PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
 
Mol Cell Biol. Feb 2009; 29(3): 861–870.
Published online Dec 1, 2008. doi:  10.1128/MCB.01524-08
PMCID: PMC2630677

A Drosophila pasha Mutant Distinguishes the Canonical MicroRNA and Mirtron Pathways[down-pointing small open triangle]

Abstract

Canonical primary microRNA (miRNA) transcripts and mirtrons are proposed to transit distinct nuclear pathways en route to generating mature ~22 nucleotide regulatory RNAs. We generated a null allele of Drosophila pasha, which encodes a double-stranded RNA-binding protein partner of the RNase III enzyme Drosha. Analysis of this mutant yielded stringent evidence that Pasha is essential for the biogenesis of canonical miRNAs but is dispensable for the processing and function of mirtron-derived regulatory RNAs. The pasha mutant also provided a unique tool to study the developmental requirements for Drosophila miRNAs. While pasha adult somatic clones are similar in many respects to those of dicer-1 clones, pasha mutant larvae revealed an unexpected requirement for the miRNA pathway in imaginal disc growth. These data suggest limitations to somatic clonal analysis of miRNA pathway components.

MicroRNAs (miRNAs) are endogenous, ~22 nucleotide (nt), regulatory RNAs that associate with Argonaute proteins to repress target transcripts posttranscriptionally (8, 9). miRNAs constitute one of the largest gene families in animal genomes, with over 600 members in humans. Although they can regulate perfectly complementary targets, the vast majority of animal miRNA targets are defined by as little as 7 nt of complementarity to positions 2 to 8 of the miRNA, also known as the miRNA seed (41). Evolutionary conservation of seed matches suggests that 20 to 30% of Drosophila and mammalian transcripts actively maintain functional target sites for one or more miRNAs, and presumably many other transcripts contain functional sites that are either not conserved and/or have seed mismatches (41).

In Drosophila, as in other animals, miRNA biogenesis proceeds in a stepwise, cell-compartmentalized manner (Fig. (Fig.1).1). Canonical miRNAs are initially transcribed, mostly by RNA polymerase II, as long primary transcripts (pri-miRNAs) bearing one or more miRNA hairpins (32). Most of these hairpins are located in the exons or introns of noncoding RNAs, but approximately one-third are located in the introns of protein-coding genes. Pri-miRNA hairpins contain >30 nt of stem, with the basal hairpin duplex serving to recruit the double-strand RNA-binding domain protein Pasha (also known as DGCR8 in mammals) (7, 14, 17, 27). Pasha binds the nuclear RNase III enzyme Drosha, which “crops” the base of the hairpin ~10 nt away from the junction of its single-stranded flanks to yield the pre-miRNA hairpin (16, 17, 31). The pre-miRNA is exported to the cytoplasm via Exportin-5, where it is cleaved by the cytoplasmic RNase III enzyme Dicer-1 (Dcr-1) (4) and its double-strand RNA-binding domain partner Loquacious (Loqs) (8). From the resultant ~22-nt duplex, one strand preferentially enters an Argonaute-1 (AGO1) complex and guides it to seed-complementary targets (8).

FIG. 1.
Canonical miRNA and mirtron pathways in Drosophila. Key protein families include RNase III endonucleases (Drosha and Dicer-1), double-stranded RNA-binding domain proteins (Pasha and Loqs) and Argonaute effectors (AGO1 and AGO2). Canonical miRNA precursors ...

Recently, the analysis of Drosophila small RNAs revealed that a subclass of miRNAs derives from atypical hairpin precursors termed mirtrons (38, 44). Their defining feature is that the ends of mirtron hairpins coincide precisely with 5′ and 3′ splice sites of introns of protein-coding genes (Fig. (Fig.1).1). Biogenesis studies carried out primarily using knockdowns of candidate factors in Drosophila S2 cells provided evidence that mirtrons use the splicing machinery to bypass Drosha cleavage. Following their linearization by intron lariat debranching enzyme, mirtrons gain access to Exportin-5 and are subsequently treated in the cytoplasm as conventional pre-miRNA hairpins. The mirtron pathway has been most thoroughly studied in Drosophila, but the analysis of large-scale small RNA sequence catalogs permitted the confident categorization of mirtrons in nematodes (44), diverse mammals (1), and most recently in chickens (13).

Although the initial studies of mirtron biogenesis were well supported, a potential caveat was their reliance on knockdown strategies. This is potentially significant in light of recent studies of Loqs. This Dcr-1 cofactor was originally classified as a core component of the miRNA biogenesis pathway based on studies of loqs knockdown in S2 cells and a hypomorphic loqs allele (10, 22, 45). Since these conditions reduced the level of at least some miRNAs and caused pre-miRNA hairpins to accumulate, one might have expected the complete loss of Loqs to confer a stronger effect on miRNA maturation. Perhaps surprisingly then, subsequent analysis of a loqs deletion revealed that the biogenesis of many miRNAs was only subtly affected in the loqs-null condition (34). This is in strong contrast to the loss of dcr-1, for which homozygous mutant cells are unable to generate miRNAs (34).

In the present study, we describe the generation of a pasha-null allele and use it to validate the hypothesis that canonical miRNAs and mirtrons transit distinct nuclear pathways. In particular, mirtrons but not canonical miRNAs are produced and can repress targets in pasha mutants. Because of its maternal contribution, homozygous pasha mutants survive embryogenesis and larval stages. This makes it a particularly useful genetic tool among mutants in core components of the Drosophila miRNA pathway. In particular, the pasha mutant demonstrates that miRNAs are strictly required for the growth of all imaginal discs, a conclusion that cannot be derived from the clonal analysis of either pasha or dcr-1. Since the postembryonic functions of invertebrate dcr-1 and vertebrate DGCR8 and Dicer must be analyzed using mosaics, our findings suggest that caution must be exercised when using this technique to infer whether a given process does or does not require miRNAs.

MATERIALS AND METHODS

Generation of pashaKO.

We used PBac{WH}f07241 and PBac{RB}e00907, which are inserted 5′ and 3′ of the pasha locus, respectively. We induced recombination in transheterozygous pBac flies that carried hs-FLP by heat shocking 48- to 72-h-old larvae at 37°C for 1 h. The F1 female progeny were crossed to w; Dr/TM3, Sb males, and individual F2 balanced males bearing putative deletion chromosomes were backcrossed to the balancer to generate stocks. We screened for deletion events using PCR, and isolated a recombinant that deleted the ~10-kb interval containing pasha; this allele retains a mini-white+ insertion at the pasha locus. We then recombined pashaKO onto a FRT82B chromosome for clonal analysis.

pasha rescue construct.

A 4.1-kb fragment including the entire pasha gene was amplified by PCR using primers with XhoI linkers: 5′ primer TTTCAAAATGGCCAATAG and 3′ primer CTCGGACTTTCTCTCTGC. After XhoI digestion, the fragment was cloned into pCasper4 in the sense orientation with respect to the P element. Transgenic lines were generated by standard methods, and a second chromosome insertion was used to rescue pashaKO. The specific effect of the deletion on the expression of pasha, but not the neighboring gene CG1792, was verified by using semiquantitative and quantitative reverse transcription-PCR (qRT-PCR) using poly(A)+ RNA and the following primer sets: pasha-F (5′-GTTCAAGGAGCTCCAAAACG-3′)/pasha-R (5′-CCTTGACATCGGGAATGAGT-3′) and CG1792-F (5′-ACGGCGATTGCTTTCCGGGAAC-3′)/CG1792-R (5′-GGTGTTCCTCGGGTAGAAGGTC-3′).

Analysis of miRNA/mirtron processing.

We used TRIzol to isolate total RNA from Canton S, ago2414 (39), loqsKO (34), dcr-2L811fs (33), and pashaKO third-instar larvae, and followed previously described methods (38) for polyacrylamide gel Northern analysis using LNA probes (Exiqon). Probes were complementary to the mature sequences of canonical miRNAs or mirtron-derived miRNAs (http://microrna.sanger.ac.uk/sequences/index.shtml).

qRT-PCR analyses of pri-miRNA segments were performed according to previously described methods (40). We performed six replicate qRT-PCR assays on each of three independently generated cDNA preparations from poly(A)+ RNA, using the following primer sets: miR-1-F (5′-GTTAGCCGCGTTGTGGAAAATC-3′)/miR-1-R (5′-CATTTCATTACGGTTCTACTTCTG-3′), miR-8-F (5′-AGAACTTTGAGCTTCCTCTGGC-3′)/miR-8-R (5′-TTTGGTGCTGCTGCTGCTGTTG-3′), miR-10-F (5′-CCGCGATTGCCTAGCGGACTTC-3′)/miR-10-R (5′-TTTCCGCTTGCCATCAGCAACAC-3′), miR-124-F (5′-ACATTGCATAACGACATAAAGCC-3′)/miR-124-R (5′-AATTTGTCTATTATGATTTCAGGC-3′), miR-263a-F (5′-AGTGCATGCGGGTGAGTAATCC-3′)/miR-263a-R (5′-TAACTTTGAAAGTTTCGGATTTCG-3′), miR-276a-F (5′-AAAAGGGAAACGCGCTGCCAAG-3′)/miR-276a-R (5′-CGTTTGTCCAGCGTTTTCTCATC-3′), miR-279-F (5′-ATTGAAATTAAAGAGGAGGCGAG-3′)/miR-279-R (5′-AAGTTTGTCAAGAAAACACGTGC-3′), miR-305-F (5′-GAAATGCTCGCAGGCGAGTCC-3′)/miR-305-R (5′-GTTGAACACTTGTATCGGTCGC-3′), miR-317-F (5′-ACGGTTTGTGTCTCTGCTGAGC-3′)/miR-317-R (5′-CTGTGGGGCATTCTCGTTATCC-3′), and miR-bantam-F (5′-CGCTCAGATGCAGATGTTGTTG-3′)/miR-bantam-R (5′-TCGACCATCGGAATGTGGAATG-3′).

For conventional Northern analysis, 20 μg of total RNA was separated on 1% agarose with 18% formaldehyde and 1× morpholinepropanesulfonic acid. After electrophoresis, the RNA was transferred by capillary flow to Hybond N+ membrane and UV cross-linked. Blot hybridizations were performed at 65°C overnight in Church hybridization buffer (0.5 M Church phosphate buffer, 1 mM EDTA, 7% sodium dodecyl sulfate [SDS]) with cDNA probes. Membranes were washed twice with 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% SDS for 15 min at room temperature and then twice with 0.5× SSC-0.1% SDS for 15 min at 65°C. [α-32P]dCTP-labeled probes were generated by using a random primed DNA labeling kit (Roche) and templates that were amplified from genomic DNA using the following primers: pri-mir-1F (5′-CAGAAGTAGAACCGTAATGAAATG-3′), pri-mir-1R (5′-TGTCGATGGAATTGCTTACGTAC-3′; a 360-nt fragment just downstream of the mir-1 hairpin), ago2-F (5′-GAGCACTTGCGCGTGTATAA-3′), and ago2-R (5′-AATCGTTCGCTTTGCGTACT-3′; a 700-nt fragment of the coding region).

Analysis of miRNA/mirtron function.

We analyzed a number of genotypes. Immunostaining was performed according to a previous report (25), using rat anti-Elav, mouse anti-β-galactosidase (Developmental Studies Hybridoma Bank), or rabbit anti-GFP, followed by Alexa 488- or 568-conjugated secondary antibodies. The following genotypes were evaluated (“X” stands for either a control third chromosome arm, dcr-1Q1147X, or pashaKO): (i) random somatic clones—hs-FLP; FRT82B, arm-lacZ/FRT82B, X; (ii) small eye clones—ey-FLP; FRT82B, arm-lacZ/FRT82B, X; (iii) large eye clones—ey-FLP; FRT82B, ubi-GFP, M(3)/FRT82B, X (Dickson method) or ey-Gal4, UAS-FLP/+; FRT82B, GMR-hid, l(3)cl/FRT 82B, X (Stowers method); (iv) notum clones—UAS-FLP/+; C684-Gal4, FRT82B/FRT82B, X; (v) expression of baculovirus p35 in dcr-1 or pasha notum clones—109-68-Gal4, UAS-p35/UAS-FLP; FRT82B, arm-lacZ/FRT82B, X; (vi) miRNA sensor test—hs-FLP; tub-GFP-miR-7 sensor, UAS-DsRed-mir-7/+; tub-Gal4, FRT82B, tub-Gal80/FRT82B, X; (vii) mirtron sensor test—hs-FLP; tub-GFP-miR-1004 sensor, UAS-DsRed-mir-1004/+; tub-Gal4, FRT82B, tub-Gal80/FRT82B, X; and (viii) germ line clones—hs-FLP; FRT82B, ovoD/FRT82B, pasha.

RESULTS

Generation of a pashaKO deletion allele.

We used FLP-mediated recombination to delete ~10 kb of genomic sequence between FRT-bearing piggyBac transposons that flank the pasha locus (51) (Fig. (Fig.2A).2A). The consequent deletion allele retains a hybrid piggyBac marked by mini-white and is not predicted to affect any other known genes (Fig. (Fig.2B).2B). PCR analysis showed that the engineered chromosome indeed contains both left and right flanks of the progenitor elements (Fig. (Fig.2C).2C). In addition, we were able to detect the unique hybrid product that bridges the hybrid piggyBac present in the deletion allele (Fig. (Fig.2C2C).

FIG. 2.
Scheme for the generation of the pashaKO deletion allele. (A) FLP-mediated recombination was used to delete genomic sequence between FRT-bearing piggyBac elements that flank the pasha locus. (B) The progenitor chromosomes bear piggyBac insertion in trans ...

Animals homozygous for the deletion died as late third-instar larvae, suggesting that pasha is an essential locus. RT-PCR analyses showed that transcription of the neighboring locus CG1792 was still detected in homozygous pasha mutants, indicating that the deletion did not adversely affect CG1792 expression (Fig. (Fig.2D).2D). In fact, qRT-PCR tests showed a slight increase (~1.65-fold) in CG1792 in the pasha mutant (see Fig. S1 in the supplemental material), suggesting that CG1792 is directly or indirectly affected by miRNA depletion. As a stringent test of whether the lethality of the mutant was specifically attributable to the loss of Pasha, we generated a 4.1-kb rescue transgene containing only the pasha gene. The pasha+ insertion rescued deletion homozygotes to adulthood in Mendelian ratio, providing firm confirmation that the deletion solely affects pasha function. Consequently, we refer to this allele as pashaKO.

pashaKO mutants are deficient in processing canonical miRNAs but not mirtrons.

The function of Drosophila pasha in small RNA biogenesis has thus far been tested only in tissue culture cells or with in vitro assays. The availability of a genuine pasha mutant allowed us to test its endogenous requirement for small RNA biogenesis in the animal. We therefore prepared RNA from Canton S and homozygous mutant pashaKO, loqsKO, dcr-2Q1147X, and ago2414 third-instar larvae and analyzed their small RNAs using Northern analysis.

Even though Loqs is often portrayed as a core component of the miRNA biogenesis pathway, the steady-state level of only some mature miRNAs is reduced in the absence of Loqs (Fig. (Fig.3A)3A) (34). Instead, the predominant effect is the accumulation of pre-miRNA hairpins (Fig. 3A and B), indicative of their suboptimal but detectable processing by Dcr-1. In contrast, we observed that all ten canonical miRNAs tested that were expressed in wild-type larvae were strongly reduced in pashaKO mutants, regardless of whether their dependence on Loqs was strong or weak (Fig. 3A and B and see Fig. S2 in the supplemental material). Both mature and pre-miRNA species were depleted, data that demonstrate Pasha to be an essential component of the Drosophila miRNA biogenesis machinery.

FIG. 3.
Small RNA expression in mutants for canonical miRNA and RNAi factors. Total RNAs were extracted from homozygous larvae of the following genotypes: Canton S (CS) = wild type, ago2 = ago2414; dcr2 = dcr-2L811fs; loqs = loqs ...

We next analyzed the accumulation of miRNAs derived from mirtron precursors. Previous analyses using dsRNA pasha knockdowns in S2 cells suggested that Pasha was not required for mirtron maturation, which was inferred to use the splicing machinery to bypass processing by the Drosha/Pasha complex (38, 44). Indeed, all three of the mirtron-derived miRNAs that we had previously shown to be detectable in third-instar larvae were still expressed in the pashaKO mutant (Fig. (Fig.3C).3C). Therefore, the pasha mutant distinguishes the processing of canonical miRNAs and mirtrons.

pashaKO mutants accumulate high levels of pri-miRNAs.

Our Northern analysis showed that all of the miRNAs tested exhibited lower levels of both mature species and pre-miRNA species, a finding consistent with a failure of “cropping.” If so, we might expect miRNA precursors to accumulate as pri-miRNA species, as shown previously using knockdown experiments (7, 14, 27). To test this, we performed qRT-PCRs of poly(A)+ RNA isolated from Canton S and pashaKO homozygous larvae. For all 10 loci tested, we observed massive elevation of pri-miRNA species in the pasha mutant, usually ranging from 100- to 1,000-fold increases (Fig. (Fig.4).4). By comparison, previous analysis of Drosophila pasha using knockdown techniques, which yield only partial target suppression, showed only three- to fivefold elevation in pri-miRNA levels (7, 27). Therefore, our analysis of bona fide pasha-null mutants more fully reports its obligate function in cleaving canonical pri-miRNA hairpins.

FIG. 4.
PashaKO larvae strongly accumulate pri-miRNA species. (A) qRT-PCR) was used to assess pri-miRNA levels in Canton S and homozygous pashaKO larvae, as normalized to rp49. Three independent RNA samples were reverse transcribed for each genotype, and six ...

A consequence of the failure to cleave primary miRNA transcripts might be their accumulation as stable full-length transcripts. Indeed, we observed that a single transcript of at least 8 kb hybridized to a pri-mir-1 probe in pashaKO, but not control Canton S larvae (Fig. (Fig.4B).4B). Curiously, this is much longer than any of the putative lengths of pri-mir-1 inferred previously using different strategies. RACE (rapid amplification of cDNA ends) experiments suggested mir-1 to produce alternative primary transcripts of 548 and 992 nt (47), while tiling microarrays detected a continuous region of transcribed mir-1 sequence extending for ~2.9 kb (2). Thus, the pasha mutant might represent a favorable background to delineate primary miRNA transcripts.

pashaKO distinguishes the activity of canonical miRNAs and mirtrons.

We have shown that the maturation of canonical miRNAs, but not mirtrons, is disrupted in pashaKO animals. We decided to perform a functional test of this distinction using a genetic assay of small RNA function. In the standard sensor assay, the Gal4-UAS system is used to ectopically express a canonical miRNA or mirtron as a fusion transcript with a DsRed-encoding marker. In this background, one introduces a ubiquitously expressed green fluorescent protein (GFP) transgene linked to binding sites for the miRNA or mirtron. Successful target repression is observed as a reduction in GFP activity specifically in DsRed-positive cells (26, 38, 48).

We adapted the standard sensor assay for use with the MARCM system (30), in which UAS transgenes can be activated specifically in homozygous somatic clones of mutations of interest. Since both dcr-1 and pasha are located on the right arm of chromosome III, we used FRT82B for somatic recombination and chose the canonical miRNA miR-7 and the mirtron miR-1004 to test against cognate sensors (38, 48). We first analyzed control clones of FRT82B. As expected, control clones expressing miR-7 repressed tub-GFP-miR-7 (Fig. (Fig.5A),5A), while miR-1004-expressing cells repressed tub-GFP-mir-1004 (Fig. (Fig.5D).5D). On the other hand, dcr-1 clones failed to exhibit either canonical miRNA-mediated or mirtron-mediated repression (Fig. 5B and E) (33). In contrast, pashaKO clones were blocked for canonical miRNA-mediated repression (Fig. (Fig.5C)5C) but showed efficient mirtron-mediated repression (Fig. (Fig.5F).5F). We observed the same trends in both eye and wing imaginal discs (Fig. (Fig.55 and data not shown), demonstrating these properties to be spatially general. These data provide convincing genetic evidence that a subclass of miRNA-family regulators remains functional in the absence of an essential component of the canonical miRNA biogenesis pathway.

FIG. 5.
pasha distinguishes the function of canonical miRNAs and mirtrons. Shown are eye imaginal discs stained for DNA and a GFP sensor, with DsRed fluorescence marking active small RNA transgenes; the right panels depict the merged GFP/DsRed channels. (A to ...

Phenotypic analysis of pasha mutants.

Homozygous mutants in dcr-1 and ago1 are lethal during embryogenesis (24, 33), while null loqs mutants survive to late pupal stages or on occasion even to adulthood (34). The differences in lethal phase appear to be a consequence of their differential requirements for miRNA biogenesis and function: Dcr-1 and AGO1 are core components, while Loqs appears to be an auxiliary factor. Homozygous pashaKO mutants survive to an intermediate stage, dying mostly as third-instar larvae. Since we showed that Pasha is a core biogenesis factor likely required for the maturation of all canonical miRNAs, survival of pashaKO animals past embryogenesis might reflect its maternal contribution. Inherited protein and RNA stores can allow mutants in essential Drosophila genes to survive to surprisingly late stages in development.

Other miRNA factors are also deposited maternally. For example, loss of both maternal and zygotic dcr-1 or ago1 causes developmental arrest at earlier times and with greater pattern abnormalities, compared to the zygotic mutants (19, 24, 36). In addition, their germ line depletion results in substantial defects during oogenesis that compromise egg production (19, 23, 53). Similarly, we observed that removal of pasha from the female germ line, using the FLP/FRT-ovoD technique, results in their inability to lay eggs (data not shown). Therefore, the later lethal phase of pasha zygotic mutants, relative to dcr-1 or ago1, is likely due to either its greater maternal deposition and/or greater stability.

To assess the effects of pasha and dcr-1 loss on adult development, we analyzed somatic FLP clones. In most respects, these revealed similar phenotypes, such as wing blistering, external loss of notum bristle sensory organs, and small, rough eyes (Fig. (Fig.66 and data not shown) (33). These rather general phenotypes resemble those caused by defects in cell viability. However, expression of the antiapoptotic baculovirus protein p35, which blocks cell death (20), did not rescue sensory organ formation in either dcr-1 or pasha clones (data not shown). This indicates that these clonal phenotypes were not simply due to cell death caused by miRNA depletion in mutant cells.

FIG. 6.
Phenotypes of dcr-1 and pashaKO adult mutant clones. Compared to wild-type, clonal loss of either Dcr-1 or Pasha results in small rough eyes (A to C) and loss of external mechanosensory bristle structures (D to F). Eye clones were made using the EGUF ...

One difference between Dcr-1 and Pasha was that pashaKO mutant eyes were smaller than those of dcr-1 (Fig. 6B and C). The apparently stronger eye phenotype of pasha was possibly consistent with its more apical position in the canonical miRNA biogenesis pathway, which might lead to a more rapid loss of miRNA production in somatic clones. An alternative interpretation is that there exist mirtrons that antagonize growth, and that these remain active in pasha but not dcr-1 eyes. However, this scenario is not obviously supported by current computational predictions of mirtron targets (e.g., http://www.targetscan.org/fly_12/).

Imaginal disc growth defects in pasha mutants.

Imaginal disc clones of dcr-1 are smaller than their wild-type twin-spot clones, implying that cells lacking miRNAs proliferate less effectively than the wild type (11). We similarly observed that random pashaKO clones generated with hs-FLP were smaller than their wild-type twin-spots (Fig. 7A to C and see Fig. S3 in the supplemental material). These imaginal disc phenotypes were reflected in the adult by the fact that both dcr-1 and pashaKO eye clones, which can be visually marked by absence or presence of eye pigmentation, contribute far less to the adult eye than their respective twin-spots (data not shown). Therefore, cells lacking Pasha or Dcr-1 exhibit similar defects in imaginal disc growth, perhaps reflecting their common loss of canonical miRNAs.

FIG. 7.
Distinct growth defects observed in mutant clones compared to homozygous pashaKO animals. (A to C) Small retinal clones in third instar eye imaginal discs stained with β-galactosidase antibody. Homozygous mutant cells have no β-galactosidase ...

Elegant techniques permit the generation of nearly completely mutant Drosophila eye tissue by flipping against a Minute mutation (37) or a GMR-hid transgene (49). Under these circumstances, twin-spots are cell lethal, while heterozygous cells are at a competitive growth disadvantage and usually contribute little to a mosaic tissue (Fig. (Fig.7D′).7D′). Such strategies make possible the generation of eyes that are nearly entirely mutant for dcr-1 (Fig. (Fig.6B)6B) (33), suggesting that cells are able to proliferate effectively in the absence of miRNAs provided that cell competition is compromised. Even under these circumstances, however, we observed that dcr-1 clones competed relatively poorly. In the third-instar eye imaginal disc, control wild-type clones eliminated almost all of the Minute heterozygous cells (Fig. (Fig.7D′),7D′), whereas discs bearing dcr-1 clones contained substantial regions of GFP+, Minute cells (Fig. (Fig.7E′).7E′). Pasha mutant cells were similarly unable to eliminate Minute heterozygous cells by the third instar (Fig. (Fig.7F7F′).

The ability of pashaKO mutants to survive to late larval stages presented a unique opportunity to assess imaginal disc development in homozygous animals. Interestingly, pasha mutant larvae essentially lack imaginal discs, with only rudiments remaining of any discs (Fig. 7G and H). The optic lobes of the brain were also strongly reduced, although the brain stem developed fairly normally. The stronger growth defect exhibited by homozygous pashaKO mutants suggests that the clonal loss of Pasha or Dcr-1 reflects only a partial loss-of-function phenotype with respect to imaginal disc growth.

DISCUSSION

Molecular and genetic analysis of a null allele of Drosophila pasha.

In the present study we described the first loss-of-function analysis of Drosophila pasha in the animal, using a deletion allele that removes this locus. We provided evidence for a nearly complete block in the production of canonical miRNAs in this mutant, similar to evidence generated for embryonic stem cells with the pasha ortholog, DGCR8, deleted (52). We found that the effects of pasha deletion on miRNA cropping are more severe than was previously observed using knockdown strategies (7, 27), validating the status of Pasha as an essential component of the canonical miRNA biogenesis pathway.

Importantly, our data provide stringent evidence for the separation of nuclear miRNA sorting pathways in Drosophila. Although Pasha is essential for processing of canonical primary miRNA transcripts, it is dispensable for the processing of mirtrons. Indeed, we showed that mirtrons were capable of potent target repression in pasha-mutant cells (Fig. (Fig.5).5). The contribution of mirtrons to the miRNA-mediated regulatory network is undoubtedly smaller than that of canonical miRNAs, owing to their generally modest expression levels (1, 44). Nevertheless, in light of the supposition that DGCR8 mutant cells are specifically lacking miRNA pathway activity (52), it is important to recognize that Pasha/DGCR8-mutant cells retain this subclass of miRNA regulators.

In theory, canonical miRNAs might be functionally reprogrammed into mirtron backbones, realizing that their 3′ ends would need to be modified into splice sites. This is plausible given that miRNA 3′ ends may be relatively subtly required for major miRNA targeting activities. Despite known roles for miRNA 3′ ends in compensatory pairing (41), all point mutants of endogenous miRNAs isolated in nematodes (lin-4, let-7, and lsy-6) (29, 42, 46) and flies (mir-278) (35, 50) invariably affect the seed region. If successful, such a scheme might enable the genetic rescue of Pasha/DGCR8-mutant phenotypes by single mirtronic-miRNA transgenes, akin to the rescue of maternal-zygotic Dicer mutants in zebrafish by injecting individual miRNA duplexes (12). It might even prove to be the case that mirtrons are especially active in Pasha/DGCR8 or Drosha mutant cells, given that Dicer would be relieved of its normal role in processing canonical pre-miRNAs in such genetic conditions.

The general fates of pri-miRNA transcripts that escape Drosha processing are incompletely understood at present. In at least some cases, stable transcripts representing full-length pri-miRNA species have been detected (5, 52). We observed the strong accumulation of many pri-miRNA fragments using qPCR analysis and also detected the stable accumulation of an ~8-kb pri-mir-1 transcript that far exceeded its previously inferred size(s) (which ranged from 0.5 to 1 kb to 3 kb [2, 47]). At the same time, it is relevant to bear in mind that heterogeneous transcripts in the process of degradation (15) are still substrates for qPCR; thus, the accumulation of pri-miRNA species as detected by qPCR need not necessarily be accompanied by single band on a Northern blot. Some invertebrate and vertebrate pri-miRNA transcripts have been suggested to be 50 to 100 kb in length (43, 50), and it would be perhaps remarkable if such long transcripts were completely immune to degradation by one or more RNases. It will therefore be interesting to examine the fates of pri-miRNA transcripts more systematically in pashaKO.

Interpreting the consequences of conditional Dicer or pasha/DGCR8 loss.

It is popularly presumed that the clonal loss of a core miRNA biogenesis component can be used to assess the consequences of removing most, if not all, miRNAs from a given developmental setting. Since mutants in core components of the miRNA biogenesis pathway are lethal in all animals, conditional loss is the only way to examine the effects of miRNA pathway loss-of-function mutations on adult tissues. The activity of residual protein and RNA/miRNA products in these conditions has not often been critically assessed. Notably, miRNAs are highly abundant species and directed tests suggested several miRNAs to be very stable and removed only by dilution in dividing cells (28). The presence of small amounts of mature miRNAs in late third-instar pashaKO larvae (Fig. (Fig.3),3), 5 days after their birth, attests to the stability of maternal Pasha and/or mature miRNAs. Double-stranded RNA-mediated knockdown studies carry similar, if not greater caveats, in light of their inherent capability for partial target suppression.

In Drosophila, an allele of dcr-1 was originally isolated in a genetic screen that assayed eye pigment levels in “whole-eye” mutant animals. The very method of its isolation meant that a substantial amount of eye tissue had to be isolated. We similarly observed that although “whole-eye” pashaKO mutant adults were substantially reduced in size, homozygous mutant adult tissue was nonetheless recovered. Such observations seemingly suggest that miRNAs are dispensable for imaginal disc growth. This seems unlikely to be the case, since it was reported many years ago that larvae deleted for the bantam miRNA lack imaginal discs (3, 21), similar to what we observed in pasha mutant larvae (Fig. (Fig.7H).7H). Thus, the strict genetic requirement for bantam does not appear to be revealed through clonal analysis of dcr-1 or pasha.

While it is conceivable that the severe pashaKO growth defects are due to a greater reduction of global miRNA activity in imaginal discs relative to clonal experiments, we cannot exclude a nonautonomous role for miRNAs in promoting imaginal disc growth. Nevertheless, our observations serve an important reminder of the necessity for caution in interpreting the consequences of conditional Dicer or Pasha/DGCR8 loss. In particular, many studies of Dicer conditional ablation concluded that many specific developmental events do not require miRNAs (6, 18). Instead, residual miRNAs may suffice to drive substantial aspects of development in early clones. Perhaps only later in the age of these clones do miRNA levels fall below a threshold that reveals a phenotype, often during differentiation or survival of mutant cells. The extant catalog of conditional Dicer phenotypes is consistent with this interpretation.

Supplementary Material

[Supplemental material]

Acknowledgments

We thank Richard Carthew, Stephen Cohen, Hugo Bellen, Qinghua Liu, and the Bloomington Stock Center for providing Drosophila stocks used in this study. Herbert Jäckle and Reinhard Lührmann provided logistical support to generate the knockout. Dae-Hwan Kim generated some of the recombinant strains.

This study was supported by the Howard Hughes Medical Institute (T.T.) and grants from the V Foundation for Cancer Research, the Sidney Kimmel Cancer Foundation, the Alfred Bressler Scholars Fund, and the U.S. National Institutes of Health (R01-GM083300) to E.C.L.

Footnotes

[down-pointing small open triangle]Published ahead of print on 1 December 2008.

Supplemental material for this article may be found at http://mcb.asm.org/.

REFERENCES

1. Berezikov, E., W.-J. Chung, J. Willis, E. Cuppen, and E. C. Lai. 2007. Mammalian mirtron genes. Mol. Cell 28328-336. [PMC free article] [PubMed]
2. Biemar, F., R. Zinzen, M. Ronshaugen, V. Sementchenko, J. R. Manak, and M. S. Levine. 2005. Spatial regulation of microRNA gene expression in the Drosophila embryo. Proc. Natl. Acad. Sci. USA 10215907-15911. [PubMed]
3. Brennecke, J., D. R. Hipfner, A. Stark, R. B. Russell, and S. M. Cohen. 2003. bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the proapoptotic gene hid in Drosophila. Cell 11325-36. [PubMed]
4. Carmell, M. A., and G. J. Hannon. 2004. RNase III enzymes and the initiation of gene silencing. Nat. Struct. Mol. Biol. 11214-218. [PubMed]
5. Chong, M. M., J. P. Rasmussen, A. Y. Rudensky, and D. R. Littman. 2008. The RNAseIII enzyme Drosha is critical in T cells for preventing lethal inflammatory disease. J. Exp. Med. 2052005-2017. [PMC free article] [PubMed]
6. Cobb, B. S., T. B. Nesterova, E. Thompson, A. Hertweck, E. O'Connor, J. Godwin, C. B. Wilson, N. Brockdorff, A. G. Fisher, S. T. Smale, and M. Merkenschlager. 2005. T-cell lineage choice and differentiation in the absence of the RNase III enzyme Dicer. J. Exp. Med. 2011367-1373. [PMC free article] [PubMed]
7. Denli, A. M., B. B. Tops, R. H. Plasterk, R. F. Ketting, and G. J. Hannon. 2004. Processing of primary microRNAs by the microprocessor complex. Nature 432231-235. [PubMed]
8. Farazi, T. A., S. A. Juranek, and T. Tuschl. 2008. The growing catalog of small RNAs and their association with distinct Argonaute/Piwi family members. Development 1351201-1214. [PubMed]
9. Flynt, A. S., and E. C. Lai. 2008. Biological principles of microRNA-mediated regulation: shared themes amid diversity. Nat. Rev. Genet. 9831-842. [PMC free article] [PubMed]
10. Forstemann, K., Y. Tomari, T. Du, V. V. Vagin, A. M. Denli, D. P. Bratu, C. Klattenhoff, W. E. Theurkauf, and P. D. Zamore. 2005. Normal microRNA maturation and germ-line stem cell maintenance requires Loquacious, a double-stranded RNA-binding domain protein. PLoS Biol. 3e236. [PMC free article] [PubMed]
11. Friggi-Grelin, F., L. Lavenant-Staccini, and P. Therond. 2008. Control of antagonistic components of the hedgehog signaling pathway by microRNAs in Drosophila. Genetics 179429-439. [PubMed]
12. Giraldez, A. J., R. M. Cinalli, M. E. Glasner, A. J. Enright, J. M. Thomson, S. Baskerville, S. M. Hammond, D. P. Bartel, and A. F. Schier. 2005. MicroRNAs regulate brain morphogenesis in zebrafish. Science 308833-838. [PubMed]
13. Glazov, E. A., P. A. Cottee, W. C. Barris, R. J. Moore, B. P. Dalrymple, and M. L. Tizard. 2008. A microRNA catalog of the developing chicken embryo identified by a deep sequencing approach. Genome Res. 18957-964. [PubMed]
14. Gregory, R. I., K. P. Yan, G. Amuthan, T. Chendrimada, B. Doratotaj, N. Cooch, and R. Shiekhattar. 2004. The Microprocessor complex mediates the genesis of microRNAs. Nature 432235-240. [PubMed]
15. Hammond, S. M., E. Bernstein, D. Beach, and G. J. Hannon. 2000. An RNA-directed nuclease mediates posttranscriptional gene silencing in Drosophila cells. Nature 404293-296. [PubMed]
16. Han, J., Y. Lee, K. H. Yeom, Y. K. Kim, H. Jin, and V. N. Kim. 2004. The Drosha-DGCR8 complex in primary microRNA processing. Genes Dev. 183016-3027. [PubMed]
17. Han, J., Y. Lee, K. H. Yeom, J. W. Nam, I. Heo, J. K. Rhee, S. Y. Sohn, Y. Cho, B. T. Zhang, and V. N. Kim. 2006. Molecular basis for the recognition of primary microRNAs by the Drosha-DGCR8 complex. Cell 125887-901. [PubMed]
18. Harfe, B. D., M. T. McManus, J. H. Mansfield, E. Hornstein, and C. J. Tabin. 2005. The RNase III enzyme Dicer is required for morphogenesis but not patterning of the vertebrate limb. Proc. Natl. Acad. Sci. USA 10210898-10903. [PubMed]
19. Hatfield, S. D., H. R. Shcherbata, K. A. Fischer, K. Nakahara, R. W. Carthew, and H. Ruohola-Baker. 2005. Stem cell division is regulated by the microRNA pathway. Nature 435974-978. [PubMed]
20. Hay, B. A., T. Wolff, and G. M. Rubin. 1994. Expression of baculovirus P35 prevents cell death in Drosophila. Development 1202121-2129. [PubMed]
21. Hipfner, D. R., K. Weigmann, and S. M. Cohen. 2002. The bantam gene regulates Drosophila growth. Genetics 1611527-1537. [PubMed]
22. Jiang, F., X. Ye, X. Liu, L. Fincher, D. McKearin, and Q. Liu. 2005. Dicer-1 and R3D1-L catalyze microRNA maturation in Drosophila. Genes Dev. 191674-1679. [PubMed]
23. Jin, Z., and T. Xie. 2007. Dcr-1 maintains Drosophila ovarian stem cells. Curr. Biol. 17539-544. [PubMed]
24. Kataoka, Y., M. Takeichi, and T. Uemura. 2001. Developmental roles and molecular characterization of a Drosophila homologue of Arabidopsis Argonaute1, the founder of a novel gene superfamily. Genes Cells 6313-325. [PubMed]
25. Lai, E. C., and G. M. Rubin. 2001. neuralized functions cell-autonomously to regulate a subset of Notch-dependent processes during adult Drosophila development. Dev. Biol. 231217-233. [PubMed]
26. Lai, E. C., B. Tam, and G. M. Rubin. 2005. Pervasive regulation of Drosophila Notch target genes by GY-box-, Brd-box-, and K-box-class microRNAs. Genes Dev. 191067-1080. [PubMed]
27. Landthaler, M., A. Yalcin, and T. Tuschl. 2004. The human DiGeorge syndrome critical region gene 8 and its Drosophila melanogaster homolog are required for miRNA biogenesis. Curr. Biol. 142162-2167. [PubMed]
28. Lau, N. C. 2004. Molecular characterization of animal microRNAs: sequence, expression, and stability. Ph.D. thesis. Massachusetts Institute of Technology, Cambridge.
29. Lee, R. C., R. L. Feinbaum, and V. Ambros. 1993. The Caenorhabditis elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 75843-854. [PubMed]
30. Lee, T., and L. Luo. 1999. Mosaic analysis with a repressible cell marker for studies of gene function in neuronal morphogenesis. Neuron 22451-461. [PubMed]
31. Lee, Y., C. Ahn, J. Han, H. Choi, J. Kim, J. Yim, J. Lee, P. Provost, O. Radmark, S. Kim, and V. N. Kim. 2003. The nuclear RNase III Drosha initiates microRNA processing. Nature 425415-419. [PubMed]
32. Lee, Y., K. Jeon, J. T. Lee, S. Kim, and V. N. Kim. 2002. MicroRNA maturation: stepwise processing and subcellular localization. EMBO J. 214663-4670. [PubMed]
33. Lee, Y. S., K. Nakahara, J. W. Pham, K. Kim, Z. He, E. J. Sontheimer, and R. W. Carthew. 2004. Distinct roles for Drosophila Dicer-1 and Dicer-2 in the siRNA/miRNA silencing pathways. Cell 11769-81. [PubMed]
34. Liu, X., J. K. Park, F. Jiang, Y. Liu, D. McKearin, and Q. Liu. 2007. Dicer-1, but not Loquacious, is critical for assembly of miRNA-induced silencing complexes. RNA 132324-2329. [PubMed]
35. Nairz, K., C. Rottig, F. Rintelen, E. Zdobnov, M. Moser, and E. Hafen. 2006. Overgrowth caused by misexpression of a microRNA with dispensable wild-type function. Dev. Biol. 291314-324. [PubMed]
36. Nakahara, K., K. Kim, C. Sciulli, S. R. Dowd, J. S. Minden, and R. W. Carthew. 2005. Targets of microRNA regulation in the Drosophila oocyte proteome. Proc. Natl. Acad. Sci. USA 10212023-12028. [PubMed]
37. Newsome, T. P., B. Asling, and B. J. Dickson. 2000. Analysis of Drosophila photoreceptor axon guidance in eye-specific mosiacs. Development 127851-860. [PubMed]
38. Okamura, K., J. W. Hagen, H. Duan, D. M. Tyler, and E. C. Lai. 2007. The mirtron pathway generates microRNA-class regulatory RNAs in Drosophila. Cell 13089-100. [PMC free article] [PubMed]
39. Okamura, K., A. Ishizuka, H. Siomi, and M. C. Siomi. 2004. Distinct roles for Argonaute proteins in small RNA-directed RNA cleavage pathways. Genes Dev. 181655-1666. [PubMed]
40. Okamura, K., M. D. Phillips, D. M. Tyler, H. Duan, Y. T. Chou, and E. C. Lai. 2008. The regulatory activity of microRNA* species has substantial influence on microRNA and 3′ UTR evolution. Nat. Struct. Mol. Biol. 15354-363. [PMC free article] [PubMed]
41. Rajewsky, N. 2006. microRNA target predictions in animals. Nat. Genet. 38(Suppl. 1)S8-S13. [PubMed]
42. Reinhart, B. J., F. Slack, M. Basson, A. Pasquinelli, J. Bettinger, A. Rougvie, H. R. Horvitz, and G. Ruvkun. 2000. The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans. Nature 403901-906. [PubMed]
43. Rodriguez, A., S. Griffiths-Jones, J. L. Ashurst, and A. Bradley. 2004. Identification of mammalian microRNA host genes and transcription units. Genome Res. 141902-1910. [PubMed]
44. Ruby, J. G., C. H. Jan, and D. P. Bartel. 2007. Intronic microRNA precursors that bypass Drosha processing. Nature 44883-86. [PMC free article] [PubMed]
45. Saito, K., A. Ishizuka, H. Siomi, and M. C. Siomi. 2005. Processing of pre-microRNAs by the Dicer-1-Loquacious complex in Drosophila cells. PLoS Biol. 3e235. [PMC free article] [PubMed]
46. Sarin, S., M. M. O'Meara, E. B. Flowers, C. Antonio, R. J. Poole, D. Didiano, R. J. Johnston, Jr., S. Chang, S. Narula, and O. Hobert. 2007. Genetic screens for Caenorhabditis elegans mutants defective in left/right asymmetric neuronal fate specification. Genetics 1762109-2130. [PubMed]
47. Sokol, N. S., and V. Ambros. 2005. Mesodermally expressed Drosophila microRNA-1 is regulated by Twist and is required in muscles during larval growth. Genes Dev. 192343-2354. [PubMed]
48. Stark, A., J. Brennecke, R. B. Russell, and S. M. Cohen. 2003. Identification of Drosophila microRNA targets. PLoS Biol. 1E60. [PMC free article] [PubMed]
49. Stowers, R. S., and T. L. Schwarz. 1999. A genetic method for generating Drosophila eyes composed exclusively of mitotic clones of a single genotype. Genetics 1521631-1639. [PubMed]
50. Teleman, A. A., S. Maitra, and S. M. Cohen. 2006. Drosophila lacking microRNA miR-278 are defective in energy homeostasis. Genes Dev. 20417-422. [PubMed]
51. Thibault, S. T., M. A. Singer, W. Y. Miyazaki, B. Milash, N. A. Dompe, C. M. Singh, R. Buchholz, M. Demsky, R. Fawcett, H. L. Francis-Lang, L. Ryner, L. M. Cheung, A. Chong, C. Erickson, W. W. Fisher, K. Greer, S. R. Hartouni, E. Howie, L. Jakkula, D. Joo, K. Killpack, A. Laufer, J. Mazzotta, R. D. Smith, L. M. Stevens, C. Stuber, L. R. Tan, R. Ventura, A. Woo, I. Zakrajsek, L. Zhao, F. Chen, C. Swimmer, C. Kopczynski, G. Duyk, M. L. Winberg, and J. Margolis. 2004. A complementary transposon tool kit for Drosophila melanogaster using P and piggyBac. Nat. Genet. 36283-287. [PubMed]
52. Wang, Y., R. Medvid, C. Melton, R. Jaenisch, and R. Blelloch. 2007. DGCR8 is essential for microRNA biogenesis and silencing of embryonic stem cell self-renewal. Nat. Genet. 39380-385. [PMC free article] [PubMed]
53. Yang, L., D. Chen, R. Duan, L. Xia, J. Wang, A. Qurashi, P. Jin, and D. Chen. 2007. Argonaute 1 regulates the fate of germline stem cells in Drosophila. Development 1344265-4272. [PubMed]

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)