Purification and analysis of endogenous human RNA exosome complexes

Abstract

As a result of its importance in key RNA metabolic processes, the ribonucleolytic RNA exosome complex has been the focus of intense study for almost two decades. Research on exosome subunit assembly, cofactor and substrate interaction, enzymatic catalysis and structure have largely been conducted using complexes produced in the yeast Saccharomyces cerevisiae or in bacteria. Here, we examine different populations of endogenous exosomes from human embryonic kidney (HEK) 293 cells and test their enzymatic activity and structural integrity. We describe methods to prepare EXOSC10-containing, enzymatically active endogenous human exosomes at suitable yield and purity for in vitro biochemistry and negative stain transmission electron microscopy. This opens the door for assays designed to test the in vitro effects of putative cofactors on human exosome activity and will enable structural studies of preparations from endogenous sources.

Keywords: RNA exosome; protein complex purification; ribonuclease.

FIGURE 1.

Purification and analysis of endogenous human exosomes. (A) (1) 3xFlag-tagged exosome component is affinity captured using anti-Flag antibodies coupled to magnetic beads. Pink and blue circles represent proteins not related to the exosome. (2) Native elution is performed with 3xFlag peptide. (3) Eluted complexes are further fractionated on a glycerol gradient and analyzed as indicated. (B) Silver stained SDS polyacrylamide gel displaying the fractionation of ExoI on a 10%–40% glycerol gradient. Exosome constituents are labeled; bands marked “core” consist of the low molecular mass components EXOSC1-8, MPHOSPH6, and C1D. Black arrow indicates the peak fraction. (C) Peak fractions from ExoI and ExoII glycerol gradients analyzed by SDS-PAGE. The ExoI peak fraction was stained with silver, ExoII with Sypro Ruby. Protein bands are labeled as in B. (D) MS-based estimate of the relative amounts of EXOSC10 and DIS3 obtained in velocity sedimented fractions of ExoI and ExoII preparations, as in C. Error bars indicate the data range. (E) Negative-stain TEM analysis of ExoI and ExoII particles. Shown are 16 representative 2D class averages for each preparation. Scale bars (black): 10 nm. (F) ExoI and ExoII 2D class averages have been enlarged to illustrate the “hole,” observed in ExoI class averages and the “lobe,” observed in ExoII class averages (white arrows).

FIGURE 2.

Preparations of endogenous human exosomes exhibit specific 3′-to-5′ distributive, exoribonucleolytic activity. (A) Structure comparison of 3′-end modified oligos utilized for RNA degradation assays. Substrate 1 is 2′-O-methylated, while substrate 2 is also 3′-phosphorylated. Both substrates are labeled with 6-FAM (6-carboxyfluorescein) at the 5′-end. (B) RNA degradation intermediates resolved by denaturing urea-polyacrylamide gel electrophoresis; ExoI was incubated with either substrate 1 or 2 for the indicated time. (C) RNA degradation intermediates produced by ExoII, incubated with RNA substrates as in B. (D) Panel of assays testing the effect of DTSSP on DIS3 enzymatic activity. (Left) Coomassie stained SDS-PAGE displaying DIS3-3xFlag purification from HEK293 cells (H.Ch. and L.Ch.: IgG heavy and light chains, respectively). (Center) DIS3 incubated with both RNA substrates, as in B. (Right) DTSSP-treated DIS3 was incubated with substrate 1 and the RNA degradation products were separated as in B.

FIGURE 3.

DTSSP-modified peptides mapped to EXOSC10 and DIS3 protein sequences and structures. (A, upper) Linear arrangement of EXOSC10 domains with DTSSP-modified peptides mapped (in purple). The domain organization is based on the protein sequences obtained from the uniprot.org and pfam.xfam.org databases. (Lower) The in silico modeled structure of EXOSC10 using I-TASSER (Yang and Zhang 2015; Yang et al. 2015). Highlighted are DTSSP-modified peptides (purple), key active site residues (red), and cocrystalized RNA (orange). The protein orientation is indicated by the miniature structure of a 12-component exosome (PDB ID: 5c0w with EXOSC10 model), upper right. (B, upper) Arrangement of DIS3 domains represented as in A. (Lower) DTSSP-modified peptides were first mapped onto the human DIS3 protein sequence and the equivalent residues were then mapped on the yeast DIS3 structure (PDB ID: 5c0w), displayed, and colored as in A. Additionally, the PIN domain was labeled in blue to distinguish it from the RNB domain. The protein orientation is indicated as in A.