Biology subjects

Li, S.

Publications and source records attributed to Li, S..

At least 469 records · Page 26Linked to original sources

bioRxiv2024

Patient-Derived Xenografts of Triple-Negative Breast Cancer Enable Deconvolution and Prediction of Chemotherapy Responses

Chemotherapy regimens for triple-negative breast cancer (TNBC) combine agents without knowing which agents drive response. Consequently, predictors derived from multi-agent regimens cannot be assumed to generalize to individual drugs or other regimens, motivating development of treatment-matched predictors. Here, we used TNBC patient-derived xenografts (PDXs) treated with carboplatin, docetaxel, or the combination to deconvolute drug-specific responses and identify associated molecular features. Combination treatment rarely improved upon the best single agent, with enhanced responses in only 13% of PDXs and antagonism in a comparable fraction. Proteogenomic analyses identified high cytokeratin-5 (KRT5) as a general marker of chemotherapy responsiveness and KRT5 immunohistochemistry discriminated responsive PDXs (AUROC, 0.83). To train treatment-specific predictors, we integrated these data with independent PDX and clinical cohorts with responses assessed after anthracycline-free platinum, taxane, or platinum-taxane therapy, ensuring response corresponded to the modeled treatment. Four feature selection strategies yielded largely nonoverlapping biomarker panels converging on treatment-relevant pathways. On independent test data, RNA-based predictors of complete response (CR) to platinum-based (carboplatin or cisplatin) and taxane-based (docetaxel or paclitaxel) chemotherapy achieved AUROCs of 0.80 and 0.86, respectively. For platinum-taxane regimens (carboplatin plus docetaxel or paclitaxel), proteomic-guided feature selection generated a 10-biomarker, protein- informed RNA predictor of pathologic complete response (pCR) that outperformed RNA-only feature selection and achieved an AUROC of 0.85 in an independent clinical cohort, while retaining practicality as an RNA-based assay. Treatment-matched integration of multi-omic PDX and clinical datasets provides a framework for chemotherapy-specific predictors with clinically relevant performance, supporting biomarker-guided precision selection and treatment optimization for patients with TNBC. Statement of significanceIntegration of multi-omic data from patient-derived xenografts with treatment-matched clinical cohorts yielded three retrospectively validated predictors of platinum, taxane, and platinum+taxane response, enabling biomarker-guided chemotherapy selection for patients with triple-negative breast cancer.

cancer biology↗

bioRxiv2024

Oncogenic RAS Induces a Distinctive Form of Non-Canonical Autophagy Mediated by the P38-ULK1-PI4KB Axis

Cancer cells with RAS mutations exhibit enhanced autophagy, essential for their proliferation and survival, making it a potential target for therapeutic intervention. However, the regulatory differences between RAS-induced autophagy and physiological autophagy remain poorly understood, complicating the development of cancer-specific anti-autophagy treatments. In this study, we identified a form of non-canonical autophagy induced by oncogenic KRAS expression, termed RAS-induced non-canonical autophagy via ATG8ylation (RINCAA). RINCAA involves distinct autophagic factors compared to those in starvation-induced autophagy and incorporates non-autophagic components, resulting in the formation of non-canonical autophagosomes with multivesicular/multilaminar structures labeled by ATG8 family proteins (e.g., LC3 and GABARAP). We have designated these structures as RAS-induced multivesicular/multilaminar bodies of ATG8ylation (RIMMBA). A notable feature of RINCAA is the substitution of the class III PI3K in canonical autophagy for PI4KB. We identified a regulatory P38-ULK1-PI4KB-WIPI2 signaling cascade governing this process, where ULK1 phosphorylation at S317, S479, S556, and S758 activates PI4KB. This activation involves PI4KB phosphorylation at S256 and T263, initiating PI4P production, ATG8ylation, and non-canonical autophagy. Importantly, elevated PI4KB phosphorylation at S256 and T263 was observed in RAS-mutated cancer cells and colorectal cancer specimens. Inhibition of PI4KB S256 and T263 phosphorylation led to a reduction in RINCAA activity and tumor growth in both xenograft and KPC models of pancreatic cancer, suggesting that ULK1-mediated PI4KB-Peptide-1 phosphorylation could represent a promising therapeutic target for RAS-mutated cancers.

cell biology↗

bioRxiv2024

STAT1 Promotes PD-L1 Activation and Tumor Growth in Lymphangioleiomyomatosis

Lymphangioleiomyomatosis (LAM) is a cystic lung disease that primarily affects women. LAM is caused by the invasion of metastatic smooth muscle-like cells into the lung parenchyma, leading to abnormal cell proliferation, lung remodeling and progressive respiratory failure. LAM cells have TSC gene mutations, which occur sporadically or in people with Tuberous Sclerosis Complex. Although it is known that hyperactivation of the mechanistic target of rapamycin complex 1 (mTORC1) due to TSC2 gene mutations contributes to aberrant cell growth in LAM lung, tumor origin and invasive mechanism remain unclear. To determine molecular drivers responsible for aberrant LAM cell growth, we performed integrative single-cell transcriptomic analysis and predicted that STAT1 interacts with Pre-B cell leukemia transcription factor (PBX1) to regulate LAM cell survival. Here, we show activation of STAT1 and STAT3 proteins in TSC2- deficient LAM models. Fludarabine, a potent STAT1 inhibitor, induced the death of TSC2- deficient cells, increased caspase-3 cleavage, and phosphorylation of necroptosis marker RIP1. Fludarabine treatment impeded lung colonization of TSC2-deficient cells and uterine tumor progression, associated with reduced percentage of PCNA-positive cells in vivo. Interestingly, IFN-{gamma} treatment increased STAT1 phosphorylation and PD-L1 expression, indicating that STAT1 aids TSC2-deficient tumor cells in evading immune surveillance in LAM. Our findings indicate that STAT1 signaling is critical for LAM cell survival and could be targeted to treat LAM and other mTORC1 hyperactive tumors.

cancer biology↗

bioRxiv2024

Coverage landscape of the human genome in nucleus DNA and cell-free DNA

For long, genome-wide coverage has been used as a measure of sequencing quality and quantity, but the biology hidden beneath has not been fully exploited. Here we performed a comparative analysis on genome-wide coverage profiles between nucleus genome DNA (gDNA) samples from the 1000 Genomes Project (n=3,202) and cell-free DNA (cfDNA) samples from healthy controls (n=113) or cancer patients (n=362). Regardless of sample type, we observed an overall conserved landscape with segmentation of coverage, where adjacent windows of genome positions present similar coverage. Besides GC-content, we identified protein-coding gene density and nucleosome density as major factors influencing the coverage of gDNA and cfDNA, respectively. Differential coverage of cfDNA vs gDNA was found in immune-receptor loci, intergenic regions and non-coding genes, reflecting distinct genome activities in different cell types. A further rise in coverage at non-coding genes and intergenic regions plus a further drop of coverage at protein-coding genes and genic regions within cancer cfDNA samples indicated a loss of contribution by normal cells. Importantly, we observed the distinctive feature of coverage convergence in cancer-derived cfDNA, with the extent of convergence positively correlated to stages. Based on the findings, we developed and validated an outlier-detection approach for cfDNA-based cancer screening without the need of cancer samples for training, outperforming current benchmarks on condition-matched and condition-unmatched cancer detection tasks.

genomics↗

bioRxiv2024

The Cardioprotective Effect of Ischemic Postconditioning is Mediated by Inhibiting RAP2C-MAP4K4 Pathway

BackgroundIschemic postconditioning (PostC) serves as a vital defense for cardiomyocytes against the deleterious effects of ischemia/reperfusion (I/R) injury, the beneficial effects could be further enhanced through pharmacological strategies. Our prior research demonstrated upregulated expression of the GTP-binding protein RAP2C in H9C2 cells post hypoxia-reoxygenation (H/R). The cardioprotective effects of RAP2C and underlying mechanisms are unclear. We therefore explored the role of RAP2C in PostC-induced cardioprotection against I/R injury. MethodsOpen-chest rat I/R and primary cultured cardiomyocytes H/R models were used. RAP2C and MAP4K4 expression was detected by immunohistochemistry and Western blotting. The BioGRID and STRING databases were tapped to predict the RAP2C-MAP4K4 binding, which was confirmed by co-immunoprecipitation and immunofluorescence. ResultsResults indicated that I/R and H/R upregulated the protein levels of RAP2C, MAP4K4, phospho-JNK, phospho-P38, and phospho-ERK, concomitant with increased apoptosis. PostC mitigated these effects. The pro-apoptotic impacts and the activation of the MAPK pathway induced by H/R were attenuated by RAP2C knockdown and intensified by RAP2C overexpression. H/R increased the interaction between RAP2C and MAP4K4, and PostC attenuated this effect. MAP4K4 knockdown reduced the pro-apoptotic and MAPK-activating effects induced by both RAP2C overexpression and hypoxia/reoxygenation (H/R). ConclusionsThese results demonstrate that PostC reduces cardiomyocyte apoptosis via modulating RAP2C/MAP4K4 pathways, suggesting their potential as therapeutic targets for the treatment of ischemic heart disease.

cell biology↗

bioRxiv2024

Arterial endothelial deletion of Alk1 causes epistaxis and cerebral microhemorrhage with aberrant arteries and defective smooth muscle coverage.

Hereditary Hemorrhagic Telangiectasia (HHT) is an autosomal dominant vascular disorder with manifestations including severe nose bleeding and microhemorrhage in brains. Despite being the second most common inherited bleeding disorder, the pathophysiological mechanism underlying HHT-associated hemorrhage is poorly understood. HHT pathogenesis is thought to follow a Knudsonian two-hit model, requiring a second somatic mutation for lesion formation. Mutations in activin receptor-like kinase 1 (ALK1) gene cause HHT type 2. We hypothesize that somatic mutation of Alk1 in arterial endothelial cells (AECs) leads to arterial defects and hemorrhage. Here, we mutated Alk1 in AECs in postnatal mice using Bmx(PAC)-CreERT2 and found that somatic arterial endothelial mutation of Alk1 was sufficient to induce spontaneous epistaxis and multifocal cerebral microhemorrhage. This bleeding occurred in the presence of tortuous and enlarged blood vessels, loss of arterial molecular marker Efnb2, disorganization of vascular smooth muscle, and impaired vasoregulation. Our data suggest that arterial endothelial deletion of Alk1 leading to reduced arterial identity and disrupted vascular smooth muscle cell coverage is a plausible molecular mechanism for HHT-associated severe epistaxis. This work provides the first evidence that somatic Alk1 mutation in AECs can cause hemorrhagic vascular lesions, offering a novel preclinical model critically needed for studying HHT-associated epistaxis, and delineating an arterial mechanism to HHT pathophysiology.

pathology↗

bioRxiv2024

Stage-specific CAR-mediated signaling generates naive-like, TCR-null CAR T cells from induced pluripotent stem cells

Genetically modified, induced pluripotent stem cells (iPSCs) offer a promising allogeneic source for the generation of functionally enhanced, chimeric antigen receptor (CAR) T cells. However, the signaling of CARs during early T cell development and the removal of the endogenous T cell receptor required to prevent alloreactivity pose significant challenges to the production of mature conventional CAR T cells from iPSCs. Here, we show that TCR-null, CD8{beta} CAR T cells can be efficiently generated from iPSCs by engineering stage-specific onset of CAR expression and signaling to both permit conventional T cell development and to induce efficient positive selection. CAR T cells produced using this approach displayed a uniform, naive T cell phenotype and demonstrated superior antigen-specific cytotoxicity compared to iPSC-derived effector memory CAR T cells. Multimodal sequencing revealed CAR-mediated positive selection induced the persistent upregulation of key transcription factors involved in naive T cell development. Achieving precise control of CAR expression and signaling in developmentally sensitive T precursors will be critical to realizing the full potential for "off-the-shelf", iPSC-derived cellular therapies.

bioengineering↗

bioRxiv2024

FIDDLE: a deep learning method for chemical formulas prediction from tandem mass spectra

Molecular identification through tandem mass spectrometry is fundamental in small molecule analysis, with formula identification serving as an initial step in the process. Current computational methods often struggle with accuracy, speed, and scalability for relatively larger molecules, limiting high-throughput workflows. We present FIDDLE (Formula IDentification by Deep LEarning), a deep learning-based method trained on over 38,000 molecules and 1 million MS/MS spectra from various Quadrupole Time-of-Flight (Q-TOF) and Orbitrap instruments. FIDDLE accelerates formula identification by more than 10-fold and achieves top-1 and top-5 accuracies of 88.3% and 93.6%, respectively, outperforming state-of-the-art methods based on top-down (SIRIUS) and bottom-up (BUDDY) approaches by over 10%. On external metabolomics datasets, FIDDLE achieves top-5 accuracies of 75.1% (positive ion mode) and 66.2% (negative ion mode), with further improvements to 80.0% and 73.8% when combined with SIRIUS and BUDDY.

bioinformatics↗

bioRxiv2024

p300/CBP-catalyzed acetylation safeguards minor zygotic genome activation via activating DUX

Zygotic genome activation (ZGA), the first transcriptional event at the onset of life, is crucial for early embryonic development. However, the mechanisms that initiate and regulate ZGA in mammals remain poorly understood. In this study, we demonstrate that p300/CBP-catalyzed acetylation plays a vital role in regulating ZGA. Inhibition of p300/CBP acetyltransferase activity results in impaired minor ZGA and a subsequent 2-cell arrest. By profiling the dynamic landscape of p300 during the ZGA stage, we elucidate the distinct patterns of p300 establishment in promoter and enhancer regions. Moreover, p300 recruits RNA polymerase (Pol II) to pre-configure in the promoter regions of ZGA genes, and orchestrates the gradual establishment of enhancer activity during the early 2-cell stage to activate ZGA. Furthermore, p300 enhances minor ZGA by promoting the expression of DUX. Restoring DUX levels, which are diminished due to p300/CBP inhibition, rescues ZGA and embryonic development by re-establishing Pol II enrichment on minor ZGA genes. Overall, our findings elucidate the mechanism by which p300/CBP regulates ZGA through acetylation and activation of the key transcription factor DUX. HighlightsO_LIp300/CBP plays a stage-specific role during preimplantation development phase. C_LIO_LIp300 displays distinctive dynamic changes in promoter and enhancer regions during ZGA stage. C_LIO_LIp300/CBP-catalyzed acetylation facilitates the pre-configuration and elongation of Pol II at ZGA gene regions. C_LIO_LIDUX recruits p300 and Pol II to activate minor ZGA genes independently of p300s catalytic activity. C_LI

developmental biology↗

bioRxiv2024

Intraspecific and intraindividual trait variability decrease with tree species richness in a subtropical tree biodiversity experiment

Phenotypic variability within tree species responds to local tree species richness. However, we lack evidence on how different sources of trait variation shape tree-tree interactions. Along a diversity gradient from one to eight tree species, we sampled 4,568 leaves from 381 trees to study changes in intraspecific and intraindividual leaf trait variability and assessed their contribution to community functional diversity. Intraspecific variability decreased with tree species richness, while intraindividual variability barely responded. Functional overlap between conspecific trees increased with tree species richness and through intraindividual variation, but was reduced through intraspecific variability, meaning that intraspecific variability may reduce intraspecific competitive interactions while intraindividual variability could arise due to varying microenvironmental conditions within the canopy. Last, intraspecific and intraindividual variability explained high community functional richness and divergence, respectively, especially in mixtures. Our findings emphasize that fine-scale variability influences tree-tree interactions and can be a driver of local functional diversity. TeaserTrait variability within tree species decreases with diversity and explains high forest functional diversity.

ecology↗

bioRxiv2024

Sirt5 preserves cardiac function in ischemia-reperfusion injury by inhibiting ANT2 lactylation

Myocardial ischemia-reperfusion (IR) injury is the most common type of heart disease. IR-disrupted mitochondrial homeostasis affects heart energy metabolism and function, but the mechanism remains poorly understood. Here we report a cardioprotective role of the mitochondrial protein Sirt5 in a mouse model of cardiac IR injury. The down-regulation of Sirt5 in IR-injured heart is associated with an increased protein lactylation and impaired mitochondrial function. Overexpression of Sirt5 alleviates, whereas conditional knockout of Sirt5 aggravates, mitochondrial damage and cardiac injury. Mechanistically, Sirt5 interacts with the inner mitochondrial membrane protein adenine nucleotide translocase 2 (ANT2), inhibiting its lysine lactylation to promote its association with the outer mitochondrial membrane protein voltage-dependent anion-selective channel 1 (VDAC1). Lactylation-resistant ANT2 efficiently complexes with VDAC1 and improves cardiac function after injury. Therefore, we uncover a Sirt5-regulated interaction of ANT2 and VDAC1 in maintaining mitochondrial homeostasis, highlighting the potential of targeting ANT2 lactylation as a therapeutic strategy in the treatment of cardiac injury. TeaserSirt5 promotes the interaction of mitochondrial proteins to protect mitochondrial homeostasis and improve heart function in ischemia-reperfusion injury.

pathology↗

bioRxiv2024

An Intermediate Resolution Model of RNA Dynamics and Phase Separation with Explicit Mg2+

RNAs are major drivers of phase separation in the formation of biomolecular condensates, and can undergo protein-free phase separation in the presence of divalent ions or crowding agents. Much remains to be understood regarding how the complex interplay of base stacking, base pairing, electrostatics, ion interactions, and particularly structural propensities governs RNA phase behavior. Here we develop an intermediate resolution model for condensates of RNAs (iConRNA) that can capture key local and long-range structure features of dynamic RNAs and simulate their spontaneous phase transitions with Mg2+. Representing each nucleotide using 6-7 beads, iConRNA accurately captures base stacking and pairing and includes explicit Mg2+. The model does not only reproduce major conformational properties of poly(rA) and poly(rU), but also correctly folds small structured RNAs and predicts their melting temperatures. With an effective model of explicit Mg2+, iConRNA successfully recapitulates experimentally observed lower critical solution temperature phase separation of poly(rA) and triplet repeats, and critically, the nontrivial dependence of phase transitions on RNA sequence, length, concentration, and Mg2+ level. Further mechanistic analysis reveals a key role of RNA folding in modulating phase separation as well as its temperature and ion dependence, besides other driving forces such as Mg2+-phosphate interactions, base stacking, and base pairing. These studies also support iConRNA as a powerful tool for direct simulation of RNA-driven phase transitions, enabling molecular studies of how RNA conformational dynamics and its response to complex condensate environment control the phase behavior and condensate material properties. SIGNIFICANCE STATEMENTDynamic RNAs and proteins are major drivers of biomolecular phase separation that has been recently discovered to underlie numerous biological processes and be involved in many human diseases. Molecular simulation has an indispensable role to play in dissecting the driving forces and regulation of biomolecular phase separation. The current work describes a high-resolution coarse-grained RNA model that is capable of describing the structure dynamics and complex sequence, concentration, temperature and ion dependent phase transitions of flexible RNAs. The study further reveals a central role of RNA folding in coordinating Mg2+-phosphate interactions, base stacking, and base pairing to drive phase separation, paving the road for studies of RNA-mediated phase separation in relevant biological contexts.

biophysics↗

bioRxiv2024

Characteristic of KPC-14, a KPC Variant Conferring Resistance to Ceftazidime-Avibactam in the Extensively Drug-resistant ST463 Pseudomonas aeruginosa Clinical Isolate

We studied two KPC-14 variants from clinical Pseudomonas aeruginosa isolates, C137 and C159, to better understand genomic diversity, mechanisms, and genes that confer antibiotic resistance and pathogenicity. C137 and C159, a sequence type 463 ExoU-positive multidrug-resistant strain, were concurrently resistant to carbapenems and ceftazidime-avibactam (CZA). Both strains possessed five intrinsic antimicrobial resistance genes (fosA, catB7, crpP, blaPAO, and a blaOXA-486 variant), as well as the blaKPC-14 gene in the chromosome. In strain C137, blaKPC-14 gene was situated on the plasmid pC137. KPC-14-harbouring transformants of pC137 exhibited resistance to CZA and restored sensitivity to carbapenems, signifying a "see-saw" effect. Both strains demonstrated the expression of the blaKPC-14 gene, concurrent inactivation of the outer membrane protein OprD, overexpression of the efflux pump MexX, and a pronounced capacity for biofilm formation. The genomic environment of KPC-14 consisted of IS26/IS26/TnpR_Tn3/ISKpn27/ISKpn6/IS26, which was classified as pseudo-compound transposons (PCTs). Plasmid pC137 closely resembled the previously described plasmid p94, which possesses the same genomic architecture, implying that IS26-mediated PCTs may store a variety of resistance genes, including blaKPC-2 and KPC variants, and were currently disseminating in the region. The KPC-14 variant presents significant challenges for clinical treatment. The blaKPC-14 gene, carried by the PCTs, was integrated into the chromosome, exhibiting stability throughout bacterial inheritance. Our research highlights the necessity for improved clinical surveillance of Klebsiella pneumoniae carbapenemase-producing P. aeruginosa.

molecular biology↗

bioRxiv2024

Nascent transcript O-MAP reveals the molecular architecture of a single-locus subnuclear compartment built by RBM20 and the TTN RNA

Eukaryotic nuclei adopt a highly compartmentalized architecture that influences nearly all genomic processes. Understanding how this architecture impacts gene expression has been hindered by a lack of tools for elucidating the molecular interactions at individual genomic loci. Here, we adapt oligonucleotide-mediated proximity-interactome mapping (O-MAP) to biochemically characterize discrete, micron-scale nuclear neighborhoods. By targeting O-MAP to introns within the TTN pre-mRNA, we systematically map the chromatin loci, RNAs, and proteins within a muscle-specific RNA factory organized around the TTN locus. This reveals an unanticipated compartmental architecture that organizes cis- and trans-interacting chromosomal domains, including a hub of transcriptionally silenced chromatin. The factory also recruits dozens of unique RNA-binding and chromatin-scaffolding factors, including QKI and SAFB, along with their target transcripts. Loss of the cardiac-specific splicing factor RBM20--a master regulator of TTN splicing that is mutated in dilated cardiomyopathy--remodels nearly every facet of this architecture. This establishes O-MAP as a pioneering method for probing single-locus, microcompartment-level interactions that are opaque to conventional tools. Our findings suggest new mechanisms by which coding genes can "moonlight" in nuclear-architectural roles.

genomics↗

bioRxiv2024

RING1 dictates GSDMD-mediated inflammatory response and host susceptibility to pathogen infection

RING1 is an E3 ligase component of the polycomb repressive complex 1 (PRC1) with known roles in chromatin regulation and cellular processes such as apoptosis and autophagy. However, its involvement in inflammation and pyroptosis needs to be better characterized. Here, we found that human RING1, but not RING2, promoted K48-linked ubiquitination of Gasdermin D (GSDMD) and acted as a negative regulator of pyroptosis and bacterial infection. Ring1 knockout mice increased the bacterial loads and mortality rate by treatment of S. typhimurium. Upon infection by M. tuberculosis (Mtb) H37Rv strain, RING1 deletion initially reduced bacterial loads but later increased lung inflammation and impaired immune defense responses in mice. Also, Ring1 knockout exacerbated acute sepsis induced by lipopolysaccharide (LPS). Mechanistically, RING1 directly interacts with GSDMD and ubiquitinates the K51 and K168 sites of GSDMD for K48-linked proteasomal degradation, thereby inhibiting pyroptosis. Inhibition of RING1 E3 ligase activity by mutation or small molecular compound increased GSDMD level and cell death during pyroptosis. Our findings reveal that RING1 dictates GSDMD-mediated inflammatory response and host susceptibility to pathogen infection, highlighting RING1 as a potential therapeutic target for combating infectious diseases. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=115 SRC="FIGDIR/small/622030v1_ufig1.gif" ALT="Figure 1"> View larger version (32K): org.highwire.dtl.DTLVardef@e6d4a6org.highwire.dtl.DTLVardef@92f137org.highwire.dtl.DTLVardef@11a69c4org.highwire.dtl.DTLVardef@115dc0b_HPS_FORMAT_FIGEXP M_FIG Graphic Abstract C_FIG HighlightsO_LIE3 ubiquitin ligase RING1 targets GSDMD for proteasomal degradation. C_LIO_LIMice lacking RING1 are more vulnerable to S. typhimurium infection and LPS- induced sepsis. C_LIO_LIIn the early stages of M. tuberculosis (Mtb) infection, Ring1-/- mice exhibit efficient bacterial clearance. C_LIO_LIChronic inflammation in Ring1-/- mice impairs the ability to control Mtb in the late stages of infection. C_LI

cell biology↗

bioRxiv2024

Control of Golgi- V-ATPase through Sac1-dependent co-regulation of PI(4)P and cholesterol

Sac1 is a conserved phosphoinositide phosphatase, whose loss-of-function compromises cell and organism viability. Here, we employed acute auxin-inducible Sac1 degradation to identify its immediate downstream effectors in human cells. Most of Sac1 was degraded in [~]1 h, paralleled by increased PI(4)P and decreased cholesterol in the trans-Golgi network (TGN) during the following hour, and superseded by Golgi fragmentation, impaired glycosylation, and selective degradation of TGN proteins by [~]4 h. The TGN disintegration resulted from its acute deacidification caused by disassembly of the Golgi V-ATPase. Mechanistically, Sac1 mediated TGN membrane composition maintained an assembly promoting conformation of the V0a2 subunit. Key phenotypes of acute Sac1 degradation were recapitulated in human differentiated trophoblasts, causing processing defects of chorionic gonadotropin, in line with loss-of-function intolerance of the human SACML1 gene. Collectively, our findings reveal that the assembly of the Golgi V-ATPase is controlled by the TGN membrane via Sac1 fuelled lipid exchange.

cell biology↗

bioRxiv2024

AutoPM3: Enhancing Variant Interpretation via LLM-driven PM3 Evidence Extraction from Scientific Literature

Rare diseases, affecting 300 million people globally, often result from genetic variants. Wholegenome sequencing has made variant detection more cost-effective, but interpreting these variants remains challenging. Current clinical practice combines quantitative evidence and literature, which is complex and time-consuming. We introduce AutoPM3, a method for automating the extraction of ACMG/AMP PM3 evidence from scientific literature using open-source LLMs. It combines an optimized RAG system for text comprehension and a TableLLM equipped with Text2SQL for data extraction. We evaluated AutoPM3 using our collected PM3-Bench, a dataset from ClinGen with 1,027 variant-publication pairs. AutoPM3 significantly outperformed other methods in variant hit and in trans variant identification, thanks to the four key modules. Additionally, we wrapped AutoPM3 with a user-friendly interface to enhance its accessibility. This study presents a powerful tool to improve rare disease diagnosis workflows by facilitating PM3-relevant evidence extraction from scientific literature.

bioinformatics↗

bioRxiv2024

MethylGPT: a foundation model for the DNA methylome

DNA methylation serves as a powerful biomarker for disease diagnosis and biological age assessment. However, current analytical approaches often rely on linear models that cannot capture the complex, context-dependent nature of methylation regulation. Here we present MethylGPT, a transformer-based foundation model trained on 226,555 (154,063 after QC and deduplication) human methylation profiles spanning diverse tissue types from 5,281 datasets, curated 49,156 CpG sites, and 7.6 billion training tokens. MethylGPT learns biologically meaningful representations of CpG sites, capturing both local genomic context and higher-order chromosomal features without external supervision. The model demonstrates robust methylation value prediction (Pearson R=0.929) and maintains stable performance in downstream tasks with up to 70% missing data. Applied to age prediction across multiple tissue types, MethylGPT achieves superior accuracy compared to existing methods. Analysis of the models attention patterns reveals distinct methylation signatures between young and old samples, with differential enrichment of developmental and aging-associated pathways. When finetuned to mortality and disease prediction across 60 major conditions using 18,859 samples from Generation Scotland, MethylGPT achieves robust predictive performance and enables systematic evaluation of intervention effects on disease risks, demonstrating potential for clinical applications. Our results demonstrate that transformer architectures can effectively model DNA methylation patterns while preserving biological interpretability, suggesting broad utility for epigenetic analysis and clinical applications.

systems biology↗