Biology subjects

Wang, B.

Publications and source records attributed to Wang, B..

At least 109 records · Page 6Linked to original sources

bioRxiv2026

Mesolimbic dopamine signaling mediates increased hedonic feeding and food seeking in lactating mice

Lactation dramatically increases energy intake to support milk production and care for the offspring. However, the behavioral and neural circuit mechanisms driving heightened feeding during lactation remain unclear. Here, we reveal that lactation increases food-seeking behavior and enhances palatable food intake in mice. Fiber photometry recordings demonstrate increased dopamine release in the nucleus accumbens in lactating animals during feeding tasks. This elevated dopamine signaling is ultimately required for promoting both food seeking and palatable food intake during lactation as pharmacological inhibition of dopamine receptors or chemogenetic inhibition of VTA dopamine neurons both reduce food seeking and palatable food intake in lactating mice to non-lactating levels. Further, selective inhibition of dopamine receptors in the nucleus accumbens produces similar results. Together, these findings provide a circuit basis mediating elevated food seeking and palatable food intake during lactation, providing novel insights into the regulation of maternal energy balance and feeding behavior. HighlightsLactation increases food seeking and hedonic feeding in mice Mesolimbic dopamine levels are enhanced during feeding in lactating mice VTA dopamine neurons mediate increased palatable food intake and food seeking during lactation VTA-NAc dopamine transmission mediates increased food seeking and palatable food intake during lactation

neuroscience↗

bioRxiv2026

A Non-Canonical Role of SMAD4 in Regulating 3D Genome Architecture to Inhibit Lung Squamous Cell Carcinoma Development

Lung squamous cell carcinoma (LUSC) lacks clearly defined key drivers and effective targeted therapies, reflecting an incomplete understanding of its molecular pathogenesis. Here, we identify SMAD4 as a critical regulator of three-dimensional (3D) genome organization in LUSC and uncover a mechanistic link between tumor suppressor loss and oncogenic transcriptional activation. By integrating clinical datasets, genetically engineered mouse models, human and murine LUSC cell lines, and multi-omics analyses, we demonstrate that SMAD4 deficiency promotes LUSC progression by unleashing EP300-mediated enhancer-promoter looping at the SOX2 locus. Mechanistically, SMAD4 does not directly bind SOX2 regulatory elements but instead constrains chromatin looping by sequestering EP300 away from loop anchor regions. Loss of SMAD4 leads to enhanced H3K27ac deposition, aberrant SOX2 activation, and increased LUSC tumor cell proliferation. Together, these findings reveal a non-canonical role for a transcription factor (e.g., SMAD4) in regulating dysregulated 3D genome architecture to inhibit tumor development.

cancer biology↗

bioRxiv2026

Polyclonal-Monoclonal Transition in Lung Squamous Cell Carcinoma Evolution

The evolutionary trajectory of lung squamous cell carcinoma (LUSC) remains poorly defined, hindering the development of effective therapies. By integrating genomic and transcriptomic sequencing from human LUSC specimens, we delineated a polyclonal-to-monoclonal evolutionary trajectory during LUSC progression. This evolutionary pattern was corroborated by single-cell RNA sequencing, which revealed consistent tumor cell heterogeneity. Specifically, the SBS5 mutational signature was enriched and correlated with poor prognosis independent of tumor stage. By further utilizing the spontaneous LUSC mouse model to identify key genomic and genetic events in LUSC progression, we observed that the JNK pathway was inhibited and that cytoskeleton-related pathways were dysregulated during LUSC development, and identified the mutations in the JNK pathway (e.g., DACT1) and cytoskeletal regulators (e.g., KIF26A). Collectively, these findings established a polyclonal-monoclonal evolution paradigm for LUSC, potentially regulated by JNK pathways, which could benefit LUSC precision therapeutics.

genomics↗

bioRxiv2026

Complete genome-derived metabolic interactions reveal impact of gut ecology on human health

Metabolic interactions govern gut microbiome assembly, yet functional rules remain obscured by genomic incompleteness and fragmentation. Here, we leverage 1,150 complete genomes to construct genome-scale metabolic models, demonstrating that draft genomes introduce systemic artifacts missing critical transport functions. We observe that metabolic interactions are intrinsically driven by genomic traits and niche specialization rather than random association. We stratifies strains into four ecological groups, including active players, resource predators, utilizers, and contributors, each of which exhibits distinct metabolite exchange/competition signatures. This functional architecture is strongly coupled to secondary metabolism. Applying this framework to inflammatory bowel disease revealed distinct temporal dynamics of ecological groups across disease subtypes, where group-specific dysbiosis predicted clinical phenotypes better than whole community profiles. Furthermore, integrating metabolic and co-occurrence networks to identify keystone for classifiers as features significantly improved disease diagnosis. Our study provides novel insights into gut metabolic interaction landscape, bridging the gap between genomic potential and clinical utility for precision microbiome therapeutics.

microbiology↗

bioRxiv2026

EvoRMD: Integrating Biological Context and Evolutionary RNA Language Models for Interpretable Prediction of RNA Modifications

RNA modifications are essential regulators of post-transcriptional gene expression, influencing RNA stability, localization, translation, and degradation. Determining the specific modification at a given nucleotide is therefore critical for understanding its regulatory role. Most computational approaches treat each modification type as an independent binary task. This strategy provides a macro-level statistical perspective, but it does not reflect that, under a defined biochemical or cellular condition, only one modification type can occur at a specific site. Current mapping assays also report a single observed modification per site, leaving all other types unlabeled rather than truly negative. These properties motivate a framework that can reason over competing modification types. We introduce EvoRMD, a unified model for biologically contextualized and interpretable prediction of RNA modification types. EvoRMD integrates contextual sequence embeddings from a large-scale RNA language model with structured biological metadata--including species, organ, cell type, and subcellular localization. A lightweight attention mechanism highlights informative sequence positions. A shared multi-class classifier then generates a context-conditioned plausibility distribution over eleven modification types (Am, Cm, Um, Gm, D, pseudouridine, m1A, m5C, m5U, m6A, m7G), consistent with the single-positive, multiple-unlabeled nature of existing datasets. Although trained in a multi-class setting, EvoRMD also produces calibrated multi-label predictions through sigmoid-transformed logits, enabling direct comparison with existing single-modification and multi-label methods. EvoRMD achieves strong predictive performance and offers interpretable insights through attention profiles and motif analyses. Together, these components establish a biologically grounded framework for identifying and prioritizing RNA modification types from sequence and context.

bioinformatics↗

bioRxiv2026

Complete Genomes of Cultivated Gut Bacteria Reveal Mobile Genetic Element-Driven Functional Diversity with Therapeutic Implications

Cultivated bacterial isolates are essential for elucidating gut microbiota functions, yet reliance on fragmented draft genomes and limited mobile genetic element (MGE) annotation has constrained high-resolution analyses. Here, we present the Cultivated Complete Genome Reference (CCGR), a comprehensive compendium of 1,150 fully circularized human gut bacterial genomes. Overcoming draft limitations, CCGR achieves chromosome-scale resolution and reveals distinct evolutionary strategies shaping gene topological organization in response to bacterial growth. Furthermore, we decode the dynamic landscape of MGEs, demonstrating that broad-host-range phages follow strict positional patterns by integrating at conserved chromosomal hotspots flanked by nutrient transporter loci. Importantly, we provide mechanistic in vivo evidence that accessory genomic elements drive strain-level functional heterogeneity. Specifically, the plasmid-encoded scrK gene in Levilactobacillus brevis involved in fructose catabolism is required to mitigate fructose-diet-exacerbated colitis. Collectively, CCGR establishes a framework linking chromosomal architecture and MGEs to bacterial evolution and host health, enabling precise, strain-resolved functional investigations.

microbiology↗

bioRxiv2026

X-Cell: Scaling Causal Perturbation Prediction Across Diverse Cellular Contexts via Diffusion Language Models

Causal models of cellular systems hold the promise to empower broad biological discovery, including the systematic identification of novel targets for drug discovery. Predicting how genetic and pathway perturbations reshape gene expression across diverse cellular contexts is a prerequisite for building generalizable cellular foundation models. However, current methods typically fail to extrapolate beyond their training distributions because they rely predominantly on observational expression atlases rather than interventional perturbation data. We present X-Atlas/Pisces, the largest genome-wide CRISPRi Perturb-seq compendium to date, comprising 25.6 million perturbed single-cell transcriptomes across 16 biologically diverse contexts, including widely used cell lines, induced pluripotent stem cells (iPSCs), resting and CD3/CD28 activated Jurkat T lymphoma cells, and multi-lineage differentiating iPSCs. Leveraging this resource, we develop X-Cell, a diffusion language model that predicts perturbation responses by iteratively refining control-to-perturbed state transitions through cross-attention to multi-modal biological priors derived from natural language, protein language models, interaction networks, genetic dependency maps, and morphological profiles. X-Cell outperforms existing state-of-the-art models by up to five-fold on key metrics such as Pearson{Delta} (correlation between predicted and observed perturbation-induced log-fold changes), and demonstrates zero-shot prediction of T cell inactivating perturbations in stimulated Jurkat cells. We scale X-Cell to 4.9 billion parameters (X-Cell-Ultra), the largest causal perturbation model to date. We demonstrate for the first time that perturbation prediction follows power-law scaling with an exponent matching large language models. X-Cell-Ultra demonstrates zero-shot generalization to novel biological contexts, including unseen iPSC-derived melanocyte progenitors and primary human CD4+ T cells from multiple donors, and outperforms all baselines after self-supervised test-time adaptation. These results demonstrate that coordinated scaling of causal perturbation data and model capacity yields foundation models capable of generalizable perturbation prediction across cellular contexts, with potential applications for improving computational target identification, validation, and context-specific therapeutic prioritization.

systems biology↗

bioRxiv2026

BioReason-Pro: Advancing Protein Function Prediction with Multimodal Biological Reasoning

Protein function annotation is fundamental to understanding biological mechanisms, designing therapeutics, and advancing biomedical research. Current computational methods either rely on shallow sequence similarity or treat function prediction as isolated classification tasks, failing to capture the integrative reasoning across sequence, structure, domains, and interactions that expert biologists perform to infer function. We introduce BioReason-Pro, the first multimodal reasoning large language model (LLM) for protein function prediction that integrates protein embeddings with biological context to generate structured reasoning traces. A key input into BioReason-Pro is the set of GO term predictions made by GO-GPT, our autoregressive transformer that captures hierarchical and cross-aspect dependencies of GO terms. BioReason-Pro is trained via supervised fine-tuning on synthetic reasoning traces generated by GPT-5 for over 130K proteins and further optimized through reinforcement learning. It achieves 73.6% Fmax on GO term prediction and an LLM judge score of 8/10 on functional summaries, substantially outperforming previous methods. Evaluations with human protein experts show that BioReason-Pro annotations are preferred over ground truth UniProt annotations in 79% of cases. Remarkably, BioReason-Pro predicted a novel interaction partner for the renal cancer biomarker RCDG1, which we confirmed in the lab by co-immunoprecipitation. In other binding-partner predictions, its per-residue attention localized to the exact contact residues resolved in cryo-EM structures. Together, GO-GPT and BioReason-Pro establish a framework for protein function prediction that combines precise ontology modeling with interpretable biological reasoning.

molecular biology↗

bioRxiv2026

Exploration of the structural and functional diversity in the metamorphic RfaH subfamily

RfaH, a member of the universally conserved NusG/Spt5 family, activates transcription and translation of virulence genes by bridging the transcribing RNA polymerase to the translating ribosome. Unlike its monomorphic, constitutively active paralog NusG, RfaH fold-switches from an autoinhibited state, wherein its C-terminal domain forms an -helical hairpin bound to the N-terminal domain, into an active NusG-like state through C-terminal domain dissociation and refolding into a {beta}-barrel. RfaH has been proposed to evolve from NusG via gene duplication by first activating genes adjacent to its production site, before developing autoinhibition to selectively control distant genes. Using AlphaFold2, we predicted the structures of thousands of RfaH homologs and identified a subset ([~]14%) that appear predominantly folded in the active state. In vivo assays confirmed that these putative monomorphic homologs exhibit constitutive activity comparable to known RfaH mutants. Phylogenetic and genomic analyses revealed that these monomorphic proteins form a distinct clade and are preferentially located within or next to long virulence-associated operons. Together, these results further support a stepwise evolutionary model of RfaH specialization through structural transformation.

biochemistry↗

bioRxiv2026

Multiscale conformational sampling of multidomain fusion proteins by a physics informed diffusion model

Multidomain fusion proteins, such as bispecific antibodies, rely on highly flexible linker regions for their therapeutic efficacy. Characterizing these vast conformational ensembles is crucial for rational drug design; however, while all-atom molecular dynamics (MD) is the traditional gold standard, its immense computational cost makes simulating large-scale domain motions prohibitive. Recently, deep generative diffusion models have emerged as a rapid alternative for sampling protein dynamics. Yet, being trained primarily on massive databases of structured, static domains, these generic models often lack the biophysical constraints required to thoroughly sample the large-scale dynamics of highly flexible multidomain architectures. To overcome this, we leverage microsecond MD trajectories of a multidomain protein construct with various linkers to train a multiscale diffusion framework utilizing an Equivariant Graph Neural Network (EGNN). To efficiently model the dynamics of the large molecular complexes, we employ a coarse-grained spatial graph that condenses rigid domains into center-of-mass anchors while preserving explicit backbone resolution for the flexible linker. By further integrating foundational rules in biophysics directly into both the training objective and the inference process, our model generates high-fidelity conformational ensembles that reproduce the thermodynamic distributions of long-timescale MD. This physics-informed approach provides a mathematically stable, highly scalable platform for the rapid multiscale characterization of flexible biologics, significantly accelerating the rational design of fusion protein therapeutics.

bioinformatics↗

bioRxiv2026

PhosSight: a Unified Deep Learning Framework Boosting and Accelerating Phosphoproteome Identification to Enable Biological Discoveries

Protein phosphorylation is a key regulator of signaling, with mass spectrometry (MS) based phosphoproteomics serving as the premier technology for its analysis. However, phosphorylation profiling is hindered by acquisition biases: Data-Dependent Acquisition (DDA) suffers from stochastic undersampling and missing values, while Data-Independent Acquisition (DIA) faces computational bottlenecks and inefficiencies from vast spectral libraries. We present PhosSight, a unified deep learning framework designed to augment identification depth and accelerate search efficiency. PhosSight features PhosDetect, a model that explicitly encodes phosphorylation specific physicochemical features to accurately predict peptide detectability. For DDA, PhosSight leverages predicted retention time, fragment intensity, and detectability to refine site localization and rescoring, recovering marginal, low-abundance spectra. For DIA, PhosSight utilizes detectability-guided library pruning to remove non-detectable noise, accelerating search speeds without compromising sensitivity. Benchmarking on synthetic and real world datasets confirms PhosSights superior performance in both modes. Applying PhosSight to a large-scale Uterine Corpus Endometrial Carcinoma (UCEC) cohort significantly reduced missing values and expanded the quantifiable phosphoproteome. This enhanced completeness enabled the discovery of novel prognosis associated kinase targets, such as MARK2, underscoring PhosSight as a powerful tool for biological discovery in precision oncology.

bioinformatics↗

bioRxiv2026

Nar1 binds the cytosolic iron sulfur cluster assembly targeting complex via a bipartite interaction interface

The cytosolic iron-sulfur cluster assembly (CIA) pathway maturates essential nuclear and cytosolic Fe-S proteins required for genome maintenance and cellular metabolism. Nar1 (also called CIAO3 or IOP1) is a conserved Fe-S protein that connects the early and late steps of the CIA pathway, yet the molecular basis for its proposed function as a metallocluster carrier remains poorly defined. In particular, the interactions responsible for Nar1 recruitment to the CIA targeting complex (CTC) during cluster delivery remain unknown. Here, we define the molecular basis for Nar1 recruitment to the CTC using biochemical reconstitution, quantitative protein-protein interaction assays, and AlphaFold modeling. Our data reveal that Nar1 binds the CTC through two distinct interfaces. A primary interface comprises an electrostatic interaction that anchors Nar1 to a conserved acidic surface on the Cia1 subunit of the CTC and a secondary interface involves binding of Nar1s divergent targeting complex recognition peptide at the Cia1-Cia2 interface. Thus, Nar1 engages a conserved CTC surface that serves as a recruitment platform for multiple binding partners, including CIA clients. Computational structural models position the putative Fe-S cluster donor site of Nar1 adjacent to a proposed acceptor site on Cia2, suggesting that this bipartite binding mechanism positions Nar1 for transfer of an Fe-S cluster to the targeting complex. Together, these findings resolve conflicting models for Nar1 recruitment and establish a mechanistic framework for understanding how the CTC engages multiple binding partners during cytosolic iron-sulfur protein maturation.

biochemistry↗

bioRxiv2026

MERFISH 2.0, an ultra-sensitive single-cell spatial transcriptomics imaging chemistry across diverse tissue types

Spatial transcriptomics has emerged as a transformative approach for elucidating tissue architecture, cellular heterogeneity, and disease mechanisms by preserving the spatial context of gene expression in cells. Despite these advances, many spatial transcriptomic methods underperform in archival or clinically relevant specimens, particularly formalin-fixed, paraffin-embedded (FFPE) tissues, where RNA degradation and crosslinking hinder transcript detection. To address these challenges, we developed Multiplexed Error Robust Fluorescence In Situ Hybridization 2.0 (MERFISH 2.0), an optimized spatial transcriptomic imaging chemistry to enhance profiling of fragmented and highly crosslinked RNA. Across diverse human and mouse tissues preserved as fresh-frozen, fixed-frozen, and FFPE specimens, MERFISH 2.0 substantially increased transcript detection sensitivity by up to [~]8-fold relative to MERFISH 1.0 while preserving quantitative concordance (Pearson r [≥] 0.8 across tissues). In archived fresh-frozen human brain samples, MERFISH 2.0s enhanced sensitivity improved transcript recovery, enhanced cell type resolution and spatial analyses. In low quality archival FFPE human breast cancer specimen, MERFISH 2.0 revealed additional cell populations, novel cell clusters, refined tumor-immune architecture, and increased detection of gene-gene and cell-cell interactions relative to MERFISH 1.0, underscoring the impact of improved sensitivity on downstream spatial analysis. By substantially expanding robust transcript detection to degraded and archival samples, MERFISH 2.0 enables scalable, cohort-level spatial transcriptomic analysis across clinically relevant tissue collections.

genomics↗

bioRxiv2026

Limited predictability of tree-level responses to drought across European forests

Climate change is increasing the frequency, duration and severity of extreme events such as heatwaves and droughts, pushing trees near or beyond their ecophysiological limits. Understanding what governs variability in how trees respond to drought - such as intrinsic factors related to their size, age, and species, or extrinsic factors shaped by their local competitive environment - is critical for predicting long-term forest resilience to climate change and developing climate-smart forest management strategies. Here, we use tree ring data from 2909 trees belonging to sixteen species distributed across Europes major forest types to comprehensively assess what factors contribute most to a trees ability to withstand and recover from extreme drought events. We found that trees with larger living crowns generally exhibited higher post-drought growth recovery and resilience, while trees exposed to lower drought intensities showed greater resistance. Conversely, neither the density nor the diversity of a trees local competitive neighbourhood had any clear influence on its response to drought. More generally, we found that our ability to predict whether a tree would exhibit resilience to drought was low (R2 = 13-21) and was largely driven by species-specific responses and topographic variation across forest types, rather than by tree- and stand-level attributes. These findings highlight that drought responses are inherently complex and strongly influenced by forest type and by heterogeneous responses among species. Integrating tree-ring, physiological, and remote-sensing data with mechanistic models represents a promising avenue for improving forecasts of future forest resilience to climate change.

ecology↗

bioRxiv2026

CBASS limits bacteriophage production while maintaining cell viability in Pseudomonas aeruginosa

CBASS is an immune pathway that recognizes phage infection and generates cyclic nucleotide signals, which directly activate effectors that stop phage replication. Membrane-acting effectors are proposed to induce cell death to prevent phage replication; however, this mechanism has not been assessed with endogenous expression levels of the effector. We therefore sought to determine the cell viability outcomes of the CBASS phospholipase effector (CapV) upon activation with 3,3-cGAMP signals in Pseudomonas aeruginosa. Here, we surprisingly observe that constitutive 3,3-cGAMP signaling from the synthase (CdnA) enables robust cell growth and viability while effectively abolishing phage production in a CapV-dependent manner. Exogenous 3,3-cGAMP also enhances CBASS anti-phage activity and cell growth. Moreover, constitutive activation of the CapV effector induces no cell fitness cost, and blocks replication of many, but not all, phages. This demonstrates that a cyclic nucleotide-activated CBASS effector possess a degree of phage specificity that has been previously overlooked. When CBASS is active, phage transcription and initiation of DNA replication proceed normally, but phages do not reach maximum DNA levels and fewer mature virions are produced. Based on these findings, we propose that CapV interferes with the early stages of phage capsid assembly at the cell membrane and resultantly disrupts DNA packaging. Collectively, we demonstrate that a successful CBASS response antagonizes a late-stage of the phage replication cycle while maintaining cell viability.

microbiology↗

bioRxiv2026

Selective targeting of a histone-like silencer Sfx to the R6K conjugal transfer operon

Conjugative plasmids drive bacterial evolution and antibiotic resistance spread, yet their gene expression must be silenced to protect the host. A histone-like protein H-NS represses many mobile and sedentary xenogenes but fails to silence the conjugal transfer vir operon of R6K, a prototype IncX plasmid. Instead, R6K encodes its own H-NS homolog, Sfx, to repress the vir operon. Here, we show that, unlike other plasmid silencers that target promoters, Sfx cooperates with Rho factor to arrest transcription elongation. ChIP-seq reveals that despite sharing similar DNA motifs and a preference for negative supercoiling, Sfx and H-NS occupy distinct niches: Sfx binds weakly to the chromosome but is enriched on the R6K vir operon, from which H-NS is excluded. We hypothesize that this selective targeting is mediated by Sfx-vir interactions and phase separation. We show that Sfx binding to vir DNA critically depends on DNA topology but not on the target location. Our results suggest that Sfx phase separates with R6K to ensure its preferential recruitment to the plasmid DNA and forms stable nucleoprotein filaments that are impermeable to competitors. These findings reveal how histone-like proteins can partition the genome into distinct regulatory niches, a strategy likely mirrored across all life. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=81 SRC="FIGDIR/small/707533v1_ufig1.gif" ALT="Figure 1"> View larger version (29K): org.highwire.dtl.DTLVardef@1ee08a9org.highwire.dtl.DTLVardef@7ab423org.highwire.dtl.DTLVardef@1c89543org.highwire.dtl.DTLVardef@1bfa132_HPS_FORMAT_FIGEXP M_FIG C_FIG

molecular biology↗

bioRxiv2026

Induction and regulation of a reversible form of suspended animation in C. elegans

Suspended animation, a state of extreme quiescence with microscopically invisible movement and development, is a remarkable yet poorly understood stress resilience strategy in animals. Here, we describe a newly discovered form of suspended animation inducible by high-population density in isosmotic liquids in C. elegans throughout larval development and adulthood. Transcriptomic, metabolomic and live-cell activity reporter imaging analyses reveal striking molecular and cellular landscape changes by such liquid-induced suspended animation (LISA), including remodeling of gene expression programs, energy metabolites, lysosomal and mitochondrial morphology. Genetic screens identify mutants with altered stress responses and survival against LISA. While key endo-lysosomal regulators promote survival during LISA, organelle remodeling and a neuronal axis via downstream neuropeptide and cAMP/PKA signaling orchestrate behavioral awakening from LISA. Our findings define a facile paradigm for reversible SA, providing a powerful model system to uncover key molecular and cellular mechanisms governing an extreme case of reversible life arrest and dormancy.

developmental biology↗

bioRxiv2026

QuantCell: machine learning based cell annotation from qualitative and quantitative imaging profiles

Recent advances in spatial omics enable high-resolution, multiplexed imaging of RNA and protein expression, but cell annotation remains challenging, particularly in complex tissues with numerous markers or rare cell types. Here, we present QuantCell, a machine learning framework that leverages quantitative imaging data to improve annotation derived from qualitative profiles. QuantCell evaluates multiple models and applies a user-defined false discovery rate to ensure high-confidence annotation. Using PhenoCycler imaging of mouse bone marrow, QuantCell increased annotated cells from 33.1% to 90.2% at 5% FDR, achieving 96.5% accuracy. QuantCell supports diverse imaging platforms and robustly detects rare cell populations.

bioinformatics↗