Distinct substrate specificities of Arabidopsis DCL3 and DCL4

In Arabidopsis thaliana, Dicer-like 3 (DCL3) and Dicer-like 4 (DCL4) cleave long, perfect double-stranded RNAs (dsRNAs) into 24 and 21 nucleotides (nt) small interfering RNAs, respectively, which in turn function in RNA-directed DNA methylation and RNA interference, respectively. To reveal how DCL3 and DCL4 individually recognize long perfect dsRNAs as substrates, we biochemically characterized DCL3 and DCL4 and compared their enzymatic properties. DCL3 preferentially cleaves short dsRNAs with 5′ phosphorylated adenosine or uridine and a 1 nt 3′ overhang, whereas DCL4 cleaves long dsRNAs with blunt ends or with a 1 or 2 nt 3′ overhang with similar efficiency. DCL3 produces 24 nt RNA duplexes with 2 nt 3′ overhangs by the 5′ counting rule. Inorganic phosphate, NaCl and KCl enhance DCL3 activity but inhibit DCL4 activity. These results indicate that plants use DCLs with distinct catalytic profiles to ensure each dsRNA substrate generates only a specific length of siRNAs that trigger a unique siRNA-mediated response.     

DRB2, DRB3 and DRB5 function in a non-canonical microRNA pathway in Arabidopsis thaliana

DOUBLE-STRANDED RNA BINDING (DRB) proteins have been functionally characterized in viruses, prokaryotes and eukaryotes and are involved in all aspects of RNA biology. Arabidopsis thaliana (Arabidopsis) encodes five closely related DRB proteins, DRB1 to DRB5. DRB1 and DRB4 are required by DICER-LIKE (DCL) proteins DCL1 and DCL4 to accurately and efficiently process structurally distinct double-stranded RNA (dsRNA) precursor substrates in the microRNA (miRNA) and trans-acting small-interfering RNA (tasiRNA) biogenesis pathways respectively. We recently reported that DRB2 is also involved in the biogenesis of specific miRNA subsets.1 Furthermore, the severity of the developmental phenotype displayed by the drb235 triple mutant plant, compared with those expressed by either drb2, drb3 and drb5 single mutants, or double mutant combinations thereof, indicates that DRB3 and DRB5 function in the same non-canonical miRNA pathway as DRB2. Through the use of our artificial miRNA (amiRNA) plant expression vector, pBlueGreen2,3 we demonstrate here that unlike DRB2, DRB3 and DRB5 are not involved in the dsRNA processing stages of the miRNA biogenesis pathway, but are required to mediate RNA silencing of target genes of DRB2-associated miRNAs.     

Specific interactions between Dicer-like proteins and HYL1/DRB- family dsRNA-binding proteins in <Emphasis Type="Italic">Arabidopsis thaliana</Emphasis>

Proteins that specifically bind double-stranded RNA (dsRNA) are involved in the regulation of cellular signaling events and gene expression, and are characterized by a conserved dsRNA-binding motif (dsRBM). Here we report the biochemical properties of nine such gene products, each containing one or two dsRBMs: four ArabidopsisDicer-like proteins (DCL1-4), ArabidopsisHYL1 and four of its homologs (DRB2, DRB4, DRB5 and OsDRB1). DCL1, DCL3, HYL1 and the four HYL1 homologs exhibit significant dsRNA-binding activity, indicating that these proteins are involved in RNA metabolism. The dsRBMs from dsRBM-containing proteins (dsRBPs) also function as a protein–protein interaction domain and homo- and heterodimerization are essential for biological functioning of these proteins. We show that DRB4 interacts specifically with DCL4, and HYL1 most strongly interacts with DCL1. These results indicate that each HYL1/DRB family protein interacts with one specific partner among the four Dicer-like proteins. Localization studies using GFP fusion proteins demonstrate that DCL1, DCL4, HYL1 and DRB4 localize in the nucleus, while DRB2 is present in the cytoplasm. Subcellular localizations of HYL1, DRB4, DCL1 and DCL4 further strengthen the notion that HYL1 and DCL1, and DRB4 and DCL4, exist as complexes. The presented data suggest that each member of the HYL1/DRB protein family may individually modulate Dicer function through heterodimerization with a Dicer-like protein in vivo.     

The dsRNA-binding protein DRB4 interacts with the Dicer-like protein DCL4 in vivo and functions in the <Emphasis Type="Italic">trans</Emphasis>-acting siRNA pathway

Arabidopsis thaliana encodes four Dicer-like (DCL) proteins and five dsRNA-binding (DRB) proteins. We have previously demonstrated that DCL4 specifically interacts with DRB4 in vitro. Here we describe the interaction between DCL4 and DRB4 in vivo. The phenotype of a mutant with a defect in DCL4 (dcl4-2) was similar to that of a mutant with a defect in DRB4 (drb4-1): both mutant plants had elongated and downwardly curled rosette leaves and over-accumulated anthocyanin. In immunoprecipitation experiments with either anti-DCL4 or anti-DRB4 antibody and crude extracts of wild-type Arabidopsis plants, co-immunoprecipitation of DCL4 and DRB4 was detected, indicating that DCL4 interacts with DRB4 in vivo. This interaction was confirmed by immunoprecipitation experiments using extracts from dcl4-2, drb4-1, or transgenic plants expressing the hemagglutinin-tagged version of DCL4 or DRB4. The results of immunoprecipitation experiments also suggest that most DCL4 is associated with DRB4, but that some DRB4 is free or associated with other proteins. Reduced accumulation of the TAS1 and TAS3 trans-acting siRNA (ta-siRNA) and over accumulation of their target mRNAs (At5g18040 and auxin response factors ARF3 and ARF4) were detected in both drb4-1 and dcl4-2 mutants. These results indicate that DRB4, together with DCL4, functions in the ta-siRNA biogenesis. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users. Yukihiro Nakazawa and Akihiro Hiraguri contributed equally to this work.     

Modified 27-nt dsRNAs with dramatically enhanced stability in serum and long-term RNAi activity

The present study describes improved properties of 27-nt dsRNAs over 21-nt siRNAs, and accents on the possibility to use their modifications and conjugates for direct long-term gene silencing in viable cells and animals, avoiding conventional transfectants. Using a Renilla Luciferase gene-silencing system and cultured cell lines, we established that 27-nt dsRNAs possessed about three to five times higher "long-term" RNAi activity than 21-nt siRNAs and 21-nt dsRNAs. Moreover, if RNA duplexes were preincubated with cell-cultured medium for several hours before their transfection in cells, 21-mer completely lost its RNAi effect, while 27-mer, its amino modifications, thiol modifications, and cholesterol conjugates manifested a strong gene silencing. In attempts to clarify the reason(s) for the higher RNAi activity of 27-nt dsRNAs, we found that they were approximately 100 times more stable than 21-nt siRNA and 21-nt dsRNA in cell-cultured medium supplemented with 10% inactivated serum, approximately 50 times more stable in 90% inactivated serum, and approximately six times more stable in active serum. The 5' sense modification was selected as the most stable, accessible to Dicer, and with highest RNAi potential. The RNAi activity of 5' sense modifications was higher even than the activity of nonmodified 27-nt dsRNA. The 5' sense amino modification also did not influence the activity of 21-nt siRNA, right overhang 25/27-nt (R25D/27), and 25D/27-nt RNAs. The stability of 5' sense modified R25D/27-nt and 25D/27-nt RNAs in serum was lower than that of blunt 27-nt dsRNA. However, these asymmetric RNAs were more active than modified and nonmodified blunt 27-nt dsRNAs, which demonstrates the superiority of the asymmetric design. The 5' sense modifications were considered as most appropriate for conjugation with small signal molecules to facilitate the intracellular delivery of RNA duplex, to preserve its RNAi capacity, and to ensure a possibility for rapid long-term gene silencing in viable cells and animals. The 5' sense conjugation with cholesterol approved this assumption.     

Arabidopsis DRB4 protein in antiviral defense against Turnip yellow mosaic virus infection

RNA silencing is an important antiviral mechanism in diverse eukaryotic organisms. In Arabidopsis DICER‐LIKE 4 (DCL4) is the primary antiviral Dicer, required for the production of viral small RNAs from positive‐strand RNA viruses. Here, we showed that DCL4 and its interacting partner dsRNA‐binding protein 4 (DRB4) participate in the antiviral response to Turnip yellow mosaic virus (TYMV), and that both proteins are required for TYMV‐derived small RNA production. In addition, our results indicate that DRB4 has a negative effect on viral coat protein accumulation. Upon infection DRB4 expression was induced and DRB4 protein was recruited from the nucleus to the cytoplasm, where replication and translation of viral RNA occur. DRB4 was associated with viral RNA in vivo and directly interacted in vitro with a TYMV RNA translational enhancer, raising the possibility that DRB4 might repress viral RNA translation. In plants the role of RNA silencing in viral RNA degradation is well established, but its potential function in the regulation of viral protein levels has not yet been explored. We observed that severe infection symptoms are not necessarily correlated with enhanced viral RNA levels, but might be caused by elevated accumulation of viral proteins. Our findings suggest that the control of viral protein as well as RNA levels might be important for mounting an efficient antiviral response.     

An Arabidopsis RNase III-like protein,AtRTL2, cleaves double-stranded RNA in vitro

Class 1 ribonuclease III (RNase III), found in bacteria and yeast, is involved in processing functional RNA molecules such as ribosomal RNAs (rRNAs). However, in Arabidopsis thaliana, the lack of an obvious phenotype or quantitative change in mature rRNAs in class 1 RNase III (AtRTL2) mutants and overexpressing plants suggests that AtRTL2 is not involved in rRNA maturation. We characterized the in vitro activity of AtRTL2 to consider its in vivo function. AtRTL2 cleaved double-stranded RNA (dsRNA) specifically in vitro, yielding products of approximately 25 nt or longer in length, in contrast to 10–20 nt long products in bacteria and yeasts. Although dsRNA-binding activity was not detected, the dsRNA-binding domains in AtRTL2 were essential for its dsRNA-cleaving activity. Accumulation of small RNAs derived from transgene dsRNAs was increased when AtRTL2 was transiently expressed in Nicotiana benthamiana leaves by agroinfiltration. These results raise the possibility that AtRTL2 has functions distinct from those of other class 1 RNase IIIs in vivo.     

A comprehensive analysis of precursor microRNA cleavage by human Dicer

Dicer cleaves double-stranded RNAs (dsRNAs) or precursor microRNAs (pre-miRNAs) to yield ∼22-nt RNA duplexes. The pre-miRNA structure requirement for human Dicer activity is incompletely understood. By large-scale in vitro dicing assays and mutagenesis studies, we showed that human Dicer cleaves most, although not all, of the 161 tested human pre-miRNAs efficiently. The stable association of RNAs with Dicer, as examined by gel shift assays, appears important but is not sufficient for cleavage. Human Dicer tolerates remarkable structural variation in its pre-miRNA substrates, although the dsRNA feature in the stem region and the 2-nt 3′-overhang structure in a pre-miRNA contribute to its binding and cleavage by Dicer, and a large terminal loop further enhances pre-miRNA cleavage. Dicer binding protects the terminal loop from digestion by S1 nuclease, suggesting that Dicer interacts directly with the terminal loop region.     

The p53-induced Wig-1 protein binds double-stranded RNAs with structural characteristics of siRNAs and miRNAs

Wig-1 is a p53-induced zinc finger protein. Here we show that human Wig-1 binds long (>or=23 bp) dsRNAs with 5'-overhangs. The first zinc finger domain is necessary but not sufficient for this dsRNA-binding in vitro. Wig-1 also binds dsRNA in living cells via zinc fingers 1 and 2. Both zinc fingers 1 and 2 are important for Wig-1-mediated growth suppression. Moreover, Wig-1 binds 21 bp dsRNAs with 3'-protruding ends. These findings demonstrate that human Wig-1 can bind different types of dsRNAs, including dsRNAs resembling small interfering RNAs (siRNAs) and microRNAs (miRNAs), and indicate that dsRNA binding has a role in Wig-1-mediated regulation of cell growth.     

RDE-4 preferentially binds long dsRNA and its dimerization is necessary for cleavage of dsRNA to siRNA

In organisms ranging from Arabidopsis to humans, Dicer requires dsRNA-binding proteins (dsRBPs) to carry out its roles in RNA interference (RNAi) and micro-RNA (miRNA) processing. In Caenorhabditis elegans, the dsRBP RDE-4 acts with Dicer during the initiation of RNAi, when long dsRNA is cleaved to small interfering RNAs (siRNAs). RDE-4 is not required in subsequent steps, and how RDE-4 distinguishes between long dsRNA and short siRNA is unclear. We report the first detailed analysis of RDE-4 binding, using purified recombinant RDE-4 and various truncated proteins. We find that, similar to other dsRBPs, RDE-4 is not sequence-specific. However, consistent with its in vivo roles, RDE-4 binds with higher affinity to long dsRNA. We also observe that RDE-4 is a homodimer in solution, and that the C-terminal domain of the protein is required for dimerization. Using extracts from wild-type and rde-4 mutant C. elegans, we show that the C-terminal dimerization domain is required for the production of siRNA. Our findings suggest a model for RDE-4 function during the initiation of RNAi.     

dsRNA binding properties of RDE-4 and TRBP reflect their distinct roles in RNAi

Double-stranded RNA (dsRNA)-binding proteins facilitate Dicer functions in RNA interference. Caenorhabditis elegans RDE-4 facilitates cleavage of long dsRNA to small interfering RNA (siRNA), while human trans-activation response RNA-binding protein (TRBP) functions downstream to pass siRNA to the RNA-induced silencing complex. We show that these distinct in vivo roles are reflected in in vitro binding properties. RDE-4 preferentially binds long dsRNA, while TRBP binds siRNA with an affinity that is independent of dsRNA length. These properties are mechanistically based on the fact that RDE-4 binds cooperatively, via contributions from multiple domains, while TRBP binds noncooperatively. Our studies offer a paradigm for how dsRNA-binding proteins, which are not sequence specific, discern dsRNA length. Additionally, analyses of the ability of RDE-4 deletion constructs and RDE-4/TRBP chimeras to reconstitute Dicer activity suggest RDE-4 promotes activity using its dsRNA-binding motif 2 to bind dsRNA, its linker region to interact with Dicer, and its C-terminus for Dicer activation.     

Characterization of DCL4 missense alleles provides insights into its ability to process distinct classes of dsRNA substrates

In the model plant Arabidopsis thaliana, four Dicer‐like proteins (DCL1–4) mediate the production of various classes of small RNAs (sRNAs). Among these four proteins, DCL4 is by far the most versatile RNaseIII‐like enzyme, and previously identified dcl4 missense alleles were shown to uncouple the production of the various classes of DCL4‐dependent sRNAs. Yet little is known about the molecular mechanism behind this uncoupling. Here, by studying the subcellular localization, interactome and binding to the sRNA precursors of three distinct dcl4 missense alleles, we simultaneously highlight the absolute requirement of a specific residue in the helicase domain for the efficient production of all DCL4‐dependent sRNAs, and identify, within the PAZ domain, an important determinant of DCL4 versatility that is mandatory for the efficient processing of intramolecular fold‐back double‐stranded RNA (dsRNA) precursors, but that is dispensable for the production of small interfering RNAs (siRNAs) from RDR‐dependent dsRNA susbtrates. This study not only provides insights into the DCL4 mode of action, but also delineates interesting tools to further study the complexity of RNA silencing pathways in plants, and possibly other organisms.     

Biochemical Requirements for Two Dicer-Like Activities from Wheat Germ

RNA silencing pathways were first discovered in plants. Through genetic analysis, it has been established that the key silencing components called Dicer-like (DCL) genes have been shown to cooperatively process RNA substrates of multiple origin into distinct 21, 22 and 24 nt small RNAs. However, only few detailed biochemical analysis of the corresponding complexes has been carried out in plants, mainly due to the large unstable complexes that are hard to obtain or reconstitute in heterologous systems. Reconstitution of activity needs thorough understanding of all protein partners in the complex, something that is still an ongoing process in plant systems. Here, we use biochemical analysis to uncover properties of two previously identified native dicer-like activities from wheat germ. We find that standard wheat germ extract contains Dicer-like enzymes that convert double-stranded RNA (dsRNA) into two classes of small interfering RNAs of 21 and 24 nt in size. The 21 nt dicing activity, likely an siRNA producing complex known as DCL4, is 950 kDa-1.2 mDa in size and is highly unstable during purification processes but has a rather vast range for activity. On the contrary, the 24 nt dicing complex, likely the DCL3 activity, is relatively stable and comparatively smaller in size, but has stricter conditions for effective processing of dsRNA substrates. While both activities could process completely complementary dsRNA albeit with varying abilities, we show that DCL3-like 24 nt producing activity is equally good in processing incompletely complementary RNAs.     

Double-stranded RNA binding may be a general plant RNA viral strategy to suppress RNA silencing

In plants, RNA silencing (RNA interference) is an efficient antiviral system, and therefore successful virus infection requires suppression of silencing. Although many viral silencing suppressors have been identified, the molecular basis of silencing suppression is poorly understood. It is proposed that various suppressors inhibit RNA silencing by targeting different steps. However, as double-stranded RNAs (dsRNAs) play key roles in silencing, it was speculated that dsRNA binding might be a general silencing suppression strategy. Indeed, it was shown that the related aureusvirus P14 and tombusvirus P19 suppressors are dsRNA-binding proteins. Interestingly, P14 is a size-independent dsRNA-binding protein, while P19 binds only 21-nucleotide ds-sRNAs (small dsRNAs having 2-nucleotide 3' overhangs), the specificity determinant of the silencing system. Much evidence supports the idea that P19 inhibits silencing by sequestering silencing-generated viral ds-sRNAs. In this study we wanted to test the hypothesis that dsRNA binding is a general silencing suppression strategy. Here we show that many plant viral silencing suppressors bind dsRNAs. Beet yellows virus Peanut P21, clump virus P15, Barley stripe mosaic virus gammaB, and Tobacco etch virus HC-Pro, like P19, bind ds-sRNAs size-selectively, while Turnip crinkle virus CP is a size-independent dsRNA-binding protein, which binds long dsRNAs as well as ds-sRNAs. We propose that size-selective ds-sRNA-binding suppressors inhibit silencing by sequestering viral ds-sRNAs, whereas size-independent dsRNA-binding suppressors inactivate silencing by sequestering long dsRNA precursors of viral sRNAs and/or by binding ds-sRNAs. The findings that many unrelated silencing suppressors bind dsRNA suggest that dsRNA binding is a general silencing suppression strategy which has evolved independently many times.     

Oryza sativa dicer-like4 reveals a key role for small interfering RNA silencing in plant development

MicroRNAs and small interfering RNAs (siRNAs) are two classes of small regulatory RNAs derived from different types of precursors and processed by distinct Dicer or Dicer-like (DCL) proteins. During evolution, four Arabidopsis thaliana DCLs and six rice (Oryza sativa) DCLs (Os DCLs) appear to have acquired specialized functions. The Arabidopsis DCLs are well characterized, but those in rice remain largely unstudied. Here, we show that both knockdown and loss of function of rice DCL4, the homolog of Arabidopsis DCL4, lead to vegetative growth abnormalities and severe developmental defects in spikelet identity. These phenotypic alterations appear to be distinct from those observed in Arabidopsis dcl4 mutants, which exhibit accelerated vegetative phase change. The difference in phenotype between rice and Arabidopsis dcl4 mutants suggests that siRNA processing by DCL4 has a broader role in rice development than in Arabidopsis. Biochemical and genetic analyses indicate that Os DCL4 is the major Dicer responsible for the 21-nucleotide siRNAs associated with inverted repeat transgenes and for trans-acting siRNA (ta-siRNA) from the endogenous TRANS-ACTING siRNA3 (TAS3) gene. We show that the biogenesis mechanism of TAS3 ta-siRNA is conserved but that putative direct targets of Os DCL4 appear to be differentially regulated between monocots and dicots. Our results reveal a critical role of Os DCL4-mediated ta-siRNA biogenesis in rice development.