Article The Small Non-coding Vault RNA1-1 Acts as a Riboregulator of Autophagy Graphical Abstract Authors Rastislav Horos, Magdalena Bu¨ scher, Rozemarijn Kleinendorst, ..., p62 N PB1 ZZ C Wolfgang Huber, Carsten Sachse, inactive vtRNA1-1 Matthias W. Hentze Correspondence [email protected] (R.H.), starvation [email protected] (M.W.H.) oligomerization In Brief N PB1 ZZ C A biological function of vault RNAs is to p62 N PB1 ZZ C directly modulate the oligomerization active state of p62, thereby controlling N PB1 ZZ C autophagy. ATG8-like binding autophagy aggregate clearance Highlights d The selective human autophagy receptor p62/ sequestosome-1 is an RNA-binding protein d p62 engages the small non-coding vault RNA1-1 as a major interacting RNA d Vault RNA1-1 riboregulates p62-dependent autophagy and aggregate clearance d Mechanistically, vault RNA1-1 interferes with p62 multimerization Horos et al., 2019, Cell 176, 1054–1067 February 21, 2019 ª 2019 Elsevier Inc. https://doi.org/10.1016/j.cell.2019.01.030 Article The Small Non-coding Vault RNA1-1 Acts as a Riboregulator of Autophagy Rastislav Horos,1,5,* Magdalena Bu¨ scher,1,4,5 Rozemarijn Kleinendorst,1 Anne-Marie Alleaume,1 Abul K. Tarafder,1 Thomas Schwarzl,1 Dmytro Dziuba,1 Christian Tischer,1 Elisabeth M. Zielonka,1 Asli Adak,1 Alfredo Castello,2 Wolfgang Huber,1 Carsten Sachse,1,3 and Matthias W. Hentze1,6,* 1European Molecular Biology Laboratory, Meyerhofstrasse 1, 69117 Heidelberg, Germany 2Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK 3Ernst Ruska-Centre for Microscopy and Spectroscopy with Electrons/ER-C3 Structural Biology, Wilhem-Johnen-Straße, 52425 Ju¨ lich, Germany 4Collaboration for joint PhD degree between EMBL and Heidelberg University, Faculty of Biosciences 5These authors contributed equally 6Lead Contact *Correspondence: [email protected] (R.H.), [email protected] (M.W.H.) https://doi.org/10.1016/j.cell.2019.01.030 SUMMARY periments showed that only a fraction of vtRNAs is incorpo- rated into vaults (Kickhoefer et al., 1998), suggesting that Vault RNAs (vtRNA) are small non-coding RNAs vtRNAs may have roles outside of vault RNPs. Overexpression transcribed by RNA polymerase III found in many eu- of vtRNA1-1 was shown to be protective against apoptosis in karyotes. Although they have been linked to drug a cellular model of Epstein-Barr virus infection (Amortetal., resistance, apoptosis, and viral replication, their mo- 2015) and to favor influenza virus replication via PKR deactiva- lecular functions remain unclear. Here, we show that tion (Li et al., 2015). Yet, the molecular functions of vault RNAs vault RNAs directly bind the autophagy receptor remain undefined. Macroautophagy (referred to further as autophagy) is an sequestosome-1/p62 in human and murine cells. vtRNA1-1 essential cellular process responsible for the recognition, Overexpression of human inhibits, while removal and degradation of intracellular components, organ- its antisense LNA-mediated knockdown enhances elles and pathogens within membrane vesicles termed p62-dependent autophagy. Starvation of cells re- autophagosomes (Klionsky et al., 2016). The molecular and duces the steady-state and p62-bound levels functional details of the multiprotein complexes that regulate of vault RNA1-1 and induces autophagy. Mechanisti- the formation and growth of autophagosomal double-mem- cally, p62 mutants that fail to bind vtRNAs branes have been intensively studied (for review, see Yin display increased p62 homo-oligomerization and et al., 2016). After encompassing the cargos, autophagosomes augmented interaction with autophagic effectors. close and fuse with lysosomes and degrade their contents to Thus, vtRNA1-1 directly regulates selective auto- supply amino acids, lipids, and nucleotides for the anabolic phagy by binding p62 and interference with oligo- needs of cells. Autophagy was considered to be a non-selective mechanism merization, a critical step of p62 function. Our data until the discovery of autophagic receptors with the ability to uncover a striking example of the potential of RNA bind specific autophagic substrates and bring them to the to control protein functions directly, as previously forming autophagosomal membranes via interaction with recognized for protein-protein interactions and Atg8-like proteins, including LC3B and GABARAP (Galluzzi post-translational modifications. et al., 2017). The protein p62 (also known as sequestosome-1 [SQSTM1]) is such an autophagic receptor with a C-terminal INTRODUCTION ubiquitin binding domain (UBA) and a LC3-interaction motif (LIR) (Pankiv et al., 2007). p62 co-localizes with LC3-positive Vault RNAs (vtRNA) have been described as small non-coding autophagosomes and is itself degraded in autophagolyso- RNA components of giant ribonucleoprotein particles (RNPs), somes (Pankiv et al., 2007; Sahani et al., 2014; Bjørkøy et al., termed vaults (Kedersha and Rome, 1986). Humans express 2005). Thus, determination of p62 protein levels can serve as four vtRNA paralogs (vtRNA1-1, vtRNA1-2, vtRNA1-3, a proxy for autophagic flux (Klionsky et al., 2016). p62 mostly vtRNA2-1), which are 88–100 nt long and transcribed by serves in selective autophagy for the removal of intracellular RNA polymerase III. Vaults are found in a broad spectrum of pathogens (Zheng et al., 2009) and the degradation of intracel- eukaryotes ranging from protists to mammals (Stadler et al., lular aggregates marked by ubiquitin (Ub) (Pankiv et al., 2007). 2009). Although vaults can occur at 10,000 to 100,000 parti- Among the autophagy receptors, p62 has the distinct property cles per cell and have been linked to cellular processes like to oligomerize via its N-terminal PB1 (Phox and Bem1p) domain drug resistance, apoptosis, and nuclear transport (Berger (Ciuffa et al., 2015). Oligomerization is functionally important, as et al., 2009), their function remains unclear. Sedimentation ex- it increases p62 affinity for LC3-positive membranes (Wurzer 1054 Cell 176, 1054–1067, February 21, 2019 ª 2019 Elsevier Inc. A B Figure 1. The Autophagy Receptor p62 Is an input eluate input IP RNA-Binding Protein UV:-+ + - (A) Western blot analysis of input and eluate sam- UV:- - + + - + antibody: IgGp62 IgG p62 ples from interactome capture experiment. TDP43 kDa RNaseA [ng/μl]: 1 0.21155 0.2 1 serves as a positive control for RNA binding, p62 50 kDa whereas actin serves as negative control. 32 75 TDP43 P: (B) Lysates from UV-treated or control cells were 37 50 treated with dilutions of RNaseA and used for ACTB 75 37 WB: p62 immunoprecipitation followed by radioactive label- 50 ing (upper panel) and western blotting (lower panel). (C) Log2 odds ratios of the enrichment of different C D RNA classes in p62 IPs over the control IPs 4 p62/control IPs (Fisher exact test, Benjamini-Hochberg [BH] 3 15 adjusted p < 0.05). 2 [ vtRNA1-2 (D) Volcano plot of differential crosslink site (CS) 10 1 10 vtRNA1-1 occurrences; each dot corresponds to a genomic [ odds ratio 0 region (exons, introns), black coloring indicates 2 vtRNA2-1 significant enrichment in p62 IPs (BH adjusted -1 5 log value -log p < 0.05). The data were normalized for background -2 p vtRNA1-3 and CS enrichment in p62 IPs over controls was -3 0 tested with DESeq2. Open circles indicate vault -4 -2 -4 0246 RNAs. fold change CDS [log ] tRNA rRNA 2 5UTR 3UTR (E) Predicted RNA secondary structures of vtRNAs. miRNA snRNA lincRNA snoRNA Mt tRNA Mt rRNA vaultRNA antisense intergenic misc RNA Mean CS count values in p62 IPs are shown by the pseudogene indicated color code. See also Figure S1 and Tables S1, S2, and S3. sense overlapping protein coding intron processed transcripts processed pseudogene polymorphic pseudogene unprocessed pseudogene E 60 U Thus, vault RNA1-1 emerges as a ribore- C U G 50 A U 30 A U C gulator of targeted autophagy. U G C C G G C G U U C U U G U C G A U 40 A C G U A C C A A G C G U RESULTS 50 A C U A A A C U C U A U G U C A G C U A U G G U C U G 40 U A U G C C A C U G C U G 50 G U A G 20 C U The Autophagy Receptor p62/ A U U C U A G G C U G 30 U U C G C U 70 U SQSTM1 Is an RNA-Binding Protein 70 A G A A C 40 C A U C G U G C C C 30 G We recently developed a method for the A C 60 U G U U A A G U G C G G C C proteome-wide identification of RNA-in- U C G U C G G C C 60 A A G G U A A U G teracting peptides in RNA-binding pro- U C U C U G U C C U U 10 U A teins (RBPs), termed RBDmap (Castello U A A G C C C G C U U U U G et al., 2016). We performed RBDmap on G C C C 20 U G C G G 30 C C G C 80 G 70 20 G C G U human hepatic HuH-7 cells and identified peptides from both known and previ- vtRNA1-1 vtRNA1-2 vtRNA1-3 vtRNA2-1 ously unknown RBPs (http://www. counts:025 0 40 020040hentze.embl.de/public/RBDmapHuh7/ vignettes/result/; Table S1). A peptide mapping to the autophagy receptor et al., 2015), and is thought to help align p62 to forming auto- p62/sequestosome-1 suggested that p62 interacts with phagosomal structures (Ciuffa et al., 2015). Importantly, oligo- RNA. While lysosome-mediated RNA degradation was merization-deficient p62 is dysfunctional in autophagy (Itakura described long ago (reviewed in Frankel et al., 2017), and and Mizushima, 2011).