DataSheet_1_Hypoxia-Associated Prognostic Markers and Competing Endogenous RNA Co-Expression Networks in Breast Cancer.docx

ObjectiveMany primary tumors have insufficient supply of molecular oxygen, called hypoxia. Hypoxia is one of the leading characteristics of solid tumors resulting in a higher risk of local failure and distant metastasis. It is quite necessary to investigate the hypoxia associated molecular hallmarks in breast cancer.
Materials and MethodsAccording to the published studies, we selected 13 hypoxia related gene expression signature to define the hypoxia status of breast cancer using ConsensusClusterPlus package based on the data from The Cancer Genome Atlas (TCGA). Subsequently, we characterized the infiltration of 24 immune cell types under different hypoxic conditions. Furthermore, the differentially expressed hypoxia associated microRNAs, mRNAs and related signaling pathways were analyzed and depicted. On this basis, a series of prognostic markers related to hypoxia were identified and ceRNA co-expression networks were constructed.
ResultsTwo subgroups (cluster1 and cluster2) were identified and the 13 hypoxia related gene signature were all up-regulated in cluster1. Thus, we defined the cluster1 as “hypoxic subgroup” compared with cluster2. The infiltration of CD8+ T cell and CD4+ T cell were lower in cluster1 while the nTreg cell and iTreg cell were higher, indicating that there was immunosuppressive status in cluster1. We observed widespread hypoxia-associated dysregulation of microRNAs and mRNAs. Next, a risk signature for predicting prognosis of breast cancer patients was established based on 12 dysregulated hypoxia associated prognostic genes. Two microRNAs, hsa-miR-210-3p and hsa-miR-190b, with the most significant absolute logFC value were related to unfavorable and better prognosis, respectively. Several long non-coding RNAs were predicted to be microRNA targets and positively correlated with two selected mRNAs, CPEB2 and BCL11A. Predictions based on the LINC00899/PSMG3-AS1/PAXIP1-AS1- hsa-miR-210-3p-CPEB2 and SNHG16- hsa-miR-190b-BCL11A ceRNA regulation networks indicated that the two genes might act as tumor suppressor and oncogene, respectively.
ConclusionHypoxia plays an important role in the initiation and progression of breast cancer. Our research provides potential mechanisms into molecular-level understanding of tumor hypoxia.

研究目的：多数原发性肿瘤存在氧分子供应不足的状况，该状况被称为缺氧（hypoxia）。缺氧是实体肿瘤的核心特征之一，会导致更高的局部复发与远处转移风险。探究乳腺癌中与缺氧相关的分子特征具有重要的研究价值。
材料与方法：基于已发表的研究成果，我们选取13个缺氧相关基因表达特征，借助ConsensusClusterPlus软件包，结合癌症基因组图谱（The Cancer Genome Atlas, TCGA）数据集，对乳腺癌的缺氧状态进行分型。随后，我们分析了不同缺氧状态下24种免疫细胞的浸润情况。此外，我们还对差异表达的缺氧相关微小RNA（microRNA, miRNA）、信使RNA（messenger RNA, mRNA）及相关信号通路进行了分析与可视化展示。在此基础上，我们筛选出一系列与缺氧相关的预后标志物，并构建了内源竞争RNA（competing endogenous RNA, ceRNA）共表达网络。
研究结果：本研究共鉴定出两个亚型（cluster1与cluster2），且13个缺氧相关基因特征在cluster1中均呈上调表达。因此，相较于cluster2，我们将cluster1定义为"缺氧亚型"。在cluster1中，CD8+T细胞与CD4+T细胞的浸润水平更低，而天然调节性T细胞（natural Treg, nTreg）与诱导性调节性T细胞（induced Treg, iTreg）的浸润水平更高，这表明cluster1存在免疫抑制状态。我们观察到广泛存在的缺氧相关微小RNA与信使RNA表达失调。基于12个失调的缺氧相关预后基因，我们构建了可用于预测乳腺癌患者预后的风险评分模型。其中，logFC绝对值最高的两个微小RNA分别为hsa-miR-210-3p与hsa-miR-190b，前者与不良预后相关，后者则与较好的预后相关。我们预测到数个长链非编码RNA（long non-coding RNA, lncRNA）可作为微小RNA的靶标，且与两个筛选出的信使RNA CPEB2及BCL11A呈正相关。基于LINC00899/PSMG3-AS1/PAXIP1-AS1 - hsa-miR-210-3p - CPEB2与SNHG16 - hsa-miR-190b - BCL11A这两条内源竞争RNA调控网络的预测结果显示，CPEB2与BCL11A分别可能作为肿瘤抑制基因与癌基因发挥作用。
研究结论：缺氧在乳腺癌的发生与发展过程中发挥重要作用。本研究为从分子层面理解肿瘤缺氧的潜在机制提供了新的思路。