Arabidopsis FIERY1, XRN2, and XRN3 are endogenous RNA silencing suppressors

Abstract

The eukaryotic defense response posttranscriptional gene silencing (PTGS) is directed by short-interfering RNAs and thwarts invading nucleic acids via the RNA slicing activity of conserved ARGONAUTE (AGO) proteins. PTGS can be counteracted by exogenous or endogenous suppressors, including the cytoplasmic exoribonuclease XRN4, which also degrades microRNA (miRNA)-guided mRNA cleavage products but does not play an obvious role in development. Here, we show that the nuclear exoribonucleases XRN2 and XRN3 are endogenous PTGS suppressors. We also identify excised MIRNA loops as templates for XRN2 and XRN3 and show that XRN3 is critical for proper development. Independently, we identified the nucleotidase/phosphatase FIERY1 (FRY1) as an endogenous PTGS suppressor through a suppressor screen in a hypomorphic ago1 genetic background. FRY1 is one of six Arabidopsis thaliana orthologs of yeast Hal2. Yeast hal2 mutants overaccumulate 3'-phosphoadenosine 5'-phosphate, which suppresses the 5'-->3' exoribonucleases Xrn1 and Rat1. fry1 mutant plants recapitulate developmental and molecular characteristics of xrn mutants and likely restore PTGS in ago1 hypomorphic mutants by corepressing XRN2, XRN3, and XRN4, thus increasing RNA silencing triggers. We anticipate that screens incorporating partially compromised silencing components will uncover additional PTGS suppressors that may not be revealed using robust silencing systems.

Figure 1.

Structure and Conservation of FRY1. (A) FRY1 (At5g63980) gene structure. Exons are represented by black boxes. The positions of the ATG start and TGA stop codons are indicated. The location of fry1 T-DNA insertions (triangles), nonsense mutations, and missense mutations with the corresponding amino acid changes are indicated above or below the gene structure. (B) Phylogeny of the FRY1 protein. In black are plants (At, Arabidopsis thaliana; Os, Oryza sativa; Zm, Zea mays), and in gray are fungi (yeast; Sc, Saccharomyces cerevisiae; Sp, Schizosaccharomyces pombe) and mammals (Mm, Mus musculus; Hs, Homo sapiens).

Figure 2.

Developmental Abnormalities of ago1, fry1, and xrn Mutants. (A) and (B) ago1-27 and ago1-33 mutant lines used for mutagenesis. Plants are dwarf and have serrated leaves. (C) to (F) fry1-4 and fry1-5 mutants recovered from the mutagenesis screen are shown in the ago1 mutant backgrounds ([C] and [D]) or alone ([E] and [F]). Plants are dwarf and have crinkly leaves, rounded leaf margins, and delayed flowering compared with wild-type plants. (G) A Columbia (Col) accession wild-type plant. (H) fry1-6 T-DNA insertion mutants display developmental anomalies identical to fry1-4 and fry1-5 mutants. (I) xrn2-1 mutants do not display obvious developmental defects. (K) Hypomorphic xrn3-3 mutants display crinkly leaves and delayed flowering. (M) xrn4-6 mutants display slightly serrated leaves. (J), (L), and (N) xrn2-1 xrn3-3 (J), xrn2-1 xrn4-6 (L), and xrn3-3 xrn4-6 (N) double mutants show additive developmental defects, and xrn2-1 xrn3-3 mutants (J) resemble fry1-4, fry1-5, and fry1-6 single mutants ([E], [F], and [H], respectively).

Figure 3.

ago1 fry1 Double Mutants Posttranscriptionally Silence L1. RNA gel blot analysis of GUS mRNA (A) and GUS siRNA (B) accumulation in L1 control and ago1, fry1, and ago1 fry1 mutant 18-d-old seedlings. (A) L1 control and L1/fry1-4 and L1/fry1-5 mutant plants are silenced and accumulate low GUS mRNA levels. ago1-27 and ago1-33 mutants release L1 silencing and accumulate high levels of GUS mRNA. Both fry1-4 and fry1-5 restore L1 silencing in ago1 mutant backgrounds and accumulate low levels of GUS mRNA, although L1 silencing is restored to a lesser extent in L1/ago1-27/fry1-4 mutants (4.2 versus 1.3). Normalized values of GUS mRNA to ethidium bromide–stained 25S RNA (bottom panel), with L1 levels set at 1.0, are indicated. (B) RNA gel blot analysis of GUS siRNA accumulation. L1 control, L1/fry1-4, L1/fry1-5, L1/ago1-27/fry1-4, and L1/ago1-33/fry1-5 mutant plants are silenced and accumulate GUS siRNAs. ago1-27 and ago1-33 mutants release L1 silencing and do not accumulate detectable GUS siRNAs. Blots were stripped and rehybridized with a probe complementary to U6 snRNA as a loading control. Normalized values of GUS siRNAs to U6 small nuclear RNA (snRNA), with L1 levels set at 1.0, are indicated. ND, not detected.

Figure 4.

fry1 Mutants Overaccumulate miRNA Target 3′ Cleavage Products. RNA gel blot analysis of AGO1 full-length mRNA and miR168 generated 3′ AGO1 cleavage products, and ARF10, ARF16, and ARF17 full-length RNA and miR160 generated 3′ ARF cleavage products in L1 control and ago1, fry1, ago1 fry1, xrn2, xrn3, and xrn4 mutant 18-d-old seedlings. (A) AGO1 full-length mRNA levels are increased in ago1 and ago1/fry1 mutants, whereas AGO1 3′ cleavage products, but not AGO1 full-length mRNA, overaccumulate in fry1 single mutants. (B) and (C) fry1 mutants and ago1 fry1 double mutants overaccumulate ARF10, ARF16, and ARF17 3′ cleavage products (B), similar to xrn4 mutants (C). Ethidium bromide–stained 25S RNA is shown as a loading control.

Figure 5.

MIRNA Loops Overaccumulate in fry1 and xrn2 xrn3 Double Mutants. (A) Diagram of the various intermediates and end products generated during MIRNA maturation. The stem-loop is the RNA sequence flanked at each end by the miRNA and miRNA* sequences; partial stem-loops are generated after either the 3′ end of the miRNA or the 5′ end of miRNA* is defined. The positions of the 21-nucleotide miRNA (miR) and miRNA* (miR*) species are indicated by gray boxes (the miR* is the 21-nucleotide RNA species that is opposite of and paired to the miRNA in the MIRNA hairpin). The loop sequence is between the miRNA and the miRNA* and is liberated during miRNA maturation. (B) RNA gel blot analysis of miRNA accumulation in Col control and fry1-4 and fry1-5 mutant inflorescences. miRNA accumulation is either unchanged or slightly reduced in fry1 mutants except for miR168, which slightly overaccumulates. Blots were stripped and rehybridized with a probe complementary to U6 RNA as a loading control. (C) RNA gel blot analysis showing that a homozygous fry1-6 T-DNA insertion mutant and fry1-4 and fry1-5 ethyl methanesulfonate mutants overaccumulate MIR168a loops in inflorescence tissues. U6 hybridization is shown as a loading control. Normalized values of the MIR168a loop RNA to U6 snRNA, with Col levels set at 1.0, are indicated. (D) RNA gel blot analyses of MIR168a and MIR164b maturation intermediates and end products in inflorescence tissues of controls (Col and MIR168a-overexpressing plants; 35S:168a), hemizygous and homozygous fry1 mutants, two homozygous xrn2 mutants, a hemizygous xrn3 mutant, and a homozygous xrn4 mutant. Homozygous fry1 mutants overaccumulate MIR168a and MIR164b loops but not stem-loops or partial stem-loops. miR164 and miR168 accumulation in fry1 homozygous mutants is shown in (B). RNA extracted from MIR168a-overexpressing inflorescences was used to indicate the positions of MIR168a stem-loops and partial stem-loops, which are undetectable in wild-type inflorescences. The positions of ³²P-labeled RNA oligonucleotides (nt) are noted. U6 hybridization is shown as a loading control. Normalized values of loop RNAs and miRNAs to U6 snRNA, with Col levels set at 1.0, are indicated. (E) RNA gel blot analysis of MIR168a and MIR164b loop accumulation in control Col, xrn2, hypomorphic xrn3, xrn4, xrn2 xrn3, xrn2 xrn4, xrn3 xrn4, and fry1 mutant inflorescences. All mutations are homozygous. xrn3-containing mutants overaccumulate loops, although to a lesser extent than fry1 mutants. U6 hybridization is shown as a loading control. Normalized values of loop RNAs to U6 snRNA, with Col levels set at 1.0, are indicated.

Figure 6.

XRN2 and XRN3 Are Endogenous Suppressors of PTGS. GUS protein activities contributed by the homozygous transgene loci L1 (A), Hc1 (B), and 6b4 (C) in control, ago1, xrn, and fry1 mutant leaves. Data are expressed as percentages of silenced plants in each genetic background. The number of individual plants analyzed is indicated above each bar.

Figure 7.

xrn4 and fry1 Mutants Have Reduced CMV Accumulation. RNA gel blot analyses of CMV RNA3 and RNA4 and CMV-derived siRNA accumulation in control L1 and xrn2, xrn3, xrn4, and fry1 mutant 29-d-old rosettes. Both fry1 and xrn4 mutants accumulate less CMV RNA and thus less CMV-derived siRNAs than control L1 plants. 25S rRNA hybridization is shown as a control for mRNA gel blots. Normalized values of CMV RNA3 and RNA4 to 25S rRNA, with L1 levels set at 1.0, are indicated. U6 hybridization is shown as a loading control for siRNA gel blots. Normalized values of CMV siRNAs to U6 snRNA, with L1 levels set at 1.0, are indicated. The number of symptomatic plants out of 24 CMV-infected plants is indicated at the bottom of each lane.

Figure 8.

Model for FRY1, XRN2, XRN3, and XRN4 Activity during PTGS. Shown at left is a simplified model for miRNA biogenesis and action. Normal miRNA maturation requires DCL1, SERRATE, HYL1, and HEN1 and liberates MIRNA loops, miRNA*, and mature miRNAs. miRNAs are exported from the nucleus by HASTY and incorporate into an AGO1-containing complex to direct the cleavage of partially complementary RNAs. MIRNA loops are substrates of both the nuclear XRN2 and XRN3, while 3′ target cleavage products are substrates of the cytoplasmic XRN4. Shown at right is a simplified model for transgene- and virus-induced silencing. Aberrant RNA deriving from transgene transcription and virus replication is converted to double-stranded RNA and processed into siRNAs by RDR6 and SGS3, and DCL2 and DCL4, respectively. siRNAs incorporate into an AGO1-containing complex to direct the cleavage of complementary RNAs. XRN2, XRN3, and XRN4 are endogenous suppressors of PTGS, and we propose that the transgene- and virus-derived aberrant RNAs that trigger PTGS are substrates of these XRNs. At center, FRY1 converts 3′-phosphoadenosine 5′-phosphate (PAP), a toxic byproduct of sulfate assimilation, into 5′ AMP + Pi. Our data, together with those from yeast, predict that XRN levels are regulated through a FRY1-dependent system that keeps PAP, an inhibitor of yeast XRN activity, levels in check. We propose that XRN activities are repressed in fry1 mutant plants, leading to the overaccumulation of aberrant RNAs that trigger PTGS.