UvCLAP is a fast and non-radioactive method to identify in vivo targets of RNA-binding proteins

Daniel Maticzka,Tugce Aktas,Rolf Backofen,Ibrahim Avsar Ilik,Asifa Akhtar

doi:10.1038/s41467-018-03575-4

Daniel Maticzka, Tugce Aktas + Show 3 more

Open Access

https://doi.org/10.1038/s41467-018-03575-4

Copy DOI

Abstract

RNA-binding proteins (RBPs) play important and essential roles in eukaryotic gene expression regulating splicing, localization, translation, and stability of mRNAs. We describe ultraviolet crosslinking and affinity purification (uvCLAP), an easy-to-use, robust, reproducible, and high-throughput method to determine in vivo targets of RBPs. uvCLAP is fast and does not rely on radioactive labeling of RNA. We investigate binding of 15 RBPs from fly, mouse, and human cells to test the method’s performance and applicability. Multiplexing of signal and control libraries enables straightforward comparison of samples. Experiments for most proteins achieve high enrichment of signal over background. A point mutation and a natural splice isoform that change the RBP subcellular localization dramatically alter target selection without changing the targeted RNA motif, showing that compartmentalization of RBPs can be used as an elegant means to generate RNA target specificity.

Highlights

ForegroundCrosslinking events replicate B G1—hg[19 ] G2—hg[19]hKHDRBS2-R489K K hKHDRBS2 hKHDRBS3hKHDRBS1−Y440F hQKI−B hKHDRBS1 hQKI−A heIF4A1 F1—hg[19 ] F3—mm[10 ] K1—dm[3 ] L1—dm[3]
Since proteins tend to run at discrete positions on a sodium dodecyl sulfate (SDS)/ lithium dodecyl sulfate-polyacrylamide gel electrophoresis setup, cutting out the labeled region removes contaminating RNA and RNA–protein complexes as they migrate to different regions of the gel
We reasoned that a stringent tandem affinity purification protocol would circumvent these time-consuming and critical steps while removing contaminating free RNA and proteins that might nonspecifically co-purify with the tagged proteins

Summary

Introduction

UvCLAP profiles of various RBPs expressed in human cells show characteristic binding on their target RNAs. We observed very weak and dispersed events from the control libraries, indicating a successful removal of free or nonspecifically bound RNA by the tandem affinity pulldown (Fig. 1c, d and Supplementary Fig. 1b). To determine to what extent the multiplexed treatment of specific and nonspecific pulldown conditions would preserve the quantity of detected RNA, we compared the total number of crosslinking events between pairs of biological replicates, each sharing the respective pulldown condition and construct expression.

Results

Conclusion