[Advances in technologies for large-scale enrichment and identification of ribonucleic acid-protein complexes]

Ribonucleic acid (RNA) rarely exists alone in the cell. RNAs interact with a variety of proteins and form RNA-protein complexes (RP-complexes) in every step of their life cycle, from transcription to degradation. These RP-complexes play key roles in regulating a variety of physiological processes. Defects in the composition and function of RP-complexes have been associated with many diseases, including metabolic disorders, muscular atrophy, autoimmune diseases, and cancer. It is hence evident that deciphering the highly complex interaction network of RNA-binding proteins (RBPs) and their RNA targets will provide a better understanding of disease development and lead to the discovery of new targets for cancer therapy. Large-scale identification of RP-complexes at the omics level is a prerequisite for obtaining insights into the complex RNA-protein interaction network. As the first step in omics-wide decoding of RP-complexes, enrichment and purification of RP-complexes is a highly challenging task. Recently, intensive efforts have been undertaken to better enrich and identify RP-complexes. Generally, the enrichment strategies can be classified into two major categories: in vitro and in vivo. Although it has been successfully applied in many studies, the in vitro transcribed bait RNA lacks modifications or structural similarity compared with its natural counterpart. Further, since the proteins relocate and remodel after cell lysis, the use of cell lysates as a protein source may result in capturing false interacting proteins that bind non-physiologically with the bait RNA. Finally, weak interactions between the non-covalently bound proteins and RNA require mild washing to remove non-specific binding, which needs careful optimization. However, substantial sample loss is inevitable. To overcome the disadvantages of in vitro approaches, in vivo cross-linking strategies that "freeze" natural RNA-protein complexes in intact cells via covalent cross-linking have become increasingly popular. The in vivo methods allow RNA to interact with proteins in the intracellular environment. Therefore, the RP-complexes formed under physiological conditions are more biologically relevant than those obtained by in vitro methods. We herein summarize recent in vivo methodological advances in the large-scale enrichment and identification of RP-complexes, including cross-linking and immunoprecipitation (CLIP) and related methods, click chemistry-assisted methods, and organic phase separations. CLIP involves irradiating living cells with 254-nm ultraviolet (UV) light to establish covalent bonds between RNA and proteins. This enables CLIP to purify RNAs bound to a specific RBP under conditions that are stringent enough to prevent co-purification of nonspecifically bound proteins or free RNAs. Since the original study, multiple variant protocols have been derived to increase both efficiency and convenience. Photoactivatable ribonucleoside-enhanced-CLIP (PAR-CLIP) introduces a variation in the crosslinking strategy. Cells were preincubated with photoactivatable ribonucleosides 4-thiouridine (4SU) or 6-thioguanosine (6SG), which enables protein-RNA crosslinking with 365-nm UV-A irradiation. It increases the efficiency of cross-linking between RNA and RBPs and is particularly valuable for studying the interactions between RBPs and nascent RNA. Using a click chemistry-assisted strategy, an alkyne modified uridine analog, 5-ethynyluridine (EU), was incorporated into nascent RNAs via metabolic incorporation in living cells. Combined with UV irradiation-based cross-linking, the alkyne-functionalized RNA and the bound proteins were purified in a poly A-independent fashion by the highly selective bioorthogonal copper (I)-catalyzed azide-alkyne cycloaddition using azide-modified beads. Thus, full lists of both coding and non-coding RNAs with their interacting proteins can be purified, which is a major methodological advance. Organic phase separation methods exploiting the physicochemical difference between cross-linked RP-complexes and free RNA and proteins do not require metabolic-based alkyne labeling or polyA-based RNA capture. Each method has unique strengths and drawbacks, which makes it important to select optimal approaches for the biological question being addressed. We hope that this review points out the current limitations and provides future directions to facilitate further development of methods for large-scale investigation of RP-complexes.