InteractiveFly: GeneBrief

zucchini: Biological Overview | References

Gene name - zucchini

Synonyms -

Cytological map position - 33B5-33B5

Function - enzyme

Keywords - RNAi and posttranscriptional gene silencing, dorsoventral patterning, translational silencing of osk mRNA, biogenesis of repeat-associated small interfering RNAs

Symbol - zuc

FlyBase ID: FBgn0261266

Genetic map position - 2L: 11,988,735..11,989,861 [-]

Classification - phospholipase-D/nuclease family

Cellular location - unknown

Recent literature
Mohn, F., Handler, D. and Brennecke, J. (2015). Noncoding RNA. piRNA-guided slicing specifies transcripts for Zucchini-dependent, phased piRNA biogenesis Science 348: 812-817. PubMed ID: 25977553
In animal gonads, PIWI-clade Argonaute proteins repress transposons sequence-specifically via bound Piwi-interacting RNAs (piRNAs). These are processed from single-stranded precursor RNAs by largely unknown mechanisms. This study shows that primary piRNA biogenesis is a 3'-directed and phased process that, in the Drosophila germ line, is initiated by secondary piRNA-guided transcript cleavage. Phasing results from consecutive endonucleolytic cleavages catalyzed by Zucchini, implying coupled formation of 3' and 5' ends of flanking piRNAs. Unexpectedly, Zucchini also participates in 3' end formation of secondary piRNAs. Its function can, however, be bypassed by downstream piRNA-guided precursor cleavages coupled to exonucleolytic trimming. These data uncover an evolutionarily conserved piRNA biogenesis mechanism in which Zucchini plays a central role in defining piRNA 5' and 3' ends.
Hayashi, R., Schnabl, J., Handler, D., Mohn, F., Ameres, S. L. and Brennecke, J. (2016). Genetic and mechanistic diversity of piRNA 3'-end formation. Nature [Epub ahead of print]. PubMed ID: 27851737
Small regulatory RNAs guide Argonaute (Ago) proteins in a sequence-specific manner to their targets and therefore have important roles in eukaryotic gene silencing. Of the three small RNA classes, microRNAs and short interfering RNAs are processed from double-stranded precursors into defined 21- to 23-mers by Dicer, an endoribonuclease with intrinsic ruler function. PIWI-interacting RNAs (piRNAs)-the 22-30-nt-long guides for PIWI-clade Ago proteins that silence transposons in animal gonads-are generated independently of Dicer from single-stranded precursors. piRNA 5' ends are defined either by Zucchini, the Drosophila homologue of mitoPLD - a mitochondria-anchored endonuclease, or by piRNA - guided target cleavage. Formation of piRNA 3' ends is poorly understood. This study report that two genetically and mechanistically distinct pathways generate piRNA 3' ends in Drosophila. The initiating nucleases are either Zucchini or the PIWI-clade proteins Aubergine (Aub) or Ago3. While Zucchini-mediated cleavages directly define mature piRNA 3' ends, Aub/Ago3-mediated cleavages liberate pre-piRNAs that require extensive resection by the 3'-to-5' exoribonuclease Nibbler (Drosophila homologue of Mut-7). The relative activity of these two pathways dictates the extent to which piRNAs are directed to cytoplasmic or nuclear PIWI-clade proteins and thereby sets the balance between post-transcriptional and transcriptional silencing. Notably, loss of both Zucchini and Nibbler reveals a minimal, Argonaute-driven small RNA biogenesis pathway in which piRNA 5' and 3' ends are directly produced by closely spaced Aub/Ago3-mediated cleavage events. These data reveal a coherent model for piRNA biogenesis, and should aid the mechanistic dissection of the processes that govern piRNA 3'-end formation.


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 (Aravin, 2003). 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 (Vagin, 2006). 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 (Saito, 2006; Pane, 2007 and references therein).

RNAi has been shown to be involved in axial polarization in the Drosophila germline (Cook, 2004; Tomari, 2004). 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 (Cook, 2004). 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 (Cook, 2004), 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 (Savitsky, 2006). 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 (Koonin, 1996; Ponting, 1996), 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 (Schüpbach, 1991). 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 (Koonin, 1996; Ponting, 1996). 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 (Sung, 1997). 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 (Itaya, 1990) 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 (Cook, 2004), 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 (Vagin, 2006). 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 (Cook, 2004). 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 (Cook, 2004). However, the results reported here together with previous data (Vagin, 2006) 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 (Aravin, 2001). 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 (Aravin, 2003). 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 (Cook, 2004). 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 (Vagin, 2006). 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 (Itaya, 1990). 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 (Sijen, 2003), 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).

The structural biochemistry of Zucchini implicates it as a nuclease in piRNA biogenesis

PIWI-family proteins and their associated small RNAs (piRNAs) act in an evolutionarily conserved innate immune mechanism to provide essential protection for germ-cell genomes against the activity of mobile genetic elements. piRNA populations comprise a molecular definition of transposons, which permits them to distinguish transposons from host genes and selectively silence them. piRNAs can be generated in two distinct ways, forming either primary or secondary piRNAs. Primary piRNAs come from discrete genomic loci, termed piRNA clusters, and seem to be derived from long, single-stranded precursors. The biogenesis of primary piRNAs involves at least two nucleolytic steps. An unknown enzyme cleaves piRNA cluster transcripts to generate monophosphorylated piRNA 5' ends. piRNA 3' ends are probably formed by exonucleolytic trimming, after a piRNA precursor is loaded into its PIWI partner. Secondary piRNAs arise during the adaptive 'ping-pong' cycle, with their 5' termini being formed by the activity of PIWIs themselves. A number of proteins have been implicated genetically in primary piRNA biogenesis. One of these, Drosophila melanogaster Zucchini, is a member of the phospholipase-D family of phosphodiesterases, which includes both phospholipases and nucleases. This study produced a dimeric, soluble fragment of the mouse Zucchini homologue (mZuc; also known as PLD6) and showed that it possesses single-strand-specific nuclease activity. A crystal structure of mZuc at 1.75 Aring; resolution indicates greater architectural similarity to phospholipase-D family nucleases than to phospholipases. Together, these data suggest that the Zucchini proteins act in primary piRNA biogenesis as nucleases, perhaps generating the 5' ends of primary piRNAs (Ipsaro, 2012).

Structure and function of Zucchini endoribonuclease in piRNA biogenesis

PIWI-interacting RNAs (piRNAs) silence transposons to maintain genome integrity in animal germ lines. piRNAs are classified as primary and secondary piRNAs, depending on their biogenesis machinery. Primary piRNAs are processed from long non-coding RNA precursors transcribed from piRNA clusters in the genome through the primary processing pathway. Although the existence of a ribonuclease participating in this pathway has been predicted, its molecular identity remained unknown. This study shows that Zucchini (Zuc), a mitochondrial phospholipase D (PLD) superfamily member, is an endoribonuclease essential for primary piRNA biogenesis. The crystal structure of Drosophila melanogaster Zuc (DmZuc) was solved at 1.75 Aring; resolution. The structure revealed that DmZuc has a positively charged, narrow catalytic groove at the dimer interface, which could accommodate a single-stranded, but not a double-stranded, RNA. DmZuc and the mouse homologue MmZuc (also known as Pld6 and MitoPLD) showed endoribonuclease activity for single-stranded RNAs in vitro. The RNA cleavage products bear a 5'-monophosphate group, a hallmark of mature piRNAs. Mutational analyses revealed that the conserved active-site residues of DmZuc are critical for the ribonuclease activity in vitro, and for piRNA maturation and transposon silencing in vivo. A model is proposed for piRNA biogenesis in animal germ lines, in which the Zuc endoribonuclease has a key role in primary piRNA maturation (Nishimasu, 2012).

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).

The epigenetic trans-silencing effect in Drosophila involves maternally-transmitted small RNAs whose production depends on the piRNA pathway and HP1

The study of P transposable element repression in Drosophila led to the discovery of the Trans-Silencing Effect (TSE), a homology-dependent repression mechanism by which a P-transgene inserted in subtelomeric heterochromatin (Telomeric Associated Sequences, 'TAS') has the capacity to repress in trans, in the female germline, a homologous P-lacZ transgene located in euchromatin. Phenotypic and genetic analysis have shown that TSE exhibits variegation in ovaries, displays a maternal effect as well as epigenetic transmission through meiosis and involves heterochromatin (including HP1) and RNA silencing. This study shows that mutations in squash and zucchini, which are involved in the piwi-interacting RNA (piRNA) silencing pathway, strongly affect TSE. In addition, a molecular analysis of TSE was carried out and it was shown that silencing is correlated to the accumulation of lacZ small RNAs in ovaries. Finally, the production of these small RNAs was shown to be sensitive to mutations affecting squash and zucchini, as well as to the dose of HP1. Thus, these results indicate that the TSE represents a bona fide piRNA-based repression. In addition, the sensitivity of TSE to HP1 dose suggests that in Drosophila, as previously shown in Schizosaccharomyces pombe, an RNA silencing pathway can depend on heterochromatin components (Todeschini, 2010)

Trans-silencing has been shown to be strongly impaired by mutations affecting several components of the piRNA silencing pathway (aubergine, armitage, homeless, piwi). By contrast, TSE was not impaired by mutations affecting R2D2, a component of the siRNA pathway, or Loquacios, a component of both the miRNA and endo-siRNA pathways. This indicates that TSE likely involves the piRNA silencing pathway, a hypothesis which is consistent with the fact that TSE is restricted to the germline, the tissue in which the “canonical” piRNA pathway functions. Further, Squash and Zucchini were found to interact with Aubergine and to localize to the nuage, a cytoplasmic organelle surrounding the nurse cell nuclei, which also contains Aubergine and Armitage and appears to be involved in RNA silencing. squ and zuc mutations were also shown to affect piRNA production in ovaries at the cytological 42AB repetitive sequence cluster, a typical piRNA-producing genomic region. Regarding TE repression in the germline, squ and zuc mutants were found to derepress transcription of the telomeric retrotransposons Het-A and TART and of the I factor, a retrotransposon involved in a Drosophila system of hybrid dysgenesis. It is noteworthy that the I factor and the Het-A retrotransposons have also been found to be sensitive to aub, armi and hls (spn-E). The genetic analysis reported in this study shows that TSE is also highly sensitive to zuc and squ mutations. TSE is therefore sensitive to mutations affecting all the genes of the germline piRNA pathway tested and thus appears to represents a bona fide piRNA-based repression (Todeschini, 2010)

The presence of lacZ small RNAs in ovaries of females carrying a TSE silencer was therefore investigated using RNase protection analysis. In addition, paternal vs maternal transmission of the telomeric silencer was compared. Indeed, TSE was previously shown to have a maternal effect, i.e. strong repression occurs only when the telomeric silencer is maternally inherited, whereas a paternally-inherited telomeric silencer has weak or null repression capacities. More precisely, it was shown genetically that TSE requires inheritance of two components, a maternal cytoplasmic component plus a chromosomal copy of the transgene, but these two components can be transmitted separately. Indeed, a paternally-inherited telomeric transgene can be 'potentiated' by a maternally-inherited cytoplasm from a female bearing a silencer. This interaction also functions between telomeric silencers located on different chromosomal arms. The RNase protection analysis reported in this study shows that: (1) P-1152, a telomeric P-lacZ silencer, produces small lacZ RNAs in ovaries; (2) P-1152 lacZ small RNA accumulation is negatively affected in squ and zuc mutants; (3) maternal transmission of P-1152 leads to accumulation of higher levels of these small RNAs than that observed upon paternal P-1152 transmission. These results were reproduced with independent RNAse protection assays. The size of the small RNAs detected in this study appears smaller (around 22-23 nt) than that corresponding to piRNAs as characterized by deep sequencing (23-28 nt), but they are consistent with piRNAs as detected by RNAse protection assays in other studies: this can result from the RNAse protection protocol which tends to reduce the size of the RNAs detected. In conclusion, the results strongly suggest that the lacZ small RNAs in P-1152 oocytes may correspond to cytoplasmically-transmitted piRNAs mediating the maternal effect of TSE, as well potentiating a paternally-inherited telomeric silencer (Todeschini, 2010)

TSE was previously shown to be sensitive to mutations affecting HP1 since a negative, dose-dependent, effect on TSE was found with two loss of function alleles of Su(var)205 (including Su(var)2-505). RNase protection analysis shows here that lacZ small RNA accumulation is also negatively affected by the dose of HP1. Although it cannot be excluded that this effect may be indirect, this opens the possibility that some piRNA-producing loci depend on the presence of HP1 itself at the locus to produce piRNAs. A similar model was recently proposed for rhino, a HP1 homolog, mutations of which strongly reduce the production of piRNAs by dual strand piRNA-producing loci (Klattenhoff, 2009). That study proposed that rhino is required for the production of the long precursor RNAs which are further processed to produce primary piRNAs. Note that in that study, rhino mutants were shown to have a drastic effect on the production of piRNAs by the X-chromosome TAS locus. A similar situation may therefore exist for HP1 at this locus and, if so, it would be interesting to characterize more precisely the function of HP1 in the production of piRNAs at the TAS locus (Todeschini, 2010).

HP1 was shown to be present at TAS. A first possibility would be that HP1 stimulates transcription of the TAS locus as a classical transcription factor, independent of any heterochromatic role at this locus. Consistent with this, it was shown that PIWI, a partner of HP1 (Brower-Toland, 2007), promotes euchromatin histone modification and piRNA transcription at the third chromosome TAS. The precise status of TAS, however, remains complex since some studies have shown that TAS exhibit some of the properties attributed to heterochromatin and carry primarily heterochromatic histone tags. Therefore, a second possibility would be that HP1 enhances the heterochromatic status of TAS in the germline, such that production of aberrant transcripts being processed into piRNAs is enhanced. This would result in a 'heterochromatin-dependent RNA silencing pathway'. Examples of heterochromatin formation that depends on RNA silencing ('RNA-dependent heterochromatin formation') have been described in numerous species including yeast and plants. In Drosophila, this type of interaction has been described for variegation of pigment production in the eye linked to the insertion of the white gene in different types of heterochromatin structures, as well as for heterochromatin formation at telomeres in the germline. Therefore, telomeric regions in fly may be submitted to both RNA-dependent heterochromatin formation and heterochromatin-dependent RNA silencing. RNA silencing may favor heterochromatin formation that in turn potentiates RNA silencing, resulting in a functional positive loop between transcriptional gene silencing and post-transcriptional gene silencing. In such cases, RNA silencing and heterochromatin may not only reinforce each other but may also be functionally interdependent. Such bidirectional reinforcement between RNA silencing and heterochromatin formation was demonstrated in S. pombe since: (1) deletion of genes involved in RNA silencing were shown to derepress transcriptional silencing from centromeric heterochromatic repeats, accompanied by loss of Histone 3 Lysine 9 methylation and Swi6 (a HP1 homolog) delocalization; (2) Swi6 was found to be required for the propagation and the maintenance of the RNA Induced Transcriptional Silencing (RITS) complex at the mat locus, a complex involved in amplification of RNA silencing. A positive loop between RNA silencing and heterochromatin formation may therefore also be at play in the Drosophila germline. According to this model, the epigenetic transmission of TSE through meiosis, ( i.e., six generations of maternal transmission of the silencer are required to elicit a strong TSE following maternal inheritance of a cytoplasm devoid of lacZ piRNAs would underlie progressive establishment of this loop. Note that RNAi-dependent DNA methylation in Arabidopsis thaliana has been shown to occur progressively over several consecutive generations (Todeschini, 2010)

Since TSE can be considered as a sub-phenomenon within P regulation, it may underlie epigenetic transmission of the P element repression. P element mobilization is responsible for a syndrome of germline abnormalities, known as the 'P-M' system of hybrid dysgenesis which includes a high mutation rate, chromosomal rearrangements, male recombination and an agametic temperature-sensitive sterility called GD sterility (Gonadal Dysgenesis). P-induced hybrid dysgenesis is repressed by a maternally inherited cellular state called the 'P cytotype'. The absence of P-repression is called M cytotype. G1 females produced from the cross (P cytotype females × M cytotype males) present a strong capacity for repression, whereas females produced from the reciprocal cross present a weak capacity for repression. In the subsequent generations, cytotype is progressively determined by the chromosomal P elements but the influence of the initial maternal inheritance can be detected for up to five generations. Therefore, P cytotype exhibits partial epigenetic transmission through meiosis. Furthermore, the identification and use of telomeric P elements as P cytotype determinants, has made it possible to show that P cytotype (like TSE) involves a strictly-maternally inherited component (called the pre-P cytotype), is sensitive to mutations affecting HP1 and aubergine and is correlated to maternal deposition of piRNAs. Some of these properties are also found for the I factor which is responsible for the occurrence of another system of hybrid dysgenesis ('I-R' system). TSE therefore parallels germline regulation of TEs (P, I), and does not resemble regulation of TEs in the somatic follicle cells (gypsy, ZAM, Idefix) for which no epigenetic transmission of repression capacities through meiosis has been described so far. It will be interesting to test if previously described cases of RNA-dependent heterochromatin formation show the reciprocal dependence, thus being able to form a positive loop (Todeschini, 2010 and references therein)

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).


Search PubMed for articles about Drosophila Zucchini

Aravin, A. A., et al. (2001). Double-stranded RNA-mediated silencing of genomic tandem repeats and transposable elements in the D. melanogaster germline, Curr. Biol. 11: 1017-1027. PubMed ID: 11470406

Aravin, A. A., et al. (2003). The small RNA profile during Drosophila melanogaster development, Dev. Cell 5: 337-350. PubMed ID: 12919683

Brower-Toland, B. et al. (2007). Drosophila PIWI associates with chromatin and interacts directly with HP1a. Genes Dev. 21: 2300-2311. PubMed ID: 17875665

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 ID: 17363252

Cook, H. H., 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: 817-829. PubMed ID: 15035984

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 ID: 20966049

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

Itaya, M. (1990). Isolation and characterization of a second RNase H (RNase HII) of Escherichia coli K-12 encoded by the rnhB gene, Proc. Natl. Acad. Sci. 87: 8587-8591. PubMed ID: 2172991

Klattenhoff, C., et al. (2009). The Drosophila HP1 homolog Rhino is required for transposon silencing and piRNA production by dual-strand clusters. Cell 138: 1137-1149. PubMed ID: 19732946

Koonin, E. V. (1996). A duplicated catalytic motif in a new superfamily of phosphohydrolases and phospholipid synthases that includes poxvirus envelope proteins, Trends Biochem. Sci. 21: 242-243. PubMed ID: 8755242

Nishimasu, H., Ishizu, H., Saito, K., Fukuhara, S., Kamatani, M. K., Bonnefond, L., Matsumoto, N., Nishizawa, T., Nakanaga, K., Aoki, J., Ishitani, R., Siomi, H., Siomi, M. C. and Nureki, O. (2012). Structure and function of Zucchini endoribonuclease in piRNA biogenesis. Nature 491: 284-287. PubMed ID: 23064230

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 ID: 17543859

Ponting, C. P. and Kerr, I. D. (1996). A novel family of phospholipase D homologues that includes phospholipid synthases and putative endonucleases: identification of duplicated repeats and potential active site residues, Protein Sci. 5: 914-922. PubMed ID: 8732763

Saito, K., et al. (2006). Specific association of Piwi with rasiRNAs derived from retrotransposon and heterochromatic regions in the Drosophila genome, Genes Dev. 20: 2214-2222. PubMed ID: 16882972

Savitsky, M., et al. (2006). Telomere elongation is under the control of the RNAi-based mechanism in the Drosophila germline. Genes Dev. 20: 345-354. PubMed ID: 16452506

Schüpbach, T. and Wieschaus, E. (1991). Female sterile mutations on the second chromosome of Drosophila melanogaster. II. Mutations blocking oogenesis or altering egg morphology, Genetics 129: 1119-1136. PubMed ID: 1783295

Sijen, T. and Plasterk, R. H. (2003). Transposon silencing in the Caenorhabditis elegans germ line by natural RNAi. Nature 426: 310-314. PubMed ID: 14628056

Sung, T. C., et al. (1997). Mutagenesis of phospholipase D defines a superfamily including a trans-Golgi viral protein required for poxvirus pathogenicity, EMBO J. 16: 4519-4530. PubMed ID: 9303296

Todeschini, A. L., Teysset, L., Delmarre, V. and Ronsseray, S. (2010). The epigenetic trans-silencing effect in Drosophila involves maternally-transmitted small RNAs whose production depends on the piRNA pathway and HP1. PLoS One. 5(6): e11032. PubMed ID: 20559422

Tomari, Y., et al. (2004). RISC assembly defects in the Drosophila RNAi mutant armitage, Cell 116: 831-841. PubMed ID: 15035985

Vagin, V. V., et al. (2006). A distinct small RNA pathway silences selfish genetic elements in the germline, Science 313: 320-324. PubMed ID: 16809489

Ipsaro, J. J., Haase, A. D., Knott, S. R., Joshua-Tor, L. and Hannon, G. J. (2012). The structural biochemistry of Zucchini implicates it as a nuclease in piRNA biogenesis. Nature 491: 279-283. PubMed ID: 23064227

Biological Overview

date revised: 25 October 2014

Home page: The Interactive Fly © 2008 Thomas Brody, Ph.D.

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