Table_1_The shared molecular mechanism of spinal cord injury and sarcopenia: a comprehensive genomics analysis.docx

IntroductionThe occurrence of Spinal cord injury (SCI) brings economic burden and social burden to individuals, families and society, and the complications after SCI greatly affect the rehabilitation and treatment of patients in the later stage.This study focused on the potential biomarkers that co-exist in SCI and sarcopenia, with the expectation to diagnose and prognose patients in the acute phase and rehabilitation phase using comprehensive data analysis.
MethodsThe datasets used in this study were downloaded from Gene Expression Omnibus (GEO) database. Firstly, the datasets were analyzed with the “DEseq2” and “Limma” R package to identify differentially expressed genes (DEGs), which were then visualized using volcano plots. The SCI and sarcopenia DEGs that overlapped were used to construct a protein–protein interaction (PPI) network. Three algorithms were used to obtain a list of the top 10 hub genes. Next, validation of the hub genes was performed using three datasets. According to the results, the top hub genes were DCN, FSTL1, and COL12A1, which subsequently underwent were Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses. We also assessed immune cell infiltration with the CIBERSORT algorithm to explore the immune cell landscape. The correlations between the hub genes and age and body mass index were investigated. To illustrate the biological mechanisms of the hub genes more clearly, a single-cell RNA-seq dataset was assessed to determine gene expression when muscle injury occurred. According to our analysis and the role in muscle, we chose the fibro/adipogenic progenitors (FAPs) cluster in the next step of the analysis. In the sub cluster analysis, we use the “Monocle” package to perform the trajectory analysis in different injury time points and different cell states.
ResultsA total of 144 overlapped genes were obtained from two datasets. Following PPI network analysis and validation, we finally identified three hub-genes (DCN, FSTL1, and COL12A1), which were significantly altered in sarcopenic SCI patients both before and after rehabilitation training. The three hub genes were also significantly expressed in the FAPs clusters. Furthermore, following injury, the expression of the hub genes changed with the time points, changing in FAPs cluster.
DiscussionOur study provides comprehensive insights into how muscle changes after SCI are associated with sarcopenia by moving from RNA-seq to RNA-SEQ, including Immune infiltration landscape, pesudotime change and so on. The three hub genes identified in this study could be used to distinguish the sarcopenia state at the genomic level. Additionally, they may also play a prognostic role in evaluating the efficiency of rehabilitation training.

**引言**
脊髓损伤（Spinal cord injury, SCI）会给个体、家庭及社会带来经济与社会负担，而脊髓损伤后并发症会严重影响患者后期的康复与治疗效果。本研究聚焦于脊髓损伤与肌少症（sarcopenia）共有的潜在生物标志物，期望通过综合数据分析实现患者急性期与康复期的诊断及预后评估。

**方法**
本研究使用的数据集均下载自基因表达综合数据库（Gene Expression Omnibus, GEO）。首先，采用R语言的"DEseq2"与"Limma"包对数据集进行分析，以筛选差异表达基因（differentially expressed genes, DEGs），并通过火山图（volcano plots）对其进行可视化。将脊髓损伤与肌少症的重叠差异表达基因构建蛋白质-蛋白质相互作用（protein–protein interaction, PPI）网络，通过三种算法筛选得到排名前10的核心基因（hub genes）。随后，利用三个独立数据集对核心基因进行验证。本研究最终筛选出的核心基因为DCN、FSTL1及COL12A1，后续对其开展基因本体（Gene Ontology, GO）与京都基因与基因组百科全书（Kyoto Encyclopedia of Genes and Genomes, KEGG）富集分析。同时，采用CIBERSORT算法评估免疫细胞浸润情况，以探究免疫细胞谱特征。此外，本研究还分析了核心基因与年龄、体质量指数的相关性。为更清晰地阐明核心基因的生物学机制，本研究纳入单细胞RNA测序（single-cell RNA-seq）数据集，以明确肌肉损伤时的基因表达模式。基于分析结果及该类细胞在肌肉损伤中的作用，本研究后续分析聚焦于成纤维/脂肪祖细胞（fibro/adipogenic progenitors, FAPs）聚类群。在亚聚类分析中，采用"Monocle"包针对不同损伤时间点与不同细胞状态开展轨迹分析（trajectory analysis）。

**结果**
本研究从两个数据集中共筛选得到144个重叠基因。经PPI网络分析与验证后，最终确定3个核心基因（DCN、FSTL1及COL12A1），其在伴肌少症的脊髓损伤患者康复训练前后均存在显著表达差异。这3个核心基因在成纤维/脂肪祖细胞聚类群中同样呈现显著表达。进一步分析发现，损伤后核心基因的表达随时间进程发生变化，且该变化发生于成纤维/脂肪祖细胞聚类群中。

**讨论**
本研究从RNA测序到单细胞RNA测序，全面解析了脊髓损伤后肌肉变化与肌少症的关联机制，涵盖免疫细胞浸润谱、拟时（pseudotime）变化等多个维度。本研究鉴定的3个核心基因可在基因组层面区分肌少症状态，此外，它们或可用于评估康复训练的疗效，发挥预后评估的作用。