Biology subjects

Gao, G.

Publications and source records attributed to Gao, G..

At least 19 recordsLinked to original sources

bioRxiv2026

TangentXY: Improved copy-number estimation of sex chromosomes reveals features associated with aneuploidies in chromosomes X and Y across cancers

Signal-to-noise ratios in copy-number estimates across genomic loci improve with the use of large panels of germline reference data. However, in the case of sex chromosomes, germline samples differ between males and females (males have one X chromosome (chrX) and one Y chromosome (chrY), while females have two chrXs), limiting the size of the reference sets to those that match the sex of the index sample. Frequent mosaic loss of chrY in normal male tissue also leads to biases when not accounted for. Here we report TangentXY, a copy-number normalization method that uses both male and female reference samples, correcting for sex differences in chrX and accounting for chrY mosaicism, to assess sex chromosome copy numbers for samples of either sex. We show that TangentXY maintains sex-specific differences in chrX and chrY while minimizing noise. We also used TangentXY to analyze associations between copy-number alterations in sex chromosomes and cancer characteristics, and found that losses in the chrX but not chrY are correlated with TP53 mutation, that gain of chrX can predict patient prognosis in certain tumor types, and that the vast majority of gains of chrX in males remain active, whereas gains of chrX in females are divided approximately equally between active and inactive chrXs. We also found that loss of chrY correlates with inactivating driver mutations in escape genes in chrX, suggesting a mechanism by which its frequent loss promotes cancers in males.

cancer biology↗

bioRxiv2026

Proteoform Barcode: An Intuitive Visualization Framework for Top-Down Proteomics

Top-down proteomics (TDP) advances biomedical research by providing a birds-eye view of proteoforms in cells, tissues, and biofluids. Thousands of proteoforms can be characterized using well-established TDP technologies, and potential proteoform biomarkers of diseases have been discovered. However, there is a lack of an easy and biologically informative approach to present the quantitative global TDP data. Here, we present proteoform barcode as a straightforward visualization approach that simultaneously displays proteoform abundance and their associated Gene Ontology (GO) biological processes, converting a list of proteoforms to a biologically informative image. The proteoform barcode allows 1) a global view of proteoforms (i.e., relative abundance and functional information) in complex biological systems (i.e., bacteria, yeast, human cells, and human plasma) and 2) the accurate distinction of samples in diverse biological conditions (i.e., control and disease) assisted by machine learning approaches. The proteoform barcode, assisted by the random forest model, accurately separated the human plasma samples of healthy controls and early-stage breast cancer. The data demonstrates the high potential of the proteoform barcode-based approach for early diagnosis of diseases in an easy and biologically informative manner.

systems biology↗

bioRxiv2026

Prototype-based AI triage for 3D pathology

Non-destructive 3D pathology enables high-resolution slide-free imaging of intact clinical specimens, providing comprehensive visualization of tissue structures beyond what conventional slide-based 2D histopathology can provide. However, the scale and complexity of volumetric datasets make exhaustive manual review impractical, motivating AI-assisted triage methods to select a small number of high-risk 2D slices for pathologist review. While prior triage models have shown promise, interpretability is poor and performance can be suboptimal, especially in the nascent field of 3D pathology in which labeled data is limited. We present SCOPE, a Segmentation-guided CrOss-slice PrototypE learning framework for comprehensive risk assessment of 2D levels within 3D pathology datasets. SCOPE combines (i) clustering-based pretraining on large-scale unlabeled volumetric data to initialize morphology-aware prototypes, (ii) segmentation-derived structural priors from publicly available models to guide proto-type learning, and (iii) cross-slice (2.5D) prototype aggregation across neighboring slices to generate slice-level risk predictions. In prostate and esophageal data cohorts, SCOPE consistently outperforms attention-based and prototype-based multiple instance learning baselines for both binary and multiclass prediction tasks, enabling depth-resolved risk profiling for 3D triage based on morphological prototypes that are interpretable to pathologists.

pathology↗

bioRxiv2026

Co-delivered PD-L1 rescues the protective efficacy mediated by an AAV-expressed HIV-1 bNAb

Adeno-associated virus (AAV)-delivered anti-HIV-1 broadly neutralizing antibodies (bNAbs) could prevent and treat HIV-1 infection but are limited by host immune responses, specifically anti-drug antibodies (ADA). We tested whether PD-L1-mediated immune shielding could improve the consistency of AAV-delivered bNAb expression from muscle tissue in rhesus macaques. AAV9.PD-L1 co-delivery with AAV9.3BNC117 reduced the occurrence of ADA and T cell responses and improved the durability of 3BNC117 expression for one year post administration. Importantly, 5 of 6 macaques that received co-delivered AAV9.PD-L1 vectors were protected against ten repeated SHIVAD8-EO challenges. Histopathological and spatial transcriptomic profiling showed that AAV9.PD-L1 co-delivery prevented severe local inflammation, muscle injury, and tertiary lymphoid structure formation at the administration site. Thus, immune shielding could serve as a strategy to prolong transgene expression from muscle-directed AAV-delivered biologics.

immunology↗

bioRxiv2026

Overcoming host immune responses to an AAV-delivered HIV-1 bNAb in rhesus macaques mediated by co-delivery of PD-L1

Adeno-associated virus (AAV)-delivered anti-HIV-1 broadly neutralizing antibodies (bNAbs) have demonstrated promise for preventing and treating HIV-1 infection in preclinical models. However, host immune responses, specifically anti-drug antibodies (ADA), limit sustained bNAb expression. We have previously shown that PD-L1-mediated immune shielding improves the consistency of AAV-delivered bNAb 3BNC117 expression from muscle tissue in rhesus macaques. Here, we test the breadth of this approach with another bNAb, 10-1074. We show that AAV9.PD-L1 co-delivery with AAV9.10-1074 reduced the occurrence of ADA responses and improved the durability of bNAb expression for one year post administration. Notably 12 of 12 macaques that received AAV9.10-1074 vectors were protected against ten repeated SHIVAD8-EO challenges. Histopathological profiling showed that AAV9.PD-L1 co-delivery prevented severe local inflammation and tertiary lymphoid structure formation at the administration site. Thus, immune shielding could serve as a broad strategy to prolong transgene expression from muscle-directed AAV-delivered biologics.

immunology↗

bioRxiv2026

Characterization of expression elements for an AAV delivered antibody in nonhuman primates when co-delivered with PD-L1

Successful AAV-expressed antibody therapy for HIV-1 requires broadly neutralizing antibody (bNAbs) concentrations and reduced immune responses to sustain viral suppression without ART. We have previously demonstrated that co-delivery of AAV-expressed PD-L1 reduces immune responses against HIV-1 bNAbs in rhesus macaques. Here we systematically evaluated six AAV9 transgene cassettes encoding 10-1074 with different promoter/intron combinations (CMV, CMV/R, CBA, CASI, CB7, EF1) across in vitro systems, immune-deficient mice, and in rhesus macaques. We show that both promoter and species selection, leads to differences in 10-1074 concentrations with the CB7 promoter leading to greatest expression in mice and CMV/R promoter in macaques. In addition to differences observed, loss of 10-1074 serum concentrations in macaques resulted in higher anti-drug antibody responses and antigen specific IFN-y T cell responses were focused on the 10-1074 heavy-chain variable region. Furthermore, inclusion of the WPRE greatly impacted 10-1074 expression leading to higher concentrations in both mice and nonhuman primates. Lastly, circulating 10-1074 in macaques retained neutralizing activity against diverse HIV-1 pseudovirus isolates. Together these results demonstrate how expression elements influence AAV-expressed antibodies in the context of co-delivery and highlight the need for further improvements to AAV transgene cassettes when co-delivered with AAV expressed PD-L1.

microbiology↗

bioRxiv2026

The Il1r2-CreERT2 knock-in mouse enables inducible labeling of slow-cycling epidermal basal cells and genetic ablation of the IL-1 decoy receptor

The interfollicular epidermis is maintained by spatially organized basal cell populations with distinct molecular signatures and division kinetics; however, the markers that define these populations remain poorly defined. In this study, we identified Il1r2, which encodes the IL-1 decoy receptor IL-1R2, as a marker of the slow-cycling basal population in the epidermis. Single-cell RNA-seq analysis of epidermal and hair follicle basal populations in the murine tail skin revealed that Il1r2 is preferentially expressed in the slow-cycling epidermal basal population, and immunofluorescence staining confirmed its protein localization in tissues. To enable fate mapping of this population, we generated an Il1r2-CreERT2 knock-in mouse line using a CRISPR-Cas9-based PITCh method. Tamoxifen induction in Il1r2-CreERT2/Rosa-tdTomato mice exhibited selective labeling of basal cells localized to the slow-cycling interscale region of the tail epidermis. Because the CreERT2 cassette was inserted into the Il1r2 coding sequence, homozygous Il1r2-CreERT2 knock-in mice can also serve as an Il1r2 knockout model through targeted gene ablation. Thus, the Il1r2-CreERT2 mouse line provides a dual genetic tool for lineage tracing of the slow-cycling epidermal basal population and for functional modulation of IL-1 signaling in vivo.

developmental biology↗

bioRxiv2026

Durable protection against SIV challenge by adeno-associated virus delivery of Env-specific antibodies

Conventional vaccines have so far failed to elicit the types of antibodies needed for protection against HIV. As an alternative, we evaluated adeno-associated virus (AAV) delivery of rhesus macaque antibodies to the SIV envelope glycoprotein for protection against SIV challenge. AAV vectors encoding a broadly neutralizing antibody (bnAb) and an antibody that only mediates antibody-dependent cellular cytotoxicity (ADCC) were administered individually or together to separate groups of rhesus macaques. Antibody expression was sustained for more than a year with minimal anti-drug antibody responses. All animals that received a control antibody or the ADCC-only antibody became infected after five low-dose, intrarectal challenges with SIVmac239. In contrast, 14 of 16 animals that received the bnAb resisted two rounds of twelve SIVmac239 challenges more than a year apart. Thus, AAV delivery of a single bnAb can afford durable protection against a pathogenic SIV strain that is notoriously difficult to protect against by vaccination.

immunology↗

bioRxiv2026

Canonical self-supervised pretraining paradigm constrains the capacity of genomic language models on regulatory decoding

Recent studies suggest that genomic language models (gLMs) could help decode genomic regulatory code. Here, we systematically evaluated 11 representative gLMs across multiple regulatory genomics applications and found that current gLMs offer limited advantages over the random baseline. Further analysis revealed a systematic misalignment between the canonical sequence-only self-supervised pretraining paradigm and the context-specific dynamic nature of gene regulation, highlighting the need for function-oriented pretraining strategies that explicitly incorporate biochemical and regulatory priors.

bioinformatics↗

bioRxiv2026

CNS diseases cerebrospinal fluid single-cell atlas reveals immune characteristics of neuropsychiatric systemic lupus erythematosus

Neuropsychiatric systemic lupus erythematosus (NPSLE) is a potentially severe complication of systemic lupus erythematosus (SLE), yet its pathogenesis remains largely elusive. By jointly probing the immune dynamics of subjects cerebrospinal fluid (CSF) and peripheral blood, we showed that both innate and adaptive immune responses jointly contribute to the pathogenesis of NPSLE. In particular, we found the remarkable enrichment of BAM-CCL3, a subtype of border-associated macrophages with strong recruitment capacity, implicating its potential role in central nervous system (CNS) inflammation. We also observed pronounced activation of memory B cells and CD4+ regulatory T cells in NPSLE CSF, along with the preferential blood-to-CSF migration and subsequent within-CSF clonal expansion of CD8+ effector memory T cells in NPSLE patients, suggesting a persistent CNS-localized adaptive immune dysregulation. Finally, we developed the single-cell CNS disease CSF-Blood Atlas (scCDCB), a comprehensive collection for CSF and peripheral blood of multiple CNS diseases, which is publicly available at (https://sccdcb.gao-lab.org) to serve as a reference for future research on CNS diseases.

genomics↗

bioRxiv2026

Postnatal gene restoration in succinic semialdehyde dehydrogenase deficiency (SSADHD) reveals phenotype reversibility

Succinic semialdehyde dehydrogenase deficiency (SSADHD) is a rare autosomal recessive metabolic disorder due to loss-of-function ALDH5A1 mutations impairing the catabolism of {gamma}-aminobutyric acid (GABA), the major inhibitory neurotransmitter in the brain. In SSADHD, pathologic accumulation of GABA and its metabolic by-product {gamma}-hydroxybutyrate (GHB) corresponds to a clinical syndrome dominated by developmental delay and epilepsy in half of patients with risk of sudden death in adolescence and adulthood. Brain-wide ALDH5A1 gene replacement for SSADHD is unavailable, and whether such treatment will reverse the SSADHD phenotype is unknown. We developed an inducible mouse SSADHD model, Aldh5a1lox-STOP, enabling Cre-dependent Aldh5a1 restoration to evaluate gene therapy feasibility. In the absence of SSADH, Aldh5a1lox-STOP mice exhibit hyperactivity and excessive serum GHB levels, culminating in death by [~]postnatal day 22, recapitulating the severe SSADHD condition. Systemic delivery of a blood-brain barrier (BBB)-penetrating adeno-associated virus (AAV) carrying a Cre gene to Aldh5a1lox-STOP mice leads to brain-wide SSADH restoration, serum GHB level reduction, normalization of hyperactivity, and substantial increase in survival. As a step toward clinical translation, we further assessed an AAV encompassing a functional native promoter (FLnP) of ALDH5A1 tethered to its human coding sequence, namely AAV-FLnP-hALDH5A1. Aldh5a1lox-STOP mice were effectively rescued when treated with AAV-FLnP-hALDH5A1 packaged in the blood-brain barrier (BBB)-penetrating capsid PHP.eB. These findings provide preclinical proof that SSADH gene replacement therapy is feasible and potentially effective.

neuroscience↗

bioRxiv2026

Decades of dreams coming true: capillary zone electrophoresis-mass spectrometry for reproducible multi-level proteomics

Capillary zone electrophoresis (CZE)-mass spectrometry (MS) has been proposed as a powerful analytical tool for bottom-up, top-down, and native proteomics (multi-level proteomics) decades ago to analyze complex biological samples at the levels of peptides (bottom-up), proteoforms (top-down), and complexoforms (native). However, its broad adoption has been impeded by the limited robustness and reproducibility. Here, we present multi-level proteomics data from nearly 170 CZE-MS runs ([~]170 hours of instrument time), demonstrating qualitatively (i.e., the number of identified peptides and proteoforms, the number of detected complexoforms, and their migration time) and quantitatively (i.e., peptide, proteoform, and complexoform intensity) reproducible measurement of complex samples with varying levels of complexity, i.e., Escherichia coli cells, HeLa cells, and human plasma. CZE-MS-based native proteomics enabled the detection of hundreds of complexoforms up to 800 kDa from the complex systems via consuming only nanograms of protein material. The results indicate that CZE-MS is sensitive and reproducible enough for broad adoption for multi-level proteomics-based biomedical research.

systems biology↗

bioRxiv2025

The biochemical nature of senile plaques: intracellular destabilized-lysosome-enriched neural processes catabolizing extracellular hemorrhagic amyloid materials

The pathological mechanism of senile plaque formation in Alzheimers disease remains a challenging unresolved problem for over one hundred years. In this study, we investigated senile plaque pathogenesis based on immunohistochemistry, fluorescence imaging and quantitative analysis of AD front brain sections with multiple senile plaque related markers such as A{beta}, MAP2, phos-Tau, Cathepsin D, Sortilin 1, Ceruloplasmin and Hemin. Firstly, we confirmed that senile plaque A{beta} deposition mostly happened in MAP2-positive but degenerative neuritic processes. Secondly, we found an interesting dual-component granular or diffusive staining pattern of all of these senile plaque related markers, matching to two states of lysosomes, relatively-stable granule lysosomes and destabilized diffusive lysosomes as shown by lysosome-related marker Cathepsin D or Sortilin 1 staining. Thirdly, dystrophic senile plaque neural processes contain hemorrhage-related markers such as Ceruloplasmin and Hemin. In addition, A{beta} deposition, Tau phosphorylation, lysosome destabilization are all enriched at the sites of microaneurysm formation. Moreover, senile plaques are also characterized by a widespread binary phos-Tau staining pattern with difference in term of staining intensity. We concluded that senile plaques are formed by destabilized-lysosome-enriched neural processes catabolizing extracellular hemorrhagic amyloid materials, with microaneurysm as an important intermediate stage.

cell biology↗

bioRxiv2025

Identification of molecular mechanisms contributing to acquired thermotolerance by transcriptome profiling

Heat stress poses a serious threat to plant survival and productivity, and has a direct influence on crop yield stability. Plants response to high temperature is tightly controlled by complex genetic networks. Plants can be acclimated through gradual pre-exposure to increasing temperatures and that in turn causes higher survival in subsequent and otherwise lethal heat stress conditions. To investigate the physiological and molecular processes underlying heat acclimation and recovery, we examined changes in Arabidopsis thaliana transcriptome throughout the acclimation and the subsequent heat shock treatment. Groups of differentially expressed genes and enriched biological pathways that constitute the heat transcriptional memory were identified. The function of flavonoids in plant heat stress were further explored experimentally. In addition, we observed altered stomata density and aperture responses in heat acclimated plants, and this might be partially controlled by AGAMOUS-LIKE16 (AGL16) transcription factor and its negative regulator microRNA824 (miR824).

plant biology↗

bioRxiv2025

Causality-oriented regulatory inference and rational transcriptomic reprogramming design with CASCADE

Systematically elucidating the causal structure of gene expression regulation is the foundation for identifying key regulators of cellular function and designing rational interventions thereof. While advances in high-content perturbation screens enables experimental probing of causal regulatory effects, the massive genomic scale, measurement noise and confounding factors pose significant challenges for existing causal discovery and inference algorithms. Here, we present CASCADE, an end-to-end framework that integrates de novo causal discovery with prior regulatory knowledge for genome-wide causal discovery and inference. In addition to its superior performance on causal regulatory structure discovery and counterfactual inference for unseen perturbation effects, CASCADE enables, for the first time, causality-oriented rational intervention design for transcriptomic programming towards various targeted cells within a unified framework.

bioinformatics↗

bioRxiv2025

3D pathology-guided microdissection

Traditional micro- and macro-dissection techniques enable the extraction of localized regions in thin tissue sections for molecular analysis. Despite the growing use of 3D microscopy, analogous methods for volumetric microdissection are lacking. We have developed a 3D microdissection method based on computer numerical controlled (CNC) milling integrated with open-top light-sheet microscopy. We demonstrate the ability to study tumor evolution along convoluted 3D branching architectures, which is inaccessible to 2D methods.

bioengineering↗

bioRxiv2025

Seipin mediates Perilipin-1 recruitment to lipid droplets to preserve human adipocyte identity

Congenital generalized lipodystrophy type 2 (CGL2) is caused by mutations in the BSCL2 gene, which encodes the protein seipin. However, how seipin loss causes adipose tissue failure remains unclear. Using human adipocyte progenitor cells capable of robust differentiation in vitro and in vivo, we reveal two unexpected findings that redefine CGL2 pathogenesis. First, seipin is dispensable for lipid droplet biogenesis but essential for recruiting the major adipocyte scaffold protein Perilipin 1 (PLIN1) to the lipid droplet surface. Second, we discover that the integrity of the lipid droplet serves as an organelle to nucleus quality control checkpoint enforcing adipocyte identity. Without seipin-dependent PLIN1 recruitment, adipocytes exhibit enhanced lipolysis and ceramide accumulation, triggering an unexpected cellular response of de-differentiation into a progenitor-like state. From this de-differentiated state, cells can undergo additional cycles of differentiation and de-differentiation upon repeated adipogenic stimuli. However, some cells escape de-differentiation, instead forming a single large droplet and displaying severe cellular structural abnormalities. Consistent with this model, we find functional adipose tissue can form in vivo from seipin-deficient cells, yet ultimately fails. These findings resolve conflicting models of CGL2 pathogenesis by reframing seipin as a regulator of PLIN1 recruitment, rather than droplet formation per se, and reveal the fundamental role of lipid droplet integrity in the development of functional human adipocytes.

cell biology↗

bioRxiv2025

Quantifying predictability of gene expression from histology image

Histopathological images are indispensable in clinical diagnosis, yet provide limited insight into underlying molecular states. Numerous computational models attempt to predict gene expression from histopathological images. However, a fundamental question remains unresolved: which genes can be accurately predicted and which cannot. Here, we introduce Expression Predictability Score (EPS), a metric that quantifies the predictability of each gene from images through expression-image mutual information. Empirical analyses across more than 500 slices further reveal consistent sets of highly predictable and unpredictable genes, as well as their underlying association with the physicochemical nature of H&E staining.

bioinformatics↗