[摘 要] 目的：探讨绿茶单体表没食子儿茶素没食子酸酯（EGCG）对大鼠肝癌发生过程中肠道菌群结构变化的影响。方法：建立二乙基亚硝胺（DEN）诱导的肝癌大鼠模型。将26只SD大鼠随机分成3组，分别为正常对照组、肝癌组和EGCG干预组。从实验开始第1天起，EGCG干预组每日给予EGCG（40 mg/kg）灌胃，正常对照组和肝癌组给予等量生理盐水灌胃，1次/日，持续至第20周。灌胃结束后，采集大鼠粪便样本，提取DNA，进行高通量16S rRNA V3-V4区测序；处死大鼠，取肝，观察肿瘤形成情况，并计算肝癌发生率。对测序数据进行生物信息学分析：经质控、聚类获得操作分类单元（OTU）表，据此计算α多样性指数（包括Observed species、Chao1、Shannon和Simpson指数），并进行β多样性分析。同时，对物种进行分类学注释，比较各组间菌群组成与丰度差异。结果： EGCG干预组大鼠（8只）肝脏肿瘤形成率明显低于肝癌组（10只）（50% vs 100%， P = 0.023），正常对照组（8只）大鼠无肿瘤发生。在肠道菌群方面，肝癌组操作分类单元（OTU）数量远低于正常对照组（P <0.001），而EGCG干预组OTU数量总体上高于肝癌组（P = 0.021）。α多样性分析显示，肝癌组Shannon指数低于正常对照组（P < 0.05）；此外，与肝癌组相比，EGCG干预组的Observed species指数、Chao1指数、Shannon指数和Simpson指数均显著提高（P < 0.05）。β多样性分析及主坐标分析（PCoA）表明，三组肠道菌群结构存在显著分离（PERMANOVA， R² = 0.3918， P = 0.001），其中EGCG干预组群落结构介于肝癌组与正常对照组之间，并更接近于正常对照组。肝癌组大鼠较正常大鼠肠道菌群中链球菌等潜在致病菌富集，丁酸产生相关菌（如丁酸球菌属、瘤胃球菌属等）丰度显著降低（P < 0.05）。相比之下，在EGCG干预肝癌发生过程中，大鼠肠道菌群结构相对稳定。厚壁菌门/拟杆菌门比值较肝癌组显著提高（P < 0.05），益生菌（如双歧杆菌属、乳杆菌属等）和丁酸产生相关菌（如丁酸球菌属）富集。结论：EGCG干预可降低DEN诱导的大鼠肝癌发生率，并有助于稳定肠道菌群结构，其作用可能与增加菌群多样性、促进益生菌及丁酸产生菌富集、恢复菌群平衡有关。

Objective To establish an analytical strategy covering multi-dataset processing, recursive feature elimination (RFE) screening and multi-model evaluation based on multiple machine learning algorithms, so as to screen radiation-sensitive genes and verify the feasibility of the evaluation strategy. Methods Qualified radiation transcriptome datasets were retrieved from public gene expression databases. Following standardized data preprocessing and feature preselection, 13 machine learning algorithms were adopted to construct models. The performance of each model was compared and validated in independent datasets. Results A total of 38 eligible datasets were included. Sixteen differentially expressed genes unreported in existing literature were screened out, among which ugcrhl, pdcl3, mct4, h2-g2 and fam120aos were correlated with radiation phenotypes. Ensemble learning algorithms including random forest and gradient boosting exhibited the optimal comprehensive performance. Independent dataset verification confirmed that the screened genes overlapped with known radiation-sensitive genes, and the model performance was consistent with the findings. Conclusion The machine learning strategy constructed in this study can effectively explore potential radiation-sensitive genes, and provides methodological support for subsequent relevant studies.

Objective To establish an analytical strategy covering multi-dataset processing, recursive feature elimination (RFE) screening and multi-model evaluation based on multiple machine learning algorithms, so as to screen radiation-sensitive genes and verify the feasibility of the evaluation strategy. Methods Qualified radiation transcriptome datasets were retrieved from public gene expression databases. Following standardized data preprocessing and feature preselection, 13 machine learning algorithms were adopted to construct models. The performance of each model was compared and validated in independent datasets. Results A total of 38 eligible datasets were included. Sixteen differentially expressed genes unreported in existing literature were screened out, among which ugcrhl, pdcl3, mct4, h2-g2 and fam120aos were correlated with radiation phenotypes. Ensemble learning algorithms including random forest and gradient boosting exhibited the optimal comprehensive performance. Independent dataset verification confirmed that the screened genes overlapped with known radiation-sensitive genes, and the model performance was consistent with the findings. Conclusion The machine learning strategy constructed in this study can effectively explore potential radiation-sensitive genes, and provides methodological support for subsequent relevant studies.

Objective To establish an analytical strategy covering multi-dataset processing, recursive feature elimination (RFE) screening and multi-model evaluation based on multiple machine learning algorithms, so as to screen radiation-sensitive genes and verify the feasibility of the evaluation strategy. Methods Qualified radiation transcriptome datasets were retrieved from public gene expression databases. Following standardized data preprocessing and feature preselection, 13 machine learning algorithms were adopted to construct models. The performance of each model was compared and validated in independent datasets. Results A total of 38 eligible datasets were included. Sixteen differentially expressed genes unreported in existing literature were screened out, among which ugcrhl, pdcl3, mct4, h2-g2 and fam120aos were correlated with radiation phenotypes. Ensemble learning algorithms including random forest and gradient boosting exhibited the optimal comprehensive performance. Independent dataset verification confirmed that the screened genes overlapped with known radiation-sensitive genes, and the model performance was consistent with the findings. Conclusion The machine learning strategy constructed in this study can effectively explore potential radiation-sensitive genes, and provides methodological support for subsequent relevant studies.

As one of the common systemic vasculitides in children， IgA vasculitis is closely associated with the interaction of "Wind pathogen" and "turbid stasis" in its pathogenesis. Based on the blood turbidity theory， this paper proposes "wind agitating blood turbidity and collateral damage" as the core pathogenesis. Regarding treatment， the sequential therapeutic principle of "stanch bleeding， activating blood circulation， and nourishing blood" is adopted， and combined with the dynamic application of wind-dispelling medicinals to "expel， search， and guard against wind"， turbidity-removing method is applied throughout all stages. In the acute stage，it is recommended to clear heat and purify turbidity， stanch bleeding and secure the collaterals. In the prolonged stage， the method of dissipating stasis and clearing up turbidity should be used， as well as activating blood circulation and unblocking the collaterals. In the recovery stage， it is advised to nourish blood to flourish the collaterals， reinforce healthy qi and resolve turbidity. This therapeutic system integrates the elimination of pathogens and the reinforcement of healthy qi to clear and resolve blood turbidity， ultimately achieving holistic collateral repair through securing， unblocking， and nourishing the collaterals， which provides new insights into the treatment of this disease in traditional Chinese medicine.

Based on the theory of "cold qi generates turbidity" described in the Inner Canon of Yellow Emperor （《黄帝内经》）， together with the physiological characteristics of children， who are considered to have an inherent insufficiency of the lung， spleen， and kidney， and the clinical features of recurrent refractory nephrotic syndrome （RNS）， this article proposes that the pathogenesis of RNS in children is characterized by yang deficiency with internal cold， turbidity arising from cold， latent turbidity transforming into toxin， activation by newly contracted pathogens， and turbid toxins damaging the collaterals. The core pathogenesis is identified as yang deficiency， turbidity coagulation， and collateral damage. Accordingly， the general treatment principle is warming yang and eliminating turbidity， with coordination of three methods including warming yang to strengthen the body's foundation， promoting yang to relieve constraint， and raising yang while replenishing qi. Zhenwu Decoction （真武汤） is employed as the basic prescription with modification according to syndrome differentiation， to achieve an effect eliminating turbidity and arresting leakage， restoring kidney function by unblocking the collaterals， and consolidating recovery to prevent relapse.

OBJECTIVE: To identify prognostic genes associated with lysosome-dependent cell death (LDCD) in patients with gastric cancer (GC). METHODS: Differentially expressed genes (DEGs) were identified using The Cancer Genome Atlas - Stomach Adenocarcinoma. Weighted gene co-expression network analysis was performed to identify the key module genes associated with LDCD score. Candidate genes were identified by DEGs and key module genes. Univariate Cox regression analysis, and least absolute shrinkage and selection operator regression and multivariate Cox regression analyses were performed for the selection of prognostic genes, and risk module was established. Subsequently, key cells were identified in the single-cell dataset (GSE183904), and prognostic gene expression was analyzed. Cell proliferation and migration were assessed using the Cell Counting Kit-8 assay and the wound healing assay. RESULTS: A total of 4,465 DEGs, 95 candidate genes, and 4 prognostic genes, including C19orf59, BATF2, TNFAIP2, and TNFSF18, were identified in the analysis. Receiver operating characteristic curves indicated the excellent predictive power of the risk model. Three key cell types (B cells, chief cells, and endothelial/pericyte cells) were identified in the GSE183904 dataset. C19orf59 and TNFAIP2 exhibited predominant expression in macrophage species, whereas TNFAIP2 evolved over time in endothelial/pericyte cells and chief cells. Functional experiments confirmed that interfering with C19orf59 inhibited proliferation and migration in GC cells. CONCLUSION C19orf59, BATF2, TNFAIP2, and TNFSF18 are prognostic genes associated with LDCD in GC. Furthermore, the risk model established in this study showed robust predictive power.
Stomach Neoplasms/pathology* ; Humans ; Prognosis ; Lysosomes/physiology* ; RNA-Seq ; Cell Death ; Single-Cell Analysis ; Gene Expression Regulation, Neoplastic ; Cell Proliferation ; Single-Cell Gene Expression Analysis

Objective:To investigate the changes in the expression levels of long non-coding ribonucleic acids (lncRNAs) in human lymphocytes induced by low-dose ionizing radiation (LDIR) and the potential of lncRNAs as radiation biomarkers.Methods:Human immortalized lymphocytes (AHH-1) were irradiated with 0, 0.05, and 0.1 Gy of γ-rays at 24 h to extract RNAs for whole transcriptome sequencing. The sequencing was performed based on the 0, 0.05, and 0.1 Gy groups. The differentially expressed lncRNAs induced by LDIR were identified. The molecular functions, biological processes, and signaling pathway enrichment of differentially expressed genes were analyzed through the Gene Ontology (GO) analysis. Candidate lncRNAs were preliminarily validated using the qRT-PCR method. AHH-1 cells were irradiated with 0, 0.02, 0.05, 0.075, 0.1, and 0.2 Gy to extract the total RNAs at 4, 24, 48, 72, 96, and 120 h. The dose-response relationship of candidate lncRNAs was detected and analyzed. Peripheral blood sampled from eight healthy persons was irradiated with 0, 0.02, 0.05, 0.075, 0.1, and 0.2 Gy in vitro, followed by culturing for 24 h and 48 h to further verify the changes in the expression levels of radiation-responsive lncRNAs at the cellular level. Results:A total of 44 lncRNAs that were significantly up- or down-regulated after 0.05 and 0.1 Gy irradiation were initially identified through transcriptome sequencing. Among them, lncRNAs with over two-fold differential expression included SNHG1, SNHG15, NEAT1, and PRC1-AS1. At the cellular level, compared to 0 Gy, the relative expression level of PRC1-AS1 after 4 h to 48 h of γ-ray irradiation, was significantly elevated at 0.05, 0.075, and 0.1 Gy( t= -3.11 to 1.23, P < 0.05). In contrast, the relative expression level of NEAT1 was significantly up-regulated in a dose range of 0.02 to 0.1 Gy ( t=-2.47 to 2.10, P < 0.05). At the level of human peripheral blood, the relative expression levels of PRC1-AS1 and NEAT1 were significantly increased at 24 h after 0 to 0.2 Gy irradiation ( t=-3.79 to -1.96, P < 0.05). Conclusion:The PRC1-AS1 and NEAT1 with significant changes in expression levels serve as potential LDIR biomarkers.

Objective:To develop a bacterial endotoxin detection method suitable for mRNA vaccines based on lipid nanoparticle（LNP）delivery system.Methods:The gel clot limit test outlined in the third section of the Chinese Pharmacopoeia（2020 edition）was utilized. Taking the human herpesvirus type 2 mRNA vaccine as the detection object，the optimal conditions for bacterial endotoxin detection were established by evaluating sample dilution factors，diluents，and other interfering elements. The results were validated using the recombinant C factor method. Finally，the applicability of the established method was validated through testing the respiratory syncytial virus mRNA vaccine and the shingles mRNA vaccine.Results:Increasing the sample dilution factor to 480 times with a magnesium-containing buffer and using 0.06 EU/ml tachypleus amebocyte lysate effectively mitigated the interference caused by LNP on the test results of the human herpesvirus type 2 mRNA vaccine，the respiratory syncytial virus mRNA vaccine，and the shingles mRNA vaccine. The results from recombinant C factor method and the gel clot limit method were consistent.Conclusion:A bacterial endotoxin gel clot limit detection method tailored for human herpesvirus type 2 mRNA vaccine，respiratory syncytial virus mRNA vaccine，and shingles mRNA vaccine is successfully developed based on existing protocols in the Chinese Pharmacopoeia（2020 edition）. This study offers insights for the detection of bacterial endotoxin in LNP-based biological products.

Objective:To explore the application and clinical significance of pathological diagnostic criteria for medically refractory epilepsy in children.Methods:Cross-sectional study.A retrospective analysis was conducted on 490 children(pathology involved) with medically refractory epilepsy treated continuously in the Pediatric Epilepsy Center of Peking University First Hospital from January 2019 to May 2022.The distribution of different pathological types was observed, and the differences in clinical characteristics among different pathological types were analyzed through Kruskal-Wallis or χ2 tests.The impact of clinical and pathological features on patient prognosis was evaluated through regression analysis. Results:Focal cortical dysplasia (FCD) was the predominant lesion (49.59%, 243/490).The electroencephalograms ( χ2=6.720, P=0.035) and clinical seizure characteristics ( χ2=26.370, P<0.001) in FCDⅡ were more focal than those in FCD Ⅰ and Ⅲ.Moreover, the proportions of focal resection in surgery ( χ2=24.286, P<0.001) and central involvement ( χ2=22.849, P<0.001) in FCDⅡ were higher than those in FCD Ⅰ and Ⅲ.Univariate and multivariate regression analyses revealed that FCD Ⅱ had a better prognosis than other dysplastic patients among the 375 cases of dysplasia ( P=0.049).Next-generation sequencing was performed on 35 cases of cortical malformations with such morphological characteristics as increased numbers of neurons in the white matter and Olig2-positive glial cell hyperplasia, and SLC35A2 mutations were detected in 2 cases (5.71%). Conclusions:Pathology of refractory epilepsy is specialized and continuously evolving.Standardized specimen processing and the accumulation of morphological, immunohistochemical, and molecular genetic data provide the foundation for clarifying the neuropathological nature of epilepsy, improving integrated classification, and advancing prognosis prediction and targeted therapy.