InteractiveFly: GeneBrief

armitage: Biological Overview | Regulation | Developmental Biology | Effects of Mutation | Evolutionary Homologs | References

Gene name - armitage

Synonyms - CG11513

Cytological map position - 63E1

Function - enzyme

Keywords - RNAi pathway, oogenesis, A/P axis polarization, silencing of Oskar mRNA

Symbol - armi

FlyBase ID: FBgn0041164

Genetic map position - 3L

Classification - RNA helicase activity

Cellular location - cytoplasmic

NCBI links: Precomputed BLAST | Entrez Gene

Polarization of the microtubule cytoskeleton during early oogenesis is required to specify the posterior of the Drosophila oocyte: this is essential for asymmetric mRNA localization during mid-oogenesis and for embryonic axis specification. The posterior determinant oskar mRNA is translationally silent until mid-oogenesis. Mutations in armitage (armi) and in three components of the RNAi pathway disrupt oskar mRNA translational silencing, polarization of the microtubule cytoskeleton, and posterior localization of oskar mRNA. armitage encodes a homolog of SDE3, a presumptive RNA helicase involved in posttranscriptional gene silencing (RNAi) in Arabidopsis, and is required for RNAi in Drosophila ovaries. Armitage forms an asymmetric network associated with the polarized microtubule cytoskeleton and is concentrated with translationally silent oskar mRNA in the oocyte. It is concluded that RNA silencing is essential for establishment of the cytoskeletal polarity that initiates embryonic axis specification and for translational control of oskar mRNA (Cook, 2004).

As a RNA helicase involved in posttranscriptional gene silencing, armi mutant male germ cells fail to silence Stellate, a gene regulated endogenously by RNAi, and lysates from armi mutant ovaries are defective for RNAi in vitro. Native gel analysis of protein-siRNA complexes in wild-type and armi mutant ovary lysates suggests that armi mutants support early steps in the RNAi pathway but are defective in the production of active RNA-induced silencing complex (RISC), which mediates target RNA destruction in RNAi. These results suggest that armi is required for RISC maturation (Tomari, 2004).

Asymmetric mRNA localization produces local protein concentrations that are critical to processes ranging from mating-type switching in yeast to synaptic plasticity in mammals. To produce protein at the right time and place within the cell, transcripts must be translationally silent during transport and remain silent until the protein products are needed. Embryonic axis specification in Drosophila is a well-studied developmental process that depends on spatial and temporal coordination of mRNA localization and translation. bicoid (bcd) mRNA encodes the primary anterior morphogen and is localized to the anterior of the developing oocyte during stage 9. However, bcd is not translated until egg activation, when the transcript is polyadenylated in the cytoplasm, recruited to polysomes, and translated to produce an anterior to posterior protein gradient. Asymmetric localization of oskar (osk) mRNA during mid-oogenesis is essential for posterior patterning and for germ cell formation. osk transcript is produced throughout oogenesis, but remains translationally silent until localization to the oocyte posterior pole during stage 9. Following localization and translational activation, Osk protein triggers the assembly of pole plasm, which specifies the germline and is required for abdominal patterning. Translational repression of osk during stages 7 and 8 is mediated by cis-elements in the osk 3' UTR that are bound by, which interacts with Cup, an eIF4E binding protein. Cup is also required for osk translational silencing and may function with Bruno to silence osk mRNA translation by blocking eIF4E interactions with other components of the translation initiation machinery (Cook, 2004).

Unlike bcd and osk mRNAs, grk mRNA is translated throughout oogenesis, producing a TGFα-related growth factor that initiates two spatially and temporally distinct signaling events that specify the anterior-posterior (A/P) and dorsal-ventral (D/V) axes. During early oogenesis, microtubules originate from the posterior of the oocyte and mediate posterior localization of grk mRNA and Grk protein. Grk signals to the overlying follicle cells to induce posterior differentiation. During mid-oogenesis, the posterior follicle cells signal back to the oocyte, inducing reorganization of the oocyte microtubule network, which is essential for the asymmetric localization of bcd, osk, and grk mRNAs. Following reorganization of the microtubule network, grk mRNA localizes to the dorsal-anterior corner of the oocyte, and local Grk signaling induces dorsal differentiation of the adjacent follicle cells. The correct organization of oocyte microtubules early in oogenesis thus initiates a series of signaling events that specify the axes of the oocyte (Cook, 2004).

The armitage gene has been shown to be required for anterior-posterior polarization of the microtubule cytoskeleton and translational silencing of osk mRNA during early oogenesis and for osk mRNA localization and posterior and dorsal-ventral patterning during mid-oogenesis. The armi gene encodes a putative RNA helicase most closely related to SDE3, which is required for posttranscriptional gene silencing (PTGS) in Arabidopsis (Dalmay, 2001; Willmann, 2001). PTGS is an evolutionarily conserved RNA silencing mechanism closely related to RNA interference (RNAi). Since armi is required for RNAi and efficient assembly of the RNA-induced silencing complex (RISC) in ovary extracts (Tomari, 2004), Armi/SDE3 class RNA helicases appear to have a conserved role in RNAi. Three additional genes shown to be implicated in RNAi are also required for osk mRNA translational silencing and microtubule reorganization during early oogenesis, indicating that the RNAi system is required for axial polarization of the oocyte. Finally, Armi protein is shown to be concentrated in the oocyte with osk mRNA. It is speculated that this asymmetric localization may spatially restrict RNA silencing activity and increase the efficiency of target recognition and thus help to establish the functional asymmetries that initiate embryonic axis specification (Cook, 2004).

Thus armi is required for initial polarization of the microtubule cytoskeleton and for temporal regulation of osk mRNA translation. Armi is required to repress osk translation during early oogenesis, but does not alter osk mRNA levels. armi is also required for Stellate silencing during spermatogenesis (Tomari, 2004), which requires small homologous miRNAs and the RNAi components Spn-E and Aub. Armi is also required for RNAi and efficient RISC assembly in ovary extracts (Tomari, 2004). These findings strongly suggest that the RNAi system is required for an early step in the axis specification pathway (Cook, 2004).

Consistent with this hypothesis, the RNAi components Spn-E, Aub, and Mael are also required for osk mRNA silencing and for polarization of the microtubule cytoskeleton during early oogenesis. Aub enhances osk translation during mid-oogenesis. However, aub disrupts posterior localization of osk mRNA during mid-oogenesis, and posterior localization is required for efficient osk translation. It is therefore speculated that the reduced Osk protein levels during later stages of oogenesis are secondary to defects in posterior patterning during early oogenesis, when Aub and other RNA silencing components are required to establish the microtubule asymmetries required to specify the posterior pole (Cook, 2004).

While the RNA silencing system represses osk during stages 1 to 6, other studies indicate that products of the bruno/arrest, cup, and Bicaudal-C genes are essential for osk silencing during stages 6 to 8. The role of Bicaudal-C in osk silencing is unclear. However, Bruno binds to three sites in the osk 3' UTR, called Bruno response elements (BREs), and deletion of the BREs leads to osk translation during stages 7 and 8 and severe patterning defects. Recent data show that Cup associates with both Bruno and the 5'-cap binding factor eIF4E (Nakamura, 2004). This suggests that Cup and Bruno may function together to repress osk translation by sequestering the 5' end of osk mRNA, thereby blocking translation initiation. Bruno and Cup do not appear to play a role in osk mRNA translational repression during stages 1-6, when the RNAi pathway is required for silencing. The biological reason for this two-step translational control mechanism is unclear, but may be linked to changing functions for the RNAi system during oogenesis (Cook, 2004).

All of the RNAi mutations examined in this study produce nearly identical defects in microtubule polarization and osk silencing during early oogenesis and posterior and D/V patterning during mid-oogenesis. The early defects in osk silencing and microtubule organization appear to reflect independent functions for the RNAi system, since mutations in osk do not suppress the cytoskeletal defects in armi mutants, and forced premature expression of Osk protein from a transgene does not induce changes in microtubule organization. The defects in anterior-posterior polarization of the microtubule cytoskeleton during early oogenesis, by contrast, may directly lead to the posterior and D/V patterning defects observed in mid-oogenesis. Microtubule-dependent posterior localization of grk mRNA in early oogenesis is thought to facilitate Grk signaling from the oocyte to the posterior follicle cells: this initiates a chain of signaling events that trigger microtubule reorganization and mRNA localization during mid-oogenesis. Thus, defects in microtubule polarization associated with RNAi mutations are likely the primary cause of later defects in axial patterning (Cook, 2004).

The spectrum of defects observed in RNAi mutants suggests that the RNA silencing machinery targets multiple processes during early oogenesis. The endogenous miRNAs that mediate RNA silencing are predicted to bind complementary sequences in the 3' UTRs of numerous target transcripts, suggesting that they may coordinate translational control of gene cassettes during complex biological processes (Stark, 2003). A computational screen for miRNA targets has identified osk mRNA, kinesin heavy chain mRNA, and transcripts encoding several other cytoskeletal proteins involved in oogenesis as targets for the miR-280 miRNA (Stark, 2003). It is interesting to note that kinesin, like the RNAi components, is required for posterior and D/V axis specification. This motor also drives ooplasmic streaming during late oogenesis, and mutations in mael lead to premature ooplasmic streaming, which could reflect overexpression of kinesin due to defects in silencing. These observations raise the possibility that the RNAi system, through miR-280 and other miRNAs, coordinates axis specification by silencing osk mRNA and simultaneously regulating genes involved in microtubule function (Cook, 2004).

Armi is asymmetrically localized during early oogenesis, when it is concentrated in the oocyte with osk mRNA. This finding raises the possibility that asymmetric Armi localization plays a role establishing developmental asymmetries during early oogeneis. Tomari (2004) has shown that Armi is required for mRNA target cleavage and RISC maturation in vitro. Armi could therefore promote local RISC assembly and thus increase the efficiency of osk mRNA silencing. This could play a role in regulating other transcripts that accumulate in the oocyte. Local increases in RISC activation could also lead to oocyte-specific silencing of transcripts that are uniformly distributed within the oocyte-nurse cell complex. The molecular, genetic, and cytological tools available in Drosophila should allow direct tests of these possibilities (Cook, 2004).


Telomeric trans-silencing: an epigenetic repression combining RNA silencing and heterochromatin formation

The study of P-element repression in Drosophila led to the discovery of the telomeric Trans-Silencing Effect (TSE), a repression mechanism by which a transposon or a transgene inserted in subtelomeric heterochromatin (Telomeric Associated Sequence or TAS) has the capacity to repress in trans in the female germline, a homologous transposon, or transgene located in euchromatin. TSE shows variegation among egg chambers in ovaries when silencing is incomplete. This study reports that TSE displays an epigenetic transmission through meiosis, which involves an extrachromosomal maternally transmitted factor. This silencing is highly sensitive to mutations affecting both heterochromatin formation (Su(var)205 encoding Heterochromatin Protein 1 and Su(var)3-7) and the repeat-associated small interfering RNA (or rasiRNA) silencing pathway (aubergine, homeless, armitage, and piwi). In contrast, TSE is not sensitive to mutations affecting r2d2, which is involved in the small interfering RNA (or siRNA) silencing pathway, nor is it sensitive to a mutation in loquacious, which is involved in the micro RNA (or miRNA) silencing pathway. These results, taken together with the recent discovery of TAS homologous small RNAs associated to PIWI proteins, support the proposition that TSE involves a repeat-associated small interfering RNA pathway linked to heterochromatin formation, which was co-opted by the P element to establish repression of its own transposition after its recent invasion of the D. melanogaster genome. Therefore, the study of TSE provides insight into the genetic properties of a germline-specific small RNA silencing pathway (Josse, 2007; full text of article).

Repression of transposable elements (TEs) involves complex mechanisms that can be linked to either small RNA silencing pathways or chromatin structure modifications depending on the species and/or the TE family. Drosophila species are particularly relevant to the study of these repression mechanisms since some families of TEs are recent invaders, allowing genetic analysis to be carried out on strains with or without these TEs. In some cases, crossing these two types of strains induces hybrid dysgenesis, a syndrome of genetic abnormalities resulting from TE mobility. In D. virilis, repression of hybrid dysgenesis has been correlated to RNA silencing since small RNAs of the retroelement Penelope, responsible for dysgenesis, were detected in nondysgenic embryos but not in dysgenic embryos. In D. melanogaster, repression of retrotransposons can be established by noncoding fragments of the corresponding element (I factor, ZAM, and Idefix) and can be in some cases (gypsy, mdg1, copia, Het-A, TART, and ZAM, Idefix) sensitive to mutations in genes from the Argonaute family involved in small RNA silencing pathways. In the same species, strong repression of the DNA P TE, by a cellular state that has been called 'P cytotype', can be established by one or two telomeric P elements inserted in heterochromatic 'Telomeric Associated Sequences' (TAS) at the 1A cytological site corresponding to the left end of the X chromosome. This includes repression of dysgenic sterility resulting from P transposition. This P cytotype is sensitive to mutations affecting both Heterochromatin Protein 1 (HP1) (Ronsseray, 1996) and the Argonaute family member AUBERGINE (Reiss, 2004). P repression corresponds to a new picture of TE repression shown, using an assay directly linked to transposition, to be affected by heterochromatin and small RNA silencing mutants (Josse, 2007).

In the course of the study of P cytotype, a new silencing phenomenon has been discovered. Indeed, a P-lacZ transgene or a single defective P element inserted in TAS can repress expression of euchromatic P-lacZ insertions in the female germline in trans, if a certain length of homology exists between telomeric and euchromatic insertions. This homology-dependent silencing phenomenon has been termed Trans-Silencing Effect (TSE) (Roche, 1998). Telomeric transgenes, but not centromeric transgenes, can be silencers and all euchromatic P-lacZ insertions tested can be targets. TSE is restricted to the female germline and has a maternal effect since repression occurs only when the telomeric transgene is maternally inherited (Ronsseray, 2001). Further, when TSE is not complete, variegating germline lacZ repression is observed from one egg chamber to another, suggesting a chromatin-based mechanism of repression. Recently, an extensive analysis of small RNAs complexed with PIWI family proteins (AUBERGINE, PIWI, and AGO3) was performed in the Drosophila female germline. The latter study showed that most of the RNA sequences associated to these proteins derive from TEs. TSE corresponds likely to such a situation (Josse, 2007).

This study analyzed the genetic properties of TSE and shows that it has an epigenetic transmission through meiosis, which involves an extrachromosomal maternally transmitted stimulating component. Further, in order to investigate the mechanism behind TSE, a candidate gene analysis was performed to identify genes whose mutations impair TSE. It was found that TSE is strongly affected both by mutations in genes involved in heterochromatin formation and in the recently discovered small RNA silencing pathway called 'repeat-associated small interfering RNAs' (rasiRNA) pathway. In contrast, this study shows that TSE is not sensitive to genes specific to the classical RNA interference pathway linked to small interfering RNAs (siRNA) or to the micro RNA (miRNA) pathway. This suggests thus that TSE involves a rasiRNA pathway linked to heterochromatin formation and that such a mechanism, working in the germline, may underlie epigenetic transmission of repression through meiosis (Josse, 2007).

A regulatory circuit for piwi by the large Maf gene traffic jam in Drosophila

PIWI-interacting RNAs (piRNAs) silence retrotransposons in Drosophila germ lines by associating with the PIWI proteins Argonaute 3 (AGO3), Aubergine (Aub) and Piwi. piRNAs in Drosophila are produced from intergenic repetitive genes and piRNA clusters by two systems: the primary processing pathway and the amplification loop. The amplification loop occurs in a Dicer-independent, PIWI-Slicer-dependent manner. However, primary piRNA processing remains elusive. This study analysed piRNA processing in a Drosophila ovarian somatic cell line where Piwi, but not Aub or AGO3, is expressed; thus, only the primary piRNAs exist. In addition to flamenco, a Piwi-specific piRNA cluster, traffic jam (tj), a large Maf gene, was determined as a new piRNA cluster. piRNAs arising from tj correspond to the untranslated regions of tj messenger RNA and are sense-oriented. piRNA loading on to Piwi may occur in the cytoplasm. zucchini, a gene encoding a putative cytoplasmic nuclease, is required for tj-derived piRNA production. In tj and piwi mutant ovaries, somatic cells fail to intermingle with germ cells and Fasciclin III is overexpressed. Loss of tj abolishes Piwi expression in gonadal somatic cells. Thus, in gonadal somatic cells, tj gives rise simultaneously to two different molecules: the TJ protein, which activates Piwi expression, and piRNAs, which define the Piwi targets for silencing (Saito, 2009).

This study has uncovered two functions of tj in the regulation of piwi's functions. In gonadal somatic cells, TJ supposedly controls transcription of various genes. This study indicates that piwi is highly likely to be one of the genes under strong TJ control because loss of tj in gonadal somatic cells abolished Piwi expression. Further support for the hypothesis that TJ controls piwi expression was provided by DNA sequences near the putative transcriptional start site of the piwi gene, which show a weak but significant similarity to the Maf binding consensus sequence, and which were bound with TJ in OSCs. Thus, the first function of tj is to activate the expression of Piwi in gonadal somatic cells. The second function is to supply piRNAs for Piwi. Without the supplement of tj piRNAs, Piwi would lose the activity to target genes that should be silenced by Piwi and the tj-piRNA complex. A likely target of such silencing is FasIII because FasIII, a cell adhesion molecule concentrated at the hub cell junction, is ectopically overexpressed in other somatic cells in tj larval testes. Indeed, the FasIII expression level was higher in piwi mutant testes than in control testes. Some of the tj piRNAs identified in this study showed strong complementarity to the FasIII primary transcript. Although all the tj-derived piRNAs are sense-oriented and thus unlikely to target the tj mRNA, given the nuclear localization of Piwi, it is conceivable that the Piwi-piRNA complex could associate with the tj gene (Saito, 2009).

These findings suggest a novel regulatory circuit where tj mRNA simultaneously produces two types of functional molecules: TJ protein, which activates expression of Piwi, and piRNAs, which are loaded on to Piwi to silence specific target genes, such as FasIII and other, as yet undiscovered, genes (Saito, 2009).

AGO3 Slicer activity regulates mitochondria-nuage localization of Armitage and piRNA amplification

In Drosophila melanogaster the reciprocal 'Ping-Pong' cycle of PIWI-interacting RNA (piRNA)-directed RNA cleavage catalyzed by the endonuclease (or 'Slicer') activities of the PIWI proteins Aubergine (Aub) and Argonaute3 (AGO3) has been proposed to expand the secondary piRNA population. However, the role of AGO3/Aub Slicer activity in piRNA amplification remains to be explored. This study shows that AGO3 Slicer activity is essential for piRNA amplification and that AGO3 inhibits the homotypic Aub:Aub Ping-Pong process in a Slicer-independent manner. It was also found that expression of an AGO3 Slicer mutant causes ectopic accumulation of Armitage, a key component in the primary piRNA pathway, in the Drosophila melanogaster germline granules known as nuage. AGO3 also coexists and interacts with Armitage in the mitochondrial fraction. Furthermore, AGO3 acts in conjunction with the mitochondria-associated protein Zucchini to control the dynamic subcellular localization of Armitage between mitochondria and nuage in a Slicer-dependent fashion. Collectively, these findings uncover a new mechanism that couples mitochondria with nuage to regulate secondary piRNA amplification (Huang, 2014).

Protein Interactions

Ovary Lysate Recapitulates RNAi In Vitro

Drosophila syncitial blastoderm embryo lysate has been used widely to study the RNAi pathway. However, armi flies lay few eggs, making it difficult to collect enough embryos to make lysate. To surmount this problem, lysates were prepared from ovaries manually dissected from wild-type or mutant females. Approximately 10 μl of lysate can be prepared from ~50 ovaries (Tomari, 2004).

The well-characterized siRNA-directed mRNA cleavage assay (Elbashir, 2001a, 2001b) was used to evaluate the capacity of ovary lysate to support RNAi in vitro. Incubation in ovary lysate of a 5' 32P-cap-radiolabeled firefly luciferase mRNA target with a complementary siRNA duplex yielded the 5' cleavage product diagnostic of RNAi. siRNAs containing 5' hydroxyl groups are rapidly phosphorylated in vitro and in vivo, but modifications that block phosphorylation eliminate siRNA activity. Replacing the 5' hydroxyl of the antisense siRNA strand with a 5' methoxy group completely blocks RNAi in the ovary lysate. In Drosophila, siRNAs bearing a single 2'-deoxy nucleotide at the 5' end are poor substrates for the kinase that phosphorylates 5' hydroxy siRNAs (Nykanen, 2001). A comparison of initial cleavage rates shows that in ovary lysate, target cleavage was slower for siRNAs with a 2'-deoxy nucleotide at the 5' end of the antisense strand than for standard siRNAs. Furthermore, the rate of target cleavage was fastest when the siRNA was phosphorylated before its addition to the reaction. A similar enhancement from pre-phosphorylation was reported for siRNA injected into Drosophila embryos (Boutla, 2001). It is concluded that lysates from Drosophila ovaries faithfully recapitulate RNAi directed by siRNA duplexes (Tomari, 2004).

Data suggest that both armi and aub are required genetically for RISC assembly, but they provide no insight into the molecular basis of their RISC assembly defect(s). At what step(s) in RISC assembly are armi and aub blocked? In order to answer this question, protein-siRNA intermediates in the RISC assembly pathway were identified. The 2'-O-methyl oligonucleotide/native gel method detects only complexes competent to bind target RNA (mature RISC). Therefore, a native gel assay designed to detect intermediates in the assembly of RISC was used. The siRNA was radiolabeled, allowing detection of complexes containing either single-stranded or double-stranded siRNA and functionally asymmetric siRNAs (Schwarz, 2003) were used to distinguish between complexes containing single- and double-stranded siRNA (Tomari, 2004).

RISC contains only a single siRNA strand. Functionally asymmetric siRNAs load only one of the two strands of an siRNA duplex into RISC and degrade the other strand (Schwarz, 2003); the relative stability of the 5' ends of the two strands determines which is loaded into RISC. siRNA 1 loads its antisense strand into RISC, whereas siRNA 2 loads the sense strand (Schwarz, 2003). The two siRNA duplexes are identical, except that siRNA 2 contains a C-to-U substitution at position 1, which inverts the asymmetry (Schwarz, 2003). For both siRNAs, the antisense strand was 3' 32P-radiolabeled and will always be present in complexes that contain double-stranded siRNA. However, RISC will contain the 32P-radiolabeled antisense strand only for siRNA 1. siRNA 2 will also make RISC, but it will contain the nonradioactive sense strand (Tomari, 2004).

When either siRNA 1 or siRNA 2 was used to assemble RISC in embryo lysate, two complexes (B and A) were detected in the native gel assay; a third complex was detected only with siRNA 1. This third complex therefore contains single-stranded siRNA and corresponds to RISC. Complexes B and A are good candidates for RISC assembly intermediates. Formation of all three complexes was dramatically reduced when the antisense siRNA strand contained a 5' methoxy group, a modification that blocks RNAi (Nykanen, 2001). When the antisense strand of the siRNA contained a single 5'-deoxy nucleotide, making it a poor substrate for phosphorylation in the lysate (Nykanen, 2001), assembly of all three complexes was reduced. Phosphorylating the 5' deoxy-substituted siRNA before the reaction restored complex assembly. Formation of complex A and of RISC required ATP. In contrast, complex B assembled efficiently in the absence of ATP, but only if the siRNA was phosphorylated prior to the reaction (Tomari, 2004).

Complexes B, A, and RISC also formed in ovary lysates. As for embryo lysate, complexes B and A contained double-stranded siRNA, whereas RISC contained single-stranded. No complexes formed in ovary lysate when siRNA 5' phosphorylation was blocked and complex assembly was reduced when siRNA phosphorylation was slow (Tomari, 2004).

To determine the relationship of complexes B, A, and RISC, the kinetics of complex formation were monitored and the data was analyzed by kinetic modeling. Of all possible models relating free siRNA, B, A, and RISC, only the simple linear pathway siRNA --> B --> A --> RISC fit well to the model. The modeled rate constants for the pathway are consistent with the observation that formation of complex B is ATP independent, but RISC is ATP dependent (Tomari, 2004).

A 'chase' experiment was also performed to confirm the prediction that complex B is a precursor to RISC (via A). Complex B was assembled by incubating 32P-radiolabeled siRNA in embryo lysate for 5 min, then a 20-fold excess of unlabeled siRNA was added to prevent further incorporation of 32P-siRNA into complex. Then the incubation was continued and the formation of complexes was monitored. Complex B disappeared with time; A increased with time then peaked at ~60 min, and RISC accumulated throughout the experiment. The amount of radiolabeled free siRNA was essentially unchanged throughout the experiment, demonstrating that the unlabeled siRNA effectively blocked incorporation of 32P-free siRNA into complex. Thus, B was chased into RISC, likely via A. Together, this kinetic modeling and chase experiment provides support for a RISC assembly pathway in which the siRNA passes through two successive, double-stranded siRNA-containing complexes, B and A, in order to be transformed into the single-stranded siRNA-containing RISC (Tomari, 2004).

Liu and colleagues have proposed that a heterodimeric complex, comprising Dicer-2 (Dcr-2) and the dsRNA binding protein R2D2, loads siRNA into RISC (Liu, 2003). Complex A contains the Dcr-2/R2D2 heterodimer. R2D2 and Dcr-2 are readily crosslinked to 32P-radiolabeled siRNA with UV light (Liu, 2003). An siRNA was synthesized containing a single photocrosslinkable nucleoside base (5-iodouracil) at position 20. The 32P-5-iodouracil siRNA was incubated with embryo lysate to assemble complexes, then irradiated with 302 nm light, which initiates protein-RNA crosslinking only at the 5-iodo-substituted nucleoside. Proteins covalently linked to the 32P-radiolabeled siRNA were resolved by SDS-PAGE. Two proteins -- ~200 kDa and ~40 kDa -- efficiently crosslinked to the siRNA. Both crosslinked proteins were coimmunoprecipitated with either α-Dcr-2 or α-R2D2 serum, but not with normal rabbit serum. Neither crosslink was observed in ovary lysates prepared from r2d2 homozygous mutant females, a result expected because Dcr-2 is unstable in the absence of R2D2 (Liu, 2003; Tomari, 2004).

The crosslinking was repeated, and the reaction analyzed by native gel electrophoresis to resolve complexes B, A, and RISC. Each complex was eluted from the gel and analyzed by SDS-PAGE. The R2D2 and Dcr-2 crosslinks were present in complexes A and RISC, but not B. In a parallel experiment, complexes B, A, and RISC were isolated (without crosslinking) and analyzed by Western blotting with either α-Dcr-2 or α-R2D2 antibodies. Again, complexes A and RISC, but not B, contained both Dcr-2 and R2D2. Finally, complex assembly was tested in ovary lysates prepared from r2d2 homozygous mutant females. Only complex B formed in these lysates. It is concluded that complex A contains the previously identified Dcr-2/R2D2 heterodimer (Liu, 2003), and that both Dcr-2 and R2D2 remain associated with at least a subpopulation of RISC, consistent with earlier reports that Dcr-2 in flies and both DCR-1 and the nematode homolog of R2D2, RDE-4, coimmunoprecipitate with Argonaute proteins (Tomari, 2004).

A distinct small RNA pathway silences selfish genetic elements in the germline requires armitage

In the Drosophila germline, repeat-associated small interfering RNAs (rasiRNAs) ensure genomic stability by silencing endogenous selfish genetic elements such as retrotransposons and repetitive sequences. Whereas small interfering RNAs (siRNAs) derive from both the sense and antisense strands of their double-stranded RNA precursors, rasiRNAs arise mainly from the antisense strand. rasiRNA production appears not to require Dicer-1, which makes microRNAs (miRNAs), or Dicer-2, which makes siRNAs, and rasiRNAs lack the 2',3' hydroxy termini characteristic of animal siRNA and miRNA. Unlike siRNAs and miRNAs, rasiRNAs function through the Piwi, rather than the Ago, Argonaute protein subfamily. These data suggest that rasiRNAs protect the fly germline through a silencing mechanism distinct from both the miRNA and RNA interference pathways (Vagin, 2006).

In plants and animals, RNA silencing pathways defend against viruses, regulate endogenous gene expression, and protect the genome against selfish genetic elements such as retrotransposons and repetitive sequences. Common to all RNA silencing pathways are RNAs 19 to 30 nucleotides (nt) long that specify the target RNAs to be repressed. In RNA interference (RNAi), siRNAs are produced from long exogenous double-stranded RNA (dsRNA). In contrast, ~22-nt miRNAs are endonucleolytically processed from endogenous RNA polymerase II transcripts. Dicer ribonuclease III (RNase III) enzymes produce both siRNAs and miRNAs. In flies, Dicer-2 (Dcr-2) generates siRNAs, whereas the Dicer-1 (Dcr-1)–Loquacious (Loqs) complex produces miRNAs. After their production, small silencing RNAs bind Argonaute proteins to form the functional RNA silencing effector complexes believed to mediate all RNA silencing processes (Vagin, 2006 and references therein).

In Drosophila, processive dicing of long dsRNA and the accumulation of sense and antisense siRNAs without reference to the orientation of the target mRNA are hallmarks of RNAi in vitro. Total small RNA was prepared from the heads of adult males expressing a dsRNA hairpin that silences the white gene via the RNAi pathway. white silencing requires Dcr-2, R2D2, and Ago2. siRNAs were detected with a microarray containing TM (melting temperature)–normalized probes, 22 nt long, for all sense and antisense siRNAs that theoretically can be produced by dicing the white exon 3 hairpin. Both sense and antisense white siRNAs were detected in wild-type flies but not in dcr-2L811fsX homozygous mutant flies. The Dcr-2–dependent siRNAs were produced with a periodicity of ~22 nt, consistent with the phased processing of the dsRNA hairpin from the end formed by the 6-nt loop predicted to remain after splicing of its intron-containing primary transcript (Vagin, 2006).

Drosophila repeat-associated small interfering RNAs (rasiRNAs) can be distinguished from siRNAs by their longer length, 24 to 29 nt. rasiRNAs have been proposed to be diced from long dsRNA triggers, such as the ~50 copies of the bidirectionally transcribed Suppressor of Stellate [Su(Ste)] locus on the Y chromosome that in testes silence the ~200 copies of the protein-coding gene Stellate (Ste) found on the X chromosome (Vagin, 2006).

Microarray analysis of total small RNA isolated from fly testes revealed that Su(Ste) rasiRNAs detectably accumulate only from the antisense strand, with little or no phasing. As expected, Su(Ste) rasiRNAs were not detected in testes from males lacking the Su(Ste) loci (cry1Y). Su(Ste) rasiRNAs were also absent from armitage (armi) mutant testes, which fail to silence Ste and do not support RNAi in vitro. armi encodes a non–DEAD-box helicase homologous to the Arabidopsis thaliana protein SDE3, which is required for RNA silencing triggered by transgenes and some viruses, and depletion by RNAi of the mammalian Armi homolog Mov10 blocks siRNA-directed RNAi in cultured human cells. Normal accumulation of Su(Ste) rasiRNA and robust Ste silencing also require the putative helicase Spindle-E (Spn-E), a member of the DExH family of adenosine triphosphatases (Vagin, 2006).

The accumulation in vivo of only antisense rasiRNAs from Su(Ste) implies that sense Su(Ste) rasiRNAs either are not produced or are selectively destroyed. Either process would make Ste silencing mechanistically different from RNAi. In support of this view, mutations in the central components of the Drosophila RNAi pathway—dcr-2, r2d2, and ago2—did not diminish Su(Ste) rasiRNA accumulation. Deletion of the Su(Ste) silencing trigger (cry1Y) caused a factor of ~65 increase in Ste mRNA, but null or strong hypomorphic mutations in the three key RNAi proteins did not (Vagin, 2006).

Fly Argonaute proteins can be subdivided into the Ago (Ago1 and Ago2) and Piwi [Aubergine (Aub), Piwi, and Ago3] subfamilies. Unlike ago1 and ago2, the aub, piwi, and ago3 mRNAs are enriched in the germline. Aub is required for Ste silencing and Su(Ste) rasiRNA accumulation. In aubHN2/aubQC42 trans-heterozygous mutants, Su(Ste) rasiRNAs were not detected by microarray or Northern analysis, and Su(Ste)-triggered silencing of Ste mRNA was lost completely. Even aubHN2/+ heterozygotes accumulated less of the most abundant Su(Ste) rasiRNA than did the wild type. That the Ago subfamily protein Ago2 is not required for Ste silencing, whereas the Piwi subfamily protein Aub is essential for it, supports the view that Ste is silenced by a pathway distinct from RNAi. Intriguingly, Su(Ste) rasiRNAs hyperaccumulated in piwi mutant testes, where Ste is silenced normally (Vagin, 2006).

Mutations in aub also cause an increase in sense, but not antisense, Su(Ste) RNA; these results suggest that antisense Su(Ste) rasiRNAs can silence both Ste mRNA and sense Su(Ste) RNA, but that no Su(Ste) rasiRNAs exist that can target the antisense Su(Ste) transcript. The finding that Su(Ste) rasiRNAs are predominantly or exclusively antisense is essentially in agreement with the results of small RNA cloning experiments, in which four of five Su(Ste) rasiRNAs sequenced were in the antisense orientation, but is at odds with earlier reports detecting both sense and antisense Su(Ste) rasiRNAs by non-quantitative Northern hybridization (Vagin, 2006).

Is germline RNA silencing of selfish genetic elements generally distinct from the RNAi and miRNA pathways? The expression of a panel of germline-expressed selfish genetic elementswas examined in mutants defective for eight RNA silencing proteins: three long terminal repeat (LTR)-containing retrotransposons (roo, mdg1, and gypsy); two non-LTR retrotransposons (I-element and HeT-A, a component of the Drosophila telomere), and a repetitive locus (mst40). All selfish genetic elements tested behaved like Ste: Loss of the RNAi proteins Dcr-2, R2D2, or Ago2 had little or no effect on retrotransposon or repetitive element silencing. Instead, silencing required the putative helicases Spn-E and Armi plus one or both of the Piwi subfamily Argonaute proteins, Aub and Piwi. Silencing did not require Loqs, the dsRNA-binding protein required to produce miRNAs (Vagin, 2006).

The null allele dcr-1Q1147X is homozygous lethal, making it impossible to procure dcr-1 mutant ovaries from dcr-1Q1147X/dcr-1Q1147X adult females. Therefore, clones of dcr-1Q1147X/dcr-1Q1147X cells were generated in the ovary by mitotic recombination in flies heterozygous for the dominant female-sterile mutation ovoD1. RNA levels, relative to rp49 mRNA, were measured for three retrotransposons (roo, HeT-A, and mdg1) and one repetitive sequence (mst40) in dcr-1/dcr-1 recombinant ovary clones and in ovoD1/TM3 and dcr-1/ovoD1 nonrecombinant ovaries. The ovoD1 mutation blocks oogenesis at stage 4, after the onset of HeT-A and roo rasiRNA production. Retrotransposon or repetitive sequence transcript abundance was unaltered or decreased in dcr-1/dcr-1 relative to ovoD1/TM3 and dcr-1/ovoD1 controls. It is concluded that Dcr-1 is dispensable for silencing these selfish genetic elements in the Drosophila female germline (Vagin, 2006).

roo is the most abundant LTR retrotransposon in flies. roo silencing was analyzed in the female germline with the use of microarrays containing 30-nt probes, tiled at 5-nt resolution, for all ~18,000 possible roo rasiRNAs; the data was corroborated at 1-nt resolution for those rasiRNAs derived from LTR sequences. As observed for Su(Ste) but not for white RNAi, roo rasiRNAs were nonhomogeneously distributed along the roo sequence and accumulated primarily from the antisense strand. In fact, the most abundant sense rasiRNA peak corresponded to a set of probes containing 16 contiguous uracil residues, which suggests that these probes nonspecifically detected fragments of the mRNA polyadenylate [poly(A)] tail. Most of the remaining sense peaks were unaltered in armi mutant ovaries, in which roo expression is increased; this result implies that they do not contribute to roo silencing. No phasing was detected in the distribution of roo rasiRNAs (Vagin, 2006).

As for Su(Ste), wild-type accumulation of antisense roo rasiRNA required the putative helicases Armi and Spn-E and the Piwi subfamily Argonaute proteins Piwi and Aub, but not the RNAi proteins Dcr-2, R2D2, and Ago2. Moreover, accumulation of roo rasiRNA was not measurably altered in loqs f00791, an allele that strongly disrupts miRNA production in the female germline (Vagin, 2006).

Loss of Dcr-2 or Dcr-1 did not increase retrotransposon or repetitive element expression, which suggests that neither enzyme acts in rasiRNA-directed silencing. Moreover, loss of Dcr-2 had no detectable effect on Su(Ste) rasiRNA in testes or roo rasiRNA in ovaries. The amount of roo rasiRNA and miR-311 was measured in dcr-1/dcr-1 ovary clones generated by mitotic recombination. Comparison of recombinant (dcr-1/dcr-1) and nonrecombinant (ovoD1/TM3 and dcr-1/ovoD1) ovaries by Northern analysis revealed that roo rasiRNA accumulation was unperturbed by the null dcr-1Q1147X mutation. Pre–miR-311 increased and miR-311 declined by a factor of ~3 in the dcr-1/dcr-1 clones, consistent with about two-thirds of the tissue corresponding to mitotic dcr-1/dcr-1 recombinant cells. Yet, although most of the tissue lacked dcr-1 function, improved, rather than diminished, silencing was observed for the four selfish genetic elements examined. Moreover, the dsRNA-binding protein Loqs, which acts with Dcr-1 to produce miRNAs, was also dispensable for roo rasiRNA production and selfish genetic element silencing. Although the possibility that dcr-1 and dcr-2 can fully substitute for each other in the production of rasiRNA in the ovary cannot be excluded, biochemical evidence suggests that none of the three RNase III enzymes in flies—Dcr-1, Dcr-2, and Drosha—can cleave long dsRNA into small RNAs 24 to 30 nt long (Vagin, 2006).

Animal siRNA and miRNA contain 5' phosphate and 2',3' hydroxy termini. Enzymatic and chemical probing was used to infer the terminal structure of roo and Su(Ste) rasiRNAs. RNA from ovaries or testes was treated with calf intestinal phosphatase (CIP) or CIP followed by polynucleotide kinase plus ATP. CIP treatment caused roo and Su(Ste) rasiRNA to migrate more slowly in polyacrylamide gel electrophoresis, consistent with the loss of one or more terminal phosphate groups. Subsequent incubation with polynucleotide kinase and ATP restored the original gel mobility of the rasiRNAs, indicating that they contained a single 5' or 3' phosphate before CIP treatment. The roo rasiRNA served as a substrate for ligation of a 23-nt 5' RNA adapter by T4 RNA ligase, a process that requires a 5' phosphate; pretreatment with CIP blocked ligation, thus establishing that the monophosphate lies at the 5' end. The rasiRNA must also contain at least one terminal hydroxyl group, because it could be joined by T4 RNA ligase to a preadenylated 17-nt 3' RNA adapter. Notably, the 3' ligation reaction was less efficient for the roo rasiRNA than for a miRNA in the same reaction (Vagin, 2006).

RNA from ovaries or testes was reacted with NaIO4, then subjected to ß-elimination, to determine whether the rasiRNA had either a single 2' or 3' terminal hydroxy group or had terminal hydroxy groups at both the 2' and 3' positions, as do animal siRNA and miRNA. Only RNAs containing both 2' and 3' hydroxy groups react with NaIO4; ß-elimination shortens NaIO4-reacted RNA by one nucleotide, leaving a 3' monophosphate terminus, which adds one negative charge. Consequently, NaIO4-reacted, ß-eliminated RNAs migrate faster in polyacrylamide gel electrophoresis than does the original unreacted RNA. Both roo and Su(Ste) rasiRNA lack either a 2' or a 3' hydroxyl group, because they failed to react with NaIO4; miRNAs in the same samples reacted with NaIO4. Together, these results show that rasiRNAs contain one modified and one unmodified hydroxyl. Because T4 RNA ligase can make both 3'-5' and 2'-5' bonds, the blocked position cannot currently be determined. Some plant small silencing RNAs contain a 2'-O-methyl modification at their 3' terminus (Vagin, 2006).

Drosophila and mammalian siRNA and miRNA function through members of the Ago subfamily of Argonaute proteins, but Su(Ste) and roo rasiRNAs require at least one member of the Piwi subfamily for their function and accumulation. To determine whether roo rasiRNAs physically associate with Piwi and Aub, ovary lysate were prepared from wildtype flies or transgenic flies expressing either myc-tagged Piwi or green fluorescent protein (GFP)–tagged Aub protein; they were immunoprecipitated with monoclonal antibodies (mAbs) to myc, GFP, or Ago1; and then the supernatant and antibody-bound small RNAs were analyzed by Northern blotting. Six different roo rasiRNAs were analyzed. All were associated with Piwi but not with Ago1, the Drosophila Argonaute protein typically associated with miRNAs; miR-8, miR-311, and bantam immunoprecipitated with Ago1 mAb. No rasiRNAs immunoprecipitated with the myc mAb when lysate was used from flies lacking the myc-Piwi transgene (Vagin, 2006).

Although aub mutant ovaries silenced roo mRNA normally, they showed reduced accumulation of roo rasiRNA relative to aub/+ heterozygotes, which suggests that roo rasiRNAs associate with both Piwi and Aub. The supernatant and antibody-bound small RNAs were analyzed after GFP mAb immunoprecipitation of ovary lysate from GFP-Aub transgenic flies and flies lacking the transgene. roo rasiRNA was recovered only when the immunoprecipitation was performed with the GFP mAb in ovary lysate from GFP-Aub transgenic flies. The simplest interpretation of these data is that roo rasiRNAs physically associate with both Piwi and Aub, although it remains possible that the roo rasiRNAs are loaded only into Piwi and that Aub associates with Piwi in a stable complex. The association of roo rasiRNA with both Piwi and Aub suggests that piwi and aub are partially redundant, as does the modest reduction in roo silencing in piwi but not in aub mutants. Alternatively, roo silencing might proceed through Piwi alone, but the two proteins could function in the same pathway to silence selfish genetic elements (Vagin, 2006).

These data suggest that in flies, rasiRNAs are produced by a mechanism that requires neither Dcr-1 nor Dcr-2, yet the patterns of rasiRNAs that direct roo and Ste silencing are as stereotyped as the distinctive siRNA population generated from the white hairpin by Dcr-2 or the unique miRNA species made from each pre-miRNA by Dcr-1. A key challenge for the future will be to determine what enzyme makes rasiRNAs and what sequence or structural features of the unknown rasiRNA precursor lead to the accumulation of a stereotyped pattern of predominantly antisense rasiRNAs (Vagin, 2006).

Synaptic protein synthesis associated with memory is regulated by the RISC pathway in Drosophila

Long-lasting forms of memory require protein synthesis, but how the pattern of synthesis is related to the storage of a memory has not been determined. This study shows that neural activity directs the mRNA of the Drosophila Ca2+, Calcium/Calmodulin-dependent Kinase II (CaMKII), to postsynaptic sites, where it is rapidly translated. These features of CaMKII synthesis are recapitulated during the induction of a long-term memory and produce patterns of local protein synthesis specific to the memory. mRNA transport and synaptic protein synthesis are regulated by components of the RISC pathway, including the SDE3 helicase Armitage, which is specifically required for long-lasting memory. Armitage is localized to synapses and lost in a memory-specific pattern that is inversely related to the pattern of synaptic protein synthesis. Therefore, it is proposed that degradative control of the RISC pathway underlies the pattern of synaptic protein synthesis associated with a stable memory (Ashraf, 2006).

The CaMKII 3'UTR is necessary and sufficient for the robust localization of CaMKII to dendritic arbors: Since the mammalian aCaMKII is found at synapses, where its synthesis is regulated by neural activity, the attention of this study turned to the CaMKII gene of Drosophila. Drosophila CaMKII has a role in neuromuscular synaptic plasticity and memory in the courtship-conditioning paradigm. CaMKII is localized to both pre- and postsynaptic sites in the adult brain. Focus was placed on the olfactory system because of its well-described neural components, circuitry, and paradigms for the establishment of memory. This system consists of sensory neurons and interneurons that form an early receptive and processing circuit with synapses organized in bilaterally symmetric centers known as the antennal lobes. The first-order interneurons (Projection Neurons; PNs) collect sensory input in a stereotyped array of multisynaptic structures known as glomeruli, where the PN dendritic synapses collect cholinergic input via nicotinic acetylcholine receptors. The PNs direct-output to two brain centers via branching axons that project to the 'calyx' of the mushroom body and to the lateral horn. These terminals release acetylcholine from choline acyltransferase (ChAT)-positive boutons. The PN dendrites also form reciprocal synapses with local interneurons. On the PN dendrites, CaMKII was localized in postsynaptic puncta with the markers Discs Large (DLG) and ARD (a nAChR β-subunit. CaMKII was also concentrated at the PN presynaptic boutons in the calyx and lateral horn. Thus, within the same neuron, CaMKII is concentrated at both pre- and postsynaptic sites (Ashraf, 2006).

The mouse aCaMKII mRNA displays dendritic localization and activity-dependent synaptic translation, features conferred by sequences in its 3'UTR. To determine whether this is the case for Drosophila CaMKII, the 3'UTR was inserted downstream of the EYFP coding sequence in the reporter, UAS-EYFP3'UTR. An additional pair of constructs was made bearing a translational fusion of EYFP to CaMKII, with the 3'UTRCaMKII present (UAS-CaMKII::EYFP3'UTR) or absent (UAS-CaMKII::EYFPNUT) (Ashraf, 2006).

When expressed specifically in PNs using the GAL4, UAS binary system, EYFP3'UTR fluorescence was strikingly localized to synapses on the PN dendritic and axonal termini and colocalized with ChAT at the presynaptic boutons and with the nAChR subunit ARD on dendritic branches in glomeruli. This distribution roughly matched that of CaMKII protein, though EYFP was somewhat more diffuse along the dendritic branches; presumably EYFP does not bind to the postsynaptic apparatus as CaMKII does. In contrast, a cytoplasmic EYFP reporter lacking CaMKII sequences was distributed poorly to the axons and dendrites (Ashraf, 2006).

The CaMKII::EYFP fusion protein synthesized from mRNA harboring the 3'UTR (CaMKII::EYFP3'UTR) displayed synaptic localization in axons and dendrites like that of EYFP3'UTR but was notably more concentrated in synaptic puncta. The same fusion protein made from mRNA lacking the 3'UTR (CaMKII::EYFPNUT) was strongly localized to axonal presynaptic sites but found at a very low level on the antennal lobe dendrites, where it was localized to synaptic puncta. Thus the CaMKII 3'UTR is necessary and sufficient for the robust localization of CaMKII to dendritic arbors but not required for axonal localization (Ashraf, 2006).

Localization and rapid induction of CaMKII in dendrites is due to 3'UTR-dependent synaptic protein synthesis; To determine how neural activity might affect CaMKII expression, brains were explanted into bath culture with acetylcholine (ACh) or nicotine (an agonist of nAChRs). After 20 min, the tissue was examined by anti-CaMKII immunohistochemistry and quantitative confocal microscopy. On average, when cholinergic synapses were activated, CaMKII immunofluorescence in the antennal lobe increased by ~3- to 4-fold. In a time-course experiment, an increase in CaMKII level was detected within 5-10 min of nicotine exposure. The increase was widespread in the brain and reflected in a ~4-fold increase of CaMKII protein on Western blot analysis. The CaMKII increase was also specific, as the levels of synaptic proteins DLG and ARD were unchanged. Consistent with the notion that this regulation occurs via translational control, the effect of cholinergic stimulation was blocked by the ribosomal inhibitor Anisomycin but not by Actinomycin D, an inhibitor of transcription (Ashraf, 2006).

These results and the requirement of the CaMKII 3'UTR for dendritic localization suggest that cholinergic activity may induce the translation of CaMKII mRNA at postsynaptic sites. This was examined by monitoring EYFP3'UTR reporter expression in explant culture. A 5 min nicotine incubation increased EYFP3'UTR expression by 30% and, after 20 min, by 250%. The induced EYFP protein was found in large punctae. Cholinergic stimulation did not increase EYFP3'UTR expression at the presynaptic terminals in the calyx. In contrast, CaMKII::EYFPNUT expression was only slightly increased by nicotine or ACh exposure. Nicotine incubation did not alter the expression of cytoplasmic EYFP, CD8::GFP, or an EGFP construct harboring an a1-tubulin 3'UTR (Hh::EGFP-3'UTRtub). These observations indicate that the localization and rapid induction of CaMKII in dendrites is due to 3'UTR-dependent synaptic protein synthesis (Ashraf, 2006).

Odor-specific induction of synaptic protein synthesis occurs when conditioned and unconditioned stimuli are presented coincidentally and with temporal spacing, the experience that establishes an LTM: In Drosophila, an olfactory LTM is induced by 'spaced training,' a protocol where an odor (CS+) and electric shock (US) are presented coincidentally at temporally spaced intervals. A second odor (CS-) follows the CS+ odor in each interval without coincident shock. An LTM appears after several hours and lasts beyond 24 hr, as assayed by tactic behavior in a T-maze. This protocol was followed and EYFP3'UTR was used to report synaptic protein synthesis in animals that developed an olfactory LTM. The analysis focused on the antennal lobe glomeruli because these structures can be reproducibly identified and display clustered synaptic activity. Furthermore, the first-order antennal lobe synapses might participate in an early stage of memory storage, including the storage of LTM. This analysis revealed an odorant-specific pattern of synaptic protein synthesis associated with the induction of a long-term memory (Ashraf, 2006).

Animals harboring the UAS-EYFP3'UTR reporter driven by the PN-specific GH146-GAL4 were trained and analyzed at times from 4 to 24 hr posttraining. The brains of trained and untrained animals were processed for microscopy in parallel. For each glomerulus, a Z stack of 6-8 confocal microscopic images was recorded and analyzed via a thresholding protocol that isolated pixel groups corresponding to synaptic puncta. An average glomerulus intensity change (ΔF/F) was calculated for 5-8 brains in each experiment. Each experiment was repeated five times. LTM was in all cases verified by T-maze performance (Ashraf, 2006).

The analysis was restricted to a set of glomeruli that included those with a primary response to the odorants octanol (OCT) and methylcyclohexanol (MCH). Only particular glomeruli displayed a training-dependent increase in EYFP3'UTR fluorescence, while others did not; their identities depended on the odorant (CS+) paired with shock. When OCT was the CS+, only glomeruli D and DL3 displayed increased fluorescence, by 115% and 108%, respectively. When MCH was the CS+, fluorescence increased significantly in glomeruli DA1 and VA1 by 95% and 70%, respectively. There were modest but possibly insignificant increases in glomeruli DM6 and VC2. The glomerulus-specific increases were noted as early as 4 hr posttraining and were not observed when odorant and/or electric shock was unpaired or left out or when temporal spacing was not employed ('massed training'). In animals that expressed a cytoplasmic EYFP reporter or CaMKII::EYFPNUT, which lack the CaMKII 3'UTR, there were no significant fluorescence changes. This analysis revealed that an odor-specific induction of synaptic protein synthesis occurred when conditioned and unconditioned stimuli were presented coincidentally and with temporal spacing, the experience that establishes an LTM. This plasticity was evidently maintained for at least 24 hr (Ashraf, 2006).

A coordinated program for synaptic gene expression occurs during the storage of a memory: If CaMKII was synthesized at the synapse, its mRNA would be localized there. To address this question, an mRNA tracking system based on the bacteriophage coat protein MS2 and its RNA binding site were utilized. The fusion protein MS2::GFP::nls is concentrated in the nucleus by nuclear localization signals (nls) but can be diverted elsewhere by binding to an MS2 binding site (MS2-bs) tagged mRNA. Three mRNAs were tagged: the Drosophila CaMKII cDNA, its 3'UTR alone, and the mouse aCaMKII 3'UTR, which mediates dendritic localization and synaptic translation in the mouse (Ashraf, 2006).

GFP fluorescence was examined when MS2::GFP::nls was expressed in projection neurons (PNs) with or without an MS2-bs tagged mRNA. Punctate fluorescence was observed in the dendrites when an MS2-bs-tagged mRNA was coexpressed with MS2::GFP::nls but not with MS2::GFP::nls alone. For example, the tagged Drosophila CaMKII mRNA increased the intensity of glomerular fluorescence by 200%. In dendrites, particular mRNAs, including the mouse aCaMKII, are localized to particles containing the motor protein Kinesin. Consistent with this observation, the GFP-positive dendritic puncta were labeled with an antibody against the major kinesin heavy chain, KHC (Ashraf, 2006).

It was asked whether the synaptic CaMKII expression induced by cholinergic activity might be associated with enhanced dendritic localization of CaMKII mRNA, as has been found for the mouse aCaMKII and Arc mRNAs. When explanted into media with nicotine or ACh, brains harboring MS2::GFP::nls and the tagged Drosophila CaMKII mRNA displayed a striking increase in dendritic GFP fluorescence. The effect with cholinergic stimulation was similar with the tagged mouse aCaMKII 3'UTR: a 70%-73% increase relative to culture without nicotine. The activity-enhanced dendritic mRNA transport was blocked by Anisomycin but not by Actinomycin D. It is supposed that mainly existing mRNA can be translocated during the short period of culture. Thus, in Drosophila, like mammals, neural activity increases the rate of mRNA movement to the synapse by a protein synthesis-dependent mechanism (Ashraf, 2006).

It was then asked whether the induction of an LTM might affect mRNA transport to the synapse. When animals expressing MS2::GFP::nls and Drosophila ms2bs-CaMKII were subjected to the spaced training protocol, the number of GFP-labeled puncta in dendrites was substantially increased. The pattern of dendritic punctae did not display evident glomerular specificity, as observed for synaptic protein synthesis. However, the induced punctae were distributed along the dendritic branches, making a determination of glomerular specificity uncertain. These observations reveal that a coordinated program for synaptic gene expression occurs during the storage of a memory (Ashraf, 2006).

Regulation of mRNA transport and synaptic protein synthesis by the RISC pathway: The RNA interference (RISC) pathway silences gene expression by the targeted degradation of mRNAs or their nondestructive silencing. In Drosophila, RISC-mediated translational silencing controls oskar expression in the developing oocyte. An SDE3-class RNA helicase, Armitage (Armi) acts as part of RISC to control oskar translation and regulate cytoskeletal organization, possibly via control of Kinesin heavy chain (Khc) translation. Both the oskar and Khc 3'UTRs have putative binding sites for the microRNA (miRNA) miR-280. The CaMKII 3'UTR has a remarkably similar miR-280 binding site. This site and a nearby site for miR-289 satisfy the predictive rule that 7 of 8 nucleotides at the 5' end of an miRNA are cognate to a target mRNA. Kinesin is also a component of the RNA-containing dendritic particles that bring mRNA to the synapse. Staufen, likewise a mediator of RNA transport, has binding sites for miR-280 and miR-305 in its mRNA 3'UTR. Thus this study explored the role of RISC in CaMKII, KHC, and Staufen expression (Ashraf, 2006).

Dicer-2 is one of two Drosophila ribonucleases that produce short RNA components of RISC. CaMKII synaptic expression was dramatically increased in a dicer-2 mutant, particularly in the antennal lobe and mushroom body. In contrast, there was no difference in the expression of the cell adhesion protein Fasciclin II in the same animals. In Western analysis, there was a striking ~25-fold increase in CaMKII protein in dicer-2 mutant brains. Synaptic CaMKII expression was also elevated in aubergine and armitage mutant brains. The aubergine locus encodes an Argonaute protein involved in RISC assembly and function. The level of Staufen protein was also increased in the armitage mutant brain (Ashraf, 2006).

Whether the miRNA binding sites in the CaMKII 3'UTR might be involved in RISC-mediated regulation was examined with the EYFP3'UTR transgene. When expressed in the PNs, EYFP fluorescence in glomeruli was 80% greater in armi than in the wild-type. The EYFP3'UTR fluorescence was localized to large dendritic puncta like those found in brains explanted into nicotine-containing media. Indeed, EYFP3'UTR expression in armi brains did not increase further upon explant with nicotine, consistent with the notion that cholinergic activation might act via antagonism of RISC. The expression of CaMKII::EYFPNUT, which lacks the 3'UTR, increased slightly in the armi mutant background, while other control constructs, such as CD8::GFP, were essentially unchanged. In addition, RT-PCR analysis of wild-type and armi mutant brains did not reveal a difference in the levels of transgenic mRNAs. There was also a substantial increase in EYFP3'UTR and CaMKII::EYFP3'UTR synaptic fluorescence in dicer-2 and aubergine mutants. Therefore it is concluded that RISC regulates CaMKII expression by a posttranscriptional mechanism, utilizing sites in the CaMKII 3'UTR (Ashraf, 2006).

Armitage expression was found in multiple neuronal populations in the brain, including the PNs and mushroom-body Kenyon cells. It is distributed in puncta in cell bodies and dendrites and to axon termini. A GFP::Armi fusion protein, when expressed in the PNs, displayed a similar punctate distribution that overlaps synaptic puncta containing CaMKII. The GFP::Armi fusion protein retains armi+ activity such that neurons with high levels of GFP::Armi expression would have increased armi+ activity. Several observations indicate that a posttranscriptional autoregulatory circuit modulates Armi expression. Nonetheless, strong transgenic expression of GFP::Armi reduced the level of CaMKII expression, as revealed by Western blot analysis and immunohistochemistry. Neurons that expressed a high level of GFP::Armi displayed reduced expression of both CaMKII and KHC. A control UAS-CD8-GFP transgene was unaffected by GFP::Armi expression (Ashraf, 2006).

Since Armi regulates KHC and Staufen expression, the possibility was considered that it might also regulate the dendritic transport of CaMKII mRNA. When examined with the MS2::GFP system, armi72.1 homozygotes indeed displayed a 78% increase in fluorescence by dendritic GFP-positive puncta, compared to an armi+ control. Therefore, Armi regulation of synaptic protein synthesis reflects a coordinated program with multiple miRNA targets, affecting both mRNA transport and translation at the synapse, where Armi protein is found (Ashraf, 2006).

Neural activity induces rapid proteasome-mediated degradation of Armitage: If mRNA silencing by RISC plays a role in LTM, this pathway would be expected to be somehow regulated by neural function. Given the inverse relationship between CaMKII expression and armi+ activity, whether Armi might be a regulatory target wad considered. The level of GFP::Armi fluorescence rapidly decreased (by 3.5-fold) in brains explanted into nicotine-containing medium. There was a correlated increase in CaMKII expression (by ~4.5-fold) in the PN dendritic arbors of the antennal lobe. A short incubation with nicotine (5 min) resulted in the complete disappearance of Armi protein in Western analysis. The GFP::Armi protein was also eliminated upon explant with nicotine. In contrast, the CaMKII protein level increased and a1-tubulin was unchanged (Ashraf, 2006).

Two experiments were performed to determine whether the activity-induced elimination of Armi required the proteasome. First, GFP::Armi was expressed along with a transgenic dominant-negative mutant of the proteasome β subunit. When the DTS5 transgene was present, the level of GFP::Armi fluorescence was elevated by 3.2-fold. In contrast, the DTS5 transgene did not alter the level of CD8::GFP. Second, incubation with the proteasome inhibitor lactacystin blocked the nicotine-induced loss of GFP::Armi and degradation of endogenous Armi protein. Preincubation with lactacystin also blocked nicotine-induced synaptic CaMKII synthesis, as determined by both Western analysis and by immunohistochemistry. Thus, cholinergic activity evidently acts via the proteasome to induce the degradation of Armitage and synaptic synthesis of CaMKII (Ashraf, 2006).

A degradative pathway for LTM: A key question is whether this degradative pathway has a role in synaptic protein synthesis associated with LTM. Animals expressing the GFP::Armi protein in projection neurons were subjected to olfactory spaced training and analyzed by the same microscopic methods used to assess LTM-associated EYFP3'UTR expression. The GFP::Armi protein was found concentrated in synaptic puncta in the glomeruli. When examined at either 3 or 24 hr posttraining, GFP::Armi fluorescence was significantly reduced in many glomeruli and most strongly reduced in the glomeruli that had displayed the greatest increase in EYFP3'UTR expression. Fluorescence in glomeruli DA1 and VA1 decreased by ~3.1- and 3.8-fold, respectively, when the odorant MCH was paired with shock. When the odorant OCT was paired with shock, the D and DL3 glomeruli displayed the most significant decreases (~2-fold). More modest losses of GFP fluorescence were observed in other glomeruli. These observations reveal an inverse relationship between synaptic Armi protein and CaMKII synthesis during the establishment of an LTM. Since these changes were still present at 24 hr posttraining, the change was evidently maintained long-term, perhaps for the term of the memory (Ashraf, 2006).

Given the role of Armi in the synaptic synthesis of CaMKII, it was wondered whether either of these genes might be required to form an olfactory LTM. Several armi hypomorphic alleles display normal adult viability and behavior, including normal odor and shock sensitivity. Given their normal performance in these tests, armi animals were examined for STM and LTM. The animals (armi72.1/armi72.1 or armi72.1/Df(3L)E1) displayed normal memory in the short-term paradigm but were profoundly deficient in LTM. Expression of the GFP::Armi transgene rescued the armi72.1/armi72.1 LTM deficiency to a normal value. A nearly complete and tissue-specific loss of CaMKII was achieved by use of a construct that generates a CaMKII hairpin RNA. Animals expressing UAS-CaMKIIhpin in all CaMKII-positive neurons (with the CaMKII-GAL4 driver) retained normal short-term memory, but displayed a near-complete loss of LTM. Thus, both CaMKII and Armitage are required for LTM but not for STM (Ashraf, 2006).

It is concluded that memory-specific patterns of synaptic protein synthesis occur with the induction of a long-term memory in Drosophila. These patterns appear to be controlled by the proteasome-mediated degradation of a RISC pathway component, Armitage, to regulate the transport of mRNA to synapses and its translation once there (Ashraf, 2006).

To visualize synaptic protein synthesis, fluorescent reporters were used based on the Drosophila CaMKII gene, which has well-described roles in synaptic plasticity and memory. The 3'UTR of CaMKII shares regulatory motifs with the mammalian aCaMKII mRNA, which mediate dendritic mRNA localization and neural activity-dependent translation. The 3'UTR of Drosophila CaMKII was also necessary and sufficient for mRNA localization to dendrites and synaptic translation. This 3'UTR sufficed for the enhanced dendritic mRNA transport and translation induced by cholinergic stimulation. Hence a simple parallel was found between the synaptic regulation of CaMKII in Drosophila and mammals (Ashraf, 2006).

When these fluorescent reporters were utilized in vivo, the induction of synaptic protein synthesis was observed in several Drosophila brain centers following the spaced training paradigm of repetitive odor paired with electric shock that establishes a long-term memory. There were local patterns of memory specificity identifiable in glomeruli of the antennal lobe where synapses of similar function are clustered. When the odorant OCT was paired with electric shock, protein synthesis was induced selectively in the D and DL3 glomeruli. When the odorant MCH was paired with shock, the DA1 and VA1 glomeruli displayed the most robust enhancement of synaptic protein synthesis. Notably, the animals were exposed to both odorants during training; the pattern of synthesis depended on coincidence with shock. There was no significant induction of protein synthesis when exposure to odor and shock was nonoverlapping, with either stimulus presented alone, or in the absence of temporal spacing ('massed training'). Thus, an odor-specific pattern of synaptic protein synthesis was induced under conditions that produce an LTM (Ashraf, 2006).

Experiments in the honeybee suggest that the antennal lobe is a 'way station' for memory where stimuli are integrated to yield plasticity more labile than a short-term memory. A long-term memory can be formed in the honeybee antennal lobe in a spaced training paradigm. Experiments have revealed plasticity in the Drosophila antennal lobe, where particular glomeruli acquired enhanced synaptic activity after a single episode of paired odor and shock. Remarkably, the enhanced synaptic protein synthesis observed with spaced training occurred in essentially the same glomeruli that displayed enhanced synaptic activity in the STM protocol. These glomeruli are distinct from those that display the greatest odor or electric shock-evoked synaptic activity. Therefore, it is supposed that the mechanism that integrates a single paired odor and shock to produce new synaptic activity might also generate the trigger for synaptic protein synthesis when the paired stimuli are repeated with temporal spacing. It is believed this trigger includes the proteasome-mediated degradation of the RISC factor Armitage (Ashraf, 2006).

Though these 'memory traces' have been recorded in the antennal lobe, there is still no evidence for their role in memory. The mushroom body, in contrast, is required for LTM. The current methods cannot resolve patterns of synaptic protein synthesis in the mushroom body because it lacks the stereotyped synaptic architecture of the antennal lobe. When determined, a global brain map of synaptic protein synthesis will provide significant insights into the mechanisms of memory storage (Ashraf, 2006).

Synaptic protein synthesis and dendritic mRNA transport are well studied for the mammalian aCaMKII gene, which bears recognition motifs in its 3'UTR for CPEB and other proteins with transport and translation control functions. The presence of potential recognition motifs for the CPEB, Pumilio, and Staufen proteins in the Drosophila CaMKII 3'UTR suggests that these mechanisms are conserved in Drosophila. Indeed, Staufen, orb (a CPEB family member), and pumilio have been identified as LTM-deficient mutants. The roles of these genes remain to be fully explored (Ashraf, 2006).

Focus was placed instead on the RISC pathway because of apparent binding motifs for microRNAs miR-280 and miR-289 in the CaMKII 3'UTR. These sites are similar to those in the 3'UTRs of oskar and Kinesin heavy chain (Khc), which are targets for translational silencing by Armitage and other RISC components in the oocyte. Armitage was found in synaptic puncta on dendrites, colocalized with CaMKII. When the level of Armitage was decreased or increased by mutation or transgenic expression, CaMKII synaptic expression was modulated in a reciprocal and cell-autonomous fashion. This regulation could be recapitulated by an EYFP reporter bearing the CaMKII 3'UTR. Mutants for the RISC components Aubergine and Dicer-2 displayed similar phenotypes. It therefore seems likely that multiple tiers of control regulate CaMKII like oskar, where two systems (Bruno/Cup and RISC) act on distinct sites in its 3'UTR (Ashraf, 2006).

A second avenue for RISC control of CaMKII synthesis is via mRNA transport. By tagging CaMKII mRNA with a GFP reporter, dendritic punctae were observed whose frequency and intensity increased under the same conditions that induced synaptic protein synthesis: cholinergic activation and olfactory spaced training. The induction of mRNA transport required new protein synthesis but not transcription. Armitage was also found to regulate the frequency and intensity of the GFP-tagged dendritic puncta. Two proteins that play a role in mRNA transport, Kinesin heavy chain (KHC) and Staufen, recapitulate this pattern of regulation by cholinergic stimulation and Armitage. Both of their mRNAs bear targets for miRNA regulation in the 3'UTR. These studies leave open the possibility that the enhanced synaptic localization of CaMKII mRNA underlies the induction of its synaptic translation. However, the presence of miRNA binding sites in the CaMKII 3'UTR, the localization of Armitage with CaMKII in synaptic punctae, and the rapid induction of CaMKII synthesis by cholinergic activity all suggest that RISC acts at the synapse. Furthermore, local translational control may be required to impose the specificity that was not evident in the pattern of mRNA transport associated with the induction of an LTM (Ashraf, 2006).

A link between the induction of memory and synaptic protein synthesis is the proteasome-mediated degradation of Armitage. In explant culture, cholinergic induction of CaMKII synthesis was accompanied by the rapid degradation of Armi; both events were blocked by inhibition of the proteasome. The relationship between Armi degradation and CaMKII synaptic translation was recapitulated in the brain as animals formed and maintained an LTM. The same glomeruli that displayed the greatest increase in CaMKII synthesis displayed the largest decline in synaptic Armi. This reciprocal relationship between the Armi and CaMKII proteins was detected as early as 3 hr after training and maintained for at least 24 hr posttraining. The training-induced change of synaptic Armi was therefore 'locked in,' possibly for the term of the memory, consistent with a role in maintaining an alteration of synaptic function (Ashraf, 2006).

Therefore a new mechanism is proposed for stable memory in which an integrated sensory trigger induces the proteasome-mediated degradation of a RISC factor, releasing synaptic protein synthesis and mRNA transport from microRNA suppression. It is supposed that this mechanism is triggered with neuronal specificity in order to produce memory-specific patterns of protein synthesis. Whether this specificity is required for memory or extends to the level of a single synapse are questions that remain to be addressed (Ashraf, 2006).

zucchini and squash encode two putative nucleases required for rasiRNA production in the Drosophila germline; Zucchini and Squash physically interact with Armitage

RNAi is a widespread mechanism by which organisms regulate gene expression and defend their genomes against viruses and transposable elements. This study reports the identification of Drosophila zucchini (zuc) and squash (squ), which function in germline RNAi processes. Zuc and Squ contain domains with homologies to nucleases. Mutant females are sterile and show dorsoventral patterning defects during oogenesis. In addition, Oskar protein is ectopically expressed in early oocytes, where it is normally silenced by RNAi mechanisms. Zuc and Squ localize to the perinuclear nuage and interact with Aubergine, a PIWI class protein. Mutations in zuc and squ induce the upregulation of Het-A and Tart, two telomere-specific transposable elements, and the expression of Stellate protein in the Drosophila germline. These defects are due to the inability of zuc and squ mutants to produce repeat-associated small interfering RNAs (Pane, 2007).

In eukaryotic organisms, RNAi, or “RNA interference,” controls a wide variety of biological processes, including development, genome organization, and virus and transposable elements defense. RNAi is triggered by small RNA molecules, which can be grouped in three classes: siRNAs, micro-RNAs (miRNAs), and repeat-associated small interfering RNAs (rasiRNAs). In Drosophila, Dcr2 is responsible for the maturation of the siRNAs from long dsRNA, while the Dcr1/Loquacious complex produces miRNAs from hairpin structures. siRNAs and miRNAs are then incorporated into specific RNP complexes, which are named, respectively, RISC (RNA-induced silencing complex) and miRNP. Core components of the RISC and miRNP complexes are members of the Argonaute (Ago) family, like Ago1 and Ago2. While RISC has been shown to target the transcripts for destruction, the miRNP complex is implicated in the control of mRNA translation. The third class of small RNAs, the so-called rasiRNAs, shares sequence complementarity with mobile elements, satellite and microsatellite DNA, and tandem repeats. It has recently been reported that the biogenesis of the rasiRNAs does not proceed through Dcr1 and Dcr2, thus pointing to a novel mechanism for the maturation of these molecules. rasiRNAs are thought to assemble into RNP complexes containing members of the PIWI family, such as Piwi and Aubergine (Aub), which are involved in chromatin organization as well as in triggering target mRNA destruction to protect the fly genome from selfish genetic elements (Pane, 2007 and references therein).

RNAi has been shown to be involved in axial polarization in the Drosophila germline. In this species, establishment of dorsal-ventral (DV) and anterior-posterior (AP) axes is achieved through the localized translation of specific mRNAs. The protein products of gurken (grk) and oskar (osk) genes are essential for this process. Early during oogenesis, grk RNA encoding a TGFα-like molecule is localized to the posterior of the oocyte, where it signals the posterior fate to the adjacent follicle cells. Following the reorganization of the microtubule cytoskeleton at stage 8, the oocyte nucleus and grk RNA are relocalized to the dorsal-anterior corner of the oocyte. Grk protein now induces dorsal cell fates in the surrounding epithelial cells. In contrast to Grk, which is expressed throughout oogenesis, osk mRNA is kept silenced early during oocyte development. At later stages, Osk protein is found at the posterior of the oocytes, where it directs the organization of the germ plasm as well as abdomen formation of the future embryo. The silencing of oskar translation from stage 1 to 6 is controlled by a set of genes, including armitage (armi), maelstrom (mael), spindle-E (spn-E), and aubergine (aub), which have been shown to be required for RNAi phenomena. Mutations in these genes induce ectopic expression of Osk at early stages of oocyte development. This observation revealed a connection between the RNAi machinery and the establishment of the AP axis during Drosophila oogenesis. armi encodes the homolog of Arabidopsis SDE-3 helicase, which plays a role in post-transcriptional gene silencing (PTGS), a mechanism closely related to RNAi. mael encodes an evolutionarily conserved protein that is required for the proper localization of Ago2 and Dicer, two components of the RNAi machinery. aub and spn-E encode a member of the PIWI class of Argonaute proteins and a DExH RNA helicase, respectively. Aub and spn-E are involved in the silencing of some classes of transposable elements and tandem repeats in the germline, in heterochromatin formation, in double-stranded RNA (dsRNA)-mediated RNAi in embryos, and in the defense against viruses. Interestingly, spn-E and aub are also involved in telomere regulation. In most eukaryotes, the telomeres are maintained through the action of telomerase, the enzyme that ensures the addition of six- to eight-nucleotide arrays to the chromosome ends. However, in Drosophila, telomere elongation occurs after the transposition of non-long-terminal repeat (non-LTR) HeT-A, TAHRE, and TART retrotransposons. Mutations in spn-E and aub cause the upregulation of Het-A and TART expression in the germline, which, in turn, increases the frequency of telomeric element attachments to chromosome ends (Pane, 2007).

This study shows that the genes zucchini (zuc) and squash (squ) are required early during oogenesis for the translational silencing of osk mRNA and at later stages for proper expression of the Grk protein. It is proposed that insufficient levels of Grk protein in zuc and squ mutants are at least partially due to activity of a checkpoint that affects Grk translation, similar to the effects of DNA repair mutants in meiotic oocytes. zuc encodes a member of the phospholipase-D/nuclease family, while squ encodes a protein with limited similarity to RNAase HII. Like Aub, Mael, and Armi proteins, Zuc and Squ localize to nuage, an electron-dense structure surrounding the nurse cell nuclei implicated in RNAi and RNA processing and transport. Zuc and Squ physically interact with Aub, thus pointing to a direct role for these proteins in the RNAi mechanisms. In further support of this conclusion, it has been demonstrated that zuc and squ are required for the biogenesis of rasiRNAs in ovaries and testes. Accordingly, mutations in these genes abolish the production of this class of siRNAs and lead to the deregulation of transposable elements and tandem repeats in the Drosophila germline (Pane, 2007).

zucchini and squash cause dorso-ventral patterning defects and egg chamber abnormalities during oogenesis: zuc and squ were identified in a screen for female sterile mutations on chromosome II of Drosophila. zuc and squ mutant females are viable, but produce eggs with a range of DV patterning defects. Flies with the most severe allele of zuc, zucHM27, lay few eggs, all of which are completely ventralized and often collapsed, whereas those with the weaker alleles, zucSG63 and zucRS49, produce some eggs with a more normal eggshell phenotype in addition to the ventralized eggs). In addition, a P element insertion in the coding region of the gene also acts as a strong loss-of-function allele with ventralized eggshell phenotypes. Three independent alleles of squ were recovered from the screen, namely squHE47, squPP32, and squHK3, and these alleles also generate a range of ventralized eggshell phenotypes (Pane, 2007).

Similar eggshell phenotypes have been described for mutations in other spindle class genes, which include both DNA repair enzymes such as spindle-B (spn-B) or okra (okr), as well as the RNAi components spn-E, aub, and mael. Similar to the spindle class mutants, several additional developmental defects can be observed in the zuc and squ mutants during oogenesis. In the wild-type oocyte, the nucleus condenses in a compact sphere, known as the karyosome. In contrast, the DNA in the nuclei of zuc and squ oocytes appears dispersed or in separate structures. Since compaction of chromatin in the karyosome occurs at stage 3, the defects observed in zuc and squ egg chambers indicate a function for the genes in the early development of the oocyte. Similar to spnE mutants, in a small number of zuc and squ egg chambers the oocyte is not positioned at the posterior as in wild-type, but is found in the middle of the egg chamber. Finally, fusion of egg chambers can also be observed in zuc mutants, resulting in egg chambers with 30 nurse cells and two oocytes. Many egg chambers in the zuc mutant undergo degeneration at different stages (Pane, 2007).

Grk expression is affected in zuc and squ mutants: The DV patterning defects suggested that the Gurken protein is not properly expressed in the mutant egg chambers. In earlier stages of oogenesis, Grk protein is detected in the oocyte similar to the wild-type egg chambers. At stage 9 in wild-type oocytes, Grk is localized in a cap above the oocyte nucleus, where it specifies the dorsal fate of the adjacent follicle cells. In zuc mutants, the amount of Grk protein found in the dorsal-anterior corner of the oocyte is strongly reduced or absent, suggesting that zuc controls the expression of Grk during mid-oogenesis. To further address this question, the distribution pattern of the grk transcript was analyzed in wild-type and zuc mutant egg chambers. In wild-type, grk mRNA localization mirrors the distribution of the protein and is found in the dorsal-anterior corner of the oocyte. Similarly, in zuc mutant egg chambers, grk mRNA is properly localized during mid-oogenesis. zuc therefore affects accumulation of the Grk protein in mid-oogenesis, most likely affecting the translation of the transcripts. This phenotype is also characteristic of the spindle class mutants in general (Pane, 2007).

In squ mutants, Grk protein also fails to accumulate properly in the oocyte at stage 9. Similar to zuc, analysis of grk transcripts in these mutants revealed that the grk mRNA is correctly localized in the majority of the squ egg chambers in mid-oogenesis. This result suggests that squ is also required for Grk translation (Pane, 2007).

zuc and squ do not belong to the spindle class of dna repair genes: The analysis of the zuc and squ egg chambers revealed defects, which place them into the spindle class genes. The spindle genes can be grouped into different categories: the DNA repair genes, the RNAi genes, and a class of translational regulators. The DNA repair genes are implicated in the repair of DNA double-strand breaks which are induced during meiotic recombination by the topoisomerase Mei-W68, a homolog of yeast Spo11. Mutations in these DNA repair genes result in the activation of a meiotic checkpoint mediated by mei-41, the Drosophila ATR homolog. Mei-41 activates the kinase Chk2 also called Mnk in Drosophila, and the activity of Chk2 results in a downregulation of Gurken translation. The resulting reduction in Gurken protein accumulation leads to the ventralized eggshell phenotype. As predicted for a mediator between DNA damage and grk translation, mutations in mei-41 and chk2 are able to suppress the phenotypes caused by mutations in the DNA repair genes. Accordingly, wild-type morphology is restored, for instance, in the eggs of flies doubly mutant for spn-B and mei-41. To assess whether zuc and squ belong to the DNA repair genes, zuc; mei-41 and squ; mei-41 double mutant flies were generated, and the eggs laid by these females were checked for the presence of DV patterning defects. In both cases, the persistence of dorso-ventral patterning defects was observed, indicating that zuc and squ do not likely belong to the class of DNA repair enzymes. Flies doubly mutant for zuc and chk2 and squ and chk2 were generated. Interestingly, it was found that while patterning defects persist in the eggs of zuc chk2 flies, wild-type morphology is restored in the eggs laid by squ chk2 homozygous females. Suppression of the eggshell ventralization phenotypes was also observed in chk2 aub mutants, but not in chk2; spn-E or chk2 piwi double mutants. This demonstrates that a checkpoint mediated by Chk2 is largely responsible for the low levels of Grk protein in aub and squ mutants. The fact that zuc, spnE, and piwi phenotypes are not suppressed by chk2 mutations suggests that they may have multiple effects on oogenesis, some of which may act independent of checkpoint activity (Pane, 2007).

Molecular analysis of the zuc and squ genes: A set of deficiencies was used to map the zuc mutation to region 33B5 of chromosome II. Transformation rescue experiments narrowed the region to a candidate region of 5 kb, containing two transcripts: CG12314 and CG16969. Sequence analysis revealed that all the zuc mutations reside in CG12314. zuc encodes a member of the phospholipase-D/nuclease family and is characterized by one copy of a conserved H(X)K(X4)D (HKD) motif. Notably, members of the family having two HKD domains are classified as phospholipase-D proteins, while members with one HKD domain have been shown to catalyze the hydrolysis of double-stranded RNA and DNA molecules in vitro. Hence, Zuc is likely to be a nuclease. The Histidine (H) residue of the HKD domain is essential for the function of the phospholipase-D/nuclease proteins, since substitution of the H residue results in a strong reduction of the catalytic activity in vitro. Interestingly, the substitution of the H of the catalytic domain with a Tyrosine in the zucSG63 allele generates a strong loss-of-function allele. zucHM27 is generated by the introduction of a stop codon at residue 5, resulting in a putative protein null allele. Finally, the zucRS49 allele contains a substitution of the Serine47 with an aspartic acid residue. Transformation rescue experiments confirmed that CG12314 corresponds to zuc (Pane, 2007).

Recombination mapping placed squ on the left arm of the second chromosome at map position 2-53. Deficiency mapping and P-element-mediated male recombination placed squ into a region containing six candidate genes including her and grp. Complementation tests and sequence analysis argued against the six genes as candidates to be squ. Upon closer inspection of the grp locus a gene, CG4711, was seen to be nested in the first intron, which had previously been predicted to encode an alternate splice exon of grp. CG4711 as sequenced in squHE47, squPP32 and squHK35 and squHE47 and squPP32 were found to both contain single nucleotide changes resulting in nonsense codons in CG4711 at residues 100 and 111, respectively. No mutations were identified in the predicted CG4711 coding region in squHK35. Transformation rescue experiments confirmed that CG4711 corresponds to squ. This gene encodes a protein with similarity to RNAase, which is known to catalyze the degradation of RNA moieties in DNA-RNA hybrids (Pane, 2007).

Zuc and Squ localize to the nuage and physically interact with Aub: The “nuage” is a cytoplasmic organelle that is widely conserved in evolution. Homologous structures exist in all eukaryotic organisms and are thought to play a fundamental role in germline functions. In Drosophila, the nuage appears as an electron-dense, punctate fibrous structure that surrounds the nuclei of the nurse cells in the egg chambers. This organelle is thought to be a staging site where ribonucleoprotein complexes originating in the nuclei are remodeled, before they are transported to specific localizations in the cells. Recent studies have also shown that the nuage is implicated in RNAi. For instance, in human cell lines Ago1 and Ago2 proteins localize to cytoplasmic bodies, called P bodies, which are thought to be homologous to the Drosophila nuage. Similar to the P bodies, Drosophila nuage hosts molecules required in RNAi phenomena like Aub, Armi, and Mael. In addition, mutations in mael, another component of the RNAi machinery, disturb the nuage granules, resulting in a displacement of the RISC components Ago2 and Dcr1. To analyze the expression pattern of Zuc during oogenesis, transgenic lines were produced that express Zuc fused to EGFP. Live imaging on ovaries dissected from these lines show a strong accumulation of Zuc in the nuage. Zuc is also found in cytoplasmic particles. Immunostaining on lines expressing Zuc fused to triple HA tag confirmed these observations. Similar to Zuc, Squ protein localizes to the nuage and in cytoplasmic particles as demonstrated by the immuno localization analysis of triple-HA-Squ transgenic lines (Pane, 2007).

These results show that Zuc and Squ localize to the nuage similar to Aub. aub encodes a member of the Piwi class of Ago proteins and has been shown to be implicated in different RNAi processes in Drosophila germline. Furthermore, the inability of aub mutants to assemble RNAi complexes in the germline led to the hypothesis that Aub might be a core component for RNAi-induced complexes in this tissue. Remarkably, both Zuc and Squ were found interact with Aub in vivo, consistent with the cellular localization of these proteins. AubGFP lines were crossed to triple-HA-Zuc and triple-HA-Squ strains, respectively. CoIP was performed with GFP- and HA-specific antibodies on ovaries of doubly transgenic flies. Bands corresponding to HA-Zuc and HA-Squ are detected in the IP lanes, while no signal above background is present in the control lanes (Pane, 2007).

Mutations in zuc and squ Activate the Expression of Osk in early oocytes: A hallmark of the spindle class genes that are involved in RNAi is the control of Osk translation at early stages of development. In wild-type oocytes, osk mRNA is silenced from stage 1 to 6 through RNAi dependent mechanisms. The translational repression of osk mRNAs at these stages is thought to involve the miRNA miR280. In contrast, ectopic translation of Osk is observed in early stages of armi, aub, spnE, and mael mutant egg chambers. To assess whether zuc and squ are involved in RNAi, the expression pattern of Osk was analyzed in zuc and squ mutant egg chambers. It was found that Osk is properly translated and localized at late stages of oogenesis, where it is found at the posterior pole of the oocyte. However, in early egg chambers Osk expression is ectopically activated, and clumps of Osk protein can be observed in the developing oocyte in zuc and squ egg chambers. Osk protein is also found in punctae surrounding the nurse cell nuclei. These results suggest that zuc and squ are involved in the RNAi silencing of osk mRNAs in the nurse cells and the oocyte (Pane, 2007).

Het-A and Tart expression is regulated by zuc and squ: To further test the involvement of zuc and squ in RNAi, the expression levels of Het-A and Tart, two telomere-specific retrotransposons, were analyzed in the ovaries of zuc and squ mutants. In Drosophila, telomere maintenance is achieved through the transposition of retrotransposons to the chromosome ends. The telomere elements in Drosophila are non-LTR-containing retrotransposons, which transpose to the chromosome ends via a poly(A)+ RNA intermediate. The mechanism of transposition is well characterized, and recent work has shown that the RNAi machinery is involved in the maintenance of the telomeres. Aub and spnE have been shown to regulate the expression of a number of transposable elements in the germline of Drosophila. In particular, mutations in aub and spnE were discovered to trigger the upregulation of the Het-A and Tart elements, two telomere-specific retrotransposons. This process occurs in the germline of Drosophila, but not in the soma, and results in the addition of extra elements to the telomere array. Since Zuc and Squ are found in a complex with Aub, whether they also share a similar function in this process was tested. To this aim, quantitative RT-PCR was performed on total RNA extracted from heterozygous zucHm27/+ and transheterozygous zucHm27/Df(2L)PRL ovaries. Df(2L)PRL is a deletion that uncovers the genomic region containing the zuc gene. Comparison of the two samples reveals more than 1000-fold upregulation of the Het-A element in the germline of zucHm27/Df(2L)PRL flies. A significant increase in the expression levels of Tart can be observed in zuc mutants, where this element is upregulated by 15-fold. Elevated levels of Het-A, but not Tart, can be observed in the ovaries dissected from squHE47/squPP32 mutant females as compared to the control squHE47/+ flies. It is possible that the levels in the heterozygous control flies are already somewhat elevated over wild-type, but since different wild-type backgrounds may vary, heterozygous flies were used as control. These results show clearly that, similar to aub and spnE, zuc and squ are required for the silencing of retrotransposons in the Drosophila germline (Pane, 2007).

Stellate silencing is impaired in testes of zuc and squ mutants: The Stellate (Ste) locus in Drosophila resides on the X chromosome and encodes a protein with homology to the β-subunit of protein kinase CK2. While the protein is normally expressed in wild-type females, it is downregulated in wild-type males through the activity of RNAi-based mechanisms. The Y chromosome of Drosophila contains the crystal locus, also called Suppressor of Stellate [Su(Ste)], which shares 90% degree of identity with Ste. The insertion of a Hoppel transposon in the region 3′ to Su(Ste) causes the transcription of antisense transcripts in addition to the sense mRNAs. Sense and antisense RNAs are thought to drive the dsRNA-mediated degradation of Ste target mRNAs. This mechanism is required in males to silence the approximately 200 repeats of the Ste locus located on the X chromosome. In males carrying a deletion of the bulk cry locus, or mutations in RNAi genes like spnE, aub, and armi, expression of Ste is relieved, which in turn leads to the accumulation of needle-shaped crystals in testes and meiotic abnormalities. To test whether zuc and squ are required for the RNAi silencing of Ste tandem repeats, testes of mutant males were stained with a Ste-specific antibody. While no signal can be detected in wild-type males, Ste crystals can be easily observed in zuc and squ mutant testes. These results demonstrate that zuc and squ are required for the silencing of tandem repeats in the Drosophila germline (Pane, 2007).

rasiRNAs biogenesis is impaired in zuc and squ mutants: The upregulation of transposable elements and tandem repeats in the germline of zuc and squ mutants pointed to a role for the Zuc and Squ proteins in the rasiRNA pathway. Hence, attempts were made determine whether these proteins are involved in the biogenesis of the rasiRNAs or rather in the mechanism which causes the silencing of selfish genetic elements. To this aim, northern blot analysis was performed on total RNA extracted from fly ovaries and testes and probed for abundant rasiRNAs. In particular, the level of expression of two recently cloned rasiRNAs, namely the roo rasi and the Su(Ste) rasi, was measured. To minimize the background effects, the production was compared of rasiRNAs in homozygous or transheterozygous mutants versus heterozygous flies. Hybridization with an antisense oligonucleotide to roo rasi reveals that rasiRNAs are not produced in the ovaries of flies mutant for zuc, aub, and spnE. A reduction of rasiRNA levels can also be observed in the ovaries of squ mutant flies, though the production of these small RNAs is not completely abolished like in zuc, aub, and spnE mutants. Hybridization of the same membranes with an antisense oligonucleotide to miR310 shows that miRNA levels are not affected in the mutants analyzed. As a loading control a final hybridization was performed with a 2S rRNA antisense probe (Pane, 2007).

Northern blots on total RNA extracted from testes were probed with an antisense oligonucleotide to Su(Ste) rasi. This experiment revealed that, similar to aub and spnE, rasiRNAs are not produced in testes of flies mutant for zuc and squ. Also in this case, hybridization with a probe corresponding to 2S rRNA was used as a loading control (Pane, 2007).

These results demonstrate a role for zuc and squ in the biogenesis of rasiRNA in the Drosophila germline (Pane, 2007).

These studies have shown that Drosophila zuc and squ control the expression of Grk and Osk, thus affecting the axial patterning of the oocyte and future embryo. The silencing of Osk at early stages is known to be controlled by RNAi-dependent mechanisms, suggesting that Zuc and Squ are involved in RNAi processes. In support, it was found that Zuc and Squ localize to the nuage and interact with Aub, a PIWI/PAZ protein that is required for the assembly of RISC complexes in the Drosophila germline. In this tissue, RNAi ensures genomic stability by silencing selfish genetic elements. Consistent with a role in a silencing RNAi process, the upregulation of some classes of transposable elements was observed in ovaries and expression of tandem repeats in testes of zuc and squ mutants (Pane, 2007).

Osk translation is silenced at early stages of oocyte development by the activity of RNAi-related proteins, namely Armi, Mael, Aub, and spn-E. Similar to armi, mael, aub, and spn-E, mutations in zuc and squ lead to early expression of Osk protein in stage 1–6 oocyte. miRNAs have been shown to mediate translational repression of target mRNAs by base-pairing with their 3′UTR. A computational approach revealed that osk 3′UTR contains a sequence complementary to miR-280, which is also found in a number of putative target genes, including kinesin heavy chain mRNA. However, the results reported here together with previous data show that miRNA biogenesis is not affected by mutations in squ, zuc, aub, armi, and spnE. Therefore, it is proposed that Zuc and Squ, together with Aub, Armi, Mael, and spn-E, might act in concert to allow the assembly of a miR-280 miRNP complex and the silencing of osk and other target genes (Pane, 2007).

Previous studies demonstrated that Aub and spn-E are implicated in the suppression of transposable element mobilization in the Drosophila germline. This process is based on RNAi mechanisms and requires a class of siRNAs called rasiRNAs. rasiRNAs are particularly abundant in the Drosophila germline and are complementary to tandem repeats, transposable elements, and satellite DNA. It was recently reported that rasiRNAs corresponding to retro-elements, like SINE, LINE and LTR retrotransposons, are also present in mouse oocytes, thus suggesting that a conserved RNAi machinery exists in eukaryotes that ensures genome stability by silencing selfish genetic elements. This study shows that, like aub and spn-E, zuc and squ regulate the expression of some classes of transposable elements and tandem repeats in the Drosophila germline. The expression of the Het-A and Tart retrotransposable elements was analyzed and it was found that they are upregulated in zuc and squ mutant egg chambers. In addition, expression of Ste protein, which is downregulated by dsRNA-mediated degradation of Ste mRNA in wild-type males, is activated in squ and zuc mutant males. Consistent with a role in RNAi, Zuc and Squ were shown to localize to the nuage together with Aub, and physically interact with Aub, a member of the PIWI class of Argonaute proteins. Interestingly Het-A and Tart are two non-LTR retrotransposable elements, which are implicated in the maintenance of telomere length in Drosophila. Upregulation of these transposons in the egg chambers of aub and spn-E mutant flies leads to a higher rate of transposition to the chromosome ends, resulting in telomere elongation and chromosomal abnormalities. This study shows that zuc and squ regulate the expression of Het-A and Tart, strongly suggesting that they might be involved in telomere regulation in the Drosophila germline (Pane, 2007).

In wild-type egg chambers, Grk localizes in a cap above the oocyte nucleus where it signals the dorsal identity to the surrounding follicle cells. In zuc and squ mutant egg chambers, Grk protein fails to accumulate properly in the dorsal-anterior corner of the oocyte, which results in the production of eggs with various degree of ventralization. A similar phenotype was reported for spn-B, spn-D,spn-A, and okra mutants, in which the DNA double-strand breaks induced during the meiotic recombination are not efficiently repaired. These mutations activate a meiotic checkpoint that involves the Drosophila ATR homolog Mei-41 and Chk-2/mnk. The latter is likely to promote the posttranslational modification of Vasa, a helicase with homology to eIF4A. This modification event is thought to cause the inhibition of Vasa activity and, consequently, the downregulation of grk translation. However, mutations in zuc and squ are not suppressed by mutations in mei-41, supporting the conclusion that these genes do not belong to the DNA repair class. Surprisingly, mutations in chk2/mnk are able to suppress the effects of mutation in squ and aub (Chen, 2007), but not zuc, spn-E, or piwi. This result indicates that squ and aub mutations activate a checkpoint mechanism that involves Chk2, but is not absolutely dependent on Mei-41. Similar to the DNA repair mutants, the checkpoint activity of Chk2 acts to cause the ventralized eggshell phenotype in these mutants. In contrast, zuc and spn-E mutants are not suppressed in combination with the chk2 mutant, even though it was found that Vas is posttranslationally modified in the zuc background, as has been reported for spnE mutations. This suggests that zuc and spnE may also activate the chk2-dependent checkpoint in oogenesis that modifies Vasa, a translational regulator of Grk, as seen in the DNA repair mutants. But Zuc and SpnE appear to affect oogenesis through additional mechanisms, acting not only through Chk-2. Similarly, mutations in armi were also observed to affect oogenesis at multiple levels. It is therefore plausible that Zuc, Squ, SpnE, Armi, and Aub all participate in the downregulation of selfish genetic elements, and that the retrotransposons and tandem repeats activity results in activation of Chk-2. Yet Zuc and Spn-E might have additional effects in oogenesis, similar to Armi, and those effects may be more direct and not mediated by a checkpoint mechanism (Pane, 2007).

Zuc is conserved in evolution and belongs to the phospholipase-D/nuclease superfamily, which contains several proteins with diverse functions. All the members share a conserved HKD domain that is fundamental for the catalytic activity. However, two different groups of proteins can be identified within this family. A group of proteins with two HKD domains includes human and plant PLD enzymes, cardiolipin synthase, phosphatidylserine synthase, and the murine toxin from Yersinia pestis. Members of the superfamily with one HKD domain include several bacterial endonucleases, like Nuc, and a helicase-like protein from E. coli. Zuc contains only one HKD domain and thus belongs to the subgroup of the nucleases. These enzymes have been shown to hydrolyze double-stranded RNA and DNA molecules in vitro, but little is known about their function in vivo. The results of this study demonstrate that zuc is involved in RNAi. Interestingly, it was shown that the biogenesis of the rasiRNAs does not require Dcr1 and Dcr2 and that this class of small RNAs has a different size and structure when compared to other siRNAs. Mutations in the zuc gene impair the production of rasiRNAs, both in ovaries and testes. Therefore, Zuc is involved in the maturation of rasiRNAs and may replace Dcr1 and Dcr2 in the germline rasiRNAs mechanisms. It was recently proposed that Aub is required for the production of the rasiRNAs 5′ ends, while the nuclease implicated in the cleavage of the 3′ termini remains elusive. Given the strong interaction between Zuc and Aub and the absence of rasiRNAs in the zuc mutants, it is tempting to speculate that Zuc might be the nuclease responsible for the production of rasiRNAs 3′ ends in Drosophila. squ encodes a protein with similarity to RNase HII, which is known to degrade the RNA moiety in RNA-DNA hybrids. Mutations in squ do not completely abolish the production of rasiRNAs in ovaries, thus suggesting that this protein might act in the actual silencing mechanism of target genes rather than in the biogenesis of the rasiRNAs. However, the analysis of Su(Ste) rasiRNAs in testes of squ mutants reveals that the Squ protein is essential for the production of rasiRNAs in this tissue. A possible explanation for these data is that Squ exerts a key function in testes together with Zuc, Aub, spnE, and Armi to ensure the proper processing of rasiRNAs. Differently, in ovaries Squ might be partially redundant since a squ paralog exists in Drosophila and might replace in part the function of Squ during oogenesis. Neither Zuc nor Squ are required for biosynthesis of microRNAs, suggesting that they are specific for the production of rasiRNAs (Pane, 2007).

In summary, this study identified the phospholipase-D/nuclease Zucchini and the RNase HII-related protein Squash as members of RNAi processes that function in the germline of Drosophila. Similar requirements for RNAi processes have also been reported for the normal development of the mammalian germline and the germline of C. elegans, and it will be interesting to determine in the future whether Zuc and Squ homologs also participate in germline RNAi in other organisms (Pane, 2007).

Probing the initiation and effector phases of the somatic piRNA pathway in Drosophila

Combining RNAi in cultured cells and analysis of mutant animals, this study probed the roles of known Piwi-interacting RNA (piRNA) pathway components in the initiation and effector phases of transposon silencing. Squash associates physically with Piwi, and reductions in its expression led to modest transposon derepression without effects on piRNAs, consistent with an effector role. Alterations in Zucchini or Armitage reduced both Piwi protein and piRNAs, indicating functions in the formation of a stable Piwi RISC (RNA-induced silencing complex). Notably, loss of Zucchini or mutations within its catalytic domain led to accumulation of unprocessed precursor transcripts from flamenco, consistent with a role for this putative nuclease in piRNA biogenesis (Haase, 2010).

Eukaryotic small RNAs regulate gene expression through various mechanisms, intervening at both transcriptional and post-transcriptional levels. Small RNAs are divided into classes according to their mechanism of biogenesis and their particular Argonaute protein partner. Piwi-interacting RNAs (piRNAs) bind Piwi-clade Argonaute proteins and act mainly in gonadal tissues to guard genome integrity by silencing mobile genetic elements (Haase, 2010).

Conceptually, the piRNA pathway can be divided into several different phases. During the initiation phase, small RNAs, called primary piRNAs, are produced from their generative loci, so-called piRNA clusters. These give rise to long, presumably single-stranded precursor transcripts, which are processed via an unknown biogenesis mechanism into small RNAs that are larger than canonical microRNAs (~24-30 nucleotides [nt]). Primary piRNAs become stably associated with Piwi proteins to form Piwi RISCs (RNA-induced silencing complexes), which also contain additional proteins that facilitate target recognition and silencing. During the effector phase, Piwi RISCs identify targets via complementary base-pairing. In some cases, for example, with Aubergine as a piRNA partner, there is strong evidence for target cleavage in vivo. This nucleolytic destruction of transposon mRNAs is probably the main Aubergine effector mechanism, although this has not been rigorously demonstrated. Piwi also conserves the Argonaute catalytic triad; however, in this case, both its nuclear localization and its association with certain chromatin proteins suggest the possibility of transcriptional and post-transcriptional effector pathways. An additional phase, adaptation, is restricted to germ cells and constitutes the ping-pong cycle. During this phase, transposon mRNA cleavage directed by primary piRNAs triggers the production of secondary piRNAs, whose 5' ends correspond to cleavage sites. These generally join Ago3 and enable it to recognize and cleave RNAs with antisense transposon content, perhaps piRNA cluster transcripts. Cleavage by Ago3 RISC again triggers piRNA production from the target, closing a loop that enables the overall small RNA population to adjust to challenge by a particular transposon. Finally, piRNA populations present in germ cells can be transmitted to the next generation to prime piRNA responses in progeny (Haase, 2010).

In Drosophila follicle cells, only the initiation and effector phases appear relevant. Here, the piRNA pathway relies on the coupling between a single Piwi protein (Piwi itself) and a principal piRNA cluster (flamenco) to silence mainly gypsy family retrotransposons. Drosophila ovarian somatic sheet cells (OSS) display many of the properties of follicle cells, and represent a convenient system to study the initiation and effector phases of the piRNA pathway without the complications inherent in the study of complex tissues in vivo. This study therefore sought to leverage information derived from the use of RNAi in OSS cells with the analysis of ovaries derived from mutant animals to probe the roles of known piRNA pathway components in the initiation and effector phases of transposon silencing (Haase, 2010).

Several prior studies have proposed models in which Piwi proteins silence targets by interfering with their transcription. Since piRNAs are largely absent from somatic tissues, impacts underlying these changes are presumed to have occurred during development and to have been epigenetically maintained in the adult. Drosophila Piwi protein is mainly localized to the nucleus and has been shown to interact with HP1, a core component of heterochromatin. Considered together, this body of evidence pointed strongly to an effector mechanism in which Piwi-associated small RNAs direct heterochromatin formation and silencing of targets (Haase, 2010).

Loss of piwi has dramatic effects on transposon expression in somatic follicle cells. Genetic mutants result in an absence of Piwi protein throughout development. This could lead to a failure to create heterochromatic marks that could have otherwise maintained epigenetic silencing of transposons in the absence of continuous Piwi expression. Alternatively, there could be an ongoing requirement for Piwi to maintain silencing, irrespective of whether it acted via transcriptional or post-transcriptional mechanisms (Haase, 2010).

To discriminate between these possibilities, OSS cells were transfected with dsRNAs corresponding to piwi, and followed impacts on Piwi mRNA and protein levels. Maximal suppression was reached by 3 d, and silencing persisted through day 6. At 6 d post-transfection, impacts on two elements known to be derepressed in the follicle cells of piwi mutant ovaries were probed: gypsy and idefix. Both showed derepression (up to 10-fold) upon piwi silencing. Additional elements were also tested, with blood being impacted strongly. Previous studies have also implicated zucchini (zuc) in the function of the somatic piRNA pathway. RNAi against this gene also increased gypsy, blood, and idefix expression. Considered together, these results demonstrate that the integrity of the piRNA pathway is essential for the ongoing repression of mobile elements and argue against a model in which silent epigenetic states, once set by the action of piwi proteins on chromatin, can autonomously maintain transposon silencing (Haase, 2010).

Nearly a dozen proteins have been linked to the fully elaborated piRNA pathway that operates in germ cells. Many of these show germ cell-specific expression patterns consistent with their selective biological effects. Mutations in armitage (armi) result in coincident loss of the characteristic nuclear accumulation of Piwi protein and a reduction in Piwi-associated piRNAs. Unlike most germline-specific pathway components, an examination of RNA-seq data from OSS cells indicated substantial armi expression. Therefore armi was suppressed by RNAi and effects on transposon expression were examined. Notably, gypsy, blood, and idefix were strongly derepressed, implying a role for armi in both the somatic and germline compartments (Haase, 2010).

The Drosophila mutant armi1 represents a P-element insertion in the 5' untranslated region (UTR) of armitage. A second allele, armi72.1, was derived from armi1 by imprecise excision. RNA-seq data covered the armi ORF in OSS, but no reads were detected corresponding to the germ cell 5' UTR. This raises the possibility that armi expression might be driven by an alternative promoter in somatic cells, and that the armi alleles examined thus far may have spared the activity of that promoter (Haase, 2010).

To investigate whether Armi and Zuc act at the initiation or effector phase of the piRNA pathway, piRNAs were examined. Silencing of piwi reduced levels of two abundant piRNAs, corresponding to gypsy, or idefix. Similar effects were noted upon silencing of armi or zuc. Aggregate OSS piRNA levels can be measured qualitatively by radioactive phosphate exchange of small RNAs in Piwi immunoprecipitates. As expected, RNAi against piwi virtually eliminated piRNAs in immunoprecipitates. Silencing of armi or zuc produced indistinguishable effects (Haase, 2010).

In germ cells, armi mutation causes loss of the prominent nuclear localization of Piwi. A similar phenotype was observed upon knockdown of armi in somatic OSS cells. Because of the mixed cell types present in ovaries, previous studies had not been able to distinguish whether Armi loss simply caused Piwi mislocalization or whether Armi influenced Piwi expression or stability. In OSS cells, knockdown of armi reduced Piwi protein levels by approximately fivefold, equivalent to a targeted knockdown of Piwi itself without affecting piwi mRNA. A similar loss of Piwi protein from the nuclei in cells exposed to zuc-dsRNAs was noted. In this case, Piwi protein but not mRNA levels also fell (Haase, 2010).

Considered together, these data strongly suggest roles of Armi and Zuc in the initiation phase of the piRNA pathway. A role for Armi, along with a previously unrecognized component, Yb, in the somatic pathway, is also supported by a recent report. Either protein could play a role in primary piRNA biogenesis, aiding piRNA production or loading, with this model resting on the presumption that association with mature piRNAs influences Piwi protein stability. Alternatively, Armi or Zuc could be core components of mature Piwi RISC, with loss of either subunit destabilizing associated components of the complex (Haase, 2010).

To investigate these alternative models, proteomic analysis of Piwi RNPs was performed. Piwi immunoprecipitates contained a number of peptides from Armi, suggesting that this protein is present in Piwi RISC. Of note, association of both Piwi and Armi with Squash (Squ), another previously identified piRNA pathway component, was also detected. Piwi could be also detected in Squ immunoprecipitates by Western blotting. Although no Zuc peptides were seen in multidimensional protein identification technology (MudPIT), Piwi could be detected to a low extent in Zuc immunoprecipitates. Overall, the emerging picture suggests that both Armi and Squ are components of Piwi RISC. Lower levels of Piwi associated with Zuc might indicate a weaker or more transient association of Zuc with Piwi RISC (Haase, 2010).

Mutations in squash (squ) show little impact on piRNA populations in mutant ovaries. Similarly, upon sequencing of small RNAs in Piwi immunoprecipitates, no differences were detected in associated piRNA populations upon comparison of squ homozygous mutant animals to heterozygous siblings. Animals harboring two squ alleles interrupted by early stop codons did, however, display an effect on transposon silencing (Haase, 2010).

As compared with heterozygous siblings, squ mutants showed significant derepression of gypsy. This occurred without any detectable change in an abundant gypsy piRNA or overall Piwi levels. In contrast, no substantial changes were detected in idefix or ZAM; however, I-element and blood were strongly derepressed (Haase, 2010).

Considered together, these results point to a role of squash in the effector phase of the piRNA pathway. A slight but reproducible reduction was noted in Piwi protein levels in homozygous squ mutants. However, this was well within the range observed in Piwi heterozygotes, where the piRNA pathway functions completely normally (Haase, 2010).

In the initial screen that placed zuc within the piRNA pathway, two alleles were identified. zucHM27 represents an early stop mutation resulting in a putative null allele (referred to as zuc mut). This mutant strongly affects piRNA silencing in both germline and somatic cells of the ovary. While somatic piRNAs are depleted in this mutant, ping-pong signatures remain intact. This places Zuc outside of the adaptive phase, consistent with accumulating evidence for a role in the initiation phase (Haase, 2010).

While the biochemical properties of Zuc have yet to be analyzed, its protein sequence places it as a member of the phospholipase D (PLD) family of phosphodiesterases. These share a HxK(x)4D motif, whose integrity is essential for catalytic activity. The second zuc mutation that emerged in the original screen, zucSG63, contains a H --> Y mutation within the phosphodiesterase motif that is predicted to render it catalytically inactive. To probe a role for Zuc catalytic activity in the piRNA pathway, the presumed null (zuc mut) and catalytically dead (zuc H --> Y) alleles were compared for their effects on piRNAs and transposon silencing (Haase, 2010).

Total ovarian small RNAs were analyzed from animals that were heterozygous or homozygous for the zuc H --> Y allele and the resulting profiles were compared to previously published analyses of the presumed zuc mut allele. In both cases, strong reductions were seen in total piRNAs and in populations that mapped uniquely to the flamenco locus, regardless of the normalization method used to compare libraries. Slightly stronger impacts were apparent when profiles of Piwi immunoprecipitates were compared. Here, piRNA populations corresponding to flamenco were almost completely lost. An accumulation was noted of 21-nt species in Piwi immunoprecipitates from both zuc mutant lines. These were enriched for a 5' U, although not to the extent for longer piRNA species. The nature of these shorter, apparently Piwi-associated RNAs remains mysterious (Haase, 2010).

Both the presumed null and H --> Y zuc alleles impacted transposon silencing. Between fivefold and 20-fold increases in gypsy, ZAM, and idefix were noted in comparison with heterozygous controls. Even stronger derepression could be observed for I-element, HeT-A, 1731, and blood. The zuc H --> Y and zuc mut alleles also showed similar impacts on piRNA populations and the overall levels of Piwi protein (Haase, 2010).

Considered together, these data point to a requirement for the presumed catalytic center of Zuc in the initiation phase of the piRNA pathway. Other PLD family nucleases that have been characterized to date cleave nucleic acids leaving 5' phosphate and 3' hydroxyl termini. These are the characteristics one might expect for a processing enzyme that catalyzed primary piRNA biogenesis. Previous studies have posited the requirement for several nucleolytic activities in the piRNA pathway. One is thought to form the 5' ends of primary piRNAs. The 3' ends of these species could be formed prior to Piwi loading or could be coupled to protein binding, as is posited for the ping-pong cycle. The nucleolytic center of Piwi proteins themselves form the 5' ends of secondary piRNAs, with their 3' ends proposed to be created by a separate enzyme. Based on its impacts in the soma on Piwi complexes, it is imagined that the Zuc catalytic center might form either the 5' or 3' ends of primary piRNAs (Haase, 2010).

To evaluate this hypothesis, RNAs derived from the flamenco locus were examined in control ovaries or in tissues from animals homozygous for either of the two zuc mutant alleles. The prevailing model holds that the flamenco locus is transcribed as a continuous, single-stranded precursor spanning >150 kb. It was reasoned that a defect in primary processing might result in an accumulation of long RNAs from this locus, since they would not be effectively metabolized into piRNAs. By quantitative PCR (qPCR) using primer pairs spanning three different regions of flamenco, 15-fold to 45-fold increases were seen in flamenco-derived long RNAs in zuc mutant ovaries (Haase, 2010).

Considered as a whole, these results strongly support a role for Zucchini in the primary processing of piRNAs from the flamenco locus. Given its size, it is virtually impossible to follow the fate of the intact flamenco transcript by Northern blotting. Three different segments of the locus do show an accumulation consistent with their failure to be parsed into piRNAs. However, several alternative explanations can also be envisioned. For example, if Zucchini impacts Piwi stability, feedback controls might operate to inhibit primary biogenesis. Without a direct, biochemical demonstration that Zucchini processes piRNA cluster transcripts, its assignment as a primary biogenesis enzyme must be viewed as provisional. However, any alternative model must account for the requirement for its phosphodiesterase active site, and, at present, a direct role in piRNA biogenesis seems the most parsimonious conclusion (Haase, 2010).

While these studies do not ascribe specific functions to Armitage and Squash, they do support their assignment to the initiator and effector phases, respectively. Armitage is a putative helicase, although no analyses as yet indicate whether this biochemical activity is required for its function. Placement of this protein in the initiation phase and its intimate association with Piwi perhaps suggest a role in loading or stability of Piwi RISC. Squash, of all the components examined in this study, had the most variable effects on transposon control in somatic cells of the ovary, but both its physical association with Piwi RISC and its impact on transposons without an effect on piRNAs imply a role in the effector phase. While the studies reported herein can suggest roles for known pathway components at specific points in the piRNA pathway, a definitive conclusion regarding the part played by any of these proteins will require reconstitution of the pathway in vitro (Haase, 2010).



To examine the distribution of endogenous Armi protein during oogenesis, polyclonal antibodies were generated against a C-terminal peptide and N-terminal fusion protein. On Western blots, the affinity-purified antibodies react with a polypeptide of the molecular weight expected for Armi, and both antibodies strongly label germline structures during oogenesis that are absent in armi mutants. These antibodies also produce diffuse labeling in the somatic follicle cells in wild-type and armi mutant egg chambers. This may be nonspecific labeling, because whole-mount in situ hybridization shows that armi transcripts are restricted to the germline and clonal analysis indicates that armi function is required in the germline for female fertility and axial patterning. The GFP-Armi fusion protein is incorporated into similar germline structures in living ovarioles, confirming that the antibody labeling is specific. Armi protein is first apparent early in oogenesis, in the cytoplasm of stem cells and mitotically dividing cystoblasts. In regions 2a and 2b of the germarium, Armi protein is most concentrated in the center of the germline cysts, where the pro-oocyte is located. In stage 1 and early stage 2 egg chambers, Armi accumulates at the anterior of the oocyte, near the ring canals. Armi also extends through the ring canals forming a branched structure that links the early oocyte with adjacent nurse cells. In stage 3 cysts, Armi accumulates at the posterior cortex and localizes to extensions that pass through the oocyte into the nurse cells (Cook, 2004).

The distribution of Armi in the germarium is reminiscent of the fusome, a vesiculated structure rich in membrane skeletal proteins that plays a critical role in orienting the cystoblast divisions. Double immunolabeling for Armi and the fusome marker Adducin demonstrate that the branched structure formed by Armi assembles as the fusome degenerates. During early oogenesis, Armi protein also accumulates in punctate structures around the nurse cell nuclei. This distribution is characteristic of nuage, amorphous perinuclear material implicated in RNA processing and transport. Through stages 4 to 7, Armi continues to be somewhat enriched at the posterior cortex of the oocyte, but at significantly lower levels. In stage 9 to 10 egg chambers, Armi is found throughout the cytoplasm of the oocyte and nurse cells, with slight enrichment at the oocyte cortex (Cook, 2004).

Armi protein and osk mRNA both accumulate in the oocyte during early oogenesis, when armi mutations lead to premature Osk protein expression. To define the spatial relationship between Armi protein and osk mRNA, wild-type egg chambers were immunostained for Armi and labeled for osk mRNA by FISH. In the germarium, osk mRNA and Armi protein are concentrated in similar regions, but the distributions do not precisely overlap. During stages 3 through 6, Armi protein and osk transcripts are concentrated near the posterior cortex of the oocyte in close proximity, but they assemble into distinct structures. These observations indicate that most of the osk mRNA in the oocyte is not physically associated with Armi protein (Cook, 2004).

The branched structure formed by Armi in early egg chambers also resembles the polarized microtubule cytoskeleton that directs asymmetric mRNA localization. Immunolabeling of wild-type ovaries for Armi and α-tubulin demonstrated that Armi protein is closely associated with the microtubule network in early egg chambers. Moreover, disruption of the microtubule network with colchicine disrupts the branched Armi network. However, Armi does not precisely colocalize with microtubules. In stage 1 egg chambers, for example, Armi accumulates near the MTOC and along microtubules that extend from the MTOC and pass through the ring canals. In single confocal optical sections, the most intense Armi labeling does not overlap with the MTOC or microtubule bundles, which appear to exclude Armi. Thus, most of the Armi protein does not directly associate with microtubules. Nonetheless, microtubules are required to establish the asymmetric distribution of Armi during early oogenesis (Cook, 2004).


To identify new genes involved in embryonic axis specification, a collection of P element transposon-induced maternal effect lethal mutations was screened for impaired localization of the pole plasm component Vasa and for disruption of D/V patterning as indicated by defects in the dorsal appendages on the eggshell. This screen identified a new mutation that produces variable appendage defects and significantly reduces Vasa accumulation at the posterior pole. Molecular and genetic analyses have demonstrated that the mutation was not in a previously identified gene. The locus was named armitage after the navigator on Robert Falcon Scott's failed Discovery expedition to the South Pole (Cook, 2004).

Excision of the P element in the original mutant, armi1, produced new armi alleles and revertant chromosomes that presumably restore gene function through precise transposon excision. Of the forty lines generated, 21 (52%) fully complemented armi1 and were homozygous fertile with normal embryonic patterning. The remaining lines failed to complement armi1, indicating that imprecise P element excision had generated new armi mutations. Most of these mutations were homozygous viable and female sterile, and the mutant females produced eggs with intermediate to weak defects in D/V patterning, suggesting that they were partial loss-of-function alleles. However, armi72.1 appears to be a strong loss-of-function allele. Homozygous armi1 females produce eggs where 67% showed strong D/V patterning defects as indicated by a complete lack of dorsal appendages. By contrast, 92% of eggs deposited by armi72.1 females completely lacked dorsal appendages, and 35% of these were also collapsed. armi72.1 over the deficiency Df(3L)E1 shows a similar range of defects, indicating that armi72.1 is a strong loss-of-function mutation. However, both armi1 and armi72.1 ovaries produce low levels of transcript, suggesting that neither allele is functionally null (Cook, 2004).

To further characterize armi patterning defects, fluorescent in situ hybridization (FISH) was performed for the three asymmetrically localized mRNAs that specify the anterior, posterior, and dorsal regions of the oocyte. In wild-type stage 9 to 11 oocytes, bcd mRNA accumulates along the anterior cortex and osk mRNA localizes to the posterior pole. In armi mutants, bcd mRNA showed a wild-type anterior distribution. However, osk RNA was either dispersed throughout the ooplasm or concentrated within the oocyte interior in 85% of armi1 egg chambers and 90% (n = 69) of armi72.1 egg chambers. The remaining egg chambers displayed weak posterior localization. osk mRNA consistently showed normal posterior localization in armirev39.2 revertants. During stages 2 through 6, osk mRNA accumulates at the posterior of wild-type oocytes. In armi mutants, osk mRNA was transported to the oocyte during these early stages, but it did not accumulate at the posterior cortex (Cook, 2004).

During stages 2 to 7, grk mRNA and protein also accumulate at the oocyte posterior. During these stages, grk mRNA is below the level of detection by FISH. However, Grk protein was detectable by antibody staining. In armi72.1 oocytes, Grk is dispersed throughout the cytoplasm. During mid-oogenesis, grk mRNA and Grk protein accumulate at the dorsal-anterior corner of the oocyte. In armi mutants, grk mRNA was undetectable by FISH. However, using a colorimetric detection method, grk mRNA was seen to form a weak ring near the anterior cortex, and immunostaining showed that Grk protein was dispersed throughout the oocyte. These observations indicate that the armi gene is required for axial polarization of the oocyte during early and mid-oogenesis (Cook, 2004).

Based on the SDE3 homology, it was postulated that armi might function in RNA silencing during oogenesis. Both osk and bcd mRNAs are produced through most of oogenesis but remain translationally silent until midoogenesis or egg activation, respectively. To determine whether armi mutations lead to premature translation of these mRNAs, egg chambers were immunolabeled for Bcd and Osk proteins. Bcd was undetectable during oogenesis in both wild-type and armi mutants, indicating that translational repression of bcd does not require armi. Wild-type embryos were immunolabeled for Bicoid as a positive control for bcd translation. In wild-type egg chambers, Osk protein is not produced until osk mRNA localizes to the posterior pole in mid-oogenesis. However, in armi72.1 mutants, Osk was produced in early oocytes before transcript localization. Osk started to accumulate as soon as transcript was apparent by FISH. Osk was also prematurely expressed in armi1 and armi72.1 hemizygous oocytes, but was not apparent in early armirev39.2 revertant oocytes (Cook, 2004).

Late stage egg chambers make up most of the ovary, and Osk is highly expressed during these stages in wild-type oocytes. It was therefore not possible to biochemically measure increased Osk protein expression in armi mutants by Western blotting. However, identical labeling and imaging procedures were used with all samples, allowing direct cytological comparison of Osk protein labeling in wild-type and mutant ovaries. Comparable levels of osk mRNA, as judged by FISH and Northern blotting, are present in armi mutants and in wild-type. armi mutants thus do not appear to affect osk transcript stability, but disrupt osk mRNA translational silencing (Cook, 2004).

In animals, miRNAs represses translation without affecting transcript stability, most likely through direct base pairing with target transcripts. miRNAs require components of the RNAi pathway for their maturation and to assemble into the RISC. To determine if components of the RNAi pathway are required for osk mRNA silencing, ovaries mutant for spindle-E (spn-E), aubergine (aub), and maelstrom (mael) were assayed for osk mRNA localization and Osk protein expression. aub and spn-E encode components of the Drosophila RNAi system, and mael is required for localization of a subset of RNAi pathway components in early Drosophila egg chambers. Mutations in all three genes lead to premature Osk protein expression without dramatically affecting the level of osk mRNA. The amount of Osk protein in aubHN2/aubQC42 oocytes was lower than in armi, spn-E, and mael mutants; however, neither aub mutation is a null allele, and the level of Osk expression is consistently above the background staining observed in wild-type controls. Significantly, mutations in aub, spn-E, and mael produce defects in osk mRNA localization during early and mid-oogenesis, as well as defects in D/V patterning that are strikingly similar to those produced by mutations in armi. Therefore, multiple components of the Drosophila RNAi system are required for osk mRNA silencing and embryonic axis specification (Cook, 2004).

To determine if armi is required for axial polarization of the microtubule cytoskeleton, armi egg chambers were immunolabeled with anti-tubulin antibodies. armi mutations do not affect initial organization of microtubules in stage 1 oocytes. During stages 2 to 7, microtubules are significantly more abundant in the oocyte than in the nurse cells and are organized by the posterior cortex of the oocyte. In armi mutants, a posterior MTOC was not apparent and microtubule levels in the oocyte and nurse cells were comparable. Armi is therefore essential for polarization of the oocyte cytoskeleton during early oogenesis. Significantly, spn-E, mael, and aub are also required for microtubule reorganization during early oogenesis (Cook, 2004).

This early polarized microtubule network is required for Grk-dependent differentiation of the posterior follicle cells, which in turn is required for triggering loss of cortical microtubules at the posterior pole during mid-oogenesis. Consistent with the observed defects in microtubule organization and Grk protein localization in early armi oocytes, cortical microtubules persist at the posterior cortex of armi mutants in mid-oogenesis. Loss of these posterior microtubules is thought to be essential for osk mRNA localization. These observations indicate that the RNAi system is required for initial anterior-posterior polarization of the microtubule cytoskeleton, which in turn is required for oskmRNA localization and posterior patterning during midoogenesis (Cook, 2004).

Mutations in armi and the other RNAi components lead to premature expression of Osk protein, raising the possibility that Osk misexpression directly or indirectly triggers the defects in microtubule organization. To test this, microtubule organization was analyzed in ovaries overexpressing Osk protein from a transgene. Microtubule organization in these ovaries was indistinguishable from wild-type controls. Moreover, armi1 osk54 and armi72.1 osk54 double mutant ovaries showed microtubule defects that were cytologically identical to the parental armi mutation. Therefore, the defects in osk mRNA silencing and microtubule polarization are genetically distinct consequences, suggesting that the RNAi pathway silences osk mRNA and additional transcripts encoding cytoskeletal regulators (Cook, 2004).

RISC assembly defects in the Drosophila RNAi mutant armitage

The RNA helicase Armitage is required to repress oskar translation in Drosophila oocytes; armi mutant females are sterile and armi mutations disrupt anteroposterior and dorsoventral patterning. armi has been shown to be required for RNAi. armi mutant male germ cells fail to silence Stellate, a gene regulated endogenously by RNAi, and lysates from armi mutant ovaries are defective for RNAi in vitro. Native gel analysis of protein-siRNA complexes in wild-type and armi mutant ovary lysates suggests that armi mutants support early steps in the RNAi pathway but are defective in the production of active RNA-induced silencing complex (RISC), which mediates target RNA destruction in RNAi. These results suggest that armi is required for RISC maturation (Tomari, 2004).

Silencing of the X-linked Ste gene by the highly homologous Y-linked Su(Ste) locus is an example of endogenous RNAi. In Drosophila testes, symmetrical transcription of Su(Ste) produces dsRNA, which is processed into siRNAs (Gvozdev, 2003). Su(Ste) siRNAs direct the degradation of Ste mRNA. Inappropriate expression of Ste protein in testes is diagnostic of disruption of the RNAi pathway. Both the Argonaute protein, aub, and the putative DEAD-box helicase, spn-E, have been shown to be required for RNAi in Drosophila oocytes. Both mutants fail to silence Ste, as evidenced by the accumulation of Ste protein crystals in the testes of aub and spn-E mutants. No Ste protein is detected in wild-type testes. Strikingly, Ste protein accumulates in testes of two different armi alleles, armi1 and armi72.1. Neither allele is expected to be a true null because armi1 is caused by a P element insertion 5' to the open reading frame, whereas armi72.1, which was created by an imprecise excision of the armi1 P element, corresponds to a deletion of sequences in the 5' untranslated region. Ste silencing is re-established in males homozygous for the revertant chromosome, armirev 39.2 (henceforth, armirev; Cook, 2004), which was generated by excision of the armi1 P element. These data suggest a role for Armi in Drosophila RNAi (Tomari, 2004).

Immunofluorescent detection of Ste protein in testes implicates both armi alleles in endogenous RNAi, but provides only a qualitative measure of allele strength. Since Ste protein in males reduces their fertility , the percent of embryos that hatch when mutant males are mated to wild-type (Oregon R) females provides a more quantitative measure of Ste dysregulation. Hatch rates were measured for the offspring of wild-type, armi1, armi72.1, and spn-E1 homozygous males mated to Oregon R females. For spn-E1 males, 82% of the progeny hatched. Seventy-five percent of the progeny of armi1 males hatched, but only 45% for armi72.1. In contrast, 97% of the offspring of wild-type males hatched. Thus, armi72.1 is a stronger allele than armi1, at least with respect to the requirement for armi in testes (Tomari, 2004).

In contrast to wild-type, lysates prepared from armi72.1 ovaries do not support siRNA-directed target cleavage in vitro: no cleavage product was observed in the armi72.1 lysate after 2 hr. This result was observed for more than ten independently prepared lysates. To determine if the RNAi defect was allele specific, ovaries from armi1 were tested. Phenotypically, this allele is weaker than armi72.1 in its effects on both male fertility and oogenesis. For armi72.1 females, 92% of the eggs lacked dorsal appendages, compared to 67% for armi1 eggs, and some armi1 eggs had wild-type or partially fused dorsal appendages. Consistent with its weaker phenotype, the armi1 allele showed a small amount of RNAi activity in vitro. The two alleles were analyzed together at least four times using independently prepared lysates. In all assays, total protein concentration was adjusted to be equal. Lysate from the revertant allele, armirev, which has wild-type dorsal appendages, showed robust RNAi, demonstrating that the RNAi defect in the mutants is caused by mutation of armi, not an unlinked gene (Tomari, 2004).

The rate of target cleavage was much slower for armi1 than for wild-type . Since the rate of target cleavage in this assay usually reflects the concentration of RISC, it was hypothesized that armi mutants are defective in RISC assembly. To test this hypothesis, a method to measure RISC was developed that requires less lysate than previously described techniques. Double-stranded siRNA was incubated with ovary lysate in a standard RNAi reaction. To detect RISC, a 5' 32P-radiolabled, 2'-O-methyl oligonucleotide complementary to the antisense strand of the siRNA was added. Like target RNAs, 2'-O-methyl oligonucleotides can bind to RISC containing a complementary siRNA, but unlike RNA targets, they cannot be cleaved and binding is essentially irreversible (Hutvagner, 2004). RISC/2'-O-methyl oligonucleotide complexes were then resolved by electrophoresis through an agarose gel (Tomari, 2004).

To validate the method, RISC formation was examined in embryo lysate. Four distinct complexes (C1, C2, C3, C4) were formed when siRNA was added to the reaction. Formation of these complexes required ATP and was disrupted by pre-treatment of the lysate with the alkylating agent N-ethylmaleimide (NEM), but it was refractory to NEM treatment after RISC assembly; these are all properties of RNAi itself. No complex was observed with an siRNA unrelated to the 2'-O-methyl oligonucleotide. The amount of complex formed by different siRNA sequences correlated well with their capacity to mediate cleavage. The four complexes were also detected in wild-type ovary lysate, suggesting that the same RNAi machinery is used during oogenesis and early embryogenesis. The lower amount of RISC formed in ovary compared to embryo lysates can be explained by the lower overall protein concentration of ovary lysates (Tomari, 2004).

The 2'-O-methyl oligonucleotide/native gel assay was used to analyze RISC assembly in armi mutant ovary lysates. armi mutants are deficient in RISC assembly. The extent of the deficiency correlated with allele strength: less C3/C4 complex formed in lysate from the strong armi72.1 allele than from armi1. Compared to the phenotypically wild-type armirev, >10-fold less RISC was produced in armi72.1 (Tomari, 2004).

The defect in RISC assembly in armi mutants is similar to that observed in lysates from aubHN2 ovaries. aub mutants do not support RNAi following egg activation and fail to silence the Ste locus in testes, and lysates from aubHN2 ovaries do not support RNAi in vitro. Aub is one of five Drosophila Argonaute proteins, core constituents of RISC. It is therefore not surprising that Aub is required for RISC assembly. Since RISC assembly in vitro was not detectable in aubHN2 lysates, the data suggest that Aub is the primary Argonaute protein recruited to exogenous siRNA in Drosophila ovaries. In contrast, ovaries from nanosBN, a maternal effect mutant not implicated in RNAi, are fully competent for both RISC assembly and siRNA-directed target RNA cleavage (Tomari, 2004).

Two intermediates have been identified in RISC assembly. Complex B forms rapidly upon incubation of siRNA in lysate, in the absence of ATP. The siRNA is then transferred to complex A, which contains the R2D2/Dcr-2 heterodimer. The siRNA is double stranded in both B and A. RISC is formed from complex A by a process that requires both siRNA unwinding and ATP. armi mutants are defective for the conversion of complex A to RISC. RISC does not form in ovary lysates from armi or aub mutants. However, both complexes B and A are readily detected in armi and aub mutants. Thus, armi and aub mutants are impaired in a step in RISC assembly after binding of the siRNA to the Dcr-2/R2D2 heterodimer (Tomari, 2004).

Armi might act after the formation of complex A to unwind siRNA duplexes prior to their assembly into RISC. To test this hypothesis, whether single-stranded siRNA circumvents the requirement for armi was tested. In vitro and in vivo, single-stranded siRNA triggers RNAi, albeit inefficiently. armi ovary lysates failed to support RNAi when the reactions were programmed with 5'-phosphorylated, single-stranded siRNA. The defect with single-stranded siRNA correlates with allele strength: some activity was seen in lysates from the weak allele, armi1, but none for the strong allele, armi72.1. The requirement for a putative ATPase -- Armi -- in RNAi triggered by single-stranded siRNA suggested the presence of an additional ATP-dependent step in the RISC assembly, after siRNA unwinding (Tomari, 2004).

To test if loading of single-stranded siRNA into RISC requires ATP, 5'-phosphorylated, single-stranded siRNA was added to embryo lysates depleted of ATP. After incubation for 2 hr, no cleavage product was detected, suggesting that there is at least one ATP-dependent step downstream of siRNA unwinding. The stability of single-stranded siRNA was not reduced by ATP depletion. In fact, single-stranded siRNA was slightly more stable in the absence of ATP. Thus differential stability cannot account for the requirement for ATP in RNAi triggered by single-stranded siRNA. In the RNAi pathway, there are at least three steps after siRNA unwinding: RISC assembly, target recognition, and target cleavage. To assess if either target recognition or cleavage was ATP dependent, single-stranded siRNA was incubated in a standard RNAi reaction with ATP to assemble RISC. Next, NEM was added to inactivate the ATP-regenerating enzyme, creatine kinase, and to block further RISC assembly. NEM was quenched with dithiothreitol (DTT), and hexokinase and glucose added to deplete ATP. Finally, mRNA target was added and the reaction incubated for 2 hr. Using this protocol, high ATP levels were maintained during RISC assembly, but less than 100 nM ATP was present during the encounter of RISC with the target RNA. Target recognition and cleavage did not require ATP when RISC was programmed with either double- or single-stranded siRNA, provided that ATP was supplied during RISC assembly (Tomari, 2004).

Four distinct RISC-like complexes common to embryo and ovary lysates have been detected. In ovaries, formation of these complexes is reduced >10-fold in armi mutants and is undetectable in aub mutants, which are RNAi defective. The requirement for Aub, an Argonaute protein, suggests that the complexes correspond to distinct RISC isoforms built on a common core of Aub and siRNA. These isoforms may play distinct regulatory roles (e.g., translational repression versus cleavage). Alternatively, the smallest, most abundant complexes may contain only the most stably associated protein constituents, whereas the larger, less abundant complexes may correspond to 'holo-RISC' that retain more weakly bound proteins. Clearly, a major challenge for the future is to define the protein constituents of each complex, their functional capacity, and their biological role. The development of a native gel assay that resolves distinct RISC complexes represents a step toward that goal (Tomari, 2004).

This study has identified two intermediates in RISC assembly. Complex B forms rapidly upon incubation of siRNA in lysate, in the absence of ATP. The siRNA is then transferred to complex A, which contains the previously identified R2D2/Dcr-2 heterodimer. The siRNA is double stranded in both B and A. RISC is formed from complex A by a process that requires both siRNA unwinding and ATP. Both aub and armi are required genetically for the production of RISC from complex A. The involvement of Armi, a putative RNA helicase protein, in the production of RISC from complex A and the finding that incorporation of single-stranded siRNA into RISC requires ATP suggest that Armi functions to incorporate single-stranded siRNA into RISC. However, the data cannot distinguish between direct and indirect roles for Armi in RISC assembly (Tomari, 2004). The Arabidopsis homolog of Armi, SDE3, together with the RNA-dependent RNA polymerase (RdRP) SDE1/SGS2, is required for PTGS triggered by transgenes that express single-stranded sense mRNA, but not silencing triggered by some RNA viruses (Dalmay, 2001). SDE3 has been proposed to facilitate the conversion of dsRNA into siRNA or the conversion of mRNA into complementary RNA by SDE1/SGS2 (Dalmay, 2001; Jorgensen, 2003). Recent studies show that SDE3 is not required for the production of siRNAs derived directly from a long dsRNA hairpin (Himber, 2003). Instead, SDE3 seems to play a role in the production of siRNAs generated by an RdRP-dependent amplification mechanism. The data from the Tamari study (2004) are not consistent with either of these functions for Armi. (1) Drosophila genomic, biochemical, and genetic data exclude a role for an RdRP in RNAi. (2) Armi is required for RISC assembly in Drosophila ovary lysates when RISC is programmed with siRNA, suggesting a role for Armi downstream of the conversion of dsRNA into siRNA, but upstream of target recognition by RISC. The apparently divergent functions of SDE3 and Armi could be reconciled if RISC is required for RdRP-mediated amplification of silencing. Alternatively, SDE3 and Armi may not have homologous functions (Tomari, 2004).

armi mRNA is abundant in oocytes and syncitial blastoderm embryos, but a low level can be detected throughout development, including in somatic tissues (Cook, 2004). While the requirement for armi in spermatogenesis and oogenesis makes Armi a good candidate for a component of the RNAi machinery in germ cells and early embryos, somatic functions for Armi are also possible. In this respect, armi is reminiscent of the maternally expressed Argonaute protein, piwi, which is also required during oogenesis. Although piwi mutants display no obvious somatic phenotype, Piwi is required in the soma both for posttranscriptional transgene silencing and for some types of transcriptional silencing. Whether Armi is likewise required for somatic transgene silencing remains to be tested (Tomari, 2004).

Drosophila rasiRNA pathway mutations disrupt embryonic axis specification through activation of an ATR/Chk2 DNA damage response

Small repeat-associated siRNAs (rasiRNAs) mediate silencing of retrotransposons and the Stellate locus. Mutations in the Drosophila rasiRNA pathway genes armitage and aubergine disrupt embryonic axis specification, triggering defects in microtubule polarization as well as asymmetric localization of mRNA and protein determinants in the developing oocyte. Mutations in the ATR/Chk2 DNA damage signal transduction pathway dramatically suppress these axis specification defects, but do not restore retrotransposon or Stellate silencing. Furthermore, rasiRNA pathway mutations lead to germline-specific accumulation of γ-H2Av foci characteristic of DNA damage. It is concluded that rasiRNA-based gene silencing is not required for axis specification, and that the critical developmental function for this pathway is to suppress DNA damage signaling in the germline (Klattenhoff, 2007).

Mutations in the Drosophila armi, aub, and spn-E genes disrupt oocyte microtubule organization and asymmetric localization of mRNAs and proteins that specify the posterior apole and dorsal-ventral axis of the oocyte and embryo. Mutations in these genes block homology-dependent RNA cleavage and RISC assembly in ovary lysates, RNAi-based gene silencing during early embryogenesis, rasiRNA production, and retrotransposon and Stellate silencing. Mutations in dcr-2 and ago-2 genes, by contrast, block siRNA function, but they do not disrupt the rasiRNA pathway or embryonic axis specification. The rasiRNA pathway thus appears to be required for embryonic axis specification. However, the function of rasiRNAs in the axis specification pathway has not been previously established (Klattenhoff, 2007).

Cytoskeletal polarization, morphogen localization, and eggshell patterning defects associated with armi and aub are efficiently suppressed by mnk and mei-41, which respectively encode Chk2 and ATR kinase components of the DNA damage signaling pathway. In addition, armi and aub mutants accumulate γ-H2Av foci characteristic of DNA DSBs and trigger Chk2-dependent phosphorylation of Vas, an RNA helicase required for posterior and dorsal-ventral specification. Mutations in spn-E also disrupt the rasiRNA pathway, trigger axis specification defects, and lead to germline-specific accumulation of γ-H2Av foci. Significantly, the mnk and mei-41 mutations do not suppress Stellate or HeT-A overexpression, indicating that axis specification does not directly require rasiRNA-dependent gene silencing. Based on these findings, it is concluded that the rasiRNA pathway suppresses DNA damage signaling in the female germline, and that mutations in this pathway disrupt axis specification by activating an ATR/Chk2 kinase pathway that blocks microtubule polarization and morphogen localization in the oocyte (Klattenhoff, 2007).

The cause of DNA damage signaling in armi, aub, and spn-E mutants remains to be established. In wild-type ovaries, γ-H2Av foci begin to accumulate in region 2 of the germarium, when the Spo11 nuclease (encoded by the mei-W68 gene) initiates meiotic breaks. The axis specification defects associated with DNA DSB repair mutations are efficiently suppressed by mei-W68 mutations, indicating that meiotic breaks are the source of DNA damage in these mutants. The axis specification defects and γ-H2Av focus formation associated with armi, by contrast, are not suppressed by mei-W68. mei-W68 double mutants with aub or spn-E have not been analyzed, but this observation strongly suggests that meiotic DSBs are not the source of DNA damage in rasiRNA pathway mutations. Retrotransposon silencing is disrupted in armi, aub, and spn-E mutants, and transcription of LINE retrotransposons in mammalian cells leads to DNA damage and DNA damage signaling. Loss of retrotransposon silencing could therefore directly induce the DSBs in rasiRNA pathway mutants. However, DNA damage can also lead to loss of retrotransposon silencing. Mutations in the rasiRNA pathway could therefore disrupt DNA repair and thus induce DNA damage, which, in turn, induces loss of retrotransposon silencing. Finally, the HeT-A retrotransposon is associated with telomeres, and overexpression of this element could reflect a loss of telomere protection and could damage signaling by chromosome ends in the rasiRNA pathway mutants. The available data do not distinguish between these alternatives (Klattenhoff, 2007).

In mouse, the piwi-related Argonauts Miwi and Mili bind piRNAs, 30 nt RNAs derived primarily from a single strand that appear to be related to rasiRNAs. Mutations in these genes disrupt spermatogenesis and lead to germline apoptosis, which can be induced by DNA damage signaling. Mammalian piRNAs and Drosophila rasiRNAs may therefore serve similar functions in suppressing a germline-specific DNA damage response (Klattenhoff, 2007).

Distinct functions for the Drosophila piRNA pathway in genome maintenance and telomere protection

Transposons and other selfish DNA elements can be found in all phyla, and mobilization of these elements can compromise genome integrity. The piRNA (PIWI-interacting RNA) pathway silences transposons in the germline, but it is unclear if this pathway has additional functions during development. This study shows that mutations in the Drosophila piRNA pathway genes, armi, aub, ago3, and rhi, lead to extensive fragmentation of the zygotic genome during the cleavage stage of embryonic divisions. Additionally, aub and armi show defects in telomere resolution during meiosis and the cleavage divisions; and mutations in ligase-IV, which disrupt non-homologous end joining, suppress these fusions. By contrast, lig-IV mutations enhance chromosome fragmentation. Chromatin immunoprecipitation studies show that aub and armi mutations disrupt telomere binding of HOAP, which is a component of the telomere protection complex, and reduce expression of a subpopulation of 19- to 22-nt telomere-specific piRNAs. Mutations in rhi and ago3, by contrast, do not block HOAP binding or production of these piRNAs. These findings uncover genetically separable functions for the Drosophila piRNA pathway. The aub, armi, rhi, and ago3 genes silence transposons and maintain chromosome integrity during cleavage-stage embryonic divisions. However, the aub and armi genes have an additional function in assembly of the telomere protection complex (Khurana, 2010).

Drosophila piRNAs have been implicated in transposon silencing and maintenance of genome integrity during female germline development. However, piRNA pathway mutations lead to complex developmental phenotypes, and piRNAs have been implicated in control of gene expression. Furthermore, the majority of piRNAs in other systems, including mouse testes, are not derived from repeated elements. The full extent of piRNA functions thus remains to be explored (Khurana, 2010).

Mutations in the majority of Drosophila piRNA pathway genes disrupt asymmetric localization of RNAs along the axes of the oocyte, and lead to maternal effect embryonic lethality. The axis specification defects linked to several of piRNA pathway mutations are dramatically suppressed by a null mutation in mnk, which encodes a Checkpoint kinase 2 (Chk2) homolog required for DNA damage signaling, indicating that the loss of asymmetric RNA localization is downstream of DNA damage. Oocyte patterning defects generally lead to embryonic lethality, but the mnk allele that suppresses the axis specification defects associated with piRNA mutations does not suppress embryonic lethality. piRNAs thus have an essential function during embryogenesis that is independent of Chk2 activation and DNA damage signaling. To gain insight into potential new functions for the piRNA pathway, the embryonic lethality associated with four piRNA pathway mutations was characterized. These studies reveal a novel function for a subset of piRNA genes in assembly of the telomere protection complex, and suggest that this process is directed by a subpopulation of 19-22 nt piRNAs (Khurana, 2010). The armi and aub genes encode a putative RNA helicase and a piRNA binding PIWI Argonaute protein, and recent studies suggest that they have distinct functions in piRNA biogenesis. Mutations in aub dramatically reduce piRNA species that overlap by 10 nt, which is characteristic of ping-pong amplification, while armi mutations reduce total piRNA production but enhance the ping-pong signature. Mutations in aub and armi lead to maternal-effect embryonic lethality, however, suggesting that these genes share an essential function. To gain insight into the lethality associated with these mutations, DNA break accumulation during oogenesis was analyzed. Germline-specific DNA breaks normally form during early oogenesis, as meiosis is initiated. In several piRNA mutants, however, DNA breaks persist, which could compromise the female pronucleus and thus lead to genetic instability in the early zygote. DNA breaks trigger phosphorylation of histone H2Av, producing γ-H2Av foci near the break sites. In wild-type ovaries, γ-H2Av foci begin to accumulate in region 2 of the germarium, as meiotic breaks are formed. These foci are significantly reduced in stage 2 egg chambers, which have completed meiotic repair and budded from the germarium. Later in oogenesis, γ-H2Av foci accumulate in the nurse cell nuclei, which undergo endoreduplication. However, these foci remain undetectable in the oocyte. In ovaries mutant for aub or armi, γ-H2Av foci appear in germarium region 2, but persist in nurse cells and the oocyte through stage 4. By stage 5, however, γ-H2Av foci are undetectable in 50% of armi and aub mutant oocytes, and are significantly reduced in the remaining oocytes. Both armi and aub mutations thus increase DNA damage during early oogenesis, but most of the damage in the oocyte appears to be repaired as oogenesis proceeds (Khurana, 2010).

As wild type oocytes mature and initiate meiotic spindle assembly, the major chromosomes form a single mass at the spindle equator and the non-exchange 4th chromosomes move toward the poles. In OregonR, distinct 4th chromosomes were observed in 79% of stage 13 oocytes. In stage 13 aub and armi mutants, by contrast, distinct 4th chromosomes were observed in only 11% and 18% of stage 13 oocytes, respectively. However, a single primary mass of chromatin was always observed. These observations are consistent with γ-H2Av data suggesting that DNA breaks formed during early oogenesis are often repaired as the oocyte matures. In addition, both aub and armi mutations appear to inhibit separation of the small 4th chromosomes, although it is also possible that this small chromosome is fragmented and thus difficult to detect cytologically (Khurana, 2010).

Drosophila oocytes are activated as they pass through the oviduct, which triggers completion of the meiotic divisions. The first meiotic division is completed in the oviduct, but meiosis II can be observed in freshly laid eggs and is characterized by four well-separated meiotic products on tandem spindles. In aub and armi mutant embryos, the meiotic chromatin was either stretched across the paired meiotic spindles, or fragmented and spread over both spindles. No wild type meiotic figures were observed. Breaks thus appear to persist in some stage 14 oocytes, although this does not disrupt the karyosome organization during earlier stages. However, other oocytes appear to have intact chromosomes that fail to resolve during the meiotic divisions (Khurana, 2010).

Chromatin fragmentation could result from replication of broken chromosomes inherited from the female, or from post-fertilization fragmentation of the zygotic genome. To directly assay zygotic genome integrity, mutant females were mated to wild type males and dual-label FISH was used to monitor physically separate regions of the Y chromosome. In male embryos derived from wild type females, the two Y chromosome probes always co-segregated through anaphase and telophase. Mutant embryos showing chromatin fragmentation, by contrast, contained chromatin clusters that did not label for either Y chromosome probe, or that labeled for only one of the two probes. In mutant embryos that proceeded through cleavage stage mitotic cycles, the majority of segregating chromatids retained both Y chromosome markers, indicating that chromosome continuity had been maintained. Chromatids with only one of two markers were observed, however, indicating that breaks had separated regions on a Y chromosome arm from the centromere. The axial patterning defects associated with piRNA mutations are suppressed by mutations in mnk, but mnk did not suppress either the chromatin fragmentation or segregation defects linked to aub and armi. Mutations in aub and armi thus destabilize the genome of the zygote and disrupt chromosome resolution during the cleavage divisions through processes that are independent of DNA damage signaling (Khurana, 2010).

Mutations in the armi and aub genes disrupt piRNA production and transposon silencing, but have also been reported to inhibit homology dependent target cleavage by siRNAs. In addition, null mutations in argonaute2 (ago2), which block siRNA based silencing, have been reported to disrupt mitosis during the syncytial blastoderm stage. These observations raise the possibility that chromatin fragmentation and fusion in aub and armi mutants result from defects in the siRNA pathway. Therefore, cleavage was analyzed in embryos from females homozygous for null mutations in ago2 and dcr2, which block siRNA production and silencing. Consistent with previous studies, it was found that embryos from ago2 and dcr2 mutant females are viable. However, neither chromosome fragmentation nor a statistically significant increase in anaphase bridge formation was found relative to wild type controls. The loquacious (loqs) gene encodes a Dicer-1 binding protein required for miRNA production, and it was found that embryos from loqs mutant females also proceed through normal cleavage stage divisions. Chromosome segregation and maintenance of zygotic genome integrity during early embryogenesis thus appear to be independent of the siRNA and miRNA pathways, but require at least two components of the piRNA pathway (Khurana, 2010).

In S. pombe, mutations in ago1, dcr1 and rdp1 disrupt kinetochore assembly and thus lead to lagging mitotic chromosomes due to defects in centromere movement to the spindle poles. To determine if Drosophila piRNA mutations disrupt kinetochore assembly, dual label FISH was performed for centromeric dodeca-satellite sequences and the telomere-specific transposon HeT-A. In aub and armi mutants, centromeric sequences segregated to the spindle poles in essentially every anaphase figure, but telomere specific sequences were consistently present at the chromatin bridges. These observations indicate that armi and aub are not required for kinetochore assembly, but are needed for telomere resolution (Khurana, 2010).

Telomeres are protected from recognition as DNA double strand breaks by the telomere-protection complex (TPC), and defects in telomere protection thus lead to covalent ligation of chromosome ends by the non-homologous end-joining (NHEJ) pathway. DNA Ligase IV is required for NHEJ, and ligase IV mutations suppress fusions that result from covalent joining of unprotected chromosome ends. To determine if chromosome fusions in aub and armi are due to NHEJ, ligIV;aub and ligIV;armi double mutant females were generated and chromosome segregation was analyzed in the resulting embryos. In aub single mutant embryos, 50% of anaphase figures show bridges, but anaphase bridges are present in only 15% of ligIV;aub double mutants. By contrast, the fraction of embryos showing chromosome fragmentation increases in ligIV;aub double mutants. Chromosome fragmentation also increased in ligIV;armi mutant embryos, and as a result morphologically normal anaphase figures could not be observed. These findings strongly suggest that lagging chromosomes result from covalent ligation of chromosome ends by the NHEJ pathway, while chromatin fragmentation results from DNA breaks that are repaired by NHEJ. Mutations in armi and aub lead to significant over-expression of transposable elements, including DNA elements that are mobilized by a 'cut and paste' mechanism that directly produces double strand breaks. In addition, NHEJ pathway has been implicated in repair of gapped retroviral integration intermediates. Chromosome fragmentation may therefore result from transposon over-expression and mobilization, which induces breaks that overwhelm the NHEJ pathway. Telomere fusions, by contrast, appear to result from defects in telomere protection, which lead to chromosome end recognition by the NHEJ pathway (Khurana, 2010).

The Drosophila TPC includes HOAP and Modigliani (Moi), which may function only at chromosome ends, and HP1a and the MRN complex, which have additional roles in heterochromatic silencing and DNA repair. To directly assay for TPC recruitment, chromatin immunoprecipitation (ChIP) was used to measure HP1a and HOAP binding to the telomere specific transposon HeT-A. In wild type ovaries, HOAP and HP1a bind to multiple regions of HeT-A. In armi and aub mutants, by contrast, HOAP and HP1a binding to the Het-A 5'-UTR and ORF are significantly reduced. The 5' end of Het-A is oriented toward the chromosome end, and is therefore likely to lie at the telomere. Ovarian tissue consists of germ cells with a surrounding layer of somatic cells, which complicates interpretation of these biochemical studies. However, ChIP on 0-3 hour old embryos from aub and mnk,aub mutant females revealed significant reduction in HOAP binding at the HeT-A 5'-UTR. The aub and armi genes thus appear to be required for TPC recruitment, consistent with ligation of chromosome ends in mutant embryos (Khurana, 2010).

To determine if other piRNA pathway mutations disrupt telomere protection, the cleavage stage embryonic divisions was analyzed in ago3 and rhi mutants. The ago3 locus encodes a PIWI clade protein that primarily binds sense strand piRNAs, and rhi encodes a rapidly evolving HP1 homologue required for production of precursor RNAs from a subset of piRNA clusters. Essentially all of the rhi and ago3 mutant embryos showed chromatin fragmentation, as observed in the majority of aub and armi mutants. Therefore TPC assembly was analyzed in ovarian chromatin using ChIP for HOAP and HP1a. Surprisingly, neither ago3 nor rhi mutations disrupt HOAP or HP1a binding to Het-A, and rhi mutants show greater than wild type levels of HOAP binding to Het-A. By contrast, these rhi alleles reduce total piRNA production by 10 fold. The ago3 mutations appear to be null, and the rhi mutations are strong hypomorphc alleles. Assembly of the TPC in the ago3 and rhi mutants is therefore unlikely to be mediated by residual protein. Instead, these findings strongly suggest that aub and armi have a function in telomere protection that is not shared by ago3 or rhi (Khurana, 2010).

In Drosophila, chromosome breaks can be converted to stable telomeres, called terminal deletions, which accumulate additional copies of the telomeric elements HeT-A and TART. When terminal deletions are passaged in animals heterozygous for aub or the piRNA pathway gene spnE, the number of terminal TART repeats increase. The defects in TPC assembly in aub and armi could therefore be triggered by increased HeT-A and TART copy number, which could titrate TPC components. Therefore telomeric transposon copy number was assayed in aub and armi mutants, which show defects in TPC assembly, and in rhi and ago3 mutants, which do not. Telomeric transposon copy number and mitotic chromosome segregation was also analyzed in a wild-type variant, Gaiano, that has been reported to carry additional HeT-A repeats. Consistent with previous reports, it was found that Gaiano has 10 to 15 fold more HeT-A copies than OregonR controls. Despite the increase in telomere length, this stock is viable and fertile, and no telomere fusions or lagging chromosomes were observed during the cleavage stage embryonic divisions. In addition, it was found that aub mutants that show defects in TPC assembly do not accumulate additional copies of HeT-A or TART, while rhi and ago3 mutants that are wild type for TPC binding show an increase in telomere-specific transposon copy number. Assembly of the TPC is therefore independent of telomere specific transposon copy number (Khurana, 2010).

piRNAs are proposed to direct PIWI clade proteins to targets through sequence specific interactions. The current observations raised the possibility that armi and aub promote production of piRNAs that direct the telomere protection complex to transposons that make up chromosome ends. Published small RNA deep sequencing data was analyzed for species derived from a fourth chromosome cluster, defined by a high density of uniquely mapping piRNAs, containing multiple repeats of the telomeric transposons. This bioinformatic analysis showed that 70-80% of telomere specific piRNAs match this cluster. Length histograms for small RNAs from wt, rhi, ago3, aub and armi mutant ovaries map to this cluster. Significantly, aub and armi mutations lead to a preferential loss of shorter piRNAs mapping to the minus genomic strand. Loss of these shorter RNAs highlights the peak at 21 nt, which is retained in all of the mutants and likely represent endogenous siRNAs. The telomeric elements (HeT-A and TART) are almost exclusively on the minus genomic strand in this cluster, and the RNAs that are lost in aub and armi thus correspond to the sense strand of the target elements. Ovaries mutant for ago3 and rhi, by contrast, retain these shorter sense strand RNAs (Khurana, 2010).

The relative abundance of typical 23-29nt long piRNAs and the shorter 19-22nt species were quantified, excluding the 21nt endo-siRNA peak. All four mutations significantly reduce 23 to 29 nt piRNAs, although rhi mutants retain approximately 50% of wild type minus strand species. Loss of these piRNAs is consistent with over-expression of transposons matching this cluster in all four mutants. By contrast, the shorter minus strand RNAs are reduced by 3 to 10 fold in armi and aub, but are expressed at 80% to 95% of wild type levels in ago3 and rhi. In addition, short piRNA species from the telomeric cluster co-immunoprecipitate with Piwi protein, which localizes to the nucleus and is a likely effector of chromatin functions for the piRNA pathway. Binding of this subpopulation of piRNAs by Piwi is retained in ago3 mutants, which assemble the TPC, but significantly reduced in armi mutants, which block assembly of the TPC (Khurana, 2010).

Taken together, these observations suggest that the piRNA pathway has two genetically distinct functions during oogenesis and early embryogenesis. The pathway prevents DNA damage during oogenesis and maintains the integrity of the zygotic genome during the embryonic cleavage divisions, which likely reflects the established role for piRNAs in transposon silencing. This function requires aub, armi, rhi and ago3, which are also required for wild type piRNA production. In addition, these studies reveal a novel function for the piRNA genes aub and armi in telomere protection, whch may be mediated by a novel class of short RNAs that bind to Piwi. Consistent with this hypothesis, it has been reported that germline clones of piwi null alleles do not significantly disrupt oogenesis, but lead to maternal effect embryonic lethality and severe chromosome segregation defects during the cleavage division. A subpopulation of Piwi-bound piRNAs may therefore direct assembly of the TPC (Khurana, 2010).

Roles for the Yb body components Armitage and Yb in primary piRNA biogenesis in Drosophila

PIWI-interacting RNAs (piRNAs) protect genome integrity from transposons. In Drosophila ovarian somas, primary piRNAs are produced and loaded onto Piwi. This study describes roles for the cytoplasmic Yb body components Armitage and Yb in somatic primary piRNA biogenesis. Armitage binds to Piwi and is required for localizing Piwi into Yb bodies. Without Armitage or Yb, Piwi is freed from the piRNAs and does not enter the nucleus. Thus, piRNA loading is required for Piwi nuclear entry. It is proposed that a functional Piwi-piRNA complex is formed and inspected in Yb bodies before its nuclear entry to exert transposon silencing (Saito, 2010).

In Drosophila, three sets of endogenous small RNAs have been identified so far: microRNAs (miRNAs), endogenous siRNAs (endo-siRNAs/esiRNAs), and PIWI-interacting RNAs (piRNAs). Of these, piRNAs are considered unique because of their germline-specific expression and specific interaction with germline-specific Argonaute proteins, PIWI proteins. The identification of the piRNAs associated with three PIWI proteins (Aubergine [Aub], Argonaute 3 [AGO3], and Piwi) has revealed distinct features of piRNAs associated with each PIWI and has led to two models for piRNA biogenesis: the primary processing pathway and the amplification loop pathway. In the amplification loop model, the Slicer (endonuclease) activity of Aub and AGO3 determines the formation of the 5' end of piRNAs. Zucchini (Zuc), a putative cytoplasmic nuclease, is involved in the primary processing pathway; however, its precise molecular function remains unclear. Furthermore, the factors other than zuc required for primary piRNA biogenesis are unknown (Saito, 2010).

The ovarian somatic cell (OSC) line consists of ovarian somas only. The expression of Aub and AGO3 is not detectable in OSCs because both proteins are germ cell-specific. This implies that the amplification loop does not operate in OSCs. However, OSCs express piRNAs and are loaded onto Piwi, indicating that the piRNAs in OSCs are generated specifically through the primary processing pathway. Thus, OSCs are an ideal tool to elucidate the molecular mechanisms of primary piRNA processing and Piwi function. Loss of zuc function drastically reduced the level of primary piRNAs in the ovaries. This was recapitulated in OSCs: Zuc depletion by RNAi caused a severe reduction in the piRNA level in OSCs. This result prompted a screen for other factors necessary for primary piRNA production using RNAi in OSCs (Saito, 2010).

To identify the genes required for somatic primary piRNA biogenesis, RNAi-based screening was performed in OSCs. The genes screened included armitage (armi), spindle-E (spn-E), and maelstrome (mael), all of which are implicated in piRNA biogenesis. However, their roles in somatic primary piRNA production remain unknown. Depletion of Armi reduced the piRNA levels to an extent very similar to that of Piwi and Zuc depletion, indicating that Armi is necessary for primary piRNA biogenesis in OSCs. Depletion of Mael and Spn-E showed little or no effect on piRNA accumulation in OSCs. Mutations in both genes have been shown to significantly reduce the piRNA levels in ovaries. Thus, spn-E and mael are factors functioning in the amplification loop. Depletion of Dicer1 and Dicer2 had little or no effect on the piRNA levels, confirming that neither protein is necessary for piRNA production (Saito, 2010).

Armi is the Drosophila ortholog of Arabidopsis Silencing-Defective 3 (SDE3) and mammalian Moloney leukemia virus 10 (MOV10). These orthologs contain a conserved ATP-dependent RNA helicase domain at their C termini and have been implicated in small RNA-mediated gene silencing. However, their precise functions remain unknown. To gain further insight into the function of Armi in somatic primary piRNA processing, a monoclonal antibody was produced against Armi. Western blotting showed a discrete band in both ovary and cultured Schneider2 (S2) cell lysates, indicating that Armi expression is not germline-specific. The ~150-kDa protein immunopurified from S2 cells with the anti-Armi antibody was confirmed to be Armi by mass spectrometry (Saito, 2010).

Immunostaining of OSCs and ovaries with the anti-Armi antibody confirmed an earlier observation that Armi is a cytoplasmic protein. The Armi signals were detected in both somatic and germ cells of ovaries. The somatic signal was considered a background signal because it did not disappear even in armi homozygous mutant egg chambers. In the present study, the cytoplasmic signal in OSCs mostly disappeared when Armi was depleted by RNAi. Thus, it is concluded that Armi is expressed in both somatic and germ cells in ovaries (Saito, 2010).

The subcellular localization of Armi in the armi trans-heterozygous mutants appeared very similar to that in the homozygous mutants. In addition, Western blotting revealed a band corresponding to Armi in the armi ovaries. By what mechanisms Armi is expressed in the mutant somas remains unclear. The simplest explanation is that the armi gene uses two distinct genomic elements as promoters in ovarian somas. In fact, armi homozygous mutants weakly express a shorter armi transcript than that expressed in the wild-type strain (Saito, 2010).

The Armi signal in germ cells was rather weak, and only a small proportion of Armi accumulated at, or near, the nuage, an electron-dense structure associated with nurse cell nuclei. Thus, Armi might not be a component of the nuage per se. This correlates well with the fact that armi mutations barely affected the ability of the ovaries to amplify endogenous piRNAs. In ovarian somas, Armi accumulated strongly at discrete cytoplasmic foci. Each somatic cell contained one or several foci. Interestingly, the Armi-positive foci were often located near the nucleus in both ovaries and OSCs (Saito, 2010)

Piwi is required for the silencing of transposons in gonads. In fact, Piwi depletion in OSCs caused derepression of transposons, as with Armi, Yb, and Zuc depletion. Under conditions where endogenous Piwi was depleted, expression of myc-Piwi-r, which was designed to be RNAi-insensitive, rescued transposon silencing. However, myc-Piwi-δN, which lacks 72 amino acids at the N terminus of Piwi and thus does not localize to the nucleus, did not rescue transposon silencing, although it does associate with piRNAs to the same extent as does the wild-type Piwi. myc-Piwi-δN13, which lacks 13 amino acids at the N terminus, behaved similarly. On the other hand, myc-Piwi-DDAA-r, a Slicer mutant of Piwi, could bind to mature piRNAs in OSCs, as does the wild-type Piwi, and rescued transposon silencing. These results might suggest that Piwi must be localized in the nucleus to silence the transposable elements, and that Piwi Slicer activity is unnecessary for its function. It is assumed that this system has evolved to prevent nascent Piwi, not loaded with piRNAs, from being imported into the nucleus. In other words, only the functional Piwi-piRNA complex (piRISC) formed at Yb bodies could be transported to the nucleus. At present, the mechanisms of this control system remain unclear. In the nongonadal somatic S2 cell line, where the expression of piRNAs is undetectable, transfected Piwi is localized to the nucleus, indicating that 'empty' Piwi can be transported to the nucleus. It seems that the machineries necessary for the nuclear transport of Piwi might recognize different features of Piwi in different cell types (Saito, 2010).

How is piRNA-free Piwi restrained in the cytoplasm in OSCs? One possibility is that some unknown protein binds the N-terminal end of Piwi, where its NLS (nuclear localization signal) resides, and interferes with the nuclear import machinery's ability to recognize Piwi as a cargo. The nuclear localization inhibitory factors may be retained on Piwi until a functional Piwi-piRNA complex is formed at Yb bodies. Once the complex is formed, a conformational change in Piwi would be induced, which would release the regulatory factors and reveal the Piwi NLS for recognition by the nuclear import machinery. It would be very interesting to determine the proteins that are associated with Piwi in OSCs under conditions of Armi or Zuc depletion, thus identifying the protein factors that restrain Piwi in the cytoplasm until it is loaded with mature piRNAs at Yb bodies (Saito, 2010).

A role for Drosophila Cyclin J in oogenesis revealed by genetic interactions with the piRNA pathway

Cyclin J (CycJ) is a poorly characterized member of the Cyclin superfamily of cyclin-dependent kinase regulators, many of which regulate the cell cycle or transcription. Although CycJ is conserved in metazoans its cellular function has not been identified and no mutant defects have been described. In Drosophila, CycJ transcript is present primarily in ovaries and very early embryos, suggesting a role in one or both of these tissues. The CycJ gene (CycJ) lies immediately downstream of armitage (armi), a gene involved in the Piwi-associated RNA (piRNA) pathways that are required for silencing transposons in the germline and adjacent somatic cells. Mutations in armi result in oogenesis defects but a role for CycJ in oogenesis has not been defined. This study assessed oogenesis in CycJ mutants in the presence or absence of mutations in armi or other piRNA pathway genes. CycJ null ovaries appeared normal, indicating that CycJ is not essential for oogenesis under normal conditions. In contrast, armi null ovaries produced only two egg chambers per ovariole and the eggs had severe axis specification defects, as observed previously for armi and other piRNA pathway mutants. Surprisingly, the CycJ armi double mutant failed to produce any mature eggs. The double null ovaries generally had only one egg chamber per ovariole and the egg chambers frequently contained an overabundance of differentiated germline cells. Production of these compound egg chambers could be suppressed with CycJ transgenes but not with mutations in the checkpoint gene mnk, which suppress oogenesis defects in armi mutants. The CycJ null showed similar genetic interactions with the germline and somatic piRNA pathway gene piwi, and to a lesser extent with aubergine (aub), a member of the germline-specific piRNA pathway. The strong genetic interactions between CycJ and piRNA pathway genes reveal a role for CycJ in early oogenesis.These results suggest that CycJ is required to regulate egg chamber production or maturation when piRNA pathways are compromised (Atikukke, 2014).


Post-transcriptional gene silencing (PTGS) provides protection in plants against virus infection and can suppress expression of transgenes. Arabidopsis plants carrying mutations at the SDE3 locus are defective in PTGS mediated by a green fluorescent protein transgene. However, PTGS mediated by tobacco rattle virus (TRV) was not affected by sde3. From these results it is concluded that SDE3, like the previously described RNA polymerase encoded by SDE1, acts at a stage in the mechanism that is circumvented when PTGS is mediated by TRV. The product of SDE3 is similar to RNA helicase-like proteins including GB110 in mouse and other proteins in Drosophila and humans. These proteins are similar to, but clearly distinct from Upf1p and SMG-2, which are required for nonsense-mediated mRNA decay in yeast and Caenorhabditis elegans and, in the case of SMG-2, for PTGS (Delmay, 2001).


Search PubMed for articles about Drosophila armitage

Ashraf, S. I., McLoon, A. L., Sclarsic, S. M. and Kunes, S. (2006). Synaptic protein synthesis associated with memory is regulated by the RISC pathway in Drosophila. Cell 124(1): 191-205. Medline abstract: 16413491

Atikukke, G., Albosta, P., Zhang, H. and Finley, R. L. (2014). A role for Drosophila Cyclin J in oogenesis revealed by genetic interactions with the piRNA pathway. Mech Dev 133: 64-76. PubMed ID: 24946235

Boutla, A., Delidakis, C., Livadaras, I., Tsagris, M. and Tabler, M. (2001). Short 5?-phosphorylated double-stranded RNAs induce RNA interference in Drosophila. Curr. Biol. 11: 1776-1780. 11719220

Chen, Y., Pane, A. and Schüpbach, T. (2007). Cutoff and aubergine mutations result in retrotransposon upregulation and checkpoint activation in Drosophila, Curr. Biol. 17: 637-642. PubMed citation: 17363252

Cook, H. A., Koppetsch, B. S., Wu, J. and Theurkauf, W. E. (2004). The Drosophila SDE3 homolog armitage is required for oskar mRNA silencing and embryonic axis specification. Cell 116(6): 817-29. 15035984

Dalmay, T., Horsefield, R., Braunstein, T. H. and Baulcombe, D. C. (2001). SDE3 encodes an RNA helicase required for post-transcriptional gene silencing in Arabidopsis. EMBO J. 20: 2069-2078. 11296239

Elbashir, S. M., Lendeckel, W. and Tuschl, T. (2001a). RNA interference is mediated by 21- and 22-nucleotide RNAs. Genes Dev. 15: 188-200. 11157775

Elbashir, S. M., Martinez, J., Patkaniowska, A., Lendeckel, W. and Tuschl, T. (2001b). Functional anatomy of siRNAs for mediating efficient RNAi in Drosophila melanogaster embryo lysate. EMBO J. 20: 6877-6888. 11726523

Gvozdev, V. A., Aravin, A. A., Abramov, Y. A., Klenov, M. S., Kogan, G. L., Lavrov, S. A., Naumova, N. M., Olenkina, O. M., Tulin, A. V. and Vagin, V. V. (2003). Stellate repeats: targets of silencing and modules causing cis-inactivation and trans-activation. Genetica 117: 239-245. 12723703

Haase, A. D., et al. (2010). Probing the initiation and effector phases of the somatic piRNA pathway in Drosophila. Genes Dev. 24(22): 2499-504. PubMed Citation: 20966049

Himber, C., Dunoyer, P., Moissiard, G., Ritzenthaler, C. and Voinnet, O. (2003). Transitivity-dependent and -independent cell-to-cell movement of RNA silencing. EMBO J. 22: 4523-4533. 12941703

Huang, H., Li, Y., Szulwach, K. E., Zhang, G., Jin, P. and Chen, D. (2014). AGO3 Slicer activity regulates mitochondria-nuage localization of Armitage and piRNA amplification. J Cell Biol 206: 217-230. PubMed ID: 25049272

Hutvagner, G., Simard, M.J., Mello, C.C. and Zamore, P.D. (2004). Sequence-specific inhibition of small RNA function. PLoS Biol. 2: e98. 15024405

Jorgensen, R. A. (2003). Sense cosuppression in plants: Past, present, and future. In RNAi: A Guide To Gene Silencing. Hannon, G.J. ed. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press), pp. 5-22

Josse, T., et al. (2007). Telomeric trans-silencing: an epigenetic repression combining RNA silencing and heterochromatin formation. PLoS Genet. 3(9): 1633-43. PubMed citation; Online text

Khurana, J. S., Xu, J., Weng, Z. and Theurkauf. W. E. (2010). Distinct functions for the Drosophila piRNA pathway in genome maintenance and telomere protection. PLoS Genet. 6(12): e1001246. PubMed Citation: 21179579

Klattenhoff, C., et al. (2007). Drosophila rasiRNA pathway mutations disrupt embryonic axis specification through activation of an ATR/Chk2 DNA damage response. Dev. Cell 12(1): 45-55. Medline abstract: 17199040

Koonin, E.V. (1992). A new group of putative RNA helicases. Trends Biochem. Sci. 17: 495-497. 1471259

Linder, P. and Daugeron, M. C. (2000). Are DEAD-box proteins becoming respectable helicases?. Nat. Struct. Biol. 7: 97-99. 10655606

Liu, Q., Rand, T. A., Kalidas, S., Du, F., Kim, H. E., Smith, D. P. and Wang, X. (2003). R2D2, a bridge between the initiation and effector steps of the Drosophila RNAi pathway. Science 301: 1921-1925. 14512631

Mooslehner, K., Muller, U., Karls, U., Hamann, L., and Harbers, K. (1991). Structure and expression of a gene encoding a putative GTP-binding protein identified by provirus integration in a transgenic mouse strain. Mol. Cell. Biol. 11: 886-893. 1899287

Nakamura, A., Sato, K. and Hanyu-Nakamura, K. (2002). Drosophila cup is an eIF4E binding protein that associates with Bruno and regulates oskar mRNA translation in oogenesis. Dev. Cell. 6(1): 69-78. 14723848

Nykanen, A., Haley, B. and Zamore, P. D. (2001). ATP requirements and small interfering RNA structure in the RNA interference pathway. Cell 107: 309-32. 11701122

Pane, A., Wehr, K., Sch├╝pbach, T. (2007). zucchini and squash encode two putative nucleases required for rasiRNA production in the Drosophila germline. Dev. Cell 12(6): 851-62. PubMed citation: 17543859

Reiss, D., Josse, T., Anxolabehere, D. and Ronsseray, S. (2004). Aubergine mutations in Drosophila melanogaster impair P cytotype determination by telomeric P elements inserted in heterochromatin. Mol. Genet. Genomics 272: 336-343. PubMed citation: 15372228

Roche, S. E. and Rio, D. C. (1998). Trans-silencing by P elements inserted in subtelomeric heterochromatin involves the Drosophila Polycomb group gene, Enhancer of zeste. Genetics 149: 1839-1855. PubMed citation: 9691041

Ronsseray, S., Lehmann, M., Nouaud, D. and Anxolabehere, D. (1996). The regulatory properties of autonomous subtelomeric P elements are sensitive to a Suppressor of variegation in Drosophila melanogaster. Genetics 143: 1663-1674. PubMed citation: 8844154

Ronsseray, S., Boivin, A. and Anxolabehere, D. (2001). P-Element repression in Drosophila melanogaster by variegating clusters of P-lacZ-white transgenes. Genetics 159: 1631-1642. PubMed citation: 11779802

Saito, K., et al. (2010). Roles for the Yb body components Armitage and Yb in primary piRNA biogenesis in Drosophila. Genes Dev. 24(22): 2493-8. PubMed Citation: 20966047

Schwarz, D. S., Hutvagner, G., Haley, B., and Zamore, P.D. (2002). Evidence that siRNAs function as guides, not primers, in the Drosophila and human RNAi pathways. Mol. Cell 10: 537-548. 12408822

Stark, A., Brennecke, J., Russell, R. B., and Cohen, R. S. (2003). Identification of Drosophila MicroRNA targets. PLoS Biology 1, e61. 14691536

Tanner, N. K. and Linder, P. (2001). DExD/H box RNA helicases: from generic motors to specific dissociation functions. Mol. Cell 8: 251-262. 11545728

Tomari, Y., Du, T., Haley, B., Schwarz, D. S., Bennett, R., Cook, H. A., Koppetsch, B. S., Theurkauf, W. E. and Zamore, P. D. (2004). RISC assembly defects in the Drosophila RNAi mutant armitage. Cell 116(6): 831-41. 15035985

Vagin, V.V., Sigova, A., Li, C., Seitz, H., Gvozdev, V. and Zamore, P. D. (2006). A distinct small RNA pathway silences selfish genetic elements in the germline. Science 313(5785): 320-4. 16809489

Weng, Y., Czaplinski, K. and Peltz, S. W. (1996). Genetic and biochemical characterization of mutations in the ATPase and helicase regions of the Upf1 protein. Mol. Cell. Biol. 16: 5477-5490. 8816461

Willmann, M. R. (2001). Unravelling PTGS: SDE3 - an RNA helicase involved in RNA silencing in Arabidopsis. Trends Plant Sci. 6: 344-345. 11495769

Biological Overview

date revised: 15 October 2014

Home page: The Interactive Fly © 2017 Thomas Brody, Ph.D.