RNAs are subject to degradation, whether for quality control, maturation or turnover. The primary bacterial exoribonuclease is PNPase. This protein contains two catalytic RNase pleckstrin-homology (PH) domains and two RNA-binding domains (S1 and KH); it assembles as a trimer. In eukaryotes and archaea, a major RNA-processing complex is the exosome, a multisubunit assembly that degrades messenger, ribosomal and noncoding RNAs from the 3′ end. Like PNPase, the exosome is composed of RNase PH-domain proteins (although in six separate subunits) and three additional S1-containing RNA-binding subunits. Exosome activity is regulated on several levels because it processes RNAs with different secondary structures, can degrade RNAs partially or completely and uses various stimulatory accessory factors.

Reprinted with permission from Mol. Cell 20, 461–471 (2005).

Two recent studies have provided new insight into archaeal exosome structure and mechanism. Büttner et al. (Mol. Cell 20, 461–471, 2005) solved the structure of two nine-subunit Archaeoglobus fulgidus exosome isoforms, and Lorentzen et al. (Mol. Cell 20, 473–481, 2005) solved the structures of the hexameric Sulfolobus solfataricus exosome-processing domain bound to either RNA or ADP.

In the A. fulgidus exosome, the RNase PH–like subunits, Rrp41 and Rrp42, form a trimer of dimers (green and blue ribbons) and constitute the hexameric RNA-processing center. The other subunits, either Csl4 or Rrp4, constitute a homotrimeric cap (orange or red ribbons) that sits on one face of the hexameric ring. Both Csl4 and Rrp4 contain an S1 domain between their N- and C-terminal domains. In both isoforms, the S1 domains are positioned in the cap's interior and are involved in restricting access through a pore to the processing core of the hexamer. The caps have a positive surface charge, and the S1 domain is even more strongly positively charged, suggesting that the RNA is attracted to the cap and then directed toward the pore by S1-domain interactions. Each cap subunit interacts with both Rrp41 and Rrp42, although the specific interactions differ between Csl4 and Rrp4. Modeling with different stoichiometries of Csl4 and Rrp4 cap subunits, which are thought to occur in eukaryotic caps, shows that this might be possible structurally without steric clash.

The hexameric ring has a narrow pore where it contacts the cap, which leads to a wider channel that constricts again as an exit pore. Lorentzen et al. show that RNA binds in this cleft and that the four 3′-terminal nucleotides interact with arginines of Rrp41 and Rrp42 in an electrostatic, non–sequence-specific manner. The Rrp41 subunits contain the sites of phosphorolysis, in a pocket located between Rrp41 and Rrp42. Cleavage of the terminal nucleotide seems to cause little conformational change, and a short sliding movement may be sufficient to reposition the newly formed 3′ end at the active site.

The overall structure of the exosome suggests that RNA processing may have similarity to proteasome-mediated protein degradation. In this case, RNA enters through a lid-like structure having a pore that can accommodate only unstructured RNA; this pore regulates passage of the RNA into the cavity formed by the hexameric ring where processing occurs.