Non-coding RNAs as Prognostic and Diagnostic Markers in Prostate Cancer

我们之前与俄亥俄州立大学的Carlo Croce合作，检测了329个人类microrna和13000个蛋白质编码基因在原发性人类前列腺肿瘤中的表达。这次合作产生了几份最早的报告，显示了前列腺癌中microRNA表达的改变。与非癌组织相比，肿瘤microrna均上调和下调。值得注意的是，前列腺肿瘤倾向于以显著高于非癌前列腺的水平表达miR-106b-25集群的所有成员，这与miR-106b-25在前列腺癌生物学中具有致癌特性是一致的。相比之下，miR-1-133簇在肿瘤中的表达始终低于非癌性前列腺，表明该microRNA簇可能具有肿瘤抑制作用。在两项后续研究中，我们现在发现miR-106b-25表达上调和miR-1-133表达抑制都是人类前列腺癌进展的关键步骤。为了进一步确定miR-106b-25簇在前列腺癌进展中的致癌作用，我们分析了一个由原发肿瘤、疾病转移和疾病复发状态组成的大型公共数据集。与我们之前的研究结果一致，所有mir -106-25编码的microrna在原发肿瘤中均显著上调。这些microrna的表达在转移性病变中进一步升高，特别是miR-106b的表达与早期疾病复发相关。为了鉴定尚不清楚的miR-106b的致癌功能，我们在人前列腺癌细胞中过表达该microRNA。该方法显示miR-106b抑制促凋亡caspase 7的表达。进一步的研究发现caspase 7是一种肿瘤抑制因子，与前列腺癌特异性生存相关。然而，最重要的是，原发肿瘤中miR-106b高表达和caspase-7低表达的结合被发现是早期疾病复发的独立预测因子（校正风险比=4.1；95%可信区间：1.6-12.3）。我们还发现，当caspase 7表达被抑制时，miR-106b-25在人前列腺癌细胞中表达诱导的表型，例如抑制凋亡和改变粘附特性，不会发生。我们的数据表明，miR-106b-25通过改变生存和细胞粘附相关途径，在前列腺癌的进展和疾病结局中发挥着突出作用。当我们研究miR-1-133簇在前列腺癌生物学中的功能时，我们使用了类似的方法。在这里，我们发现miR-1表达在远处转移中进一步降低，导致肿瘤中miR-1低表达的患者早期疾病复发。此外，我们还进行了体外实验来探索miR-1的肿瘤抑制功能。基于细胞的分析表明，miR-1在人类前列腺癌中是表观遗传沉默的。miR-1在这些细胞中的过度表达导致生长抑制和调控细胞周期进程、有丝分裂、DNA复制/修复和肌动蛋白动力学的基因下调。蛋白表达分析和基于3′- utr的报告基因实验进一步证实了这一观察结果，表明这些通路中的基因是miR-1的直接或间接靶点。一项基因集富集分析显示，mir -1介导的肿瘤抑制作用与组蛋白去乙酰化酶抑制剂在全球范围内相似。最后，我们获得了初步证据，证明miR-1可以改变F-actin的细胞组织，抑制肿瘤细胞的侵袭和filipodia的形成。总之，我们的研究结果首次表明，miR-1通过影响多种癌症相关过程和抑制细胞增殖和运动，在前列腺癌中发挥肿瘤抑制作用。最近，Croce小组报告说，来自超保守区域的RNA表达在人类癌症中发生了改变。超保守区（ultra - conserved regions， UCR）是指具有200多个碱基对的基因组片段，在哺乳动物中是绝对保守的。这种保守性表明这些序列编码了一个重要的调控功能。因此，我们假设ucr编码的转录本可能在人类前列腺癌的发病机制中具有未知的作用。利用微阵列，研究人员在正义和反义方向上表达了481种ucr衍生转录本，发现了这些转录本的新表达模式，这些转录本与前列腺癌的发展、Gleason评分和前列腺外延伸有关。为了寻找ucRNA作为非编码RNA可能的功能相互作用，我们对RNA环环相互作用进行了计算建模，以发现ucRNA:mRNA结合对。这种方法在ucRNAs和mrna之间产生了新的候选相互作用。我们还测试了雄激素暴露或表观遗传药物治疗是否会影响ucRNA转录物的表达。这些研究确定了几种对表观遗传药物或合成雄激素R1881治疗有反应的ucRNAs。最后，我们将选择的ucRNAs的表达与肿瘤mRNA和microRNA的表达模式进行了比较，发现了它们之间的全局关系。这是对人类前列腺癌中ucRNA表达的首次研究，表明这些rna在该疾病中表达失调。

We previously examined expression of 329 human microRNAs and 13,000 protein-coding genes in primary human prostate tumors in collaboration with Carlo Croce at Ohio State University. This collaboration generated several of the earliest reports that showed altered microRNA expression in prostate cancer. Tumor microRNAs were both up- and down-regulated when compared with non-cancerous tissue. Notably, prostate tumors tended to express all members of the miR-106b-25 cluster at significantly higher levels than the non-cancerous prostate, which is consistent with miR-106b-25 having oncogenic properties in prostate cancer biology. In contrast, the expression of the miR-1-133 cluster was consistently lower in tumors than in the non-cancerous prostate, indicating that this microRNA cluster may act as a tumor suppressor. In two follow-up studies, we now showed that both up-regulation of miR-106b-25 expression and suppression of miR-1-133 expression are key steps in the progression of human prostate cancer. To further define the oncogenic role of the miR-106b-25 cluster in prostate cancer progression, we analyzed a large public dataset consisting of primary tumors, disease metastases, and disease recurrence status. Concordant with our previous findings, all miR-106-25-encoded microRNAs were significantly up-regulated in primary tumors. Expression of these microRNAs further increased in metastatic lesions, and specifically the expression of miR-106b was associated with early disease recurrence. To identify yet unknown oncogenic functions of miR-106b, we overexpressed this microRNA in human prostate cancer cells. The approach revealed that miR-106b suppressed the expression of pro-apoptotic caspase 7. Additional investigations made the novel observation that caspase 7 is a tumor suppressor and is associated with prostate cancer-specific survival. Most importantly, however, the combination of high miR-106b and low caspase-7 expression in primary tumors was found to be an independent predictor of early disease recurrence (adjusted hazard ratio=4.1; 95% confidence interval: 1.6-12.3). We also showed that phenotypes induced by miR-106b-25 expression in human prostate cancer cells, e.g., inhibition of apoptosis and altered adhesion properties, did not develop when caspase 7 expression was inhibited. Our data indicate a prominent role of miR-106b-25 in prostate cancer progression and disease outcome by altering survival- and cell adhesion-related pathways. We used a similar approach when we investigated the functions of the miR-1-133 cluster in prostate cancer biology. Here, we found that miR-1 expression is further reduced in distant metastasis, leading to early disease recurrence among patients with low miR-1 expression in their tumors. Moreover, we performed in vitro experiments to explore the tumor suppressor function of miR-1. Cell-based assays showed that miR-1 is epigenetically silenced in human prostate cancer. Overexpression of miR-1 in these cells led to growth inhibition and down-regulation of genes in pathways regulating cell cycle progression, mitosis, DNA replication/repair and actin dynamics. This observation was further corroborated with protein expression analysis and 3'-UTR-based reporter assays, indicating that genes in these pathways are either direct or indirect targets of miR-1. A gene set enrichment analysis revealed that the miR-1-mediated tumor suppressor effects are globally similar to those of histone deacetylase inhibitors. Lastly, we obtained preliminary evidence that miR-1 alters the cellular organization of F-actin and inhibits tumor cell invasion and filipodia formation. In conclusion, our findings showed, for the first time, that miR-1 acts as a tumor suppressor in prostate cancer by influencing multiple cancer-related processes and by inhibiting cell proliferation and motility. More recently, the Croce group reported that RNA expression from ultraconserved regions is altered in human cancer. Ultraconserved regions (UCR) are genomic segments of more than 200 base pairs that are absolutely conserved among mammalian species. This conservation suggests an important regulatory function that is encoded by these sequences. Thus, we hypothesized that UCR-encoded transcripts may have unidentified roles in the pathogenesis of human prostate cancer. Using microarrays representing 481 UCR-derived transcripts in sense and antisense direction, we discovered novel expression patterns for these transcripts that were associated with prostate cancer development, Gleason score, and extraprostatic extension. To search for possible functional interactions of ucRNAs as non-coding RNAs, RNA loop-loop interactions were computationally modeled to discover ucRNA:mRNA binding pairs. This approach yielded novel candidate interactions between ucRNAs and mRNAs. We also tested whether androgen exposure or epigenetic drug therapy may affect ucRNA transcript expression. These studies identified several ucRNAs that were responsive to treatment with either epigenetic drugs or a synthetic androgen, R1881. Lastly, we compared expression of selected ucRNAs with tumor mRNA and microRNA expression patterns and found global relationships between them. This first study of ucRNA expression in human prostate cancer indicates a dysregulated expression of these RNAs in the disease.