Reiterated WG/GW motifs form functionally and evolutionarily conserved ARGONAUTE-binding platforms in RNAi-related components

Abstract

Two forms of RNA Polymerase IV (PolIVa/PolIVb) have been implicated in RNA-directed DNA methylation (RdDM) in Arabidopsis. Prevailing models imply a distinct function for PolIVb by association of Argonaute4 (AGO4) with the C-terminal domain (CTD) of its largest subunit NRPD1b. Here we show that the extended CTD of NRPD1b-type proteins exhibits conserved Argonaute-binding capacity through a WG/GW-rich region that functionally distinguishes Pol IVb from Pol IVa, and that is essential for RdDM. Site-specific mutagenesis and domain-swapping experiments between AtNRPD1b and the human protein GW182 demonstrated that reiterated WG/GW motifs form evolutionarily and functionally conserved Argonaute-binding platforms in RNA interference (RNAi)-related components.

Figure 1.

The SD of the NRPD1b CTD binds AGO4 and is required for RdDM in Arabidopsis. (A) AtNRPD1a/b and NRPD1b/DSD proteins are shown schematically. The SD and CD are shown in red and green, respectively. The conserved box H and the CTD repeats are indicated. (B) Myc-AGO4 extract was applied onto equimolar amounts of GST and GST-based fusion protein beads, and the bound proteins were detected by immunoblotting with anti-Myc and anti-RPB1 antibodies. (C) AtSN1 DNA methylation in nrpd1b-11 is rescued by transgenes expressing NRPD1b but not NRPD1b/DSD (three T1 transgenic lines for each construct are shown). AtSN1 retroelement and controls (Cut 1 and control lanes) were amplified from HaeIII-digested genomic DNA as described (Pontier et al. 2005). (D) Proteins were extracted from pooled samples of the T1 transgenic lines shown in C, quantified, and subjected to immunoblot analysis using the anti-Flag M2 monoclonal antibodies (Sigma-Aldrich).

Figure 2.

The SD of the NRPD1b CTD has a WG/GW repeat signature sequence. (A) Dot plot analysis of AtNRPD1b versus SoNRPD1b highlighting the divergent region of the CTD. Positions of the divergent and AGO4-binding regions are indicated in relation to a schematic representation of AtNRPD1b. The SD shows sequence identity to the N-terminal region of HsGW182. The WG motifs are indicated with blue dots. The RNA recognition motif (RRM) domain of HsGW182 is indicated in brown. (B) Relative distribution of the WG/GW/GWG motifs within the SDs of plant NRPD1b proteins. (C) Myc-AGO4 or HA-AGO2 extracts were applied onto equimolar amounts of GST and GST-based fusion protein beads, and bound proteins were detected by immunoblotting with anti-Myc and anti-HA antibodies. (D) Alignment of AtNRPD1b and HsGW182. The WG/GW repeats are highlighted in red.

Figure 3.

Reiterated WG/GW motifs form conserved AGO-binding platform. (A) Sequences of wild-type and mutant AtNRPD1b consensus monomers. (B) Myc-AGO4 and HA-AGO1/2 extracts were applied onto equimolar amounts of GST and GST-based fusion protein beads, and bound proteins were detected by immunoblotting with anti-Myc and anti-HA antibodies. (C) Myc-AGO4 extract was applied onto equimolar amounts of GST, GST-WG1/WG8 beads, and bound proteins were detected by immunoblotting and quantified using ImageJ software. (D) Sequences of wild-type and mutant SoNRPD1b consensus monomers. (E) Myc-AGO4 extract was applied onto equimolar amounts of GST, GST-GW8/GF8 beads, and bound proteins were detected by immunoblotting with anti-Myc antibodies.

Figure 4.

Reiterated WG/GW motifs form functionally conserved AGO-binding platforms. (A) AtNRPD1b and NRPD1b/GW182 mutant are shown schematically. The N-terminal WG/GW-rich region of HsGW182 is shown in brown. (B) AtSN1 DNA methylation in nrpd1b-11 mutant is partially restored by transgenes expressing NRPD1b/GW182 (two T1 transgenic lines are shown). (C) 5S DNA methylation in nrpd1b-11 mutant is partially restored by transgenes expressing NRPD1b/GW182 (DNA was extracted from pooled samples of four T1 transgenic lines).

Figure 5.

Exhaustive search for conserved WG/GW platforms. Proteins containing conserved WG/GW motifs from different organisms were identified by PSI-BLAST alignment of the SD from AtNRPD1b against protein databases. An all-against-all alignment of the identified proteins was carried out using the Dotter program. The figure presents a twofold mirror image of the comparison of each sequence against all of the others, the diagonal representing self-versus-self alignments highlighting intramolecular repetitions. Sequences used were Arabidopsis thaliana NRPD1b; conceptual translations of Arabidopsis genes At5g04290 and At1g65440; P. tetraurelia; NOWA1/2; Homo sapiens proteins GW182, TNRC6B, and TNRC6C; Drosophila melanogaster protein GAWKY; T. thermophila cnjB; A. thaliana SDE3; C. elegans AIN-1; H. sapiens PrP27-30; and C. elegans PrP-like.