Biology subjects

Wong, L.

Publications and source records attributed to Wong, L..

At least 19 recordsLinked to original sources

bioRxiv2026

Integrated analysis of ribosomal DNA copy number and methylation using nanopore long-read sequencing

Ribosomal RNA (rRNA) provides the structural and catalytic core of ribosomes and is encoded by ribosomal RNA genes (rDNA) arranged in tandem repeat arrays. rDNA copy number (CN) is highly dynamic, representing a clinically relevant form of structural variation, but its accurate quantification has been challenging due to its highly repetitive and GC-rich nature. Here, we present RICO (Ribosomal DNA Integrated Copy Number and Methylation Analysis), a novel computational pipeline for integrated estimation of rDNA CN and methylation using nanopore long-read sequencing. RICO leverages long sequencing reads that span entire rDNA repeats, mapped to an rDNA-augmented reference genome, and normalizes coverage using an array of single-copy genes. We show that RICO provides accurate rDNA CN estimates in simulated datasets and reproducible measurements across human samples, with strong agreement to short-read sequencing and PCR-based methods. As biological validation, RICO detects a [~]40% reduction in rDNA CN in Atrx-knockout mouse cells, consistent with established effects of ATRX loss on rDNA CN, and captures detected increased total and active rDNA CN in malignant cells from a MYC-driven B-cell lymphoma mouse model, in line with prior psoralen-based chromatin studies. Applying RICO to independent human cohorts, we uncover that individuals with higher total rDNA CN consistently exhibited higher fractions of high-methylated rDNA copies, suggesting a dosage compensation mechanism that potentially maintains a similar number of active rDNA copies across individuals. Together, RICO enables integrated analysis of rDNA CN and methylation state, providing a scalable framework for investigating rDNA regulation across population and disease studies.

bioinformatics↗

bioRxiv2026

Comprehensive evaluation of LLM capabilities for interpretation and analysis of genome-scale metabolic models in metabolic engineering

Genome-scale metabolic models (GSMs) underpin pathway and strain engineering by linking genes to metabolic reactions and enabling system-level simulation of cellular fluxes and intervention effects, yet end-to-end analysis workflows remain fragmented, expert-demanding, and slow to adapt. Large language models (LLMs) could transform this landscape, lowering the barrier by explaining concepts, interpreting GSM files, and turning natural-language instructions into valid analysis code, thereby substantially mitigating the time, effort, and expertise required. However, their reliability for domain-specific tasks remains unexplored. Here, we delivered a systematic benchmark of four leading LLMs (GPT-4, Gemini, Claude, DeepSeek-R1) across four task areas central to metabolic engineering: domain knowledge, metabolic flux prediction, pathway construction, and flux optimization. For benchmarking, we introduced a standardized, rubric-based evaluation framework that uses multi-LLM automated scoring (an ensemble of LLM-as-a-judge assessments) and two distinct sets of nine task-tailored metrics (domain vs coding-focused tasks), rated on a 1-5 scale (up to 45 per task), covering scientific validity and code executability where applicable. Across tasks, we reveal consistent strengths (conceptual explanation, code synthesis) and critical failure modes (e.g., context window limitations, incorrect identifier assumptions, strain-dependent reasoning errors, and errors in domain-specific algorithms). In aggregate, DeepSeek-R1 led in domain tasks, narrowly edging GPT-4, Claude, and Gemini, demonstrating that conceptual biological logic remains highly invariant across architectures. In contrast, Gemini achieved the highest score for coding tasks, distinguished by functional execution and excelled in error handling, documentation, and readability, followed by GPT-4, Claude, and DeepSeek. We also evaluated LLM self-inspection capability by injecting subtle, consequential faults: a stoichiometric sign error causing mass imbalance and an omitted pathway reaction. We reveal that conversational "blind search" prompting completely fails to localize these network faults. Instead, robust error localization requires prompts reframed with domain-informed constraints that force the LLM to leverage tool-assisted code procedures, such as COBRApy mass-balance functions. Together, this work establishes an evidence-based baseline for LLM-enabled GSM analysis, providing actionable guidance for building reliable, automation-ready workflows for pathway and strain design. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=88 SRC="FIGDIR/small/730004v1_ufig1.gif" ALT="Figure 1"> View larger version (28K): org.highwire.dtl.DTLVardef@413e29org.highwire.dtl.DTLVardef@1581de2org.highwire.dtl.DTLVardef@11ef7borg.highwire.dtl.DTLVardef@1817653_HPS_FORMAT_FIGEXP M_FIG C_FIG

systems biology↗

bioRxiv2026

Skiver: Alignment-free Estimation of Sequencing Error Rates and Spectra using (k, v)-mer Sketches

BackgroundQuality control of sequencing datasets is an important first step in numerous bioinformatics pipelines such as mapping, variant calling, and assembly. Existing methods typically rely on alignment results or quality scores. However, the reference genome is not always available for mapping, and uncalibrated quality scores may yield biased estimates of error rates. ResultsWe present skiver, a reference-free and alignment-free algorithm that estimates sequencing error rates and calibrates Phred quality scores using (k, v)-mer sketches. By identifying the consensus from the sketched (k, v)-mers, skiver estimates survival and hazard rates that capture positional information of sequencing errors. Across simulated and real datasets from various sequencing platforms, skiver accurately recovers sequencing error rates and the proportion of different error types. We further demonstrate its ability to calibrate Phred scores. It also reliably handles complex datasets containing multiple strains, alleles, and repetitive regions through an iterative outlier filtering strategy. Skiver is computationally efficient and supports tools that need accurate sequencing error rate estimates or quality scores as prior knowledge. Availability and ImplementationAn implementation of skiver is available at https://github.com/GZHoffie/skiver, and dataset and scripts for reproducibility are available at https://github.com/GZHoffie/skiver-test.

bioinformatics↗

bioRxiv2025

Efficient trace reconstruction in DNA storage systems using Bidirectional Beam Search

MotivationAs DNA data storage systems gain popularity, the need for an efficient trace reconstruction algorithm becomes increasingly important. These algorithms aim to reconstruct the original encoded sequence from its noisy sequenced copies (or "traces"), enabling a faster and more reliable decoding process. Previous works have often been adaptations of methods for multiple sequence alignment or read error correction, typically operating under strict assumptions such as fixed error rates. However, such methods demonstrate limited generalizability to real datasets with higher error rates and suffer from slow processing times when dealing with a large number of traces. ResultsWe propose a new probabilistic formulation of the trace reconstruction problem. Instead of optimizing alignment among traces, we model the traces as observations of a k-th order Markov chain and try to predict the sequence that is generated by the Markov chain with the highest probability. Such a formulation inspires a novel solution, i.e. Bidirectional Beam Search (BBS), whose reconstruction phase operates in linear time with respect to the length of the encoded sequences. Experiments on multiple public Nanopore datasets demonstrate that BBS achieves top-tier accuracy compared with the state-of-the-art methods while being [~]20x faster, showing its potential to enhance the efficiency of DNA data storage systems. Availability and ImplementationThe implementation of BBS is available at https://github.com/GZHoffie/bbs, and the dataset and scripts for reproducibility are available at https://github.com/GZHoffie/bbs-test.

bioinformatics↗

bioRxiv2024

Human class B1 GPCR modulation by plasma membrane lipids

The class B1 G protein-coupled receptor (GPCR) subfamily is a class of receptors known for their regulatory roles in metabolism and neuronal activity and as important drug targets. Lipids play key functional roles in modulation of GPCR signalling, yet our understanding of the molecular level detail of specific lipid interactions with class B1 GPCRs remains limited. Here we present coarse-grained molecular dynamics (MD) simulations of the active and inactive states of 15 human class B1 family members and use aiida-gromacs to capture full provenance for the set-up of simulations in complex plasma membranes. Receptors exhibit state-dependent lipid interactions with the regulatory sterol cholesterol and phospholipid phosphatidylinositiol-3,4-bisphosphate (PIP2) at defined locations on the receptor surface. Global analysis of trends across the subfamily reveals conserved patterns of lipid interaction dynamics. The glycosphingolipid GM3 exerts a modulatory influence on the dynamics of class B1 extracellular domains in both simulations and in vitro time-resolved FRET assays.

molecular biology↗

bioRxiv2024

GCfix: A Fast and Accurate Fragment Length-SpecificMethod for Correcting GC Bias in Cell-Free DNA

MotivationCell-free DNA (cfDNA) analysis has wide-ranging clinical applications due to its non-invasive nature. However, cfDNA fragmentomics and copy number analysis can be complicated by GC bias. There is a lack of GC correction software based on rigorous cfDNA GC bias analysis. Furthermore, there is no standardized metric for comparing GC bias correction methods across large sample sets, nor a rigorous experiment setup to demonstrate their effectiveness on cfDNA data at various coverage levels. ResultsWe present GCfix, a method for robust GC bias correction in cfDNA data across diverse coverages. Developed following an in-depth analysis of cfDNA GC bias at the region and fragment length levels, GCfix is both fast and accurate. It works on all reference genomes and generates correction factors, tagged BAM files, and corrected coverage tracks. We also introduce two orthogonal performance metrics for (1) comparing the fragment count density distribution of GC content between expected and corrected samples, and (2) evaluating coverage profile improvement post-correction. GCfix outperforms existing cfDNA GC bias correction methods on these metrics. AvailabilityGCfix software and code for reproducing the figures are publicly accessible on GitHub: https://github.com/Rafeed-bot/GCfix_Software.

genomics↗

bioRxiv2024

Benchmarking Recent Computational Tools for DNA-binding Protein Identification

Identification of DNA-binding proteins (DBPs) is a crucial task in genome annotation, as it aids in understanding gene regulation, DNA replication, transcriptional control and various cellular processes. In this paper, we conduct an unbiased benchmarking of eleven state-of-the-art computational tools as well as traditional tools such as ScanProsite, BLAST, and HMMER for identifying DBPs. We highlight the data leakage issue in conventional datasets leading to inflated performance. We introduce new evaluation datasets to support further development. Through a comprehensive evaluation pipeline, we identify potential limitations in models, feature extraction techniques and training methods; and recommend solutions regarding these issues. We show that combining the predictions of the two best computational tools with BLAST based prediction significantly enhances DBP identification capability. We provide this consensus method as user-friendly software. The datasets and software are available at: https://github.com/Rafeed-bot/DNA_BP_Benchmarking. 1. Key PointsO_LIWe designed a comprehensive evaluation pipeline which systematically evaluates eleven recent machine learning (ML) based DBP identification tools. C_LIO_LIWe analyzed the test prediction mistakes made by top-performing tools identifying their potential limitations in terms of model architecture, feature extraction and class balancing. C_LIO_LIWe showed that although the best of these tools do not convincingly outperform BLAST, they still provide substantial value when integrated together with BLAST into a simple majority-voting ensemble. C_LIO_LIWe provide recommendations on more robust development & evaluation and better usability of future tools. C_LIO_LIWe provide the two best-performing ML-based tools, BLAST and the ensemble method as user-friendly software, as well as our proposed datasets, publicly available via GitHub. C_LI

bioinformatics↗

bioRxiv2024

Genome-guided, field-deployable loop-mediated isothermal amplification (LAMP) assay for specific detection of Dickeya dadantii

Potatoes, among the most economically significant crops worldwide, are susceptible to various plant pathogens that significantly impact their propagation, production, storage, and distribution. Soft rot disease, caused primarily by Dickeya and Pectobacterium, results in substantial economic losses to the agricultural industry annually. In this study, we developed a rapid, reliable, and field-deployable loop-mediated isothermal amplification (LAMP) assay for detecting D. dadantii, a common soft rot causing bacteria. The D. dadantii-specific LAMP primers were designed targeting a highly conserved genomic region within D. dadantii, the TetR/AcrR family transcriptional regulator CDS and its flanking sequences. This assay was thoroughly validated with the members of inclusivity (nine strains of D. dadantii) and exclusivity panels (85 strains, including all Dickeya species, related taxa, and host DNA), detecting no false positives or negatives. The limit of detection (LOD) was established by performing assays with 10-fold serially diluted pure gDNA of D. dadantii and gDNA spiked with host crude extract; the assay detected the target pathogen down to 1 pg (188 copies) without being adversely affected by the host crude extract. The developed LAMP assay specifically detected the target pathogen in infected plant materials. Additional multi-operator blind and multi-instrument tests were conducted to assess the assays robustness and applicability, consistently yielding accurate results without false positives or negatives. These findings demonstrate the assays potential utility for biosecurity, routine diagnostics, and epidemiological studies.

microbiology↗

bioRxiv2024

Long-read sequencing and genome assembly of natural history collection samples and challenging specimens

Museum collections harbor millions of samples, largely unutilized for long-read sequencing. Here, we use ethanol-preserved samples containing kilobase-sized DNA to show that amplification-free protocols can yield contiguous genome assemblies. Additionally, using a modified amplification-based protocol, employing an alternative polymerase to overcome PCR bias, we assembled the 3.1 Gb maned sloth genome, surpassing the previous 500 Mb protocol size limit. Our protocol also improves assemblies of other difficult-to-sequence molluscs and arthropods, including millimeter-sized organisms. By highlighting collections as valuable sample resources and facilitating genome assembly of tiny and challenging organisms, our study advances efforts to obtain reference genomes of all eukaryotes.

genomics↗

bioRxiv2023

Quantification of circulating tumor DNA using deep learning

Quantification of circulating tumor DNA (ctDNA) levels in blood enables non-invasive surveillance of cancer progression. Fragle is an ultra-fast deep learning-based method for ctDNA quantification directly from cell-free DNA fragment length profiles. We developed Fragle using low-pass whole genome sequence (lpWGS) data from multiple cancer types and healthy control cohorts, demonstrating high accuracy, and improved lower limit of detection in independent cohorts as compared to existing tumor-naive methods. Uniquely, Fragle is also compatible with targeted sequencing data, exhibiting high accuracy across both research and commercial targeted gene panels. We used this method to study longitudinal plasma samples from colorectal cancer patients, identifying strong concordance of ctDNA dynamics and treatment response. Furthermore, prediction of minimal residual disease in resected lung cancer patients demonstrated significant risk stratification beyond a tumor-naive gene panel. Overall, Fragle is a versatile, fast, and accurate method for ctDNA quantification with potential for broad clinical utility.

bioinformatics↗

bioRxiv2023

Similar deamination activity but different phenotypic outcomes induced by APOBEC3 enzymes in breast epithelial cells

APOBEC3 (A3) enzymes deaminate cytosine to uracil in viral single-stranded DNA as a mutagenic barrier for some viruses. A3-induced deaminations can also occur in human genomes resulting in an endogenous source of somatic mutations in multiple cancers. However, the roles of each A3 are unclear since few studies have assessed these enzymes in parallel. Thus, we developed stable cell lines expressing A3A, A3B, or A3H Hap I using non-tumorigenic MCF10A and tumorigenic MCF7 breast epithelial cells, to assess their mutagenic potential and cancer phenotypes in breast cells. The activity of these enzymes was characterized by {gamma}H2AX foci formation and in vitro deamination. Cell migration, and soft agar colony formation assays assessed cellular transformation potential. We found that all three A3 enzymes had similar {gamma}H2AX foci formation, despite different deamination activity in vitro. Notably, in nuclear lysates the in vitro deaminase activity of A3A, A3B, and A3H did not require digestion of cellular RNA, in contrast to A3B and A3H in whole cell lysates. Their similar activities in cells nonetheless resulted in distinct phenotypes where A3A decreased colony formation in soft agar, A3B decreased colony formation in soft agar after hydroxyurea treatment, and A3H Hap I promoted cell migration. Overall, we show that in vitro deamination data does not always reflect in cell deamination, all three A3s induce somatic mutagenesis, and the impact of each is different.

cancer biology↗

bioRxiv2023

Frontotemporal Dementia Patient Neurons With Progranulin Deficiency Display Protein Dyshomeostasis

Haploinsufficiency of progranulin (PGRN) causes frontotemporal dementia (FTD), a devastating neurodegenerative disease with no effective treatment. PGRN is required for efficient proteostasis, as loss of neuronal PGRN results in dysfunctional lysosomes and impaired clearance and cytoplasmic aggregation of TDP-43, a protein involved in neurodegeneration in FTD. These and other events lead to neurodegeneration and neuroinflammation. However, the detailed mechanisms leading to protein dyshomeostasis in PGRN-deficient cells remain unclear. We report here the development of human cell models of FTD with PGRN-deficiency to explore the molecular mechanisms underlying proteostasis breakdown and TDP-43 aggregation in FTD. Neurons differentiated from FTD patient induced pluripotent stem cells (iPSCs) have reduced PGRN levels, and the neurons recapitulate key disease features, including impaired lysosomal function, defective TDP-43 turnover and accumulation, neurodegeneration, and death. Proteomic analysis revealed altered levels of proteins linked to the autophagy-lysosome pathway (ALP) and the ubiquitin-proteasome system (UPS) in FTD patient neurons, providing new mechanistic insights into the link between PGRN-deficiency and disease pathobiology.

neuroscience↗

bioRxiv2022

HydRA: Deep-learning models for predicting RNA-binding capacity from protein interaction association context and protein sequence

RNA-binding proteins (RBPs) control RNA metabolism to orchestrate gene expression, and dysfunctional RBPs underlie many human diseases. Proteome-wide discovery efforts predict thousands of novel RBPs, many of which lack canonical RNA-binding domains. Here, we present a hybrid ensemble RBP classifier (HydRA) that leverages information from both intermolecular protein interactions and internal protein sequence patterns to predict RNA-binding capacity with unparalleled specificity and sensitivity using support vector machine, convolutional neural networks and transformer-based protein language models. HydRA enables Occlusion Mapping to robustly detect known RNA-binding domains and to predict hundreds of uncharacterized RNA-binding domains. Enhanced CLIP validation for a diverse collection of RBP candidates reveals genome-wide targets and confirms RNA-binding activity for HydRA-predicted domains. The HydRA computational framework accelerates construction of a comprehensive RBP catalogue and expands the set of known RNA-binding protein domains. HighlightsO_LIHydRA combines protein-protein interaction and amino acid sequence information to predict RNA binding activity for 1,487 candidate genes. C_LIO_LIHydRA predicts RNA binding with higher specificity and sensitivity than current approaches, notably for RBPs without well-defined RNA-binding domains. C_LIO_LIOcclusion Mapping with HydRA enables RNA-binding domain discovery. C_LIO_LIEnhanced CLIP confirms HydRA RBP predictions with RNA-binding domain resolution. C_LI

bioinformatics↗

bioRxiv2021

Microglia-specific promoter activities of human HexB gene

Adeno-associated virus (AAV)-mediated genetic targeting of microglia remains challenging. Overcoming this hurdle is essential for gene editing in the central nervous system (CNS). Here, we characterized the minimal/native promoter of the HexB gene, which is known to be specifically and stably expressed in the microglia during homeostatic and pathological conditions. Dual reporter and serial deletion assays identified the critical role of the natural 5 untranslated region (-97 bp related to the first ATG) in driving transcriptional activity of the mouse HexB gene. The native promoter region of mouse, human and monkey HexB located at -135, -134 and -170 bp to the first ATG, respectively. These promoters were highly active and specific in microglia with strong cross-species transcriptional activities, but had no activities in primary astrocytes. In addition, we identified a 135 bp promoter of CD68 gene was also highly active in microglia but not in astrocytes. Considering that HexB is specifically expressed in microglia, not in monocytes/macrophages or other neuronal cells, these data suggest that the newly characterized 134 bp microglia-specific HexB promoter can be an ideal candidate for microglia-targeting AAV gene therapy, which could be developed for HIV eradication in the brain wherein microglia harbor the main HIV reservoirs in the CNS. SummaryIt is hard to overstate the importance of gene therapy that can remove viral genes from human cells. A cure for HIV would mean a lifetime free of treatment for patients who now must maintain a strict regimen of ART indefinitely. In order to develop a cure using AAV delivery, payload DNA must meet the AAV vector size limitations, and the payload genes must be expressed appropriately. Previous studies have identified microglia-specific HexB gene that shows stable expression during neural homeostasis and pathogenesis. Our study identified the essential HexB gene promoter (134 bp) as a strong candidate for AAV gene therapy to specifically target the brain microglia, the main cellular reservoirs of HIV in the central nervous system. Our studies continue to move us closer to identifying target-specific gene therapy for NeuroAIDS.

neuroscience↗

bioRxiv2021

High-density spinal cord stimulation selectively activates lower urinary tract afferents

Epidural spinal cord stimulation (SCS) has recently been reported as a potential intervention to improve limb and autonomic functions, with lumbar stimulation improving locomotion and thoracic stimulation regulating blood pressure. We asked whether sacral SCS could be used to target the lower urinary tract. Here we show that high-density epidural SCS over the sacral spinal cord and cauda equina of anesthetized cats evokes responses in nerves innervating the bladder and urethra and that these nerves can be activated selectively. Sacral epidural SCS always recruited the pelvic and pudendal nerves and selectively recruited these nerves in all but one animal. Individual branches of the pudendal nerve were always recruited as well. Electrodes that selectively recruited specific peripheral nerves were spatially clustered on the arrays, suggesting anatomically organized sensory pathways. This selective recruitment demonstrates a mechanism to directly modulate bladder and urethral function through known reflex pathways, which could be used to restore bladder and urethral function after injury or disease.

bioengineering↗

bioRxiv2021

Selective stimulation of the ferret abdominal vagus nerve with multi-contact nerve cuff electrodes

Dysfunction and diseases of the gastrointestinal (GI) tract are a major driver of medical care. The vagus nerve innervates and controls multiple organs of the GI tract and vagus nerve stimulation (VNS) could provide a means for affecting GI function and treating disease. However, the vagus nerve also innervates many other organs throughout the body, and off-target effects of VNS could cause major side effects such as changes in blood pressure. In this study, we aimed to achieve selective stimulation of populations of vagal afferents using a multi-contact cuff electrode wrapped around the abdominal trunks of the vagus nerve. Four-contact nerve cuff electrodes were implanted around the dorsal (N=3) or ventral (N=3) abdominal vagus nerve in six ferrets, and the response to stimulation was measured via a 32-channel microelectrode array (MEA) inserted into the nodose ganglion. Selectivity was characterized by the ability to evoke responses in MEA channels through one bipolar pair of cuff contacts but not through the other bipolar pair. We demonstrated that is was possible to selectively activate subpopulations of vagal afferents using abdominal VNS. Additionally, we quantified the conduction velocity of evoked responses to determine what types of nerve fibers (i.e. A{delta} vs. C) responded to stimulation. We also quantified the spatial organization of evoked responses in the nodose MEA to determine if there is somatotopic organization of the neurons in that ganglion. Finally, we demonstrated in a separate set of three ferrets that stimulation of the abdominal vagus via a four-contact cuff could selectively alter gastric myoelectric activity, suggesting that abdominal VNS can potentially be used to control GI function.

bioengineering↗

bioRxiv2020

SPRDA: a matrix completion approach based on the structural perturbation to infer disease-associated Piwi-Interacting RNAs

Emerging evidence suggests that PIWI-interacting RNAs (piRNAs) are one of the most influential small non-coding RNAs (ncRNAs) that regulate RNA silencing. piRNA and PIWI proteins have been confirmed for disease diagnosis and treatment as novel biomarkers due to its abnormal expression in various cancers. However, the current research is not strong enough to further clarify the functions of piRNA in cancer and its underlying mechanism. Therefore, how to provide large-scale and serious piRNA candidates for biological research has grown up to be a pressing issue. The main motivation of this work is tantamount to fill the gap in research on large-scale prediction of disease-related piRNAs. In this study, a novel computational model based on the structural perturbation method is proposed, called SPRDA. In detail, the duplex network is constructed based on the piRNA similarity network and disease similarity network extracted from piRNA sequence information, Gaussian interaction profile kernel similarity information and gene-disease association information. The structural perturbation method is then used to predict the potential associations on the duplex network, which is more predictive than other network structures in terms of structural consistency. In the five-fold cross-validation, SPRDA shows high performance on the benchmark dataset piRDisease, with an AUC of 0.9529. Furthermore, the predictive performance of SPRDA for 10 diseases shows the robustness of the proposed method. Overall, the proposed approach can provide unique insights into the pathogenesis of the disease and will advance the field of oncology diagnosis and treatment.Competing Interest StatementThe authors have declared no competing interest.View Full Text

bioinformatics↗

bioRxiv2020

Induced mutagenesis by the DNA cytosine deaminase APOBEC3H Haplotype I protects against lung cancer

The DNA cytosine deaminases APOBEC3A, APOBEC3B, and APOBEC3H Haplotype I can induce mutations in cells that lead to cancer evolution. The database of cancer biomarkers in DNA repair genes (DNArCdb) identified a single nucleotide polymorphism (rs139298) of APOBEC3H Haplotype I that is involved in lung cancer1. Here, we show that this single nucleotide polymorphism causes the destabilization of APOBEC3H Haplotype I. Computational analysis suggests that the resulting K121E change affects the structure of APOBEC3H leading to active site disruption and destabilization of the RNA-mediated dimer interface. A K117E mutation in a K121E background stabilized the APOBEC3H Haplotype I, enabled biochemical study, and showed that the K121E affected catalytic activity, single-stranded DNA binding, and oligomerization on single-stranded DNA. That the destabilization of a DNA mutator would be associated with lung cancer suggests that too much mutation could result in immune recognition or death of tumor cells, suggesting that multiple APOBEC3s would not be expressed in the same tumor cells. In support of this hypothesis, stably expressed APOBEC3H Haplotype I caused a high amount of double-stranded DNA breaks in a lung cancer cell line that endogenously expressed APOBEC3B. Altogether, the data support the model that high APOBEC3 mutations in tumors are protective.

biophysics↗