Contributing factors to the oxidation-induced mutational landscape in human cells

Cordero, Cameron; Mehta, Kavi P. M.; Weaver, Tyler M.; Ling, Justin A.; Freudenthal, Bret D.; Cortez, David; Roberts, Steven A.

doi:10.1038/s41467-024-55497-z

Download PDF

Article
Open access
Published: 23 December 2024

Contributing factors to the oxidation-induced mutational landscape in human cells

Nature Communications volume 15, Article number: 10722 (2024) Cite this article

5901 Accesses
4 Citations
28 Altmetric
Metrics details

Subjects

Abstract

8-oxoguanine (8-oxoG) is a common oxidative DNA lesion that causes G > T substitutions. Determinants of local and regional differences in 8-oxoG-induced mutability across genomes are currently unknown. Here, we show DNA oxidation induces G > T substitutions and insertion/deletion (INDEL) mutations in human cells and cancers. Potassium bromate (KBrO₃)-induced 8-oxoGs occur with similar sequence preferences as their derived substitutions, indicating that the reactivity of specific oxidants dictates mutation sequence specificity. While 8-oxoG occurs uniformly across chromatin, 8-oxoG-induced mutations are elevated in compact genomic regions, within nucleosomes, and at inward facing guanines within strongly positioned nucleosomes. Cryo-electron microscopy structures of OGG1-nucleosome complexes indicate that these effects originate from OGG1’s ability to flip outward positioned 8-oxoG lesions into the catalytic pocket while inward facing lesions are occluded by the histone octamer. Mutation spectra from human cells with DNA repair deficiencies reveals contributions of a DNA repair network limiting 8-oxoG mutagenesis, where OGG1- and MUTYH-mediated base excision repair is supplemented by the replication-associated factors Pol η and HMCES. Transcriptional asymmetry of KBrO₃-induced mutations in OGG1- and Pol η-deficient cells also demonstrates transcription-coupled repair can prevent 8-oxoG-induced mutation. Thus, oxidant chemistry, chromatin structures, and DNA repair processes combine to dictate the oxidative mutational landscape in human genomes.

Structural basis for human OGG1 processing 8-oxodGuo within nucleosome core particles

Article Open access 31 October 2024

Watching a double strand break repair polymerase insert a pro-mutagenic oxidized nucleotide

Article Open access 06 April 2021

Human 8-oxoguanine glycosylase OGG1 binds nucleosome at the dsDNA ends and the super-helical locations

Article Open access 28 September 2024

Introduction

Reactive oxygen species (ROS) react with nucleotide bases in DNA to form a variety of mutagenic lesions, including 8-oxoG adducts¹. ROS are generated in cells by endogenous processes like lipid peroxidation and cell metabolism^2,3 or through exposure to exogenous agents such as potassium bromate (KBrO₃)^1,4, a former food additive, or UVA exposure⁵. During carcinogenesis, oncogene activation can also drive ROS formation through changes in metabolic oxidation⁶. Due to the prevalence of exogenous and endogenous agents that induce oxidative stress, oxidative lesions are the second most common DNA lesion following abasic sites (AP-sites)⁷. Moreover, mutations caused by oxidative damage are a common feature of human cancer genomes with the Catalogue of Somatic Mutations in Cancer (COSMIC) single base substitution signature 18 (SBS18) hypothesized to arise from unrepaired 8-oxoG lesions. This hypothesis is supported by experimental evidence indicating that deficiency in 8-oxoG repair mechanisms⁸ or UVA exposure⁵, which generates cellular ROS, result in SBS18-like mutations. SBS18 occurs in ~50% of sequenced human tumors and contributes an average of 300 mutations per genome⁹. SBS18 consists of G to T (and complementary C to A) substitutions and are distributed non-uniformly across the human genome^9,10. What factors dictate the sequence and topological specificity of 8-oxoG-induced mutation in the human genome are unknown.

Mutations caused by oxidative damage are primarily prevented by the activity of base excision repair (BER), which eliminates 8-oxoG in duplex DNA. BER can be initiated by OGG1, a glycosylase that recognizes 8-oxoG across from cytidine (C) and excises the adducted base leaving an AP-site that is subsequently processed by downstream enzymes in the BER pathway^1,11. If OGG1 fails to remove an 8-oxoG adduct, the adducted guanidine can mutagenicaly Hoogsteen base pair with an adenine inserted by multiple polymerases (i.e. Pol δ, η, κ, and ζ) during DNA synthesis^12,13,14,15. These 8-oxoG:dA mispairs are identified by a second DNA glycosylase, MUTYH, which cleaves the adenine leaving an AP-site that is further processed by the BER pathway¹⁶. Due to their direct role in removing 8-oxoG or Hoogsteen paired adenines, loss of OGG1 and MUTYH results in increased mutation rates and an altered mutation spectrum consisting of higher numbers of G to T (and complementary C to A) substitutions^8,17,18. This elevated mutation rate is believed to cause MUTYH-associated polyposis syndrome¹⁸, where individuals inheriting germline MUTYH mutations develop higher incidences of gastro-intestinal cancers throughout life^19,20,21,22. Accordingly, human cells or cancer genomes with bi-allelic OGG1 or MUTYH mutations display respective SBS18^8,18,23 and SBS36²⁴ mutation signatures, both of which are dominated by C to A substitutions. In other organisms, like yeast, BER can be supported by additional mechanisms to limit 8-oxoG mutagenesis. For example, some eukaryotic trans-lesion synthesis polymerases, like Pol η, preferentially insert C across from 8-oxoG in template DNA^12,25,26, allowing error-free bypass of the lesion²⁷. In addition, yeast also utilize mismatch repair to remove adenines mispaired with 8-oxoG lesions^27,28,29. Whether similar alternative 8-oxoG repair mechanisms limit mutagenesis in human cells remains to be determined.

Here, we unravel the sequence and topological determinants of 8-oxoG formation, repair, and mutagenesis, and decipher how different DNA repair and tolerance pathways coordinate to produce the oxidation-induced mutational landscapes observed in human cancer genomes. We found that KBrO₃-treatment produces 8-oxoG lesions and mutations with similar, trinucleotide preferences indicating that ROS chemistry is the primary cause of mutational sequence specificity. Additionally, KBrO₃ produced a unique INDEL signature that was also observed in human cancers, providing evidence that ROS induces other mutation types beyond the canonical G > T substitutions associated with 8-oxoG. Beyond sequence determinants, we identified that chromatin structure is a key topological determinant of oxidation-induced mutations in human cancer genomes. KBrO₃-treatment resulted in mutations that were enriched in heterochromatin, nucleosome bound DNA, and bases facing the histone octamer, a similar phenomenon observed in the SBS18 mutational signature³⁰. Cryo-EM structures of OGG1 bound to nucleosomes containing 8-oxoG revealed the enzyme uses a DNA sculpting and base flipping mechanism for repairing 8-oxoG in the nucleosome, providing a mechanistic basis for the elevated mutational density at bases facing the histone octamer. Finally, analysis of mutation spectra from OGG1-, MUTYH-, Pol η-, and HMCES-deficient cells determined the human 8-oxoG repair network includes OGG1 and MUTYH performing primary mutation avoidance while Pol η and HMCES function in secondary roles mediating tolerance of unrepaired 8-oxoG or 8-oxoG-derived AP sites. Subsequent analysis of Pol η-deficient cells has unveiled the presence of transcription-coupled repair of 8-oxoG on the transcribed strand of genes.

Results

To characterize processes that modulate 8-oxoG mutagenesis and thereby dictate its distribution in human cells, we propagated independent clonal isolates of wild-type immortalized human retinal epithelial cells (hTERT-RPE-1) in the absence and presence of 250 µM KBrO₃ for 100 days, to mimic an exogenous 8-oxoG-producing exposure. RPE-1 cells were chosen for their diploid genome status, which facilitates mutation calling, and non-cancerous origin making these cells a closer model to normal cells in the body. KBrO₃ treatment resulted in only a modest increase in cellular ROS ( ~ 2-fold; Supplementary Fig. 1), This result is consistent with previous experiments showing that KBrO₃ induces 8-oxoG by a chemical reaction that requires glutathione and is resistant to catalase and superoxide dismutase ROS scavengers, therefore occurring distinctly from traditional ROS⁴. Surviving KBrO₃-treated clonal isolates were obtained following this outgrowth and genomic DNA was isolated for Illumina whole genome sequencing. Whole genome sequencing of outgrowth clones was compared to that of corresponding pre-outgrowth populations to identify mutations accumulated during propagation (Fig. 1A). A total of 19684 and 128366 mutations were identified from untreated and KBrO₃-treated cells, respectively, using the consensus calls of three probabilistic variant callers: VarScan2³¹, SomaticSniper³², and Strelka2³³. Cells treated with KBrO₃ had a 23.5-fold increase in substitutions and 3.35-fold increase in small insertion/deletion (INDEL) mutations per sequenced genome when compared to the non-treated cells (Fig. 1B).

**Fig. 1: Mutagenesis in hTERT RPE-1 p53^-/- cells untreated and treated with 250 µM KBrO₃ after 100 cell divisions.**

Spontaneous and KBrO₃-induced mutation spectra in human retinal pigment epithelial cells

Based upon this large increase in mutation load, we assumed most mutations within the KBrO₃-treated cells were induced by 8-oxoG. We therefore produced a de novo KBrO₃ specific mutation signature using SigProfilerExtractor³⁴ set to detect 2 signatures from our dataset: one corresponding to the KBrO₃-induced mutations (SBS96A) and the second representing the spontaneously acquired mutations during untreated outgrowth (SBS96B) (Supplementary Fig. 2A, B). Deconvolution of these signatures into known COSMIC SBS signatures revealed that untreated RPE-1 cells contained a broad spectrum of base substitution mutations most consistent with SBS40, SBS5, and a small percentage of SBS18 (Supplementary Fig. 2C), which are all consistent with spontaneously accumulated mutations during cell culture⁵. Contrastingly, KBrO₃-treated cells were dominated by C > A substitutions, as expected for mutations derived from unrepaired 8-oxoG (Fig. 1C and Supplementary Fig. 2D).

SigProfilerExtractor also identified 2 signatures for INDEL mutations. Untreated cells contained primarily 1 bp T insertions and deletions in long homopolymer runs and thus appears to be a combination of COSMIC INDEL (ID) signatures ID1 and ID2 (Fig. 1D). These mutation signatures are associated with replication slippage events that would be expected to arise spontaneously through cell division⁹. KBrO₃-treated cells also contained the ID1- and ID2-like mutations, though we observed a significant number of 1 bp deletions of C and T nucleotides (Fig. 1D). These deletions were most common when not in homopolymer runs suggesting that they are likely induced by DNA damage and independent of polymerase slippage. While a 1 bp deletion of C bases could logically stem from error-prone replication past a KBrO₃-induced 8-oxoG, the presence of a similar number of 1 bp T base deletions was surprising. KBrO₃ largely produces 8-oxoG lesions through a reaction with glutathione that generates an unknown oxidant^4,35, and no T based lesion has been specifically identified. The sequences flanking 1 bp T deletions displayed a random distribution of C:G and A:T base pairs, indicating that the T deletions were unlikely to be collateral mutations caused by extended synthesis via deletion-prone TLS polymerases recruited to bypass an 8-oxoG³⁶ (Supplementary Fig. 3A).

We next evaluated whether a similar INDEL signature is potentially associated with endogenous ROS during cancer development. We obtained mutation calls for whole genome sequenced primary tumors from the International Cancer Genome Consortium (ICGC) and identified 68 tumors containing greater than 25% of their substitutions stemming from SBS18, meaning that ROS is a major mutagen in these samples. Subsequently, we utilized SigProfilerExtractor on these samples to produce de novo INDEL signatures. This analysis determined that the fourth-most abundant INDEL signature (constituting ~12% of INDEL mutations) had high similarity to our KBrO₃-induced INDEL signature (Fig. 1E and Supplementary Fig. 4; cosine similarity = 0.751). We also utilized mutationalpatterns.R³⁷ to re-assign mutations in ICGC to the entire catalog of COSMIC signatures with the addition of our KBrO₃-induced INDEL signature. Following this process, the number of mutations in our KBrO₃-induced INDEL signature correlated with the number of SBS18 mutations (Supplementary Fig. 3B), indicating that the two signatures are likely linked, and that endogenous ROS produces insertion/deletion mutations in addition to well characterized substitutions in tumors. Recent sequencing of normal epithelial crypts in human colon samples also observed the presence of the INDEL signature ID5 in association with SBS18 mutations³⁸. The COSMIC ID5 signature closely resembles our KBrO₃-induced INDEL signature and correlates strongly with SBS18 in tumors (Supplementary Fig. 3C, D). These similarities strongly suggest that COSMIC ID5 is an oxidation-induced mutation signature.

Endogenous and exogenous oxidants produce different mutation signatures

We next compared the KBrO₃-induced SBS signature to SBS18, which is proposed to originate from ROS producing 8-oxoG in human cancers (Fig. 2A). While both the KBrO₃ and SBS18 signatures were dominated by C > A substitutions, the dominant sequence contexts at which mutations occur were different, producing a cosine similarity of only 0.812. This difference was most pronounced at the sequences CCA, CCT, GCA, and GCT. KBrO₃-treatment produced a greater proportion of mutations at CCA and CCT sequences and a corresponding lower proportion of mutations in GCA and GCT sequences compared to SBS18. We wondered whether this difference in sequence specificity resulted from differences in the ___location of 8-oxoG formation caused by endogenous ROS compared to KBrO₃. We therefore obtained CLAPS-seq reads, which identify the genomic positions (at single nucleotide resolution) of 8-oxoG lesions formed during culture, from HeLa cells grown in the presence and absence of KBrO₃³⁹. Like mutations, KBrO₃-induced 8-oxoG occurred at a different distribution of sequence contexts compared to 8-oxoG caused by endogenous ROS in untreated HeLa cells (cosine similarity = 0.909) (Fig. 2B). KBrO₃-induced lesions occurred in contexts highly similar to KBrO₃-induced mutations, except for a higher proportion of lesions occurring in the context of CCC, suggesting that these lesions may be either more accurately bypassed or preferentially repaired prior to mutagenesis. 8-oxoG generated by endogenous ROS is also over-represented in the CCC context compared to mutations in SBS18. However, these lesions also occurred in CCT and GCC contexts more frequently than SBS18 mutations. Still, the difference of the KBrO₃ and endogenous 8-oxoG proportional sequence contexts displayed a striking similarity to that of the KBrO₃-induced mutations and SBS18, suggesting that differences in lesion formation largely account for the differences between the KBrO₃ and SBS18 mutation signatures.

**Fig. 2: C > A mutation spectra and 8-oxodG lesion spectra in human cells under endogenous or KBrO₃-induced DNA damage.**

BER reduces 8-oxoG mutations in solvent exposed, less chromatin compacted DNA

In human cancers, mutation densities caused by a variety of DNA damages are dictated in part by chromatin structure with heterochromatic regions having higher mutation rates arising from reduced DNA repair efficiency^40,41. We sought to determine whether BER of 8-oxoG lesions was similarly impacted by chromatin resulting in a non-random distribution of mutations. To accomplish this, we profiled where KBrO₃ treatment forms 8-oxoG lesions and mutations (using HeLa cell CLAPS-seq data and WT RPE-1 cell variant calls, respectively) relative to different chromatin states derived by Hidden Markov Modeling (HMM) of eight histone modifications and CTCF^42,43,44 (Fig. 3A). This modeling results in 15 chromatin states with different extents of euchromatic character. 2 of these states are associated with highly repetitive sequences and therefore were excluded from our analysis. By stratifying heterochromatin, promoters, enhancers, and transcribed regions into different states, we found that KBrO₃-induced mutations decreased in less compact regions. Interestingly, 8-oxoG lesions were evenly distributed across all states indicating that inhibited DNA repair in heterochromatin likely underlies the higher mutation rates in these regions. As the repressive nature of heterochromatin is largely generated by tightly packed nucleosomes within these regions, we also profiled 8-oxoG mutations and lesions around strongly positioned nucleosomes within the human genome (Fig. 3B). Consistent with our prior analysis, we observed 8-oxoGs formed relatively evenly across nucleosome bound regions. However, KBrO₃-induced mutations oscillated with a ~ 192 bp periodicity peaking within histone bound DNA, while lower mutagenicity was observed in linker DNA between nucleosomes. This finding is consistent with DNA repair being inhibited by tightly bound histones that obscure access to 8-oxoG during repair. Repair inhibition also extended within individual nucleosomes as KBrO₃-induced mutations displayed a strong 10.3 bp oscillation (Fig. 3B). The peaks of this oscillation occurred at inward facing nucleotides closest to the histones, whereas the troughs occurred at the most solvent exposed nucleotides. This result indicates that either 8-oxoG preferentially forms at histone proximal nucleotides or that repair of 8-oxoG by BER is more efficient at solvent exposed lesions. 8-oxoG lesions displayed a similar pattern as KBrO₃-induced mutations, but with a significantly lesser amplitude. We therefore conclude that efficient BER at outward facing bases is the primary factor influencing 8-oxoG mutations within nucleosomes, though a subtle lesion formation preference may exist. To determine if other DNA proteins beyond histones can block the repair of 8-oxoG, we profiled KBrO₃ mutations and lesions at active transcription factor binding sites (Fig. 3C). We found that neither mutations nor lesions were elevated at these sites in contrast to other types of DNA damage like CPDs⁴⁵, suggesting that the inhibition of 8-oxoG repair is specific for nucleosome structure.

**Fig. 3: Chromatin state, nucleosome binding, and transcription factor binding’s impact on 8-oxodG mutagenesis and lesion formation.**

The mapping of 8-oxoG lesions and KBrO₃-induced mutagenesis indicate that 8-oxoG undergoes preferential repair at solvent-exposed positions in the nucleosome (Fig. 3B). Consistent with these findings, previous work identified that the DNA glycosylase OGG1 excises 8-oxoG from solvent-exposed positions more efficiently than histone-occluded positions in recombinant nucleosomes in vitro^46,47. To obtain mechanistic insight into the preferential repair of solvent-exposed 8-oxoG in the nucleosome, we determined a 3.3 Å cryo-EM structure of OGG1 bound to a nucleosome containing a solvent-exposed 8-oxoG at superhelical ___location (SHL) − 6, referred to as OGG1-8-oxoG-nucleosome core particle (NCP) − 6 (Fig. 4A, Supplementary Figs. 5–7, and Supplementary Table 1). We utilized a catalytically dead variant of OGG1 (K249Q) that maintains the ability to specifically recognize 8-oxoG ensuring we captured an 8-oxoG substrate recognition complex^48,49. The local resolution of the nucleosome was 3–4 Å and the local resolution of OGG1 was 5–7 Å (Supplementary Fig. 6F), which was sufficient to unequivocally dock the previously determined high-resolution X-ray crystal structure of OGG1 (PDB: 1EBM)⁴⁸ into the cryo-EM map (Supplementary Fig. 6H). Although the local resolution of OGG1 (5–7 Å) was not sufficient for determining the exact position of OGG1 side chains, the side chain conformations presented below represent those from the high-resolution X-ray crystal structure of OGG1 (PDB: 1EBM)⁴⁸.

**Fig. 4: Single particle analysis of OGG1-8-oxoG-NCP − 6.**

In the OGG1-8-oxoG-NCP − 6 substrate recognition complex, OGG1 is engaged with ~5 base pairs of nucleosomal DNA at SHL − 6, which buries ~1086 Å² of surface area (Fig. 4A). The interaction of OGG1 with the nucleosomal DNA is mediated by a network of non-specific interactions with the phosphate backbone of the damaged nucleosomal DNA strand, as well as extensive contacts with the orphan cytosine and 8-oxoG (Fig. 4, B C). Interestingly, we did not observe a direct interaction between OGG1 and the histone octamer, indicating that nucleosome binding by OGG1 is primarily driven by the interactions with nucleosomal DNA. At the center of the OGG1 binding footprint lies the nucleosomal 8-oxoG, which has been evicted from the DNA helix and positioned into the OGG1 active site (Fig. 4B C). In this conformation, the extrahelical nucleosomal 8-oxoG is positioned in proximity to key OGG1 amino acid residues that are important for 8-oxoG binding specificity (G42 carbonyl), stabilization of the extrahelical 8-oxoG (C253, F319, and Q315), and 8-oxoG excision (K249Q and D268) (Fig. 4C)⁴⁸. Cumulatively, this data shows OGG1 is in a conformation poised for 8-oxoG excision.

To position the 8-oxoG into the catalytic active site, OGG1 binding induces significant structural changes in the nucleosomal DNA during 8-oxoG recognition. These structural changes include a 1 bp register shift in the nucleosomal DNA, significant minor groove widening at SHL − 6, and nucleosomal DNA bending around SHL − 5.5 to SHL − 6.5 when compared to 8-oxoG-NCP − 6 (Fig. 4D). The OGG1-induced minor groove widening and nucleosomal DNA bending facilitate extrusion of the 8-oxoG from the nucleosomal DNA into the OGG1 active site (Fig. 4C). Ultimately, the mode of 8-oxoG recognition and the OGG1-induced structural changes in the nucleosomal DNA are similar to those seen for OGG1 bound to 8-oxoG in non-nucleosomal DNA (RMSDDNA - 1.621) (Supplementary Fig. 8A)⁴⁸, indicating a conserved 8-oxoG recognition mechanism in chromatin and non-chromatinized DNA.

To determine whether OGG1 uses the same mechanism for 8-oxoG recognition at different translational locations in the nucleosome, we determined a 3.6 Å cryo-EM structure of OGG1 K249Q bound to a nucleosome containing a solvent-exposed 8-oxoG at SHL + 4, referred to as OGG1-8-oxoG-NCP + 4 (Fig. 4E, Supplementary Figs. 9–11, and Supplementary Table 1). The local resolution of the nucleosome was 3–6 Å and the local resolution of OGG1 was 5–8 Å (Supplementary Fig. 10F). Importantly, the cryo-EM map was sufficient to dock the previously determined high-resolution X-ray crystal structure of OGG1 (PDB: 1EBM)⁴⁸ into the cryo-EM map (Supplementary Fig. 10H), and the side chain conformations were kept from the high-resolution X-ray crystal structure of OGG1 (PDB: 1EBM)⁴⁸.

The general mechanism of nucleosome binding and 8-oxoG recognition by OGG1 at SHL + 4 are similar to those observed for OGG1 bound to 8-oxoG at SHL-6 (Supplementary Fig. 8B, C). However, OGG1 binding induces modest structural rearrangements in the nucleosomal DNA during 8-oxoG recognition at SHL + 4, which includes minor groove widening of the nucleosomal DNA without significant DNA bending. The lack of OGG1-induced DNA bending is likely due to the inherently bent conformation of the nucleosomal DNA near SHL + 4 (Supplementary Fig. 8D). Despite these subtle differences, the final conformation of OGG1 and the nucleosomal DNA in the OGG1-8-oxoG-NCP + 4 and OGG1-8-oxoG-NCP − 6 structures are very similar (Fig. 4F and Supplementary Fig. 8A). This data strongly suggests that OGG1 uses the same general mechanism for 8-oxoG recognition and repair at solvent-exposed positions throughout the nucleosome. Notably, the structural changes observed during the recognition of solvent-exposed 8-oxoG by OGG1 are incompatible with binding 8-oxoG proximal to the histone octamer, as this would result in significant clashes between OGG1 and the core histone octamer (Supplementary Fig. 8E, F). Together, this data provides a strong structural rationale for the preferential repair of solvent-exposed 8-oxoG in the nucleosome in vitro and in vivo, and the elevated levels of KBrO₃-induced mutagenesis at nucleotides proximal to the histone octamer (Fig. 3B)^46,47.

Replication and transcription-associated mechanisms limit 8-oxoG mutagenesis

In various species, 8-oxoG mutagenicity is limited by multiple, redundant DNA repair and damage tolerance pathways. This includes the activities of OGG1-initiated BER, MutY-initiated BER, mismatch repair, nucleotide excision repair (NER), and accurate TLS bypass by DNA polymerase η. We therefore compared spontaneous and KBrO₃-induced mutation spectra among WT human cell lines and those lacking OGG1⁸, MUTYH⁸, Pol η⁵⁰, or HMCES, a recently identified replication-associated factor that participates in bypass of ssDNA lesions^{51,52,53,54,55} and protects cells from cytotoxicity associated with KBrO₃ exposure^51,54. Loss of OGG1, MUTYH, and HMCES resulted in moderate ~2 to 3–fold increases in the amount of spontaneously acquired mutations per genome compared to corresponding WT lines, while Pol η-deficiency failed to increase spontaneous mutagenesis (Supplementary Fig. 12A). Changes in mutagenesis in these repair deficient cells were primarily due to increased substitutions as, spontaneous INDEL frequency was only increased in HMCES^-/- cells, which displayed less than 2-fold increase in 1 bp T insertions and deletions (Supplementary Fig. 13). RPE-1 cells lacking HMCES maintained similar substitution spectra compared to WT cells, suggesting the increased spontaneous mutation load results from a general reduction in error-free lesion bypass (Supplementary Fig. 12B). Contrastingly, loss of either OGG1 or MUTYH produced spectra consisting primarily of SBS18-like mutations, indicating these glycosylases are the primary mechanism for preventing 8-oxoG mutagenesis and HMCES is likely involved more generally in lesion bypass (Supplementary Fig. 12B).

To directly evaluate the role of Pol η and HMCES in 8-oxoG bypass, we compared mutation spectra from RPE-1 knockout lines following prolonged KBrO₃ exposure to those in WT RPE-1 cells (evaluated in Fig. 1). Deficiency in Pol η or HMCES resulted in 1.5– and 1.3–fold increases in total KBrO₃-induced mutations, respectively, although statistical significance of increased mutation load in the Pol η knockout cells was not possible due to the lack of replicate samples sequenced (Fig. 5A). These mild increases are likely underestimates of the true augmentation of KBrO₃ mutagenesis as HMCES^-/- cells showed a significant growth delay upon initial KBrO₃ exposure likely resulting in fewer cell divisions for these lines. Despite the lack of statistical significance for total mutations load, loss of either HMCES or Pol η significantly altered the mutation spectra for both substitutions and INDEL mutation types (Fig. 5A). The substitution spectra of Pol η- and HMCES-deficient cells were nearly identical to that of KBrO₃-treated WT cells (Fig. 5B), indicating that both enzymes likely participate in some form of error-free bypass of 8-oxoG or derived repair intermediates. Interestingly, loss of Pol η increased not only G > T substitutions predicted to be caused by 8-oxoG bypass, but also G > C, and G > A substitutions (p < 1×10^-5, by Chi-square) as well (Supplementary Fig. 14), suggesting that another polymerase may insert G or T across from 8-oxoG in Pol η’s absence. Neither gene disruption changed the impact of chromatin compaction on 8-oxoG induced substitution frequency (Supplementary Figs. 15 and 16), suggesting they are primarily operating in contexts without nucleosome involvement. INDEL spectra from KBrO₃-treated Pol η- and HMCES-deficient cells, were also like that of WT cells, except for several subtle differences. HMCES^-/- cells displayed a small increase in 1 bp T insertions at shorter homopolymer lengths, while Pol η-deficiency resulted in a general loss of 1 bp T insertions and a preference for 1 bp C or T deletions occurring in 2–3 bp homopolymer repeats (Fig. 5C). Ultimately, these differences in mutation spectra, particularly for INDELs, further supports roles for both Pol η and HMCES in oxidative lesion bypass.

**Fig. 5: KBrO₃-induced mutagenesis in human WT, *POLH*^-/-, or *HMCES*^-/- RPE-1 cells lines.**

To determine whether mismatch repair (MMR) limits 8-oxoG mutagenesis in human cells, we evaluated replication strand asymmetry of KBrO₃-induced mutations and 8-oxoG-induced G to T substitutions in untreated OGG1^-/- and MUTYH^-/- cells using methods similar to AsymTools2 software⁵⁶ that determines leading and lagging strand association of mutations based upon the directionality of the replication fork movement in the mutated region. Applying this analysis to CLAPS-seq reads in KBrO₃-treated WT cells revealed an equal distribution of 8-oxoG lesions on the leading and lagging template strands (Fig. 6A). KBrO₃-induced mutations however displayed slightly more G > T substitutions on the leading strand. While statistically significant (p = 0.0302 by two-sided paired t-test comparing the number of G > T substitutions on the leading and lagging strands in individual samples normalized to the strand specific sequence composition), the leading strand mutational bias induced by KBrO₃ treatment was only 9.35% that of other replication-associated mutagenic processes, like APOBEC signature mutations (Supplementary Fig. 17A), suggesting that any preferential removal of 8-oxoG from the lagging strand template by DNA repair (potentially MMR) is limited. HMCES deletion exacerbated the 8-oxoG leading strand bias (p = 1.08×10^-4 by two-sided paired t-test), indicating that HMCES may favor bypass of leading strand lesions. Interestingly, loss of Pol η removed the replication strand asymmetry. This indicates that Pol η likely mediates error-free bypass of 8-oxoG in the lagging strand template in human cells. Similar results indicate Pol η TLS functions preferentially during lagging strand synthesis for bypass of UV photoproducts in human melanomas and fibroblasts⁵⁷, suggesting a general lagging strand preference for this TLS polymerase. Similar mild leading strand bias was observed for G > T substitutions in untreated MUTHY^-/- and OGG1^-/- neuroblastoma cells (Supplementary Fig. 17B) that are defective in 8-oxoG repair, indicating a general better bypass of 8-oxoG on the lagging strand independent of the chemical species creating the lesion or cell type context.

**Fig. 6: KBrO₃-induced 8-oxodG mutation and lesion strand bias on leading/lagging and transcribed/non-transcribed strands in human cells.**

We also assessed whether KBrO₃-induced mutations displayed transcriptional asymmetry, which would be indicative of transcription coupled repair of 8-oxoG. Transcription-coupled NER and BER have been suggested to be involved in 8-oxoG removal^58,59. However, little transcriptional mutation asymmetry has been reported for ROS-associated SBS18 mutations¹⁰, suggesting that transcription coupled repair of 8-oxoG may be limited. We found that G > T substitutions in WT and HMCES^-/- cells only slightly favored the transcribed strand of genes (Fig. 6B). This bias was also observed in CLAPS-seq reads for 8-oxoG lesions, suggesting this effect is caused by a preference in lesion formation instead of a repair process. Transcriptional asymmetry in KBrO₃-treated Pol η-deficient cells, however, was very pronounced favoring the non-transcribed strand (p-value < 1×10^-5 by Chi-square), which impacted all G > T, G > C, and G > A substitution types. The effect size of this asymmetry is on par with other mutational processes limited by transcription-coupled nucleotide excision repair⁵⁶ (Supplementary Fig. 18A) and provides strong evidence that transcription-coupled repair can remove 8-oxoG lesions but its impact in mutational data is obscured by the error-free bypass of the lesion by Pol η. Interestingly, increased mutation of the non-transcribed strand was also observed for G > T substitutions in OGG1^-/- neuroblastoma cells (Supplementary Fig. 18B), indicating that TC-NER can also remove 8-oxoG to help compensate for the absence of BER.

Discussion

Mutations from endogenously and exogenously derived 8-oxoG

We observed that long-term treatment of RPE-1 cells with the oxidant KBrO₃ increases substitutions and produces a mutational signature like COSMIC SBS18 that is proposed to be caused by endogenous ROS. Both signatures are composed of C > A mutations, however, the preferred trinucleotide sequences that these substitutions occurred in differ. Specifically, KBrO₃ exposure produced an over-representation of substitutions at CCA and CCT and under-representation at GCA, and GCT compared to SBS18. Other studies have demonstrated the same KBrO₃ signature for both long term exposure in RPE-1 cells⁵⁰ or short term exposure of human iPSCs⁶⁰, indicating that differences in exposure protocol and/or cell lines are not responsible for the differences in SBS18 and KBrO₃-induced substitution specificity. Additionally, CLAPs-seq mapping of KBrO₃-induced 8-oxoG lesions in HeLa cells produced similar sequence preferences as KBrO₃-induced mutations, providing evidence that the oxidizing agent can dictate the sequences most likely to form 8-oxoG. Interestingly, CCA and CCT mutation contexts correspond to trinucleotides with low vertical ionization potential (VIP), which sensitizes these motifs (i.e. TGG and AGG sequences) to long-range guanine oxidation by charge transfer⁶¹. Reciprocally, TGC and AGC have higher VIPs, indicating that GCA and GCT sequences would have fewer mutations produced by this mechanism. This correlation suggests that KBrO₃ may induce more guanine oxidation through charge transfer than endogenous ROS, leading to a KBrO₃ specific 8-oxoG mutation pattern. An alternative possibility is that specific oxidants produce 8-oxoG at different sequences. While the mechanism by which KBrO₃ generates 8-oxoG is unknown, its requires the presence of a reducing agent, like glutathione, and is insensitive to traditional cellular ROS scavengers, indicating a different oxidation chemistry than for endogenous ROS⁴. By extension, the DNA damage induced by endogenous ROS could result from the combined activity of multiple different species (e.g. peroxide, superoxide, etc.), which may all have different sequence preferences in forming 8-oxoG. In the future, utilizing human cell systems to determine the mutation signatures of individual endogenous reactive oxygen species will be beneficial in determining which sources of ROS are most relevant for inducing mutation in human cancers.

KBrO₃ exposure primarily produces DNA damage in the form of 8-oxoG³⁵, which canonically produces G > T substitutions through Hoogsteen base pairing of 8-oxoG with dA⁶² during DNA synthesis. We were therefore surprised to observe that KBrO₃ treatment also increased INDEL mutations and that a large percentage of these mutations occurred at A:T pairs. While a 1 bp deletion of C bases could logically stem from error-prone replication past an KBrO₃-induced 8-oxoG, the presence of a similar number of 1 bp T base deletions suggested that these mutations were caused either through collateral mutagenesis³⁶ adjacent to an 8-oxoG or by a second KBrO₃-induced DNA lesion. T deletions lacked an enrichment of C:G base pairs flanking the mutation, indicating that they were unlikely incurred as collateral mutations during 8-oxoG bypass. The lack of a reasonable connection of T deletions to 8-oxoG suggests that KBrO₃ also causes at least one other mutagenic DNA lesion and targets T or A bases, such as thymine glycol^4,63. Our analysis of INDEL signatures in human tumors displaying high levels of SBS18 mutations indicated that a similar INDEL process occurs in these tumors. Moreover, the number of SBS18 mutations correlates with the number of mutations attributed to our KBrO₃-induced INDEL signature and the similar COSMIC ID5 signature, both suggesting that endogenous ROS produces these mutations. Additional research is needed to identify the specific DNA lesions causing these signatures.

Oxidation-induced mutagenesis within the context of chromatin

Chromatin structure is another major influence on the density of 8-oxoG-induced mutations. Higher mutation densities were observed in more compact regions of the genome, which likely stems from the density of nucleosomes within these regions. Within nucleosomes, mutations had a 10.3 bp periodicity when treated with KBrO₃, occurring primarily on nucleotides proximal to the histone octamer. This suggests that DNA repair mechanisms may be excluded from the inward facing positions of the nucleosomal DNA. This is in stark contrast to UV-induced cyclo-pyrimidine dimer positioning at outward facing nucleotides at nucleosomes due to preferential lesion formation at these sites⁶⁴. Prior work was unable to obtain a cryo-EM structure of OGG1 engaged with an 8-oxoG in the nucleosome⁶⁵, which was hypothesized to result from an inability of OGG1 to flip the 8oxoG embedded in the nucleosome into the enzyme active site. However, our structural data and another recently published OGG1-8-oxoG-NCP structure⁶⁶ clearly demonstrate that OGG1 accesses outward facing 8-oxoG in the nucleosome by sculpting nucleosomal DNA and flipping the 8-oxoG base into its active site for catalysis, similar to the DNA sculpting mechanisms previously observed for the DNA glycosylase AAG⁶⁷ and APE1⁶⁸. This leaves inward facing 8-oxoG in the nucleosome more prone to mutation, as OGG1 lacks the ability to recognize 8-oxoG in these positions without massive changes in nucleosome structure^46,47. Repair of these sites is likely significantly delayed and may require active nucleosome remodeling in response to DNA damage by additional cellular factors, such as the BER-associated nucleosome remodeler ALC1⁶⁹. Interestingly, other protein-DNA interactions appear to have little to no impact on 8-oxoG-induced mutagenesis. Transcription factors bound to gene promoter regions produce no change in the density of KBrO₃-induced mutations in contrast to other types of DNA damage, like cyclobutane pyrimidine dimers. 8-oxoG lesions at these sites are greatly reduced compared to neighboring DNA, suggesting that they are rapidly repaired. However, we are unable to exclude the possibility that transcription factor binding protects their binding sites from formation of 8-oxoG. In contrast to the variable sequence preferences for 8-oxoG formation induced by KBrO₃ or ROS, chromatin impacts on 8-oxoG induced mutation appear to be largely conserved for each method of lesion formation as the observed distributions of mutation with respect to chromatin state, nucleosome occupancy, replication and transcriptional strand bias, and transcription factor binding sites for KBrO₃-induced mutation largely mirror those reported for SBS18 mutations in human tumor genomes^10,30.

An expanded 8-oxoG repair network in human cells

Our data indicates at least three mechanisms of 8-oxoG repair influence its mutagenicity (Fig. 7). OGG1 and MUTYH activities provide the first line of defense against 8-oxoG mutagenesis as their deletion results in spontaneous mutator phenotypes displaying SBS18-like mutation spectra. 8-oxoG escaping BER can be bypassed by the replication-associated damage tolerance mechanisms of Pol η and HMCES. The 1.5-fold increase in mutation load in KBrO₃-treated Pol η-deficient cells is significantly lower than the impact of Pol η loss in yeast²⁵. This suggests either lesser reliance on Pol η due to redundancy with other human DNA polymerases in 8-oxoG bypass, or less accurate Pol η bypass in humans compared to yeast. Supporting the latter, biochemical experiments have shown human Pol η correctly inserts dC across from 8-oxoG to a lesser extent than yeast Pol η²⁵. HMCES-deficiency also resulted in elevated KBrO₃-induced mutagenesis, consistent with the previously reported sensitivity of HMCES^-/- cells to KBrO₃⁵¹. KBrO₃-treated HMCES-deficient cells displayed a mutation spectrum identical to KBrO₃-treated wildtype cells suggesting that HMCES aids in error-free bypass of 8-oxoG or derived repair intermediates, despite its previously demonstrated role preventing mutagenesis^51,55,70,71 by cross-linking to AP-sites⁵¹. Such a mutation spectrum could be reconciled with HMCES’s known biochemical activity toward AP-sites if OGG1 removal of 8-oxoG can be uncoupled from the rest of BER, resulting in AP-sites occurring specifically at dG nucleotides. Subsequent TLS-based bypass of these AP-sites could either be error-free (in the case of REV1-mediated bypass via C-insertion) or produce C > A substitutions (by A-rule AP-sites site bypass), which would recapitulate the KBrO₃ substitution signature. The error-prone bypass of AP-sites may also cause the oxidation-induced 1 bp deletions we identified and that are elevated in HMCES^-/- cells⁷². By analogy with Ung1 generated AP-sites sites from dU in the lagging strand template⁷³, uncoupling of OGG1 glycosylase activity from BER would be expected to occur more frequently at sites of DNA replication (where HMCES functions) as the glycosylase could recognize 8-oxoG in ssDNA, but subsequent steps of BER would be inhibited do to the lack of a complementary DNA strand. KBrO₃-treated HMCES^-/- cells also displayed enhanced leading strand replication strand-bias compared to similarly treated WT cells suggesting HMCES may function more during leading strand synthesis. The underlying reason for this preference is currently unclear, especially considering the synthetic lethal phenotype between HMCES-deficiency and APOBEC3A expression^53,74, which damages the lagging strand template⁷⁵. Transcriptional strand asymmetries in KBrO₃-treated Pol η−deficient cells and untreated OGG1^-/- cells ultimately revealed a substantial decrease in the mutational burden on the transcribed strand, suggesting that TC-NER can remove oxidative damage when other repair processes are overwhelmed or defective. Previous studies indicate that CSB, a major component of TC-NER, can be recruited to transcription sites upon oxidative damage and is important for fitness following oxidative damage^59,76.

**Fig. 7: Mechanisms that limit KBrO₃-induced 8-oxodG mutagenesis in human cells.**

In conclusion, this study outlines the multi-dimensional mutational landscape of exogenously and endogenously induced oxidative damage and the consequences of topology on this landscape while providing mechanistic insight into primary, secondary, and tertiary strategies to limit 8-oxoG mutagenesis in human cells. Variants and loss of multiple of the factors in this study lead to cancer and neurodegenerative disease including OGG1 and Pol η. The robust bioinformatics pipeline and exhaustive topological analysis can also be used as a blueprint and foundation to develop a database of holistic multi-dimensional mutational signatures to explore mechanism and drug targets. Future research should explore topological mutagenesis studies of various types of damage and genetic conditions in human cells. This research should also aim to understand the interplay between these pathways and identify potential therapeutic targets for future interventions.

Methods

Cell Culture

hTERT-RPE-1 p53^-/- cells were cultured in DMEM (Thermo Fisher Cat No. 11965092 supplemented with 7.5% FBS, 1X Glutamax (Thermo Fisher Cat No. 35050061), 1X Non-Essential Amino Acids (Thermo Fisher Cat No. 11140035, and 1X Penicillin-Streptomycin (Thermo Fisher Cat No. 15140122) at 37° C in 5% CO2. Cells were single-cell cloned with cloning rings. Single-cell clone parental cells were transfected with pSpCas9(BB)-2A-Puro 2 (Addgene Cat No. 48139) that contain guide RNAs that target the intron-exon junction of the second exon of HMCES (5’-TTGCGCCTACCAGGATCGGC and 5’-ACTTTAGACGGTGGTCACGG). Cells were selected with 15 μg/mL puromycin for two days prior to plating for individual clones because hTERT-RPE-1 p53^-/- are already mildly puromycin resistant. Clones were screened for deletion by PCR (using primers HMCES F2: 5’-GCATTTGCAGAGCTCCTCTC and HMCES R2: 5’-GACAGAAGCACTGGGCTG) and by immunoblotting for loss of HMCES expression with antibodies raised against the middle and C-terminus of the protein. HMCES^-/- cells were confirmed to be hypersensitive to KBrO₃, consistent with previous results⁵¹.

Measurement of cellular ROS

For the reactive oxygen species assay, a single-cell clone (Clone 3) of WT hTERT-RPE-1 p53^-/- was seeded into an optically clear 96-well plate. 24 hours later cells were treated with 250 µM KBrO₃ or vehicle for an hour, followed by treatment with 5 µM of CellRoX 488 reagent and NucBlue for 30 minutes. Cells were fluorescently imaged live with a Nikon Ti2E microscope and analyzed for integrated nuclear fluorescent intensity using Nikon NIS Elements. CellRoX488 fluorescence intensities were plotted using GraphPad Prism.

Long-term Mutagenesis Assay

For long-term mutagenesis assays, pooled parental and a single-cell clone (Clone 3) of WT hTERT-RPE-1 p53^-/- (generously provided by Daniel Durocher, University of Toronto) as well as three individual single-cell HMCES^-/- clones of HMCES knockouts (Clone3.1, Clone 3.3, and Clone 3.4) were seeded into 10 cm dishes and carried continuously in the presence or absence of 250 µM KBrO₃ for 100 generations (3 months). Each passage, cells were seeded at similar cell numbers (20% confluency) and carried until 80% confluent at which point they were passaged again. After 100 generations (24 passages), each cell line was single-cell cloned and two of each clone (WT pool, WT Clone 3, HMCES^-/- Clones 3.1, 3.3, 3.4) were harvested for genomic DNA (Promega Cat No. A1120). Genomic DNA was submitted for 150 bp paired-end Illumina dep-sequencing sequencing targeting 30X depth at Vanderbilt University’s VANTAGE Next Generation-Sequencing core.

System Information

All computational analyses were performed on Linux, specifically Ubuntu 22.04.03 LTS. Data analysis was conducted using Python v3.10.12, Python v2.7.18, Perl v5.34.0 and R v4.1.2 (unless a virtual environment was required). Further system, software, library versions, and hardware information is available on request.

Sequencing Alignment

Results were aligned to the Genome Reference Consortium Human Build 37 (GRCh37/hg19) using the Burrows-Wheeler Aligner (BWA) mem algorithm on default parameters (BWA v0.7.17). The resulting Sequence Alignment/Map (SAM) files, which contain aligned sequence reads, were compressed into Binary Alignment/Map (BAM) format using samtools view (samtools v1.13 using htslib v1.13+ds). [Note: All samtools steps were run using default parameters to maintain a standard approach] After compression, the BAM files were sorted based on genomic coordinates using samtools sort to prepare for removal of duplicate reads which can arise from PCR amplification artifacts during sequencing. These were removed using samtools rmdup so it would not have an impact on downstream variant calling and analysis. These final BAM files were converted to MPILEUP files using samtools mpileup. The final BAM and MPILEUP files were used to call mutations from multiple mutation callers.

Mutation Calling

The BAM files were processed with Strelka2 (v2.9.10), Manta (v1.6.0), and Somatic Sniper (v1.0.5.0) while the MPILEUP files were processed using VarScan2 (v2.3). SNVs and INDELs were called using VarScan somatic comparing treated cells to untreated counterparts with the following parameters changed -min-coverage 10 -min-var-freq 0.2 -somatic-p-value 0.05 -min-freq-for-hom 0.9 -min-avg-qual 30 to reduce artifacts of mutation calling. The resulting SNV and INDEL files were split into germline, somatic, and loss of heterozygosity (LOH) files using VarScan processSomatic on default parameters, to split the results and isolate the high confidence somatic SNV and INDEL mutation calls which were used to identify consensus mutations.

The BAM files were initially compared to their corresponding normal counterparts utilizing Manta’s structural variant pipeline⁷⁷, employing default parameters to detect small INDEL candidates for input into Strelka2. Strelka2 was run on default parameters comparing tumor to normal using hg19 and Manta’s INDEL candidates for the tumor/normal pair. The resulting SNV and INDEL mutation calls were used to identify consensus mutations. The BAM files were also used to create a third set of SNV calls using Somatic Sniper on default parameters except -Q 40 -G -L which requires a minimum somatic score of 40 as recommended by the developers for BWA aligned reads, and not report loss of heterozygosity (LOH) and gain of reference (GOR) mutations in the final output to reduce the likelihood of false positives. The resulting SNV mutation calls were also used to identify consensus mutations.

To account for artifacts of mutation calling and sequencing from different callers, we took the consensus from all three callers (Strelka2, Somatic Sniper, and VarScan2) for SNVs and the consensus from both Strelka2 and VarScan2 for INDELs. This was done using a custom Python script requiring mutations to be present in all sets of mutation calls for the sample. Then all the separate consensus mutations were pooled, and mutations present in more than one sample were omitted due to a high likelihood of being a germline mutation or artifact of sequencing and mutation calling. The concatenated mutation calls were then split into separate sets based on treatment, genotype, or both depending on the analysis.

Processing of 8-oxoG Lesion Data

CLAPS-seq FASTQ files were aligned to hg19 using the bwa-mem algorithm on default parameters. The resulting SAM files were processed using a custom Python script to convert the SAM file into a BED file. The script filtered out reads that did not align with a CIGAR score of 150 M. It also filtered reads keeping ones that aligned to chromosome 1-22, X, or Y. It took the reads passing this filter and checked the bitwise flag for 0 (complemented) or 16 (reverse complemented) and processed the proper alignment position (either 5’ or 3’ of the top strand) to determine the base pair position where the lesion occurred. We then filtered the custom BED file for positions where there was a G at that context which removed reads which were assumed to be false positives reported by the authors. The resulting BED files were converted to a VCF format using a custom Python script to process this data through other programs, like vcf2maf and nucleosome profiling.

Mutation Signature Generation

We generated mutation signatures from cell mutations using SigProfilerExtractor³⁴ (v1.1.23) on default parameters with a minimum and maximum of 1 and 5 signatures respectively. The most stable number of signatures KBrO₃-associated was 2 for both SNV and INDEL mutation signatures, which was used for all subsequent comparisons. All other signatures generated used the most stable number of signatures.

Correlation of INDEL signature with SBS18

The PCAWG data was analyzed using the MutationalPatterns package in R³⁷ to conduct non-negative matrix factorization (NMF). This analysis incorporated the COSMIC SBS and INDEL signatures, along with an additional custom INDEL signature derived from the NMF results of KBrO₃ treated cells. Tumors were deemed positive for SBS18 and the custom INDEL signature if they exhibited a minimum of 20 mutations associated with each signature. These samples were plotted with the log₂ transformed number of mutations. The Pearson correlation coefficient was computed based on the mutation count per sample for each signature. The SigProfiler signatures in samples data provided with PCAWG were used for the correlation of ID5 with SBS18.

Strand Asymmetry

Replication strand asymmetry was calculated similarly to AsymTools2⁵⁶ on default parameters. Custom Python scripts were generated to calculate replication strand asymmetry for individual samples using map of left or right replicating regions in the hg19 reference genome provided within AsymTools2. Mutational strand preference among replicate samples was assessed by two-sided paired t-test. To produce a replication asymmetry of APOBEC-induced mutation in tumors for comparison to 8-oxoG mutational asymmetry, C to G substitutions in TCW contexts of BRCA-proficient breast cancers were filtered from ICGC mutational data as described in⁷⁸. Transcribed strand asymmetry was calculated using a custom Python script using a similar approach to AsymTools2. The script takes an RPE-1 transcribed gene list from GEO accession number GSE146121 and cross-references the list with the UCSC hg19 gene list. This provided us with a gene list that was actively transcribed in RPE-1 cells, which we then compared with mutations and lesions. Mutations mapping to the top strand with a G base were considered to be on the (+) strand and mutations mapping with a C base were considered to be on the (−) strand. By analyzing the gene’s orientation, we were able to ascertain the strand on which the event took place and subsequently compare the occurrences of each event on both the transcribed and non-transcribed strands. To normalize the events, the event counts were divided by the guanine base count on that transcribed or non-transcribed strand, respectively, resulting in the unit of events/Mb. The results were similar to what was represented in the AsymTools2 results, however, were specific to the cell line and had a higher resolution since transcribed regions were not binned but were measured at single-nucleotide resolution.

Chromatin State & Nucleosome Profiling

Chromatin states were assessed by mapping mutations and lesions to chromatin states from the epithelial cell HMM chromatin maps (https://genome.ucsc.edu/cgi-bin/hgFileUi?db=hg19&g=wgEncodeBroadHmm) using bedtools intersect^79,80. Subsequently, we standardized the results to events/Mb based on the HMM map’s region size. The order of heterochromatin to euchromatin was determined by the map construction.

Mutations and lesions were intersected with strongly positioned nucleosome dyads following the protocol outlined in⁸¹ in a 1000 base-pair (bp) window. Expected counts were calculated using genomic trinucleotide mutation or lesion frequencies multiplied by the occurrence of those contexts at each position in the dyad map. The observed counts were divided by the expected counts and log₂ transformed to generate the graphs.

$${\log }_{2}\left(\frac{{actual\; number\; of\; AAA\; evebts\; at\; map\; postition}+\ldots+{actual\; number\; of\; TTT\; events\; at\; map\; position}}{\left({genomewide\; AAA\; event\; frequency} \, * \, {AAA\; context\; at\; map\; position}\right)+\ldots+\left({genomewide\; TTT\; event\; frequency} \, * \, {TTT\; context\; at\; map\; position}\right)}\right)$$

The data was smoothed using a Savitzky–Golay filter with a 200 bp window with a polynomial order of 3.

Transcription Factor Profiling

Mutations and lesions were intersected with known active transcription factor binding sites using a map generated from previous work⁴⁵ in a 1000 bp window. Events were counted and graphed using a custom Python script and smoothed using a Savitzky–Golay filter with a 200 bp window with a polynomial order of 3.

Purification of H. sapiens OGG1 K249Q

A pGEX6P1 vector (N-terminal GST tag) with the H. sapiens OGG1 gene bearing the K249Q mutation was obtained from GenScript. For protein expression, the pGEX6P1-OGG1-K249Q vector was transformed into BL21-CondonPlus (DE3) RIPL cells (Agilent). The transformed cells were grown in 2x YT media at 37 °C until an OD₆₀₀ of 0.8 and protein expression induced with 0.5 mM IPTG overnight at 18 °C. The cells were harvested by centrifugation and resuspended in a buffer containing 50 mM HEPES (pH-7.5), 150 mM NaCl, 1 mM DTT, and a protease inhibitor cocktail (Benzamidine, Leupeptin, AEBSF, Pepstatin A). The resuspended cells were lysed by sonication and the lysate clarified by centrifugation. The clarified lysate was loaded onto a GSTrap HP column (Cytiva) equilibrated with 50 mM HEPES (pH-7.5), 150 mM NaCl, and 1 mM DTT, and the protein was eluted in a buffer containing 50 mM HEPES (pH-7.5), 150 mM NaCl, 1 mM DTT, and 50 mM reduced glutathione. Fractions containing GST-OGG1 were loaded onto a Resource S cation exchange column (Cytiva) equilibrated with 50 mM HEPES (pH-6.8), 50 mM NaCl, 1 mM DTT, and 1 mM EDTA, and eluted in a high salt buffer containing 50 mM HEPES (pH-6.8), 1 M NaCl, 1 mM DTT, and 1 mM EDTA. OGG1 was then liberated from the GST-tag by incubation with PreScission Protease for 4 hours in a buffer containing 50 mM HEPES (pH-7.5), 150 mM NaCl, and 1 mM DTT. The cleaved OGG1 protein was rerun over a Resource S cation exchange column (Cytiva), and the eluted protein loaded on a Sephacryl S-200 HR (Cytiva) equilibrated with 50 mM HEPES (pH-7.5), 150 mM NaCl, and 1 mM TCEP. The purified OGG1 fractions were combined, concentrated to 10 mg ml^-1, and stored at -80 °C.

Preparation of oligonucleotides

DNA oligonucleotides (oligos) containing 8-oxoG were obtained from TriLink BioTechnologies, and non-damaged oligos were obtained from Integrated DNA Technologies. Each oligo was resuspended at 1 mM in a buffer containing 10 mM Tris (pH-8.0) and 1 mM EDTA. Complimentary oligos (see Supplementary Table 2) were mixed at a 1:1 ratio and annealed by heating to 90 °C followed by a stepwise cooling to 4 °C using a linear gradient at -1 °C min^-1. The annealed oligos were stored long-term at -20 °C.

Purification of recombinant human histones

The genes encoding H. sapien histones H2A, H2B, H3.2 (C110A), and H4 were cloned into a pet3a expression vector. For histone H2A, H3.2, and H4 expression, vectors were transformed into T7 Express lysY competent cells (New England Biolabs). For histone H2B expression, the vector was transformed into BL21-CodonPlus (Agilent) competent cells. The cells were grown in minimal media at 37 °C until an OD₆₀₀ of 0.4 was reached, and protein expression induced with 0.4 mM IPTG (H2A, H2B, and H3.2) or 0.3 mM IPTG (H4) for 3-4 hours at 37 °C. The cells were harvested by centrifugation and resuspended in a buffer containing 50 mM Tris (pH-7.5) 100 mM NaCl, 1 mM benzamidine, 1 mM DTT, and 1 mM EDTA. The histones were purified under denaturing conditions using an established method^82,83. In brief, the resuspended cells were lysed by sonification, inclusion bodies isolated by centrifugation, and the histones extracted from the inclusion bodies under denaturing conditions (6 M Guanidinium chloride). After extraction, the histones were purified using subtractive anion-exchange chromatography and cation-exchange chromatography using gravity flow columns. The purified histones were then dialyzed into H₂O, lyophilized, and stored at -20 °C.

Preparation of H2A/H2B Dimers and H3/H4 Tetramers

H2A/H2B dimers and H3/H4 tetramers were prepared using an established method^82,83. In brief, each individual histone was resuspended in a buffer containing 20 mM Tris (pH-7.5), 6 M guanidinium chloride, and 10 mM DTT. For H2A/H2B dimers, H2A and H2B were mixed at a 1:1 ratio and dialyzed three times against a buffer containing 20 mM Tris (pH-7.5), 2 M NaCl, and 1 mM EDTA. For H3/H4 tetramers, H3 and H4 were mixed at a 1:1 ratio and dialyzed three times against a buffer containing 20 mM Tris (pH-7.5), 2 M NaCl, and 1 mM EDTA. The H2A/H2B dimers and H3/H4 tetramers were subsequently purified over a Sephacryl S-200 HR (Cytiva) in a buffer containing 20 mM Tris (pH-7.5), 2 M NaCl, and 1 mM EDTA. The purified H2A/H2B dimers and H3/H4 tetramers were stored in 50% glycerol at -20 °C.

Nucleosome assembly and purification

Recombinant nucleosomes were assembled by an established salt-dialysis method^82,83. In brief, H2A/H2B dimers and H3/H4 tetramers were mixed with DNA in a 2:1:1 molar ratio, respectively, in a buffer containing 20 mM Tris (pH 7.5), 2 M NaCl, and 1 mM EDTA. Stepwise nucleosome assembly was then performed by decreasing the amount of NaCl from 2.0 M NaCl to 1.5 M NaCl, 1.0 NaCl, 0.66 M NaCl, 0.5 M NaCl, 0.25 M NaCl, 0.125 M, and 0 M NaCl over a period of 24 - 26 hours. The reconstituted nucleosomes were heat shocked at 37 °C for 15 minutes to generate uniform DNA positioning and purified by ultracentrifugation over a 10% - 40% sucrose gradient. Final nucleosome purity was determined using native polyacrylamide gel electrophoresis (5%, 59:1 acrylamide:bis-acrylamide), and the purified nucleosomes were stored at 4 °C.

Cryo-EM sample and grid preparation

For cryo-EM sample preparation, 8-oxoG-NCP (5 μM) was mixed with OGG1 K249Q (7.5 μM - 10 μM) in a buffer containing 25 mM HEPES (pH-7.1), 25 mM NaCl, 1 mM TCEP, and 1 mM EDTA. The OGG1-8-oxoG-NCP complexes were then incubated at 4 °C for 10 minutes and fixed with glutaraldehyde (0.1%) for 20 minutes. The samples were loaded onto a Superdex S200 Increase 10/300 GL (Cytiva) equilibrated with a buffer containing 50 mM HEPES (pH-7.1), 100 mM NaCl, 1 mM TCEP, and 1 mM EDTA. Fractions containing OGG1-NCP were identified via native polyacrylamide gel electrophoresis (5%, 59:1 acrylamide:bis-acrylamide). The fractions containing the OGG1-8-oxoG-NCP complex were then combined and concentrated to 1.5 μM for short-term storage. Gels corresponding to the 8-oxoG-NCP − 6 and 8-oxoG-NCP + 4 samples used for cryo-EM grid preparation can be found in Supplementary Fig. 5A and 8 A. The samples (3 μL, 1.5 μM) were then applied to a Quantifoil R2/2 300 mesh copper cryo-EM grid at 8 °C and 95% humidity, and the grids plunge frozen in liquid ethane using a Vitrobot Mark IV (Thermo Fisher).

Cryo-EM Data collection and processing

All cryo-EM data collections were performed on a Titan Krios G3i equipped with Gatan K3 direct electron detector and BioContinuum energy filter at the University of Chicago Advanced Electron Microscopy Core Facility (RRID:SCR_019198). All cryo-EM datasets were processed with cryoSPARC⁸⁴ using the workflows outlined in Supplementary Fig. 5 and 9. In brief, the micrographs were corrected for beam-induced drift using Patch Motion Correction and contrast transfer function (CTF) fit using Patch CTF Estimation. The micrographs were then manually curated to exclude micrographs of poor quality. Following micrograph curation, a subset of micrographs was subjected to blob picking to generate initial templates, which were then used for automated template picking. The particle stacks were then extracted from the micrographs and multiple rounds of 2D classification were performed. Ab-initio models were then generated using the final particle stacks and several rounds of heterogeneous refinement performed to initially separate 8-oxoG-NCP and OGG1-8-oxoG-NCP maps.

To improve the interpretability of the 8-oxoG-NCP maps, additional 3D-classification was performed using a focus mask for the entry/exit site nucleosomal DNA, which is prone to partially unwrapping from the histone octamer. Following 3D classification, the final particle stacks for each 8-oxoG-NCP structure were re-extracted to full box size (600 pixels), and the re-extracted particles subjected to local CTF refinement and non-uniform refinement. The final 8-oxoG-NCP maps were then subjected to a B-factor sharpening using PHENIX autosharpen. The final 8-oxoG-NCP maps were deposited into the electron microscopy data bank under accession numbers EMD-43595 for 8-oxoG-NCP − 6 and EMD-43600 for 8-oxoG-NCP + 4.

To improve interpretability of the OGG1-8-oxoG-NCP maps, 3D-classification was performed using a focus mask for OGG1 and the surrounding nucleosomal DNA. Following 3D-classification, the final particle stacks for each OGG1-8-oxoG-NCP structure were re-extracted to full box size (600 pixels), and the re-extracted particles subjected to local CTF refinement and non-uniform refinement. To further improve interpretability of the maps, local refinement (without particle subtraction) was performed using a focus mask for OGG1 and the surrounding nucleosomal DNA or a focus mask for the NCP. A composite map for the OGG1-8-oxoG-NCP − 6 structure was then generated by combining the maps from a non-uniform refinement and two local refinement (OGG1/DNA and NCP local refine) jobs using PHENIX combine focused maps. A composite map for the OGG1-8-oxoG-NCP + 4 structure was then generated by combining the maps from the non-uniform refinement and local refinement (OGG1/DNA local refine) jobs using PHENIX combine focused maps. The final cryo-EM maps were deposited into the Electron Microscopy Data Bank under accession numbers EMD-43600 for OGG1-8-oxoG-NCP − 6 (composite), EMD-43597 for OGG1-8-oxoG-NCP − 6 (consensus), EMD-43598 for OGG1-8-oxoG-NCP − 6 (NCP local refine), EMD-43599 for OGG1-8-oxoG-NCP − 6 (OGG1/DNA local refine), EMD-43601 for OGG1-8-oxoG-NCP + 4 (composite), EMD-43602 for OGG1-8-oxoG-NCP + 4 (consensus), and EMD-43603 for OGG1-8-oxoG-NCP + 4 (OGG1/DNA local refine).

Model building and refinement

All model building and refinement was performed iteratively using University of California San Francisco (UCSF) Chimera⁸⁵, PHENIX⁸⁶, and COOT⁸⁷. An initial nucleosome model was generated using a previously determined cryo-EM structure of a nucleosome containing an AP-site (PDB: 7U52)⁶⁸. The initial OGG1 model was generated from a previously determined X-ray crystal structure of an OGG1-8-oxoG-DNA complex (PDB:1EBM)⁴⁸. The models for each respective structure were rigid body docked into the cryo-EM map using UCSF Chimera⁸⁵. The models were then refined in PHENIX⁸⁶ using protein and nucleic acid secondary structure restraints, and manual adjustments to the models made in COOT⁸⁷. All final models were validated using MolProbity⁸⁸, and model coordinates for each structure were deposited in the Protein Data Bank (PDB) under accession numbers 8VWS for 8-oxoG-NCP − 6, 8VWT for OGG1-8-oxoG-NCP − 6, 8VWU for 8-oxoG-NCP + 4, 8VWV for OGG1-8-oxoG-NCP + 4.

Statistics & Reproducibility

The number of independent WT hTERT-RPE-1 p53^-/- and hTERT-RPE-1 p53^-/- HMCES^-/- clones selected for sequencing was chosen to allow statistical comparisons in mutations per genome between genotypes and treatments by Mann-Whitney U test. Additionally, the length of passaging was chosen to acquire over 10,000 aggregate mutations in each treatment type based on previously established mutation rates of hTERT-RPE-1 cells in culture⁸⁹. This number of mutations allows for robust statistical analysis comparing the density of mutations in different genome features. The number of analyzed POLH^-/-, MUTYH^-/-, and OGG1^-/- clones was determined by the public availability of the sequencing data. No power calculation was used to predetermine sample size. No data were excluded from the analyses, the experiments were not randomized, and the Investigators were not blinded to allocation during experiments and outcome assessment.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The next generation sequencing data generated in this study for untreated and KBrO₃-treated hTERT-RPE-1 p53^-/- and hTERT-RPE-1 p53^-/- HMCES^-/- cells have been deposited as FASTQ files at the NCBI SRA database under accession code PRJNA1100509. Full mutation lists used for analysis are provided in Supplementary Data 1. hTERT-RPE-1 POLH^-/- VCF files used in this study are available from⁵⁰ available on Mendeley Data server (https://doi.org/10.17632/jkjkpvgxyd.1). FASTQ files for KBrO₃-treated hTERT-RPE-1 POLH^-/- cells used I this study are available in NCBI SRA database under accession code PRJNA940340. MUTYH^-/- and OGG1^-/- VCF files can be obtained from the supplementary dataset S01 from⁸. CLAPS-seq 8-oxoG lesion mapping data from³⁹ are available from the Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/) under accession code GSE181312. Publicly available lists of tumor mutations were obtained from the International Cancer Genome Consortium (ICGC) from consensus_snv_indel/final_consensus_passonly.snv_mnv_indel.icgc.public.maf.gz and simple_somatic_mutation.open.BRCA-EU.tsv.gz. Corresponding tumor mutation lists can be downloaded from ICGC using the linked download instructions. The final cryo-EM maps are available from the Electron Microscopy Data Bank under accession numbers EMD-43600 for OGG1-8-oxoG-NCP − 6 (composite), EMD-43597 for OGG1-8-oxoG-NCP − 6 (consensus), EMD-43598 for OGG1-8-oxoG-NCP − 6 (NCP local refine), EMD-43599 for OGG1-8-oxoG-NCP − 6 (OGG1/DNA local refine), EMD-43601 for OGG1-8-oxoG-NCP + 4 (composite), EMD-43602 for OGG1-8-oxoG-NCP + 4 (consensus), and EMD-43603 for OGG1-8-oxoG-NCP + 4 (OGG1/DNA local refine). The model coordinates for each structure are available from the Protein Data Bank (PDB) under accession numbers 8VWS for 8-oxoG-NCP − 6, 8VWT for OGG1-8-oxoG-NCP − 6, 8VWU for 8-oxoG-NCP + 4, 8VWV for OGG1-8-oxoG-NCP + 4. All data is publicly available and accessible without restriction. Values underlying all graphs in figures are provided in the Source Data file. Source data are provided with this paper.

Code availability

All custom scripts for mutation and lesion analyses⁹⁰ are available at the S-RobertsLab GitHub (https://github.com/S-RobertsLab/Cordero-et-al.-2024).

References

Dizdaroglu, M. Oxidatively induced DNA damage and its repair in cancer. Mutat. Res Rev. Mutat. Res. 763, 212–245 (2015).
Article PubMed CAS Google Scholar
Nakamura, J. et al. The endogenous exposome. DNA Repair (Amst.) 19, 3–13 (2014).
Article PubMed PubMed Central CAS Google Scholar
Swenberg, J. A. et al. Endogenous versus exogenous DNA adducts: their role in carcinogenesis, epidemiology, and risk assessment. Toxicol. Sci. 120, S130–S145 (2011).
Article PubMed CAS Google Scholar
Kawanishi, S. & Murata, M. Mechanism of DNA damage induced by bromate differs from general types of oxidative stress. Toxicology 221, 172–178 (2006).
Article PubMed CAS Google Scholar
Zhivagui, M. et al. DNA damage and somatic mutations in mammalian cells after irradiation with a nail polish dryer. Nat. Commun. 14, 276 (2023).
Article ADS PubMed PubMed Central CAS Google Scholar
Weinberg, F. et al. Mitochondrial metabolism and ROS generation are essential for Kras-mediated tumorigenicity. Proc. Natl Acad. Sci. USA 107, 8788–8793 (2010).
Article ADS PubMed PubMed Central CAS Google Scholar
Lindahl, T. Instability and decay of the primary structure of DNA. Nature 362, 709–715 (1993).
Article ADS PubMed CAS Google Scholar
van den Boogaard, M. L. et al. Defects in 8-oxo-guanine repair pathway cause high frequency of C > A substitutions in neuroblastoma. Proc Natl Acad Sci USA 118 https://doi.org/10.1073/pnas.2007898118. (2021).
Alexandrov, L. B. et al. The repertoire of mutational signatures in human cancer. Nature 578, 94–101 (2020).
Article ADS PubMed PubMed Central CAS Google Scholar
Otlu, B. et al. Topography of mutational signatures in human cancer. Cell Rep. 42, 112930 (2023).
Article PubMed PubMed Central CAS Google Scholar
Klungland, A. et al. Accumulation of premutagenic DNA lesions in mice defective in removal of oxidative base damage. Proc. Natl Acad. Sci. USA 96, 13300–13305 (1999).
Article ADS PubMed PubMed Central CAS Google Scholar
Einolf, H. J. & Guengerich, F. P. Fidelity of nucleotide insertion at 8-oxo-7,8-dihydroguanine by mammalian DNA polymerase delta. Steady-state and pre-steady-state kinetic analysis. J. Biol. Chem. 276, 3764–3771 (2001).
Article PubMed CAS Google Scholar
Zhang, Y. et al. Error-prone lesion bypass by human DNA polymerase eta. Nucleic Acids Res. 28, 4717–4724 (2000).
Article PubMed PubMed Central CAS Google Scholar
Haracska, L., Prakash, L. & Prakash, S. Role of human DNA polymerase kappa as an extender in translesion synthesis. Proc. Natl Acad. Sci. USA 99, 16000–16005 (2002).
Article ADS PubMed PubMed Central CAS Google Scholar
Haracska, L., Prakash, S. & Prakash, L. Yeast DNA polymerase zeta is an efficient extender of primer ends opposite from 7,8-dihydro-8-Oxoguanine and O6-methylguanine. Mol. Cell Biol. 23, 1453–1459 (2003).
Article PubMed PubMed Central CAS Google Scholar
Woods, R. D. et al. Structure and stereochemistry of the base excision repair glycosylase MutY reveal a mechanism similar to retaining glycosidases. Nucleic Acids Res. 44, 801–810 (2016).
Article PubMed CAS Google Scholar
Al-Tassan, N. et al. Inherited variants of MYH associated with somatic G:C->T:A mutations in colorectal tumors. Nat. Genet. 30, 227–232 (2002).
Article PubMed CAS Google Scholar
Robinson, P. S. et al. Inherited MUTYH mutations cause elevated somatic mutation rates and distinctive mutational signatures in normal human cells. Nat. Commun. 13, 3949 (2022).
Article ADS PubMed PubMed Central CAS Google Scholar
Collaborative Group on Duodenal Polyposis in, M. A. P. et al. Duodenal Adenomas and Cancer in MUTYH-associated Polyposis: An International Cohort Study. Gastroenterology 160, 952–954.e954 (2021).
Article Google Scholar
Sampson, J. R. et al. Autosomal recessive colorectal adenomatous polyposis due to inherited mutations of MYH. Lancet 362, 39–41 (2003).
Article PubMed CAS Google Scholar
Theodoratou, E. et al. A large-scale meta-analysis to refine colorectal cancer risk estimates associated with MUTYH variants. Br. J. Cancer 103, 1875–1884 (2010).
Article PubMed PubMed Central CAS Google Scholar
Vogt, S. et al. Expanded extracolonic tumor spectrum in MUTYH-associated polyposis. Gastroenterology 137, 1976–1985.e1971-1910 (2009).
Article PubMed CAS Google Scholar
Zou, X. et al. A systematic CRISPR screen defines mutational mechanisms underpinning signatures caused by replication errors and endogenous DNA damage. Nat. Cancer 2, 643–657 (2021).
Article PubMed PubMed Central CAS Google Scholar
Viel, A. et al. A Specific Mutational Signature Associated with DNA 8-Oxoguanine Persistence in MUTYH-defective Colorectal Cancer. EBioMedicine 20, 39–49 (2017).
Article PubMed PubMed Central Google Scholar
Haracska, L., Yu, S. L., Johnson, R. E., Prakash, L. & Prakash, S. Efficient and accurate replication in the presence of 7,8-dihydro-8-oxoguanine by DNA polymerase eta. Nat. Genet. 25, 458–461 (2000).
Article PubMed CAS Google Scholar
McCulloch, S. D., Kokoska, R. J., Garg, P., Burgers, P. M. & Kunkel, T. A. The efficiency and fidelity of 8-oxo-guanine bypass by DNA polymerases delta and eta. Nucleic Acids Res. 37, 2830–2840 (2009).
Article PubMed PubMed Central CAS Google Scholar
Mudrak, S. V., Welz-Voegele, C. & Jinks-Robertson, S. The polymerase eta translesion synthesis DNA polymerase acts independently of the mismatch repair system to limit mutagenesis caused by 7,8-dihydro-8-oxoguanine in yeast. Mol. Cell Biol. 29, 5316–5326 (2009).
Article PubMed PubMed Central CAS Google Scholar
Boiteux, S. & Jinks-Robertson, S. DNA repair mechanisms and the bypass of DNA damage in Saccharomyces cerevisiae. Genetics 193, 1025–1064 (2013).
Article PubMed PubMed Central CAS Google Scholar
Ni, T. T., Marsischky, G. T. & Kolodner, R. D. MSH2 and MSH6 are required for removal of adenine misincorporated opposite 8-oxo-guanine in S. cerevisiae. Mol. Cell 4, 439–444 (1999).
Article PubMed CAS Google Scholar
Pich, O. et al. Somatic and Germline Mutation Periodicity Follow the Orientation of the DNA Minor Groove around Nucleosomes. Cell 175, 1074–1087.e1018 (2018).
Article PubMed CAS Google Scholar
Koboldt, D. C. et al. VarScan 2: somatic mutation and copy number alteration discovery in cancer by exome sequencing. Genome Res. 22, 568–576 (2012).
Article PubMed PubMed Central CAS Google Scholar
Larson, D. E. et al. SomaticSniper: identification of somatic point mutations in whole genome sequencing data. Bioinformatics 28, 311–317 (2012).
Article PubMed CAS Google Scholar
Kim, S. et al. Strelka2: fast and accurate calling of germline and somatic variants. Nat. Methods 15, 591–594 (2018).
Article PubMed CAS Google Scholar
Islam, S. M. A. et al. Uncovering novel mutational signatures by de novo extraction with SigProfilerExtractor. Cell Genom. 2, None https://doi.org/10.1016/j.xgen.2022.100179 (2022).
Ballmaier, D. & Epe, B. Oxidative DNA damage induced by potassium bromate under cell-free conditions and in mammalian cells. Carcinogenesis 16, 335–342 (1995).
Article PubMed CAS Google Scholar
Poti, A., Szikriszt, B., Gervai, J. Z., Chen, D. & Szuts, D. Characterisation of the spectrum and genetic dependence of collateral mutations induced by translesion DNA synthesis. PLoS Genet. 18, e1010051 (2022).
Article PubMed PubMed Central CAS Google Scholar
Manders, F. et al. MutationalPatterns: the one stop shop for the analysis of mutational processes. BMC Genomics 23, 134 (2022).
Article PubMed PubMed Central Google Scholar
Lee-Six, H. et al. The landscape of somatic mutation in normal colorectal epithelial cells. Nature 574, 532–537 (2019).
Article ADS PubMed CAS Google Scholar
An, J. et al. Genome-wide analysis of 8-oxo-7,8-dihydro-2’-deoxyguanosine at single-nucleotide resolution unveils reduced occurrence of oxidative damage at G-quadruplex sites. Nucleic Acids Res 49, 12252–12267 (2021).
Article PubMed PubMed Central CAS Google Scholar
Polak, P. et al. Cell-of-origin chromatin organization shapes the mutational landscape of cancer. Nature 518, 360–364 (2015).
Article ADS PubMed PubMed Central CAS Google Scholar
Makova, K. D. & Hardison, R. C. The effects of chromatin organization on variation in mutation rates in the genome. Nat. Rev. Genet. 16, 213–223 (2015).
Article PubMed PubMed Central CAS Google Scholar
Ernst, J. et al. Mapping and analysis of chromatin state dynamics in nine human cell types. Nature 473, 43–49 (2011).
Article ADS PubMed PubMed Central CAS Google Scholar
Ernst, J. & Kellis, M. ChromHMM: automating chromatin-state discovery and characterization. Nat. Methods 9, 215–216 (2012).
Article PubMed PubMed Central CAS Google Scholar
Ernst, J. & Kellis, M. Chromatin-state discovery and genome annotation with ChromHMM. Nat. Protoc. 12, 2478–2492 (2017).
Article PubMed PubMed Central CAS Google Scholar
Mao, P. et al. ETS transcription factors induce a unique UV damage signature that drives recurrent mutagenesis in melanoma. Nat. Commun. 9, 2626 (2018).
Article ADS PubMed PubMed Central Google Scholar
Bilotti, K., Kennedy, E. E., Li, C. & Delaney, S. Human OGG1 activity in nucleosomes is facilitated by transient unwrapping of DNA and is influenced by the local histone environment. DNA Repair (Amst.) 59, 1–8 (2017).
Article PubMed CAS Google Scholar
Bilotti, K., Tarantino, M. E. & Delaney, S. Human Oxoguanine Glycosylase 1 Removes Solution Accessible 8-Oxo-7,8-dihydroguanine Lesions from Globally Substituted Nucleosomes Except in the Dyad Region. Biochemistry 57, 1436–1439 (2018).
Article PubMed CAS Google Scholar
Bruner, S. D., Norman, D. P. & Verdine, G. L. Structural basis for recognition and repair of the endogenous mutagen 8-oxoguanine in DNA. Nature 403, 859–866 (2000).
Article ADS PubMed CAS Google Scholar
Nash, H. M., Lu, R., Lane, W. S. & Verdine, G. L. The critical active-site amine of the human 8-oxoguanine DNA glycosylase, hOgg1: direct identification, ablation and chemical reconstitution. Chem. Biol. 4, 693–702 (1997).
Article PubMed CAS Google Scholar
Yurchenko, A. A. et al. Genomic mutation landscape of skin cancers from DNA repair-deficient xeroderma pigmentosum patients. Nat. Commun. 14, 2561 (2023).
Article ADS PubMed PubMed Central CAS Google Scholar
Mohni, K. N. et al. HMCES Maintains Genome Integrity by Shielding Abasic Sites in Single-Strand DNA. Cell 176, 144–153 e113 (2019).
Article PubMed CAS Google Scholar
Thompson, P. S., Amidon, K. M., Mohni, K. N., Cortez, D. & Eichman, B. F. Protection of abasic sites during DNA replication by a stable thiazolidine protein-DNA cross-link. Nat. Struct. Mol. Biol. 26, 613–618 (2019).
Article PubMed PubMed Central CAS Google Scholar
Mehta, K. P. M., Lovejoy, C. A., Zhao, R., Heintzman, D. R. & Cortez, D. HMCES Maintains Replication Fork Progression and Prevents Double-Strand Breaks in Response to APOBEC Deamination and Abasic Site Formation. Cell Rep. 31, 107705 (2020).
Article PubMed PubMed Central CAS Google Scholar
Srivastava, M. et al. HMCES safeguards replication from oxidative stress and ensures error-free repair. EMBO Rep. 21, e49123 (2020).
Article PubMed PubMed Central CAS Google Scholar
Semlow, D. R., MacKrell, V. A. & Walter, J. C. The HMCES DNA-protein cross-link functions as an intermediate in DNA interstrand cross-link repair. Nat. Struct. Mol. Biol. 29, 451–462 (2022).
Article PubMed PubMed Central CAS Google Scholar
Haradhvala, N. J. et al. Mutational Strand Asymmetries in Cancer Genomes Reveal Mechanisms of DNA Damage and Repair. Cell 164, 538–549 (2016).
Article PubMed PubMed Central CAS Google Scholar
Seplyarskiy, V. B. et al. Error-prone bypass of DNA lesions during lagging-strand replication is a common source of germline and cancer mutations. Nat. Genet. 51, 36–41 (2019).
Article PubMed CAS Google Scholar
Guo, J., Hanawalt, P. C. & Spivak, G. Comet-FISH with strand-specific probes reveals transcription-coupled repair of 8-oxoGuanine in human cells. Nucleic Acids Res. 41, 7700–7712 (2013).
Article PubMed PubMed Central CAS Google Scholar
Kumar, N. et al. Global and transcription-coupled repair of 8-oxoG is initiated by nucleotide excision repair proteins. Nat. Commun. 13, 974 (2022).
Article ADS PubMed PubMed Central CAS Google Scholar
Kucab, J. E. et al. A Compendium of Mutational Signatures of Environmental Agents. Cell 177, 821–836.e816 (2019).
Article PubMed PubMed Central CAS Google Scholar
Bacolla, A. et al. Guanine holes are prominent targets for mutation in cancer and inherited disease. PLoS Genet. 9, e1003816 (2013).
Article PubMed PubMed Central CAS Google Scholar
Kouchakdjian, M. et al. NMR structural studies of the ionizing radiation adduct 7-hydro-8-oxodeoxyguanosine (8-oxo-7H-dG) opposite deoxyadenosine in a DNA duplex. 8-Oxo-7H-dG(syn).dA(anti) alignment at lesion site. Biochemistry 30, 1403–1412 (1991).
Article PubMed CAS Google Scholar
Burrows, C. J. & Muller, J. G. Oxidative Nucleobase Modifications Leading to Strand Scission. Chem. Rev. 98, 1109–1152 (1998).
Article PubMed CAS Google Scholar
Brown, A. J., Mao, P., Smerdon, M. J., Wyrick, J. J. & Roberts, S. A. Nucleosome positions establish an extended mutation signature in melanoma. PLoS Genet 14, e1007823 (2018).
Article PubMed PubMed Central Google Scholar
You, Q., Feng, X., Cai, Y., Baylin, S. B. & Li, H. Human 8-oxoguanine glycosylase OGG1 binds nucleosome at the dsDNA ends and the super-helical locations. Commun. Biol. 7, 1202 (2024).
Article PubMed PubMed Central CAS Google Scholar
Ren, M. et al. Structural basis for human OGG1 processing 8-oxodGuo within nucleosome core particles. Nat. Commun. 15, 9407 (2024).
Article PubMed PubMed Central CAS Google Scholar
Zheng, L., Tsai, B. & Gao, N. Structural and mechanistic insights into the DNA glycosylase AAG-mediated base excision in nucleosome. Cell Discov. 9, 62 (2023).
Article PubMed PubMed Central CAS Google Scholar
Weaver, T. M. et al. Structural basis for APE1 processing DNA damage in the nucleosome. Nat. Commun. 13, 5390 (2022).
Article ADS PubMed PubMed Central CAS Google Scholar
Hewitt, G. et al. Defective ALC1 nucleosome remodeling confers PARPi sensitization and synthetic lethality with HRD. Mol. Cell 81, 767–783 e711 (2021).
Article PubMed PubMed Central CAS Google Scholar
Wu, L. et al. HMCES protects immunoglobulin genes specifically from deletions during somatic hypermutation. Genes Dev. 36, 433–450 (2022).
Article PubMed PubMed Central CAS Google Scholar
Shukla, V. et al. HMCES Functions in the Alternative End-Joining Pathway of the DNA DSB Repair during Class Switch Recombination in B Cells. Mol. Cell 77, 384–394.e384 (2020).
Article PubMed CAS Google Scholar
Kokoska, R. J., McCulloch, S. D. & Kunkel, T. A. The efficiency and specificity of apurinic/apyrimidinic site bypass by human DNA polymerase eta and Sulfolobus solfataricus Dpo4. J. Biol. Chem. 278, 50537–50545 (2003).
Article PubMed CAS Google Scholar
Hoopes, J. I. et al. Avoidance of APOBEC3B-induced mutation by error-free lesion bypass. Nucleic Acids Res. 45, 5243–5254 (2017).
Article PubMed PubMed Central CAS Google Scholar
Biayna, J. et al. Loss of the abasic site sensor HMCES is synthetic lethal with the activity of the APOBEC3A cytosine deaminase in cancer cells. PLoS Biol. 19, e3001176 (2021).
Article PubMed PubMed Central CAS Google Scholar
Hoopes, J. I. et al. APOBEC3A and APOBEC3B Preferentially Deaminate the Lagging Strand Template during DNA Replication. Cell Rep. 14, 1273–1282 (2016).
Article PubMed PubMed Central CAS Google Scholar
de Waard, H. et al. Different effects of CSA and CSB deficiency on sensitivity to oxidative DNA damage. Mol. Cell Biol. 24, 7941–7948 (2004).
Article PubMed PubMed Central Google Scholar
Chen, X. et al. Manta: rapid detection of structural variants and indels for germline and cancer sequencing applications. Bioinformatics 32, 1220–1222 (2016).
Article PubMed CAS Google Scholar
Mertz, T. M. et al. Genetic inhibitors of APOBEC3B-induced mutagenesis. Genome Res. 33, 1568–1581 (2023).
Article PubMed PubMed Central Google Scholar
Quinlan, A. R. BEDTools: The Swiss-Army Tool for Genome Feature Analysis. Curr Protoc Bioinformatics 47, 11 12 11-34 https://doi.org/10.1002/0471250953.bi1112s47 (2014).
Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010).
Article PubMed PubMed Central CAS Google Scholar
Morledge-Hampton, B., Kalyanaraman, A. & Wyrick, J. J. Analysis of cytosine deamination events in excision repair sequencing reads reveals mechanisms of incision site selection in NER. Nucleic Acids Res. 52, 1720–1735 (2024).
Article PubMed CAS Google Scholar
Dyer, P. N. et al. Reconstitution of nucleosome core particles from recombinant histones and DNA. Methods Enzymol. 375, 23–44 (2004).
Article PubMed CAS Google Scholar
Ryan, B. J., Weaver, T. M., Spencer, J. J. & Freudenthal, B. D. Generation of Recombinant Nucleosomes Containing Site-Specific DNA Damage. Methods Mol. Biol. 2701, 55–76 (2023).
Article PubMed PubMed Central CAS Google Scholar
Punjani, A., Rubinstein, J. L., Fleet, D. J. & Brubaker, M. A. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods 14, 290–296 (2017).
Article PubMed CAS Google Scholar
Pettersen, E. F. et al. UCSF Chimera-a visualization system for exploratory research and analysis. J. Comput Chem. 25, 1605–1612 (2004).
Article PubMed CAS Google Scholar
Adams, P. D. et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D. Biol. Crystallogr 66, 213–221 (2010).
Article ADS PubMed PubMed Central CAS Google Scholar
Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr D. Biol. Crystallogr 60, 2126–2132 (2004).
Article ADS PubMed Google Scholar
Davis, I. W. et al. MolProbity: all-atom contacts and structure validation for proteins and nucleic acids. Nucleic Acids Res. 35, W375–W383 (2007).
Article ADS PubMed PubMed Central Google Scholar
Brody, Y. et al. Quantification of somatic mutation flow across individual cell division events by lineage sequencing. Genome Res. 28, 1901–1918 (2018).
Article PubMed PubMed Central CAS Google Scholar
S-RobertsLab/Cordero-et-al.-2024: Publication Release v2 v. v1.0.1 (Zenodo, 2024).
Hayward, N. K. et al. Whole-genome landscapes of major melanoma subtypes. Nature 545, 175–180 (2017).
Article ADS PubMed CAS Google Scholar
Adasme, M. F. et al. PLIP 2021: expanding the scope of the protein-ligand interaction profiler to DNA and RNA. Nucleic Acids Res. 49, W530–W534 (2021).
Article PubMed PubMed Central CAS Google Scholar

Download references

Acknowledgements

Whole genome sequencing in this manuscript was conducted by the Vanderbilt Technologies for Advanced Genomics (VANTAGE) core. We thank Eric J. Crow for assistance establishing portions of our variant calling and mutation signature pipeline. This work was supported by the following National Institutes of Health (NIH) grants: F32GM140718 (from NIGMS to TMW), R35GM128562 (from NIGMS to BDF), F32GM136096 and K99/R00ES034058 (from NIGMS and NIEHS to KPMM), R01ES030575 (From NIEHS to DC) and R01ES032814 and R01CA269784 (from NIEHS and NCI to SAR). KPMM was additionally supported by the Office of the Vice Chancellor for Research and Graduate Education, University of Wisconsin-Madison and a Vanderbilt University Destination Biochemistry Advanced Postdoctoral Scholars Award. DC obtained funding from the Vanderbilt-Ingram Cancer Center. SAR received startup funds from the University of Vermont and the University of Vermont Cancer Center.

Author information

These authors contributed equally: Cameron Cordero, Kavi P. M. Mehta, Tyler M. Weaver.

Authors and Affiliations

Department of Microbiology and Molecular Genetics, University of Vermont, Burlington, VT, 05405, USA
Cameron Cordero & Steven A. Roberts
University of Vermont Cancer Center, University of Vermont, Burlington, VT, 05405, USA
Cameron Cordero & Steven A. Roberts
School of Molecular Biosciences, Washington State University, Pullman, WA, 99164, USA
Cameron Cordero & Steven A. Roberts
Department of Biochemistry, Vanderbilt University School of Medicine, Nashville, TN, 37232, USA
Kavi P. M. Mehta & David Cortez
Department of Comparative Biosciences, School of Veterinary Medicine, University of Wisconsin, Madison, WI, 53706, USA
Kavi P. M. Mehta
Department of Biochemistry and Molecular Biology, University of Kansas Medical Center, Kansas City, KS, 66160, USA
Tyler M. Weaver, Justin A. Ling & Bret D. Freudenthal
Department of Cancer Biology, University of Kansas Medical Center, Kansas City, KS, 66160, USA
Tyler M. Weaver, Justin A. Ling & Bret D. Freudenthal
University of Kansas Cancer Center, Kansas City, KS, 66160, USA
Tyler M. Weaver & Bret D. Freudenthal
Center for Reproductive Biology, Washington State University, Pullman, WA, 99164, USA
Steven A. Roberts

Authors

Cameron Cordero
View author publications
Search author on:PubMed Google Scholar
Kavi P. M. Mehta
View author publications
Search author on:PubMed Google Scholar
Tyler M. Weaver
View author publications
Search author on:PubMed Google Scholar
Justin A. Ling
View author publications
Search author on:PubMed Google Scholar
Bret D. Freudenthal
View author publications
Search author on:PubMed Google Scholar
David Cortez
View author publications
Search author on:PubMed Google Scholar
Steven A. Roberts
View author publications
Search author on:PubMed Google Scholar

Contributions

CC, KPMM, TMW, BDF, DC, and SAR designed the project. KPMM and TMW developed experimental systems. CC developed computational pipelines. CC, KPMM, TMW, JAL, BDF, and SAR performed experiments and analyzed data. CC, KPMM, TMW, BDF, DC, and SAR interpreted results. CC and TMW created figures. CC, KPMM, TMW, and SAR wrote the manuscript. CC, KPMM, TMW, BDF, DC, and SAR edited the manuscript.

Corresponding authors

Correspondence to Kavi P. M. Mehta, Bret D. Freudenthal, David Cortez or Steven A. Roberts.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Maria Zhivagui, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Description of Additional Supplementary Information

Supplementary Data 1

Reporting Summary

Transparent Peer Review file

Source data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Cordero, C., Mehta, K.P.M., Weaver, T.M. et al. Contributing factors to the oxidation-induced mutational landscape in human cells. Nat Commun 15, 10722 (2024). https://doi.org/10.1038/s41467-024-55497-z

Download citation

Received: 09 May 2024
Accepted: 10 December 2024
Published: 23 December 2024
DOI: https://doi.org/10.1038/s41467-024-55497-z

This article is cited by

Suppressed DNA repair capacity in flight attendants after air travel
- Sneh M. Toprani
- Irina Mordukhovich
- Zachary D. Nagel
Scientific Reports (2025)
Structural basis of gap-filling DNA synthesis in the nucleosome by DNA Polymerase β
- Tyler M. Weaver
- Benjamin J. Ryan
- Bret D. Freudenthal
Nature Communications (2025)