The exosome complex has key roles in RNA processing and quality control. Single-particle EM analyses now provide compelling evidence for two distinct pathways by which substrate RNAs can pass through the exosome structure to reach the catalytic site for exonuclease digestion.
The exosome complexes in archaea and eukaryotes have strikingly similar barrel-like structures with an internal channel through which target RNAs are fed before degradation. In archaea, the exosome complex functions only by channeling the single-stranded 3′ ends of substrate RNAs toward three internal active sites with exonuclease activity1,2. However, the eukaryotic complex turns out to be more versatile. The core barrel structure has, surprisingly, lost its exonuclease activities3,4. Nonetheless, it channels RNA via a pathway remarkably similar to that of the archaeal complex, even including a 'fly-by' of the location of the former active site of the Rrp41 component5,6,7,8. This long pathway requires around 31 nt of single-stranded RNA and leads substrates to the exonuclease active site of Rrp44 (Dis3). However, the observation that several well-characterized substrates for the exonuclease activity of Rrp44 apparently lack single-stranded tails of sufficient length to follow this pathway strongly suggested that another shorter path must lead substrates to the Rrp44 active site. Previous 3D reconstructions of EM images and the X-ray structure of Rrp44 indicated that a direct route to Rrp44, bypassing the exosome barrel, was feasible9,10. In the current issue, Wang and colleagues11 confirm this prediction by further EM analyses, which also show that, more surprisingly, the long and short pathways appear to be mutually exclusive and separated by a large structural rearrangement.
The 11-component nuclear form of the eukaryotic exosome (Exo11) contains three different nuclease activities. Rrp44 has both an RNase II−like hydrolytic 3′-exonuclease activity and a PIN-domain endonuclease activity, whereas Rrp6 provides an additional RNase D−like hydrolytic 3′-exonuclease activity. The Exo11 core structure comprises a nine-protein barrel that closely resembles the nine-component archaeal exosome and is related to the bacterial PNPase trimer. The nucleases Rrp44 and Rrp6 associate with this core barrel structure to provide the enzymatic functionalities.
Comparison of the structure of an isolated exosome subcomplex containing Rrp44 and two core components, Rrp41 and Rrp45 (ref. 6), with the structure of the intact exosome that included an RNA molecule threaded through the central barrel8 revealed substantial differences in the position of Rrp44 relative to that of the Rrp41−Rrp45 dimer. It was, however, unclear whether these changes had been induced by the presence of the RNA molecule threaded through the exosome barrel to the Rrp44 active site in the intact exosome or were simply due to the binding of Rrp41−Rrp45−Rrp44 to the other protein components. In this issue, Wang and colleagues11 analyze differences in the conformation of the intact exosome bound to long or short RNA substrates and confirm that these indeed represent functionally distinct conformers.
Previous structural analyses of the eukaryotic exosome demonstrated that, despite the loss of the internal exonuclease active sites in the barrel, RNA can still be threaded through the central channel to reach the exonuclease active site of Rrp44 (refs. 3,5,6,7,8). This route traverses the entire exosome and is remarkably well conserved between the archaeal and eukaryotic complexes. In this pathway, some ∼31−33 nt of single-stranded RNA are located within the core of the exosome and Rrp44 and are therefore protected from nuclease digestion. In the RNA-bound Exo11 complex, the RNA pathway runs directly from the exosome core toward the exonuclease active site of Rrp44, with 12 nt at the 3′ end of the RNA located within Rrp44 (ref. 8). In the recent EM analysis11, the major form of the exosome complex visualized on long RNA substrates closely resembles this crystal structure. In contrast, exosome complexes with shorter RNA molecules can adopt a substantially different conformation, in which the exonuclease domain of Rrp44 is rotated by ∼120°. This dramatic structural change has the effect of moving the cold-shock RNA-binding domain 2 (CSD2) of Rrp44 into the route, thereby disconnecting the base of the central exosome channel from the exonuclease site (Fig. 1). At the same time, a short direct route for RNA into the exonuclease active site of Rrp44 is opened, and this also brings the S1 RNA-binding domain of Rrp44 into proximity with the substrate RNA.
In the closed-channel or 'short-RNA' conformation, there are two solvent-accessible RNA entry sites to the exonuclease active site of Rrp44, plus an additional pathway to the endonuclease active site (Fig. 2a). The authors visualized the locations of RNA substrates on the exosome complex, revealing that long substrates mainly bind, as expected, at the entry pore into the exosome lumen (top of complex in Fig. 2b). In contrast, short substrates preferentially bind on the 'front' and 'back' of the exonuclease domain (Fig. 2c). This provides good evidence for direct access to the exonuclease active site, although the identification of two apparent binding sites on Rrp44 also suggests that more than one direct-access route might exist. Notably, very few complexes showed RNA binding at both sites, thus supporting the model that binding to the alternative pathways is mutually exclusive. In time courses with long substrate, the authors initially saw a high frequency of binding at the long pathway entry site; this binding declined over time, whereas binding at the short pathway remained quite constant. This may reflect truncation of long substrates followed by release and rebinding to the short entry site. In addition, some exosome complexes were observed with RNA bound at the PIN domain, thus indicating direct recruitment to the endonuclease active site of the exosome (Fig. 2a).
Recent functional analyses revealed that nuclear surveillance and degradation of tRNA precursors is surprisingly prevalent12, whereas in vivo cross-linking indicated that pre-tRNAs access the active site of Rrp44 without making substantial contacts to the exosome core13. In agreement with this finding, a 3′-extended tRNA transcript was localized to the direct-access route by EM analysis (Fig. 2c).
The observations strongly suggest that the two major conformers of the exosome normally exist in equilibrium, although this was not directly demonstrated by the data. The authors suggest, on the basis of EM and functional analyses, an equilibrium of around 70:30 in favor of the long conformation in the presence of RNAs that can follow either pathway. An incoming RNA molecule being fed into the central channel of the exosome by the Mtr4 or Ski2 helicases14,15,16,17 would presumably initially encounter the complex in which the long passage to the Rrp44 exonuclease site is blocked by CSD2. However, spontaneous conformational exchange by 'breathing' of the structure may periodically open the channel to Rrp44, thus allowing progress of the substrate RNA. After passage of the 3′ end of the RNA through to the Rrp44 region of the complex, this conformer would be locked in until degradation of the RNA to less than ∼12 nt or until stalling of degradation due to the presence of RNA or ribonucleoprotein structures in the substrate that would prevent further translocation into the exosome channel.
Wang and colleagues11 further propose that the intersubunit movement of the exosome domains might contribute to processivity in RNA degradation by forming a ratchet that helps to pull the RNA through the complex. Conversely, this movement may facilitate release of stalled substrates by offsetting the binding energy of truncated RNAs that are long enough to extend into the Rrp44 active site cleft through the channel but are too short to reach the actual cleavage site.
The new findings underline the versatility of the eukaryotic exosome relative to the ancestral archaeal form, perhaps helping to explain why this multifaceted version was selected.
Lorentzen, E. et al. Nat. Struct. Mol. Biol. 12, 575–581 (2005).
Lorentzen, E. & Conti, E. Mol. Cell 20, 473–481 (2005).
Dziembowski, A. et al. Nat. Struct. Mol. Biol. 14, 15–22 (2007).
Schneider, C. et al. Mol. Cell 27, 324–331 (2007).
Malet, H. et al. EMBO Rep. 11, 936–942 (2010).
Bonneau, F. et al. Cell 139, 547–559 (2009).
Wasmuth, E.V. & Lima, C.D. Mol. Cell 48, 133–144 (2012).
Makino, D.L. et al. Nature 495, 70–75 (2013).
Wang, H.-W. et al. Proc. Natl. Acad. Sci. USA 104, 16844–16849 (2007).
Lorentzen, E. et al. Mol. Cell 29, 717–728 (2008).
Liu, J.-J. et al. Nat. Struct. Mol. Biol. 21, 95–102 (2014).
Gudipati, R.K. et al. Mol. Cell 48, 409–421 (2012).
Schneider, C. et al. Mol. Cell 48, 422–433 (2012).
Jia, H. et al. Proc. Natl. Acad. Sci. USA 109, 7292–7297 (2012).
Halbach, F. et al. Cell 154, 814–826 (2013).
Jackson, R.N. et al. EMBO J. 29, 2205–2216 (2010).
Weir, J.R. et al. Proc. Natl. Acad. Sci. USA 107, 12139–12144 (2010).
Schaeffer, D. et al. Nucleic Acids Res. 40, 9298–9307 (2012).
This work was supported by a Royal Society University Research Fellowship (UF100666) to C.S. and a Wellcome Trust Fellowship (077248) to D.T. Work in the Wellcome Trust Centre for Cell Biology is supported by Wellcome Trust core funding (092076).
The authors declare no competing financial interests.
About this article
Cite this article
Schneider, C., Tollervey, D. Looking into the barrel of the RNA exosome. Nat Struct Mol Biol 21, 17–18 (2014). https://doi.org/10.1038/nsmb.2750
Nucleic Acids Research (2020)
The SKI complex is a broad-spectrum, host-directed antiviral drug target for coronaviruses, influenza, and filoviruses
Proceedings of the National Academy of Sciences (2020)
Journal of Eukaryotic Microbiology (2018)
The RNA Exosome Adaptor ZFC3H1 Functionally Competes with Nuclear Export Activity to Retain Target Transcripts
Cell Reports (2018)
Wiley Interdisciplinary Reviews: RNA (2015)