Super-lncRNA与super-enhancer的相互作用在乳腺癌发生发展中作用机制的研究

Breast cancer is the most diagnosed cancer and the leading cause of cancer deaths in women. By applying basic scientific principles to the clinic, the past 50 years witnessed great improvements in patient care with reduced morbidity and mortality. Many remaining challenges call for a better understanding of molecular mechanisms involved in different subtypes of breast cancer. Long noncoding RNAs (lncRNAs) represent a new category of oncogenes and tumor suppressor genes. Numerous lncRNAs with subtype specificity have been identified in breast cancer. However, how to sort through the large number of lncRNAs for addictive oncogenic drivers remains a challenging task..Here, we propose the existence of a set of lncRNAs (which we term super-lncRNAs) that regulate subtype-specific super-enhancers via RNA:DNA:DNA triplex formation in breast cancer. Using a logistic regression model that we established previously in identifying super-lncRNAs in 27 human cell and tissue types, we will analyze whole genome datasets from The Cancer Genome Atlas (TCGA) and datasets generated from patient-derived xenografts at Baylor College of Medicine to identify super-enhancers and associated super-lncRNAs. Such lncRNAs may recruit transcription factors and other regulatory complexes to appropriate genomic locations within the super-enhancers and may also contribute to the topological and spatial organization of chromatin for optimal transcriptional activity. Next, we will objectively test our hypothesis using biochemical and genetic approaches. First, we ask how ablation of super-lncRNAs may affect the epigenetic marks of the related super-enhancers and a variety of hallmarks of breast cancer. Next, we test if the repeat domain on super-lncRNAs mediates RNA:DNA:DNA triplex formation..We expect to provide initial and unequivocal evidence that super-lncRNA represents a novel group of addictive oncogenic drivers. It will provide a new way in deciphering the non-coding genome and it may be targeted for further improvement of breast cancer therapy.

乳腺癌是女性常见的恶性肿瘤。最新研究发现多种乳腺癌分型的产生与发展与lncRNA这类癌基因和抑癌基因密切相关。然而，如何筛选具有转录因子样作用的乳腺癌驱动lncRNA尚未充分研究。我们前期研究发现一类super-lncRNA，它能够与乳腺癌亚型相关super-enhancer形成RNA:DNA:DNA三合体并参与乳腺癌的产生和发展。基于此，本项目采用不同分型PDX来源的乳腺癌细胞构建基因表达和表观遗传数据库，再通过逻辑回归模型获取不同分型中特异性表达的super-enhancer和相关super-lncRNA，最终使用CRISPR/Cas9技术敲除super-lncRNA/super-enhancer相互作用序列，从而明确RNA:DNA:DNA的形成对相关癌基因的表达作用。本项目旨在通过揭示super-lncRNA这种新型驱动因子在乳腺癌发展中的作用机制，为乳腺癌的治疗提供新思路与新靶点。

乳腺癌是女性常见的恶性肿瘤。在过去50年间，通过将科研结果应用于乳腺癌的临床治疗，使患者的治疗效果得到很大改善，明显降低了发病率和死亡率，显著提高了患者的生存。然而更加深入地探究不同乳腺癌分型发生发展的分子机制仍然极具挑战性。长链非编码RNA（lncRNA）可以扮演癌基因与抑癌基因的角色，最新研究发现多种乳腺癌分型的产生与发展与lncRNA密切相关。本研究以寻找与乳腺癌发生发展相关的lncRNA为出发点，以探究乳腺癌特异性super-lncRNA并进一步揭示其与乳腺癌分型的相关性为目标开展研究。在项目执行期间，我们针对上述研究内容首先验证了相关深度学习模型的可靠性，并用体外研究证实了lncRNA可以与目标DNA以RNA:DNA:DNA三合体的形式发生相互作用，应用乳腺癌单细胞测序分析证实了lncRNA与蛋白质编码基因联合能够对乳腺癌的分子分型产生预测作用；为了识别 super-lncRNA 的存在，我们分析了来自于 27 个人类细胞以及组织类型的 RNA-Seq 数据，分别在每组中通过Triplexator 预测表达的lncRNA与 super-enhancer 基因序列的三合体结合位点，再经过逻辑回归模型分析得出：有 30%的 super-lncRNA 局限于单一的细胞或组织类型，同时有趣的是，绝大部分（80%）super-lncRNA 仅含有一个重复序列，所预测的 super-lncRNA/super-enhancer相互作用对的表达也是有相关性的。在ER+乳腺癌细胞中我们筛选出super-lncRNA RP11-158K1.3的特异性表达，并发现其与ER+乳腺癌的癌基因标记物ESR1和XBP1相关。上述研究结论在一定程度上为乳腺癌的治疗提供了新思路与新靶点。同时在项目执行期间，我们也探索了lncRNA H19X所编码的miR322(424)/503与肌肉萎缩的相关性，证明了该miRNA簇所导致肌肉萎缩的机制在于其可以通过靶向多种翻译起始因子发挥作用，故而最终证实了靶向miR-322(424)/miR-503 簇能够促进肌肉生长，并抑制肌肉萎缩，有望为多种病因所导致的肌肉萎缩提供新的治疗靶点。