U2 small nuclear RNA auxiliary factor 2 (U2AF2), a key pre‑mRNA splicing factor whose role in NSCLC progression remains incompletely understood, was previously found by us to be significantly upregulated in non‑small cell lung cancer (NSCLC) tissues and to contribute to NSCLC progression‑related phenotypes. In this study, we observed that knockdown of U2AF2 induced exon 4–6 skipping in UPP1. Consistent with the role of U2AF2 in metabolic regulation, UPP1 was also found to modulate glycolysis‑related metabolic activity, with the UPP1‑L isoform exerting a more pronounced enhancement of glycolytic parameters, particularly under high glucose conditions. In vitro and in vivo functional assays showed that knockdown of UPP1‑L more strongly inhibited lung cancer cell growth and more robustly increased caspase‑dependent apoptosis than either control or UPP1‑S knockdown; Similarly, UPP1‑L knockdown caused a greater reduction in cancer cell migration and invasion and more clearly attenuated EMT‑associated phenotypes compared with UPP1‑S knockdown. Co‑IP assays indicated that UPP1‑L and UPP1‑S interact with partially distinct sets of proteins, which may at least partially underlie their differential functional effects in lung cancer cells. Analysis of clinical NSCLC samples further revealed a significant positive correlation between U2AF2 and UPP1‑L expression levels and showed that elevated UPP1‑L expression was significantly associated with poorer prognosis. Collectively, these findings identify a previously unrecognized role of U2AF2 in regulating UPP1 alternative splicing during NSCLC progression and suggest that modulation of UPP1 splicing may represent a potential therapeutic vulnerability in NSCLC.

Apurinic/apyrimidinic endonuclease 1 (APE1) is a pivotal biomarker frequently overexpressed in cancer cells, hence in situ monitoring of its dynamic alterations during cell death is crucial for therapeutic evaluation. Herein, we designed highly sensitive and selective fluorescent nanoprobes based on DNA-functionalized gold nanoparticles (AuNPs) for real-time imaging and monitoring of APE1 activity in living cells. The nanoprobes, composed of APE1-responsive double-stranded DNA containing apurinic/apyrimidinic (AP) sites and conjugated with black hole quencher 2 (BHQ2), are immobilized on the surface of AuNPs. Our results demonstrate that electrical stimulation (ES) selectively induces pyroptosis of HeLa cells, and nearly complete suppression of APE1 activity as ES-induced pyroptosis progressed was revealed via confocal microscopy analyses. The ES treatment leads to a marked accumulation of intracellular reactive oxygen species (ROS) and the resulting oxidative stress causes a pronounced up-regulation of γ-H2AX, indicating severe DNA double-strand breaks. We propose that the ROS-mediated down-regulation of APE1 contributes to impaired base excision repair (BER) and exacerbates genomic instability during ES-induced pyroptosis. Notably, this phenomenon did not occur in normal H8 cells. This study establishes a nanoprobe sensing platform for real-time monitoring of APE1 dynamics and identifies the down-regulation of APE1 as an indicator of effective ES-induced pyroptosis in cancer cells, providing new insights into therapeutic assessment in cancer treatment.

Introduction Nasopharyngeal carcinoma (NPC) is an aggressive malignancy with a poor prognosis. Ubiquitination is a complex post translational modification involved in cancer progression. However, ubiquitination related genes (URGs) in immunotherapy of NPC remains largely unexplored.Methods Differentially expressed URGs were screened based on the single-cell RNA sequencing (scRNA-seq) dataset and a risk model of NPC was constructed and evaluated for prognostic significance. The oncogenic role of RNF149 in NPC was investigated through in vitro and in vivo experiments, including tumor cells, NPC-like organoids, and tumor-bearing mice.Results scRNA-seq data showed that URGs scores were higher in cancer cells than in normal epithelial cells. We identified 216 differentially expressed URGs between cancer and normal epithelial cells, but only 33 differentially expressed URGs associated with prognosis. Based on 33 URGs, TCGA-HNSC samples were classified into two distinct subtypes with significant differences in the tumor immune microenvironment, immunotherapy effect, and survival-prognostic genes. Using LASSO algorithm, 13 URGs were selected to construct a risk model, which demonstrated high predictive performance. The expression profiles of these 13 URGs were analyzed in TCGA-HNSC tumor and adjacent non-cancerous samples, and six URGs (BSPRY, OTUB1, PJA1, RNF149, RNF181, USP10) exhibited consistent expression trends. Moreover, quantitative real- time PCR revealed that RNF149 was up-regulated expression in NPC cells compared to the NP69 cells. RNF149 knockdown significantly impeded the proliferative, migratory, and invasive capabilities and exaggerated apoptosis of NPC cells. RNF149 knockdown cells exhibited a reduced capacity to form NPC organoids in a 3D culture system. shRNA-RNF149 diminished subcutaneous tumorigenic capacity of HK-1 cells compared to the control group.Discussion The URGs-based prognostic risk model offers a robust tool for predicting immunotherapy efficacy in NPC and RNF149 promotes NPC progression.Conclusion A URGs-related prognostic risk model capable of predicting clinical outcomes in NPC patients and RNF149 promotes NPC progression. Our findings are expected to provide new strategies to improve outcomes for NPC patients.

Endoplasmic reticulum stress-related cancer-associated fibroblasts (ERS–CAF) remodel the tumor microenvironment and drive immune exclusion and therapy resistance in chordoma, yet routine and non-invasive readouts of this biology are lacking. We hypothesized that standard pre-operative MRI and H E whole-slide images (WSI) encode image-based surrogates of ERS–CAF-driven immunoregulation that can be learned and generalized across cancers. Three bulk-transcriptomic reference scores were defined for surrogate supervision, capturing ERS-program activity, ERS–CAF-immuneligand-receptor crosstalk and microenvironmental heterogeneity. In 126 chordoma cases, a stage-wise multimodal framework integrating calibrated WSI attention, gated radiopathomic fusion and domain alignment showed strong concordance with molecular profiles, independent prognostic value and biologically specific localization to fibrotic immune-excluded regions. These associations were generalized in zero-shot analyses to the TCGA pan-cancer. An MRI-only distilled model preserved most predictive performance with substantial gains in efficiency, supporting scalable non-invasive clinical application.

Alzheimer's disease (AD) is a neurodegenerative disorder with a complex, multifactorial pathogenesis. Growing evidence implicates disturbances in cellular energy metabolism and iron dyshomeostasis as interlinked contributors to pathology. Within this framework, iron accumulation may act as an upstream regulator in certain contexts and stages, while in others it emerges downstream and amplifies ongoing injury. As iron is an essential cofactor for mitochondrial respiration and the tricarboxylic acid cycle, iron imbalance can compromise ATP production and disrupt glucose metabolism, exacerbating neuronal energy deficits. The interplay among iron accumulation, oxidative stress, and neuroinflammation can create vicious cycles that reprogram cellular metabolism and disrupt the critical metabolic coupling between neurons and glial cells. This review synthesizes recent advances in understanding the iron-energy metabolism axis in AD, delineates mechanisms by which iron imbalance precipitates mitochondrial dysfunction and glucose metabolic impairments, and evaluates how these deficits synergize with neuroinflammation and proteinopathy across disease stages. Finally, we appraise emerging therapeutic strategies targeting iron overload and metabolic pathways, discuss their stage-dependent risks and benefits, and outline the need for biomarker-guided approaches to optimize patient selection and treatment timing.

This study systematically evaluated the anti-breast cancer potential and mechanisms of the cannabidiol (CBD) derivative MCPB-21. The structure of MCPB-21 was confirmed by nuclear magnetic resonance (NMR). The study analyzed differentially expressed genes associated with breast cancer using public databases and verified the binding affinity of MCPB-21 to glycogenin-2 (GYG2) through molecular docking. Additionally, the effects of MCPB-21 on apoptosis, invasion capacity, and lipid metabolism were evaluated in MDA-MB-231 and MCF-7 breast cancer cells using flow cytometry, Transwell invasion assays, cell proliferation assays, and Oil Red O staining. Western blot was employed to examine expression changes in proteins related to fatty acid β-oxidation and ferroptosis, including Acyl-CoA Oxidase 1 (ACOX1), ATP Binding Cassette Subfamily D Member 3 (ABCD3), ATP Binding Cassette Subfamily D Member 4 (ABCD4), Peroxisomal l-bifunctional enzyme (EHHADH), Carnitine palmitoyltransferase 1α (CPT1α), Glutathione Peroxidase 4 (GPX4), Solute Carrier Family 7 Member 11 (SLC7A11), and Acyl-CoA Synthetase Long Chain Family Member 4 (ACSL4). The role of fatty acid oxidation in ferroptosis was further analyzed using the ACOX1 inhibitor 10,12-Tricosadiynoic acid (500 nM). Additionally, the effects of MCPB-21 on fatty acid oxidation and ferroptosis were evaluated by interfering with GYG2 expression. Ferrostatin-1 (Fer-1) rescue experiments were conducted to verify the dependence of MCPB-21-induced cell death. Finally, the anti-tumor efficacy of various doses of MCPB-21 was compared to the control drug CBD. In vitro experimental results showed that MCPB-21 can affect the behavior of breast cancer cells by inducing cancer cell apoptosis, increasing reactive oxygen species (ROS) levels, and promoting lipid accumulation. At the same time, Western blot detection showed that MCPB-21 could downregulate key enzymes of fatty acid β-oxidation (ACOX1, ABCD3, ABCD4, EHHADH, CPT1α) and antioxidant factors (GPX4, SLC7A11), and upregulate the enzyme ACSL4 that promotes lipid peroxidation. Mechanistic studies further showed that MCPB-21 affects the expression of ACOX1 by regulating GYG2, inhibits fatty acid β-oxidation, and induces ferroptosis. At the same time, the combined use of ACOX1 inhibitors enhanced lipid accumulation and ROS levels, verifying its role in regulating fatty acid oxidation. In animal experiments, MCPB-21 (10 and 40 mg/kg) significantly inhibited the growth of nude mouse transplanted tumors, caused tumor tissue necrosis, inhibited the proliferation marker Ki67, and regulated the expression of ferroptosis-related proteins (GPX4 and SLC7A11 decreased, and ACSL4 increased). Immunohistochemical analysis showed that MCPB-21 had a stronger anti-tumor effect than CBD, mainly by regulating the fatty acid β-oxidation pathway to promote ferroptosis. In summary, MCPB-21 exhibits excellent anti-breast cancer potential. Its mechanism of action is mainly to achieve anti-tumor effects by inhibiting fatty acid β-oxidation and activating ferroptosis, which provides a theoretical basis and potential therapeutic strategy for the development of new anti-breast cancer drugs.

Background: Obesity is a heterogeneous chronic disease driven by interacting genetic, metabolic, inflammatory, and environmental factors, yet clinically useful blood-based markers for early risk stratification remain limited. Methods: Using plasma proteomic data from the UK Biobank, we assessed 2,923 circulating proteins in relation to incident obesity. Findings: We identified 98 proteins associated with future risk, with changes detectable years before clinical onset. These proteins were enriched in pathways related to metabolism, low-grade inflammation, and tissue remodeling and were linked to diverse systemic phenotypes. A protein-based model achieved an AUC of 0.809 for predicting incident obesity. Integrative analyses highlighted ADM, NCAN, APOM, and LEP as key candidates, suggesting distinct biological contributions to obesity development. Interpretation: Collectively, these findings delineate early circulating proteomic alterations preceding obesity and support their potential utility for risk prediction and mechanistic prioritization.

Lung cancer bone metastasis presents a major clinical challenge due to therapeutic resistance and severe morbidity. Although disrupting tumor-bone microenvironment crosstalk is a promising strategy, clinically actionable targets remain limited. Here, by analyzing bulk RNA sequencing data from bone metastatic tumors across multiple cancer types, we identified PARP10 as a gene consistently upregulated in bone metastases. High PARP10 expression in primary tumors was correlated with poor patient survival. Functional studies demonstrated that PARP10 promoted lung cancer growth and bone metastasis both in vitro and in vivo. Mechanistically, multi-omics integrated analyses revealed that PARP10 deletion induced DNA damage and oxidative stress, and upregulated BCAT2 expression in a MYC-dependent manner to enhance BCAA catabolism. This metabolism exerts an adaptive compensatory effect on tumor cells via boosting mitochondrial oxidative phosphorylation, yet depletes bone microenvironmental BCAA and consequently suppresses osteoclast differentiation, thereby inhibiting bone metastasis. Importantly, pharmacological inhibition of PARP10 with OUL232 mitigated bone metastatic burden in mice without observable toxicity, demonstrating its therapeutic potential by concurrently inducing tumor cell apoptosis and disrupting the pro-metastatic niche. Our findings establish PARP10 as a central regulator of a targetable metabolic competition axis and propose its inhibition as a dual-mechanism strategy that simultaneously attacks tumor cells and disrupts the pro-metastatic niche.

Precise regulation of the activating H3K4me3 and repressive H3K27me3 histone modifications at bivalent promoters is essential for normal development but is frequently disrupted in cancer. Among the polycomb group (PCGF) family members, which are key components of polycomb repressive complex 1 (PRC1), PCGF1 emerged as the factor most strongly associated with poor prognosis in non-small cell lung cancer (NSCLC) based on analyses of The Cancer Genome Atlas (TCGA) cohort. In lung cancer cells, PCGF1 upregulation enhanced the deposition of H2AK119ub and H3K27me3 at chromatin. These depositions inhibit the cytokine-cytokine receptor interaction pathway, especially CCL5, CXCL10, CD40, and FAS. Single-cell RNA sequencing further indicated that PCGF1 acts as a negative regulator of natural killer (NK) cell effector function. When NK cell-derived cytokines attempted to activate the cytokine-cytokine receptor interaction pathway in tumor cells, this repressive chromatin state attenuated pathway activation. Consequently, reduced expression of these genes weakened NK cell recruitment and cytotoxic responses. Collectively, this study uncovers a previously unrecognized mechanism by which PCGF1-driven disruption of bivalent promoter balance silences immune signaling cascades, enabling tumor cells to evade NK cell-mediated immunity in NSCLC. These findings highlight bivalent chromatin as a critical regulatory node in tumor immune escape and establish PCGF1 as a promising epigenetic target for immunotherapeutic intervention.PCGF1 promotes immune evasion in NSCLC by suppressing cytokine-cytokine receptor interaction pathway. Upregulation of PCGF1 in NSCLC cells enhances the deposition of the repressive histone modifications H2AK119ub and H3K27me3 while reducing the enrichment of the transcriptionally active histone modification H3K4me3 at target chromatin regions, thereby suppressing cytokine-cytokine receptor interaction pathway. This epigenetic repression reduces the expression of immune-related genes, including CCL5, CXCL10, CD40, and FAS. Consequently, tumor-cell responses to NK cell-derived cytokines are attenuated, leading to impaired NK cell recruitment and cytotoxicity. The schematic diagram was created using BioRender.

This study investigated the involvement of Oridonin-related prognostic genes in programmed cell death (PCD) in primary lung cancer and constructed a corresponding prognostic model. Transcriptomic, mutation, and clinical data from the TCGA database, including the TCGA-LUSC (Lung Squamous Cell Carcinoma) and TCGA-LUAD (Lung Adenocarcinoma) cohorts with 989 primary lung cancer samples, were analyzed. A total of 38 potential Oridonin-related target genes were identified from the PubChem database, among which 15 demonstrated prognostic significance. Weighted Gene Co-expression Network Analysis (WGCNA) revealed correlations between these genes and 13 types of PCD. LASSO and random forest algorithms constructed risk score models and identified FLNC and FOSL1 as independent prognostic biomarkers. Molecular docking analysis confirmed strong binding affinities between FLNC, FOSL1, and Oridonin, suggesting their potential as therapeutic targets. The results showed that the tumor mutation burden (TMB) of the high-risk group was significantly higher than that of the low-risk group. Analysis of immune profiles revealed that the heterogeneity in risk scores was closely linked to distinct patterns of immune cell infiltration and differential activation of immune responses in the tumor microenvironment. Additionally, experimental validation through Western blot demonstrated elevated protein expression of FLNC and FOSL1 in tumor cell lines, and immunohistochemistry (IHC) confirmed their high expression in tumor tissues. The results highlight the critical role of FLNC and FOSL1 as prognostic markers in lung cancer, while also providing new perspectives on the development of Oridonin-based therapeutic interventions.

Background:Increasing evidence has suggested that the early initiation of antiviral treatment promotes better treatment responses in young children with chronic hepatitis B (CHB), including functional cure. But there are limited studies regarding the clinical responses of antiviral treatment for these children. This study aims to explore the relationship between age at the initiation of antiviral therapy and hepatitis B surface antigen (HBsAg) loss among pediatric patients. Methods:We systematically searched five electronic databases and reviewed relevant research, and then conducted two-arm and single-arm meta-analyses of treatment responses among children with CHB to explore the effect of baseline age on treatment responses. Finally, based on the different clinical phases of CHB, treatment regimens, and observation end points, subgroup analysis was also conducted to find possible influence factors during the treatment. The primary outcome was HBsAg loss. Results:Eighteen studies involving 2,459 children with CHB were analyzed in this systematic review and meta-analysis. Children aged 1-7 years with CHB exhibited better treatment responses than those aged 8-18 years (odds ratios: HBsAg loss, 5.584 [4.069-7.664] ; HBV DNA suppression, 1.946 [1.356-2.793]; hepatitis B e antigen [HBeAg] loss, 2.369 [1.612-3.482]; HBeAg seroconversion, 3.094 [2.196-4.360]; and HBsAg seroconversion, 4.970 [1.503-16.439]), especially among children with CHB in the immune-clearance phase. The pooled rates of HBV DNA suppression, HBeAg loss, HBeAg seroconversion, HBsAg loss, and HBsAg seroconversion of children aged 1-7 versus 8-18 years were 85.7% versus 71.2% (P > 0.05), 74.4% versus 52.0% (P < 0.01), 70.6% versus 38.9% (P < 0.01), 54.5% versus 18.8% (P < 0.001), and 62.0% versus 17.5% (P < 0.05), respectively. Conclusions:Baseline age at the initiation of antiviral treatment is associated with HBsAg loss in children with CHB. Since all the available data came from observational studies, our findings should be validated by further multicohort randomized controlled trials. Registration:This study was registered in PROSPERO (CRD42024483967).

BACKGROUND:In the budding yeast Saccharomyces cerevisiae, the widespread adoption of ribosome profiling technology has allowed the discovery of evidence of transcription and translation for thousands of small proteins or microproteins whose importance was once disregarded. Both conserved and evolutionarily short-lived microproteins have demonstrated relevant involvement in biological functions. However, sequences exist in a broad spectrum of conservation. Here, we tested whether these small proteins in yeast detected by ribosome profiling technology have different properties across their levels of conservation, and how do these properties compare with the canonical small protein-coding sequences. RESULTS:Here, we applied a phylostratigraphic approach to peptides encoded by small open reading frames. We compared 20,023 ribo-seq-detected small peptides against annotated small proteins belonging to reference annotations on the basis of their respective conservation patterns. We identified 1134 unannotated microproteins that, despite their difficulty in being detected by methods other than ribosome profiling, display hallmarks of functionality such as conservation across many taxonomical levels and signals of purifying selection not dissimilar to those of canonical proteins of comparable length. Sequences that initially did not show evidence of belonging to any gene family were found to possess signals of homology traceable mostly at genus level when compared against noncoding regions and using TBLASTN, but also, to a lesser extent, to species belonging to the phyla Basidiomycota and Microsporidia. In addition, we show an analysis of the mutations behind the origin of small open reading frames exclusive to S. cerevisiae and identified changes in the initiation codon as the most common group of mutations when compared to Saccharomyces paradoxus, the closest species to S. cerevisiae. CONCLUSIONS:Our work, by presenting robust analysis of the extended landscape of small proteins in yeast, suggests that small conserved sequences, either canonical or not, possess a shared evolutionary trajectory, as demonstrated by their properties. These results shed some light into the evolutionary processes behind the extended landscape of small proteins in yeast.

Wilms tumor (WT), the most common pediatric renal malignancy, exhibits a relatively low mutational burden compared to adult cancers, which hinders the development of targeted therapies. To elucidate the molecular landscape of WT, we perform integrative proteomic, phosphoproteomic, transcriptomic, and whole-exome sequencing analyses of WT and normal kidney tissue adjacent to tumor. Our multi-omics approach uncovers prognostic genetic alterations, distinct molecular subgroups, immune microenvironment features, and potential biomarkers and therapeutic targets. Proteome- and transcriptome-based stratification identifies three molecular subgroups with unique signatures, correlating with different histopathological subtypes and putative cellular origins at different stages of embryonic kidney development. Notably, we identify EHMT2 as a promising prognostic biomarker and therapeutic target associated with epigenetic regulation and Wnt/β-catenin pathway. In this work, we provide a comprehensive molecular characterization of WT, offering valuable insights into its pathogenesis and a foundational resource for future therapeutic development.

Metatranscriptomic analysis is increasingly performed in environments to provide dynamic gene expression information on ecosystems, responding to their changing conditions. Many computational methods have undergone remarkable development in the past years, but a comprehensive benchmark study is still lacking. There are concerns regarding the accuracies of the qualitative and quantitative profilers obtained from metatranscriptomic analysis, especially for the microbiota in extreme environments, most of them are unculturable and lack well-annotated reference genomes. Here, we presented a benchmark experiment that included 10 single-species and their cell or RNA-admixtures with the predefined species compositions and varying evenness, simulating the low annotation rate and high heterogeneity. In total, 1 metagenome sample and 24 metatranscriptome were sequenced for the comparisons of 36 combination of analysis methods for tasks ranging from sample preparation, quality control, rRNA removal, alignment strategies, taxonomic profiling, and transcript quantification. For each part of the workflow mentioned above, corresponding metrics have been established to serve as standards for assessment and comparison. Evaluation revealed the performances and proposed an optimized pipeline named MT-Enviro (MetaTranscriptomic analysis for ENVIROnmental microbiome). Our data and analysis provide a comprehensive framework for benchmarking computational methods with metatranscriptomic analysis. MT-Enviro is implemented in Nextflow and is freely available from https://github.com/Li-Lab-SJTU/MT-Enviro.

The mechanisms via which inflammatory macrophages mediate intestinal inflammation are not completely understood. Herein, using merged analysis of RNA sequencing and mass spectrometry-based quantitative proteomics, we detected differences between proteomic and transcriptomic data in activated macrophages. Dipeptidase-2 (DPEP2), a member of the DPEP family, was highly expressed and then downregulated sharply at the protein level but not at the mRNA level in macrophages in response to inflammatory stimulation. Suppression of DPEP2 not only enhanced macrophage-mediated intestinal inflammation in vivo but also promoted the transduction of inflammatory pathways in macrophages in vitro. Mechanistically, overexpressed DPEP2 inhibited the transduction of inflammatory signals by resisting MAK3K7 in inactivated macrophages, whereas DPEP2 degradation by activated Trim32 resulted in strong activation of NF-κB and p38 MAPK signaling via the release of MAK3K7 in proinflammatory macrophages during the development of intestinal inflammation. The Trim32-DPEP2 axis accumulates the potential energy of inflammation in macrophages. These results identify DPEP2 as a key regulator of macrophage-mediated intestinal inflammation. Thus, the Trim32-DPEP2 axis may be a potential therapeutic target for the treatment of intestinal inflammation.

Colorectal liver metastasis (CRLM) is one of the leading death causes among colorectal cancer (CRC) patients, yet its underlying molecular events remain poorly understood, particularly at the proteomic and phosphoproteomic levels. A proteogenomic analysis combining genomics, transcriptomics, proteomics, and phosphoproteomics is performed on 102 samples from 34 treatment-naïve CRLM patients, including primary CRC, adjacent normal colorectal, and matched liver metastasis tissues. CRC cell lines, organoids, mouse models, and an independent patient cohort are used to validate the findings. Proteomics and phosphoproteomics show profoundly dysregulated pathways in liver metastasis tissues, notably disruptions in carbon metabolism. Functional validation using CRC organoids and mouse models demonstrates that the one-carbon metabolism enzyme SHMT1 promotes CRC tumorigenesis and metastasis via formate-mediated AMPK inhibition, whereas PIM kinase-dependent NDRG1 Ser330 phosphorylation exacerbates liver metastasis by promoting ubiquitin-dependent degradation of NDRG1. Unsupervised clustering identifies two proteomic subtypes of liver metastasis samples with distinct clinical outcomes: a poor-prognosis C1 (metabolism) subtype and a better-prognosis C2 (RNA function) subtype. Considering expression frequency, specificity, and functional relevance, FTCD, GPD1, SOD2, and EIF4B Ser422 phosphorylation are further identified and validated as subtype prognostic biomarkers. This study provides critical insights into the molecular mechanisms underlying CRLM and offers resources for high-risk metastatic CRC.

The escalating global threat of multidrug-resistant pathogens underscores the urgent demand for innovative strategies in antimicrobial peptide (AMP) discovery. Notably, deep-sea-related data resources remain underexplored, despite their potential as valuable sources of novel AMPs. Current AMP prediction methods, however, are limited by dataset biases such as sequence length imbalance between AMPs and non-AMPs, N-terminal methionine artifacts in non-AMPs, and microbial origin specificity. To overcome these constraints, we developed a dual-engine predictor named XAMP, which integrates two complementary architectures: XAMP-E, built on ESM-2 for high-accuracy feature representation, and XAMP-T, built on one-layer Transformer for accelerating large-scale screening. This dual-engine design ensures both robust feature learning and enhanced generalization capability. By constructing length-balanced datasets, removing N-terminal methionine from non-AMPs, and training microbial-specific variants, XAMP achieved a median area under the receiver-operating characteristic curve (AUC) of 0.972, representing an approximately 10% improvement over state-of-the-art predictors. Using an integrated AMP mining pipeline incorporating these models, we screened large-scale deep-sea metagenomic data and identified 2,355 promising AMP candidates. This study establishes a robust deep-learning framework that facilitates targeted discovery of bioactive peptides from extremophile microbiomes through systematic multi-omics validation. ### Competing Interest Statement The authors have declared no competing interest. The National Natural Science Foundation of China, 32570783, 32170664 The Key Project for Computational Biology of Shanghai, 23JS1400800 The Fundamental Research Funds for the Central Universities, YG2023ZD11

For metaproteomics data derived from the collective protein composition of dynamic multi-organism systems, the proportion of missing values and dimensions of data exceeds that observed in single-organism experiments. Consequently, evaluations of differential analysis strategies in other mass spectrometry (MS) data (such as proteomics and metabolomics) may not be directly applicable to metaproteomics data. In this study, we systematically evaluated five imputation methods [sample minimum, quantile regression, k-nearest neighbors (KNN), Bayesian principal component analysis (bPCA), random forest (RF)] and six imputation-free methods (moderated t-test, two-part t-test, two-part Wilcoxon test, semiparametric differential abundance analysis, differential abundance analysis with Bayes shrinkage estimation of variance method, and Mixture) for differential analysis in simulated metaproteomic datasets based on both data-dependent acquisition MS experiments and emerging data-independent acquisition experiments. The simulation datasets comprised 588 scenarios by considering the impacts of sample size, fold change between case and control, and missing value ratio at random and nonrandom. Compared to imputation-free methods, KNN, bPCA, and RF imputation performed poorly in datasets with a high missingness ratio and large sample size and resulted in a high false-positive risk. We made empirical recommendations based on the balance of sensitivity in analysis and control of false positives. The moderated t-test was optimal in scenarios of large sample size with a low missingness ratio. The two-part Wilcoxon test was recommended in scenarios of small sample size with a low missingness ratio or large sample size with a high missingness ratio. The comprehensive evaluations in our study can provide guidance for the differential abundance analysis in metaproteomics.

IntroductionThe delivery of nucleic acid into cells using polyethylenimine (PEI) as non-viral carrier is a potential candidate technique for the treatment of hepatitis B virus (HBV) infection.MethodsIn the present study, PEI was used as cationic polymers and transfected with unmodified oligodeoxynucleotides in cell cultures and the BALB/c mouse model to investigate its efficiency in blocking HBV surface antigen (HBsAg) secretion.Results and discussionPEI/oligonucleotide complexes selectively inhibited HBsAg secretion in the culture supernatant, while there were no evident alterations in HBeAg and HBV DNA levels, thereby suggesting its potential inhibitory activity against the production of HBsAg. The complexes formed by PEI with double-stranded decoy oligonucleotides also suppressed HBsAg secretion but showed no expected interference with the intermediate levels of HBV transcription or replication. Furthermore, PEI/plasmid-DNA complexes demonstrated no influence on the expression levels of HBsAg, thus highlighting the specific effects of PEI/oligonucleotides exerted on HBsAg release. PEI-oligonucleotides transfection prior to the viral inoculation impaired HBV infection in HepG2-NCTP cells. Importantly, the PEI/oligonucleotide complex also induced the decline of HBsAg in hydrodynamically injected BALB/c mice. These findings demonstrate that transfection of PEI/oligonucleotide complexes can help effectively reduce HBsAg level and may offer a new potential avenue for the development of anti-HBV treatment.