Radiation Damage Travels: How Bystander Cells Rewrite the DNA Story

For nearly a century, radiobiology rested on a deceptively simple idea: radiation kills cells by hitting DNA directly, and the damage stops where the energy deposition stops. A new narrative review by Carmel Mothersill, Andrej Rusin and Colin Seymour of McMaster University, published in Frontiers in Oncology, argues that this target-theory framework, formalized in the 1930s, has been quietly undermined by evidence that has accumulated since 1905. Non-targeted effects, or NTEs, are biological responses detected in cells, tissues and even organisms that never received a direct radiation hit. The review, drawing on almost 50 years of work from the authors’ laboratories in Ireland and Canada, makes the case that these effects are not curiosities at the margins of the field but system-level mechanisms that coordinate DNA damage responses across populations of cells, with profound implications for cancer therapy and low-dose radiation protection.

The modern story begins in the mid-1980s, when a series of papers challenged the assumption that radiation damage expressed itself only in the first divisions after exposure. Researchers demonstrated that lethal mutations and chromosomal aberrations could appear in distant progeny of irradiated cells, after many apparently normal divisions. Because the damage was lethal and non-clonal, it could not have been carried by the cells themselves; instead, the genome had become less able to eliminate or correct mistakes. This phenomenon is now known as radiation-induced genomic instability, or RIGI. The second pillar of NTE research came in 1992, when Nagasawa and Little exposed Chinese hamster ovary cells to such a low fluence of alpha particles that fewer than 1 percent of cell nuclei were directly traversed, yet sister chromatid exchanges appeared in roughly 30 percent of the population. Radiation-induced bystander effects, or RIBEs, were born: signals produced by irradiated cells could induce damage in unexposed neighbors.

Subsequent decades established the reality of bystander signaling using a suite of experimental protocols. Medium transfer experiments showed that culture medium harvested from irradiated cells could reduce the clonogenic survival of unirradiated recipients. Microbeam studies, in which single cells were targeted with counted particles, revealed that damage signals spread up to a millimeter away from the hit site. Transwell systems, in which irradiated and unirradiated cultures shared medium but not physical contact, confirmed that soluble factors were involved. Whole-animal models extended the phenomenon beyond the dish. Strikingly, the review documents that effects of this kind were recorded as early as 1968, when plasma from patients receiving X-ray therapy or total-body irradiation was shown to induce chromosomal aberrations in cultured lymphocytes from healthy donors, and again in 1977 with plasma from atomic bomb survivors. The phenomena were variously called indirect, abscopal, clastogenic or out-of-field effects, but all involved damage at sites in space or time that received no energy deposition.

How does a radiation hit in one cell produce DNA damage in another? The first mechanistic answer came from Little’s laboratory, which showed that gap junction intercellular communication, or GJIC, was involved in bystander signaling after alpha-particle irradiation. The inhibitor lindane suppressed the effect, and connexin proteins forming the junctions were upregulated in cultures exposed to ultra-low fluences. But GJIC could not explain medium transfer experiments, where the signal survived in culture medium for hours. The leading explanation there involved reactive oxygen species and reactive nitrogen species generated by direct irradiation, which induced secondary damage in surrounding cells. Yet ROS and NOS are short-lived, so they too could not fully account for medium transfer, which triggers apoptosis, calcium release, mitochondrial membrane depolarization and caspase activation in recipients. The review notes that no single magic factor was ever characterized; some experiments from the authors’ group suggested the bystander effect might actually arise from a loss of factors rather than active secretion, with serotonin naturally present in serum bound by irradiated cells and triggering downstream calcium release in recipients.

Mitochondria have emerged as central players in this signaling web. Lyng and colleagues reported mitochondrial membrane depolarization as a robust marker of bystander signaling, and altered energy production was documented in both targeted and non-targeted cells. Nugent and colleagues measured direct effects on mitochondrial DNA and increased mitochondrial mass in cells with functionally compromised mitochondria after both direct gamma radiation and bystander factors. A particularly striking line of work from the McMaster group established a connection between mitochondrial complex 1 function and radiation-induced UVA biophoton emission from directly irradiated cells. This ultraweak ultraviolet emission proved both necessary and sufficient to produce bystander effects in unirradiated cells held in separate flasks with no medium contact. In parallel, other researchers showed that irradiated cells release exosomes containing miRNAs and even mitochondrial DNA that can communicate information to bystander cells. Reconciling the two hypotheses, Le and colleagues extracted exosomes from cultures that had sat above tritium-irradiated flasks, where beta tracks could not reach them, and demonstrated that the purified exosome fraction induced mitochondrial damage and reproductive cell death in a completely separate set of cells, but only when UVA photons had not been blocked by a yellow cellophane filter.

Perhaps the most provocative claim in the review is that NTEs, long measured using endpoints seen as bad at the level of the individual cell, are actually protective. Bystander signaling can induce adaptive responses, in which signals from irradiated cells prime unirradiated neighbors to resist subsequent genotoxic insults through enhanced DNA repair, altered cell-cycle regulation, antioxidant responses and elimination of damaged cells by apoptosis. The review links this to the phenomenon of low-dose hypersensitivity and induced radioresistance, or HRS/IRR, a dose-response pattern in which cells show a steep response below about 100 mGy followed by a plateau and a shallower curve above roughly 0.5 Gy. Joiner’s group attributed the transition to a threshold dose above which DNA repair becomes inducible, involving PARP and ATR signaling. Collaborative experiments showed that bystander effects could be demonstrated in the HRS region of the dose-response curve but were neutralized when the dose applied in the bystander assay fell in the IRR range, suggesting that RIBE is a low-dose phenomenon that can be counteracted if DNA repair is induced, and that it may account for part of the enigmatic shoulder of the classical survival curve.

Dose and dose rate studies sharpen the picture. Experiments by the authors’ group and Prise’s group determined that NTEs are triggered by very low doses, around 2 to 3 mGy of low-LET radiation, and appear to saturate above 0.5 Gy. A survival curve reconstructed over seven log cycles of dose showed three distinct phases: below 2 mGy no bystander effect, between 2 and 500 mGy no difference between directly irradiated cells and cells receiving irradiated-cell conditioned medium, and above that a clear divergence where directly irradiated cells followed the classic dose response while medium-transfer cells showed no further increase in reproductive death. The calcium spike that Lyng identified as the first sign of bystander signaling was detected only when conditioned medium came from cells receiving 2 mGy or more, indicating a mechanistic threshold. Radiation quality adds further nuance: alpha particles, X-rays and gamma rays all produce bystander effects and genomic instability, but the evidence suggests that pure neutrons, when the gamma contamination is shielded out, do not produce NTE at low doses, a conclusion supported by zebrafish experiments in which robust gamma bystander signaling occurred but neutron exposure produced no effect.

Critics once dismissed NTEs as artifacts of cell culture, but the review assembles ample in vivo evidence. Cross-sex bone marrow transplant studies in mice demonstrated transmissible chromosomal instability in regenerating marrow. Abscopal effects in lung tissue after radiotherapy showed in vivo bystander responses, with the immune system implicated by Demaria and others. Bystander effects were found in serum from Chernobyl-affected populations as long as 20 years after the reactor explosion, and multigenerational genomic instability was documented in bank voles in the Belorussian Polessky reserve. Russian researchers showed that irradiated mice produce volatile signals causing chromosomal damage in unirradiated littermates, and the McMaster group extended inter-animal signaling to fish, earthworms, crickets and mice. Proteomic analysis of bystander fish revealed a protective proteome enriched in antioxidant and DNA repair proteins, which the directly irradiated fish conspicuously lacked, reinforcing the view that NTE signaling functions as an early-warning system coordinating population-level responses.

The implications ripple outward in several directions. For radiotherapy, especially new modalities such as FLASH protocols using ultra-high dose rates, out-of-field and abscopal effects raise questions about how bystander signaling plays out during treatment planning, and whether signals from the tumor ecosystem, including fibroblasts, immune cells and endothelial support vessels, could be exploited as biomarkers or therapeutic targets based on mitochondrial metabolism, the apoptotic cascade or serotonin receptors. For radiation protection, the review argues that NTEs are relevant to low-dose and chronic exposures because they involve signaling between and within populations of species in ecosystems, coordinating adaptation and evolution, concepts such as resilience, adaptation and historic dose. In the human protection field, however, BEIR and UNSCEAR reports merely suggest that no conclusions can be drawn about the relevance of NTE at this time. The authors conclude that imaging and therapy would both benefit from greater mechanistic understanding of NTE, and that these communication mechanisms are likely to be highly relevant to population responses to low-dose radiation at the level of cell, tissue and organism signaling.

Subject of Research: Non-targeted effects of radiation and DNA damage signaling

Article Title: Non-targeted effects and DNA: vulnerabilities at the interface

Article References: Mothersill, C., Rusin, A., & Seymour, C. (2026). Non-targeted effects and DNA: vulnerabilities at the interface. Frontiers in Oncology, 16, Article 1952660. https://doi.org/10.3389/fonc.2026.1952660

Image Credits: AI Generated

DOI: 10.3389/fonc.2026.1952660

Keywords: radiobiology, bystander effect, DNA repair, non-targeted effects, genomic instability, radiotherapy, radiation protection, mitochondria, exosomes, adaptive response, Non-targeted, effects