Biology subjects

Kumar, S.

Publications and source records attributed to Kumar, S..

At least 145 records · Page 8Linked to original sources

bioRxiv2025

Engineering single-AAV CRISPR-Cas13 RNA base editors for treatment of inherited retinal diseases

Site-directed RNA editing, especially RNA base editing, allows for specific manipulation of RNA sequences, making it a useful approach for the correction of pathogenic mutations. Correction of RNA transcripts allows therapeutic gene editing in a safe and reversible manner and avoids permanent alterations in the genome. RNA-targeting CRISPR-Cas nucleases (e.g., CRISPR-Cas13) enable delivery within a single adeno-associated virus (AAV) vector for RNA base editing, making the approach clinically feasible. Here, we used the inactive CRISPR-Cas13bt3 (also known as Cas13X.1) fused to the ADAR2 deaminase domain (ADAR2DD) for targeted correction of inherited retinal disease (IRD) mutations. First, we show in vitro that dCas13bt3-ADAR2DD can efficiently correct a pathogenic nonsense mutation (c.130C>T [p.R44X]) found in the mouse Rpe65 gene and recover protein expression in retinal pigment epithelium cells (RPEs). Across clinically reported RPE65 mutations, we observed editing efficiencies ranging from 0% to 60%. In the Rpe65-deficient mouse model of retinal degeneration (rd12), we observed that RNA base editing can recover Rpe65 expression in RPEs and rescue retinal function with no observable adverse effects. We further employed our RNA base editor against the large USH2A gene to assess the promise of RNA base editing for addressing untreatable IRDs caused by genes too large for AAV gene delivery. Against the human USH2A in vitro, we observed up to 60% on-target efficiency. We further found that gRNA mismatches, domain-inlaid ADAR2DD design and nucleocytoplasmic shuttling of the RNA base editor optimised on-target and bystander editing for a highly precise base editor. Against the mouse Ush2a in vitro, we similarly observed up to 60% on-target editing in mammalian cells, while in the Ush2aW3947X mice, we observed [~]12% on-target editing, with no impact on retinal structure or function, or transcriptome-wide editing. Overall, our findings demonstrate dCas13bt3-ADAR2DD as a potent tool for gene therapy against IRDs, addressing a significant unmet clinical need in ophthalmology.

bioengineering↗

bioRxiv2025

Barrier Function of the Extracellular Matrix in AAV Gene Therapy

PurposeThe extracellular matrix (ECM) is a major component of the tissue microenvironment which may pose a barrier to the distribution of AAV in target organs, preventing delivery of therapeutic cargo. We sought to address this potential barrier to AAV gene therapy by furthering our understanding of AAV-ECM interactions. We hypothesized that both the AAV serotype and ECM composition will impact AAV transport and gene delivery. MethodsAAV2, AAV6, and AAV8 viral vectors were fluorescently labeled to allow for visualization of their diffusion through the ECM. Lung, liver, and small intestinal submucosal dECM hydrogels were formulated as models of the ECM with tissue-specific biomolecular content. We then characterized AAV and nanoparticle diffusion within decellularized ECM using fluorescent video microscopy and multiple particle tracking. Additionally, we evaluated AAV transduction in dECM-incorporated 2D and 3D spheroid tissue culture models. ResultsAll AAV displayed reduced diffusivity through ECM as compared to similarly sized nanoparticles. AAV2 diffusion was least affected by the presence of ECM across tissue types as compared to AAV6 and AAV8. AAV transduction in dECM incorporated in vitro models was significantly reduced in both a 2D and 3D setting. ConclusionsThese results suggest binding of AAV to the ECM may decrease their therapeutic effect in target tissues throughout the body. The barrier function of the ECM should be considered in development of AAV for gene therapy applications.

bioengineering↗

bioRxiv2025

Assessing the Potential of Switchgrass (Panicum virgatum L.) for Storing Carbon Belowground: Insights from a Multi-Site US Study

Agricultural intensification depletes belowground carbon (C) stocks, partly due to the exclusion of deep-rooted perennial species from landscapes where they were once dominant. Reintroducing deep-rooted perennials to cultivated land may help to mitigate SOC loss and restore ecosystem function. We quantified the effect of replacing shallow-rooted annual crops with a deep-rooted perennial grass, switchgrass (Panicum virgatum L.), by comparing 10 to 30 year-old switchgrass stands with paired annual row crop fields across the central and eastern USA. We hypothesized that switchgrass would store more root C and SOC than neighboring shallow-rooted annual crops, and that these effects would extend deeper than 30 cm. We also evaluated whether switchgrass stimulates decomposition of SOC at depth using radiocarbon (14C) to quantify replacement of slow cycling isotopically depleted SOC. Finally, we explored whether the effect of switchgrass on SOC is moderated by soil chemical and physical properties. While the effect of switchgrass on SOC in the surface 100 cm was positive at most sites, the average effect was highly uncertain and statistically indistinguishable from zero (mean difference in SOC = 0.6 kg C m-2 [95% CI -0.8 to +1.9 kg C m-2]). By contrast, we found that root biomass C was consistently more deeply distributed and more abundant under switchgrass, yielding an estimated additional 0.6 kg C m-2 in the surface 100 cm of soil under switchgrass [95% CI +0.5 to +0.7 kg C m-2]. 14C measurements suggested that root C inputs were adding to existing SOC without stimulating decomposition. The effect of switchgrass on belowground C was not strongly related to any of the environmental factors that we evaluated. Our observations show that root biomass C can contribute substantially to belowground C stocks when deep-rooted perennial grasses replace shallow-rooted annual crops.

ecology↗

bioRxiv2025

Identification and Characterization of Outer Membrane Proteins and Membrane Spanning Protein Complexes in Brucella melitensis

Brucellosis (Malta fever) is a zoonotic disease that affects both humans and animals, including cattle, sheep, and goats. Brucella melitensis is the most virulent and clinically significant species in humans. It is a Gram-negative bacterium with three groups of outer membrane proteins (OMPs): minor OMPs (Group 1), and major OMPs (Groups 2 and 3). OMPs with {beta}-barrel architecture play important roles in nutrient transport, efflux, adhesion, and membrane biogenesis. Despite their importance, the structure, function, and interaction dynamics of several B. melitensis {beta}-barrel OMPs and associated protein complexes remain mostly unexplored. In this study, we conducted a comprehensive in silico analysis to characterize known outer membrane {beta}-barrel (OMBB) proteins and identify novel OMBBs in B. melitensis 16M. Proteins were modelled using five computational tools: AlphaFold 3, ESMFold, SWISS-MODEL, RoseTTAFold, and TrRosetta. Outer-membrane insertion of the novel OMBBs was confirmed using PPM 3.0, Protein GRAVY, DREAMM, and MemProtMD_Insane. Putative functions were predicted using structure- and sequence-based annotations. Sequence variation across 46 B. melitensis strains were identified and mapped onto the structural models. OMBB-associated protein complexes - the RND (Resistance-Nodulation-Division) efflux pumps, the lipopolysaccharide transport (Lpt) complex, and the {beta}-barrel assembly machinery (BAM) complex - were modelled, and protein-protein interactions (PPIs) were analyzed to confirm thermodynamically stable assemblies. This study presents a robust in silico strategy for exploring OMP architecture and provides valuable structural insights to support the development of diagnostics, targeted therapeutics, and vaccines against B. melitensis.

bioinformatics↗

bioRxiv2025

What's next for Indian Ornithology? 101 key research questions

India has a rich history of ornithology, and bird research in this field has expanded considerably in recent decades, spurred by a growing number of birdwatchers and ornithologists. Despite this progress, critical gaps remain. This paper highlights key areas where further research is needed and identifies pressing questions that will shape Indian ornithology in the coming years. Drawing from diverse inputs, we present a curated list of 101 research questions spanning across the various disciplines in ornithology - Natural History, Physiology and Disease Ecology, Behaviour, Population and Community Ecology, Habitat Ecology, Macroecology and Biogeography, Population Genetics and Evolution, Applied/Economic Ornithology and Conservation. The list was compiled through a multi-stage process, starting with a public open call for questions, followed by review and curation by a smaller panel of subject specialists. Each question was independently scored by a panel of experts based on three criteria: generality, novelty, and relevance. To account for variation in scoring styles, scores were normalised using Z-scores. The top 101 research questions were then selected based on these standardised scores. Each question is accompanied by an annotation that describes the significance of the question, and highlights opportunities to address it. While our list of research questions highlights significant research priorities, it is not intended to be exhaustive. Rather, it reflects the perspectives of those involved in its curation, who deemed these questions particularly relevant and impactful towards advancing Indian ornithology. We expect these questions to spark new project ideas among students, researchers, and citizen scientists, while guiding funders, managers, and policymakers toward priority research areas.

ecology↗

bioRxiv2025

Charged Scanning Mutagenesis as a High Throughput Approach for Epitope Mapping

Identifying neutralizing epitopes is important for developing vaccines and inhibitors against viral pathogens. We describe a rapid method for epitope mapping, employing barcoded charged scanning mutagenesis libraries displayed on the yeast surface, and screening using flow cytometry coupled with deep sequencing. Prior scanning mutagenesis data suggest that mutations to a charged residue, such as Aspartic acid or Arginine, will be well tolerated at exposed positions of an antigen, and minimally affect protein stability and expression. Yet such substitutions at epitope residues strongly perturb binding to a cognate partner. We constructed an Aspartate scanning library of SARS-CoV-2 RBD and linked every mutation in the library to a defined unique barcode. The approach was used to map epitopes targeted in polyclonal sera of mice immunized with different SARS-CoV-2 immunogens. In contrast to complete mutational scans, charged scanning mutagenesis with the introduced barcoding strategy employs libraries with >50-fold lower diversity, facilitating library construction, screening, and downstream analysis, and also allowing for further multiplexing of samples, thus accelerating interaction site identification, as well as vaccine and inhibitor development.

biophysics↗

bioRxiv2025

Advantages of outcrossing in Plasmodium falciparum: insights from genetic crosses using fluorescent labelled parasites

Malaria parasites are obligately sexual hermaphrodite protozoans with gamete fusion occurring in the mosquito midgut, followed by meiosis and recombination. Malaria parasite populations show a spectrum of populations structures ranging from predominantly selfing to highly outcrossed. However, the fitness consequences of selfing and outcrossing for malaria parasites are poorly understood. This project was designed to investigate the dynamics of gamete fusion within the mosquito midgut and the relative fitness of selfed and outcrossed zygotes. We generated florescent-labelled clones of NF54 (mCherry), an African parasite, and NHP4026 (GFP), a Thai parasite, crossed these parasites, and scored genotypes of 8540 oocysts from 435 mosquitoes sampled from 7 to 14 days post infection. We observed decreasing proportions of outcrossed oocysts and increasing levels of inbreeding over the course of the infection in two independently replicated crosses. These results are consistent with the faster maturation of transmissible sporozoites derived from outcrossed compared with selfed oocysts. Our results suggest a substantial outcrossing advantage, perhaps because this allows for the removal of deleterious mutations accumulated during asexual parasite replication in the vertebrate host. We also found that selfed NF54 oocysts were larger than outcrossed or selfed NHP4026 oocysts, which may influence production of sporozoites and onward transmission. We conclude that fluorescent labelled parasites provide clear resolution of mating patterns, temporal dynamics and transmission potential of malaria parasites in mosquitoes. Importantly, faster maturation of outcrossed parasites can maximize levels of recombination in transmitted malaria parasite populations.

genetics↗

bioRxiv2025

The HOTAIRM1-miR-222 Axis Regulates Venetoclax Resistance and Defines a High-Risk Subset in Pediatric t(8;21) Acute Myeloid Leukemia

Although acute myeloid leukemia (AML) with the RUNX1::RUNX1T1 fusion [t(8;21)(q22;q22.1)] defines a distinct cytogenetic subtype, differences in treatment response suggest additional molecular contributors beyond chromosomal abnormalities. Deregulated hematopoietic lineage-specific long non-coding RNAs (lncRNAs) contribute to leukemogenesis and therapy resistance. To investigate their role in t(8;21) AML, we performed whole-transcriptome sequencing of pediatric patients and age-matched healthy controls, identifying significant downregulation of lncRNA HOTAIRM1, a regulator of myeloid differentiation (adjusted P < 0.05). This was confirmed in a single-cell RNA-sequencing dataset (GSE116256) and the Leukemia MILE dataset (GSE13159, P=0.03). Validation of expression in our study cohort using qPCR specifically demonstrated significant downregulation of the myeloid specific isoform, HOTAIRM1 - HM1V2 (P<0.0001). Analysis of downstream pathways activated by HM1V2 loss identified miR-222, an oncomiR, as a de-repressed target (P=0.01). Elevated miR-222 expression was observed across AML cell lines (P<0.05), leukemic stem and progenitor cells (GSE117090, P<0.05), AML plasma-derived exosomes (GSE142699, P<0.0001), the current study dataset (P<0.0001), and the TARGET AML dataset (P<0.0001). Restoring HM1V2 expression with epigenetic agents azacytidine and panobinostat induced apoptosis in venetoclax-resistant Kasumi-1 cells (P < 0.01), through suppression of miR-222 (P < 0.01) and downregulation of anti-apoptotic proteins BCL-xL and MCL-1 (P < 0.05), key mediators of the venetoclax resistance mechanism. Machine learning based feature selection and Cox regression analysis showed that high miR-222 expression predicts poor outcome in pediatric t(8;21) AML, validated in both our institutional pediatric AML cohort (P < 0.05) and the multi-institutional TARGET cohort (P < 0.0001). Together, our findings highlight an epigenetic based approach to restore isoform-specific HM1V2 pathway function in venetoclax-resistant AML cells, and identifies miR-222 as a prognostic marker to refine risk stratification within the traditionally favorable-risk t(8;21) AML subgroup. Key PointsO_LILoss of myeloid lineage specific isoform of lncRNA HOTAIRM1 - HOTAIRM1 variant 2, results in de-repression of microRNA miR-222, and contributes to venetoclax resistance in pediatric AML patients harbouring the t(8;21)(q22;q22.1)/RUNX1::RUNX1T1 fusion. C_LIO_LIMicroRNA miR-222 shows potential as a single marker predictor that complements current risk stratification by identifying a subset of pediatric t(8;21) AML patients with poor prognosis. C_LI O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=156 SRC="FIGDIR/small/663834v1_ufig1.gif" ALT="Figure 1"> View larger version (53K): org.highwire.dtl.DTLVardef@19a0931org.highwire.dtl.DTLVardef@1d1a51aorg.highwire.dtl.DTLVardef@ae6783org.highwire.dtl.DTLVardef@de78e_HPS_FORMAT_FIGEXP M_FIG C_FIG

cancer biology↗

bioRxiv2025

The role of attention in the generation of anticipatory potentials to affective stimuli: an ERP and source analysis study

Anticipatory EEG signals are characterised by the occurrence of negative slow cortical potentials. This negativity is posed to be enhanced when expecting highly emotional stimuli; however, the specific role attention plays in its generation is unclear, as emotional content is more salient and arousing, and thus recruits higher attentional resources. Here, affective anticipation signals were recorded in 35 participants with EEG, and their brain sources elucidated using multiple sparse priors algorithm. Using a cued-paradigm, the category of a sound being negatively valenced or neutral could be predicted with a 68% accuracy. To shift attentional resources away from the emotional content, participants were instructed to listen and respond to a burst of white noise that occurred on 9% of trials. Results showed slower reaction times following the aversive cue, yet no difference in EEG amplitude between aversive and neutral anticipation. Response times positively correlated with EEG amplitude - participants with stronger negativity were faster to respond. EEG source reconstruction demonstrated no differences between conditions, and showed activation of areas within the salience network including insula, somatosensory cortex, and thalamus. The current results suggest that anticipatory EEG negativity is an index of attentional resource-allocation during the anticipation period and does not reflect the emotional content of upcoming stimuli.

neuroscience↗

bioRxiv2025

Design of a Multi-Epitope Vaccine using β-barrel Outer Membrane Proteins Identified in Chlamydia trachomatis

Chlamydia trachomatis is an obligate intracellular Gram-negative pathogen responsible for causing sexually transmitted infections (STIs) and trachoma. Current interventions, including screening and antibiotics, are limited due to the widespread nature of asymptomatic infections, and the absence of licensed vaccine exacerbates the challenge. In this study, we predicted outer membrane {beta}-barrel (OMBB) proteins and designed a multi-epitope vaccine (MEV) construct using identified proteins. We employed a consensus-based computational framework on the C. trachomatis D/UW-3/CX proteome and identified 17 OMBB proteins, including well-known Pmp family members and MOMP. Eight OMBB proteins were computationally characterized, which showed significant structural homology with known outer membrane proteins from other bacteria. Sequence-based annotation tools were used to determine their putative functions. B-cell and T-cell epitopes were predicted from the selected proteins. The MEV construct was designed using four cytotoxic T lymphocyte (CTL) epitopes and 29 helper T lymphocyte (HTL) epitopes predicted from six OMBB proteins, which were conserved across 106 C. trachomatis serovars. The vaccine was supplemented at the N-terminus with Cholera enterotoxin subunit B and PADRE sequence to enhance its immunogenicity. The MEV construct of 780 amino acids was antigenic, non-allergenic, non-toxic, and soluble. Secondary structure analysis revealed 95% random coils. The 3D structural model of MEV was generated and validated, confirming its structural reliability. Molecular docking between MEV and Toll-like receptor 4 (TLR4) revealed strong and stable binding interactions, supporting its potential to elicit a strong immune response. This study highlights OMBB proteins as promising immunogenic targets and presents a computationally designed MEV candidate for C. trachomatis infection.

bioinformatics↗

bioRxiv2025

POmAb, an antibody targeting open prothrombin, results in anticoagulation without excessive bleeding in mice

Anti-prothrombin antibodies are commonly found in patients with Antiphospholipid Syndrome (APS), yet their role in clinical manifestations remains unclear. We recently identified two classes of anti-prothrombin antibodies based on their ability to recognize closed and open forms of prothrombin. Type-I antibodies bind to the open form, while Type-II antibodies bind to both forms. POmAb is a prototypical Type-I antibody that specifically targets kringle-1 of prothrombin, maintaining it in an open state. In this study, we assess the effects of POmAb in mice using the cremaster arteriole laser-induced injury model. POmAb bound mouse prothrombin and decreased thrombin generation in mouse plasma. When administered intravenously shortly before the injury, POmAb quickly accumulated on the damaged vessel wall. This accumulation significantly reduced fibrin generation with a modest effect on platelet accumulation and without causing excessive bleeding. Results obtained with POmAb offer insights into the potential roles of the anti-prothrombin antibodies in APS. They also provide proof of concept for a new class of anticoagulants that, by specifically targeting open prothrombin, could mitigate thrombosis with reduced bleeding risk.

molecular biology↗

bioRxiv2025

Investigations into genetic control of six spike traits with a focus on breeding for terminal heat stress tolerance in common wheat

To unravel the genetic architecture of six spike traits under heat stress, we used a doubled-haploid (DH) population (177 lines), developed from a cross between a heat-sensitive cultivar (PBW343) and a heat-tolerant genotype (KSG1203). This DH population and the two parents were phenotyped under timely, late, and very late sown conditions across 15 environments over three years and two locations. Best linear unbiased estimates and a high-density genetic map (5,710 SNP markers) were used for QTL mapping. A total of 51 QTL were detected in timely (17), late (10), and very late (18) sown conditions, with six common across environments with phenotypic variation explained ranging from 7.1% to 23.6%. A set of 14 stable, major QTL were validated in high-yielding DH lines and are recommended for marker-assisted recurrent selection. Several QTL co-localized with known genes responsible for important traits including grain yield (TaGW2-B1, PI1-1B). Seventy heat-responsive candidate genes associated with QTL were identified, encoding 33 proteins. A KASP marker for floret fertility QTL (QFf.ccsu-3A), and gene-based SSR markers for five key genes were developed and validated alongside Indels and SNPs. The generated genomic resources could be used in future studies and to breed heat-tolerant, high-yielding wheat varieties and germplasm. HighlightIdentified and validated stable QTL for spike traits across 15 environments, promising candidate genes for thermotolerance and spike traits, novel KASP and gene-based markers, providing genomic resources for breeding high-yielding, heat-tolerant wheat. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=153 HEIGHT=200 SRC="FIGDIR/small/662399v1_ufig1.gif" ALT="Figure 1"> View larger version (67K): org.highwire.dtl.DTLVardef@1e3cc33org.highwire.dtl.DTLVardef@734b17org.highwire.dtl.DTLVardef@146e3a5org.highwire.dtl.DTLVardef@10b9fbc_HPS_FORMAT_FIGEXP M_FIG C_FIG

genetics↗

bioRxiv2025

Frequency-dependent cellular microrheology with pyramidal atomic force microscopy probes

Atomic force microscopy (AFM) is widely used to measure the elastic properties of living cells, typically by extracting the Youngs (elastic) modulus from force-indentation curves acquired at a fixed loading rate. However, it is increasingly clear that cells relax applied stresses over time and exhibit significant frequency-dependent loss properties, a phenomenon which figures centrally in many biological problems. AFM has been employed to capture these dynamic mechanical properties via microscale frequency sweeps, where the AFM probe indents the sample in an oscillatory fashion over a range of frequencies and the oscillatory sample response is recorded. The relationship between dynamic stress and strain is then used to extract frequency-dependent microrheological properties. Despite the power of AFM oscillatory microrheology, the method remains surprisingly underutilized. This may be due to the traditional call for sphere-tipped AFM probes, which produce comparatively low indentation pressures yet are often expensive to purchase and difficult to fabricate. There remains a need for a framework in which standard blunt pyramidal AFM probes can be used "off the shelf" for oscillatory microrheology. In this study, we present such a framing. We derive expressions to extract rheological moduli from the data and explore practical experimental issues such as calibration and parameter optimization, validating the method using agarose hydrogel standards. Finally, we perform mechanical measurements on cultured cells treated with cytoskeletal inhibitors nocodazole and cytochalasin D, which yield rheological changes consistent with expected contributions of the corresponding cytoskeletal networks. Statement of SignificanceAtomic force microscopy (AFM) is often used in a force spectroscopy mode to measure an apparent elastic modulus at a single indentation rate, but in reality the mechanical behavior of cells is strongly rate-dependent, reflecting cellular viscoelasticity. While cellular viscoelastic properties can be measured by AFM oscillatory microrheology, the method traditionally calls for sphere-tipped probes, which are expensive to purchase and difficult to reproducibly fabricate. Fewer studies have used pyramidal AFM probes. Here, we examine this method in detail. We walk through data analysis expressions, compare force spectroscopy and frequency sweep measurements of a soft hydrogel, and discuss experimental setup. We then demonstrate measurement and interpretation of viscoelastic parameters in a model cell line treated with cytoskeletal inhibitors.

biophysics↗

bioRxiv2025

Unlocking the Power of CXCR2 Inhibition to Overcome Gemcitabine Resistance in Pancreatic Cancer

Pancreatic ductal adenocarcinoma (PDAC) is the fourth leading cause of cancer-related mortality, characterized by intrinsic resistance to conventional therapies and limited effective treatment options. In this study, we investigated the role of the CXCR2 axis in PDAC therapy resistance. CXCR2, a chemokine receptor, is actively involved in inflammation, tumor angiogenesis, and metastasis. Our working hypothesis is that CXCR2 contributes to PDAC chemotherapy resistance. To test this, we generated gemcitabine-resistant (GemR) lines using T3M4 and CD18/HPAF (CD18) cell lines. Baseline expression of CXCL1, CXCL5, and CXCL8 ligands were higher in GemR cells compared to parental cells. Upon gemcitabine treatment, parental cells exhibited a greater increase in CXCL1 and CXCL8 expression than GemR cells. Further analysis in T3M4 cells revealed a dose- and time-dependent increase in CXCL1 and CXCL8 expression following gemcitabine exposure. Next, we assessed whether targeting CXCR2 could enhance the therapeutic response. We treated parental and GemR cell lines with gemcitabine in combination with the CXCR2 antagonist SCH 479833 (SCH 47). Notably, lower concentrations of gemcitabine combined with SCH 47 were more effective than higher concentrations of gemcitabine alone in GemR cell lines. In both parental and GemR xenograft models, combination therapy with Navarixin (a CXCR2 antagonist) and gemcitabine demonstrated superior antitumor and antimetastatic activity compared to either treatment alone. In conclusion, these findings highlight the critical role of the CXCR2 axis in PDAC therapy resistance. Targeting CXCR2 enhances gemcitabine efficacy, offering a potential therapeutic strategy to overcome resistance in PDAC.

cancer biology↗

bioRxiv2025

CRISPR-Based Gene Dependency Screens reveal Mechanism of BRAF Inhibitor Resistance in Anaplastic Thyroid Cancer

Anaplastic thyroid cancer (ATC) is the most aggressive form of thyroid cancer. Despite recent advances in treating BRAFV600E-driven ATC, therapy resistance remains a significant challenge, often resulting in disease progression and death. Leveraging a focused CRISPR/KO screen in parallel with a CRISPR/activation screen, both tailored on response to BRAFV600E inhibitor treatment, we identified TAZ (encoded by the WWTR1 gene) deficiency as synthetically lethal with BRAF inhibitor in ATC. TAZ is overexpressed in ATC compared to well-differentiated thyroid tumors. We demonstrate that TAZ-deficient ATC cells display heightened sensitivity to BRAF inhibitors both in vitro and in vivo. Using gene essentiality score across a large panel of cancer cell lines, we found that BRAFV600E-driven cancers are highly sensitive to TAZ loss, unlike their counterparts with wild-type BRAF and non-BRAFV600E. Mechanistically, we demonstrate that dabrafenib triggers the Unfolded Protein Response (UPR) under ER stress and suppresses protein synthesis. TAZ loss represses the UPR, reverses the inhibition of protein synthesis, and triggers increased cell death by ferroptosis in dabrafenib-treated ATC. Collectively, our findings unveil TAZ as a new target to overcome resistance to BRAF inhibitors in undifferentiated thyroid cancer.

cancer biology↗

bioRxiv2025

Development of PROTACs for targeted degradation of oncogenic TRK fusions

Chromosomal translocations leading to the fusion of tropomyosin receptor kinases (TRK) with diverse partner proteins have been identified as oncogenic drivers in many adult and pediatric cancers. While first-generation TRK kinase inhibitors, such as entrectinib and larotrectinib, have shown positive responses in TRK fusion-positive cancers, resistance mutations against these inhibitors in the kinase domain limit their efficacy. Second-generation inhibitors are in clinical evaluation, highlighting a need for novel therapeutic modalities to achieve durable suppression of the oncogenic activity of TRK fusions. Here, we developed heterobifunctional small molecule degraders (PROTACs) to achieve targeted degradation of TRK fusions. By conjugating entrectinib to thalidomide, we identified JWJ-01-378 as a potent and selective CRBN-recruiting degrader of the TPM3-TRKA fusion. JWJ-01-378 induced TPM3-TRKA degradation through the ubiquitin-proteasome system and proteomics analysis confirmed the acute selectivity of JWJ-01-378 for achieving TPM3-TRKA degradation with minimal off-target effects. While JWJ-01-378 was also able to degrade wild-type TRK, it was unable to degrade TRK inhibitor resistant mutants and ALK fusions. Importantly, TPM3-TRKA degradation by JWJ-01-378 suppressed downstream signaling and reduced cancer cell viability, with improved responses compared to heterobifunctional control compounds that cannot degrade TPM3-TRKA. Together, our study expands the toolbox of compounds for evaluating targeted degradation of TRK fusions in cancer. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=97 SRC="FIGDIR/small/660465v1_ufig13.gif" ALT="Figure 13"> View larger version (19K): org.highwire.dtl.DTLVardef@6c78f7org.highwire.dtl.DTLVardef@179dcdeorg.highwire.dtl.DTLVardef@1939988org.highwire.dtl.DTLVardef@144c718_HPS_FORMAT_FIGEXP M_FIG C_FIG JWJ-01-378 recruits cereblon (CRBN) to induce potent and selective degradation of oncogenic TRK fusions, leading to a collapse in downstream signaling and loss of cancer cell viability. Graphical abstract was created using Biorender.com.

cancer biology↗

bioRxiv2025

Data Representation Bias and Conditional Distribution Shift Drive Predictive Performance Disparities in Multi-Population Machine Learning

Machine learning frequently encounters challenges when applied to population-stratified datasets, where data representation bias and data distribution shifts substantially impact model performance and generalizability across different population groups. These challenges are well illustrated in the context of polygenic prediction for diverse ancestry groups, and the underlying mechanisms are broadly applicable to machine learning with population-stratified data across domains. Using synthetic genotype-phenotype datasets representing five continental populations, we evaluate three approaches for utilizing population-stratified data, mixture learning, independent learning, and transfer learning, to systematically investigate how data representation bias and distribution shifts influence multi-population machine learning. Our results show that conditional distribution shifts, in combination with data representation bias, significantly influence machine learning performance across diverse populations and the effectiveness of transfer learning as a disparity mitigation strategy, while the effect of marginal distribution shifts is limited. The joint effects of data representation bias and distribution shifts demonstrate distinct patterns under different multi-population machine learning approaches, providing critical insights for the development of effective and equitable machine learning models for population-stratified data.

bioinformatics↗

bioRxiv2025

Multi-omics analysis uncovers differential growth kinetics of Japanese Encephalitis Virus in human and porcine macrophages

Japanese encephalitis virus (JEV), a mosquito-borne neurotropic virus, poses a significant threat to juvenile humans, while causing mild symptoms in pigs, except piglets. Pigs with high viremia are amplifier hosts, whereas humans with lower viremia are dead-end hosts. The molecular mechanisms behind the varying viremia levels in pigs and humans remain poorly understood. A comparative multi-omics study was conducted on JEV-infected macrophages, the early site of JEV replication in both species. We identified 24 key determinants potentially regulating viral growth. Functional validation of three antiviral factors - LGALS1, ssc-miR-133a-3p, and hsa-miR-3614-5p, and two proviral factors - MIF and VMP1, in macrophages showed that the upregulation of MIF and VMP1, and downregulation of ssc-miR-133a-3p, enhanced JEV replication in pigs. In contrast, the downregulation of MIF and VMP1 and upregulation of hsa-miR-3614-5p reduced JEV replication in human macrophages. These findings provide critical insights into JEV infection and its species-specific dynamics.

microbiology↗