Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Oct 10.
Published in final edited form as: Mol Cell. 2013 Sep 19;52(1):113–123. doi: 10.1016/j.molcel.2013.08.023

Functionally diverse microRNA effector complexes are regulated by extracellular signaling

Pei-Hsuan Wu 1, Mamiko Isaji 1, Richard W Carthew 1,*
PMCID: PMC3797187  NIHMSID: NIHMS518048  PMID: 24055343

Summary

Since microRNAs (miRNAs) influence the expression of many genes in cells, understanding how the miRNA pathway is regulated is an important area of investigation. We found that the Drosophila miRNA-induced silencing complex (miRISC) exists in multiple forms. A constitutive form, called G-miRISC, is comprised of Ago1, miRNA, and GW182. Two distinct miRISC complexes that lack GW182 are regulated by mitogenic signaling. Exposure of cells to serum, lipids, or the tumor promoter PMA suppressed formation of these complexes. P-miRISC is comprised of Ago1, miRNA, and Loqs-PB, and it associates with mRNAs assembled into polysomes. The other regulated Ago1 complex associates with membranous organelles, and is likely an intermediate in miRISC recycling. The formation of these novel complexes is correlated with a five- to ten-fold stronger repression of target gene expression inside cells. Taken together, these results indicate that mitogenic signaling regulates the miRNA effector machinery to attenuate its repressive activities.

Introduction

Sudden environmental changes can lead cells to responses that either re-establish homeostasis or adapt cells to an altered state. MicroRNAs (miRNAs) have been documented to frequently mediate these responses by altering gene expression programs (Leung and Sharp, 2010). As such, environmental change often alters the status quo of the miRNA pathway. This effect can occur at one of several steps in the pathway: biogenesis of miRNAs, expression of miRNA target messages, or activity of the miRNA Induced Silencing Complex (miRISC) (Leung and Sharp, 2010). For example, changes in nutrient condition have been found to affect miRISC activity against certain mammalian genes. Under conditions of amino acid starvation, CAT-1 mRNA is relieved from miR-122-mediated repression (Bhattacharyya et al., 2006). This de-repression requires binding of HuR protein to an AU-rich element (ARE) within the CAT-1 mRNA 3′ UTR. AREs are signals present in the 3′ UTRs of short-lived mammalian mRNA transcripts, and through interactions with RNA-binding proteins, they usually promote transcript turnover. Studies of TNFα mRNA found that an ARE in the message 3′ UTR regulates the effect of miRISC on TNFα expression (Vasudevan and Steitz, 2007; Vasudevan et al., 2007). This occurs specifically under conditions of serum starvation, and switches miRISC from acting as a repressor to an activator. The switch requires the ARE-binding protein FXR1.

These studies uncovered regulation of miRISC activity that was incumbent upon RNA binding proteins that presumably do not interact with most mRNAs in the cell. However, they did not address whether miRISC activity is more generally regulated by changes in nutritional status. Here, we explore the issue using Drosophila S2 cells. In this hematocyte cell line, a mature miRNA associates directly with the Argonaute protein Ago1 to form miRISC (Carthew and Sontheimer, 2009). GW182 protein is another subunit of miRISC, and acts downstream of Ago1 to repress mRNA transcripts complementary to the loaded miRNA (Behm-Ansmant et al., 2006; Eulalio et al., 2008). Repression is exerted by transcript destabilization, involving deadenylation and decapping, and by inhibition of protein translation (Behm-Ansmant et al., 2006; Eulalio et al., 2008). Studies have found that S2 cell miRISC inhibits translation initiation in a manner dependent upon GW182 (Zdanowicz et al., 2009; Zekri et al., 2009). miRISC can inhibit translation initiation in a GW182-independent manner as well (Fukaya and Tomari, 2012). Studies in other model systems have confirmed that translation initiation is a step commonly targeted by miRISC (Ding and Grosshans, 2009; Humphreys et al., 2005; Pillai et al., 2005). However, miRISC has been found to repress translation elongation in other studies (Maroney et al., 2006; Nottrott et al., 2006; Petersen et al., 2006; Seggerson et al., 2002). The reasons for these conflicting results are not entirely clear. It might be that translation is rate-limiting at different steps depending on the message and/or cells, and miRISC regulates the rate limiting step. Alternatively, it has been suggested that different Ago proteins might mediate inhibition at different steps (Iwasaki et al., 2009).

In the present study, we transiently altered the nutritional environment of S2 cells by serum withdrawal. We find this treatment has little effect on miRISC complexes that contain GW182. However, it stimulates rapid induction of two novel miRISC complexes, neither of which contains GW182. One complex associates with mRNAs on polysomes and exerts repression of elongation. The other complex associates with membranous cellular structures and likely is an intermediate in miRISC-target recycling. Overall, the formation of these novel complexes is correlated with a five- to ten-fold stronger repression of target gene expression. Lipid signaling transduced via PKC blocks formation of the inducible miRISC complexes while insulin signaling inhibits miRISC recycling as evident by the miRISC intermediate.

Results

Serum withdrawal enhances miRNA-mediated gene silencing

Withdrawal of serum from cell culture leads to complex changes in cell physiology, notably the induction of cell cycle arrest and apoptosis. As cells undergo serum depletion, they display a rapid response in their transcriptome pattern, and this early expression program evolves into a late program as serum depletion continues (Liu et al., 2007). We switched S2 cells to serum-free medium, and observed them to exhibit reduced viability within one day of incubation (Fig. S1A). However, serum depletion that lasted less than six hours elicited no visible defects in cell viability, nor did it elicit the production of cytoplasmic stress granules (Fig. S1B–D). These characteristics made brief serum withdrawal (< 6 hr) an ideal model to study its effects on the miRNA pathway.

We first investigated the effect of serum withdrawal on miRNA-mediated regulation of protein synthesis. Several criteria influenced our choice of target and detection method (Fig. 1A). We wanted to avoid the compound effects of ARE-binding proteins or other UTR-binding proteins, and so we expressed a reporter transcript with a minimal 3′ UTR containing artificial miRNA (miR-CXCR4) binding sites. Detection of reporter protein synthesis during the specific period of serum depletion was also a goal, and we therefore immunoprecipitated the GFP reporter protein after labeling cells with 35S-methionine during serum depletion. To measure direct miRNA repression, we co-transfected a control with the same 3′UTR except that it lacked the miR-CXCR4 binding sites. The control encoded a GFP protein that was distinguishable by SDS-PAGE from the reporter GFP because of a difference in their molecular weights (Fig. 1A). Importantly, we wanted to detect nutrient-induced changes from steady-state repression, and so we first expressed control, reporter and miR-CXCR4 for 36 hours before withdrawing serum. As anticipated, reporter protein synthesis was modestly diminished after serum withdrawal regardless of the presence or absence of miR-CXCR4 (Fig. 1B–D), since lack of serum is known to slow down overall protein synthesis (Soeiro and Amos, 1966; Thomas et al., 1980).

Figure 1. Serum withdrawal enhances miRISC silencing.

Figure 1

Open in a new tab

(A) Experimental scheme to measure CXCR4 miRNA repression activity specifically after serum withdrawal. Control and reporter plasmids were co-transfected with or without the miR-CXCR4 expression plasmid. Control and reporter protein synthesis were measured by radiolabeled methionine incorporation during the final two hours of serum withdrawal. (B) Radiolabeled reporter and control proteins from cells expressing a low level of miR-CXCR4, as indicated. 100 ng miRNA expression plasmid was transfected per sample. Cells underwent serum depletion as indicated. (C–D) Quantitation of radiolabeled reporter and control proteins from cells expressing a high level of miR-CXCR4, as indicated. 500 ng miRNA expression plasmid was transfected per sample. (C) Reporter protein normalized to control protein labeling in serum-fed conditions. Shown are two experimental replicates, and the fold-repression caused by miR-CXCR4 is indicated. (D) Normalized reporter protein labeling in serum-depleted conditions. Shown are two experimental replicates, and the fold-repression caused by miR-CXCR4 is indicated. (E) Reporter mRNA levels normalized to control in serum-fed and serum-depleted conditions, as assayed by RT-qPCR. Standard deviations are shown as error bars. (F) Western blot of Ago1 and α-tubulin proteins after serum-depletion for the time intervals indicated. (G) Cells were treated with cycloheximide for times as indicated to arrest protein synthesis. Ago1 and α-tubulin protein were detected by Western blot to measure protein turnover. (H) Western blot of GW182 protein from titrated extract derived from serum-fed cells or cells depleted of serum for four hours. (I) Northern blots of miR-2a, miR-CXCR4, and 2S ribosomal RNA from cells after two hours serum depletion, as indicated. 2S rRNA is a maturation product of 28S rRNA that is found in Drosophila. See also Figures S1 and S2.

We first induced low levels of miR-CXCR4 in cells, and under serum-fed conditions, reporter protein synthesis was very weakly repressed (Fig. 1B). However during serum depletion (two-to-four hours), reporter protein synthesis was repressed five-fold by miR-CXCR4. Thus, miR-CXCR4 repressed reporter protein synthesis much more effectively when serum had been briefly deprived from cells. To confirm this effect, we induced higher levels of miR-CXCR4 in serum-fed cells, resulting in ten-fold repression (Fig. 1C and Fig. S2). Under serum depleted conditions, repression was strongly enhanced to a level between 50- and 100-fold (Fig. 1D and Fig. S2). Thus, serum withdrawal caused a large increase in miRNA-mediated repression over a significant dynamic range.

The impact of the CXCR4 miRNA on reporter protein synthesis could have been a consequence of mRNA decay or translation repression. To determine the contribution of mRNA decay, we measured reporter transcript levels by RT-qPCR. Transcript levels were weakly sensitive to the effects of miR-CXCR4 whether or not cells were incubated in serum (Fig. 1E). This result argues that the primary effect of miR-CXCR4 is on reporter translation and not mRNA decay. The relative effect was unchanged during serum withdrawal, and together, the data implies that serum withdrawal greatly enhanced miRNA-mediated translation repression.

Stronger miRNA-mediated repression upon serum withdrawal could stem from changes in the level of Ago1 protein. However, Ago1 protein abundance was unchanged with or without serum (Fig. 1F). Furthermore, the turnover rate of Ago1 was unaffected by serum depletion (Fig. 1G). We also looked at the steady state level of GW182 and found it to be unaffected by serum withdrawal (Fig. 1H). Finally, an examination of the abundance of mature miR-CXCR4 and an endogenous miRNA (miR-2a) revealed that increased miRNA biogenesis was not the cause of enhanced repression after serum withdrawal (Fig. 1I).

Serum withdrawal generates novel miRISC complexes

We ruled out that the effect of serum withdrawal was due to altered miRNA biogenesis, target mRNA levels, or abundance of miRISC components. It therefore seemed possible that the effect was caused by a change in miRISC activity. To examine the biochemical properties of miRISC from S2 cells, we separated high molecular weight complexes by iodixanol density gradient centrifugation. Iodixanol’s low viscosity and iso-osmolarity allow for efficient separation of protein complexes and subcellular organelles (Graham et al., 1994; Sheff et al., 1999). Cytoplasmic extract from S2 cells was fractionated on a 5–20% iodixanol density gradient. This resolved cytosolic complexes (light fractions), membranous organelles (middle fractions), and endoplasmic reticulum and ribosomes (heavy fractions) (Fig. 2A,B). Fractions were then assayed for the presence of miRISC subunits. Extract from serum-fed cells showed a simple fractionation profile, in which Ago1, GW182, and endogenous miRNA miR-2a co-sedimented in a single peak (Fig. 2C,E). miR-2a co-immunoprecipitated with Ago1 protein present in this peak (Fig. 2F). These results are consistent with studies that found S2 cell miRISC to be composed of mature miRNA associated with Ago1 and GW182 (Behm-Ansmant et al., 2006; Eulalio et al., 2008; Miyoshi et al., 2009). We refer to this complex as GW182-miRISC or G-miRISC.

Figure 2. Novel miRISC complexes are induced by serum withdrawal.

Figure 2

Open in a new tab

(A–E, G–I) Iodixanol gradient sedimentation of cytoplasmic extracts from serum-fed and serum-depleted cells. (A) Sedimentation of 2S rRNA to mark ribosomes. 2S rRNA is a maturation product of 28S ribosomal RNA that is found in Drosophila. Neighboring fractions were pooled and total 2S rRNA in each pool was measured. (B) Sedimentation of Cactus (55 kDa cytosolic protein), Rab7 (23 kDa endosome-associated protein), and Calnexin (68 kDa endoplasmic reticulum-associated protein) to mark different cytoplasmic compartments. (C) Distribution of Ago1, GW182, and miR-2a from cells grown under serum-fed conditions. Every fraction was assayed to illustrate the resolution of protein and RNA peaks. (D, E) Distribution of Ago1, GW182, and miR-2a in which alternating fractions ranging from 4 to 24 were assayed on the same blot to control for variability in loading and transfer. (D) Sample from serum-depleted cells. (E) Sample from serum-fed cells. (F) Immunoprecipitation of Ago1 from light (fraction 8) and heavy (fraction 24) peaks after sedimentation of extract from serum-depleted cells in an iodixanol gradient. Co-precipitated miR-2a RNA is shown below. (G, H) Profile of Ago1 and GW182 in fractions 10–21 of the iodixanol gradient to highlight the shoulder of Ago1 sedimentation that is present in serum-depleted cells (G) and absent in serum-fed cells (H). (I) Average Ago1 levels in fractions from serum-fed and serum-depleted samples. Averages are from four independent experiments. Standard deviations are shown as error bars. Asterisks denote a significant difference between the values from corresponding fractions (* indicates P < 0.05; ** indicates P < 0.01, t-test). (J) Levels of GW182 protein in serum-depleted cells expressing high levels of miR-CXCR4 as indicated (500 ng transfected). Cells experienced RNAi knockdown of GW182 as indicated. Control RNAi treatment involved soaking cells in lacZ dsRNA. (K) Effects of highly expressed miR-CXCR4 on radiolabeling of reporter and control proteins in serum-depleted cells. Cells experienced RNAi knockdown of GW182 or control dsRNA. See also Figure S3.

We then sedimented extract from cells that had been serum-depleted for two hours. A peak corresponding to G-miRISC was detected, but additional fractions also contained miRISC subunits (Fig. 2D). There was a new peak of Ago1 and miR-2a near the bottom of the gradient, and strikingly, this peak had virtually no GW182 protein. To ascertain if miR-2a and Ago1 were physically associated in this peak, we immunoprecipitated Ago1 from the peak fraction and observed co-immunoprecipitation of miR-2a (Fig. 2F). Thus, a novel Ago1-miRNA complex was induced by serum withdrawal.

Under serum-depleted conditions, the G-miRISC peak of Ago1 and miR-2a showed a shoulder of heavier fractions (fr. 12–17) that contained these subunits (Fig. 2D,G). GW182 did not sediment in the shoulder fractions, indicating that these complexes were not stably associated with GW182. The Ago1 shoulder in fractions 12–17 appeared minor but summed together, constituted 17% of total Ago1 protein. In contrast, serum-fed cells had only 4% of total Ago1 protein in fractions 12–17 (Fig. 2H). Moreover, the detection of this shoulder from serum-depleted cells was highly reproducible and significant (Fig. 2I).

The novel Ago1-miRNA complexes were induced by conditions that also enhanced miRNA-mediated gene silencing. A reasonable hypothesis was that the novel complexes were responsible for the increase in reporter gene silencing during serum withdrawal. Unlike G-miRISC, the novel complexes did not appear to contain GW182. If these GW182-free complexes were capable of silencing gene expression, we predicted that enhanced silencing of reporter gene expression in serum-depleted cells would be independent of GW182. To test this prediction, we knocked down GW182 protein by soaking cells with GW182 dsRNA (Fig. 2J). Similar knockdown of GW182 in serum-fed S2 cells resulted in impaired silencing of reporter genes by various miRNAs (Behm-Ansmant et al., 2006). We then measured reporter gene silencing by miR-CXCR4 under serum-depleted conditions. Strikingly, the reporter remained strongly silenced in cells lacking GW182 (Fig. 2K). Thus, the enhanced repression induced by serum withdrawal cannot be explained by augmented repressive activity of G-miRSC. We conclude that serum withdrawal induces a form of miRNA-mediated silencing that is not wholly dependent on GW182.

Polyribosome Associated Ago1-miRNA Complexes

The heaviest Ago1-miRNA complex co-sedimented with ribosomal RNA in the iodixanol gradient (Fig. 2A,D). We wondered if this complex was physically associated with mRNA transcripts undergoing translation. If true, we predicted that complexes would co-sediment with polysomes during sucrose gradient centrifugation. We subjected cell extracts to centrifugation through a 15–50% sucrose gradient and identified fractions that contained polyribosomes (Fig. 3A). Polysomal fractions from serum-fed cells contained little Ago1 and miRNA (Fig. 3B). In contrast, polysomal fractions from serum-depleted cells contained significant levels of Ago1 and miRNA (Fig. 3C,D). Thus, Ago1 and miRNA co-fractionated with ribosomes by two distinct centrifugation methods.

Figure 3. Serum-free miRISC associates with polysomes.

Figure 3

Open in a new tab

(A, B) Sucrose gradient sedimentation of extract from serum-fed cells. Fractions 1–4 are omitted since they contain the bulk of miRISC but lack ribosomes/polysomes. (A) Light absorbance at 254nm to visualize ribosomal RNA. Note the monosome peak in fraction 9. (B) Distribution of Ago1, Loqs-PB, and miR-2a across fractions. (C, D) Sucrose gradient sedimentation of extract from serum-depleted cells. (C) Light absorbance at 254nm to visualize ribosomal RNA. (D) Distribution of Ago1, Loqs-PB, and miR-2a across fractions. Fractions 1–4 (not shown) contain the bulk of these factors, and their profiles in fr. 1–4 are similar to their profiles in fr. 1–4 from the serum-fed sample. (E, F) Iodixanol gradient sedimentation of Ago1 from serum-depleted cells. (E) Cells were treated with cycloheximide prior to and subsequent to homogenization, and untreated cells were analyzed in parallel. (F) Cells were treated with puromycin prior to and subsequent to homogenization, and untreated cells were analyzed in parallel. (G) Sucrose gradient sedimentation of Ago1 from serum-depleted cells that were treated with puromycin prior to and subsequent to homogenization. Untreated cells were processed in parallel. (H, I) Iodixanol gradient sedimentation of Ago1 and miR-2a from serum-depleted cells. (H) Extract was mock-nuclease-treated prior to centrifugation. (I) Extract was treated with micrococcal nuclease prior to centrifugation.

Cycloheximide was added to extracts before the sucrose sedimentation to stabilize polysomes; cycloheximide arrests ribosome elongation. If sucrose and iodixanol gradients were resolving the same Ago1 complexes, we predicted that cycloheximide would have no effect on Ago1 sedimentation in the iodixanol gradient. Indeed, extract sedimented from serum-depleted cells showed that these complexes remained intact in the presence of cycloheximide (Fig. 3E).

Drosophila transcripts subjected to miRNA repression have been observed to co-sediment with polysomes in distinct structures called pseudo-polysomes (Thermann and Hentze, 2007; Fukaya and Tomari, 2012). These structures are insensitive to puromycin, which causes premature termination of translation and disassembly of polysomes. To determine if the Ago1-miRNA complexes induced by serum withdrawal were pseudo-polysomes, we treated cells with puromycin either for the entire two hours of serum withdrawal or the final 20 minutes of the withdrawal treatment. Sedimentation of extract from such cells through iodixanol and sucrose gradients showed that Ago1 no longer sedimented in the same regions of the gradients but shifted to lighter sedimenting fractions (Fig. 3F,G). Polysomes were also disrupted by the puromycin treatment, as predicted (data not shown). Therefore, the Ago1 complex induced by serum withdrawal was associated with polysomes and not pseudo-polysomes.

If the Ago1 complex was associated with polysomes, then the complex would be sensitive to nuclease digestion of the transcript RNA undergoing translation. We briefly treated cytoplasmic extract with micrococcal nuclease, which degrades RNA. Strikingly, the heavy peak of Ago1 was greatly reduced after this treatment (Fig. 3H,I). The effect was not due to nuclease digestion of miRNA since blots of miR-2a showed normal levels of the miRNA after nuclease treatment, although its sedimentation profile was altered in parallel to Ago1 (Fig. 3H,I). Based on these collective data, we conclude that the heaviest Ago1-miRNA complex is associated with ribosome-bound transcripts. For this reason, we hereafter refer to it as Polyribosomal miRISC (P-miRISC). We surmise that serum withdrawal potentiates the formation of P-miRISC.

Membrane association of Ago1-miRNA complexes

As noted earlier, fractions 12–17 in the iodixanol gradient from serum-depleted cells contained a shoulder of Ago1 and miRNA but not GW182 (Fig. 2G). This shoulder co-sedimented with intracellular membranous organelles (Fig. 2B), suggesting that Ago1 was associated with such bodies. To test this hypothesis, we treated extract with the detergent Triton X-100, which disrupted the sedimentation of membranous organelles in iodixanol (Fig. 4A). The shoulder of Ago1 was entirely abolished by this treatment, indicating that the Ago1 was associated with detergent-sensitive membranes (Fig. 4B). As predicted, G-miRISC from serum-fed cells was unaffected by detergent (Fig. 4C). Ago1 has been found associated with late endosomes, which are labeled with Rab7 protein (Feng et al., 1995; Lee et al., 2009). When extract from serum-depleted cells was fractionated by iodixanol or sucrose gradient sedimentation, there was only partial overlap between Ago1 and Rab7 (Figs. 2B,D and 4D,E). Therefore, late endosomes are not the only organelles that account for the appearance of membrane-associated Ago1.

Figure 4. Membrane association of novel miRISC complexes.

Figure 4

Open in a new tab

(A–C) Iodixanol gradient sedimentation of extract that was treated with Triton X-100 detergent before centrifugation. (A) Distribution of the endosomal protein Rab7 and endoplasmic reticulum protein Calnexin from serum-fed cells to demonstrate the efficacy of the detergent treatment. Compare their profiles to the untreated samples in Figure 2B. (B) Ago1 protein distribution from serum-fed cells. Detergent treatment as indicated. (C) Ago1 protein distribution from serum-depleted cells. Detergent treatment as indicated. (D–E) Sucrose gradient sedimentation of Ago1 and Rab7 in relation to ribosomal RNA, as measured by light absorbance at 254 nm. (D) Distribution of Ago1, Rab7, and ribosomes from serum-fed cells. (E) Distribution of Ago1, Rab7, and ribosomes from serum-depleted cells. (F) Distribution of Ago1 from serum-depleted cells in which extract was treated with Triton X-100 detergent prior to sucrose gradient centrifugation. The distribution from an untreated sample is also shown.

We observed that the heavy peak in iodixinol gradients was mildly disrupted by detergent treatment (Fig. 4B). These fractions contain components of the endoplasmic reticulum (ER) (Fig. 2B), and detergent treatment strongly disrupted ER since the ER protein calnexin did not sediment after Triton X-100 treatment (Fig. 4A). Since P-miRISC appeared to be polysome-associated, we wondered if some of these complexes were membrane bound. We sedimented detergent-treated extract through a sucrose gradient. Detergent treatment reduced the levels of ribosomal RNA and Ago1 that were present in polysomal fractions, consistent with membrane association (Fig. 4F and data not shown). Thus, some P-miRISC/polysome complexes are associated with membrane, most likely ER.

Assembly of Novel Ago1-miRNA Complexes

Distinct Ago1-miRNA complexes appear after serum withdrawal, but it was unclear how these complexes are made. Do they arise de novo or does serum withdrawal stimulate dissociation of GW182 from miRISC, liberating the Ago1-miRNA heterodimer to engage with targets in a novel manner? To address this question, we immunoprecipitated Ago1 and examined the levels of miRNAs and GW182 associated with Ago1. The amount of associated miRNA was weakly increased after serum withdrawal (Fig. 5A), indicating that at most, a little more Ago1 was loaded with miRNA. The amount of associated GW182 did not change after serum withdrawal (Fig. 5B). Therefore, GW182 dissociation was not correlated with formation of the inducible complexes. We knocked down GW182 protein from serum-fed cells by soaking them with GW182 dsRNA. This resulted in reduction of GW182 without affecting Ago1 levels (Fig. 5C). Extract from these cells was then sedimented through an iodixanol gradient. The novel miRISC complexes were not detected after such treatment (Fig. 5D). However, the G-miRISC peak shifted to lighter fractions, indicating that GW182-free Ago1 complexes were generated by the knockdown. We conclude that GW182 dissociation from Ago1 is not sufficient to form the novel miRISC complexes. These novel complexes likely form de novo by a different mechanism.

Figure 5. Subunit constituency of miRISC.

Figure 5

Open in a new tab

(A) Immunoprecipitation of Ago1 protein from serum-fed and serum-depleted cells, and detection of co-precipitated miRNAs, as indicated. As a control for antibody specificity, anti-V5 antibody was used to immunoprecipitate from extract. Below is shown the average level of co-immunopreciptated miRNA from three independent experiments. Standard deviations are shown as error bars. (B) Immunoprecipitation of Ago1 protein from serum-fed and serum-depleted cells, and detection of co-precipitated GW182. As a control for antibody specificity, anti-V5 antibody was used. (C) GW182 and Ago1 proteins from serum-fed cells that were soaked in GW182 dsRNA or control dsRNA for three days. (D) Cells soaked in GW182 or control dsRNA were maintained in serum-fed conditions before sedimentation on an iodixanol gradient. The sedimentation of Ago1 in fraction 24 is unchanged whereas the upper peak of Ago1 shifts to lighter fractions when cells were depleted of GW182. (E) Iodixanol gradient sedimentation of Loqs-PB protein from serum-fed and serum-depleted cells.

We next sought to explore the relationship between the novel Ago1 complexes and the miRNA Loading Complex (miRLC). miRLC contains Ago1, Dicer-1 (Dcr-1), and its partner protein Loquacious (Loqs-PB) (Miyoshi et al., 2009). miRLC normally does not contain mature miRNA but instead contains pre-miRNA. It is believed that miRLC functions to process and load mature miRNA into Ago1. The loading of mature miRNA coincides with Ago1 dissociation from Dcr-1/Loqs-PB and association with GW182 (Miyoshi et al., 2009). We examined the profile of Loqs-PB from extracts sedimented in iodixanol. When analyzed from serum-fed cells, Loqs-PB sedimented in a peak that overlapped with Ago1 (Fig. 5D,E). When we assayed fractions from serum-starved cells for Loqs-PB, we observed a reduction in the lighter peak and the presence of a new peak of Loqs-PB that co-sedimented with P-miRISC (Fig. 5E).

This result suggested that Loqs-PB is associated with P-miRISC. To validate this observation, we sedimented extract from serum-depleted cells through a 15–50% sucrose gradient, which finely resolves the P-miRISC complex. Strikingly, the levels of Ago1, miRNA, and Loqs-PB proteins in polysomal fractions precisely correlated (Fig. 3D). Under serum-fed conditions, virtually no Loqs-PB was detectable in polysomal fractions (Fig. 3B). This result indicates that P-miRISC contains Ago1/miRNA/Loqs-PB. Although we could not assay for Dcr-1 protein in these fractions due to poor Dcr-1 antigenicity, we surmise that P-miRISC might carry a Dcr-1 subunit as well.

Lipid and Insulin Signaling Regulate Ago1 Complexes

P-miRISC and membrane-associated Ago1 complexes were detectable from cells that were withdrawn from serum for less than one hour, and they remained present for at least six hours of serum withdrawal (Fig. S3, data not shown). Their formation was reversible, since re-addition of serum reduced their levels (Fig. S3). Thus, they are highly dynamic complexes that are dependent on serum conditions.

Serum is a complex mixture of biomolecules, and it elicits diverse effects on many aspects of cell biology. The protein kinase C (PKC) family is an extensively studied group of kinases that integrate multiple lipid- and insulin-activated signals from serum (Griner and Kazanietz, 2007; Newton, 1995). To test whether PKC mediates the effect of serum on miRISC, we used a phorbol ester that constitutively activates PKC called 12-O-tetradecanoylphorbol-13-acetate (PMA). Cells grown in serum-supplemented medium were transferred to serum-free medium containing PMA and were cultured for two hours before extract was sedimented by centrifugation. This treatment was sufficient to greatly inhibit the formation of membrane-associated Ago1 and P-miRISC complexes without affecting G-miRISC (Fig. 6A). PKC is activated by different extracellular signals including lipids (Hannun and Obeid, 2008). We transferred S2 cells from serum-containing medium to serum-free medium that was supplemented with lipid-rich albumin and a cocktail of free fatty acids and cholesterol. After two hours, cell extracts were sedimented by gradient centrifugation to resolve miRISC complexes. Strikingly, the lipid supplements were also sufficient to inhibit the formation of membrane-associated Ago1 and P-miRISC (Fig. 6B).

Figure 6. Extracellular lipids and insulin regulate miRISC complexes.

Figure 6

Open in a new tab

(A–C) Iodixanol gradient sedimentation of Ago1 from cells incubated in serum-free medium for two hours. (A) Serum-free medium was either supplemented with the phorbol ester PMA or its carrier solvent DMSO. (B) Serum-free medium was supplemented with lipids or was unsupplemented. (C) Serum-free medium was supplemented with insulin or was unsupplemented. (D) Iodixanol gradient sedimentation of Ago1 from cells incubated in lipid-supplemented serum-free medium that was further supplemented with insulin as indicated. (E) A proposed model for lipid and insulin-mediated regulation of miRISC formation. See text for details.

Next, we assessed whether growth factors were involved. Although serum contains multiple growth factors, only insulin is known to affect S2 cells (Britton et al., 2002; Puig et al., 2003). We transferred cells from serum-containing medium to serum-free medium that was supplemented with insulin. This treatment inhibited formation of membrane-associated Ago1, but had no impact on P-miRISC formation (Fig. 6C). We noticed that formation of membrane-associated Ago1 was partially but not completely inhibited by lipids. This pointed to the possibility that insulin acted additively with lipids to inhibit this complex. To test the hypothesis, we supplemented serum-depleted cells with insulin and lipids simultaneously. This resulted in a sedimentation profile that closely resembled that of serum-fed cells (Fig. 6D). It suggests that insulin acts in parallel with lipids to repress formation of membrane-associated Ago1.

Discussion

Here, we find that different miRISC complexes are present in S2 cells, depending upon extracellular signals received by the cells. A constitutive G-miRISC complex composed of Ago1, miRNA, and GW182 is present under all signaling conditions tested. Other groups have shown that G-miRISC in S2 cells suppresses target mRNAs via inhibition of translation initiation and enhanced mRNA decay (Behm-Ansmant et al., 2006; Eulalio et al., 2008; Zdanowicz et al., 2009; Zekri et al., 2009). We found that lipid signaling does not affect G-miRISC but blocks other miRISC complexes from forming. This signaling is likely mediated by PKC since a phorbol ester mimics the effect of lipids on miRISC formation. Signaling blocks the formation of P-miRISC, which contains Ago1, miRNA, and Loqs-PB but not GW182. P-miRISC represses translation of target mRNAs, which is manifested in polysome association of the complex. Thus, our work reveals a mechanistic shift in miRISC-executed translation repression under the influence of extracellular lipid signals (Fig. 6E). In the presence of lipid signaling, initiation is inhibited and this occurs by G-miRISC. In the absence of lipid signaling, we propose that cells generate two levels of translational repression: one mediated by G-miRISC that inhibits initiation, and one mediated by P-miRISC that inhibits elongation. We propose that each miRISC complex independently represses the same target, and because they act in series (initiation - elongation), the net result on protein synthesis is the product (not sum) of each inhibitory step. This would provide the strongly synergized repression of reporter protein synthesis that was observed after serum withdrawal.

P-miRISC resembles the miRLC complex in terms of subunit composition (Ago1, Loqs-PB) but the two differ in one important way. Whereas miRLC contains pre-miRNA, P-miRISC contains mature miRNA. Thus, P-miRISC has an inherent potential to engage target mRNAs via base pairing interactions. We suggest that P-miRISC is formed by the processing and loading of mature miRNA into Ago1 within the miRLC (Fig. 6E). Rather than releasing Loqs-PB/Dcr-1 and recruiting GW182, the loaded Ago1 retains Loqs-PB and never recruits GW182. P-miRISC can then engage target mRNAs but its subunit composition dictates a different mode of repression upon the target.

Although GW182 and Loqs-PB binding to Ago1 are mutually exclusive (Miyoshi et al., 2009), P-miRISC is not simply a default state when GW182 recruitment fails to occur. Knockdown of GW182 was insufficient to induce formation of P-miRISC. Moreover, formation of P-miRISC did not appear to occur at the expense of G-miRISC levels, as measured in sedimentation and immunoprecipitation experiments. This suggests a mechanism in which stable loading of miRNA is limited by the availability of cofactors for Ago1. Under serum-fed conditions, only GW182 is available whereas both GW182 and Loqs-PB are available under serum-free conditions. Possibly this offers a rapid way to modulate miRISC levels without the need for synthesis of more cofactors.

The switch in miRISC formation is regulated by PKC but it is not clear how this switch occurs. A recent study demonstrated that the mammalian homolog of Drosophila Ago1 can be phosphorylated by Akt3, which contributes to increased miRISC-mediated translation repression (Horman et al., 2013). However, we did not find evidence for differential phosphorylation of Ago1 in S2 cells (data not shown). A study of the mammalian ortholog of Loqs-PB, called TRBP, found it to be phosphorylated by ERK kinase in response to PKC (Paroo et al., 2009). Phosphorylation stabilized miRLC and increased processing of growth-promoting miRNAs. The same mechanism was not shown for Loqs-PB, and our examination of the Loqs-PB sequence failed to find strict conservation of those sites.

A second Ago1 complex also appears when lipid signaling is absent. Membrane-associated Ago1 likely contains miRNA but not Loqs-PB or GW182. Association of mammalian Ago proteins with late endosomes has been previously observed (Gibbings et al., 2009). Drosophila Ago1 has also been observed to associate with endosomes in vivo (Lee et al., 2009). Endosomes have been proposed to serve as sites for miRISC turnover whereby miRISC continuously associates and releases from endosomes, constituting a mechanism that promotes miRISC recycling onto new targets (Gibbings et al., 2009; Lee et al., 2009). Thus, membrane-associated Ago1 may represent an intermediate in miRISC turnover. If so, where does the membrane-associated Ago1 originate? Several lines of evidence suggest that it originates from P-miRISC. First, its appearance precisely correlates with P-miRISC. Second, it is sensitive to puromycin treatment, which also disrupts association of P-miRISC with polysomes. However, membrane-associated Ago1 does not sediment in ribosome-containing fractions. Third, insulin specifically inhibits membrane-associated Ago1, arguing that membrane-associated Ago1 is not an obligate precursor of P-miRISC. The simplest interpretation of these data is that membrane-associated Ago1 is formed from a P-miRISC precursor. If so, then Loqs-PB dissociation must be involved in the conversion since Loqs-PB is not found in the membrane-associated complex. A similar manner of cofactor stripping was observed for GW182, which dissociated from Ago-miRNA complexes when they associated with endosomes (Gibbings et al., 2009). Perhaps cofactor dissociation is a fundamental part of the recycling mechanism.

Our model might provide some insights into a long-standing controversy in the miRNA field. Some studies have found evidence for translation initiation as the regulated step while others have found evidence for translation elongation (Carthew and Sontheimer, 2009). Our work provides a potential explanation for these differences. That is, experimental model systems experiencing diverse extracellular signals might respond accordingly to form distinct types of miRISC complexes, which regulate different steps of translation. Thus, all studies have depicted an accurate picture of miRISC activity because signals that dictate miRISC subunit composition affect its mode of action.

Experimental Procedures

Cell culture

S2 cells were cultured in Schneider’s Drosophila medium (Invitrogen) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco). Cultures were never allowed to grow to saturation phase while passaging. For serum depletion experiments, cells were only used when their density was less than 2 × 106 cells/ml. On the day of a depletion experiment, cells were incubated for 1–2 hr in fresh medium supplemented with 10% FBS. Cells were then briefly washed with and resuspended in serum-free medium (SFM). Unless otherwise noted, serum depletion was for 2 hr. For serum-fed controls, cells were washed and resuspended in medium supplemented with 10% FBS. For lipid-supplemented medium, we added 1% AlbuMAX (Invitrogen) and 1X Chemically Defined Lipid Concentrate (Invitrogen) to SFM. For PKC activation, cells were incubated in SFM containing 100 nM PMA (Sigma). For insulin treatment, SFM was supplemented with 1 μg/ml bovine insulin (Sigma).

miR-CXCR4 reporter assay

The miR-CXCR4 expression plasmid pMT-D05 (Zhou et al., 2008) contains the CXCR4 miRNA sequence embedded in the pri-let7 miRNA. The pMK-3xHA-GFP-CXCR4-BS reporter and pMK-GFP control genes were constructed from pMK-3xHA-GFP (Lee at al., 2009). This parent plasmid has the metallothionein promoter driving HA-tagged GFP. The SV40 3′UTR with four imperfect CXCR4 miRNA binding sites (Doench et al., 2003) was inserted into pMK-3xHA-GFP. To make the control, 3xHA was removed from pMK-3xHA-GFP and an unmodified SV40 3′UTR was inserted. The only difference between the modified and unmodified SV40 3′UTRs are the 20-nt CXCR4 binding sites.

To assay CXCR4 miRNA repression, 1 × 106 cells were transfected with 500 ng pMK-3xHA-GFP-CXCR4-BS, 500 ng pMK-GFP, and either 100 ng or 500 ng pMT-D05 using Cellfectin II (Invitrogen). CuSO4 was added to medium 24 hr after transfection to induce control, reporter and CXCR4 miRNA gene expression. Cells were incubated for a further 36 hr until serum depletion.

Metabolic labeling was initiated 2 hr after cells had been transferred to SFM. After a brief wash with methionine-free SFM, cells were incubated for 2 hr in methionine-free SFM containing 100 μCi/ml 35S-methionine (Perkin Elmer). Control incubation was accomplished by supplementing methionine-free SFM with 10% methionine-free FBS, which was generated by dialysis against PBS. Cells were homogenized, and GFP was immunoprecipitated using an anti-GFP monoclonal antibody (Wako) and DynaBeads (Invitrogen). Eluted precipitate was resolved by SDS-PAGE, and labeled proteins were detected on a PhosphoImager (Molecular Dynamics) and quantified by ImageQuant 5.

Stress granule analysis

As described (Farny et al., 2008), cells were xed in 4% paraformaldehyde and washed in PBS + 1% Triton X-100. This was followed by changes in methanol, 70% ethanol, and 1M Tris pH 8.0. Cells were incubated with 1 μg/ml Cy3-oligo dT(30) (IDT) at 37°C overnight, and were washed twice in 2 x SSC before mounted for imaging with a Zeiss fluorescence microscope.

RNAi Knockdowns

MEGAscript (Ambion) was used to synthesize dsRNA against GW182 and LacZ using PCR-generated DNA templates. Cells were switched to SFM 15 min prior to dsRNA soaking. We soaked 1 × 106 cells with 10 μg dsRNA. Cells were gently agitated at room temperature for 30 minutes, and FBS was then added to a final concentration of 10%. After three days, cells were subjected to subsequent analysis. To assay miRNA repression activity after GW182 knockdown, dsRNA soaking and plasmid transfection were carried out sequentially on the same day. The procedures for each were as described.

Iodixanol Gradient Centrifugation

Cells were homogenized by passage through a 261/2 G syringe in HB buffer composed of 250 mM sucrose, 30 mM HEPES pH 7.4, 100 mM KOAc, 2 mM Mg2OAc, 5 mM DTT, and proteinase inhibitors. Homogenate was centrifuged at 1000 × g for 5 minutes, and the supernatant (3 mg protein) was layered on top of a 5%–20% linear density gradient prepared with Optiprep media (AxisShield). This was centrifuged in a SW41Ti rotor (Beckman) at 90,000 × g for 18 hours at 4°C. Fractions were collected from top to bottom.

Sucrose Gradient Centrifugation

Cycloheximide was added to the cell culture at a final concentration of 100μg/ml 20 min before harvesting. Cells were homogenized in HB buffer + 100 μg/ml cycloheximide. The homogenate was centrifuged at 1000 × g for 5 minutes at 4°C, and the supernatant (3mg of total protein) was layered on a 15%–50% (w/v) linear sucrose gradient. Fractionation was carried out in a SW41Ti rotor (Beckman) at 40,000 rpm for 90 minutes at 4°C. Fractions were collected from top to bottom. Polysome pro les were obtained by monitoring fractions for 254 nm light absorbance (A254).

Cell and Extract Manipulations

To assay sensitivity to translation inhibitors, cells were incubated in medium supplemented with 50 μg/ml cycloheximide (Sigma) or 300 μM puromycin (Sigma) for 20 to 120 min before harvesting. To assay sensitivity to nuclease, extract was treated with micrococcal nuclease (1 unit/mg protein) (Sigma) for 2 min at 28°C. EGTA was added to 2 mM final concentration to stop the nuclease reaction, and was present in the gradient as well. Mock treatment involved the same steps except that nuclease addition was omitted. Detergent sensitivity was assayed by treating extract with 1% Triton X-100 on ice for 30 minutes prior to centrifugation.

Antibodies

Ago1, GW182, and Loqs antibodies were gifts from the Siomi labs (Keio University and University of Tokyo, Japan). GW182, Rab7, and Calnexin antibodies were gifts from A. Simmonds (University of Alberta), A. Nakamura (RIKEN Center for Developmental Biology, Japan), and N. Colley (University of Wisconsin), respectively. Antibodies for Cactus and α-tubulin were from the Developmental Studies Hybridoma Bank. GFP antibody was from Wako.

RNA and Protein Analysis

To detect miRNAs in iodixanol fractions, pairs of adjacent fractions were pooled together, and RNA was extracted using phenol-chloroform. Northern blotting was carried out as described (Marques et al., 2010) with LNA probes (Exiqon). Probe for 2S rRNA has been described (Czech et al., 2008). A PhosporImager (Molecular Dynamics) and ImageQuant 5 were used to quantitate signal. RNA from sucrose fractions was extracted and analyzed in the same way except that fractions were not pooled. Quantitative RT-PCR was carried out as previously described (Marques et al., 2010). Relative mRNA levels were obtained by normalizing to RpL32 using the ΔΔCt method. Western blotting to nitrocellulose was detected by ECL (ThermoScientific) and quantified by scanning. PhotoShop CS4 was used to measure signal intensity and gradient fraction signals were normalized to the signal from input extract that was used to load each gradient.

Immunoprecipitation

Immunoprecipitation of Ago1 from S2 cell extract was as previously described (Miyoshi et al., 2009). We immunoprecipitated Ago1 from iodixanol fractions using the same procedure, with Ago1 antibody directly added to the fractions. Immunoprecipitation of reporter GFP was performed in a similar way, except that 5% glycerol was added to the IP buffer

Oligonucleotides

GW182 and GFP dsRNA:

  • T7-GW182 (forward): 5′-TAATACGACTCACTATAGGGGGTTCTGGAGCAACTTCGAG-3′

  • T7-GW182 (reverse): 5′-TAATACGACTCACTATAGGGGGACCACTTACACCAACGCT-3′

  • T7-GFP (forward): 5′-TAATACGACTCACTATAGGGCACATGAAGCAGCACGACTT-3′

  • T7-GFP (reverse): 5′-TAATACGACTCACTATAGGGTGTTCTGCTGGTAGTGGTCG-3′

  • T7-LacZ (forward): 5′CCCGGGTAATACGACTCACTATAGGGTGAAACGCAGGTCGCCAG-3′

  • T7-LacZ (reverse): 5′-CCCGGGTAATACGACTCACTATAGGGTGATTACGATCGCGCTGC-3′

RT-qPCR:

  • RpL32 (forward): 5′-GACGCTTCAAGGGACAGTATCTG-3′

  • RpL32 (reverse): 5′-AAACGCGGTTCTGCATGAG-3′

  • GFP (forward): 5′-GAAGTTCGAGGGCGACAC-3′

  • GFP (reverse): 5′-CCGTCCTCCTTGAAGTCG-3′

Northern blotting probes:

  • miR-2a: 5′-GCTCATCAAAGCTGGCTGTGATA-3′

  • CXCR4 miRNA: 5′-AAGTTTTCACTCCAGCTAACA-3′

Supplementary Material

01

Highlights.

  • Lipid signaling through PKC inhibits formation of P-miRISC

  • P-miRISC associates with polysomes and represses translation of miRNA targets

  • A constitutive form of miRISC represses translation independent of signaling

  • Loss of signaling causes much stronger repression of miRNA targets

Acknowledgments

We thank N. Colley, G. Hannon, C. Horvath, A. Nakamura, C. Petersen, A. Simmons, H. Siomi, M. Siomi for gifts of reagents. We also thank the O. Uhlenbeck and R. Lamb labs for sharing centrifuges, and A. Wanless and R. Patterson for their valuable assistance. We are indebted to the Developmental Hybridoma Studies Bank and Flybase for their resources. We thank our lab colleagues for advice and help. Funding support is gratefully acknowledged from the American Heart Association to P.-H.W., and National Institutes of Health (GM068743 and GM077581) to R.W.C.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Bagga S, Bracht J, Hunter S, Massirer K, Holtz J, Eachus R, Pasquinelli AE. Regulation by let-7 and lin-4 miRNAs results in target mRNA degradation. Cell. 2005;122:553–563. doi: 10.1016/j.cell.2005.07.031. [DOI] [PubMed] [Google Scholar]
  2. Behm-Ansmant I, Rehwinkel J, Doerks T, Stark A, Bork P, Izaurralde E. mRNA degradation by miRNAs and GW182 requires both CCR4:NOT deadenylase and DCP1:DCP2 decapping complexes. Genes & Development. 2006;20:1885–1898. doi: 10.1101/gad.1424106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bhattacharyya SN, Habermacher R, Martine U, Closs EI, Filipowicz W. Relief of microRNA-mediated translational repression in human cells subjected to stress. Cell. 2006;125:1111–1124. doi: 10.1016/j.cell.2006.04.031. [DOI] [PubMed] [Google Scholar]
  4. Britton JS, Lockwood WK, Li L, Cohen SM, Edgar BA. Drosophila’s insulin/PI3-kinase pathway coordinates cellular metabolism with nutritional conditions. Developmental cell. 2002;2:239–249. doi: 10.1016/s1534-5807(02)00117-x. [DOI] [PubMed] [Google Scholar]
  5. Carthew RW, Sontheimer EJ. Origins and Mechanisms of miRNAs and siRNAs. Cell. 2009;136:642–655. doi: 10.1016/j.cell.2009.01.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Czech B, Malone CD, Zhou R, Stark A, Schlingeheyde C, Dus M, Perrimon N, Kellis M, Wohlschlegel JA, Sachidanandam R, et al. An endogenous small interfering RNA pathway in Drosophila. Nature. 2008;453:798–802. doi: 10.1038/nature07007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Ding XC, Grosshans H. Repression of C. elegans microRNA targets at the initiation level of translation requires GW182 proteins. EMBO J. 2009;28:213–222. doi: 10.1038/emboj.2008.275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Doench JG, Petersen CP, Sharp PA. siRNAs can function as miRNAs. Genes & Development. 2003;17:438–442. doi: 10.1101/gad.1064703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Eulalio A, Huntzinger E, Izaurralde E. GW182 interaction with Argonaute is essential for miRNA-mediated translational repression and mRNA decay. NatureStructural & Molecular Biology. 2008;15:346–353. doi: 10.1038/nsmb.1405. [DOI] [PubMed] [Google Scholar]
  10. Farny NG, Hurt JA, Silver PA. Definition of global and transcript-specific mRNA export pathways in metazoans. Genes & Development. 2008;22:66–78. doi: 10.1101/gad.1616008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Feng Y, Press B, Wandinger-Ness A. Rab 7: an important regulator of late endocytic membrane traffic. The Journal of Cell Biology. 1995;131:1435–1452. doi: 10.1083/jcb.131.6.1435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fukaya T, Tomari Y. MicroRNAs mediate gene silencing via multiple different pathways in Drosophila. Molecular Cell. 2012;48:825–836. doi: 10.1016/j.molcel.2012.09.024. [DOI] [PubMed] [Google Scholar]
  13. Gibbings DJ, Ciaudo C, Erhardt M, Voinnet O. Multivesicular bodies associate with components of miRNA effector complexes and modulate miRNA activity. Nature Cell Biology. 2009;11:1143–1149. doi: 10.1038/ncb1929. [DOI] [PubMed] [Google Scholar]
  14. Graham J, Ford T, Rickwood D. The preparation of subcellular organelles from mouse liver in self-generated gradients of iodixanol. Analytical Biochemistry. 1994;220:367–373. doi: 10.1006/abio.1994.1351. [DOI] [PubMed] [Google Scholar]
  15. Griner EM, Kazanietz MG. Protein kinase C and other diacylglycerol effectors in cancer. Nature Reviews Cancer. 2007;7:281–294. doi: 10.1038/nrc2110. [DOI] [PubMed] [Google Scholar]
  16. Hannun YA, Obeid LM. Principles of bioactive lipid signalling: lessons from sphingolipids. Nature Reviews Molecular Cell Biology. 2008;9:139–150. doi: 10.1038/nrm2329. [DOI] [PubMed] [Google Scholar]
  17. Horman SR, Janas MM, Litterst C, Wang B, Macrae IJ, Sever MJ, Morrissey DV, Graves P, Luo B, Umesalma S, et al. Akt-mediated phosphorylation of Argonaute 2 downregulates cleavage and upregulates translational repression of microRNA targets. Molecular Cell. 2013;50:356–367. doi: 10.1016/j.molcel.2013.03.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Humphreys DT, Westman BJ, Martin DI, Preiss T. MicroRNAs control translation initiation by inhibiting eukaryotic initiation factor 4E/cap and poly(A) tail function. Proceedings of the National Academy of Sciences of the United States of America. 2005;102:16961–16966. doi: 10.1073/pnas.0506482102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Iwasaki S, Kawamata T, Tomari Y. Drosophila Argonaute1 and Argonaute2 employ distinct mechanisms for translational repression. Molecular Cell. 2009;34:58–67. doi: 10.1016/j.molcel.2009.02.010. [DOI] [PubMed] [Google Scholar]
  20. Lee YS, Pressman S, Andress AP, Kim K, White JL, Cassidy JJ, Li X, Lubell K, Lim do H, Cho IS, et al. Silencing by small RNAs is linked to endosomal trafficking. Nature Cell Biology. 2009;11:1150–1156. doi: 10.1038/ncb1930. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Leung AK, Sharp PA. MicroRNA functions in stress responses. Molecular Cell. 2010;40:205–215. doi: 10.1016/j.molcel.2010.09.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Liu H, Adler AS, Segal E, Chang HY. A transcriptional program mediating entry into cellular quiescence. PLoS Genet. 2007;3:e91. doi: 10.1371/journal.pgen.0030091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Maroney PA, Yu Y, Fisher J, Nilsen TW. Evidence that microRNAs are associated with translating messenger RNAs in human cells. Nature Structural & Molecular Biology. 2006;13:1102–1107. doi: 10.1038/nsmb1174. [DOI] [PubMed] [Google Scholar]
  24. Marques JT, Kim K, Wu PH, Alleyne TM, Jafari N, Carthew RW. Loqs and R2D2 act sequentially in the siRNA pathway in Drosophila. Nature Structural & Molecular Biology. 2010;17:24–30. doi: 10.1038/nsmb.1735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Miyoshi K, Okada TN, Siomi H, Siomi MC. Characterization of the miRNA-RISC loading complex and miRNA-RISC formed in the Drosophila miRNA pathway. RNA. 2009;15:1282–1291. doi: 10.1261/rna.1541209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Newton AC. Protein kinase C: structure, function, and regulation. The Journal of Biological Chemistry. 1995;270:28495–28498. doi: 10.1074/jbc.270.48.28495. [DOI] [PubMed] [Google Scholar]
  27. Nottrott S, Simard MJ, Richter JD. Human let-7a miRNA blocks protein production on actively translating polyribosomes. Nature Structural & Molecular Biology. 2006;13:1108–1114. doi: 10.1038/nsmb1173. [DOI] [PubMed] [Google Scholar]
  28. Paroo Z, Ye X, Chen S, Liu Q. Phosphorylation of the human microRNA-generating complex mediates MAPK/Erk signaling. Cell. 2009;139:112–122. doi: 10.1016/j.cell.2009.06.044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Petersen CP, Bordeleau ME, Pelletier J, Sharp PA. Short RNAs repress translation after initiation in mammalian cells. Molecular Cell. 2006;21:533–542. doi: 10.1016/j.molcel.2006.01.031. [DOI] [PubMed] [Google Scholar]
  30. Pillai RS, Bhattacharyya SN, Artus CG, Zoller T, Cougot N, Basyuk E, Bertrand E, Filipowicz W. Inhibition of translational initiation by Let-7 MicroRNA in human cells. Science. 2005;309:1573–1576. doi: 10.1126/science.1115079. [DOI] [PubMed] [Google Scholar]
  31. Puig O, Marr MT, Ruhf ML, Tjian R. Control of cell number by Drosophila FOXO: downstream and feedback regulation of the insulin receptor pathway. Genes & Development. 2003;17:2006–2020. doi: 10.1101/gad.1098703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Seggerson K, Tang L, Moss EG. Two genetic circuits repress the Caenorhabditis elegans heterochronic gene lin-28 after translation initiation. Developmental Biology. 2002;243:215–225. doi: 10.1006/dbio.2001.0563. [DOI] [PubMed] [Google Scholar]
  33. Sheff DR, Daro EA, Hull M, Mellman I. The receptor recycling pathway contains two distinct populations of early endosomes with different sorting functions. The Journal of Cell Biology. 1999;145:123–139. doi: 10.1083/jcb.145.1.123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Soeiro R, Amos H. Arrested protein synthesis in polysomes of cultured chick embryo cells. Science. 1966;154:662–665. doi: 10.1126/science.154.3749.662. [DOI] [PubMed] [Google Scholar]
  35. Thermann R, Hentze MW. Drosophila miR2 induces pseudo-polysomes and inhibits translation initiation. Nature. 2007;447:875–878. doi: 10.1038/nature05878. [DOI] [PubMed] [Google Scholar]
  36. Thomas G, Siegmann M, Kubler AM, Gordon J, Jimenez de Asua L. Regulation of 40S ribosomal protein S6 phosphorylation in Swiss mouse 3T3 cells. Cell. 1980;19:1015–1023. doi: 10.1016/0092-8674(80)90092-6. [DOI] [PubMed] [Google Scholar]
  37. Vasudevan S, Steitz JA. AU-rich-element-mediated upregulation of translation by FXR1 and Argonaute 2. Cell. 2007;128:1105–1118. doi: 10.1016/j.cell.2007.01.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Vasudevan S, Tong Y, Steitz JA. Switching from repression to activation: microRNAs can up-regulate translation. Science. 2007;318:1931–1934. doi: 10.1126/science.1149460. [DOI] [PubMed] [Google Scholar]
  39. Zdanowicz A, Thermann R, Kowalska J, Jemielity J, Duncan K, Preiss T, Darzynkiewicz E, Hentze MW. Drosophila miR2 primarily targets the m7GpppN cap structure for translational repression. Molecular Cell. 2009;35:881–888. doi: 10.1016/j.molcel.2009.09.009. [DOI] [PubMed] [Google Scholar]
  40. Zhou R, Hotta I, Denli AM, Hong P, Perrimon N, Hannon GJ. Comparative analysis of argonaute-dependent small RNA pathways in Drosophila. Molecular Cell. 2008;32:592–599. doi: 10.1016/j.molcel.2008.10.018. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS518048-supplement-01.doc (773.5KB, doc)

RESOURCES