KEY POINTS
- This experimental study investigated radiation-induced bystander effects (RIBEs) mediated by soluble factors released from proton-irradiated SW1353 human chondrosarcoma cells. Conditioned medium from irradiated tumor cells was transferred to two nonirradiated recipient models: T/C-28A2 immortalized human chondrocytes and EA.hy926 endothelial-like cells, a hybrid line derived from endothelial and A549 lung carcinoma cells. SW1353 cells received 0.1, 0.5, or 2 Gy of 10.6-MeV protons at LET 12.6 keV/μm. Conditioned medium was collected 24 hours after irradiation, followed by recipient-cell exposure and functional assessment. Most experiments involved three independent biological replicates.
- Bystander signaling significantly reduced clonogenic survival, with different donor irradiation doses producing the strongest responses in the two cell models. In T/C-28A2 chondrocytes, conditioned medium from tumor cells irradiated with 0.1 Gy reduced survival to 79 ± 4.7% of control (p=0.017). In EA.hy926 endothelial cells, the greatest reduction occurred after exposure to conditioned medium from 2-Gy-irradiated tumor cells, with survival decreasing to 59 ± 9.5% of control (p=0.0069). Significant between-cell-line differences were observed at both 0.1 Gy (p=0.023) and 2 Gy (p<0.001). Importantly, the recipient cells were not directly irradiated, and no parallel photon-irradiation comparator was included.
- Micronucleus formation confirmed genotoxic effects in nonirradiated recipient cells. Chondrocytes exposed to conditioned medium from 0.1-Gy-irradiated SW1353 cells developed approximately 120 micronuclei per 1,000 binucleated cells (p=0.018), while endothelial cells exposed to medium from 2-Gy-irradiated SW1353 cells developed approximately 200 micronuclei per 1,000 binucleated cells (p<0.001). No significant genotoxic response was observed at 0.5 Gy in either model. These findings indicate a nonlinear relationship between donor irradiation dose and bystander genotoxicity rather than a simple increase in injury with increasing dose.
- Cell-cycle responses differed substantially between the two models. T/C-28A2 chondrocytes showed no statistically significant changes in G1, S, or G2 cell-cycle distribution at three or 24 hours after exposure to conditioned medium from 0.1-Gy-irradiated chondrosarcoma cells. In contrast, EA.hy926 endothelial cells exposed to medium from 2-Gy-irradiated cells demonstrated a significant reduction in the G1 population at three hours (p=0.012) and in the G2 population at 24 hours (p=0.002). These changes suggest greater cell-cycle disruption under the selected endothelial experimental condition, without establishing a specific cell-cycle checkpoint mechanism.
- Oxidative stress responses followed opposite patterns in chondrocytes and endothelial-like cells. After exposure to conditioned medium from 0.1-Gy-irradiated tumor cells, chondrocytes demonstrated an early reduction in mitochondrial superoxide of approximately 10% (p=0.015) and total ROS-related signal of approximately 18% (p<0.001) at three hours, followed by partial mitochondrial ROS recovery. In endothelial cells exposed to 2-Gy-derived conditioned medium, mitochondrial superoxide increased significantly at 24 hours (p=0.0014), while total ROS remained elevated at both three and 24 hours. These observations are consistent with different redox responses but cannot establish whether the changes are protective, damaging, or adaptive in intact tissues.
- Quantitative proteomics identified extensive but largely nonoverlapping changes in protein abundance. Chondrocytes exposed to conditioned medium from 0.1-Gy-irradiated SW1353 cells showed 85 differentially abundant proteins, including 50 increased and 35 decreased, whereas endothelial-like cells exposed to 2-Gy-derived medium showed 427 differentially abundant proteins, including 238 increased and 189 decreased. Only three proteins were shared between the two profiles. Functional enrichment implicated pathways associated with cartilage development, extracellular matrix organization, IL-17, PI3K-AKT, and TNF signaling in chondrocytes, compared with mitochondrial metabolism, oxidative phosphorylation, cell-cycle regulation, and DNA repair in endothelial cells. These pathway associations do not establish direct causal mechanisms.
- The proteomic findings were highly sensitive to the statistical correction method. The reported 85 and 427 proteins were identified using an empirical false-discovery-rate approach and a fold-change threshold. Under the more conservative Benjamini–Hochberg correction, only two proteins remained significant in chondrocytes and none in endothelial cells; a paired-replicate analysis with the same correction yielded no significant proteins in either model. Furthermore, the detailed proteomic experiments compared different donor irradiation doses (0.1 vs 2 Gy), preventing separation of cell-type effects from dose effects. The small biological sample, absence of primary healthy tissue models, and lack of in-vivo experiments limit mechanistic and clinical interpretation.
CLINICAL TAKEAWAY
Conditioned medium from proton-irradiated chondrosarcoma cells induced distinct genotoxic, redox, and survival responses in nonirradiated chondrocyte and endothelial-like cell models, supporting further investigation of radiation-induced bystander signaling. However, the proteomic results were highly dependent on statistical methodology, and different irradiation doses were used for the detailed cellular comparisons. These findings do not establish normal-tissue toxicity in patients or identify validated therapeutic targets, and require confirmation using primary cells, matched-dose experiments, and in-vivo models.