PLoS ONEplosplosonePLOS ONE1932-6203Public Library of ScienceSan Francisco, CA USA10.1371/journal.pone.0258269PONE-D-21-18200Research ArticleBiology and life sciencesGeneticsDNADNA repairBiology and life sciencesBiochemistryNucleic acidsDNADNA repairResearch and analysis methodsAnimal studiesExperimental organism systemsModel organismsCaenorhabditis elegansResearch and analysis methodsModel organismsCaenorhabditis elegansResearch and analysis methodsAnimal studiesExperimental organism systemsAnimal modelsCaenorhabditis elegansBiology and life sciencesOrganismsEukaryotaAnimalsInvertebratesNematodaCaenorhabditisCaenorhabditis elegansBiology and life sciencesZoologyAnimalsInvertebratesNematodaCaenorhabditisCaenorhabditis elegansBiology and life sciencesGeneticsDNADNA damageBiology and life sciencesBiochemistryNucleic acidsDNADNA damageBiology and life sciencesGeneticsMutationMedicine and health sciencesOncologyCancer treatmentRadiation therapyMedicine and health sciencesClinical medicineClinical oncologyRadiation therapyMedicine and health sciencesOncologyClinical oncologyRadiation therapyBiology and life sciencesGeneticsMutationSubstitution mutationBiology and life sciencesGeneticsMutagenesisMedicine and health sciencesOncologyCancers and neoplasmsMalignant tumorsC. elegans genome-wide analysis reveals DNA repair pathways that act cooperatively to preserve genome integrity upon ionizing radiationCooperative action of DNA repair pathways to preserve genome integrity upon ionizing radiationMeierBettinaFormal analysisSoftwareVisualizationWriting – original draftWriting – review & editing1VolkovaNadezda V.Data curationFormal analysisInvestigationSoftwareVisualizationWriting – original draftWriting – review & editing2WangBinResources13https://orcid.org/0000-0002-6977-1060González-HuiciVíctorInvestigationResources1¤BertoliniSimoneResources1CampbellPeter J.ConceptualizationFunding acquisitionSupervision456GerstungMoritzConceptualizationData curationFormal analysisFunding acquisitionInvestigationMethodologyResourcesSupervisionVisualizationWriting – original draftWriting – review & editing27*https://orcid.org/0000-0003-4639-9902GartnerAntonFormal analysisFunding acquisitionInvestigationMethodologyProject administrationSupervisionWriting – original draftWriting – review & editing189*Centre for Gene Regulation and Expression, University of Dundee, Dundee, United KingdomEuropean Molecular Biology Laboratory, European Bioinformatics Institute, Hinxton, United KingdomNational Engineering Research Center for Non-Food Biorefinery, Guangxi Academy of Sciences, Nanning, ChinaCancer, Ageing and Somatic Mutation, Wellcome Sanger Institute, Hinxton, United KingdomDepartment of Haematology, University of Cambridge, Cambridge, United KingdomDepartment of Haematology, Addenbrooke’s Hospital, Cambridge, United KingdomEuropean Molecular Biology Laboratory, Genome Biology Unit, Heidelberg, GermanyCenter for Genomic Integrity, Institute for Basic Science, Ulsan, Republic of KoreaDepartment of Biological Sciences, School of Life Sciences, Ulsan National Institute of Science and Technology, Ulsan, Republic of KoreaCotterillSueEditorSaint George’s University, UNITED KINGDOM
The authors have declared that no competing interests exist.
Current address: Institute for Research in Biomedicine, Parc Científic de Barcelona, Barcelona, Spain
* E-mail: tgartner@ibs.re.kr (AG); moritz.gerstung@ebi.ac.uk (MG)610202120211610e025826926202122920212021Meier et alThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Ionizing radiation (IR) is widely used in cancer therapy and accidental or environmental exposure is a major concern. However, little is known about the genome-wide effects IR exerts on germ cells and the relative contribution of DNA repair pathways for mending IR-induced lesions. Here, using C. elegans as a model system and using primary sequencing data from our recent high-level overview of the mutagenic consequences of 11 genotoxic agents, we investigate in detail the genome-wide mutagenic consequences of exposing wild-type and 43 DNA repair and damage response defective C. elegans strains to a Caesium (Cs-137) source, emitting γ-rays. Cs-137 radiation induced single nucleotide variants (SNVs) at a rate of ~1 base substitution per 3 Gy, affecting all nucleotides equally. In nucleotide excision repair mutants, this frequency increased 2-fold concurrently with increased dinucleotide substitutions. As observed for DNA damage induced by bulky DNA adducts, small deletions were increased in translesion polymerase mutants, while base changes decreased. Structural variants (SVs) were augmented with dose, but did not arise with significantly higher frequency in any DNA repair mutants tested. Moreover, 6% of all mutations occurred in clusters, but clustering was not significantly altered in any DNA repair mutant background. Our data is relevant for better understanding how DNA repair pathways modulate IR-induced lesions.
IBS Institute for Basic ScienceIBS-R022-A2-2021https://orcid.org/0000-0003-4639-9902GartnerAntonhttp://dx.doi.org/10.13039/100004440Wellcome TrustRG70175https://orcid.org/0000-0003-4639-9902GartnerAntonhttp://dx.doi.org/10.13039/100011713Worldwide Cancer Research18-0644https://orcid.org/0000-0003-4639-9902GartnerAntonThis work was supported by the Wellcome Trust [COMSIG consortium grant RG70175 to A.G., Senior Research award 090944/Z/09/Z to A.G.]; Worldwide Cancer Research [18-0644 to A.G.]; and the Korean Institute for Basic Science [IBS-R022-A2-2021 to A.G.]. Nadezda Volkova is a member of Lucy Cavendish College, University of Cambridge. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.Data AvailabilitySequencing data are available under ENA Study Accession Numbers ERP000975 and ERP004086 with ENA sample IDs annotated in Supplementary S1 Table. The R code for analysis of mutation rates, mutation signatures, and mutation clustering is available on GitHub under http://github.com/gerstung-lab/radiation.Introduction
Genome integrity is essential for cellular and organismal survival. One of the most important genotoxic agents is IR. Irradiation occurs during environmental or accidental exposure, during diagnostic radiology or cancer radiotherapy. Exposure to radon, a naturally occurring radioactive noble gas produced by the decay of Uranium-226 in the earth’s crust, is reported to be associated with ~10% of lung cancers [1]. Health consequences of IR exposure were studied on survivors of atomic bomb explosions [2–4] and the Chernobyl nuclear reactor accident that led to the release of large quantities of Iodine-131 and Caesium-137 [5]. Long-term effects of radiation exposure are evaluated epidemiologically as increased cancer incidence, an undertaking hampered by latency periods that can span several decades and the difficulty to establish causal relationships when the exposure to radiation is low. Utilizing the genotoxic potential of radiation, radiotherapy is a highly effective cancer treatment, first employed within a year after the discovery of IR in the late 19th century, and currently applied to treat ~40% of all UK cancer patients [6].
The effects of IR on inheritance were first determined by Herman Müller in the 1920s, recording increased frequencies of lethal and visible mutations in fruit flies [7]. In C. elegans, counting the number of lethal mutations associated with ‘chromosomal translocations’ allowed for the frequency of such events to be gauged at 0.12 per gamete per 15 Gy of IR [8, 9]. More recently, genome-wide sequence analyses of IR exposure have started to emerge. In budding yeast, whole genome sequencing detected ~2 single nucleotide substitutions following exposure to 100 Gy of γ-rays [10]. The use of IR to induce mutations for plant breeding motivated studies in several plant species including Arabidopsis, rice and banana. In Arabidopsis, 200 Gy of carbon ion beam radiation yielded 20–60 single nucleotide variants (SNVs) [11], and 60 Gy of fast neutron radiation led to a range of 8–32 mutations in 6 sequenced lines [12]. In rice, sequencing of 41 lines derived from irradiated parental plants treated with 20 Gy of fast neutrons yielded an average of ~31 SNVs, ~3.5 deletions, ~3.5 insertions, ~2 inversions and ~4.2 translocations per line [13]. In banana plants, large deletions were observed in ~⅔ of lines subjected to 20 Gy or 40 Gy of γ-radiation from a cobalt-60 source [14]. A comparison of the mutation rates in the progeny of mice exposed to 3 Gy of X-rays and unirradiated controls did not identify elevated numbers of SNVs, but a ~2 fold increased frequency of indels and an average of 0.08 copy number variants (CNVs) [15]. These results are broadly in line with a recent study confirming the absence of IR-induced SNVs, but reporting on average 9.6 radiation-induced indels in the progeny derived from mice spermatogonia irradiated with 4 Gy [16]. In contrast, the analysis of 12 human secondary malignancies associated with radiation treatment revealed an average of ~4000 SNVs with all four DNA bases mutated at equal frequency [17]. In addition, ~200 deletions smaller than 100 bp in size, often with microhomology at their breakpoints, were observed [17]. The radiation doses applied to treat the corresponding primary malignancies were not reported, but cumulative cancer radiotherapy doses typically range from 20 to 80 Gy [18]. More recent studies, analysing thyroid carcinomas, linked to Iodine-131 exposure related to the Chernobyl disaster failed to detect increased SNV levels, indels and SVs being increased [19]. Human organoid exposure to IR similarly only led to increased indels and SVs [20]. In radiation induced tumours induced in p53 mutant mice SNVs levels were reported to be rare, indels and SV being detectable [21].
In summary, while these analyses provide insights into mutation rates and mutation types, their comparison is complicated by the types of analysis applied, differences in effects of chosen radiation sources, DNA repair status, tissue type, and possible organismal differences. Also, these studies do not provide insight into the relative contribution of various DNA repair pathways counteracting IR induced mutagenesis.
IR directly damages DNA by causing breaks in the DNA backbone or hydrogen bonds between bases, or by damaging bases. In addition, approximately two thirds of all damage caused by X- and γ-rays is thought to be caused by reactive oxygen species that form when water or organic molecules are hit [22, 23]. In human cells, 1 Gy of radiation is thought to induce 2000 base damages, 1000 DNA single-strand breaks, 40 DNA double-strand breaks, and 150 DNA-protein crosslinks [24, 25]. Modelling of radiation track structures with X- or γ-rays, led to two or more ionizations within 1–4 nm a distance that corresponds to the diameter of the DNA double helix and associated water layers, thus potentially inducing clustered DNA lesions [23–25]. The high quantity and multifarious nature of DNA damage inflicted by IR, in contrast to the relatively low mutation rates reported, imply that cells utilize their arsenal of DNA damage repair pathways to efficiently restore their genetic information from the vast majority of IR-induced lesions. To the best of our knowledge, the contributions of various DNA repair and damage response pathways to mending DNA damage caused by IR has not been systematically investigated on a genome-wide scale.
We recently reported a global, high-level analysis encompassing 2700 C. elegans genomes treated with 11 genotoxic agents [26]. This study demonstrated that DNA repair activities play a major role in shaping mutagenesis. Altered mutation rates and spectra were observed in 40% of analysed DNA repair mutants from various DNA repair pathways [26]. Here, using the same primary sequence data, we expand the analysis to provide a detailed and comprehensive investigation of mutagenesis induced by exposure to Cs-137. We dissect the role of DNA repair and DNA damage response factors in preventing or modulating IR-induced mutagenesis through mutation rate calculations and mutation profile analysis. Our study provides a comprehensive catalogue of radiation damage in wild-type and DNA repair deficient C. elegans. Our study highlights that the number of mutations (especially SVs) inflicted by IR is surprisingly low and that determining the contribution of various DNA repair pathways in preventing SVs is challenging. Our data are consistent with the notion that the vast majority of IR-induced lesions are repaired by several, redundant DNA repair pathways. Our results are relevant for a better understanding of the genome-wide effects of accidental or intended IR exposure and may influence clinical strategies to sensitize tumour cells to IR-induced cell death by targeting the most protective DNA repair pathways.
Materials and methodsC. elegans strains
C. elegans mutants used in this study (S1 Table) were backcrossed 6 times against the wild-type N2 reference strain TG1813 [27] and frozen as glycerol stocks.
Treatment of C. elegans strains and sample preparation for sequencing
L4 stage animals were irradiated with doses of 20 Gy, 40 Gy, 60 Gy and 80 Gy using a Cs-137 source (IBL 437C, CIS Bio International), and doses of 50 J, 100 J, 200 J and 500 J using a Waldman UV 236 B device with a spectrum of 280 nm to 360 nm (UV-B / UV-A light). Cisplatin exposure was performed as described [27]. In short, animals were incubated under gentle shaking for 16h at 20°C in M9 liquid culture with cisplatin (Sigma-Aldrich, P4394) concentrations from 0 to 400 μM (S1 Table). Treated animals were allowed to recover and 3 times 3 adults per genotype and dose were transferred onto fresh plates 24h post treatment. Adults were removed after 4h and eggs laid within this time were allowed to hatch and scored for viability [26, 27]. 2 F1 L4s per plate were then singled. For higher doses of radiation, not all F1 animals produced progeny. Progeny of only one of each set of 2 F1 lines was frozen, providing 3 independent lines per genotype and treatment. Genomic DNA was isolated using Invitrogen ChargeSwitch® gDNA Mini Tissue Kit (Thermo Fisher Scientific, CS11204) and sent for sequencing.
Variant analysis and mutation rate estimation
DNA samples were sequenced using Illumina HiSeq 2000 100 bp short read paired-end sequencing. Alignment, variant calling, post-processing filtering, and data analysis were performed, as described previously [26, 28]: raw reads were aligned against WBcel235.74.dna.toplevel.fa reference genome (http://ftp.ensembl.org/pub/release-74/fasta/caenorhabditis_elegans/dna/) using BWA [29], and variant analysis was performed using CaVEMan, Pindel and DELLY [30–32]. A parental wild-type line was used as a control for variant calling. Additional filtering included applying cut-offs on coverage, read support of the variant in the test and control samples, and overlap with other genomic variants (particularly applicable for base changes and indels in homopolymer runs), as previously described [26]. Variants of each sample were additionally filtered against a panel of unrelated samples to exclude technical artifacts (for more details, see S4 File). Detailed information of samples used in this study and their corresponding ENA accession codes are provided in S1 Table.
Statistical analysis of mutation rates and signatures
Average numbers of SNVs, MNVs, indels, SVs and overall numbers of mutations per dose were calculated using non-zero intercept additive Poisson regression across all samples and radiation doses for a given genotype. Effects of different genetic backgrounds, the mutational signature of IR, as well as the mutation rate fold-changes per genotype were inferred following a model almost identical to the one described in [26]: we modelled observed mutation counts in 119 mutation types (96 SNVs based on their trinucleotide contexts, 14 indels of different types and sizes, and 7 types of SVs) using additive Poisson regression with separate components for genomic contribution and IR: for a genetic background G, YG ~ Poisson(G + dose∙IRwt∙exp(LFCG), where G is a 119-long vector denoting the background contribution shared across all samples of the same genotype, IRwt denotes mutational signature of IR in wild-type, and LFCG ∈ ℝ119 denotes log fold-change between IR signature in background G and in wild-type. Estimates for the mean and standard deviation of the parameters were obtained from HMC sampling of the observed counts.
Significance of the differences between mutation rates in different mutation classes was assessed based on z-test comparing the coefficients estimated from additive Poisson model for each DNA repair deficient background and that for wild-type [33]. Statistical comparisons across genotypes were done applying false discovery rate (FDR) control using the Benjamini-Hochberg procedure to correct for multiple testing [34].
Analysis of clustered mutations
Clustering of mutations was assessed using the start points of all base substitutions and indels across samples of the same genotype and generation. Clustered status was assigned based on a sliding window of 1000 bps.
To obtain the rates of clustering per dose, we used a linear model: P ~ proportions + ε, ε ~ N(0,σ2). Clustering estimates in DNA repair deficient backgrounds were compared to those in wild-type by the following Z test:
Z=rg-rwtSE(rg)2+SE(rwt)2 [33]. False discovery rate among the resulting p-values was corrected for multiple testing using Benjamini-Hochberg procedure [34].
Cancer data analysis
Analysis of cancer samples associated with radiation exposure was performed using the WGS data for 12 radiation-associated secondary malignancies from [17]. Data for bone and breast cancers without an obvious association to IR exposure was obtained from the ICGC data portal (https://icgc.org) by selecting WGS-analysed primary tumour samples with less than 10,000 mutations/genome from BRCA-UK (33 samples) and BOCA-UK (62 samples) projects. Mutation cluster analysis was performed following the same procedure as described above using different transition probabilities: 0.01 for transition from clustered to non-clustered state and 0.1 for an inverse transition. For mutational spectra comparison, C. elegans mutation counts were adjusted to the differences in human trinucleotide composition [28].
ResultsIR-induced mutations in C. elegans wild-type
Mutation spectra derived from wild-type C. elegans exposed to increasing doses of IR in independently conducted, triplicate experiments, revealed a linear and dose-dependent increase of single nucleotide variants (SNVs) (Figs 1B and 2A and S1 Table, list of C. elegans strains and samples). Indels and structural variants (SVs) were also increased, albeit to a lesser extent, occurring with a lower incidence (Figs 1B and 2A). Using additive Poisson regression on mutations across all wild-type samples, we calculated an average number of 36.63 (SD = 1.13) SNVs, 1.3 (SD = 0.2) DNVs, and 4 (SD = 0.4) indels and 1.4 (SD = 0.13) SVs per 80 Gy (Fig 1B) (Materials and Methods). Mutation distribution was uniform across the 6 possible SNVs C>A, C>G, C>T, T>A, T>C and T>G (Figs 1B and 3B, top panel) with no overt influence of adjacent 5’ or 3’ nucleotides [26]. A similarly flat SNV pattern has previously been reported in radiation-associated human cancers [17]. To enable comparison between our results and human cancer profiles, we adjusted the C. elegans mutational SNV pattern to trinucleotide frequencies in the human genome, as described previously (Fig 1C) [28]. Using the humanized C. elegans pattern, we identified three human cancer mutation profiles with high similarity; the SNV pattern observed in radiation-associated secondary tumours (cosine similarity = 0.85) [17], COSMIC signature 3 (cosine similarity = 0.84), and COSMIC signature 40 (cosine similarity = 0.92). COSMIC signatures were computationally extracted from thousands of cancer genomes encompassing most cancer types [35]. COSMIC signature 3 has been associated with HR deficiency, while signature 40 is increased with age and is detectable in a large variety of tumours where it typically only contributes a small proportion of mutations. The reduced frequency of C>A, C>G, C>T changes in the context of G as the 3’ nucleotide visible in the humanised C. elegans patterns and three related cancer signatures (Fig 1C) can be explained by the reduced abundance of CG sequences in the mammalian genome resulting from 5-methylcytosine deamination. Interestingly, mutational signatures derived from HR defective C. elegans lines or DT40 cells propagated over multiple generations without exposure to DNA damaging agents, also show a uniform SNV profile in conjunction with small deletions and SVs [28, 36]. Among deletions, 1 bp deletions were most common, followed by 2–5 and 6–50 bp deletions, and >50 bp deletions in C. elegans and radiation-associated secondary tumours (S1A Fig) [17]. Turning to SVs, we detected 0–2 SVs in irradiated wild-type lines (Figs 1B and S1B). The proportion of SV types varied between irradiated nematodes and radiation-associated secondary tumours; deletions occurred more frequently in C. elegans whereas inversions and translocations were more prominent in human cancer (S1B Fig). In summary, the pattern of IR-induced SNVs is evenly spread across all base changes in C. elegans wild-type and no overt influence of 5’ and 3’ bases could be observed. In addition, IR-exposure induced indels and SVs. The comparison between the γ-ray treated C. elegans and x-ray treated human cancer samples may thus suggest that both IR sources induce base damage and subsequent substitutions that are largely independent of sequence context.
10.1371/journal.pone.0258269.g001
Experimental outline and mutation patterns induced by ionising radiation in C. elegans wild-type compared to human cancer samples.
A. Experimental outline: Parental (P0) animals were exposed to a Cs137 radiation source at the L4 developmental stage. 3 times 3 P0 animals were allowed to lay eggs and two L4s of the first filial (F1) generation per P0 plate were individually transferred to new plates and allowed to clonally expand by self-fertilization. One F1 line per P0 plate was used for DNA preparation and whole genome sequencing (Materials and Methods). B. The number of mutations observed in two biological replicates exposed to increasing doses of Cs-137 radiation are shown. Mutations are categorized into base substitutions (upper panel) as 6 possible single nucleotide variants (SNV), dinucleotide variants (DNV) and multi-nucleotide variants (MNVs), indels (center panel) as insertions (INS), deletions (DEL) and deletions with insertions (INDEL), and structural variants (SVs) (lower panel) encompassing tandem duplications (TD), large deletions (DEL), inversions (INV), complex (COMPLEX) SVs, translocations (TRSL), interchromosomal rearrangements (INTCHR) and foldback structures (FOLDBACK). C. Comparison of the IR-induced C. elegans SNV profile adjusted to the human trinucleotide content with mutation profiles observed in radiation-exposed human secondary tumours (human) and COSMIC signatures 3 (Sig 3) and 40 (Sig 40) [17, 35]. SNV profiles are depicted as the relative contribution of the six possible SNVs in their 5’ and 3’ base context. Cosine similarities between the humanized C. elegans signature and human cancer profiles are shown.
10.1371/journal.pone.0258269.g002
Dose dependent mutation burden in wild-type and DNA repair mutants.
A. Number of mutations observed across Cs-137-irradiated wild-type samples by dose and mutation type; single (SNVs) and multi- nucleotide variants (MNVs) (left panel), indels (center panel) and structural variants (left panel). Black dots represent mutations observed in individual samples at a given dose, red lines represent a best fit linear regression. The values of r and p denote the Pearson correlation coefficient and its two-sided p-value between the number of mutations and radiation dose for each mutation type, respectively. B. Numbers of IR-induced mutations in DNA repair mutants compared to wild-type (dashed line); single nucleotide variants (SNVs) (top left), multi-nucleotide variants (MNVs) (top right), indels (bottom left) and structural variants (SVs) (bottom right). Blue and red dots indicate the average number of mutations per 80 Gy IR; red dots represent samples that exhibit statistically significantly different mutation numbers to wild-type (FDR adjusted p-value < 0.05). Grey bars denote 95% confidence intervals.
10.1371/journal.pone.0258269.g003
IR-induced mutation profiles of DNA repair mutants significantly different from wild-type.
A. Similarity of IR-induced mutation profiles in DNA repair mutants compared to wild-type. The level of divergence from the wild-type mutation profile is indicated by cosine distance. Black bars represent mean cosine distances, blue shaded areas confidence intervals. Dark blue shaded bars indicate mutation profiles which are significantly different from wild-type. B. Mutation profiles that are different from wild-type as determined in panel A. Each barplot represents the number of Cs-137 radiation induced mutations per average dose of 80 Gy. Coloured bars represent the mean number of mutations per mutation type, black bars indicate 95% confidence intervals.
Contribution of DNA repair deficiencies to IR-induced mutation patterns
To systematically explore how various DNA repair pathways contribute to the repair of Cs-137-inflicted DNA damage, we investigated mutation rates and spectra from 48 DNA repair and DNA damage response deficient C. elegans strains (S1 Table) [26]. These include strains defective for apoptosis, p53, base excision repair (BER), single-strand break repair (SSBR), nucleotide excision repair (NER), mismatch repair (MMR), direct damage reversal mediated by O6-methyl guanine methyl transferases (MGMTs), double-strand break repair (DSBR) by homologous recombination (DSBR-HR), non-homologous end-joining (DSBR-NHEJ) and microhomology-mediated DNA end-joining (DSBR-MMEJ), DNA crosslink repair (CL), and mutants defective for translesion synthesis (TLS) polymerases [26] (S1 Table). Overall, all DNA repair mutants proliferated normally under unchallenged conditions. Upon irradiation, we observed a high level of variation when comparing mutation numbers between individual irradiated lines, especially for indels and structural variants (Fig 2A, wild-type; S1 File, all genotypes). We thus applied rigorous criteria to assess which genotypes led to significantly increased or decreased rates of SNVs, indels and SVs in response to IR, requiring that at least two classes of mutations show a dose dependent increase (Fig 2A, wild-type; S1 File, all genotypes). 35/44 mutants fulfilled these dose response criteria and were subsequently analysed in detail (S2–S5 Figs, for raw mutation count data for all 44 strains by DNA repair and damage response pathway). We found that mutation numbers and mutational patterns induced by Cs-137 did not vary greatly across most genotypes. Of the 35 mutants analysed, 10 displayed significantly elevated mutagenesis (at 5% FDR) (Figs 2B and 3, for significantly different mutation profiles, S2 File, for all mutation profiles). Strains with mutational patterns significantly different from wild-type are indicated by an asterisk (*) (P ≤ 0.05), strains excluded from our detailed analysis are indicated by ∉.
DNA repair defects associated with increased levels of SNVs
SNV numbers and indels were significantly elevated in xpa-1, xpc-1, and xpf-1 NER mutants (Figs 2B and S2A). XPF-1 and XPA-1 contribute to both global genome (GG-NER) and transcription coupled NER (TC-NER), XPC-1 is solely involved in GG-NER, and CSB-1 specifically contributes to TC-NER. Our finding that mutagenesis was increased in xpa-1, xpc-1, and xpf-1 but not in csb-1 mutants (Figs 2B and S2 and S2 File) indicates that a high proportion of IR-induced base changes and indels were prevented by error-free GG-NER.
SNV numbers and indels were also increased in brc-1 HR defective lines (Figs 2B and 3B and S3). Indels (but not base substitutions) were also increased in rad-54, slx-1 and mrt-2 HR defective mutants (Figs 2B and 3B and S2 File and S3 Fig) and a strain defective for the ung-1 base uracil glycosylase. This enzyme facilitates base excision repair by generating an abasic site upon uracil removal, uracil arising from the deamination of cytosine (Fig 2B and S2 File and S4 Fig). Finally, the number of indels was increased in rev-1 and polh-1 translesion synthesis mutants (Figs 2B and 3B and S2 File and S5 Fig). None of the mutants examined showed a statistically increased rate of SVs.
Of the direct DNA repair pathways, BER and SSBR, mutants of agt-2 encoding for a putative alkyl-guanine DNA methyltransferase and parp-2, encoding for a Poly ADP ribose polymerase did not show altered rates of mutagenesis (S4 Fig). agt-1, apn-1, exo-1, ndx-1, parp-1 and tdpo-1 BER and SSBR mutants were not included into our stringent dataset. While base substitutions tended to be increased in these lines similar to wild-type irradiated with 40 Gy, no further dose response was observed at 80 Gy possibly due to an experimental error (S4 Fig). All in all, our data indicate that, except for ung-1 mutants, which showed a higher rate of indels, BER and SSBR mutants did not exhibit IR-induced mutagenesis above wild-type levels.
DNA repair defects associated with decreased levels of SNVs
Interestingly, him-6 mutants, defective for the C. elegans homolog of the BLM helicase involved in multiple steps of HR (Figs 2B and S4B) [37], polh-1 and rev-1 translesion polymerase mutants (Figs 2B and S5), as well as ced-4 mutants, defective for the C. elegans Apaf-1 like apoptosis gene, showed a reduced number of SNVs upon IR exposure (Figs 2B and S6). The evidence of decreased SNVs in him-6 is supported by dose-dependent wild-type levels of indels and SVs (Figs 2B and S4). POLH-1 translesion synthesis polymerase mutants were unique, showing a significantly altered mutational profile (Fig 2C) with reduced number of SNVs concomitant with increased numbers of indels (Figs 2C and 3 and S2 File and S5 Fig). Reduced SNV numbers were also observed in polh-1 and rev-3 mutants treated with alkylating agents, in line with the role of these polymerases in error-prone bypass of damaged bases [26].
Dinucleotide variants
Dinucleotide substitutions or variants (DNVs) occur more frequently in cancers than expected from calculating random SNV distributions, and thus likely reflect distinct mutational processes [35]. Investigating the contribution of DNA repair pathways to Cs-137-induced DNVs, we found that DNVs were observed infrequently in wild-type upon Cs-137 treatment (Figs 1B and 2B and 3 and 3B and S2 and S3). Among DNA repair mutants tested, DNVs were most prevalent in xpa-1, xpc-1 and xpf-1 NER and brc-1 HR mutants (Figs 2B and 4B), indicating that NER and HR are required to repair DNA lesions that otherwise lead to dinucleotide substitutions. Interestingly, csb-1 mutants, specifically defective for transcription-coupled NER, did not show increased levels of DNVs (Figs 2B and 4B), suggesting that GG-NER plays a major role in preventing dinucleotide substitutions upon IR treatment. The pattern of IR-induced DNVs in C. elegans is not closely related to any DNV pattern deduced from cancer genomes, including DNV signatures DBS2 and DBS11, both of which are largely composed of CC>NN changes [35] (Fig 4B, lower panels).
10.1371/journal.pone.0258269.g004
Dinucleotide substitution profiles induced by ionising radiation, UV-light and cisplatin exposure in C. elegans and human cancers.
A. Relative contribution of dinucleotide substitutions observed in C. elegans wild-type following Cs-137, UV-B, and cisplatin exposure. B. Cs-137-induced dinucleotide substitutions in C. elegans wild-type and DNA repair mutants compared to closely related COSMIC dinucleotide substitution signatures described in human cancers. C. UV-induced dinucleotide substitutions in C. elegans wild-type and DNA repair mutants compared to closely related COSMIC dinucleotide substitution signatures described in human cancers. D. Cisplatin-induced dinucleotide substitutions in C. elegans wild-type and DNA repair mutants compared to a closely related COSMIC dinucleotide substitution signature described in human cancers.
In C. elegans DNVs arise from IR, UV and cisplatin exposure, however on different bases and with different mutation outcomes (Fig 4A). UV exposure predominantly led to CC>AT and CC>TT changes (Fig 4A), the relative proportion being increased in xpf-1 mutants (Fig 4A and 4C). CC>TT changes are the hallmark of cancer DNV signature DBS11, associated with UV-induced melanoma [35]. Cisplatin exposure, predominantly led to CT>AC changes [27] (Fig 4D) a signature related to the cancer DNV signature DBS5, associated with cisplatin-induced tumours and characterised by CT>AA and CT>AC changes [35]. In summary, our data shows that dinucleotide variants arise following a variety of DNA insults. This is consistent with the high propensity of dinucleotide substitutions in a variety of human cancer types. DNVs likely occur during the repair of DNA crosslinks between adjacent bases, DNA lesions that are recognized and can be repaired error-free by the NER pathway.
IR-induced mutation clusters
Both the direct action of IR, which can induce base damage and breaks in the DNA backbone, as well as the indirect action via reactive oxygen species has the potential to create clustered DNA damage. Furthermore, DNA repair synthesis in HR and DSB repair by non-homologous end-joining or microhomology-mediated end-joining are error-prone, potentially resulting in clustered mutations [23–25]. We thus analysed the clustering of IR-induced mutations. Globally, mutations were randomly distributed across chromosomes throughout all wild-type (Fig 5A and S3 File) and DNA repair deficient lines analysed (S3 File). Analysing local mutation distribution, we observed significant mutation clustering in irradiated wild-type lines compared to untreated wild-type. Overall, 6% of all mutations occurred in clusters (P < 1.7x10-7) with ~ 1.2 clusters induced per 80 Gy per genome (P < 1x10-7) (Figs 5A and 5B and S7 and S2 Table) Clustering was not observed in unchallenged C. elegans strains propagated over generations [38]. Clusters typically spanned 10–20 bp and on average consisted of 2–3 mutations (Fig 5C). Interestingly, clustered mutations did not occur more frequently in DNA repair mutants including those deficient in error-free and error-prone DSBR (Figs 5A–5C and S7). Sequence analysis of 12 radiation-associated secondary malignancies from patients with different tumour types was recently reported [17]. Using the same primary data, we observed 10–150 mutation clusters in each of these genomes with 1–10% of all mutations occurring in clusters (S8A and S8B Fig). As in C. elegans, most clusters spanned less than 50 bp and encompassed predominantly 2 and rarely more than 4 bases (S8C and S8D Fig). However, the frequency and type of mutation clusters were not significantly different between radiation-induced cancers and cancers of the same type not associated with a history of radiation exposure (S8E and S8F Fig). Thus, these data do not provide firm evidence for mutational clustering linked to IR exposure in human secondary malignancies.
10.1371/journal.pone.0258269.g005
IR-induced mutation clusters.
A. Number of mutation clusters in IR-exposed C. elegans strains per 80 Gy of Cs-137 radiation. Grey dots indicate the number of mutation clusters in different C. elegans mutants. The red dot and dashed line indicate the number of mutation clusters observed in wild-type. Black bars indicate the median mutation cluster size, squares the interquartile range and the error bar the low and high 1.5*interquartile ranges. B. Proportion of IR-induced mutations within clusters for wild-type (red dot, dashed line) and DNA repair deficient C. elegans lines (grey dots) as depicted in A. C. Boxplots representing the sizes of IR-induced mutation clusters in bp for wild-type and DNA repair mutants. Grey dots represent the size of individual mutation clusters observed for the respective genotype. D. Number of mutations per mutation cluster (black dots) in IR-exposed wild-type and DNA repair mutants.
Discussion
Using C. elegans as a model, our study is the first to systematically describe IR-induced mutation rates and patterns in wild-type and DNA repair and damage response defective strains encompassing the vast majority of known pathways needed for mending DNA lesions. In line with expectations of IR induced mutation rate estimates in mouse and C. elegans germ cells [8, 9, 39] and from sequencing the progeny of irradiated mice, budding yeast and several plant species (for instance [10–16]), we confirmed that IR-induced mutagenesis is surprisingly low, with 80 Gy inducing an average of 36.6 SNVs, 1.3 DNVs, 4 indels and 1.4 (SD = 0.13) SVs in the 100 million base C. elegans genome. The highest proportion of mutants being SNVs is in line with the analysis of IR induced secondary tumours [17]. In contrast, increased SNVs could not be detected in thyroid carcinomas, linked to the Chernobyl disaster [19] and irradiated human organoids [20].
We typically did experiments in triplicates, assessing samples without, and with an intermediate and high dose of IR. Clearly, increased statistical power would require the analysis of a larger number of samples. It is impossible to significantly increase the dose of IR as this would lead to 100% embryonic lethality. Broadly, 80 Gy leads to a 30 to 50% reduction of survival of irradiated progeny, with up to 10-fold lower survival in C. elegans DNA repair mutants [40–45]. Radiation sensitivity inversely correlates with genome size: Only 8–12 Gy are needed to cause 50% lethality in mice 5–8 weeks after whole body irradiation [46], mice having a 25 times larger genome than C. elegans with its 100 million base haploid genome. In contrast, ~1000 Gy are needed for 50% lethality in diploid budding yeast cells carrying a 12 million base haploid genome, haploid cells being ~20 times more sensitive [47]. Assuming that SVs are the most deleterious mutations, it appears unlikely that the reduced survival can solely be ascribed to IR-induced DNA lesions. This hypothesis is supported by the presence of predominantly heterozygous SVs detected in our experimental system; heterozygous SVs generally not causing phenotypes (also see discussion below) (Fig 1A). Also, it is unlikely us having missed complex chromosomal rearrangements, us having detected those previously in C. elegans lines treated with the DNA crosslinking agents cisplatin and nitrogen mustard using the same analysis pipeline [27]. On the other hand, while a large number of C. elegans fusion chromosomes are known to be viable in a heterozygous state and are being used as balancer chromosomes [48], aneuploidy of autosomes, typically caused by meiotic segregation defects, leads to embryonic lethality [49]. It is thus possible that we fail to detect a subset of chromosomal fusions or the loss of large chromosomal fragments resulting from unrepaired DSBs, if these events lead to a near aneuploid state and consequently to embryonic or larval lethality. In other words, we might fail to detect highly mutagenised genomes because affected embryos die, and do not sire progeny needed for whole genome sequence analysis. Detection of mutational events leading to embryonic lethality would require single embryo or single zygote whole genome sequencing, studies we will pursue in the future.
When the consequence of germline mutagenesis is scored in the subsequent generation, it is important to consider the nature of the germ cells exposed to IR, mindful that mutagenic outcomes of individual germ cell cells are scored. Such a quantal approach likely explains the relatively large variation we observe in parallel experiments (Fig 1B and S1 File). Our experimental system, which mimics the procedure used in mutagenesis C. elegans genetic screens over the past six decades [9], takes advantage of C. elegans’ self-fertilizing reproduction to clonally amplify the genome from single zygotes, derived from male and female germ cells treated with IR (Fig 1B). Based on the timing of germline differentiation, including the staging of mitotic and meiotic cell cycles, the timing after IR treatment is chosen such that female germ cells are irradiated in meiotic pachytene. In pachytene, the initiation of meiotic recombination is completed and maternal and paternal chromosomes tightly align into the so-called synaptonemal complex [49, 50]. Male germ cells are produced in C. elegans hermaphrodites before female germ cells are formed and are stored in the spermatheca after the completion of meiotic divisions [51]. It remains to be investigated if mutation rates differ between male and female or between mitotic and meiotic germ cells. A meiotic checkpoint, which only acts on female germ cells, leads to the elimination of a large proportion of female pachytene cells in C. elegans by apoptosis [40]. Apoptosis is compromised in ced-4 mutants, defective for the C. elegans Apaf-1 like gene [40, 52]. CEP-1, the sole p53-like gene [53–55] is essential for DNA damage induced apoptosis triggered by IR [40]. Surprisingly, we did not observe increased mutagenesis in either mutant. Thus, our analysis does not provide evidence that apoptosis is specifically used to eliminate germ cells subjected to excessive DNA damage.
Our data supports previous evidence that only a minority of DNA double-strand breaks inflicted by ionizing radiation is converted into more severe genomic changes. In mammalian cells, about 40 DNA double-strand breaks are induced per Gy of IR [24, 25]. In C. elegans, antibody-stained RAD-51 recombinase foci indicate that ~26 DSBs are present in mitotic germ cells 6 hours after treatment with 120 Gy [41]. About 10 times more DSBs are induced in meiotic cells (24/10 Gy), as estimated by the dose of IR required to bypass a defect in the SPO-11 nuclease essential for meiotic DSB formation [56]. All in all, these results indicate that the vast majority of DNA lesions, including DNA double-strand breaks which can lead to SVs such as insertion, deletions, duplications, and translocations are repaired by various DNA repair pathways. It appears surprising that mutations defective in DNA end-joining (EJ) and microhomology mediated end-joining (MMEJ) do not exhibit increased IR-induced mutagenesis. However, these results are consistent with EJ’s predominant function in somatic cells and a likely redundancy of MMEJ with HR [57–60]. More surprising is our finding that mutations affecting checkpoint signalling or HR, including complete loss of function mutants of brd-1 and brc-1 corresponding to human BRCA1 and BARD1 genes, as well as him-6, the C. elegans BLM helicase homolog, do not show increased numbers of SVs. In some of these mutants, numbers of SVs tended to be increased, but did not reach statistical significance (Fig 2). Importantly, HR is essential for meiotic recombination. However, brc-1, brd-1 and him-6 null mutants are viable and only partially compromised for meiotic HR suggesting genetic redundancy in HR. Nevertheless, brc-1, brd-1 and him-6 deficiencies show reduced survival upon exposure to IR [37, 61]. Also, these three HR mutants show increased numbers of SVs when propagated over 40 generations in the absence of genotoxin exposure [62]. Finally, brc-1, brd-1, and slx-1, a nuclease involved in HR intermediate resolution, [63, 64], show increased numbers of indels, which might also be products of error-prone DSB repair. All in all, these data indicate that DSB repair pathways are highly redundant and that redundant mechanisms also act during HR.
We were intrigued to observe that the number of IR-induced single nucleotide variants was strongly reduced in mutants defective for him-6, even more so than in polh-1 translesion synthesis polymerase mutants. Reduced SNV numbers in translesion synthesis mutants suggest that a sizable proportion of SNVs are generated by POLH-1 likely by incorporation of incorrect nucleotides opposite damaged bases. The failure to bypass damaged bases in the absence of POLH-1 can explain the observed increase in small deletions, which are also present in polh-1 mutants treated with a variety of alkylating agents, UV, and the bulky DNA adduct forming agents aflatoxin B1 and aristolochic acid [26]. At present, we can only speculate why less SNVs are present in him-6 BLM mutants. The late stages of HR are highly redundant and two major pathways, Holliday Junction dissolution mediated by BLM and Holliday Junction resolution mediated by resolving enzymes, such as the GEN-1 nuclease or the SLX-4/SLX-1/MUS-81 nuclease complex, are needed to resolve double Holliday Junctions, that covalently link resected DNA double-strand breaks with template DNA [65–67]. HR is also used to mend DNA base damage that halts DNA replication. Such a situation can lead to fork reversal, the formation of a ‘chicken foot structure’ [68–70], and the subsequent error-free bypass of the damaged base through recombinogenic pathways, which likely involves GEN1 nuclease and/or the SLX-4/SLX-1/MUS-81 resolvases. It is possible that in the absence of HIM-6 more lesions are funnelled into the error-free GEN1 and SLX-4/SLX-1/MUS-81 dependent pathways. Indeed, the increased level of SNVs we observe in brc-1 mutants might be indicative of error-prone translesion synthesis, employed when damaged bases are neither repaired by BER or NER nor by recombination linked to replication fork reversal [70].
The increase of SNVs and DNVs in NER mutants indicates an important role of this pathway in preventing single and dinucleotide changes. These results appear surprising at first glance, as base damage inflicted by reactive oxygen species is largely not helix distorting and thus generally thought to be mended by BER, the deficiency of which in our system had only minimal effects. However, radiation induced reactive oxygen species are known to induce bulky cyclodeoxynucleosides by affecting intramolecular crosslinks between the C-8 position of adenine or guanine and the 5’ position of 2-deoxyribose, cyclodeoxynucleosides being excised by NER [71, 72]. It is possible that the increased rate of C>T changes (G>A on the other strand) observed in xpa-1 and xpf-1 mutants may be related to the failure to repair cyclodeoxyguanine. In addition, ionizing radiation is known to cause multiple types of intrastrand crosslinks [73–77], which may account for radiation-induced dinucleotide changes repaired by NER.
Radiation tracks typically generate more than one reactive oxidative species when penetrating DNA and surrounding water layers [23–25], implying that clustered DNA damage should be highly prominent. While we do observe an increase in clustered mutagenesis, only 6% of all mutations are clustered. Thus, ionizing events in the majority of cases may only cause single mutations or DNA repair of clustered DNA damage is highly efficient and largely error-free. Since we did not find any DNA repair pathway whose inactivation substantially increased clustered mutations, the former possibility seems more likely. Alternatively, the repair of clustered DNA lesions might be highly redundant.
Taken together, we show that IR induces multiple types of mutations. Our data also implies that DNA repair pathways mend the vast majority of DNA lesions caused by IR. Moreover, DNA repair pathways are highly redundant for IR-induced damage repair as single DNA repair mutants, if at all, generally only show a modest increase of mutagenesis. The most dramatic effect was observed in xpf-1 mutants, likely reflecting the function of the XPF-1 nuclease in both nucleotide excision and DNA double-strand break repair [78, 79] and possibly also in DNA crosslink repair [80, 81]. Our data suggest that drugs targeting this nuclease might be useful to sensitise cancer killing by IR.
Supporting information
Strain and sample description.
Details of C. elegans strains used in and sequencing samples of this study. For sample description unless stated otherwise primary whole genome sequencing data were deposited (Filtered VCF files) in Volkova, Meier et al., Nat Commun. 2020 (Supplementary Data 6 file collection) [1].
(XLSX)
Clustered mutations and cluster summary.
Details of all mutations reported as clustered across genotypes and different irradiation doses, as well as a summary of clustering for all samples where any clustered mutations were found.
(XLSX)
Comparative analysis of indels and structural variants observed in radiation-associated secondary malignancies and Cs-137-irradiated C. elegans.
A. Average number of indels observed in secondary malignancies (1) (top panel) and Cs-137 irradiated C. elegans (bottom panel). Indels are classified as deletions (DEL), deletions with insertions (INDEL) and insertions (INS) with indicated size ranges. Error bars represent confidence intervals with 2 standard errors of the mean (SEM) (top panel) and 95% credible intervals (bottom panel). B. Average numbers of structural variants observed in secondary malignancies (1) (top panel) and in Cs-137 irradiated C. elegans (bottom panel) by type. Error bars represent confidence intervals with 2 standard errors of the mean (SEM) (top panel) and 95% credible intervals (bottom panel).
(TIFF)
Mutation numbers and types observed in NER and CLR mutants following irradiation with Cs-137.
A. Mutation numbers by class and type observed in wild-type and nucleotide excision repair (NER) mutants for indicated radiation doses and 3 irradiated lines per genotype. Mutation types are shown as the 6 possible single nucleotide variants, dinucleotide variants (DNV), multi-nucleotide variants (MNVs) (top panel), insertions (INS) and deletions (DEL) and deletions with insertions (INDEL) (centre panel), and structural variants (SVs) encompassing tandem duplications (TD), large deletions (DEL), inversions (INV), translocations (TRSL), foldback (FOLDBACK), interchromosomal (INTCHR) and complex (COMPLEX) SVs (lower panel). B. Mutation numbers by class and type observed in wild-type and DNA crosslink repair (CLR) mutants for indicated radiation doses and 3 irradiated lines per genotype. Mutation types as described in A. N.D. indicates lines for which no DNA sequencing information could be obtained. Strains with mutational patterns statistically significantly different from wild-type are indicated by an asterisk ‘*’ (P ≤ 0.05). Strains lacking a dose-response and thus were excluded from our detailed analysis are indicated with ∉.
(TIFF)
Mutation numbers and types observed in DSBR and helicase mutants following irradiation with Cs-137.
A. Mutation numbers by class and type observed in wild-type and double-strand break repair (DSBR) mutants for 3 irradiated lines per genotype and indicated radiation dose. B. Mutation numbers by class and type observed in wild-type and helicase mutants for indicated radiation doses and 3 irradiated lines per genotype. N.D. indicates lines for which no DNA sequencing information could be obtained. Strains with mutational patterns statistically significantly different from wild-type are indicated by an asterisk ‘*’ (P ≤ 0.05). Strains lacking a dose-response and thus excluded from our detailed analysis are indicated with ∉.
(TIFF)
Mutation numbers and types observed in MGMT and BER mutants following irradiation with Cs-137.
A. Mutation numbers by class and type observed in wild-type and putative direct damage reversal repair 06-Methylguanine methyltransferase (MGMT) mutants for indicated radiation doses and 3 irradiated lines per genotype. B. Mutation numbers by class and type observed in wild-type and base excision repair (BER) mutants for indicated radiation doses and 3 irradiated lines per genotype. N.D. indicates lines for which no DNA sequencing information could be obtained. Strains with mutational patterns statistically significantly different from wild-type are indicated by an asterisk ‘*’ (P ≤ 0.05). Strains lacking a dose-response thus excluded from our detailed analysis are indicated with ∉.
(TIFF)
Mutation numbers and types observed in TLS mutants following irradiation with Cs-137.
A. Mutation numbers by class and type observed in wild-type and translesion synthesis (TLS) mutants for indicated radiation doses and 3 irradiated lines per genotype. N.D. indicates lines for which no DNA sequencing information could be obtained. Strains with mutational patterns statistically significantly different from wild-type are indicated by an asterisk ‘*’ (P ≤ 0.05). Strains lacking a dose-response thus excluded from our detailed analysis are indicated with ∉.
(TIFF)
Mutation numbers and types observed in SAC and apoptosis mutants following irradiation with Cs-137.
A. Mutation numbers by class and type observed in wild-type and spindle assembly checkpoint (SAC) mutants for indicated radiation doses and 3 irradiated lines per genotype. B. Mutation numbers by class and type observed in wild-type and apoptosis mutants for indicated radiation doses and 3 irradiated lines per genotype. N.D. indicates lines for which no DNA sequencing information could be obtained. Strains with mutational patterns statistically significantly different from wild-type are indicated by an asterisk ‘*’ (P ≤ 0.05). Strains lacking a dose-response thus excluded from our detailed analysis are indicated with ∉.
(TIFF)
IR-induced mutation clusters in DR, BER and RecQ helicase mutants.
Sizes of IR-induced mutation clusters shown in bp spanned by the mutation cluster (top panel) and as the number of mutations per cluster (bottom panel). Grey or black dots indicate bp spanned or mutations by cluster, respectively, observed in individual lines of the indicated genotypes. Black bars of the boxplot (right top panel) indicate the median mutation cluster size, squares the interquartile range and error bars 1.5*interquartile ranges.
(TIFF)
Mutation clusters in human tumours with and without association to IR-exposure.
A. Observed number of mutation clusters in 12 IR-associated human secondary malignancies (Material and Methods) [1]. B. Proportion of clustered mutations in 12 IR-associated human secondary malignancies. C. Boxplots depicting the sizes of mutation clusters in 12 IR-associated human secondary malignancies in bp. Black bars indicate the median mutation cluster size, squares the interquartile range, and error bars 1.5* interquartile ranges. Grey dots represent the size of individual mutation clusters observed in the respective tumour. D. Number of mutations per mutation cluster in 12 IR-associated human secondary malignancies. Black dots represent the number of mutations observed in individual clusters of the respective tumour. E. Number of mutation clusters in different human cancers. Mutation clusters are shown as the relative number of mutation clusters to total mutation burden. Grey dots represent the number of mutation clusters in 33 individual non-IR associated BRCA negative breast cancers, 3 IR-associated secondary breast cancers (PD8618a, PD8622a, PD8623a, Panel A-D), 62 non-IR associated bone cancers, and 9 IR-associated bone cancers including angio-, osteo-, and spindle cells sarcomas (Panel A-D). Outliers are highlighted by a black outline. Black bars indicate the median mutation cluster size, squares the interquartile range and error bars 1.5* interquartile ranges. F. Proportion of clustered mutations in different human cancers shown in non-IR and IR-associated human cancers as described in E.
(TIFF)
Wild-type and DNA repair mutant dose response.
Details of the observed number of mutations in all samples of C. elegans wild-type and mutant strains across radiation doses. Number of mutations observed across all Cs-137-irradiated samples by dose and mutation type: single nucleotide variants (left panel), indels (center panel) and structural variants (left panel). Black dots represent mutations observed in individual samples at a given dose, red lines represent a best fit linear regression.
(PDF)
IR-induced mutation profiles of wild-type and DNA repair mutants.
Each barplot reflects the number of mutations per average dose of 80 Gy of Cs-137-radiation. Three stars indicate samples with SNVs, MNV, indels, or SV significantly different to wild-type (FDR < 5%). Bold lines below a mutation class indicate that a specific substitution type with the above classes differs from wild-type (FDR < 5%).
(PDF)
Chromosomal location of mutations in wild-type and DNA repair mutants.
Chromosomal location of mutations observed after exposure to the indicated radiation doses across the 5 C. elegans autosomes (I-V) and the X chromosome for all genotypes. Small circles, triangles, and squares, indicate single SNVs, indels, and DNVs, respectively. Larger circles, triangles, and squares, indicate clustered SNVs, indels, and DNVs, respectively.
(PDF)
(DOCX)
We thank Orlando D. Schärer for helpful discussions on the NER pathway and Dmitry Ivanov for proofreading. We are grateful to the Mitani Lab (National Bio-Resource Project, Japan), and the Caenorhabditis Genetics Center (P40 OD010440) for providing strains.
ReferencesAjroucheR, IelschG, CléroE, RoudierC, GayD, GuillevicJ, et al. Quantitative Health Risk Assessment of Indoor Radon: A Systematic Review. . 2017. pp. 69–77. doi: 10.1093/rpd/ncx15229036475PrestonDL, RonE, TokuokaS, FunamotoS, NishiN, SodaM, et al. Solid Cancer Incidence in Atomic Bomb Survivors: 1958–1998. . 2007. pp. 1–64. doi: 10.1667/RR0763.117722996RichardsonD, SugiyamaH, NishiN, SakataR, ShimizuY, GrantEJ, et al. Ionizing radiation and leukemia mortality among Japanese Atomic Bomb Survivors, 1950–2000. . 2009;172: 368–382. doi: 10.1667/RR1801.119708786DoupleEB, MabuchiK, CullingsHM, PrestonDL, KodamaK, ShimizuY, et al. Long-term radiation-related health effects in a unique human population: lessons learned from the atomic bomb survivors of Hiroshima and Nagasaki. . 2011;5Suppl 1: S122–33.CardisE, KrewskiD, BoniolM, DrozdovitchV, DarbySC, GilbertES, et al. Estimates of the cancer burden in Europe from radioactive fallout from the Chernobyl accident. . 2006;119: 1224–1235. doi: 10.1002/ijc.2203716628547MaylesWPM, Radiotherapy Development Board. Survey of the availability and use of advanced radiotherapy technology in the UK. . 2010;22: 636–642. doi: 10.1016/j.clon.2010.06.01420667700MullerHJ. ARTIFICIAL TRANSMUTATION OF THE GENE. . 1927;66: 84–87. doi: 10.1126/science.66.1699.8417802387RosenbluthRE, CuddefordC, BaillieDL. Mutagenesis in Caenorhabditis elegans. II. A spectrum of mutational events induced with 1500 r of gamma-radiation. . 1985;109: 493–511. doi: 10.1093/genetics/109.3.4933979812KutscherLM. Forward and reverse mutagenesis in C. elegans. . 2014. pp. 1–26. doi: 10.1895/wormbook.1.167.124449699St CharlesJ, Hazkani-CovoE, YinY, AndersenSL, DietrichFS, GreenwellPW, et al. High-resolution genome-wide analysis of irradiated (UV and γ-rays) diploid yeast cells reveals a high frequency of genomic loss of heterozygosity (LOH) events. . 2012;190: 1267–1284. doi: 10.1534/genetics.111.13792722267500YoshiharaR, NozawaS, HaseY, NarumiI, HidemaJ, SakamotoAN. Mutational effects of γ-rays and carbon ion beams on Arabidopsis seedlings. . 2013;54: 1050–1056. doi: 10.1093/jrr/rrt07423728320BelfieldEJ, GanX, MithaniA, BrownC, JiangC, FranklinK, et al. Genome-wide analysis of mutations in mutant lineages selected following fast-neutron irradiation mutagenesis of Arabidopsis thaliana. . 2012;22: 1306–1315. doi: 10.1101/gr.131474.11122499668LiG, ChernM, JainR, MartinJA, SchackwitzWS, JiangL, et al. Genome-Wide Sequencing of 41 Rice (Oryza sativa L.) Mutated Lines Reveals Diverse Mutations Induced by Fast-Neutron Irradiation. . 2016;9: 1078–1081. doi: 10.1016/j.molp.2016.03.00927018389DattaS, Jankowicz-CieslakJ, NielenS, IngelbrechtI, TillBJ. Induction and recovery of copy number variation in banana through gamma irradiation and low-coverage whole-genome sequencing. . 2018;16: 1644–1653. doi: 10.1111/pbi.1290129476650AdewoyeAB, LindsaySJ, DubrovaYE, HurlesME. The genome-wide effects of ionizing radiation on mutation induction in the mammalian germline. . 2015;6: 6684. doi: 10.1038/ncomms768425809527SatohY, AsakawaJ-I, NishimuraM, KuoT, ShinkaiN, CullingsHM, et al. Characteristics of induced mutations in offspring derived from irradiated mouse spermatogonia and mature oocytes. . 2020;10: 37. doi: 10.1038/s41598-019-56881-231913321BehjatiS, GundemG, WedgeDC, RobertsND, TarpeyPS, CookeSL, et al. Mutational signatures of ionizing radiation in second malignancies. . 2016;7: 12605. doi: 10.1038/ncomms1260527615322SchwarzR, BrulandO, CassoniA, SchombergP, BielackS. The role of radiotherapy in oseosarcoma. . 2009;152: 147–164. doi: 10.1007/978-1-4419-0284-9_720213389MortonLM, KaryadiDM, StewartC, BogdanovaTI, DawsonET, SteinbergMK, et al. Radiation-related genomic profile of papillary thyroid carcinoma after the Chernobyl accident. . 2021;372. doi: 10.1126/science.abg253833888599YoukJ, KwonHW, LimJ, KimE, KimR, ParkS, et al. Mutational impact and signature of ionizing radiation. . 2021. Available: https://www.biorxiv.org/content/10.1101/2021.01.12.426324v1.abstractRose LiY, HalliwillKD, AdamsCJ, IyerV, RivaL, MamunurR, et al. Mutational signatures in tumours induced by high and low energy radiation in Trp53 deficient mice. . 2020;11: 394. doi: 10.1038/s41467-019-14261-431959748WardJF. The complexity of DNA damage: relevance to biological consequences. . 1994;66: 427–432. doi: 10.1080/095530094145514017983426RootsR, OkadaS. Estimation of life times and diffusion distances of radicals involved in x-ray-induced DNA strand breaks of killing of mammalian cells. . 1975;64: 306–320. 1197641SageE, ShikazonoN. Radiation-induced clustered DNA lesions: Repair and mutagenesis. . 2017;107: 125–135. doi: 10.1016/j.freeradbiomed.2016.12.00827939934GoodheadDT. Initial events in the cellular effects of ionizing radiations: clustered damage in DNA. . 1994;65: 7–17. doi: 10.1080/095530094145500217905912VolkovaNV, MeierB, González-HuiciV, BertoliniS, GonzalezS, VöhringerH, et al. Mutational signatures are jointly shaped by DNA damage and repair. . 2020;11: 2169. doi: 10.1038/s41467-020-15912-732358516MeierB, CookeSL, WeissJ, BaillyAP, AlexandrovLB, MarshallJ, et al. C. elegans whole-genome sequencing reveals mutational signatures related to carcinogens and DNA repair deficiency. . 2014;24: 1624–1636. doi: 10.1101/gr.175547.11425030888MeierB, VolkovaNV, HongY, SchofieldP, CampbellPJ, GerstungM, et al. Mutational signatures of DNA mismatch repair deficiency in C. elegans and human cancers. . 2018;28: 666–675. doi: 10.1101/gr.226845.11729636374LiH, DurbinR. Fast and accurate short read alignment with Burrows-Wheeler transform. . 2009;25: 1754–1760. doi: 10.1093/bioinformatics/btp32419451168Nik-ZainalS, AlexandrovLB, WedgeDC, Van LooP, GreenmanCD, RaineK, et al. Mutational processes molding the genomes of 21 breast cancers. . 2012;149: 979–993. doi: 10.1016/j.cell.2012.04.02422608084YeK, SchulzMH, LongQ, ApweilerR, NingZ. Pindel: a pattern growth approach to detect break points of large deletions and medium sized insertions from paired-end short reads. . 2009;25: 2865–2871. doi: 10.1093/bioinformatics/btp39419561018RauschT, ZichnerT, SchlattlA, StützAM, BenesV, KorbelJO. DELLY: structural variant discovery by integrated paired-end and split-read analysis. . 2012;28: i333–i339. doi: 10.1093/bioinformatics/bts37822962449CloggCC, PetkovaE, HaritouA. Statistical Methods for Comparing Regression Coefficients Between Models. . 1995. pp. 1261–1293. doi: 10.1086/230638BenjaminiY, HochbergY. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. . 1995. pp. 289–300. doi: 10.1111/j.2517-6161.1995.tb02031.xAlexandrovLB, KimJ, HaradhvalaNJ, HuangMN, Tian NgAW, WuY, et al. The repertoire of mutational signatures in human cancer. . 2020;578: 94–101. doi: 10.1038/s41586-020-1943-332025018PótiÁ, GyergyákH, NémethE, RuszO, TóthS, KovácsháziC, et al. Correlation of homologous recombination deficiency induced mutational signatures with sensitivity to PARP inhibitors and cytotoxic agents. . 2019. doi: 10.1186/s13059-019-1867-031727117WickyC, AlpiA, PassannanteM, RoseA, GartnerA, MüllerF. Multiple genetic pathways involving the Caenorhabditis elegans Bloom’s syndrome genes him-6, rad-51, and top-3 are needed to maintain genome stability in the germ line. . 2004;24: 5016–5027. doi: 10.1128/MCB.24.11.5016-5027.200415143192MeierB, VolkovaNV, HongY, BertoliniS, González-HuiciV, PetrovaT, et al. Protection of the C. elegans germ cell genome depends on diverse DNA repair pathways during normal proliferation. . 2021;16: e0250291. doi: 10.1371/journal.pone.025029133905417RussellWL, KellyEM. Mutation frequencies in male mice and the estimation of genetic hazards of radiation in men. . 1982;79: 542–544. doi: 10.1073/pnas.79.2.5426952206GartnerA, MilsteinS, AhmedS, HodgkinJ, HengartnerMO. A conserved checkpoint pathway mediates DNA damage—induced apoptosis and cell cycle arrest in C. elegans. . 2000;5: 435–443. doi: 10.1016/s1097-2765(00)80438-410882129González-HuiciV, WangB, GartnerA. A Role for the Nonsense-Mediated mRNA Decay Pathway in Maintaining Genome Stability in Caenorhabditis elegans. . 2017;206: 1853–1864. doi: 10.1534/genetics.117.20341428634159BertoliniS, WangB, MeierB, HongY, GartnerA. . 2. 2013. Available: https://discovery.dundee.ac.uk/ws/portalfiles/portal/19356330/g3.117.1122.full.pdfBaillyAP, FreemanA, HallJ, DéclaisA-C, AlpiA, LilleyDMJ, et al. The Caenorhabditis elegans homolog of Gen1/Yen1 resolvases links DNA damage signaling to DNA double-strand break repair. . 2010;6: e1001025. doi: 10.1371/journal.pgen.100102520661466BaeW, HongS, ParkMS, JeongH-K, LeeM-H, KooH-S. Single-strand annealing mediates the conservative repair of double-strand DNA breaks in homologous recombination-defective germ cells of Caenorhabditis elegans. . 2019;75: 18–28. doi: 10.1016/j.dnarep.2019.01.00730710866RoerinkSF, KooleW, StapelLC, RomeijnRJ, TijstermanM. A broad requirement for TLS polymerases η and κ, and interacting sumoylation and nuclear pore proteins, in lesion bypass during C. elegans embryogenesis. . 2012;8: e1002800. doi: 10.1371/journal.pgen.100280022761594RyuS-H, ParkJ-H, JeongE-S, ChoiS-Y, HamS-H, ParkJ-I, et al. Establishment of a mouse model of 70% lethal dose by total-body irradiation. . 2016;32: 116–121. doi: 10.5625/lar.2016.32.2.11627382380SaekiT, MachidaI, NakaiS. Genetic control of diploid recovery after gamma-irradiation in the yeast Saccharomyces cerevisiae. . 1980;73: 251–265. doi: 10.1016/0027-5107(80)90192-x7007877EdgleyML, BaillieDL, RiddleDL, RoseAM. Genetic balancers. . 1995;48: 147–184. 8531724HillersKJ, JantschV, Martinez-PerezE, YanowitzJL. (May4, 2017), WormBook, ed. The C. elegans Research Community, WormBook, doi: 10.1895/wormbook.1.178.126694509CraigAL, MoserSC, BaillyAP, GartnerA. Methods for studying the DNA damage response in the Caenorhabdatis elegans germ line. . 2012;107: 321–352. doi: 10.1016/B978-0-12-394620-1.00011-422226529StromeS. . WormBook. 2005; 1–10.YuanJ, HorvitzHR. The Caenorhabditis elegans cell death gene ced-4 encodes a novel protein and is expressed during the period of extensive programmed cell death. . 1992;116: 309–320. 1286611DerryWB, PutzkeAP, RothmanJH. Caenorhabditis elegans p53: role in apoptosis, meiosis, and stress resistance. . 2001;294: 591–595. doi: 10.1126/science.106548611557844SchumacherB, HofmannK, BoultonS, GartnerA. The C. elegans homolog of the p53 tumor suppressor is required for DNA damage-induced apoptosis. . 2001;11: 1722–1727. doi: 10.1016/s0960-9822(01)00534-611696333RutkowskiR, HofmannK, GartnerA. Phylogeny and function of the invertebrate p53 superfamily. . 2010;2: a001131. doi: 10.1101/cshperspect.a00113120595397YokooR, ZawadzkiKA, NabeshimaK, DrakeM, ArurS, VilleneuveAM. COSA-1 reveals robust homeostasis and separable licensing and reinforcement steps governing meiotic crossovers. . 2012;149: 75–87. doi: 10.1016/j.cell.2012.01.05222464324LowdenMR, MeierB, LeeTW-S, HallJ, AhmedS. End joining at Caenorhabditis elegans telomeres. . 2008;180: 741–754. doi: 10.1534/genetics.108.08992018780750VermezovicJ, StergiouL, HengartnerMO, d’Adda di FagagnaF. Differential regulation of DNA damage response activation between somatic and germline cells in Caenorhabditis elegans. . 2012;19: 1847–1855. doi: 10.1038/cdd.2012.6922705849PontierDB, TijstermanM. A robust network of double-strand break repair pathways governs genome integrity during C. elegans development. . 2009;19: 1384–1388. doi: 10.1016/j.cub.2009.06.04519646877KampJA, van SchendelR, DilwegIW, TijstermanM. BRCA1-associated structural variations are a consequence of polymerase theta-mediated end-joining. . 2020;11: 3615. doi: 10.1038/s41467-020-17455-332680986BoultonSJ, MartinJS, PolanowskaJ, HillDE, GartnerA, VidalM. BRCA1/BARD1 orthologs required for DNA repair in Caenorhabditis elegans. . 2004;14: 33–39. doi: 10.1016/j.cub.2003.11.02914711411MeierB, VolkovaNV, HongY, BertoliniS, González-HuiciV, PetrovaT, et al. . doi: 10.1101/2020.06.04.133306WyattHDM, SarbajnaS, MatosJ, WestSC. Coordinated actions of SLX1-SLX4 and MUS81-EME1 for Holliday junction resolution in human cells. . 2013;52: 234–247. doi: 10.1016/j.molcel.2013.08.03524076221CastorD, NairN, DéclaisA-C, LachaudC, TothR, MacartneyTJ, et al. Cooperative control of holliday junction resolution and DNA repair by the SLX1 and MUS81-EME1 nucleases. . 2013;52: 221–233. doi: 10.1016/j.molcel.2013.08.03624076219WestSC, BlancoMG, ChanYW, MatosJ, SarbajnaS, WyattHDM. Resolution of Recombination Intermediates: Mechanisms and Regulation. . 2015;80: 103–109. doi: 10.1101/sqb.2015.80.02764926370409DehéP-M, GaillardP-HL. Control of structure-specific endonucleases to maintain genome stability. . 2017;18: 315–330. doi: 10.1038/nrm.2016.17728327556BizardAH, HicksonID. The dissolution of double Holliday junctions. . 2014;6: a016477. doi: 10.1101/cshperspect.a01647724984776CoxMM. The nonmutagenic repair of broken replication forks via recombination. . 2002;510: 107–120. doi: 10.1016/s0027-5107(02)00256-712459447LiX, HeyerW-D. Homologous recombination in DNA repair and DNA damage tolerance. . 2008;18: 99–113. doi: 10.1038/cr.2008.118166982QuinetA, LemaçonD, VindigniA. Replication Fork Reversal: Players and Guardians. . 2017;68: 830–833. doi: 10.1016/j.molcel.2017.11.02229220651KuraokaI, BenderC, RomieuA, CadetJ, WoodRD, LindahlT. Removal of oxygen free-radical-induced 5’,8-purine cyclodeoxynucleosides from DNA by the nucleotide excision-repair pathway in human cells. . 2000. pp. 3832–3837. doi: 10.1073/pnas.07047159710759556BrooksPJ. The cyclopurine deoxynucleosides: DNA repair, biological effects, mechanistic insights, and unanswered questions. . 2017;107: 90–100. doi: 10.1016/j.freeradbiomed.2016.12.02828011151BellonS, RavanatJ-L, GasparuttoD, CadetJ. Cross-Linked Thymine-Purine Base Tandem Lesions: Synthesis, Characterization, and Measurement in γ-Irradiated Isolated DNA. . 2002;15: 598–606. doi: 10.1021/tx015594d11952347BoxHC, BudzinskiEE, DawidzikJB, GobeyJS, FreundHG. Free radical-induced tandem base damage in DNA oligomers. . 1997;23: 1021–1030. doi: 10.1016/s0891-5849(97)00166-49358245BoxHC, BudzinskiEE, DawidzikJB, WallaceJC, IijimaH. Tandem lesions and other products in X-irradiated DNA oligomers. . 1998;149: 433–439. 9588353BoxHC, BudzinskiEE, DawidzikJD, WallaceJC, EvansMS, GobeyJS. Radiation-induced formation of a crosslink between base moieties of deoxyguanosine and thymidine in deoxygenated solutions of d(CpGpTpA). . 1996;145: 641–643. 8619032GuC, WangY. LC-MS/MS identification and yeast polymerase eta bypass of a novel gamma-irradiation-induced intrastrand cross-link lesion G[8–5]C. . 2004;43: 6745–6750. doi: 10.1021/bi049774915157108LiS, LuH, WangZ, HuQ, WangH, XiangR, et al. ERCC1/XPF Is Important for Repair of DNA Double-Strand Breaks Containing Secondary Structures. . 2019;16: 63–78. doi: 10.1016/j.isci.2019.05.01731153042FaridounniaM, FolkersGE, BoelensR. Function and Interactions of ERCC1-XPF in DNA Damage Response. . 2018;23. doi: 10.3390/molecules2312320530563071RahnJJ, AdairGM, NairnRS. Multiple roles of ERCC1-XPF in mammalian interstrand crosslink repair. . 2010;51: 567–581. doi: 10.1002/em.2058320658648KashiyamaK, NakazawaY, PilzDT, GuoC, ShimadaM, SasakiK, et al. Malfunction of nuclease ERCC1-XPF results in diverse clinical manifestations and causes Cockayne syndrome, xeroderma pigmentosum, and Fanconi anemia. . 2013;92: 807–819. doi: 10.1016/j.ajhg.2013.04.0072362338910.1371/journal.pone.0258269.r001Decision Letter 0CotterillSueAcademic Editor2021Sue CotterillThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Submission Version0
21 Jul 2021
PONE-D-21-18200
C. elegans genome-wide analysis reveals DNA repair pathways that act cooperatively to preserve genome integrity upon ionizing radiation
PLOS ONE
Dear Dr. Gartner,
Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but requires some adjustments to fully meet PLOS ONE’s publication criteria. These are discussed in detail in the 2 sets of reviewers comments attached with this letter.Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised.
Please submit your revised manuscript by Sep 04 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.
Please include the following items when submitting your revised manuscript:
A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.
If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.
If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.
We look forward to receiving your revised manuscript.
Kind regards,
Sue Cotterill
Academic Editor
PLOS ONE
Journal Requirements:
When submitting your revision, we need you to address these additional requirements.
1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at
https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and
2. We noted in your submission details that a portion of your manuscript may have been presented or published elsewhere. [We note that our current manuscript uses the same primary sequence source data we deposited as part of our Nat Commun. 2020 May 1;11(1):2169. doi: 10.1038/s41467-020-15912-7. We clearly indicate this in the abstract, the results section, and the discussion. In the Nature Communication paper, we provided a high-level overview (and comparison) of the mutagenic consequences of exposure to 11 genotoxic agents, one agent being ionizing radiation (IR).] Please clarify whether this publication was peer-reviewed and formally published. If this work was previously peer-reviewed and published, in the cover letter please provide the reason that this work does not constitute dual publication and should be included in the current manuscript.
3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information
Additional Editor Comments:
Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.
[Note: HTML markup is below. Please do not edit.]
Reviewers' comments:
Reviewer's Responses to Questions
Comments to the Author
1. Is the manuscript technically sound, and do the data support the conclusions?
The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.
Reviewer #1: Yes
Reviewer #2: Yes
**********
2. Has the statistical analysis been performed appropriately and rigorously?
Reviewer #1: Yes
Reviewer #2: Yes
**********
3. Have the authors made all data underlying the findings in their manuscript fully available?
The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.
Reviewer #1: Yes
Reviewer #2: Yes
**********
4. Is the manuscript presented in an intelligible fashion and written in standard English?
PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.
Reviewer #1: Yes
Reviewer #2: Yes
**********
5. Review Comments to the Author
Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)
Reviewer #1: This study provides interesting observations about the effect of ionizing radiation (IR) on mutational signatures, with respect to different DNA repair and damage response defects, in a model organism C. elegans. Although the sequencing data and some general observations have been already published by the authors as part of a broader study (in Nature Communications), in this study, the authors focus only on the IR-treated samples and more in-depth analysis of the mutational signatures in these samples. The results support a role of GG-NER pathway (XPF-1, XPA-1, and XPC-1) in error-free repair of IR-induced damage and the role of TLS (POLH-1 and REV-1) in error-prone repair of IR-induced damage.
Comments
1. The dose-dependent increase should have statistics performed and reported (such as the correlation coefficient and p-value).
2. “Structural variants (SVs) were augmented with dose”: Was this statistically significant? (The numbers look very low, so maybe the sentence in abstract should be adjusted accordingly.)
3. In the cluster analysis, did you observe any dose effect?
4. “In C. elegans, ~6 % of all IR-induced mutations are clustered, an effect generally not observed in unchallenged C. elegans strains propagated over generations”: Can you make a direct comparison? What proportion of mutations are clustered in unchallenged strains?
5. The link of the github repository refers to a different project.
6. Could you please share the VCF files with the readers? This will simplify reproducibility and potential future use of the data set in the community. It those files are already available in your previous publication, could you please make a simple list of how to find the VCF files relevant for this publication?
7. Would it be possible to include in the github repository everything needed to easily reproduce (at least the main) manuscript figures? (E.g., it seems the scripts expect various input files and parameterisation files, but I did not find those files in the repository.)
8. The methods should be ideally summarized in the methods/supplementary text of this paper in order to enable reproducibility without the need to look for many details and parameterization described in the previous paper. In particular, it would be good if the authors could describe in more detail how the variant filtering against a panel of unrelated samples was performed, as well as the cut-offs on coverage, read support of the variant in the test and control samples, and overlap with other genomic variants.
9. Could you please (in the discussion) briefly comment of similarities and differences of your results to those of mutational signatures of ionizing radiation observed in the following publications:
o (Li et al., 2020) https://www.nature.com/articles/s41467-019-14261-4
o (Kageyama et al., 2020) https://www.thegreenjournal.com/article/S0167-8140(20)30856-2/fulltext
o (Youk et al., 2021) https://www.biorxiv.org/content/10.1101/2021.01.12.426324v1.full
Minor comments
1. Top of the page 10: the formula has some missing characters.
2. Page 13: incorrect font.
3. Page 14: a typo in “A Indeles”.
4. Page 14: “Finally the proportion of indels was increased in rev-1 and pol-h translesion synthesis defective strains”: Why proportion? (The figure seems to show mutation counts.)
Reviewer #2: In this study, authors irradiate the germ cells of C.elegans in wild-type and DNA repair or damage response mutant animals and perform whole genome sequencing on the progeny which have undergone clonal expansion of mutations from zygotes. The study is methodologically sound and the analysis that was performed to describe the nature of these mutations generated by ionizing radiation (IR) is thorough. This is certainly an important and comprehensive study and the data is a valuable resource to the genomics community. While overall it is surprising that the radiation does not have a more substantial effect on mutagenesis in C.elegans, some interesting observations were made. The data suggests that global NER is important for repair of IR induced damage and that error prone bypass of IR lesions by translesion synthesis polymerases cause SNVs. A specific dinucleotide substitution pattern of mutations from irradiation was also identified. Minor comments are made below.
Minor
1. As the human genome is ~30x larger than the C.elegans genome, 3 replicates per parent might not result in enough mutations to make conclusions for lower frequency mutations such as SVs. This possibility should be added to the discussion.
2. Since irradiation affects survival of progeny, can you discuss the possibility that the model system selects for progeny that accumulated less mutations? In the discussion on page 20, authors say this is not the case as apoptosis compromised ced-4 mutants do not have increased mutation burden. Can you preclude other possibilities such as necrosis or loss of capacity for self-fertilization?
3. Figure 1C – Since COSMIC signature 3 and SBS40 are non-distinct signatures, the cosine similarity may appear similar to that of the humanized C.elegans irradiation signature without a genuine biological relationship between the signatures. You should compare the cosine similarity of simulated mutations to signature 3 and 40 and see if the value is lower.
4. Figure 2A – Add p-value and R-squared for line of best fit linear regression. It does not look like irradiation significantly induces indels and SNVs.
5. Figure 2B – The figure legend describes values as fold change but the y-axis is number of mutations. Can you add ‘wild-type level’ on the dotted line (like in figures 5A and 5B) to make the comparison to wild-type clearer.
6. Figure 5 – Can you perform simulations or bootstrapping analysis to see what proportion of mutations would fall in clusters by chance such as randomly shuffling mutations in trinucleotide preserved manner? Or simply, do non-irradiated animals also display mutation clusters?
7. Figure 5 – Where in the genome do mutation clusters lie? For example, are they associated with heterochromatin or other epigenetic marks? Were there any genomic sites of mutation clustering that occurred in more than 1 sample tested.
**********
6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.
If you choose “no”, your identity will remain anonymous but your review may still be made public.
Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.
Reviewer #1: No
Reviewer #2: Yes: Jayne Barbour
[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]
While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.
10.1371/journal.pone.0258269.r002Author response to Decision Letter 0Submission Version1
1 Sep 2021
(same letter as ‘coverletter’)
Dear Mrs Cotteril,
We herewith submit our revised manuscript ‘PONE-D-21-18200 ‘C. elegans genome-wide analysis reveals DNA repair pathways that act cooperatively to preserve genome integrity upon ionizing radiation’.
We are grateful for the support and the helpful comments of the two reviewers. We addressed all of their comments; thereby improving our manuscript. Below, A) please find point to point response to the reviewers. Further below, B) as requested (Journal requirement 2) we provide a detailed case for our manuscript not constituting dual publication; it being distinct from our previous publication.
All submitted files are in the correct format, and captions for Supporting Information files are provided.
We are confident that the manuscript is now ready for publication.
best regards,
Anton Gartner and Moritz Gerstung
A) Response to reviewers:
Reviewer #1: This study provides interesting observations about the effect of ionizing radiation (IR) on mutational signatures, with respect to different DNA repair and damage response defects, in a model organism C. elegans. Although the sequencing data and some general observations have been already published by the authors as part of a broader study (in Nature Communications), in this study, the authors focus only on the IR-treated samples and more in-depth analysis of the mutational signatures in these samples. The results support a role of GG-NER pathway (XPF-1, XPA-1, and XPC-1) in error-free repair of IR-induced damage and the role of TLS (POLH-1 and REV-1) in error-prone repair of IR-induced damage.
Comments
1. The dose-dependent increase should have statistics performed and reported (such as the correlation coefficient and p-value).
Answer: Following this request, we have added the correlation coefficient and p-value to Figure 2A to report the statistics behind the dose-dependent effect in wild-type. The mutational effects of ionising radiation in other genotypes (Figure 2B) are reported with their average values, CIs and p-values all in comparison to the wild-type effect.
2. “Structural variants (SVs) were augmented with dose”: Was this statistically significant? (The numbers look very low, so maybe the sentence in the abstract should be adjusted accordingly.)
Answer: As reported now in Figure 2A, the correlation between the number of SVs and irradiation dose was weaker than for SNVs, but nevertheless statistically significant.
3. In the cluster analysis, did you observe any dose effect?
Answer: We did observe a weak dose-dependent effect in the wild-type strains when comparing the number of clusters per sample across different doses, but due to the overall low number of clusters this link could only be reliably established in the wild-type (as it had the highest amount of samples tested). We have added a relevant Figure to Sup Figure S7.
4. “In C. elegans, ~6 % of all IR-induced mutations are clustered, an effect generally not observed in unchallenged C. elegans strains propagated over generations”: Can you make a direct comparison? What proportion of mutations are clustered in unchallenged strains?
Answer: As we reported in a publication studying mutation accumulation in DNA repair deficient strains (https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0250291), we did not observe any clustering in wild-type lines even after propagating nematodes for 40 generations. Hence we did not expect any clustering in the untreated strains after just one generation. Also mutation numbers in control samples not treated with ionizing irradiation are too low to assess clustering. We have amended a relevant statement in the text and added clustering information across different genetic backgrounds to Supplementary Data (Supplementary Figure S7 (top left panel), Supplementary Table S2).
5. The link of the github repository refers to a different project.
Answer: Thank you, we have amended the link path.
6. Could you please share the VCF files with the readers? This will simplify reproducibility and potential future use of the data set in the community. If those files are already available in your previous publication, could you please make a simple list of how to find the VCF files relevant for this publication?
Answer: Following this comment, we have added a mutation counts table to the Supplementary File 1, and explicitly stated the link to the relevant VCF collection uploaded as part of our previous publication in the Supplementary Information section.
7. Would it be possible to include in the github repository everything needed to easily reproduce (at least the main) manuscript figures? (E.g., it seems the scripts expect various input files and parameterisation files, but I did not find those files in the repository.)
Answer: We have streamlined the codes appearing on the Github page and added more instructions on data preparation.The scripts in the Github repository expects the user to have downloaded the above mentioned VCF data as well as the supplementary table of this publication.
8. The methods should be ideally summarized in the methods/supplementary text of this paper in order to enable reproducibility without the need to look for many details and parameterization described in the previous paper. In particular, it would be good if the authors could describe in more detail how the variant filtering against a panel of unrelated samples was performed, as well as the cut-offs on coverage, read support of the variant in the test and control samples, and overlap with other genomic variants.
Answer: As mentioned in the Methods section, all the variant calling and filtering for this manuscript was performed in accordance with the procedure used in our previous publication (https://www.nature.com/articles/s41467-020-15912-7 ), Nevertheless, we have now added the relevant details to the Supplementary Information section of this manuscript (clearly referencing our previous paper).
9. Could you please (in the discussion) briefly comment of similarities and differences of your results to those of mutational signatures of ionizing radiation observed in the following publications:
o (Li et al., 2020) https://www.nature.com/articles/s41467-019-14261-4
o (Kageyama et al., 2020) https://www.thegreenjournal.com/article/S0167-8140(20)30856-2/fulltext
o (Youk et al., 2021) https://www.biorxiv.org/content/10.1101/2021.01.12.426324v1.full
These new papers and preprints as well as a recent paper on thyroid carcinomas induced by the Chernobil disaster are now cited in the introduction by adding the underlined sentences below. Also in the discussion (text further below), we add a sentence highlighting that IR induced SV (in contrast to SV, and indels) could not be detected in some recent studies.
More recent studies, analysing thyroid carcinomas, linked to Iodine-131 exposure related to the Chernobyl disaster failed to detect increased SNV levels, indels and SVs being increased [19]. Human organoid exposure to IR similarly only led to increased indels and SVs [20]. Radiation induced tumours induced in p53 mutant mice SVs levels were reported to be rare, indels and SV being detectable [21] .
‘In summary, while these analyses provide insights into mutation rates and mutation types, their comparison is complicated by the types of analysis applied, differences in effects of chosen radiation sources, DNA repair status, tissue type, and possible organismal differences. Also, these studies do not provide insight into the relative contribution of various DNA repair pathways counteracting IR induced mutagenesis.’
added to the discussion:
‘The highest proportion of mutants being SVs is in line with the analysis of IR induced secondary tumours [17]. In contrast, increased SVs could not be detected in thyroid carcinomas, linked to the Chernobyl disaster [19] and irradiated human organoids [20]. ‘
Minor comments
1. Top of the page 10: the formula has some missing characters.
2. Page 13: incorrect font.
thanks, fixed
3. Page 14: a typo in “A Indeles”.
thanks, fixed
4. Page 14: “Finally the proportion of indels was increased in rev-1 and pol-h translesion synthesis defective strains”: Why proportion? (The figure seems to show mutation counts.)
thanks, we now say that ‘the number of indels’ was increased...
Reviewer #2: In this study, authors irradiate the germ cells of C.elegans in wild-type and DNA repair or damage response mutant animals and perform whole genome sequencing on the progeny which have undergone clonal expansion of mutations from zygotes. The study is methodologically sound and the analysis that was performed to describe the nature of these mutations generated by ionizing radiation (IR) is thorough. This is certainly an important and comprehensive study and the data is a valuable resource to the genomics community. While overall it is surprising that the radiation does not have a more substantial effect on mutagenesis in C.elegans, some interesting observations were made. The data suggests that global NER is important for repair of IR induced damage and that error prone bypass of IR lesions by translesion synthesis polymerases cause SNVs. A specific dinucleotide substitution pattern of mutations from irradiation was also identified. Minor comments are made below.
Minor
1. As the human genome is ~30x larger than the C.elegans genome, 3 replicates per parent might not result in enough mutations to make conclusions for lower frequency mutations such as SVs. This possibility should be added to the discussion.
Thanks for pointing this out, we now added two sentences (these are underlined below) acknowledging that our study is limited by us only analysing three replicates of each sample. The relevant section of the amended discussion is pasted below:
‘In line with expectations of IR induced mutation rate estimates in mouse and C. elegans germ cells [8,9,36] and from sequencing the progeny of irradiated mice, budding yeast and several plant species (for instance [10–16]), we confirmed that IR-induced mutagenesis is surprisingly low, with 80 Gy inducing an average of 36.6 SNVs, 1.3 DNVs, 4 indels and 1.4 (SD=0.13) SVs in the 100 million base C. elegans genome. We typically did experiments in triplicates, assessing samples without, and with an intermediate and high dose of IR. Clearly, increased statistical power would require the analysis of a larger number of samples. It is impossible to significantly increase the dose of IR as this would lead to 100% embryonic lethality. Broadly, 80 Gy leads to a 30 to 50% reduction of survival of irradiated progeny, with up to 10-fold lower survival in C. elegans DNA repair mutants [37–42].’
2. Since irradiation affects survival of progeny, can you discuss the possibility that the model system selects for progeny that accumulated less mutations? In the discussion on page 20, authors say this is not the case as apoptosis compromised ced-4 mutants do not have increased mutation burden. Can you preclude other possibilities such as necrosis or loss of capacity for self-fertilization?
Answer: We now added a sentence (underlined) and slightly changed our wording (inderlines) on pape 19, second last paragraph. We now state much clearer that we might underestimate mutation rates; us, in our experimental procedure, not sequencing gDNA from single embryos.
‘On the other hand, while a large number of C. elegans fusion chromosomes are known to be viable in a heterozygous state and are being used as balancer chromosomes [48], aneuploidy of autosomes, typically caused by meiotic segregation defects, leads to embryonic lethality [49]. It is thus possible that we fail to detect a subset of some chromosomal fusions or the loss of large chromosomal fragments resulting from unrepaired DSBs, if these events lead to a near aneuploid state and consequently to embryonic or larval lethality. In other words, we might fail to detect highly mutagenised genomes because affected embryos die, and do not sire progeny needed for whole genome sequence analysis. Detection of mutational events leading to embryonic lethality would require single embryo or single zygote whole genome sequencing, studies we will pursue in the future.’
Figure 1C – Since COSMIC signature 3 and SBS40 are non-distinct signatures, the cosine similarity may appear similar to that of the humanized C.elegans irradiation signature without a genuine biological relationship between the signatures. You should compare the cosine similarity of simulated mutations to signature 3 and 40 and see if the value is lower.
Answer: We appreciate that flat mutational signatures without distinctive features would show high similarity to each other. We decided to draw value from the fact that the base substitution signature we extracted from IR-treated C. elegans lines also did not exhibit strong imbalances in mutation preference, making it similar to those flat signatures often observed in cancers. It is not clear to us what ‘simulated mutations’ refers to. We think that the reviewer meant comparing the IR signature we observe to a random mutation distribution. We note, as elaborated in our 2018 (Meier at al., ) paper, that trinucleotide prevalences between humans and C.elegans genomes vary, thus a random distribution of mutations would be rather different between the two species.
4. Figure 2A – Add p-value and R-squared for line of best fit linear regression. It does not look like irradiation significantly induces indels and SNVs.
Answer: We have added the Pearson correlation coefficients and their corresponding p-value to Figure 2A.
5. Figure 2B – The figure legend describes values as fold change but the y-axis is number of mutations. Can you add ‘wild-type level’ on the dotted line (like in figures 5A and 5B) to make the comparison to wild-type clearer.
Thanks, we added this to Figure 2B and amended the legend.
6. Figure 5 – Can you perform simulations or bootstrapping analysis to see what proportion of mutations would fall in clusters by chance such as randomly shuffling mutations in trinucleotide preserved manner? Or simply, do non-irradiated animals also display mutation clusters?
As outlined in the response to the other reviewer (4) we previously reported in a publication studying mutation accumulation in DNA repair deficient strains (https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0250291) that we did not observe any clustering in the wild-type lines even after propagating lines for 40 generations. Hence we did not expect any clustering in the untreated strains after just one generation. We have amended a relevant statement in the text and added clustering information across different genetic backgrounds to Supplementary Data.
7. Figure 5 – Where in the genome do mutation clusters lie? For example, are they associated with heterochromatin or other epigenetic marks? Were there any genomic sites of mutation clustering that occurred in more than 1 sample tested.
The number of mutational clusters observed in the experiments was too low to try to make an inference about the genomic location of the clusters. We did not observe recurrent mutation clusters (we have added another Supplementary Data object describing mutations in clusters across different genomic backgrounds which demonstrates this (also see response 4 to previous reviewer)).
B) Reasons for our work not constituting dual publication:
We note that our current manuscript uses the same primary sequence source data we deposited as part of our Nat Commun. 2020 May 1;11(1):2169. doi: 10.1038/s41467-020-15912-7. We clearly indicate this in the abstract, the results section, and the discussion. In the Nature Communication paper, we provided a high-level overview (and comparison) of the mutagenic consequences of exposure to 11 genotoxic agents, one agent being ionizing radiation (IR).
Given the immense importance of understanding IR damage, we here provide a much more detailed and comprehensive analysis of the mutagenic consequences of IR as part of this manuscript. Such detailed, in-depth analysis of IR-induced mutagenesis was not possible in our Nat Communication paper and is a valuable resource for researchers studying IR response and mutagenesis. Our five main figures, the three supplementary files and the 8 supplementary figures are all new and are all based on new analysis. None of the figures of this manuscript was shown in our Nature Communications publication. Introductory Figure 1 provides a refined analysis of the IR-induced mutational signature in DNA repair proficient C. elegans and a comparison to radiation-associated signatures observed in human cancer. Importantly, absolute mutation numbers and the variation between IR exposure experiments are clearly visualized. Figure 2 provides a refined analysis of DNA repair mutants that showed increased numbers of single nucleotide variants (SNVs), multi nucleotide variants (MNVs), indels and structural variants (SVs) compared to wild-type. The analysis is new and (in contrast to our Nature Communications paper) only considers IR exposed samples. Figure 3 provides a refined analysis of the extent of change in mutational signatures in various DNA repair mutants. Importantly, Figure 4 provides a new, comprehensive analysis of dinucleotide variants induced by radiation, compared to those induced by UV and cisplatin exposure. Figure 5, summarizes for the first time the extent of mutational clustering upon IR exposure, which is surprisingly low and not overtly affected by the deficiency of various DNA repair pathways analysed in our studies. Suppl Files 1-3 show a comprehensive analysis of the genomic location of mutations, mutational profiles and dose response in wild-type and DNA repair mutants. Suppl Fig S1 compares indel frequencies in our experimental system and IR induced cancers, Suppl Fig S2-6 provide a comprehensive overview of absolute numbers and types of mutation resulting from IR exposure, stratified by DNA repair pathway. Suppl Fig S7 and 8 provide supporting data of our analysis on mutational clustering (Figure 5).
Importantly, in this manuscript we also discuss our data in detail, providing a comprehensive review of previous literature and also providing a detailed and comprehensive discussion of our data and its relevance to the field. The in-depth description of IR induced mutagenesis will provide a basis for a more targeted design of analogous studies in mammalian cells.
Finally, we point out that both reviewers strongly supported our work, and both second that our manuscript is novel and distinct from our previous publication.
Submitted filename: Response to Reviewers.Meier-Volkova. revision.docx
10.1371/journal.pone.0258269.r003Decision Letter 1CotterillSueAcademic Editor2021Sue CotterillThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Submission Version1
23 Sep 2021
C. elegans genome-wide analysis reveals DNA repair pathways that act cooperatively to preserve genome integrity upon ionizing radiation
PONE-D-21-18200R1
Dear Dr. Gartner,
We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.
Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.
An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.
If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.
Kind regards,
Sue Cotterill
Academic Editor
PLOS ONE
Additional Editor Comments (optional):
Reviewers' comments:
10.1371/journal.pone.0258269.r004Acceptance letterCotterillSueAcademic Editor2021Sue CotterillThis is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
28 Sep 2021
PONE-D-21-18200R1
C. elegans genome-wide analysis reveals DNA repair pathways that act cooperatively to preserve genome integrity upon ionizing radiation
Dear Dr. Gartner:
I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.
If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.
If we can help with anything else, please email us at plosone@plos.org.
Thank you for submitting your work to PLOS ONE and supporting open access.