Epigenetic Mechanisms of Gene Expression in Lung Cancer Cells

通过初步的qRT-PCR实验，我们选择了四种肺癌细胞系（H1299、H841、A549、Calu-6），这些细胞系表现出nyeso -1、MAGE-A1和MAGE-A3 CT-X基因表达的异质模式，以及不表达任何CT-X基因的正常人支气管上皮细胞（NHBE）和人小气道上皮细胞（SAEC）。免疫印迹分析证实了qRT-PCR结果。广泛的pyrosequencing和染色质免疫沉淀（ChIP）实验表明，肺癌细胞中NY-ESO-1、MAGE-A1或MAGE-A3表达的缺乏可归因于正常异染色质结构的持续存在，而不是通过启动子低甲基化激活这些基因，然后选择性沉默与二价染色质相关。进一步的实验确定组蛋白赖氨酸甲基化的调节是否会改变肺癌细胞中NY-ESO-1、MAGE-A1和MAGE-A3的表达。简而言之，在不表达NY-ESO-1、MAGE-A1或MAGE-A3的H841细胞中，使用慢病毒shRNA技术敲除介导单、二和三甲基化H3K4（组蛋白激活标记）去甲基化的LSD-1（KDM1）和JARID1B (KDM5B)，或介导H3K27（组蛋白抑制标记）三甲基化的组蛋白赖氨酸甲基转移酶EZH2 （KMT6）。ChIP实验显示，激活和抑制标记的全局变化与类似的变化相一致，并且相对于对照细胞，敲除中NY-ESO-1， MAGE-A1和MAGE-A3启动子中各自组蛋白修饰子的占用减少。虽然单独敲除EZH2、LSD1或JARID1B不足以激活NY-ESO-1、MAGE-A1或MAGE-A3，但抑制EZH2、LSD1或JARID1B的表达可显著增强dac介导的肺癌细胞中这些CT基因的诱导。靶向调节组蛋白赖氨酸甲基化的效果比敲除III类组蛋白去乙酰化酶SirT1后观察到的效果更明显。另外的qRT-PCR和免疫blot实验表明，3-deazaneplanocin A （DZNep）是一种新的EZH2表达药理学抑制剂，在癌细胞中，浓度比该药物的细胞毒剂量低1倍时，显著增强了dac介导的H841细胞中NY-ESO-1、MAGE-A1和MAGE-A3的激活（图2）。这种现象扩展到其他CT-X基因，如MAGE-A6和MAGE-A12，并在许多其他肺癌系中观察到。DZNep对肺癌细胞中dac介导的CT-X基因去抑制的增强程度明显高于SAEC或NHBE，接近或超过连续使用DAC-DP或DAC-trichostatin A （TSA）治疗后的效果。Pyrosequencing和ChIP实验表明，这一现象并非归因于去甲基化增强，而是与NY-ESO-1、MAGE-A1和MAGE-A3启动子中EZH2和H3K27Me3水平降低有关。与这些观察结果一致，EZH2的组成性表达显著减弱了DZNep对dac介导的肺癌细胞NY-ESO-1、MAGE-A1或MAGE-A3的增强作用。随后的细胞因子和铬释放实验表明，DZNep在HLA-A*0201环境下，通过表达NY-ESO-1或MAGE-A3特异性重组T细胞受体的异体PBL，显著增强了dac介导的肺癌细胞识别和裂解。效应细胞与经药物处理的HLA-A*0201转导的正常气道上皮细胞或真皮成纤维细胞共培养后，未观察到细胞因子释放或溶解。这些实验的全部细节首次证明了组蛋白赖氨酸甲基化的调节可能是癌症免疫治疗的一种新的表观遗传策略，已发表在《癌症研究》上。在进一步的研究中，Affymetrix微阵列被用于检测全球基因表达谱，并特异性地鉴定了一组恶性胸膜间皮瘤（MPM）细胞系中编码PcG蛋白的基因，相对于培养的正常间皮瘤。该分析表明，在MPM细胞系中，编码多梳抑制复合物-2 （PRC-2）核心成分的EZH2 （KMT6）和EED和SUZ12（较小程度）过表达。利用识别EZH2剪接变异体的引物进行定量RT-PCR （qRT-PCR）实验和免疫印迹分析证实，相对于培养的正常间皮细胞，MPM系中EZH2过表达，而EED或SUZ12未过表达；EZH2的上调与PRC-2介导的抑制染色质标记H3K27Me3在MPM细胞中的整体增加相一致。随后的qRT-PCR和免疫组织化学实验证实，与正常胸膜相比，80%-85%的原发性MPM标本中EZH2过表达。一般来说，EZH2水平倾向于与mRNA拷贝数一致，尽管注意到一些变化，这表明转录后机制也有助于EZH2在MPM中的过度表达。与这些发现一致，通常靶向EZH2 3' UTR的miR-101或miR-26的水平在MPM标本中与正常胸膜相比显着降低。随后的研究表明EZH2蛋白水平与疾病分期无关；然而，Illumina阵列技术评估的两种EZH2变体中的任何一种的瘤内表达与局部晚期MPM患者接受潜在治愈性切除的生存率相关。我们还进行了其他实验，以检验异常的PRC-2活性是否直接导致胸膜间皮瘤细胞的恶性表型。简单地说，shRNA技术在培养的MPM细胞中敲除EZH2；用于这些实验的shRNA针对EZH2的两个剪接变体。我们进行了类似的实验来敲除EED，尽管它在原发性MPM中没有过度表达，但对维持PRC-2的稳定性和EZH2的组蛋白甲基转移酶活性至关重要。特异性敲除这些PRC-2组分可显著抑制MPM细胞的增殖、迁移、软琼脂克隆原性和致瘤性。在MPM细胞中，EED敲除对H3K27Me3整体水平、增殖、迁移、克隆原性和致瘤性的影响比EZH2敲除更明显；这些结果可能是由于shrna的相对敲除效果，EZH1对EZH2敲除的补偿，或者是由于EED的耗尽导致PRC-2更严重的不稳定。随后的免疫印迹和焦磷酸测序实验表明，DZNep介导EZH2和EED的时间和剂量依赖性消耗，并降低H3K27Me3水平，而不诱导培养的MPM细胞的整体DNA去甲基化；这些影响与这些细胞的增殖、迁移和软琼脂克隆性显著降低相一致。此外，腹腔注射DZNep可显著抑制胸腺裸鼠已建立的MPM异种移植物的生长，且无明显的全身毒性。微阵列、qRT-PCR和ChIP实验表明，靶向破坏PRC-2或DZNep治疗在MPM细胞中的生长抑制作用与H3K27Me3启动子占用减少和多种肿瘤抑制子上调相一致，这些抑制子调节了癌细胞的多能性、细胞周期进展和凋亡。这些实验的详细结果最近发表在《临床癌症研究》杂志上。

Preliminary qRT-PCR experiments allowed us to select for further study four lung cancer lines (H1299, H841, A549, Calu-6) exhibiting heterogeneous patterns of NY-ESO-1, MAGE-A1, and MAGE-A3 CT-X gene expression, as well as normal human bronchial epithelia (NHBE) and human small airway epithelial cells (SAEC), which do not express any CT-X genes. Immunoblot analysis confirmed qRT-PCR results. Extensive pyrosequencing and chromatin immunoprecipitation (ChIP) experiments demonstrated that lack of NY-ESO-1, MAGE-A1 or MAGE-A3 expression in lung cancer cells was attributable to persistence of normal heterochromatin structure, rather than global activation of these genes via promoter hypomethylation, followed by selective silencing associated with bivalent chromatin.Additional experiments were performed to ascertain if modulation of histone lysine methylation altered NY-ESO-1, MAGE-A1, and MAGE-A3 expression in lung cancer cells. Briefly, lentiviral shRNA techniques were used to knock-down LSD-1(KDM1) and JARID1B (KDM5B) that mediate demethylation of mono, di-, and trimethylated H3K4 (histone activation marks, or the histone lysine methyltransferase EZH2 (KMT6) that mediates trimethylation of H3K27 (histone repressive mark) in H841 cells that do not express NY-ESO-1, MAGE-A1, or MAGE-A3. ChIP experiments revealed that global changes in activation and repression marks coincided with similar alterations, and decreased occupancy of the respective histone modifiers within the NY-ESO-1, MAGE-A1 and MAGE-A3 promoters in knock-downs relative to control cells. Whereas knock-down of EZH2, LSD1, or JARID1B alone was insufficient to activate NY-ESO-1, MAGE-A1, or MAGE-A3, inhibition of EZH2, LSD1, or JARID1B expression significantly enhanced DAC-mediated induction of these CT genes in lung cancer cells. The effects of targeted modulation of histone lysine methylation were more pronounced than those observed following knock-down of the class III histone deacetylase, SirT1. Additional qRT-PCR and immunoblot experiments demonstrated that 3-deazaneplanocin A (DZNep), a novel pharmacologic inhibitor of EZH2 expression, at a concentration one log lower than the cytotoxic dose of this agent in cancer cells, significantly enhanced DAC-mediated activation of NY-ESO-1, MAGE-A1 and MAGE-A3 in H841 cells (Figure 2). This phenomenon extended to other CT-X genes such as MAGE-A6 and MAGE-A12, and was observed across numerous other lung cancer lines. The magnitude of enhancement of DAC-mediated de-repression of CT-X genes in lung cancer cells by DZNep was markedly higher than that observed in SAEC or NHBE, and approximated or exceeded that observed following sequential DAC-DP or DAC-trichostatin A (TSA) treatment. Pyrosequencing and ChIP experiments demonstrated that this phenomenon was not attributable to enhanced demethylation, but instead coincided with decreased EZH2 and H3K27Me3 levels within the NY-ESO-1, MAGE-A1 and MAGE-A3 promoters. Consistent with these observations, constitutive expression of EZH2 significantly attenuated the enhancement effect of DZNep on DAC-mediated induction of NY-ESO-1, MAGE-A1 or MAGE-A3 in lung cancer cells. Subsequent cytokine and chromium release assays demonstrated that DZNep significantly enhanced DAC-mediated recognition and lysis of lung cancer cells by allogeneic PBL expressing recombinant T cell receptors specific for NY-ESO-1 or MAGE-A3 in the context of HLA-A*0201. No cytokine release or lysis was observed following co-culture of effector cells with drug treated HLA-A*0201-transduced normal airway epithelial cells or dermal fibroblasts. Full details of these experiments, which were the first to demonstrate that modulation of histone lysine methylation may be a novel epigenetic strategy for cancer immunotherapy, have been published in Cancer Research.In additional studies, Affymetrix microarrays were used to examine global gene expression profiles, and specifically identify genes encoding PcG proteins in a panel of malignant pleural mesothelioma (MPM) lines relative to cultured normal mesothelia. This analysis demonstrated over-expression of EZH2 (KMT6) and to a lesser extent, EED and SUZ12, which encode core components of polycomb repressor complex-2 (PRC-2), in MPM lines. Quantitative RT-PCR (qRT-PCR) experiments using primers recognizing both EZH2 splice variants, and immunoblot analysis confirmed over-expression of EZH2, but not EED or SUZ12 in MPM lines relative to cultured normal mesothelial cells; up-regulation of EZH2 coincided with a global increase in the PRC-2 mediated repressive chromatin mark, H3K27Me3 in MPM cells. Subsequent qRT-PCR and immunohistochemistry experiments confirmed over- expression of EZH2 in 80%-85% of primary MPM specimens relative to normal pleura. In general, EZH2 levels tended to coincide with mRNA copy numbers, although some variations were noted, suggesting that post-transcriptional mechanisms also contribute to EZH2 over-expression in MPM. Consistent with these findings, levels of miR-101 or miR-26, which normally target the 3' UTR of EZH2, were significantly decreased in MPM specimens compared to normal pleura. Subsequent studies demonstrated that EZH2 protein levels did not correlate with stage of disease; however, intratumoral expression of either of the two EZH2 variants assessed by Illumina array techniques correlated with survival in patients with locally advanced MPM undergoing potentially curative resections. Additional experiments were performed to examine if aberrant PRC-2 activity directly contributes to the malignant phenotype of pleural mesothelioma cells. Briefly, shRNA techniques were used to knock-down EZH2 in cultured MPM cells; the shRNA used for these experiments targeted both splice variants of EZH2. Similar experiments were undertaken to knock-down EED, which although not over-expressed in primary MPM, is critical for maintaining stability of PRC-2, and histone methyltransferase activity of EZH2. Specific knock-down of these PRC-2 components significantly inhibited proliferation, migration, soft agar clonogenicity, as well as tumorgenicity of MPM cells. The effects of EED knock-down on global H3K27Me3 levels, as well as proliferation, migration, clonogenicity and tumorgenicity were more pronounced than EZH2 knock-down in MPM cells; these results may have been due to relative knock-down efficacies of the shRNAs, compensation of EZH2 knock-down by EZH1, or more profound destabilization of PRC-2 by depletion of EED. Subsequent immunoblot and pyrosequencing experiments demonstrated that DZNep mediated time and dose-dependent depletion of EZH2 and EED, and decreased H3K27Me3 levels without inducing global DNA demethylation in cultured MPM cells; these effects coincided with significantly decreased proliferation, migration and soft agar clonogenicity of these cells. In addition, intraperitoneal (IP) administration of DZNep significantly inhibited growth of established MPM xenografts in athymic nude mice without obvious systemic toxicities. Micro-array, qRT-PCR, and ChIP experiments demonstrated that the growth inhibitory effects of targeted disruption of PRC-2 or DZNep treatment in MPM cells coincided with decreased promoter occupancy of H3K27Me3, and up-regulation of numerous tumor suppressors modulating pluripotency, cell cycle progression and apoptosis in cancer cells. Detailed results of these experiments have been published recently in Clinical Cancer Research.