A prokaryotic twist on argonaute function

Abstract

Argonaute proteins can be found in all three domains of life. In eukaryotic organisms, Argonaute is, as the functional core of the RNA-silencing machinery, critically involved in the regulation of gene expression. Despite the mechanistic and structural similarities between archaeal, bacterial and eukaryotic Argonaute proteins, the biological function of bacterial and archaeal Argonautes has remained elusive. This review discusses new findings in the field that shed light on the structure and function of Argonaute. We especially focus on archaeal Argonautes when discussing the details of the structural and dynamic features in Argonaute that promote substrate recognition and cleavage, thereby revealing differences and similarities in Argonaute biology.

Figure 1

Overall architecture of Argonaute from the three domains of life. The domain composition (left) and structures (middle) of the bacterial (based on Thermus thermophilus, PDB: 3DLH), the archaeal (based on Pyrococcus furiosus, PDB: 1U04) and the eukaryotic (based on human Argonaute 2, PDB: 4EI3) Argonaute reveal an evolutionarily conserved architecture. Differences can be found in the surface charge distribution of Argonaute proteins (negatively-charged surfaces in red; positively-charged surfaces in blue). The binding pocket for the 5'-end of the guide in the MID domain is highlighted with a green circle.

Figure 2

Important structural, functional and dynamic features of Argonaute. Structural elements that are important for Argonaute function are highlighted based on the human Argonaute 2 (hAgo2) structure in complex with a guide (red) and target (blue) strand (PDB: 4W5T). (a) The 5'-end is buried in a binding pocket in the MID domain (orange), where specific interactions with the terminal phosphate of the guide strand and interactions between the protein backbone of the specificity loop (highlighted in purple or orange) contribute to the specific recognition of the first nucleotide (PDB: 3LUD). This interaction network leads to the stable positioning of UTP in hAgo2. In contrast, the Argonaute structure from Pyrococcus furiosus (PfAgo) shows that the specificity loop (orange) is pulled away from the first nucleotide (PDB: 1U04). (b) The PAZ domain (pink) of all Argonaute variants is a mobile element, as revealed by structural, kinetic and single-molecule studies. Shown are the conformational changes (highlighted by a broken arrow) of the PAZ domain between the RNA guide-associated hAgo2 (pink, PDB: 4EI3) and hAgo2 in complex with an RNA guide and an 11-nucleotide RNA target (grey, PDB: 4W5T). The movement of the PAZ domain is more pronounced when comparing the structure of DNA guide-associated Thermus thermophilus Ago (TtAgo, PDB: 3DLH) and the ternary TtAgo complex, which also includes a 19-nucleotide RNA target (PDB: 3HVR). Progression to the ternary complex leads to the release of the 3'-end of the guide from its binding pocket in the PAZ domain. Another flexible element that undergoes a structural change upon ternary complex formation is helix α7 (boxed), which is only found in archaeal-eukaryotic Argonautes. (c) The PIWI domain (green) harbors the active site where the glutamate finger can be found in an “unplugged” or “plugged” conformation (PfAgo in its free state (mint green) with the “unplugged” glutamate finger, PDB: 1U04; cleavage-incompatible ternary TtAgo complex with “unplugged” glutamate finger (PDB: 3F73, corn blue); cleavage-compatible ternary TtAgo complex with “plugged” glutamate finger (PDB: 3DLH, orange); ternary hAgo2 complex with “plugged” glutamate finger (PDB: 4W5T, grey). In the “plugged” conformation, an invariant glutamate sidechain is inserted to complete the tetrad in the catalytic pocket (the broken arrow indicates the relocation of E512).

Figure 3

Putative mechanisms of bacterial Ago-mediated silencing pathways. (I) Guide sequences of Thermus thermophilus Ago (TtAgo) and Rhodobacter sphaeroides Ago (RsAgo) are derived from plasmid DNA or RNA transcripts, respectively. (II) TtAgo is loaded with a 13–25-nt guide DNA and RsAgo with a 15–19-nt guide RNA. (III) The target substrates are (a) ssRNA, (b) negatively-supercoiled plasmid DNA and (c) ssDNAs in the case of TtAgo. RsAgo binds plasmid DNA, which will be cleaved, yielding 22–24-nt DNA fragments. Binding of the guide-Ago complex to plasmid DNA furthermore possibly leads to an inhibition of plasmid transcription. The short fragments either stay bound to Argonaute (d) or interact with other RsAgo molecules to constitute DNA-RsAgo complexes and regulate plasmid transcription (e). Figure in part modified from Olovnikov et al. [8].