Modulating Cancer Stem Cell Signaling in Thoracic Malignancies

已经开展了一系列实验，以全面检查有助于烟草诱导的人类肺癌的发生和发展的mirna。简单地说，阵列技术用于检测吸烟者和非吸烟者建立的肺癌细胞系以及在存在或不存在CSC的情况下培养的正常呼吸道上皮细胞的miRNA表达谱。该分析显示，miRNA特征与人类肺癌的进展相吻合。此外，miRNA谱区分了吸烟者和非吸烟者的肺癌。在相关暴露条件下，CSC在培养的正常呼吸上皮细胞和肺癌细胞中一致上调miR-31。qRT-PCR和western blot实验证实，CSC显著提高了miR-31的表达，激活了正常呼吸道上皮细胞和肺癌细胞中miR-31的宿主基因LOC554202；停止CSC暴露后，miR-31和LOC554202的表达持续存在，表明这些细胞中存在重编程。qRT-PCR实验显示，肺癌标本中miR-31和LOC554202的表达水平相对于邻近的正常肺组织显著升高。RNA交联免疫沉淀（CLIP）和3' UTR报告基因检测显示miR-31与Dickkopf-1 （Dkk-1）和DACT-3直接相互作用。过表达miR-31可显著降低正常呼吸道上皮细胞和肺癌细胞中Dkk-1和DACT3的表达水平。miR-31的敲低增加了Dkk-1和DACT3的水平，并消除了csc介导的Dkk-1和DACT-3表达的下降。此外，miR-31的过表达降低了其他几种Wnt拮抗剂的表达，包括SFRP1、SFRP4和WIF-1，并增加了Wnt-5a的表达，Wnt-5a是一种非规范的Wnt配体，与癌症干细胞的维持有关，从而增强了体外和体内肺癌的恶性表型。染色质免疫沉淀（ChIP）实验表明，CSC增加了LOC554202启动子内的H3K4Me3、H3K9/14Ac和C/ ebp - β水平。敲除C/ ebp - β可消除csc介导的LOC554202激活。过表达miR-31可显著增强肺癌细胞的增殖和致瘤性；miR-31的敲低抑制了这些细胞的生长。这些实验表明，miR-31在烟草诱导的人类肺癌发生过程中作为一种肿瘤起作用，这些实验最近发表在PLoS One上。最近使用相同体外模型系统的研究表明，CSC在培养的正常呼吸上皮细胞和肺癌细胞中显著抑制miR-487b。有趣的是，与观察到的miR-31相似，在停止CSC暴露后，培养细胞中miR-487b的抑制持续存在。qRT-PCR实验表明，miR-487b在原发性肺癌中的表达水平明显低于邻近正常肺实质，尤其是吸烟者。软件引导的分析揭示了miR-487b的许多潜在靶点，包括Wnt5a， SUZ12, BMI1， c-MYC和k -ras-干细胞多能性的介质。过表达miR-487b抑制了SAEC和Calu-6细胞中Wnt5a、BMI1、SUZ12、c-MYC和K-ras的表达，而内源性miR-487b的缺失则增强了Wnt5a、BMI1、SUZ12和K-ras的表达。ChIP分析显示，在正常呼吸上皮细胞和肺癌细胞中，miR-487b的抑制与SUZ12和BMI1在Dkk-1、SFRP1、SFRP4和WIF-1启动子区域的募集增加以及这些基因的下调相一致。CLIP和3' UTR实验证实miR-487b直接干扰Wnt5a、BMI1、SUZ12、c-MYC和K-ras转录本。亚硫酸氢钠测序、甲基化DNA沉淀（MeDIP）和ChIP以及核小体定位实验表明，miR-487b的抑制与DNA甲基化、去novo核小体占用以及SUZ12和bmi1的募集一致，同时miR-487b启动子区域内H2AZ和TCF1水平降低。脱氧氮胞苷诱导miR-487b表达，并减弱csc介导的miR-487b抑制。tgf - β 1概括了csc介导的对miR-487b的影响。miR-487b的组成表达取消了Wnt信号；抑制CSC介导的肺癌侵袭，或c-MYC、k-ras或tgf - β 1的过表达；减少了肺癌细胞的增殖和致瘤性。这些发现表明，在烟草诱导的人肺癌发生过程中，miR-487b的表观遗传沉默与mir -31的上调协同激活介导多能性的关键途径，并表明DNA去甲基化剂可能有助于恢复miR-487b的表达，以用于肺癌治疗。这些研究是第一个直接涉及miR-487b在人类肺癌发病机制中的抑制，有关这些研究的手稿已提交发表。为了进一步研究吸烟增加肺癌和食管癌细胞恶性表型的机制，还进行了其他实验。简单地说，肺癌和食管癌细胞（A549、Calu-6、NCI-SB-ESC1和NCI-SB-ESC2）在有或没有CSC的正常培养基（NM）中，在临床相关的暴露条件下培养。微阵列分析显示，5天的CSC暴露显著上调了ABCG2，编码一种在癌症干细胞中高度表达的外源泵。定量逆转录- pcr （qRT-PCR）、western blot和免疫组化实验证实，在培养的癌细胞中，ABCG2表达上调，但暴露于CSC的正常小气道上皮细胞（SAEC）或永生化食管鳞状细胞（HET1A）不表达上调。流式细胞术实验表明，CSC增加了培养癌细胞的侧群（SP）。利用ABCG2启动子报告基因构建的瞬时转染实验表明，删除外源应答元件和SP-1位点可显著减弱CSC对ABCG2的诱导作用。ChIP实验显示csc介导的ABCG2的诱导与ABCG2启动子内芳烃受体（AHR）、SP-1和Nrf2的占用增加以及RNA pol II和H3K9Ac的水平增加相一致。在可能在临床环境中实现的条件下，米霉素降低了ABCG2启动子内AHR、SP-1和Nrf2水平的基础和csc介导的升高，显著下调ABCG2表达，降低SP，并显著抑制肺癌和食管癌细胞的增殖和致瘤性。微阵列分析显示，米霉素治疗显著抑制细胞周期和癌症相关基因。在两种米霉素剂量下，Calu-6和A549细胞中大约有1260个基因普遍改变；这些基因中的绝大多数都被药物治疗抑制了。例如，38个癌症转化基因中有37个被米霉素下调。进一步的分析显示，暴露于米霉素后，AKT、tgf - β和Wnt等关键癌症通路受到显著抑制。总的来说，这些发现提供了吸烟状况与肺癌和食管癌患者预后之间的潜在机制联系，并支持了米特霉素在胸部恶性肿瘤中靶向癌症干细胞信号的临床评价。与这些研究相关的手稿即将完成，一项利用米霉素根除胸部恶性肿瘤患者癌症干细胞的新临床方案将在未来几个月内启动。

A series of experiments have been undertaken to comprehensively examine miRNAs that contribute to initiation and progression of tobacco-induced human lung cancers. Briefly, array techniques were used to examine miRNA expression profiles in lung cancer lines established from smokers and non-smokers as well as normal respiratory epithelia cultured in the presence or absence of CSC. This analysis revealed that miRNA signatures coincided with human lung cancer progression. Furthermore, miRNA profiles distinguished lung cancers derived from smokers relative to nonsmokers. Under relevant exposure conditions, CSC consistently up-regulated miR-31 in cultured normal respiratory epithelia and lung cancer cells. qRT-PCR and western blot experiments confirmed that CSC significantly increased miR-31 expression, and activated LOC554202 (the host gene for miR-31) in normal respiratory epithelia and lung cancer cells; miR-31 and LOC554202 expression persisted following discontinuation of CSC exposure suggesting reprogramming in these cells. qRT-PCR experiments revealed that miR-31 and LOC554202 expression levels were significantly elevated in lung cancer specimens relative to adjacent normal lung tissues. RNA cross-link immunoprecipitation (CLIP) and 3' UTR reporter assays demonstrated direct interaction of miR-31 with Dickkopf-1 (Dkk-1) and DACT-3. Over-expression of miR-31 markedly diminished Dkk-1 and DACT3 expression levels in normal respiratory epithelia and lung cancer cells. Knock-down of miR-31 increased Dkk-1 and DACT3 levels, and abrogated CSC-mediated decreases in Dkk-1 and DACT-3 expression. Furthermore, over-expression of miR-31 diminished expression of several other Wnt antagonists including SFRP1, SFRP4, and WIF-1, and increased expression of Wnt-5a, a non-canonical Wnt ligand implicated in maintenance of cancer stem cells, which enhances the malignant phenotype of lung cancers in-vitro and in-vivo. Chromatin immunoprecipitation (ChIP) experiments demonstrated that CSC increased H3K4Me3, H3K9/14Ac and C/EBP-beta levels within the LOC554202 promoter. Knock-down of C/EBP-beta abrogated CSC-mediated activation of LOC554202. Over-expression of miR-31 significantly enhanced proliferation and tumorigenicity of lung cancer cells; knock-down of miR-31 inhibited growth of these cells. These experiments demonstrating that miR-31 functions as an oncomir during tobacco-induced human pulmonary carcinogenesis were published recently in PLoS One. More recent studies utilizing the same in-vitro model system revealed that CSC significantly repressed miR-487b in cultured normal respiratory epithelia and lung cancer cells. Interestingly, analogous to what was observed for miR-31, repression of miR-487b in cultured cells persisted following cessation of CSC exposure. qRT-PCR experiments demonstrated that miR-487b expression was significantly lower in primary lung cancers particularly those from smokers relative to adjacent normal lung parenchyma. Software-guided analysis revealed numerous potential targets for miR-487b including Wnt5a, SUZ12, BMI1, c-MYC and K-ras-mediators of stem cell pluripotency. Constitutive over-expression of miR-487b inhibited, whereas depletion of endogenous miR-487b enhanced expression of Wnt5a, BMI1, SUZ12, c-MYC and K-ras in SAEC and Calu-6 cells. ChIP analysis revealed that repression of miR-487b coincided with increased recruitment of SUZ12 and BMI1 to Dkk-1, SFRP1, SFRP4, and WIF-1 promoter regions, and down-regulation of these genes in normal respiratory epithelia and lung cancer cells. CLIP and 3' UTR experiments confirmed direct interference of miR-487b with Wnt5a, BMI1, SUZ12, c-MYC, and K-ras transcripts. Sodium bisulfite sequencing, methylated DNA precipitation (MeDIP) and ChIP, and nucleosome positioning experiments demonstrated that repression of miR-487b coincided with DNA methylation, de-novo nucleosome occupancy, and recruitment of SUZ12 and BMI1with decreased H2AZ and TCF1 levels within the miR-487b promoter region. Deoxyazacytidine induced miR-487b expression, and attenuated CSC-mediated repression of miR-487b. TGF-beta1 recapitulated CSC-mediated effects on miR-487b. Constitutive expression of miR-487b abrogated Wnt signaling; inhibited lung cancer invasion mediated by CSC, or over-expression of c-MYC, k-ras, or TGF-beta1; and decreased proliferation, and tumorigenicity of lung cancer cells. These findings indicate that epigenetic silencing of miR-487b cooperates with up-regulation of miR-31to activate critical pathways mediating pluripotency during tobacco-induced human pulmonary carcinogenesis, and suggest that DNA demethylating agents may be useful for restoring miR-487b expression for lung cancer therapy. A manuscript pertaining to these studies, which are the first to directly implicate repression of miR-487b in the pathogenesis of human lung cancers, has been submitted for publication. Additional experiments have been performed to further examine mechanisms by which cigarette smoke increases the malignant phenotype of lung and esophageal cancer cells. Briefly, lung and esophageal cancer cells (A549, Calu-6, NCI-SB-ESC1 and NCI-SB-ESC2), were cultured in normal media (NM) with or without CSC under clinically relevant exposure conditions. Microarray analysis revealed that five day CSC exposure significantly up-regulated ABCG2, encoding a xenobiotic pump highly expressed in cancer stem cells. Quantitative reverse transcription-PCR (qRT-PCR), western blot, and immunohistochemistry experiments confirmed up-regulation of ABCG2 in cultured cancer lines, but not normal small airway epithelial cells (SAEC) or immortalized esophageal squamous cells (HET1A) exposed to CSC. Flow cytometry experiments demonstrated that CSC increased the side population (SP) of cultured cancer cells. Transient transfection experiments using ABCG2 promoter reporter constructs revealed that deletion of xenobiotic response elements as well as SP-1 sites markedly attenuated ABCG2 induction by CSC. ChIP experiments revealed that CSC-mediated induction of ABCG2 coincided with increased occupancy of aryl hydrocarbon receptor (AHR), SP-1, and Nrf2, as well as increased levels of RNA pol II and H3K9Ac within the ABCG2 promoter. Under conditions potentially achievable in clinical settings, mithramycin diminished basal as well as CSC-mediated increases in AHR, SP-1, and Nrf2 levels within the ABCG2 promoter, markedly down-regulated ABCG2 expression, decreased SP, and dramatically inhibited proliferation and tumorigenicity of lung and esophageal cancer cells. Micro-array analysis revealed that mithramycin treatment significantly repressed cell cycle and cancer-related genes. Approximately 1260 genes were commonly altered in Calu-6 and A549 cells at two mithramycin doses; the vast majority of these genes were repressed by drug treatment. For example, 37 of 38 cancer transformation genes were down-regulated by mithramycin. Additional analysis revealed significant inhibition of critical cancer pathways including AKT, TGF-beta, and Wnt following mithramycicn exposure. Collectively, these findings provide a potential mechanistic link between smoking status and outcome of patients with lung and esophageal cancers, and support clinical evaluation of mithramycin for targeting cancer stem cell signaling in thoracic malignancies. A manuscript pertaining to these studies is nearing completion, and a novel clinical protocol utilizing mithramycin to eradicate cancer stem cells in patients with thoracic malignancies will be initiated in the next several months.