smaug

REGULATION

Post-transcriptional regulation of nanos

Proper deployment of Nanos protein at the posterior of the Drosophila embryo, where it directs posterior development, requires a combination of RNA localization and translational controls. These controls ensure that only the posteriorly-localized nanos mRNA is translated, whereas unlocalized nanos mRNA is translationally repressed. This study describes cloning of the gene encoding Smaug, an RNA-binding protein that interacts with the sequences, SREs, in the nanos mRNA that mediate translational repression. Using an in vitro translation assay, it is demonstrated that SRE-dependent repression occurs in extracts from early stage embryos. Immunodepletion of Smaug from the extracts eliminates repression, consistent with the notion that Smaug is involved. Smaug is a novel gene and the existence of potential mammalian Smaug homologs raises the possibility that Smaug represents a new class of conserved translational repressor (Smibert, 1999; full text of article).

The Nanos translational control element (TCE) function requires formation of a bipartite secondary structure that is recognized by Smaug repressor and at least one additional factor. Translational activation requires the interaction of localization factors with sequences that overlap TCE structural motifs. The identification of separate but overlapping recognition motifs for translational repressors and localization factors provides a molecular mechanism for the switch between translational repression and activation (Crucs, 2000).

Translational repression of NOS mRNA is mediated by a bipartite cis-acting translational control element consisting of two stem-loop structures. One stem-loop, stem-loop II, is the target for binding by the previously identified Smg protein. The other, which is not required for Smg interaction but is essential for translational repression, contains a unique double-stranded recognition motif. The essential requirement for both of these stem-loops indicates that binding by one or more factors in addition to Smg is necessary for TCE-mediated translational repression. Translational activation of NOS mRNA requires recognition by localization factors of NOS 3'UTR motifs distinct from, but overlapping those that mediate translational repression. This arrangement of recognition motifs can explain the ability of a single region of the NOS 3'UTR to mediate these two mutually exclusive functions. These results provide direct molecular evidence for a model in which translational activation of nos RNA results when binding of translational repressors is prevented by binding of localization factors at distinct but overlapping sites (Crucs, 2000).

NOS TCE function requires formation in vivo of secondary structure predicted both by phylogenetic conservation and by RNA folding algorithms. Two stem-loops make unique contributions to TCE function by providing distinct recognition motifs for trans-acting translational repressors. The requirement for formation of stem-loop II can be explained entirely by its role in presenting a single-stranded binding site for Smg protein. The Smg RNA-binding domain does not interact with stem-loop III. By contrast, the specific sequence as well as structural requirements for stem-loop III activity suggest that stem-loop III may be recognized by a double-stranded RNA-binding protein. One such protein, Staufen (Stau), is required for localization of BCD RNA and an indirect in vivo assay suggests that Stau interacts with a double-stranded region of the BCD RNA localization signal. Another double-stranded RNA-binding protein, Prbp, binds to a conserved stem-loop in the 3'UTR of the mouse protamine 1 (Prm1) mRNA in vitro and is required in vivo for translational activation of Prm1 during spermatogenesis (Crucs, 2000).

Stau and Prbp belong to a large family of proteins that bind double-stranded RNA through a conserved amino acid motif, the dsRBD. Biochemical and structural studies of dsRBD family members indicate that binding is highly specific for double-stranded RNA but largely independent of RNA sequence. Since the function of stem-loop III requires alternating U:A and A:U base pairs, recognition of the helix appears to be sequence-dependent. It is possible, therefore, that TCE stem-loop III is recognized by a double-stranded RNA-binding protein with a novel sequence-specific binding domain. Alternatively, local sequence-dependent structural variation may confer specificity to an interaction with a more typical ds-RBD. The partial reduction in TCE function caused by replacement of the poorly conserved III loop with a helix-stabilizing tetraloop further suggests that TCE function is sensitive to the geometry of the stem III helix (Crucs, 2000).

The ability of TCE stem-loops II and III to repress translation when separated by 52 nucleotides argues that coaxial stacking of the stem II and III helices, potentiated by the phylogenetically conserved junctional structure, is not required for TCE function. Furthermore, the junction does create a binding site for translational repressors. However, the ability of a bulged nucleotide within the stem III helix to disrupt TCE function suggests that the position of the distal half of the stem III helix with respect to the stem II helix is indeed critical. It is possible, therefore, that translational repression requires interaction between factors bound to stem-loops II and III or between these factors and a third component. When the two stem-loops are separated in a rearranged variant, increased flexibility in the RNA molecule may continue to allow this interaction (Crucs, 2000).

Posterior localization and translational activation can be largely recapitulated by the 180 nucleotide +2 element of the NOS RNA localization signal. This element can be further subdivided into the +1 element, which is coincident with the TCE, and the +2' element. While the +2' element also contains a Smg-binding site, it confers minimal translational repression. On their own, the +1 and +2' elements each provide weak localization function. However, when combined to form the +2 element, they collaborate to produce substantial localization function. RNA localization function of the +1 element is highly sensitive to mutation. All mutations analyzed here, including compensatory mutations and the single U51A mutation that retains TCE function, eliminate localization by the +1 element. Thus, localization requires recognition of multiple sequence and possibly structural motifs distributed throughout this element. The possibility that localization requires formation of an alternate secondary structure cannot be eliminated. While all suboptimal structures predicted by RNA folding algorithms closely resemble the optimal structure and cannot explain the effect of these mutations on localization function, it is possible that binding by localization factors stabilizes an alternate structure with a low propensity to form on its own (Crucs, 2000).

A synergistic interaction between the +1 and +2' elements permits the +2 element to recapitulate much of wild-type NOS localization function. The effect of TCE/+1 element mutations on localization by the +2 element indicates that the TCE/+1 sequences are recognized in two different ways, perhaps by two different proteins or protein complexes. While each recognition event corresponds to one TCE stem-loop, it is separable from the recognition event required for the translational repression function of that stem-loop. Presumably, contacts between proteins bound to these +1 element sites and proteins bound to the +2' element or contacts between +1 and +2' element binding proteins and a common component, such as a complex of Osk, Vas, and Tud proteins, underlie the interaction (Crucs, 2000).

Biochemical experiments have shown that Osk protein interacts directly with Smg, although the functional significance of this interaction is not known. A recently proposed model, in which a ternary complex of Osk, Smg, and the TCE serves both to localize NOS RNA to the posterior and to activate NOS translation (Dahanukar, 1999), is not supported by the data presented here. Rather, these results indicate that direct interaction of localization factors with NOS 3'UTR sequences is obligate for translational activation. The recruitment of NOS mRNA, via factors bound to localization element sites, to a localization complex with Osk, Vas, and Tud proteins may facilitate a Smg:Osk interaction that inactivates Smg function. However, the close apposition or overlap of localization factor recognition sites to the motifs required for translational repressor recognition supports an alternative model in which binding of translational repressors and localization factors is mutually exclusive. In this model, the Smg:Osk interaction may serve as a fail-safe mechanism to sequester Smg and inhibit its activity. Given the structural requirements for TCE function, binding by localization factors to motifs within stem-loops II and III may prevent binding of repressors by preventing formation of TCE structure. Thus, the structure of the TCE may lie at the heart of the switch between translational repression and activation (Crucs, 2000).

Temporal complexity within a translational control element in the nanos mRNA by Smaug

Translational control of gene expression plays a fundamental role in the early development of many organisms. In Drosophila, selective translation of nanos mRNA localized to the germ plasm at the posterior of the embryo, together with translational repression of nanos in the bulk cytoplasm, is essential for development of the anteroposterior body pattern. Both these components of spatial control of nanos translation initiate during oogenesis and translational repression is initially independent of Smaug, an embryonic repressor of nanos. Repression during oogenesis and embryogenesis are mediated by distinct stem loops within the nanos 3' untranslated region; the Smaug-binding stem-loop acts strictly in the embryo, whereas a second stem-loop functions in the oocyte. Thus, independent regulatory modules with temporally distinct activities contribute to spatial regulation of nanos translation. It is proposed that nanos evolved to exploit two different stage-specific translational regulatory mechanisms (Forrest, 2004).

Selective translation of posteriorly localized nos mRNA achieves the restricted distribution of Nos protein required to regulate hb and cycB mRNAs in the posterior of the embryo without affecting hb or bcd translation in the anterior. Although some maternal mRNAs, including osk and gurken are translated in the oocyte, others, such as hb and bcd, are translationally repressed during oogenesis and activated only after fertilization. For nos, translational activation during oogenesis would permit accumulation of Nos protein in the posterior of the embryo early enough to block hb translation. However, the translational status of nos in late stage oocytes has remained enigmatic, owing to the impermeability of these tissues to immunostaining. Use of a gfp-nos fusion gene as a reporter for nos translation has allowed this issue to be addressed; it has been established that translation of localized nos mRNA indeed begins during oogenesis (Forrest, 2004).

Achieving the restricted Nos protein distribution in the embryo requires that translational activity of nos in the oocyte be spatially regulated. However, the only known repressor of nos translation, Smg, is absent from the ovary. This dilemma has been solved by showing that a distinct, Smg-independent mechanism mediates translational repression of unlocalized nos mRNA in late oocytes. Translational repression of unlocalized nos is mediated by a 90 nucleotide translational control element (TCE), the function of which requires formation of two stem-loops (II and III). Failure to repress nos in late oocytes, as exemplified by the behavior of the nos-tub:TCEIIIA transgene (mutated in stem-loop III of the RNA), results in unrestricted production of Nos protein that perdures to embryogenesis. The resulting embryos die, lacking anterior structures. Thus, by showing that the program for spatially restricted synthesis of Nos operates during oogenesis, these results reveal how temporal demands are reconciled with spatial constraints on nos translation needed for embryonic patterning (Forrest, 2004).

The elucidation of temporally distinct functions of the two TCE stem-loops explains the enigmatic structural complexity of this regulatory element. Both stem-loops (II and III) retain function, despite their separation by a 52-nucleotide spacer, suggesting that they operate independently. Indeed, phylogenetic analysis of the nos 3'UTR reveals that TCE stem-loops II and III are not juxtaposed in all Drosophilid species, indicating that the distance between stem II and III is not under tight evolutionary constraint (Forrest, 2004).

After fertilization, Smg binds to TCE stem-loop II to mediate repression in the preblastoderm embryo. It is not known if the ovarian repressor remains bound to stem-loop III in the embryo, but its function is superceded by Smg. A minor requirement for stem-loop III in the embryo suggested by in vitro translation experiments may reflect the need to maintain the ovarian repression mechanism until Smg reaches sufficient levels in the embryo. Accordingly, the requirement for stem-loop III would decrease over time after fertilization. A more significant contribution by stem-loop III might have been missed, however, if the stem-loop III-dependent repressor is unstable in the embryonic extract (Forrest, 2004).

The smg mutant phenotype indicates that nos is not the only target of Smg in the embryo. Although the ovarian repressor has not yet been identified, a candidate ovarian stem-loop III binding factor has been isolated recently that appears to regulate multiple maternal mRNAs. Thus, it seems likely that nos has evolved to co-opt existing stage-specific regulatory proteins for its advantage. The nos TCE can repress translation in subsets of cells in both the central and peripheral nervous systems. Although the repressors are not known in these cases either, it is possible that the ability to interact with yet additional proteins will underlie the multifunctionality of the TCE (Forrest, 2004).

Other RNAs may use a similar strategy of recognition by stage-specific factors to maintain translational regulation across developmental transitions. In the Drosophila oocyte, translational repression of unlocalized osk mRNA occurs through the interaction of Bruno (Bru) with specific sequence motifs in the osk 3'UTR. Since Bru is not present in the embryo, where the majority of osk mRNA remains unlocalized, an embryonic repressor may be required to maintain the repression initiated by Bru. Intriguingly, the existence of binding sites for multiple, distinct microRNAs within individual 3'UTRs suggests a similar paradigm for controlling translation through multiple developmental stages or in different tissues (Forrest, 2004).

The translational quiescence of unlocalized nos in late oocytes contrasts sharply with its translational activity in the nurse cells. Deposition of both actively translated nos mRNA and the previously synthesized Nos protein into the oocyte during nurse cell dumping presents a challenge for restricting Nos to the posterior of the oocyte. Although it cannot be determined whether nos is repressed in oocytes prior to stage 10, the results indicate that the majority of nos RNA, which enters the oocyte during dumping, must switch from a translationally active state in the nurse cells to an inactive state in the oocyte. This switch could be mediated by interaction of nos with an ovarian repressor restricted to the oocyte. Alternatively, a repressor may bind to nos RNA in the nurse cells, but become activated during or after passage into the oocyte (Forrest, 2004).

Translationally repressed nos RNA is associated with polysomes, indicating that repression is imposed at a late step in the translation cycle. However, recent evidence that Smg interacts with Cup to prevent recruitment of eIF-4G by eIF-4E suggests that translation is blocked at the initiation step. The identification of a Smg-independent mechanism for translational repression during oogenesis may explain these divergent results. Indeed, a post-initiation mechanism may be ideally suited to rapidly repress polysomal nos RNA entering the oocyte from the nurse cells (Forrest, 2004).

In addition to translationally active nos RNA, substantial amounts of Nos protein enter the oocyte during nurse cell dumping. Perdurance of this protein up to embryogenesis would probably disrupt anterior development. Nos protein entering the oocyte from stage 10 nurse cells is cleared from the oocyte by stage 13. Nos protein made in the nurse cells may therefore be specifically targeted for degradation. Alternatively, Nos might have a short half-life regardless of its site of synthesis. Despite considerable effort, no ubquitinated forms of Nos protein have been detected, although the transient nature of ubquitinated intermediates may preclude detection. Similarly, a genetic interaction has not been detected between mutations in numerous components of the ubiquitin degradation pathway and nos transgenes. Thus, how Nos is degraded remains an unanswered question. Regardless of mechanism, however, continuous translation of wild-type nos RNA at the posterior pole or of unlocalized nos-tub3'UTR RNA throughout the oocyte would result in accumulation of Nos protein. Thus post-translational control of Nos protein stability as well as post-transcriptional regulation of nos RNA contribute to the correct spatial distribution of Nos in the early embryo (Forrest, 2004).

Rapid ATP-dependent deadenylation of nanos mRNA in a cell-free system from Drosophila embryos

Shortening of the poly(A) tail (deadenylation) is the first and often rate-limiting step in the degradation pathway of most eukaryotic mRNAs and is also used as a means of translational repression, in particular in early embryonic development. The nanos mRNA is translationally repressed by the protein Smaug in Drosophila embryos. The RNA has a short poly(A) tail at steady state and decays gradually during the first 2-3 h of development. Smaug has also been implicated in mRNA deadenylation. To study the mechanism of sequence-dependent deadenylation, a cell-free system was developed from Drosophila embryos that displays rapid deadenylation of nanos mRNA. The Smaug response elements contained in the nanos 3'-untranslated region are necessary and sufficient to induce deadenylation; thus, Smaug is likely to be involved. Unexpectedly, deadenylation requires the presence of an ATP regenerating system. The activity can be pelleted by ultracentrifugation, and both the Smaug protein and the CCR4.NOT complex, a known deadenylase, are enriched in the active fraction. The same extracts show pronounced translational repression mediated by the Smaug response elements. RNAs lacking a poly(A) tail are poorly translated in the extract; therefore, SRE-dependent deadenylation contributes to translational repression. However, repression is strong even with RNAs either bearing a poly(A) tract that cannot be removed or lacking poly(A) altogether; thus, an additional aspect of translational repression functions independently of deadenylation (Jeske, 2006; full text of article).

Smaug recruits the CCR4/POP2/NOT deadenylase complex to trigger maternal transcript localization in the early Drosophila embryo

Asymmetric localization of mRNAs within cells promotes precise spatio-temporal control of protein synthesis. Although cytoskeletal transport-based localization during Drosophila oogenesis is well characterized, little is known about the mechanisms that operate to localize maternal RNAs in the early embryo. One such mechanism—termed “degradation/protection”—acts on maternal Hsp83 transcripts, removing them from the bulk cytoplasm while protecting them in the posterior pole plasm. The RNA binding protein, Smaug, previously known as a translational repressor of nanos, has been identified as a key regulator of degradation/protection-based transcript localization. In smaug mutants, degradation of Hsp83 transcripts is not triggered, and, thus, localization does not occur. Hsp83 transcripts are in an mRNP complex containing Smaug, but Smaug does not translationally repress Hsp83 mRNA. Rather, Smaug physically interacts with the CCR4/POP2/NOT deadenylase, recruiting it to Hsp83 mRNA to trigger transcript deadenylation and degradation. When Smaug is targeted to heterologous stable reporter transcripts in vivo, these are deadenylated and destabilized. A deletion that removes the gene encoding CCR4 exhibits dose-sensitive interactions with Smaug in both a loss-of-function and a gain-of-function context. Reduction of CCR4 protein levels compromises Hsp83 transcript destabilization. It is concluded that Smaug triggers destabilization and localization of specific maternal transcripts through recruitment of the CCR4/POP2/NOT deadenylase. In contrast, Smaug-mediated translational repression is accomplished via an indirect interaction between Smaug and eIF4E, a component of the basic translation machinery. Thus, Smaug is a multifunctional posttranscriptional regulator that employs distinct mechanisms to repress translation and to induce degradation of target transcripts (Semotok, 2005; full text of article).

The Drosophila PNG kinase complex regulates the translation of Cyclin B; Png regulates smaug translation in a poly(A)-dependent and -independent manner

The Drosophila Pan Gu (Png) kinase complex regulates the developmental translation of cyclin B. cyclin B mRNA becomes unmasked during oogenesis independent of Png activity, but Png is required for translation from egg activation. Although polyadenylation of cyclin B augments translation, it is not essential, and a fully elongated poly(A) is not required for translation to proceed. In fact, changes in poly(A) tail length are not sufficient to account for Png-mediated control of cyclin B translation and of the early embryonic cell cycles. Evidence is presented that Png functions instead as an antagonist of Pumilio-dependent translational repression. The data argue that changes in poly(A) tail length are not a universal mechanism governing embryonic cell cycles, and that Png-mediated derepression of translation is an important alternative mechanism in Drosophila (Vardy, 2007).

These data indicate that in Drosophila translation of cyclin B can proceed in the absence of polyadenylation in ovaries and syncytial embryos. Polyadenylation is not required for the unmasking of the mRNA, but it may play a role in fine-tuning translation. A model is proposed in which Png has dual roles—poly(A) dependent and independent—in promoting cyclin B translation. At egg activation, Png is needed for full-length poly(A) tails, and this may augment translation. The data indicate, however, that Png also promotes translation independently of poly(A) tail length, most likely by overcoming repressive action by Pum. Removal of Pum in a png mutant therefore allows translation of the target mRNA even if the poly(A) tails are not fully elongated (Vardy, 2007).

SMAUG (SMG) protein levels and the poly(A) tail are decreased in the png mutant, and restoration of poly(A) tail length by overexpressing poly(A) polymerase is not enough to promote smaug translation in a png mutant. Thus, Png also appears to regulate smaug translation in a poly(A)-dependent and -independent manner. This indicates that multiple levels of translational regulation may be a common strategy in development (Vardy, 2007).

If the Png kinase complex functions ultimately to regulate the activity of Pum, one of the key questions concerns the nature of this regulation. Since NANOS is found predominantly at the posterior pole, Pum may be able to recruit different factors to the target mRNA. The defect seen in png is likely due to a failure to relieve translational repression as opposed to a failure in activation, because injection of excess amounts (1 μg/μl) of the 5′FF3′−cycB transcript into png mutant embryos allows translation to proceed at much stronger levels. This suggests that the repressor is limiting in this reaction, thus allowing translation to proceed, and is consistent with Pum being downstream of Png. It will be interesting to determine the nature of the Png kinase substrates and how they relate to Pum function (Vardy, 2007).

Png likely has multiple targets, because removal of Pum in a png mutant does not completely restore the embryonic divisions. While the early syncytial cycles can progress in a png;pum double mutant, mitosis still fails in the later S-M cycles. Smg protein levels are decreased in the png mutant due to a failure in translation. Smg protein levels are not restored in a png;pum double mutant, and given Smg's role in progression through cycle 11/12, it provides an explanation for why the embryos fail at this stage. An independent role for the Png kinase complex later in embryogenesis has been described in the maternal mRNA degradation pathway. Although these pathways appear to be independent, in each case translational regulation of a key component appears to be the mechanism: Cyclin B to control the S-M cycles and SMG to control maternal transcript destruction (Vardy, 2007).

This study has established a role for the Png kinase complex in the translational regulation of cyclin B in the early syncytial cycles of Drosophila. A role for the Png kinase complex in protein stability has been suggested, and it is possible that it could act to regulate Cyclin B levels by multiple mechanisms. These data, however, reveal it to be a key regulator of cyclin B translation, ensuring coordinated passage through these early cycles (Vardy, 2007).

The Smaug RNA-binding protein is essential for microRNA synthesis during the Drosophila maternal-to-zygotic transition

Metazoan embryos undergo a maternal-to-zygotic transition (MZT) during which maternal gene products are eliminated and the zygotic genome becomes transcriptionally active. During this process RNA-binding proteins (RBPs) and the microRNA-induced silencing complex (miRISC) target maternal mRNAs for degradation. In Drosophila, the Smaug (SMG), Brain tumor (BRAT) and Pumilio (PUM) RBPs bind to and direct the degradation of largely distinct subsets of maternal mRNAs. SMG has also been shown to be required for zygotic synthesis of mRNAs and several members of the miR-309 family of microRNAs (miRNAs) during the MZT. This study carried out global analysis of small RNAs both in wild type and in smg mutants. It was found that 85% all miRNA species encoded by the genome are present during the MZT. Whereas loss of SMG has no detectable effect on Piwi-interacting RNAs (piRNAs) or small interfering RNAs (siRNAs), zygotic production of more than 70 species of miRNAs fails or is delayed in smg mutants. SMG is also required for the synthesis and stability of a key miRISC component, Argonaute 1 (AGO1), but plays no role in accumulation of the Argonaute-family proteins associated with piRNAs or siRNAs. In smg mutants, maternal mRNAs that are predicted targets of the SMG-dependent zygotic miRNAs fail to be cleared. BRAT and PUM share target mRNAs with these miRNAs but not with SMG itself. The study hypothesizes that SMG controls the MZT, not only through direct targeting of a subset of maternal mRNAs for degradation but, indirectly, through production and function of miRNAs and miRISC, which act together with BRAT and/or PUM to control clearance of a distinct subset of maternal mRNAs (Luo, 2016).

To identify small RNA species expressed during the Drosophila MZT and to assess the role of SMG in their regulation 18 small-RNA libraries were produced and sequenced: nine libraries from eggs or embryos produced by wild-type females and nine from smg-mutant females. The 18 libraries comprised three biological replicates each from the two genotypes and three time-points: (1) 0-to-2 hour old unfertilized eggs, in which zygotic transcription does not occur and thus only maternally encoded products are present; (2) 0-to-2 hour old embryos, the stage prior to large-scale zygotic genome activation; and (3) 2-to-4 hour old embryos, the stage after to large-scale zygotic genome activation. After pre-alignment processing, a total of ~144 million high quality small-RNA reads was obtained and 110 million of these perfectly matched the annotated Drosophila genome (Luo, 2016).

Loss of SMG had no significant effect on piRNAs and siRNAs, or on the Argonaute proteins associated with those small RNAs: Piwi, Aubergine (AUB), AGO3, and AGO2, respectively. In contrast, loss of SMG resulted in a dramatic, global reduction in miRNA populations during the MZT as well as reduced levels of AGO1, the miRISC-associated Argonaute protein in Drosophila (Luo, 2016).

A pre-miRNA can generate three types of mature miRNA: (1) a canonical miRNA, which has a perfect match to the annotated mature miRNA; (2) a non-canonical miRNA, which shows a perfect match to the annotated mature miRNA but with additional nucleotides at the 5'- or 3'- end that match the adjacent primary miRNA sequence, and (3) a miRNA with non-templated terminal nucleotide additions (an NTA-miRNA), which has nucleotides at its 3'-end that do not match the primary miRNA sequence (Luo, 2016).

In these libraries a total of 364 distinct miRNA species were identified that mapped to miRBase, comprising 85% (364/426) of all annotated mature miRNA species in Drosophila. Thus, the vast majority of all miRNA species encoded by the Drosophila genome are expressed during the MZT. Overall, in wild type, an average of 75% of all identified miRNAs fell into the canonical category. The remaining miRNAs were either non-canonical (10%) or NTA-miRNAs (15%) (Luo, 2016).

To validate these sequencing results, those mature miRNA species identified in the data that perfectly matched the Drosophila genome sequence (i.e., canonical and non-canonical) were compared with a previously published miRNA dataset from 0 to 6 hour old embryos. To avoid differences caused by miRBase version, data sets from previous study were remapped to miRBase Version 19 and f99% of their published miRNA species were found to be on the miRNA list (176/178 mature miRNA species comprising 161 canonical miRNA s and 94 non-canonical miRNA s) . There were an additional 181 mature miRNA species in the library that had not been identified as expressed in early embryos in the earlier study (Luo, 2016).

As a second validation, the list of maternally expressed miRNA species (those present in the 0-to-2 hour wild-type unfertilized egg samples) were compared with the most recently published list of maternal miRNAs, which had been defined in the same manner. 99% of the 86 published maternal miRNA species were on this study's maternal miRNA list (85/86). An additional 144 maternal miRNA species in the library were identified that had not been observed in the previous study. Identification of a large number of additional miRNA species in unfertilized eggs and early embryos can be attributed to the depth of coverage of the current study. The current dataset, therefore, provides the most complete portrait to date of the miRNAs present during the Drosophila MZT (Luo, 2016).

Next, global changes in miRNA species during the MZT were analyzed in wild-type embryos. A dramatic increase was observed in the proportion of miRNAs relative to other small RNAs that was due to an increase in absolute miRNA amount rather than a decrease in the amount of other types of small RNAs. In wild-type 0-to-2 hour unfertilized eggs, the proportion of the small RNA libraries comprised of canonical and non-canonical miRNAs was 12.8%. These represent maternally loaded miRNAs since unfertilized eggs do not undergo zygotic genome activation. The proportion of small RNAs represented by miRNAs increased dramatically during the MZT, reaching 50.7% in 2-to-4 hour embryos. The other abundant classes of small RNAs underwent either no change or relatively minor changes over the same time course. It is concluded that there is a large amount of zygotic miRNA synthesis during the MZT in wild-type embryos (Luo, 2016).

For more detailed analysis of the canonical, non-canonical and NTA isoforms focus was placed on 154 miRNA species that possessed an average of > 10 reads per million (RPM) for all three isoform types in one or more of the six sample sets. A focus was placed on changes in wild type. Among all miRNAs, in wild type the proportion of canonical isoforms increased over the time-course from 69% to 83%, the proportion of non-canonical miRNAs remained constant (from 9% to 10%) , and the proportion of the NTA-miRNAs decreased (from 22% to 7%). These results derive from the fact that, during the MZT, the vast majority of newly synthesized miRNAs were canonical, undergoing a more than seven-fold increase from 103,105 to 744,043 RPM; that non-canonical miRNAs underwent a comparable, nearly seven-fold, increase from 13,902 to 92,199; whereas NTA-miRNAs underwent a less than two-fold increase, from 32,840 to 63,847, thus decreasing in relative proportion (Luo, 2016).

Whereas the proportion of the small-RNA population that was comprised of miRNAs increased fourfold over the wild-type time-course, concomitant with increases in overall miRNA abundance, there was no such increase in the smg mutant embryos: 21.9% of the small RNAs were miRNAs in 0-to-2 hour unfertilized smg mutant eggs (mean RPM = 203,415) and 20.5% (mean RPM = 196,110) were miRNAs in 2-to-4 hour smg mutant embryos (Luo, 2016).

This difference between wild type and smg mutants could have resulted from the absence of a small number of extremely highly expressed miRNA species in the mutant. Alternatively, it may have been a consequence of a widespread reduction in the levels of all or most zygotically synthesized miRNAs in smg mutants. To assess the cause of this difference, canonical miRNA reads were graphed in scatter plots. These showed that a large number of miRNA species had significantly reduced expression levels in 0-to-2 and in 2-to-4 hour smg-mutant embryos relative to wild type. Most of the down-regulated miRNA species exhibited a more than four-fold reduction in abundance relative to wild type. Furthermore, this reduction occurred for miRNA species expressed over a wide range of abundances in wild type (Luo, 2016).

Box plots were then used to analyze the canonical, non-canonical and NTA isoforms of the 154 miRNA species identified in the previous section. These showed that, in wild type, the median abundance of canonical, non-canonical and 3' NTA miRNAs increased significantly in 0-to-2 and in 2-to-4 hour embryos relative to 0-to-2 hour unfertilized eggs. In contrast, there was no significant increase in the median abundance of any of the three isoforms of miRNAs in the smg-mutant embryos. Also for all three isoform types, when each time point was compared between wild type and smg mutant, there was no difference between wild type and mutant in 0-to-2 hour unfertilized eggs but there was a highly significant difference between the two genotypes at both of the embryo time-points. Whereas the abundance of miRNAs differed between wild-type and mutant embryos, there was no difference in length or first-nucleotide distribution of canonical miRNAs, nor in the non-templated terminal nucleotides added to NTA-miRNAs (Luo, 2016).

As described above, during the wild-type MZT canonical miRNAs comprised the major isoform that was present (69% to 83% of miRNAs). It was next asked whether miRNA species could be categorized into different classes based on their expression profiles during the wild-type MZT. 131 canonical miRNA species that had > 10 mean RPM in at least one of the six datasets were analyzed. Hierarchical clustering of their log 2 RPM values identified five distinct categories of canonical miRNA species during the MZT. The effects of smg mutations on each of these classes were analyzed (Luo, 2016).

The data are consistent with a model in which SMG degrades its direct targets without the assistance of miRNAs whereas a large fraction of the indirectly affected maternal mRNAs in smg mutants fails to be degraded by virtue of being targets of zygotically produced miRNA species that are either absent or present at significantly reduced levels in smg mutants. Thus, SMG is required both for early, maternally encoded decay and for late, zygotically encoded decay. In the former case SMG is a key specificity component that directly binds to maternal mRNAs; in the latter case SMG is required for the production of the miRNAs (and AGO1 protein) that are responsible for the clearance of an additional subset of maternal mRNAs (Luo, 2016).

In Drosophila, the stability of miRNAs is enhanced by AGO1 and vice versa. Since miRNA levels are dramatically reduced in smg mutants, Ago1 mRNA and AGO1 protein levels were assessed during the MZT both in wild type and in smg mutants. In wild type, AGO1 levels were low in unfertilized eggs and 0-to-2 hour embryos but then increased substantially in 2-to-4 hour embryos. These western blot data are consistent with an earlier, proteomic, study that reported a more than three-fold increase in AGO1 in embryos between 0-to-1.5 hours and 3-to-4.5 hours. In contrast to AGO1 protein, it was found using RT-qPCR that Ago1 mRNA levels remained constant during the MZT. Taken together with a previous report that Ago1 mRNA is maternally loaded, the increase in AGO1 protein levels in the embryo is, therefore, most likely to derive from translation of maternal Ago1 mRNA rather than from newly transcribed Ago1 mRNA (Luo, 2016).

Next, AGO1, AGO2, AGO3, AUB and Piwi protein levels were analyzed in eggs and embryos from mothers carrying either of two smg mutant alleles: smg¹ and smg⁴⁷. The smg mutations had no effect on the expression profiles of AGO2, AGO3, AUB or Piwi. In contrast, in smg-mutant embryos, the amount of AGO1 protein at both 0-to-2 and 2-to-4 hours was reduced relative to wild type and this defect was rescued in embryos that expressed full-length, wild-type SMG from a transgene driven by endogenous smg regulatory sequences. The reduction of AGO1 protein levels in smg mutants was not a secondary consequence of reduced Ago1 mRNA levels since Ago1 mRNA levels in both the smg-mutant and the rescued-smg-mutant embryos were very similar to wild type (Luo, 2016).

A plausible explanation for the decrease in AGO1 levels in smg mutants is the reduced levels of miRNAs, which would then result in less incorporation of newly synthesized AGO1 into functional miRISC and consequent failure to stabilize the AGO1 protein. To assess this possibility, a time-course in wild-type unfertilized eggs was analyzed in which zygotic genome activation and, therefore, zygotic miRNA synthesis, does not occur. It was found that AGO 1 levels were reduced in 2-to-4 hour wild-type unfertilized eggs compared with wild-type embryos of the same age. This result is consistent with a requirement for zygotic miRNAs in the stabilization of AGO1 protein (Luo, 2016).

Next, wild-type unfertilized egg and smg-mutant unfertilized egg time-courses were compared, and AGO1 levels were found to be further reduced in the smg mutant relative to wild type. This suggests that SMG protein has an additional function in the increase in AGO1 protein levels that is independent of SMG's role in zygotic miRNA production (since these are produced in neither wild-type nor smg-mutant unfertilized eggs) (Luo, 2016).

To assess whether this additional function derives from SMG's role as a post-transcriptional regulator of mRNA, smg¹ mutants were rescued either with a wild-type SMG transgene driven by the Gal4:UAS system (SMG^WT) or a GAL4:UAS-driven transgene encoding a version of SMG with a single amino-acid change that abrogates RNA-binding (SMG^RBD) and, therefore, is unable to carry out post-transcriptional regulation of maternal mRNAs. It was found that, whereas AGO1 was detectable in both unfertilized eggs and embryos from SMG^WT-rescued mothers, AGO1 was undetectable in unfertilized eggs from SMG^RBD-rescued mothers and was barely detectable in embryos from these mothers. Thus, SMG's RNA-binding ability is essential for its non-miRNA-mediated role in regulation of AGO1 levels during the MZT (Luo, 2016).

Since the abundance of SMG^WT and SMG^RBD proteins is very similar, the preceding result excludes the possibility that it is physical interaction between SMG and AGO1 that stabilizes the AGO1 protein. It was previously shown that the Ago1 mRNA is not bound by SMG. Thus, SMG must regulate one or more other mRNAs whose protein products, in turn, affect the synthesis and/or stability of AGO1 protein. It is known that turnover of AGO1 protein requires Ubiquitin-activating enzyme 1 (UBA1) and is carried out by the proteasome . It was previously shown that the Uba1 mRNA is degraded during the MZT in a SMG-dependent manner and that both the stability and translation of mRNAs encoding 19S proteasome regulatory subunits are up-regulated in smg-mutant embryos. It is speculated that increases in UBA1 and proteasome subunit levels in smg mutants contribute to a higher rate of AGO1 turnover and, thus, lower AGO1 abundance than in wild type (Luo, 2016).

AGO1 physically associates with BRAT. It is not known whether AGO1 interacts with PUM but it has been reported that, in mammals and C. elegans , Argonaute-family proteins interact with PUM/PUF-family proteins. Recent studies identified direct target mRNAs of the BRAT and PUM RBPs in early Drosophila embryos and showed through analysis of brat mutants that, during the MZT, BRAT directs late (i.e., after zygotic genome activation) decay of a subset of maternal mRNAs. These data permitted asking whether the maternal mRNAs that are predicted to be indirectly regulated by SMG via its role in miRISC production might be co-regulated by BRAT and/or PUM (Luo, 2016).

A highly significant overlap was found between the predicted miRNA-dependent indirect targets of SMG and both BRAT-and PUM-bound mRNAs in early embryos. This suggests that BRAT and PUM might function together with miRISC during the MZT to direct decay of maternal mRNAs (Luo, 2016).

Given that BRAT and PUM bind to largely non-overlapping sets of mRNAs during the MZT, there are three types of hypothetical BRAT-PUM-miRISC-containing complexes: one with both BRAT and PUM, one with BRAT only, one with PUM only. To assess this possibility for a specific set of zygotically produced miRNAs, the lists of mRNAs stabilized in 2-to-3 hour old embryos from miR-309 deletion mutants were compared to the lists of BRAT and PUM direct-target mRNAs. There was no significant overlap of PUM-bound mRNAs with those up-regulated in miR-309 mutants. However, there was a highly significant overlap of mRNAs up-regulated in miR-309-mutant embryos with BRAT-bound mRNAs. These results lead to the hypothesis that BRAT (but not PUM) co-regulates clearance of miR-309-family miRNA target maternal mRNAs during the MZT (Luo, 2016).

Protein Interactions

Translational regulation plays an essential role in development and often involves factors that interact with sequences in the 3' untranslated region (UTR) of specific mRNAs. For example, Nanos protein at the posterior of the Drosophila embryo directs posterior development, and this localization requires selective translation of posteriorly localized nanos mRNA. Spatial regulation of nanos translation requires Smaug protein bound to the nanos 3' UTR; binding represses the translation of unlocalized nanos transcripts. While the function of 3' UTR-bound translational regulators is, in general, poorly understood, these regulators presumably interact with the basic translation machinery. Smaug is shown to interact with the Cup protein and Cup is an eIF4E-binding protein that blocks the binding of eIF4G to eIF4E. Cup mediates an indirect interaction between Smaug and eIF4E, and Smaug function in vivo requires Cup. Thus, Smaug represses translation via a Cup-dependent block in eIF4G recruitment (Nelson, 2004),

To understand the mechanisms that underlie Smg's ability to repress translation, attempts were made to identify Smg-binding proteins. Initial work focused on proteins that would interact with amino acids 583-763. This region contains the Smg SAM domain, which is the protein's RNA-binding domain. An affinity resin carrying covalently coupled GST-Smg^583-763 was mixed with early embryo extracts. After extensive washing, bound proteins were eluted and detected via silver staining following SDS-PAGE. Several proteins were eluted from both the GST-Smg^583-763 resin and a resin carrying covalently coupled GST-Smg^179-307. However, an `80 kDa protein and an `140 kDa protein were specifically eluted from the Smg^583-763 resin. Both proteins were subjected to MALDI-TOF mass spectrometry, and while the smaller protein was not identified the larger was identified as Cup, which plays an essential but ill-defined role during oogenesis and early embryogenesis. To confirm that Cup interacts with Smg^583-763, Cup was generated via in vitro translation in rabbit reticulocyte lysate. This protein interacted with GST-Smg^583-763, as assayed by capture of Cup on glutathione agarose in the presence of GST-Smg^583-763 (Nelson, 2004),

Biochemical and genetic evidence is presented that is consistent with Cup functioning as an eIF4E-binding protein that mediates an interaction between Smg and eIF4E. Cup blocks the eIF4E/eIF4G interaction, suggesting that Smg-dependent translational repression of SRE-containing mRNAs results from a Cup-mediated block in the recruitment of eIF4G. Cup's role in Smg function is therefore similar to that played by Maskin in translational repression mediated by CPEB. Given that Maskin and Cup are not homologous, this suggests that other undiscovered adaptor eIF4E-binding protein/3' UTR-binding protein pairs will employ this mechanism to regulate translation (Nelson, 2004),

Cup interacts with eIF4E using both an eIF4E-binding motif and a second site that interacts with eIF4E through a distinct mechanism. Despite this difference, the second site is still able to inhibit the eIF4E/eIF4G interaction in vitro. Further work will be required to assess the significance of this site to Cup function in vivo (Nelson, 2004),

This model for Cup suggests that Smg represses translation at the level of initiation. However, the association of repressed nos mRNA with polysomes indicates that translational repression is achieved at a step after initiation. This apparent contradiction may reflect the fact that repression of nos translation is mediated by at least two trans-acting factors: Smg and a yet to be identified factor that functions through sequences in the nos 3' UTR that are distinct from the SREs. Thus, while Smg regulates translation at the level of initiation, additional factors may function at other levels. Similarly, Smg itself may utilize multiple mechanisms to repress nos expression, only one of which is Cup dependent (Nelson, 2004),

Regulation of translation during development often involves both translational repression and translational activation. The combination of these controls can spatially or temporally restrict the expression of an mRNA, thereby directing the proper development of a cell type or tissue. For example, nos translation is spatially regulated allowing for the proper development of the posterior of the Drosophila embryo. Smg plays an essential role in this process by repressing the translation of unlocalized nos mRNA, while nos mRNA localized to the posterior escapes this repression allowing for the accumulation of Nos protein specifically at the posterior. Given that Smg protein is distributed throughout the embryo, this suggests that Smg function must be over-ridden at the posterior. Cup is also distributed throughout the embryo, suggesting that spatial regulation of nos translation may involve disrupting Cup and/or Smg function specifically at the posterior. Osk protein, which is localized to the posterior, is required for nos translation and Osk interacts with Smg. Thus translational activation could involve Osk binding to Smg thereby blocking Smg function. Interestingly, Cup and Osk interact with the same region of the Smg protein. This might imply that Osk's interaction with Smg could disrupt the Cup/Smg complex and in so doing play a role in activating nos translation at the posterior (Nelson, 2004),

In Xenopus, temporal regulation of translation involves Maskin-mediated repression of target mRNAs in immature oocytes. Upon oocyte maturation, this repression is disrupted resulting in the activation of translation . This activation of translation involves a CPEB-mediated increase in the length of the transcript's poly(A) tail and subsequent recruitment of poly(A)-binding protein (PABP) to the message. PABP brings eIF4G to the mRNA, which in turn disrupts the Maskin/eIF4E complex resulting in translational activation. Measurement of the length of the nos poly(A) tail suggests that regulation of nos translation does not involve changes in poly(A) tail length. Thus, activation of nos translation does not likely involve disruption of the Cup/eIF4E complex through poly(A)-dependent eIF4G recruitment. Taken together, these results also suggest that the use of adaptor proteins such as Cup in translational regulation mediated by sequence-specific RNA-binding proteins is not restricted to mRNAs whose translation is regulated through their poly(A) tail (Nelson, 2004),

The data demonstrate that the same region of Smg that has previously been shown to function in sequence-specific RNA binding also interacts with Cup. The model therefore suggests that this region of the protein would be sufficient to repress translation. However, a transgene that expresses the Smg RNA-binding domain plus a short carboxy-terminal extension fails to rescue the smg mutant phenotype. These results would suggest that Smg has other essential functions in the early embryo in addition to Cup-dependent translational repression. Smg has been suggested to induce the degradation of target mRNAs in a process that may be distinct from its ability to repress translation. Perhaps this ability to induce mRNA degradation is essential and requires regions of Smg outside of amino acids 583-763 (Nelson, 2004),

Phenotypic analysis of several cup mutant alleles highlights Cup's involvement in a number of different biological processes during oogenesis and early embryogenesis, including oocyte growth, maintenance of chromosome morphology, and the establishment of egg chamber polarity. However, the molecular mechanisms that underlie Cup function have not been characterized. The demonstration that Cup is an eIF4E-binding protein suggests that at least some of the defects associated with mutations in the cup gene result from misregulation of translation. Consistent with this possibility is the fact that Cup has been previously shown to interact with Nos protein, which is itself a translational repressor. Genetic experiments suggest that Cup negatively regulates Nos activity during oogenesis, but the molecular mechanisms are not understood. This contrasts with Cup's positive effect on Smg-mediated translational repression. Thus Cup might utilize different molecular mechanisms to influence different translational repressors. The pleiotropic nature of the cup mutant phenotype suggests that Cup may serve as an adaptor protein that is utilized by multiple translational repressors to interact with eIF4E (Nelson, 2004),

Cup is homologous to 4E-T, a human nucleocytoplasmic shuttling protein that employs an eIF4E-binding motif to transport eIF4E into the nucleus (Dostie, 2000). The similarity between these proteins may suggest that Cup also functions to transport eIF4E into the nucleus. Thus some of the phenotypes associated with cup mutants may be related to a defect in eIF4E shuttling during oogenesis. The similarity between Cup and 4E-T also suggests that 4E-T might function in translational repression as an adaptor protein that mediates interactions between eIF4E- and 3' UTR-binding proteins. Specifically, 4E-T could function in translational repression mediated by the human Smg homolog. Similarly, additional RNA-binding proteins that interact with other eIF4E-binding proteins could function to regulate translation spatially or temporally. These protein pairs could control the translation of different mRNAs in various cell types throughout development (Nelson, 2004).

Oskar allows nanos mRNA translation in Drosophila embryos by preventing its deadenylation by Smaug/CCR4

Anteroposterior patterning of the Drosophila embryo depends on a gradient of Nanos protein arising from the posterior pole. This gradient results from both nanos mRNA translational repression in the bulk of the embryo and translational activation of nanos mRNA localized at the posterior pole. Two mechanisms of nanos translational repression have been described, at the initiation step and after this step. This study identifies a novel level of nanos translational control. The Smaug protein bound to the nanos 3' UTR recruits the deadenylation complex CCR4-NOT, leading to rapid deadenylation and subsequent decay of nanos mRNA. Inhibition of deadenylation causes stabilization of nanos mRNA, ectopic synthesis of Nanos protein and head defects. Therefore, deadenylation is essential for both translational repression and decay of nanos mRNA. A mechanism is proposed for translational activation at the posterior pole. Translation of nanos mRNA at the posterior pole depends on oskar function. Oskar prevents the rapid deadenylation of nanos mRNA by precluding its binding to Smaug, thus leading to its stabilization and translation. This study provides insights into molecular mechanisms of regulated deadenylation by specific proteins and demonstrates its importance in development (Zaessinger, 2006).

Post-transcriptional mechanisms of gene regulation play a prominent role during early development. Because the oocyte and developing embryo go through a phase in which no transcription takes place, gene expression relies on a pool of maternal mRNAs accumulated during oogenesis and is regulated at the level of translation or mRNA stability. It has been shown in several biological systems that poly(A) tail shortening contributes to translational silencing, whereas translational activation requires poly(A) tail extension. Poly(A) tail shortening, or deadenylation, is also the first step in mRNA decay. Subsequent steps occur only after the poly(A) tail has been shortened beyond a critical limit. Rapid deadenylation of unstable RNAs is caused by destabilizing elements, for example AU-rich elements (AREs) found in the 3' UTRs of several mRNAs. A number of proteins have been identified that bind to destabilizing RNA sequences and accelerate deadenylation as well as subsequent steps of decay (Zaessinger, 2006).

In yeast, deadenylation is mostly catalyzed by the multi-subunit CCR4-NOT complex, and this complex is also involved in deadenylation in Drosophila (Temme, 2004) and in mammalian cells. A second conserved deadenylase, the heterodimeric PAN2-PAN3 complex, appears to act before the CCR4-NOT complex. A third enzyme, the poly(A)-specific ribonuclease (PARN) is present in most eukaryotes but has not been found in yeast and Drosophila (Zaessinger, 2006).

Translational regulation of maternal mRNAs in Drosophila is essential to the formation of the anteroposterior body axis of the embryo. During embryogenesis, a gradient of the Nanos (Nos) protein arises from the posterior pole and organizes abdominal segmentation. This gradient results from translational regulation of maternal nos mRNA. The majority of nos transcripts is uniformly distributed throughout the bulk cytoplasm and is translationally repressed and subsequently degraded during the first 2-3 hours of embryonic development. A small proportion of nos transcripts is localized in the pole plasm, the cytoplasm at the posterior pole that contains the germline determinants. This RNA escapes repression and degradation, and its translation product forms a concentration gradient from the posterior pole. Both translation activation at the posterior pole and repression elsewhere in the embryo are essential for abdominal development, and head and thorax segmentation, respectively (Zaessinger, 2006 and references therein).

Translation of nos mRNA is repressed in the embryo by Smaug (Smg), which binds two Smaug response elements (SREs) in the proximal part of the nos 3' UTR. The SREs are also essential for the decay of nos mRNA. Repression of nos translation appears to be a multistep process, involving at least one level of regulation at the initiation step and another after nos mRNA has been engaged on polysomes. Repression at the initiation step is thought to involve an interaction between Smg and the protein Cup. The latter associates with the cap-binding initiation factor eIF4E, displacing the initiation factor eIF4G. Translation of nos mRNA at the posterior pole depends on Oskar (Osk) protein, although its mechanism of action has remained unknown (Zaessinger, 2006).

Bulk nos mRNA has a short poly(A) tail, and it was thought that nos translational control was independent of poly(A) tail length regulation. More recently, Smg and its yeast homologue Vts1 were shown to be involved in the degradation of mRNAs. Smg induces degradation and deadenylation of Hsp83 mRNA during early embryogenesis. This appears to result from recruitment by Smg of the CCR4-NOT deadenylation complex on Hsp83 mRNA, although the Smg-binding sites in this mRNA have not been identified. However, Hsp83 mRNA deadenylation was reported not to repress its translation. This study shows that nos mRNA is subject to regulation by active deadenylation by the CCR4-NOT deadenylation complex. This deadenylation depends on Smg and on the SREs in the 3' UTR of nos mRNA. The model is confirmed of the CCR4-NOT complex recruitment by Smg, in this case, onto nos mRNA, using genetic interactions between mutants affecting smg and the CCR4 deadenylase (properly named Twin and not to be confused with Twins), and showing the presence in a same protein complex of endogenous Smg and CAF1 (properly termed Pop2 and not to be confused with Caf1), a protein of the CCR4-NOT complex. Active deadenylation of nos mRNA contributes to its translational repression in the bulk embryo and is essential for the anteroposterior patterning of the embryo. Moreover, Osk activates translation of nos by preventing the specific binding of Smg protein to nos mRNA, thereby precluding active deadenylation and destabilization of nos mRNA (Zaessinger, 2006).

This paper shows that poly(A) tail length regulation is central to nos translational control. Poly(A) tail length regulation is a major mechanism of translational control, particularly during early development. nos translational control has been reported to be independent of poly(A) tail length. This conclusion came from the absence of nos poly(A) tail elongation between ovaries and early embryos, and the lack of nos poly(A) tail shortening between wild-type and osk mutant embryos in which nos mRNA is not translated at the posterior pole. However, later studies suggested that this lack of poly(A) tail change was not unexpected, as nos mRNA translation starts in ovaries, and the pool of translationally active nos mRNA in embryos is very small (4%) and remains undetected among the amount of translationally repressed nos in whole embryos. It has now been found that nos mRNA deadenylation by the CCR4-NOT complex, recruited to the 3' UTR by Smg, is required for nos translational repression in the bulk embryo. In addition, these data also suggest that nos translation at the posterior pole depends on the prevention of this deadenylation. nos mRNA is regulated at several levels, including localization, degradation, translational repression and translational activation. Localization at the posterior pole depends on two mechanisms: an actin-dependent anchoring at late stages of oogenesis, after nurse cells dumping and localized stabilization. Localization and translational control are coupled in that the localized RNA escapes both translational repression and degradation. A mechanism is proposed for this coupling. Translational repression and RNA degradation both involve Smg-dependent deadenylation. Deletion of the SREs in a nos transgene, as well as mutations in smg or in twin, which encodes the major catalytic subunit of the deadenylating CCR4-NOT complex, abrogate poly(A) tail shortening. Lack of deadenylation prevents the timely degradation of the RNA and also relieves translational repression. Deadenylation could repress nos mRNA translation by two mechanisms. Interaction of the cytoplasmic poly(A) binding protein (PABP) with mRNA poly(A) tails is important for the activation of translation initiation. Therefore, poly(A) shortening of nos mRNA would lead to PABP dissociation and inhibition of translation. In addition, deadenylation leads eventually to nos mRNA decay, which should also contribute to translational repression. Consistent with the Smg-dependent deadenylation of nos mRNA, describe in embryos, a recent study documented SRE-dependent deadenylation of chimeric transcripts containing the 3' UTR of nos mRNA in cell-free extracts from Drosophila embryos (Jeske, 2006). In this system, deadenylation of the chimeric RNAs also strongly contributes to translational repression, along with at least another deadenylation-independent mechanism (Zaessinger, 2006).

In this analysis, twin and smg mutants, although both impaired in nos mRNA poly(A) tail shortening, did not show the same defects. twin mutants fail to show nos poly(A) tail shortening during embryogenesis, whereas in smg mutant embryos or when poly(A) tails are measured from nos(ΔTCE) transgene, a poly(A) tail elongation is visible. This suggests that nos mRNA is also regulated by cytoplasmic polyadenylation which balances the deadenylation reaction, and that Smg binding to the RNA reduces the polyadenylation reaction. Consistent with a dynamic regulation of poly(A) tail length of maternal mRNAs resulting from a tight balance between regulated deadenylation and polyadenylation, it was found that in mutants for the GLD2 poly(A) polymerase, which is involved in cytoplasmic polyadenylation, nos mRNAs are precociously degraded in 0-1 hour embryos (Zaessinger, 2006).

Ectopic expression of osk in the bulk cytoplasm of the embryo is sufficient to impair nos mRNA binding to Smg and its deadenylation and destabilization. Therefore, it is proposed that, in wild-type embryos, Osk at the posterior pole inhibits Smg binding to the anchored nos mRNA, preventing deadenylation, decay and translational repression. This results in localized nos stabilization and translation. Osk might achieve this by a direct binding to Smg; it was shown to interact with Smg in vitro, through a region overlapping the RNA-binding domain in Smg. Alternatively, Osk could prevent Smg function independently of its binding to Smg, through its recruitment by another protein in nos-containing mRNPs. Consistent with a potential presence of Smg and Osk in the same protein complex, it was possible to co-immunoprecipitate Osk with Smg in embryos overexpressing Osk (Zaessinger, 2006).

Two mechanisms of nos translational repression have already been described. A first mode of translation inhibition appears to act during elongation, as suggested by polysome analysis and by the involvement of the Bicaudal protein, which corresponds to a subunit of the nascent polypeptide associated complex. The second mode of repression involves Smg and is thought to affect initiation. It requires the association of Smg with the protein Cup, which also binds eIF4E. The association of Cup with eIF4E competes with the eIF4E/eIF4G interaction, which is essential for translation initiation. This study identifies deadenylation by the CCR4-NOT complex as a novel level of nos translational repression, also involving Smg. Smg protein synthesis is probably induced by egg activation during egg-laying. Smg is absent in ovaries and accumulates during the first hours of embryogenesis, with a peak at 1-3 hours. Its amount is low during the first hour and possibly nonexistent during the first 30 minutes. This correlates with the presence at that time of high levels of nos mRNA in the bulk embryo that are not destabilized. nos translational repression is active, however, as this pool of mRNA is untranslated. Thus a Smg-independent mode of repression must be efficient during the first hour of development. This might correspond to repression at the elongation step and/or involve the Glorund protein, a Drosophila hnRNP F/H homologue newly identified as a nos translational repressor in the oocyte (Kalifa, 2006). Glorund has a role in repression of unlocalized nos mRNA in late oocytes and has been suggested to also act at the beginning of embryogenesis while Smg is accumulating to ensure the maintenance of translational repression at the oogenesis to embryogenesis transition (Kalifa, 2006). Analysis of glorund mutants revealed that the embryonic phenotypes are less severe than expected and led to the proposal that at least an additional level of nos translational repression is active in oocytes (Kalifa, 2006). Overexpression of Osk in the germline with nos-Gal4 results in long poly(A) tails of nos mRNA, even in 0-1 hour embryos in which Smg protein is poorly expressed. This suggests that the short poly(A) tail of nos mRNA in 0-1 hour wild-type embryos could in part result from active deadenylation during oogenesis, which would depend on a regulatory protein different from Smg. Deadenylation could therefore be involved in nos regulation during oogenesis, and would also be prevented by Osk in the pole plasm, as in embryos (Zaessinger, 2006).

Genetic evidences indicate that all three levels of translational repression are additive. Although the importance of the Smg/Cup/eIF4E mode of nos translational repression for the anteroposterior patterning of the embryo has not been addressed, the other two levels of repression are essential, as ectopic Nos protein leads to disruption of the embryo anteroposterior axis in twin or bicaudal mutants. This demonstrates that none of the three levels of repression is sufficient by itself and suggests that all three regulations are required to achieve complete translational repression of nos. As Osk acts by preventing the binding of Smg to the nos 3' UTR, it is likely to inhibit both Smg-dependent mechanisms of translational repression (Zaessinger, 2006).

The presence of Smg in discrete cytoplasmic foci and its partial colocalization in these foci with components of the CCR4-NOT deadenylation complex, and with components of P bodies, suggest that Smg-dependent deadenylation and translational control of nos occur in P bodies. P body dynamics and function have not been addressed in a complete organism during development. Consistent with the apparent complexity of P body function, including mRNA decay and translational repression, embryos different subsets of Smg-containing structures were identified in embryos: these subsets either do or do not contain the CCR4-NOT deadenylation complex and the Xrn1 5'-3' exonuclease. This suggests the existence of different types of P bodies that may have distinct functions (Zaessinger, 2006).

The CCR4-NOT complex is involved in default deadenylation of bulk mRNAs in somatic cells (Temme, 2004). This study finds that the same deadenylation complex has a role in active, sequence-specific deadenylation of a particular mRNA. Activation of deadenylation by CCR4-NOT results from the recruitment of the deadenylation complex by a regulatory RNA-binding protein to its specific mRNA target (this study) (Semotok, 2005). Several RNA-binding proteins are expected to interact with the CCR4-NOT complex to regulate the deadenylation of different pools of mRNAs in different tissues. CCR4 controls poly(A) tail lengths of Cyclin A and B mRNAs during oogenesis (Morris, 2005); the regulatory protein has not been identified, but it cannot be Smg, which is not expressed in ovaries. A similar mode of active deadenylation involving the recruitment of the deadenylation complex by ARE-binding proteins has been proposed in mammalian cells. A study in yeast has identified the PUF (Pumilio/FBF) family of RNA-binding proteins as activators of CCR4-NOT-mediated deadenylation through a direct interaction between PUF and POP2 (the CAF1 homologue). Although default deadenylation by CCR4 is not essential for viability (Temme, 2004), active deadenylation by CCR4 of specific mRNAs is essential for certain developmental processes, in particular during early development (Zaessinger, 2006).

SMAUG is a major regulator of maternal mRNA destabilization in Drosophila and its translation is activated by the PAN GU kinase

In animals, egg activation triggers a cascade of posttranscriptional events that act on maternally synthesized RNAs. In Drosophila, the Pan Gu (Png) kinase sits near the top of this cascade, triggering translation of Smaug (Smg), a multifunctional posttranscriptional regulator conserved from yeast to humans. Although Png is required for cytoplasmic polyadenylation of smg mRNA, it regulates translation via mechanisms that are independent of its effects on the poly(A) tail. Analyses of mutants suggest that Png relieves translational repression by Pumilio (Pum) and one or more additional factors, which act in parallel through the smg mRNA's 3' untranslated region (UTR). Microarray-based gene expression profiling shows that Smg is a major regulator of maternal transcript destabilization. Smg-dependent mRNAs are enriched for gene ontology annotations for function in the cell cycle, suggesting a possible causal relationship between failure to eliminate these transcripts and the cell cycle defects in smg mutants (Tadros, 2007).

Gene expression profiling analyses have shown that, in Drosophila, a remarkably high fraction (55%) of encoded mRNAs is expressed and loaded into mature oocytes. An earlier estimate of 30% was derived from methods biased toward identification of RNAs that are strictly maternally expressed, whereas, in principle, the method used identifies all maternally expressed genes, including those also expressed at other stages or in other cell types. The predicted number of maternal RNAs is likely to increase as more sensitive in situ hybridization methods are used to determine the maternal versus nonmaternal cutoff (Tadros, 2007).

In Drosophila, elimination of a subset of maternal transcripts is accomplished through the joint action of two pathways: one is maternally encoded and active in unfertilized eggs; the second requires fertilization and zygotic transcription. This study shows that 20% of the maternal mRNAs (more than 1600) are destabilized by the 'maternal' pathway. The actual number of maternal transcripts that are destabilized in embryos is, thus, expected to be significantly larger (Tadros, 2007).

Maternal mRNA destabilization in zebrafish depends on miR-430, which is absent from the oocytes and is transcribed only after fertilization. miR-430 therefore functions in a zebrafish pathway equivalent to the Drosophila 'zygotic' pathway (Bashirullah, 1999). These analyses suggest that, in Drosophila, the earlier, 'maternal' destabilization pathway does not require miRNAs, as known miRNA binding sites are not enriched in the 3′ UTRs of unstable transcripts. Nonetheless, the fact that several miRNA target sites are enriched in the maternal class as a whole suggests that miRNAs may function in the translational regulation of these transcripts rather than their degradation. It also remains possible that miRNAs participate in the 'zygotic' pathway; because this pathway is expected to affect a significantly larger subset of maternal mRNAs than the 'maternal' pathway, this would explain the observed enrichment of miRNA targets in the maternal class as a whole (Tadros, 2007).

Quite unexpected was the discovery that Smg is a major regulator of maternal transcript destabilization, being required for elimination of two thirds of the mRNAs that degrade upon egg activation. Smg regulates translation through cis elements known as SREs. However, SREs do not mediate Smg-dependent degradation of endogenous transcripts. For example, although both nanos (Smg-independent) and Hsp83 (Smg-dependent) mRNAs contain SREs, degradation is SRE dependent only in the case of the former. Therefore, it is not surprising that SREs are enriched in the unstable class of transcripts but not further enriched in the Smg-dependent subclass. In summary, though Smg may trigger the degradation of endogenous transcripts through an SRE-independent mechanism, so also SREs may bind a degradation factor other than Smg (Tadros, 2007).

The discovery that the Png kinase complex coordinates translation of smg mRNA through its 3′ UTR is reminiscent of the role of the Aurora A kinase in translational unmasking of maternal mRNAs during Xenopus oocyte maturation. In the frog system, Aurora A phosphorylates CPE binding protein (CPEB), which is bound to a 3′ UTR element known as the cytoplasmic polyadenylation element (CPE). CPEB then promotes lengthening of the poly(A) tail, thus facilitating binding of poly(A) binding protein (PABP). PABP in turn binds eIF4G, bringing it into proximity with eIF4E, thus disrupting eIF4E's interaction with the repressor, Maskin. This permits recruitment of the 40S ribosomal complex and initiation of translation. CPEB-mediated regulation of polyadenylation and translation is also crucial later, during early embryogenesis, for cell cycle progression (Tadros, 2007).

Though frog Aurora A and fly Png are both Ser/Thr kinases that function through 3′ UTRs to translationally activate maternal mRNAs, their modes of action differ. Although Png is required for the polyadenylation of its target transcripts, the data suggest that its role in promoting translation is either 'downstream' of or runs 'in parallel' to polyadenylation. The distinction between these two mechanisms lies in the interpretation of the fact that lengthening smg poly(A) tails results in increased translation in wild-type but not in png embryos. A downstream role for Png could be to transduce a signal linking polyadenylation and translation. For example, in plants, phosphorylation of PABP increases its cooperative binding to poly(A) RNA. Alternatively, Png might function in a pathway independent of polyadenylation. For example, during Xenopus oocyte maturation, Ser/Thr phosphorylation of Maskin is crucial for its dissociation from eIF4E and subsequent translational activation of CPE-bearing transcripts. Png could phosphorylate and cause dissociation of an analogous eIF4E binding protein (there is no clear Maskin ortholog in Drosophila) (Tadros, 2007).

Png has been shown to promote translation of Cyclin B and together with these results on smg, it is suggested that the previously surmised independent regulation of destabilization and the cell cycle by Png lies at the level of its targets: smg mRNA in the case of destabilization, and cyclin B mRNA in that of the early embryonic cell cycle. Though Png regulates cyclin B mRNA translation through the proposed relief of Pum-mediated translational repression, for smg mRNA, Png acts to relieve repression by Pum and one or more proteins that act in parallel. Because there are no canonical Nanos response elements (NREs) in the smg 3′ UTRs, regulation of smg translation by Pum must be indirect or occur via noncanonical NREs. Consistent with either of these possibilities is the recent finding that smg mRNA is associated with a transgenic Pum protein fragment in embryonic extracts (Gerber, 2006). Also noteworthy is the fact that Pum's repression of one of its known target mRNAs, hb, occurs through both polyadenylation-dependent and -independent mechanisms (Tadros, 2007).

Expression of Smg protein in png mutants is not sufficient to restore instability to Hsp83, a Smg-dependent maternal mRNA. Thus, destabilization of Smg-dependent maternal mRNAs in eggs from png mutant mothers requires one or more additional proteins. Png may promote the translation of Y, an essential component of the destabilization machinery, or may phosphorylate Y, thus activating the degradation machinery. Global analyses of maternal RNA stability in png mutants expressing UAS-smg-bcd3′UTR will identify whether any of the Png-dependent transcripts that are Smg dependent are Y independent (Tadros, 2007).

It is noted that a third of the unstable maternal mRNAs are Smg independent. Png function is likely to be required to destabilize a subset of these Smg-independent maternal transcripts. This is suggested by the fact that nanos mRNA is fully stabilized in png mutants but is only partially stabilized in smg mutants. Global analyses of maternal RNA stability in png mutants will identify all Png-dependent transcripts (Tadros, 2007).

In Drosophila embryos, the transition from maternal to zygotic control of development has been hypothesized to require two processes: elimination of maternal mRNAs and synthesis of zygotic mRNAs. Zygotic transcription is required for cellularization, the hallmark of the Drosophila MBT. However, the functional significance of maternal transcript elimination has remained largely unexplored. smg mutants have been shown to fail to progress beyond nuclear cycle 12, never reaching the MBT, and computational analyses have shown that the Smg-dependent unstable maternal transcripts are enriched for GO terms related to mitosis and the cell cycle. This enables the presentation of a model in which elimination of maternal cell cycle mRNAs by Smg is essential for progression through the final syncytial nuclear divisions and, ultimately, the MBT. Detailed cellular and molecular analysis of the smg mutant phenotype will be required to test this hypothesis (Tadros, 2007).

Smg homologs exist from yeast to humans, where they function in posttranscriptional regulation. Furthermore, the budding yeast homolog Vts1 has been shown to interact with the same cis element as Smg. Since turnover of maternal mRNAs occurs prior to the MBT in all metazoa, Smg homologs may fulfill a conserved developmental function: targeting a subset of maternal mRNAs for elimination and thus permitting the MBT to occur (Tadros, 2007).

Maternal mRNA deadenylation and decay by the piRNA pathway in the early Drosophila embryo

Piwi-associated RNAs (piRNAs), a specific class of 24- to 30-nucleotide-long RNAs produced by the Piwi-type of Argonaute proteins, have a specific germline function in repressing transposable elements. This repression is thought to involve heterochromatin formation and transcriptional and post-transcriptional silencing. The piRNA pathway has other essential functions in germline stem cell maintenance and in maintaining germline DNA integrity. This study uncovered an unexpected function of the piRNA pathway in the decay of maternal messenger RNAs and in translational repression in the early embryo. A subset of maternal mRNAs is degraded in the embryo at the maternal-to-zygotic transition. In Drosophila, maternal mRNA degradation depends on the RNA-binding protein Smaug and the deadenylase CCR4, as well as the zygotic expression of a microRNA cluster. Using mRNA encoding the embryonic posterior morphogen Nanos (Nos) as a paradigm to study maternal mRNA decay, it was found that CCR4-mediated deadenylation of nos depends on components of the piRNA pathway including piRNAs complementary to a specific region in the nos 3' untranslated region. Reduced deadenylation when piRNA-induced regulation is impaired correlates with nos mRNA stabilization and translational derepression in the embryo, resulting in head development defects. Aubergine, one of the Argonaute proteins in the piRNA pathway, is present in a complex with Smaug, CCR4, nos mRNA and piRNAs that target the nos 3' untranslated region, in the bulk of the embryo. It is proposed that piRNAs and their associated proteins act together with Smaug to recruit the CCR4 deadenylation complex to specific mRNAs, thus promoting their decay. Because the piRNAs involved in this regulation are produced from transposable elements, this identifies a direct developmental function for transposable elements in the regulation of gene expression (Rouget, 2010).

In Drosophila embryos, Nos is expressed as a gradient that emanates from the posterior pole and organizes abdominal segmentation. The majority of nos mRNA is distributed throughout the bulk cytoplasm, translationally repressed and subsequently degraded during the first 2-3 h of development. This repression is essential for head and thorax segmentation. A small amount of nos transcripts, localized at the posterior pole of the embryo, escapes degradation and is actively translated, giving rise to the Nos protein gradient. nos mRNA decay in the bulk cytoplasm depends on the CCR4-NOT deadenylation complex and its recruitment onto nos by Smaug (Smg). This contributes to translational repression in the bulk of the embryo and is required for embryonic antero-posterior patterning (Zaessinger, 2006; Rouget, 2010 and references therein).

Smg has been suggested to be not the only activator of nos mRNA decay during early embryogenesis. Zygotically expressed miRNAs have been reported to activate maternal mRNA deadenylation in zebrafish embryos and decay in Drosophila embryos. This study investigated the potential involvement of other classes of small RNAs in mRNA deadenylation and decay before zygotic expression. Because piRNAs are expressed maternally in the germ line and are present in early embryos, the possible role of the piRNA pathway in maternal mRNA deadenylation was analyzed. Piwi, Aubergine (Aub) and Ago3 are specific Argonaute proteins, Armitage (Armi) and Spindle-E (Spn-E) are RNA helicases, and Squash (Squ) is a nuclease involved in piRNA biogenesis and function. Poly(A) test assays were performed to measure nos mRNA poly(A) tail length in embryos spanning 1-h intervals during the first 4 h of embryogenesis. In contrast to the progressive shortening of nos mRNA poly(A) tails observed in wild-type embryos correlating with mRNA decay during this period, nos poly(A) tail shortening was affected in embryos from females mutant for the piRNA pathway (herein referred to as mutant embryos). This defect in deadenylation correlated with higher amounts of nos mRNA in mutant embryos, as quantified by reverse transcription-quantitative PCR (RT-qPCR). In situ hybridization revealed stabilized nos mRNA in the bulk cytoplasm of mutant embryos where it is normally degraded in the wild type. Consistent with previous data showing that nos mRNA deadenylation is required for translational repression (Zaessinger, 2006), defective deadenylation in mutant embryos resulted in the presence of ectopic Nos protein throughout the embryo. The presence of Nos in the anterior region results in the repression of bicoid and hunchback mRNA translation and in affected head skeleton. It was found that the piwi¹ mutant embryos that were able to produce a cuticle had head defects (Rouget, 2010).

The piRNA pathway has a role during early oogenesis in preventing DNA damage, possibly through the repression of transposable element transposition. DNA double-strand breaks arising in mutants of the piRNA pathway correlate with affected embryonic axis specification, and this developmental defect is suppressed by mutations in the Chk2 DNA-damage signal transduction pathway. This study found that defects in nos mRNA deadenylation and decay observed in aub or armi mutants were not suppressed by Chk2 (mnk^P6) mutations, indicating that these defects did not result from activation of the Chk2 pathway earlier during oogenesis. Moreover, affected deadenylation of nos mRNA in piRNA pathway mutants did not depend on oskar (Rouget, 2010).

A potential direct role of the piRNA pathway in the regulation of nos mRNA deadenylation and decay in the embryo was addressed. Aub and Piwi accumulate in the pole plasm and in pole cells of the embryo. However, lower levels of Aub and Piwi are found throughout the entire embryo. Ago3 is also present throughout the embryo. Aub and Ago3 are found in the cytoplasm and accumulated in discrete foci, a distribution similar to that of CCR4 and Smg. CCR4 and Smg were reported to partially colocalize in small cytoplasmic foci (Zaessinger, 2006). Aub and Ago3 also partially colocalize with Smg and CCR4 in the bulk of syncytial embryos, in both cytoplasmic foci and a diffusely distributed cytoplasmic pool. Importantly, the distributions of CCR4 and Smg depended on the piRNA pathway; they are strongly affected in aub and spn-E mutant embryos. Although global amounts of CCR4 and Smg do not decrease in mutant embryos, CCR4 foci strongly increase in size, whereas Smg foci decreases in size or disappears. This suggests that subsets of CCR4 and Smg foci have different functions and that deadenylation may take place diffusely in the cytoplasm. These results demonstrate a functional link between CCR4-mediated deadenylation and the piRNA pathway (Rouget, 2010).

Co-immunoprecipitation experiments showed that Aub co-precipitate Smg, CCR4 and Ago3 in the absence of RNA, indicating the presence of these proteins in a common complex. Smg also co-precipitates CCR4, Aub and Ago3; however, Piwi is not found to co-precipitate Smg or CCR4. Importantly, Smg, CCR4 and Ago3 also co-precipitate with Aub in osk⁵⁴ mutant embryos that are defective in pole plasm assembly, indicating the presence of this complex outside the pole plasm. Next it was shown that nos mRNA co-precipitate with Aub in both wild-type and osk⁵⁴ embryos. The amount of nos mRNA is similar in Aub and Smg immunoprecipitates (Rouget, 2010).

These findings show that the Argonaute proteins Aub and Ago3 associate with Smg and the CCR4 deadenylase complex to directly regulate nos mRNA in the bulk cytoplasm of early embryos (Rouget, 2010).

The nos 3'untranslated region (UTR) contains Smg-binding sites located in its 5'-most region [referred to as the translational control element (TCE)]. piRNAs sequenced from early embryos and presumed capable of targeting nos 3'UTR were sought based on their sequence complementarity. Notably, a specific region located in the 3'-most part of the 3'UTR could be targeted by over 200 copies of piRNAs originating from two transposable elements, 412 and roo. piRNAs complementary to nos 3'UTR were visualized by northern blots. In addition, piRNAs predicted to target nos 3'UTR co-immunoprecipitated with Aub. nos genomic transgenes deleted for different parts of the 3'UTR were used to address the requirement of the corresponding regions for nos mRNA deadenylation. It was shown previously that the TCE (nucleotides 1-184) is required for nos mRNA poly(A) tail shortening, consistent with the role of Smg in this process. Deletion of region 184-403 (nos(delta1)) had no effect, whereas poly(A) tails from the transgene deleted for the region 403-618 (nos(delta2)) were elongated in 3-4-h embryos. This could indicate regulation by the miRNA predicted by miRBase to target this region. Deletion of 618-844 in the nos 3'UTR (nos(delta3)) had a strong effect on nos deadenylation. Consistent with this, nos mRNA levels produced by this transgene remained mostly stable. This resulted in defects in embryo patterning: a total of 35% of embryos from nos(delta3) females did not hatch and among them 86% showed head skeleton defects. Next specific sequences complementary to 412 (15 nucleotides) and roo (11 nucleotides) retrotransposon piRNAs were deleted. These short deletions, either independently or in combination, affected nos mRNA deadenylation (Rouget, 2010).

To support further the role of retrotransposon piRNAs in nos mRNA regulation, 412 and roo piRNAs were blocked by injecting specific 2'-O-methyl anti-piRNA in embryos, and cuticles were recorded as a functional assay of Nos ectopic synthesis at the anterior pole. Injection of anti-piRNA(412) or anti-piRNA(roo) resulted in specific head development defects (Rouget, 2010).

Together, these results provide strong evidence that an interaction between piRNAs and nos mRNA is required for nos mRNA deadenylation and translational repression in the first hours of embryogenesis (Rouget, 2010).

This study has identified a new function of the piRNA pathway in the regulation of maternal mRNAs. Recently, piRNAs derived from the 3'UTRs of cellular transcripts have been identified in gonadal somatic cells, although their biological role has not been clarified. It is proposed that piRNAs, in complex with Piwi-type Argonaute proteins Aub and Ago3, target nos maternal mRNAs and recruit or stabilize the CCR4-NOT deadenylation complex together with Smg. These interactions induce rapid mRNA deadenylation and decay. Thus, activation of mRNA deadenylation represents a new direct mechanism of action for the piRNA pathway with an essential developmental function during the first steps of embryogenesis (Rouget, 2010).

Smg is a general factor for mRNA decay during early embryogenesis. Because Aub and Ago3 are present in a complex with Smg in early embryos, a proportion of Smg mRNA targets could be regulated by the piRNA pathway. Consistent with this, other maternal mRNAs that are destabilized during early embryogenesis are targeted by abundant piRNAs and their deadenylation depends on the piRNA pathway (Rouget, 2010).

These piRNAs involved in gene regulation are generated from transposable element sequences. Although transposable elements have been described to be essential for genome dynamics and evolution, their immediate function within an organism has remained rather elusive. This study provides evidence for a co-evolution between transposable elements and the host genome and reveals the direct developmental function of transposable elements in embryonic patterning, through the regulation of gene expression (Rouget, 2010).

microRNA-independent recruitment of Argonaute 1 to nanos mRNA through the Smaug RNA-binding protein

Argonaute (Ago) proteins are typically recruited to target messenger RNAs via an associated small RNA such as a microRNA (miRNA). This study describe a new mechanism of Ago recruitment through the Drosophila Smaug RNA-binding protein. Smaug interacts with the Ago1 protein, and Ago1 interacts with and is required for the translational repression of the Smaug target, nanos mRNA. The Ago1/nanos mRNA interaction does not require a miRNA, but it does require Smaug. Taken together, these data suggest a model whereby Smaug directly recruits Ago1 to nanos mRNA in a miRNA-independent manner, thereby repressing translation (Pinder, 2013).

The data indicate that Smaug directly recruits Ago1 to nos mRNA in a miRNA-independent fashion. Interestingly, previous work suggests similarities between the mechanisms that Smaug and Ago use to regulate mRNAs. For example, translationally repressed nos mRNA is polysome associated similar to some transcripts that are repressed by Ago proteins. However, it is unclear if Smaug protein participates in repression of nos mRNA that is polysome associated or if this is mediated the RNA-binding protein Glorund, which is another regulator of nos mRNA. Another similarity between Smaug and Ago-mediated mechanisms comes from in vitro translation extracts showing that both Smaug and Ago1 repress translation through steps that function downstream of translation initiation and require ATP. While these similarities are consistent with Smaug-mediated translational repression functioning, at least to some extent, through Ago1, it is noteworthy that several other mechanisms have been shown to have roles in translational repression by Ago protein (Pinder, 2013).

The data also indicate that while Ago1 functions in nos translational repression it does not induce nos mRNA degradation. Thus, in the context of nos mRNA, Ago1 does not induce transcript decay. In contrast, Smaug has a role in both translational repression and degradation of nos message, indicating that some of Smaug's functions are Ago1-independent (Pinder, 2013).

The interaction of Smaug with both Ago1 and Ago2 could reflect a common binding site on Ago proteins that mediates their interaction with Smaug. However, no ectopic Nos protein was seen in Ago2 mutant embryos. This could suggest that while the Smaug/Ago1 interaction is functional, the interaction with Ago2 is not, perhaps because of the different mechanisms that are used by Ago1 and Ago2 to repress translation (Pinder, 2013).

The data presented in this study, along with previous work, indicate that Smaug uses several mechanisms to regulate its target mRNAs. In addition to translational repression mediated by Smaug/Ago1, Smaug interacts with the Cup protein, which in turn interacts with the cap-binding protein eIF4E. eIF4E bound to the 5' cap of an mRNA indirectly recruits the 40S ribosomal subunit to an mRNA through eIF4E's interaction with eIF4G. As Cup blocks the eIF4E/eIF4G interaction, formation of the Smaug/Cup/eIF4E complex on a target mRNA represses translation initiation. Smaug also recruits the Ccr4/Not deadenylase to target mRNAs resulting in removal of the transcripts poly(A) tail, thus repressing translation and/or inducing transcript degradation. Interestingly, Smaug has also been shown to interact with the Piwi-type Ago proteins, Aubergine and Ago3. That study proposed that a complex consisting of Smaug, Aubergine with associated piwiRNAs and the Ccr4/Not deadenylase is recruited to nos mRNA through binding of both Smaug to the nos SREs and piwiRNAs to complementary to sequences in the nos 3' UTR (Pinder, 2013).

These models of Smaug function raise the question as to why one RNA-binding protein uses several mechanisms to repress its target transcripts? One possibility is that by using several mechanisms, Smaug ensures that its targets are efficiently silenced. This might be especially important for a transcript, such as nos, where even low levels of inappropriate expression are lethal to the embryo. Alternatively, Smaug might use different mechanisms to regulate different target mRNAs. Indeed, while Smaug induces the degradation of Hsp83 mRNA, it does not repress Hsp83 translation. In contrast, while Smaug represses nos translation, it has only a modest role in destabilizing nos mRNA. This differential regulation could reflect the fact that the location of the SREs in the target mRNA (e.g., 3' UTR for nos versus open reading frame for Hsp83) and/or more cis-elements within these target mRNAs along with the trans-acting factors that recognize these influence the mechanisms of Smaug function (Pinder, 2013).

While the data indicate that Smaug recruits Ago1 to nos mRNA via a miRNA-independent mechanism, it is still possible that a miRNA bound to Ago1, but not base-paired with nos mRNA, could be important for nos regulation. In this model, the miRNA would be required for Ago1 to repress nos translation at a step downstream of recruitment. For example, miRNA binding to Ago1 might be required to induce allosteric changes in Ago1 that facilitate its interaction with other factors that are required for Ago1-mediated repression. In this model, any miRNA would be sufficient for repression as it is not base-paired with the target transcript. Indeed, it has been suggested that miRNA binding might be involved in allosteric changes that are important for Ago-mediated repression, while other studies have shown that miRNA binding confers substantial structural stability to an Ago protein. It is noted that the Ago1 miRNA-binding mutant used in the present study also fails to interact with GW182, an Ago-binding protein that is essential for translational repression; therefore it could not be used to test for miRNA-independent translational repression of nos mRNA in vivo (Pinder, 2013).

A variety of RNA-binding proteins have been shown to function at numerous steps in miRNA/Ago-mediated post-transcriptional regulation. For example, both a mammalian and Caenorhabditis elegans Pum form a complex that includes an Ago protein, and in vitro experiments showed that the mammalian version of this complex represses translation of a reporter mRNA that carried only Pum-binding sites. The lack of miRNA-binding sites within this reporter suggests that in this in vitro system Pum is able to recruit Ago to an mRNA in a miRNA-independent manner. It is, however, unclear whether this Pum/Ago complex is able to recognize target mRNAs in vivo in the absence of miRNA binding. Indeed, it is suggested that in vivo targeting of the Pum/Ago complex might involve the presence of both Pum and miRNA-binding sites within a transcript. Thus, the data indicating that, in vivo, Smaug can recruit Ago1 to an mRNA in a miRNA-independent fashion, combined with other in vitro data suggesting that Pum can also directly recruit an Ago protein, suggests that other RNA-binding proteins might also function in a similar manner (Pinder, 2013).

DEVELOPMENTAL BIOLOGY

Embryonic

To determine when and where Smaug is expressed, polyclonal antibodies against different portions of the protein were prepared. In extracts prepared from early embryos, these antibodies specifically recognize a protein with an apparent molecular weight of 120 kDa that is encoded by smg. Analysis of Western blots reveals that Smaug protein is undetectable in ovaries. Translation of maternal SMG mRNA during early embryogenesis results in the accumulation of protein 0-3 hr postfertilization. As is the case for many other maternal mRNAs, SMG transcripts are degraded prior to formation of the cellular blastoderm, and Smaug protein levels fall dramatically at this point in development. Next, antibodies that recognize Smaug were used to examine its distribution during early embryonic development. Smaug protein is distributed uniformly throughout the bulk cytoplasm and pole plasm of the syncitial embryo. At no point in early development is any asymmetry in the distribution of Smaug observed. Two lines of evidence demonstrate that these antibodies specifically recognize Smaug in situ: (1) embryos derived from females with extra copies of a smg⁺ transgene exhibit the same monotonous distribution of antigen, albeit at higher levels; (2) smg mutant embryos have no detectable protein (Dahnukar, 1999).

Effects of Mutation or Deletion

To examine the role of smaug in vivo, a loss-of-function mutation was isolated in smg by the following strategy. The gene was mapped to polytene band 66F, and deficiencies that uncover smg were identified by quantitative Southern blot analysis. Embryos from females bearing one of these Df(Scf) deficiencies have approximately one-half the amount of Smaug present as do embryos from wild-type females, collected under similar conditions, as measured by Western blot. Flies with a wild-type third chromosome were mutagenized with EMS and then crossed to flies bearing Df(Scf). Recessive mutations that cause lethal or female-sterile phenotypes in trans to Df(Scf) were identified by analysis of approximately 5000 crosses. One candidate from this screen proved to be a recessive, loss-of-function mutation in smg. While embryos from smg/+ heterozygous females are indistinguishable from wild type, embryos from smg/Df(Scf) females (hereinafter, smg mutant embryos) die. This embryonic lethality is rescued by maternal expression from a smg⁺ transgene but not by maternal expression from various smg deletion derivatives. In particular, expression of either a Smaug derivative lacking the RNA-binding domain or one consisting of the RNA-binding domain plus a short C-terminal extension does not rescue embryonic viability. Western blot analysis reveals that each derivative is stable in vivo, and, thus, Smaug appears to have separable essential functions in the N-terminal region and in the RNA-binding domain. smg mutant embryos contain approximately normal levels of full-length SMG mRNA, but essentially no Smaug protein is detected by Western blot using an antibody directed against the RNA-binding domain. The smg¹ allele appears to be null (at least for early embryonic function), based on the criterion that the terminal phenotype of embryos from smg/Df(Scf) and from smg/smg females are indistinguishable (Dahanukar, 1999).

The embryonic lethality caused by loss of Smaug function is not a result of defects in NOS mRNA regulation, suggesting that Smaug has additional roles during early embryonic development. Unlike nos mutant embryos, which complete embryonic development and die with abdominal segmentation defects, smg mutant embryos arrest very early in embryonic development. To investigate the genesis of these early defects, development was monitored by preparing timed collections of wild-type and mutant embryos that were fixed and subsequently stained with the DNA-binding dye DAPI to reveal the distribution of nuclei. Smaug is required for normal progression beyond nuclear division cycles 11-12 in the precellular blastoderm stage. From fertilization through cycle 10, the nuclear divisions appear to be essentially indistinguishable in wild-type and smg embryos. However, beginning in cycle 11, the pattern of nuclei in smg mutant embryos becomes increasingly irregular, with a variety of defects including asynchronous mitoses and gaps in the otherwise regular array of nuclei at the surface of embryo. The proportion of mitotic nuclei in smg embryos suggests that progression through mitosis is retarded. The terminal phenotype appears to result from the aggregation of nuclei that fall from the surface into the middle of the embryo. smg mutant embryos never achieve the nuclear density at the surface of cycle 14 wild-type embryos and never cellularize (Dahanukar, 1999).

These phenotypes are similar in some respects to those associated with mutations in genes involved in regulating cell cycle progression, cytoskeletal integrity, or the coordination between the two. For example, mutations in grapes (grp) or nuclear-fallout (nuf) cause similar nuclear phenotypes and arrest development between cycles 10 and 14. grp encodes a S/T kinase that is thought to participate in a DNA damage checkpoint, and nuf encodes a protein that cycles between the cytoplasm and the centrosomes. While the genesis of the defects in smg mutant embryos has not been investigated further, Smaug evidently is required for normal progression through the late nuclear division cleavages in the syncitial blastoderm (Dahanukar, 1999 and references).

Temporal reciprocity of miRNAs and their targets during the maternal-to-zygotic transition in Drosophila

During oogenesis, female animals load their eggs with messenger RNAs (mRNAs) that will be translated to produce new proteins in the developing embryo. Some of these maternally provided mRNAs are stable and continue to contribute to development long after the onset of transcription of the embryonic (zygotic) genome. However, a subset of maternal mRNAs are degraded during the transition from purely maternal to mixed maternal-zygotic gene expression. In Drosophila, two independent RNA degradation pathways are used to promote turnover of maternal transcripts during the maternal-to-zygotic transition. The first is driven by maternally encoded factors, including SMAUG, whereas the second is activated about 2 hr after fertilization, coinciding with the onset of zygotic transcription. This paper reports that a cluster of zygotically expressed microRNAs (miRNAs) targets maternal mRNAs for turnover, as part of the zygotic degradation pathway. miRNAs are small noncoding RNAs that silence gene expression by repressing translation of their target mRNAs and by promoting mRNA turnover. Intriguingly, use of miRNAs to promote mRNA turnover during the maternal-to-zygotic transition appears to be a conserved phenomenon because a comparable role was reported for miR-430 in zebrafish (Giraldez, 2006). The finding that unrelated miRNAs regulate the maternal to zygotic transition in different animals suggests convergent evolution (Bushati, 2008).

The Drosophila miR-309 cluster contains eight microRNA (miRNA) genes, which encode six different miRNAs. Nucleotides 2 to 8 at the miRNA 5' end comprise the 'seed' region, which serves as the primary determinant of target specificity. The cluster encodes miRNAs with five distinct seed sequences, and so has the potential to regulate a broad spectrum of target messenger RNAs (mRNAs) (Bushati, 2008).

By using homologous recombination, a mutant was generated in which the 1.1 kb comprising the miR-309 cluster was deleted and replaced with green fluorescent protein (GFP). Northern-blot analysis was used to verify that the first and last miRNAs in the cluster, miR-309 and miR-6, were not produced in the mutant. Homozygous mutant animals completed embryogenesis with no apparent defects in patterning, but approximately 20% died as larvae at different larval stages. Some individuals stopped growing at the size of L2 larva and arrested at this developmental stage for a few days before dying. Approximately 80% of mutants survived to adulthood and were viable and fertile. Introduction of a transgene containing a 2.6 kb fragment of genomic DNA spanning the miRNA cluster restored survival of the mutants to normal levels. The mutant animals showed a developmental delay during larval stages. This delay was suppressed in simultaneously collected and staged mutant larvae carrying the rescue transgene. The phenotypes that result from complete deletion of the three miR-6 miRNA genes (together with the rest of the cluster mRNAs) contrast with the severe embryonic defects that were reported with the use of antisense 2'-O-methyl oligonucleotide injection to deplete miR-6 or miR-286 (Bushati, 2008).

RNA samples from precisely staged embryos were used to examine the expression of the miR-309 cluster during early embryogenesis. The levels were compared of mature miR-6 and miR-309 in these samples by quantitative real-time polymerase chain reaction (qPCR). Samples were normalized to two reference miRNAs, miR-310 and miR-184, which were found to be expressed at constant levels when normalized to total RNA. miR-6 and miR-309 were expressed at barely detectable levels in RNA collected from embryos during a 30 min period before the onset of zygotic transcription. The miRNAs were then strongly induced coincident with the onset of zygotic transcription. In situ hybridization analysis at this stage, showed expression of the miR-309 cluster primary transcript throughout the embryo, except in pole cells. This transcript was not detectable in miR-309 cluster mutant embryos (Bushati, 2008).

Although the mature miRNA products persist for some time, the expression of the primary transcript shows a dynamic spatial pattern by in situ hybridization. At the midpoint of cellularization, expression of the cluster is turned off at the posterior pole and in a stripe in the anterior region of the embryo. During gastrulation, expression is lost ventrally and laterally, resulting in transient stripes in the dorsal ectoderm. By the onset of germ-band elongation, the primary transcript was essentially undetectable, but in Northern blots, the mature miRNAs are detectable until larval stages (Bushati, 2008).

The miR-309 cluster is predicted to target many mRNAs, including those of several genes implicated in embryo patterning. However, immunolabelling for the detection of these proteins did not reveal alterations in their expression levels or patterns in the miR-309 cluster mutant. For example, the expression of the predicted miR-3-miR-309 target Ftz was compared with Even Skipped (which is not a predicted target). There was no striking difference between mutant and control embryos, consistent with the observation that miR-309 cluster mutant embryos did not show discernable embryonic patterning defects. The significance of the dynamics of spatial expression of the cluster miRNAs and the implied potential to regulate genes involved in embryonic patterning remains unclear (Bushati, 2008).

Given that the early onset of cluster miRNA expression does not appear to play a role in regulating zygotic mRNAs involved in patterning, attention was turned to their potential to regulate the maternal-to-zygotic-transition. Expression was compared of the miR-309 cluster to a high-resolution temporal gene expression profile of early embryonic development. mRNAs with a temporal expression profile most similar to that of the miR-309 cluster contained significantly fewer 7-mers complementary to miR-309 cluster miRNAs in their 3'untranslated regions (UTRs) than would be expected by chance. This suggests that these mRNAs have been under selection to reduce their regulation by the cluster miRNAs with which they are coexpressed. Reciprocally, 7-mer seed matches complementary to cluster miRNAs were enriched in the 3'UTRs of maternal transcripts that were strongly downregulated as miRNA expression increased. The same trends hold true for 6-mer seed matches to cluster miRNAs. For the 6-mer set, the correlation data are more significant because of overall larger numbers of miRNA targets in each bin (Bushati, 2008).

To investigate whether early zygotic miR-309 cluster miRNA expression might contribute to this downregulation, microarray analyses were performed of control and mutant embryos at 0-1 hr and 2-3 hr of embryonic development. During the first hour, miR-309 cluster miRNAs are expressed at barely detectable levels, whereas they are strongly induced during the 2-3 hr interval. Messenger RNA levels in control and miRNA mutant embryos were compared. Messenger RNAs whose expression was upregulated in the absence of the cluster miRNAs were examined with reference to two sets of maternal mRNAs that had previously been classified as being moderately or strongly downregulated during the maternal-to-zygotic transition. Forty-two of the 291 mRNAs (14%) that normally decrease by more than 3-fold between 2 and 3 hr of embryonic development were upregulated by over 1.5-fold in mutant embryos at this stage. This represents a 5-fold enrichment among the upregulated mRNAs and is statistically significant. The effect of the removal of the miRNAs was stronger in the group of the 32 maternal transcripts annotated to decrease by more than 10-fold at this stage. Thirty-five percent of these were upregulated in the mutant (12/32), a 12.5-fold enrichment (Bushati, 2008).

The degree of enrichment of these annotated gene sets among upregulated transcripts is likely to underestimate the true degree of correlation, because only 30% of the genome was included in the original classification of moderately or strongly downregulated maternal gene sets. To get a more complete picture, a similar analysis was performed on the larger set of maternal mRNAs. One thousand sixty-five mRNAs were classified as unstable maternal transcripts on the basis of expression profiling of RNA from unfertilized wild-type eggs and assessment of the degree of their destabilization over time. One hundred thirty-eight of the 1065 unstable maternal mRNAs were among the 410 mRNAs upregulated in cluster mutant embryos at 2-3 hr. This represents more than 4-fold enrichment and is statistically highly significant. There was no significant enrichment in 0-1 hr embryos (before the miRNAs are expressed). Much less enrichment was seen in the stable maternal class, which contains both stable transcripts and transcripts that are stable in unfertilized eggs but likely degraded by the zygotic pathway in fertilized embryos. For example, some of the stable maternal class mRNAs have been classified as 3× down or 10× down. Sixteen of these mRNAs were upregulated in the miRNA mutant and probably contribute to the 1.2-fold enrichment of mRNAs classified as maternal stable in this set. This analysis indicates that downregulation of maternal transcripts is impaired in the miRNA cluster mutant, suggesting that these miRNAs play a role in the zygotic pathway of maternal mRNA turnover (Bushati, 2008).

The foregoing observations suggest that the miRNA cluster and its targets have largely reciprocal temporal expression patterns, a situation analogous to the spatially reciprocal relationship between many miRNAs and their targets at later stages of embryogenesis and to the temporal relationship between the C. elegans heterochronic miRNAs and their targets. To assess the significance of these observations, the occurrence of miRNA cluster target sites among the regulated mRNAs was compared with what would be expected to occur by chance. Among the 410 transcripts upregulated in the miRNA cluster mutant, 96 contained 7-mers complementary to the seed of one or more cluster miRNAs. This represents a statistically significant enrichment of 1.8-fold (Bushati, 2008).

Among the mRNAs upregulated in cluster mutant embryos at 2-3 hr, mRNAs from a set of maternal mRNAs, which contained such 7-mer sites, were enriched 3.6-fold. The enrichment was 6.4-fold in the class of maternal mRNAs 3× downregulated containing such 7-mers and 48-fold in 10× downregulated set containing miR-309 cluster 7-mer sites. Importantly, no significant enrichment of 7-mers was observed in 0-1 hr embryos, prior to the onset of miRNA cluster expression. Comparable analysis for the larger set of mRNAs produced similar results. Maternal mRNAs containing target sites were enriched 2.5-fold and the set of unstable maternal mRNAs carrying target sites 6-fold among the mRNAs upregulated in cluster mutant embryos at 2-3 hr. Again, no significant enrichment was seen in the 0-1 hr samples (Bushati, 2008).

These statistical relationships suggest that the regulation of these mRNAs depends on the presence of the miRNA sites. To confirm that such sites are indeed functional, luciferase reporter constructs containing the 3' UTRs of 32 of the affected maternal mRNAs were prepared from the different functional categories mentioned above and expressed together with the miR-309 cluster in Drosophila S2 cells. Twenty-nine of the 32 reporters were statistically significantly downregulated upon miR-309 cluster expression, indicating that they carry functional miR-309 cluster target sites (Bushati, 2008).

The cluster encodes miRNAs with five different seed sequences, reflecting the capacity to regulate different sets of target mRNAs. To assess the contribution of individual miRNAs to the effects of the cluster as a whole, 7-mer seed matches complementary to individual miR-309 cluster miRNAs were analyzed. Four of the five unique seeds (miR-3 and 309 have the same seed sequence) were significantly enriched among the upregulated mRNAs at 2-3 hr but not at 0-1 hr. The magnitude of the enrichment and the statistical significance were stronger for miR-6, suggesting that it may contribute disproportionately to the effects of the cluster. This might be in part because miR-6 is present in three copies and so might be expressed at a higher level than the others. These data suggest that, with the possible exception of miR-286, the five distinct miRNAs encoded in the cluster act in concert to regulate a broad spectrum of mRNAs during the maternal-to-zygotic transition (Bushati, 2008).

SMAUG has been identified as a key component of the maternal system for maternal mRNA turnover in the embryo (Tadros, 2007), whereas the evidence presented above suggests that the miR-309 cluster acts zygotically to promote turnover of maternal mRNAs. A priori, these systems might be functionally related, acting in concert. Alternatively, they might represent independent systems. To explore these possibilities, the degree to which the sets of targets regulated by these two systems overlap was examined (Bushati, 2008).

Of the 1065 unstable maternal transcripts identified by Tadros (2007), 710 were identified as SMAUG targets by expression profiling of RNA from unfertilized eggs laid by smaug mutant flies (note: SMAUG is deposited maternally and acts on maternally deposited mRNAs). As mentioned before, 138 of the transcripts upregulated in the miR-309 cluster mutant at 2-3 hr were classified as unstable maternal transcripts, which represents more than 4-fold enrichment. Ninety-two of these transcripts were also targeted by SMAUG, which represents more than 4-fold enrichment. Of these, 20 (21.7%) had 7-mer seed matches complementary to cluster miRNAs in their 3' UTRs and so might represent a set of mRNAs potentially coregulated by the maternal and zygotic systems. Other mRNAs among the SMAUG targets were not affected in the miRNA cluster mutants -- for example, Hsp83, whose downregulation depends strongly on the SMAUG system. Of the 355 unstable transcripts that had been reported to be SMAUG independent, 46 were among the 410 mRNAs upregulated in the miR-309 cluster mutant embryos. This represents a more than 4-fold enrichment. Eighteen (39%) of these carry 7-mers complementary to miR-309 cluster miRNAs, an 8-fold enrichment. This set includes mRNAs such as orb, oskar, and exuperantia and may represent the set of mRNAs regulated mainly by the zygotic system. Together, these data suggest that the maternal and zygotic systems regulate distinct but overlapping sets of maternal mRNAs (Bushati, 2008).

The observation that some SMAUG targets also appear to be targets of the zygotic system raised the question of whether there might be a genetic interaction between the two systems. It can be expected that there might be an additive effect of removing two systems that share some common targets (if it is assumed that the common targets contribute to the mutant phenotype). To address this, it was asked whether removing one copy of maternal SMAUG would enhance the severity of the zygotic miR-309 cluster mutant phenotypes. No difference was observed in embryonic survival rates between miR-309 cluster mutants and those also lacking one copy of maternal SMAUG. However, there appeared to be a small reduction in survival of miR-309 cluster mutant larvae whose mothers lacked one copy of SMAUG, from 85% ± 5% to 69% ± 12%. This difference was, however, not statistically significant (t test = 0.06). The marginal reduction in survival might reflect an additive effect of perturbing both systems on their common targets. It is possible that a further reduction of SMAUG activity might result in a statistically significant effect. At present, though, it is not possible to conclude that there is an interaction that is more than additive between the two systems (Bushati, 2008).

These findings indicate that the early zygotic onset of miR-309 cluster miRNA expression acts to promote the turnover of many maternally deposited mRNAs. Failure to downregulate maternal mRNAs by this zygotic mechanism has knock-on effects on zygotic gene expression and may result in a late onset phenotype reflected by reduced survival and delayed larval development for many of the surviving animals. Elimination of the early zygotic expression of the miR-430 miRNA gene family also led to substantial misregulation of maternal mRNAs and to a late onset zygotic defect in Zebrafish (Giraldez, 2006). Although miRNAs have been shown to act to ensure a proper transition between maternal and zygotic gene expression programs in flies and fish, the miRNAs involved are not conserved. Perhaps the fact that miRNAs act in part by leading to mRNA deadenylation, and subsequent destabilization, provided a means to promote turnover of a selected set of maternally deposited mRNAs. miRNAs may have been co-opted independently during evolution to fulfill a comparable function in different animals. The mechanistic basis for their action and the biological output are both conserved, but the miRNAs themselves and the identity of their targets are not. This may be an example of convergent evolution (Bushati, 2008).

Spatial regulation of nanos is required for its function in dendrite morphogenesis: A role for Glorund and Smaug

Spatial control of mRNA translation can generate cellular asymmetries and functional specialization of polarized cells like neurons. A requirement for the translational repressor Nanos (Nos) in the Drosophila larval peripheral nervous system (PNS) implicates translational control in dendrite morphogenesis. Nos was first identified by its requirement in the posterior of the early embryo for abdomen formation. Nos synthesis is targeted to the posterior pole of the oocyte and early embryo through translational repression of unlocalized nos mRNA coupled with translational activation of nos mRNA localized at the posterior pole. Abolishment of nos localization prevents abdominal development, whereas translational derepression of unlocalized nos mRNA suppresses head/thorax development, emphasizing the importance of spatial regulation of nos mRNA. Loss and overexpression of Nos affect dendrite branching complexity in class IV dendritic arborization (da) neurons, suggesting that nos also might be regulated in these larval sensory neurons. This study shows that localization and translational control of nos mRNA are essential for da neuron morphogenesis. RNA-protein interactions that regulate nos translation in the oocyte and early embryo also regulate nos in the PNS. Live imaging of nos mRNA shows that the cis-acting signal responsible for posterior localization in the oocyte/embryo mediates localization to the processes of class IV da neurons but suggests a different transport mechanism. Targeting of nos mRNA to the processes of da neurons may reflect a local requirement for Nos protein in dendritic translational control (Brechbiel, 2008).

Translational activation of nos at the posterior pole is tightly coupled to translational repression of unlocalized nos mRNA to prevent accumulation of Nos in the anterior of the embryo, where Nos suppresses anterior development. Because nos localization during oogenesis is inefficient, this linkage is essential to silence nos mRNA that remains distributed throughout the bulk cytoplasm. Translational repression of nos mRNA is mediated by a structural motif, the translational control element (TCE), within the nos 3'UTR. TCE function requires the formation of two stem loops, designated as II and III, that have temporally distinct activities. Whereas stem-loop III mediates repression of nos during oogenesis, through its interaction with Glorund (Glo), stem loop II is responsible for repression of nos in the early embryo, through its interaction with a different repressor, Smaug (Smg) (Brechbiel, 2008).

Replacement of the nos 3'UTR by α-tubulin 3'UTR sequences (nos-tub3'UTR) abolishes nos localization and translational repression, leading to unrestricted synthesis of Nos and defects in anterior development. GAL4 mediated overexpression of a UAS-nos-tub3'UTR transgene in class IV da neurons also is deleterious, causing decreased branching complexity. This overexpression phenotype is ameliorated by reinsertion of the nos TCE. The observation that both loss and overexpression of nos cause similar defects indicates that although nos is required for dendrite morphogenesis, the level of Nos protein must be carefully modulated in da neurons. Moreover, the ability of the TCE to suppress the toxicity of nos mRNA overexpression in da neurons suggests that it may normally function to control Nos levels in the PNS. Attempts were therefore made to determine whether endogenous nos is regulated by the TCE in da neurons (Brechbiel, 2008).

Ectopic expression studies have identified several additional somatic cell types in which the TCE can repress translation, including neuroendocrine cells and the dorsal pouch epithelium. However, TCE function in the dorsal pouch does not depend on the Glo or Smg binding sites but requires a distinct sequence motif with homology to the Bearded (Brd) box. Mutation of the Brd box-like motif does not abrogate the ability of the TCE to suppress excess nos activity in da neurons. Consequently, to determine whether endogenous nos mRNA might be regulated by the TCE, da neurons were analyzed in glo and smg mutant larvae (Brechbiel, 2008).

Larvae mutant for glo or smg survive until third instar stage, permitting examination of the effect of eliminating either repressor on dendrite morphology of da neurons. Compared to wild-type class IV da neurons, glo mutant larvae show a significant decrease in the number of higher order dendritic branches as reflected by a decreased number of terminal dendritic processes. Because glo mutant larvae exhibit additional defects, glo function was disrupted specifically in class IV da neurons either by using GAL4⁴⁷⁷ to express a UAS-gloRNAi transgene or by using the MARCM method to generate mosaic animals. In both cases glo mutant da neurons show decreased branching complexity. Mutation of smg or GAL4⁴⁷⁷-mediated overexpression of a UAS-smg transgene also causes loss of high-order branches. Larvae doubly mutant for glo and smg do not show a more severe phenotype than larvae mutant for either gene alone, suggesting that each repressor contributes independently. Thus, defects due to loss or overexpression of the repressors are consistent with defects caused by loss or overexpression of nos. Due to the inadequacy of anti-Nos antibodies, changes in Nos protein levels could not be monitored in glo and smg mutant da neurons. However, when combined with the analysis of Glo and Smg binding site mutations, these results strongly support a role for glo and smg in regulation of nos for dendrite morphogenesis (Brechbiel, 2008).

In the oocyte, Glo binds specifically to the distal double-stranded helix of TCE stem-loop III (the Glo Recognition Helix or GRH. In the embryo Smg interacts with nos TCE stem loop II via nucleotides within the loop designated as the Smg Recognition Element (SRE). A second SRE located downstream of the TCE in the nos 3'UTR appears to act redundantly. To determine whether the defects observed in glo and smg mutant da neurons are due to loss of TCE-mediated repression, whether mutation of the nos GRH or SREs produces a similar phenotype was tested. Mutations that disrupt both SREs (SREs^-), the binding site for Glo (GRH^-), or the SREs and GRH (SREs^-GRH^-) together were introduced into the gnos transgene. The resulting gnosSREs^-, gnosGRH^-, and gnosSREs^-GRH^- transgenes all produce mRNAs that show wild-type localization in the early embryo but whose translation is not restricted to the posterior pole. When compared to larvae expressing the wild-type gnos transgene, branching complexity is significantly reduced in da neurons of larvae expressing gnosSREs^-, gnosGRH^-, and gnosSREs^-GRH^- transgenes. Moreover, each of these transgenes behaves similarly to the gnos-tub3'UTR transgene, which lacks the entire nos 3'UTR, indicating that mutation of the GRH and/or SREs is sufficient to disrupt nos regulation in the PNS. Together, these results show that TCE-mediated regulation of nos in da neurons is essential for dendrite morphogenesis. Furthermore, the finding that the same phenotype is produced by either eliminating the repressors or mutating their binding sites provides strong evidence that this regulation is mediated by Glo and Smg (Brechbiel, 2008).

In many cell types protein synthesis is spatially regulated through the transport of translationally silent mRNAs and activation of these mRNAs at the target destination. Linkage of translation and localization serves not only to prevent premature accumulation of nos during transit to the oocyte posterior but also to silence the large pool of nos that remains unlocalized due to inefficient posterior localization. It cannot yet be distinguish whether localization of nos in da neurons is similarly inefficient or whether translational repression of nos serves primarily to repress translation during transport. However, the deleterious effect on dendrite morphogenesis caused by mutations that disrupt TCE function show that, as for maternally synthesized nos mRNA, localization alone is not sufficient to modulate its activity (Brechbiel, 2008).

It is concluded that nos plays an important role in dendrite morphogenesis, and this study shows that nos function in da neurons requires spatial regulation of nos mRNA. Cis-acting sequences and two cognate factors that control nos mRNA localization and/or translation in the oocyte and early embryo are redeployed during larval stages to regulate localization and translation of nos in da neurons. Localization of nos mRNA to the processes of class IV da neurons is essential for dendritic branching. Movement of RNA particles in neurons of intact animals was shown to take place, and analysis of nos mRNA particle movement suggests that nos localization occurs by different mechanisms depending on cellular context. Taken together, these results support a role for Nos as a local regulator of translation in the PNS (Brechbiel, 2008).

In the early embryo Nos functions in a complex with the RNA-binding protein Pumilio (Pum) to repress hunchback mRNA translation, thereby promoting abdominal development. Whereas Pum is produced throughout the embryo, restriction of Nos synthesis to the posterior limits the spatial domain of the repressor complex. Mutations in nos and pum produce similar defects in dendrite morphogenesis, suggesting that Nos and Pum also act together to repress translation in da neurons. Thus, spatial regulation of nos may serve a similar function in the PNS as it does in the early embryo, by restricting the activity of the Nos/Pum repressor complex to dendrites (Brechbiel, 2008).

Aviv, T., Lin, Z., Lau, S., Rendl, L. M., Sicheri, F. and Smibert, C. A (2003). The RNA-binding SAM domain of Smaug defines a new family of post-transcriptional regulators. Nat. Struct. Biol. 10: 614-621. 12858164

Bashirullah, A., et al. (1999). Joint action of two RNA degradation pathways controls the timing of maternal transcript elimination at the midblastula transition in Drosophila melanogaster. EMBO J. 18(9): 2610-20. Medline abstract: 10228172

Benoit, B., et al. (2009). An essential role for the RNA-binding protein Smaug during the Drosophila maternal-to-zygotic transition. Development 136(6): 923-32. PubMed Citation: 19234062

Bergsten, S. E. and Gavis, E. R. (1999). Role for mRNA localization in translational activation but not spatial restriction of Nanos RNA. Development 126: 659-669. PubMed Citation: 9895314

Brechbiel, J. L. and Gavis, E. R. (2008). Spatial regulation of nanos is required for its function in dendrite morphogenesis. Curr. Biol. 18(10): 745-50. PubMed Citation: 18472422

Breitwieser, W., Markussen, F.H., Horstmann, H., and Ephrussi, A. (1996). Oskar protein interaction with Vasa represents an essential step in polar granule assembly. Genes Dev. 10: 2179-2188. PubMed Citation: 8804312

Bushati, N., Stark, A., Brennecke, J. and Cohen, S. M. (2008). Temporal reciprocity of miRNAs and their targets during the maternal-to-zygotic transition in Drosophila. Curr. Biol. 18(7): 501-6. PubMed Citation: 18394895

Crest, J., Oxnard, N., Ji, J. Y. and Schubiger, G. (2007). Onset of the DNA replication checkpoint in the early Drosophila embryo. Genetics 175: 567-584. PubMed Citation: 17151243

Crucs, S., Chatterjee, S. and Gavis, E. R. (2000). Overlapping but distinct RNA elements control repression and activation of nanos translation. Molec. Cell 5: 457-467. PubMed Citation: 10882131

Dahanukar, A. and Wharton, R. P. (1996). The Nanos gradient in Drosophila embryos is generated by translational regulation. Genes Dev. 10: 2610-20. PubMed Citation: 8895662

Dahanukar, A., Walker, J. A. and Wharton, R. P. (1999). Smaug, a novel RNA-binding protein that operates a translational switch in Drosophila. Mol. Cell 4: 209-218. PubMed Citation: 10488336

Dostie, J., Ferraiuolo, M., Pause, A., Adam, S. A., Sonenberg, N. (2000). A novel shuttling protein, 4E-T, mediates the nuclear import of the mRNA 5' cap-binding protein, eIF4E. EMBO J. 19: 3142-3156. 10856257

Forrest, K. M., Clark, I. E., Jain, R. A. and Gavis, E. R. (2004). Temporal complexity within a translational control element in the nanos mRNA. Development 131: 5849-5857. 15525666

Gavis, E. R. and Lehmann, R. (1994). Translational regulation of nanos by RNA localization. Nature 369: 315-8. PubMed Citation: 7514276

Gerber, A. P., Luschnig, S., Krasnow, M. A., Brown, P. O. and Herschlag, D. (2006). Genome-wide identification of mRNAs associated with the translational regulator PUMILIO in Drosophila melanogaster. Proc. Natl. Acad. Sci. 103(12): 4487-92. Medline abstract: 16537387

Giraldez, A.J., Mishima, Y., Rihel, J., Grocock, R.J., Van Dongen, S., Inoue, K., Enright, A.J. and Schier, A.F. (2006). Zebrafish MiR-430 promotes deadenylation and clearance of maternal mRNAs. Science 312: 75-79. PubMed Citation: 16484454

Green, J. B., Gardner, C. D., Wharton, R. P. and Aggarwal, A. K. (2003). RNA recognition via the SAM domain of Smaug. Mol. Cell 11: 1537-1548. 12820967

Jeske, M., Meyer, S., Temme, C., Freudenreich, D. and Wahle, E. (2006). Rapid ATP-dependent deadenylation of nanos mRNA in a cell-free system from Drosophila embryos. J. Biol. Chem. 281: 25124-25133. Medline abstract: 16793774

Kalifa, Y., Huang, T., Rosen, L. N., Chatterjee, S. and Gavis, E. R. (2006). Glorund, a Drosophila hnRNP F/H Homolog, is an ovarian repressor of nanos translation. Dev. Cell 10: 291-301. Medline abstract: 16516833

Luo, H., Li, X., Claycomb, J.M. and Lipshitz, H.D. (2016). The Smaug RNA-binding protein is essential for microRNA synthesis during the Drosophila maternal-to-zygotic transition. G3 (Bethesda) [Epub ahead of print]. PubMed ID: 27591754

Morris, J. Z., Hong, A., Lilly, M. A. and Lehmann, R. (2005). twin, a CCR4 homolog, regulates cyclin poly(A) tail length to permit Drosophila oogenesis. Development 132(6): 1165-74. Medline abstract: 15703281

Nelson, M. R., Leidal, A. M. and Smibert, C. A. (2004). Drosophila Cup is an eIF4E-binding protein that functions in Smaug-mediated translational repression. EMBO J. 23: 150-159 . 14685270

Pinder, B. D. and Smibert, C. A. (2013). microRNA-independent recruitment of Argonaute 1 to nanos mRNA through the Smaug RNA-binding protein. EMBO Rep 14: 80-86. PubMed ID: 23184089

Rouget, C., et al. (2010). Maternal mRNA deadenylation and decay by the piRNA pathway in the early Drosophila embryo. Nature 467(7319): 1128-32. PubMed Citation: 20953170

Semotok, J. L., et al. (2005). Smaug recruits the CCR4/POP2/NOT deadenylase complex to trigger maternal transcript localization in the early Drosophila embryo. Curr. Biol. 15(4): 284-94. Medline abstract: 15723788

Smibert, C. A., Wilson, J. E., Kerr, K. and Macdonald, P. M. (1996). smaug protein represses translation of unlocalized Nanos mRNA in the Drosophila embryo. Genes Dev. 10: 2600-2609. PubMed Citation: 8895661

Smibert, C. A., et al. (1999). Smaug, a novel and conserved protein, contributes to repression of nanos mRNA translation in vitro. RNA 5(12): 1535-47. Medline abstract: 10606265

Tadros, W., et al. (2007). SMAUG is a major regulator of maternal mRNA destabilization in Drosophila and its translation is activated by the PAN GU kinase. Dev. Cell 12(1): 143-55. Medline abstract: 17199047

Temme, C., Zaessinger, S., Meyer, S., Simonelig, M. and Wahle, E. (2004). A complex containing the CCR4 and CAF1 proteins is involved in mRNA deadenylation in Drosophila. EMBO J. 23(14): 2862-71. Medline abstract: 15215893

Vardy, L. and Orr-Weaver, T. (2007). The Drosophila PNG kinase complex regulates the translation of Cyclin B. Dev. Cell 12: 157-166. Medline abstract: 17199048

Zaessinger, S., Busseau, I. and Simonelig, M. (2006). Oskar allows nanos mRNA translation in Drosophila embryos by preventing its deadenylation by Smaug/CCR4. Development 133(22): 4573-83. Medline abstract: 17050620

date revised: 21 November 2016

Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.

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