Epigenetic Regulation of Normal and Pathologic CTCF Functions by BORIS

(1)Boris和生殖系发育。我们继续对Boris(印迹位点调节器的兄弟)-CTCF-Paralog进行研究，我们发现了这一点。Boris与CTCF有一个几乎相同的11锌指(11ZF)DNA结合域(DBD)，但他们侧翼的NH2-和COOH-末端区域是不同的。11ZF区先前在实验室中被鉴定为多价DBD，它能够识别和结合延长的(约50bp)靶序列。通过共享相同的DBD，CTCF和Boris可以识别相同的DNA序列，但可能具有不同的调节作用，并与蛋白质辅助因子形成不同的关联。此外，由于Boris基因在雄性生殖细胞中的组织特异性表达，它可能通过利用新的CTCF/Boris位点，通过特定的环形成，参与在Igf2/H19基因座的特定印迹位置重建父系特有的DNA甲基化模式。基于我们的研究，我们预测大多数ICR序列将包含meCpG敏感的CTCF/Boris靶点，这在几个不相关的印迹基因座上得到了验证。除了在发育中的作用外，Boris还可能在肿瘤发生中发挥关键作用。事实上，虽然Boris在正常体细胞中的表达是沉默的，但在癌细胞中它是被激活的；即Boris是所谓的癌症-睾丸(CT)基因。我们和其他人之前描述了Boris在子宫癌、乳腺癌、肺癌、前列腺癌、骨肉瘤和其他肿瘤中的表达。由于Boris本身是一个基因表达调控因子，因此假设Boris介导的启动子调控是负责多个CT基因表达的调控网络。 (2)Boris与癌症：Boris与CTCF的对抗性。使用Boris KO模型，我们证明了Boris直接调控睾丸特异的蛋白酶基因TSP50，而TSP50又受到P53的负调控。我们先前发现DNA甲基化在人类端粒酶基因hTERT的调控中具有双重作用，hTERT是细胞永生化的关键因素之一。甲基化阻止了具有抑制活性的CTCF的结合，但核心启动子的部分低甲基化是hTERT表达所必需的。然而，在淋巴样细胞中，端粒酶似乎是通过甲基化独立的机制激活的。在我们的后续工作中，我们发现在B细胞、一些T细胞淋巴瘤和非肿瘤性淋巴结中，hTERT启动子没有甲基化。在端粒酶阳性的B细胞中，B细胞特异性转录因子PAX5可以通过甲基化非依赖性的机制，推翻CTCF的抑制功能，激活hTERT。最近的研究表明，在生殖细胞癌中，甲基化本身并不是抑制hTERT上CTCF结合的主要机制。我们测试了一个假设，即这些癌细胞中Boris的异常激活阻止了CTCF与一些关键位点的结合，包括hTERT启动子。利用已经证实Boris异常表达的人类癌细胞系，以及瞬时表达Boris编码载体的细胞，我们发现Boris在第一个外显子内结合了hTERT基因，并促进了其转录。在瞬时转染实验中，Boris基因的下调导致hTERT转录减少。然而，在睾丸和卵巢细胞系中，Boris下调并不影响内源性hTERT转录。因此，Boris可能扮演CTCF拮抗剂的角色，使hTERT在癌细胞和永生化细胞中得以表达，但这不是一个简单的二元系统。通过使用一个突变体来取消hTERT外显子1内的CTCF结合，揭示了hTERT调控的复杂性。在这个突变体中，Boris能够激活hTERT转录。这些结果表明，该基因中存在神秘的Boris结合位点，该位点(S)被Boris异构体(S)结合，或者这种调节是由其他未知因素介导的。 (3)关于Boris在干细胞中的作用的研究也导致了几个重要的发现。Boris在小鼠和人类ES细胞中都有表达，分化后其表达被关闭。基于RNA保护分析，胚胎干细胞中的Boris信息似乎与成人睾丸中的不同。我们已经成功地生产出了带有Boris基因座的小鼠ES细胞，使我们能够去除小鼠Boris基因，并用人的cDNA取而代之。使用这些细胞，我们将能够使用我们的单克隆细胞对在芯片中工作得非常好的人类Boris进行芯片序列分析。

(1) BORIS and germline development. We continued our studies of BORIS (Brother Of the Regulator of Imprinted Sites) - a CTCF-paralog, which we discovered. BORIS shares with CTCF a nearly identical 11 Zn-finger (11ZF) DNA binding domain (DBD), but their flanking NH2- and COOH-terminal regions are divergent. The 11ZF region was previously identified in the lab as a multivalent DBD, which is able to recognize and bind extended (around 50bp) target sequences. By virtue of sharing the identical DBD, CTCF and BORIS can recognize the same DNA sequences, but likely have distinct regulation and form different associations with protein cofactors. Furthermore, due to the tissue-specific expression of BORIS in male germ cells, it is likely involved in the re-establishment of paternal-specific DNA methylation patterns at particular imprinted sites of the Igf2/H19 locus through specific loop formation, by utilizing novel CTCF/BORIS sites. Based on our studies we predicted that most ICR sequences would contain meCpG-sensitive CTCF/BORIS target sites, which was validated for several unrelated imprinted loci. In addition to its role in development, BORIS likely plays a key role in oncogenesis. Indeed, while BORIS expression is silenced in normal somatic cells, it is activated in cancer cells; i.e. BORIS is a so-called cancer-testis (CT) gene. We and others previously characterized BORIS expression in uterine cancers, breast, lung, prostate cancers, osteosarcomas and others. As BORIS is itself a gene expression regulator, it was hypothesized that BORIS-mediated regulation of promoters is the regulatory network responsible for the expression of multiple CT genes. (2) BORIS and cancer: antagonism of BORIS and CTCF. Using the Boris KO model we demonstrated that BORIS directly regulates the testis-specific protease gene TSP50, which is in turn negatively regulated by p53. We previously discovered that DNA methylation plays a dual role in the regulation of human telomerase gene, hTERT, one of the key cell immortalization factors. Methylation prevents binding of CTCF, which has repressor activity, but partial hypomethylation of the core promoter is necessary for hTERT expression. In lymphoid cells, however, telomerase appears to be activated through a methylation-independent mechanism. In our follow up work we found that in B cells, some T cell lymphomas, and in non-neoplastic lymph nodes, the hTERT promoter is unmethylated. The B cell-specific transcription factor PAX5 can override the repressive function of CTCF and activate hTERT in telomerase-positive B cells by a methylation-independent mechanism. The sum of recent studies suggests that methylation per se is not the chief mechanism inhibiting CTCF binding at hTERT in germ cell cancers. We tested a hypothesis that abnormal activation of BORIS in those cancer cells prevents CTCF binding to some key sites, including hTERT promoter. Using human cancer cell lines where abnormal expression of BORIS was already documented, as well as cells with transient expression of BORIS-coding vectors, we showed that BORIS binds the hTERT gene within the first exon and facilitates its transcription. Down-regulation of BORIS led to a decrease of hTERT transcription in transient transfection experiments. However, in testicular and ovarian cell lines BORIS down-regulation did not affect endogenous hTERT transcription. Thus, BORIS might play the role of CTCF antagonist, enabling the expression of hTERT in cancer and immortalized cells, but it is not a simple binary system. The complexity of hTERT regulation was revealed by using a mutant which abolished CTCF binding within the exon1 of hTERT. In this mutant, BORIS was able to activate hTERT transcription. These results suggest that either cryptic BORIS-binding sites exist in this gene, the site(s) are bound by BORIS isoform(s), or that the regulation is mediated by other, yet unknown factors. (3) Studies on the BORIS role in stem cells have also led to several significant findings. BORIS is expressed in both mouse and human ES cells and its expression is shut down after differentiation. Based on RNA protection assays, the BORIS message in embryonic stem cells seems to be different from the one in adult testis. We have successfully produced mouse ES cells with floxed BORIS loci enabling us to remove murine BORIS and substitute with the human cDNA. Using these cells we will be able to perform ChIP-seq using our monoclonals to human BORIS that worked very well in ChIP.