Barcode-Based Mutagenesis Assay Reveals Strand-Bias Rules Behind Oxidative DNA Damage

Every day, the DNA inside our cells sustains tens of thousands of chemical injuries, and among the most consequential of these are oxidative lesions produced as a byproduct of normal metabolism and of exposure to ionizing radiation. One of the best-studied of these lesions is 8-oxo-7,8-dihydroguanine, commonly abbreviated 8-oxo-G, an oxidized form of the DNA base guanine that can mispair with adenine during replication and thereby give rise to characteristic base substitutions. Yet a deceptively simple question has remained stubbornly difficult to answer: when a single oxidized base sits at one position in a DNA molecule, where exactly do the resulting mutations appear, and why do they cluster at some positions while sparing others? A new study from researchers at Hiroshima University, published in BMC Cancer, offers an unusually precise answer by combining a barcoded shuttle-vector sequencing assay with chronic low-dose-rate gamma irradiation and site-specific lesion placement.

The team, led by Hidehiko Kawai and Hiroyuki Kamiya, set out to dissect the mutational consequences of reactive oxygen species generated under chronic irradiation conditions that mimic low-dose, low-dose-rate exposures. This regime is biologically important because cancers arising after such exposures are largely indistinguishable from spontaneous tumors, making it extremely hard to define the mutational mechanisms at work. Under these conditions, radiation-generated reactive oxygen species are thought to inflict persistent low-level oxidative DNA damage rather than the dramatic double-strand breaks associated with high-dose irradiation. The central challenge has been to connect that diffuse chemical damage to specific, quantifiable mutational outcomes at defined DNA positions, something conventional mutagenesis assays have struggled to achieve with single-molecule resolution.

To overcome this limitation, the researchers employed a next-generation sequencing version of the classical supF shuttle vector assay, enhanced with a random twelve-nucleotide barcode incorporated into each vector molecule. This nucleotide-barcoding strategy is the technical heart of the study. Because every vector molecule carries a unique barcode, each independent mutational event can be identified and counted individually, allowing the team to distinguish true single-molecule mutation events from sequencing artifacts or clonal amplification. The assay was run under both standard supF-selection conditions and without selection, and mutation frequencies and spectra were quantified across the entire supF reporter gene. In parallel, the team prepared shuttle vector libraries carrying a single, site-specific 8-oxo-G lesion at defined positions, providing a controlled benchmark against which the irradiation-induced mutational patterns could be compared.

The results from chronic gamma irradiation were striking. Irradiation significantly increased mutations at C:G base pairs located within 5′-TCN-3′:5′-NGA-3′ sequence contexts, where N denotes any nucleotide. This trinucleotide-level specificity echoes the kind of sequence-context dependence that has become familiar from large-scale cancer genomics, in which different mutational processes leave characteristic footprints in particular sequence neighborhoods. But the more unexpected finding was positional: many of the positions that responded to irradiation were already spontaneous mutation hotspots, and they also coincided with distant C:G sites that were preferentially mutated in response to a single 8-oxo-G lesion placed elsewhere in the vector.

That coincidence points to the study’s most conceptually provocative result, a phenomenon the authors describe as strand-biased action-at-a-distance mutagenesis. When a single 8-oxo-G lesion was introduced at one defined site, mutations did not simply accumulate at or immediately adjacent to the lesion. Instead, the lesion induced extensive mutations at distant C:G sites, and these mutations showed a pronounced strand bias, preferentially affecting one strand of the duplex over the other. Chronic gamma irradiation further enhanced these distant mutations without substantially altering their positional distribution. In other words, irradiation did not create a new and distinct set of mutation sites; it amplified a pre-existing pattern of susceptible positions that was already latent in the molecule.

This amplification model has significant implications for how scientists interpret mutational signatures in cancer genomes. If radiation-induced mutagenesis largely reinforces spontaneous hotspots rather than generating an independent radiation-specific landscape, then the mutational signature of chronic low-dose radiation may be subtle and entangled with background processes, which is consistent with the clinical observation that radiation-associated cancers are so hard to distinguish from spontaneous ones. The finding also suggests that the identity of the damaged base may matter less than the structural and sequence context in which it sits, since a lesion at one position can shape mutational outcomes many bases away.

What could mechanistically explain action-at-a-distance mutagenesis and its strand bias? The authors found important clues in the local secondary structures that single-stranded DNA can adopt. During replication, transcription, or repair, stretches of DNA transiently become single-stranded, and these single-stranded regions are chemically vulnerable and can fold into hairpins and other structures. The study’s hotspot and coldspot behaviors, as well as position-dependent substitution patterns, were associated with predicted local secondary structures of single-stranded DNA. This suggests that structural context contributes to mutagenesis beyond the trinucleotide sequence context alone, adding a layer of physical geometry to the sequence-based models that currently dominate the field of mutational signature analysis.

Strand bias itself is a well-recognized feature of mutational processes in cancer genomics, often reflecting whether a lesion sits on the leading or lagging strand during replication, or on the transcribed versus non-transcribed strand during transcription-coupled repair. The strand-biased mutations observed here, driven by a single oxidative lesion and modulated by chronic irradiation, provide a controlled experimental system in which such biases can be studied mechanistically rather than inferred statistically from tumor sequencing data. The authors note that processes such as translesion DNA synthesis, in which specialized polymerases copy across damaged bases, and base excision repair, in which glycosylases such as OGG1 remove oxidized guanine, are natural candidates for the enzymatic pathways that convert a localized lesion into distant, strand-biased mutations, although the present study defines the patterns rather than fully resolving the responsible mechanisms.

Methodologically, the barcoded supF NGS platform represents a versatile new framework for mutagenesis research. Classical shuttle vector assays, which date back decades, allowed mutated plasmids recovered from cells to be sequenced, but they could not easily distinguish independent events or achieve the throughput needed to map mutation positions comprehensively. The addition of random twelve-nucleotide barcodes and next-generation sequencing changes that calculus, enabling precise quantification of mutation frequencies and spectra at single-molecule resolution. The authors position the platform as a general tool for mechanistic studies of oxidative, radiation-associated, and cancer-related mutagenesis, and it is easy to imagine its application to other lesions, other exposure regimes, and cells deficient in specific repair pathways.

The broader significance of the work lies in reframing how low-dose radiation risk might be understood at the molecular level. Rather than radiation writing a wholly new mutational script into the genome, the study suggests it turns up the volume on a script that oxidative metabolism has already written, preferentially at pre-existing susceptible sites whose locations are governed by sequence context and single-stranded DNA structure. For cancer epidemiology, this offers a potential explanation for why chronic low-dose exposures are so difficult to trace in tumor genomes, and for mutagenesis biology, it elevates DNA secondary structure from a curiosity to a first-class determinant of where mutations strike. As the authors conclude, chronic low-dose-rate gamma irradiation and a single 8-oxo-G lesion promote closely related mutational processes that act preferentially at pre-existing susceptible sites and generate strand-biased action-at-a-distance mutations, a set of strand-bias laws that may ultimately sharpen both radiation risk assessment and the interpretation of mutational signatures across human cancers.

Subject of Research: Strand-biased mutagenesis induced by oxidative DNA damage and chronic low-dose-rate gamma irradiation

Article Title: Systematic mutagenesis assay promotes comprehension of the strand-bias laws for mutations induced by oxidative DNA damage

Article References: Kawai, H., Ebi, S., Sugihara, R., Fujiwara, C., Fujikawa, Y., Kimura, S., & Kamiya, H. (2026). Systematic mutagenesis assay promotes comprehension of the strand-bias laws for mutations induced by oxidative DNA damage. BMC Cancer. https://doi.org/10.1186/s12885-026-17081-0

Image Credits: AI Generated

DOI: 10.1186/s12885-026-17081-0

Keywords: 8-oxo-G, oxidative DNA damage, ionizing radiation, mutagenesis, strand bias, mutation hotspots, supF shuttle vector, next-generation sequencing, reactive oxygen species, DNA secondary structure, cancer, base excision repair