DNA Sequence Context Research Articles

Towards the genomic sequence code of DNA fragility for machine learning.

Genomic DNA breakages and the subsequent insertion and deletion mutations are important contributors to genome instability and linked diseases. Unlike the research in point mutations, the relationship between DNA sequence context and the propensity for strand breaks remains elusive. Here, by analyzing the differences and commonalities across myriads of genomic breakage datasets, we extract the sequence-linked rules and patterns behind DNA fragility. We show the overall deconvolution of the sequence influence into short-, mid- and long-range effects, and the stressor-dependent differences in defining the range and compositional effects on DNA fragility. We summarize and release our feature compendium as a library that can be seamlessly incorporated into genomic machine learning procedures, where DNA fragility is of concern, and train a generalized DNA fragility model on cancer-associated breakages. Structural variants (SVs) tend to stabilize regions in which they emerge, with the effect most pronounced for pathogenic SVs. In contrast, the effects of chromothripsis are seen across regions less prone to breakages. We find that viral integration may bring genome fragility, particularly for cancer-associated viruses. Overall, this work offers novel insights into the genomic sequence basis of DNA fragility and presents a powerful machine learning resource to further enhance our understanding of genome (in)stability and evolution.

Variation in the Spectrum of New Mutations among Inbred Strains of Mice.

The mouse serves as a mammalian model for understanding the nature of variation from new mutations, a question that has both evolutionary and medical significance. Previous studies suggest that the rate of single-nucleotide mutations (SNMs) in mice is ∼50% of that in humans. However, information largely comes from studies involving the C57BL/6 strain, and there is little information from other mouse strains. Here, we study the mutations that accumulated in 59 mouse lines derived from four inbred strains that are commonly used in genetics and clinical research (BALB/cAnNRj, C57BL/6JRj, C3H/HeNRj, and FVB/NRj), maintained for eight to nine generations by brother-sister mating. By analyzing Illumina whole-genome sequencing data, we estimate that the average rate of new SNMs in mice is ∼μ = 6.7 × 10-9. However, there is substantial variation in the spectrum of SNMs among strains, so the burden from new mutations also varies among strains. For example, the FVB strain has a spectrum that is markedly skewed toward C→A transversions and is likely to experience a higher deleterious load than other strains, due to an increased frequency of nonsense mutations in glutamic acid codons. Finally, we observe substantial variation in the rate of new SNMs among DNA sequence contexts, CpG sites, and their adjacent nucleotides playing an important role.

DNA sequence and chromatin differentiate sequence-specific transcription factor binding in the human malaria parasite Plasmodium falciparum.

Development of the malaria parasite, Plasmodium falciparum, is regulated by a limited number of sequence-specific transcription factors (TFs). However, the mechanisms by which these TFs recognize genome-wide binding sites is largely unknown. To address TF specificity, we investigated the binding of two TF subsets that either bind CACACA or GTGCAC DNA sequence motifs and further characterized two additional ApiAP2 TFs, PfAP2-G and PfAP2-EXP, which bind unique DNA motifs (GTAC and TGCATGCA). We also interrogated the impact of DNA sequence and chromatin context on P. falciparum TF binding by integrating high-throughput in vitro and in vivo binding assays, DNA shape predictions, epigenetic post-translational modifications, and chromatin accessibility. We found that DNA sequence context minimally impacts binding site selection for paralogous CACACA-binding TFs, while chromatin accessibility, epigenetic patterns, co-factor recruitment, and dimerization correlate with differential binding. In contrast, GTGCAC-binding TFs prefer different DNA sequence context in addition to chromatin dynamics. Finally, we determined that TFs that preferentially bind divergent DNA motifs may bind overlapping genomic regions due to low-affinity binding to other sequence motifs. Our results demonstrate that TF binding site selection relies on a combination of DNA sequence and chromatin features, thereby contributing to the complexity of P. falciparum gene regulatory mechanisms.

G-Quadruplex Forming DNA Sequence Context Is Enriched around Points of Somatic Mutations in a Subset of Multiple Myeloma Patients.

Multiple myeloma (MM) is the second most common hematological malignancy, which remains incurable despite recent advances in treatment strategies. Like other forms of cancer, MM is characterized by genomic instability, caused by defects in DNA repair. Along with mutations in DNA repair genes and genotoxic drugs used to treat MM, non-canonical secondary DNA structures (four-stranded G-quadruplex structures) can affect accumulation of somatic mutations and chromosomal abnormalities in the tumor cells of MM patients. Here, we tested the hypothesis that G-quadruplex structures may influence the distribution of somatic mutations in the tumor cells of MM patients. We sequenced exomes of normal and tumor cells of 11 MM patients and analyzed the data for the presence of G4 context around points of somatic mutations. To identify molecular mechanisms that could affect mutational profile of tumors, we also analyzed mutational signatures in tumor cells as well as germline mutations for the presence of specific SNPs in DNA repair genes or in genes regulating G-quadruplex unwinding. In several patients, we found that sites of somatic mutations are frequently located in regions with G4 context. This pattern correlated with specific germline variants found in these patients. We discuss the possible implications of these variants for mutation accumulation and specificity in MM and propose that the extent of G4 context enrichment around somatic mutation sites may be a novel metric characterizing mutational processes in tumors.

Development and testing of a versatile genome editing application reporter (V-GEAR) system

CRISPR/Cas9 and novel cas fusion proteins leveraging specific DNA targeting ability combined with deaminases or reverse transcriptases has revolutionized genome editing. However, their efficacy heavily relies upon protein variants, targeting sgRNAs, and surrounding DNA sequence context within the targeted loci. This necessitates the need for efficient and rapid screening methods to evaluate these editing reagents and designs. Existing plasmid-based reporters lack flexibility, being fixed to specific DNA sequences, hindering direct comparisons between various editing approaches. To address this, we developed the versatile genome editing application reporter (V-GEAR) system. V-GEAR comprises genes detectable after desired editing via base editing, prime editing, or homology-directed repair within relevant genomic contexts. It employs a detectable synthetic cell surface protein (RQR8) followed by a customizable target sequence resembling genomic regions of interest. These genes allow for reliable identification of corrective editing and cell enrichment. We validated the V-GEAR system with base editors, prime editors, and Cas9 mediated homology-directed repair. Furthermore, the V-GEAR system offers versatility by allowing transient screening or stable integration at the AAVS1 safe harbor loci, rapidly achieved through immunomagnetic isolation. This innovative system enables direct comparisons among editing technologies, accelerating the development and testing of genome editing approaches.

Abstract IA021: Novel genomic features of POLE-mutant tumors: the utility for tumor classification and identification of new driver alleles

Abstract Alterations in the exonuclease domain of POLE drive the development of ultramutated endometrial tumors with better prognosis. A hallmark of these tumors is the accumulation of G>T mutations in AGA sequences; however, understanding of their genomic features beyond the trinucleotide motifs is limited. We applied a novel computational framework to the whole-exome sequencing data of 524 endometrial tumors reported by The Cancer Genome Atlas network to analyze the extended DNA sequence context of ultramutation. We found that the presence of POLE driver alleles is associated with an increased frequency of G>T mutations in polypurine tracts. Sequences containing three consecutive purines, including the classic AGA context, are only moderately mutable, but the mutability increased abruptly as the tract length reached six or more purines. Moreover, mutations showed a strong preference for certain positions within the tracts, creating a characteristic quantifiable pattern. Using this signature, we developed a machine learning classifier to identify tumors with hitherto unknown POLE drivers and validated two new drivers, POLE-E978G and POLE-S461L, by functional assays in yeast. Unlike previously known pathogenic variants, the E978G substitution affects the DNA polymerase rather than exonuclease domain of POLE. We further show that the extended genomic signature of POLE ultramutation is shared by tumors with POLD1 drivers. These findings uncover new mechanisms of ultramutation and highlight the limitations of the current focus on POLE exonuclease domain variants when stratifying patients. Citation Format: Daria Ostroverkhova, Kathryn Tyryshkin, Annette K. Beach, Elizabeth A. Moore, Yosef Masoudi-Sobhanzadeh, Stephanie R. Barbari, Igor B Rogozin, Konstantin V Shaitan, Anna R Panchenko, Polina V. Shcherbakova. Novel genomic features of POLE-mutant tumors: the utility for tumor classification and identification of new driver alleles [abstract]. In: Proceedings of the AACR Special Conference on Endometrial Cancer: Transforming Care through Science; 2023 Nov 16-18; Boston, Massachusetts. Philadelphia (PA): AACR; Clin Cancer Res 2024;30(5_Suppl):Abstract nr IA021.

Hydrogen bonding in duplex DNA probed by DNP enhanced solid-state NMR N-H bond length measurements.

Numerous biological processes and mechanisms depend on details of base pairing and hydrogen bonding in DNA. Hydrogen bonds are challenging to quantify by X-ray crystallography and cryo-EM due to difficulty of visualizing hydrogen atom locations but can be probed with site specificity by NMR spectroscopy in solution and the solid state with the latter particularly suited to large, slowly tumbling DNA complexes. Recently, we showed that low-temperature dynamic nuclear polarization (DNP) enhanced solid-state NMR is a valuable tool for distinguishing Hoogsteen base pairs (bps) from canonical Watson-Crick bps in various DNA systems under native-like conditions. Here, using a model 12-mer DNA duplex containing two central adenine-thymine (A-T) bps in either Watson-Crick or Hoogsteen confirmation, we demonstrate DNP solid-state NMR measurements of thymine N3-H3 bond lengths, which are sensitive to details of N-H···N hydrogen bonding and permit hydrogen bonds for the two bp conformers to be systematically compared within the same DNA sequence context. For this DNA duplex, effectively identical TN3-H3 bond lengths of 1.055 ± 0.011Å and 1.060 ± 0.011Å were found for Watson-Crick A-T and Hoogsteen A (syn)-T base pairs, respectively, relative to a reference amide bond length of 1.015 ± 0.010Å determined for N-acetyl-valine under comparable experimental conditions. Considering that prior quantum chemical calculations which account for zero-point motions predict a somewhat longer effective peptide N-H bond length of 1.041Å, in agreement with solution and solid-state NMR studies of peptides and proteins at ambient temperature, to facilitate direct comparisons with these earlier studies TN3-H3 bond lengths for the DNA samples can be readily scaled appropriately to yield 1.083Å and 1.087Å for Watson-Crick A-T and Hoogsteen A (syn)-T bps, respectively, relative to the 1.041Å reference peptide N-H bond length. Remarkably, in the context of the model DNA duplex, these results indicate that there are no significant differences in N-H···N A-T hydrogen bonds between Watson-Crick and Hoogsteen bp conformers. More generally, high precision measurements of N-H bond lengths by low-temperature DNP solid-state NMR based methods are expected to facilitate detailed comparative analysis of hydrogen bonding for a range of DNA complexes and base pairing environments.

Interpretable predictive models of genome-wide aryl hydrocarbon receptor-DNA binding reveal tissue-specific binding determinants.

The aryl hydrocarbon receptor (AhR) is an inducible transcription factor whose ligands include the potent environmental contaminant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Ligand-activated AhR binds to DNA at dioxin response elements (DREs) containing the core motif 5'-GCGTG-3'. However, AhR binding is highly tissue specific. Most DREs in accessible chromatin are not bound by TCDD-activated AhR, and DREs accessible in multiple tissues can be bound in some and unbound in others. As such, AhR functions similarly to many nuclear receptors. Given that AhR possesses a strong core motif, it is suited for a motif-centered analysis of its binding. We developed interpretable machine learning models predicting the AhR binding status of DREs in MCF-7, GM17212, and HepG2 cells, as well as primary human hepatocytes. Cross-tissue models predicting transcription factor (TF)-DNA binding generally perform poorly. However, reasons for the low performance remain unexplored. By interpreting the results of individual within-tissue models and by examining the features leading to low cross-tissue performance, we identified sequence and chromatin context patterns correlated with AhR binding. We conclude that AhR binding is driven by a complex interplay of tissue-agnostic DRE flanking DNA sequence and tissue-specific local chromatin context. Additionally, we demonstrate that interpretable machine learning models can provide novel and experimentally testable mechanistic insights into DNA binding by inducible TFs.

Genome-wide impact of cytosine methylation and DNA sequence context on UV-induced CPD formation.

Exposure to ultraviolet (UV) light is the primary etiological agent for skin cancers because UV damages cellular DNA. The most frequent form of UV damage is the cyclobutane pyrimidine dimer (CPD), which consists of covalent linkages between neighboring pyrimidine bases in DNA. In human cells, the 5' position of cytosine bases in CG dinucleotides is frequently methylated, and methylated cytosines in the TP53 tumor suppressor are often sites of mutation hotspots in skin cancers. It has been argued that this is because cytosine methylation promotes UV-induced CPD formation; however, the effects of cytosine methylation on CPD formation are controversial, with conflicting results from previous studies. Here, we use a genome-wide method known as CPD-seq to map UVB- and UVC-induced CPDs across the yeast genome in the presence or absence in vitro methylation by the CpG methyltransferase M.SssI. Our data indicate that cytosine methylation increases UVB-induced CPD formation nearly 2-fold relative to unmethylated DNA, but the magnitude of induction depends on the flanking sequence context. Sequence contexts with a 5' guanine base (e.g., GCCG and GTCG) show the strongest induction due to cytosine methylation, potentially because these sequence contexts are less efficient at forming CPD lesions in the absence of methylation. We show that cytosine methylation also modulates UVC-induced CPD formation, albeit to a lesser extent than UVB. These findings can potentially reconcile previous studies, and define the impact of cytosine methylation on UV damage across a eukaryotic genome.

Human DNA ligases I and IIIα as determinants of accuracy and efficiency of base excision DNA repair

Mammalian Base Excision Repair (BER) DNA ligases I and IIIα (LigI, LigIIIα) are major determinants of DNA repair fidelity, alongside with DNA polymerases. Here we compared activities of human LigI and LigIIIα on specific and nonspecific substrates representing intermediates of distinct BER sub-pathways. The enzymes differently discriminate mismatches in the nicked DNA, depending on their identity and position, but are both more selective against the 3′-end non-complementarity. LigIIIα is less active than LigI in premature ligation of one-nucleotide gapped DNA and more efficiently discriminates misinsertion products of DNA polymerase β-catalyzed gap filling, that reinforces a leading role of LigIIIα in the accuracy of short-patch BER. LigI and LigIIIα reseal the intermediate of long-patch BER containing an incised synthetic AP site (F) with different efficiencies, depending on the DNA sequence context, 3′-end mismatch presence and coupling of the ligation reaction with DNA repair synthesis. Processing of this intermediate in the absence of flap endonuclease 1 generates non-canonical DNAs with bulged F site, which are very inefficiently repaired by AP endonuclease 1 and represent potential mutagenic repair products. The extent of conversion of the 5′-adenylated intermediates of specific and nonspecific substrates is revealed to depend on the DNA sequence context; a higher sensitivity of LigI to the sequence is in line with the enzyme structural feature of DNA binding. LigIIIα exceeds LigI in generation of potential abortive ligation products, justifying importance of XRCC1-mediated coordination of LigIIIα and aprataxin activities for the efficient DNA repair.

Comparison of tissue culture-induced variation in triticale regenerants obtained by androgenesis and somatic embryogenesis

Triticale is becoming an increasingly important livestock crop production. This is evidenced by increasing triticale-producing areas and by improved yields. In addition, meeting the increasing demand for cereals involves the introduction of high-yielding and stress-resistant varieties into breeding. In vitro culture techniques can accelerate the development of new varieties. Therefore, it seems extremely important to develop efficient plant regeneration methods through in vitro cultures and to understand the mechanisms involved in gaining regenerants. Obtaining regenerants of triticale through somatic embryogenesis and androgenesis may lead to tissue culture-induced variation. In the present study, we compared regenerants obtained in both regeneration systems (anther and immature zygotic embryo cultures), considering the level of genetic and epigenetic changes observed in different DNA sequence contexts for methylated cytosine (CG, CHG, CHH). The changes concerning the DNA sequence (so-called sequence variation) and the changes concerning the DNA methylation patterns, i.e., the removal of methylated cytosine (DNA demethylation) and the introduction of methylation to cytosine (de novo DNA methylation), were analyzed. We observed that regenerants derived via somatic embryogenesis and androgenesis differ notably for demethylation in the symmetrical CG sequence context and de novo methylation in the asymmetrical CHH context. These changes may be related to the reprogramming of microspore development from gametophytic to sporophytic and lack of such process in zygotic embryos.

TP53-dependent toxicity of CRISPR/Cas9 cuts is differential across genomic loci and can confound genetic screening

CRISPR/Cas9 gene editing can inactivate genes in a precise manner. This process involves DNA double-strand breaks (DSB), which may incur a loss of cell fitness. We hypothesize that DSB toxicity may be variable depending on the chromatin environment in the targeted locus. Here, by analyzing isogenic cell line pair CRISPR experiments jointly with previous screening data from across ~900 cell lines, we show that TP53-associated break toxicity is higher in genomic regions that harbor active chromatin, such as gene regulatory elements or transcription elongation histone marks. DSB repair pathway choice and DNA sequence context also associate with toxicity. We also show that, due to noise introduced by differential toxicity of sgRNA-targeted sites, the power of genetic screens to detect conditional essentiality is reduced in TP53 wild-type cells. Understanding the determinants of Cas9 cut toxicity will help improve design of CRISPR reagents to avoid incidental selection of TP53-deficient and/or DNA repair deficient cells.

N1-methyl-pseudouridine is incorporated with higher fidelity than pseudouridine in synthetic RNAs

In vitro transcribed synthetic messenger RNAs (mRNAs) represent a novel therapeutic modality. To overcome the inherent immunogenicity, as well as to increase the therapeutic efficacy of the molecules, uridine analogs—such as pseudouridine (Ψ) and N1-methyl-pseudouridine (m1Ψ), are incorporated in the synthetic mRNA. To decipher the fidelity with which these modifications are incorporated during the in vitro transcription (IVT) process, we compared the incorporation fidelity of uridine analogs with different RNA polymerases. We demonstrate that m1Ψ is incorporated with higher fidelity than Ψ. The fidelity of nucleotide incorporation differs between RNA polymerases; however, the spectrum of mutations observed between the RNAPs is similar. We also show that the array of nucleotide misincorporation is not dependent on the template DNA sequence context and that the distribution of these misincorporated nucleotides is not localized to any specific region along the length of the RNA. Based on our findings, we introduce a novel method to improve uridine analog incorporation fidelity during IVT. Our proof-of-concept experiments for higher-fidelity incorporation of uridine analogs during IVT provide guidelines when choosing RNAPs for the generation of modified uridine-containing mRNAs in vitro.

Establishing Linkages Among DNA Damage, Mutagenesis, and Genetic Diseases.

DNA damage by chemicals, radiation, or oxidative stress leads to a mutational spectrum, which is complex because it is determined in part by lesion structure, the DNA sequence context of the lesion, lesion repair kinetics, and the type of cells in which the lesion is replicated. Accumulation of mutations may give rise to genetic diseases such as cancer and therefore understanding the process underlying mutagenesis is of immense importance to preserve human health. Chemical or physical agents that cause cancer often leave their mutational fingerprints, which can be used to back-calculate the molecular events that led to disease. To make a clear link between DNA lesion structure and the mutations a given lesion induces, the field of single-lesion mutagenesis was developed. In the last three decades this area of research has seen much growth in several directions, which we attempt to describe in this Perspective.

Mutational signatures and processes in hepatobiliary cancers.

The evolutionary history of hepatobiliary cancers is embedded in their genomes. By analysing their catalogue of somatic mutations and the DNA sequence context in which they occur, it is possible to infer the mechanisms underpinning tumorigenesis. These mutational signatures reflect the exogenous and endogenous origins of genetic damage as well as the capacity of hepatobiliary cells to repair and replicate DNA. Genomic analysis of thousands of patients with hepatobiliary cancers has highlighted the diversity of mutagenic processes active in these malignancies, highlighting a prominent source of the inter-cancer-type, inter-patient, intertumour and intratumoural heterogeneity that is observed clinically. However, a substantial proportion of mutational signatures detected in hepatocellular carcinoma and biliary tract cancer remain of unknown cause, emphasizing the important contribution of processes yet to be identified. Exploiting mutational signatures to retrospectively understand hepatobiliary carcinogenesis could advance preventative management of these aggressive tumours as well as potentially predict treatment response and guide the development of therapies targeting tumour evolution.

The rate of formation and stability of abasic site interstrand crosslinks in the DNA duplex

DNA interstrand crosslinks (ICLs) strands pose an impenetrable barrier for DNA replication. Different ICLs are known to recruit distinct DNA repair pathways. NEIL3 glycosylase has been known to remove an abasic (Ap) site derived DNA crosslink (Ap-ICL). An Ap-ICL forms spontaneously from the Ap site with an adjacent adenine in the opposite strand. Lack of genetic models and a poor understanding of the fate of these lesions leads to many questions about the occurrence and the toxicity of Ap-ICL in cells.Here, we investigate the circumstances of Ap-ICL formation. With an array of different oligos, we have investigated the rates of formation, the yields, and the stability of Ap-ICL. Our findings point out how different bases in the vicinity of the Ap site change crosslink formation in vitro. We reveal that AT-rich rather than GC-rich regions in the surrounding Ap site lead to higher rates of Ap-ICL formation. Overall, our data reveal that Ap-ICL can be formed in virtually any DNA sequence context surrounding a hot spot of a 5′-Ap-dT pair, albeit with significantly different rates and yields. Based on Ap-ICL formation in vitro, we attempt to predict the number of Ap-ICLs in the cell.

Sequence context effects of replication of Fapy•dG in three mutational hot spot sequences of the p53 gene in human cells

Fapy•dG and 8-OxodGuo are formed in DNA from a common N7-dG radical intermediate by reaction with hydroxyl radical. Although cellular levels of Fapy•dG are often greater, its effects on replication are less well understood than those of 8-OxodGuo. In this study plasmid DNA containing Fapy•dG in three mutational hotspots of human cancers, codons 248, 249, and 273 of the p53 tumor suppressor gene, was replicated in HEK 293T cells. TLS efficiencies for the Fapy•dG containing plasmids varied from 72 to 89%, and were further reduced in polymerase-deficient cells. The mutation frequency (MF) of Fapy•dG ranged from 7.3 to 11.6%, with G→T and G→A as major mutations in codons 248 and 249 compared to primarily G→T in codon 273. Increased MF in hPol ι-, hPol κ-, and hPol ζ-deficient cells suggested that these polymerases more frequently insert the correct nucleotide dC opposite Fapy•dG, whereas decreased G→A in codons 248 and 249 and reduction of all mutations in codon 273 in hPol λ-deficient cells indicated hPol λ’s involvement in Fapy•dG mutagenesis. In vitro kinetic analysis using isolated translesion synthesis polymerases and hPol λ incompletely corroborated the mutagenesis experiments, indicating codependence on other proteins in the cellular milieu. In conclusion, Fapy•dG mutagenesis is dependent on the DNA sequence context, but its bypass by the TLS polymerases is largely error-free.

Androgenic-Induced Transposable Elements Dependent Sequence Variation in Barley.

A plant genome usually encompasses different families of transposable elements (TEs) that may constitute up to 85% of nuclear DNA. Under stressful conditions, some of them may activate, leading to sequence variation. In vitro plant regeneration may induce either phenotypic or genetic and epigenetic changes. While DNA methylation alternations might be related, i.e., to the Yang cycle problems, DNA pattern changes, especially DNA demethylation, may activate TEs that could result in point mutations in DNA sequence changes. Thus, TEs have the highest input into sequence variation (SV). A set of barley regenerants were derived via in vitro anther culture. High Performance Liquid Chromatography (RP-HPLC), used to study the global DNA methylation of donor plants and their regenerants, showed that the level of DNA methylation increased in regenerants by 1.45% compared to the donors. The Methyl-Sensitive Transposon Display (MSTD) based on methylation-sensitive Amplified Fragment Length Polymorphism (metAFLP) approach demonstrated that, depending on the selected elements belonging to the TEs family analyzed, varying levels of sequence variation were evaluated. DNA sequence contexts may have a different impact on SV generated by distinct mobile elements belonged to various TE families. Based on the presented study, some of the selected mobile elements contribute differently to TE-related SV. The surrounding context of the TEs DNA sequence is possibly important here, and the study explained some part of SV related to those contexts.

Mining mutation contexts across the cancer genome to map tumor site of origin

The vast preponderance of somatic mutations in a typical cancer are either extremely rare or have never been previously recorded in available databases that track somatic mutations. These constitute a hidden genome that contrasts the relatively small number of mutations that occur frequently, the properties of which have been studied in depth. Here we demonstrate that this hidden genome contains much more accurate information than common mutations for the purpose of identifying the site of origin of primary cancers in settings where this is unknown. We accomplish this using a projection-based statistical method that achieves a highly effective signal condensation, by leveraging DNA sequence and epigenetic contexts using a set of meta-features that embody the mutation contexts of rare variants throughout the genome.

DNA Methylation, Deamination, and Translesion Synthesis Combine to Generate Footprint Mutations in Cancer Driver Genes in B-Cell Derived Lymphomas and Other Cancers.

Cancer genomes harbor numerous genomic alterations and many cancers accumulate thousands of nucleotide sequence variations. A prominent fraction of these mutations arises as a consequence of the off-target activity of DNA/RNA editing cytosine deaminases followed by the replication/repair of edited sites by DNA polymerases (pol), as deduced from the analysis of the DNA sequence context of mutations in different tumor tissues. We have used the weight matrix (sequence profile) approach to analyze mutagenesis due to Activation Induced Deaminase (AID) and two error-prone DNA polymerases. Control experiments using shuffled weight matrices and somatic mutations in immunoglobulin genes confirmed the power of this method. Analysis of somatic mutations in various cancers suggested that AID and DNA polymerases η and θ contribute to mutagenesis in contexts that almost universally correlate with the context of mutations in A:T and G:C sites during the affinity maturation of immunoglobulin genes. Previously, we demonstrated that AID contributes to mutagenesis in (de)methylated genomic DNA in various cancers. Our current analysis of methylation data from malignant lymphomas suggests that driver genes are subject to different (de)methylation processes than non-driver genes and, in addition to AID, the activity of pols η and θ contributes to the establishment of methylation-dependent mutation profiles. This may reflect the functional importance of interplay between mutagenesis in cancer and (de)methylation processes in different groups of genes. The resulting changes in CpG methylation levels and chromatin modifications are likely to cause changes in the expression levels of driver genes that may affect cancer initiation and/or progression.