CLIP-seq

For CLIP-seq 100 million cells were UV cross-linked using 2000 µJoules/cm2 followed by extraction of nuclear proteins. UV cross-linked lysates from wild-type, TGFR2^G357W and ACTA2^R179H cells were pre-supplemented with RNAse inhibitor (1:50), protease inhibitor cocktail, phenylmethylsulfonyl fluoride (1 mM, PMSF) and treated with turbo DNAse I for 30 min at 37 C. Lysates were then incubated with 5 µg of HDAC9 or BRG1 antibodies at 4 C overnight. RNA extraction from HDAC9 and BRG1 pull downs were performed using miRNeasy kit (Qiagen, USA) according to the manufacturer’s instruction. RNA sequencing for human mRNA and lncRNAs were performed using Ribo-Zero™ Magnetic Kit and RnaseH is used to remove rRNA, remained RNA is then retrieved by ethonal precipitation. The retrieved RNA is fragmented using Ambion Fragmentation Solution. And according to Illumina solexa transcriptome sequencing protocol, random hexamer-primer was used to synthesize the first-strand cDNA. The second-strand cDNA is synthesized using buffer, dNTPs, RNaseH and DNA polymerase I. Fragments are purified with QiaQuick PCR extraction kit and resolved with EB buffer for end reparation and adding A at 3’ end. After that, the fragments are ligated with sequencing adaptors. UNG is added before synthesizing the second cDNA strand. After the second strand degraded, dUTP is then converted to dTTP. Y-adaptor is added afterwards. Suitable fragments are selected as templates according to the agarose gel electrophoresis results for PCR amplification. Generated libraries were sequenced using Illumina HiSeq™ technology. CLIP-seq reads were aligned to human UCSC h19 using standard Illumina library-type. Cufflinks 2.0.2 was run on merged bam files to assemble of transcripts and estimation of FPKM values. FPKM ≤ 1.0 cutoff was used as detection threshold to generate the list of genes. For coverage of MALAT1 transcript the fastqs files were aligned to the human genome (hg19) using STAR⁶⁸. Bedtools CoverageBed (v2.26 http://bedtools.readthedocs.io/en/latest/content/tools/coverage.html) is used for each of the sorted (by chromosome position) alignment bam files to obtain the red coverage for each base in the MALAT1 gene region. The BED file for MALAT1 was obtained from ENSEMBL 75 for the canonical transcript that has a length of 8708 bases. The read coverage for each sample is then normalized to reads per million mapped to the human genome (RPM) by considering only the reads that map uniquely to the human genome respectively. Genomic data from CLIP-seq experiments was deposited to GEO under the access number {"type":"entrez-geo","attrs":{"text":"GSE41607","term_id":"41607"}}GSE41607.