Objective: This study aims to identify Alzheimer’s disease (AD)-related miRNAs according to meta-analysis of miRNA controlled profiling studies of AD. Methods: MiRNA controlled profiling studies of AD were systematically searched and screened in the PubMed. Meta-analysis was conducted on eligible studies under a random effects model. Dysregulated miRNAs of AD were reported with odd ratios, 95% confidence intervals, P values and mean fold changes. Subgroup analyses based on species, tissues, post-mortem intervals, diagnostic criteria and Braak stag examined the sources of heterogeneity among studies. Sensitivity analysis based on miRNA detection methods tested the robustness of meta-analysis results. Dysregulated miRNAs identified in meta-analysis were visualized in miRNA-target interaction networks. Further gene set functional enrichment analysis predicted the targets function of dysregulated miRNAs in AD. Results: A total of 54 miRNA controlled profiling studies with 469 AD-related miRNAs were included in this meta-analysis. Fifty six miRNAs (36 up-regulated miRNAs and 20 down-regulated miRNAs) were identified as dysregulated miRNAs (P<0.05) by meta-analysis. MiR-34c and miR-384 were the most frequently studied and definite up-regulated and down-regulated miRNAs, respectively. Subgroup analysis found eight miRNAs (miR-9, miR-125b, miR-34c, miR-342-5p, miR-98, miR-181a-5p, miR-384 and miR-146a) dysregulated in at least two tissues; while miR-34c and miR-384 were consistently dysregulated in the brain and blood. Sensitivity analysis showed 32 dysregulated miRNAs consistent with overall effects of meta-analysis. Gene set functional enrichment analysis showed that the target genes of the dysregulated miRNAs were involved in regulation of protein phosphorylation, cellular and macromolecular metabolic processes. Conclusion: This study identified 56 dysregulated miRNAs of AD. Among them, miR-34c and miR-384 would be potential blood biomarkers of AD.

目的: 進行阿爾茲海默症 miRNA 表達譜研究的薈萃分析，確定與阿爾茲海默症 相關的 miRNA。 方法: 在 PubMed 中系統檢索文獻，篩選阿爾茲海默症 miRNA 表達譜研究，採 用隨機效應模型對差異表達的 miRNA 進行薈萃分析，報告每個 miRNA 表達的 優勢比、95%置信區間、P 值和平均的改變倍數。以種族、樣本組織、解剖間隔 時間、診斷標準、神經原纖維化改變嚴重程度為基礎進行亞組分析檢驗不同阿爾 茲海默症 miRNA 表達譜研究結果異質性的來源，以 miRNA 檢測方法為基礎的 敏感性分析檢驗薈萃分析結果的穩定性。構建薈萃分析得到的 miRNA 及其靶點 相互作用網絡圖，並對差異表達 miRNA 的靶點集進行功能富集分析。 結果: 共54篇符合納入的阿爾茲海默症miRNA表達譜研究共涉及469個miRNA。 薈萃分析確定其中 56 個(36 個上調，20 個下調)差異性表達的 miRNA (P<0.05)。 其中 miR-34c 和 miR-384 分別是被研究最多最明確的上調和下調 miRNA。在基 於樣本組織的亞組分析中，8 個 miRNA (miR-9, miR-125b, miR-34c, miR-342-5p, miR-98, miR-181a-5p, miR-384 和 miR-146a)至少在兩種組織中呈現出差異性表 達，其中 miR-34c 和 miR-384 在腦組織和血液組織中具有一致的表達方向。敏感 性分析顯示有 32 個 miRNA 與薈萃分析結果異常表達的方向一致。靶基因集功 能富集分析顯示差異性表達 miRNA 的靶基因可能與蛋白質磷酸化、調節細胞代 謝過程和大分子代謝過程等生化過程有關。 結論: 薈萃分析確認了在阿爾茲海默症中56個異常表達的miRNA，其中miR-34c 和 miR-384 可能成為阿爾茲海默症的血液生物標記。 關鍵字: 阿爾茲海默症，薈萃分析，miRNA。