Long noncoding RNA CCTT recruits CENP-C to centromeres by directly binding to centromeric DNA [ChIRP-Seq]

To investigate the exact locations of CCTT-chromatin interaction on a genome-wide scale, we modified previously reported ChIRP-seq (chromatin isolation by RNA purification followed by deep-sequencing) with a crosslinker 4'- aminomethyltrioxalen (AMT, a psoralen derivative), which allows fixation of nucleic acid interaction by ultraviolet light without crosslinking proteins. We designed 8 complementary DNA oligonucleotides that tiled the Alu-depleted part of CCTT. Affinity-purified CCTT-binding DNAs were sequenced and mapped to the HuRef genome hg38, which contains human a-satellite sequence models in each centromeric region. In contrast with 5.38% of input reads mapping to a-satellite sequence, 16.89% of CCTT-binding reads were enriched at a-satellites. Peaks bound by lnc-CCTT were identified by MACS2 algorithm, which compared the reads in CCTT-captured samples with that in input ones. CCTT-binding peaks mapped across the centromeric regions of all 23 reference centromeres. Next, we validated the enrichment of centromeric peaks located at all chromosomes via ChIRP-qPCR. To validate our ChIRP-seq results, we selected representatives of three major subpopulations: multimapping peaks in one specific chromosome and several chromosomes, as well as single-copy peaks. We performed ChIRP-qPCR and found abundant CCTT-binding centromeric peaks throughout all 23 chromosomes. Importantly, CCTT ChIRP-seq profile was highly correlated with the previously reported ChIP-seq profiles of CENP-C (Pearson correlation R = 0.82), both of which showed very strong, extensive signals across entire centromeric regions in HeLa cells. Overall design: Identification and characterization of DNA regions bound by lnc-CCTT in Hela cells

为在全基因组范围内精准定位CCTT-染色质互作位点，我们对已有报道的ChIRP-seq（染色质RNA纯化分离测序，chromatin isolation by RNA purification followed by deep-sequencing）技术进行改良：采用交联剂4'-氨甲基三氧alen（4'-aminomethyltrioxalen，AMT，一种补骨脂素衍生物），该试剂可通过紫外线介导核酸互作的固定，且无需对蛋白质进行交联。我们设计了8条覆盖CCTT序列中Alu缺失区域的互补DNA寡核苷酸探针。将经亲和纯化得到的CCTT结合DNA进行测序，并将测序读段比对至HuRef参考基因组hg38——该基因组的每个着丝粒区域均包含人类α卫星序列模型。与仅5.38%的输入读段比对至α卫星序列相比，CCTT结合读段中有16.89%富集于α卫星区域。我们采用MACS2算法，通过对比CCTT捕获样本与输入样本的测序读段，鉴定得到lnc-CCTT结合峰。CCTT结合峰分布于全部23个参考着丝粒的着丝粒区域。随后，我们通过ChIRP-qPCR验证了所有染色体上着丝粒结合峰的富集情况。为验证本研究的ChIRP-seq结果，我们选取了三类主要亚群的代表性峰：仅比对至单条特定染色体的峰、比对至多条染色体的峰，以及单拷贝峰。经ChIRP-qPCR实验验证，我们发现全部23条染色体上均存在大量CCTT结合的着丝粒峰。值得注意的是，CCTT的ChIRP-seq信号谱与此前报道的CENP-C（着丝粒蛋白C）的ChIP-seq（染色质免疫共沉淀测序，chromatin immunoprecipitation followed by sequencing）信号谱高度相关（皮尔逊相关系数R=0.82），二者在HeLa细胞的全着丝粒区域均呈现出强烈且广泛的信号。实验整体设计：鉴定并表征HeLa细胞中lnc-CCTT结合的DNA区域。