Objective: To explore the characteristics of the diversity and composition of oral microbial flora with age, and to provide a reference for understanding the succession of oral microecology at different ages. Methods: Oral rinse 16S rRNA (V4 region) sequencing data from 9 021 participants 14-69 years of age in the 2009-2012 National Health and Nutrition Examination Survey (NHANES) were analyzed. Alpha diversity (Observed OTUs, Faith’s PD, Shannon Index), beta diversity (Bray-Curtis and UniFrac), and genus-level composition were examined using weighted generalized linear models (GLMs), including quadratic terms for age and adjusting for key covariates (gender, race/ethnicity, BMI, smoking status, and periodontitis severity). Results: Alpha diversity demonstrated a clear inverted U-shaped trajectory across age, peaking at 25-30 years old and declining thereafter. This trend remained consistent across sex, race, smoking, and periodontal health strata. Beta diversity analyses revealed a modest but steady age-related shift in community structure. Genus-level analyses revealed that Rothia, Prevotella_6, and Lactobacillus increased steadily with age, while Haemophilus, Porphyromonas, and Corynebacterium declined significantly. Notably, potential periodontopathogens, such as Fusobacterium and Treponema_2, peaked in early adulthood before declining with age. Conclusion Age is an important driver of oral microbial succession, and the oral microbiome exhibits dynamic changes across different life stages. Future longitudinal and multi-omic studies are warranted to elucidate the mechanisms underlying these age-related trajectories.

Chinese hamster ovary (CHO) cells are the most established and versatile mammalian expression system for the large-scale production of recombinant therapeutic proteins, owing to their genetic stability, adaptability to serum-free suspension culture, and ability to perform human-like post-translational modifications. More than 70% of biologics approved by the U.S. Food and Drug Administration rely on CHO-based production platforms, underscoring their central role in modern biopharmaceutical manufacturing. Despite these advantages, CHO systems continue to face three persistent bottlenecks that limit their potential for high-yield, reproducible, and cost-efficient production: excessive metabolic burden during high-density culture, heterogeneity of glycosylation patterns, and progressive loss of long-term expression stability. This review provides an integrated analysis of recent advances addressing these challenges and proposes a forward-looking framework for constructing intelligent and sustainable CHO cell factories. In terms of metabolic regulation, excessive lactate and ammonia accumulation disrupts energy balance and reduces recombinant protein synthesis efficiency. Optimization of culture parameters such as temperature, pH, dissolved oxygen, osmolarity, and glucose feeding can effectively alleviate metabolic stress, while supplementation with modulators including sodium butyrate, baicalein, and S-adenosylmethionine promotes specific productivity (qP) by modulating apoptosis and chromatin structure. Furthermore, genetic engineering strategies—such as overexpression of MPC1/2, HSP27, and SIRT6 or knockout of Bax, Apaf1, and IGF-1R—have demonstrated significant improvements in cell viability and product yield. The combination of multi-omics metabolic modeling with artificial intelligence (AI)-based prediction offers new opportunities for building self-regulating CHO systems capable of dynamic adaptation to environmental stress. Regarding glycosylation uniformity, which determines therapeutic efficacy and immunogenicity, gene editing-based glycoengineering (e.g., FUT8 knockdown or ST6Gal1 overexpression) has enabled the humanization of CHO glycan profiles, minimizing non-human sugar residues and enhancing drug stability. Process-level strategies such as galactose or manganese co-feeding and fine control of temperature or osmolarity further allow rational regulation of glycosyltransferase activity. Additionally, in vitro chemoenzymatic remodeling provides a complementary route to construct human-type glycans with defined structures, though industrial applications remain constrained by cost and scalability. The integration of model-driven process design and AI feedback control is expected to enable real-time prediction and correction of glycosylation deviations, ensuring batch-to-batch consistency in continuous biomanufacturing. Long-term expression stability, another critical challenge, is often impaired by promoter silencing, chromatin condensation, and random genomic integration. Molecular optimization—such as the use of improved promoters (CMV, EF-1α, or CHO endogenous promoters), Kozak and signal peptide refinement, and incorporation of chromatin-opening elements (UCOE, MAR, STAR)—helps maintain durable transcriptional activity, while site-specific integration systems including Cre/loxP, Flp/FRT, φC31, and CRISPR/Cas9 can enable single-copy, position-independent gene insertion at genomic safe-harbor loci, ensuring stable, predictable expression. Collectively, this review highlights a paradigm shift in CHO system optimization driven by the convergence of genome editing, synthetic biology, and artificial intelligence. The transition from empirical optimization to rational, data-driven design will facilitate the development of programmable CHO platforms capable of autonomous regulation of metabolic flux, glycosylation fidelity, and transcriptional activity. Such intelligent cell factories are expected to accelerate the transformation from laboratory-scale research to industrial-scale, high-consistency, and economically sustainable biopharmaceutical manufacturing, thereby supporting the next generation of efficient and customizable biologics manufacturing.

Chinese hamster ovary (CHO) cells are the most established and versatile mammalian expression system for the large-scale production of recombinant therapeutic proteins, owing to their genetic stability, adaptability to serum-free suspension culture, and ability to perform human-like post-translational modifications. More than 70% of biologics approved by the U.S. Food and Drug Administration rely on CHO-based production platforms, underscoring their central role in modern biopharmaceutical manufacturing. Despite these advantages, CHO systems continue to face three persistent bottlenecks that limit their potential for high-yield, reproducible, and cost-efficient production: excessive metabolic burden during high-density culture, heterogeneity of glycosylation patterns, and progressive loss of long-term expression stability. This review provides an integrated analysis of recent advances addressing these challenges and proposes a forward-looking framework for constructing intelligent and sustainable CHO cell factories. In terms of metabolic regulation, excessive lactate and ammonia accumulation disrupts energy balance and reduces recombinant protein synthesis efficiency. Optimization of culture parameters such as temperature, pH, dissolved oxygen, osmolarity, and glucose feeding can effectively alleviate metabolic stress, while supplementation with modulators including sodium butyrate, baicalein, and S-adenosylmethionine promotes specific productivity (qP) by modulating apoptosis and chromatin structure. Furthermore, genetic engineering strategies—such as overexpression of MPC1/2, HSP27, and SIRT6 or knockout of Bax, Apaf1, and IGF-1R—have demonstrated significant improvements in cell viability and product yield. The combination of multi-omics metabolic modeling with artificial intelligence (AI)-based prediction offers new opportunities for building self-regulating CHO systems capable of dynamic adaptation to environmental stress. Regarding glycosylation uniformity, which determines therapeutic efficacy and immunogenicity, gene editing-based glycoengineering (e.g., FUT8 knockdown or ST6Gal1 overexpression) has enabled the humanization of CHO glycan profiles, minimizing non-human sugar residues and enhancing drug stability. Process-level strategies such as galactose or manganese co-feeding and fine control of temperature or osmolarity further allow rational regulation of glycosyltransferase activity. Additionally, in vitro chemoenzymatic remodeling provides a complementary route to construct human-type glycans with defined structures, though industrial applications remain constrained by cost and scalability. The integration of model-driven process design and AI feedback control is expected to enable real-time prediction and correction of glycosylation deviations, ensuring batch-to-batch consistency in continuous biomanufacturing. Long-term expression stability, another critical challenge, is often impaired by promoter silencing, chromatin condensation, and random genomic integration. Molecular optimization—such as the use of improved promoters (CMV, EF-1α, or CHO endogenous promoters), Kozak and signal peptide refinement, and incorporation of chromatin-opening elements (UCOE, MAR, STAR)—helps maintain durable transcriptional activity, while site-specific integration systems including Cre/loxP, Flp/FRT, φC31, and CRISPR/Cas9 can enable single-copy, position-independent gene insertion at genomic safe-harbor loci, ensuring stable, predictable expression. Collectively, this review highlights a paradigm shift in CHO system optimization driven by the convergence of genome editing, synthetic biology, and artificial intelligence. The transition from empirical optimization to rational, data-driven design will facilitate the development of programmable CHO platforms capable of autonomous regulation of metabolic flux, glycosylation fidelity, and transcriptional activity. Such intelligent cell factories are expected to accelerate the transformation from laboratory-scale research to industrial-scale, high-consistency, and economically sustainable biopharmaceutical manufacturing, thereby supporting the next generation of efficient and customizable biologics manufacturing.

ObjectiveProtein-protein interactions (PPIs) are fundamental to the execution of biological functions within living cells. However, traditional biochemical methods, such as co-immunoprecipitation (Co-IP), often fail to capture transient, weak, or membrane-associated interactions due to the stringent detergent requirements for cell lysis. Proximity labeling (PL) has emerged in recent years as a transformative technology for mapping the proteomes of specific subcellular compartments and identifying dynamic interactomes in situ. Golgi protein 73 (GP73, also known as GOLPH2), a resident type II Golgi transmembrane protein, is a well-recognized clinical biomarker for liver diseases, including hepatocellular carcinoma (HCC). Despite its clinical significance, the comprehensive physiological and pathological functions of GP73 remain partially understood. This study aims to establish an APEX2-mediated proximity labeling system specifically targeting GP73 to map its interactome in a living cellular environment, thereby providing new insights into its molecular roles and regulatory mechanisms. MethodsTo achieve spatial specificity, we first constructed a stable cell line expressing a fusion protein consisting of GP73 and the engineered soybean peroxidase APEX2. The localization of the GP73-APEX2 fusion protein was validated to ensure it correctly targeted the Golgi apparatus. The proximity labeling reaction was initiated by incubating the cells with biotin-phenol (BP) for 30 min, followed by a brief (1 min) treatment with1 mmol/L hydrogen peroxide (H₂O₂). This catalytic reaction converts BP into highly reactive, short-lived biotin-phenoxyl radicals that covalently attach to endogenous proteins within a small labeling radius of the GP73-APEX2 enzyme. Subsequently, the cells were quenched, and biotinylated proteins were enriched using high-affinity streptavidin-coated magnetic beads. The captured “neighbor” proteins were subjected to on-bead digestion and analyzed via liquid chromatography-tandem mass spectrometry (LC-MS/MS) for high-throughput identification. Rigorous bioinformatics analysis, including Gene Ontology (GO) enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and protein-protein interaction network mapping, was performed to interpret the biological significance of the identified candidates. ResultsOur results demonstrate the successful establishment of a robust and sensitive APEX2-based proximity labeling system for GP73. We identified a total of 95 high-confidence interacting proteins that were significantly enriched in the GP73 proximity proteome compared to control groups. Bioinformatics analysis revealed that these interactors were predominantly associated with biological processes such as vesicular transport, protein localization, and, most notably, molecular functions related to “ribosome binding” and “translation regulation”. This suggested an unexpected role for the Golgi-resident GP73 in the cellular translation machinery. To validate these findings, we performed targeted biochemical assays which confirmed a direct interaction between GP73 and the subunits of the eukaryotic translation initiation factor 3 (eIF3) complex, specifically EIF3G and EIF3I. Furthermore, functional validation using the surface sensing of translation (SUnSET) assay—a non-radioactive method to monitor protein synthesis—revealed that the overexpression of GP73 significantly promoted global protein translation levels in the cell, whereas its depletion or inhibition resulted in reduced translation efficiency. ConclusionThis study successfully utilized APEX2-mediated proximity labeling to provide the first systematic map of GP73 interactome in living cells. Our findings uncover a novel, unconventional function of GP73 as a regulator of cellular protein translation, likely mediated through its interaction with the eIF3 complex. This discovery significantly broadens our understanding of the biological roles of GP73 beyond its traditional function in the Golgi apparatus and suggests that it may act as a bridge between Golgi-related trafficking and the protein synthesis machinery. Furthermore, the technical framework established in this study provides a valuable template for investigating other complex organelle-associated protein networks and resolving transient macromolecular interactions in various physiological and pathological contexts.

Objective: To identify early clinical predictors associated with dysphagia and delayed swallowing recovery in patients with traumatic brain injury (TBI). Methods: In this retrospective study, we enrolled adult TBI patients admitted to the rehabilitation unit of a tertiary medical center between June 2019 and June 2023. Data on baseline characteristics, neurological status, imaging findings, and rehabilitation-related variables were collected. Swallowing function was assessed using two indicators: (1) nasogastric (NG) tube retention and (2) the Functional Oral Intake Scale (FOIS) scores at 1, 4, and 12 weeks post-injury. Regression analyses were conducted to identify predictors associated with dysphagia and swallowing recovery. Results: A total of 160 patients were included. At 1 week post-injury, longer intensive care unit (ICU) stay, poor initial sitting balance and use of sedative medication in ICU were associated with NG tube retention. At 4 weeks, lower initial Rancho Los Amigos Scale (RLAS) scores, immobility-related complications, longer hospitalization, and temporal lobe hematomas were associated with persistent NG tube dependence. By 12 weeks, older age, delayed ability to follow commands, and poor initial sitting balance remained associated with NG tube retention. FOIS outcomes were also associated with older age, delayed time to follow commands, impaired initial sitting balance, prolonged ICU stay, temporal lobe hematomas, lower initial RLAS scores, immobility-related complications, prolonged endotracheal tube placement and extended hospital stays. Conclusion Impaired cognitive status, poor physical function, immobility-related complications, and temporal lobe hematomas were key factors associated with dysphagia and delayed oral intake in individuals with TBI.

Objective To investigate the early clinical efficacy of a 3D-printed external fixation guide combined with video-assisted thoracic surgery (VATS) in the treatment of flail chest, and to provide evidence for its clinical application. Methods Patients with flail chest admitted to Xiamen University Affiliated Chenggong Hospital between January 2010 and January 2023 were retrospectively selected as the study subjects. Based on the surgical methods, the patients were divided into two groups: patients treated with the 3D-printed external fixation guide combined with VATS were assigned to a 3D group, and those who underwent open reduction and internal fixation were assigned to an internal fixation group. The operative time, intraoperative blood loss, duration of chest tube drainage, recovery of thoracic volume, visual analogue scale (VAS) scores at 1 month postoperatively, and complications were compared between the two groups. Results A total of 40 patients were included, with 20 patients in each group. The 3D group consisted of 13 males and 7 females, with a mean age of (45.7±3.8) years; the internal fixation group consisted of 14 males and 6 females, with a mean age of (47.3±4.1) years. There were no statistical differences between the two groups in terms of gender, age, number of rib fractures, or preoperative VAS scores (P>0.05). The surgeries were performed successfully in both groups, primary wound healing was achieved in all patients, and pain symptoms were significantly alleviated compared to preoperation. No postoperative complications occurred in the 3D group, whereas 1 patient each of chronic postoperative pain, fracture malunion, and incision infection occurred in the internal fixation group, resulting in a complication rate of 15.0%. The operative time, intraoperative blood loss, and duration of chest tube drainage in the 3D group were significantly shorter or less than those in the internal fixation group (P<0.05). There were no statistical differences in the recovery of thoracic volume or the VAS scores at 1 month postoperatively between the two groups (P>0.05). Conclusion The early clinical efficacy of a 3D-printed external fixation guide combined with VATS in the treatment of flail chest is satisfactory. This technique offers the advantages of being minimally invasive, highly efficient, promoting rapid recovery, and resulting in fewer postoperative complications, and can effectively reconstruct the thoracic contour and restore thoracic volume.

Electroencephalography (EEG) is a non-invasive neurophysiological monitoring technique. It records the electrical activity of the cerebral cortex using electrodes placed on the scalp surface. Owing to its high safety, portability, and millisecond-level temporal resolution, EEG has been widely utilized in a variety of fields, including clinical diagnosis, brain-computer interfaces (BCIs), and cognitive neuroscience research. However, due to its microvolt-level amplitude, EEG is highly susceptible to various artifacts, including electrooculographic (EOG), electrocardiographic (ECG), electromyographic (EMG), and power line interference (PLI). These artifacts can obscure genuine neural activity and introduce spurious electrophysiological features. Consequently, they may compromise EEG signal quality, thereby reducing the reliability of downstream analyses. To address this issue, numerous EEG artifact removal methods have been developed, including both traditional denoising techniques and deep learning-based approaches. Traditional EEG denoising methods have long served as the primary solutions for artifact removal. Representative approaches include filtering, regression, and blind source separation. Although these methods have demonstrated effectiveness in specific scenarios, they suffer from several inherent limitations. Filtering assumes that artifacts and EEG signals can be separated in the frequency domain, but many artifacts, such as EOG and EMG, overlap with EEG spectra, which may lead to the loss of valuable neural information. Regression methods require high-quality artifact references to estimate and subtract contaminations, limiting their effectiveness in reference-free scenarios. Blind source separation can remove artifacts without external references, but it typically requires the number of EEG channels to exceed the number of sources, restricting its application in single- or low-channel EEG recordings. Deep learning-based EEG denoising methods address these limitations effectively. First, they learn the nonlinear mapping between contaminated and clean EEG directly from data in an end-to-end manner. This approach does not rely on assumptions about spectral separability, thereby preserving neural activity more completely. Second, the reference information is incorporated during the training phase, allowing the trained model to perform artifact removal independently without external references. Third, deep learning models can be flexibly designed to accommodate various recording setups, achieving robust denoising for both high-density and single-channel EEG. Collectively, these advantages enable deep learning-based methods to overcome the main challenges of traditional approaches, providing more accurate and reliable EEG signal recovery. The superior denoising performance of deep learning-based EEG denoising methods has attracted increasing attention in EEG artifact removal research. As a result, many deep learning-based denoising methods have been developed and successfully applied in neural engineering areas. However, a systematic review of the techniques and applications in this field is still lacking. To address this gap, this paper reviews recent advances in deep learning-based EEG denoising from four perspectives: technical principle, benchmark dataset, denoising model, and evaluation method. Representative applications in neural signal analysis and BCI decoding are also summarized. Furthermore, the advantage, existing challenge, and future research direction of deep learning-based EEG denoising are discussed. This review aims to provide valuable theoretical insights and technical guidance for researchers. It is also expected to promote further advances and broader applications of deep learning-based EEG denoising techniques.

Objective : To observe the brain tissue injury during hypoxia-ischemia, as well as the pathological changes and the expression of ferroptosis-related factors after the use of Shenfu injection(SFI), and to explore the protective effect of SFI on hypoxic-ischemic brain injury(HIBD) by inhibiting ferroptosis. Methods : An animal model of HIBD in SD rats was constructed and intervened with SFI. Pathologic changes in brain tissue were observed by HE staining methods. Nissen staining was used to observe neuron survival. Glutathione Peroxidase 4(GPX4) and Divalent Metal Transporter 1(DMT1) expression were detected in brain tissue by Western blot, immunohistochemistry and immunofluorescence. Reduced Glutathione(GSH), Lactate Dehydrogenase(LDH), Malondialdehyde(MDA), Superoxide Dismutase(SOD) and tissue iron content were determined with the kits. BV-2 microglial cell line(BV2) cells were culturedin vitroand divided into control group(Ctrl group), oxygen-glucose deprivation group(OGD group), iron ferroptosis-inducing group(Erastin group), iron ferroptosis-inhibiting group(Fer-1 group), Shenfu injection group(SFI group), and Erastin+Shenfu injection group(Erastin+SFI group). 2′,7′-Dichlorodihydrofluorescein diacetate(DCFH-DA) reactive oxygen species(ROS) fluorescent probe was used to detect the ROS release level; Immunofluorescence was used to observe intracellular GPX4, DMT1 expression. Results : Compared with the Sham group, rats in the HIBD group showed significant neuronal cell damage in brain tissue, decreased GPX4 expression(P<0.01), increased DMT1 expression(P<0.01), decreased GSH and SOD levels(P<0.01), and increased LDH, MDA and tissue iron levels(P<0.05,P<0.05,P<0.01). In contrast, after the intervention of SFI, GPX4 expression was elevated(P<0.01), DMT1 expression decreased(P<0.01), GSH and SOD levels were elevated(P<0.01), and LDH, MDA, and tissue iron levels decreased(P<0.05,P<0.05,P<0.01). The cells experiments showed that compared with the Ctrl group, the OGD group had a significantly higher ROS content and a decrease in the expression of GPX4 fluorescence intensity, and an increase in the fluorescence intensity of DMT1(P<0.01), compared with the OGD group, the ROS content was reduced in the SFI group, while the expression of GPX4 was elevated and the expression of DMT1 was reduced(P<0.01). Conclusion Hippocampal and cortical regions are severely damaged after HIBD in neonatal rats, and their brain tissues show decreased expression of GPX4 and increased expression of DMT1. The above suggests that ferroptosis is involved in HIBD brain injury in neonatal rats. In contrast, Shenfu injection has a protective effect on HIBD experimental animal model and BV2 cell injury model by reducing iron aggregation and ROS production.

In recent years, the deep integration of artificial intelligence (AI) into medical education has created new opportunities for teaching Biochemistry and Molecular Biology, while also offering innovative solutions to the pedagogical challenges associated with protein structure and function. Focusing on the case of anaplastic lymphoma kinase (ALK) gene mutations in non-small-cell lung cancer (NSCLC), this study integrates AI into case-based learning (CBL) to develop an AI-CBL hybrid teaching model. This model features an intelligent case-generation system that dynamically constructs ALK mutation scenarios using real-world clinical data, closely linking molecular biology concepts with clinical applications. It incorporates AI-powered protein structure prediction tools to accurately visualize the three-dimensional structures of both wild-type and mutant ALK proteins, dynamically simulating functional abnormalities resulting from conformational changes. Additionally, a virtual simulation platform replicates the ALK gene detection workflow, bridging theoretical knowledge with practical skills. As a result, a multidimensional teaching system is established—driven by clinical cases and integrating molecular structural analysis with experimental validation. Teaching outcomes indicate that the three-dimensional visualization, dynamic interactivity, and intelligent analytical capabilities provided by AI significantly enhance students’ understanding of molecular mechanisms, classroom engagement, and capacity for innovative research. This model establishes a coherent training pathway linking “fundamental theory-scientific research thinking-clinical practice”, offering an effective approach to addressing teaching challenges and advancing the intelligent transformation of medical education.

Objective: To explore the relationship of physical activity (PA) and sedentary behavior (SB) with cardiorespiratory fitness (CRF) among middle schoold students in Tibet, so as to provide empirical references for improving the cardiorespiratory fitness and health levels of adolescents in Tibet. Methods: From August to December 2020, 1 225 junior and senior high school students were selected from 2 middle schools in Lhasa, Tibet Autonomous Region, using the stratified cluster random sampling method. Triaxial accelerometers were used to evaluate PA and SB behaviors, and the 20 meter shuttle run was employed to assess CRF among the middle school students. Isochronous substitution modeling was used to analyze the associations of SB, low intensity physical activity (LPA), and moderate vigorous physical activity (MVPA) with CRF, and the saturation threshold effect in the dose response relationship between MVPA and CRF was analyzed through restricted cubic spline and two stage linear regression. Results: After adjusting for covariates such as gender, body mass index and sleep quality score, isotemporal substitution analysis showed that among junior high school students aged 13-15, replacing 30 minutes of SB ( B =1.73) or LPA ( B =2.38) with MVPA were positively associated with CRF (both P <0.05). Among senior high school students aged 16-18, replacing SB ( B =0.99) or LPA ( B =1.38) with MVPA were also positively associated with CRF (both P <0.05). Restricted cubic spline and two piecewise linear regression analyses indicated that only middle school girls aged 13-18 exhibited a saturation threshold effect between MVPA and CRF (logarithmic likelihood ratio test=0.03), with the optimal CRF improvement observed at 60 minutes of MVPA per day ( B=0.13, P ＜ 0.01). Conclusions Reducing SB and LPA while increasing MVPA can improve CRF in Tibetan middle school students. To maximize CRF improvement, middle school girls should engage in at least 60 minutes of MVPA daily.