dsRNA with 5' overhangs contributes to endogenous and antiviral RNA silencing pathways in plants

Abstract

In plants, SGS3 and RNA-dependent RNA polymerase 6 (RDR6) are required to convert single- to double-stranded RNA (dsRNA) in the innate RNAi-based antiviral response and to produce both exogenous and endogenous short-interfering RNAs. Although a role for RDR6-catalysed RNA-dependent RNA polymerisation in these processes seems clear, the function of SGS3 is unknown. Here, we show that SGS3 is a dsRNA-binding protein with unexpected substrate selectivity favouring 5'-overhang-containing dsRNA. The conserved XS and coiled-coil domains are responsible for RNA-binding activity. Furthermore, we find that the V2 protein from tomato yellow leaf curl virus, which suppresses the RNAi-based host immune response, is a dsRNA-binding protein with similar specificity to SGS3. In competition-binding experiments, V2 outcompetes SGS3 for substrate dsRNA recognition, whereas a V2 point mutant lacking the suppressor function in vivo cannot efficiently overcome SGS3 binding. These findings suggest that SGS3 recognition of dsRNA containing a 5' overhang is required for subsequent steps in RNA-mediated gene silencing in plants, and that V2 functions as a viral suppressor by preventing SGS3 from accessing substrate RNAs.

Figure 1

AtSGS3 constructs. (A) Schematic diagram of AtSGS3 constructs. (B) Coomassie brilliant blue-stained SDS–PAGE gel of purified AtSGS3 and AtSGS3ΔN.

Figure 2

RNA/DNA structures used in this study. Cartoon representations for the RNA/DNA structures used in this study. Forms 1–4 is a 37 nt single strand. Form 5 is a 35 bp long double-strand with 2 nt 3′ overhang on each strand. Form 6 is a 37 bp double-strand without overhang. Form 7 is a 35 bp long double-strand with 2 nt 5′ overhang on each strand. Form 8 is a 33 bp long double-strand with both 2 nt 5′ overhang and 2 nt 3′ overhang on one strand. Form 9 is a 25 bp long double-strand with 12 nt 5′ overhang on each strand. Form 10 is a 23 bp long double-strand with 12 nt 5′ overhang and 2 nt 3′ overhang on each strand. Form 11 is a 33 bp long double-strand with both 3 nt 5′ overhang and 1 nt 3′ overhang on each strand. Form 11 is designed by breaking terminal base pairs in form 4. Form 9 is a 21 nt single strand. RNA form 12 represents a typical siRNA/miRNA single strand. Form 13 is a 19 bp long double-strand with 2 nt 3′ overhang on each strand. RNA form 14 represents a typical siRNA duplex. Form 14 is a 21 bp double-strand without overhang. Form 15 is a 19 bp long double-strand with 2 nt 5′ overhang on each strand. Form 16 is a 12 bp long double-strand with 2 nt 3′ overhang on each strand. Form 17 is a 12 bp double-strand without overhang. Form 18 is a 12 bp long double-strand with 2 nt 5′ overhang on each strand. Form 1 has no modification at 5′ end. Form 2 has 5′ triphosphate. Form 3 has 5′ cap modification. Forms 4–18 has 5′ terminal monophosphate on each strand. Form 19 is same as form 18, except that it has no terminal phosphate on either strand. Form 20 is same as form 19 except that it has 3′ terminal phosphate on each strand. Form 20 is same as form 18, except that it has 3′ end methylation modification at ribose 2′ position. See Materials and methods for the sequences.

Figure 3

RNA-binding study of AtSGS3 and AtSGS3ΔN by electrophoretic gel mobility shift analyses (EMSA). (A, B) EMSAs of binding of AtSGS3 (A) and AtSGS3ΔN (B) to RNAs form 6, 7 and 9. (C, D) RNA-binding curves for AtSGS3 (C) and AtSGS3ΔN (D) to RNAs form 1–11, determined by EMSAs. Note that K_d was determined according to protein monomer concentrations, which are shown. See Figure 2 for the RNA structures.

Figure 4

RNA-binding study of AtSGS3 and AtSGS3ΔN by gel mobility shift analyses with shorter RNA and analyses of the effect of terminal phosphate. (A, B) RNA-binding curves for AtSGS3 (A) and for AtSGS3ΔN (B) to RNAs form 12–18, determined by EMSAs. (C, D) EMSAs of binding of AtSGS3 (C) and AtSGS3ΔN (D) to RNAs form 18–20. The non-radioactive RNAs (1 μM) were used, and the gels were stained with SYBR Gold. Protein monomer concentrations are shown.

Figure 5

RNA-binding analyses of V2. (A) Coomassie brilliant blue-stained SDS–PAGE gel of purified wild-type V2 and the C84S/C86S mutant V2 (V2SS). (B) RNA-binding curves for V2 to RNAs form 1–11, determined by EMSAs. (C) RNA-binding curves for V2 to RNAs form 12–18 and for V2SS mutant to RNA form 18, determined by EMSAs. (D) EMSAs of binding of V2 (top panel) and V2SS mutant (bottom panel) to RNA form 18. (E) EMSAs of binding of V2 to RNAs form 18–20. The non-radioactive RNAs (1 μM) were used, and the gels were stained with SYBR Gold. Protein monomer concentrations are shown.

Figure 6

RDR6 activity assay in the presence and absence of AtSGS3, AtSGS3ΔN or V2. (A) Complementary RNA strand synthesised by RDR6 from ssRNA template was resolved by denaturing polyacrylamide gel electrophoresis. The synthesised RNA is shown by arrow. (B) Band intensities of the synthesised RNAs were quantified by phosphoimager and the averaged amounts (n=3, error bar represents standard error) were plotted. The arbitrary unit is used for synthesised RNA amount, where the averaged value for AtRDR6-only condition at 3 h is defined as 100.

Figure 7

RNA-binding competition analyses between V2 and SGS3 proteins. EMSAs of binding of V2/SGS3 proteins to RNA form 18. (A) Binding competition between AtSGS3 and V2 (right panel) or V2SS (left panel). (B) Binding competition between SlSGS3 and V2 (right panel) or V2SS (left panel). A measure of 1 μM each of indicated protein was incubated with RNA form 18. For reaction condition with single protein (V2-alone, AtSGS3-alone, SlSGS3-alone or V2SS-alone), the reaction mixtures were incubated for 1 h at 4 °C after protein addition. For reaction conditions with two proteins (V2/V2SS + AtSGS3/SlSGS3), five different orders of protein additions and incubation time were tested. (a) Add V2 and incubate for 55 min, then add SGS3 protein and incubate for 5 min more; (b) add V2 and incubate for 30 min, then add SGS3 protein and incubate for 30 min more; (c) add SGS3 protein and incubate for 55 min, then add V2 and incubate for 5 min more; (d) add SGS3 and incubate for 30 min, then add V2 and incubate for 30 min more; (e) add both V2 and SGS3 at the same time and incubate for 5 min. The samples were run on the gel for long enough so that the shifted bands can be resolved; unbound RNA was not retained on the gel.