Entry created on 1 July 2019 (Revision 1.0) Annotator: Rita Pancsa
This entry is part of a multi-component system encompassing the following entries: Q9UPQ9 Q9UKV8
Basic protein information
Accession Q9UKV8
Common name AGO2
Gene AGO2
Organism Homo sapiens
Uniprot name Protein argonaute-2
Basic LLPS information
Organelle RISC complex; P-body; micro-ribonucleoprotein complex; miRISC; GW-body
Type of experimental evidence
Joined entry Q9UPQ9 Q9UKV8
Protein region(s) mediating LLPS
517
-
818
PIWI domain
Based on the experimental results of the following publication: 29576456
Molecular features viewer
PDB structures
Extended LLPS information
Functional description
miRISC is a multi-protein assembly that uses microRNAs (miRNAs) to identify mRNAs targeted for repression. Two core protein components of human miRISC, Argonaute2 (Ago2) and TNRC6B, condense into phase-separated droplets in vitro and in live cells. Phase separation is promoted by multivalent interactions between the glycine/tryptophan (GW)-rich domain of TNRC6B and three evenly spaced tryptophan-binding pockets in the Ago2 PIWI domain. When exposed to mammalian cell lysate, miRISC droplets also recruit components of the CCR4-NOT complex as well as deadenylase activity, which can be accelerated by more than two orders of magnitude upon phase separation. These observations suggest a model in which miRISC uses molecular condensation to sequester miRNA targets and concentrate them with factors that mediate mRNA decay. The use of phase separating systems to control RNA localization and stability may be a feature common to many aspects of RNA metabolism (PMID:29576456).
Literature supporting the LLPS: 29576456
Functional class of membraneless organelle: activation/nucleation/signal amplification/bioreactor
Binding partners (at biological protein concentrations)
1) TNRC6B (strictly required for LLPS)
Type of RNA(s) required/used for the LLPS at biological protein concentrations
other type of RNA: miRNAs
Molecular interaction types contributing to LLPS
multivalent domain-motif interactions (PMID:29576456)
Determinants of phase separation and droplet properties
1) protein concentration of Ago2 2) salt concentration 3) crowding agent concentration 4) valency of Ago2
Membrane cluster No
Partner-dependent Yes
RNA-dependent No
PTM required No
Domain-motif interactions Yes
Discrete oligomerization No
Regulation and disease
Post-translational modifications affecting LLPS
Position Residue PTM Effect Reference Modifying enzyme Notes
Isoforms known to affect LLPS
Isoform Effect Reference
All known isoforms containing sequence changes in the LLPS region(s)
Position type Isoform names from UniProt
Disease mutations affecting LLPS
Mutation dbSNP Disease OMIM Effect Reference Notes
Experimental information
Experimental techniques applied to prove/investigate LLPS
The combined results of in vitro protein-protein interaction detection assays and X-ray crystallography with multiple Ago2 and TNRC6B mutants indicate that all three pockets in the trp-binding region of Ago2 contribute to physical interaction with TNRC6B. Specific Trp residues in TNRC6B Ago-binding domain (ABD) have pocket preferences, suggesting that the three trp-binding pockets are not perfectly redundant with each other, however, each pocket is able to interact with multiple different trp residues in TNRC6B, thus the interaction is multivalent and the structures formed may be heterogeneous and complex. Under physiological salt concentrations, in vitro, solutions containing both Ago2 and the TNRC6B-ABD quickly became opaque (change in optical properties), indicative of the formation of massive particles. Concentrated solutions of the ABD alone (but not that of Ago2) also became turbid at temperatures below 15°C. The fluorescently labeled TNRC6B ABD and Ago2 formed liquid droplets (morphology) in a protein concentration-dependent and salt concentration-dependent manner (particle size and count) as assessed by confocal microscopy. Ago valency (finetuned with mutations of the Trp-binding pockets) had a strong effect on droplet formation (particle size and count by microscopy). In a 293 HEK cell line stably overexpressing GFP-fused TNRC6B (genetic transformation) GFP-labeled, dynamic foci formed in vivo with 0.2 to 1mm in size (particle size and count by microscopy). Transfecting cells with a plasmid encoding mCherry-fused Ago2 revealed that Ago2 co-localized with TNRC6B foci in vivo. Ago2-TNRC6B droplets selectively sequester miRNA targets from the bulk solution: when mixed with droplets containing Ago2-let7 (Ago2 loaded with the miRNA let-7) in vitro, a target RNA with eight seed-matched let-7 binding sites (8xlet7) appeared almost exclusively in the pellet fraction (co-localization). In contrast, the 8xlet7 target RNA remained in the supernatant when added to droplets formed with Ago2-miR122. Ago2 kept its endonucleolytic cleavage (termed‘‘slicing’’) activity towards target RNAs within the condensates in the presence of a small RNA guide and the required ions in vitro (enzymatic activity assay). Ago2 and TNRC6B were mixed in the presence of soluble lysate from HEK293 cells, and the resulting droplets were isolated by centrifugation. Analysis by proteomics techniques (western blot) revealed that subunits of the CCR4-NOT deadenylase complex (a known miRISC component) co-pelleted with TNRC6B, while actin, which is not a component of miRISC, did not. (PMID:29576456).
Experimental observations supporting the liquid material state of the condensate
dynamic movement/reorganization of molecules within the droplet (PMID:29576456) dynamic exchange of molecules with surrounding solvent (PMID:29576456) morphological traits (PMID:29576456)