RNA silencing phenomena, either the regulation of mRNA translation or regulation of mRNA degradation, intersect at the ribonuclease Dicer. In animals, the double-stranded RNA-specific endonuclease Dicer produces two classes of functionally distinct, tiny RNAs: microRNAs (miRNAs) and small interfering RNAs (siRNAs). miRNAs regulate mRNA translation, whereas siRNAs direct RNA destruction via the RNA interference (RNAi) pathway. siRNAs and miRNAs then direct a RNA-induced silencing complex (RISC) to cleave mRNA or block its translation (RNAi). Mutations have been characterized in the Drosophila dicer-1 and dicer-2 genes. Mutation in dicer-1 blocks processing of micro RNA precursors, whereas dicer-2 mutants are defective for processing siRNA precursors. It has been recently found that Drosophila Dicer-1 and Dicer-2 are also components of siRNA-dependent RISC (siRISC). Dicer-1 and Dicer-2 are required for siRNA-directed mRNA cleavage, though the RNase III activity of Dicer-2 is not required. Dicer-1 and Dicer-2 facilitate distinct steps in the assembly of siRISC. However, Dicer-1 (but not Dicer-2) is essential for miRISC-directed translation repression. Thus, siRISCs and miRISCs are different with respect to Dicers in Drosophila (Lee, 2004).

Small RNAs influence a wide variety of biological processes by silencing the expression of genes within organisms. These RNAs, with a size of about 22 nucleotides, influence development, genome organization, viral and transposon defense, and disease (Hannon, 2002). There are two classes of small RNAs, and they exert their powers of silencing differently. One class is processed from longer double-stranded (dsRNA) precursor molecules with perfect complementarity. The dsRNAs are cleaved into small interfering RNAs (siRNAs) that are 21-23 nucleotide duplexes. They act as guides for a siRNA-induced silencing complex (siRISC) to target complementary mRNAs. If such an mRNA molecule is found, the base pairing interactions between siRNA and mRNA lead to cleavage of the mRNA molecule and its degradation. A second class of small RNAs, the microRNAs (miRNAs), is processed from stem-loop RNA precursors (pre-miRNAs) that are encoded within plant and animal genomes. The known functions of a few of these miRNAs indicate that they play widespread roles in growth and development (Abrahante, 2003; Brennecke, 2003; Lee, 1993; Lin, 2003; Llave, 2002; Palatnik, 2003; Reinhart, 2000). Animal miRNAs silence gene expression primarily by blocking the translation of mRNA transcripts into protein. They act as guides for a multiprotein complex, miRISC, which identifies mRNAs with imperfect complementarity in the 3' untranslated region of the message (Lee, 2004 and reference therein).

The extent of base pairing between small RNA and mRNA determines the outcome of silencing. An miRNA will direct mRNA cleavage if the target transcript is perfectly complementary in sequence (Hutvagner, 2002). Conversely, an siRNA will block protein synthesis if the target transcript has partial complementarity (Doench, 2003; Zeng, 2003). These observations imply that the extent of base pairing between small RNA and mRNA determines the outcome of silencing. It is unclear whether a single silencing complex is competent to both cleave mRNA and block translation, or whether an miRNA (or siRNA) associates with two biochemically distinct RISC complexes -- one able to cleave mRNA and another able to block translation (Lee, 2004 and reference therein).

Although dsRNAs and pre-miRNAs are structurally distinct, they are both processed into siRNAs and miRNAs, respectively, by the Dicer class of RNase III enzymes (Bernstein, 2001; Grishok, 2001; Ketting, 2001; Lee, 2002). Dicer makes staggered cuts in dsRNA to form siRNA duplexes with 3' overhangs, each strand bearing 5' phosphate and 3' hydroxyl termini (Myers, 2003; Provost, 2002). Dicer exhibits little sequence specificity for cleavage, though it favors processing from the end of a dsRNA substrate (Elbashir, 2001). The siRNA product then assembles into a siRISC that retains either the sense or antisense strand of the duplex (Hammond, 2000; Nykanen, 2001). Dicer also processes pre-miRNA (Hutvagner, 2001; Ketting, 2001; Lee, 2002). However, in the course of assembly into miRISC, one strand is preferentially retained from the siRNA-like duplex. Dicer mutants are defective for both transcript destruction and translational repression, suggesting that Dicer is required in both the siRNA and miRNA pathways (Grishok, 2001; Ketting, 2001; Knight, 2001). This dual role has made its genetic analysis more complicated (Lee, 2004 and reference therein).

In addition to dsRNA processing, Dicer appears to play some other, as yet ill-defined role in the siRNA pathway. Dicer functions downstream of siRNA production, as depletion of Dicer in mammalian cells reduces the effectiveness of added siRNAs (Doi, 2003). Dicer physically associates with protein components of RISC, and it binds siRNAs tightly in vitro (Doi, 2003; Hammond, 2001; Liu, 2003; Tabara, 2002; Tang, 2003). In Drosophila, this latter interaction is enhanced by an auxiliary dsRNA binding protein, R2D2 (Liu, 2003). Dicer-1 and Dicer-2 are shown to facilitate distinct steps in the assembly of siRISC. However, Dicer-1 but not Dicer-2 is essential for miRISC-directed translation repression. Thus, siRISCs and miRISCs are different with respect to Dicers in Drosophila (Lee, 2004).

Thus Dcr-1 and Dcr-2 generate different classes of small RNAs in Drosophila. Dcr-1 processes pre-miRNAs while Dcr-2 processes dsRNAs. This specificity may reflect the distinct structural properties of the two types of substrates. miRNA precursors are imperfectly paired stem loops, whereas siRNA precursors are typically long dsRNA helices with at least one blunt end. Dcr-1 might preferentially bind and attack imperfectly paired helices characteristic of miRNAs. Dcr-2 might prefer dsRNA with perfect complementarity. Another difference between siRNA and miRNA precursors is their abundance. Typically, dsRNA substrates for siRNA processing are highly abundant in cells, resulting from viral infection or promiscuous transcription. Precursors of miRNAs are expressed from endogenous genes and are not highly abundant. Dcr-1 and Dcr-2 could utilize one substrate over another according to differences in their kinetic and thermodynamic properties for each type of precursor. Finally, Dcr-1 and Dcr-2 might commit to different substrates because they contain different biochemical activities. For example, only Dcr-2 contains a DExH helicase domain and only Dcr-1 contains a PAZ domain. The PAZ domain might help link Dcr-1 to miRNA precursor molecules as they are shuttled from the cell's nucleus. These experiments clearly indicate that Dcr-2 DExH helicase activity is required for dsRNA processing. The RNA helicase domain might be needed to move Dcr-2 along the dsRNA substrate or displace Dcr-2 from the dsRNA substrate upon cleavage (Lee, 2004).

In contrast to their processing specificities, both Dcr-1 and Dcr-2 are required for siRNA-directed transcript cleavage and gene silencing. In both cases the requirement is not absolute, arguing that there is some overlapping redundancy between them. Both Dcr-1 and Dcr-2 are required for assembly of siRNA into siRISC, but play distinct roles. Dcr-2 is required to form a stable siRNA-protein complex, which contains Dcr-2 and R2D2 (Liu, 2003; Pham, 2004). This complex initiates siRISC assembly. Dcr-1 has a distinct role from Dcr-2. Dcr-1 is not necessary to form a stable initiator complex but instead functions to form a stable intermediate in siRISC assembly. It is likely that Dcr-1 is directly involved since Dcr-1 protein directly associates with siRNA in initiator complexes, intermediate complexes, and assembled siRISC (Pham, 2004). It is suggested that Dcr-1 within the initiator complex facilitates the stable association of other factors and formation of an intermediate complex (Lee, 2004).

Each Dicer has distinct qualities with regards to siRISC assembly and yet they are somewhat redundant. They may function analogously to TBP binding to a TATA sequence, initiating assembly of a transcription complex. Each component alone (Dicer or siRNA) is not sufficient to start assembly, but the combination provides enough interaction energy to drive the process. Dcr-1 and Dcr-2 remain closely associated with siRNA in assembled siRISC. This could also be analogous to TBP, which remains a component of the assembled transcription complex. However, in the process of siRISC assembly, siRNA duplex is unwound and a single strand is retained. Once this strand finds a perfect mRNA complement, the transition back to a dsRNA state initiates RNA cleavage. Clearly, Dcr-2 is not directly required for siRNA duplex unwinding, since the RNA helicase activity of Dcr-2 is not necessary for siRNA-dependent mRNA cleavage. One tantalizing notion is that the nuclease activities of the Dicers are used for cleavage of the mRNA strand within the hybrid duplex. But if so, such activities do not require the RNase III activity of both Dicers, since point mutants that abolish Dcr-2 RNase III activity still promote mRNA cleavage. One possibility is that a nonconventional catalytic site within Dcr-2 cleaves the hybrid duplex, or that Dcr-1 can efficiently cleave the duplex in place of Dcr-2. Alternatively, the Dicers might hold a siRNA-mRNA duplex in place for attack by the cleavage enzyme. The position of the mRNA cleavage site is highly exact (Elbashir, 2001), corresponding to the bond precisely ten bases from the corresponding 5' end of the siRNA strand (Lee, 2004).

Genetic analysis demonstrates that Dcr-1 but not Dcr-2 is required for gene silencing by miRNAs. Loss of Dcr-2 has a profound effect on dsRNA processing but no significant effect on Drosophila development. Although this suggests that endogenous dsRNAs do not play a critical role in development, it is possible that Dcr-1 has weak dsRNA processing activity and in a Dcr-2 mutant this weak activity might process enough endogenous dsRNA to fulfill possible developmental functions. Loss of Dcr-1 derepresses miRNA target genes and causes profound changes in development and patterning. It cannot be definitely said whether Dcr-1 and mature miRNAs form effector complexes, analogous to those formed by Dcr-1 and siRNAs. However, Dicer coimmunoprecipitates with miRNAs (Lee, 2003), suggesting that a similar mechanism is at work. Because siRISC contains Dcr-2, and because Dcr-2 is dispensable for miRISC function, it argues that siRISCs and miRISCs are inherently different. This might correlate with the functional differences seen between siRISC and miRISC with regard to mRNA cleavage and translation. That said, the capacity for Dcr-1 to act in both siRNA and miRNA pathways could explain how small RNAs of either class can cleave mRNA or block its translation, depending on the degree of complementarity between the small RNA and the mRNA target (Hutvagner, 2002; Doench, 2003; Zeng, 2003). Dcr-1 could recruit a miRNA into a RISC with cleavage activity, and Dcr-1 could recruit a siRNA into a RISC that represses translation (Lee, 2004).

GENE STRUCTURE

cDNA clone length - 5665 base pairs

Bases in 5' UTR - 80

Exons - 8

Bases in 3' UTR - 419

PROTEIN STRUCTURE

Amino Acids - 1722

Structural Domains

Most Dicer orthologs contain a DExH-type ATP-dependent RNA helicase domain at their amino termini, and a PAZ domain that is also found in some protein components of RISC. Interestingly, Drosophila Dcr-1 lacks a functional helicase domain, whereas Dcr-2 lacks a PAZ domain. This suggests that the two enzymes might have different or complementary biochemical activities. However, like other members of the Dicer family, both Dcr-1 and Dcr-2 contain two RNase III domains and a dsRNA binding domain at their carboxy-termini (Lee, 2004)

date revised: 30 June 2003

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.