Requirement for CRIF1 in RNA interference and Dicer-2 stability

Abstract

RNA interference (RNAi) is a eukaryotic gene-silencing system. Although the biochemistry of RNAi is relatively well defined, how this pathway is regulated remains incompletely understood. To identify genes involved in regulating the RNAi pathway, we screened for genetic mutations in Drosophila that alter the efficiency of RNAi. We identified the Drosophila homolog of the mammalian CR6-interacting factor 1 (CRIF1), also known as growth arrest and DNA-damage-inducible 45-gamma interacting protein (Gadd45GIP1), as a potential new regulator of the RNAi pathway. Loss-of-function mutants of Drosophila CRIF1 (dCRIF) are deficient in RNAi-mediated target gene knock-down, in the biogenesis of small interfering RNA (siRNA) molecules, and in antiviral immunity. Moreover, we show that dCRIF may function by interacting with, and stabilizing, the RNase III enzyme Dicer-2. Our results suggest that dCRIF may play an important role in regulating the RNAi pathway.

Abbreviations

CRIF1

Cytokine response 6 (CR6)-interacting factor 1

Gadd45GIP1

growth arrest and DNA-damageinducible 45-gamma interacting protein 1

Dcr-2

RNAase III enzyme Dicer-2

Introduction

RNA interference (RNAi) is a double-stranded RNA (dsRNA)-induced gene silencing system in eukaryotes. It plays important roles in organismal development and physiology, including the defense against invading pathogens.^1-3 The RNAi response can be divided into 2 phases, in both of which the RNAase III enzyme Dicer-2 (Dcr-2) plays a critical role.^4,5 In Drosophila, the first phase, the “initiator” phase, of RNAi involves the cleaving of dsRNA into small interfering RNAs (siRNAs) by the RNAase III enzyme Dicer-2 (Dcr-2).⁴ The second phase, the “effector” phase, begins with the assembly of the RNA-Induced Silencing Complex (RISC). Dcr-2 and the Dcr-2 associated protein r2d2 play a role in loading the siRNA guide strand into this complex ^6-8. Dcr-2, r2d2, and the siRNA then interact with additional RISC proteins, particularly Argonaute 2 (AGO2). The mature RISC binds to the target mRNA by base paring of the guide strand of the siRNA with the mRNA, inducing degradation of the mRNA in a sequence-specific manner.⁹

While the RNAi response is evolutionarily conserved in eukaryotes from fission yeast to plants and humans, the scope of its activity varies from organism to organism. For example, lower eukaryotes and plants make use of RNA-dependent RNA Polymerases (RdRP) to amplify the RNAi response and also to send a signal to neighboring cells, effecting transitive and systemic silencing. Conversely, Drosophila and mammals neither have RdRPs encoded in their genomes nor have the capability to induce RNAi silencing in cells that do not themselves produce siRNAs.¹⁰ Moreover, it appears that the RNAi machinery is not involved in the heterochromatic silencing of multiple copies of transgenes in mammalian cells.¹¹ While some of this diversity can be attributed to different components of the pathway in different species, some of the diversity may be due to different mechanisms of pathway regulation.

In order to understand better how RNAi is regulated, we sought to identify additional components of the RNAi pathway in Drosophila. Previous screens based on functional criteria have identified many components of the RNAi pathway in Drosophila and C. elegans.^4,12-15 Using this approach, we have identified a novel candidate gene in Drosophila, CG7172, that is important for Drosophila RNAi. We subsequently renamed the CG7172 gene dCRIF, due to its homology to the mammalian CR6-interacting factor 1 (CRIF1), also known as growth arrest and DNA-damage-inducible proteins-interacting protein 1 (Gadd45GIP1) ¹⁶ (Fig. S1). Our biochemical studies have shown that loss of dCRIF reduces the efficiency of RNAi-mediated silencing and siRNA production, and that dCRIF likely functions to stabilize Dcr-2 protein levels.

Results and Discussion

A genetic screen for components of the RNAi pathway

To carry out a genetic screen for new RNAi components, we utilized a Drosophila line that carries the transgene p[GMR-wIR], which causes RNAi-mediated silencing of the white⁺ (w⁺) gene.¹⁷ The w⁺ gene is a dosage-sensitive determinant of the red eye pigmentation in wild-type flies. The p[GMR-wIR] transgene expresses an mRNA with inverted repeats of exon 3 of the w⁺ gene, driven by the eye-specific promoter/enhancer GMR, which is controlled by the eye-specific transcription factor Glass.¹⁷ The inverted repeats are separated by a functional intron. When the intron is spliced out, a loopless hairpin RNA is produced, which can be used efficiently as a substrate for the production of small interfering RNA (siRNA) molecules, 22 nucleotides in length, that effect post-transcriptional gene silencing.⁴

The p[GMR-wIR] transgene provides dosage-dependent silencing of the w⁺ gene. Two copies of the p[GMR-wIR] transgene lead to an almost complete loss of eye pigmentation, with eyes resembling the white eyes of w^−/− flies (Fig. 1A, panel 2), while a single copy of p[GMR-wIR] results in a less stringent eye phenotype, with a light yellow or orange eye color, indicating incomplete silencing of w⁺ (Fig. 1A, panel 3; also see.¹⁷ When dosage of the glass⁺ gene, an essential transcription factor for the GMR promoter, was reduced by half, presumably reducing p[GMR-wIR] transcription, RNAi-induced w⁺ gene silencing was significantly suppressed, leading to a higher level of red eye pigmentation (Fig. 1A, panel 4). A previous genetic screen for suppressors of GMR-wIR has identified the RNAi component Dcr-2.¹⁷ Thus, the eye color of p[GMR-wIR]-bearing flies is sensitive to the dosage and strength of the p[GMR-wIR] transgene; this allows for dosage-sensitive genetic screens to isolate modifiers of RNAi.

Figure 1.

dCRIF mutation suppresses p[GMR-wIR]-induced RNAi silencing (A) The levels of eye pigmentation of flies of the indicated genotypes are shown. Note that RNAi produced by 2 copies of the p[GMR-wIR] transgene (panel 2) caused almost complete loss of red pigmentation, and that one copy of p[GMR-wIR] (panel 3) caused much reduced red eye pigmentation when compared with the wild-type eye (WT). Also note that the RNAi effects were suppressed by halving the dose of the gl⁺ gene, restoring significant levels of red eye pigmentation (panel 4). (B) Schematic representation of the genomic region 78D to 79A. Deficiencies and gene coding regions tested are represented as horizontal boxes, approximately representing the covered genomic regions. Red indicates suppression of p[GMR-wIR]-induced loss of eye pigmentation; colorless boxes indicate no suppression. (C) Representative eye of 2-day old adult males of the indicated genotypes are shown. Note that heterozygous Df(3L)ED4978 or heterozygous dCRIF^EY03252 suppressed p[GMR-wIR]-induced RNAi, resulting in an increase in red eye pigmentation. (D) Total RNA was isolated from adult siblings with the indicated genotypes, and RT-PCR for mRNA of w and rp49 (control) were performed. Representative RT-PCR results for each genotype are shown. Note that in the presence of p[GMR-wIR], w mRNA is undetectable (lane 3), but is detectable in dCRIF^+/− flies (lane 4). Scale bar = 200 μm.

As the first step in screening, we crossed isogenic females homozygous for a p[GMR-wIR] transgene located on the X chromosome to individual males carrying autosomal deletions (deficiencies or Df) and examined the F1 males for changes in eye color. The autosomal deficiency stocks used in the screen were from the deficiency kit from the Bloomington Drosophila Stock Center, which consists of 226 stocks with unique autosomal deletions that altogether uncover over 70% of the autosomes. Siblings that did not inherit the chromosomal deficiencies were used as controls. Eyes that showed lighter pigmentation than the controls indicated enhancing of RNAi silencing, while eyes that turned darker (to orange-red) indicated that RNAi silencing was suppressed, potentially signifying the loss of a gene that participates in the RNAi pathway. Crosses were repeated to confirm the observed changes in eye color.

Out of the 226 Df lines we studied, 56 deficiencies (∼25%), constituting 46 unique regions of the genome, showed changes in eye color and warranted further testing, while the rest (∼75% of the strains tested) showed no change in the F1 eye color (Fig. S2; Table S1). Of the 56 deficiencies that altered the eye phenotype, 6 enhanced and 50 suppressed the RNAi silencing effects (Fig. S2; Table S1). We tested additional overlapping deficiencies and available mutations localized to the genomic regions uncovered by the deficiencies in order to narrow down the gene(s) responsible for the modification of RNAi. In doing so, we identified a few genes known to encode components of the RNAi pathway, such as piwi, aub, Dcr-2, Ago2, and R2D2, validating the specificity of the genetic screen (Table S1), although the effects of these mutations on RNAi in this particular screen could be indirect. Importantly, we also identified a few regions that might contain novel components of RNAi (Table S1).

Identification of dCRIF as potentially a novel RNAi component

Df(3L)ED4978 was among the strongest suppressors of the 50 deficiencies that suppressed the RNAi effects, causing an increase in eye pigmentation. Further testing of overlapping deficiencies narrowed down the suppressor mutation to an area extending from cytogenetic region 78D5 to 78F4 on the left arm of chromosome 3 (Fig. S2; Table S1). Although there are as many as 50 genes located in the 78D5 to 78F4 genomic region, mutant alleles for only 10 genes were available (Fig. 1B). We tested these 10 genes, and found that disruption of only one of them, CG7172, recapitulated the suppressive effects on RNAi of Df(3L)ED4978 (Fig. 1B, C). Interestingly, a previous screening for RNAi components using Drosophila S2 cells identified CG7173, which is also located in this genomic region.¹² However, because no mutant alleles of this gene were available, we were unable to test CG7173 in our system. Since a CG7172 mutation suppresses RNAi to the same extent as Df(3L)ED4978, we decided to investigate the functions of CG7172 in RNAi.

The mutation that disrupts CG7172 is designated as EY03252, which is associated with a P-element insertion in the 5’ untranslated region of the CG7172 gene (Fig. S3A).¹⁶ Animals homozygous for dCRIF^EY03252 or in transheterozyous with Df(3L)ED4978 die at the second instar larval stage (Fig. S3B, C). We subsequently renamed CG7172 dCRIF due to its homology to mammalian CRIF (CR6-interacting factor) ¹⁸ (Fig. S1), and named the mutant allele dCRIF^EY03252. Precise excision of the P element, mediated by expression of the P element transposase (Δ2–3), reversed the lethality and the suppressive effects on RNAi. Imprecise excision, in which part of the dCRIF coding region was deleted with the loss of the P element, maintained the lethality and the suppressive effects on RNAi. These results suggest that the observed genetic phenotypes were due to disruption of dCRIF function by the P element insertion.

We confirmed that the increased red eye pigmentation in the presence of dCRIF^EY03252 was due to an increase in the levels of w⁺ mRNA. Using RT-PCR to amplify w⁺ mRNA, we found that the p[GMR-wIR] transgene effectively reduced w⁺ mRNA to undetectable levels (Fig. 1D, lane 3). In a dCRIF^+/− background, however, the effect of p[GMR-wIR] was attenuated, such that w⁺ mRNA became detectable (Fig. 1D, lane 4), consistent with the increase in the red eye pigmentation (Fig. 1C).

To confirm that dCRIF^+/− indeed suppressed RNAi rather than transcription from the eye-specific GMR promoter, we tested dCRIF^+/− in the background of the p[GMR>hid] transgene, which expresses the pro-apoptosis gene hid under the same GMR promoter, causing loss of ommatidia in the eye.¹⁹ We found that, although gl^+/− effectively suppressed the cell death effects caused by hid expression from p[GMR>hid], dCRIF^+/− had no effect (Fig. S4). These results suggest that dCRIF^EY03252 does not interfere with transcription from the GMR promoter, and that the increased eye pigmentation observed in p[GMR-wIR]/+; dCRIF^+/− flies is attributable to dCRIF's function in the RNAi pathway.

To investigate whether dCRIF plays a role in RNAi in general, we tested the effects of dCRIF mutation on an independent RNAi transgene in another tissue. We used Scalloped-Gal4 (Sd-Gal4) to drive the expression of a D-Fos RNAi transgene in the developing wing. Knocking down D-Fos in the fly wing with this RNAi transgene results in partial lethality and in uninflated wings in the survivors.²⁰ In a dCRIF^+/− background, however, the effects of expressing D-Fos RNAi were ameliorated, as evidenced by a partial suppression of the wing (Fig. 2A) and lethality (Fig. 2B) phenotypes. These results indicate that mutation of dCRIF suppresses RNAi not only in the eye but also in the wing.

Figure 2.

dCRIF mutation reduces levels of target mRNA and siRNA (A) An adult female of each indicated genotype is shown. Note that knocking down dFOS by sd>dFOS RNAi causes uninflated wings, which is partially suppressed by dCRIF^EY03252/+. Scale bar = 500 μm. (B) Percent surviving flies and total numbers of surviving flies (N) are shown. All of the flies were siblings, grown in the same vials. Three independent trials were done and P values determined by Student's t-Test. Error bars are standard deviations. (C) Total RNA was isolated from adult siblings of the indicated genotypes and the levels of w and rp49 (control) mRNA were measured by real-time qRT-PCR. Expression levels of w relative to rp49 are shown. Error bars indicate standard deviation from the mean. P values were determined by Student's t-Test. (D) Total RNA isolated from adult siblings of the indicated genotypes was subjected to sequential Northern blotting to detect siRNA and tRNA^val (control) using p³² labeled RNA probes specific for w and tRNA^val, respectively. Note that dCRIF^EY03252/+ flies have reduced levels of w siRNA when compared with their wild-type siblings. The histogram in the bottom shows averages of w siRNA levels, normalized to those of tRNA^val, for 3 independent Northern blots (n = 3). Error bars indicate standard deviation. * indicates P < 0.05. P values were determined by Student's t-Test. (E) S2 cells were incubated with dsRNAs targeting luciferase (luc; negative control), Dcr-2, and dCRIF, respectively, and the levels of the endogenous siRNA esi-2.1 were detected by Northern blotting, with 2S rRNA as loading control. Numbers under the lanes are esi-2.1 levels normalized to luc knockdown. (F) S2 cells were treated with dsRNAs targeting luciferase (luc; negative control), Dcr-2, and dCRIF, respectively, and then the cell lysates were incubated with p³²-labeled dsRNA substrate. The levels of the resulting 21 nt siRNA reflect Dcr-2 activity in each lysate. Numbers under the lane are normalized levels of 21 nt siRNA. The histogram to the right shows relative Dicer activities for 4 independent experiments (n = 4). Standard deviation and p values are shown. P values were determined by Student's t-Test. (G) Wild-type, Dcr-2^L811fsx, dCRIF^EY03252 heterozygous flies of the same age were infected with DCV and the viral load was quantified by qPCR amplification of viral RNA. The rp49 gene was used as a control. Viral RNA levels are normalized to those in infected wild-type control flies. N indicates the number of independent experiments performed. *** indicates P < 0.001, as determined by Student's t-Test.

In addition to these qualitative observations, we used quantitative real-time PCR to test the efficiency of knocking down a non-essential gene, Ptp61F, in wild type and dCRIF^EY mutant genetic backgrounds. Since the loss of Ptp61F does not affect animal viability, large quantities of flies carrying the RNAi construct can be obtained for mRNA collection and quantification. When a Ptp61F RNAi transgene was expressed using a ubiquitous driver, Tubulin 80-gal4 (T80-Gal4), Ptp61F was efficiently knocked down in otherwise wild-type flies (Fig. 2C). However, knock down was less efficient in their dCRIF^+/− mutant siblings (Fig. 2C). Taken together, these results indicate that dCRIF is required in general for robust RNAi gene silencing.

Role of dCRIF in siRNA biogenesis and antiviral immunity

RNAi-mediated post-transcriptional gene silencing can be divided into 2 stages, the initiation stage, i.e., siRNA production, and the effector stage, where siRNA guides the degradation of the cognate mRNA. As described above, loss of dCRIF reduced the effectiveness of the RNAi pathway. However, the stage at which this occurred was unclear. To investigate at what stage dCRIF is required for RNAi, we used Northern blotting to examine the levels of siRNA produced from the hairpin dsRNA substrate expressed from p[GMR-wIR] in wild-type vs. dCRIF^+/− flies. We found that the levels of the ∼21 nt siRNA were reduced in dCRIF^+/− flies when compared to wild type flies carrying p[GMR-wIR] (Fig. 2D), consistent with the idea that dCRIF might be important for the initiation or the siRNA production stage in the RNAi pathway.

To investigate if dCRIF is also required for biogenesis of endogenous siRNAs, we knocked down dCRIF in Drosophila S2 cells and used Northern blotting to measure the levels of the endogenous siRNA esi-2.1.²¹ We used dsRNAs against luciferase (luc) and Dcr-2, respectively, as controls. Knocking down dCRIF did not affect Dcr-2 mRNA levels, and vice versa (Fig. S6A). We found that knocking down dCRIF led to a moderate reduction in the levels of esi-2.1, whereas knocking down Dcr-2 had a more dramatic effect on esi-2.1 levels (Fig. 2E).

To further investigate the role of dCRIF in siRNA production, we used the “dicing assay” ²² to measure Dcr-2 activity in cell extracts following knocking down dCRIF. Lysates from S2 cells treated with dsRNAs as described above were incubated with radiolabeled dsRNA molecules, and the amount of siRNAs produced reflects the levels of Dcr-2 activity present in the lysate. We found that knocking down dCRIF indeed reduced Dcr-2 activity in S2 cell lysates (Fig. 2F).

To assess the physiological role of dCRIF in antiviral immunity, which requires systemic RNAi ²³, we infected dCRIF mutants and control flies with Drosophila C virus (DCV) and measured viral RNA levels by qPCR 28 hours post-infection. We found that compared with infected wild-type flies, Dcr-2 homozygous mutants and dCRIF mutant heterozygotes had a much higher viral load, with approximately 40 fold higher viral RNA levels (Fig. 2G). Taken together, these results support a role of dCRIF in siRNA biogenesis.

dCRIF interacts with Dicer-2 and contains a putative PIWI domain

We next investigated the molecular mechanism by which dCRIF functions in the RNAi pathway. We first examined the morphology of dCRIF^EY03252 mutants. The homozygous dCRIF^EY03252 mutants were not viable, arresting development and dying at the second instar larval stage, but without any discernable morphological defects (Fig. S3B, C). We then studied the subcellular localization of dCRIF. Since there were no available anti-dCRIF antibodies, we expressed a hemaglutinin (HA)-tagged dCRIF (dCRIF-HA) in salivary gland cells using the salivary gland-specific driver Sgs-GAL4. Salivary gland cells are the largest cells in Drosophila, allowing for convenient examination of the subcellular distribution of dCRIF-HA. We found that dCRIF-HA was localized mainly in peri-nuclear regions of the cytoplasm, in a punctate pattern (Fig. 3A). By co-immunostaining, we found that dCRIF-HA and Dcr-2 showed similar, punctate staining and partially co-localized in the cytoplasm (Fig. 3A).

Figure 3.

Co-localization and co-immunoprecipitation of Dcr-2 and dCRIF (A) HA-tagged dCRIF was overexpressed in larval salivary gland cells using Sgs-GAL4, a salivary gland-specific driver. Immunostaining was performed to visualize the localization of dCRIF-HA, Dcr-2, and HP1. HP1 staining was used as control to visualize the nucleus. The samples were analyzed using confocal microscopy. Note that dCRIF-HA partially co-localizes with Dcr-2 in the cytoplasm. Scale bar = 10 μm. (B) HA-tagged dCRIF and Flag-tagged Dcr-2 were co-transfected into S2 cells and Flag antibody was used to immunoprecipitate Dcr-2-Flag and associated proteins. The pull-down fraction was then analyzed by Western blotting using anti-HA to detect the presence of dCRIF-HA. Note that dCRIF-HA was co-immunoprecipitaed with Dcr-2-Flag.

To determine whether dCRIF and Dcr-2 indeed interact physically, we carried out co-immunoprecipitation experiments using cultured Drosophila S2 cells co-transfected with Dcr-2-Flag and dCRIF-HA. Consistent with their partial colocalization as detected by immunostaining, a high level of dCRIF-HA was detected in the immunoprecipitates when anti-Flag antibodies were used to pull down Dcr-2 from the cell lysates (Fig. 3B). Thus, the results suggest that dCRIF physically associates with Dcr-2 in vivo.

dCRIF encodes a novel peptide of 27 kD. Based on protein alignments, dCRIF shares a low but significant degree of homology with Argonaute proteins in the PIWI domain (Fig. S5), which has been implicated in binding to Dcr proteins,²⁴ and which is the catalytic domain used for RNA cleavage.²⁵ However, the catalytic residues are not conserved in dCRIF, thus dCRIF is unlikely to have any “slicer” activity. Taken together with the colocalization and co-immunoprecipitation results described above, we suggest that dCRIF may directly interact with Dcr-2.

dCRIF maintains Dcr-2 protein stability

Since dCRIF appeared to interact with Dcr-2, we speculated that loss or over-expression of dCRIF might affect Dcr-2 in some way. To investigate this possibility, we used immunostaining with antibodies directed against Dcr-2, to analyze mitotic dCRIF^−/− clones that were generated in dCRIF^+/− flies by means of the FLP-FRT system.²⁶ Strikingly, we found that Dcr-2 levels were greatly reduced in dCRIF^−/− cells (Fig. 4A, GFP⁻) when compared with flanking wild-type cells (Fig. 4A, GFP+), suggesting that dCRIF may be important for Dcr-2 expression or stability. We then over-expressed dCRIF in marked clonal cells (Fig. 4B, GFP+) and observed Dcr-2 levels by immunostaining. Over-expressing dCRIF resulted in similar levels, or only a slight increase, in Dcr-2 immunostaining when compared with wild-type cells (Fig. 4B), suggesting that physiological levels of dCRIF may be sufficient for optimal expression or stabilization of Dcr-2. These results suggest that although loss of dCRIF leads to Dcr-2 instability, over-expressing dCRIF may not lead to increased Dcr-2 levels.

Figure 4.

dCRIF is required for maintaining Dcr-2 protein levels (A) Mitotic dCRIF^EY03252 homozygous clones were generated in dCRIF^EY03252/+ larvae using the FLP-FRT system (see Methods). Larval salivary glands were immuno-stained with anti-Dcr-2 antibodies and analyzed using confocal microscopy. GFP expression (green) marks the wild type cells; the non-GFP expressing cells are dCRIF^−/− cells. Note that the level of Dcr-2 immunostaining is much reduced in the 2 dCRIF^−/− cells (outlined with white dotted lines) in comparison with the flanking dCRIF^+/− cells (green). A total of 12 dCRIF^−/− clones were found in salivary glands, all qualitatively similar. A representative image is shown. See Figure S7 for additional example. (B) dCRIF over-expression was induced in random cells marked with GFP (green). Salivary glands were dissected and immunostained with anti-Dcr-2 antibodies. Note that when dCRIF was overexpressed (GFP+ cells), the Dcr-2 levels were similar or slightly higher than those in flanking (GFP⁻) cells. (C) Protein extracts were prepared from second instar dCRIF^+/+ (WT), dCRIF^EY03252 heterozygous (dCRIF^+/−), or homozygous (dCRIF^−/−) larvae and were subjected to SDS-PAGE, blotting and incubation with anti-Dcr-2 antibodies. Note that dCRIF^−/− larvae contain much reduced levels of full-length Dcr-2 protein (approx. 200 k_D) compared to dCRIF^+/− siblings or wild-type controls. Numbers under the lanes represent Dcr-2 levels normalized to WT control. Scale bar = 10 μm.

To further test the importance of dCRIF in maintaining Dcr-2 protein levels, we performed total protein immunoblots using extracts from dCRIF^+/− or dCRIF^−/− larvae. We found that Dcr-2 protein levels were indeed much reduced in dCRIF^−/− larvae (Fig. 4C). Since we did not detect significant differences in Dcr-2 mRNA expression in these samples (Fig. S6), the reduced Dcr-2 protein levels in dCRIF^−/− mutants suggest that dCRIF might be required for maintaining Dcr-2 protein stability.

Concluding Remarks

In summary, by genetic means, we have identified the gene dCRIF as a player in the RNAi pathway in Drosophila. We have further investigated the effects of loss or overexpression of dCRIF on RNAi. Our results indicate that loss of dCRIF reduces siRNA biogenesis, and that dCRIF may play a role in maintaining Dcr-2 protein stability. Since Dcr-2 is also crucial for loading siRNAs into siRISC, reducing Dcr-2 will profoundly diminish RNAi efficiency. Interestingly, mammalian CRIF1 has been shown to physically interact with and negatively regulate NRF2 protein stability by proteasome-mediated degradation.²⁷ In mammals, it has been shown that trans-activation-responsive RNA binding protein (TRBP) associates with Dcr-2 and enhances its stability and RNAi efficiency.²⁸ It is not clear whether dCRIF functions similarly in regulating Dcr-2 stability. Despite the evidence presented here, we cannot rule out the possibility that dCRIF is indirectly required for Dcr-2 protein stability. Lastly, since dcr-2 null mutant flies are viable, whereas dCRIF mutant flies die as early larvae, dCRIF must have essential biological functions other than stabilizing Dcr-2. Indeed, many different functions have been reported for mammalian CRIF1, including functions as a transcription co-factor and functions in mitochondria.^29-33 We are currently investigating the functions of dCRIF and its role in Dcr-2 stability.

Methods

Fly strains and genetics

The fly strains used in this study CG7172^EY03252/TM3, hsFLP; FRT2A, FRT2A Ubi-EGFP, UAS-Dicer-2, hsp70-flp; Act5C>y⁺>Gal4 UAS-EGFP/CyO, Sbv, D, FRT2A/TM3, GAL4 driver lines (Sgs-GAL4, en-GAL4, Tub-GAL4, T80-Gal4, Sd-GAL4, and ey-GAL4), and the deficiency kit used for the genetic screen were obtained from the Bloomington Drosophila Stock Center (Bloomington, IN). The RNAi lines UAS-CG7172 RNAi and UAS-Ptp61F RNAi were from the Vienna Drosophila RNAi Center (VDRC; Vienna, Austria). UAS-dCRIF-HA was cloned and injected into germ cells to generate transgenic flies. The UAS-DFos RNAi line was a gift from the laboratory of Dirk Bohmann (University of Rochester, NY). GMR-wIR and Dcr-2^L811fsx flies were kindly provided by the laboratory of Richard Carthew (Northwestern University, IL).¹⁷ UAS-dCRIF-HA and UAS-Dcr-2-Flag were constructed by cloning the full-length genes into pUAST-HA or pUAST-Flag vector respectively and injected into fly embryos to obtain transgenic lines.

Genetic mosaic analysis in salivary glands

Genetic mosaic analysis using salivary glands was carried out as previously described.^34,35 Briefly, we crossed hsFlp; FRT^2A ubq-GFP flies to FRT^2A-dCRIF^EY03252 flies. Embryos (0 – 22 h after egg laying) were collected and heat-shocked for 1 hour at 37°C and raised at 25°C. To express dCRIF in random clones, hsp70-flp; Act5C>y⁺>Gal4 UAS-GFP flies were crossed to UAS-dCRIF-HA flies. The progeny were heat-shocked at 37°C for 2 hrs in late embryo and early larval stages. Subsequently, crawling third instar larvae were collected and dissected to obtain salivary glands. Dissected tissues were then fixed in 4% paraformaldehyde and washed in 0.3% Triton-X in PBS, followed by overnight incubation in primary antibodies at 4°C, and then by washing and incubation in secondary antibody labeled with fluorescence for visualization. The nucleic acid stains Toto3 or DAPI were used to visualize the nucleus. The samples were mounted on glass slides and analyzed using confocal microscopy.

Cell culture, transfection, and RNAi

Drosophila S2 and S2-NP cells were maintained in Schneider's medium (Invitrogen) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Invitrogen). RNAi was performed as described previously.^21,22 Briefly, ∼2 × 10⁶ S2-NP cells were seeded in 6-well plates for 24 h and then transfected with 3 μg of the appropriate dsRNA using calcium phosphate method. Two days later, cells were harvested, replated in 6-cm plates for 24 h, transfected again with 9 μg dsRNA, and harvested 3 d later.

Immunofluorescence and Antibodies

Immunostaining of salivary glands was performed as described previously.^34,35 Briefly, larval tissues were dissected, placed in phosphate buffered saline (PBS), fixed in 4% paraformaldehyde for 20 minutes, and washed briefly with 0.3% Triton-X in PBS. Tissues were then submerged in blocking buffer, consisting of 5% normal goat serum (NGS), for half an hour at room temperature, followed by overnight incubation with primary antibody. Tissues were then gently washed with PBS containing 0.1% Triton-X, and incubated overnight at 4°C with secondary fluorescent antibody. After a few gentle washes with 0.1% Triton-X/PBS, the samples were mounted on a glass slide in Mowiol medium.

Antibodies specific for Dicer-2 were from Abcam (Cambridge, MA; Cat No. AB4732; 1:250). Nucleic acid staining was done using Toto3 or DAPI. Stained tissues were photographed with a Leica confocal microscope. Images were processed and analyzed using Adobe Photoshop.

Immunoprecipitation and Western blot analysis

Co-immunoprecipitation was done as previously described.³⁶ For Western blot studies, adult flies were allowed to lay eggs on apple juice/agar plates. Larvae were collected for preparation of cell lysates. Aliquots of samples were separated on 10% SDS-PAGE gels, transferred to nitrocellulose membranes, and blocked in 5% BSA solution in TBS-T (Tris-buffered saline with 0.1% tween-20). The membranes were incubated overnight at 4°C with the designated antibodies against the target proteins. The membranes were then washed 3 times in TBS-T and incubated for another hour at room temperature with secondary antibodies conjugated to horseradish peroxidase, washed in TBS-T for 3 times, treated with SuperSignal West Pico Chemiluminescent subsbrate (Pierce) and placed on an X-ray film for different exposure times. Antibodies used for these studies were specific for Dicer-2 (Abcam, Cat No. AB4732) or gamma-tubulin (Sigma, Cat No. 5192).

Viral infection

Similarly aged flies (4–7d) of the appropriate genotype were injected with 46 nl of PBS suspension of 10 pfu of Drosophila C virus (DCV; gift from Anette Schneemann, Scripps Research Institute, La Jolla, CA) using Nanoject II (Drummond), kept at 25°C and harvested 28 hours later. Total RNA was extracted and levels of DCV genomic RNA were measured using RT-qPCR using the following primers.

DCV_F ATGTTGAATTGGTCTCGCcG

DCV_R CCGTTTCAATGTTCGACATCAAG

Northern blot

Northern blotting was performed as previously described.^21,22,37 In brief, total cellular RNA was isolated with TRIzol (Invitrogen) from S2 cells or adult flies and was subsequently enriched for low molecular weight (LMW) RNAs by using the miRNA isolation kit mirVana from Ambion (Grand Island, NY). Samples of 15 μg of RNA were separated on 15% denaturing polyacrylamide gels and transferred to Hybond-N+ membranes (Amersham Biosciences) in 1X TBE buffer. Small RNAs were UV crosslinked to the membranes, and the membranes were prehybridized in hybridization buffer for 2 h. DNA probes complementary to the appropriate strands were 5′ radiolabeled with ³²P and incubated with membranes overnight at 37°C. Membranes were washed twice in 1X SSC with 0.1% SDS at 42°C, and then exposed to Phosphorimager screens for 12–48 h. Membranes were stripped by the addition of boiling 0.1% SDS solution and incubated for 30 min. The primers used for probe synthesis were the following.

White exon 3 with T7 promoter sequence forward primer:

TAATACGACTCACTATAGGGAGAATCCGGGCGAACTTTTG

White exon 3 with SP6 promoter sequence reverse primer:

ATTTAGGTGACACTATAGAAGAGCTTCGCTGGGAGTGCC

tRNA-Val: TGGTGTTTCCGCCCGGGAA

esi-2.1 probe: GGAGCGAACTTGTTGGAGTCAA

2S rRNA probe: TACAACCCTCAACCATATGTAGTCCAAGCA

Dicing assay

Dicing assay was performed as previously described, with minor modifications.²² Briefly, cytoplasmic extracts from frozen S2-NP cells were prepared by thawing cells in a hypotonic buffer composed of 10 mM HEPES-KOH (pH 7.0), 2 mM magnesium acetate, 0.1% β-mercaptoethanol, and 1X EDTA-free protease inhibitors (Roche). Long dsRNAs were synthesized using a T7 MEGAscript in vitro transcription kit (Ambion) with α-³²P-GTP, and purified using RNAeasy column (Qiagen). For the dicing assay, aliquots of 6 μl of cell lysates containing the same amount of total protein were incubated in a final volume of 10 μl reaction mixture (20 mM HEPES-KOH (pH 7.0), 2 mM DTT, 2 mM magnesium chloride, 1 mM ATP, 25 mM creatine phosphate, 0.06 U/μl creatine kinase (Roche), 0.8 U/μl ribonuclease inhibitor (Promega), and 2000–10,000 cpm dsRNA substrate) at 25°C for 2 h. Subsequently 200 μl of proteinase K buffer (200 mM Tris-HCl (pH 7.5), 25 mM EDTA, 300 mM sodium chloride, 2% w/v SDS, and 50 μg/mL proteinase K) was added and the mixture was incubated at 65°C for 30 min, and extracted with phenol/chloroform (1:1). RNA was precipitated from the supernatant and resolved by 15% urea-PAGE.

For amplifying dsRNA template:

T7-dCRIF-F: TAATACGACTCACTATAGGGCGTTGTCACCAGACTTCCT

T7-dCRIF-R: TAATACGACTCACTATAGGGAGTTGCATAACCTGCCTGCT

Quantitative Real-time polymerase chain reaction

RNA was extracted using Invitrogen's Trizol reagent. Reverse transcription was carried out using Invitrogen's Super Script III Reverse Transcription kit, 3μg of DNase-treated RNA, and oligo-dT primers. The resulting cDNA served as template for qPCR reactions using Invitrogen's Sybr Green mastermix. The expression levels of target mRNA were normalized to RP49 as an internal control. qPCR primers used are:

Rp49 forward primer: TCCTACCAGCTTCAAGATGAC

Rp49 reverse primer: CACGTTGTGCACCAGGAACT

Ptp61F forward primer: TAACCGATCTCGAAACGCAGCAGA

Ptp61F reverse primer: ATTGAGCGAAAGTATACCGCCGGA

Dcr-2 forward primer: GCGCGGAATGTGGAGTGATGAAAT

Dcr-2 reverse primer: TGATGACAACGCTCAGGGAGCTTA

dCRIF forward primer: TAAATGCCCGTGAAGAGGAC

dCRIF reverse primer: TATCCACCTTGAACCCGAAG

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Richard Carthew (Northwestern University), Dirk Bohmann (University of Rochester, NY), Anette Schneemann (Scripps Research Institute), the Vienna Drosophila RNAi Center (VDRC; Vienna, Austria), and the Bloomington Drosophila Stock Center (Bloomington, IN) for various Drosophila strains and reagents. We thank Kimberly Larson and Guowei Wu for technical assistance.

Funding

This study was supported, in part, by grants from the National Institutes of Health (W.X.L. and R.Z.), the University of California Cancer Research Coordinating Committee (CRCC) (W.X.L.), the American Heart Association (R.Z.) and the start-up fund from the Sanford-Burnham Medical Research Institute (R.Z.).

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.