Distinct RNA-dependent RNA polymerases are required for RNAi triggered by double-stranded RNA versus truncated transgenes in Paramecium tetraurelia

Abstract

In many eukaryotes, RNA-dependent RNA polymerases (RdRPs) play key roles in the RNAi pathway. They have been implicated in the recognition and processing of aberrant transcripts triggering the process, and in amplification of the silencing response. We have tested the functions of RdRP genes from the ciliate Paramecium tetraurelia in experimentally induced and endogenous mechanisms of gene silencing. In this organism, RNAi can be triggered either by high-copy, truncated transgenes or by directly feeding cells with double-stranded RNA (dsRNA). Surprisingly, dsRNA-induced silencing depends on the putatively functional RDR1 and RDR2 genes, which are required for the accumulation of both primary siRNAs and a distinct class of small RNAs suggestive of secondary siRNAs. In contrast, a third gene with a highly divergent catalytic domain, RDR3, is required for siRNA accumulation when RNAi is triggered by truncated transgenes. Our data further implicate RDR3 in the accumulation of previously described endogenous siRNAs and in the regulation of the surface antigen gene family. While only one of these genes is normally expressed in any clonal cell line, the knockdown of RDR3 leads to co-expression of multiple antigens. These results provide evidence for a functional specialization of Paramecium RdRP genes in distinct RNAi pathways operating during vegetative growth.

Figure 1.

RdRPs in P. tetraurelia. Sequences of RdRP catalytic domains from P. tetraurelia and other organisms were aligned using the MUSCLE v4 software (67). Conserved residues are highlighted black and grey; the individual position in the protein is given by the position in amino acids. The asterisk indicates the aspartatic acid which was found to be necessary for in vitro catalytic activity of Qde1 (Neurospora crassa), Rdr6 (A. thaliana) and Rdp1 (S. pombe) (10,33,34). Accession numbers are as follows: A. thaliana, Rdr1-6: Q8H1K9, Q82504, O82190, O82189, O82188, Q9LKP0; C. elegans, Rrf1-3: Q9NDH1, Q9BH56, Q19285, Ego1: Q93593; Dictyostelium discoideum, RrpA & RrpB: Q95ZG7, Q95ZG6; N. crassa, Qde1: Q9NDH1; S. pombe, Rdp1: O14227; T. thermophila, Rdr1: QOMSN7; P. tetraurelia; Rdr1-4: Q3SE67, Q3SE69, Q3SE68, A0DMU3.

Figure 2.

Rdr1 and Rdr2 are involved in dsRNA-mediated silencing. (A and B) Phenotypic analysis. Silencing induced by dsRNA feeding of the two reporter genes ND169 (ND) (A) and A⁵¹ (B) was compared when they were co-silenced either with RDR1-RDR4 (R1-4) or control genes (ICL7a; A⁵¹ or ND169). Silencing efficiency is indicated as percentage of cells which (A) discharged <50 % of trichocysts compared with negative control cells or which (B) were not immobilized by anti-A⁵¹ serum. Values result from 13 (A) and 9 (B) independent experiments (±standard deviation). Phenotypes were determined 48 h after first dsRNA feeding. Significance levels (P-values) between RDR1 or RDR2 samples and controls ranged between 0.04 and 0.0001 (one-way ANOVA). (C–F) Northern blot analysis of silencing associated sRNAs. Total RNA was isolated 48 h after first dsRNA feeding and run on 15% polyacrylamide–urea gels. Above each blot, the individual probes as well as the dsRNA feeding combinations are indicated: (C) A⁵¹ was co-silenced with RDR1-RDR4 (R1-4) or the control gene ND169. (D and E) ND169 (ND) was co-silenced with RDR1-RDR4 (R1-4) or the control gene ICL7a. (F) Plasmid map (L4440) for T7 polymerase-mediated dsRNA synthesis in E. coli. Polylinker sequences (white) were located at both ends of the gene-specific fragment (grey) between the T7 promotors. Probes used in (C–E) are indicated as black bars. The lower panels show hybridization to glutamine tRNA as a loading control. RDR4 silencing revealed lower amounts of polylinker siRNAs in Figure 2E, whereas the same blot showed no reduction of secondary ∼22-nt siRNAs (D).

Figure 3.

Characteristics of dsRNA-induced primary and secondary siRNAs. (A and B) Strand bias of siRNA. Blots described in Figure 2D and E were hybridized with two adjacent 50-nt strand-specific oligonucleotide probes located in the centre of the ND169 dsRNA fragment (A) or a single oligo in the polylinker sequence (B) (upper blot: antisense-orientated probe, lower blot: sense-orientated probe). Arrowheads indicate small amounts of ∼23-nt ND169 siRNAs from both strands. (C) Properties of 5′- and 3′-ends of dsRNA-induced siRNA were analysed by treatment with CIP, Terminator (Ter) and periodate followed by β-elimination (P/β). Treatment of total RNA with CIP alkaline phosphatase, removing all 5′ phosphates, resulted in a ∼0.5-nt slower migration of siRNA in comparison to untreated samples. This was found for polylinker-specific ∼23-nt siRNA (upper blot) and ND169-specific ∼22-nt siRNA (middle blot). Treatment of total RNA with Terminator 5′-monophosphate-specific exonuclease (Ter) degraded both classes of siRNA. Periodate treatment and subsequent β-elimination (P/β) resulted in ∼1.5-nt faster migration of both classes of siRNAs as it was also observed for the 3′-unmodified control oligo. The second P/β-lane (right) represents the latter one with increased contrast. A 5′-monophosphorylated (grey arrowhead) and a 5′-unphosphorylated (black arrowhead) 22-nt RNA oligonucleotide, both lacking a 3′ modification, were added to each reaction as a control (lower blot). The lower panels show hybridization to glutamine tRNA as a loading control.

Figure 4.

Transgene-mediated silencing involves Rdr3 and is independent from the dsRNA-induced pathway. (A) Linearized transgene plasmid construct to induce ND169 silencing. The pTI− construct contains a 3′-truncated version of the ND169 gene. GFP on the opposite site of a bidirectional promoter served as a control for presence of the construct in the macronucleus. The pTI+ construct containing the entire ND169 gene served as a negative control in the following experiments. Probes for detection of siRNA and long transgenic transcripts are indicated as black bars (B = BmgBI; S = SmaI; +1 = first base of bidirectional promoter). Experiments were carried out with three injected cell clones; all displayed results from northern blots originate from the same clone. (B) Phenotypic analysis. Each of the RDR genes (R1-4) and two control genes (ICL7a; A⁵¹) were co-silenced by dsRNA feeding. Silencing efficiency was measured as percentage of cells which discharged <50% of trichocysts compared with non-injected wild-type cells or the pTI+ control. Phenotypes were determined 96 h after first dsRNA feeding. Values are generated from 11 independent experiments (±standard deviation); values did not differ significantly between injected clones. (C and D) Northern blot analysis of silencing-associated siRNAs. Total RNA was isolated 96 h after the first dsRNA feeding and loaded on 15% polyacrylamide–urea gels. Probes were corresponding to the entire ND169 cds (1700 bp) (C) and to the 3′ plasmid part of the transgene construct (D) (for probes see A). The lower panels show hybridization to glutamine tRNA as a loading control. (E) Properties of 5′- and 3′-ends of transgene-induced siRNA (upper blot). Removal of 5′ phosphates with CIP alkaline phosphatase resulted in a ∼0.5-nt slower migration of ND169-siRNA in comparison to the untreated sample. sRNAs showed sensitivity to Terminator 5′-monophosphate-specific exonuclease (Ter) but were resistant to periodate treatment and subsequent β-elimination (P/β). The second P/β-lane (right) represents the latter one with increased contrast. Controls (lower blot) were added in the same way as described for Figure 3C. The lower panels show hybridization to glutamine tRNA as a loading control. (F) Independency of dsRNA- and transgene-induced silencing. pTI− injected cells were cultivated to develop a trichocyst non-discharge phenotype (100%). Transgene-induced silencing was then inhibited by feeding RDR3 (R3) dsRNA and R3 + ICL7a dsRNA as a double feeding control. A⁵¹ silencing by dsRNA was induced to determine dsRNA-silencing efficiency when transgene-induced silencing was inhibited. Phenotypes were observed after 48 hof dsRNA feeding. (G) Long sense and antisense transcripts from the transgene. RNA samples shown in (C and D) were separated on a denaturing 1.2% agarose gel. Northern analysis was carried out with strand-specific riboprobes corresponding to the 1700 nt ND169 cds (above: antisense-orientated probe, below: sense-orientated probe). ND169 mRNA occurs in non-injected and pTI+ controls (overexpressed) and in RDR3-silenced pTI− injected cells (indicated by the arrowhead). In each of the pTI− samples aberrantly sized ND169 RNA of both strands was detected (arrows). The lower panels show ethidium bromide staining of total RNA as a loading control.

Figure 5.

Rdr3 is involved in accumulation of endogenous siRNAs. (A) Average division rate of RDR1-RDR4 (R1-4) and control (ND169; ICL7a; A⁵¹; GFP) silenced cells (±standard deviation). Single cells were grown in silencing medium and individualized every day. Reduction of division rate of RDR3 silenced cells started on the fourth day of the experiment. (B and C) Rdr3 dependency of endogenous siRNAs. Total RNA was isolated on days 3, 5 and 9 of RDR3 (R3) and ICL7a (control) silencing. Northern blots were probed with two adjacent 50-nt oligonucleotides corresponding to endogenous siRNAs produced from an intergenic region of scaffold 22. Probes were orientated top (B) and bottom (C), relative to transcription of the 5′-marginal ORF. The lower panels show hybridization to glutamine tRNA as a loading control. (D) Properties of 5′- and 3′-ends of endogenous siRNAs. Removal of 5′ phosphates with CIP alkaline phosphatase resulted in a ∼0.5-nt slower migration in comparison the untreated sample. Endogenous siRNAs showed sensitivity to Terminator exonuclease (Ter) and were sensitive to periodate treatment and subsequent β-elimination (P/β), indicated by ∼1.5-nt faster migration (upper blot). The second P/β-lane (right) represents the latter one with increased contrast. Controls (lower blot) were added in the same way as described for Figure 3C. The lower panels show hybridization to glutamine tRNA as a loading control.

Figure 6.

Rdr3 is involved in exclusive expression of surface antigens. Long-term silencing of RDR3 leads to a destabilization of the currently expressed surface antigen (SAg). (A) Protein expression was monitored in immobilization reactions using anti-sera raised against surface antigen A⁵¹, B⁵¹, C⁵¹, D⁵¹, E⁵¹, G⁵¹, H⁵¹, I⁵¹, J⁵¹, N⁵¹ or Q⁵¹. Starting with a SAg A⁵¹-expressing culture, cells began to immobilize in response to anti-D⁵¹ serum after 48 h. In the following, reaction to anti-D⁵¹ serum was more and more reduced. After 8 days (data not shown) the cultures reacted to none of the available antisera. (B) Monitoring of the transcriptional level by real-time PCR revealed that other antigens, B⁵¹ and D⁵¹, are strongly upregulated during silencing of RDR3. Expression data were normalized to GAPDH-mRNA level, which was found to be constant during RDR3 knockdown (noRT-control not shown). (C) After 9 days of RDR3 (R3) silencing surface proteins were analysed on silver-stained SDS–polyacrylamide gels showing several large proteins present on the surface (∼200 kDa, indicated by arrows). ICL7a silenced cells expressing surface antigen A⁵¹ were used as a control. M = size marker (D) Western blots identify the surface proteins expressed in RDR3 (R3) silenced cells to be A⁵¹, B⁵¹, D⁵¹ and H⁵¹. A smaller 80 kDa protein, which was constitutively expressed (bottom blot) and cross-reacted with anti-A⁵¹ serum, served as a loading control. (E) Indirect immunofluorescence staining of surface proteins A⁵¹ and D⁵¹ revealed their co-expression on individual cells. Antibodies used were Y4 mouse monoclonal (primary) and Alexa anti-mouse (secondary) for SAg A⁵¹; anti-serum (primary) and TexasRed anti-rabbit for SAg D⁵¹.