Before, cells or tissues are irradiated with UV light, which creates covalent bonds between proteins and RNAs that are in direct contact (step 1). a A simplified schematic of the iCLIP protocol. We conclude that the use of the iCLIP cDNA-starts is appropriate to assign the protein–RNA crosslink sites in iCLIP and related methods.Īn overview of methods and experiments. We show that the broad size range of iCLIP cDNAs in these new experiments allows the cDNA-starts to assign binding sites that align with the expected binding motifs (PTBP1) or binding regions (eIF4A3). To experimentally validate this finding, we produced additional PTBP1 and eIF4A3 iCLIP experiments, which demonstrate that the prevalence of the non-coinciding cDNA-starts is directly correlated with the extent of cDNA-end constraints. These results demonstrate that the cDNA-starts can reliably be used to determine crosslink sites in iCLIP, regardless of the crosslinking method.įurther analyses demonstrated that presence of sequence and structural constraints at cDNA-ends is the cause of the non-coinciding cDNA-starts. Interestingly, when using the photoactivatable 4-thiouridine (4SU)-based crosslinking in combination with iCLIP, the motifs were more highly enriched at cDNA-starts than at T-to-C transitions. Moreover, we identified RNA motifs that are commonly associated with crosslink sites and found them most highly enriched at cDNA deletions in CLIP, and cDNA-starts in iCLIP, eCLIP and irCLIP, even if non-coinciding cDNA-starts are prevalent.

We first examined the position and prevalence of crosslink-induced mutations to confirm previous findings, showing that such mutations are generally >5-fold less common within iCLIP than CLIP cDNAs, regardless of the presence of non-coinciding cDNA-starts. To understand if cDNA-starts can be used to assign RNA-binding sites, we further analysed the iCLIP data with high frequency of non-coinciding cDNA-starts. The study concluded that the use of cDNA-starts may not be appropriate in iCLIP whenever non-coinciding cDNA-starts are prevalent. The non-coinciding cDNA-starts in eIF4A3 iCLIP data produced by the previous study were shifted upstream of this expected region and it was proposed that the presence of non-coinciding cDNA-starts might be related to this shift. However, further studies showed that the sequence and structure of a nascent messenger RNA (mRNA) can shift EJC deposition as far as 10 nt away from this expected site. In vitro biochemical experiments with several splicing substrates demonstrated that the site of EJC deposition is normally expected at nucleotides −20 to −24 upstream of the exon-exon junction (–24.–20 nt). eIF4A3 is a component of the exon junction complex (EJC). Here, we focused on experiments produced for polypyrimidine tract binding protein 1 (PTBP1), eukaryotic initiation factor 4A-III (eIF4A3) and the splicing factor U2 auxiliary factor 65 kDa subunit (U2AF2), which represent examples of non-coinciding or coinciding cDNA-starts in introns or exons. However, a recent study observed that the starts of long and short iCLIP cDNAs often map to different genomic positions for several RBPs, which leads to non-coinciding cDNA-starts.

The computational methods that use cDNA-starts to assign RNA-binding sites have been developed along with iCLIP. Therefore, understanding the proportion and characteristics of truncated cDNAs in these protocols is essential. Recently, further variants were developed that also amplify truncated cDNAs, including BrdU-CLIP, eCLIP and irCLIP. Even though iCLIP amplifies both truncated and readthrough cDNAs, computational comparisons of CLIP and iCLIP cDNAs estimated that over 80% of iCLIP cDNAs truncate at the crosslink sites of most RBPs. The cDNA-starts of these truncated cDNAs identify the nucleotide just downstream of the crosslinked peptide. Therefore, individual-nucleotide resolution CLIP (iCLIP) was also developed to exploit the ‘truncated cDNAs’. The CLIP protocol prepares the cDNA library in a way that requires the reverse transcriptase to read through this peptide, thereby generating only ‘readthrough cDNAs’. This is followed by reverse transcription, during which the bound peptide can lead to truncation of complementary DNAs (cDNA) at the crosslink site. During the CLIP protocol, crosslinked protein–RNA complexes are purified and the RNA fragments are released by digesting the protein, resulting in RNAs with a covalently bound peptide at the crosslink site.

To understand the mechanisms of their action, it is essential to identify the endogenous sites of protein–RNA interactions, which has been aided by the development of ultraviolet (UV) crosslinking and immunoprecipitation (CLIP). RNA-binding proteins (RBPs) play crucial roles in all aspects of post-transcriptional gene regulation.

Iclip germany