Protein-centric Approaches: Cross-linking and
Immunoprecipitation
Even though detection of RBPs in datasets derived from several
RNA-centric RBPome studies is frequently a reliable indicative of RNA
recognition, candidate RBPs should not be considered bona-fideRNA-binders until they have been shown to bind RNA in their native
systems, ideally using a variety of orthogonal in vitro andin vivo methods. The discovery of non-canonical RNA-binding
activity among metabolic enzymes and proteins which are known to bind
DNA emphasises the importance of verifying direct RNA binding and
excluding false-positives introduced by experimental artifacts.
Furthermore, protein-centric studies can provide pivotal evidence for
determining the physiological role of such RNA-binding events.
The most widely used techniques for globally identifying the RNAs bound
to RBPs are CLIP and related protocols, such as CRAC, eCLIP, iCLIP and
PAR-CLIP 22–26. Like RBPome capture, these approaches
rely on UV cross-linking of RBPs to their target transcripts. To enrich
for the cross-linked (and, therefore, likely directly bound RNAs),
immunoprecipitations are generally performed under (semi-)denaturing
conditions 22–25. Following RNase trimming of
unshielded ribonucleotide sequences, adapter ligation and proteinase K
digestion of the preserved RNPs, high-throughput sequencing allows
transcriptome-wide mapping of the sites to which the tested RBP was
binding (Fig. 1). Despite its technical power and widespread use,
cross-linking and immunoprecipitation (CLIP) presents some limitations.
A family of CLIP methods has emerged to provide solutions for some of
these challenges in specific biological contexts 27.