Mechanisms underlying the development of malignant retinoblastoma (RB) remain largely unknown. The purpose of this study was to identify weighted genes that are associated with the progression of RB and to assess the usefulness of bioinformatic analysis in RB research. Bioinformatic analysis was performed to construct weighted gene co-expression and protein-protein interaction (PPI) networks and to predict long non-coding RNA (lncRNA)-microRNA (miRNA)-mRNA regulatory networks. RNA extracted from RB and adjacent retinal tissue was used to validate the results obtained from bioinformatic analysis, using a semi-quantitative PCR (qPCR) assay. Twenty-one modules were generated from 5000 most variably expressed genes. Both the light-yellow and red modules were significantly associated with the cellular anaplastic grade of RB. The genes clustered in the light-yellow module included protocadherin beta (PCDHBs) family members. The red module included 5 hub genes involved in cell division. According to the hypothesis that lncRNA may serve as a competing endogenous RNA (ceRNA) for miRNAs and modulates mRNA expression, a network was constructed between lncRNA metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) and cell division-related mRNAs. PCR analysis using 23 tumor tissues and 5 adjacent retinal tissue showed increased expression of PCDHB5 in tumor samples, and supported the predicted upregulation of mitotic checkpoint serine/threonine kinase (BUB1) by MALAT1 via miR-495-3p. Our study highlights the importance of bioinformatic analysis in identifying potential markers and mechanisms associated with the malignant transformation of RB, and provides evidence to suggest that PCDHB5 and the ceRNA regulatory network of MALAT1/miR-495-3p/BUB1 are involved in the progression of RB.

恶性视网膜母细胞瘤 (RB) 发展的潜在机制在很大程度上仍然未知。本研究的目的是确定与 RB 进展相关的加权基因，并评估生物信息学分析在 RB 研究中的有用性。进行生物信息学分析以构建加权基因共表达和蛋白质-蛋白质相互作用（PPI）网络并预测长链非编码RNA（lncRNA）-微RNA（miRNA）-mRNA调控网络。从 RB 和邻近视网膜组织中提取的 RNA 用于验证从生物信息学分析中获得的结果，使用半定量 PCR (qPCR) 测定。21 个模块由 5000 个表达变化最大的基因生成。浅黄色和红色模块均与 RB 的细胞间变性等级显着相关。聚集在浅黄色模块中的基因包括原钙粘蛋白 (PCDHBs) 家族成员。红色模块包括参与细胞分裂的 5 个中枢基因。根据 lncRNA 可作为 miRNA 的竞争性内源 RNA (ceRNA) 并调节 mRNA 表达的假设，在 lncRNA 转移相关肺腺癌转录物 1 (MALAT1) 和细胞分裂相关 mRNA 之间构建了一个网络。使用 23 个肿瘤组织和 5 个相邻视网膜组织的 PCR 分析显示肿瘤样品中 PCDHB5 的表达增加，并支持 MALAT1 通过 miR-495-3p 对有丝分裂检查点丝氨酸/苏氨酸激酶 (BUB1) 的预测上调。我们的研究强调了生物信息学分析在识别与 RB 恶性转化相关的潜在标志物和机制方面的重要性，