A fungal Argonaute interferes with RNA interference

Abstract

Small RNA (sRNA)-mediated gene silencing phenomena, exemplified by RNA interference (RNAi), require a unique class of proteins called Argonautes (AGOs). An AGO protein typically forms a protein-sRNA complex that contributes to gene silencing using the loaded sRNA as a specificity determinant. Here, we show that MoAGO2, one of the three AGO genes in the fungus Pyricularia oryzae (Magnaporthe oryzae) interferes with RNAi. Gene knockout (KO) studies revealed that MoAGO1 and MoAGO3 additively or redundantly played roles in hairpin RNA- and retrotransposon (MAGGY)-triggered RNAi while, surprisingly, the KO mutants of MoAGO2 (Δmoago2) showed elevated levels of gene silencing. Consistently, transcript levels of MAGGY and mycoviruses were drastically reduced in Δmoago2, supporting the idea that MoAGO2 impeded RNAi against the parasitic elements. Deep sequencing analysis revealed that repeat- and mycovirus-derived small interfering RNAs were mainly associated with MoAGO2 and MoAGO3, and their populations were very similar based on their size distribution patterns and positional base preference. Site-directed mutagenesis studies indicated that sRNA binding but not slicer activity of MoAGO2 was essential for the ability to diminish the efficacy of RNAi. Overall, these results suggest a possible interplay between distinct sRNA-mediated gene regulation pathways through a competition for sRNA.

Figure 1.

Phylogenetic analysis and schematic representation of 3 AGO genes in Pyricularia oryzae. (A) A neighbor-joining tree was constructed by alignment of Piwi domain sequences from AGO proteins in Ascomycete fungi. Ago1 in Schizosaccharomyces pombe (NP_587782) was used as an outgroup. Bootstrap values are shown from 1000 replicates. Accession numbers of sequences are as follows: Cryphonectria parasitica AGL3 (ACY36941), Aspergillus niger EHA21073 (CAK38206), Histoplasma capsulatum HCAG 06198 (XP_001538593), Blumeria graminis f. sp. hordei (B. hordei) CCU75107 (CCU75107), B. hordei CCU79215 (CCU79215), A. niger PIWIL3 (XP_001398045), Aspergillus nidulans AN1519 (XP_659123), C. parasitica AGL2 (ACY36940), H. capsulatum HCAG 09806 (XP_001544570), Fusarium oxysporum FOMG 16409 (EXK27157), Trichoderma reesei ETS02811 (ETS02811), Colletotrichum sublineola KDN60156 (KDN60156), Gaeumannomyces graminis 09431 (XP_009225543), MoAGO2 (XP_003717504), G. graminis 02126 (XP_009218161), Magnaporthe poae 06244 (KLU87243), C. parasitica AGL1 (ACY36939), C. sublineola KDN61505 (KDN61505), F. oxysporum FOMG 11108 (EXK33907), T. reesei ETS00705 (ETS00705), Neurospora crassa QDE-2 (XP_011394903), MoAGO3 (XP_003714217), G. graminis 00042 (XP_009216045), M. poae 05413 (KLU86399), H. capsulatum, HCAG 08776 (XP_001536454), A. niger EHA23002 (CAL00657), B. hordei CCU74167 (CCU74167), N. crassa SMS-2 (XP_958586), C. sublineola KDN63192 (KDN63192), F. oxysporum FOMG 01451 (XP_018232952), T. reesei ETS02384 (ETS02384), C. parasitica ACY36942 (ACY36942), MoAGO1 (XP_003716704), G. graminis 05449 (XP_009221516) and M. poae 03783 (KLU84747). (B) Schematic representation of InterPro-predicted protein domains for three AGO genes in P. oryzae. Conserved residues for nucleic acid binding site, 5′ RNA guide strand anchoring site and enzymatic active site are indicated as bars. L1, linker 1 domain; L2, linker 2 domain.

Figure 2.

Gene silencing in AGO KO mutants of Pyricularia oryzae. (A) Transformants with a pSilent2- or pSilent-MG-based vector to knockdown the hygromycin resistance (hph) gene were constructed in various genetic backgrounds as indicated, and cultured on media with or without 400 μg/ml Hygromycin B (hyg). The rate of growth reduction on hyg-containing media relative to that on hyg-less media was calculated for each transformant, and their average was plotted on the graph. cAgo1, cAgo2 and cAgo3 indicate gene complemented strains of Δago1, Δago1 and Δago3, respectively. Asterisks indicate a significant difference from the wild-type (WT) strain (two-tailed t-test after angular transformation; *P < 0.05; **P < 0.01) ND, not determined. (B) qRT-PCR analysis of the hph gene in AGO KO mutants transformed with the above-mentioned hph silencing vectors. The parent AGO KO mutants without the silencing vectors were also employed in the analysis to serve as control for normalization. The actin gene (MGG_03982) was used as an internal control. Asterisks indicate a significant difference from WT (two-tailed t-test after angular transformation; *P < 0.05; **P < 0.01).

Figure 3.

Effects of AGO KO on the activity of TEs and mycoviruses. (A) qRT-PCR analysis of TEs (MAGGY and MGL) in AGO KO mutants. The actin gene (MGG_03982) was used as an internal control. Asterisks indicate a significant difference from WT (two-tailed t-test; *P < 0.05; **P < 0.01). (B) Intron-excision assay was performed to examine the transposition activity of MAGGY in the AGO mutants. A relative ratio of intron-less or intron-containing MAGGY DNA was measured by qPCR using sets of primers specific to ‘exon’ junction or intron internal sequences. Asterisks indicate a significant difference from WT (two-tailed t-test; *P < 0.05; **P < 0.01). (C) qRT-PCR analysis of mycoviruses (PoOLV1 and PoOLV2) in AGO KO mutants and an AGO2 overexpression (OE) strain. The actin gene was used as an internal control. Asterisks indicate a significant difference from WT (two-tailed t-test; *P < 0.05; **P < 0.01). (D) qRT-PCR analysis of mycoviruses (PoOLV1 and PoOLV2) in the wild-type, Δdcl2 and Δdcl2/Δago2 strains. The actin gene was used as an internal control. Asterisks indicate a significant difference (two-tailed t-test; **P < 0.01).

Figure 4.

Characteristics of small RNAs (sRNAs) associated with each AGO protein in Pleomorphomonas oryzae. (A) The relative frequency of each 5′ terminal nucleotide of sRNAs in input RNA (no immunoprecipitation) and those associated with MoAGO1, MoAGO2 and MoAGO3 proteins. (B) Size distribution of sRNAs in the input (gray line)-, MoAGO1 (red line)-, MoAGO2 (yellow line)- and MoAGO3 (blue line)-cDNA libraries. (C) The relative frequency of 5′-U at each sRNA length in the sRNA cDNA libraries as mentioned in (B). (D) Analysis of potential genomic origins of sRNAs associated with each AGO protein in P. oryzae. The P. oryzae reference genomic sequence was divided into 100-nt blocks and the number of mapped sRNAs in each block was counted. The resulting 410 272 genome blocks were classified based on whether they contained more than 10 mapped RPMs in each library. The number of blocks that met the criteria and the total number of mapped reads in the blocks were plotted for each group of genome blocks.

Figure 5.

MoAGO2 and MoAGO3 are associated with highly similar siRNA populations derived from TEs and mycoviruses. (A) Fractions of siRNAs derived from TEs (MAGGY and MGL) and mycoviruses (PoOLV1 and PoOLV2) in the MoAGO1-, MoAGO2- and MoAGO3- sRNA libraries. (B) Size distribution of TE- and mycovirus-derived sRNAs in the input (gray line)-, MoAGO1 (red line)-, MoAGO2 (yellow line)- and MoAGO3 (blue line)-libraries. (C) The relative frequency of 5′ and 3′ terminal adenine (A) and thymine (T) of siRNAs derived from TEs and mycoviruses in the input-, MoAGO2- and MoAGO3-sRNA libraries.

Figure 6.

sRNA binding but not slicer activity of MoAGO2 is essential for inhibiting RNAi. (A) In vitro target RNA cleavage assay. FLAG-tagged MoAGO3, MoAGO2 and MoAGO2^AEAD (slicer mutant) were expressed in Δmoago3 cells under the control of a constitutive promoter (Aspergillus nidulans gpdA), and purified by immunoprecipitation with anti-FLAG agarose beads. Eluted protein was incubated with FITC-labeled target and control RNAs. Cleaved RNAs were analyzed by a denaturing 12% polyacrylamide gel. The numbers under the bands indicate the intensity of the target bands normalized to the control bands. (B) Hygromycin-sensitive growth reduction assay was used to assess the ability of MoAGO2^AEAD (slicer mutant) and MoAGO2^Y619E+K623A (sRNA binding mutant) to diminish RNAi. A Δmoago2 strain showing a hygromycin-hypersensitive phenotype was used as a parent strain. A plasmid expressing FLAG-tagged MoAGO2, MoAGO2^AEAD or MoAGO2^Y619E+K623A under the control of the A. nidulans gpdA promoter was introduced into the parent strain. Five to ten transformants were cultured on media with or without 400 μg/ml Hygromycin B (hyg). The rate of growth reduction on hyg-containing media relative to that on hyg-less media was calculated for each transformant, and their average was plotted on the graph. Asterisks indicate a significant difference from the parent strain (two-tailed t-test after angular transformation; *P < 0.05; **P < 0.01).

Figure 7.

MoAGO2 and MoAGO3 are localized in the cytoplasm and cytoplasmic granules. GFP-tagged MoAGO2 and mCherry-tagged MoAGO3 were co-expressed in the wild-type Pyricularia oryzae cells. Images were captured using the KEYENCE BZ-9000 epifluorescent microscope and analysed using BZ-9000 software. Nuclei were visualized by DAPI staining.