes were observed in the animal spleen such as a significant loss of global DNA methylation 6 hr, 96 hr, and 14 days after irradiation, resembling those induced in whole body irradiated rats. These changes also include DNA binding protein methylation, expression of methylotransferases and the methyl group binding retrotransposomal element LINE-1, and overexpression of micro RNA, miR-194. Therefore, these transcriptionally regulated epigenetic changes seem undoubtedly to be related to the radiation induced bystander effect, although they may be specific to certain tissues, because similar changes were absent in the dermal tissue (Koturbash et al, 2007, Ilnytskyy et al., 2009). *TP53* overexpression, change of proliferation rate measured as Ki67 antigen expression, as well as the increase in the percentage of apoptosis and DNA double strand breaks, the marker of which was the histone H2AX phosphorylation were also observed in bystander spleen of mice exposed to 1 Gy X-rays to their heads. These changes persisted from 24 hours to seven months (Koturbash et al, 2008). All of those experiments indicate that cells and tissues irradiated *in vivo* send signals which are transmitted by paracrine and endocrine systems and are able to induce damage in DNA, apoptosis, clastogenic effects, and epigenetic changes that lead to genetic instability. The consequence of the long-persisting changes may be the late effects including mutation and induction of second primary cancer. In good agreement with data presented above are results of the elegant study on bystander effect in human tissue models, which preserve the threedimensional structure and communication of cells present in tissues *in vivo* (Sedelnikova et al. 2007). The artificial skin which is able to survive 2-3 weeks in culture was irradiated with microbeam helium ions (7 MeV 4He, range in tissue 31 μm). The beam size was restricted to a 1 to 2 nuclei width along the line of irradiation. Bystander effect was studied on histological slides prepared at various time post irradiation (up to seven days). The authors observed increases in bystander cells the double strand breaks formation, followed by increased levels of apoptosis and micronucleus frequency, hypomethylation of nuclear DNA, and by an increased fraction of senescent cells. These findings point out the DNA DSBs induced by bystander signals as precursors of different cellular consequences in human tissues.
