Structural insights into Rhino‐Deadlock complex for germline piRNA cluster specification
2018; Springer Nature; Volume: 19; Issue: 7 Linguagem: Inglês
10.15252/embr.201745418
ISSN1469-3178
AutoresBowen Yu, Yu An Lin, Swapnil S. Parhad, Zhaohui Jin, Jinbiao Ma, William E. Theurkauf, ZZ Zhao Zhang, Ying Huang,
Tópico(s)Genomics and Phylogenetic Studies
ResumoScientific Report1 June 2018free access Transparent process Structural insights into Rhino-Deadlock complex for germline piRNA cluster specification Bowen Yu State Key Laboratory of Molecular Biology, National Center for Protein Science Shanghai, Shanghai Science Research Center, Shanghai Key Laboratory of Molecular Andrology, CAS Center for Excellence in Molecular Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Shanghai, China Search for more papers by this author Yu An Lin Department of Embryology, Carnegie Institution for Science, Baltimore, MD, USA Search for more papers by this author Swapnil S Parhad Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, MA, USA Search for more papers by this author Zhaohui Jin State Key Laboratory of Molecular Biology, National Center for Protein Science Shanghai, Shanghai Science Research Center, Shanghai Key Laboratory of Molecular Andrology, CAS Center for Excellence in Molecular Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Shanghai, China Search for more papers by this author Jinbiao Ma State Key Laboratory of Genetic Engineering, Collaborative Innovation Center for Genetics and Development, Department of Biochemistry, Institute of Plant Biology, School of Life Sciences, Fudan University, Shanghai, China Search for more papers by this author William E Theurkauf Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, MA, USA Search for more papers by this author ZZ Zhao Zhang Corresponding Author [email protected] Department of Embryology, Carnegie Institution for Science, Baltimore, MD, USA Search for more papers by this author Ying Huang Corresponding Author [email protected] orcid.org/0000-0002-2806-2874 State Key Laboratory of Molecular Biology, National Center for Protein Science Shanghai, Shanghai Science Research Center, Shanghai Key Laboratory of Molecular Andrology, CAS Center for Excellence in Molecular Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Shanghai, China Search for more papers by this author Bowen Yu State Key Laboratory of Molecular Biology, National Center for Protein Science Shanghai, Shanghai Science Research Center, Shanghai Key Laboratory of Molecular Andrology, CAS Center for Excellence in Molecular Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Shanghai, China Search for more papers by this author Yu An Lin Department of Embryology, Carnegie Institution for Science, Baltimore, MD, USA Search for more papers by this author Swapnil S Parhad Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, MA, USA Search for more papers by this author Zhaohui Jin State Key Laboratory of Molecular Biology, National Center for Protein Science Shanghai, Shanghai Science Research Center, Shanghai Key Laboratory of Molecular Andrology, CAS Center for Excellence in Molecular Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Shanghai, China Search for more papers by this author Jinbiao Ma State Key Laboratory of Genetic Engineering, Collaborative Innovation Center for Genetics and Development, Department of Biochemistry, Institute of Plant Biology, School of Life Sciences, Fudan University, Shanghai, China Search for more papers by this author William E Theurkauf Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, MA, USA Search for more papers by this author ZZ Zhao Zhang Corresponding Author [email protected] Department of Embryology, Carnegie Institution for Science, Baltimore, MD, USA Search for more papers by this author Ying Huang Corresponding Author [email protected] orcid.org/0000-0002-2806-2874 State Key Laboratory of Molecular Biology, National Center for Protein Science Shanghai, Shanghai Science Research Center, Shanghai Key Laboratory of Molecular Andrology, CAS Center for Excellence in Molecular Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Shanghai, China Search for more papers by this author Author Information Bowen Yu1,‡, Yu An Lin2,‡, Swapnil S Parhad3, Zhaohui Jin1, Jinbiao Ma4, William E Theurkauf3, ZZ Zhao Zhang *,2 and Ying Huang *,1 1State Key Laboratory of Molecular Biology, National Center for Protein Science Shanghai, Shanghai Science Research Center, Shanghai Key Laboratory of Molecular Andrology, CAS Center for Excellence in Molecular Cell Science, Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Shanghai, China 2Department of Embryology, Carnegie Institution for Science, Baltimore, MD, USA 3Program in Molecular Medicine, University of Massachusetts Medical School, Worcester, MA, USA 4State Key Laboratory of Genetic Engineering, Collaborative Innovation Center for Genetics and Development, Department of Biochemistry, Institute of Plant Biology, School of Life Sciences, Fudan University, Shanghai, China ‡These authors contributed equally to this work *Corresponding author. Tel: +1 4102463092; E-mail: [email protected] *Corresponding author. Tel: +86 2120778200; E-mail: [email protected] EMBO Rep (2018)19:e45418https://doi.org/10.15252/embr.201745418 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract PIWI-interacting RNAs (piRNAs) silence transposons in germ cells to maintain genome stability and animal fertility. Rhino, a rapidly evolving heterochromatin protein 1 (HP1) family protein, binds Deadlock in a species-specific manner and so defines the piRNA-producing loci in the Drosophila genome. Here, we determine the crystal structures of Rhino-Deadlock complex in Drosophila melanogaster and simulans. In both species, one Rhino binds the N-terminal helix–hairpin–helix motif of one Deadlock protein through a novel interface formed by the beta-sheet in the Rhino chromoshadow domain. Disrupting the interface leads to infertility and transposon hyperactivation in flies. Our structural and functional experiments indicate that electrostatic repulsion at the interaction interface causes cross-species incompatibility between the sibling species. By determining the molecular architecture of this piRNA-producing machinery, we discover a novel HP1-partner interacting mode that is crucial to piRNA biogenesis and transposon silencing. We thus explain the cross-species incompatibility of two sibling species at the molecular level. Synopsis Rhino recruits Deadlock through a novel interacting mode that is crucial to piRNA biogenesis and transposon silencing. Key amino acid differences determine the cross-species incompatibility between Drosophila melanogaster and Drosophila simulans. The crystal structures of Rhino-Deadlock complex from melanogaster and simulans indicate that one Rhino CSD domain binds the HhH motif of one Deadlock through a novel interface. Disrupting the interface leads to infertility and transposon hyperactivation in flies, indicating the crucial role of Rhino-Deadlock machinery in piRNA biogenesis. The structures highlight that the cross-species incompatibility is due to electrostatic repulsion caused by amino acid differences between the two species. Introduction In animal gonads, the PIWI-interacting RNA (piRNA) pathway suppresses transposons during germline development 1-8. The piRNA precursors are first transcribed from discrete heterochromatic loci in the genome called piRNA clusters 9. They are eventually chopped and trimmed into 22–30 nucleotide-long piRNAs 9-12. In Drosophila, an heterochromatin protein 1 (HP1) homolog protein, Rhino (Rhi), is required to specify piRNA clusters for piRNA biogenesis 13-15. Because Rhi is essential to the formation of piRNA precursors, mutating rhi unleashes transposons and results in female sterility 15. Despite its important role in piRNA biogenesis and transposon silencing, how Rhi distinguishes itself from other proteins in the HP1 family to fulfill its unique function is still largely unknown. Similar to other HP1 family proteins, Rhi consists of an N-terminal chromodomain (CD), a C-terminal chromoshadow domain (CSD), and a hinge region in between these two domains 16. Rhi binds heterochromatic piRNA clusters by recognizing the H3K9me3 marks in part via its CD 14, 17, 18. The CSD domain of Rhi is predicted to anchor a protein complex that encompasses Deadlock (Del) and Cutoff (Cuff) through protein–protein interactions 14. Previous yeast two-hybrid assay data showed that Del may function as a scaffold protein that interacts with Rhi-CSD via its N-terminal 660 amino acids and recruits Cuff via the C-terminal 321 amino acids 14. Moreover, Deadlock interacts with Moonshiner, a TFIIA-L paralog of RNA polymerase II preinitiation complex, to trigger the transcription of piRNA precursors 19. Depletion of any of the three proteins in the Rhino-Deadlock-Cutoff complex leads to mislocalization of the other two proteins and disruption in piRNA production, indicating that the interactions among these three proteins are indispensable for their functions 14. The molecular basis of these interactions, however, remains unclear. Unlike most proteins in the HP1 family, rhi is rapidly evolving and has undergone repeated episodes of positive selections, implicating its essential role in transposon silencing 16. Moreover, simulans Rhi (Ds-Rhi) cannot bind melanogaster Del (Dm-Del), or rescue the melanogaster rhi null allele, largely because Ds-Rhi-CSD and Dm-Del cannot interact, suggesting cross-species incompatibility between melanogaster and simulans. Here, we describe the crystal structures of Rhi-CSD-Del complex in Drosophila melanogaster and in Drosophila simulans, respectively. We found that, in both species, the N-terminal 60 residues of Del (Del-N), which fold into a helix–hairpin–helix (HhH) motif, are essential and sufficient to interact with the β-sheet of Rhi-CSD in a non-canonical 1:1 binding mode. Residues involved in the interactions are highly conserved among Drosophila Rhi proteins but not in classical HP1s. Moreover, mutating the key residues at the interaction interface results in transposon hyperactivation and animal sterility. Structural analyses of the melanogaster and simulans Rhi-Del complexes reveal that individual amino acid differences in Rhi-CSD between the two species lead to local conformational change and electrostatic repulsion that render simulans Rhi-CSD unable to interact with melanogaster Del. Thus, our findings uncover a novel HP1-partner binding mode that is essential for piRNA biogenesis and provide molecular insights to understand the cross-species incompatibility between two sibling species. Results and discussion Deadlock binds Rhino-CSD via the N-terminal 60 amino acids Interaction between Rhi and Del is essential to anchor the protein complex on clusters to drive the production of piRNA precursor 14. To explore the molecular mechanism underlying the physical interactions between Rhi and Del, we first performed yeast two-hybrid assays to map the minimal interaction regions from each protein. By testing three fragments from the CSD of Dm-Rhi and ten fragments from Dm-Del, we found that residues 353–418 from Dm-Rhi-CSD and the N-terminal 60 amino acids of Dm-Del (Dm-Del-N) are essential and sufficient for the physical binding (Figs 1A and EV1A–C). His-tagged Dm-Rhi-CSD (His-Dm-Rhi-CSD) co-migrated with GST-fused Dm-Del-N (GST-Dm-Del-N) on size-exclusive chromatography (SEC) earlier than His-Dm-Rhi-CSD alone (Fig EV1E and F), validating that Dm-Rhi-CSD forms a complex with Dm-Del-N in solution. We also detected that simulans Rhi-CSD interacts with simulans Del from the same regions but not with the corresponding melanogaster partner (Figs 1B and EV1C–D), indicating that there was no cross-species interaction between melanogaster Del-N and simulans Rhi-CSD. Interestingly, we detected interaction between melanogaster Rhi-CSD with simulans Del (Fig EV1D), consistent with the previous finding that these two proteins could form non-functional complex 20. Figure 1. Overall structure of melanogaster Rhi-CSD and Del-N A, B. Top: Domain architecture of Drosophila melanogaster Deadlock (Dm-Del) and D. simulans Deadlock (Ds-Del). The N-terminal 60 amino acids are colored in pink and teal, respectively. Bottom: Schematic of D. melanogaster Rhino (Dm-Rhi) and D. simulans Rhino (Ds-Rhi). The chromodomain (CD) is colored in yellow. Dm-Rhi-CSD (residue 353–418) is colored in green, and Ds-Rhi-CSD (residue 436–500) is colored in violet. Dashed lines indicate the interaction region between Rhi and Del. C. The overall structure of the Dm-Rhi-CSD in complex with Dm-Del-N shown in a ribbon representation and view of 90° rotation around horizontal axis. Dm-Rhi-CSD (A: green, B: light blue) form a homodimer. Two molecules of Dm-Del-N (A: pink, B: yellow) bind to each monomer of a Dm-Rhi-CSD dimer. D. Surface representation of Dm-Rhi-CSD in complex with Dm-Del-N. Color is shown as (C). E. Interaction details of Dm-Rhi-CSD dimerization. F. Dm-Del-N represents a helix–hairpin–helix fold. Schematic diagram showing the hydrophobic interaction details between α1 and α2 of Dm-Del-N. Download figure Download PowerPoint Click here to expand this figure. Figure EV1. Mapping the minimum interaction region between Rhi and Del (related to Fig 1A and B) A–D. Yeast two-hybrid assays mapped the interaction regions between Rhi and Del. Interactions were determined by measuring the β-galactosidase activity produced by the reporter gene. Data are averages of three independent β-galactosidase measurements (n = 3) : − (0.00–0.99), + (1.00–9.99), ++ (10.00–99.99), and +++ (100.00–999.99). (A) Dm-Rhi-CSD (residues 353–418, colored in green) is sufficient to interact with the N-terminal 660 amino acids of Del (Dm-Del-ΔC). (B) The N-terminal 60 amino acids of Del (Dm-Del-N, colored in pink) is sufficient to interact with Dm-Rhi-CSD. (C) Ds-Rhi-CSD (residues 436–500, colored in wine) showed no interaction with Dm-Del-N. (D) Both Ds-Rhi-CSD and Dm-Rhi-CSD could interact with Ds-Del-N (residues 1–60, colored in teal). E. His-Dm-Rhi-CSD and GST-Dm-Del-N form a complex in solution. Left: His-Dm-Rhi-CSD and GST-Dm-Del-N co-migrated as a single peak on size-exclusive chromatography (Superdex 75 increase 10/300 column) earlier than His-Dm-Rhi-CSD alone. The elution profiles are indicated in the key. F. Top and bottom panels showed the SDS–PAGE results of the peak fractions as indicated in (E). Download figure Download PowerPoint Overall structure of melanogaster Rhino-CSD and Deadlock complex To gain insights into this essential interaction, we crystallized melanogaster Rhi and Del complex. We fused Dm-Del-N to the C terminus of Dm-Rhi-CSD using a (Gly-Ser)5 linker to generate a fusion protein Dm-Rhi-CSD-(GS)5-Del-N and obtained crystals with 2.1 Å diffraction quality. The overall structure of Dm-Rhi-CSD adopts an OB-fold consisting of four β-strands (β1–β4) and two α-helices (α1 and α2). β1–β3 were found to form an anti-parallel β-sheet packed against α1 and α2 (Fig 1C). Two Dm-Rhi-CSD molecules (A–B) forms a homodimer related by a non-crystallographic twofold axis similar to other HP1 family proteins (Figs 1E and EV2A–D) although Dm-Rhi-CSD shares limited sequence identity with human HP1β-CSD (22.7%; PDB 3Q6S) and Drosophila HP1a-CSD (24.2%; PDB 3P7J). However, residues involved in the dimerization interface are conserved among HP1 proteins (Fig EV3A). Click here to expand this figure. Figure EV2. Dm-Rhi-CSD exists a homodimer in crystal (related to Fig 1E) A. Superimposition of Dm-Rhi-CSD (light blue), Dm-HP1a-CSD (3P7J, yellow), and Hs-HP1β-CSD (3Q6S, wheat). B–F. Homodimer formed by CSD domain of HP1 family proteins. Structures are shown with ribbon cartoon. (B) Dm-Rhi-CSD in complex with Del. Hydrophobic interface I has not been shown. (C) Dm-HP1a-CSD: 3P7J. (D) Hs-HP1β-CSD in complex with Shugoshin 1: 3Q6S. (E) Hs-HP1β-CSD in complex with EMSY: 2FMM. (F) Hs-HP1γ-CSD in complex with histone H3: 5T1I. G. Detailed intermolecular contacts on interface I' formed by Dm-Del-N (B: yellow) and Dm-Rhi-CSD (B: light blue). H. Superimposition of interface I and I'. Download figure Download PowerPoint Click here to expand this figure. Figure EV3. Sequence alignments of Rhi-CSD and Del-N Sequence alignment of CSDs in Rhino and other HP1 family members. Amino acids that are identical and conserved are shown in red and pink. Residues on the Rhi-CSD dimeric interface are indicated by blue squares. Residues involved in interface I and interface II are labeled by green triangles and purple pentagons, respectively. The amino acid difference causing conformational change between Dm-Rhi-CSD and Ds-Rhi-CSD is framed in red. Secondary structural elements of Dm-Rhi-CSD are indicated above. Sequence alignment of Del-N in Drosophila species. Color code is indicated as in (A). Residues forming the interface I and interface II are marked by green triangles and purple pentagons, respectively. Secondary structural elements of Dm-Del-N are indicated above. Download figure Download PowerPoint Full-length Dm-Del is predicted not to have any known domain structure. We found that Dm-Del-N consists of two α-helices (α1 and α2) and a C-terminal β-strand (β1; Fig 1F). The two α-helices fold into a helix–hairpin–helix motif (HhH) 21 and pack almost perpendicularly against the β-sheet of Dm-Rhi-CSD (Fig 1C). Two molecules of Dm-Del-N (A–B) bind to each monomer of the Dm-Rhi-CSD dimer (Fig 1C and D), while α1 and α2 are mostly stabilized by hydrophobic interactions (Fig 1F). In detail, L4, I7 on the N-terminal loop, M9, L13, W16, I19, L20, and L23 on α1 form a hydrophobic cluster with W33, F36, F37, F40, L41, and W44 on α2. Moreover, W33 and W44 contact K12 and K28 via cation-π interactions, respectively (Fig 1F). Rhino-CSD interacts with Del-N with a novel interface different from classical HP1-CSDs Two molecules of Dm-Del-N bind to each monomer of the Dm-Rhi-CSD dimer through novel interaction interfaces I and I' (Fig 2A and B). At this unique interface (we named it interface I), six aromatic residues including W44, F40 from Dm-Del-N (A), H369, Y371, F378, and F380 from Dm-Rhi-CSD (A) form a hydrophobic cluster (Fig 2A and B). C15 and A391 further enhance the interactions. In addition, residues L4 and I7 from Dm-Del-N (A) contact residues L368 and I389 through hydrophobic interactions, and M373 is surrounded by I19, L23, and F36. K28 contacts D375 and D376 through electrostatic interactions, and H18 π-π stacks against F356 (Fig 2B). Interface I', which is formed by Dm-Del-N (B) and Dm-Rhi-CSD (B), is highly similar to interface I (Fig EV2G–H). For convenience, we refer interface I' as interface I, unless specified. Figure 2. Interaction interfaces between melanogaster Rhi-CSD and Del-N A. Interaction interfaces between Dm-Del-N (A: pink), Dm-Del-N (B: yellow), and Dm-Rhi-CSD homodimer (A: green, B: light blue). Interface I is formed by the HhH motif of Dm-Del-N (A) and the β-sheets of Dm-Rhi-CSD (A). Interface I' is formed by Dm-Del-N (B) and Dm-Rhi-CSD (B). Interface II consists of the β-sandwich structure formed by the C-terminal extension of Dm-Del-N (A) and the C-terminal β-strands of Dm-Rhi-CSD dimer. B, C. Stereo views of detailed intermolecular contacts on interface I and interface II, respectively. Yellow dashed lines indicate the hydrogen bonds. D. A surface representation showing the C-terminal β-strand of Dm-Del-N (A: pink) is located at the center of Dm-Rhi-CSD dimer. I50 occupies the central position and is named as position 0, Y48, C49, Q51, and T52 are designated as position −2, −1, +1, and +2. Dm-Del-N (A) and Dm-Del-N (B) are shown in a cartoon model. Download figure Download PowerPoint The melanogaster Rhi-Del complex also has the canonical HP1-partner binding mode. The C-terminal extension of Dm-Del-N (A) emanates from the α-helices platform to form the β-sandwich structure with the Dm-Rhi-CSD dimer (we named it interface II; Fig 2A and C). In detail, I50, which locates at the center of the β-sandwich structure, is symmetrically surrounded by L412 and I414 from both monomers of the Dm-Rhi-CSD. Thus, we specified I50 as position 0 (Fig 2C and D). Moreover, Y48 is situated in a hydrophobic pocket formed by H369, Y408, L412, and V416′ while C49 occupies position −1, which forms a hydrophobic cluster with R413, I415, P417′, and I415′(Fig 2C). Sequence alignment indicated that residues involved in interface I are highly conserved among the Drosophila species, indicating that the intermolecular contacts on interface I are highly specific in flies (Fig EV3A and B). The β-sandwich structure on interface II requires participation of both molecules of Rhi-CSD dimer. In contrast, interface I is constituted by one molecule of Rhi-CSD and one molecule of Del-N. However, both interfaces are equally important for the binding of Rhi-CSD to Del, which differs to classic HP1s that recruit their binding partners via the formation of β-sandwich structure (Fig EV2D–F) 22-24. Both interfaces are important for intermolecular contacts and the protein function in vivo To validate the key residues that are important for Rhino function, we constructed two triple mutants, YFF (Y371T/F378S/F380S) and HYM (H369G/Y371T/M373S), which substitute the hydrophobic residues on interface I with either hydrophilic residues or the corresponding residues from human HP1β, as well as a quintuple mutant LRIIV, where residues 412–416 on β4 were replaced with five alanine residues to disrupt interface II. All mutants lost the ability to interact with Dm-Del-N in the yeast two-hybrid assay (Fig 3A). Figure 3. Both interfaces are essential for Rhi function in vivo Yeast two-hybrid assays indicate the binding of wild-type or mutant Dm-Rhi-CSD with Dm-Del-N. Mutations of key residues on both interface I or interface II abolish the binding. The averages of three independent β-galactosidase measurements (n = 3) with SD values (error bar) are shown. The detailed β-gal units are listed as follows: WT, 7.95 ± 3.10; YFF, 0.28 ± 0.02; HYM, 0.61 ± 0.03; LRIIV, 0.60 ± 0.09; NDD, 0.24 ± 0.07; HIND, 0.44 ± 0.37; K28M, 1.85 ± 0.20. Reverse mutant of Ds-Rhi-CSD could interact with Dm-Del-N. The averages of three independent β-galactosidase measurements (n = 3) with SD values (error bar) are shown. The detailed β-gal units are listed as follows: WT, 0.40 ± 0.04; RM, 60.50 ± 4.64. Fertility of the rhi mutant flies carrying one copy of transgene expresses either wild-type Rhi or Rhi with amino acid mutations. The transgene is driven by rhi promoter and has an eGFP tag at N-terminus. The localization of wild-type Rhi or Rhi with amino acid mutations. RNA-Seq to quantify the expression of Rhi transcripts from each transgene. RhoGAP54D and CG30105 are two genes locate upstream of rhi. The transcripts from them serve as normalization control. rhi2/KG flies only produce a few truncated rhi transcripts. Note: The GFP tagged rhi construct does not contain 3'UTR of rhi. Western blot to quantify the GFP::Rhi proteins from 60 fly ovaries. Download figure Download PowerPoint To validate our findings in vivo, we introduced each of the three variants and wild-type rhi (as control) into rhi null mutant flies. All four transgenes, which are driven by rhi endogenous promoter and express N-terminus eGFP tag, are located at the same genomic locus to ensure the same expression level. As an essential piRNA pathway component, mutating rhi leads to female sterility 15, 25. The positive control transgene rescued fly sterility (Fig 3C). In contrast, flies carrying either YFF or HYM mutant were still sterile (Fig 3C); flies that had quintuple amino acid substitution, which may have disrupted the protein interaction on interface II, were not completely sterile although their fertility was dramatically compromised (Fig 3C; eggs laid by 14 female in 10 days: 10,104 from wild-type controls versus 1,868 from LRIIV mutants; hatching rate: 78.29% from wild-type controls versus 17.13% from LRIIV mutants). Both Rhi and Del localize to piRNA clusters and display characteristic puncta in the germ cell nuclei by immunostaining 14, 15. Depleting Del in nurse cells leads to the disappearance of Rhi foci 14, suggesting that Del-binding is required for Rhi localization. Indeed, while eGFP-tagged wild-type Rhi displayed the distinct staining pattern, no signal from any of the three Del-binding mutation constructs was observed (Fig 3D). By assaying the Rhi transcript level using RNA-seq, we confirmed that all four transgenes express at similar level (Fig 3E). Next, we sought to quantify the protein abundance and stability for these mutants by Western blot. However, we were unable to detect any signal even from eGFP-tagged wild-type Rhi likely because Rhi maintains at a low steady-state level in Drosophila ovaries. As a result, we utilized immunoprecipitation to enrich Rhi concentration for Western blot and found that Del binding appears to stabilize Rhi in vivo. Only wild-type Rhi displayed a strong band, while the signals from other three mutants are weaker (Fig 3F). Loss of Rhi results in transposon hyperactivation in germ cells 15. For this reason, we attempted to determine whether the transgenes could rescue transposon silencing defects in rhi mutants by measuring the transcriptome of ovaries. Out of the 92 transposon families examined 26, mutating Rhi causes 23 to increase their transcript abundance more than fivefold (Fig 4), indicating that these transposons are largely silenced by the piRNAs derived from the dual-strand clusters marked by Rhi. Consistent with the results from fertility and localization assays (Fig 3D and E), introducing one copy of wild-type transgene into rhi mutants restored transposon silencing for all of these 23 families (Fig 4). However, at least 20 of these 23 transposon families showed more than fivefold increase (22 of them showed more than twofold increase) on transposon mRNA when any of the mutant constructs was introduced into the rhi null mutants (Fig 4). The R2-element was the only family that appeared to be affected by complete loss of Rhi, but was still suppressed by introducing of one of the three mutant constructs (Fig 4). This may reflect that R2-element suppression is achieved by Rhi in a manner independent of Del-binding. We also noticed that loss of Rhi had no effect on blood silencing, but introducing any of the three mutant constructs resulted in an increase of blood mRNA for more than sixfold (Fig 4). This suggests that these mutant versions of transgenes may have a pronounced effect on blood derepression. Figure 4. Disrupting Rhi and Del interaction leads to transposon overexpressionRNA-Seq to quantify the transposon mRNAs from fly ovaries. Upon the loss of Rhi (rhi2/KG), transposon families that have more than fivefold increase at mRNA level are marked as red. These transposons are likely suppressed by piRNAs derived from the dual-strand clusters that marked by Rhi. Download figure Download PowerPoint Collectively, our structural data indicated that, in melanogaster, Rhi interacts with Del through one classical and one novel HP1 binding modes, and both interaction interfaces are required for Rhino function in vivo. Simulans Rhi-Del complex structure resembles the Rhi-Del structure in melanogaster Previous studies have shown that there is a cross-species barrier for simulans Rhi and melanogaster Del interaction 20. The yeast two-hybrid assay concurred that there is no interaction between Ds-Rhi-CSD and Dm-Del-N (Fig EV1C). Therefore, we crystallized the simulans Rhi-Del complex using the same strategy, where Ds-Del-NΔβ (residues 1–46) is fused to the C terminus of Ds-Rhi-CSDΔβ (residues 436–493) with a (Gly-Ser)3 linker (Table EV1). Both simulans Del-NΔβ and Rhi-CSDΔβ show highly structural homology to melanogaster Del-N and Rhi-CSD (Fig EV4A–D) as well as to the interaction interface I (Fig EV4C and D). Click here to expand this figure. Figure EV4. Interface I is conserved in simulans Rhi and Del complex (related to Fig 4C) Stereo views of the hydrophobic interaction details between two α-helices of Ds-Del-NΔβ. Residues involved in the hydrophobic interaction are labeled. I7, M9, L13, W16, I19, L20, L23 on α1 and M28, W33, F36, F37, F40, L41, W44 on α2 form a hydrophobic cluster. W44 and K12 form a cation-π interaction. Superimposition of Ds-Del-NΔβ (lime) and Dm-Del-N (pink). Stereo views of the interface I between Ds-Rhi-CSDΔβ (wine) and Ds-Del-NΔβ (lime). Residues involved in the interface are labeled. F40, W44 on Ds-Del-NΔβ, and H452, Y454, F460, F462 on Ds-Rhi-CSDΔβ form an aromatic cluster. C15 on Ds-Del-NΔβ and A473 on Ds-Rhi-CSDΔβ further enhance the aromatic cluster. I7 on Ds-Del-NΔβ has hydrophobic interactions with L451 and I471 on Ds-Rhi-CSDΔβ. M456 on Ds-Rhi-CSDΔβ is surrounded by I19, L23, F36 on Ds-Del-NΔβ. Superimposition of interface I between Rhi-CSD and Del-N in Drosophila melanogaster and Drosophila simulans. The 2Fo-Fc
Referência(s)