应用单细胞RNA测序分析椎间盘退变中纤维环细胞-巨噬细胞的相互作用及其关键通路

Single-cell RNA sequencing reveals the interaction between annulus fibrosus cells and macrophages and its key pathways in intervertebral disc degeneration

中文关键词:  椎间盘退变  巨噬细胞  纤维环细胞  单细胞测序  SPP1-CD44通路

英文关键词:Intervertebral disc degeneration  Macrophage  Annulus fibrosus cell  Single-cell RNA sequencing  SPP1-CD44 pathway

中文摘要:

  【摘要】 目的：通过单细胞RNA测序阐明免疫细胞与纤维环细胞相互作用的时空动态特征。方法：实验采用12只8周龄雄性Sprague Dawley大鼠，并随机分为3组：正常对照组（NT）、针刺退变7d组（IDD7）及针刺退变14d组（IDD14），每组三只大鼠的椎间盘组织合并作为一个scRNA-seq样本（即NT/IDD7/IDD14各1个样本，n=3大鼠/组）。椎间盘组织经过混合的多种胶原酶消化后，并未进行流式分选，直接利用BD Rhapsody平台，对3个样本开展了单细胞RNA测序（scRNA-seq）。通过以下流程解析细胞动态变化：采用Seurat对原始计数矩阵进行质量控制、标准化及批次校正（Harmony算法整合），保留基因表达均值用于后续分析。通过FindVariableFeatures筛选2000个高变基因，经主成分分析（principal component analysis，PCA）降维后，采用 Louvain聚类算法（分辨率0.2）对细胞进行分群，并结合SingleR工具进行细胞类型注释，获取各亚群的标志基因。对比正常对照组（NT）、针刺退变7d组（IDD7）及针刺退变14d组（IDD14）间的差异表达基因（筛选阈值：|logFC|>0.5，P<0.05）。使用基于R语言开发的生物信息学软件包ClusterProfiler与enrichR软件分别对差异基因进行GO（gene ontology）生物学过程与KEGG（kyoto encyclopedia of genes and genomes）通路富集分析（P<0.05，基因数≥2），筛选显著相关通路。基于Monocle3构建伪时间轨迹，将整合数据转换为cell_data_set对象，经UMAP（uniform manifold approximation and projection）降维与Louvain聚类后，利用learn_graph与order_cells函数揭示细胞状态演化路径。运用CellChat与CellPhoneDB分析配体-受体相互作用，计算细胞间通讯概率，可视化关键信号通路（如细胞因子、生长因子）在NT与IDD组间的动态变化。采用免疫荧光技术对 CD68（巨噬细胞标记物）、SPP1、CD44等蛋白的共定位表达进行定量分析。结果：单细胞数据鉴定出9个主要细胞群：纤维环细胞（NP）、纤维环细胞（AF）、巨噬细胞（Mφ）、单核细胞、中性粒细胞、T细胞、成纤维细胞、内皮细胞及平滑肌细胞。74个基因持续差异表达，主要与免疫细胞趋化相关。并系统解析了纤维环细胞的时序性分子变化。纤维环细胞存在修复型（Fibro-1/3）与耗竭型（Fibro-2）两极分化，为精准干预提供靶群依据。拟时序T1期可能（对应IDD7d）是应激保护向纤维化转化的关键节点，为阻断不可逆损伤提供治疗窗口期。细胞互作动态演变：纤维环亚群中耗竭型Fibro-2与巨噬细胞的互作数最多，SPP1信号轴由Fibro-2主导发送，Mφ-3为主要接收者，该通路可能驱动纤维环细胞向促纤维化表型转化，提示其可能参与基质破坏和纤维化重塑进程。结论：纤维环细胞可能通过SPP1-CD44信号轴增强与巨噬细胞的相互作用，进而激活炎症反应并加速基质降解。这一发现揭示了椎间盘退变过程中的关键细胞受体-配体对，为IDD的精准干预提供了新的潜在治疗靶点。

英文摘要:

  【Abstract】 Objectives: To elucidate the spatiotemporal dynamics of immune cell-annulus fibrosus interactions through single-cell RNA sequencing(scRNA seq). Methods: Twelve 8-week-old male Sprague Dawley rats were randomly divided into three groups, normal control(NT), needle puncture-induced degeneration 7d(IDD7), and needle puncture-induced degeneration 14d(IDD14) groups. Intervertebral disc tissues from three rats per group were pooled as one scRNA-seq sample(i.e., one NT/IDD7/IDD14 sample each, n=3 rats/group). Following digestion with a mixture of collagenases without flow cytometry sorting, scRNA-seq was performed for the three samples using the BD Rhapsody platform. Cellular dynamics were analyzed via the following pipeline: Quality control, normalization, and batch-effect correction(using Harmony algorithm) of the raw count matrix were performed with Seurat, retaining mean gene expression for subsequent analysis. 2000 highly variable genes were selected using the FindVariableFeatures method. Following PCA dimensionality reduction, cells were clustered via the Louvain algorithm (resolution=0.2) and subsequently annotated using SingleR to identify marker genes for each subpopulation. Differentially expressed genes (DEGs) among the NT, IDD7, and IDD14 groups were identified(thresholds: |logFC|>0.5, P<0.05). Functional enrichment analyses for gene ontology(GO) biological processes and KEGG pathways were performed on the DEGs using ClusterProfiler and enrichR(P<0.05, gene count ≥2) to screen for significantly associated pathways. A pseudotemporal trajectory was constructed with Monocle3 by converting the integrated data into a cell_data_set object, performing UMAP dimensionality reduction and Louvain clustering, and then employing the learn_graph and order_cells functions to delineate the cell state evolution path. Cell-cell communication was analyzed using CellChat and CellPhoneDB to calculate interaction probabilities and visualize the dynamic changes of key signaling pathways(e.g., cytokines, growth factors) between NT and IDD conditions. Co-localization and quantitative expression analysis of proteins such as CD68(a macrophage marker), SPP1, and CD44 were conducted using immunofluorescence staining. Results: The scRNA-seq data identified nine major cell populations: nucleus pulposus cells(NP), annulus fibrosus cells(AF), macrophages(Mφ), monocytes, neutrophils, T cells, fibroblasts, endothelial cells, and smooth muscle cells. 74 genes were consistently differentially expressed, predominantly associated with immune cell chemotaxis. Temporal molecular changes in AF cells were systematically delineated. AF cells exhibited polarization into reparative(Fibro-1/3) and exhausted(Fibro-2) subtypes, providing potential target subpopulations for precise intervention. The T1 phase in pseudotime(corresponding to IDD7d) likely represented a critical juncture transitioning from stress-protective responses to fibrosis, offering a potential therapeutic window to prevent irreversible damage. Analysis of dynamic cellular interactions revealed that the exhausted Fibro-2 subtype engaged in the highest number of interactions with macrophages. The SPP1 signaling axis was predominantly initiated by Fibro-2 cells, with Mφ-3 as the primary recipient, suggesting this pathway may drive AF cells toward a pro-fibrotic phenotype and potentially contribute to matrix degradation and fibrotic remodeling. Conclusions: AF cells may enhance their interaction with macrophages via SPP1-CD44, activating inflammatory responses and accelerating matrix degradation. This reveals the temporal regulation of cellular interactions in intervertebral disc degeneration models and provides new targets for precise intervention in IDD.