A quantitative analysis of CLIP methods for identifying binding sites of RNA-binding proteins (mRNA-seq)

Crosslinking and immunoprecipitation (CLIP) is increasingly used to map transcriptome-wide binding sites of RNA-binding proteins (RBPs). We developed a method for CLIP data analysis and applied it to compare 254 nm CLIP with PAR-CLIP, which involves crosslinking of photoreactive nucleotides with 365 nm UV light. We found small differences in the accuracy of these methods in identifying binding sites of HuR, a protein that binds low-complexity sequences and Argonaute 2, which has a complex binding specificity. We show that crosslink-induced mutations lead to single-nucleotide resolution for both PAR-CLIP and CLIP. Our results confirm the expectation from original CLIP publications that RNA-binding proteins do not protect sufficiently their sites under the denaturing conditions used during the CLIP procedure, and we show that extensive digestion with sequence-specific ribonucleases strongly biases the set of recovered binding sites. We finally show that this bias can be substantially reduced by milder nuclease digestion conditions. Overall design: We performed duplicate mRNA-Seq experiments for cells grown in the presence of modified nucleotides or in normal medium, and following crosslinking at 365nm, 254nm or without crosslinking. In addition, we performed single mRNA-Seq experiments of cells transfected with anti-GFP or anti-HuR siRNA.