The role of TGF-|[beta]|1|[ndash]|miR-21|[ndash]|ROS pathway in bystander responses induced by irradiated non-small-cell lung cancer cells
yuyang jiang,xin chen,weidong tian,xuedong yin,jin wang,huanghao yang
DOI: https://doi.org/10.1038/bjc.2014.368
IF: 9.075
2014-01-01
British Journal of Cancer
Abstract:The traditional dogma claims that biological effects of ionising radiation require direct damage to DNA. However, the occurrence of radiation-induced bystander effects (RIBEs) challenges this dogma. RIBEs refer to the biological responses occurred in unirradiated cells through intercellular communication when the neighbouring cells are traversed by ionising radiation. RIBEs have attracted an intense interest since 1992, when Nagasawa and Little (1992) provided direct evidence for the existence of this effect. To date, RIBEs have been demonstrated both in vitro (Azzam et al, 2001; Zhou et al, 2001; Lyng et al, 2002; Shao et al, 2004; Yang et al, 2005) and in vivo (Koturbash et al, 2006; Mancuso et al, 2008; Jain et al, 2011). Besides normal cells, cancer cells have recently been found to express RIBEs (Burdak-Rothkamm et al, 2007; Harada et al, 2009; Rzeszowska-Wolny et al, 2009; Herok et al, 2010). Moreover, a recent study (McMahon et al, 2013) suggests that bystander effects may be essential in producing a robust decrease in cancer cell survival in directly irradiated populations. Another study (Aravindan et al, 2014) strongly suggests that acquired tumour cell radiation resistance at the treatment site is mediated through radiation-orchestrated intercellular communication. All these results indicate a potential important clinical implication of bystander effects in the context of radiotherapy. Although the existence of RIBEs has been well recognised and documented, the underlying molecular mechanisms are poorly understood. So far, intercellular gap junctions (Azzam et al, 2001; Zhou et al, 2001), reactive oxygen species (ROS) (Azzam et al, 2002; Yang et al, 2005) and soluble signalling molecules such as cytokines (Ivanov et al, 2010; Hei et al, 2011) have been demonstrated to mediate bystander response. Among cytokines, tumour growth factor-beta1 (TGF-β1) has been found to be an important mediator of bystander effects (Iyer et al, 2000; Shao et al, 2008; Gow et al, 2010; Temme and Bauer, 2013). However, the detailed mechanisms for how TGF-β1 mediates RIBEs remain to be elucidated. It has recently been proposed that bystander effects may be epigenetically mediated (Kovalchuk and Baulch, 2008; Ilnytskyy and Kovalchuk, 2011; Mothersill and Seymour, 2012). microRNAome changes have been demonstrated in bystander cells and tissues (Koturbash et al, 2007, 2008; Kovalchuk et al, 2010; Chaudhry and Omaruddin, 2012), suggesting the involvement of miRNA modulation in RIBEs. However, through miRNA knockdown, Dickey et al (2011) found no difference in DNA damage in bystander cells between wild type and knockdown cells with depleted levels of mature miRNAs, suggesting that miRNAs were not the primary signalling factor in the induction of bystander effect, although they had a role in RIBE manifestation. Thus, the roles of miRNAs in bystander signalling need to be further explored. In the present study, we investigated the roles of TGF-β1 and miR-21 in medium-mediated RIBEs in H1299 human non-small-cell lung cancer cells. We found that different biological changes were induced in bystander cells at different times post radiation. The TGF-β1 signalling pathways in both irradiated and bystander cells were critical to the induction of RIBEs. TGF-β1-dependent ROS appeared to be involved in early but not in late bystander responses. We also found that miR-21 in bystander cells was regulated by the TGF-β1 pathway and was an important mediator of bystander signalling. 8-hydroxy-2′-deoxyguanosine detection was performed as previously described (Moiseeva et al, 2009). In brief, bystander cells were fixed with 4% paraformaldehyde for 15 min at room temperature (RT) after treatment with conditioned medium. Then cells were sequentially incubated with 2 M HCl for 20 min and 0.1 mM sodium borate (pH 8.5) for 2 min at RT. After wash, cells were permeabilized and incubated with primary antibody against 8-hydroxy-2′-deoxyguanosine (Abcam) overnight at 4 °C, followed by incubation with secondary antibody. Then cells were counter stained with 4′, 6′-diamidimo-2-phenylindole and mounted on slides. The fluorescence intensity of at least 100 cells was recorded using TCS SP2 Laser Scanning Confocal Microscope (Leica, Solms, Germany), the data were analysed with the Leica-SP2 quantification software (Leica Microsystems). Cells treated with 100 μM H2O2 were used as positive control. By pretreating signalling cells with SB431542 before irradiation, we found that the percentage of bystander cells with 53BP1 foci in 1-h RCM was back to the level in SCM (Figure 2B), and the proliferation inhibition in bystander cells in 18-h RCM was eliminated from 15% to 3% (Figure 2C). The findings suggested that inhibiting the TGF-β1 pathway in signalling cells abolished DNA damage and proliferation inhibition in bystander cells. After adding SB431542 into conditioned medium before transfer, we found that SB431542 at 10 μM alone did not significantly affect the percentage of H1299 cells with 53BP1 foci (Figure 2B) and cell proliferation (data not shown). However, in the presence of SB431542, the percentage of bystander cells with 53BP1 foci in 1-h RCM was back to the level in SCM (Figure 2B), and the proliferation inhibition in H1299 bystander cells was attenuated from 15% to 5% (Figure 2C). The data suggested that inhibiting the TGF-β1 pathway in bystander cells also abolished the DNA damage and proliferation inhibition in bystander cells. These results indicated that the TGF-β1 signalling pathways in both irradiated signalling and bystander cells were critical to induction of bystander effects. We also measured the level of oxidative DNA damage in bystander cells. As shown in Figure 3C, compared with in fresh medium, the level of oxidative DNA damage of cells decreased obviously after cultured in 1-h SCM for 1 h, but remained similar in 1-h RCM. Bystander cells in 1-h RCM showed more significant oxidative DNA damage than cells in 1-h SCM. Thus, the changes in oxidative DNA damage were in accordance with the changes in ROS level in cells cultured with 1 h conditioned medium. In addition, our unpublished data showed no obvious oxidative DNA damage in bystander cells in 18-h RCM (data not shown), which agreed with no increase in ROS levels with 18-h RCM culture (Figure 3A). In addition, when signalling cells were treated with SB431542 1 h before irradiation, the generated conditioned medium failed to cause an increase in the ROS levels in bystander cells (Figure 3A), suggesting that activating the TGF-β1 pathway in signalling cells was critical to the increase in ROS levels in bystander cells. Moreover, when SB431542 was added into 1-h RCM, no increase in the ROS levels in bystander cells was detected (Figure 3A), indicating that the increase in ROS levels in bystander cells was dependent on activation of their TGF-β1 pathway after cultured in RCM. We then assessed whether alterations in miR-21 expression in bystander cells were dependent on the TGF-β1 pathway. When the TGF-β1 pathway in signalling cells was inhibited by SB431542, the generated 1-h SCM and 1-h RCM increased miR-21 expression in bystander cells compared with in the absence of SB431542, but the difference of miR-21 expression in bystander cells in 1-h RCM and in 1-h SCM was significantly attenuated (Figure 4A). The generated 18-h SCM and 18-h RCM also increased miR-21 expression in bystander cells; however, 18-h RCM caused a greater increase than 18-h SCM, thus the reduction of miR-21 expression in bystander cells in 18-h RCM compared with in 18-h SCM observed in the absence of SB431542 disappeared (Figure 4B). When the TGF-β1 pathway in bystander cells was inhibited by adding SB431542 into conditioned medium, we found that SB431542 alone caused about 2.5-fold increase in miR-21 expression in cells, but the difference of miR-21 expression in bystander cells in 1-h SCM and 1-h RCM observed in the absence of SB431542 decreased by 50% (Figure 4A). Similarly, inhibition of the TGF-β1 pathway in bystander cells resulted in abolishment of the reduction of miR-21 expression in bystander cells in 18-h RCM compared with in 18-h SCM in the absence of SB431542 (Figure 4B). These data suggested that the alterations of miR-21 expression in bystander cells were dependent on the TGF-β1 pathways in both signalling and bystander cells. On one hand, upregulation of miR-21 alone increased the percentage of cells with 53BP1 foci by 17% (Figure 5C). This was in accordance with the increased miR-21 level and induction of 53BP1 foci in bystander cells in 1-h RCM. When the bystander cells transfected with miR-21 inhibitor were cultured in 1-h RCM, no induction of 53BP1 foci was observed (Figure 5C). Moreover, downregulation of miR-21 significantly attenuated the ROS levels in bystander cells cultured with 1-h RCM back to that with 1-h SCM (Figure 5D). The results indicated that elevated miR-21 expression in bystander cells led to the increase in the ROS levels and DNA damage. On the other hand, when the miR-21 levels in cells were downregulated by transfection with a miR-21 inhibitor, the proliferation was reduced by 18% compared with the relative control cells transfected with negative inhibitor (Figure 5E). This was in accordance with the decreased miR-21 levels and proliferation inhibition in bystander cells cultured in 18-h RCM. Furthermore, when the bystander cells transfected with miR-21 mimics were cultured in 18-h RCM, the proliferation was back to the level in cells cultured in 18-h SCM (Figure 5E). The results indicated that the decreased miR-21 levels in bystander cells inhibited the cell proliferation. In the present study, we demonstrated that X-irradiation could induce medium-mediated bystander effects in H1299 cells. Moreover, RCM harvested at different times post irradiation caused completely different biological changes. One-hour RCM induced elevated miR-21 expression, increased ROS levels and DNA damage in bystander cells, while 18-h RCM resulted in reduced miR-21 level and inhibited proliferation in bystander cells. This result is similar to that in a previous study (Zhang et al, 2009) showing that the radiation-induced bystander mutation was dependent on the time point at which conditioned medium was harvested. Furthermore, our results suggested that at different times post irradiation, bystander cells received different signals from irradiated cells that were in different stages of DNA damage response, thus activating different pathways and yielding different changes at the cellular level. In spite of a large body of evidence supporting the occurrence of RIBEs, there is some debate regarding whether RIBEs are universal (Groesser et al, 2008; Sowa et al, 2010). Besides the possible explanations for why RIBEs were not observed in some tested systems, such as the epigenetic status of a specific cell line or the precise culture conditions and medium supplements (Groesser et al, 2008; Sowa et al, 2010), our results suggested that appropriate end point examined at appropriate time could be an important factor. TGF-β1 has been shown to mediate RIBEs (Iyer et al, 2000; Shao et al, 2008; Gow et al, 2010; Temme and Bauer, 2013). TGF-β1 can be released from irradiated cells into conditioned medium and can diffuse freely to unirradiated cells to induce bystander effects (Shao et al, 2008). This may be the case in our experiment with 18-h RCM, where elevated TGF-β1 levels were detected. However, when bystander cells were cultured in 1-h RCM, where no TGF-β1 release from irradiated cells was observed, bystander DNA damage occurred. It suggests that active TGF-β1 in conditioned medium was not necessary for the induction of bystander DNA damage. When we pretreated irradiated cells with SB431542, the bystander effects were completely abolished. This result suggests that the TGF-β1 signalling pathway in irradiated cells was critical to RIBEs regardless of their status of active TGF-β1 secretion. Similarly, when we added SB431542 into RCM, the bystander effects were attenuated. This result suggests that the TGF-β1 signalling pathway in bystander cells could be activated by bystander signalling factors in RCM, thereby mediating RIBEs. miRNA modulation has been shown to be involved in RIBEs (Koturbash et al, 2008; Kovalchuk et al, 2010; Chaudhry and Omaruddin, 2012). But Dickey et al (2011) reported that miRNAs did not seem to be the primary signalling factor associated with bystander DNA damage. Therefore, the roles of miRNAs in bystander signalling remain to be explored. We report here for the first time that miR-21 acted as an important mediator of RIBEs. Interestingly, miR-21 of bystander cells dysregulated differently after cultured in RCM harvested at different times post irradiation (that is, it was elevated when bystander cells were in 1-h RCM, but was decreased in 18-h RCM). Most importantly, elevated miR-21 levels of bystander cells cultured in 1-h RCM seemed to be critical to the increase in ROS levels and bystander DNA damage, as these effects were abolished when miR-21 was downregulated in bystander cells before cultured in RCM. It has been reported that overexpression of miR-21 can significantly repress the production of superoxide dismutase 2 in cells, which results in increased intracellular ROS levels (Fleissner et al, 2010). Moreover, Zhang et al (2012) recently revealed that miR-21 regulates cellular ROS levels by directly attenuating superoxide dismutase 3 or by indirectly reducing superoxide dismutase 2 levels through tumour necrosis factor-α. Our data also show a reduced superoxide dismutase 2 expression in cells in 1-h RCM compared with in 1-h SCM (Jiang et al, 2014). Furthermore, when bystander cells were under oxidative stress, DNA damage occurred. On the other hand, the proliferation inhibition in bystander cells cultured in 18-h RCM was related to the decreased miR-21 expression. This is in agreement with previous reports demonstrating that the upregulation of miR-21 promotes cell growth and blocking miR-21 leads to proliferation inhibition (Gong et al, 2011; Huang et al, 2013). Therefore, differential dysregulation of miR-21 may explain why bystander cells cultured in RCM harvested at different times post irradiation expressed different changes at the cellular level. This indicates an important mediating role for miR-21 in the induction of RIBEs. We have also found that dysregulation of miR-21 in bystander cells was dependent on TGF-β1 signalling pathways in both signalling and bystander cells. The relationship between TGF-β1 and miR-21 seems complicated. On one hand, miR-21 can have both positive and negative effects on the regulation of TGF-β1 signalling (Liu et al, 2011; Bhagat et al, 2013). On the other hand, miR-21 can be induced by TGF-β1 (Liu et al, 2011), TGF-β1/smad signalling can act as a critical upstream regulator of miR-21 (Wang et al, 2012). Our results indicated that the TGF-β1 signalling pathway in bystander cells regulated miR-21 differently at different times. While 1-h RCM caused miR-21 upregulation in bystander cells, 18-h RCM induced miR-21 downregulation in bystander cells. Since the addition of SB431542 into RCM diminished bystander effects in both situations, it is suggested that the TGF-β1 pathway was activated in bystander cells cultured with either 1-h or 18-h RCM. However, the reason why the activated TGF-β1 pathway caused different changes in miR-21 expression in bystander cells is unknown. One possible explanation is that the signals triggering activation of the TGF-β1 pathway in bystander cells at different time may be different. Moreover, Figure 4 shows that while the downregulation of miR-21 induced by 18-h RCM is abolished by adding SB431542 into RCM, 1-h RCM-induced miR-21 upregulation is only partially attenuated in the presence of SB431542, suggesting that the upstream regulators other than the TGF-β1 pathway may be involved in the increase in miR-21 expression of bystander cells cultured in 1-h RCM. In summary, the conditioned medium from irradiated H1299 human lung cancer cells induces different biological changes in bystander cells at different times post irradiation. The activation of the TGF-β1 pathway in irradiated cells is involved in bystander signalling. Moreover, the TGF-β1–miR-21–ROS pathway in bystander cells is critical for the manifestation of bystander responses. Figure 6 illustrates our proposed working model of the radiation-induced bystander signalling pathway involved. When cells are irradiated, the TGF-β1 pathway becomes activated and cells release some signalling molecules including TGF-β1 into the medium. These signalling molecules diffuse to bystander cells and activate the TGF-β1 pathway in bystander cells. Depending on the time at which RCM is harvested post irradiation, the activated TGF-β1 pathway can either upregulate or downregulate miR-21 expression in bystander cells. The increased miR-21 levels then cause oxidative stress in bystander cells, leading to DNA damage. On the other hand, decreased miR-21 levels induce cell cycle delay and proliferation inhibition. However, there are still some unanswered questions regarding this bystander signalling pathway. For example, what is (are) the signalling molecule (s) in 1-h RCM responsible for activating the TGF-β1 pathway in bystander cells? Although the TGF-β1 pathway can be activated in bystander cells cultured in RCM harvested at different times post irradiation, why does it dysregulate miR-21 differently, which in turn causes different changes at the cellular level? These questions warrant further investigation. We thank Professor Tom K. Hei of Columbia University for his critical reading of this manuscript. This research was supported in part by the National Natural Science Foundation of China (grant No. 31270898 and 11335011), the Key programs of Natural Science Foundation of Jiangsu Educational Committee (12KJA310005), Beijing Key Laboratory of DNA Damage Response (BKL-DDR) open funding and the Priority Academic Program Development of Jiangsu Higher Education Institution (PARD).